World Journal of Microbiology and Biotechnology

World Journal of Microbiology and Biotechnology July 2005 Volume 21, Number 5 Prevalence and antibiotic resistance of Escherichia coli in tropical seafood

(619 - 623)

H. S. Kumar, A. Parvathi, I. Karunasagar, I. Karunasagar DOI: 10.1007/s11274-004-3555-8 Rapid detection of Salmonella typhi in foods by combination of immunomagnetic separation and polymerase chain reaction

(625 - 628)

S Kumar, K Balakrishna, G.P. Singh, H.V. Batra DOI: 10.1007/s11274-004-3553-x DNA amplification fingerprinting as a tool for checking genetic purity of strains in the cyanobacterial inoculum

(629 - 634)

B. Jeberlin Prabina, K. Kumar, S. Kannaiyan DOI: 10.1007/s11274-004-3566-5 Utilization of a polyphasic approach in the taxonomic reassessment of antibiotic- and enzyme-producing Bacillus spp. isolated from the Philippines

(635 - 644)

Marie Antonette Ruth V. Guerra-Cantera and Asuncion K. Raymundo DOI: 10.1007/s11274-004-3567-4 Relationship among acidophilic bacteria from diverse environments as determined by randomly amplified polymorphic DNA analysis (RAPD)

(645 - 648)

Tanveer Akbar, Kalsoom Akhtar, Muhammad A. Ghauri, Munir A. Anwar, Moazur Rehman, Mehboobur Rehman, Yusuf Zafar, Ahmad M. Khalid DOI: 10.1007/s11274-004-3568-3 Purification and characterization of a thermoalkalophilic xylanase from Bacillus sp.

(649 - 654)

M. P. Sapre, H. Jha, M. B. Patil DOI: 10.1007/s11274-004-3569-2 Liquid–liquid extraction of an extracellular alkaline protease from fermentation broth using aqueous two-phase and reversed micelles systems

(655 - 659)

T. S. Porto, T. I. R. Monteiro, K. A. Moreira, J. L. Lima-Filho, M. P. C. Silva, A. L. F. Porto, M. G. Carneiro-da-Cunha DOI: 10.1007/s11274-004-3570-9 Bioleaching of nickel from equilibrium fluid catalytic cracking catalysts Oguz Bayraktar DOI: 10.1007/s11274-004-3573-6

(661 - 665)

Decolourization of textile dye Reactive Violet 5 by a newly isolated bacterial consortium RVM 11.1

(667 - 672)

Safia Moosvi, Haresh Keharia, Datta Madamwar DOI: 10.1007/s11274-004-3612-3 Exopolysaccharide production by Lactobacillus delbruckii subsp. bulgaricus and Streptococcus thermophilus strains under different growth conditions

(673 - 677)

Belma Aslim, Zehra Nur Yüksekdag˘, Yavuz Beyatli, Nazime Mercan DOI: 10.1007/s11274-004-3613-2 Isolation and characterization of actinomycetes antagonistic to pathogenic Vibrio spp. from nearshore marine sediments

(679 - 682)

J. L. You, L. X. Cao, G. F. Liu, S. N. Zhou, H. M. Tan, Y. C. Lin DOI: 10.1007/s11274-004-3851-3 Mathematical description of bikaverin production in a fluidized bed bioreactor

(683 - 688)

Ma. del Carmen Chávez-Parga, Omar González-Ortega, Guadalupe SánchezCornejo, Ma. de la Luz X. Negrete-Rodríguez, Guillermo González-Alatorre, Eleazar M. Escamilla-Silva DOI: 10.1007/s11274-004-3854-0 Adsorption-elution purification of chimeric Bacillus stearothermophilus leucine aminopeptidase II with raw-starchbinding activity

(689 - 694)

Yu-Wen Hua, Meng-Chun Chi, Huei-Fen Lo, Lih-Ying Kuo, Kuo-Lung Ku, Long-Liu Lin DOI: 10.1007/s11274-004-3853-1 Continuous citric acid fermentation by Candida oleophila under nitrogen limitation at constant C/N ratio

(695 - 705)

