World Journal of Microbiology and Biotechnology

World Journal of Microbiology and Biotechnology Volume 21, Number 2 (101 - 106) Biotransformation of biphenyl by the filamentous fungus Talaromyces helicus Maria C. Romero, Elke Hammer, Renate Hanschke, Angelica M. Arambarri, Frieder Schauer DOI: 10.1007/s11274-004-2779-y Solid-state fermentation production of tetracycline by Streptomyces strains using some agricultural wastes as substrate

(107 - 114)

Agnes E. Asagbra, Abiodun I. Sanni, Olusola B. Oyewole DOI: 10.1007/s11274-004-2778-z Influence of glucose and oxygen on the production of ethyl acetate and isoamyl acetate by a Saccharomyces cerevisiae strain during alcoholic fermentation

(115 - 121)

C. Plata, J. C. Mauricio, C. Millán, J. M. Ortega DOI: 10.1007/s11274-004-2780-5 Rapid screening of Aspergillus terreus mutants for overproduction of lovastatin

(123 - 125)

M. A. Vilches Ferrón, J. L. Casas López, J. A. Sánchez Pérez, J. M. Fernández Sevilla, Y. Chisti DOI: 10.1007/s11274-004-3045-z Impact of balanced substrate flux on the metabolic process employing fuzzy logic during the cultivation of Bacillus thuringiensis var. Galleriae

(127 - 134)

R. K. I. Anderson and Kunthala Jayaraman DOI: 10.1007/s11274-004-3043-1 Degradation and corrosive activities of fungi in a diesel–mild steel– aqueous system

(135 - 142)

Fátima Menezes Bento, Iwona Boguslava Beech, Christine Claire Gaylarde, Gelsa Edith Englert, Iduvirges Lourdes Muller DOI: 10.1007/s11274-004-3042-2 Changes in lignocellulolytic enzyme activites in six Pleurotus spp. strains cultivated on coffee pulp in confrontation with Trichoderma spp.

(143 - 150)

G. Mata, D. M. Murrieta Hernández, L. G. Iglesias Andreu DOI: 10.1007/s11274-004-3041-3 Milk-clotting protease production by Nocardiopsis sp. in an inexpensive medium

(151 - 154)

M. T. H. Cavalcanti, C. R. Martinez, V. C. Furtado, B. B. Neto, M. F. Teixeira, J. L. Lima Filho, A. L. F. Porto DOI: 10.1007/s11274-004-3470-z Diversity and dynamics of bacterial species in a biofilm at the end of the Seoul water distribution system Dong-Geun Lee, Jung-Hoon Lee, Sang-Jong Kim DOI: 10.1007/s11274-004-2890-0

(155 - 162)

Utilization in alginate beads for Cu(II) and Ni(II) adsorption of an exopolysaccharide produced by Chryseomonas luteola TEM05

(163 - 167)

Guven Ozdemir, N. Ceyhan, E. Manav DOI: 10.1007/s11274-004-1563-3 Production of extracellular alkaline proteases by Aspergillus clavatus

(169 - 172)

Célia R. Tremacoldi and Eleonora Cano Carmona DOI: 10.1007/s11274-004-2724-0 Barley-based medium for the cost-effective production of Bacillus thuringiensis

(173 - 178)

P. S. Vimala Devi, T. Ravinder, C. Jaidev DOI: 10.1007/s11274-004-2217-1 Induction, and production studies of a novel glucoamylase of Aspergillus niger

(179 - 187)

M. Ibrahim Rajoka and Amber Yasmeen DOI: 10.1007/s11274-004-1766-7 Characterization of Mucor miehei lipase immobilized on polysiloxanepolyvinyl alcohol magnetic particles

(189 - 192)

L. M. Bruno, J. S. Coelho, E. H. M. Melo, J. L. Lima-Filho DOI: 10.1007/s11274-004-3321-y Lipase-catalyzed naproxen methyl ester hydrolysis in water-saturated ionic liquid: significantly enhanced enantioselectivity and stability

(193 - 199)

Jia-Ying Xin, Yong-Jie Zhao, Yan-Guo Shi, Chun-Gu Xia, Shu-Ben Li DOI: 10.1007/s11274-004-3108-1 Antimicrobial screening and active compound isolation from marine bacterium NJ6-3-1 associated with the sponge Hymeniacidon perleve Li Zheng, Haimin Chen, Xiaotian Han, Wei Lin, Xiaojun Yan DOI: 10.1007/s11274-004-3318-6

(201 - 206)

World Journal of Microbiology & Biotechnology 2005 21: 101–106 DOI: 10.1007/s11274-004-2779-y

 Springer 2005

Biotransformation of biphenyl by the filamentous fungus Talaromyces helicus Maria C. Romero1, Elke Hammer2,*, Renate Hanschke2, Angelica M. Arambarri1 and Frieder Schauer2 1 Instituto Botanica Spegazzini, Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata, La Plata, Argentina 2 Institut fu¨r Mikrobiologie, Ernst-Moritz-Arndt-Universita¨t Greifswald, F.-L.-Jahn-Str. 15, 17487 Greifswald, Germany *Author for correspondence: Tel.: +49-3834-864211, Fax: +49-3834-864202, E-mail: [email protected] Received 24 December 2003; accepted 15 June 2004

Keywords: Biphenyl degradation, filamentous fungi, oxidation products, ring fission, Talaromyces

Summary The filamentous fungus Talaromyces helicus, isolated from oil-contaminated sludge, oxidizes biphenyl via 4hydroxybiphenyl to the dihydroxylated derivatives 4,4¢-dihydroxybiphenyl and 3,4-dihydroxybiphenyl, which, to a certain extent, are converted to glycosyl conjugates. The sugar moiety of the conjugate formed from 4,4¢dihydroxybiphenyl was identified as glucose. Further metabolites: 2-hydroxybiphenyl, 2,5-dihydroxylated biphenyl, and the ring cleavage product 4-phenyl-2-pyrone-6-carboxylic acid accumulated only in traces. From these results the main pathway for biotransformation of biphenyl in T. helicus could be proposed to be the excretion of dihydroxylated derivatives (>75%) and their glucosyl conjugates (90%) and compound I (traces); 4,4¢-dihydroxybiphenyl was slowly transformed to product II (10%). Identification of products I and II Both products (I and II) seemed to be more hydrophilic, as indicated by their lower retention times on reverse phase material compared to the parent compounds. Neither of these two products was detectable after extraction of the culture supernatant by ethyl acetate followed by GC–MS analysis. In view of the high similarity of u.v. spectra to those of hydroxylated biphenyls and in comparison to the behaviour of products found after biotransformation of dibenzofuran by Penicillium canescens (Hammer et al. 2001), these compounds were assumed to be conjugates. This assumption should be proven by deconjugation experiments carried out with glucuronidase and arylsulphatase. Culture supernatant containing products I and II was concentrated by lyophilization, and individual 50mg-samples were treated with each of the enzymes. Hydrolysis with glucuronidase led to the depletion of the products I and II and produced increasing amounts of 3,4-dihydroxybiphenyl (from product I) and 4,4¢-dihydroxybiphenyl (from product II). In contrast, both the control without enzyme and the sample treated with arylsulphatase showed no hydrolysis and both conjugates remained intact. Because of the unspecific reaction of glucuronidase towards several sugar conjugates, the sugar moiety was analysed by a more specific enzymatic assay for glucose with glucose oxidase. After glucuronidase treatment of separated product II, the assay showed accumulation of glucose in the samples, whereas the sugar was not detectable in the controls (medium,

Discussion The fungus T. helicus cannot use biarylic compounds like biphenyl, dibenzofuran, or naphthalene for growth. However, the fungus is able to co-metabolize these compounds. Cells grown in mineral salts medium with glucose or grown on complex medium in addition, transformed the hydrophobic parent compound to more hydrophilic products: hydroxylated products in high amounts and more hydrophilic sugar conjugates as well as ring-cleavage products in lower amounts. The fungus produced relatively large amounts of hydroxylated intermediates, which included monohydroxylated and dihydroxylated compounds, indicating the involvement of monoxygenases in the biotransformation pathway. From biphenyl, 4-hydroxybiphenyl was the major product, whereas production of 2-hydroxybiphenyl was rather low. Hydroxylation of the C4- position was also described as the main pathway for other filamentous fungi, yeasts, and mammals (Meyer & Scheline 1976; Smith et al. 1980; Golbeck et al. 1983; Lange et al. 1998; Sietmann et al. 2000). A second hydroxylation by T. helicus can occur in the same ring as well as in the second ring. As a result, the dihydroxylated derivatives 3,4-dihydroxybiphenyl and 4,4¢-dihydroxybiphenyl are produced in a ratio 1:3. Strong accumulation of para, para¢-hydroxylated biphenyl was also observed for Aspergillus strains by Mobley et al. (1993). In contrast, hydroxylation at the unsubstituted ring of 4-hydroxybiphenyl occurred only in traces or could not be observed at all in yeast. In strains of the genus Trichosporon (Sietmann et al. 2002) or Debaryomyces vanrijiae (Lange et al. 1998) the main transformation pathway of biphenyl goes via 4-hydroxybiphenyl and 3,4-dihydroxybiphenyl up to ring cleavage. Formation of considerable amounts of 2,5-dihydroxybiphenyl as described for Trichosporon mucoides (Sietmann et al. 2000) was not observed in T. helicus. The strong inhibition of hydroxylation reactions in the presence of the cytochrome-P-450-inhibiting substance 1-aminobenzotriazole points to an involvement of such enzymes in the transformation reactions. The differences in the level of inhibition of the first and the second hydroxylation may indicate the existence of different

Biphenyl degradation by Talaromyces

105 A

HO

OH OH

D

C

B

HO OH HO

OH

E

OH

G

F

OH

CO OH

O

HO

O

O

HO

OH

H

O

O

O

HO OH

HO

OH OH

OH OH

Metabolite II

Metabolite I

Figure 4. Proposed pathway for the biotransformation of biphenyl by the ascomycete T. helicus. (A) Biphenyl; (B) 2-hydroxybiphenyl; (C) 3hydroxybiphenyl; (D) 4-hydroxybiphenyl; (E) 2,5-dihydroxybiphenyl; (F) 3,4-dihydroxybiphenyl; (G) 4,4¢-dihydroxybiphenyl.

monooxygenases in the fungus. Moreover, the substrate specificity of the cytochrome P-450 enzymes involved in unspecific oxidation of hydrophobic compounds in various fungi and yeast seems to be rather different. As mentioned above, the hydroxylation pattern can differ drastically in various eukaryotic organisms. Although dihydroxylated biphenyl derivatives accumulated in high amounts in the culture supernatant of T. helicus, they are not dead-end products, but were further transformed. As found in yeasts – mainly described for Trichosporon strains (Sietmann et al. 2002) and Debaryomyces vanrijiae (Lange et al. 1998) but also for the imperfect filamentous fungus Paecilomyces lilacinus (Gesell et al. 2001), 3,4-dihydroxybiphenyl can be oxidized up to ring cleavage. As a result 4-phenyl-2-pyrone-6-carboxylic acid was formed (Figure 4). In T. helicus this reaction only occured to a low extent when biphenyl or 4-hydroxybiphenyl were used as substrates. If 3,4-dihydroxybiphenyl was used as substrate, nearly complete conversion to the pyrone was observed. Perhaps the dihydroxylated biphenyl induces a ring-cleaving enzyme, which does not occur at the low level of 3,4-dihydroxylated biphenyl that accumulated by transformation of biphenyl or 4-hydroxybiphenyl. Furthermore, under these conditions 3,4-dihydroxybiphenyl and 4,4-dihydroxybiphenyl, were partially converted to glucose conjugates. Therefore, detoxification of harmful compounds in T. helicus seems to occur via the so-called phase I (hydroxylation)/phase II transformation system (Zhang et al. 1996), through which excretion of toxic intermediates becomes possible. In contrast, in yeast detoxification of such compounds is

achieved by ring cleavage of the dihydroxylated derivatives produced (Lange et al. 1998; Sietmann et al. 2000, 2002). Surprisingly, conjugation of 4-hydroxybiphenyl was not observed. Furthermore, it remains unclear why hydroxylated biphenyl derivatives can be excreted partially directly, whereas another part seems to be only excreted after conversion to sugar conjugates. Here, it is proven for the first time that a Talaromyces strain isolated from oil- and PAH-contaminated water shows higher potential to tolerate and to transform biphenyl and its derivatives in comparison to Talaromyces isolates from natural sites.

Acknowledgements This research was supported by the National Research Council, Argentina and the Bundesministerium fu¨r Forschung und Technologie, Germany. We thank M. Specht, Institute for Organic Chemistry, University of Hamburg for providing 4-phenyl-2-pyrone-6-carboxylic acid and H. Lehnherr for revising the manuscript.

References Bergmeyer, H.U., Bernt, E., Schmidt, F. & Stork, H. 1974 In Bergmeyer, H.U. Methoden der enzymatischen Analyse, vol. 2, 3rd edn. pp. 1241–1246. Weinheim: Verlag Chemie. ISBN 3-527-25530-3. Cerniglia, C.E. 1997 Fungal metabolism of polycyclic aromatic hydrocarbons: past, present and future applications in bioremediation. Journal of Industrial Microbiology and Biotechnology 19, 324–333.

106 Cerniglia, C.E., Freeman, J.P. & Mitchum, R.K. 1982 Glucuronide and sulfate conjugation in the fungal metabolism of aromatic hydrocarbons. Applied and Environmental Microbiology 43, 1070–1075. Cox, J.C. & Golbeck, J.H. 1985 Hydroxylation of biphenyl by Aspergillus parasiticus: approaches to yield improvement in fermentor cultures. Biotechnology and Bioengineering 27, 1395– 1402. De Boer, T.D. & H.J. Backer. 1956 Diazomethane. Organic Synthesis 36, 14–16. Dodge, R.H., Cerniglia, C. & Gibson, D.T. 1979 Fungal metabolism of biphenyl. Biochemical Journal 178, 223–230. Gesell, M., Hammer, E., Specht, M., Francke, W. & Schauer, F. 2001 Biotransformation of biphenyl by Paecilomyces lilacinus and characterization of ring cleavage products. Applied and Environmental Microbiology 67, 1551–1557. Golbeck, J.H., Albaugh, S.A. & Radmer, R. 1983 Metabolism of biphenyl by Aspergillus toxicaricus: induction of hydroxylating activity and accumulation of water-soluble conjugates. Journal of Bacteriology 156, 49–57. Hammer, E. & Schauer, F. 1997 Fungal hydroxylation dibenzofuran. Mycological Research 101, 433–436. Hammer, E., Schoefer, L., Schaefer, A., Hundt, K. & Schauer, F. 2001 Formation of glucoside conjugates during biotransformation of dibenzofuran by Penicillium canescens. Applied Microbiology and Biotechnology 57, 390–394. Hakkinen, I., Hernberg, S., Karli, P. & Vikkula, E. 1973 Diphenyl poisening in fruit paper production. Archives of Environmental Health 26, 70–74. Hanschke, R. & Schauer, F. 1996 Improved ultrastructural preservation of yeast cells for scanning electron microscopy. Journal of Microscopy 184, 81–87. Kaufman, D.D. & Blake, J. 1973 Microbial degradation of several acetamide, acylanilide, carbamate, toluidine, and urea pesticides. Soil Biology and Biochemistry 5, 297–308. Lange, J., Hammer, E., Specht, M., Francke, W. & Schauer, F. 1998 Biodegradation of biphenyl by the ascomycetous yeast Debaryomyces vanrijiae. Applied Microbiology and Biotechnology 50, 364– 368.

M.C. Romero et al. Meyer, T. & Scheline, R.R. 1976 The metabolism of biphenyl. II. Phenolic metabolites in the rat. Acta Pharmacologica and Toxicologica 39, 419–432. Mobley, D.P., Finkbeiner, H.L., Lockwood, S.H. & Spivack, J. 1993 Synthesis of 3-arylmuconolactones using biphenyl metabolism in Aspergillus. Tetrahedron 49, 3273–3780. Schultz, T.W., Kraut, D.H., Sayler, G.S. & Layton, A.C. 1998 Estrogenicity of selected biphenyls evaluated using a recombinant yeast assay. Environmental Toxicology and Chemistry 17, 1727– 1729. Seigle-Murandi, F.M., Krivobok, S.M.A., Steinman, R.L., BenoitGuyod, J.L. & Thiault, G. 1991 Biphenyl oxide hydroxylation by Cunninghamella echinulata. Journal of Agricultural and Food Chemistry 39, 428–430. Sietmann, R., Hammer, E., Moody, J., Cerniglia, C.E. & Schauer, F. 2000 Hydroxylation of biphenyl by the yeast Trichosporon mucoides. Archives of Microbiology 174, 353–361. Sietmann, R., Hammer, E. & Schauer, F. 2002 Biotransformation of biarylic compounds by yeasts of the genus Trichosporon. Systematic and Applied Microbiology 25, 332–339. Sietmann, R., Hammer, E., Specht, M., Cerniglia, C.E. & Schauer, F. 2001 Novel ring cleavage products in the biotransformation of biphenyl by the yeast Trichosporon mucoides. Applied and Environmental Microbiology 67, 4158–4165. Smith, R.V. & Rosazza, J.P. 1974 Microbial models of mammalian metabolism. Aromatic hydroxylation. Archives of Biochemistry and Biophysics 161, 551–558. Smith, R.V., Davis, P.J., Clark, A.M. & Glover-Milton, S. 1980 Hydroxylations of biphenyl in fungi. Journal of Applied Bacteriology 49, 65–73. Sutherland, J.B., Selby, A.L., Freeman, J.P., Fu, P.P., Miller, D.W. & Cerniglia, C.E. 1992 Identification of xyloside conjugates formed from anthracene by Rhizoctonia solani. Mycological Research 96, 506–517. Zhang, D., Yang, Y, Leakey, J.E. & Cerniglia, C.E. 1996 Phase I and phase II enzymes produced by Cunninghamella elegans for the metabolism of xenobiotics. FEMS Microbiology Letters 138, 221– 226.

World Journal of Microbiology & Biotechnology 2005 21: 107–114 DOI: 10.1007/s11274-004-2778-z

 Springer 2005

Solid-state fermentation production of tetracycline by Streptomyces strains using some agricultural wastes as substrate Agnes E. Asagbra1,*, Abiodun I. Sanni2 and Olusola B. Oyewole3 1 Biotechnology Division, Federal Institute of Industrial Research, Oshodi, Nigeria 2 Department of Botany and Microbiology, University of Ibadan, Nigeria 3 Department of Food Science and Technology, University of Agriculture Abeokuta, Nigeria *Author for correspondence: Tel.: +234-1-080-23001173, E-mail: [email protected] Received 23 December 2003; accepted 15 June 2004

Keywords: Optimisation, peanut shells, solid-state fermentation, Streptomyces, tetracycline

Summary The ability of Streptomyces sp. OXCI, S. rimosus NRRL B2659, S. rimosus NRRL B2234, S. alboflavus NRRL B1273 S. aureofaciens NRRL B2183 and S. vendagensis ATCC 25507 to produce tetracycline using some local agricultural wastes as solid state media, were assessed. The wastes employed include peanut (groundnut) shells, corncob, corn pomace and cassava peels. Bacillus subtilis ATCC 6633 was used to assay antimicrobial activity. All the strains produced tetracycline in a solid-state fermentation process containing peanut (groundnut) as the carbohydrate source. Streptomyces sp. OXC1 had the highest ability for tetracycline production with peanut shells as the substrate in solid fermentation (13.18 mg/g), followed by S. vendagensis ATCC 25507 (11.08 mg/g), S. rimosus NRRL B1679 (8.46 mg/g), S. alboflavus NRRL B1273 (7.59 mg/g), S. rimosus NRRL B2234 (6.37 mg/ g), S. aureofaciens NRRL B2183 (4.27 mg/g). Peanut (groundnut) shells were the most effective substrate (4.36 mg/ g) followed by corncob (2.64 mg/g), cassava peels (2.16 mg/g) and corn pomace (1.99 mg/g). The composition of the peanut (groundnut) shell medium optimal for tetracycline production were peanut shells 100 g, organic nitrogen (peanut meal) 10 g, (NH4)2SO4 1 g, KH2PO4 0.5 g, CaCO3 0.5 g, NaCl 0.5 g, MgSO4
7H2O 0.5 g, soluble starch 10 g, peanut oil 0.25 ml with initial moisture content of 65–68%, and initial pH 5.3–6.3. Substrate (1 g dry weight) was inoculated with 1.0 · 108 conidia per ml and incubated at 28–31
C for 5–7 days, producing 13.18 mg/g of total tetracycline. Tetracycline detection started on day 3 and attained its maximum level on day 5.

Introduction Streptomyces is the largest antibiotic-producing genus in the microbial world so far (Watve et al. 2001). However it is becoming increasingly apparent that 99% of the diverse bacterial species with antibiotic potentials are still unexplored (Ward et al. 1990; Watve et al. 2000). The need to continue to explore for more antibioticproducing microbial strains of microorganisms has been stressed because of the resistance of disease-causing microorganisms to known antibiotics and the varying diversities in the properties of antibiotics produced by different organisms (Pelczar et al. 1993). Antibiotic imports, including tetracycline and ampicillin constitute a major percentage of medical imports in Nigeria. Unfortunately, while the raw materials for antibiotic production are readily available in the country, there is no single facility for commercial antibiotic production (Ifudu 1986a, b). Tetracyclines are broad-spectrum antibiotics with an octahydronaphthacene structure with four annelated sixmembered rings. They are useful in a variety of infections

caused by bacteria, rickettsias, trachoma, coccidia, amoebae, balantida and mycoplasma (Yang & Swei 1996). They are used for the control of plant diseases, stimulation of acid fermentation and inhibition of material biodeterioration (Somerson & Phillips 1961; Huang 1972). Submerged cultures (Bhatnagar et al. 1988) and culture media are usually used to produce antibiotics; culture conditions affect the kind and quantity of antibiotic production (Komatsu et al. 1975; Yang and Ling 1989). Solid-state fermentation gives a product which is more stable than that of submerged culture and also requires less energy input (Wang 1989). Cellulolytic materials are abundantly available globally and can be used by a number of microorganisms (Yang & Swei 1996). Peanut (groundnut) and its residues are abundantly available in Nigeria, as at 1996 the annual peanut (groundnut) production was 7608 tons (Rake 1998). This study was embarked upon to compare the antibiotic-producing potentials of some reference standard Streptomyces spp. with a locally isolated strain, using different agricultural wastes in Nigeria as substrates in tetracycline production.

108 Materials and methods Agricultural wastes The agricultural wastes used for the work include peanut shells, corn cob, corn pomace and cassava peels. They were obtained from the pilot plant of the Federal Institute of Industrial Research Oshodi, Lagos, Nigeria. The wastes were dried separately and were screened with 4–16 mesh sizes to remove the dust and large aggregates. Determination of moisture contents of substrates Samples were dried at 60
C under vacuum for 8–12 h to a constant weight. The weight difference after drying was considered as the moisture content. pH Initial pH of substrates was determined directly by immersing the electrode into the substrate, but the final value was determined after mixing a sample with 4 volume distilled water (Yang & Swei 1996). Bulk density The dry weight or wet weight of samples per unit volume (1 ml) was the bulk density in dry weight or wet weight, respectively (Baver 1956). Water activity (Aw) Samples with different moisture contents were placed in a sealed container at 25
C and water activity was determined by a hygrometer (Yang 1977). Microorganisms Streptomyces sp. OXC1 was obtained from topsoil around the pilot plant building of the Federal Institute of Industrial Research Oshodi and was characterized by 16SrRNA gene sequencing. S. rimosus NRRL B2659, S. rimosus NRRL B2234, S. alboflavus NRRL B1273 S. aureofaciens NRRL B2183 and S. vendagensis ATCC 25507 were obtained from the Northern Regional Agricultural Research Center in USA. All the Streptomyces isolates produced antibiotics. Bacillus subtilis ATCC 6633 was used to assay antimicrobial activity.

A.E. Asagbra et al. of the suspension contained about 108 spores. The solid state medium contained (g): peanut (groundnut) shells 100, (NH4)2SO4 1.5, CaCO3 1.0, NaCl 0.5, KH2PO4 0.5 and MgSO4
7H2O. The medium was mixed thoroughly with seed culture and the initial moisture content was then adjusted to 65% with distilled water. Incubation was at 28
C for 5–7 days. Extraction of antibiotics After fermentation, the culture mass was extracted with four times volume of distilled water with shaking at room temperature for 5 min (Yang & Ling 1989; Yang & Swei 1996). Determination of antibiotic activity Antimicrobial activity of the extract was measured by the paper disc method in antibiotic medium 1 (DIFCO Laboratory, USA) at 30
C as described by Yang & Ling (1989). Total tetracycline equivalent potency was calculated from the clear zones for tetracycline in the range of 1 lg/g to 15 mg/g. The regression equation for this curve was y ¼ 0.349485x + 0.6 where y was the log concentration of tetracycline (mg/g and x was the diameter of the zone of inhibition (mm) (r2 ¼ 0.950). Effect of substrate The peanut shells in the solid state medium were replaced by other agricultural substrates using the Streptomyces strains. The waste producing the highest tetracycline level in the solid state fermentation process was chosen. The substrates investigated included corn cob, peanut shells, corn pomace and cassava peels. The composition of the medium is as stated earlier. Peanut shells were chosen because they produced the highest tetracycline concentration. Process optimization Following earlier investigations to choose the best producing strain and the best substrate, the optimum conditions for tetracycline production were determined by employing the methods of Yang & Ling (1989) and Yang & Swei (1996). pH and antibiotic bioassay were monitored in all cases. Effect of initial moisture content

Culture media and culture conditions The Streptomyces strains were initially grown at 26
C in solid medium containing (g 11): soluble starch 10; (NH4)2SO4 2; CaCO3 2; NaCl 1; K2HPO4 1; MgSO4
7H2O 1, trace elements 1 ml (contained FeSO4
7H2O 1.0 g, CuSO4
5H2O 0.5 g) ZnSO4
7H2O 1.0 g and MnSO4
2H2O 1 g) and agar 20 g. Spores were washed with 5 ml 0.05% Tween-80 in sterile water and harvested by centrifugation at 3000 · g for 10 min. Each ml

The effects of different initial moisture contents on the production of tetracycline by Streptomyces sp. OXC1 were investigated. The modified solid-state fermentation medium described by Yang & Swei (1996) was used: peanut (groundnut) shell 100 g, CaCO3, 1.0 g; KH2PO4, 0.5 g; MgSO4
7H2O, 2 g; (NH4)2SO4, 2 g was used for all the tests. The moisture contents tested were between 50 and 80%. The medium was sterilized at 121
C for 20 min. The sterile medium was inoculated and the

109

Tetracycline production on wastes appropriate volume of sterile distilled water was added to make up the desired moisture content. These were then incubated statically in flasks (the thickness of medium was about 2 cm). Incubation was at 31
C for 5–7 days with stirring once a day. Effect of pH pH values of 4.3, 5.3, 6.3, 7.3 and 8.3 were used. The medium was as stated above. The sterile medium was inoculated and the moisture content adjusted to 68% moisture. Duplicate sets of experiments were carried out. Incubation was at 31
C for 5–7 days. The statically incubated flasks were stirred once daily. Effect of temperature The effect of different temperature ranges on the production of tetracycline was investigated. The modified medium as described by Yang & Ling (1989) was used. The different temperatures tested were 28, 31, 34, 37 and 40
C. The medium were sterilized at 121
C for 20 min. The sterile media were inoculated with 1 · 108 conidia and the moisture content made up with sterile distilled water to 68%. Duplicate flasks were incubated at all the above stated temperatures. Effect of inoculum size Different inoculum sizes of Streptomyces sp. OXC1 were used to inoculate sterile medium, to produce a high yield of tetracycline. The inoculum sizes used were 1.0 · 105, 1.0 · 106, 1.0 · 107, 1.0 · 108, and 1.0 · 109. After the addition of the inoculum in duplicates, the moisture content was made up to 68%. Incubation was at 31
C for 5–7 days with stirring once daily. The samples were analysed for pH change and bioassay.

sources on pH and antimicrobial compound production were investigated. A modification of the solid state medium as described by Yang & Swei (1996) was used. The organic nitrogen sources were used to replace inorganic sources in the following concentrations (10, 20, 30, 40 and 50%) and urea was used at the following concentrations (0.25, 0.5, 0.75, 1%). The medium was adjusted with the required organic nitrogen and sterilized at 121
C for 20 min. The sterile medium was inoculated and the moisture content adjusted to 68% with incubation carried out statically at 28
C for 5– 7 days with stirring once a day. Effect of combined nitrogen sources Combined inorganic and organic nitrogen sources were studied. A modification of the solid state medium as described by Yang & Ling (1989) was used. The medium comprised 100 g of peanut shells, CaCO3 1 g, NaCl 0.2 g and 0.5% (NH4)2SO4. The organic nitrogen sources (peanut meal, soybean meal and rice bran) were used in addition to the inorganic nitrogen source [1% (NH4)2SO4]. The concentrations of the organic nitrogen used were 10%. The medium was treated as stated above. Effect of additional carbon source Effect of additional carbon sources on the antimicrobial compound production was investigated. A modification of the solid state medium described by Yang & Ling (1989) was used: peanut shells 100 g, (NH4)2SO4 2.4 g, CaCO3 1 g, NaCl 0.2 g, and 1 · 108 conidia at 65% moisture content was the inoculum load. The carbon sources tested were glucose, sucrose, maltose and starch at 10% concentration. The sugars were filter sterilized while the medium was sterilized at 121
C for 20 min. Effect of inorganic salts

Effect of inorganic nitrogen sources The effects of the nitrogen sources on pH and tetracycline production were investigated. Inorganic nitrogen sources such as ammonium chloride (NH4Cl), ammonium nitrate (NH4NO3) and ammonium sulphate (NH4)2SO4 were used in the following concentrations (0.25 0.5 0.75, 1%). A modification of the solid state medium as described by Yang & Swei (1996) was used. 100 g of the medium containing each concentration of the nitrogen source was distributed into flasks in duplicates. The medium was sterilized at 121
C for 20 min and moisture level adjusted to 68% prior to inoculation with Streptomyces sp. OXC1 in different flasks. The inoculated flasks were incubated statically at 31
C with stirring once a day. Effect of organic nitrogen sources Rice bran, soybean meal and peanut meals were used as organic nitrogen sources. The effects of the nitrogen

The effects of different concentrations of inorganic salts on the production of antimicrobial compound by Streptomyces sp. OXC1 were investigated. The modified medium described by Yang & Swei (1996) was used as earlier stated. The salts tested were KH2PO4 CaCO3, MgSO4
7H2O and NaCl at 0.0, 0.5, 1.0, 1.5 and 2.0% concentrations. The sterile medium was thoroughly mixed with 1 · 108 conidia and the moisture content adjusted to 68% with sterile distilled water. The experiments were carried out in duplicate. Effect of oils on production of antimicrobial compound Vegetable oils such as palm oil, peanut (groundnut) oil, melon oil, soybeans oil and palm kernel oil (PKO) was used in the experiment. Ifudu (1986a, b) described the method used. To the medium were added different concentrations of the oils (0.25, 0.50, 0.75 and 1.0%). The medium was sterilized at 121
C for 20 min. The sterile medium was inoculated and the appropriate

110

A.E. Asagbra et al.

volume of sterile distilled water was added to raise the moisture content to 68%. These were incubated statically in flasks (the thickness of medium being about 2 cm). Incubation was at 31
C for 5–7 days with stirring once a day. The experiments were carried out in duplicates. Results Table 1 shows the tetracycline production of the different strains when cultivated on the basal medium. Streptomyces sp. OXC1 demonstrated the potential for tetracycline production using peanut (groundnut) shells as the substrate in solid fermentation (13.18 mg/g), followed by S. vendagensis ATCC 25507 (11.08 mg/g), S. rimosus NRRL B1679 (8.46 mg/g), S. alboflavus NRRL B1273) (7.59 mg/g), S. rimosus NRRL B2234 (6.37 mg/g), S. aureofaciens NRRL B2183 (4.27 mg/g). Therefore Streptomyces sp. OXC1 was chosen as the best strain in this study. Agricultural wastes Results of proximate analyses of the substrates used revealed that peanut shells had total carbohydrate 16.99%, crude protein 6.77%, crude fibre 70.38% and ash 2.65%, corn cob (total carbohydrate 49.64%, crude protein 5.30%, crude fibre 38.18% and ash 2.53%), corn pomace (total carbohydrate 75.46%, crude protein 5.65%, crude fibre 12.45% and ash 2.93%) and cassava peels (total carbohydrate 76.47%, crude protein 4.11%, crude fibre 10.35% and ash 3.79%). Effect of substrate on isolates for tetracycline production Experimental results indicated that peanut shells was the most effective substrate (4.36 mg/g) followed by corn cob (2.64 mg/g), followed by cassava peels (2.16 mg/g) and corn pomace (1.99 mg/g). Therefore peanut shells were chosen as the substrate for the model system of antibiotic production. Process optimization Initial moisture content The effect of initial moisture content of substrate on tetracycline production is shown in Table 2. The initial Table 1. Tetracycline production by the different strains when cultivated on peanut shells at 68% moisture content at 31
C for 5 days. Streptomyces spp.

Initial pH

Final pH

Total tetracycline equivalent (mg/g)

S. vendagensis ATCC 25507 S. rimosus NRRL B1679 S. alboflavus NRRL B1273 S. rimosus NRRL B2234 S. aureofaciens NRRL B2183 Streptomyces sp. OXC1

5.40 6.29 5.80 5.15 5.30 5.60

6.60 7.05 6.30 7.40 5.50 5.90

11.08 8.46 7.59 6.37 4.27 13.18

Table 2. Tetracycline determination with bioassay method at different initial moisture content. Initial moisture content

% increase in Final pH moisture content

Bioassay (mg/g)

50 55 60 65 67 68 69 70 75 80

7.58 6.16 5.05 1.53 1,66 1.58 1.49 5.10 4.49 3.73

0.82 1.65 2.70 3.40 3.75 4.10 3.57 3.40 0.69 0.01

5.35 5.35 5.31 5.70 6.00 6.05 6.20 6.01 6.33 6.36

moisture content of the substrate increased between 1.53% and 1.66% from an initial of 65% whereas it increased between 4.49% and 5.10% for initial moisture content of 75%. During fermentation, tetracycline was first detected on day 3, reached a maximum yield on day 5 and gradually decreased. Tetracycline production increased with initial moisture content of 50–70%, having its maximum at 68% (4.1 mg/g). When the initial moisture content was less than 50%, tetracycline production was low as the substrate was too dry for cell growth and antibiotic production. At initial moisture of 68%, Aw of the substrate was 0.995, and bulk densities on dry and wet weight bases were 0.20 and 0.39 g/m3, respectively. At moisture contents higher than 68% tetracycline production gradually decreased while at initial moisture content of 80%, total tetracycline equivalent was only 0.01 mg/g dry substrate, since the togetherness of the substrate prevented gas exchange. Initial pH Tetracycline production at different initial substrate pH is shown in Figure 1. The optimal pH for tetracycline production was the same as the original pH of peanut shells (5.35–5.60). When the pH was lower than 5.3, tetracycline production was low and each gram of the dry substrate produced about (0.15 mg/g) total tetracycline equivalent, when the initial pH was higher than 6.3, tetracycline production, a slight decrease was observed. At initial pH of 8.3 each gram of dry substrate produced 1.70 mg/g of total tetracycline equivalent during fermentation. The pH change of substrate was not significant when the initial pH was lower than 6.3 but the reverse was observed at higher pH. As the original pH of the peanut shells ranged between 5.35 and 5.55, it was not necessary to adjust the pH for tetracycline production in solid state fermentation. Incubation temperature Tetracycline production was optimal at 31
C, and decreased sharply when incubation temperature was higher than 40
C or less than 28
C. Each gram of dry substrate produced 3.60 and 2.52 mg/g of total tetracycline equivalent at 31
C and 34
C respectively.

111

Tetracycline production on wastes

Table 3. Effect of inorganic salts on tetracycline production in solidstate fermentation of peanut shells.

Figure 1. Effect of pH on tetracycline production in solid fermentation of peanut shells at 31
C for 5 days by Streptomyces sp. OXCI.

Inorganic salts Addition of CaCO3 regulated the substrate pH value and stimulated tetracycline production. Addition of 0.5% CaCO3 resulted in the optimal production of 3.40 mg/g tetracycline equivalent. Supplementation with 1.0% MgSO4
7H2O produced 2.70 mg/g of total tetracycline equivalent. NaCl and MgSO4
7H2O stimulated antimicrobial production. However NaCl stimulated it by 140% while KH2PO4 stimulated tetracycline production by 40% (Table 3). Nitrogen sources The effect of inorganic nitrogen source on tetracycline production and substrate pH showed that 1.0% of (NH4)2SO4 was the best inorganic nitrogen source and it produced 4.62 mg/g of total tetracycline equivalent per gram of dry substrate. NH4NO3 was the next best at 2.0% concentration giving 4.27 mg/g total tetracycline

Inorganic salts

Final pH

Total tetracycline equivalent (mg/g)

KH2PO4 0.0 0.5 1.0 1.5 2.0

5.50 5.50 5.90 5.90 6.05

2.35 3.05 2.90 2.85 2.70

CaCO3 0.0 0.5 1.0 1.5 2.0

4.90 5.90 6.05 5.15 6.35

0 3.40 2.75 2.00 1.70

MgSO4 Æ 7H2O 0.0 0.5 1.0 1.5 2.0

5.50 6.30 6.40 6.10 6.10

0 2.08 2.70 2.53 2.47

NaCl 0.0 0.5 1.0 1.5 2.0

5.80 5.90 5.90 5.90 5.90

1.70 4.80 2.70 2.70 2.35

equivalent per gram of dry substrate. Agricultural wastes were tested as alternative organic nitrogen sources. Of the wastes tested 10.0% soybean meal and 10.0% peanut meal enhanced production of tetracycline producing 4.27 and 4.62 mg/g tetracycline equivalent, respectively. To improve the utilization of the agricultural wastes the combined organic nitrogen source and inorganic source was found to have a complementing effect on tetracycline production. When 1.0% of (NH4)2SO4 was added to 10.0% of peanut meal there was an increase in the tetracycline produced. Each gram substrate produced 11.78 mg/g of tetracycline (Table 4).

Table 4. Effect of different nitrogen sources on tetracycline production in solid-state fermentation.a Nitrogen source (%)

Initial pH

Final pH

Bioassay method (total tetracycline equivalent) (mg/g)

None (NH4)SO4 (1.0%) NH4Cl (1.0%) Urea (0.5%) NH4NO3 (0.5%) Rice bran (10%) Soybean meal (10%) Peanut meal (10%) Peanut meal (10%) + (NH4)2SO4 (1.0%) Rice bran (10%) + (NH4)2SO4 (1.0%) Soybean meal (10%) + (NH4)2SO4 (1.0%)

5.60 5.35 5.20 6.10 5.30 5.30 5.30 5.30 5.30 5.50 5.30

6.90 5.80 5.35 8.20 5.65 5.40 5.60 6.00 5.80 5.80 5.60

1.47 4.02 2.00 3.49 4.27 6.28 6.28 10.21 11.78 8.64 8.99

*a

Peanut shells 100 g, CaCO3 0.5 g, NaCl 0.5 g and 1 · 108 were mixed thoroughly, and the initial moisture content was adjusted with distilled water to 68%. The substrate was incubated at 31
C for 5 days.

112

A.E. Asagbra et al.

Table 5. Effect of additional carbon source on tetracycline production by Streptomyces sp. OXC1. Carbon source (10%)

Initial pH

Final pH

Bioassay (mg/g)

None Glucose Maltose Sucrose Soluble starch

5.40 5.50 5.30 5.35 5.30

4.00 5.50 5.20 5.90 4.40

2.35 2.17 2.78 1.70 4.01

Additional carbon source Effect of additional carbon source on tetracycline production is shown in Table 5. Maltose and soluble starch stimulated tetracycline production by 1.29–1.95 times. Glucose and sucrose inhibited the production of tetracycline by 10.0 and 40.0% respectively. Additional vegetable oil The experiments showed that peanut oil was the best oil for tetracycline production. 0.25% peanut oil produced 4.02 mg/g of tetracycline equivalent. At 0.25%, palm kernel oil and palm oil did not synthesize any antimicrobial compound but rather inhibited the production. As the concentration of the oil increased the rate of tetracycline production reduced (Figure 2). Inoculum size Each gram of dry substrate inoculated with 1.0 · 108 conidia could produce 3.22 mg total tetracycline equivalent when the inoculum size was 1.0 · 1010 (2 mg of

Figure 2. Biosynthesis of tetracycline produced by Streptomyces sp. OXCI in groundnut shells supplimented with various vegetable oils.

total tetracycline equivalent) or less than 1.0 · 106 (0.86 mg of total tetracycline equivalent) conidia. Tetracycline production decreased after 5 days incubation. The above results reveal the following. The optimum conditions for tetracycline production using soybean meal as the organic nitrogen source were peanut shells at an initial moisture content of 65–70%, initial pH 5.6– 5.8, (NH4)2SO4 0.5%, CaCO3 0.5%, NaCl 0.5%, KH2PO4 0.55%, MgSO4
7H2O 1.0%, incubation at 31
C for 5 days. Each gram of the dry substrate produced 11.78 mg/g of total tetracycline equivalent. In the case of peanut meal as the organic nitrogen source: the optimum conditions for tetracycline production by Streptomyces sp. OXC1 were peanut shells at initial moisture content of 65–70%, initial pH 5.6–5.8, supplemented with (NH4)2SO4 0.5%, CaCO3 0.5%, NaCl 0.5%, KH2PO4 0.55%, MgSO4
7H2O 1.0%, soluble starch 10%, peanut oil 0.25% and incubated at 31
C for 5 days. Each gram of the dry substrate produced 13.18 mg/g of tetracycline (Table 6).

Discussion Tetracycline was detected on the third day of incubation and had maximal activity at 5 days and decreased gradually within a 1 month period. Tetracycline in Streptomyces sp. OXC1 was a secondary metabolite synthesized and secreted in the late lag phase or in the stationary phase (Okami & Oomura 1979). In submerged fermentation, antibiotic activity sharply decreased after prolonged incubation due to cell autolysis (Yang &Ling 1989). During this study it is observed that the moisture content of the substrate increased, possibly due to the production of metabolic water by the Streptomyces. This result compares with that obtained by Yang & Ling (1989), Yang & Cheng (1991) in the solid-state fermentation of sweet potatoes and when sweet potatoes were enriched with amylolytic fungi. It also compares with corncob fermentation by Trichoderma and oxytetracycline by S. rimosus in solid-state fermentation of corncob (Yang 1993; Yang & Swei 1996) and during protease production from sweet potato residue by amylolytic fungi (Yang & Huang 1994). The maximal initial moisture content for tetracycline production was at 65– 68%, and the final moisture content was 67–70%. The substrate pH was regular and did not become acidic; this could have been due to the initial pH of the substrate, and the high level of nitrogen source. Yang (1988) observed that the presence of other nitrogen sources apart from inorganic nitrogen source helped to maintain substrate pH. Calcium carbonate also played a role in counteracting acidity and enhancing tetracycline production (Yang & Ling 1989). The optimal temperature for tetracycline production was dependent on the test organism (Yang 1993). In this study Streptomyces sp. OXC1 had an optimal temperature of 31
C for tetracycline production in solid state

0.5 0.5 0.5 10.0 0.0 20.0 0.0 0.0 0.25 0.0 5.60 5.90 11.78 0.5 0.5 0.5 0.0 0.0 0.0 20.0 0.0 0.0 0.0 5.90 7.10 8.89 0.5 0.5 0.5 0.0 0.0 10.0 0.0 0.0 0.25 0.0 5.50 6.30 10.37 0.5 0.5 0.5 0.5 10.0 10.0 0.0 0.0 0.25 0.0 5.40 5.80 11.78 0.5 0.5 0.5 0.5 10.0 0.0 0.0 0.0 0.0 0.0 5.80 6.00 3.73 0.5 0.5 0.5 0.5 0.0 0.0 20.0 1.0 0.0 0.0 5.50 5.90 6.80 0.5 0.5 0.5 0.5 10.0 10.0 0.0 0.0 0.0 0.25 5.60 6.00 13.18 0.5 0.5 0.5 0.5 0.0 10.0 0.0 0.0 0.25 0.0 5.50 5.90 6.19 0.5 0.5 0.5 0.5 0.0 0.0 20.0 0.0 0.25 0.0 5.60 6.00 11.08 0.5 0.5 0.5 0.5 0.0 0.0 0.0 0.0 0.0 0.0 5.60 6.00 3.05 0.5 0.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.60 6.10 1.47

0.5 0.5 0.5 0.5 0.0 10.0 0.0 1.0 0.0 0.0 5.50 6.00 7.59

100 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 1.0 100 1.0

Peanut shells Inorganic nitrogen source(NH4)2SO4 CaCO3 NaCl KH2PO4 MgSO4 Soluble starch Peanut meal Soybean meal Melon oil Soybean oil Peanut oil Initial pH Final pH Bioassay method (Total tetracycline equivalent) (mg/g)

L K J I H G F E D C B A

Different combinations of medium items Medium item

Table 6. Effect of combined ingredients on tetracycline production by Streptomyces sp. OXC1 by solid-state fermentation of peanut shells with initial moisture content of 68% at 31
C for 5 days.

Tetracycline production on wastes

113 fermentation. In 1989 Yang & Ling found that S. viridifaciens had an optimal temperature of 26
C for tetracycline production using sweet potato residue. Also Yang & Swei (1996) recorded an optimal temperature of 25–30
C for oxytetracycline production using corn cob. This study also shows that (NH4)2SO4 was the best inorganic nitrogen source. This result was consistent with Yang & Ling’s (1984) result during tetracycline production with sweet potato residue. This study revealed that supplementation with high concentrations of (NH4)2SO4 (1.5%) inhibited tetracycline production. This result is consistent with Yang & Swei’s (1996) finding with oxytetracycline production using corncob. High concentration of nitrogen sources also inhibited b-lactam production by Cephalosporium acremonium (Shen et al. 1984) and spiramycin production by S. ambofaciens (Ahmed et al. 1987). This study showed that combination of inorganic nitrogen [(NH4)2SO4] and organic nitrogen source (peanut meal and soybean meal) enhanced tetracycline production. Yang & Ling (1989), Yang & Yuan (1990) using sweet potato residue observed similar results, while Yang & Swei (1996) observed the same with corn cob. Phosphate had a stimulatory effect when its concentrations was not greater than 0.5% but became inhibitory in higher concentrations. Similar results were obtained in nikkomycin production by S. tendae (Treskatis et al. 1992) and cephamycin production by S. clavuligerus (Ahmed et al. 1987). NaCl stimulated tetracycline production in this study. This could be attributed to the absorption of Cl ions, which are essential for antibiotic production (Yang & Ling 1989). Monosaccharides were good for cell growth, but inhibited the production of tetracycline in this study. Disaccharides and polysaccharide stimulated its production. Yang & Ling (1989) obtained similar results when they observed that a small amount of soluble starch or other fermentable polysaccharide was good for secondary metabolites production. Ifudu (1986a, b) indicated that the use of 0.5% vegetable oils was important in biosynthesis of antibiotics since they functioned as antifoaming agents and as a source of carbon and energy for the microorganism. In this study it was found that 0.25% of peanut oil enhanced tetracycline production. Solid-state fermentation gave a product which was more stable than that from submerged culture, and also required less energy input (Wang 1989). Each gram of substrate produced 13.18 mg of total tetracycline equivalent while the submerged produced 3.05 mg of total tetracycline equivalent. The solid-state product also had an advantage that it could be stored temporarily without significant loss of activity. It is concluded that solid-state fermentation maybe an economic alternative in the production of value added pharmaceuticals and agriculture chemicals in the Nigerian economy. An added advantage of this study was that the local strain isolated was even more effective than the standard strains.

114 References Ahmed, L., Germain, P. & Lefebvre, G. 1987 Phosphate repression of cephamycin and clavulanic acid production by Streptomyces clavuligerus. Applied Microbiology and Biotechnology 26, 130– 135. Baver, L.D. 1956 Soil Physics, 178 pp. 3rd edn. New York: Wiley. Bhatnagar, R.K., Doull, J.L. & Vining, L.C. 1988 Role of the carbon source in regulating chloramphenicol production by Streptomyces venezuelae studies in batch and continuous cultures. Canadian Journal of Microbiology 34, 1217–1233. Huang, J.H. 1972 Antibiotics. pp. 234–268. Taipei, Tafu. Ifudu, N.D. 1986a Indigenous resources for antibiotic production. pp. 27–32. Aug/Sept edition. Expansion Today (Nigeria). Nigeria: AU Press Ltd. Ifudu, N.D. 1986b Indigenous resources for antibiotic production pp. 52–53. Nov/Dec edition. Expansion Today (Nigeria). Nigeria: AU Press Ltd. Komatsu, K.I., Mizuno, M. & Kodaira, R 1975 Effect of methionines on cephalosporin C and penicillin N production by a mutant of Cephalosporium acremonium. Journal of Antibiotics 28, 881– 888. Okami, Y. & Oomura, O. 1979 Production of Antibiotic Substances. Tokyo: Kyoritsu Press Ltd. Pelczar, M.J., Chan, E.C.S. & Krieg, N.R. 1993 Microbiology Concepts and Applications. NY: McGraw-Hill Inc. ISBN 0-07049258-1. Rake, A. 1998 New African Year Book, 12th edn. London, UK: I.C Publication. ISBN 0-90526862-8. Shen, Y.Q., Heim, J., Solomon, N.A. Wolfe, S. & Demain, A.L. 1984 Repression of beta-lactam production in Cephalosporium acremonium by nitrogen sources. Journal of Antibiotics 37, 503– 511. Somerson, N.L. & Phillips, T. 1961 Production of glutamic acid. U.S Patent 3,089,297, 37-1, 695. Treskatis, D.H. King, R., Wolf, H. & Galles, E.D. 1992 Nutritional Control of nikkomycin and juglomycin production by Streptomyces tendae in continuous culture. Applied Microbiology and Biotechnology 36, 440–445.

A.E. Asagbra et al. Wang, H.H. 1989. Utilization of particulate agricultural products through solid state fermentation. Proceedings of the National Science Council, Republic of China, Part B, 13, 145–159. Ward, D.M., Weller, R. & Bateson, M.M. 1990 16SrRNA sequence reveal numerous uncultured microorganisms in a natural community. Nature 345, 63–65. Watve, M.G., Shejval, V., Sonawane, C., Rahalkar, M., Matapurkar, M., Shouche, Y., Patole, M., Phadnis, N., Champhekar, A., Damle, K., Karandikar, S., Kshiragar, V. & Jog, M. 2000 The ‘K’ selected oligiophilic bacteria: a key to uncultured diversity? Current Science 78, 1535–1542. Watve, M.G., Tickoo, R., Jog, M.M. & Bhole, B.D. 2001. How many antibiotics are produced by the genus Streptomyces? Archives of Microbiology 177, 86–90. Yang, S.S. 1977 Quantitative determination of soil gas with regard to soil microbial activities. National Science Council Monthly 5, 478–502. Yang, S.S. 1988 Protein enrichment of sweet potato residue with amylolytic yeasts by solid state fermentation. Biotechnology and Bioengineering 32, 886–890. Yang, S.S. 1993. Protein enrichment of sweet potato residue with coculture of amylolytic fungi by solid state fermentation. Biotechnology Advances 11, 495–505. Yang, S.S. & Cheng, Z.J. 1991. Protein enrichment of corncob with Trichoderma by solid-state fermentation. Chinese Journal of Microbiology and Immunology 24, 177–195. Yang, S.S. & Huang, C.I. 1994 Protease production by amylolytic fungi in solid state germentation. Biotechnology and Bioengineering 32, 886–890. Yang, S.S. & Ling, M.Y. 1989 Tetracycline production with sweet potato residues by solid state fermentation. Biotechnology and Bioengineering 33, 1021–1028. Yang, S.S. & Swei, W.J. 1996 Cultural condition and oxytetracycline production by Streptomyces rimosus in solid state fermentation of corncob. World Journal of Microbiology and Biotechnology 12, 43– 46. Yang, S.S. & Yuan, S.S. 1990 Oxytetracycline production by Streptomyces rimosus in solid state fermentation of sweet potato residue. World Journal of Microbiology and Biotechnology 6, 236–244.