S. Anastassiadis, C. Wandrey, H. -J. Rehm DOI: 10.1007/s11274-004-3850-4 The influence of different yeasts on the fermentation, composition and sensory quality of cachaça

(707 - 715)

Evelyn Souza Oliveira, Helena Maria André Bolini Cardello, Elisangela Marques Jeronimo, Elson Luiz Rocha Souza, Gil Eduardo Serra DOI: 10.1007/s11274-004-4490-4 Biological control of sheep parasite nematodes by nematodetrapping fungi: in vitro activity and after passage through the gastrointestinal tract

(717 - 722)

Érika B. N. Graminha, Alvimar J. Costa, Gilson P. Oliveira, Antonio C. Monteiro, Solange B. S. Palmeira DOI: 10.1007/s11274-004-4045-8 Micro-encapsulation of Bifidobacterium lactis for incorporation into soft foods L. D. McMaster and S. A. Kokott DOI: 10.1007/s11274-004-4798-0

(723 - 728)

Evaluation of fluorescent Pseudomonads and Bacillus isolates for the biocontrol of a wilt disease complex of pigeonpea

(729 - 732)

Shazia Siddiqui, Zaki A. Siddiqui, Iqbal Ahmad DOI: 10.1007/s11274-004-4799-z Purification and characterization of a thermostable chitinase from Bacillus licheniformis Mb-2

(733 - 738)

Aris Toharisman, Maggy Thenawidjaja Suhartono, Margarethe Spindler-Barth, Jae-Kwan Hwang, Yu-Ryang Pyun DOI: 10.1007/s11274-004-4797-1 Use of alginate and cryo-protective sugars to improve the viability of lactic acid bacteria after freezing and freeze-drying

(739 - 746)

B. De Giulio, P. Orlando, G. Barba, R. Coppola, M. De Rosa, A. Sada, P. P. De Prisco, F. Nazzaro DOI: 10.1007/s11274-004-4735-2 Bio-production of compost with low pH and high soluble phosphorus from sugar cane bagasse enriched with rock phosphate

(747 - 752)

Gaber Zayed and Heba Abdel-Motaal DOI: 10.1007/s11274-004-5407-y Purification and characterization of naringinase from a newly (753 - 758) isolated strain of Aspergillus niger 1344 for the transformation of flavonoids Munish Puri and Sukirti Kalra DOI: 10.1007/s11274-004-5488-7 Galacto-oligosaccharide production by a thermostable recombinant β-galactosidase from Thermotoga maritima

(759 - 764)

Eun-Su Ji, Nyun-Ho Park, Deok-Kun Oh DOI: 10.1007/s11274-004-5487-8 Production of poly(3-hydroxybutyrate) and poly(3hydroxybutyrate-co-3-hydroxyvalerate) by Rhodopseudomonas palustris SP5212

(765 - 769)

Mahuya Mukhopadhyay, A. Patra, A. K. Paul DOI: 10.1007/s11274-004-5565-y Effect of sugar-feeding strategies on astaxanthin production by Xanthophyllomyces dendrorhous Zhong-Ce Hu, Yu-Guo Zheng, Zhao Wang, Yin-Chu Shen DOI: 10.1007/s11274-004-5566-x

(771 - 775)

Ó Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:619–623 DOI 10.1007/s11274-004-3555-8

Prevalence and antibiotic resistance of Escherichia coli in tropical seafood H.S. Kumar, A. Parvathi, I. Karunasagar and I. Karunasagar* Department of Fishery Microbiology, University of Agricultural Sciences, College of Fisheries, Mangalore 575 002, India *Author for correspondence: Tel.: +91-824-2246384, Fax: 91-824-2246384, E-mail: [email protected]

Keywords: antibiotic resistance, Escherichia coli, faecal coliforms, plasmid, seafood

Summary The occurrence and antibiotic resistance of Escherichia coli in tropical seafood was studied. A 3-tube MPN method was used for determining the level of faecal contamination of fresh and processed seafood. Of the 188 samples tested which included finfish, shellfish, water and ice, 155 were positive for the presence of faecal coliforms following incubation at 44.5 °C. However, E. coli was isolated from only 47% of the samples positive for faecal coliforms. The antibiotic resistance of 116 strains isolated from seafood was tested using 14 different antibiotics including ampicillin, cephalothin, chloramphenicol, ciprofloxacin, gentamycin, nalidixic acid, streptomycin and vancomycin. Seven strains were resistant to more than five antibiotics of which one was resistant to eight antibiotics. The multiple drug resistant strains harboured plasmids of varying sizes. Antibiotic susceptibility studies revealed that seafood from India contains multiple antibiotic resistant strains of E. coli which may serve as a reservoir for antibiotic resistance genes in the aquatic environment. All the strains used in this study did not harbour any virulence genes commonly associated with pathogenic E. coli, when tested by polymerase chain reaction (PCR).

Introduction Seafood is a major vehicle for transmission of several bacterial diseases. The faecal contamination of natural water bodies has emerged as a major challenge in developing and densely populated countries like India. Estuaries and coastal water bodies, which are the major sources of seafood in India, are often contaminated by the activities of adjoining populations and partially treated or untreated sewage from the townships is released into these water bodies. The fish harvested from such areas often contain human pathogenic microorganisms. In addition, poor sanitation in landing centers and the open fish markets exacerbates the situation. Escherichia coli has been traditionally recognized as an indicator organism of faecal contamination of water and seafood (Geldreich 1997). Testing of seafood for the presence of E. coli is still a gold standard used to assess the faecal contamination in seafood processing plants in India and elsewhere. E. coli is a normal inhabitant of the intestinal tracts of all warm blooded animals. However, strains of human pathogenic E. coli have evolved that are recorded as causative agents of a broad range of human diseases compared to any other pathogenic bacteria (Nataro & Kaper 1998; Paton & Paton 1998).

Antimicrobial resistance in bacteria associated with food and water has been a global concern. It is now widely accepted that there is an association between the use of antimicrobial agents and the occurrence of resistance. Antimicrobials exert a selective pressure on microorganisms and therefore their use is considered a key issue in epidemiological studies (McGeer 1998). The disease threat from antibiotic resistant strains of pathogens has increased in recent years (Williams & Heymann 1998). Antimicrobial resistance can spread through horizontal transfer of resistance genes from one type of bacteria to another. The presence of resistance, together with the acquisition of virulence genes can lead to clonal expansion and spread of a particular disease-causing agent. Hence it is considered important to study the antimicrobial resistance in pathogenic as well as indicator bacteria associated with food animals (OIE 1999). The study reported here was undertaken to determine the prevalence of E. coli in seafood, the occurrence of any pathogenic genotypes in seafood isolates of E. coli and the distribution of antibiotic resistance in them. The results demonstrate the widespread occurrence of E. coli in tropical seafood and highlight the associated health risks due to the distribution of antibiotic resistance in seafood-associated strains of E. coli in India.

620

H.S. Kumar et al.

Materials and methods Samples A total of 188 samples were analysed which included finfish, shellfish, ice and water (Table 1) for total and faecal coliforms using the 3-tube MPN method (FDA 1998). Samples of fish and shellfish included both fresh (n ¼ 128) and frozen (n ¼ 23), collected from different fish markets, landing centres and seafood processing plants, in and around Mangalore. Twelve water samples and 10 ice samples were collected from the fresh fish market and 15 ice samples were collected from shrimpprocessing plants. The samples were transported to the laboratory immediately and analysed within an hour of collection. All dehydrated media used in this study for microbiological analysis were obtained from Himedia, Mumbai, India and prepared according to manufacturer’s instructions. Briefly, the samples were homogenized in a sterile blender and inoculated into lauryl sulphate tryptose broth (LSTB) and incubated at 37 °C for 24–48 h. The LSTB tubes showing turbidity and gas in Durham tubes were recorded as positive. Two loopfuls from positive LSTB tubes were inoculated into corresponding labelled tubes with 5 ml of EC (E. coli) broth medium. EC broth tubes exhibiting turbidity and gas production following 24 h incubation at 44.5 °C in a water bath were considered positive for the presence of faecal coliforms. For isolation of E. coli, two loopfuls from positive EC broth tubes were streaked onto eosin methylene blue (EMB) agar plates. A minimum of five typical colonies were picked up, purified on tryptone soya agar (TSA) plates and subjected to standard biochemical tests for the identification of E. coli. Antimicrobial susceptibility testing A total of 116 E. coli strains were tested for antibiotic resistance by standard agar disc diffusion technique (Bauer et al. 1966) on Mueller Hinton agar using commercial discs (HiMedia, Mumbai, India). The following antibiotics with the disc strength in parentheses were used: ampicillin (10 mcg), amoxycillin (30 mcg), ceftriaxone (30 mcg), cephalothin (30 mcg),