World Journal of Microbiology & Biotechnology 2005 21: 115–121 DOI: 10.1007/s11274-004-2780-5


Springer 2005

Influence of glucose and oxygen on the production of ethyl acetate and isoamyl acetate by a Saccharomyces cerevisiae strain during alcoholic fermentation C. Plata, J.C. Mauricio*, C. Milla´n and J.M. Ortega Department of Microbiology, Faculty of Sciences, University of Cordoba, Campus Universitario de Rabanales, Edificio Severo Ochoa, 14014 Co´rdoba, Spain *Author for correspondence: Tel.:+34-957-218640, Fax: +34-957-218650, E-mail: [email protected] Received 29 January 2004; accepted 15 June 2004

Keywords: Alcohol acetyltransferase, esterases, ethyl acetate, isoamyl acetate, Saccharomyces cerevisiae

Summary The effect of glucose and dissolved oxygen in a synthetic medium simulating the standard composition of grape juice on the production of ethyl acetate and isoamyl acetate by a Saccharomyces cerevisiae strain during alcoholic fermentation was studied. The specific in vitro activity of alcohol acetyltransferase (AATase, EC 2.3.1.84) and esterases (ESase, EC 3.1.1.1; hydrolysis and synthesis of esters) in cell-free extracts was also examined. The specific activity of AATase for ethyl acetate was found to peak at the beginning of the exponential growth phase and that for isoamyl acetate at its end. While the glucose concentration only affected the maximum specific activity of AATase, and only slightly, oxygen inhibited such activity, to a greater extent for isoamyl acetate than for ethyl acetate. On the other hand, esterases were found to catalyse the synthesis of ethyl acetate only at a low or medium glucose concentration (50 or 100 g l)1, respectively), and to reach their maximum hydrolytic activity on isoamyl acetate during the stationary growth phase. The highest ethyl acetate and isoamyl acetate concentrations in the medium were obtained with a glucose concentration of 250 g l)1 and semianaerobic conditions.

Introduction Esters are byproducts of the alcoholic fermentation of sugars by wine yeasts. The factors most strongly affecting the ester content in wine are the particular yeast species and strain, the must composition and the fermentation conditions (Soles et al. 1982; Mauricio et al. 1993; Rojas et al. 2001; Plata et al. 2003). After higher alcohols, esters constitute the family of major compounds accounting to the greatest extent for wine aroma. The ethyl esters of succinic and lactic acids are the most abundant among them, followed by acetates and the ethyl esters of fatty acids (Schreier 1979, 1984; Soles et al. 1982). Among acetates in wine, ethyl acetate is the most abundant and isoamyl acetate that most markedly contributing to wine aroma (Van Der Merwe & Van Wyk 1981). With a given yeast species, their production is governed by the must composition and fermentation conditions (Soles et al. 1982). Saccharomyces cerevisiae, the well known principal wine yeast species, produces esters via an intracellular process that is catalysed by an alcohol acetyltransferase (EC 2.3.1.) using energy provided by the acyl-coenzyme A compounds. The synthesis of acetate esters during fermentation of wine has been widely studied and ascribed to the activity of at least three acetyltransferases (AATase, EC 2.3.1.84), namely: alcohol acetyl-

transferase, ethanol acetyltransferase and isoamyl alcohol acetyltransferase (Lilly et al. 2000). Two distinct AATase activities for isoamyl alcohol and other alcohols have been studied in S. cerevisiae (encoded by ATF1 and ATF2 genes) that exhibit different mechanisms of regulation and, probably, also different physiological roles (Fujii et al. 1996; Lilly et al. 2000; Mason & Dufour 2000). The overexpression of the ATF1 gene in wine yeasts was reported to significantly increase the concentrations of ethyl acetate, ethyl caproate, hexyl acetate, isoamyl acetate and 2-phenylethyl acetate in wine produced by these transgenic microorganisms (Lilly et al. 2000). Recently, analysis of the fermentation products confirmed that the expression levels of ATF1 and ATF2 greatly affect the production of ethyl acetate and isoamyl acetate (Verstrepen et al. 2003). The atf1D atf2D double deletion strain did not form any isoamyl acetate, showing that together, Atf1p and Atf2p are responsible for the total cellular isoamyl alcohol acetyltransferase activity. However, the double deletion strain still produced considerable amounts of certain other esters, such as ethyl acetate (50% of the total concentration produced by the wild-type strain), suggesting the presence of as yet-unknown ester synthases in the yeast proteome (Verstrepen et al. 2003). It has also been shown that S. cerevisiae AATase is strongly repressed under highly aerobic conditions and by the addition of

116 unsaturated fatty acids to the culture medium (Malcorps et al. 1991; Fujii et al. 1997). Esterases (ESase, EC 3.1.1.1.) function mainly by hydrolysing esters; in S. cerevisiae, however, esters can also be synthesized via the reverse reaction in the absence of coenzyme A. In contrast, the relevance attributed to the ester synthetase as an ester-synthesizing activity is rather limited: two esters (ethyl caprylate and ethyl acetate) have been reported as being produced by breadmaking and beer yeast strains, respectively, of S. cerevisiae from ethanol and the respective acids (Rojas et al. 2002). Campbell et al. (1972) succeeded in electrophoretically separating two esterase activities in S. cerevisiae, probably corresponding to the enzymes encoded by the EST1 and EST2 genes identified by Schermers et al. (1976). The latter gene was cloned and sequenced by Fukuda et al. (1996) and its product is a carboxyesterase that hydrolyses mainly isoamyl acetate. Thus, the production of esters is widely believed to rely on the balance of ester synthesis by AATase and ester hydrolysis by ESase (Inoue et al. 1997; Fukuda et al. 1998). In winemaking yeasts, isoamyl acetate is only synthesized in the presence of acetyl-CoA by acetyltransferases; by contrast, ethyl acetate is produced both in the presence of acetyl-CoA (by acetyltransferases) and in the presence of ethanol and acetic acid, by a reverse reaction of esterases (Plata et al. 1998). Both can be hydrolysed in the wine, whether spontaneously or under the action of esterases. The production of esters during alcoholic fermentation depends on the extent of yeast growth and increases during the second half of the growth phase as the synthesis of lipids and sterols stops and the availability of acetyl-CoA increases. After the fermentation ends, the esterase activity increases and the production of esters in the wine decreases during the cell lysis phase (Mauricio et al. 1993). In previous work we studied the potential of various wine yeast species for producing ethyl acetate and isoamyl acetate, which contribute substantially to the aroma of wine (Plata et al. 2003), in a model grape must. In the present work, we examined the influence of the glucose and dissolved oxygen in the must on the activity of alcohol acetyltransferases and esterases in S. cerevisiae, as well as on the production of ethyl acetate and isoamyl acetate during alcoholic fermentation.

C. Plata et al. The total number of cells was determined by counting under a light microscope in a Thoma cell-counting chamber. Culture medium, fermentation conditions and sampling The synthetic fermentation medium used to simulate a standard natural grape juice was that reported by Singh & Kunkee (1976), which was modified by adding 250, 100 or 50 g glucose l)1. Medium containing (g l)1): malic acid, 3; potassium bitartrate, 5; citric acid, 2; casamino acids, 3; K2HPO4, 1; MgSO4Æ7H2O, 0.25; CaCl2Æ2H2O, 0.50; as well as vitamins and trace elements. The medium was adjusted to pH 3.5 with 50% KOH, sterilized by passage through a filter of 0.45 lm pore size . Inoculum was prepared in a flask with cotton plug containing 75 ml culture grown in the same synthetic fermentation medium and incubated at 28 C for 48 h without shaking. Fermentation tests were conducted in 2-l flat-bottomed flasks that were filled with 1800 ml of the abovedescribed synthetic medium, inoculated with 1 · 106 cells ml)1 and incubated at 28 C under different oxygen availability conditions, namely: semiaerobic, semianaerobic and strict anaerobic. Semiaerobic conditions were obtained by continuous shaking at 150 rev min)1 on an orbital shaker from New Brunswick Scientific (Edison, NJ, USA). Semianaerobic conditions were established by allowing the flasks to stand with their mouths stopped by cotton plugs, these conditions are similar to the usually carried out for winemaking. Finally, strict anaerobic conditions were accomplished by using a three-mouthed flask, one mouth being fitted with a special air lock filled with mercury and the other two with two taps connected to vacuum and to a nitrogen gas supply, respectively. The fermentation vessel was previously degassed and then refilled with nitrogen by bubbling for 15–20 min. Samples were collected for analysis at 0, 6, 12, 24, 48, 72 and 240 h of fermentation, and immediately supplied with 20 lg cycloheximide ml)1 and 20 lg chloramphenicol ml)1 to prevent the potential synthesis of proteins upon contact of the cells with oxygen during the later procedure. Each sample was adjusted by cellular recount in a microscope to obtain 1 · 108 to 1 · 109 cells that were collected by centrifugation at 3.500 · g at 4 C for 5 min.

Materials and methods

Analytical procedures

Yeast strain

The supernatant of the fermentation medium was used to determine the amount of ethanol produced, using the method of Crowell & Ough (1979). Reducing sugars were determined enzymatically, using specific kits from Boehringer–Mannheim GmbH (Mannheim, Germany) as recommended by the manufacturer. Isoamyl alcohol, ethyl acetate and isoamyl acetate were extracted with Freon-11 in a continuous extractor for 24 h, and concentrated to 200 ll in a microconcentrator.

The yeast strain used was Saccharomyces cerevisiae E-1 (ATCC Number MYA 425), isolated from spontaneously fermenting must in the Montilla–Moriles designation of origin (southern Spain). The strain was stored in Agar YM (0.3% yeast extract, 0.3% malt extract, 0.5% peptone, 1% glucose, and 2.5% agar; pH 6.5) at 4 C prior to use.

117

Glucose and oxygen effects on ester production 2-Octanol at a 481 lg l)1 concentration was added as an internal standard in all samples. These compounds were quantified by gas chromatography (GC). A volume of 2 ll was injected in the injector, with a split ratio of 60:1, from a Hewlett-Packard 5980 gas chromatograph equipped with a Supelco SP-1000, 60 m · 0.32 mm i.d. fused silica capillary column. The oven temperature programmer was as follows: 9 min at 50 C, followed by a 6 C min)1 ramp to 185 C, which was held for 1 min. The injector and detector temperatures were kept at 275 and 300 C, respectively, and nitrogen at a flow-rate of 26 ml min)1 was used as carrier gas. Determination of enzyme activities The enzyme activities studied were determined as described in a previous paper by the authors (Plata et al. 2003).

Figure 1. Growth kinetic of Saccharomyces cerevisiae strain in the fermentation conditions used in this study: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 of glucose g l)1; (j) anaerobic and 250 g of glucose l)1.

Results and discussion Yeast growth, glucose consumption and ethanol production Figure 1 shows the variation of yeast growth over a period of 10 days under the different fermentation conditions used. As can be seen, the glucose concentration had no substantial effect on yeast growth. On the other hand, as expected, growth was influenced by the available oxygen in the medium; thus, it was maximal under semianaerobic conditions, followed by semiaerobic and, finally, anaerobic conditions, where growth ceases after 12 h presumably due to a lack of essential lipids (Mauricio et al. 1998). Table 1 shows the glucose consumption and ethanol production after 10 days under the operating conditions assayed. In the semianaerobic tests with 50 and 100 g l)1 glucose, the sugar was completely fermented to ethanol by the S. cerevisiae strain used by the end of the experiment. In the same period, with 250 g l)1 glucose, however, the sugar was incompletely fermented to ethanol: to a greater extent under semianaerobic conditions than under semiaerobic and anaerobic conditions. Ester production Figure 2A and B shows plots of the data of AATase specific activities corresponding to the assayed fermenta-

tions. In all cases, the specific activity for the synthesis of ethyl acetate by AATase (Figure 2A) was high at the beginning of fermentation, after which it decreased exponentially (the curve was a hyperbola running along the x-axis and y-axis). The highest value in activity was observed at hour 6 under semiaerobic conditions, but later on diminished more drastically than the rest of the other cases assayed. The anaerobic fermentation tests and those conducted under semianaerobic conditions with 50 or 100 g glucose l)1 showed intermediate values of activity, the lowest values being those obtained with 250 g l)1 under semianaerobic conditions. The specific activity for the synthesis of isoamyl acetate by AATase (Figure 2B) peaked during the exponential growth phase under semianaerobic conditions (12–24 h), and 24–48 h (stationary phase) under anaerobic conditions. This differential behaviour of the kinetics of AATase activity for the synthesis of the two acetates may be consistent with the presence of various AATases (Lilly et al. 2000). Although a number of AATases have been reported for S. cerevisiae, their activity cannot be unequivocally ascribed to a specific acetate ester as the different AATases are somewhat non-specific for the different alcohols. Oxygen was found to inhibit the synthetic activity of AATase for the two acetates, particularly that for isoamyl acetate (see Figure 2). According to Fujiwara et al. (1998), this inhibitory effect is a result of the increased

Table 1. Glucose consumed, ethanol produced and ethanol yield coefficient (g ethanol produced per g glucose consumed) after 10 days of fermentation. Fermentation Conditions

Initial glucose (g l)1)

Glucose consumed (g l)1)

Ethanol produced (% v/v) Ethanol yield coefficient

Semianaerobic Semianaerobic Semianaerobic Semiaerobic Anaerobic

50 100 250 250 250

50.0 99.9 182.2 86.0 46.0

3.30 6.82 11.10 5.25 3.24

0.52 0.53 0.48 0.48 0.55

118

C. Plata et al.

(A) Figure 3. Esterase (ESase) specific activity for ethyl acetate synthesis in the Saccharomyces cerevisiae strain tested during the different fermentation conditions: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 g of glucose l)1; (j) anaerobic and 250 g of glucose l)1.

Figure 2. Alcohol acetyltransferase (AATase) specific activity in the Saccharomyces cerevisiae strain tested during the different fermentation conditions: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 g of glucose l)1; (j) anaerobic and 250 g of glucose l)1. Ethanol/acetyl-CoA or isoamyl alcohol/ acetyl-CoA were, respectively, used as substrates corresponding to the (A) ethyl acetate and (B) isoamyl acetate formed.

synthesis of unsaturated fatty acids and their consequently increased concentration in the plasma membrane, which, according to Fujii et al. (1997) and Yoshimoto et al. (1998), inhibits the transcription of the ATF1 gene. Also, according to Fujiwara et al. (1998), the effect is more marked in synthetic than in naturally rich culture media. By contrast, the initial concentration of glucose had little effect on the synthetic activity of AATase. Esterases were only found to synthesize ethyl acetate in the fermentation tests involving the lower glucose concentrations (viz. 50 and 100 g l)1) and their development was similar in both. The activity of synthesis was high at the beginning of fermentation (6 h) and decreased through the end of the exponential growth phase (24 h), after which it remained at very low levels (Figure 3). Whatever the fermentation conditions, no reverse esterase activity was detected with 250 g glucose l)1: probably because its synthesis was somehow repressed. Also,

µ

(B)

consistent with previous results of Mauricio et al. (1993), no esterase activity for synthesis of isoamyl acetate was detected in S. cerevisiae, whatever the fermentation conditions. Thus, according to Schermes et al. (1976) and Yoshioka & Hashimoto (1981), at low and medium glucose concentrations (50 and 100 g l)1), ethyl acetate synthesis results at least from the action of AATases and the reverse reaction of esterases; on the other hand, ethyl acetate synthesis at high glucose concentrations (250 g l)1) and isoamyl acetate synthesis results from the action of AATases only. Figures 4 and 5 show the concentrations of each acetate measured in the different fermentation tests. Under semianaerobic conditions, the acetate concentra-

Figure 4. Ethyl acetate concentration in the medium during the different fermentation conditions by Saccharomyces cerevisiae strain: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 g of glucose l)1; (j) anaerobic and 250 g of glucose l)1.

µ

Glucose and oxygen effects on ester production

Figure 5. Isoamyl acetate concentration in the medium during the different fermentation conditions by Saccharomyces cerevisiae strain: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 g of glucose l)1; (j) anaerobic and 250 g of glucose l)1.

µ

tion was proportional to the initial glucose concentration and, as found in previous work (Plata et al. 2003), directly related to the production of the corresponding alcohols (Table 1, Figure 6). The anaerobic and semiaerobic conditions provided lower acetate concentrations; although the corresponding alcohols were also produced in smaller amounts, the origin of the decrease was different (viz. inhibition of AATase activity under semiaerobic conditions and delayed or halted fermentation under anaerobic conditions). In previous work, we investigated the relationship of AATase and esterases to the concentration of acetate esters in two S. cerevisiae strains during the fermentation of grape must. After 24 h of fermentation, ethyl, isoamyl and hexyl acetates were found to be produced in

Figure 6. Isoamyl alcohol concentration in the medium during the different fermentation conditions by Saccharomyces cerevisiae strain: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 g of glucose l)1; (j) anaerobic and 250 g of glucose l)1.

119 large amounts, consistent with an increased AATase activity (Mauricio et al. 1993). The new findings in the synthetic medium used in this work are also consistent with these results. Thus, the in vitro specific activities of the enzymes correlate with the production of ethyl acetate. The specific production rate (Figure 7A) peaked at the beginning of fermentation, coinciding with a high AATase activity (Figure 2A). Most of the ethanol produced at that stage was probably converted into ethyl acetate by AATase and esterases in order to detoxify cells (Leao & Van Uden 1985; Pampulha & Loureiro-Diaz 1990). After 12 h of fermentation, the enzymatic activity dropped in the fermentation with media initially containing low and intermediate glucose concentrations (Figure 2A); this decreased the production of ethyl acetate. By contrast in the fermentations with 250 g glucose l)1, the activity was higher (particularly under semianaerobic conditions), and so was the production of ethyl acetate as a result. In the stationary phase (after 72 h), the enzymatic activity was higher under anaerobic conditions, as reflected in a higher specific ester production rate under these conditions (Figure 7A); this was possibly a consequence of the increased availability of acetyl-CoA resulting from the absence of dissolved oxygen in the medium. This would give rise to a low metabolic demand of yeasts for the production of unsaturated fatty acids and sterol, and acetyl-CoA not used for this purpose would be available for alternative processes such as the synthesis of acetate esters. However, the acetate production was greatest under semianaerobic conditions, possibly because the amount of ethanol produced at this stage is usually the factor governing the synthesis of ethyl acetate: particularly once a high concentration is reached (Minetoki 1992). The production of isoamyl acetate was also influenced by the enzyme activity. Thus, the highest specific production rate (Figure 7C) was reached immediately after the enzyme activity peaked under semianaerobic conditions; this coincided with the highest production rate for isoamyl alcohol (Figure 7B). The decreased production under semiaerobic conditions was a result of AATase activity probably being inhibited by the oxygen present, as it seems unlikely that isoamyl alcohol, the concentration of which was very close to that reached under semianaerobic conditions (Figure 6), may have been the limiting factor. The AATase-encoding gene is not transcribed in the presence of oxygen or unsaturated fatty acids (Fujii et al. 1997). Also, acetyl-CoA is less available for ester synthesis under these conditions, due to the great metabolic demands of the yeast for this coenzyme. Unlike Nyka¨nen (1986), we found that anaerobic conditions do not cause an increase of the synthesis of isoamyl acetate, but rather a decrease in relation to the semianaerobic conditions; this finding is consistent with our own results in previous work (Mauricio et al. 1997), which revealed that ester production was maximal under semianaerobic conditions (particularly during the

120

C. Plata et al.

(A)

.

.

.

.

.

(B)

The extent to which esters accumulate in the wine depends on the relationship between the synthetic activity of AATase and the hydrolytic activity of esterases (Wakai et al. 1990). Usually, the ester concentration decreases near the end of fermentation due to the effect of the strong action of hydrolases (Mauricio et al. 1993; Inoue et al. 1997; Fukuda et al. 1998). The hydrolytic activity of esterases for isoamyl acetate (Figure 8) peaked at the beginning of the stationary growth phase and then decreased throughout the process under all fermentation conditions. An equilibrium was reached between the synthetic and hydrolytic activities for this ester, as a result of which its concentration remained constant through the end of fermentation. Ester production is influenced by a number of factors, the most important of which are the conditions used during the fermentation process (Mallouchos et al. 2003). In fact, the production of the acetate esters studied varied with the fermentation conditions; thus, the highest concentrations of ethyl acetate and isoamyl acetate were obtained with a high glucose concentration (250 g l)1) under semianaerobic conditions and the greatest production of both occurred during the stationary growth phase. In addition, the production of these acetate esters may be influenced by the activity of the corresponding enzymes (viz. AATase and esterases) and by the availability of the corresponding alcohols: the highest concentrations of which were also obtained under the same conditions. The last is in accordance with some studies that have shown a certain degree of correlation between substrate concentrations and corresponding ester formation, and it has been generally accepted that substrate concentrations control ester formation (Thurston et al. 1982; Calderbank & Hammond 1994; Younis & Stewart 2003).

(C) Figure 7. Ethyl acetate (A), isoamyl alcohol, (B) and isoamyl acetate, (C) specific production rates during the different fermentation conditions by Saccharomyces cerevisiae strain: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 g of glucose l)1; (j) anaerobic and 250 g of glucose l)1.

exponential growth phase), and with those of Calderbank & Hammond (1994), who found low oxygen levels (0.75 mg l)1) to delay cell growth and reduce ester production.

Figure 8. Esterase (ESase) specific activity of Saccharomyces cerevisiae strain during the different fermentation conditions for isoamyl acetate hydrolysis: (d) semianaerobic and 250 g of glucose l)1; ( ) semianaerobic and 100 g of glucose l)1; (.) semianaerobic and 50 g of glucose l)1; (,) semiaerobic and 250 g of glucose l)1; (j) anaerobic and 250 g of glucose l)1.

Glucose and oxygen effects on ester production Acknowledgements This work was supported by a grant from the Government of Spain. Project VIN 00-039-C2-01. References Calderbank, J. & Hammond, J.R.M. 1994 Influence of higher alcohol availability on ester formation by yeast. Journal of the American Society of Brewing Chemists 52, 84–90. Campbell, Y., Gilmour, R.H. & Rous, P.R. 1972 Differentiation of yeasts by electrophoretic methods. Journal of the Institute of Brewing 78, 491–496. Crowell, E.A. & Ough, C.S. 1979 A modified procedure for alcohol determination by dichromate oxidation. American Journal of Enology and Viticulture 30, 61–63. Fujii, T., Yoshimoto, H. & Tamai, Y. 1996 Acetate ester production by Saccharomyces cerevisiae lacking the ATF1 gene encoding the alcohol acetyltransferase. Journal of Fermentation and Bioengineering 81, 538–542. Fujii, T., Kobayashi, O., Yoshimoto, H., Furukawa, S. & Tamai, Y. 1997 Effect of aeration and unsaturated fatty acids on expression of the Saccharomyces cerevisiae alcohol acetyltransferase gene. Applied and Environmental Microbiology 63, 910–915. Fujiwara, D., Yoshimoto, H., Sone, H., Harashima, S. & Tamai, Y. 1998 Transcriptional coregulation of Saccharomyces cerevisiae alcohol acetyltransferase gene, ATF1 and n-9 fatty acid desaturase gene, OLE1 by unsaturated fatty acids. Yeast 14, 711–721. Fukuda, K., Kuwahata, O., Kiyokawa, Y., Yanagiuchi, T., Wakai, Y., Kitamoto, K., Inoue, Y. & Kimura, A. 1996 Molecular cloning and nucleotide sequence of the isoamyl acetate hidrolyzing esterase gene (EST2) from Saccharomyces cerevisiae. Journal of Fermentation and Bioengineering 81, 8–15. Fukuda, K., Yamamoto, N., Kiyokawa, Y., Yanagiuchi, T., Wakai, Y., Kitamoto, K., Inoue, Y. & Kimura, A. 1998 Balance of activities of alcohol acetyltransferase and esterase in Saccharomyces cerevisiae is important for production of isoamyl acetate. Applied and Environmental Microbiology 64, 4076–4078. Inoue, Y., Trevanichi, S., Fukuda, K., Izawa, S., Wakai, Y. & Kimura, A. 1997 Roles of esterase and alcohol acetyltransferase on production of isoamyl acetate in Hansenula mrakii. Journal of Agricultural and Food Chemistry 45, 644–649. Leao, C. & Van Uden, N. 1985 Effects of ethanol and other alkanols on the relations of glucose transport and fermentation in Saccharomyces cerevisiae. Applied Microbiology and Biotechnology 22, 359–363. Lilly, M., Lambrechts, M.G. & Pretorius, I.S. 2000 Effect of increased yeast alcohol acetyltransferase activity on flavor profiles of wine and distillates. Applied and Environmental Microbiology 66, 744–753. Malcorps, P., Cheval, J.M., Jamil, S. & Dufour, J.P. 1991 A new model for the regulation of ester synthesis by alcohol acetyltransferase in Saccharomyces cerevisiae during fermentation. Journal of the American Society of Brewing Chemists 49, 47–53. Mallouchos, A., Komaitis, M., Koutinas, A. & Kanellaki, M. 2003 Wine fermentations by immobilized and free cells at different temperatures. Effect of immobilization and temperature on volatile by-products. Food Chemistry 80, 109–113. Mason, A.B. & Dufour, J.P. 2000 Alcohol acetyltransferase and the significance of ester synthesis in yeast. Yeast 16, 1287–1298. Mauricio, J.C., Moreno, J., Valero, E., Zea, L., Medina, M. & Ortega, J.M. 1993 Ester formation and specific activities of in vitro alcohol acetyltransferase and esterase by Saccharomyces cerevisiae during grape must fermentation. Journal of Agricultural and Food Chemistry 41, 2086–2091. Mauricio, J.C., Moreno, J., Zea, L., Ortega, J.M. & Medina, M. 1997 The effects of grape must fermentation conditions on volatile alcohols and esters formed by Saccharomyces cerevisiae. Journal of the Science of Food and Agriculture 75, 155–160.

121 Mauricio, J.C., Milla´n, C. & Ortega, J.M. 1998 Influence of oxygen on the biosynthesis of cellular fatty acids, sterols and phospholipids during alcoholic fermentation by Saccharomyces cerevisiae and Torulaspora delbrueckii. World Journal of Microbiology and Biotechnology 14, 405–410. Minetoki, T. 1992 Alcohol acetyltransferase of sake yeast. Journal of Brewing Society of Japan 87, 334–340. Nyka¨nen, L. 1986 Formation and occurrence of flavour compounds in wine and distilled alcoholic beverages. American Journal of Enology and Viticulture 37, 84–96. Pampulha, M.E. & Loureiro-Diaz, M.C. 1990 Activity of glycolitic enzymes of Saccharomyces cerevisiae in the presence of acetic acid. Applied Microbiology and Biotechnology 34, 375–380. Plata, M.C., Mauricio, J.C., Milla´n, C. & Ortega, J.M. 1998 In vitro specific activity of alcohol acetyltransferase and esterase in two flor yeast strains during biological aging of Sherry wines. Journal of Fermentation and Bioengineering 85, 369–374. Plata, M.C., Mauricio, J.C., Milla´n, C. & Ortega, J.M. 2003 Formation of ethyl acetate and isoamyl acetate by various species of wine yeasts. Food Microbiology 20, 217–224. Rojas, V., Gil, J.V., Pin˜aga, F. & Manzanares, P. 2001 Studies on acetate ester production by non-Saccharomyces wine yeast. International Journal of Food Microbiology 70, 283–289. Rojas, V., Gil, J.V., Manzanares, P., Gavara, R., Pin˜aga, F. & Flors, A. 2002 Measurement of alcohol acetyltransferase and ester hydrolase activities in yeast extracts. Enzyme and Microbial Technology 30, 224–230. Schermers, F.H., Duffus, J.H. & McLeod, A.M. 1976 Studies on yeast esterase. Journal of the Institute of Brewing 82, 170–184. Schreier, P. 1979 Flavor composition of wines: a review. CRC Critical Reviews in Food Science and Nutrition 12, 59–111. Schreier, P. 1984 Formation of wine aroma. In Proceding of Alko Symposium of Flavour Research of Alcoholic Beverages – Instrumental and sensory Analysis, eds. Nyka¨nen, L. & Lehtonen, P. pp. 9–37. Helsinki: Foundation for Biotechnical and Industrial Fermentation Research. Singh, R. & Kunkee, R. 1976 Alcohol dehydrogenase activities of wine yeasts in relation to higher alcohol formation. Applied and Environmental Microbiology 32, 666–670. Soles, R.M., Ough, C.S. & Kunkee, R.E. 1982 Ester concentration differences in wines fermented by various species and strains of yeasts. American Journal of Enology and Viticulture 33, 94–98. Thurston, P.A., Quain, D.E. & Tubb, R.S. 1982 Lipid metabolism and the regulation of volatile synthesis in Saccharomyces cerevisiae. Journal of the Institute of Brewing 88, 90–94. Van Der Merwe, C.A. & Van Wyk, C.J. 1981 The contribution of some fermentation products to the odour of dry white wines. American Journal of Enology and Viticulture 32, 41–46. Verstrepen, K.J., Van Laere, S.D.M., Vanderhaegen, B.M.P., Derdelinckx, G., Dufour, J.P., Pretorius, I.S., Winderickx, J., Thevelein, J.M. & Delvaux, F.R. 2003 Expression levels of the yeast alcohol acetyltransferase genes ATF1, Lg-ATF1, and ATF2 control the formation of a broad range of volatile esters. Applied and Environmental Microbiology 69, 5228–5237. Wakai, Y., Yanagiuchi, T. & Kiyokawa, Y. 1990 Properties of an isoamyl acetate hydrolytic enzyme from sake yeast strain. Hakkokogaku 68, 101–105. Yoshimoto, H., Fujiwara, D., Momma, T., Ito, C., Sone, H., Kaneko, Y. & Tamai, Y. 1998 Characterization of the ATF1 and Lg-ATF1 genes encoding alcohol acetyltransferases in the bottom fermenting yeast Saccharomyces pasteurianus. Journal of Fermentation and Bioengineering 86, 15–20. Yoshioka, K. & Hashimoto, H. 1981 Ester formation by alcohol acetyltransferase from brewer’s yeast. Agricultural and Biological Chemistry 45, 2183–2190. Younis, O.S. & Stewart, G.G. 2003 The effect of worth maltose content on volatile production and fermentation performance in brewing yeast. In Brewing Yeast Fermentation Performance, ed. Smart, K. vol. 1, pp. 170–176. Oxford, United Kingdom: Blackwell Science. ISBN 0632064986.

World Journal of Microbiology & Biotechnology 2005 21: 123–125 DOI: 10.1007/s11274-004-3045-z

 Springer 2005

Rapid screening of Aspergillus terreus mutants for overproduction of lovastatin M.A. Vilches Ferro´n1, J.L. Casas Lo´pez1,*, J.A. Sa´nchez Pe´rez1, J.M. Ferna´ndez Sevilla1 and Y. Chisti2 1 Department of Chemical Engineering, University of Almerı´a, 04120 Almerı´a, Spain 2 Institute of Technology and Engineering, Massey University, Palmerston North, New Zealand *Author for correspondence: Tel.: +34-950-015314; Fax: +34-950-015484; E-mail: [email protected] Received 26 February 2004; accepted 21 June 2004

Keywords: Aspergillus terreus, Candida albicans, lovastatin, mutation, screening

Summary A novel rapid screening method is demonstrated for isolating lovastatin-overproducing strains of Aspergillus terreus. The screening methodology, based on the activity of lovastatin against the yeast Candida albicans, is nearly three times as fast as the selection methods used earlier. The new 6-h assay shows a linear correlation between the quantity of lovastatin generated by A. terreus isolates and the inhibition zones obtained on plates of C. albicans. The new technique is less expensive and requires less labour.

Introduction Lovastatin is a potent cholesterol-lowering drug. Lovastatin acts by competitively inhibiting the enzyme 3hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA) which catalyses the rate-limiting step of cholesterol biosynthesis. Lovastatin is produced as a secondary metabolite by a variety of filamentous fungi, including Penicillium sp. (Endo et al. 1976a), Monascus ruber (Endo et al. 1976b; Juzlova et al. 1996), and Aspergillus terreus (Alberts et al. 1980). Commercial production of lovastatin uses A. terreus batch fermentation which has been investigated extensively (Novak et al. 1997; Manzoni et al. 1998, 1999; Szaka´cs et al. 1998; Kumar et al. 2000a). Random mutation and selection strategies have been reported for obtaining overproducing isolates of A. terreus (Vinci et al. 1991). A particularly useful rapid screening method for identifying lovastatin overproducers was described by Kumar et al. (2000b). The method was based on the anti-fungal properties of the bhydroxyacid form of lovastatin and its inhibitory effect on the mycelial fungus Neurospora crassa. This bioassay required 16–18 h of plate incubation to measure zones of inhibition. This paper describes a new method for screening for overproduction of lovastatin. Activity of lovastatin against the yeast Candida albicans is used as the basis for this method. Zones of inhibition proportional to lovastatin concentration were generated on agar plates that had been previously inoculated with C. albicans. This new 6-h assay allowed relatively rapid identification of lovastatin-overproducing strains of A. terreus. A mutation-selection program was carried out with

A. terreus ATCC 20542. After two mutation runs, a 4fold enhancement in production of lovastatin was achieved compared to the original strain.

Materials and methods Microorganism and inoculation The fungus used was obtained from the American Type Culture Collection, as Aspergillus terreus ATCC 20542. A suspension of spores was prepared by washing petri dish (potato dextrose agar, PDA) cultures with a sterile aqueous solution of 2% Tween
20. The spore concentration was determined spectrophotometrically at 360 nm by reference to a standard curve that had been made by direct counts (Coulter EPICS XL-MCL flow cytometer) of spores in suspension. Mutation with EMS Aspergillus terreus spores were exposed to methanesulphonic acid ethyl ester (EMS) at three concentrations (0.15, 0.3 and 0.5 M) for four durations of exposure (30, 60, 90, and 120 min). The spores were grown for 4 days on PDA slants. Colonies were collected in agar plugs (3 mm diameter · 7 mm height) when more than 90% mortality had been attained in a dish. Each agar plug was transferred to an Eppendorf tube containing 500 ll of the lactose-yeast-extract medium and incubated at 28 C for 5 days. This medium consisted of (per litre): 120 g lactose, 1.85 g yeast extract, 2 g KH2PO4, 0.52 g MgSO4 Æ 7H2O, 0.40 g NaCl, 2 mg Fe(NO3)3 Æ 9H2O, 1 mg ZnSO4 Æ H2O, 0.04 mg biotin and 1 ml of a trace

124

Bioassay with Candida albicans After incubation, A. terreus colonies from one of the duplicate dishes (A) were isolated in agar plugs punched using a sterile 10-mm stainless steel cork borer and transferred to a screw-capped test tube. One millilitre of ethyl acetate was added for lovastatin extraction at 50 C (15 min; with vortex agitation at 2 min intervals). The lovastatin extract was recovered by centrifugation (2800 · g, 5 min). The other slant (B) was stored at 4– 6 C for further culture. Candida albicans was grown for 12 h on PDA dishes at 28 C, then harvested and transferred at a concentration of 5–7 · 102 cells per ml on fresh PDA slants. Fifty microliters of lovastatin extract were transferred onto a 6 mm diameter paper disk and placed on the surface of a 90 mm diameter C. albicans plate. The spacing between lovastatin impregnated disks on a plate was at least 15 mm. Negative standard disks were prepared with ethyl acetate alone. Positive standard discs were made by impregnating the paper with 50 ll of a solution of known concentration of lovastatin in ethyl acetate. The plates were incubated for 6 h and zones of inhibited growth were recorded. A large diameter of the inhibition zone indicated a high titre of lovastatin. Lovastatin analysis Lovastatin was quantified as its b-hydroxyacid form, by HPLC. Because the fungus secretes lovastatin in the b-hydroxyacid form, the assay eliminated the conversion step to the active lactone form of the drug. Using the b-hydroxyacid permitted rapid analysis because this form elutes earlier from a chromatography column than does the lactone form of lovastatin. Also, the b-hydroxyacid is quite stable in solution. HPLC was performed on a Beckman Ultrasphere ODS (250 · 4.6 mm I.D., 5 lm support diameter) column. A diode array detector was used. The eluent was acetonitrile/0.1% phosphoric acid (60:40, by vol). The eluent flow rate was 1.5 ml/min. The detection wavelength was 238 nm. The sample injection volume was 20 ll.

ously described. The culture lasted up to 7 days. The flasks were inoculated with 1 ml of a spore suspension which had been standardized to contain 106 spores/ml.

Results and discussion In submerged cultures of C. albicans, addition of the bhydroxyacid form of lovastatin caused growth inhibition at lovastatin concentrations greater than 0.06 g/l. On solid medium, when a given amount of lovastatin was placed on the agar surface using a paper disk, zones of no growth were obtained on plates of C. albicans. The diameter of the inhibition zones correlated linearly (Figure. 1) with the quantity of lovastatin impregnated in the paper disc, as follows: Inhibition zone diameter (mm) ¼ 0:195
lovastatin dose ðlgÞ The value of the slope in the above equation will depend on the nature of the agar gel. The slope will be lower for gels containing a high concentration of agar because the agar content is well known to affect the rate of diffusion in the gel. Preliminary experiments showed a 10–20% variation in the inhibition zone diameters for different batches of agar gel. Considering this, a positive standard with a given amount of b-hydroxyacid lovastatin was placed on each plate and the measured diameters of inhibition zones were normalized relative to this standard for comparing with the calibration curve shown in Figure 1. After establishing the bioassay protocol, two twostage mutation runs were carried out starting from the parent strain (A. terreus, ATCC 20542). Although in routine work only the high-yielding isolates (plate B) would be cultured in a liquid medium to assess the lovastatin production in submerged fermentation, in this study all the isolates were grown in shake flasks to illustrate the feasibility of the proposed methodology.

12

Inhibition zone diameter (mm)

element solution. The trace element solution contained (per l of solution): Na2B4O7 Æ 10H2O, 100 mg; MnCl2 Æ 4H2O, 50 mg; Na2MoO4 Æ 2H2O, 50 mg and CuSO4 Æ 5H2O, 250 mg. The pH of the medium was adjusted to 5.5 before sterilization. Spores were harvested and plated at a concentration of 106 spores per ml on duplicate fresh petri dishes with the same medium that had been solidified with 15 g of agar/l. The plates were incubated for 7 days at 28 C.

M.A. Vilches Ferro´n et al.

10

8

6

4

y = 0.195x

2

r2 = 0.963

Shake flask cultures 0

All fermentations were carried out at 28 C in 100 ml shake flasks filled with 20 ml of the medium and held on a rotary platform shaker (150 rev/min, 2.6 cm stroke). The culture medium used was the lactose-yeast-extract previ-

0

10

20

30

40

50

60

Lovastatin dose (µg)

Figure 1. Standard curve for dose of the b-hydroxyacid form of lovastatin vs. inhibition zone diameter on C. albicans plates.

125

Isolation of lovastatin-overproducing strains

of A. terreus in a significantly shorter period compared to the best existing methods. Also, the new technique was less expensive and required less labour. Currently, we have a substantial mutation/selection program underway for isolating better overproducer strains than those obtained already.

6

Inhibition zone diameter (mm)

5

CONTROL

4 E2B4 E2F6 WT E2J2 E2E9 E2A5

3

2

E2J8

E2K6

Acknowledgement

1

0 0

5

(a)

10

15

20

25

30

Lovastatin concentration (mg/L)

This research was supported by the Ministerio de Ciencia y Tecnologı´ a (MYCT) and FEDER (PPQ2000-0032-P402), Spain.

14

CONTROL

Inhibition zone diameter (mm)

12

References

E354

10 E122 8

E123

6

E351 E352

4 WT 2

E121

E124

0 0

(b)

20

40

60

80

Lovastatin concentration (mg/L)

Figure 2. Inhibition zone diameter on plates of C. albicans vs. lovastatin concentration (analysed by HPLC) from submerged fermentation of A. terreus isolates (duplicate cultures): (a) first stage mutation run; (b) second stage mutation run.

The first run of mutagenesis gave rise to mutants with a wide range of lovastatin production titres (Figure 2a). The parent culture displayed a lovastatin production level that was midway compared to the range for the isolates. There was a positive correlation (correlation coefficient >0.98) between the measured diameter of inhibition zone on agar (plate A) and the lovastatin titre obtained in liquid cultures of the isolate (plate B). The isolate E2B4 was selected for a second mutagenesis run (Figure 2b). As expected, most of isolates from this second mutagenesis stage gave lovastatin titres higher than that of the parent culture (ATCC 20542). The highest titre (60.3 mg/l), obtained with the isolate E354, was four times the lovastatin production level of the original culture. Conclusions The screening methodology demonstrated in this study permitted isolation of lovastatin overproducer mutants

Alberts, A.W., Chen, J., Kuron, G., Hunt, V., Huff, J., Hoffman, C., Rothrock, J., Lopez, M., Joshua, H., Harris, E., Patchett, A., Monaghan, R., Currie, S., Stapley, E., Albers-Schonberg, G., Hensens, O., Hirshfield, J., Hoogsteen, K., Liesch, J. & Springer, J. 1980 Mevinolin: a highly potent competitive inhibitor of hydroxymethylglutaryl-coenzyme A reductase and a cholesterol-lowering agent. Proceedings of the National Academy of Sciences of the USA 77, 3957–3961. Endo, A., Kuroda, M. & Tsujita, Y. 1976a ML-236 A, ML236 B and ML-236 C, new inhibitors of cholesterogenesis produced by Penicillium citrinum. Journal of Antibiotics 29, 1346– 1348. Endo, A., Kuroda, M. & Tanazawa, K. 1976b Competitive inhibition of 3-hydroxy-3-methyl glutaryl coenzyme A reductase by ML-236 A and ML236 B fungal metabolites having hypocholesterolemic activity. FEBS Letters 72, 323–326. Juzlova, P., Martinkova, L. & Kren, V. 1996 Secondary metabolites of the fungus Monascus: a review. Journal of Industrial Microbiology 16, 163–170. Kumar, M.S., Jana, S.K., Senthil, V., Shashanka, V., Kumar, S.V. & Sadhukhan, A.K. 2000a Repeated fed-batch process for improving lovastatin production. Process Biochemistry 36, 363–368. Kumar, M.S., Kumar, P.M., Sarnaik, H.M. & Sadhukhan, A.K. 2000b A rapid technique for screening of lovastatin-producing strains of Aspergillus terreus by agar plug and Neurospora crassa bioassay. Journal of Microbiological Methods 40: 99–104. Manzoni, M., Rollini, M., Bergomi, S. & Cavazzoni, V. 1998 Production and purification of statins from Aspergillus terreus strains. Biotechnology Techniques 12, 529–532. Manzoni, M., Bergomi, S., Rollini, M. & Cavazzoni, V. 1999 Production of statins by filamentous fungi. Biotechnology Letters 21, 253–257. Novak, N., Gerdin, S. & Berovic, M. 1997 Increased lovastatin formation by Aspergillus terreus using repeated fed-batch process. Biotechnology Letters 19, 947–948. Szaka´cs, G., Morovja´n, G. & Tengerdy, R.P. 1998 Production of lovastatin by a wild strain of Aspergillus terreus. Biotechnology Letters 20, 411–415. Vinci, V.A., Hoerner, T.D., Coffmann, A.D., Schimmel, T.G., Dabora, R.L., Kirpeker, A.C., Ruby, C.L. & Stieber, R.W. 1991 Mutants of a lovastatin hyperproducing Aspergillus terreus deficient in the production of sulochrin. Journal of Industrial Microbiology 8, 113–120.

World Journal of Microbiology & Biotechnology 2005 21: 127–133 DOI: 10.1007/s11274-004-3043-1


Springer 2005

Impact of balanced substrate flux on the metabolic process employing fuzzy logic during the cultivation of Bacillus thuringiensis var. Galleriae R.K.I. Anderson1,* and Kunthala Jayaraman2 1 TICEL Bio Park Ltd., Taramani Road, Taramani, Chennai - 600 113, India 2 Department of Biotechnology, Vellore Institute of Technology, Vellore - 632 014, India *Author for correspondence: Tel.: +91-44-22542061, Fax: +91-44-22542055, E-mail: [email protected] Received 23 December 2003; accepted 21 June 2004

Keywords: Bacillus thuringiensis subsp. galleriae, fed-batch, fuzzy logic control, insecticidal crystal protein, LabVIEW, online substrate monitoring, process optimization

Summary The insecticidal crystal protein (ICP) synthesized at the onset of sporulation by Bacillus thuringiensis var. galleriae (Btg) is lethal against specific pests. Attempts were made to enhance the synthesis of biomass and ICP by Btg employing process optimization strategies. The process optimization was carried out with residual glucose concentration control in a bench scale bioreactor. A fuzzy logic-based feedback control system for maintaining the residual glucose concentration at a constant level during cultivation was developed in LabVIEW. This control system indicated the possibilities in providing a balanced substrate flux during cultivation. The identified optimum level of 2.72 g/l in residual glucose concentration was maintained by fed-batch cultivation with glucose and yeast extract fed at equal concentration with the above control system. High cell density of 16.0 g/l with specific growth rate of 0.69 h)1 was obtained during the cultivation. The balanced flux of substrate during cultivation has resulted in the enhanced synthesis of biomass and ICP. This optimized process could be commercially exploited by comparing the fluxes of basal compounds in different media sources used in fermentation.

Introduction Effective biotechnological process development depends on accurate on-line monitoring of the metabolic state of the culture during fermentation. Identification of a common process parameter and development of a flexible monitoring system suitable for any bioprocess is gaining importance in biotechnology. With advancement in research, measuring elements such as biosensors and analytical equipments interfaced with computers provided a solution for continuous monitoring in fermentation processes. Control of Saccharomyces cerevisae concentration using a laser turbidometer (Cho & Chang 1995) and continuous pH monitoring in a perfusion bioreactor using an optical pH sensor (Jeevarajan et al. 2002) were reported. Controlled environmental conditions in terms of nutrients play a crucial role in the regulation of growth and product formation by microorganisms. Reliable mathematical models for describing bioprocesses, appropriate online sensors for detecting cellular growth information and state of the bioprocess (Feng & Glassey 2000) and knowledge regarding the relationships between controlled process variables and the process outputs are necessary for the development of reliable control systems for fermentation applications.

Knowledge-based control of fermentation processes has resulted in the solution to control problems associated with uncertainties and non-quantitative knowledge of biochemical systems (Guthke et al. 1998; Suzuki et al. 2000). This approach provides flexibility to handle bioprocess information, in both mathematical and linguistic forces, into a computer implementable system. A fuzzy logic controller (FLC) is one of the simple architectures of knowledge-based control systems (Konstantinov & Yoshida 1992). Further, fuzzy logic in combination with neural networks (Shioya et al. 1999; Ronen et al. 2002) and genetic algorithm (Ranganath et al. 1999) has become a powerful tool for the control of biochemical processes. Fuzzy logic finds extensive application in microbial and mammalian cell cultivation. A prototype neural-network-supervised control system for Bacillus thuringiensis fermentation has been reported with manipulations in pH and cultivation temperature for maintaining constant specific growth rate (Zhang et al. 1994). The fuzzy data in the neural network showed promising results in improving representation and learning accuracy. A recurrent trainable neural network was also proposed to be used as a process predictor in a predictive control system during the fermentation of Bacillus thuringiensis subsp. kurstaki (Valdez et al. 2003). Dynamic parameter adjustment

128

R.K.I. Anderson and K. Jayaraman

with fuzzy logic was carried out for optimizing the feed rates for hybridoma growth in a bioreactor (Dhir et al. 2000). The wide application of fuzzy logic in biochemical systems for providing a controlled environment facilitated a condition towards understanding the flux of substrate(s) in process optimization.

surement was interfaced with the computer and used for the online monitoring of residual glucose concentration in the bioreactor. The analyser facilitates the programming of sampling interval, purging time, sample flow rate and calibration time as 5, 1 min, 2500 ll/min and every two samples respectively.

Materials and methods

Filtration system

Microorganism and media

Cell separation was achieved using a tangential flow filtration module (SPIROSEP 90TM). The exposed membrane area of the filter was 45 cm2 and it was made from polyethersulphone packed with a non-woven polypropylene matrix of 0.22 lm pore size. The retentate from the module was recycled back into the bioreactor. During sampling, permeate flow was provided by means of a peristaltic pump in the analyser.

Bacillus thuringiensis var. galleriae (Btg) was obtained from the laboratory strain collection of the Centre for Biotechnology, Anna University, Chennai, India and maintained on nutrient broth (NB) agar plates at 4 C. Medium composition, nutrient broth (NB), glucose yeast extract (GY) medium (Sachidanandham & Jayaraman 1993) and Glucose–YE–Aminoacids medium (GYA) (Sachidanandham et al. 1996). Feed solutions for fed-batch cultivation were glucose: 80.0 g/l; yeast extract: 80.0 g/l and ammonium sulphate: 13.6 g/l.

Computer and accessories

10 ml of NB medium was inoculated with a single colony of Bacillus thuringiensis var. galleriae. Exponentially growing culture at 30 C for 6 h at 125 rev/min in a shaker (ORBITEK) was used as inoculum. Microscopic observation was carried out periodically to monitor the sterility of the culture.

The feed pump (Matson Warlow 503U) and the biochemistry analyser were interfaced with the computer (Digital Celebris GL 5133) through the DAQ channel board (AT-MIO-16DE-10, National Instruments) and interfacing hardware (NB-MID-32X, National Instruments). Data acquisition from the bioreactor was carried out with the above hardware. Graphical programming for Instrumentation, LabVIEW (National Instruments) was used for the development of data acquisition and control application programs.

Cultivation conditions

Off-line analytical methods

Cultivations were carried out in 1.5-l bioreactor (Bioengineering AG, Wald, Switzerland) with working volume of 1.0 l. During cultivation, pH and temperature was maintained at 7.0 and 30 C respectively. Cultivation was started with 500 rpm as agitation and 1 vvm as aeration. Dissolved oxygen concentration in the bioreactor was maintained above 30% of saturation by adjusting aeration and agitation manually. 10% polypropylene glycol (PPG) was used for controlling the foam during cultivation.

Biomass concentration in terms of optical density was measured using spectrophotometer (Hitachi U2000). Cell dry weight was measured and correlated with the optical density (O.D.) by, Cell density, g/l ¼ 0.1933 · O.D + 0.0444 [R2: 0.9963]. The offline analysis of residual glucose concentration was carried out in the biochemistry analyser (Yellow Springs Instruments Inc.). Spore concentrations were estimated from the colonies developed on NB agar plates by the heattreated (70 C for 10 min) culture on incubation. The total protein was estimated according to the Lowry method. The insecticidal crystal protein content was estimated using enzyme-linked immunosorbent assay (ELISA)-based quantification method (Smith & Ulrich 1983).