chloramphenicol (30 mcg), ciprofloxacin (5 mcg), gentamycin (10 mcg), kanamycin (30 mcg), nalidixic acid (30 mcg), penicillin G (10 U), rifampicin (5 mcg), streptomycin (10 mcg), tetracycline (30 mcg) and vancomycin (30 mcg). Characterization of E. coli strains The four STEC strains used in this study were those previously described (Kumar et al. 2001). These included one each of O76:H21 (stx1+, stx2+, ehxA+) and ONT:H21 (stx1+, stx2+, ehxA+) and two O110:NM (stx1+, ehxA+). The remaining 112 strains were tested for the presence of virulence genes by polymerase chain reaction (PCR) using primers previously described for labile toxin (lt) and stable toxin (st) genes of enterotoxigenic E. coli (Olsvik & Strockbine 1990), shiga toxin genes stx1 and stx2 (Pal et al. 1999), E. coli attachment and effacement (eae) gene (Paton & Paton 1998) and enterohaemorrhagic E. coli haemolysin A (ehxA) gene (Fratamico et al. 1995). Extraction of plasmids Plasmids were extracted by the alkaline lysis method (Sambrook et al. 1989) from 10 strains exhibiting multiple resistance to the antimicrobials tested. Briefly, 1.5 ml of the overnight culture in Luria Bertani (LB) broth was centrifuged at 10,000 rev/min for 10 min in a biofuge (Heraeus, Germany) and the pellet resuspended in 100 ll of solution I (50 mM glucose; 25 mM Tris–Cl, pH 8.0; 10 mM EDTA, pH 8.0). The cells were lysed by the addition of 200 ll of solution II (0.2 N NaOH; 1% SDS). The genomic DNA was precipitated by the addition of solution III (5 M potassium acetate and 5 M glacial acetic acid) and pelleted by centrifugation at 10,000 rev/min for 10 min. The supernatant containing plasmid DNA was extracted twice with phenol: chloroform, ethanol precipitated, treated with RNAase (20 ll/ml) and finally dissolved in 50 ll TE buffer (10 mM Tris, pH 8.0; 1 mM EDTA). The extracted plasmids were electrophoresed on 0.8% agarose gel, stained with ethidium bromide and photographed using gel documentation system (Herolab, Weisloch, Germany).

Table 1. Prevalence of faecal coliforms and E. coli in seafood, ice and water samples collected from different sources. Sample type

Source

No. of samples

No. positive for faecal coliforms

No. positive for E. coli

% Prevalence of E. coli

Fin fish Fin fish Shrimp Shrimp Shrimp Clams Oysters Water Ice Ice Total

Fresh fish market Landing centre Fresh fish market Landing centre Processing plants Fresh fish market Estuaries Fresh fish market Fresh fish market Processing plants