Inoculum development

Fed-batch operation Btg was cultivated in GY media with constituents concentration as mentioned in Results and Discussion. The feeding was done through the control system to maintain the residual glucose concentration at specified levels. In addition, yeast extract was pulsed three times at 10-min interval for 0.25 min during the cultivation resulting in the addition of 0.82 g. In addition, 0.14 g of ammonium sulphate was also pulsed once for 0.25 min. Biochemistry analyser Biochemistry analyser YSI SELECT 2730 (Yellow Springs Instruments Inc.) configured for online mea-

Fuzzy logic control algorithm A fuzzy logic feedback controller was developed with error (E) and change in error (CE) as input variables. The residual glucose concentration was used as both measured and controlled variable. Error was represented in the membership function diagram with five fuzzy subsets (Figure 1a) and labeled as Z (zero), S (small), M (medium), L (large) and X (very large). Similarly, representation of the change in error was done

129

Impact of balanced substrate flux

Figure 1. The membership functions for the fuzzy logic control of residual glucose concentration during fed-batch cultivation of Btg. (a) Input membership function for error in residual glucose concentration. (b) Input membership function for change in error in residual glucose concentration. (c) Output membership function for output voltage to the feed pump.

with three fuzzy subsets labeled as N (negative), Z (zero) and P (positive) (Figure 1b). These two input membership function diagrams were employed for the fuzzification of input variables from their crisp numerical values. The output membership function diagram was also constructed with five fuzzy subsets labeled as Z (zero), S (small), M (medium), L (large) and X (very large) with output voltage 0–10 V (Figure 1c). A rule base for fuzzy inference was designed as shown in Table 1. The membership value of the output fuzzy subset was determined as the minimum value among the membership values of the input fuzzy subset by each selected rule defined in the rule base. This inference procedure was according to min–max algorithm developed by Mamdani (1976). The centre of area method was used for the defuzzification of the output membership value into a crisp real value (Pedryz 1991).

The signal voltage from the fuzzy logic control system (Vc) corresponds to 0–7.5 V with baseline at 1.5 V. During the execution of control, sterile medium was fed into the bioreactor for a constant time at a flow rate corresponding to the voltage supplied to the feed pump, ml/s ¼ 0.1779*V ) 0.0345 with V indicating control voltage i.e. V ¼ Vc+1.5. The mass flowrate of glucose was, g/s ¼ 2.224*V ) 0.431. With feed containing glucose and yeast extract in equal concentration, total mass flowrate was, g/s ¼ 2.224*V ) 0.431. The algorithm for the control system was written in LabVIEW. The output control voltage was simulated for different values of error and change in error within the experimental range. The simulated data, output control voltage was plotted in MATLAB 4.0 as shown in Figure 2. The 3D plot indicates the response of the control system in output voltage for changes in the input variables.

Results and Discussion Table 1. Fuzzy inference control rule base for the fuzzy logic control system. Change in error

N Z P

ERROR Z

S

M

L

X

Z Z Z

Z S M

S M L

M L X

L X X

Symbols are explained in Materials and Methods.

The increased awareness and application of biopesticides in the agricultural sector has motivated research towards its economical production. Process optimization towards the enhancement of biomass and ICP synthesis was focussed on providing balanced substrate flux through the macroscopic environment during the cultivation of Btg. This necessitated a control system for maintaining the residual glucose concentration at constant level in the bioreactor. A fuzzy logic-based

130

Figure 2. Simulation of the control system for the output control voltage with changes in error and change in error in residual glucose concentration.

feedback control system was developed with online monitoring of residual glucose concentration during cultivation. The fuzzy logic algorithm and accessories needed for this system have been explained in Materials and Methods. Control system performance with cultivation The experimental setup depicting the control system with accessories is schematically represented in Figure 3. With the earlier medium development studies for Btg, it was observed that GYA medium resulted in relatively high cell density (Sachidanandham et al. 1996). Fed-batch cultivations were performed in GYA medium with control extending on residual glucose concentration at different levels for various time periods. The cell density and residual glucose concentration profiles during these cultivations are shown in Figure 4. Due to different substrate utilization rates at various phases of growth, depletion of residual glucose was observed to vary in a

R.K.I. Anderson and K. Jayaraman nonlinear fashion. Significantly, final cell density during the cultivation with control at 2.0 g/l for 10 h (Figure 4(d)) was found to be relatively high. This indicated the inhibitory effect of glucose at elevated levels in previous cultivations. Also, cell density was observed to increase linearly due to the conditions imposed on the growing culture by providing nutrients at low concentrations. The substrate flux at this level has provided a balanced metabolism by preventing the accumulation of metabolic wastes. The maintenance of residual glucose concentration at constant level has provided the environment for balanced substrate flux. Hence, the developed fuzzy logic-based feedback control system has facilitated in maintaining the residual glucose concentration at desired levels by operating the bioreactor in fed-batch mode. Subsequently, it was necessary to identify the optimum level of residual glucose concentration for enhancing the cell density and ICP during cultivation. Fed-batch cultivation using fuzzy logic control The experimental data from batch cultivations of Btg at varying initial concentration of glucose and yeast extract (Anderson & Jayaraman 2003) were utilized to understand the influence of residual glucose concentration on the instantaneous growth rate and instantaneous substrate utilization rate. With microbial cultures subjected to transient conditions in batch cultivation, instantaneous growth rate and instantaneous substrate utilization rate during the growth phase were observed to follow the variation in the macroscopic environment. Cell density (X ) and residual glucose concentration (S ) were correlated with time in terms of a higher order polynomial using GRAPHER 1.09. The polynomial for cell density with time was differentiated once with respect to time in order to obtain an expression relating the instantaneous growth rate of the culture with time [dX/dt ¼ f(t)]. Similarly, the polynomial depicting the

Figure 3. Schematic diagram of the setup for the fed-batch cultivation of Btg employing online monitoring and control of residual glucose concentration by fuzzy logic-based feedback control system.

131

Impact of balanced substrate flux

3

4 3

2

2

1.5

Cell Density, g/l

2.5

1 1

0.5 Cell density

0

0 5

10

15

2 2

1.5 1

1 0

20

0 0

5

10

2

2

1

1

Cell density 0

0 12

15

Residual glucose conc., g/l

3

Residual glucose conc.

9

25

4

Cell Density, g/l

Residual Glucose conc., g/l

4

6

20

(b)

4

3

15

Time, hrs

(a)

0

0.5

Cell density

Time, hrs

3

2.5

3

5 4

3

Residual glucose conc.

3 2 2 1

1 Cell density

0

18

Cell density, g/l

0

3 Residual glucose conc.

Cell density, g/l

Residual glucose conc., g/l

Residual glucose conc. Residual glucose conc., g/l

4

3.5

5

0

5

10

Time, hrs

15

20

0

25

Tim e, hrs

(d)

(c)

Figure 4. Profiles of residual glucose concentration and cell density during the fed-batch cultivations of Btg with control on residual glucose concentration. The initial glucose concentration was 3.4 g/l in all cultivations. (a) Control for a period of 10 h at the setpoint of 3.4 g/l. (b) Control for a period of 10 h at the setpoint of 1.0 g/l. (c) Control for a period of 5 h at the setpoint of 2.0 g/l. (d) Control for a period of 10 h at the setpoint of 2.0 g/l.

4

15 YE

AS

3

9 2

Residual glucose conc.

6 1

3 Cell density

0

0 0

5

10

15

20

Residual glucose conc., g/l

Cell Density, g/l

12

25

(a) Time, hrs 38

10 5 90 DO

60 37

Reactor weight

45 30

36 .5

Reactor weight, Kgs

37 .5

75

DO, %

relationship of instantaneous substrate utilization rate with time [dS/dt ¼ f(t)] was obtained on differentiation. Cell density, residual glucose concentration, instantaneous growth rate and instantaneous substrate utilization rate were simulated in MATLAB 4.0 with time interval of 0.01 h to identify the optimum environment in terms of glucose and yeast extract concentrations. Earlier batch cultivations indicated a relatively high specific growth rate with 3.4 g/l and 20 g/l as initial concentrations of glucose and yeast extract respectively (Anderson & Jayaraman 2003). The analysis of instantaneous substrate utilization rate and growth rate during cultivation indicated their maxima at 1.5 g/l and 0.97 g/l of residual glucose concentration respectively. With metabolic activities dependent on the flux of substrates/ nutrients through the metabolic network, the possibility of enhancing glucose utilization by maintaining the residual glucose concentration at 1.5 g/l was identified. Hence, Btg was cultivated in GY medium with 20 g/l of Yeast extract and 1.5 g/l of glucose. The residual glucose concentration was maintained at 1.5 g/l by feeding concentrated glucose solution. Cell density and residual glucose concentration profiles during this fedbatch cultivation were shown in Figure 5a. This cultivation has resulted in a specific growth rate of 1.2 h)1 with 10.8 g/l of cell density and spore count of 11.0 · 1012 spores/ml. The net increase in cell density with counteracting factor of cell concentration by

15 0

36 0

4

8

12

16

20

24

(b) Time, hrs

Figure 5. Fed-batch cultivation of Btg with residual glucose concentration maintained at 1.5 g/l by feeding glucose solution with fuzzy logic control system. (a) Profiles of cell density and residual glucose concentration. The arrows indicate the pulsing of nitrogen sources during cultivation. YE – yeast extract; AS – ammonium sulphate. (b) Profiles of dissolved oxygen concentration (DO) and reactor weight.

132 20

5 YE AS

Cell Density, g/l

3 Residual glucose conc.

10

2 5 1

Cell density

0

Residual glucose conc., g/l

4 15

0 0

10

20

(a)

30

40

Time, hrs

105

42 DO

41

75

40

60 39 45 38

Reactor weight

30

Reactor weight, Kgs

90

D O ,%

filtration module in the setup is shown in Figure 5b. In addition, heterogenous population of cells with ICP concentration of 12.8 g/l was observed during sporulation. The oxygen uptake during the cultivation was observed to be relatively high. This could be due to the consumption of glucose as energy source through fermentative pathway (Mignone & Rossa 1996) and utilization of yeast extract towards biomass formation. Non-limiting conditions of organic and inorganic nitrogen sources were also established during the cultivation by pulsing the respective nitrogen sources. Hence, fedbatch cultivation with residual glucose concentration maintained at 1.5 g/l and high initial concentration of yeast extract has resulted in the redirection of substrate flux through the fermentative pathway. In order to avoid glucose being consumed as an energy source, experimental data from batch cultivations (Anderson & Jayaraman 2003) with initial concentrations of glucose/yeast extract at 3.4/1.0, 18.7/1.0 and 34/1.0 were analysed. During the cultivation at 3.4/ 1.0, instantaneous growth rate and substrate utilization rate have attained their maxima relatively faster at 2.72 g/l and 2.3 g/l of residual glucose concentration respectively. Residual glucose concentration at 2.72 g/l could have favoured the substrate flux towards biomass synthesis resulting in maximum instantaneous growth rate. Hence, Btg was cultivated in GY medium with 2.72 g/l of glucose and 1.0 g/l of yeast extract and residual glucose concentration maintained at 2.72 g/l by feeding glucose and yeast extract at equal concentration. Feeding the substrates at equal concentration was made on the basis of providing balanced condition. Cell density and residual glucose concentration profiles during the cultivation are shown in Figure 6a. Sporulation phase with uniform release of spores at the cell density of 16.0 g/l with 90.77 · 1013 spores/ml and 15.7 g/l of ICP was observed. The achievement of enhanced cell density by growth of the culture alone was observed in Figure 6b. The specific growth rate was observed to be 0.69 h)1, lower in comparison with the previous fed-batch cultivation. In fact, yeast extract and ammonium sulphate were fed into the bioreactor for ensuring the non-limiting conditions. This demonstrated the effective utilization of glucose as carbon source with significant increase in substrate flux towards biomass synthesis. The need of optimum specific growth rate in attaining enhanced substrate flux towards biomass synthesis and subsequently in spore formation and ICP synthesis was observed. External feeding performed till the initiation of sporulation and then continued as batch cultivation indicated the influence of batch operation on the sporulation process towards attaining elevated concentration of ICP. The external feeding of glucose and yeast extract at equal concentration during the cultivation of Btg has resulted in a balanced substrate flux upon maintaining the residual glucose concentration at 2.72 g/l. The balanced substrate flux at the optimum residual glucose concentration provided by the fuzzy logic-based control

R.K.I. Anderson and K. Jayaraman

37

15 0

36 0

5

10

(b)

15

20

25

30

35

40

Time, hrs

Figure 6. Fed-batch cultivation of Btg with residual glucose concentration maintained at 2.72 g/l by feeding glucose and yeast extract at equal concentration with fuzzy logic control system. (a) Profiles of cell density and residual glucose concentration. The arrows indicate the pulsing of nitrogen sources during cultivation. YE – yeast extract; AS – ammonium sulphate. (b) Profiles of dissolved oxygen concentration (DO) and reactor weight.

system has alleviated the inhibiting conditions and favoured the enhanced flux of substrates/nutrients towards biomass and ICP synthesis. In addition, supply of amino acids from the hydrolysis of added yeast extract could have influenced the cultivation. Hence, fluxes of different amino acids during the condition of balanced substrate flux could be studied to facilitate the comparison of different constituents in complex sources towards the economical production of biopesticides.

Acknowledgements R.K.I. Anderson thanks CSIR, India for his SRF fellowship. The authors also thank Dr P. Kaliraj for the cooperation and support extended during this study at Centre for Biotechnology, Anna University, Chennai, India. References Anderson, R.K.I. & Jayaraman, K. 2003 Influence of carbon and nitrogen sources on the growth and sporulation of Bacillus thuringiensis var galleriae for biopesticide production. Chemical and Biochemical Engineering Quarterly 17, 225–231.

Impact of balanced substrate flux Cho, Y.N. & Chang, Y.K. 1995 On-line measurement and control of cell concentration of Saccharomyces cerevisiae using a laser turbidometer. Biotechnology techniques 9, 557–562. Dhir, S., Morrow, Jr. J., Rhinehart, R.R. & Wiesner, T. 2000 Dynamic optimization of hybridoma growth in a fed-batch bioreactor. Biotechnology and Bioengineering 67, 197–205. Feng, M. & Glassey, J. 2000 Physiological state-specific models in estimation of recombinant Escherichia coli fermentation performance. Biotechnology and Bioengineering 69, 495–503. Guthke, R., Schmidt, H.W. & Pfaff, M. 1998 Knowledge acquisition and knowledge-based control in bioprocess engineering. Journal of Biotechnology 65, 37–46. Jeevarajan, A.S., Vani, S., Taylor, T.D. & Anderson, M.M. 2002 Continuous pH monitoring in a perfused bioreactor system using an optical pH sensor. Biotechnology and Bioengineering 78, 467–472. Konstantinov, K.B. & Yoshida, T. 1992 Knowledge-based control of fermentation processes. Biotechnology and Bioengineering 39, 479– 486. Mamdani, E.H. 1976 Advances in the linguistic synthesis of fuzzy controllers. International Journal of Man-Machine Studies 8, 669– 678. Mignone, C.F. & Rossa, C.A. 1996 Analysis of glucose carbon fluxes in continuous cultures of Bacillus thuringiensis. Applied Microbiology and Biotechnology 46, 78–84. Pedryz, W. 1991 Fuzzy Control and fuzzy Systems New York: Research Studies Press Ltd., John Wiley & Sons Inc. ISBN 0-471-. Ranganath, M., Renganathan, S. & Rao, C.S. 1999 Genetic algorithm based fuzzy logic control of a fed-batch fermentor. Bioprocess Engineering 21, 215–218.

133 Ronen, M., Shabtai, Y. & Guterman, H. 2002 Optimization of feeding profile for a fed-batch bioreactor by an evolutionary algorithm. Journal of Biotechnology 97, 253–263. Sachidanandham, R. & Jayaraman, K. 1993 Formation of spontaneous asporogenic variants of Bacillus thuringiensis var galleriae in continuous cultures. Applied Microbiology and Biotechnology 40, 504–507. Sachidanandham, R., Jenny, K., Fiechter, A. & Jayaraman, K. 1996 Stabilization and increased production of insecticidal crystal proteins of Bacillus thuringiensis var galleriae in steady and transient-state continuous cultures. Applied Microbiology and Biotechnology 47, 12–17. Shioya, S., Shimizu, K. & Yoshida, T. 1999 Knowledge-based design and operation of bioprocess systems. Journal of Fermentation and Bioengineering 87, 261–266. Smith, R.A. & Ulrich, J.T. 1983 Enzyme – linked immunosorbent assay for quantitative detection of Bacillus thuringiensis crystal protein. Applied and Environmental Microbiology 45, 586–590. Suzuki, H., Kishimoto, M., Kamoshita, Y., Omasa, T., Katakura, Y. & Suga, K.I. 2000 On-line control of feeding of medium components to attain high cell density. Bioprocess Engineering 22, 433–440. Valdez, C.L., Baruch, I. & Barrera, C.J. 2003 Neural networks applied to the prediction of fed-batch fermentation kinetics of Bacillus thuringiensis. Bioprocess and Biosystems Engineering 25, 229–233. Zhang, Q., John, F.R., Bruce, L.J., Jinliang, R. & Chang, S.W. 1994 A prototype neural network supervised control system for Bacillus thuringiensis fermentations. Biotechnology and Bioengineering 43, 483–489.

World Journal of Microbiology & Biotechnology 2005 21: 135–142 DOI: 10.1007/s11274-004-3042-2

 Springer 2005

Degradation and corrosive activities of fungi in a diesel–mild steel–aqueous system Fa´tima Menezes Bento1, Iwona Boguslava Beech2, Christine Claire Gaylarde3,*, Gelsa Edith Englert4 and Iduvirges Lourdes Muller4 1 Department of Soil, Faculty of Agronomy, UFRGS, 7712 Bento Goncßalves Avenue, CEP 91540-001, POA, RS, Brazil 2 School of Pharmacy and Biomedical Sciences, Microbiology Research Laboratory, St Michael’s Building, White Swan Road, Portsmouth PO1 2DT,UK 3 Department of Soil, MIRCEN, UFRGS, 7712 Bento Gonc¸alves Avenue, CEP: 90001-970, POA, RS, Brazil 4 Department of Metallurgy, Biocorrosion and Biofilms Lab, UFRGS, 99 Osvaldo Aranha Avenue s.615D, CEP: 90035-190, POA, RS, Brazil *Author for correspondence: Tel.: +55-51-33-16-6026, Fax: +55-51-3316-6029, E-mail: [email protected] Received 22 December 2003; accepted 22 June 2004

Keywords: Aspergillus fumigatus, biocorrosion, biodegradation, diesel oil, fungi, mild steel, propionic acid, storage tanks

Summary The fungi Aspergillus fumigatus, Hormoconis resinae and Candida silvicola were isolated from the fuel/water interfacial biomass in diesel storage tanks in Brazil. Their corrosive activities on mild steel ASTM A 283-93-C, used in storage tanks for urban diesel, were evaluated after various times of incubation at 30
C in a modified Bushnell– Haas mineral medium (without chlorides) with diesel oil as sole source of carbon. Their ability to degrade diesel oil was evaluated after growth for 30 and 60 days. The fungus Aspergillus fumigatus and the consortium of all three organisms showed the highest production of biomass; A. fumigatus gave the greatest value for steel weight loss and produced the greatest reduction in pH of the aqueous phase. Solid phase microextraction (SPME) showed that the main acid present in the aqueous phase after 60 days incubation with A. fumigatus was propionic acid. Polarization curves indicated that microbial activity influenced the anodic process, probably by the production of corrosive metabolites, and that this was particularly important in the case of A. fumigatus. This fungus preferentially degraded aliphatic hydrocarbons of chain lengths C11AC13 in the diesel, producing 47.7, 37.5 and 51% reductions in C11, C12 and C13, respectively. It produced more degradation than the consortium after 60 days incubation. It is likely that the presence of other species in the consortium inhibited the growth of A. fumigatus, thus resulting in a lower rate of diesel fuel degradation.

Introduction Biocorrosion or microbiologically influenced corrosion (MIC) is corrosion that is influenced by the presence and activities of microorganisms or their metabolites. MIC is recognized by many industries as a significant factor in reducing the useful life of equipment (Westbrook et al. 1988; Rosales et al. 1994; Machado et al. 1998; Beech & Gaylarde 1999; Gaylarde et al. 1999; Videla 2001) and has been documented for metals exposed to crude and distillate fuels during storage (Videla et al. 1994; Stoecker 1995; Machado et al. 1998). Microbial contamination of hydrocarbon fuels is the cause of serious problems in the quality of the product, as well as the corrosion of metallic structures in contact with the fuels. Microbial growth facilitates passivity breakdown and induces localized corrosion of fuel storage tanks through the action of organic acidic metabolites derived from

hydrocarbon degradation. Some hydrocarbons, such as n-alkanes, are easily degraded by a large number of microorganisms which use them as sources of carbon and energy. Their corrosive effect depends on the pH and electrolyte composition of the medium, especially chloride and phosphate levels. According to Walker & Cooney (1975), monoterminal oxidation appears to be the major pathway of n-alkane oxidation; this involves conversion to the homologous primary alcohol, aldehyde and monoic acid, and the fatty acid is oxidized via b-oxidation rather than being incorporated directly into cell components. However, in addition to the structure of the compound, the composition of the hydrocarbon mixture also affects the degradability of individual components (Olson et al. 1999). Diesel oil is a complex mixture of hydrocarbons, which varies according to the production process. Many

136 publications describe the degradation of diesel oil by microorganisms (Wright & Ratledge 1988; Margesin & Schinner 1996; MacCormack et al. 1998; Olson et al. 1999; Richard & Vogel 1999), but there have been few studies on the ability of fungi isolated in Brazil to degrade Brazilian diesel oil (Gaylarde et al. 1999). There are few references on the effect of fungal contaminants of diesel oil on the electrochemical behavior of mild steel, although this is one of the most common materials employed in the construction of fuel storage and distribution systems. These experiments were designed to determine the degradation activities of three fungi isolated from the fuel/water interfacial biomass in diesel storage tanks in Brazil, together with the corrosion behavior of mild steel exposed to these organisms.

Materials and methods Microorganisms The fungi Aspergillus fumigatus, Hormoconis resinae and Candida silvicola were isolated from the fuel/water interfacial biomass in a diesel storage tank in Rio Grande do Sul, Brazil. They were isolated by filtration of 500 ml oil through 0.45 lm pore-size filters followed by culture of the filters on malt agar (Bento & Gaylarde 1996). Growth assays The filamentous fungi were cultured on malt agar and incubated at 30
C for 7 days; for the yeast, incubation was for 48 h. After growth, spore and cell suspensions of the filamentous fungi and yeast respectively were prepared in distilled water and counted in a Petroff– Hauser chamber. The final concentrations used were 103 spores ml)1 for filamentous fungi and 102 cells ml)1 for yeast. Growth assays were carried out in 100 ml flasks with 30 ml of modified mineral medium BH* (Bushnell–Haas medium modified by the omission of chlorides, containing magnesium sulphate 0.2 g/l; calcium nitrate 0.02 g/l; ammonium molybdate 0.001 g/l; potassium dihydrogen phosphate 1 g/l; dipotassium monohydrogen phosphate 1 g/l; ammonium nitrate 1 g/l) plus 5 ml metropolitan diesel, sterilized by filtration through a Millipore membrane (0.22 lm), as the sole source of carbon. After various times, the biomass of filamentous fungi (A. fumigatus and H. resinae) was filtered, dried to constant weight and the final weight recorded. The yeast C. silvicola was serially diluted in distilled water and enumerated as c.f.u. Five replicates were set up for each treatment. The experiments were carried out with each microorganism separately and the mixed culture (consortium). Control samples were not inoculated and contained only diesel oil and mineral medium BH*.

C.C. Gaylarde et al. Water phase analyses After 60 days of incubation with and without microorganisms, pH, surface tension and acid metabolites were measured. The Solid Phase Microextraction (SPME) technique was used for the extraction of the main metabolites from the degraded diesel oil after 60 days incubation. Surface tension was determined using a DuNouy Tensiometer. All measurements were made on cell-free broth obtained by filtration through a Millipore membrane (0.22 lm). Diesel oil analyses After 30 and 60 days incubation, the hydrocarbon layer was extracted using a separating funnel and to each 50 ll one ml of dichloromethane was added. Hydrocarbons were detected by gas chromatography with mass spectroscopy (GC-MS). Operating conditions for GC-MS were: HP 5890II GC with HP5972 MS (Mass Selective Detector); Scan Mode: 35–450 AMU (atomic mass units); Column: SGE BPX5 25m · 0.2 mm, 0.25 lm film thickness, bonded phase capillary; Carrier gas: He @ 1 ml/min flow rate Injector temp: 250
C, split injection @ 20:1; Injection size 1.0 ll; Type (MSD) Detector temp 280
C; Oven program: initial temp 60
C, held for 3 min, then ramped at 7
C /min to 280
C and held for 10 min. Duration of the run: approximately 40 min. The percentage degradation was calculated for each peak, assigning 100% to C17. The peak areas from the control flasks were compared with the peak areas from the inoculated flasks. Three samples were run and the values averaged. Electrochemical assays The electrochemical tests were carried out in a special cell containing aqueous medium (BH*) taken from flasks after 60 days incubation with and without microoganisms, using a coupon of mild steel (ASTM A 283-93 C) as working electrode, a standard calomel electrode (SCE) and a platinum wire as counter electrode. Prior to each test the working electrodes were polished with emery paper (100–600 mesh). The polarization curves were obtained by scanning the potential from )1000 mVSCE to 700 mVSCE. The scan rate was 20 mV/min. This method was selected as the best to characterize the differences in behavior of the steel in distinct electrolytes. Beginning at negative potentials, previous air-formed films are reduced, and in the anodic scan the tendency to form passive films or, alternatively, the active dissolution of the metal, is seen. Weight loss assays Mild steel coupons were suspended in flasks containing 50 ml BH* and 5 ml metropolitan diesel, sterilized by filtration through a Millipore membrane (0.22 lm). The

137

Degradation and corrosive activities of fungi

Results and discussion Aspergillus fumigatus and the consortium produced the greatest biomass (Figure 1). The consortium appeared to enter the stationary phase at 30 days while A. fumigatus only reached the same weight (approx. 60 mg) on day 60. During the first month, the production of biomass by A. fumigatus and Hormoconis resinae was similar. However, at 42 days H. resinae reached the highest weight recorded (42 mg), and at the end of the incubation period entered the stationary phase. The behavior of A. fumigatus and H. resinae in consortium was different when compared with the growth curves of single species. The consortium showed the highest values of biomass throughout the whole incubation time, including in the stationary phase. Candida silvicola showed the maximum biomass on day 7. The presence of fungal metabolites had a marked effect on the anodic potentiostatic polarization curves of mild steel in a system containing the same medium, after 60 days of incubation with microorganisms. Conditions

in the sterile medium (control, uninoculated medium) favoured passivation of the metal, observed as the lowest values of the anodic currents (1 lA) (Figure 2). The procedure used to obtain the polarization curves (by scanning the potential from )1000 to 700 mVSCE) can affect the results of the anodic scan, but was useful to characterize the influence of the metabolites (Figure 3). In the sterile medium, (Figure 2) the scan beginning from )1000 mV first showed a cathodic branch, characteristic of hydrogen ion reduction and dissolved oxygen reduction. This was followed by an anodic peak indicating active dissolution of the iron after the removal of the previous air-formed film by reduction at the lower potentials. Continuing in the noble direction, the steel tended to passivate and thus a cathodic branch was once more seen, since the cathodic current is greater than the anodic one at a certain potential. When the cathodic reduction of dissolved oxygen became sufficiently low, a final anodic branch was found where low anodic currents typical of passive

800 600 400

0 -200 -400 -600 -800 -1000 -1200 10-9

600

-1

4 Consortium H. resinae A.fumigatus C.silvicola

0 0

10

20

30

40

50

3 2

60

Time (days)

Figure 1. Growth curves of Hormoconis resinae, Aspergillus fumigatus, Candida silvicola and consortium, in medium BH*/diesel oil. The filamentous fungi were enumerated as dry weight (left vertical axe) and the yeast as c.f.u. (right vertical axe).

200

E(mV vs ECS)

c.f.u. ml

Dry Weight (mg)

400

20

10-6

1x10-5

1x10-4

10-3

800

7

5

10-7

Figure 2. Potentiostatic curve of ASTM A284 steel in aqueous medium (BH*) after 60 days of incubation in presence of diesel oil without microorganisms (Control).

8

6

10-8

log i ( A/cm2)

60

40

Control

200

E(mV vs ECS )

coupons were placed vertically in the aqueous phase. Flasks were inoculated with A. fumigatus, H. resinae and C. silvicola at the same concentrations utilized in the growth assays, as single species and as a consortium (all species together). The control was not inoculated. Four metal samples were removed from each treatment after 15, 30, 60, 80 and 100 days of incubation. They were weighed and observed in the scanning electron microscope (SEM). For this purpose the samples were washed in Clark’s Solution (20 g Sb2O3/50 g SnCl2 in 1l HCl P.A.) for 1 min to remove the corrosion products. Samples were washed with sterile distilled water, dried, weighed and then coated with gold prior to SEM observation. The weight loss results were analysed statistically, using the non-parametric Kruskal–Wallis test with a 1% significance level.

Aspergillus fumigatus Hormoconis resinae Candida silvicola Consortium

Hormoconis

0 -200 Candida

-400 -600 Aspergillus

-800

Consortium

-1000 -1200 10-7

10-6

1x10 -5

1x10 -4

10-3

10-2

2

log i ( A/cm )

Figure 3. Potentiostatic curves of ASTM A284 steel in aqueous medium (BH*) after 60 days of incubation in presence of diesel oil with Aspergillus fumigatus, Hormoconis resinae, Candida silvicola, Consortium (all microorganisms).

138 behavior were seen. Three open-circuit potentials were thus to be found during the potential scan: one around )750 mV , the second one at )600 mV and the third one at )400 mV. This typical electrochemical behavior in BH* is due mainly to the presence of phosphate (Bento et al. 2004). The solution was chosen to allow the influence of the microorganisms to be evaluated. In the presence of the metabolic products of the fungi, the anodic potentiostatic polarization curves of mild steel showed active dissolution (high anodic currents) (Figure 3) as compared with the passive behavior in the sterile electrolyte (Figure 2). The active peak seen in the sterile medium was no longer followed by a passivation, showing that the microorganisms modified the electrolyte. The mineral medium used contained sulphate, nitrate and phosphate ions at concentrations of 0.16, 0.79 and 1.242 g l)1, respectively. Sulphate could facilitate pitting at higher potentials (chloride was not present) but phosphate induces passivation by forming barrier precipitates such as Fe3(PO4)2, which was detected by X-ray diffraction (results not shown), and which inhibit pit propagation in sterile medium (Nanchollas 1983; Franklin et al. 1990). Franklin et al. (2000) showed that the presence or the metabolic activity of bacteria affects the corrosion of carbon steel in a system containing phosphates. The observed acidification of the electrolyte was not expected to produce a great change in corrosion since it is well known that corrosion of mild steel in static aerated freshwater is not dependent on pH between the values of 4 and 9. However, this is only true so long as the rate-determining reaction of corrosion is diffusioncontrolled oxygen reduction. The behavior can undergo important modifications due to the presence of substances which may complex metal ions, induce pitting, inhibit dissolution, strongly adsorb on the surface, or promote alternative cathodic processes. In the present case, the chosen electrolyte with high phosphate levels induced passivation, as seen in the corresponding polarization curve. Acidity can prevent repassivation, hindering protective oxide film formation (Rozenfeld

C.C. Gaylarde et al. 1981; Nanchollas 1983; Videla et al. 1994). Acidity, however, is not the only factor accounting for the increase in aggressivity observed in the presence of fungal contaminants. The uptake of nitrogen and phosphate from the medium diminishes the level of corrosion inhibitors, which can result in catastrophic corrosion of aluminium alloys (Rosales et al. 1994; Videla et al. 1994) and a similar effect can occur with phosphate for mild steel (Franklin et al. 2000). The uptake of phosphate by microorganisms, which has been demonstrated (Bento et al. 2004), can thus hinder the passivation of the steel, while sulphate and nitrate do not protect steel from corrosion. The various metabolites, especially organic acids, may also stimulate corrosion by the adsorption and complexing of iron. The cathodic curve was also different for each fungal species. The control showed the lowest value of a cathodic limiting current (30 lA) (Figure 2) and A. fumigatus produced the highest (110 lA) (Figure 3). The other values detected were 70 lA for C. silvicola and H. resinae and for the consortium 80 lA. The most probable explanation for the increased cathodic current is the reduction of metabolites such as organic acids and carbon dioxide, but in the filtered solution used for polarization curves the CO2 produced will have been lost to the atmosphere. Thus it is more likely that organic metabolic products are reduced at cathodic potentials, thus increasing cathodic currents, which otherwise are due mainly to dissolved oxygen reduction or, at lower potentials, to hydrogen ion reduction. From the polarization curves obtained it is therefore possible to conclude that an increase in corrosion in the presence of the metabolites is due to an increase in both the anodic and the cathodic reactions. The presence of metabolic products after 60 days incubation had a marked effect on the corrosion of the carbon steel. Low values of pH may be the main reason for this increased aggressivity (Table 1). Propionic acid was identified, among other metabolites (a series of alcohols and ketones), in the aqueous phase after 60 days incubation with A. fumigatus (Figure 4). It has

Figure 4. Chromatogram of water phase after 60 days of incubation with Aspergillus fumigatus showing the main metabolites as alcohols, ketones and propionic acid.

139

Degradation and corrosive activities of fungi been reported that the low pH in the aqueous phase of microbially infected fuels is mainly due to organic acid production (Videla et al. 1994). The rate of acid production is a function of the rate of hydrocarbon degradation, which is largely dependent on the availability of nitrogen and phosphorus in the medium (Schiapparelli & Meybaum 1980; Videla et al. 1988). The amount of hydrocarbon degraded by microorganisms is an important factor in explaining microbial activity in relation to corrosion (Videla et al. 1994). After 30 and 60 days of growth, the diesel oil was degraded in the inoculated flasks (Tables 2 and 3). The yeast C. silvicola produced a lower degree of degradation after 60 days than the other microorganisms, but secreted a surfactant (decreasing surface tension in the aqueous medium) (Table 1) and formed emulsions in the oil phase, suggesting that it could, under the right conditions, increase the bioavailability of the hydrocarbons. After 60 days, surface tension and pH measurements were reduced for all cultures, suggesting the production of surfactant and organic acids derived from hydrocarbon degradation. A. fumigatus produced the greatest reduction in pH and surface tension (Table 1), Table 1. pH and surface tension measurements from aqueous phase growth curves after 60 days of incubation with and without microorganisms in diesel oil, as sole carbon source. Microorganisms

pH Initial: 7.0

Surface tension (mN/m) Initial: 74

Aspergillus fumigatus Hormoconis resinae Candida silvicola Consortium Control

4.8 6.4 6.6 6.4 6.9

48.5 52.5 60 51 70

and was able to degrade the hydrocarbons more efficiently than other microorganisms tested (Tables 2 and 3), while the fungus H. resinae showed low degradation activity at 30 days, but at 60 days the percentage degradation, mainly for C11, was the highest (65.8%, Table 2). Reductions in some carbon chainlengths were observed in the controls (not shown), probably caused by abiotic processes of inorganic oxidation, transformation, sorption to glass walls, interactions with medium components and volatilization (Batts & Fathoni 1991; Margesin & Schinner 1996). After 60 days incubation, hydrocarbons above C17 were seen to be more resistant to biodegradation. The accumulation of toxic products that takes place during the growth of microorganisms on fuels in closed systems may be responsible for the limited degradation of these hydrocarbons, but it is more likely that these components are naturally more recalcitrant. Mixed populations meet the conditions of the natural environment while pure cultures do not. Microorganisms may act synergistically, increasing their overall capacity. Using pure cultures in degradation experiments is far from representative, but in order to analyse the results such simplifications are necessary. Indeed it would be difficult to prepare a representative mixed culture, since there would be so many interacting variables to consider. However, the results presented here show that single microbial cultures may induce higher degradation than mixed cultures. The selected consortium did not efficiently reduce the level of any hydrocarbon detected in the diesel oil. The results show the high biodegradability of aliphatic (chain lengths C11AC13, Figure 5) and low biodegradability of aromatic hydrocarbons. Above C17 the hydrocarbons appear to be resistant to degradation

Table 2. Degradation of diesel oil after 30 days in mineral medium Bushnell–Haas with microorganisms. The percentage degradation (%) was calculated for each peak with relation to the control. Carbon chain length

Aspergillus fumigatus (%)

Hormoconis resinae (%)

Candida silvı´cola (%)

Consortium (%)

C11 C12 C13 C14 C15 C16

15.69 31.32 42.10 29.84 18.28 13

22.62 2.80 6.16 5.38 3.69 1.22

)5.32 24.33 26.61 8.97 12.68 1.81

14.18 9.12 )2.37 )3.26 )0.43 1.45

Table 3. Degradation of diesel oil after 60 days in mineral medium Bushnell–Haas with microorganisms. The percentage degradation (%) was calculated for each peak with relation to the control. Carbon chain length

Aspergillus fumigatus (%)

Hormoconis resinae (%)

Candida silvı´cola (%)

Consortium (%)

C11 C12 C13 C14 C15 C16

47.72 37.48 51.06 35.22 25 14.7

65.87 44.18 32.76 16.77 12.01 12.82

28.12 32.64 33.29 19.20 13.82 14.55

36.28 14.03 17.26 14.18 9.22 3.56

140

C.C. Gaylarde et al.

Figure 5. Chromatogram of degraded of diesel oil (aliphatics C11, C12 and C13) after 60 days of incubation with Aspergillus fumigatus, Control (uninoculated) and Consortium.

and the area under this peak in the chromatogram does not vary significantly between control and inoculated samples. This suggests that after significant depletion of the short-chain alkane compounds, the microorganisms were unable to degrade the aromatic compounds and this has previously been shown by other workers (Fedorak et al. 1984; MacCormack et al. 1998; Olson et al. 1999; Richard & Vogel 1999). The results of the weight loss measurements are shown in Figure 6. After 30 and 60 days the steel weight losses for single cultures of A. fumigatus and H. resinae were similar and significantly different from those of the control and C. silvicola. The potentiostatic polarization curves after 60 days were coherent with these results. Neither the control, nor the metal incubated with C. silvicola, showed significant differences of weight loss with time (P > 0.1). However, all the other microorganisms produced increased weight loss with time and this was significant at the 1% level. The fungus A. fumigatus produced the highest weight loss at 80 and 100 days in immersion tests and this measurement

of generalized corrosion was confirmed by the SEM picture, in addition to the appearance of localized pitting. This is shown in Figure 7 on the surface of a coupon that had been incubated with A. fumigatus for 100 days. Apart from the phosphate consumption and production of acid metabolites capable of stimulating iron corrosion, it must be taken into account that carbon dioxide is one of the main final metabolites of hydrocarbon degradation. It is well known in the petroleum industry that CO2 greatly stimulates corrosion of mild steel in neutral or slightly acid deaerated solutions (Schmitt & Rothmann 1977). Several mechanisms have been proposed to explain the fact that the limiting cathodic current observed in the presence of CO2 is much higher than in its absence at the same pH (Schmitt 1984). Under biofilms or adhered microorganisms a relative deaeration is known to be produced and CO2 corrosion could result. CO2 corrosion is frequently

20

Weight Loss (mg)/cm 2

18

Aspergillus fumigatus Control Hormoconis resinae Candida silvicola Consortium

16 14 12 10 8 6 4 2 0 0

20

40

60

80

100

Time (Days)

Figure 6. Weight loss of mild steel ASTM A 283 after various times of incubation with microorganisms in modified Bushnell-Haas mineral medium/diesel oil.

Figure 7. SEM of mild steel ASTM A 283 after 100 days of incubation with Aspergillus fumigatus (magnification: ·1800) in modified Bushnell-Haas mineral medium/diesel oil.

Degradation and corrosive activities of fungi in the form of pits, which were found in the present work. All microorganisms produced a general attack while the control showed only a little intergranular corrosion. The reduced pH and the presence of different anions, such as sulphate and nitrate, in the mineral medium may explain the corrosion occurring in the control (Salvarezza et al. 1983). Phosphate films, such as those formed on the steel surface, are generally considered slightly protective, but are not thought to inhibit corrosion over a long period (Shreir 1979). The present results indicate that, under the conditions of our experiments in the presence of the other anions (sulphate, nitrate), these phosphate films are degraded with time resulting in incipient corrosion.

Conclusions 1. The strains studied were able to grow on diesel oil and the metabolites resulting from degradation influenced the corrosive behavior of mild steel. 2. The microorganisms isolated from diesel oil storage tank sludge degraded the hydrocarbons of diesel oil, especially the aliphatic fractions. Aspergillus fumigatus was more efficient than the other microorganisms tested. 3. Propionic acid was identified, among other metabolites, in the water phase from growth assays after 60 days incubation with A. fumigatus 4. A. fumigatus produced greatest weight loss of the mild steel coupons, and this corrosion was confirmed by electrochemical measurements.

Acknowledgements The authors wish to thank Mr M.A.W. Hill for his technical assistance with GC–MS, Dr Graham Mills for his technical assistance with Solid Phase Microextraction from headspace, the University of Portsmouth for the use of the equipment and UNESCO for a Biotechnology Fellowship to Fatima Menezes Bento.

References Batts, B.D. & Fathoni, Z.A. 1991 A literature review on fuel stability studies with particular emphasis on diesel Oil. Energy and Fuels 5, 2–21. Beech, I.B. & Gaylarde, C.C. 1999 Recent advances in the study of biocorrosion – an overview. Revista de Microbiologia 30, 177–190. Bento, F.M. & Gaylarde, C.C. 1996 Microbial contamination of stored diesel oil in Brazil. Revista de Microbiologia 27, 71–75. Bento, F.M., Englert, G.E., Gaylarde, C.C. & Muller, I.L. 2004 Influence of aqueous phase on electrochemical biocorrosion tests in diesel/water systems. Materials and Corrosion, (in press). Fedorak, P.M., Semple, K.M. & Westlake, D.W.S. 1984 Oil-degrading capabilities of yeasts and fungi isolated from coastal marine environments. Canadian Journal of Microbiology 30, 565–571.

141 Franklin, M.J., White, D.C. & Isaacs, H.S. 1990 Effect of Bacterial Biofilms on carbon steel pit propagation in phosphate containing medium. In Proceedings of International Congress on Microbially Influenced Corrosion and Biodeterioration, eds. NACE. Houston, pp. 3-35-41. Franklin, M.J., White, D.C., Little, B, Ray, R. & Pope, R. 2000 The role of bacteria in pit propagation of carbon steel. Biofouling 15, 13–23. Gaylarde, C.C., Bento, F.M. & Kelley, J. 1999 Microbial contamination of stored hydrocarbon fuels and its control. Revista de Microbiologia 30, 1–10. MacCormack, W.P., Rios Merino, L.N. & Fraile, E.R. 1998 Diesel oil biodegradation in antarctic soils. In III Latin American Biodeterioration and Biodegradation, LABS3, eds. Gaylarde, C.C., Barbosa, T.C.P. & Gabilan, N.H. UK: The Phycological Society. Published on CD-ROM, Paper No. 51, Florianopo´lis. Machado, P.F.L., Gaylarde, C.C. & Muller, I.L. 1998 Microbial influenced corrosion of mild steel ASTM A283 in a marine diesel– aqueous system. In III Latin American Biodeterioration and Biodegradation, LABS3, eds. Gaylarde, C.C., Barbosa, T.C.P. & Gabilan, N.H. UK: The Phycological Society. Published on CDROM, Paper No. 54, Florianopo´lis. Margesin, R & Schinner, F. 1996 Biodegradation of diesel oil in liquid cultures and in soils at low temperatures. In Papers of the 10th International Biodeterioration and Biodegradation Symposium, Hamburg, 15–18 september. DECHEMA Monographs, vol. 133, pp. 295–301. Nanchollas, G.H. 1983 Phosphate precipitation in corrosion protection: reaction mechanisms. Corrosion 39, 77–82. Olson, J.J., Mills, G.L., Herbert, B.E. & Morris, P.J. 1999 Biodegradation rates of separated diesel components. Environmental Toxicology and Chemistry 18, 2448–2453. Richard, J.Y. & Vogel, T.M. 1999 Characterization of a soil bacterial consortium capable of degrading diesel fuel. International Biodeterioration and Biodegradation 44, 93–100. Rosales, B.M., Chichizola, S. & Baleani, D. 1994 Corrosion por el hongo Hormoconis resinae de as aleaciones de Al de uso aeronautico AA 7075 y 2024. In First NACE Latin American Region Corrosion Congress and First Venezuelan Corrosion Congress, ed. NACE, Paper No. 94054. Maracaibo, 06–11. November. Rozenfeld, I.L. 1981 Corrosion Inhibitors pp. 175–181. New York: McGraw-Hill. ISBN 0-07054170-1. Salvarezza, R.C., Mele, M.F.L., Videlu, H.A. 1983 Mechanisms of the microbiol corrosion of aluminium alloys. Corrosion v.39(1): 26–32. Schiapparelli, E.R. & Meybaum, B.R. 1980 The role of dodecanoic acid in the microbiological corrosion of jet aircraft integral fuel tanks. International Biodeterioration Bulletin 16, 61–66. Schmitt, G. 1984 Fundamentals of CO2 Corrosion, In Advances in CO2 Corrosion, Vol. 1:10, ed. Hausler R.H.; Houston: NACE. ISBN 0915567-10-5. Schmitt, G. & B. Rothmann. 1977 Werkstoffe und Korrosion 28, 816. Shreir, L.L. 1979 In Corrosion vol 2. pp. 16–28. London: Ed NewnesButterworths. ISBN 0 408 00267 0. Stoecker, J.G. 1995 Microbiological and electrochemical types of corrosion: Back to basics. Materials Performance 34, 49–52. Videla, H. Guiamet, P.S., Do Valle, S. & Reinoso, E.H. 1988 Effects of fungal and bacterial contaminants of kerosene fuel. In CORROSION 88, Paper No 91., St. Louis. Videla, H.A., Guiamet, P.S. & Reinoso, E.H. 1994 Biocorrosion of structural materials by fungal contaminants of jet fuels. A State of the art. In First NACE Latin American Region Corrosion Congress and First Venezuelan Corrosion Congress, Maracaibo, November 06–11. Paper No. 94094. Videla, H.A. 2001 Microbially induced corrosion: an updated overview. International Biodeterioration and Biodegradation 48, 176– 201. Walker, J.D & Cooney, J.J. 1973 Pathway of n-alkane oxidation in Cladosporium resinae. Journal of Bacteriology 15, 635–639. Westbrook, S.R., Barbee, J.G., Stavinoha, L.L., Lepera, M.E. & Mengenhauser, J.V. 1988 Methodology for Identification of diesel

142 fuel system contaminants related to problems in the field. In Distillate Fuel: Contamination, Storage and Handling, ASTM STP 1005, eds. Chesneau H.L. & Dorris, M.M. pp. 37–47. Philadelphia: American Society For Testing and Materials.

C.C. Gaylarde et al. Wright, J.D. & Ratledge, C. 1988 Degradation of aromatic compounds by yeasts and moulds isolated from diesel oil contaminated soil. In Biodegradation of Lubricants. pp. 423–425.

World Journal of Microbiology & Biotechnology 2005 21: 143–150 DOI: 10.1007/s11274-004-3041-3


Springer 2005

Changes in lignocellulolytic enzyme activites in six Pleurotus spp. strains cultivated on coffee pulp in confrontation with Trichoderma spp. G. Mata1,*, D.M. Murrieta Herna´ndez1 and L.G. Iglesias Andreu2 1 Instituto de Ecologı´a A.C., Apartado Postal 63, Xalapa 91000, Veracruz, Me´xico 2 Instituto de Gene´tica Forestal, U.V., Apartado Postal 551, Xalapa 91000, Veracruz, Me´xico *Author for correspondence: Tel.: +52-2-288-421828, Fax: +52-2-288-187809, E-mail: [email protected] Received 7 November 2003; accepted 22 June 2004

Keywords: Coffee pulp, endoglucanase, laccase, Mn peroxidase, Pleurotus, Trichoderma

Summary The antagonistic effect of six Pleurotus spp. strains was studied in confrontation with three strains of Trichoderma spp. Pleurotus strains were cultivated on sterile coffee pulp, with and without a Trichoderma inoculant. Laccase, Mn peroxidase and endoglucanase activities were determined during incubation. Laccase production was also studied by PAGE analysis to detect enzymatic isoforms. Results show that the presence of Trichoderma induced a significant increase in oxidase production by the Pleurotus strains. Nevertheless, Trichoderma was not observed to induce laccase isoforms.

Introduction Degradation of lignocellulosic substrates by mushrooms of the genus Pleurotus depends on the production and secretion of such enzymes as the cellulases, hemicellulases and ligninases. The production of these enzymes is important in substrate colonization, as well as decisive in fruit body production (Giardina et al. 1999; Okamoto et al. 2000; Ohga et al. 2000). The production of emergent hyphae, the formation of a dark line, and the overproduction of laccase are all reactions observed when the mycelia of Lentinula edodes (shiitake) or Pleurotus spp. are confronted with mycelia of antagonistic moulds of the genus Trichoderma (Mata & Savoie 1998b). Confrontation of Trichoderma spp. with Agaricus bisporus does not produce a dark line (Ohmasa & Cheong 1999; Mamoun et al. 2000; Savoie & Mata 2003). Although the increase in laccase production by Pleurotus mycelia is a well-studied phenomenon, there is not yet enough information on the origin of the overproduction of this enzyme. Mexico is the fifth-largest producer of coffee in the world (Consejo Mexicano del Cafe´ 2000), and additionally produces more than 100,000 tons annually of coffee pulp, an organic waste by-product resulting from coffee production. This waste could be useful because of its high content of carbohydrates and proteins. However, the presence of caffeine, tannins and polyphenols limits its utilization (Martı´ nez et al. 1984; Zuluaga 1989). Coffee pulp has little commercial application, and it is considered the main polluting agent of rivers and lakes located near coffee-producing regions (Russos et al.

1995). Nevertheless, the fermentation and/or vermicomposting of this waste offer attractive recycling alternatives (Aranda & Barois 2000; Pandey et al. 2000). Coffee pulp has also been used experimentally in the cultivation of mushrooms (Martı´ nez-Carrera 1987; Bermu´dez et al. 1994; Martı´ nez-Carrera et al. 1996; Leifa et al. 1999; Pandey et al. 2000). However, when this substrate is not handled appropriately, moulds (mainly Trichoderma) can produce considerable losses. Coffee pulp is not used on an industrial level for Pleurotus production, although the yields obtained on this substrate can be relatively high, compared to other agroindustrial products (Martı´ nez-Carrera 1987). Although different aspects of Pleurotus production on coffee pulp have been studied, little it is known about the antagonistic relationship between Pleurotus and Trichoderma. The objective of this study was to determine the capacity of some Pleurotus strains, when cultivated on coffee pulp, to resist the attacks of Trichoderma moulds, and establish a method to select Pleurotus strains that are adapted to growth on coffee pulp.

Materials and methods Six strains of the genus Pleurotus, and three strains of the genus Trichoderma, were studied (Table 1). Pleurotus strains were selected for their high capacity for laccase production and resistance to Trichoderma, according the results of Salmones et al. (2000). The following strains of Trichoderma were isolated: (1) T. viride 1 was obtained from shiitake samples (L. edodes)

144

G. Mata et al.

Table 1. Species and strains under study. Species

Strain number

Origin

Pleurotus djamor (Fr.) Boedjin

IE-121 IE-218 IE-38 IE-49 IE-137 IE-225 IE-639 IE-637a IE-638b

Mexico Mexico Hong Kong Guatemala Czechoslovakia Mexico Mexico Mexico Mexico

Pleurotus ostreatus (Jaq. ex Fr.) Kumm. Pleurotus pulmonarius (Fr.) Que´l. Trichoderma reesei E.G. Simmons Trichoderma viride Pers. ex SF. Gray

a b

Trichoderma viride 1 (V1). T. viride 2 (V2).

cultivated on coffee pulp; (2) T. viride 2 was taken from Pleutrotus samples cultivated on coffee pulp; and (3) T. reesei was harvested from Pleurotus samples cultivated on barley straw. These strains were deposited in the strain collection of the Instituto de Ecologia (IE), where they are maintained in Petri dish cultures [malt extract agar medium (MEA); aseptic conditions; 28 C; total darkness]. Spawn and cultivation substrate preparation Spawn was prepared using moistened sorghum grains (Sorghum vulgare Pers.) in plastic bags (100 g/bag). Bags were autoclaved for 1 h at 121 C (Guzma´n et al. 1993). Once cooled, 1 cm2 of mycelium cultivated in MEA was placed into each spawn bag and incubated at 28 C for 15 days in total darkness. This time period was sufficient to allow the mycelium to completely cover the sorghum grains. Coffee pulp was soaked in water for 12 h at room temperature, and then 250 g were placed in plastic bags and autoclaved for 1 h at 121 C. Coffee pulp was inoculated by adding 12.5 g (equivalent to 5%) of the Pleurotus spawn. For the confrontation studies, once the bagged coffee pulp samples had been inoculated with Pleurotus strains, fragments of MEA (5 mm diameter) covered with Trichoderma mycelium were placed inside. Samples inoculated only with Trichoderma or Pleurotus, as well as samples of non-inoculated coffee pulp (the control group), were also prepared. Ten replicates of each condition were prepared. Samples were incubated at 28 C in total darkness for 16 days (this time period allowed the incubation phase to be completed). Enzymatic activity determination Substrate samples containing mycelia were taken at 0, 4, 8, 12 and 16 days of cultivation. For laccase determination, fresh substrate samples were used according to the methods of Vela´zquez-Ceden˜o et al. (2002), while endoglucanase and Mn peroxidase were determined form lyophilized samples (Mata & Savoie 1998a). Extracts of crude enzyme were prepared in 100 ml flasks each containing 0.7 g of fresh or lyophilized

substrate. Ten ml of distilled water were added to each flask, and these were then rotated end over end at 45 rev min)1, for 30 min, in a roto-torque (Equipar Mod. N. 7637-75). The supernatant was filtered through nylon mesh (porosity 0.37 mm) and centrifuged twice at 3 C and 8500 rev min)1 (10,016 · g) for 15 min (Mata & Savoie 1998a). The recovered liquid, know as crude enzyme extract, was used to evaluate enzymatic activities. Laccase (EC 1.10.3.2) activity was determined from the oxidation of syringaldazin in a phosphate–citrate buffer solution (0.1 M; pH 5.0), changes in the absorbance at 526 nm (e ¼ 6.5 · 104 M )1 cm)1) were monitored for 2 min (Leonowicz & Grzywnowicz 1981). Mn peroxidase (EC 1.11.1.13) activity was determined from the oxidation of 3-dimethylaminobenzoic acid (DMAB) and 3-methyl-2-benzothiazolinone hydrazone hydrochloride (MBTH) in a phosphate–citrate buffer solution (0.1 M; pH 5.0). Changes in the absorbance at 590 nm (e ¼ 3.29 · 104 M)1 cm)1) were monitored for 2 min (Lonergan & Baker 1995). MnSO4 and H2O2 were added to the mixture to start the reaction. Blanks lacking H2O2 and MnSO4 were also prepared. Endoglucanase (cellulase, EC 3.2.1.4) activity was evaluated through liberation of reducing sugars (Miller 1959), using carboxymethylcellulose (CMC 2%) as the enzymatic substrate and incubating the samples for 30 min. Enzyme activities were determined with a spectrophotometer (Spectronic Genesis 5). All enzyme assays were performed at 30 C. Enzyme activities were expressed as U g)1 of cultivation substrate, where U g)1 was defined as the amount of enzyme needed to produce 1 lmol of product min)1 g)1 of substrate extracted. Three replicates were prepared per condition and enzymatic activity. Folin–Ciocalteu reagent (Box 1983) was used for determining phenolic concentration in the substrates. For each variable, a multivariate analysis of variance (MANOVA) was carried out. The effect of antagonism on laccase production Induction of laccase production Based on the results obtained above, antagonistic effects on the extracellular production of laccase were studied

Changes in enzyme activities of Pleurotus spp. using P. pulmonarius (IE 225) and T. viride (IE 637) cultured in a liquid medium. Both strains were precultivated for 2 days on a nylon membrane (2 cm in diameter) that had been placed on MEA. Then each colonized membrane was used to inoculate 50 ml of malt extract medium (ME) (Savoie et al. 1998). Samples were incubated at 25 C for 11 days. Laccase activity was measured in the ME medium, in all cases, on days 7 and 11. The following treatments were evaluated. Treatment 1. This control group consisted of samples that had been inoculated with either P. pulmonarius (C1) or T. viride (C2). Treatment 2. The talus confrontation (Tc) group was composed of 7-day-old P. pulmonarius mycelia to whose margins had been added the mycelia of T. viride previously grown on nylon membranes. Treatment 3. This ‘extracellular metabolites’ group required that the effects of the two antagonists be determined in a reciprocal way. First, cell-free culture fluids were collected from pure cultures, following 7 days of incubation, using a nylon filter (0.2 lm; Whatmann). Then these culture fluids were inoculated with an antagonistic mycelia that had been previously cultivated on a nylon membrane. P. pulmonarius cultured in metabolites derived from T. viride was denominated MT, and T. viride cultured upon metabolites corresponding to P. pulmonarius was named MP. Treatment 4. This group was prepared in order to evaluate the effect of lysing enzymes produced by T. harzianum. A dilution of 0.1 g ml)1 of lysing enzymes (Sigma L-1412) was prepared in distilled water and then sterilized by filtration using a nylon filter (0.2 lm; Whatmann). One ml of this solution was added to the culture flasks containing 7-day-old P. pulmonarius. This treatment was labelled LE. Treatment 5. The YME treatment group was used to investigate the induction of laccase activity by phenols in P. pulmonarius cultured on a 2% yeast plus 1% ME. A solution of water-soluble lignin derivates was prepared (Indulin AT, Sigma) and the phenol concentration adjusted to 0.1 mM (Mata et al. 1997). This solution was added to the YME culture medium (described above) in order to prepare a distinct culture medium for P. pulmonarius that was denominated WSLD. Five replicates were prepared for all treatment groups. Results were subjected to an analysis of variance (ANOVA) and to Tukey’s multiple-range test. Laccase isoforms produced during antagonism Extracts for electrophoresis were prepared by taking 1 ml of liquid media from each of the treatments cited above, and then filtering them by centrifugation (using ‘Ultrafree-C’, Millipore filters) for 90 min at 13,000 rev min)1 (16,060 · g) and 3 C. The pellets obtained by

145 this procedure were washed with 20 ll of Tris–HCl Millipore (pH 6.8) and stored at )20 C until needed. This same procedure was carried out for the crude enzymatic extract obtained from P. pulmonarius, following 12 days of culturing on coffee pulp (with and without T. viride from the previous experiment). Electrophoresis was undertaken on denaturing polyacrylamide gels (Pharmacia, SDS-PAGE; 8–25% gradient) in a Phast system (Pharmacia Biotech). Laccase bands were revealed with a 0.09% solution of 4-Chloro1-naphthol in sodium acetate buffer solution (0.1 M, pH 5.0). Gels were fixed in 7% solution of acetic acid (Mata et al. 1997, Savoie et al. 1998).

Results The results of laccase production during the vegetative growth of six Pleurotus strains in confrontation with three Trichoderma strains cultured on coffee pulp are shown in Figure 1. Laccase increased at day 8, showing its maximum value at 12 days, and then falling off in most of the strains at day 16, when the mycelia had completely covered the substrates. The only exception to this trend occurred when the Pleurotus strains were placed in confrontation with T. reesei. In this case, laccase production remained the same through to day 16. Highest activity was recorded for the P. pulmonarius strains, IE 137 and IE 225. In confrontation with T. viride 1 and T. reesei, these two Pleurotus strains showed significant increases (MANOVA P < 0.001) in laccase production on days 12 and 16, when compared to the control groups. The P. ostreatus strains, IE 38 and IE 49, showed important increases in laccase production, but these were not statistically significant. The P. djamor strains, IE 121 and IE 218, showed the lowest values of laccase production. Furthermore, these increases were not statistically significant in confrontations, especially in the case of IE 218. Mn peroxidase activity showed a similar pattern to laccase production, having its maximum activity on day 12, and then falling off in most strains at day 16. The exception to this trend in Pleurotus was the treatment confrontation with T. reesei, in which production continued to increase until day 16 (Figure 2). Maximum values for all confrontations were recorded for the IE 137 and IE 225 strains of P. pulmonarius (MANOVA P < 0.001). Endoglucanase activity remained low throughout the vegetative development of all Pleurotus strains (Figure 3); therefore, it was not possible to establish a clear pattern of production. However, in confrontation with T. viride 2 a significant increase was observed for P. djamor strains (IE 121, IE 218) at day 16 (MANOVA P < 0.001). Results for the degradation of phenols by the six strains of Pleurotus can be seen in Figure 4. In all cases, phenol concentration diminished to less than 5 lmol ml)1. The lowest values for the Pleurotus control group and for Pleurotus in confrontation with T. viride

146

G. Mata et al. 200

U g-1

150

100

IE38 50

IE49 IE121 IE137 IE218

0 4

8

12

16

4

8

12

16

4

V1

Control

8

12

16

4

8

V2

12

16

IE225

Tr

Figure 1. Kinetics of the enzymatic activity of laccase during the vegetative development of six strains of Pleurotus (IE 38, IE 49, IE 121, IE 137, IE 218, IE 225), when confronted with three strains of Trichoderma (IE 637, IE 638, IE 639) (Control ¼ Pleurotus without confrontation, v1 = Trichoderma viride IE 637, v2 ¼ Trichoderma viride IE 638, and Tr ¼ Trichoderma reesei IE 639).

9000

8000

7000

U g-1

6000

5000

4000

3000

IE 38 IE 49

2000

IE 121 1000

IE 137 IE-218

0 4

8

12

Control

16

4

8

12

16

V1

4

8 V2

12

16

4

8

12

16

IE 225

Tr

Figure 2. Kinetics of the enzymatic activity of Mn peroxidase during the vegetative development of six strains of Pleurotus (IE 38, IE 49, IE 121, IE 137, IE 218, IE 225), when confronted with three strains of Trichoderma (IE 637, IE 638, IE 639) (Control ¼ Pleurotus without confrontation, v1 ¼ Trichoderma viride IE 637, v2 ¼ Trichoderma viride IE 638, and Tr ¼ Trichoderma reesei IE 639).

were observed on day 8, while in the confrontations with T. viride 1 and T. reesei the lowest values for Pleurotus were registered on day 12 (MANOVA P < 0.001). Phenol degradation was not observed for the Trichoderma control group, nor was it registered for the noninoculated control samples (in which case, no enzymatic activity was observed; data not shown). Analysis of the effects of antagonism on laccase production Results of the laccase activity obtained between P. pulmonarius (IE 225) and T. viride (IE 637) are shown in

Figure 5. In the Pleurotus control group (C1), no significant increase in laccase activity was observed between days 7 and 11. On the other hand, laccase activity was not observed in the Trichoderma control group (C2), thus suggesting an inability to produce this enzyme. In the Tc treatment we observed that, upon adding T. viride mycelia to 7-day-old samples of P. pulmonarius, laccase production was increased significantly (ANOVA P < 0.001) (Figure 5), yielding more than three times the laccase produced by the P. pulmonarius controls. Since the metabolites of T. viride did not allow for the development of P. pulmonarius strains, there are no data for laccase activity associated with the MT treatment.

Changes in enzyme activities of Pleurotus spp.

147

1.4

1.2

U g-1

1.0

0.8

0.6

IE 38

0.4

IE9 IE 121

0.2

IE 137 IE 218

0.0 4

8

12

16

4

Control

8

12

16

4

8

V1

12

16

8

4

V2

12

16

IE 225

Tr

Figure 3. Kinetics of the enzymatic activity of endoglucanase during the vegetative development of six strains of Pleurotus (IE 38, IE 49, IE 121, IE 137, IE 218, IE 225), when confronted with three strains of Trichoderma (IE 637, IE 638, IE 639) (Control ¼ Pleurotus without confrontation, v1 ¼ Trichoderma viride IE 637, v2 ¼ Trichoderma viride IE 638 and Tr ¼ Trichoderma reesei IE 639).

25

Phenols µmol/ml

20

15

10

IE 38 IE 49 5

IE 121 IE 137 IE 218

0 4

8

12

Control

16

4

8

12

16

4

v1

8

12 v2

16

4

8

12

16

IE 225

Tr

Figure 4. Degradation of water-soluble phenols during the vegetative development of six strains of Pleurotus (IE 38, IE 49, IE 121, IE 137, IE 218, IE 225), when confronted with three strains of Trichoderma (IE 637, IE 638, IE 639) (Control ¼ Pleurotus without confrontation, v1 ¼ Trichoderma viride IE 637, v2 ¼ Trichoderma viride IE 638, and Tr ¼ Trichoderma reesei IE 639).

However, in the MP treatment, laccase production by P. pulmonarius fell off, possibly because T. viride degraded the laccase in the culture medium that had been produced by P. pulmonarius (Figure 5). In the LE treatment (lytic enzyme addition) a significant increase in laccase activity was observed (ANOVA P < 0.001) (Figure 5), representing more than an eightfold increase in production compared to the P. pulmonarius controls. In the WSLD treatment, the presence of water-soluble lignin derivatives produced a light but significant increase in laccase activity in 7-day-old, P. pulmonarius strains (ANOVA P < 0.001), although this decreased

on day 11 (Figure 5). In the YME treatment, a similar situation that obtained for C1 in ME was observed. The results of the electrophoretic analysis are presented in Figure 6. Two patterns of main bands were observed; one pattern was formed by Pleurotus treatments grown on the liquid culture media and, the other pattern was associated with the coffee pulp cultures, with or without Trichoderma. However, the WSLD treatment showed a similar pattern to the one observed for coffee pulp. The presence of Trichoderma did not seem to induce new laccase isoforms; but rather, it only seemed to stimulate the overproduction of this enzyme.

148

G. Mata et al. 5000

mU g-1

4000

3000 7 Days 2000

11 Days

1000

0 C1

C2

TC

MT

MP

LE

WLSD

YME

Treatments

Figure 5. Enzymatic activity of laccase in liquid media (C1 ¼ Pleurotus control, C2 ¼ Trichoderma control, Ct ¼ confrontation of talus of Pleurotus and Trichoderma, MT ¼ Pleurotus in metabolites of Trichoderma, MP ¼ Trichoderma in metabolites of Pleurotus, LE ¼ Pleurotus with lysing enzymes, WSLD ¼ Pleurotus in water soluble lignine derivatives, YME ¼ extract of malt with yeast).

No. of bands

1 2 3

+

1

2

3

4

5

6

7

8

Treatmemts

Figure 6. Laccase zymogram. 1 ¼ Pleurotus with lysing enzymes, 2 ¼ Pleurotus YME, 3 ¼ Pleurotus in WLSD, 4 ¼ Pleurotus and Trichoderma in ME, 5 ¼ Pleurotus in ME, 6 ¼ Pleurotus on coffee pulp, 7 ¼ Pleurotus and Trichoderma on coffee pulp, 8 ¼ Pleurotus and Trichoderma in ME. Arrow indicates protein migration. Note: treatments 4 and 8 are the same.

Discussion Results indicate that the Pleurotus strains under investigation had similar pattern of enzyme secretion. Oxidase activity agreed with that reported by Mata & Savoie (1998a) for L. edodes, by Savoie (1998) for A. bisporus, and by Vela´zquez-Ceden˜o et al. (2002) for P. ostreatus and P. pulmonarius. According to Buswell et al. (1996), changes in enzymatic activity and phenol concentration observed during vegetative development suggest that Mn peroxidase might be involved in the depolymerization of substrates, while laccase might be associated with the detoxification of soluble phenolic compounds, as well as the degradation of lignin (Medeiros et al. 1999; Hublik & Schinner 2000; Lucas et al. 2001). Laccase activity and the degradation of phenols were negatively correlated (r ¼ )0.61), except in the case

of the P. djamor strains (IE 121, IE 218) in which activity levels were low (r ¼ )0.41) but phenols were nevertheless degraded. This result suggests that P. djamor may possess another enzymatic system for oxidase production that is different from the one that was under investigation in this study. Moulds of the genus Trichoderma are cellulolytic and produce neither laccase nor Mn peroxidase; thus, they were unable to degrade phenols in the coffee pulp medium (Vela´zquez-Ceden˜o et al. 2002). When Pleurotus is cultured on pasteurized substrates, inoculation is generally not carried out under sterile conditions, so that there is a certain risk of contamination by spores of moulds such as those of Trichoderma. This contamination is clearly visible after 2 weeks, with the appearance of the characteristic green spores of Trichoderma (Guzma´n et al. 1993). The three strains of Trichoderma studied here showed different competitive abilities; probably related to their differential susceptibility to the metabolites produced by Pleurotus for T. harzianum (see Savoie et al. 2001). Endoglucanase production has been associated with the stages of primordia formation and fructification, in which significant increases of enzymatic activity have been registered, declining in the crop stage (Kaviyarasan & Natarajan 1997; Mata & Savoie 1998a). Results in the present study did not show increased endoglucanase activity, probably because these mushrooms were studied only during incubation, a period in which antagonistic moulds should have caused considerable damage. However, in confrontation with T. viride 2, the P. djamor strains (IE 121 and IE 218) showed significant increases in endoglucanase production that might have been associated with previous states of fructification, an especially likely possibility as these strains are known to be precocious. Although Trichoderma produces polysaccharidases, in this study endoglucanase activity was relatively low. This result agrees with those obtained in studies of confrontations between Trichoderma and L. edodes (Savoie & Mata 1999). Enzymes produced by T. harzianum that are capable of degrading the cell wall are important elements in

Changes in enzyme activities of Pleurotus spp. antibiosis and in the parasitism of plant pathogens (Savoie et al. 200l). Nevertheless, in this study the presence of other metabolites in the lytic solutions and in the liquid culture media might have been responsible for the induction of laccase production. Other elements, like those derived from water-soluble lignin in the WSLD treatment, can also induce laccase activity. The main function of the enzyme, in this case, is to detoxify the culture medium of inhibitory compounds (Savoie et al. 1995). The possibility of inducing laccase production might have certain biotechnological applications; for example, in facilitating the detoxification of agroindustrial residuals with a high phenolic content. Phenols and components rich in lignin, when used in the inocula of L. edodes on pasteurized wheat straw, induce and improve this organism’s ability to withstand the attacks of Trichoderma spp. (Mata et al. 1998; Savoie et al. 2000). Results of this study indicated a high sensitivity of Pleurotus mycelia to the presence of lytic enzymes in the LE treatment. Although this condition inhibited the growth of Pleurotus strains, it did induce laccase production at a rate more than twice that observed for Pleurotus strains in the presence of T. viride mycelia (Tc treatment). Savoie et al. (1998) obtained similar results in confrontations between Trichoderma spp. and Lentinula edodes. However, Savoie & Mata (2003) observed somewhat different effects on the mycelial growth of A. bisporus, L. edodes and Pleurotus spp., after adding the lytic enzymes produced by T. harzianum to two, solid culture media (YMEA and WSLD). In general, they found that although mycelial growth was definitely affected, some strains of L. edodes and Pleurotus spp. were able to rapidly adapt to the presence of the lytic enzymes and thereby improve their resistance to this antagonistic mould. In this study, three laccase isoforms were detected in P. pulmonmarius. Palmieri et al. (1997), however, reported five laccase isoforms in P. ostreatus, while Lo et al. (2001) detected five in P. sajor-caju. On the other hand, Vela´zquez-Ceden˜o (2002) found two main laccase isoforms in confrontations between P. eryngii and Trichoderma sp., the isoform with higher molecular weight appeared during the first few days of confrontation, but later disappeared. In this study, the laccase isoforms obtained on coffee pulp were the same those found on the WSLD culture medium. From this finding, it might be inferred that Pleurotus strains adapted to growing on WSLD should also be able to resist Trichoderma on the coffee pulp substrate. Accordingly, it could be considered that this method constitutes a simple test for selecting strains with a capacity to resist antagonistic, Trichoderma moulds. Nevertheless, to confirm this hypothesis, it will be necessary to carry out an additional study in which a larger number of strains are analysed. The results of this study failed to demonstrate the production of laccase isoforms by P. pulmonarius, either when grown in the presence of lytic enzymes produced by T. harzianum or when cultured on coffee pulp in the presence of T. viride.

149 Acknowledgements The authors are grateful to CONACyT (Proyect 28 530N) and to the Instituto de Ecologı´ a, A.C. for financial support. Thanks are also due to Cornelia Dickel for revising this paper. Finally, Dan Bennack is acknowledged for translating portions of the original manuscript from Spanish to English.

References Aranda, E. & Barois, I. 2000 Coffee pulp in vermicomposting treatment. In Coffee Biotechnology and Quality, eds. Sera, T., Soccol, R.C., Pandey, A. & Roussos, S. pp. 489–506. Dordrecht: Kluwer Academic Publishers. ISBN 0-7923-6582-8. Bermu´dez, R.C., Traba, J.A., Verdecia, M.J. & Gropss, P. 1994 Produccio´n de Pleurotus sp. cfr. Florida sobre residuales de la agroinsdustria cafetalera en Cuba. Micologı´a Neotropical Aplicada 7, 47–50. Box, J.D. 1983 Investigation of the Folin Ciocalteu’s reagent for determination of polyphenolic substances in natural waters. Water Research 17, 501–525. Buswell, J.A., Cai, Y.J. & Chang, S. 1996 Lignolytic enzyme production and secretion in edible mushroom fungi. In Mushroom Biology and Mushroom Products, ed. Royse, D.J. pp. 113–122. University Park: Pennsylvania State University. ISBN 1-88395601-3. Consejo Mexicano del Cafe´, 2000 http://www.sagar.gob.mx/cmc/ cafe015p.htm. Giardina, P., Palmieri, G., Scaloni, A., Fontanella, B. & Faraco, V. 1999 Protein and gene structure of a blue laccase from Pleurotus ostreatus. Biochemical Journal 341, 655–663. Guzma´n, G., Mata, G., Salmones, D., Soto-Velazco, C. & Guzma´nDa´valos, L. 1993 El cultivo de los hongos comestibles. Con especial atencio´n a especies tropicales y subtropicales en esquilmos y residuos agro-industriales. 245 pp. Me´xico, D.F: IPN. ISBN 968-29-4492-9. Hublik, G. & Schinner, F. 2000 Characterization and immobilization of the laccase from Pleurotus ostreatus and its use for the continuous elimination of phenolic pollutants. Enzyme and Microbial Technology 27, 330–336. Kaviyarasan, V. & Natarajan, K. 1997 Changes in extracellular enzyme activities during growth and fruiting of Pleurotus cornucopiae var. citrinopileatus. In Advances in Mushroom Biology and Production, eds. Rai, R.D., Dhar, B.L. & Verna, R.N. pp. 309–319. Solan: National Research Centre for Mushroom. Leifa, F., Pandey, A. & Soccol, C.R. 1999 Cultivation of Pleurotus sp. on coffee residues. In 3rd International Conference on Mushroom Biology & Mushroom Products. eds. Broderick, A. & Nair, T. pp. 274–284. Sidney: WSMBMP. ISBN 1-86341-879-2. Leonowicz, A. & Grzywnowicz, K. 1981 Quantitative estimation of laccase forms in some white-rot fungi using syringaldazine as a substrate. Enzyme and Microbial Technology 33, 55–58. Lo, S.C., Ho, Y.S. & Buswell, J.A. 2001 Effect of phenolic monomers on the production of laccases by the edible mushroom Pleurotus sajor-caju, and partial characterization of a major laccase component. Mycologia 93, 413–421. Lonergan, G. & Baker, W.L. 1995 Comparative study of substrates of fungal laccase. Letters in Applied Microbiology 21, 31–33. Lucas, L.R., Robles Go´mez, A.M., Ga´lvez Del Postigo Ruiz, A., Garcı´ a Gutie´rrez, T., Pe´rez Pulido, R. & A´lvarez De Cienfuegos Lo´pez, G. 2001 Biodegradacio´n de la celulosa y la lignina. 148 pp. Jae´n: Universidad de Jae´n. ISBN 84-8439-059-4. Mamoun, M.L., Savoie, J.M. & Olivier, J.M. 2000 Interactions between the pathogen Trichoderma harzianum Th2 and Agaricus bisporus in mushroom compost. Mycologia 92, 233–240. Martı´ nez-Carrera, D. 1987 Design of a mushroom farm for growing Pleurotus on coffee pulp. Mushroom Journal for the Tropics 7, 13–23.

150 Martı´ nez-Carrera, D., Morales, P., Martı´ nez, W., Sobal, M. & Aguilar, A. 1996 Large-scale drying of coffee pulp and its potential for mushroom cultivation in Me´xico. Micologı´a Neotropical Aplicada 9, 43–52. Martı´ nez, D., Quirarte, M., Soto, C., Salmones, D. & Guzma´n, G. 1984 Perspectivas sobre el cultivo de hongos comestibles en residuos agroindustriales en Me´xico. Boletı´n de la Sociedad Mexicana de Micologı´a 19, 207–219. Mata, G. & Savoie, J.M. 1998a Extracellular enzyme activities in six Lentinula edodes strains during cultivation in wheat straw. World Journal of Microbiology and Biotechnology 14, 513–519. Mata, G. & Savoie, J.M. 1998b Screening of Lentinula edodes strains by laccase induction and resistance to Trichoderma sp. Revista Mexicana de Micologı´a 14, 29–32. Mata, G., Savoie, J.M. & Delpech, P. 1997 Variability in laccase production by mycelia of Lentinula boryana and Lentinula edodes in presence of soluble lignin derivates in solid media. Material und Organismen 31, 109–122. Mata, G., Savoie, J.M., Delpech, P. & Olivier, J.M. 1998 Reduction in the incidence of Trichoderma spp. Using substrate supplementation with peat and an alternative spawn during cultivation of Lentinula edodes on pasteurized wheat straw. Agronomie 18, 515– 520. Medeiros, M.B., Beneto, A.V., Nunes, A.L.L. & Oliveira, S.C. 1999 Optimization of some variables that affect the synthesis of laccase by Pleurotus ostreatus. Bioprocess Engineering 21, 483–487. Miller, G.L. 1959 User of dinitrosalicylic acid reagent for determination of reducing sugars. Analytical Chemistry 31, 424–426. Ohga, S., Cho, N., Thurston, C.F. & Wood, D.A. 2000 Transcriptional regulation of laccase and cellulase in relation to fruit body formation in the mycelium of Lentinula edodes on a sawdust-based substrate. Mycoscience 41, 149–153. Okamoto, K., Yanagi, S.O. & Sakai, T. 2000 Purification and characterization of extracellular laccase from Pleurotus ostreatus. Mycoscience 41, 7–13. Ohmasa, M. & Cheong, M.L. 1999 Effects of culture conditions of Lentinula edodes, ‘Shiitake mushroom’, on the disease resistance of Lentinula edodes against Trichoderma harzianum in the sawdust cultures. In 3rd International Conference on Mushroom Products. eds. Broderick, A. & Nair, T. Sydney: WSMBMP. ISBN 1-86341879-2. Palmieri, G., Giardina, P., Bianco, C., Scaloni, A., Capasso, A. & Sannia, G. 1997 A novel white laccase from Pleurotus ostreatus. Journal of Biological Chemistry 272, 31301–31307. Pandey, A., Soccol, C.R., Nigan, P., Brand, D., Mohan, R. & Roussos, S. 2000 Biotechnological potential of coffee pulp and coffee husk for bioprocesses. Biochemical Engineering Journal 6, 53–162.

G. Mata et al. Roussos, S., Aquia´huatl, M. De Los A., Trejo-Herna´ndez, M. Del R., Gaime Perraud, I., Favela, E., Ramakrishna, M., Raimbult, M. & Viniegra-Gonza´lez, G. 1995 Biotechnological management of coffee pulp-isolation, screening, characterization, selection of caffeine-degrading fungi and natural microflora present in coffee pulp an husk. Applied Microbiology and Biotechnology 42, 756– 762. Salmones, D., Mata, G. & Walizewski, C. 2000 Seleccio´n y adaptacio´n de cepas de Pleurotus spp. a la pulpa de cafe´. Memorias del VII Congreso Mexicano de Micologı´a. Me´xico: Quere´taro, Qro., 43 pp. Savoie, J.M. 1998 Changes in enzyme activities during early growth of the edible mushroom, Agaricus bisporus, in compost. Mycological Research 102, 1113–1118. Savoie, J.M. & Mata, G. 1999 The antagonistic action of Trichoderma sp. hyphae to Lentinula edodes hyphae changes lignocellulotytic activities during cultivation in wheat straw. World Journal of Microbiology and Biotechnology 15, 369–373. Savoie, J.M. & Mata, G. 2003 Trichoderma harzianum metabolites preadapt mushrooms to Trichoderma agressivum antagonism. Mycologia 95, 191–199. Savoie, J.M., Cesbron V. & Delpech, P. 1995 Induction of polyphenoloxidasas in the mycelium of Lentinula edodes. Mushroom Science 14, 787–793. Savoie, J.M., Mata, G. & Billete, C. 1998 Extracellular laccase production during hyphal interactions between Trichoderma spp. and shiitake, Lentinula edodes. Applied Microbiology and Biotechnology 49, 589–593. Savoie, J.M., Mata, G., Delpech, P. & Billette, C. 2000 Inoculum adaptation changes the outcome of the competition between Lentinula edodes and Trichoderma spp. During shiitake cultivation on pasteurised wheat straw. Mushroom Science 15, 667–674. Savoie, J.M., Mata, G. & Mamoun, M. 2001 Variability in brown line formation and extracellular laccase production during interaction between white-rot basidiomycetes and Trichoderma hazarium biotype Th2. Mycologia 93, 243–248. Vela´zquez-Ceden˜o, M.A. 2002 Interactions entre Pleurotus eryngii et les microorganismes pre´sents dans une culture sur substrat ligno-cellulosique. DEA Rapport Faculte´ des Sciences et Techniques de Saint-Je´roˆme. Marseille, France: Universite D’aix Marseille III. Vela´zquez-Ceden˜o, M.A., Mata, G. & Savoie, J.M. 2002 Wastereducing cultivation of Pleurotus ostreatus and Pleurotus pulmonarius on coffee pulp: changes in the production of some lignocellulolytic enzymes. World Journal of Microbiology and Biotechnology 18, 201–207. Zuluaga, V.J. 1989 Utilizacio´n integral de los subproductos del cafe´. Memorias I Seminario Internacional sobre Biotecnologı´a en la Agroindustria Cafetalera. pp. 63–76. Xalapa: INMECAFE.

World Journal of Microbiology & Biotechnology 2005 21: 151–154 DOI: 10.1007/s11274-004-3470-2

 Springer 2005

Milk-clotting protease production by Nocardiopsis sp. in an inexpensive medium M.T.H. Cavalcanti1,2, C.R. Martinez1, V.C. Furtado1, B.B. Neto3, M.F. Teixeira4, J.L. Lima Filho1,2 and A.L.F. Porto1,5,* 1 Laborato´rio de Imunopatologia Keizo Asami, LIKA/UFPE, Brazil 2 Departamento de Bioquı´mica, UFPE, Brazil 3 Departamento de Quı´mica Fundamental, UFPE, Brazil 4 Departamento de Patologia, UA, Brazil 5 Departamento de Morfologia e Fisiologia Animal, UFRPE, Brazil *Author for correspondence: Tel.: +55-2181-3301-2504, Fax: +55-2181-3271-8485, E-mail: analuporto@yahoo. com.br Received 18 February 2004; accepted 9 July 2004

Keywords: Milk-clotting protease, Nocardiopsis sp., production of milk-clotting protease

Summary The milk-clotting and proteolytic activities of extracellular enzyme preparations from Nocardiopsis sp. were investigated under different culture conditions. A soybean flour medium was used, with concentrations of soybean flour and of glucose varying from 0.25 to 1% w/v and from 0 to 1% w/v, respectively. Growth was monitored with 2 ml samples withdrawn from the culture medium at 8-h intervals, for determination of total protein, proteolytic activity, milk-clotting activity and sugar reduction. The best milk-clotting protease production, with a specific activity of 24.49 U/mg at 40 h, was obtained in the glucose-free medium containing soybean flour 1% w/v.

Introduction Programs for selecting new microorganisms for enzyme production are increasing in number around the world. One of the most investigated groups is that of the Actinomycetes, because they constitute a potential source of biotechnologically interesting substances (Lealem & Gashe 1994). Actinomycetes are particularly efficient in breaking down soil proteins, and also produce proteases (Tsujibo et al. 1990), of which the milk-clotting enzymes are a special class. The main milk-clotting enzyme found in animal tissues, higher plants and microorganisms is rennin (EC 3.4.23.4). The production of milk-clotting enzymes from microbial sources for use as rennin substitutes has been receiving increasing attention (Hashem 1999). For a long time calf rennet has been the traditional source of milk-clotting enzymes for cheese manufacturing, but decreasing worldwide rennet supplies, accompanied by ever increasing cheese production and consumption, made necessary the utilization of rennet substitutes (Areces et al. 1992). Nocardiopsis sp. has been reported as a potential source of milk-clotting proteases (Cavalcanti et al. 2004) and of proteases for the detergent industry (Moreira et al. 2002). Although the partial purification and characterization of milk-clotting proteases have been

reported (Cavalcanti et al. 2004), the study of protease production seems to be still absent from the literature. This motivates the present work, where we report on a study of the production of milk-clotting and proteolytic proteases from Nocardiopsis sp. cultivated under different media.

Materials and methods Microorganism The Nocardiopsis sp. was isolated from a soil sample collected in Northeastern Brazil. The microorganism was maintained at 28
C on ISP-2 agar slants (Pridham et al. 1957), containing 1.0% (w/v) malt extract, 0.4% (w/v) yeast extract, and 2.0% (w/v) agar. Culture media and condition of culture Four modifications of a soybean flour medium previously described (Porto et al. 1996) were used, resulting in the following media: M1: 1% (w/v) soybean flour; M2: 1% (w/v) soybean flour + 0.25% (w/v) glucose; M3: 0.5% (w/v) soybean flour + 0.5% (w/v) glucose, M4: 0.25% soybean flour (w/v) + 1% (w/v) glucose. To all media, 0.1 ml of a mineral salt solution was added (100 mg FeSO4 Æ 7H2O, 100 mg MnCl2 Æ 4H2O, 100 mg

152

Enzyme assay The proteolytic activity was determined as described previously (Moreira et al. 2001), in cell-free supernatants clarified by centrifugation (12,000 · g) for 5 min at 25
C. Azocasein, 1.0% (w/v) (Sigma Chemical Co., St. Louis, Mo USA) in 0.1 M Tris–HCl, pH 7.6, containing 1 mM CaCl2, was used as substrate. One activity unit was defined as the amount of enzyme producing in 1 h a 1-unit increase in the absorbance at 440 nm. The milk-clotting activity was measured as described before (Cavalcanti et al. 2003), in the culture medium previously clarified by centrifugation at 35
C (20,200 · g, 10 min). Skim milk 10% (w/v in deonized water) containing 10 mM CaCl2 was used as substrate. One activity unit was defined as the amount of enzyme that clotted 1 ml of substrate within 40 min at 35
C. Total protein concentration was determined according to the Bradford method using bovine serum albumin as a standard protein. Determination of the sugar reduction Sugar consumption was determined with DNSA (dinitrosalicylic acid) in the culture medium previously clarified by centrifugation (20,200 · g, 10 min) (Miller 1960).

Results and discussion Effect of medium composition on protease and milkclotting protease production and biomass production The levels of proteolytic and milk-clotting proteolytic activities of Nocardiopsis sp. were evaluated on several growth media based on the nitrogen source (soybean flour). These hydrolytic activities were monitored in the supernatant of each medium during a 56-h growth

40 M1 M2 M3 M4

35

Milk-Clotting Activity (U/mL)

ZnSO4 Æ H2O, 100 mg CaCl2 Æ H2O, distilled water 100 ml) and they contained in addition 0.3 M NH4Cl, 0.04 M MgSO4 Æ 7H2O, 0.063 M K2HPO4, and the pH was adjusted to 7.2. Growth curve experiments were carried out using Erlenmeyer flasks (50 ml) containing 20 ml of culture medium starting with 10% (v/v) of inoculum, incubated for 56 h at 28
C with orbital shaker (200 rev min)1). The total volume of each Erlenmeyer flask (20 ml) was collected every 8 h of growth, samples used for biomass and for the determinations: pH, total protein, milk-clotting and proteolytic activities. The biomass concentration was determined as mycelial dry weight of culture broth, mycelial pellets were dried (110
C for 2 h), and the biomass dry weight was determined gravimetrically (g l)1). (All measurements were carried out in duplicate) Data analysis, which included a principal component analysis (Beebe et al. 1998) of the milk-clotting enzyme production, was performed with the software Statistica 5.0 (StatSoft, Tulsa).

M.T.H. Cavalcanti et al.

30

56 48 5640

25

40

48 20 32 15 48 56

32 10 5 16 1 4 16 816 1 16 448 881

32 40

24 48 40 56 24 24 32 24

0 0

1

2

3

4

5

Proteolytic Activity (U/mg)

Figure 1. Influence of medium composition and culture age on proteolytic and milk-clotting activity of Nocardiopsis sp. M1: 1% (w/ v) soybean flour; M2: 1% (w/v) soybean flour + 0.25% (w/v) glucose; M3: 0.5% (w/v) soybean flour + 0.5% (w/v) glucose, M4: 0.25% soybean flour (w/v) + 1% (w/v) glucose. Point labels are the respective cultivation times, in hours.

period. The results, plotted in Figure 1, show that the M1 and M2 media (soybean flour 1% w/v and soybean flour 1% w/v + 0.25% w/v glucose, respectively) resulted in 33 and 28 U/ml of milk-clotting activities, respectively. The milk-clotting activities determined in M3 and M4 were low (11 and 2.55 U/ml, respectively), suggesting that induction effects of the soybean flour may be an important factor. The synthesis and secretion of protease are induced by peptides or other proteinaceous substrates, such as soybean flour (Porto et al. 1996) and wheat bran (Areces et al. 1992). Depending on the peptide nature and level, protease synthesis and secretion may be induced or repressed (Porto et al. 1996). Glucose, on the other hand, did not appear to influence protease production in this study. Other authors are in agreement with these results. Hashem (1999) and Areces et al. (1992), working respectively with Penicillium oxalicum and Mucor baciliformis, noted that glucose was not the most favourable carbon source for milk-clotting enzyme production. Figure 2 shows the milk-clotting activity curves of Nocardiopsis sp. in the four culture media. For milkclotting protease production, the best protease production was obtained in medium M1, from which glucose is absent. Effect of culture time on protease and milk-clotting enzyme production The production of extracellular milk-clotting enzymes changes according to the age of the culture (Figure 1). The results show that the optimum cultivation periods for milk-clotting enzyme productions were 56 h (M1) and 48 h (M2). These results are in agreement with those of other authors (Porto et al. 1996; Moreira et al. 2001), which showed that protease synthesis starts in Streptomyces clavuligerus in the post-exponential phase of growth. A higher production of milk-clotting enzymes between 72 and 192 h of growth has been reported for

Nocardiopsis milk-clotting protease

153

30

2 48

1

32 40 1 M1 M2 M3 M4

20

Scores on PC2 (24%)

Milk-clotting activity (U/mg)

25

15

10

8 1 1 16 44 16 8 4 816

56

-1

16

48

48 40

32 40

32

56

56

48 40 24 32

8 M1 M2 M3 M4

4 56

0

0

2

4

6

8

10

-3 -1.5

-1.0

-0.5

Culture Time (hours)

P. oxalicum, M. baciliformis and Mucor miehei (Areces et al. 1992; Escobar & Barnett 1993; Hashem 1999). The fast production of milk-clotting enzymes by Nocardiopsis sp. reported here constitutes an obvious advantage of this microorganism for industrial purposes, as compared to other sources of milk-clotting enzymes. Principal component analysis of milk-clotting enzyme production Figures 3 and 4 present the results of a principal component analysis of all experimental responses. The first two principal components (PC) account for 72% of the total variance. The loadings plot (Figure 3) shows that the biomass and the milk-clotting enzyme production parameters are located on different components, indicating a lack of correlation between biomass and milk-clotting enzyme production. Hashem (1999) working with P. oxalicum, reported similar results. The proteolytic and milk-clotting activities are strongly correlated, and also negatively correlated to protein

0.8

GLIGML

0.4

PMGML

0.0

0.5

1.0

1.5

2.0

2.5

Scores on PC1 (49%)

Figure 2. Milk-clotting activity curves of Nocardiopsis sp. M1: 1% (w/ v) soybean flour; M2: 1% (w/v) soybean flour + 0.25% (w/v) glucose; M3: 0.5% (w/v) soybean flour + 0.5% (w/v) glucose, M4: 0.25% soybean flour (w/v) + 1% (w/v) glucose.

PC2 Loadings (23%)

24

24 1

0

-2

5

24

Figure 4. Scores on the first two principal components of all production results. M1: 1% (w/v) soybean flour; M2: 1% (w/v) soybean flour + 0.25% (w/v) glucose; M3: 0.5% (w/v) soybean flour + 0.5% (w/v) glucose, M4: 0.25% soybean flour (w/v) + 1% (w/v) glucose.

and glucose amounts in the culture filtrate, as shown by the respective loadings on the first PC axis, which are located on opposite sides of the plot. The scores plot (Figure 4) shows that larger milk-clotting activities are associated mainly to media M1 and M2, which have the largest positive scores on PC1. Media M4, on the other hand, has the weakest milk-clotting activity values but the largest biomass, as shown by the scores on PC2. The best conditions for milk-clotting protease production obtained in this work employed a medium with 1% w/v soybean flour and a growth time of 56 h. These results suggest the importance of induction effects on this process, because higher soybean flour concentrations yielded the best results for milk-clotting protease production. Glucose did not show any noticeable influence on the milk-clotting protease production. As this study shows, the soybean medium may be considered a promising and inexpensive alternative for milkclotting protease production. For Brazil, the utilization of soybean flour is especially important, because the country is the world’s largest soybean producer.

Acknowledgements This work was supported by CNPq. Federal University of Pernambuco and Rural Federal University of Pernambuco.

AEUMG ATUML ACUML 0.0

-0.4

References GL

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

PC1 Loadings (49%)

Figure 3. Loadings on the first two principal components of all production results. GLIGML: glucose (g/ml), PMGML: total protein (mg/ml), GL: biomass (g/l), AEUMG: specific proteolytic activity (U/ mg), ATUML: total proteolytic activity (U/ml), ACUML: milkclotting activity (U/ml).

Areces, L.B., Biscoglio de Jime´nez Bonino, M.J., Parry, M.A.A., Fraile, E.R., Ferna´ndez, H.M. & Cascone, O. 1992 Purification and characterization of a milk-clotting protease from Mucor baciliformis. Applied Biochemistry and Biotechnology 37, 283–294. Beebe, K.R., Pell, R.J. & Seasholtz, M.B. 1998 Chemometrics – A Practical Guide. Wiley: New York. ISBN 0471124516. Cavalcanti, M.T.H., Teixeira, M.F.S., Lima Filho, J.L. & Porto, A.L.F. 2004 Partial purification of new milk-clotting enzyme produced by Nocardiopsis sp. Bioresource Technology 93, 29–35.

154 Escobar, J. & Barnett, S.M. 1993 Effect of agitation speed on the synthesis of Mucor miehei acid protease. Enzyme and Microbial Technology 15, 1009–1013. Hashem, A.M. 1999 Optimization of milk-clotting enzyme productivity by Penicillium oxalicum. Bioresource Technology 70, 203– 207. Lealem, F. & Gashe, B.A. 1994 Amylase production by a Grampositive bacterium isolated from fermenting tef (Eraglostis tef). Journal of Applied Bacteriology 77, 348–352. Miller, G.L. 1960 Dinitrosalicylic acid reagent for determination of reducing sugar. Analytical Chemistry 31, 426–428. Moreira, K.A., Cavalcanti, M.T.H., Duarte, H.S., Tambourgi, E.B., Melo, E.H.M., Silva, V.L., Porto, A.L.F. & Lima-Filho, J.L. 2001 Characterization partial of proteases from Streptomyces clavuligerus using an inexpensive medium. Brazilian Journal of Microbiology 32, 215–220.

M.T.H. Cavalcanti et al. Moreira, K.A., Albuquerque, B.F., Teixeira, M.F.S., Porto, A.L.F. & Lima-Filho, J.L. 2002 Application of protease from Nocardiospis sp. as a laundry detergent additive. World Journal of Microbiology and Biotechnology 18, 307–312. Porto, A.L.F., Campos-Takaki, G.M. & Lima-Filho, J.L. 1996 Effects of culture conditions on protease production by Streptomyces clavuligerus growing soybean flour medium. Applied Biochemistry and Biotechnology 60, 115–122. Pridham, T.G., Anderson, P., Foley, C., Lindenfelser, L.A., Hesseltine, C.W. & Bendict, R.G. 1957 A selection of media for maintenance and taxonomic study of Streptomycetes. Antibiotics Annual 1, 947–953. Tsujibo, H., Miyamoto, K., Hasegawa, T. & Inamori, Y. 1990 Purification and characterization of two types of alkaline serine proteases produced by an alkalophilic actinomycete. Journal of Applied Bacteriology 69, 520–529.

World Journal of Microbiology & Biotechnology 2005 21: 155–162 DOI: 10.1007/s11274-004-2890-0


Springer 2005

Diversity and dynamics of bacterial species in a biofilm at the end of the Seoul water distribution system Dong-Geun Lee1, Jung-Hoon Lee2 and Sang-Jong Kim2,* 1 Department of Bioscience and Biotechnology, College of Engineering, Silla University, Kwaebop-dong 1-1, Sasang-Gu Busan 617-736, Republic of Korea 2 School of Biological Sciences, College of Natural Sciences, Seoul National University, Seoul 151-742, Republic of Korea *Author for correspondence: Tel.: +82-2-880-6704, Fax: +82-2-889-9474, E-mail: [email protected] Received 21 April 2004

Keywords: 16S rDNA, biofilm, denaturant gradient gel electrophoresis (DGGE), drinking water distribution system, E. coli, fatty acid methyl ester (FAME)

Summary To investigate changes in the bacterial species and hygienic safety of the biofilm at the end of the drinking water distribution system in Seoul (Korea), denaturing gradient gel electrophoresis (DGGE) and DNA sequencing were used to analyse the bacterial population in the biofilm of a semi-pilot galvanized iron pipe model. The presence of sequences from aerobic Sphingomonas sp., anaerobic Rhodobacter sp., and unculturable bacteria indicated that these organisms coexisted after 1 day of model operation, demonstrating the ease of biofilm formation on galvanized iron pipes in the end region of the water distribution system studied. Sequences similar to those of unculturable bacteria, E. coli, and anaerobic bacteria were detected during the course of succession on the biofilm. More complicated band patterns were observed after 70 days of operation. PCR-DGGE illustrated changes in the biofilm during succession as well as the possibilities of anaerobic conditions and faecal contamination of the drinking water system. PCR-DGGE and culture-dependent fatty acid methyl ester (FAME) analysis showed different patterns for the same samples (Lee & Kim 2003); however, PCR-DGGE showed less diversity than did FAME analysis. This study compared the culture-dependent FAME and culture-independent PCR-DGGE methods directly, and their use in promoting the hygienic safety of drinking water.

Introduction In drinking water distribution systems, bacteria can live in the water phase in a planktonic state and on surfaces as a biofilm. Biofilms have been described as Ôcities of microbesÕ (Watnick & Kolter 2000). Bacterial regrowth and biofilms are issues of concern with regard to drinking water. Although chlorine disinfection virtually eliminates the possibility of an outbreak of waterborne bacterial disease, pathogenic bacteria have been detected in drinking water biofilms (Park et al. 2001). Culture-dependent methods have been used to study microbial biofilms in drinking water distribution systems (Norton & LeChevallier 2000; Lee & Kim 2003). A considerable handicap in the investigation of water microbes has been the inability to cultivate many environmental bacteria by conventional laboratory techniques (Lee et al. 1996), necessitating the development of new analytical methods to overcome this complication. The polymerase-chain reaction (PCR) is one solution for overcoming the culture problem in describing microbial community structure and diversity

(Lee et al. 1996; Cho & Kim 2000; Ralebitso et al. 2003). Although the PCR products of organisms might be the same size, differences in the nucleotide sequences between organisms can be used to resolve electrophoretic PCR fingerprints. This method has the advantage of accommodating many samples at once, and it can be extended to unculturable bacteria (Muyzer et al. 1993; Lee et al. 1996; Clement et al. 1998). Since the development of denaturant gradient gel electrophoresis (DGGE) for the study of microbial communities (Muyzer et al. 1993) researchers have used DGGE to study microbial mats (Norris et al. 2002), soil (Ralebitso et al. 2003), wastewater (Rowan et al. 2003), grassland (Clegg et al. 2003), drinking water biofilters (Fonseca et al. 2001), and to compare bacterial species (Rowan et al. 2003). The dynamics of biofilms and the recovery of coliform and Enterococcus species at the end of a drinking water distribution system were examined in an extensive culture-dependent fatty acid methyl ester (FAME) study (Lee & Kim 2003). Micrococcus, Bacillus, and

156 Pseudomonas sp. were detected more frequently in the FAME analysis. Staphylococcus, Arthrobacter, Acinetobacter, and Cellulomonas sp. were also detected frequently in the identification of more than 1500 isolates; however, 4–43% of bacteria remained unidentified on different sampling days (Lee & Kim 2003). The presence of many unidentified bacteria in culturedependent FAME analysis (Lee & Kim 2003) suggests the need for other methods of analysis of the bacterial community in the biofilm. Ralebitso et al. (2003) detected changes that were not seen using the conventional plate culture method. In addition, the recovery of coliforms and Enterococcus (Lee & Kim 2003) suggests the presence of pathogenic bacteria, which may not be detected using culture-dependent methods (Rozak et al. 1984; Kapley et al. 2000). The combination of 16S rDNA analysis and DGGE has been demonstrated to be a powerful tool for monitoring community shifts and environmental bacteria, as mentioned above. Hence, this study aimed to describe bacterial species and the state of hygiene in a biofilm by PCR-DGGE targeted at 16S rDNA in a culture-independent manner, using the same samples as used previously (Lee & Kim 2003).