21 20 20 25 23 32 10 12 10 15 188

21 20 20 20 8 32 10 12 7 5 155

8 5 3 4 2 25 10 8 7 1 73

38 25 15 16 8.6 78 100 66 70 6.66 38.8

E. coli in tropical seafood Results Incidence of faecal coliforms and E. coli in fish and shellfish Of the total 188 samples tested, 155 were found to be positive for the presence of faecal coliforms in EC broth when incubated at 44.5 oC suggesting a prevalence of 82.4% but interestingly only 73 of 155 (47%) samples positive for faecal coliforms were in fact positive for E. coli (Table 1). Overall, 38.8% (73 of 188) samples were positive for E. coli after isolation and biochemical characterization. Though all the samples of fresh finfish collected from landing centre and fresh fish market were positive for the presence of faecal coliforms by MPN method, only 8/21 (38.8%) finfish samples from fresh fish market and 5/20 (25%) from landing centre were confirmed as E. coli. All of the 20 shrimp samples collected from market and 20 of 25 samples from landing centre were positive for the presence of faecal coliforms (Table 1) but the prevalence of E. coli in these samples was just 15% and 16%, respectively. In the case of frozen shrimp samples collected from processing plants in Mangalore, 8/23 (34.78%) were found to be positive for the presence of faecal coliforms of which only 2 (8.6%) contained E. coli. Results in Table 2 show that among freshly caught shrimp samples, P. monodon and P. indicus were negative for the organism while E. coli was detected in Metapenaeus spp. and Parapenaeopsis spp. While none of the frozen P. monodon samples were positive for E. coli, 2% of frozen P. indicus was found to harbour E. coli. All of the 32 clam samples examined in the present study were positive for the presence of faecal coliforms and 25 (78%) of these were positive for E. coli. Both the species of clams used in this study, Meretrix meretrix and M. casta harboured E. coli as seen from Table 2. In the case of the oyster sample, Crassostrea madrasensis, all of the 10 samples examined in this study were positive for faecal coliforms and E. coli, recording a cent percent level of prevalence in the samples undertaken in this study.

621 Table 2. Incidence of E. coli in different species of finfish and shellfish. Sample types

Finfish Sardines Mackerles White bait Soles Pink perch Scianids Fresh water fish Total Shell fish Fresh shrimps P. monodon P. indicus Metapenaeus spp. Parapenaeopsis spp. Frozen shrimps P. monodon P. indicus Total Clams Meretrix meretrix M. casta Total Oysters Crassostrea madrasensis Total

No. tested

No. positive for E. coli

8 7 5 5 7 5 4 41

3 2 2 – 2 2 2 13

10 15 15 5

– – 5 2

8 15 68

– 2 9

22 10 32

18 7 25

10 10

10 10

bials tested. However, 29 were resistant to one antibiotic, 25 to 3 antibiotics, 14 to 4 antibiotics, one was resistant to 5, 3 were resistant to 6, and 1 was resistant to 8 antibiotics used. The 8 antibiotics to which one of the strains was resistant included ampicillin, cephalothin, kanamycin, nalidixic acid, penicillin, streptomycin, trimethoprim, and vancomycin. Among 6 STEC strains tested, one was resistant to 6 antibiotics (ampicillin, amoxycillin, cephalothin, ceftriaxone, penicillin, and vancomycin). Significantly 3 of the 6 strains tested were resistant to penicillin and cephalothin while two were resistant to amoxycillin. Table 3 describes the percentage of strains resistant to individual antibiotic types. Resistance to vancomycin

Prevalence of E. coli in ice and water

Table 3. Percentage of E. coli strains resistant to antibiotics tested in this study.

All of the 12 samples of water collected from fresh fish market were faecal coliform positive, of which 8 contained E. coli (Table 1). Seven of 10 ice samples from fresh fish market were positive for faecal coliforms and E. coli. Ice samples from processing plants were also analysed and the results showed 5 of 15 samples positive for faecal coliforms and interestingly only one positive for E. coli.

Antibiotic used

% Resistant

Amoxycillin Ampicillin Ceftriaxone Cephalothin Chloramphenicol Ciprofloxacin Gentamycin Kanamycin Nalidixic acid Penicillin G Rifampicin Streptomycin Tetracycline Vancomycin

2.5 4.3 0.86 67.2 0 0 0 2.5 0.86 42 0 6 0.86 94

Antibiotic susceptibility testing Forty two of the 116 strains of E. coli analysed during this study were resistant to at least two antimicrobials tested and no strain was sensitive to all the antimicro-

622

Figure 1. Agarose gel electrophoresis analysis of plasmids extracted from multiple antibiotic resistant E. coli strains. Lane 1, DNA marker (k DNA digested with EcoRI & Hind III); lanes 2, 3, 4 & 5, plasmids from different E. coli strains; lane 6, 100 bp DNA ladder.

was observed in 94% of the strains followed by cephalothin (67.2%) and penicillin 42%. No resistance was observed to chloramphenicol, ciprofloxacin, rifampicin and gentamycin. Six percent of the strains were resistant to streptomycin while 4.3% were resistant to ampicillin. Resistance to amoxycillin and kanamycin was observed in 2.5% of the strains. A small percentage (0.86) of the strains was resistant to tetracycline and nalidixic acid. The resistant strains of E. coli were found to harbour plasmids of varying sizes (Figure 1).