Materials and methods Semi-pilot model system A semi-pilot model system, with an internal diameter of 13 mm and a length of 8.8 m, was constructed from pipes and fittings of galvanized iron (Figure 1) and connected to a tap at Seoul National University, which is located at the end of a water distribution system. The water distributed through this system had undergone conventional water treatment (e.g., coagulation, precipitation, filtration, and disinfection) at a water treatment

Dong-Geun Lee et al. plant (Lee & Kim 2003). Twelve galvanized iron pipe (GIP) wafers (13 · 75 · 1.0 mm) were sonicated, autoclaved, and aseptically inserted into the pipe system. Both sides of the wafers were exposed to tap water. Water flowed along the wafers lengthwise in a continuous one-way flow at a dilution rate of 0.43–0.51 min)1 (0.5–0.6 l min)1). The sequence of sampling was from the effluent point of the GIP model system (Lee & Kim 2003). Sampling and recovery of the bacterial community All of the chemicals used for sampling and recovery of the bacterial community, DNA extraction, DNA purification, and DGGE were purchased from Sigma or Amersham. A single wafer was obtained for each sampling day (days 1, 3, 7, 14, 21, 28, 35, 42, 49, 56, 70, and 84). Wafers were washed with detaching buffer solution [10)3 M ethylene-glycol-bis( ß-aminoethyl ether)-tetraacetic-acid (EGTA), 0.01% (w/v) peptone] to detach reversibly bound bacteria after aseptic sampling. To detach bacteria from wafers, three cycles of mild sonication (Branson 3210, Krautkramer Branson Inc., Cincinnati, OH, USA), chilling on ice, vortexing, and further standing on ice, for 1 min per step, were performed in detaching buffer (Lee & Kim 2003). We confirmed the effectiveness of this detachment cycle by the stagnation of the increase in heterotrophic plate count (HPC) after three cycles and by electron microscopic observation (data not shown). HPC were done by standard plate-count agar (PCA) and minimal R2A agar. Colony-forming units (c.f.u.) were counted after culture at 37 C for 48 h in PCA and at 20 C for 1 week in R2A agar. After acridine orange staining, total bacteria were counted directly by fluorescence microscopy (Zeiss, Go¨ttingen, Germany) following standard methods (APHA 1995). After treatments for HPC and direct count (Lee & Kim 2003), the bacterial fractions were recovered by centrifugation (10,000 · g for 20 min), washed with 10 ml of phosphate-buffered saline (10 mM, pH 7.0), and collected by a second centrifugation. The resulting pellet was stored in a deep freezer ()76 C, Sanyo, Japan) until used for DNA extraction. DNA extraction and purification and PCR

Figure 1. Schematic diagram of the galvanized-iron-pipe (GIP) model system (diameter ¼ 13 mm). The influent was tap water at a dilution rate of 0.43–0.51 min)1 (0.5–0.6 l min)1). Each wafer was inserted separately (arrow point in GIP); the direction of the arrow represents the direction of water flow. The isolated coupon at each sampling time is marked in italics (Lee & Kim 2003).

Bacterial cells were destroyed with a bead beater and three freeze–thaw cycles in boiling water and liquid nitrogen (Miller et al. 1999). Cell destruction was confirmed by microscopic observation. DNA was purified on a Sephadex G-50 column after phenol–chloroform extraction. DNA was amplified with the primers 534r (5¢-ATTACCGCGGCTGCTGG-3¢) and gc338f, which span the V3 region of the 16S rDNA. The forward primer consisted of primer 338f (5¢-ACTCCTACGGGAGGCAGCAG-3¢) attached to the 3¢-end of a GC clamp (5¢-CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGG-3¢). PCR

Bacteria in biofilm of drinking water by DGGE

157

was performed with a Perkin-Elmer model 2400 thermal cycler. Touchdown PCR was performed as described by Muyzer et al. (1993). We used a commercial mixture of Pfu and Taq polymerases (Promega) for fidelity. PCR was performed in triplicate at different times, and a mixture of the PCR products was purified. For DGGE, purified PCR products from the same sampling day were mixed to diminish PCR bias. DGGE After measuring the concentrations of the purified PCR products, an equal mass (about 200 ng) of each purified PCR product was mixed with 20 ll of DGGE loading dye (70% (w/v) glycerol). DGGE was performed using the Bio-Rad DGene apparatus according to the procedure first described by Muyzer et al. (1993) with 8% (w/v) polyacrylamide gels. The range of denaturant [100% denaturant corresponded to 7 M urea (Sigma) and 40% (vol/vol) deionized formamide (Sigma)] was from 35 to 70%. Electrophoresis was run at 60 C for 15 min at 70 V for stacking and then for 6 h at 120 V. Staining and statistical analysis of DGGE bands The gels were stained with ethidium bromide and photographed using a u.v. transilluminator (Eagle Eye II, Stratagene, USA); the stored gel image files were analysed with Bio 1D++ (Bioprofil, Paris, France). Dice’s similarity coefficients were determined to compare all profiles. The unweighted pair group method with mathematical averages (UPGMA) was used to create a dendrogram describing the pattern analysis (Sokal & Sneath 1963; Cho & Kim 2000). Band identification and nucleotide sequence accession numbers DNA fragments to be sequenced (small letters in Figure 2b) were excized from the gel following elution with 25 ll of sterilized deionized water at 4 C. The eluent was reamplified with the primer set 338f-534r (no GC clamp), purified, and ligated into a T-vector (Promega) for sequencing by a capillary-type automatic DNA sequencer (ABI PRISM 3100 Genetic Analyser, Applied Biosystems, California). Partial 16S rDNA sequences were compared with known sequences in the BLAST database (http://www.ncbi.nlm.nih.gov/). Clone sequences have been deposited in GenBank under the accession numbers given in Table 1.

Results and discussion Physicochemical properties Previously, we reported the water qualities of influent and effluent and the results of culture-dependent FAME analysis of the biofilm using the same samples studied

Figure 2. Negative image of DGGE (35–70% denaturant) (a), schematic pattern of DGGE bands (b), and UPGMA tree of similarity (%) among sampling days based on Dice’s method (c), representing biofilm succession on galvanized iron wafers in tap water (0.3 mg l)1 or less residual chlorine). Black triangles in (a) and (b) represent the same positions in each figure, and the small letters in each profile (Figure 2b) are the sequenced bands. PCR-DGGE was performed for the V3–16S rDNA region. Abbreviation: SD – reference standard for DGGE.

here (Lee & Kim 2003). Briefly, the residual chlorine concentration was below 0.3 mg l)1, and the water temperatures were in the range of 16.0–25.1 C (mean 20.1 C). The total organic carbon concentrations of the influent and effluent were similar (mean, 1.96 mg l)1; range, 1.27–2.65 mg l)1). A low chlorine concentration and a high water temperature imply a high possibility of microbiological water deterioration.

158

Dong-Geun Lee et al.

Table 1. Sequence analyses of bands derived from the bacterial 16S rDNA V3 region on PCR-DGGE gels of biofilm samples from drinking water containing 0.3 mg l)1 or less free chlorine. DGGE band (Accession no.)

Phylogenetic group

Most closely related bacterial sequence

Sequence length (bp)

% Similarity

GenBank accession no.

D1-a, D7-a (AY214392) D1-b (AY214393) D1-c (AY214394) D1-d (AY214395) D1-e (AY214396) D1-f (AY214397) D3-a (AY214398) D3-b (AY214399) D3-c, D21-c (AY214400) D3-d (AY214401) D7-b

a

Uncultured alpha proteobacterium Sphingomonas spp. (248–420) Uncultured Duganella spp. (307–504) Uncultured beta proteobacterium clone Uncultured beta proteobacterium (322–490)

173

98

198

98

169

99

AY144206 AY026948 AY129794 AY043674 AY043784

185

97

(AY214402) D14-a (AY214403) D21-a (AY214404) D21-b (AY214405) D42-a (AY214406) D42-b (AY214407) D42-c

b b

170

99

b

Uncultured beta proteobacterium (292–477) Zoogloea sp. (302–483) Uncultured bacterium Rhodobacter gluconicum (301–470) Uncultured beta proteobacterium (293–490)

198

97

AY043784 D84590 AJ504015 AB077986 AY043784

b

Uncultured beta proteobacterium (293–490)

198

98

AY043784

b

Uncultured beta proteobacterium (306–490)

186

98

AY043784

b

Uncultured beta proteobacterium (292–490)

197

98

AY043784

b

198

98

198

98

b

Uncultured beta proteobacterium (293–490) Janthinobacterium agaricidamnosum (304–501) Uncultured Duganella sp. (307–504) Uncultured beta proteobacterium clone Agricultural soil bacterium Uncultured bacterium clone Can-H19(2)-A-2

165

98

AY043784 Y08845 AY129794 AY043784 AS0252575 AF203838

b

Uncultured beta proteobacterium (322–519)

198

99

AF445688

c

Pseudomonas spp. (269–466) Uncultured gamma proteobacterium clone Uncultured beta proteobacterium (305–486)

198

99

182

96

AF456220 AY133072 AY043784

173

98

198

99

197

98

198

100

179

92

173

100

171

95

174

97

201

99

174

9291

173

97

199

99

181

94

b a

b

b a c

(AY214408) D56-a (AY214409) D70-a (AY214410) D70-b (AY214411) D70-c (AY214412) D70-d

a

(AY214413) D70-e

a

(AY214414) D70-f (AY214415) D70-g (AY214416) D84-a (AY214417) D84-b (AY214418) D84-c (AY214419)

b c b a

c Clostridia a c b

Uncultured bacterium Rhodobacter gluconicum (301–473) E. coli Shigella flexneri Uncultured bacterium clone Uncultured beta proteobacterium (293–491) Pseudomonas spp. (269–466) Uncultured gamma proteobacterium Zoogloea spp. (303–479) Uncultured beta proteobacterium (293–469) Uncultured bacterium clone 1CBTE8 Sphingomonas spp. (278–450) Hyphomicrobium spp. (279–449) Uncultured Eubacterium WD238. Acidocella spp. Sphingomonas spp. (242–415)

Novosphingobium aromaticivorans (5823–5650) Pseudomonas spp. (268–466) Uncultured gamma proteobacterium Uncultured bacterium (326–498) Eubacterium formicigenerans (353–525) Uncultured bacterium Rhodobacter gluconicum (300–470) Pseudomonas spp. (269–466) Uncultured gamma proteobacterium Uncultured beta proteobacterium clone (175–355)

AJ504515 AB077986 AE016770 AE015391 AY135915 AY043784 AF456220 AY133072 D84590 AY043784 AF390907 AB047364 AF534572 AJ292597 X91797 AF361178 NZ_AAAV01000 115 AF408901 AY048886 AF371595 L34619 AJ504015 AB077986 AF456220 AY133072 AF35156920

159

Although DGGE discriminates mainly among sequence differences, the lengths of the PCR products might interfere with the electrophoretic movement if many bands and a wide denaturant gradient range are present. Owing to its sequence variability, the V3 region of the 16S rRNA gene might give sufficient information and might permit the discrimination of bands with a single run of a wide denaturant gradient. Hence, until recently, this short sequence has been used to assess the biodiversity of environmental samples by DGGE (Kurisu et al. 2002; Plant et al. 2003). For this reason, it was also used in this study. The negative image of the DGGE profile and the schematic pattern of the DGGE bands are shown in Figure 2. Intense bands created difficulties in band discrimination and were detected throughout the experimental period under the same denaturant condition (upper arrow in Figure 2a). The same sequences were derived from the multiple sequencing T-vector of each intense band in Figure 2a. The photograph of the DGGE profile is unclear, as the intense bands overlapped other bands upon long exposure (Figure 2a). Replicated DGGE patterns were similar, and DGGE using a smaller amount of purified PCR product led to the disappearance of many weak bands. Figure 2a shows the best example of both the intensity and number of bands simultaneously. Unique intense bands were observed at days 1, 42, and 84 (arrows in Figure 2b). Many bands were present even in the day-1 sample. Therefore, biofilm formation might begin as soon as the pipe materials are exposed to water (O’Toole et al. 2000; Lee & Kim 2003). The DGGE profiles from day 3 to day 35 were somewhat similar, except for that of day 14 (Figure 2a). The PCR-DGGE profile was very similar from day 21 to day 35 (Figure 2), and the total bacterial number (TBN) showed an increase during this period (Figure 3a). These findings imply that similar populations, or similar major populations, showed growth during this period. However, a higher number of bacterial species was detected by FAME (Figure 3b). Different DGGE profiles were observed during the succession of the bacterial biofilm, especially from day 35 to day 70 (Figure 2a). The DGGE profiles were more diverse at days 70 and 84 than at any other time during the study. Remarkable changes in the band patterns (Figure 2a) and bacterial concentrations (Figure 3a) were observed in the samples taken on days 35, 42, and 49. A 1-log decrease in the heterotrophic plate count (HPC) was observed in nutrient-rich PCA medium, while 2-log increases in TBN and HPC were seen in nutrient-poor R2A medium. Reasoner (2004) stated that R2A is of limited use as quality control data and is very variable (1–5 log10), although R2A is much more sensitive than PCA for measuring biological activity. Hence, it is not proper to discuss the HPC level only. Perhaps the microbial population fluctuated with the result of DGGE (Figure 2a) and the variability of TBN (Fig-

10

11

10

6

`

10

2

0 7 14 21 28 35 42 49 56 63 70 77 84 day

Number of identified species and DNA bands

Analysis of biofilm microbial community structure by PCR-DGGE

Bacterial concentraion in biofilm (CFU/cm2, cells/cm2)

Bacteria in biofilm of drinking water by DGGE

20

15

10 5

0 0

7 14 21 28 35 42 49 56 63 70 77 84 day

Figure 3. The bacterial concentration of biofilm (a). Bacterial concentrations were estimated by eutrophic PCA medium (s, c.f.u. cm)2) and minimal R2A medium (x, c.f.u. cm)2) and by total bacterial number (d, cells cm)2) during the succession of the biofilm on galvanized iron wafers. The number of identified bacterial species was determined by FAME (r, Lee & Kim 2003), and DNA bands were determined by DGGE ()) (b).

ure 3a) during this period. Bacterial growth rates are temperature-dependent and the water temperature exceeded 20 C from day 28. Moreover, it exceeded 24 C from day 70 on, when the TBN decreased (Figure 3a). Other changes could also have resulted in fluctuation. Many reports have related the formation and dissolution of biofilms to the trophic state of the biofilm and the environment (Sartory & Holmes 1997; Allison et al. 1998; O’Toole et al. 2000). However, we could not confirm the trophic state of the biofilm or identify other factors that contributed to the detachment of bacteria. Figure 2c shows the unweighted pair group method with mathematical averages (UPGMA) tree of similarity (%) based on Dice’s method across sampling days. It was impossible to do a bootstrap analysis, as the program Bio 1D++ does not offer the bootstrap option. There was high similarity on days 21, 28, and 35. Slight similarity was observed on days 1, 3, and 56, days 14, 42, and 49, and days 70 and 84. Analysis of the predominant bacterial species by PCR-DGGE The dominant and unique bands of each profile were analysed to provide general information on the

160 dominant and unique species, respectively (Table 1). Migration in DGGE does not depend on fragment size. The similarities between the closest sequences ranged from 92 to 100%. Many sequences were similar to those of unculturable or unidentified bacteria; they were closest to the 16S rDNA sequences of a-, b-, and cProteobacteria. D1-a (band a of the day-1 sample, Figure 2b) and D1-e (band e of the day-1 sample, Figure 2b), sequences closest to Sphingomonas (Accession : AY026948) and anaerobic Rhodobacter (AB077986), respectively, were detected in the day-1 sample (Table 1). It has been reported that Sphingomonas is a pioneer group in biofilm formation (Gauthier et al. 1999). Rhodobacter is an anaerobic bacterium, so it was surprising that the sequence closest to it was detected on day 1. However, Rhodobacter can produce capsules and slime, which may account for its presence. Rhodobacter may have originated upstream of the semipilot model, in the water distribution system under anaerobic conditions. Clustering may facilitate downstream delivery of this anaerobe. Park (1993) observed clusters of Rhodobacter cells in drinking water. Kapley et al. (2000) tried using a PCR technique to detect pathogenic bacteria in drinking water, since water-borne pathogens are difficult to detect using culture-dependent methods (Rozak et al. 1984). It has been reported that E. coli is the best biological indicator of faecal contamination (Edberg et al. 2000). A sequence (D42-c) close to those of E. coli (AE016770) and Shigella (AE015391) was detected at day 42. Hence, our results might be regarded as indicating possible of faecal contamination. Such bacteria would serve as a persistent source of water contamination in a distribution system. Selective media and FAME analysis indicated the presence of Enterobacteriaceae instead of E. coli in the same samples (Lee & Kim 2003), but the origin of these organisms might be environmental, and not faecal material. The patterns observed on days 70 and 84 were similar, with more diverse bands than those of other samples. Sequences of oligotrophic a-Proteobacteria (D70-c, D70-d, and D70-e), eutrophic c-Proteobacteria (D70-a, D70-f), and b-Proteobacteria (D70-b) were observed in the day-70 sample. Also present were sequences of Zoogloea (D70-b) and Hyphomicrobium (D70-d), which are stalked bacteria believed to be abundant only under nutrient-limited conditions. These species have been observed in samples of drinking water by electron microscopy (Ridgway & Olson 1981). Molecules close to those of Sphingomonas species were observed again on day 70. On day 84, sequences similar to aerobic bacteria (D84-b, D84-c) and the anaerobic Rhodobacter genus (D84-a) were detected. The presence of many diverse bands and the coexistence of sequences similar to those of aerobic and anaerobic bacteria suggests that the biofilm had a complex structure. There may have been anaerobic and aerobic conditions with high bacterial diversity (Table 1). The similarities of the findings on days 70 and 84, combined with the low similarities

Dong-Geun Lee et al. among samples taken at other times (Figure 2c), suggest that there may be different stages in biofilm development on different sampling days. Comparisons of FAME and DGGE Ralebitso et al. (2003) found that DGGE detected changes that were not detected using conventional plate culture, as we also report. Table 2 shows the bacteria identified by FAME (Lee & Kim 2003) and the sequences found in this study. Micrococcus, Bacillus, and Pseudomonas sp. were more frequently detected in FAME analysis. Staphylococcus, Arthrobacter, Acinetobacter, and Cellulomonas sp. were also detected frequently. However, of these organisms, DGGE detected only the sequence of Pseudomonas species. The sequence closest to Pseudomonas was first detected on day 21 (D21-b, Figure 2b) and appeared again on days 70 (D70-a, f) and 84 (D84-b) with DGGE, while FAME detected it continuously except on days 21, 42, and 84 (Lee & Kim 2003). One frequently detected band (D1-b) was closest to the sequence of uncultured Duganella sp. (GenBank accession:AY129794) or uncultured b-Proteobacteria (AY043674) (Table 1). Manz et al. (1993) found that bacteria belonging to the b-Proteobacteria group were predominant in a water distribution system, while c-Proteobacteria made up less than 5% of the community in the biofilm when assessed by a culture method. The dynamics of detectable bacterial species and sequences were different in FAME and DGGE (Figure 3b). FAME showed more diversity than did DGGE, and DGGE showed an increased number of bands on days 70 and 84, unlike FAME. Actinobacteridae, Bacillales, and Lactobacillales were not detected by DGGE. Only one sequence (D70-g) close to Eubacterium sp. (L34619, Clostridia) was detected with 91% similarity in DGGE. Destruction of most cells was observed by microscopic observation before the DNA extraction step. Hence, these findings might reflect the true state of the samples studied. Namely, some bacteria in the samples might have become concentrated during culture, although they were not abundant in the natural state. Members of the Actinobacteridae, Bacillales, and Lactobacillales might belong to this category in these samples (Lee & Kim 2003). Lee et al. (1996) found that the microbial community differed between culturedependent and culture-independent methods and they postulated that this was the result of a bias in cultures. However, PCR-DGGE could also be biased in terms of the primer specificity of PCR or there might be a bias in PCR, although we collected three PCR different samples to overcome this problem. Molecules close to b-Proteobacteria were frequently detected (Table 1). Oligotrophic a- and eutrophic c-Proteobacteria were identified by DGGE. DGGE also detected E. coli and the anaerobic Rhodobacter genus, which were not detected by FAME analysis. These findings might reflect anaerobic conditions, potential problems, complexity, or high bacterial diversity of the

161

Bacteria in biofilm of drinking water by DGGE Table 2. Comparison of culture-dependent FAME and culture-independent DGGE analyses of the same biofilms. Bacterial identification

Recovered bacteria or sequences on biofilm FAME (Lee & Kim 2003)

DGGE (in this study)

a-Proteobacteria

Brevundimonas Methylobacterium Xanthobacter

Rhodobacter Sphingomonas

b-Proteobacteria

Acidovorans Alcaligenes Aquaspirillium Burkholderia Comamonas Hydrogenophaga Variovorax

Duganella Janthinobacterium Zoogloea 11 uncultured bacteria

c-Proteobacteria

Acinetobacter Hafnia Pantoea Proteus Pseudomonas Stenotrophomonas Xenorhabdus

Escherichia Pseudomonas

Clostridia

Eubacterium

Actinobacteridae

Amycolatopsis Arthrobacter Aureobacterium Cellulomonas Corynebacterium Curtobacterium Gordona Micrococcus Nocardia Rathayibacter Rhodococcus

Bacillales

Bacillus Paenibacillus Staphylococcus

Lactobacillales

Enterococcus Lactobacillus Pediococcus Weisella

The bacteria listed in the DGGE column are those closest to the sequence of the band.

biofilm in this study. FAME and DGGE provided different views of bacterial diversity and composition (Table 2). Here, we tried to obtain more information by using culture-independent PCR-DGGE profiling and sequencing. This study presents an opportunity to directly compare culture independent PCR-DGGE and the culture-dependent FAME method, and their employment in deciding the hygienic safety of drinking water. The combination of culture-independent and culturedependent methods may be complementary. FAME revealed more diverse bacteria than did DGGE in these samples, although this may relate to the number of samples analysed; more than 1500 isolates were analysed using FAME, which was impossible using DGGE. However, DGGE saves time and can contribute to the understanding of bacterial diversity by detecting unculturable bacteria (Ralebitso et al. 2003).

References Allison, D.G., Ruiz, B., SanJose, C., Jaspe, A. & Gilbert, P. 1998 Extracellular products as mediators of the formation and detachment of Pseudomonas fluorescence biofilms. FEMS Microbiology Letters 167, 179–184. APHA, 1995 Standard Methods for the Examination of Water and Wastewater, 19th edn. Washington DC: American Public Health Association. ISBN 0875532233. Cho, J.-C. & Kim, S.-J. 2000 Computer-assisted PCR-single-strandconformation polymorphism analysis for assessing shift in soil bacterial community structure during bioremediational treatments. World Journal of Microbiology and Biotechnology 16, 231–235. Clegg, C.D., Lovell, R.D.L. & Hobbs, P.J. 2003 The impact of grassland management regime on the community structure of selected bacterial groups in soils. FEMS Microbiology Ecology 43, 263–270. Clement, B.G., Kehl, L.E., DeBord, K.L. & Kitts, C.L. 1998 Terminal restriction fragments patterns TRFPs, a rapid, PCR-based method for the comparison of complex bacterial communities. Journal of Microbiological Methods 31, 135–142.

162 Edberg, S.C., Rice, E.W., Karlin, R.J. & Allen, M.J. 2000 Escherichia coli, the best biological drinking water indicator for public health protection. Journal of Applied Microbiology Symposium Supplement 88, 106S–116S. Fonseca, A.C., Summers, R.S. & Hernandez, M.T. 2001 Comparative measurements of microbial activity in drinking water biofilters. Water Research 35, 3817–3824. Gauthier, V., Redercher, S. & Block, J.-C. 1999 Chlorine inactivation of Sphingomonas cells attached to goethite particles in drinking water. Applied and Environmental Microbiology 65, 355–357. Kapley, A., Lample, K. & Purohit, H.J. 2000 Development of duplex PCR for the detection of Salmonella and Vibrio in drinking water. World Journal of Microbiology and Biotechnology 16, 457–458. Kurisu, F., Satoh, H., Mino, T. & Matsuo, T. 2002 Microbial community analysis of thermophilic contact oxidation process by using ribosomal RNA approaches and the quinone profile method. Water Research 36, 429–438. Lee, D.-G. & Kim, S.-J. 2003 Bacterial species in biofilm cultivated from the end of the Seoul water distribution system. Journal of Applied Microbiology 95, 317–324. Lee, D.-H., Zo, Y.-G. & Kim, S.-J. 1996 Nonradioactive method to study genetic profiles of natural bacterial community by PCRsingle-strand-conformation polymorphism. Applied and Environmental Microbiology 62, 3112–3120. Manz, W., Szewzyk, U., Ericsson, P., Amann, R., Schleifer, K.H. & Stenstrom, T.A. 1993 In situ identification of bacteria in drinking water and adjoining biofilms by hybridization with 16S and 23S rRNA-directed fluorescent oligonucleotide probes. Applied and Environmental Microbiology 59, 2293–2298. Miller, D.N., Bryant, J.E., Madsen, E.L. & Giorse, W.C. 1999 Evaluation and optimization of DNA extraction and purification procedures for soil and sediment sample. Applied and Environmental Microbiology 65, 4715–4724. Muyzer, G., de Wall, E.C. & Uitterinden, A.G. 1993 Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chin reaction-amplified genes coding for 16S rRNA. Applied and Environmental Microbiology 59, 695–700. Norris, T.B., McDermott, T.R. & Castenholz, R.W. 2002 The longterm effects of U.V. exclusion on the microbial composition and

Dong-Geun Lee et al. photosynthetic competence of bacteria in hot-spring microbial mats. FEMS Microbiology Ecology 39, 193–209. Norton, C.D. & LeChevallier, M.W. 2000 A pilot study of bacteriological population changes through potable water treatment and distribution. Applied and Environmental Microbiology 66, 268–276. O’Toole, G., Kaplan, H.B. & Kolter, R. 2000 Biofilm formation as microbial development. Annual Review of Microbiology 54, 49–79. Park, S.J. 1993 Bacterial regrowth and biofilm formation in water distribution systems. PhD thesis, Seoul National University, Seoul. Park, S.R., Mackay, W.G. & Reid, D.C. 2001 Helicobacter sp. recovered from drinking water biofilm sampled from a water distribution system. Water Research 35, 1624–1626. Plant, L., Lam, C., Conway, P.L. & O’Riordan, K. 2003 Gastrointestinal microbial community shifts observed following oral administration of a Lactobacillus fermentum strain to mice. FEMS Microbiology Ecology 43, 133–140. Ralebitso, T.K., Yamazoe, A., Ro¨ling, W.F.M., Braster, M., Senior, E. & van Verseveld, H.W. 2003 Insights into bacterial associations catabolizing atrazine by culture-dependent and molecular approaches. World Journal of Microbiology and Biotechnology 19, 59–67. Reasoner, D.J. 2004 Heterotrophic plate count methodology in the United States. International Journal of Food Microbiology 92, 307–315. Ridgway, H.F. & Olson, B.H. 1981 Scanning microscopic evidence for bacterial colonization of a drinking water distribution system. Applied and Environmental Microbiology 41, 274–287. Rowan, A.K., Snape, J.R., Fearnside, D., Barer, M.R., Curtis, T.P. & Head, I.M. 2003 Composition and diversity of ammonia-oxidising bacterial communities in wastewater treatment reactors of different design treating identical wastewater. FEMS Microbiology Ecology 43, 195–206. Rozak, D.B., Grimes, D.J. & Colwell, R.R. 1984 Viable but nonrecoverabe stage of Salmonella enteritidis in aquatic systems. Canadian Journal of Microbiology 30, 334–338. Sartory, D.P. & Holmes, P. 1997 Chlorine sensitivity of environmental, distribution system and biofilm coliforms. Water Science and Technology 35, 289–292. Sokal, R.R. & Sneath, P.H.A. 1963 Principles of Numerical Taxonomy. San Francisco, CA: Freeman. Watnick, P. & Kolter, R. 2000 Biofilm, city of microbes. Journal of Bacteriology 182, 2675–2679.

World Journal of Microbiology & Biotechnology 2005 21: 163–167 DOI: 10.1007/s11274-004-1563-3

 Springer 2005

Utilization in alginate beads for Cu(II) and Ni(II) adsorption of an exopolysaccharide produced by Chryseomonas luteola TEM05 Guven Ozdemir1,*, N. Ceyhan2 and E. Manav3 1 Faculty of Science, Department of Biology, Ege University, 35100 Izmir, Turkey 2 Faculty of Science and Letters, Department of Biology, Mugla University, 48147 Mugla, Turkey 3 Faculty of Engineering, Department of Bioengineering, Ege University, 35100 Izmir, Turkey *Author for correspondence: Tel.: +90-232-3884000-1519, Fax: +90-232-3881036, E-mail: [email protected] Received 21 April 2004

Keywords:

Alginate, biosorption, Chryseomonas luteola, copper, extracellular polysaccharide, nickel

Summary Copper and nickel adsorption onto calcium alginate, sodium alginate with an extracellular polysaccharide (EPS) produced by the activated sludge bacterium Chryseomonas luteola TEM05 and the immobilized C. luteola TEM05 from aqueous solutions were studied. After that, the multi metal ions containing these ions together were prepared and partial competitive adsorptions of these mixtures were also investigated. The metal adsorption of gel beads were carried out at pH 6.0, 25
C. The maximum adsorption capacities in Langmuir isotherm for calcium alginate, calcium alginate + EPS, calcium alginate + C. luteola TEM05 and calcium alginate + EPS + C. luteola TEM05 were 1.505, 1.989, 1.976, 1.937 mmol/g dry weight for Cu(II) and 0.996, 1.224, 1.078, 1.219 mol/g dry weight for Ni(II), respectively. The competitive biosorption capacities of the carrier for all metal ions were lower than single conditions.

Introduction Biosorption may be a suitable wastewater technology to remove heavy metals as demonstrated by several researchers because it is possible the use cheap adsorption materials that can be competitive with the conventional technologies (Volesky 1989; Kratochvil & Volesky 1998). Biopolymers, which can be used for entrapping microorganisms, are known to bind metal ions (Lazaro et al. 2003). Thus far, isolated biopolymers for heavy metal remediation have not been applied on a large scale, although synthetic polymers have been used for various precipitation treatments. However, polyelectrolyte complexes formed by mixing polysaccharides of opposite charge, have recently attracted considerable attention because of their potential in various biotechnological applications (Hugerth et al. 1997; Jianlong & Yi 1999). It seems likely that incorporating exocellular polysaccharides or polyelectrolyte complexes into biofilter technology may provide applications for remediation, although much depends on the economics of such treatments (Gutnick & Bach 2000). The main difficulty in using dead microbial biomass as a biosorbent is the small particle size and the low mechanical strength of the native biomass. It has been reported that biomass immobilisation into particles of a desirable size, mechanical strength and biosorptive

characteristics is the best way to apply the process of biosorption for metal value recovery from process or waste solutions (Ferguson et al. 1989; Tsezos et al. 1989). Therefore, polysaccharide gel immobilized microorganisms can be used to remove heavy metal ions from aqueous solutions, providing an alternative to single dead microbial biomass for wastewater treatment (Sag et al. 1995; Veglio et al. 2002; Arca et al. 2003). Natural polymers such as alginate, chitosan, chitin and cellulose derivates have been mostly used as the matrix for the immobilization of microbial cells via an entrapment technique. These polymers are also known to bind metal ions strongly. Entrapment of microbial cells in these polymer supports could also enhance microbial cell performance and adsorptive capacity of the biosorbent system for the heavy metal ions (Christ et al. 1994; Sag et al. 1995; Jianlong et al. 2000). In this study, the use together with alginate of EPS as a new biomaterial for metal adsorbent was investigated. This choice has been dictated by the ability of microorganism to form a polysaccharide capsule, which is responsible for the binding and accumulation metal ions in the form of superficial mucilage layer. Since exocellular polysaccharide (EPS) plays a role in the increase of heavy metal adsorption capacity, the EPS, biomass and alginate mixtures were made into a gel bead and the adsorption of metal ions by these EPS-alginate gel beads was analyzed.

164 Materials and methods Bacterial strain A floc-forming bacterium used in this work was C. luteola. This strain was previously described by Ozdemir & Baysal (2004) and deposited in the Microbial Culture Collection of the Basic and Industrial Microbiology Section, in Department of Biology, Ege University, Turkey (Izmir), with the code TEM05. Isolation of crude exocellular polysaccharide (EPS) The strain was cultivated aerobically in 500 ml conical flasks containing sterile nutrient broth (Difco) on a rotary shaker (100 rev min)1) at 30
C. Cells were harvested at the end of exponential phase, i.e. after 48 h incubation. After cultivation, the culture of the most EPS producing bacterium was centrifuged at 10,000 g for 20 min at room temperature (25
C) and supernatant liquid were then decanted in to three volumes of propan-2-ol, shaken vigorously and held at 4
C for 4 h. Precipitated polysaccharide was freezedried to obtain a crude EPS preparation (Tago & Aida 1977; Ozdemir et al. 2003) Sugar and protein contents of EPS The sugar content was determined by phenol-H2SO4 method (Dubois et al. 1956); the protein content was determined by the Bradford method (Bolling & Edelstein 1991). Alginate and alginate-EPS beads preparation and their characterization Calcium alginate and calcium alginate-EPS beads were prepared by dropping 2% (w/v) aqueous solutions of sodium alginate by a peristaltic pump into 5% (w/v) CaCl2 solution under magnetic stirring at 4–7
C. The beads were stirred in this solution for 2 h. Successively; they were then collected by filtration, washed three times with distilled water and stored in a 2% (w/v) CaCl2 solution at 4
C. The second procedure is the same as described above except that 2% sodium alginate was replaced by mixture solution of 1.5% sodium alginate and 0.5% crude EPS. Preparation of the microorganisms for biosorption In this study, C. luteola TEM05 was cultivated aerobically in 500 ml conical flasks containing sterile nutrient broth (Difco) on a rotary shaker (100 revolutions per min) at 30
C. Cells were harvested at the end of exponential phase, i.e. after 48 h incubation. After cultivation, the cells were centrifuged at 10,000 · g for 20 min for inactivation of the cells, the cultures were autoclaved (121
C, 15 min) before harvested by centrifugation (10,000 g for 20 min at room temp.) and finally

Guven Ozdemir et al. freeze-dried. 0.67% (w/v) of freeze-dried cells was resuspended in 2% Na-alginate or Na-alginate-EPS mixtures. Na-alginate-biomass or Na-alginate-EPS-biomass slurries were then extruded into 5% (w/v) CaCl2 for polymerization and bead formation. Biosorption studies The biosorption of the metal ions on the alginate, alginate-EPS beads and on the immobilized C. luteola TEM05 from aqueous solutions was investigated in batch biosorption-equilibrium experiments. The biosorption time and the initial concentrations of heavy metal ions on the biosorption rate and capacity were studied. The effect of the initial metal ions concentration on the biosorption was studied at pH 6.0 as described above except that the concentration of each metal ion in the adsorption medium was varied between 0.393 and 4.721 mM for Cu(II); 0.426 and 5.112 mM for Ni(II). Competition in metal ion uptake In order to ascertain whether there was any competition between the different metal ions for uptake by a particular biomass, multiple metal ion solutions were prepared. The experiment was experienced the almost same molar ratio (1.574 and 1.704 mM for Cu and Ni, respectively) of two competitive metal ions; so that where two metals were present the total metal concentration was 3.278 mM. The abilities of the biomasses to adsorb one metal were compared with the adsorption of that same metal where one was present in the same solution. Analysis of metal ions The concentration of unadsorbed Ni(II) and Cu(II) ions in the biosorption medium were determined by using a Varian SpectrAA-220 FS atomic absorption spectrometer (in flame mode). Air-acetylene flame was employed and the working currents/wavelengths for Ni(II) and Cu(II) ions were 4 mA/351.5 nm and 4 mA/249.2 nm respectively. Deuterium background correction was used. Data treatment During the biosorption, a rapid equilibrium is established between adsorbed metal ions on biosorbent (qeq) and unadsorbed metal ions in solution (Ceq). This equilibrium can be represented by the Langmuir adsorption isotherm, which is widely used to analyze data for water and wastewater treatment applications. The Langmuir equation which is valid for monolayer sorption onto a surface a finite number of identical sites and is given by Equation (1) qeq ¼

Q0 bCeq 1 þ bCeq

ð1Þ

Cu(II) and Ni(II) adsorption by Chryseomonas luteola TEM05

Results Selection of microorganism The most EPS producing bacterium selected for this study was a gram-negative bacterium, which was further identified as Chryseomonas luteola TEM05, which produces a non-diffusible yellow pigment. Characteristics of extracellular polysaccharide and properties of the beads Total sugar and total protein analysis indicated that was composed of 33% total sugar and 26% EPS total protein. Alginate is a natural polymer and may be converted into hydrogels via crosslinking with divalent calcium ions. It was preferred over other materials because of advantages including biodegradability, hydrophilicity, presence of carboxylic groups, and natural origin. One of the most important disadvantages of cell immobilization is the increase in mass transfer resistance due to the polymeric matrix. From this point of view, alginate beads have advantages when compared with support materials such as polyvinyl alcohol and 2-hydroxylethylmethacrylate, because the presence of carboxylic groups in the alginate structure enhances heavy metal ions adsorption. All beads were spherical shaped with approx. 2.8– 3 mm. Effect of pH on the biosorption capacity PH of the biosorption studies was selected as 6.0 because the maximum biosorption of Cu(II) and Ni(II) on the immobilized biomass was observed at around pH 6.0. (Al-Saraj et al. 1999; Blanco et al. 1999; Yan & Viraraghavan 2001).

Biosorption time The effect of the time of exposure of the beads to metals on the biosorption characteristics was investigated in a batch system. The measured concentrations were plotted as a function of time, as shown in Figure 1. Biosorption of metal ions was rapid during the first 30 min for Ni(II) and the first 80 min for Cu(II) and continued at a slower rate for the following several minutes and reached a saturation value after 90 min. After this period, the concentration of adsorbed metal ions did not significantly change further with time. Statistically, there was no significant difference between preparations in case Cu(II) or Ni(II) uptake (ANOVA, P > 0.05). However, a significant difference was estimated between Cu(II) and Ni(II) uptake of these preparations ( P < 0.005). From the Figure 1, it is evident that alginate beads adsorbed Cu(II) and Ni(II) ions far less than that of EPS-alginate, alginate-biomass and alginate-EPS-biomass beads. Note that in such a biosorption process, there are several parameters that determine the biosorption rate, including structural properties biosorbent (e.g. protein and carbohydrate composition and surface charge density, topography and surface area). All these studies published in the literature have been carried out under different experimental conditions (Sag et al. 1995; Saglam et al. 1999; Say et al. 2001). Effects of initial concentration of metals ions on the biosorption capacity The biosorption capacities of all the beads are given in Figure 2 as a function of the initial concentrations of Cu(II) and Ni(II) ions within the aqueous phase, respectively. The biosorption capacities of alginate beads were near to that of the other beads. The amount of adsorbed metal increased with the initial metal concentration in the solution and reached a saturation value (3.147 mmol for copper (as an exception, alginatecopper; 2.36 mmol) and 3.408 mmol for nickel).

1.4

Biosorbed metal (mmol/g)

where Q0 is the maximum amount of the metal ion per unit weight of various preparations to form a complete monolayer on the surface bound at high Ceq (mmol g)1) and b is a constant related to the affinity of the binding sites, Q0(mmol g)1 dry weight) represents a practical limiting adsorption capacity when the surface is fully covered with metal ions and assists in the comparison of adsorption performance, particularly in cases where the sorbent did not reach its full saturation in experiments. Q0 and b can be determined from the linear plot of Ceq/ qeq vs. Ceq. A known quantity of wet Ca-alginate and Ca-alginate-EPS preparations was used in the adsorption test. After the adsorption process, the beads were dried in an oven at 50
C overnight and the dry weight of the preparations was used in the above equation. Each experiment was repeated three times and the results given are the average values.

165

1.2 1

Cu (Alg)

Cu (Alg-EPS)

Cu (Alg-cell)

Cu (Alg-EPS-Cell)

Ni (Alg)

Ni (Alg-EPS)

Ni (Alg-cell)

Ni (Alg-EPS-cell)

0.8 0.6 0.4 0.2 0 0

2

5

15

30

60

90

120

time (min)

Figure 1. Biosorption times of Cu(II) and Ni(II) ions on biosorbents from aqueous solutions at pH 6.0, at 100 rev/min, at 25
C and initial concentration of 1.57 mM of Cu(II) and 1.70 mM Ni(II).

166

Guven Ozdemir et al. 6

1.5

1

0.5

Cu (Alg)

Cu (Alg-EPS)

Cu (Alg-cell)

Cu (Alg-EPS-cell)

Ni (Alg)

Ni (Alg-EPS)

Ni (Alg-cell)

Ni (Alg-EPS-cell)

4 3 2 1 0 0

0 0

1

2

3

4

5

Ni (Alg-EPS) Ni (Alg-cell-EPS) Cu (Alg-EPS) Cu (Alg-cell-EPS)

Ni (Alg) Ni (Alg-cell) Cu (Alg) Cu (Alg- cell)

5

Ceq/qeq (mmol)

Biosorbed metal (mmol/g)

2

1

2

5

4

3

Ceq (mmol)

6

Biosorbent concentration (mmol)

Figure 2. Biosorption was carried out at pH 6.0, at various concentrations of Cu(II) (0.39–4.72 mM) and Ni(II) (0.43–5.11 mM), at 100 rev/min and at 25
for 120 min.

Figure 3. The linearized Langmuir adsorption isotherms of Cu(II) and Ni(II) by various preparations.

Table 1. Isotherm model constants for adsorption of copper and nickel on C. luteola TEM05.

Langmuir adsorption isotherms Figure 3 shows the Langmuir plots for Cu(II) and Ni(II) biosorption by beads of Ca-alginate, Ca-alginate-EPS, Ca-alginate-cell and Ca-alginate-EPS-cell. The Langmuir constants (Q0 and b) along with correlation coefficients (R2) have been the plots (Figure 3) for biosorption of Cu(II) and Ni(II) on the biosorbents and the results are presented in Table 1.

Metal ion Biosorbent type

Langmuir adsorption isotherm

Cu(II) (Alg) Cu(II) (Alg-EPS) Cu(II) (Alg-cell) Cu(II) (Alg-EPS-cell) Ni(II) (Alg) Ni(II) (Alg-EPS) Ni(II) (Alg-cell) Ni(II) (Alg-EPS-cell)

Q0

b

R2

1.505 1.989 1.976 1.937 0.996 1.224 1.078 1.219

5.926 2.001 3.261 4.180 1.470 0.895 2.295 1.309

0.993 0.977 0.989 0.991 0.993 0.972 0.997 0.980

Competitive biosorption Competitive biosorption of Cu(II) and Ni(II) ions were also studied. The medium containing 1.573 mmol Cu(II) and 1.704 mmol of Ni(II) was incubated with the biomass in batch fashion. The Competitive biosorption capacities were 0.551, 0.618, 0.605, 0.628 for Cu(II); 0.356, 0.369, 0.374, 0.389 mmol for Ni(II) per g alginate, alginate-EPS, alginate-biomass and alginate-EPS-biomass, respectively. As seen in Figure 4, the Competitive

0.8

Biosorbed metal (mmol/g)

It probably means that the cross-linking of potential metal binding-sites among microbial biomass, EPS and alginate gel occurred. Bioremediation studies have been conducted using microbial polysaccharides and inactivated microbial cell systems from contaminated waters by several researchers (Fourest & Volesky 1997). In the systems of natural polysaccharides used for removal of heavy metal ions from industrial wastewater, the metal removal process is based on solid-liquid contacting and separation process. Such preparations offer advantages in terms of mechanical strength and durability, handling and ease of scale up. On the other hand, all these adsorbents are expensive and required several preparation steps. The metal removal were 1.427, 1.801, 1.777, 1.787 mmol for 4.721 mmol copper; 0.838, 0.941, 0.961, 0.982 mmol for 5.112 mmol of nickel per g alginate, alginate-EPS, alginate-biomass and alginateEPS-biomass, respectively. The magnitude of changes in metal ion binding capacity of C. luteola TEM05 may due to the properties of the metal sorbates (e.g. ionic size, atomic weight, or reduction potential of the metal) and the properties of the bacterium (e.g. structure, functional groups, and surface area). Capsules and slime layers of bacteria contain polysaccharides as basic building blocks, which have ion exchange properties, and also proteins and lipids and therefore offer a host of functional groups capable of binding to heavy meals. These functional groups such as amino, carboxylic, sulphydryl and phosphate groups differ in their affinity and specificity for metal binding (Nourbakhsh et al. 1994).

Cu (Alg-EPS) Cu (Alg-EPS-cell) Ni (Alg-EPS) Ni (Alg-EPS-cell)

Cu (Alg) Cu (Alg-cell) Ni (Alg) Ni (Alg-cell)

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0

2

5

15

30

60

90

120

time (min)

Figure 4. Competitive biosorption times of Cu(II) and Ni(II) ions on biosorbents from aqueous solutions at pH 6.0; at 100 rev/min, at 25
C and at initial concentration of 1.57 mM of Cu(II) and 1.70 mM Ni(II).

Cu(II) and Ni(II) adsorption by Chryseomonas luteola TEM05 biosorption capacities of the beads for all metal ions were lower than non-competitive conditions. Observation of the figure confirms that copper ions are the most effectively sequestered by the four biomass compared with nickel ions. The order of affinity for competitive conditions was as follows (based on a lmol accumulation): Cu(II) > Ni(II). This order is the same as in the non-competitive condition.

Conclusions The aim of this work was to find the biosorption characteristics of selected a new biomaterial against to heavy metals for the removal of copper and nickel ions. Experiments were performed as a function of pH, initial metal ion concentration and time. The obtained results showed that the Ca-alginate-EPS beads improved performance in the batch system for the treatment of wastewater containing copper and nickel ions. The equilibrium was well described by the Langmuir adsortion isotherm for the mathematical description of the biosorption of copper and nickel ions to EPS and the isotherms constants were evaluated to compare the biosorptive capacity of EPS for metal ions. Consequently, the use of biomaterial biosorption then may provide an attractive alternative to use of conventional ion exchange resins. However, biomaterial biosorption technologies are still being developed and much more work is required.

References Arca, M.Y., Arpa, C., Ergene, A., Bayramoglu, G. & Genc, O. 2003 Ca-alginate as a support for Pb(II) and Zn(II) biosorption with immobilized Phanerochaete chrysosporium. Carbohydrate Polymers 52, 167–174. Blanco, A., Sanz, B., Llama, M.J. & Serra, J.L. 1999 Biosorption of heavy metals to immobilised Phormidium laminosum biomass. Journal of Biotechnology 69, 227–240. Bolling, D.W. & Edelstein, S.J. 1991 Protein Methods. New York, pp. 50–55. Christ, R.H., Martin, J.R., Carr, D., Watson, J.R. & Clarke, H.J. 1994 Interaction of metals and protons with algae. 4. Ion-exchange vs adsorption models and a reassessment of scatchard plots – ion – exchange rates and equilibria compared with calcium alginate. Environmental Science and Technology 28, 1859–1866. Dubois, M., Gilles, A.K., Hamilton, J.K., Rebers, P.A. & Smith, F. 1956 Colorometric method for determination of sugars and related substances. Analytical Chemistry 28, 350–356.

167

Ferguson, C.R., Peterson, M.R. & Jeffers T.H. 1989 Removal of metal contaminants from waste waters using biomass immobilized in polysulfone beads. In Biotechnology in Minerals and Metal Processing, eds. Scheiner, B.J., Doyle, F.M., Kawatras, S.K. pp. 193–199. Littleton, CO: Society of Mining Engineers. Fourest, E. & Volesky, B. 1997 Alginate properties and heavy metal biosorption by marine algae. Applied Biochemistry and Biotechnology 67, 215–226. Gutnick, D.L. & Bach, H. 2000 Engineering bacterial biopolymers for the biosorption of heavy metals; new products and novel formulations. Applied Microbiology and Biotechnology 54, 451–460. Hugerth, A., Caram-Lelham, N. & Sundelof, L.O. 1997 The effect of charge density and conformation on the polyelectrolyte complex formation between carageenan and chitosan. Carbohydrate Polymers 34, 149–156. Jianlong, W., Horan, N., Stentiford, E. & Yi, Q. 2000 The radial distribution and bioactivity of Pseudomonas sp immobilized in calcium alginate gel beads. Process Biochemistry 35, 465–469. Kratochvil, D. & Volesky, B. 1998 Advances in biosorption of heavy metals. Trends In Biotechnology 16, 291–300. Lazaro, N., Sevilla, A.L., Morales, S. & Marques, A.M. 2003 Heavy metal biosorption by gellan gum gel beads. Water Research 37, 2118–2126. Nourbakhsh, M., Sag, Y., Ozer, D., Aksu, Z., Kutsal, T. & Caglar, A. 1994 A Comparative study of various biosorbents for removal of chromium(VI) ions from industrial waste waters, Process Biochemistry 29, 1–5. Ozdemir, G. & Baysal, S.H. 2004 Chromium and aluminum biosorption on Chryseomonas luteola TEM05. Applied Microbiology and Biotechnology 64, 599–603. Ozdemir, G., Ozturk, T., Ceyhan, N., Isler, R. & Cosar, T. 2003 Heavy metal biosorption by biomass of Ochrobactrum anthropi producing exopolysaccharide in activated sludge. Bioresource Technology 90, 71–74. Sag, Y., Nourbakhsh, M., Aksu, Z. & Kutsal, T. 1995 Comparasion of Ca-alginate and immobilized Z. ramigera as sorbents for copper(II) removal. Process Biochemistry 30, 175–181. Saglam, N., Say, R., Denizli, A., Patir, S. & Arica, M.Y. 1999 Biosorption of inorganic mercury and alkylmercury species on to by Phanerochaete chrysosporium mycellium. Process Biochemistry 34, 725–730. Say, R., Denizli, A. & Arica, M.Y. 2001 Biosorption of cadmium(II), lead(II) and Copper(II) with the filamentous fungus P. Chrysosporium. Bioresource Technology 76, 67–70. Tago, Y. & Aida, K. 1977 Exocellular mucopolysaccharide closely related to bacteria floc formation. Applied and Environmental Microbiolology 34, 308–314. Tsezos, M., McCready, R.G.L. & Bell, J.P. 1989 The continuous recovery of uranium from biologically leached solutions using immobilized biomass. Biotechnology and Bioengineering 34, 10–17. Veglio, F., Esposito, A. & Reverberi, A.P. 2002 Copper adsorption on calcium alginate beads: equilibrium pH-related models. Hydrometallurgy 65, 43–57. Volesky, B. 2001 Detoxification of metal-bearing effluents; biosorption for the next century. Hydrometallurgy 59, 203–216. Yan, G. & Viraraghavan, T. 2001 Heavy metal removal in a biosorption column by immobilized M. rouxii biomass. Bioresource Technology 78, 243–249.

World Journal of Microbiology & Biotechnology 2005 21: 169–172 DOI: 10.1007/s11274-004-2724-0

 Springer 2005

Production of extracellular alkaline proteases by Aspergillus clavatus Ce´lia R. Tremacoldi and Eleonora Cano Carmona* Department of Biochemistry and Microbiology, Universidade Estadual Paulista, Av: 24 A, no. 1515, Bela Vista, Rio Claro, Sa˜o Paulo 13506-900, Brazil *Author for correspondence: Tel.: +55-19-3526-4179, Fax: +55-3526-4176, E-mail: [email protected] Received 22 April 2004

Keywords: Alkaline proteases, Aspergillus clavatus, optimization, protease characterization, protease production

Summary The production of extracellular alkaline proteases from Aspergillus clavatus was evaluated in a culture filtrate medium, with different carbon and nitrogen sources. The fungus was cultivated at three different temperatures during 10 days. The proteolytic activity was determined on casein pH 9.5 at 37
C. The highest alkaline proteolytic activity (38 U/ml) was verified for culture medium containing glucose and casein at 1% (w/v) as substrates, obtained from cultures developed at 25
C for 6 days. Cultures developed in Vogel medium with glucose at 2% (w/v) and 0.2% (w/v) NH4NO3 showed higher proteolytic activity (27 U/ml) when compared to the cultures with 1% of the same sugar. Optimum temperature was 40
C and the half-lives at 40, 45 and 50
C were 90, 25 and 18 min, respectively. Optimum pH of enzymatic activity was 9.5 and the enzyme was stable from pH 6.0 to 12.0.

Introduction Extracellular proteases have high commercial value and multiple applications in various industrial sectors, such as the detergent, food, pharmaceutical, leather and diagnostics industries, waste management and silver recovery (Godfrey & West 1996). Currently, the largest share of the enzyme market has been held by detergent proteases active and stable at alkaline pH (Rao et al. 1998). Microorganisms account for a two-thirds share of commercial protease production worldwide and the alkaline serine proteases (EC.3.4.21.) are the most important group of commercial enzymes (Kumar & Takagi 1999). Alkaline proteases from Aspergillus species are used in leather treatment (Godfrey & West 1996), endo- and exoproteases from A. oryzae have been used to modify wheat gluten, an insoluble protein, by limited proteolysis thereby facilitating its handling and machining, and these proteases are also important in the pharmaceutical industry (Chiplonkar et al. 1985). Genes encoding alkaline proteases from A. oryzae, A. sojae, A. fumigatus, A. flavus and A. nidulans have been cloned, sequenced and expressed (Hodgson 1994). An A. clavatus strain was isolated from Brazilian soil as the best protease producer (Attili 1994). Recently, we reported the acid protease production by this strain and some of its properties (Tremacoldi et al. in press). The aim of the present work was to study the production of extracellular alkaline proteases from this isolate, grown on medium with different nitrogen and carbon sources at different temperatures. The influence of pH and tem-

perature on the alkaline proteolytic activity was determined, since several biotechnological applications of proteases require thermal resistance and high stability at alkaline pH.

Material and methods Microorganism Aspergillus clavatus (CCT2759 – Fundac¸a˜o Tropical de Pesquisas e Tecnologia Andre´ Tosello Collection, Campinas, Brazil) was isolated from soil of the Atlantic forest, Peruı´ be city, SP, Brazil. This strain was selected from among 879 fungal strains because it showed the highest extracellular proteolytic activity in preliminary tests (Attili 1994). In the lab, it was maintained in Vogel solid medium (Vogel 1956) at 25
C, in the absence of light. Protease production A suspension (1 ml) containing 107 spores/ml from 7-day-old colonies was inoculated into Erlenmeyer flasks containing 50 ml of Vogel medium. The cultures were developed in a shaker at 120 rev/min. The tested carbon sources were sucrose or glucose at 1% (w/v) at 20, 25 and 30
C. These sugars were also tested at 2% (w/v) at 25
C, because this is the optimum temperature for growing this strain. The nitrogen sources were 0.2% (w/v) NH4NO3 and NaNO3, or casein and gelatin at 1% (w/v) at 20, 25 and 30
C. All cultures were developed at

170

The alkaline proteolytic activity was determined in triplicate through the incubation of 250 ll of culture filtrate with 1 ml of casein (Sigma) at 2% (w/v) in 0.1 M glycine–NaOH buffer pH 9.5, for 90 min at 37
C. The reaction was stopped by the addition of 2 ml of 10% (w/v) trichloroacetic acid, followed by centrifugation at 6000 · g for 20 min at 4
C. The proteolytic activity was determined by the absorbance reading of the supernatant at 280 nm, against an appropriate blank. One unit of enzyme activity (U) was defined as the amount catalysing an increase of 0.1 in the absorbance at 280 nm, per ml of sample during 1 h incubation (Van Jaarsveld et al. 1997; Sundd et al. 1998; Kundu et al. 2000). The protein concentration of each sample was estimated by the Sedmack & Grossberg (1977) method, using bovine serum albumin as standard. Enzyme precipitation The protein from the crude culture filtrates from A. clavatus grown under the best protease production conditions was precipitated with ammonium sulphate from 40% to 75% of saturation (Scopes 1994). After centrifugation at 6000 · g for 20 min at 4
C, the precipitate was resuspended in an appropriate volume of 50 mM ammonium acetate buffer, pH 6.8. The sample was dialysed overnight (membrane of 12 up to 16 kDa), against 50 volumes of the same buffer and employed in experiments for the characterization of the alkaline proteolytic complex. Effect of temperature and pH on the alkaline proteolytic activity and stability The optimum pH was determined using casein at 2% (w/v) as substrate in McIlvaine buffer (pH 3.0–8.0), and in 0.1 M glycine–NaOH buffer (pH 9.5–12.0) at 37
C. The pH stability was determined by incubation of the enzyme for 24 h, in those buffers, at 4
C, before the enzymatic activity assay at 37
C. Optimum temperature was determined by activity assay on casein at pH 9.5 from 20 to 50
C. Thermal stability was investigated by preincubating the enzyme without substrate at 40, 45, 50 and 55
C from 0 to 180 min, before determination of proteolytic activity on casein at 2% (w/v) at pH 9.5.

protease activity (U/ml)

Enzyme assay

levels depending on the carbon and nitrogen sources and temperatures, during the culture incubation period, were lower than those observed with the acid protease from this same strain (Tremacoldi et al. in press). Cultures grown on the medium containing NH4NO3 and sucrose at 1% showed the highest proteolytic activity (4.5 U/ml) at day 8 of incubation at 25
C (Figure 1A) and those developed in NaNO3 and glucose 1% showed 11 U/ml at day 7 also at 25
C (Figure 1B). The highest value verified for alkaline protease in a shorter incubation period in the medium containing glucose is due to the fact that such a monosaccharide is readily available for the metabolism of the fungus (Ashour et al. 1996). This conclusion is supported by the results from Figure 3B that shows a similar profile with the other nitrogen source, thereby excluding the influence of ammonium. The lower protease production in medium containing sucrose could also have been limited by its hydrolysis, since the invertase activity was not evaluated. An alkaline proteolytic activity of 38 U/ml was observed for 6- and 7-day-old cultures grown on glucose and casein at 1% at 25
C (Figure 2A) and it was also the highest value when a comparison among all cultures was performed. The alkaline protease activity verified in 8-day cultures grown on glucose and gelatin at 1% (Figure 2B) showed a maximum of 24 U/ml at 25
C. In casein medium the maximum could already be observed at day 6, while in gelatin medium it occurred only at day 8. Gelatin is a poorer nutritional source than casein and even though fungi are able to synthesize the amino acids that they need, when available in the culture medium,

15 A 10

5

0

protease activity (U/ml)

initial pH of 7.0. The sampling flasks were taken out daily, in triplicate, over 10 days and the culture filtrates were used as enzymatic source. The results of the alkaline proteolytic activity were expressed as the mean of the three experiments.

Ce´lia R. Tremacoldi and Eleonora Cano Carmona

0

1

2

3

2

3

4 5 6 7 8 culture time (days)

9

10

11

4

9

10

11

15 B 10 5

0 0

1

5

6

7

8

culture time (days)

Results and discussion The alkaline proteolytic activity was detected in all culture filtrates. However, the variations in the activity

Figure 1. Alkaline proteolytic activity by Aspergillus clavatus cultivated in liquid Vogel medium containing 0.2% NH4NO3 and 1% sucrose (A) or 0.2% NaNO3 and 1% glucose (B), at 20 ( ), 25 (h) and 30 (n)
C. Bars represent mean ± SE. *SE ¼ standard error.




A

0

protease activity (U/ml)

protease activity (U/ml)

40 35 30 25 20 15 10 5 0

171

40 35 30 25 20 15 10 5 0

1

2

3

4 5 6 7 8 culture time (days)

9

10

1

2

3

4

5

6

7

8

25

A

20 15 10 5 0 0

1

2

3

4

5

6

7

8

9

10

11

8

9

10

11

culture time (days)

B

0

30

11

protease activity (U/ml)

protease activity (U/ml)

Aspergillus alkaline protease

9

10

30 25

B

20 15 10 5 0 0

11

1

2

3

culture time (days)

the energy needed for the metabolism is lower. The necessity of synthesizing the amino acids required for protein biosynthesis might explain the low values of enzymatic activity observed with the employment of gelatin. The highest production of proteolytic enzymes in medium containing casein as nitrogen source was also observed for A. niger (Chakraborty et al. 1995) and A. oryzae (Battaglino et al. 1991). For testing the effect of the carbon source concentration on the proteolytic activity, glucose or sucrose at 2% was also added to the Vogel basal medium with NH4NO3 as nitrogen source at 25
C. The alkaline protease production was similar for cultures developed in the medium with sucrose at 1% or 2%, with maxima ranging from 3.5 to 4.5 U/ml at day 8 (Figure 3A). The activity curves for sucrose as well as glucose at 1% and 2% (Figure 3B) showed almost the same shape until day 7 of incubation. In medium with glucose at 2% the enzymatic activity was increased, reaching 27 U/ml at day 8. Repression of synthesis and secretion of proteases by glucose are described in many species of proteaseproducer microorganisms (Cohen 1973; Klapper et al. 1973; Tsuchiya & Kimura 1984; Taragano et al. 1997). The use of higher glucose concentration increased the alkaline proteolytic activity from A. clavatus, however glucose concentration should have significantly decreased before the protease production. The precipitated and dialysed sample of culture filtrates from A. clavatus grown on Vogel medium containing 1% glucose and 1% casein, at day 6 of incubation at 25
C, was employed for the characterization of the proteolytic complex regarding the optimum pH and temperature. The alkaline proteolytic

6

7

activity showed an optimum activity temperature of 40
C (Figure 4A), and half-lives of 90, 25 and 18 min at 40, 45 and 50
C, respectively (Figure 4B). Several alkaline proteases from Aspergillus species have an optimum temperature at 40
C (Ohara & Nasuno 1972; Kundu & Manna 1975; Larcher et al. 1992). The thermal stability of a microbial protease may be increased with the addition of Ca2+ ions to the enzymatic extract or through mutagenesis techniques.

relative activity (%)




5

Figure 3. Alkaline proteolytic activity by Aspergillus clavatus cultivated in liquid Vogel medium containing 0.2% NH4NO3 and 1% (j) or 2% (() of sucrose (A) and 1% (j) or 2% (() of glucose (B) at 25
C. Bars represent mean ± SE.

125 A

100 75 50 25 0 10

20

30

40

50

60

temperature (oC) relative activity (%)

Figure 2. Alkaline proteolytic activity by Aspergillus clavatus cultivated in liquid Vogel medium containing 1% glucose and 1% casein (A) or 1% glucose and 1% gelatin (B), at 20 ( ), 25 (h) and 30 (n)
C. Bars represent mean ± SE.

4

culture time (days)

125 B

100 75 50 25 0 0

60

120 time (min)

180

240

Figure 4. Effect of temperature on activity (A) and stability (B) of alkaline proteases from Aspergillus clavatus. B: at 40
C (r), at 45
C (j) and at 50
C (m).

172 relative activity (%)

Ce´lia R. Tremacoldi and Eleonora Cano Carmona 125 100 75 50 25 0 2

4

6

8 pH

10

12

14

Figure 5. Effect of pH on activity (d) and stability (O) of alkaline proteases from Aspergillus clavatus. McIlvaine (pH 3.0–8.0) and 0.1 M glycine–NaOH (pH 9.5–12.0) buffers.

Alkaline proteases from many microorganisms have been engineered to make them more stable, to suit their utilization to their respective industrial application (Gupta et al. 2002). The highest protease activity at 37
C was verified for pH 9.5 and it was also stable over a wide range of pH from 6.0 to 12.0 (Figure 5). The highest alkaline protease activity from A. flavus was observed at pH 7.5 and was stable from pH 8.0 to 11.0 (Mellon & Cotty 1996), whereas the protease from A. tamarii was active and stable from pH 5.0 to 9.5 (Boer & Peralta 2000). Alkaline protease from Fusarium culmorum showed optimum pH from 8.3 to 9.6 (Pekkarinen et al. 2002). Our results indicate possible employment of such enzymatic complex in processes requiring stability in alkaline pH. This is the first report on the production of alkaline protease from A. clavatus aiming at its optimization, and the purification of the main component of this complex and its biochemical characterization are in development in our laboratory.

Acknowledgement This work was supported by Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo (FAPESP).

References Ashour, S.A., El Shora, H.M., Metwally, M. & Habib, S.A. 1996 Fungal fermentation of whey incorporated with certain supplements for the production of proteases. Microbios 86, 59–69. Attili, D.S. 1994 Isolation, Identification and Ecology of Cellulolytic Fungi from Jureia Ecological Station Soil – Itatins, SP. PhD Thesis, Rio Claro, UNESP – Bioscience Institute, Rio Claro, Brazil. Battaglino, R.A., Huergo, M., Pilosof, A.M.R. & Bartholomai, G.B. 1991 Culture requirements for the production of protease by Aspergillus oryzae in solid-state fermentation. Applied Microbiology and Biotechnology 35, 292–296. Boer, C.G. & Peralta, R.M. 2000 Production of extracellular protease by Aspergillus tamarii. Journal of Basic Microbiology 40, 75–81. Chakraborty, R., Srinivasan, M., Sarkar, S.K. & Raghavan, K.V. 1995 Production of acid protease by a new Aspergillus niger by solid substrate fermentation. Journal of Microbial Biotechnology 10, 17– 30.

Chiplonkar, J.M., Gangodkar, S.V., Wagh, U.V., Ghadge, G.D., Rele, M.V. & Srinivasan, M.C. 1985 Applications of alkaline protease from Conidiobolus in animal cell culture. Biotechnology Letters 7, 665–668. Cohen, B.L. 1973 Regulation of intracellular and extracellular neutral and alkaline proteases in Aspergillus nidulans. Journal of General Microbiology 79, 311–320. Godfrey, T. & West, S. 1996 Industrial Enzymology, 2nd edn. p. 3. NY: Macmillan Publishers, Inc. ISBN 0-33359464-9. Gupta, R., Beg, Q.K. & Lorenz, P. 2002 Bacterial alkaline proteases: molecular approaches and industrial applications. Applied Microbiology and Biotechnology 59, 15–32. Hodgson, J. 1994 The changing bulk catalysis market: recombinant DNA techniques have changed bulk enzyme production dramatically. Biotechnology 12, 789–790. Klapper, B.F., Jameson, D.M. & Mayer, R.M. 1973 Factors affecting the synthesis and release of the extracellular protease of Aspergillus oryzae NRRL 2160. Biochimica et Biophysica Acta 304, 513–519. Kumar, C.G. & Takagi, H. 1999 Microbial alkaline proteases: from a bioindustrial viewpoint. Biotechnology Advances 17, 561–594. Kundu, A.K. & Manna, S. 1975 Purification and characterization of extracellular proteinases of Aspergillus oryzae. Applied Microbiology 30, 507–513. Kundu, S., Sundd, M. & Jagannadham, M.V. 2000 Purification and characterization of a stable cysteine protease ervatamin B, with two disulfide bridges, from the latex of Ervatamia coronaria. Journal of Agricultural and Food Chemistry 48, 171–179. Larcher, G., Bouchara, J.-P., Annaix, V., Symoens, F., Chabasse, D. & Tronchin, G. 1992 Purification and characterization of a fibrinolytic serine proteinase from Aspergillus fumigatus culture filtrate. FEBS Letters 308, 65–69. Mellon, J.E. & Cotty, P.J. 1996 Purification and partial characterization of an elastinolytic proteinase from Aspergillus flavus culture filtrates. Applied Microbiology and Biotechnology 46, 138– 142. Ohara, T. & Nasuno, S. 1972 Enzymatic properties of alkaline proteinase from Aspergillus candidus. Agricultural and Biological Chemistry 36, 1797–1802. Pekkarinen, A.I., Jones, B.L. & Niku-Paavola, M.L. 2002 Purification and properties of an alkaline proteinase of Fusarium culmorum. European Journal of Biochemistry 269, 798–807. Rao, M.B., Tanskale, A.M., Ghatge, M.S. & Deshpande, V.V. 1998 Molecular and biotechnological aspects of microbial proteases. Microbiology and Molecular Biology Reviews 62, 597–635. Scopes, R. 1994 Protein Purification: Principles and Practice, 380p, 3rd edn. NY: Springer-Verlag. ISBN 0-38794072-3. Sedmack, J.J. & Grossberg, S.E. 1977 A rapid, sensitive and versatile assay for protein using Coomassie Brilliant Blue G250. Analytical Biochemistry 79, 544–552. Sundd, M., Kundu, S., Pal, G.P. & Medicherla, J.V. 1998 Purification and characterization of a highly stable cysteine protease from the latex of Ervatamia coronaria. Bioscience Biotechnology and Biochemistry 62, 1947–1955. Taragano, V., Sanchez, V.E. & Pilosof, A.M.R. 1997 Combined effect of water activity depression and glucose addition on pectinases and protease production by Aspergillus niger. Biotechnology Letters 19, 233–236. Tremacoldi, C.R., Watanabe, N.K. & Carmona, E.C. in press Production of extracellular acid proteases by Aspergillus clavatus. World Journal of Microbiology and Biotechnology. Tsuchiya, K. & Kimura, T. 1984 Decrease of protease activity by addition of glucose to the culture of Cephalosporium sp. Journal of Fermentation Technology 62, 35–39. Van Jaarsveld, F.P., Naude, R.J. & Oelofsen, W. 1997 Optimization of calcium-dependent protease and cathepsin D assays in ostrich muscle and the effect of chemical and physical dry-curing parameters. Meat Science 47, 287–299. Vogel, H.J. 1956 A convenient growth medium for Neurospora (Medium N). Microbial Genetics Bulletin 13, 42–43.

World Journal of Microbiology & Biotechnology 2005 21: 173–178 DOI: 10.1007/s11274-004-2217-1


Springer 2005

Barley-based medium for the cost-effective production of Bacillus thuringiensis P.S. Vimala Devi*, T. Ravinder and C. Jaidev Directorate of Oilseeds Research, Rajendranagar, Hyderabad 500 030, Andhra Pradesh, India *Author for correspondence: Tel.: +91-40-2401-5345/6141, Fax: +91-40-2401-7969, E-mail: psvimaladevi@ rediffmail.com Received 27 April 2004

Keywords: Bacillus thuringiensis, barley, castor semilooper, cost-effective production

Summary Studies conducted for the multiplication of Bacillus thuringiensis (Bt) using barley Hordeum vulgare as the carbon source led to the development of a protocol for the cost-effective, mass production of Bt. The production employs the simple shake flask method and can be easily adopted with a production potential of 1.5 kg Bt per day approximately at an overall production cost of Rs. 360 kg)1 (8 US dollars). The protocol is suitable for promoting localized production of Bt at the village/district level. The product when tested as 0.1% (w/v) spray against the castor semilooper, Achoea janata proved highly effective, causing immediate feeding cessation of the larvae followed by 85% and 100% mortality by 48 and 72 h after treatment, respectively. Introduction The insect pathogenic bacterium Bacillus thuringiensis (Bt) is the only microbe which has been successfully exploited commercially for the management of insect pests. Bt accounts for 95% of the world market of microbial pest control agents due to the twin advantages of safety to natural enemies, honey bees, etc. and its rapid action against target insect pests (Vimala Devi et al. 2001). It is highly effective against several lepidopteran pests of economic importance. Castor, pigeonpea and peanut are the crops cultivated in kharif season predominantly in Mahaboobnagar and Nalgonda districts of Andhra Pradesh in India. Castor cultivation in these dryland areas is totally rainfed and constrained by the vulnerability of the improved cultivars to several insect pests of which the castor semilooper Achaea janata, Helicoverpa armigera and Spodoptera litura are economically important. The potential of Bt in the management of A. janata and H. armigera is amply proven with the commercial formulations. However, large scale exploitation of Bt in pest management has not gained momentum, due to the prohibitive cost and restricted availability of the commercial products, particularly so for crops of lowincome returns. The Andhra Pradesh–Netherlands Biotechnology Programme under implementation in Andhra Pradesh has the primary objective of bringing potential eco-friendly pest management technologies to the doorstep of poor dryland farmers in Mahaboobnagar and Nalgonda districts as well as generating employment in these areas. In tune with these objectives, a study was carried out to develop a cost-effective

multiplication protocol for Bt using locally available materials. Several studies have been reported on the standardization of media composition for increasing the endotoxin production in Bt (Salama et al. 1983; Mummigatti & Raghunathan 1990; Morris et al. 1997; Vora & Shethna 1999). However these procedures are still not cost-effective and will not enable promotion of Bt production as a cottage industry in the developing countries. The commonly used sources of sugar are corn steep liquor and cane sugar molasses while several leguminous seeds, pulses, agro-industrial byproducts have been demonstrated as good sources of protein. However in India several restrictions exist with regard to sale and transport of molasses which makes it difficult for its commercial exploitation in Bt production by small entrepreneurs. Use of raw materials in fermentation media is therefore greatly influenced by the availability, cost and local conditions. Barley, Hordeum vulgare contains on an average 63–65% starch, 8–13% protein, 2–3% fat, 1–1.5% soluble gums, 8–10% cellulose and 2–2.5% ash (Pomeranz 1973). Based on the nutritional value of barley as well as its easy availability, the present study was undertaken to exploit the potential of barley in the multiplication of Bt.

Material and methods Bt strains were obtained from the Bacillus Genetic Stock Center (BGSC), Ohio State University, USA. Two kurstaki strains 4D17 and 4D21 found effective against

174

P.S. Vimala Devi et al.

the castor semilooper, Achoea janata (Vimala Devi et al. 2001) were selected for the study in addition to four local isolates of Bt. Bt isolations were carried out by the method of Travers et al. 1987, from dead larvae/soil samples collected from the farmers fields in Mahaboobnagar district of Andhra Pradesh.

room temperature with slow shaking. The solution was then centrifuged and the supernatant was used for estimation of the crystal protein employing the Lowry method.

Bt multiplication

Efficacy of Bt strain 4D21 multiplied on the media was tested through bioassays against 7 day old castor semilooper larvae. Bt spray solution was prepared in 0.02% Tween-80 single distilled water and sprayed at 0.1 and 0.2% concentrations on castor leaves whose stalks were dipped in wet cotton. After the solution dried up, the larvae were released on the sprayed leaves. Observations of larval mortality were recorded every day up to 5 days after spraying

(a) Five combinations (T1–T5) of barley based media were tested for the growth and multiplication of Bt (Table 1). Five grams of powdered barley (particle size 500 lm) was taken in a Erlenmeyer conical flask of 250ml capacity. The remaining ingredients were dissolved in 50-ml distilled water and this liquid mixture was added to barley. pH of the medium was adjusted to 7.2. Flasks containing the media were sterilized at 15 psi for 20 min, cooled and inoculated with the seed culture 2% (v/v) of BGSC strain 4D21 multiplied in nutrient broth (NB) and incubated for 48 h at 30 C on a shaker at 200 rev/ min. The medium from the flasks was centrifuged, the pellet was dried in a laminar air flow, powdered and used for larval bioassays. (b) Using the combination T4, Bt multiplication was standardized in 1000-ml flasks with three different volumes of the medium per flask – 200, 300 and 400ml. The inoculation and incubation conditions were the same as mentioned above. (c) The strain 4D21 was also multiplied in 300-ml volume of NB and molasses medium (Mummigatti & Raghunathan 1990) and barley medium of T4 combination in 1000-ml conical flasks. The products were bioassayed against the castor semilooper larvae and the LC50 values were determined.

Bioassays

Field studies The Bt strain 4D21, var kurstaki multiplied in 1000 ml flasks was tested against 7–8-day-old castor semilooper larvae on castor (var Jyothi) in the research farm of the Directorate of Oilseeds Research during the rainy season (kharif ) of the year 2001. Ten third instar larvae were released per plant. Ten plants were used for each replicate and 3 replicates were maintained for each treatment. One treatment with 0.02% Tween-80 solution was used as control. Statistical analysis Data was subjected to analysis of variance using the statistical software MSTATC. Probit analysis for calculation of LC50 values was carried out using the statistical software SPSS 8.0 for windows.

Toxin estimation For extraction of crystal protein, 100 mg of the Bt powder was taken in a centrifuge tube of 50-ml capacity. To this, 10 ml of 1 M sodium chloride solution was added to remove the extracellular and cell associated metalloproteases. The solution was mixed thoroughly and centrifuged at 10,000 rev/min for 10 min at 4 C in a Heraeus Biofuge. The supernatant was discarded and the pellet was given two washes with sterile distilled water. To this pellet, 10 ml of 100 mM sodium chloride solution was added. The tubes were incubated for 2 h at

Results Bt multiplication on media Multiplication of the Bt strain 4D21 on the five media combinations (Table 1) did not show significant differences in the yield of the product. However, the spore count and toxin content varied significantly and increased only with the addition of a nitrogen source and salts. The toxin content and spore count in T1 were

Table 1. Composition of media used for Bt multiplication. Media combinations

T1 T2 T3 T4 T5

Components Barley (g)

Yeast extract (mg)

Green gram powder (mg)

CaCl2 (mg)

MgSO4 (mg)

K2HPO4 (mg)

KH2PO4 (mg)

Water (ml)

5 5 5 5 5

– – – 63 63

– 500 500 – –

– 24 12 24 12

– 60 30 60 30

50 50 50 50 50

50 50 50 50 50

50 50 50 50 50

Barley-based medium for Bacillus thuringiensis

175

12 mg g)1 and 4.2 · 1010 g)1, respectively. In the T2 and T3 combinations where green gram powder as a nitrogen source and salts (calcium chloride and magnesium sulphate) were added, the toxin content and spore count increased to 30–32 mg g)1 and 7.4–7.6 · 1010 g)1, respectively. In the T4 and T5 combinations where the nitrogen was provided in the form of yeast extract powder (Hi-Media), the toxin content and spore count further increased to 40–44 mg g)1 and 8.4–8.5 · 1010 g)1, respectively (Table 2). Bioassays Bioassays conducted with the Bt powder obtained through multiplication on the five media combinations at 0.1% concentration resulted in high larval mortality ranging from 68.33 to 71.67% by 2 days after treatment only in the combinations T4 and T5 along with complete feeding cessation while a low mortality ranging from 3.33 to 15.0% coupled with low feeding in larvae was observed in the T2 and T3 combinations. Feeding was relatively high in T1 but less when compared to the control. The combination T4 yielded 100% mortality by 3 days after treatment with combination T5 being on

par. A cumulative mortality of 83.33% was recorded in T1 by 5 days after spraying whereas 100% mortality was recorded in treatments T2 and T3 by 5 days after spraying. However, only a slight increase in mortality was observed in bioassays at 0.2% concentration with combinations T1, T2 and T3 (Table 2). In the second experiment wherein Bt multiplication was standardized using different volumes of medium in 1000-ml flasks, the yield increase was directly proportional to the increase in the volume of the medium while the toxin content as well as spore count g)1 decreased (Table 3). Bioassays conducted at 0.1% concentration resulted in 100% mortality of larvae with Bt powder from 200 ml medium within two days after spraying while Bt powder from the 300 ml medium resulted in a significant mortality of 85% at 2 days and 100% by 3 days. No feeding was observed in the first two treatments while low feeding was observed with Bt powder obtained from the third treatment which resulted in 86.67% mortality of larvae at 4 days after spraying (Table 3). Based on the above study, 2 BGSC strains 4D17, 4D21 and 4 local isolates were multiplied in 1000 ml flasks on 300 ml medium. All strains/isolates were

Table 2. Yield, toxin content, spore counts and efficacy of Bt strain 4D21 multiplied on five combinations of barley based media. Combination of Spores g)1 Toxin (mg g)1) barley medium (·1010)

Yield (g)

% of mortality 0.1% (w/v) Concentration

T1

4.2

12

2.71

T2

7.4

30

2.30

T3

7.6

32

2.07

T4

8.5

40

2.12

T5

8.4

44

2.15

S.E. mean ± C.D. (5%)

– –

– –

– –

0.2% (w/v) Concentration

2 days

3 days

4 days

5 days

2 days

3 days

4 days

5 days

3.33 (6.14) 15.00 (22.50) 15.00 (22.50) 71.67 (58.00) 68.33 (55.89) 3.423 7.052

35.00 (36.22) 46.67 (43.08) 38.33 (38.14) 100.00 (90.00) 93.33 (77.71) 3.098 6.383

65.00 (53.78) 78.33 (62.57) 73.33 (59.11) 100.00 (90.00) 100.00 (90.00) 2.132 4.391

83.33 (66.14) 100.00 (90.00) 100.00 (90.00) 100.00 (90.00) 100.00 (90.00) 1.08 2.23

6.67 (12.29) 25.00 (29.89) 16.67 (23.86) 78.33 (62.57) 68.33 (55.89) 3.40 7.013

16.67 (23.86) 55.00 (47.88) 45.00 (42.12) 100.00 (90.00) 100.00 (90.00) 1.58 3.261

66.67 (54.78) 95.00 (80.78) 85.00 (67.50) 100.00 (90.00) 100.00 (90.00) 1.084 6.098

83.33 (66.14) 100.00 (90.00) 100.00 (90.00) 100.00 (90.00) 100.00 (90.00) 1.084 2.233

Table 3. Yield, toxin content, spore counts and efficacy of Bt strain 4D21 multiplied on different volumes of T4 medium in 1000 ml flasks. Volume of T4 medium

Yield (g) % Mortality at indicated days after treatment Spores g)1 Toxin (·1010) content (mg g)1) 0.1% (w/v) Concentration 0.2% (w/v) Concentration

200 ml

8.7

42

9.49

300 ml

7.1

24

19.63

400 ml

5.2

13

29.54

– –

– –

S.E. mean ± – C.D. (5%) –

1 day

2 days

3 days

4 days

1 day

2 days

3 days

4 days

43.33 (41.15) 26.67 (31.00) 6.67 (12.29) 3.516 7.662

100.00 (90.00) 85.00 (67.50) 18.33 (25.21) 1.851 4.034

100.00 (90.00) 100.00 (90.00) 53.33 (46.92) 0.993 2.164

100.00 (90.00) 100.00 (90.00) 86.67 (68.86) 1.399 3.049

66.67 (54.78) 55.00 (47.88) 16.67 (23.86) 2.035 4.434

100.00 (90.00) 100.00 (90.00) 36.67 (37.22) 1.036 2.257

100.00 (90.00) 100.00 (90.00) 73.33 (59.00) 1.144 2.493

100.00 (90.00) 100.00 (90.00) 90.00 (75.00) 4.058 8.843

176

P.S. Vimala Devi et al.

capable of multiplying on the medium. The yield as well as the toxin content and spore counts varied with the isolate (Table 4). However, majority of the strains/ isolates excluding DOR-2 resulted in 96–100% mortality by 3 days after treatment in the laboratory bioassays (Table 4). Bt product of 4D21 strain multiplied in 300 ml medium of NB, barley medium and molasses medium in 1000-ml flasks when subjected to bioassays against 6-day-old castor semilooper larvae resulted in LC50 values of 83.07, 116.8 and 25.5 mg ml)1, respectively at 24 h after treatment (Table 5). The effective dose of Bt strain 4D21 multiplied in NB and molasses medium was 1.48 and 4.58-fold lower when compared to that on barley medium. Field study The BGSC strain 4D21 multiplied on 300 ml barley based medium in 1000 ml flask was tested in the castor field at 0.1 and 0.2% concentrations against the castor semilooper, Achoea janata. A high mortality of 61.11 and 71.11% was recorded with 0.1 and 0.2%, respectively at 2 days which was complete by 3 days. No feeding was observed from the day 1 after spraying. There was absolutely no mortality of larvae in the control plants sprayed with 0.02% Tween-80 solution (Table 6).

Discussion The study aims at development of a protocol for increasing and optimizing the production of Bt through

Table 6. Field testing of Bt strain 4D21 multiplied on 300 ml barley medium in 1000 ml flask. Treatment

Mortality at indicated days after spraying

BGSC 4D21 – 0.1% BGSC 4D21 – 0.2% Water spray control S.E. mean ± CD (0.05%)

2

3

61.11 71.11 0.0 2.31 7.51

98.89 100.00 0.0 0.64 2.10

shake cultures at the village level with minimum infrastructure requirement. Shake flask method has been employed only for bench scale multiplication of Bt to support research since the yield of Bt by multiplication on standard media like Luria-Bertani broth (LB), NB, etc. is very low and cannot be exploited for field scale multiplication. The study showed that barley could be successfully used as a carbon source. Since barley is low in protein content, an external nitrogen source was provided. The nitrogen source played a major role in the toxin production since the toxin production increased 2.5fold over the control with the addition of green gram powder while the increase was 3.33-fold with the addition of yeast extract. Within the two nitrogen sources, the difference in toxin content as well as the spore count did not vary significantly with the concentration of calcium and magnesium salts (Table 2). However, the larval mortality was significantly higher with Bt multiplied on T4 and T5 combinations where yeast extract was used as the nitrogen source when compared to the mortality obtained from the T2 and T3

Table 4. Yield, toxin content, spore counts and efficacy of different strains/isolates of Bt multiplied on 300 ml T-4 medium in 1000 ml flasks. Strain/isolate

Spores g)1 (·1010)

Toxin content g)1

Yield of Bt (g) % of mortality at indicated days after treatment 0.1% (w/v) Concentration 1 day

4D21 4D17 DOR-1 DOR-2 DOR-3 DOR-4 S.E. mean ± C.D.(5%)

8.70 6.40 5.76 5.56 6.32 6.20 – –

24 20 18 15 22 18 – –

19.63 18.24 18.36 19.25 18.43 18.92 – –

35.00 25.00 35.00 15.00 16.67 23.33 2.163 4.417

(36.22) (29.89) (36.22) (22.50) (23.86) (28.78)

0.2% (w/v) Concentration

2 days

3 days

1 day

81.67 70.00 78.33 55.00 66.67 63.33 2.596 5.300

100.00 (90.00) 100.00 (90.00) 100.00 (90.00) 83.33 (66.14) 96.67 (66.14) 96.67 (66.14) 3.324 6.788

53.33 43.33 50.00 25.00 21.67 36.67 2.042 4.169

(65.04) (56.89) (62.57) (47.88) (54.78) (52.82)

2 days (46.92) (41.15) (45.00) (29.89) (27.67) (37.18)

100.00 (90.00) 100.00 (90.00) 100.00 (90.00) 90.00 (75.00) 95.00 (80.78) 93.33 (79.43) 4.673 9.543

Table 5. Efficacy of Bt strain 4D21 multiplied on three different media multiplied on 300 ml medium in 1000 ml flasks against 7 days old castor semilooper larvae at 24 h after treatment. Medium

LC50 (mg ml)1)

Fiducial limits

Regression equation

Effective dosage in comparison to Bt multiplied on barley

Nutrient broth Molasses Barley

83.07 25.50 116.84

70.96–101.16 19.97–34.69 101.50–136.47

Y = )1.75 + 0.021X Y = )1.83 + 0.072X Y = )1.69 + 0.014X

1.48-fold lower 4.58-fold lower

Barley-based medium for Bacillus thuringiensis

177

combinations where green gram powder was used as the nitrogen source. The laboratory bioassays as well as the field study show that Bt produced from 300 ml medium in 1000-ml flasks could effectively manage the castor semilooper at 0.1% concentration. The procedure generally adopted for screening the efficacy of Bt isolates involves multiplication of Bt isolates in LB or NB and testing the Bt powder obtained against insect larvae. Mummigatti & Raghunathan (1990) have reported high toxin production through incorporation of molasses and green gram powder as the carbon and nitrogen sources, respectively in basal media. However, the yield from the above two protocols is relatively low and cannot cater to localized production using the simple equipment employed for bench scale multiplication. The yield of Bt from barley medium optimized in this study is relatively higher. The cost of Bt production based on the LC50 values of Bt from 300 ml of nutrient broth as well as the yield from molasses medium and barley medium in eighty one 1000 ml Erlenmeyer flasks which can be accommodated on a platform of 48¢¢ · 48¢¢ in one run, including the expenditure on running costs like electricity, manpower, media components, etc. has been presented in Tables 7 and 8. The cost of production (excluding labour which can vary from place to place) is highest in NB while the cost of production by molasses-based medium and barley-based medium is more or less same but much lower when compared to the production on NB. How-

ever, the product from one run is sufficient for 1031 l [@ 0.021% spray] when multiplied on molasses medium while the product from one run on barley medium is sufficient for 1539 l [@ 0.1%]. Thus multiplication on barley medium (Rs. 36/- for 100 l spray solution) is not only cost-effective but also can cover a much larger area when compared to other bench scale multiplication protocols. At the same time, the protocol has the potential for exploiting the simple shake flask method of multiplication at a cottage industry level due to the high yield of the Bt product. Placing of two shakers in a room with one batch inoculation (one shaker) per day can further lower the cost of production (Rs. 360 kg)1) since the expenditure towards the electricity for the air conditioner is halved. This would encourage localized production of Bt at village level with low capital investment coupled with employment generation which in turn would ensure the ready availability of quality product to the farmers.

Acknowledgements The authors thank the Project Director, DOR and Head, Crop Protection Section, DOR for providing the facilities. The research for this publication was conducted as part of the programme ÔBiotechnology for Dryland Agriculture in Andhra PradeshÕ with financial support

Table 7. Cost of seed culture, labour and electricity charges. Component

Quantity (g)

Cost (Rs.)

Seed culture Electricity charges @ Rs. 4.50 unit)1 Autoclave Shaker incubator Room temp. with AC at 30 Ca Centrifugation Labour @ Rs. 60/- per head

486 ml (@ Rs. 20/- for 800 ml)

12.15

4.0 units 48 units 30 units 48 h)1 2 units 10 h)1 for 2 centrifuges 2 labour

18.00 216.00 135.00 18.00 120.00

a

Since two shakers will be employed to run one batch per day, the cost of electricity towards AC per batch will be halved to Rs. 67.5 and the total expenditure towards seed culture electricity and labour would be Rs. 451.65.

Table 8. Cost of production of Bt multiplied on different media. Sl. no.

Heads of expenditure

Nutrient broth medium Molasses medium

Barley medium

1

Yield (g) of Bt from 8 l flasks of 1000 ml capacity Cost (Rs.)a of medium used in 8 1 flasks of 1000 ml capacity Material cost (Rs.)/l of spray solution at the effective dose (%w/v) Cost (Rs.) of production including material cost, seed culture, labour and electricity charges Amount of spray solution (l) from the Bt yield at Sl. no. 1 at the effective dose (%w/v) Cost (Rs.) of 1 l of spray solution

89.10

216.68

1539.00

405.00

55.40

109.60

3.24 at 0.071%

0.054 at 0.021%

0.071/l at 0.1%

405.00 + 451.65b = 856.65 125.5 @ 0.071%

55.40 + 451.65 = 507.05 1031 @ 0.021%

109.60 + 451.65 = 561.25 1539 @ 0.1%

Rs. 6.83/l @ 0.071% for 125.5 l

Rs. 0.49/l @ 0.021% for 1031 l

Rs. 0.36 (0.1%) for 1539 l

2 3 4 5 6

a b

One US dollar = Forty five Indian rupees approximately. Details at Table 7.

178 of the Research Communications Division, Ministry of Foreign Affairs, The Government of Netherlands. Responsibility for the contents and for the opinions expressed rests solely with the authors, publication does not constitute an endorsement by BTU or the funding agency.

References Morris, O.N., Kanagaratnam, P. & Converse, V. 1997 Suitability of 30 agricultural products and by-products as nutrient sources for laboratory production of Bacillus thuringiensis subsp. aizawai (HD133). Journal of Invertebrate Pathology 70, 113–120. Mummigatti, S.G. & Raghunathan, A.N. 1990 Influence of media composition on the production of endotoxin by Bacillus thuringiensis var. thuringiensis. Journal of Invertebrate Pathology 55, 147– 151.

P.S. Vimala Devi et al. Pomeranz, Y. 1973 Industrial uses of barley. In Industrial Uses of Cereals: Symposium Proceedings Held in Conjunction with 58th Annual Meeting of American Association of Cereal Chemists, Nov 4–8, 1973. St. Louis, Missouri. Y. Pomeranz, Chairman. St. Paul, AACC, 1973, XI, pp. 371–392. Salama, H.S., Foda, M.S., Dulmage, H.T. & El-Sharaby, A. 1983 Novel fermentation media for production of delta-endotoxins from Bacillus thuringiensis. Journal of Invertebrate Pathology 41, 8–19. Travers, R.S., Martin, P.A.W. & Reichelderfer, C.F. 1987 Selective process for efficient isolation of soil Bacillus spp. Applied and Environmental Microbiology 53, 1263–1266. Vora, D. & Shethna, Y.L. 1999 Enhanced growth, sporulation and toxin production by Bacillus thuringiensis subsp. kurstaki in oil seed meal extract media containing cystine. World Journal of Microbiology & Biotechnology 15, 747–740. Vimala Devi, P.S., Balakrishnan, K., Ravinder, T. & Prasad, Y.G. 2001 Identification of potent strains of Bacillus thuringiensis for the management of the castor semilooper Achaea janata (Linn) and optimization of production. Entomon 26(Silver Jubilee Special Issue), 98–103.

World Journal of Microbiology & Biotechnology 2005 21: 179–187 DOI: 10.1007/s11274-004-1766-7


Springer 2005

Induction, and production studies of a novel glucoamylase of Aspergillus niger M. Ibrahim Rajoka* and Amber Yasmeen National Institute for Biotechnology and Genetic Engineering (NIBGE), P.O. Box 577, Jhang Road, Faisalabad 651475, Pakistan *Author for correspondence: Tel.: +92-41-651475/550815; Fax: +92-41-651472; E-mail: [email protected] Received 28 April 2004

Keywords: Aspergillus niger, glucoamylase, induction, regulation, repression

Summary The influence of carbon and nitrogen sources on the production of a raw-starch-digesting glucoamylase was investigated. The enzyme production was variable according to the carbon source. Levels of glucoamylase were minimal in the presence of even low concentrations of glucose while its production was stimulated by other carbohydrates. Wheat bran and cellulose were the most effective inducers of glucoamylase activities, followed by rice bran. Exogenously supplied glucose inhibited the synthesis of the enzyme in cultures of A. niger growing on wheat bran. In defined medium with maltose, the glucoamylase titres were 5.2- to 16.7-fold higher with cells growing on monomeric sugars and 1.5 times higher than cells growing on other disaccharides. It appeared that synthesis of glucoamylase varied under an induction mechanism, and a repression mechanism which changed the rate of synthesis of enzyme in induced over non-induced cultures. In this organism, substantial synthesis of glucoamylase could be induced by maltose, cello-dextrin, cellulose or cellulose- and hemicellulose-containing substrates which showed low volumetric substrate uptake rate. During growth of A. niger on wheat bran, maximum volumetric productivity (Qp) of glucoamylase was 274 IU l)1 h)1 and is significantly higher than the values reported for some other potent fungi. The addition of actinomycin D (a repressor of transcription) and cycloheximide, (a repressor of translation) completely repressed glucoamylase biosynthesis, suggested that the regulation of glucoamylase synthesis in this organism occurs at both transcriptional and translational level. Thermodynamic studies revealed that the culture exerted protection against thermal inactivation when exposed to different fermentation temperatures.

Nomenclature Qx Qs qs l Qp Yp/s Yp/x qp DHx DH
D DS* DS
D

rate of cell mass formation (g cells l)1 h)1) rate of substrate consumption (g substrate l)1 h)1) specific rate of substrate consumption (g substrate consumed g)1 cells h)1) specific growth rate (h)1) rate of a-glucoamylase formation (IU l)1 h)1) a-glucoamylase yield (IU g)1 substrate utilized) specific yield of enzyme production (IU g)1 cells) specific rate of enzyme production (IU g)1 cells h)1) activation enthalpy for production formation (kJ mol)1) activation enthalpy for inactivation formation (kJ mol)1) action entropy for product formation (kJ mol)1 K)1) action entropy for product inactivation (kJ mol)1 K)1)

Introduction Glucoamylase (amyloglucosidase) (a-1,4-glucan glucohydrolase EC 3.2.1.3) is used in production of high glucose syrup, which is used to produce crystalline glucose and high fructose syrup. Other applications include the production of different antibiotics, amino acids, ethanol and organic edible acids (Nigam & Singh 1995). A successful process for enzymic conversion of different starches depends on the enzymes’ ability to remove single glucose residues from a soluble oligosaccharide from the non-reducing end until all of the starch is converted into glucose for subsequent application. In this process, glucoamylase is usually used in combination with an amylopectin-debranching enzyme, such as pullulanase or isoamylase. Although a number of glucoamylases are known to hydrolyse both the a-1,4- and a-1,6-glucosidic linkage, their activity towards the latter is very low. We screened a large number of fungi and finally found that A. niger NIAB 280 was capable of producing a raw-starch-digesting

180 enzyme. Among moulds, Aspergillus spp. and Rhizopus spp. are the main glucoamylase producers on industrial scale. Just like other enzymes, synthesis of glucoamylase can be induced by many oligomeric and dimeric sugars (Hrmova et al. 1991; Magnelli & Forchiassin 1999). Starch and maltose components in starchy substrates are essential for formation of mRNA to support maximum formation of glucoamylase at the transcriptional level. Recently, numerous studies have been carried out concerning the molecular biology of glucoamylase in various organisms such as bacteria and fungi (Fowler et al. 1990; Devchand & Gwynne1991; Verdoes et al. 1994; Cha et al. 1997; Morlino et al. 1999; Ma et al. 2000). It was realized that A. niger, given its environment and the decay it produces on vegetation, was likely to produce glucoamylase in substantial amounts. Once this was confirmed, it was decided to use this microorganism in these investigations to produce glucoamylase not only from starchy wastes but also from lignocellulosic (LC) wastes which are cheap resources for production of vendible products (Van Wyk 2001). A study of the production of glucoamylase, measured on soluble starch correlating with substrate utilization parameters by A. niger was performed. This work also reports the induction, repression, and production of the enzyme. Efforts to establish the optimal culture conditions for production of glucoamylase are also reported.

Materials and methods Organism Aspergillus niger NIAB 280 was used throughout these studies. The strain was maintained on potato-dextrose agar plates and slants as described earlier (Siddiqui et al. 1997). All chemicals were purchased from Sigma Chemical Co., Missouri, USA. Growth media Aspergillus niger was grown in glucose Vogel’s salt medium supplemented with 0.2% yeast extract (Siddiqui et al. 1997) containing acid-washed glass beads. The latter were added to achieve uniform turbidity. The pH of the medium was adjusted to 6.5 with 1 M HCl. For other growth studies, the seed culture developed on glucose was used as inoculum and washed twice with sterile saline before use. This inoculum was used to study the influence of other carbon sources, (mentioned in the text and in tables) on glucoamylase (GAM) synthesis. Actinomycin and cycloheximide were added at 15 and 20 lg ml)l concentrations, respectively, as earlier tests indicated that higher concentrations inhibited the growth of the organism while lower concentrations had no significant influence on the synthesis of the enzyme.

M.I. Rajoka and A. Yasmeen Substrates and their preparation All lignocellulosic (LC) substrates were obtained from local sources. The dry powder of LC biomass was alkalitreated as described earlier (Rajoka & Malik 1997). The treated biomass of bagasse, rice husk and corncobs had 84  1.2, 82  1.0 and 81  1.5% total saccharides respectively determined using standard methods (Latif et al. 1994). Batch-culture studies The ability of the organism to utilize rice husk, corn cobs, bagasse, wheat bran, mono-saccharides and disaccharides for improved production of GAM with reference to soluble starch, was examined in basal Vogel salts’ medium containing 0.2% yeast extract and 0.2% (v/v) Tween 80 as described earlier (Rajoka & Malik 1997). Carbon sources were added individually to batches of basal medium to give a saccharide level of 20 g l)1 (found to be optimal). All media were adjusted to pH 6.5 with 1 M NaOH or 1 M HCl and were dispensed in 50 ml aliquots into 250-ml Erlenmeyer flasks in triplicate. The time course of GAM production in shake-flask batch cultures was examined at 30 C on a gyratory shaker at 150 rev min)1. Sample flasks in triplicate were withdrawn after predetermined time intervals and processed. The amount of growth, reducing sugars released from polysaccharides, protein production and enzyme activities present in the extra cellular fraction were assayed. When the test organism was grown on insoluble substrates, the culture medium after growth was centrifuged (10 min, 4000 · g) to remove substrate. The residue was shaken with chilled water containing 1% (v/v) Tween 80 for 30 min at 4 C and clear supernatant was obtained by centrifugation (15,000 · g, 15 min). Further washings were collected until enzyme activity was not observed in the supernatant. All washings were pooled for determining enzyme activity and compensated for the adsorbed portion of enzyme. The washed substrate was oven-dried to constant weight for further processing. The amount of growth was measured gravimetrically as dry cell mass. For this purpose, 50 ml portions of cell suspension were also centrifuged (15,000 · g, 30 min). The cell-free supernatant was preserved for enzyme assays and cell pellets were washed twice with physiological saline, suspended in 10 ml distilled water and dried at 70 C to constant mass. Clear supernatant from 50 ml original culture broth was obtained by centrifugation (15,000 · g, 30 min, 10 C). The cell pellet was used to extract cellular fractions as described earlier (Rajoka et al. 1998a). The enzyme activity present in the cell-free supernatant or cell extract was assayed. Enzyme assays For GAM assays (Iqbal et al. 2003), the appropriately diluted culture supernatant or cell extract was incubated

Glucoamylase of Aspergillus niger with soluble starch (1%) as enzyme substrate. Glucose released was measured using a glucose oxidase kit. One unit of enzyme activity is defined as the amount of enzyme which releases 1 lmol glucose per ml per min. Saccharide determination In these tests, reducing sugars were estimated colorimetrically with 3,5-dinitrosalicylic acid after Miller (1959) using glucose as standard. Glucose was determined using Human (Germany) glucose kit following instructions of the suppliers. Polysaccharides, disaccharides and glucose in fermentation broth were also determined using HPLC (Rajoka et al. 2003). Protein determination The protein was determined by the Bradford method using bovine serum albumin as the standard. The protein content in the substrate and the spent dry matter was determined by multiplying the nitrogen content determined by Kjeldahl’s method by 6.25 (Duenas et al. 1995). Effect of varying pH and temperature on enzyme production The effect of initial pH of the fermentation medium on enzyme production parameters was studied in shake flasks by varying pH (5.0–7.5) and maintaining optimum temperature and other growth supporting conditions. For studying the effect of temperature, the experiments were repeated in shake flasks (250 ml each) and incubated on an orbital shaker (150 rev min)1) at either 20, 22, 24, 26, 28, 30, 32, 34, 36 or 38 C for up to a period of 96 h (in time course study). The enzyme preparations were analysed for enzyme activities as described earlier. The enzyme prepared as above was concentrated using an Amicon concentrator with a hollow fibre cartridge (cut-off size ¼ 10–15 kDa). Portions of 7.5 g of commercial maize starch and 50 ml of concentrated glucoamylase (containing 200 IU ml)1, pH 5.5) were dispensed into 250 ml Erlenmeyer flasks in triplicate and incubated at 35 C on shaking incubator. Flasks (in triplicate) were harvested periodically and properly diluted aliquots were used to determine glucose, maltose and oligosaccharides by HPLC. The above process was repeated with maize starch (7.5 g flask)1). Considering the optimal enzymatic activity and thermostability of a-amylase and glucoamylase (Sigma Chemical Company, USA), 100 U a-amylase ml)1 was used at 95 C for 30 min for gelatinization. The saccharification was brought about with 100 IU ml)1 a-amylase and 200 IU ml)1 glucoamylase at 60 C. The sugars released were measured periodically on triplicate flasks. Determination of kinetic parameters Dry cell mass (g l)l) of A. niger, after growth on different carbon sources, in time course study, was determined on

181 triplicate samples. Enzyme activities (product l)1) were determined as mentioned earlier in this section. The maximum volumetric substrate uptake rate (Qs), cell mass formation rate (Qx) and enzyme production rate (Qp), and other kinetic parameters were estimated after Pirt (1975). Statistical analysis Treatment effects on well replicated treatments (in triplicate) were compared by the protected least significant difference method as described earlier (Rajoka et al. 1998a). Significance of difference has been presented as one-factor completely randomized factorial design in the form of probability (P) values using MstatC software.

Results and discussion Time course of glucoamylase (GAM) production and substrate utilization The cultures were grown in time course studies to determine the minimum time for induction, other kinetic parameters and biosynthesis of enzyme and cell mass; optimum time was 80–96 h while lag phase was 8 h before GAM increased appreciably. The organism grown in Vogel’s medium for different time intervals was processed for substrate, cell mass and assays of glucoamylase (Tables 1–5). The enzyme was produced using 2% saccharides present in the substrate or equivalent saccharides as described in Materials and methods. Enzyme production, cell mass biosynthesis and substrate utilization kinetics of representative substrates namely soluble starch (a), maltose (b), cellobiose (c) and CMC (d) under the conditions of shake-flask cultures are shown in Figure 1. Maximum cell mass was supported by maltose. Corn cobs, rice husk as well as wheat bran were easily degraded e.g. Qs of corn cobs was 0.40 g l)1 h)1 and was significantly (P £ 0.05) higher and different from those of other carbon sources (Table 3). Enhanced substrate consumption and cell mass formation rates by A. niger could lead to increased productivities in commercial glucoamylase production from agro-industrial residues available in Pakistan. Evaluation of carbon sources for induction/repression of glucoamylase Initially the concentration of substrate (maltose) was optimized; 2% substrate concentration was optimum to provoke maximum production of glucoamylase. To evaluate the carbohydrates to cause induction or repression of glucoamylase, A. niger was grown on many monosaccharides, disaccharides, and carboxymethylcellulose. Following growth on different substrates, there was significant (P £ 0.05) variation in Qs with different carbon sources with corresponding variation of gluco-

182

M.I. Rajoka and A. Yasmeen

Table 1. Comparative fermentation kinetic parameters of A. niger for glucoamylase production and substrate utilization parameters following growth on different substrates in submerged fermentation. Carbon source

Qp (IU l)1 h)1)

Arabinose Fructose Galactose Glucose Xylose Lactose Maltose Sucrose Cellobiose Dextrin CMC Soluble starch

9.7 31.0 11.0 1.9 20.0 45.2 67.4 44.0 47.0 90.0 84.5 104.9

LSD (P £ 0.05) Significance (P £ 0.05)*

0.458 HS

± ± ± ± ± ± ± ± ± ± ± ±

0.04k 0.12g .05j 0.001j 0.11h 0.21f 0.12e 0.11f 1.0d 0.64b 1.1c 3.1a

Yp/s (IU g)1)

Specific act. (IU mg)1 prot.)

21.0 ± 1.0i 30.5 ± 1.2h 15.0 ± 1.0k 2.1 ± 0.02l 30.0 ± 1.2h 67.0 ± 1.5g 120 ± 2.1e 97.0 ± 1.2f 123.0 ± 3.5d 152.0 ± 3.8c 155.0 ± 2.3b 266.3 ± 10a

0.8 3.5 1.7 0.08 2.5 6.5 7.9 5.9 9.3 10.2 10.4 12

2.145 HS

0.05404 HS

0.01k 0.02h 0.01j 0.001l 0.06i 0.08f 0.05e 0.06g 0.09d 0.07c 0.04g 0.10a

± ± ± ± ± ± ± ± ± ± ± ±

Qs (g l)1 h)1) 0.47 0.31 0.44 0.50 0.39 0.28 0.40 0.59 0.46 0.45 0.26 0.26

± ± ± ± ± ± ± ± ± ± ± ±

0.05404 HS

0.02bc 0.015f 0.021cde 0.029b 0.012e 0.011f 0.016de 0.022a 0.015bc 0.014bcd 0.012f 0.11f

qs (g g)1 h)1) 0.44 0.54 0.64 0.74 0.43 0.37 0.33 0.44 0.48 0.45 0.39 0.37

± ± ± ± ± ± ± ± ± ± ± ±

0.021cd 0.026b 0.015cf 0.021a 0.024cd 0.018ef 0.014f 0.021cd 0.023c 0.022c 0.019dc 0.10ef

0.05404 HS

Each value is a mean of three replicates. ± shows standard error among replicates. Means followed by different letters in each column differ significantly at P £ 0.05 using one factor factorial design in MstatC software. Qp = Glucoamylase formation rate, Yp/s = GAM produced g)1 substrate utilized, Qs = substrate consumption rate, qs = specific substrate consumption rate. * HS is highly significant.

Figure 1. Kinetics of glucoamylase (GAM) production in shake flask fermentation of four representative substrates namely (a) soluble starch, (b) maltose, (c) cellobiose, and (d) CMC. The initial pH of the medium was 6.5, inoculum size 10%, on 2% (w/v) substrates, and temperature 30 C. ¼ GAM, n ¼ cell mass, and h ¼ substrate present in the fermentation medium. Error bars show standard deviation among three replicates.



amylase biosynthesis. The values of Qs (table 1) on monomeric sugars (0.43 ± 0.084 g l)1 h)1), from dimeric sugars (0.347 ± 0.011 g l)1 h)1), CMC )1 (0.26 ± 0.011 g l h)1), and soluble starch (0.26 ± 0.021 g l)1 h)1) were different and exerted inverse relationship on synthesis of glucoamylase. Fructose among monomeric saccharides was found to be the

best inducer. This study substantiates the work of Nochure et al. (1993) who demonstrated fructose as the best inducer of Avicellase in Clostridium thermocellum. Bagga et al. (1989) identified lactose as the best inducer of endoglucanase and cellobiohydolase. Trehalose has been demonstrated as the best inducer of cellulases in a Clostridium sp. (Thirumale et al. 2001).

Glucoamylase of Aspergillus niger

183

Table 2. Comparative fermentation kinetic parameters of A. niger for glucoamylase formation and substrate consumption parameters following growth on different insoluble substrates in submerged fermentation. Carbon source

Qp (IU l)1 h)1)

Yp/s (IU g)1 substrate)

Corn cobs Rice husk Rice polishing CMC a-cellulose Wheat bran Bagasse Maize starch

109.2 ± 0.8b 64.8 ± 0.2f 98.0 ± 1.1d 84.5 ± 0.7g 92.0 ± 1.5b 176.8 ± 1.8a 104.31.0e 46.6 ± 2.0de

362 203 270 126.0 346 375 330 363

LSD (P £ 0.05) Significance (P £ 0.05)*

0.5799 HS

0.6796 HS

± ± ± ± ± ± ± ±

11.2c 11.3f 12.8e \2.0g 13.5c 24.1a 24.8d 19.2b

Specific act. (IU mg)1 protein) 19.1 11.5 15.3 10.4 19.2 21.4 18.7 11.7

± ± ± ± ± ± ± ±

b

Yp/x (g g)1 cells) 565 400 512 250 480 745 550 566

f

0.1d 0.01g 0.2e 0.15a 0.2c b

0.323 HS

± ± ± ± ± ± ± ±

61.2b 31.2f 51.2d 21.2f 35.2e 50.2a 51.2c 63.2b

1.361 HS

Each value is a mean of three replicates. ± shows standard error among replicates. Means followed by different letters in each column differ significantly at P £ 0.05 using one factor factorial design in MstatC software. Qp = Glucoamylase formation rate, Yp/s = GAM produced g)1 substrate utilized, Yp/x = GAM produced g)1 cells. * HS is highly significant.

From soluble carbohydrates, low levels of GAM, were attributed to the organism’s low requirement of the enzyme for growth and metabolism. There was greater enhancement in GAM productivity (2.6-fold enhancement) following growth on wheat bran (Table 2) over that obtained from maltose (Table 1). Additionally the level of GAM varied over a 5.2- to 16.7-fold range with the non-inducing carbon sources namely, xylose, galactose, arabinose. compared with that on glucose and 24.7- to 56.5-fold higher on inducing carbon sources namely maltose, cellobiose, cello-dextrin and a-cellulose and showed inverse relationship with Qs. The difference in synthesis of GAM by the carbon sources is mainly due to repression of mRNA formation (Fowler 1993; Li & Ljungdahl 1994) because all carbon sources suppressed the GAM synthesis only when they were present in greater excess (>1%) than the amount expected to be released from complex carbon sources during growth state. The glucoamylase activities (Tables 1–3) accumulated in the culture supernatants; they were scarce in cell extracts. Enzyme (12.5 ± 0.75%) was attached to the solid surface of the insoluble substrates and could be effectively eluted with Tween 80 in three washings. Values presented in Tables 2–6 have been compensated

to include these activities. A. niger produced maximum GAM from wheat bran medium followed by a-cellulose and minimum levels on CMC medium. Productivity of GAM synthesis from wheat bran was significantly (P £ 0.05) higher than that from other substrates (Tables 2 and 3). In all substrates, more than 98–100% carbohydrates were utilized. In all carbon sources, the level of expression of GAM was distinctly different. During growth of the organism on different cellulosic and LC substrates, reducing sugars accumulated slowly in the growth medium as unmetabolized substances and induced GAM. A. niger released relatively less reducing sugars (Table 3) from some LC substrates and secreted the highest amount of GAM while from some other substrates released the highest amount of sugars, and synthesized the lowest amount of GAM (Tables 2 and 3), e.g., CMC. This may have occurred due to production of gentiobiose, which is a strong inducer of cellulases in fungi and release of glucose is slow to cause induction of enzyme syntheses (Suto & Tomita 2001). Volumetric productivity of GAM and growth parameters of the organism in VYE-wheat bran medium when glucose was added (1, 2.5, 5 and 10 g l)1 final concentration) at the time of inoculation (Table 4), showed

Table 3. Comparative fermentation kinetic parameters of A. niger for substrate consumption parameters following growth on different insoluble substrates in submerged fermentation. Carbon source

QEP (mg l)1 h)1)

Corn cobs Rice husk Rice polishing CMC a-cellulose Wheat bran Bagasse Maize starch

5.82 5.72 6.52 5.51 6.68 8.42 5.6 5.41

± ± ± ± ± ± ± ±

0.4e 0.3f 0.3c 0.4g 0.5b 0.8a 0.4d 0.4d

Yx/s (g g)1)

RS (mg l)1)

0.46f 0.35 0.45d 0.46e 0.52b 0.55a 0.45c 0.45e

70 83 75 85 80 52 72 71

± ± ± ± ± ± ± ±

5.1d 5.1d 5.1d 7.1d 6.2a 4.15b 5.2c 5.1d

Qs (g l)1 h)1) 0.40 0.39 0.26 0.26 0.39 0.51 0.41 0.39

± ± ± ± ± ± ± ±

.0.01c 0.02c 0.01d 0.01d 0.02c 0.03a 0.02b .0.01c

Qx (g cells l)1 h)1) 0.43 0.37 0.40 0.33 0.42 0.53 0.43 0.43

± ± ± ± ± ± ± ±

0.05b 0.01c 0.02c 0.02b 0.02b 0.012a 0.01b 0.05b

Each value is a mean of three replicates. ± shows standard error among replicates. Means followed by different letters in each column differ significantly at P £ 0.05 using one factor factorial design in MstatC software. QEP = Extracellular protein formation rate, Yx/s = Cells produced g)1 substrate utilized. For all other parameters, see Tables 1 and 2.

184

M.I. Rajoka and A. Yasmeen

Table 4. Effect of addition of glucose to wheat bran medium on glucoamylase formation and substrate consumption parameters. Glucose concn. (g 100 ml)1)

Qp (IU l)1 h)1)

0.0 1.0 2.5 5.0 10.0

176.8 171.9 139.0 125.4 105.9

LSD (P £ 0.05) Significance (P £ 0.05)*

0.665 HS

± ± ± ± ±

12.1a 11.3b 11.5c 11.0d 10.7e

Qx (g l)1 h)1) 0.29 0.31 0.33 0.34 0.38

± ± ± ± ±

0.02c 0.021bc 0.023bc 0.21b 0.023a

0.0407 HS

Qs (g l)1 h)1) 0.34 0.31 0.26 0.23 0.21

± ± ± ± ±

0.019a 0.023a 0.014b 0.012c 0.011c

0.0407 HS

Each value is a mean of three replicates. ± Shows standard error among replicates. Means followed by different letters in each column differ significantly at P £ 0.05 using one factor factorial design in MstatC software. * HS is highly significant.

significant decrease in the enzyme synthesis. All treatments had statistically significant (P £ 0.05) influence on GAM productivity, substrate consumption and cell mass formation rates. It was found that glucose enhanced the cell mass productivity, but suppressed GAM productivity and substrate consumption rate. Mixed inductive or repressive effects have been observed in other organisms (Magnelli & Forchiassin 1999). This effect has not been observed in production of other enzymes by other organisms (Rajoka & Malik 1997). Catabolite repression plays an important role in the regulation and secretion of inducible enzymes. Such a repression effect has been observed in other organisms (Morlino et al. 1999; Peixoto et al. 2003). The effect of different glucose concentrations on GAM activity was determined in order to differentiate it from the effect of glucose on the synthesis of this enzyme; the same enzymatic activity values being obtained with these glucose concentrations studied, thus suggesting that the decrease in the content of enzyme, together with the increase in the initial concentration of glucose in the fermentation medium, were due to the negative effect of sugar on the synthesis of this enzyme. Moreover, when the level of available sugar decreased as a result of culture growth, the synthesis of enzyme increased up to 96 h of culturing. From there onwards, there was a decrease in enzyme synthesis. The effect of glucose was investigated with mycelium harvested during stationary phase in order to minimize

the influence of growth. When glucose was added at the beginning of the studies, GAM production ceased even though the inducer (wheat bran) was also present. However, when after 8 h of incubation, the mycelium was washed free of glucose and placed in glucose-free medium containing wheat bran, the synthesis of GAM started again, thus proving the reversibility of the repression mechanism of the synthesis of GAM by glucose. It can be concluded that GAM is an inducible enzyme. Inducer plus glucose strongly affected GAM synthesis. Even when the wheat bran concentration was the same as that of glucose, GAM level was lower than control, a fact that showed that the rate of production of GAM was not completely restored. However, apparently there is a basal level of constitutive enzyme, since it was produced in a medium containing sugars other than maltose. Similar results have been reported by other workers (Chiquetto et al. 1992). The addition of a repressor of transcription such as actinomycin D reduced GAM biosynthesis during the first 8 h, after which time, synthesis of the enzyme ceased altogether. This observation suggested that translation occurs as long as there is mRNA available in the system. When cycloheximide, an inhibitor of translation, was added instead of actinomycin D, GAM production decreased markedly. The above observation suggested that the regulation of GAM synthesis in this organism occurs at both transcriptional and transnational level. The enzyme was provoked by wheat bran and repressed by glucose.

Table 5. Comparative fermentation kinetic parameters of A. niger for GAM formation following growth on wheat bran in the presence of different nitrogen sources in shake-flask cultures under optimized culture conditions. Nitrogen source

Qp (IU l)1 h)1)

Yp/s (IU g)1 substrate)

Ammonium nitrate Ammonium sulphate Corn steep liquor Sodium glutamate Sodium nitrate Urea

97.1l 135.0d 274.0a 201.7c 139.0f 212.8b

275 299 512 450 267 475

LSD (P £ 0.05) Significance (P £ 0.05)*

3.1 HS

1.921 HS

± ± ± ± ± ±

45l 46d 50a 43c 31f 41b

Yp/x (IU g)1 cells) 550 ± 50d 598 ± 45c 1024 ± 75a 950 ± 75b 534 ± 50e 950 ± 70b 3.85 HS

Each value is an average of two replicates. ± Shows standard error between the replicates. Means followed by different letters in each column differ significantly at P £ 0.05 using one factor factorial design in MstatC software. * HS is highly significant.

Glucoamylase of Aspergillus niger

185

Effect of nitrogen sources Varying the nitrogen source to give an equimolar amount of nitrogen maintained at 1.2 g l)1, indicated that soy bean, and corn-steep liquor were the best nitrogen sources followed by glutamate and ammonium sulphate. Nakamura & Kitamura (1988) found that peptone was the most favorable nitrogen source for protein and cellulases synthesis but Rajoka (1990) observed that an increase in NaNO3 concentration greatly increased cellulase synthesis in Cellulomonas biazotea. The variations in values of GAM suggest that there is another regulatory mechanism for GAM synthesis in addition to induction. This regulation mechanism lowers the biosynthesis of GAM when the organism is grown on easily metabolizable substrates and different nitrogen sources. Derepression of GAM production was obtained by limiting carbon and nitrogen sources up to a certain level; phosphorous sources at 3-fold concentration of that in the fermentation medium, did not show any repressive effect. Aspergillus niger supported maximum volumetric productivity, GAM yield and specific yield of 274 IU l)1 h)1, 512 IU g)1 substrate and 1024 IU g)1 cells on wheat bran medium (supplemented with corn steep liquor) respectively. These levels are higher than those reported in yeast cultures (Mase et al. 1996; Cha et al. 1997; Furuta et al. 1997) and filamentous fungal cultures (Papagianni et al. 2002; Stamford et al. 2002; Peixoto et al. 2003) grown on different substrates. The specific productivity of glucoamylase from A. niger (127 IU g)1 h)1) on maltose (results not presented) is 16.4-fold of that from a hyper-glucoamylase producing A. niger strain (35 IU g)1 h)1; Metwally 1998) grown in continuous culture. Effect of pH and temperature on glucoamylase production/induction The optimum pH and temperature for maximum production of GAM kinetic parameters were 6.5 (Figure 2a) and 30 C (Figure 2b). Maximum activity in bacteria was obtained around neutral pH of the medium and temperature 30 C but fungi vary with respect to pH and temperature to support maximum production of glucoamylase In these studies, 30 C temperature was found optimum to support maximum production of GAM as observed by other workers (Ma et al. 2000; Papgianni et al. 2002). At higher or lower temperature, the organism has to spend a lot of energy for maintenance purposes (Pirt 1975). Thermodynamics of GAM production The approach of Arrhenius (Aiba et al. 1973) was used to describe the relationship of temperature-dependent reversible and irreversible inactivation kinetics of GAM production. The activation enthalpy of GAM production was graphically calculated by application of the

Figure 2. Effect of fermentation medium pH (a) and fermentation temperature (b) on volumetric (Qp) and product yield coefficient (Yp/s) of glucoamylase following fermentation of wheat bran in Vogel’s medium . ¼ Volumetric productivity (Qp) and d ¼ product yield coefficient (Yp/s). Error bars show standard deviation among three replicates.



Arrhenius approach extended to microbial processes (Arni et al. 1999). Application of the Arrhenius type equation (Equation (1) below) to estimate enthalpies and entropies of GAM formation and thermal inactivation using specific productivity values is shown in Figure 3 qp ¼ TkB =heDS=R eDH




=RT

ln qp =T ¼ ln kB =h þ DS
=R  DH
=RT

ð1Þ ð2Þ

The values of the thermodynamic parameters estimated with above model are presented in Table 6. The activation enthalpy of GAM formation (DH* ¼ 92.7 kJ mol)1) compares well with that for phytase (70–80 kJ mol)1) reported by Al-Asheh & Duvniak (1994). The phenomena responsible for thermal inactivation of enzyme are characterized by an activation enthalpy (DHD
¼ 107:7 kJ mol)1 K)1) that is remarkably higher than that for GAM production. This means that its rate decreases much faster with temperature than product formation rate. Therefore, the overall specific productivity falls sharply above 30 C. The value of DH
D is significantly lower than the values reported for an

186

M.I. Rajoka and A. Yasmeen

Figure 3. Arrhenius plot to calculate enthalpy (DH*) and entropy (DS*) of glucoamylase formation and its deactivation applying relationship: ln(qp/T) ¼ ln(kB/h) + DS*/R ) DH*/R.1/T (for details see Table 6).

Table 6. Thermodynamic parameters estimated by Arrhenius approach for batch formation of GAM from wheat bran by Aspergillus niger. GAM formation Activation enthalpy (kJ mol)1) Activation entropy (J mol)1 K)1)

Thermal inactivation

92.7 ± 4b

107.7 ± 6a

108.0 ± 5a

)470.0 ± 12b

Enthalpy values were calculated from slopes of straight lines in Figure 3 (slope = )DH/RT) and entropies were calculated from the intercepts on Y-axis (intercept = ln (KB/h)+DS/RT) using Eyring’s equation: ln(qp/T) = ln(KB/h)+DS/RT) ) DH/RT where qp, KB, h, DS, R, T and DH are specific rate of product formation, Boltzmann constant, Planck’s constant, entropy of activation, gas constant, absolute temperature and enthalpy of activation respectively. Each value is a mean of three replicates. Values followed by different letters in each row differ significantly at P £ 0.05 applying one factor factorial design using MstatC software.

glucose isomerase (160–235 kJ mol)1) (Converti & DelBorghi 1997). The activation entropy of GAM formation (108 J mol)1 K)1) is very low and compares favorably with those of glucose isomerase (Converti & DelBorghi 1997). The entropy value of thermal inactivation ()470 J mol)1 K)1) is also very low and compares favorably with b-galacosidase from a thermotolerant yeast (Rajoka et al. 2003) (and bears a negative symbol). This reflects that this inactivation phenomenon implies a little randomness increase during the activated state formation. This suggests a sort of protection exerted by cell system against thermal inactivation.

Figure 4. Saccharification of raw starch using glucoamylase (200 IU g)1) from A. niger (top graph) and commercial a-amylase and glucoamylase (200 IU g)1 starch each) (bottom graph) under optimal working conditions (Ma et al. 2000) as described in Materials and methods.

obtained (90%) using Sigma enzyme preparations under their optimal working conditions (Figure 4: Sigma enzymes) as given in Materials and methods.

Conclusion These studies led us to conclude that the organism may serve as a good source for raw-starch-digesting GAM production, often deficient in many organisms. The scope to increase the production by manipulating other cultural conditions cannot be ignored. Addition of glucose in growing cultures indicated that production is regulated by both induction and repression. Isolation of a derepressed mutant after Rajoka et al. (1998a) might enhance the potential of this organism for GAM formation. Thermodynamic studies provided insight into the mechanism of product formation and led us to conclude that the organisms exerted protection against thermal inactivation. Thus organism can be used for large scale production of GAM at a wide range of temperatures. Genetic engineering can further increase the productivity of the enzyme (Rajoka et al. 1998b; Ma et al. 2000).

Saccharification of raw starch

Acknowledgements

The saccharification was carried out at 35 C using concentrated GAM with shaking up to 24 h. The enzyme could liberate 80% sugars (Figure 4: Test enzyme) with in 24 h. The reaction mixture contained 80% glucose and remaining oligosaccharides. The maximum hydrolysis rate was comparable with that

These studies were supported in part by the Pakistan Atomic Energy Commission which provided facilities to complete this work. This work formed a part of M. Phil thesis of Amber Yasmeen who was financially supported by Fauji Foundation Some chemicals were purchased from USAID Proposal PSTC 6.163.

Glucoamylase of Aspergillus niger References Aiba, S., Humphrey, A.E. & Millis, N.F. 1973 Biochemical Engineering, 2nd edn. New York: Academic Press. pp. 92–127. Al-Asheh, S. & Duvniak, Z. 1994 Characteristics of phytase produced by Aspergillus carbonarius NRC 401121 in canola meal. Acta Biotechnologica 14, 223–237. Arni, S., Molinari, F., DelBorghi, M. & Converti, A. 1999 Improvement of alcohol fermentation of a corn starch hydrolysate by viscosity raising additives. Sta¨rke/Starch 51, 218–224. Bagga, P.S., Sandhu, D.K. & Sharma, S. 1989 Catabolite repression of cellulase production in Aspergillus nidulans. Process Biochemistry 24, 41–45. Cha, H.I., Ghoi, S.S., Yoo, Y.J. & Bentley, W.E. 1997 Enhancement of production of cloned glucoamylase under conditions of low aeration from recombinant yeast using a SUC 2 promoter. Process Biochemistry 32, 679–684. Chiquetto, M.L., Facciotti, M.C.R., Kilikian, B.V. & Schmidell, W. 1992 Influence of carbon and nitrogen sources on glucoamylase production by Aspergillus in batch process. Biotechnology Letters 14, 465–470. Converti, A. & Del Borghi, M. 1997 Simultaneous effects of immobilization and substrate protection on the thermodynamics of glucose isomerase activity and inactivation. Enzyme and microbiol Technology 21, 511–517. Devchand, M. & Gwynne, D. I991. Expression of heterologous proteins in Aspergillus. Journal of Biotechnology 17, 3–10. Duenas, R., Tengerdy, R. & Gutierrez-Corea, M. 1995 Cellulase production by mixed fungi in solid-substrate fermentation of bagasse. World Journal of Microbiology and Biotechnology 11, 333– 337. Fowler, T. 1993 Deletion of the Trichoderma reesei b-glucosidase gene, bglI. In b-Glucosidase, Biochemistry, and Molecular Biology, ed. Esen, A. pp. 56–65. Washington, DC, USA: American Chemical Society. ISBN 0-84122697-0. Fowler, T., Berka, R.M. & Ward, M. 1990 Regulation of the glaA gene of Aspergillus niger. Current Genetics 18, 537–545. Furuta, H., Arai, T., Hama, H., Shiomi, N., Kondo, A. & Fukuda, H. 1997 Production of glucoamylase by passively immobilized cells of a flocculant yeast, Saccharomyces diasticus. Journal of Fermentation and Bioengineering 84, 169–171. Hrmova, M., Petrakova, E. & Biely, P. 1991 Induction of celluloseand xylan- degrading enzyme systems in Aspergillus terreus by homo- and hetero-disaccharides composed of glucose and xylose. Journal of General Microbiology 137, 541–547. Iqbal, Z., Rashid, M.H., Jabbaar, A., Malana, M.A., Khalid, A.M. & Rajoka, M.I. 2003 Kinetics of enhanced thermostability of extracellular glucoamylase from Arachniotus sp. Biotechnology Letters 25, 1667–1670. Latif, F., Rajoka, M.I. & Malik, K.A. 1994 Saccharification of Leptochloa fusca (kallar grass straw) by thermostable cellulases. Bioresource Technology 50, 107–112. Li, X.-L. & Ljungdahl, L.G. 1994 Cloning, sequencing and regulation of a xylanase gene 2 from the fungus Aureobasidium pullulans Y-2311-1. Applied and Environmental Microbiology 60, 3160–3166. Ma, Y.J., Lin, LL., Chien, H.R. & Hsu, W.H. 2000 Efficient utilization of starch by a recombinant strain of Saccharomyees cerevisiae producing glucoamylase and isolamylase. Applied and Environmental Microbiology 53, 1352–1357. Magnelli, P. & Forchiassin, F. 1999 Regulation of the cellulase complex production by Saccobolus saccoboloides: induction and repression by carbohydrates. Mycologia 91, 359–364. Mase, T., Matsumiya, Y., Mori, S. & Matsuura, A. 1996 Purification and characterization of a novel glucoamylase from Acremonium sp. YT-78. Journal of Fermentation and Bioengineering 81, 347– 350. Metwally, M. 1998 Glucoamylase production in continuous cultures of Aspergillus niger with special emphasis on growth parameters. World Journal of Microbiology and Biotechnology 14, 113–118.

187 Miller, G.L. 1959 Use of dinitrosalisylic acid (DNS) for determination of reducing sugars. Analytical Chemistry 31, 426–428. Morlino, G.B., Tizzani, L., Fleer, R., Frontali, L. & Bianchi, M.M. 1999 Inducible amplification of gene copy number and heterogonous protein production in the yeast Kluyveromyces lactis. Applied and Environmental Microbiology 65, 4808–4813. Nakamura, K. & Kitamura, K. 1988 Cellulases of Cellulomonas uda. Methods in Enzymology 160, 211–216. Nigam, P. & Singh, D. 1995 Enzyme and microbial systems involved in starch processing. Enzyme and Microbial Technology 17, 770– 778. Nochure, S.V., Roberts, M.F. & Demain, A.L. 1993 True cellulase production by Clostridium thermocellum grown on different carbon sources. Biotechnology Letters 15, 641–646. Papagianni, M., Joshi, N. & Moo-Young, M. 2002 Comparative studies on extracellular protease secretion and glucoamylase production by free and immobilized Aspergillus niger cultures. Journal of Industrial Microbiology and Biotechnology 5, 259–263. Peixoto, S.C., Jorge, J.A., Terenzi, H.F. & Polizeli Mde, L. 2003 Rhizopus microsporus var rhizopodiformis: a thermotolerant fungus with potential for production of thermostable amylases. International Microbiology 4, 269–273. Pirt, S.J. 1975 Principles of Microbe and Cell Cultivation. Oxford: Blackwell Scientific. ISBN 0-63208150-3. Rajoka, M.I. 1990 Bioconversion of lignocellulosic materials raised from saline lands for production of biofuels using Cellulomonas species (as cellulolytic organisms). PhD thesis, University of the Punjab, Lahore, Pakistan. Rajoka, M.I., Bashir, A., Hussain, M.-R.A. & Malik, K.A. 1998a Mutagenesis of Cellulomonas biazotea for improved production of cellulases. Folia Microbiologica 43, 15–22. Rajoka, M.I., Bashir, A., Hussain, M.-R.A., Parvez, S., Ghauri, M.T. & Malik, K.A. 1998b Cloning and expression of bgl genes in Escherichia coli and Saccharomyces cerevisiae using shuttle vector pYES2.0. Folia Microbiologica 43, 129–135. Rajoka, M.I., Khan, S. & Shahid, R. 2003 Kinetics and regulation of the production of b-galactosidase from Kluyveromyces marxianus grown on different substrates. Food Technology and Biotechnology 41, 315–320. Rajoka, M.I. & Malik, K.A. 1997 Enhanced production of cellulases by Cellulomonas strains grown on different cellulosic residues. Folia Microbiologica 42, 59–64. Ramadas, M., Holst, O. & Mattiasson, B. 1996 Production of amyloglucosidase by Aspergillus niger under different cultivation regimens. World Journal of Microbiology and Biotechnology 12, 267–271. Siddiqui, K.S., Azar, M.J., Rashid, M.H. & Rajoka, M.I. 1997. Purification and the effect of Mn++ on the activity of carboxymethyl cellulase from Aspergillus niger and Cellulomonas biazotea. Folia Microbiologica 42, 303–311. Stamford, T.L., Stamfor, N.P., Coelho, L.C. & Araujo, J.M. 2002 Production and characterization of a thermostable glucoamylase from Streptosporangium sp. endophyte of maize leaves. Bioresource Technology 83, 105–109. Suto, M. & Tomita, F. 2001 Induction and catabolite repression mechanisms of cellulase in fungi. Journal of Bioscience and Bioengineering 92, 305–311. Thirumale, S., Rani, D.S. & Nand, K. 2001 Control of cellulase formation by trehalose in Clostridium papyrosolvens CFR-703. Process Biochemistry 37, 241–245. Van Wyk, J.P.H. 2001 Biotechnology and utilization of biowaste as a resource for bioproduct development. Trends in Biotechnology 19, 172–177. Verdoes, J.C., van Diepeningen, A.D., Punt, P.J., Debets, A.J.M., Stouthamer, A.H. & van den Hondel, C.A.M.J.J. 1994 Evolution of molecular and genetic approaches to generate glucoamylase overproducing strains of Aspergillus niger. Journal of Biotechnology 43, 195–205.

World Journal of Microbiology & Biotechnology 2005 21: 189–192 DOI: 10.1007/s11274-004-8321-4

Ó Springer 2005

Characterization of Mucor miehei lipase immobilized on polysiloxane-polyvinyl alcohol magnetic particles L.M. Bruno1,*, J.S. Coelho2, E.H.M. Melo3 and J.L. Lima-Filho3 1 Embrapa Tropical Agroindustry, CP 3761, 60511-110, Fortaleza, CE, Brazil 2 LIKA, Federal University of Pernambuco, Recife, PE, Brazil 3 Department of Biochemistry, Federal University of Pernambuco, Recife, PE, Brazil *Author for correspondence: Tel.: +55-85-2991800, Fax: +55-85-2991833, E-mail: [email protected] Received 29 April 2004

Keywords: Covalent immobilization, enzyme immobilization, hybrid material, lipase activity, sol–gel technique

Summary Mucor miehei lipase was immobilized on magnetic polysiloxane-polyvinyl alcohol particles by covalent binding. The resulting immobilized biocatalyst was recycled by seven assays, with a retained activity around 10% of its initial activity. Km and Vmax were respectively 228.3 lM and 36.1 U mg of protein)1 for immobilized enzyme. Whereas the optimum temperature remained the same for both soluble and immobilized lipase (45 °C), there was a shift in pH profiles after immobilization. Optimum pH for the immobilized lipase was 8.0. Immobilized enzyme showed to be more resistant than soluble lipase when assays were performed out of the optimum temperature or pH.

Introduction Although the basic function of lipases (glycerol ester hydrolases E.C. 3.1.1.3) is to hydrolyze triglycerides, they may also be used to catalyze many different reactions in vitro, such as synthesis of esters from glycerol and long-chain fatty acids (Jaeger & Reetz 1998). These make lipases an important group of biocatalysts, not only to modify oils and fats, but also for organic chemistry. Nevertheless the current market price of lipases is about one order of magnitude higher than the energy costs associated with the standard processes. Therefore, in order to reduce overall process costs, efficient methods for lipase immobilization are needed since immobilization allows enzyme to be reused, extending its useful active life (Balca˜o et al. 1996; Villeneuve et al. 2000; Soares et al. 2002). Several methods for lipase immobilization have been reported, such as deposition onto solid supports (Oliveira et al. 2000; Persson et al. 2000), covalent binding (Soares et al. 1999) and entrapment within a polymer matrix or hydrophobic sol–gel materials (Reetz et al. 1998; Keeling-Tucker et al. 2000). The latter method can be applied to a variety of lipases, yielding immobilized systems with 80-fold esterification activity as compared to the free enzyme (Reetz et al. 1996). The sol–gel process involves the transition of a system from a liquid ‘sol’ (mostly colloidal) into a solid ‘gel’ phase (Reetz et al. 1996). According to Ingersoll & Bright (1997) it is a simple three-step reaction sequence. In the first step, hydrolysis of a metal or semi-metal

alkoxide precursor forms the hydroxylated product and the corresponding alcohol. Next, condensation between an unhydrolyzed alkoxide group and a hydroxyl group or between two hydroxyls eliminates the solvent (water and alcohol) and forms a colloidal mixture called the sol. In the third and final step, polycondensation between these colloidal sols and additional networking eventually results in a porous, glasslike, three-dimensional network. Sol–gel technique has been employed especially to protein entrapment, since it is chemically inert, thermally stable and transparent, enabling spectroscopic monitoring of the entrapped sample (Dı´ az & Peinado 1997). However, the major disadvantage of this kind of immobilization is that it requires a careful control of pore size to avoid mass-transfer problems and leakage of the biological recognition element (Dı´ az & Peinado 1997; Faber 1997). Thus, covalent immobilization could be an attractive alternative to overcome such limitations. It has been reported that sol–gel technique is also an excellent method to prepare hybrid material. Silica glass beads were synthesized by alkoxide sol–gel process and were used as solid phase in ELISA for experimental plague studies (Barros et al. 2002; Coelho et al. 2003). Those authors used tetraethoxysilane (TEOS) and polyvinyl alcohol (PVA) for the matrix formation followed by activation with glutaraldeyde, to render more biocompatible surface for covalent immobilization (Barros et al. 2002; Bruno 2003). The resulting polysiloxane-polyvinyl alcohol (POS-PVA) composite

190 combined the PVA property to covalently retain proteins, via glutaraldehyde, with excellent optical, thermal and chemical stability of the host silicon oxide matrix. Moreover, the resulting matrix can be conjugated to magnetite (Fe3O4), allowing the separation of water insoluble derivative, that can be readily recovered by magnetic force without loss of enzymatic activity. This paper reports the use of a POS-PVA magnetic particles as support for Mucor miehei lipase immobilization, employing glutaraldehyde as bifunctional agent to promote covalent attachment. The kinetic characteristics of M. miehei lipase derivative are also described.

L.M. Bruno et al. Lipase assay The activity of both free and immobilized lipase were determined as described by Winkler & Stuckman (1979) with the following modifications: one milliliter of isopropanol containing 3 mg of p-nitrophenylpalmitate (pNPP) was mixed with 9 ml of 0.05 M Tris–HCl buffer (pH 8.0), containing 40 mg of Triton X-100 and 10 mg of arabic gum. Liberation of p-nitrophenol at 28 °C was detected at 410 nm (Hitachi spectrophotometer). One enzyme unit was defined as 1 lmol of p-nitrophenol enzymatically released from the substrate per minute. Protein measurement

Materials and methods Materials Mucor miehei lipase (4.79 ± 0.21 U mg of protein)1) was kindly donated by Novozymes (Auracaria, PR, Brazil). TEOS and ethanol were purchased from Merck (Darmstadt, Germany). Glutaraldehyde (25% [w/v]), HCl, ethanol, arabic gum and polyvinyl alcohol (MW 72000) were supplied by Reagen (Rio de Janeiro, RJ, Brazil). Triton X-100 and p-nitrophenylpalmitate (pNPP) were obtained from Sigma (St. Louis, MO, USA). POS-PVA synthesis A POS-PVA hybrid composite was prepared by the hydrolysis and polycondensation of TEOS as described by Barros et al. (2002). The reagents TEOS (5 ml), ethanol (5 ml), and PVA solution 2% (w/v) (6 ml) were carefully mixed and stirred for 5 min at 60 °C, followed by the addition of 2–3 drops of concentrated HCl, in order to catalyze the reaction. After an incubation period of 40 min, the material was transferred to microwells of tissue culture plates and kept at 25 °C until complete gel solidification (formation of the interpenetrated network of POS-PVA). Then, the spheres were ground in a ball mill to attain 37 lm diameter particles, which were magnetized according to Carneiro-Lea˜o et al. (1991), based on the coprecipitation from a solution of FeCl3
6H2O and FeCl2
4H2O. Activation of POS-PVA particles was carried out with 2.5% (w/v) glutaraldehyde at pH 7.0 for 1 h at room temperature, followed by exhaustive washings with distilled water. Lipase immobilization onto magnetic POS-PVA particles Magnetic POS-PVA particles were incubated with lipase solution containing 0.05 U mg of protein)1 in phosphate buffer (0.1 M, pH 7.0) under low stirring for 16 h at 4 °C. The immobilized lipase derivative was recovered by applying a magnetic field (6000 Oe). Afterwards, the derivative was washed with 1 M NaCl, and maintained in 50 mM Tris–HCl buffer (pH 8.0) at 4 °C.

Protein concentration was determined according to Lowry method (1951) using bovine serum albumin (BSA) as a standard. The amount of bound protein was determined indirectly by difference between the amount of protein introduced into the coupling reaction mixture and the amounts of protein in filtrate and in washing solutions.

Results and discussion Enzyme immobilization Table 1 shows the results of Mucor miehei lipase immobilization on POS-PVA hybrid composite. It can be observed that there were considerable losses of enzyme activity after its first use. Desorption keeps occurring until the last assay, although its intensity gradually decreases. This phenomenon was also observed by Carneiro da Cunha et al. (2002) working with Candida rugosa lipase covalently immobilized onto cellulose membranes. According to them, the loss of activity resulting from successive reuses reflects the amount of physically adsorbed lipase that is released into the assay emulsion. The retained activity for M. miehei lipase immobilized onto POS-PVA was almost the same found when M. miehei lipase was immobilized onto PVA (Bruno et al. 2004). The enzyme derivative produced using the first procedure, however, presented higher absolute activity after repetitive uses.

Table 1. Covalent immobilization of lipase onto POS-PVA No. of cycles

Specific activity (U mg of protein)1 g of support)

Retention of activity (%)

1st 2nd 3rd 4th 5th 6th 7th

22.5 11.1 4.0 4.2 3.1 2.7 2.5

100 49.3 17.8 18.7 13.8 12.0 11.1

191

Lipase immobilization on POS-PVA magnetic particles 120 Relative activity (%)

In order to improve retention of activity of this immobilized system, we suggest that future works involve, for instance, addition of stabilizing agents that protect the enzyme during immobilization steps. Macromolecular additives, such as proteins and polyethylene glycol have demonstrated stabilizing effects in the activity of the enzyme and are frequently employed (Soares et al. 2001). Effect of substrate concentration on the activity of soluble and immobilized lipase

Effect of temperature Figure 1 shows the effect of temperature on the lipase activity. It can be seen that, for both free and immobilized lipases, the optimum temperature was 45 °C. Immobilization, however, seems to protect the enzyme activity for temperatures over 45 °C, since POS-PVAlipase derivative presented higher relative activity than soluble lipase at 50, 57, and 65 °C. Literature relates changes on optimum temperature after immobilization. Montero et al. (1993) verified a shift of the optimum temperature from 37 to 45 °C, after immobilization of C. rugosa on polypropylene; and Fadiloglu & So¨ylemez (1998) also observed an alteration of optimum temperature to 45 °C for C. rugosa lipase immobilized on Celite. As each immobilized system has its own characteristics, changes in kinetics parameters can occur or not, being influenced by several variables, such as enzyme source, kind of support, immobilization method and enzyme-support interactions.

80 60 40 20 0 25

35

45 55 Temperature (oC)

65

75

Figure 1. Effect of temperature on the soluble (j) and immobilized ( ) Mucor miehei lipase. Activity was assayed with pNPP incubated from 30 to 65 °C at pH 8.0.

Effect of pH Determination of optimum pH for both soluble and immobilized lipase was performed by incubation of reaction mixture in a pH range of 5.0–9.0, employing citrate-phosphate buffer (pH 5.0–6.0, 50 mM), phosphate buffer (pH 7.0–8.0, 50 mM), and Tris–HCl buffer (pH 9.0, 50 mM), at 30 °C. Figure 2 shows the pH dependence of the enzyme activity for both free and immobilized lipase. Optimum pH values for soluble and immobilized enzyme were, respectively, 7.0 and 8.0.

120 Relative activity (%)

Table 2 shows estimates of Km and Vmax for free lipase and for immobilized lipase derivative, obtained in this work and by other authors: Montero et al. (1993) working with C. rugosa lipase immobilized on polypropylene and using p-nitrophenyl acetate as substrate and Nadruz et al. (1994) studying the immobilization of C. rugosa lipase, but employing alkylamine glass beads as support and pNPP as substrate. Km values can significantly vary from enzyme to enzyme and even for different substrates of the same enzyme, explaining the large range of Km values found in literature for immobilized lipases.

100

100 80 60 40 20 0 4

5

7 pH

6

9

8

10

Figure 2. Effect of pH on the soluble (j) and immobilized ( ) Mucor miehei lipase. Activity assays pH ranged from 5.0 to 9.0, employing pNPP as substrate at 30 °C.

Table 2. Estimated kinetic parameters (Km and Vmax) to free and immobilized enzymes Parameter

Free enzyme This work

Km (lM) 390.4 Vmax (U mg of protein)1) 55.2 a

Immobilized lipase Montero et al. (1993)b

Nadruz et al. (1994)c

This work

Montero et al. (1993)b

Nadruz et al. (1994)c

2700 39.2

– –

228.3 36.1

2600 9.7

88 615

Lipase immobilized on POS-PVA and using pNPP as substrate. Assays were performed at 30 °C and pH 8.0, employing pNPP solutions ranging from 100 to 1000 lM. b C. rugosa lipase immobilized on polypropylene and using p-nitrophenyl acetate as substrate. c C. rugosa lipase immobilized on alkylamine glass beads and using pNPP as substrate.

192 However it can be noticed that at pH values of 5.0, 6.0, and 9.0 lipase-POS-PVA relative activity was higher than that of soluble enzyme, suggesting that immobilization also seemed to confer some kind of protection to enzyme when the reaction media presented a pH value different from the optimum. Fadiloglu & So¨ylemez (1998) observed a reduction of the optimum pH from 7.0 to 6.5 for C. rugosa lipase after immobilization on Celite, whereas Nadruz et al. (1994), working with the same microbial lipase immobilized on alkylamine beads, verified an increase on optimum pH from 7.2 to 8.0. Nadruz et al. (1994) also related that the immobilized lipase demonstrated a wider pH activity profile than the free enzyme, what they assumed to be due to stabilization of lipase by covalent coupling.

Conclusion The results obtained pointed the use of magnetic polysiloxane polyvinyl alcohol particles as an attractive support for enzyme immobilization, specially lipases. Immobilization on POS-PVA allowed repetitive use of lipase derivative (seven cycles) and facility to separate enzyme system from reaction media. When assays were performed out of the optimum operational conditions, immobilized enzyme showed higher activities than free lipase. Moreover easiness of support synthesis and immobilization procedure, justify the improvement of the process to achieve better results and find other applications for this system, as for ester synthesis.

Acknowledgements The authors would like to thank the Brazilian researchfunding agency Conselho Nacional de Desenvolvimento Tecnolo´gico (CNPq) for financial support.

References Balca˜o, V.M., Paiva, A.L. & Malcata, F.X. 1996 Bioreactors with immobilized lipases: state of the art. Enzyme and Microbial. Technology 18, 392–416. Barros, A.L., Almeida, A.M.P., Carvalho, Jr., L.B. & Azevedo, W.M. 2002 Polysiloxane/PVA-glutaraldehyde hybrid composite as solid phase for immunodetections by ELISA. Brazilian Journal of Medical and Biological Research 35, 459–463. Bruno, L.M. 2003 Desenvolvimento de sistemas enzima´ticos (lipases) para aplicac¸a˜o na hidro´lise e sı´ntese de e´steres PhD thesis, Universidade Federal de Pernambuco, PE, Brazil. Bruno, L.M., Pinto, G.A.S., de Castro, H.F., Lima-Filho, J.F. & Melo, E.H.M. 2004 Variables that affect immobilization of Mucor miehei lipase on nylon membrane. World Journal of Microbiology and Biotechnology 20, 371–375. Carneiro da Cunha, M.G.C., Rocha, J.M.S., Cabral, J.M.S., Gil, M.H. & Garcia, F.A.P. 2002 Covalent immobilisation of lipase on different supports. Latin American Applied Research 32, 69–72.

L.M. Bruno et al. Carneiro-Lea˜o, A.M.A., Oliveira, E.A. & Carvalho, Jr., L.B. 1991 Immobilization of protein on ferromagnetic dacron. Applied Biochemistry and Biotechnology 31, 53–58. Coelho, R.A.L., Yamasaki, H., Perez, E. & Carvalho, Jr., L.B. 2003 The use of polysiloxane/polyvinyl alcohol beads as solid phase in IgG anti-toxocara canis detection using a recombinant antigen. Memo´rias do Instituto Oswaldo Cruz 98, 391–393. Dı´ az, A.N. & Peinado, M.C.R. 1997 Sol–gel cholinesterase biosensor for organophosphorus pesticide fluorimetric analysis. Sensors and Actuators B 38–39, 426–431. Faber, K. 1997 Biotransformations in Organic Chemistry: A Textbook. pp. 345–357. Berlin: Springer-Verlag. ISBN 3-540-61688-8. Fadiloglu, S. & So¨ylemez, Z. 1998 Olive oil hydrolysis by celiteimmobilized Candida rugosa lipase. Journal of Agricultural and Food Chemistry 8, 3411–3414. Ingersoll, C.M. & Bright, F.V. 1997 Using sol–gel-based platforms for chemical sensors. Chemtech 27, 26–31. Jaeger, K.E. & Reetz, M.T. 1998 Microbial lipases form versatile tools for biotechnology. Trends in Biotechnology 16, 396–403. Keeling-Tucker, T., Rakic, M., Spong, C. & Brennan, J.D. 2000 Controlling the material properties and biological activity of lipase within sol–gel derived bioglasses via organosilane and polymer doping. Chemistry of Materials 12, 3695–3704. Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J. 1951 Protein measurement with the folin phenol reagent. Journal of Biological Chemistry 193, 265–275. Montero, S., Blanco, A., Virto, M.D., Landeta, L.C., Agude, I., Solozabal, R., Lascaray, J.M., Renobales, M., Llama, M.J. & Serra, J.L. 1993 Enzyme and Microbial Technology 15, 239–247. Nadruz, W., Lea˜o, I.C., Krieger, N., Pimentel, M.C.B., Ledingham, W.M., Melo, E.H.M., Lima-Filho, J.L. & Kennedy, J.F. 1994 Characterisation of Candida rugosa lipase immobilised on alkylamine glass beads. The Genetic Engineer and Biotechnologist 14, 143–148. Oliveira, P.C., Alves, G.M. & Castro, H.F. 2000 Sı´ ntese do butirato de n-butila empregando lipase microbiana imobilizada em copolı´ mero de estireno-divinilbenzeno. Quı´mica Nova 23, 632–636. Persson, M., Wehtje, E. & Adlercreutz, P. 2000 Immobilisation of lipases by adsorption and deposition: high protein loading gives lower water activity optimum. Biotechnology Letters 22, 1571–1575. Reetz, M.T., Zonta, A. & Simpelkamp, J. 1996 Efficient immobilization of lipases by entrapment in hydrophobic sol–gel materials Biotechnology and Bioengineering 49, 527–534. Reetz, M.T., Zonta, A., Vijayakrishnan, V. & Schimossek, K. 1998 Entrapment of lipases in hydrophobic magnetite-containing sol– gel materials: magnetic separation of heterogeneous biocatalysts. Journal of Molecular Catalysis A: Chemical 134, 251–258. Soares, C.M.F., Castro, H.F., Moraes, F.F. & Zanin, G.M. 1999 Characterization and utilization of Candida rugosa lipase immobilized on controlled pore silica. Applied Biochemistry and Biotechnology 77/79, 745–757. Soares, C.M.F., Castro, H.F., Santana, M.H.A. & Zanin, G.M. 2001 Selection of stabilizing additive for lipase immobilization on controlled pore silica by factorial design. Applied Biochemistry and Biotechnology 91/93, 703–718. Soares, C.M.F., Castro, H.F., Santana, M.H.A. & Zanin, G.M. 2002 Intensification of lipase performance for long-term operation by immobilization on controlled pore silica in presence of polyethylene glycol. Applied Biochemistry and Biotechnology 98/100, 863–874. Villeneuve, P., Muderhwa, J.M., Graille, J. & Haas, M.J. 2000 Customizing lipases for biocatalysis: a survey of chemical, physical and molecular biological approaches. Journal of Molecular Catalysis B: Enzymatic 9, 113–148. Winkler, U.K. & Stuckmann, M. 1979 Glycogen, hyaluronate, and some other polysaccharides greatly enhance the formation of exolipase by Serratia marcescens. Journal of Bacteriology 138, 663– 670.

World Journal of Microbiology & Biotechnology 2005 21: 193–199 DOI: 10.1007/s11274-004-3108-1

Ó Springer 2005

Lipase-catalyzed naproxen methyl ester hydrolysis in water-saturated ionic liquid: significantly enhanced enantioselectivity and stability Jia-Ying Xin1,2,*, Yong-Jie Zhao2, Yan-Guo Shi1, Chun-Gu Xia2 and Shu-Ben Li2 1 College of Food Science and Pharmacy, Harbin University of Commerce, No. 138 Tongda Street, Daoli District, Harbin 150076, China 2 State Key Laboratory for Oxo Synthesis & Selective Oxidation, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, China *Author for correspondence: Fax: +86-45184838194, E-mail: [email protected] Received 14 May 2004

Keywords: 1-Butyl-3-methylimidazolium hexafluoro-phoshate, Candida rugosa lipase, enantioselectivity, ionic liquid, Naproxen, Naproxen methyl ester, selective hydrolysis

Summary The lipase selective hydrolysis of Naproxen methyl ester was explored in both water-saturated isooctane and watersaturated ionic liquid 1-butyl-3-methylimidazolium hexafluoro-phoshate ([bmim]PF6) to see any significant differences in terms of enantioselectivity and stability between two different classes of reaction media. It is shown that polar and hydrophobic of [bmim]PF6 made it an unearthly reaction medium for hydrolysis of Naproxen methyl ester. It not only decreases the equilibrium constant (K) and enhances the enantiomeric ratio (E ), consequently improves the equilibrium conversion (CEq) of the hydrolysis reaction and enantiomeric excess of product (eep), but also maintains the lipase activity. Because the lipase would not dissolve in the 1-butyl-3-methylimidazolium hexafluoro-phoshate, it can be filtrated up from 1-butyl-3-methylimidazolium hexafluoro-phoshate and recycled for several runs. The stability of lipase was improved due to the higher solubility of methanol in 1-butyl-3methylimidazolium hexafluoro-phoshate than in isooctane.

Introduction (S )-(+)-2-(6-Methoxy–2-naphthyl) propionic acid (Naproxen) is a nonsteroidal antiinflammatory drug which belongs to the family of 2-aryl propionic acid derivatives, it is widely used as a drug for human connective tissue diseases. The physiological activity of the S-form Naproxen is 28-fold that of the R-form (Pandey et al. 1999). Hence, only the S-form is used as a drug for humans. Naproxen can be resolved by chemical means, enzymatic means, and asymmetric synthesis. Among these methods, the application of enzyme in the resolution of racemate is becoming increasingly prominent. This is obviously due to the high stereoselectivity of the enzymes. Enzymatic resolution of Naproxen can be performed by hydrolysis (Battistel et al. 1991; Chang et al. 1999) or esterification (Tsai et al. 1996; Duan et al. 1997). From the standpoint of high productivity, easy product separation, and few reaction steps, lipase-catalyzed asymmetric hydrolysis has been judged to be superior for the practical resolution of racemic Naproxen (Cambou & Klibanov 1984). For enzymatic hydrolysis, the resolution of Naproxen can be performed in aqueous

system or in aqueous–organic biphasic system. In the former system, the relatively low solubility of substrate causes a low reaction rate. And the separation of substrate and product is also difficult (Kise & Tomiuchi 1991; Caron & Kazlauskas 1993; Akita et al. 1998). In the later system, reactor design for the use of biocatalysts in biphase system presents several complications compared with simple aqueous or organic solvent system. Agitation conditions may affect the interfacial area between the two phases, by which the reaction rate may be affected. On the other hand, the formation of an emulsion in an aqueous–organic biphase system may complicate work-up (Halling 1987). Another possibility is to perform the hydrolysis reaction in water-saturated organic solvent (Kise & Tomiuchi 1991; Akita et al. 1998), provided that the enzyme remains active and enantioselective. Substrate solubility may be high, and enzyme may be easily recovered and recycled, the reaction system is very suitable for water-sensitive substrate. However, three problems may be hampering the achievement of the desired extent of conversion and enantiomeric excess: Firstly, the hydrolysis reaction in water-saturated organic solvent often suffer from slow reaction rates and low yields, which are caused by the very low water

194 capacity of most conventional nonpolar organic solvents (Rakels et al. 1994). Secondly, one of the hydrolysis products of Naproxen methyl ester, methanol, may inhibit or inactivate the lipase that interconverts it. The inhibition of lipase by methanol due to the low solubility of methanol in the most conventional nonpolar organic solvents becomes significant. In other words, methanol, at a relatively low yield in aqueous-saturated organic solvent system, will severely inhibit lipase activity (Xu et al. 2001). Thirdly, another product of the Naproxen methyl ester hydrolysis, Naproxen, may be precipitated in aqueous-saturated organic solvent system due to the very low solubility of Naproxen in the most conventional nonpolar organic solvents. The precipitation of Naproxen would blend with organic solvent-immiscible lipase particles and make its recovery uneasy and cumbersome (Xin et al. 2000). Ionic liquids may be particularly attractive as a new type of alternative media for the enzymatic hydrolysis reactions of Naproxen methyl ester because their polarity may enhance the capacity of water and the solubility of Naproxen and methanol. Although more polar organic solvents inactivate enzymes, surprisingly ionic liquids do not (Susheel et al. 2003). In this paper, a water-saturated ionic liquid system was developed to overcome the aforementioned problem. The lipasecatalyzed reaction in both water-saturated isooctane and water-saturated ionic liquids 1-butyl-3-methylimidazolium hexafluoro-phoshate [bmim]PF6 (Scheme 1) were comparatively studied with a special attention to the capacity of water and the activity maintaining of lipase. The goal is to enhance the enantioselectivity and stability of lipase-catalyzed naproxen ester hydrolysis.

Materials and methods Materials 1-Butyl-3-methylimidazoliumhexafluoro-phoshate ([bmim]PF6) (Figure 1) was synthesized as described by Jairton et al. (Huddleston et al. 1998; Jairton et al. 2002). Lipase from Candida rugosa (780 U/mg) was purchased from Sigma Chemical Co. (S)-Naproxen was purchased from international pharmaceutical factory (Shanghai, China). All other reagent used was obtained from commercial suppliers and were of reagent grade.

J.-Y. Xin et al. N

N PF6

Figure 1. Room-temperature ionic liquid 1-butyl-3-methylimidazolium hexafluoro-phoshate ([bmim]PF6).

Synthesis of ester The methyl ester of (R,S)-Naproxen was prepared by the classical methodology using thionyl chloride and methanol. Thionyl chloride, 15 ml (0.2 mol), was added dropwise to a cooled, stirred suspension of Naproxen (0.12 mol) in methanol (250 ml). The reaction mixture was refluxed for 2.5 h, and then the solvent was evaporated and the residue purified by column chromatography using SiO2 as adsorbent and isooctane as eluant. Enzymatic hydrolysis of (R,S)-Naproxen methyl ester in water-saturated ionic liquid and isooctane Experiment was performed in a 50-ml shaken bottle containing 25 ml of isooctane or ionic liquid (the concentration of Naproxen methyl ester was 10 mg/ ml) presaturated with 0.2 mol/l phosphate buffer (pH 7.0), and the reaction had been started by adding 200 mg of lipase. The reaction mixture was stirred with a magnetic stirrer at 30 °C. In both cases, samples were withdrawn at specified time intervals for measurement of the conversion (C) and enantiomeric excess of the reaction. Analytical procedure The enantiomeric excess value of Naproxen was determined by HPLC (HP1090) by using a chiral column (Chirex R-NGLY & DNB, phenomenex) capable of separating the R and S-isomers of Naproxen, and the mobile phase was methanol solution (0.03 mol/l ammonium acetate), at a flow rate of 1.0 ml/min. The enantiomeric excess value of Naproxen methyl ester was determined by HPLC (HP1090) with a chiral column (coated with cellulose tris (3,5-dimethylphenyl carbamate)) capable of separating the R- and S-isomers of Naproxen ester, and the mobile phase was a volume/ volume mixture of 99.5% n-hexane 0.5% 2-propanol, at a flow rate of 0.5 ml/min. UV detection at 254 nm was used for quantification at the 25 °C. The HPLC analyses

Scheme 1. Lipase catalyzed hydrolysis of Naproxen methyl ester in water-saturated isooctane or water-saturated ionic liquid [bmim]PF6.

195

Lipase-catalyzed naproxen methyl ester hydrolysis were standardized by an external standard method with standard Naproxen (Aldrich) and Naproxen methyl ester prepared as described in the paper. Standard curves were obtained from peak areas of (S)- and (R)- of racemic Naproxen or Naproxen methyl ester. The experimental data were obtained from four independent repeated experiments. The content of methanol was analyzed by gas chromatography; gas chromatography was performed in a GC-204 gas chromatograph equipped with a flame ionization detector (FID), and a PorparkQ column (3 mm 32,000 mm). Injector temperature was 90 °C and the detector temperature was 176 °C, and the carrier gas was nitrogen. A column temperature was 90 °C and a N2 stream was 7.0 ml/min. Calculations The equilibrium constant (K) is defined by using the equation reported by Chen et al. (1987) for reversible enantioselective reactions. K¼

½ðRÞ-Naproxen methyl ester
Eq ½ðRÞNaproxen
Eq

½ðSÞ-Naproxen methyl ester
Eq ¼ ½ðSÞNaproxen
Eq

ð1Þ

K was determined by allowing the reaction to proceed until the fast-reacting (S)-isomer attained equilibrium. The value of K could be estimated from the concentrations of the remaining substrate and product of the fastreacting (S)-isomer. From the experimentally determined data of eeS, eeP, C (the extent of conversion, C ¼ eeS/ (eeS + eeP)) and K, values for the enantiomeric ratio (E ), (support A and B are the fast- and slow-reacting enantiomers that compete for the same site on the enzyme, E is definitively by (2) (Chen et al. 1987): InðA=A0 Þ InðB=B0 Þ In½1  ð1 þ KÞðC þ eeS f1  CgÞ
¼ In½1  ð1 þ KÞðC  eeS f1  CgÞ
In½1  ð1 þ KÞCð1 þ eep Þ
¼ In½1  ð1 þ KÞCð1  eep Þ


E¼

ð2Þ

and the equilibrium conversion (CEq, CEq is the conversion as the equilibrium of fast- and slow-reacting enantiomers all is established, at the moment, eeP ¼ eeS ¼ 0%) were obtained by using the nonlinear regression according to the equation reported by Chen (Chen et al. 1987) for reversible enantioselective reaction obeying uni–uni kinetics.

Results and discussion According to Chen (Chen et al. 1987), the value of K is the concentration ratio of the remaining substrate and

product of fast-reacting isomer (or slow-reacting isomer) at the equilibrium states, it is a thermodynamic functions whose value is independent of the properties of enzyme. The value of E is a kinetic parameter whose value will be various with different catalysts and reaction medium. The enantiospecificity of enzymecatalyzed asymmetric hydrolysis depends on the complex interaction of kinetic and thermodynamic functions. K and E are sensitive to water content of the medium. For the reversible system, at the initial stages of reaction, the lipase preferentially attacks the (S)-enantiomer and transforms into (S)-Napeoxen. When the fast-reaction isomer approaches equilibrium, the enantiomeric excess of substrate (eeS) and the product (eeP) fractions begin to fall due to concentration changes of the slow-reacting (R)-enantiomer. So eeS and eeP are not only dependent on the extent of conversion (C) and enantiomeric ratio (E ), but also dependent on the equilibrium constant (K). To correlate the value of eeS and eeP with the extent of conversion (C), enantomeric ratio (E ) and the equilibrium constant (K), the theoretical curves (Figure 2) were generated by computer according (2) and provide a useful overview of the interrelationships between various C, eeS, and eeP for fixed values of E and K. It is evident that the optical purity of product and remaining substrate is related to the K and E. From the standpoint of lipase activity maintaining, the solvent isooctane works well as a reaction medium for the hydrolysis of Naproxen (Xin et al. 2000). Unfortunately the very low water capacity of isooctane hampers the achievement of the desired extent of conversion and enantiomeric excess. As shown in Figure 3a, because of the very low water capacity of isooctane and consequently unfavorable reaction equilibrium, the lipase catalyzed enantioselective hydrolysis of Naproxen methyl ester in water-saturated isooctane leads to low values of eeS and eeP at a particular conversion. Also, the effect of the water capacity of the reaction medium on the progress of a racemate conversion is shown in Figure 4. The very low water capacity of isooctane leads to a sharp bending of the conversion value at very low value of conversion. More polar solvents such as toluene, cyclohexane may be employed to overcome the low water capacity problem, however, at the price of decreasing enzyme activity and stability (Xin et al. 2000). Increasing the solvent polar resulted in a decrease in enzyme activity. It may be impossible to find a solvent that both maintains lipase activity and have higher water capacity. It had been reported that ionic liquids with highly hydrophilic in nature would also likely inactivate enzyme. However, the water-immiscible ionic liquid [bmim]PF6 (1-butyl-3-methylimidazolium hexafluorophosphate) (Joel et al. 2003), in particular, despite being polar due to its ionic nature, is hydrophobic and can maintain the activity of enzyme. In some cases, its use enhanced the enzyme activity and selectivity (Kim et al.

196

J.-Y. Xin et al. Enantiomeric excess of substrate (%)

c 80

60

a

b

E=5

c

60 40 40 20 20

b a

0

0 0

20

40

60

80

Enantiomeric excess of product (%)

80

100

100

Conversion (%)

Enantiomeric excess of substrate (%)

c 80

80

a

b

E=100

c

60

60

40

40

b

20

20

a 0

0 0

20

40

60

80

Enantiomeric excess of product (%)

100

100

100

Conversion (%)

Figure 2. Plot of percent enantiomeric excess as a function of the percent conversion for various enantiomeric (E ) and equilibrium constant (K). The curves were computer generated from (2). The value of K were (a) 10, (b) 1 and (c) 0.

2003). Also, it is worth noting that [bmim]PF6 can dissolve up to 5% of water (v/v) (Erbeldinger et al. 2000). This feature of [bmim]PF6 made it an unearthly reaction medium for hydrolysis of Naproxen methyl ester. It is not only favorable to the equilibrium yield of the hydrolysis reaction, but also maintains the lipase activity. As shown in Table 1, compare with the result of hydrolysis reaction in water-saturated isooctane, in water-saturated BMIMPF6, the values of K for the hydrolysis reaction was decreased from 12 to 0.7 and the values of E for the hydrolysis reaction was increased from 88 to 356, respectively. This also increased the equilibrium conversion (CEq) of the hydrolysis reaction; as a consequence, a higher eeS and eeP were obtained at a particular conversion (Figure 3b). The enantioselectivity was improved strikingly. At the same time, the increasing values of K and E also lead to the sharp bending of the conversion curve at higher value of conversion (Figure 4). In the enantioselective hydrolysis reaction of Naproxen methyl ester, the goal compound is the (S)-Naproxen, the product of the steroselective hydrolysis reaction catalyzed by lipase. As shown in Figure 3, it was clear that the enantiomeric excess of Naproxen (eeP) is higher in BMIMPF6 than in isooctane at comparable conversion (C). To obtain similar eeP, the conversion must be lower in isooctane than in BMIMPF6. The results demonstrated that the enzymatic hydrolysis reaction of Naproxen methyl ester in water-saturated ionic liquid

BMIMPF6 provided higher yield and enantiomeric excess than those in water-saturated isooctane. The results indicate that ionic liquid BMIMPF6 is particularly useful as the media for enzymatic hydrolysis reaction of Naproxen methyl ester that is difficult to achieve the desired extent of conversion and enantiomeric excess in conventional water-saturated organic solvents. In the water-saturated isooctane system, the catalytically active lipase does not dissolve in the organic solvent, but remained suspended as a powder. At the same time, the solubility of polar product (methanol and Naproxen) is low (Xin et al. 2000). Due to most of methanol is adsorbed by the polar solid lipase particle, methanol at a relatively low yield will severely inhibit lipase activity (Xu et al. 2001). Also, Naproxen precipitated in aqueous-saturated organic solvent system will blend with lipase particles and make its recovery uneasy and cumbersome. From Figure 5, it is clear that the activity of the lipase is dropped partially in watersaturated isooctane. This drop in reactivity may be caused by the inhibitory action of methanol which was adsorbed by the polar solid lipase particle. As comparing, the same reaction was also performed in water-saturated [bmim]PF6. As a reaction medium, ionic liquids exhibit excellent physical characteristics including the ability to dissolve polar and nonpolar organic or inorganic compound (Itoh et al. 2003). So in the water-saturated [bmim]PF6, the solubility of polar product (methanol and Naproxen) may be higher than

197

Lipase-catalyzed naproxen methyl ester hydrolysis

Figure 3. Comparison of the lipase-catalyzed enantiospecific hydrolysis of Naproxen methyl ester in water-saturated isooctane (a) and watersaturated ionic liquid [bmimPF6] (b). The curves depicting the relationship between ee of substrate or product and conversion were computergenerated by using the non-linear regress from the equation reported by Chen (Chen et al. 1987) for reversible enantioselective reaction obeying uni–uni kinetics. The symbols were the experimental data. The reaction was started by adding 200 mg of lipase (780 U/mg).

40

2

35

Conversion (%)

30 25 20 15

1

10 5 0 0

100

200

300

400

500

600

700

800

Reaction time (h)

Figure 4. The progress curve of the lipase-catalyzed reaction in water-saturated isooctane (1) and water-saturated ionic liquid [bmimPF6] (2).

that in water-saturated isooctane. Also, it is worth noting that the limited soluble water (up to 5% (v/v) of water was dissolved in [bmim]PF6) may greatly enhance the solubility of methanol and Naproxen. Hence, the water-saturated [bmim]PF6 can supply substrate (Naproxen and H2O) and recover products (methanol and Naproxen). After completion of the reaction, the lipase

particles was filtered off and then added into fresh water-saturated BMIMPF6 medium containing Naproxen methyl ester to catalyze the second run reaction. No substantial precipitation of Naproxen and Naproxen methyl ester was found to blend with solid lipase particles by HPLC analysis. Also, no methanol was determined in solid lipase particles by GC. The lipase-

198

J.-Y. Xin et al.

Table 1. Effect of reaction medium on the equilibrium constant (K ) and enantiomeric ratio (E ) and the equilibrium conversion (CEq) in hydrolysis of Naproxen methyl ester by lipase. Reaction medium

K

E

CEq (%)

Water-saturated isooctane Water-saturated [bmimPF6]

12.0 0.7

88 356

7.6 58.8

References

330

Relative activity (%)

Water-saturated isooctane Water-saturated [bmimPF6] 300

90

80 0

3

2

3

4

Major State Basic Research of China and the Royal Society of UK for the financial supporting.

5

run

Figure 5. The relative activity of lipase. Repetitive lipase-catalyzed reactions were performed in water-saturated isooctane and watersaturated ionic liquid [bmimPF6] at 30 °C for 72 h, respectively. The first run was carried out for 72 h with fresh lipase. The value of conversion 30.6% (in water-saturated ionic liquid [bmimPF6]) and 3.8% (in water-saturated isooctane) was used as the 100%. The second to forth run were done with recycled lipase. The conversion percent in each run was compared with that in the first reaction to obtain the relative activity.

catalyzed reactions were repeated four times with the recycling of enzyme. The result showed that it was possible to use the enzyme repeatedly in this system.

Conclusion [bmim]PF6 is polar and hydrophobic, it can dissolve up to 5% of water (v/v) (Erbeldinger et al. 2000). This feature of [bmim]PF6 made it an unearthly reaction medium for hydrolysis of Naproxen methyl ester. In this paper, the lipase selective hydrolysis of Naproxen methyl ester was explored in both water-saturated isooctane and water-saturated [bmim]PF6 to see any significant differences in terms of enantioselectivity and reactivity. It is shown that BMIMPF6 not only decreases the equilibrium constant (K) and enhances the enantiomeric ratio (E ), consequently improves the equilibrium conversion (CEq) of the hydrolysis reaction and enantiomeric excess of product (eep), but also maintains the lipase activity. The enhancement of stability of lipase was attributed to the higher solubility of methanol in [bmim]PF6 than in isooctane. Acknowledgements The authors would like to thank the National Nature Science Foundation of Chinese, the Special Funds for

Akita, H., Enoki, Y. & Yamada, H. 1998 Enzymatic hydrolysis in organic solvents for kinetic resolution of water-insoluble a-acyloxy esters with immobilized lipases. Chemical and Pharmaceutical Bulletin 37, 2876–2878. Battistel, E., Bianchi, D., Cesti, P. & Pina, C. 1991 Enzymatic resolution of (S)-(+)-Naproxen in a continuous reactor. Biotechnology and Bioengineering 38, 659–664. Cambou, B. & Klibanov, A.M. 1984 Comparison of different strategies for the lipase-catalyzed resolution of racemic acid and alcohols: asymmetric hydrolysis, esterification, and transesterification. Biotechnology and Bioengineering 26, 1449–1454. Caron, G. & Kazlauskas, R.J. 1993 Sequential resolution of (2,3)butanediol in organic solvent using lipase. Tetrahedron Asymmetry 4, 1995–2000. Chang, C.S., Tsai, S.W. & Jimmy, K. 1999 Lipase-catalyzed dynamic resolution of Naproxen 2,2,2-trifluoroethyl thioester by hydrolysis in isooctane. Biotechnology and Bioengineering 64, 120– 126. Chen, C.S., Wu, S.H., Girdauks, G. & Sih, C.J. 1987 Quantitative analyses of biochemical kinetic resolution of enantiomers. 2. Enzyme-catalyzed esterifications in water-organic solvent biphase system. Journal of the American Chemical Society 109, 2812– 2817. Duan, G., Ching, C.B., Lim, E. & Ang, C.H. 1997 Kinetic study of enantioselective esterification of ketoprofen with n-propanol catalysed by an lipase in an organic medium. Biotechnology Letters 19, 1051–1055. Erbeldinger, M., Mesiano, A.J. & Russell, A.J. 2000 Enzymatic catalysis of formation of Z-aspartame in ionic liquid – an alternative to enzymatic catalysis in organic solvents. Biotechnology Progress 16, 1129–1131. Halling, P.J. 1987 Biocatalysis in multi-phase reaction mixtures containing organic liquids. Biotechnology Advances 5, 47–84. Huddleston, J.G., Willauer, H.D, Swatloski, R.P, Visser, A.E. & Rogers, R.D. 1998 Room temperature ionic liquids as novel media for ‘Clean’ liquid–liquid extraction. Chemical Communication 1765–1766. Itoh, T., Nishimura, Y., Ouchi, N. & Hayase, S. 2003 1-Butyl-2,3dimethylimidazolium tetrafluoroborate: the most desirable ionic liquid solvent for recycling use of enzyme in lipase-catalyzed transesterification using vinyl acetate as acyl donor. Journal of Molecular Catalysis B: Enzymatic 26, 41–45. Jairton, D., Roberto, F.D.S. & Paulo, A.Z.S. 2002 Ionic liquid (molten salt) phase organometallic catalysis. Chemical Review 102, 3667– 3692. Joel, L.K., Anita, M.J., Jason, A.B., Roger, M. & Alan J.R. 2003 Impact of ionic liquid physical properties on lipase activity and stability. Journal of the American Chemical Society 125, 4125– 4131. Kim, M.-J., Choi, M.Y., Lee, J.K. & Ahn, Y. 2003 Enzymatic selective acylation of glycosides in ionic liquids: significantly enhanced reactivity and regioselectivity. Journal of Molecular Catalysis B: Enzymatic 26, 115–118. Kise, H. & Tomiuchi, Y. 1991 Unusual solvent effect on protease activity and effective optical resolution of amino acids by hydrolytic reactions in organic solvents. Biotechnology Letters 13, 317– 322. Pandey, A., Benjamin, S. & Soccol, C.R. 1999 The realm of microbial lipase in biotechnology. Biotechnology and Applied Biochemistry 29, 119–131. Rakels, J.L.L., Straathof, A.J.J. & Heijnen, J.J. 1994 Improvement of enantioselective enzymatic ester hydrolysis in organic solvent. Tetrahedron Asymmetry 5, 93–100.

Lipase-catalyzed naproxen methyl ester hydrolysis Susheel, J.N., Jitendra, R.H. & Manikrao, M.S. 2003 Lipase-catalysed polyester synthesis in 1-butyl-3-methylimidazolium hexafluorophosphate ionic liquid, Tetrahedron Letters 44, 1371–1373. Tsai, S.W., Liu, B.Y. & Chang, C.S. 1996 Enhancement of (S)Naproxen ester productivity from racemic Naproxen by lipase in organic solvents. Journal of Chemical Technology and Biotechnology 65, 156–162.

199 Xin, J.Y., Li, S.B., Xu, Y. & Wang, L.L. 2000 Enzymatic resolution of S-(+)-Naproxen in a trapped aqueous–organic solvent biphase continuous reactor. Biotechnology and Bioengineering 68, 78– 83. Xu, Y., Chen, J.B., Xin, J.Y. & Li, S.B. 2001 Efficient microbial elimination of methanol inhibition for Naproxen resolution by a lipase. Biotechnology Letters 23, 1975–1979.

World Journal of Microbiology & Biotechnology 2005 21: 201–206 DOI: 10.1007/s11274-004-3318-6

 Springer 2005

Antimicrobial screening and active compound isolation from marine bacterium NJ6-3-1 associated with the sponge Hymeniacidon perleve Li Zheng1, Haimin Chen1, Xiaotian Han1, Wei Lin1 and Xiaojun Yan2,* 1 Institute of Oceanology, Chinese Academy of Sciences, Graduate School of The Chinese Academy of Science, Qingdao 266071, P.R. China 2 Marine Biotechnology Laboratory, Post Box 71 Ningbo University, Ningbo 315211, P.R. China *Author for correspondence: Tel.: +86-574-87600458, Fax: +86-574-87600590, E-mail: [email protected] Received 27 May 2004

Keywords: Antimicrobial activity, beta-carboline, marine bacteria, sponge Summary Twenty-nine marine bacterial strains were isolated from the sponge Hymeniacidon perleve at Nanji island, and antimicrobial screening showed that eight strains inhibited the growth of terrestrial microorganisms. The strain NJ6-3-1 with wide antimicrobial spectrum was identified as Pseudoalteromonas piscicida based on its 16S rRNA sequence analysis. The major antimicrobial metabolite, isolated through bioassay-guide fractionation of TLC bioautography overlay assay, was identified as norharman (a beta-carboline alkaloid) by EI-MS and NMR. Introduction Marine invertebrates have developed highly specific relationships with numerous associated microorganisms and these associations are of recognized ecological and biological importance (Sponga et al. 1999; Armstrong 2001; Strahl et al. 2002). It has been reported that the ratio of microorganisms with antimicrobial activity from invertebrates was higher than from other sources (Ivanova et al. 1998; Burgess et al. 1999), which suggests that invertebrate-associated microorganisms might play a chemical defence role for their hosts. This kind of microorganism as a sustainable resource has a high potential to biosynthesize novel biologically active secondary metabolites. Sponges are primitive marine invertebrates and contain more natural products than any other marine phylum. Many of their products have strong bioactivities including anticancer, antimicrobial and anti-inflammatory activities, and are often applicable for medical use (Faulkner 2002; Pettit et al. 2004; Sonnenschein et al. 2004). Unlike other invertebrates, sponges harbour extraneous microorganisms on their surface, in their canal systems and in the intercellular matrix, which constitute a large part of the body, as much as up to 40% of the total biomass (Wilkinson 1978; Santavy et al. 1990). Some evidence indicates that many compounds previously found in sponges are biosynthesized through microorganisms associated with them or indeed produced by microorganisms (Unson & Faulkner 1993; Unson et al. 1994; Bewley & Faulkner 1998). To confirm this hypothesis, there has been a great deal of interest in isolating microorganisms with bioactivities from

sponges, and in recent years, a number of novel compounds with such activities have been discovered through cultivation of sponge-associated microorganisms (Jayatilake 1996; Mitova et al. 2003; Suzumura et al. 2003). With the aim of finding new bioactive compounds from marine microorganisms and to investigate the real origin of natural products, we started a programme to isolate bacteria with antimicrobial activity associated with the sponge Hymeniacidon perleve and further to elucidate their active metabolites.

Material and methods Isolation sponge-associated bacteria The sponge Hymeniacidon perleve was collected in the intertidal zone during low tide on Nanji island (Eastern China Sea, Zhejiang province). The living sponge material was immediately rinsed with sterile seawater to remove the non-attached bacteria, 0.5 g sample was triturated with sterile seawater and spread on the entire surface of 1/10 Marine Agar (peptone 0.5 g, yeast extraction 0.1 g, FePO4 0.1 g and agar 15 g dissolved in 1 l seawater, pH 7.2–7.6). After incubation at 25
C for 20 days, all colonies with different pigmentation and morpha were picked out. Crude extract preparation from marine bacteria Marine bacteria isolated as above were cultured in 300 ml Marine Broth (peptone 5 g, yeast extract 1 g and

202 FePO4 0.1 g, dissolved in 1 l seawater, pH 7.2–7.6) for the production of secondary metabolites in 500 ml Erlenmeyer flasks. Flasks were incubated on a rotatory shaker at 220 rev/min at 25
C. After 7 days of cultivation, the broth was centrifuged at 5000 · g for 30 min to remove the cell and extracted three times with ethyl acetate (EtOAc) (100 ml · 3). The EtOAc extracts were used as the crude samples for antimicrobial activity screening after solvent removal under reduced pressure at 37
C. Agar diffusion assay Antimicrobial activities were assayed in duplicate using a standard paper disc assay (Mearns-Spragg et al. 1998) and the test microorganisms were Bacillus subtilis CMCC 63003, Staphylococcus aureus CMCC 26001, Escherichia coli CMCC 44102, Agrobacterium tumefaciens AS 1.1416 and the yeast Saccharomyces cerevisiae ACCC 2.1882. The crude extracts dissolved in EtOAc at a concentration of 100 mg/ml, 2 · 10 ll samples were used to saturate antimicrobial assay paper disks (6 mm) with a period of drying between each application and 20 ll EtOAc was used as the control. The disks were placed onto the agar surface containing the test microorganisms, and incubated at 37
C for 24 h after a diffusion process for 10 h at 8
C. The diameters of any inhibition zones that had formed around the paper disks were measured. Identification of strains The strains which showed significant antimicrobial activity were identified to species level by PCR amplification of the 16S rRNA gene, BLAST analysis, and comparison with sequences in the GenBank nucleotide database. Specifically, the 16S rRNA gene from strains was amplified using universal primers 27f (5¢AGAGTTTGATCCTGGCTCAG-3¢) and 1492r (5¢-GGTTACCTTGTTACGACTT-3¢) and PCR conditions were as described previously (Acinas et al. 1999). PCR products were purified and sequenced by the Dingan Bio-company (Shanghai, China). The sequences were compared with known sequences in the GenBank nucleotide database and the species level was identified as the nearest phylogenetic neighbor with sequences >99% sequence similarity (Hentschel et al. 2001). Fermentation and isolation of metabolites of NJ6-3-1 The strain NJ6-3-1 was cultured in 25-l fermentation media (peptone 5 g, yeast extraction 1 g, FePO4 0.1 g and glucose 1 g dissolved in 1 l seawater, pH 7.2–7.6) at 25
C in a bioreactor with constant aeration. Culture broth samples of 100 ml were collected for the measurement of cell growth in a bioreactor during different fermentation time courses. Cell concentrations were estimated by measuring the dry cell weight, and dissolved oxygen (DO) and pH values were also

Li Zheng et al. recorded. After 6 days, the culture broth was centrifuged at 5000 · g for 30 min to remove the cells. The liquid phase was adsorbed on to 1200 g non-polar macroporous resin, washed with deionized water, and eluted with 2 l MeOH. Then the concentrated syrup was extracted with dichloromethane (DCM) after evaporating off the MeOH under reduced pressure at 37
C. The DCM extract (1.34 g) was loaded onto a silica gel column and eluted with Petrol/DCM (1:1), DCM, DCM/EtOAc (1:1), EtOAc, EtOAc/MeOH (1:1), MeOH, subsequently. All fractions were collected and subjected to TLC bioautography overlay assay. TLC bioautography overlay assay The antimicrobial activities of the fractions were detected using TLC bioautography overlay assay (Gibbons & Gray 1998). Briefly, every fraction was developed using DCM: MeOH (10:1, v/v) on TLC silica gel plates (TLC aluminium sheets, 20 · 20 cm, Silica Gel 60F254, Merck Co, USA), followed by detection at 254 and 365 nm. The developed TLC plates were sterilized by u.v. lamp for 30 min before encasing in nutrient agar in petri dishes (9mm) and then covered with another layer of molten nutrient agar (45
C) containing the test microorganism Staphylococcus aureus. After 10 h of diffusion at 8
C, the plates were incubated for 24 h at 37
C and then the upper agar was sprayed with 5 mg methylthiazoletetrazolium (MTT)/ml which was converted to a formazan dye by the test microorganism. Inhibition zones were observed as clear spots against a purple background. Purification of the antimicrobial compounds For separation and purification of the antimicrobial compounds from fractions with antimicrobial activity, the fractionations were performed by semipreparative HPLC (Waters Co., USA) using C18 reverse column (6 lm, 7.8 · 300 mm, Waters). Structure elucidation NMR [1H, 13C, homonuclear correlation spectroscopy (COSY), heteronuclear single quantum coherence spectroscopy (HSQC), and heteronuclear multiple bond correlation spectroscopy (HMBC)] spectra were recorded on a Varian INOVA 400 spectrometer (Varian Inc, Palo Alto, CA, USA), operating at 400 MHz for 1 H and 100 MHz for 13C, respectively, with tetramethylsilane (TMS) as internal standard. The molecular weight of each compound was detected by mass spectrometer (QP2010 GC-MS, Shimadzu Co., Tokyo, Japan) at 70 eV using the DI method. U.v. spectra were recorded on a photodiode array detector linked with HPLC.

Isolation of marine bacterial strains from sponge Hymeniacidon perleve Results

203

Genbank database with the following accession numbers: AY621063 and AY621379.

Antimicrobial activity screening of marine bacteria associated with the sponge

Fermentation of strain NJ6-3-1

Among twenty-nine marine strains isolated from the sponge Hymeniacidon perleve, eight showed antimicrobial activity to at least one terrestrial microorganism (Table 1). Strain NJ6-3-1 had a wide antimicrobial spectrum to all the test microorganisms and strain NJ63-2 had significant antimicrobial activity against four test microorganisms except for E. coli. The 16S rRNA sequence analysis indicated that NJ6-3-1 and NJ6-3-2 were most closely related to the members of the genus Pseudoalteromonas and Bacillus, respectively. The highest sequence similarity values (99.1%, 99.5%) were obtained between NJ6-3-1 and Pseudoalteromonas piscicida (Genbank accession number ABO 90232 and ABO 90233), NJ6-3-2 and Bacillus megaterium (Genbank accession number AY030338). The 16S rRNA sequences of NJ6-3-1 and NJ6-3-2 have been deposited in the

Fermentation of the marine bacterial strain NJ6-3-1 was carried out in a 30-l bioreactor in order to obtain enough biomass for further secondary bioactive metabolites study. As shown in Figure 1, DO and pH values had a close correlation with the growth conditions of NJ6-3-1. At the beginning of fermentation, the biomass content was low and the DO value was high at 97.2% indicating a low consumption of oxygen. Then, along with the growth of the bacterium, the pH value decreased gradually and the DO value fell sharply to zero due to the active consumption of a large amount of oxygen during the rapid proliferation process. During the exponential phase (days 2 and 3), the DO value remained at nearly zero, but subsequently rose to 100% on day 4 of fermentation, indicating the attainment of the stationary phase. The trend of pH value changes was the same, from

Table 1. Results of antimicrobial activity of marine bacteria associated with sponge Hymeniacidon perleve using agar diffusion assay. Strain

NJ6-3-1 NJ6-3-2 NJ6-8-1 NJ6-10-1 NJ6-14 NJ6-20 NJ6-22 NJ6-25

Antimicrobial activity BS

SA

EC

AT

SC

+++ ++ + ) + ) + +

++ ++ + ) ) ) + +

+ ) ) ) ) ) ) )

++ +++ + + + + ) )

++ ++ ) ) ) ) ) )

The test microorganisms are BS: Bacillus subtilis CMCC 63003, SA: Staphylococcus aureus CMCC 26001, EC: Escherichia coli CMCC 44102, AT: Agrobacterium tumefaciens AS1.1416, SC: fungus Saccharomyces cerevisiae ACCC 2.1882. ‘)’: No inhibition. ‘+’: Inhibition zone was 1– 3 mm. ‘++’: Inhibition zone was 3–5 mm. ‘+++’: Inhibition zone was ‡5 mm.

9 1.2

100

0.8 60 0.6

∆ Biomass 40 Ο DO pH 20

0.4 0.2

7

pH

8

80

DO (%)

Biomass (g/L)

1.0

6

0

0.0

5 0

1

2

3

4

5

6

Time (day) Figure 1. Time course of the growth of NJ6-3-1. (n) Biomass (g l)1) of NJ6-3-1. (s) DO value monitored in medium. (h) pH value monitored in medium.

204

Li Zheng et al.

lowest 5.9 at day 1.5–8.1 at day 3.5. During the stationary phase, the growth of cells slowed down, and the DO and pH values did not change again.

6

Isolation and structure elucidation of active compound

7

A DCM extract of fermentation medium was chromatographed on a silica gel column and 30 fractions were collected. Antimicrobial activities in all fractions were detected by TLC bioautography overlay assay using Staphylococcus aureus as the test microorganism, and fraction numbers 13–18 showed a strong antimicrobial inhibition zone with the same Rf value of 0.42 (developing solvent, DCM:MeOH, 10:1 v/v). These fractions were combined and subjected to further separation by Sephadex LH-20 gel permeation chromatography, eluting with MeOH. One major compound with antimicrobial activity was purified by semi-preparative HPLC eluting with 80% MeOH (Figure 2), the retention time was 13.90 min. The compound 1 with antimicrobial activity was a yellow-brown amorphous solid. The Rf value on silica gel TLC was 0.42 (developing solvent, DCM:MeOH, 10:1 v/v), three peaks showed in the HPLC-PDA (PhotoDiode Array) chromatogram with maximum absorption wavelengths of 246.5, 281.9 and 337.7 nm. The pseudo-molecular ion [M]+m/z was 168.02 in EIMS. Analysis of the 1H NMR spectrum of compound 1 showed only 7 aromatic proton signals (d ¼ 7–9 ppm) and another one free proton signal (d ¼ 10.6 ppm), the 13 C NMR spectrum showed eleven carbon signals around 110–145 ppm. Therefore, the molecular formula was calculated as C11H8N2 (calculated molecular weight: 168.07), and the structure was deduced as a carbolinelike alkaloid through mass spectrum similarity structure retrieval. The existence of only one singlet proton (8.888 ppm) implies it to be a beta or delta carboline.

5

3 N

8a 8

9a

N H

1

Figure 3. Structure of norharman (9H-Pyrido(3,4-b)indole) (1).

Comprehensive analysis of 2D NMR data, including the results of H-H COSY, HSQC and HMBC experiments enable us to assign compound 1 to norharman (bCarboline, 9H-Pyrido[3,4-b]indole) as shown in Figure 3. The assignment of chemical shifts of each proton and carbon is described in Table 2, consistent with the data reported (http://www.aist.go.jp/RIODB/ SDBS/ sdbs/owa/sdbs_sea.cre_frame_disp?sdbsno=9314).

Discussion Bacteria associated with marine invertebrates have been shown to have significant bioactivity including antifouling, antibacterial and cytotoxic acitivities and there have already been numerous studies on marine medical exploration and ecology investigation about spongeassociated bacteria and fungi (Holler et al. 2000; Burja et al. 2001; Osinga et al. 2001). In this present work, 28% (8/29) of the total bacterial strains that we isolated associated with the marine sponge Hymeniacidon perleve showed antimicrobial activity. The results confirmed that invertebrate-associated microorganisms are highly potential resources of bioactive natural products due to their competition for nutrition, space and light (Ivanova et al. 1998; Burgess et al. 1999; Armstrong et al. 2001). For antimicrobial assays, TLC autobiographic overlay assay has been proved exceptionally popular because it is easy to perform, of low cost, speedy, and has the ability to be scaled up to assess the antimicrobial activity of a large number of samples. It is a very sensitive assay and gives accurate localization of active compounds. This technique was used to track activity through the separation process in our work. Table 2. NMR spectral data of norharman in acetone-d6.

Figure 2. TLC bioautographic overlay assay of compounds separated by HPLC. The compound showing antimicrobial activity to Staphylococcus aureus was a b-carboline (norharman).

4

4a

4b

1

H and

Atom number

13

C chemical shifts (ppm)

1

1 3 4 5 6 7 8 8a 4b 4a 9a

134.961 139.434 115.068 122.320 120.274 128.869 112.565 141.663 137.190 112.565 122.076

8.888 8.335 8.025 8.197 7.231 7.514 7.597 – – – –

13

C chemical shifts of

H chemical shifts (ppm)

(1H, (1H, (1H, (1H, (1H, (1H, (1H,

s) d, J = 5.2 Hz ) d, J = 5.6 Hz) d, J = 8.0 Hz) ddd, J = 6.8 Hz) ddd, J = 7.2 Hz) d, J = 8.0 Hz)

Isolation of marine bacterial strains from sponge Hymeniacidon perleve Carboline was initially isolated from terrestrial plants and possesses antimicrobial activities (Agurell et al. 1968; Al-Shamma et al. 1981). Along with the exploration of marine natural products, some carboline alkaloids have been found in marine organisms such as bryozoan, ascidian and sponges and more biological activities have been discovered (Ang et al. 2000; Rashid et al. 2001; Harwood et al. 2003). There was no report of b-carboline from a marine bacterium until last year when the antibiotic harman (1-methyl-9H-pyrido[3,4b]indole) produced by a tunicate-associated bacterium was identified (Aassila et al. 2003). In our studies, this is the first report of norharman isolated from a marine bacterium associated with sponge. Aassila et al. (2003) reported that harman has significant antimicrobial activity against the marine bacteria Vibrio anguillarum, Vibrio alginolyticus, Vibrio harveyi, and Vibrio carchariae with MIC values of 3.12, 12.5, 25 and 25 lg/ml, respectively. Our results show that norharman has antimicrobial activity against Bacillus subtilis, Staphylococcus aureus, and the plant pathogen Agrobacterium tumefaciens with MIC values of 50, 50 and 100 lg/ml, respectively. Besides this antimicrobial activity, both harman and norharman also had activity against the cyanobacteria Anabaena cylindrical, Anabaena variabilis, Oscillatoria agardhii, Anacystis marina, Microcystis aeruginosa and Microcystis viridis, at a concentration of 30 lg/disk (Kodani et al. 2002). The similarity bioactivities of these two compounds were probably due to their similar structures. Previous studies have indicated that sponge-associated microorganisms are the origin of bioactive metabolites in sponges (Unson & Faulkner 1993; Unson et al. 1994; Bewley & Faulkner 1998), our study supports this hypothesis, since norharman has been found in an Indonesian sponge (Rao et al. 2003). The work of isolating bioactive compounds from the sponge Hymeniacidon perleve is underway and we expect to find exact evidence to confirm this viewpoint. Acknowledgements This work was supported by grants from National Natural Science Foundation of China (29932030), and China New Drug Project (2003AA2Z3511), and RC02059. The authors would like to thank Professor Jinhe Li of Institute of Oceanology, Chinese Academy of Sciences, for identification of the sponge. References Aassila, H., Bourguet-Kondracki, M.L., Rifai, S., Fassouane, A. & Guyot, M. 2003 Identification of harman as the antibiotic compound produced by a tunicate-associated bacterium. Marine Biotechnology 5, 163–166. Acinas, S.G., Anton, J. & Rodriguez-Valera, F. 1999 Diversity of freeliving and attached bacteria in offshore western Mediterranean waters as depicted by analysis of genes encoding 16s rRNA. Applied and Environmental Microbiology 65, 514–522.

205

Agurell, S., Holmstedt, B., Lindgren, J.E. & Schultes, R.E. 1968 Identification of two new beta-carboline alkaloids in South American hallucinogenic plants. Biochemical Pharmacology 17, 2487–2488. Al-Shamma, A., Drake, S., Flynn, D.L., Mitscher, L.A., Park, Y.H., Rao, G.S., Simpson, A., Swayze, J.K., Veysoglu, T. & Wu, S.T. 1981 Antimicrobial agents from higher plants. Antimicrobial agents from Peganum harmala seeds. Natural Products 44, 745– 747. Ang, K.K., Holmes, M.J., Higa, T., Hamann, M.T. & Kara, U.A. 2000 In vivo antimalarial activity of the beta-carboline alkaloid manzamine A. Antimicrobial Agents and Chemotherapy 44, 1645–1649. Armstrong, E., Yan, L., Boyd, K.G., Wright, P.C. & Burgess, J.G. 2001 The symbiotic role of marine microbes on living surfaces. Hydrobiologia 461, 37–40. Bewley, C.A. & Faulkner, D.J. 1998 Lithistid sponges: star performers or hosts to the stars. Angewandte Chemie-International Edition 37, 2162–2178. Burgess, J.G., Jordan, E.M., Bregu, M., Mearns-Spragg, A. & Boyd, K.G. 1999 Microbial antagonism: a neglected avenue of natural products research. Journal of Biotechnology 70, 27–32. Burja, A.M. & Hill, R.T. 2001 Microbial symbionts of the Australian Great Barrier Reef sponge, Candidaspongia flabellate. Hydrobiologia 461, 41–47. Faulkner, D.J. 2002 Marine natural products. Natural Products Report 19, 1–48. Gibbons, S. & Gray, A.I. 1998 Isolation by polanar chromatography. In Natural Products Isolation, ed. Cannell, RJP. pp. 209– 245. Humana press, Totowa, New Jersey. ISBN 0896033627. Harwood, D.T., Urban, S., Blunt, J.W. & Munro, M.H. 2003 Betacarboline alkaloids from a New Zealand marine bryozoan, Cribricellina cribraria. Natural Products Research 17, 15–19. Hentschel, U., Schmid, M., Wagner, M., Fieseler, L., Gernert, C. & Hacker, J. 2001 Isolation and phylogenetic analysis of bacteria with antimicrobial activities from the Mediterranean sponges Aplysina aerophoba and Aplysina cavernicda. FEMS Microbiology Ecology 35, 305–312. Holler, U., Wright, A.D., Matthee, G.F., Konig, G.M., Draeger, S., Hans-Jurgen, A. & Schulz, B. 2000 Fungi from marine sponges: diversity, biological activity and secondary metabolites. Mycological Research 104, 1354–1365. Ivanova, E.P., Nicolan, D.V., Yumoto, N., Taguchi, T., Okamoto, K., Tatsu, Y. & Yoshikawa, S. 1998 Impact of conditions of cultivation and adsorption on antimicrobial activity of marine bacteria. Marine Biology 130, 545–551. Jayatilake, G.S., Thornton, M.P., Leonard, A.C., Grimwade, J.E. & Baker, B.J. 1996 Metabolites from an Antarctic sponge-associated bacterium, Pseudononas aeruginosa. Journal of Natural Products 59, 293–296. Kodani, S., Imoto, A., Mitsutani, A. & Murakami, M. 2002 Isolation and identification of the antialgal compound, harmane (1-methylb-carboline), produced by the algicidal bacterium, Pseudomonas sp. K44-1. Journal of Applied Phycology 14, 109–114. Mearns-Spragg, A., Bregu, M., Boyd, K.G. & Burgess, J.G. 1998 Cross-species induction and enhancement of antimicrobial activity produced by epibiotic bacteria from marine algae and invertebrates, after exposure to terrestrial bacteria. Letters in Applied Microbiology 27, 142–146. Mitova, M., Tommonaro, G. & Rosa, D.S. 2003 A novel cyclopeptide from a bacterium associated with the marine sponge Ircinia muscarum. Zeitschrift fu¨r Naturforschung 58, 740–745. Osinga, R., Armstrong, E., Burgess, G.J., Hoffmann, F., Reitner, J. & Schumann-Kindel, G. 2001 Sponge-microbe associations and their importance for sponge bioprocess engineering. Hydrobiologia 461, 55–62. Pettit, G.R., Xu, J.P., Chapuis, J.C., Pettit, R.K., Tackett, L.P., Doubek, D.L., Hooper, J.N. & Schmidt, J.M. 2004 Antineoplastic agents. 520. Isolation and structure of irciniastatins A and B from the Indo-Pacific marine sponge Ircinia ramosa. Journal of Medicinal Chemistry 47, 1149–1152.

206 Rao, K.V., Santarsiero, B.D., Mesecar, A.D., Schinazi, R.F., Tekwani, B.L. & Hamann, M.T. 2003 New manzamine alkaloids with activity against infectious and tropical parasitic diseases from an Indonesian sponge. Journal of Natural Products 66, 823–828. Rashid, M.A., Gustafson, K.R. & Boyd, M.R. 2001 New cytotoxic Nmethylated beta-carboline alkaloids from the marine ascidian Eudistoma gilboverde. Journal of Natural Products 64, 1454–1456. Santavy, D.L., Willenz, P. & Colwell, R.R. 1990 Phenotypic study of bacteria associated with the Caribbean sclerosponge, Ceratoporella nicholsoni. Applied and Environmental Microbiology 56, 1750– 1762. Sonnenschein, R.N., Farias, J.J., Tenney, K., Mooberry, S.L., Lobkovsky, E., Clardy, J. & Crews, P.A. 2004 Further study of the cytotoxic constituents of a milnamide-producing sponge. Organic Letters 6, 779–782. Sponga, F., Cavaletti, L., Lazzarini, A., Borghi, A., Ciciliato, I., Losi, D. & Marinelli, F. 1999 Biodiversity and potentials of marinederived microorganisms. Journal of Biotechnology 70, 65–69.

Li Zheng et al. Strahl, E.D., Dobson, W.E. & Lundie, J.L.L. 2002 Isolation and screening of brittlestar-associated bacteria for antibacterial activity. Current Microbiology 44, 450–459. Suzumura, K., Yoko, T., Funatsu, M., Nagai, K., Tanaka, K., Zhang, H. & Suzuki, K. 2003 YM-266183 and YM-266184, novel thiopeptide antibiotics produced by Bacillus cereus isolated from a marine sponge II. Structure elucidation. Journal of Antibiotics (Tokyo) 56, 129–134. Unson, M.D. & Faulkner, D.J. 1993 Cyanobacterial symbiont biosynthesis of chlorinated metabolites from Dysidea herbacea. Experientia 49, 349–353. Unson, M.D., Holland, N.D. & Faulkner, D.J. 1994 A brominated secondary metabolite synthesized by the cyanobacterial symbiont of marine sponge and accumulation of the crystalline metabolite in the sponge tissue. Marine Biology 119, 1–11. Wilkinson, C.R. 1978 Microbial association in sponges. II: numerical analysis of sponge and water bacterial populations. Marine Biology 49, 169–176.