Discussion A large percentage of samples tested (82%) were positive for the presence of faecal coliforms by the MPN method (Table 1) as indicated by the production of acid and gas in EC broth incubated at an elevated temperature of 44.5 oC but only 38.8% of the samples positive for faecal coliforms contained E. coli, strengthening the earlier observation by Jeyasekaran et al. (1990). The prevalence of E. coli was the highest (78%) in clams from the fresh fish market. This is to be expected since these shellfish are present in the estuarine environment where contamination with faecal matter is often recorded. It is of significance that water and ice from the fish market had a high prevalence of E. coli (66% and 70%, respectively) which is in contrast to ice from processing plants where the prevalence was only 6.6%. This shows that water and ice used in the fresh fish markets and landing centre in Mangalore is of poor quality and are thus important sources of contamination for fish. E. coli does not survive in the marine environment for long and therefore this organism cannot be expected in fish harvested off-shore. E. coli was isolated from pelagic fish such as sardines, mackerels, whitebaits and scianids, while the soles, a demersal fish group, were negative for the presence of E. coli (Table 2). However, the presence of E. coli

H.S. Kumar et al. detected in some marine fish in the present study might represent post-harvest contamination such as in the landing centre and fish market through use of unpotable ice and water. Interestingly, fresh shrimps such as P. monodon and P. indicus were free from E. coli. In view of their high market value, it is possible that these items are carefully handled to prevent any contamination. In line with the above observation, low value shrimps such as Metapenaeus spp. and Parapenaeopsis spp. showed contamination by E. coli, perhaps due to poor handling. Among frozen shrimp samples, P. monodon was free from E. coli while 13% of P. indicus had this organism. From the results presented in Table 1 where 8 of 23 frozen shrimp samples from processing plants were positive for faecal coliforms but the recovery of E. coli from only 2 samples (3.6%), it is inferred that the low temperature sensitivity of E. coli is responsible for its low positivity. However when it is recovered from processed seafood, it could be attributed to contamination during post-process handling and transportation. In spite of the observations that the faecal coliform test often leads to overestimation of the contamination in tropical seafood (Knutton et al. 1977), it remains the most convenient method in the absence of any other reliable indicator organism. Results of our study indicate that it is necessary to isolate E. coli from faecal coliform-positive samples to confirm faecal contamination. Kumar et al. (2001) have reported the occurrence of pathogenic E. coli strains in Indian seafood. No separate enrichment methods are available to selectively isolate various pathogenic groups of E. coli. Isolation of E. coli will help further characterization of such strains. The prevalence of various pathogenic strains in Indian seafood needs to be studied to estimate the disease burden from E. coli contamination. Antibiotic resistance in E. coli isolated from seafood The results of the antibiotic resistance study (Tables 3 and 4) indicate that E. coli strains resistant to more than one drug are widespread in seafood included in this study. Resistant strains of E. coli were recorded to 9 out of 13 antibiotics tested. Apart from a large percentage of strains (67%) that were resistant to the cephalosporin antibiotic, cephalothin, low levels of resistance was observed against ampicillin and the fluoroquinolone antibiotic, nalidixic acid. The presence of multiple antibiotic resistance in seafood strains of E. coli is of grave concern. In Mangalore, the untreated or partially treated domestic sewage is released into open estuaries. The presence of antibiotic residues in such wastes cannot be ruled out since these also contain hospital wastes, though no comprehensive studies have been conducted to support this hypothesis. In addition to the sources of contamination already mentioned, the surface runoff during the monsoon might introduce resistant strains of E. coli into the estuarine environment. It is therefore to be expected

E. coli in tropical seafood that the fish and shellfish harvested from such environments would be contaminated with E. coli. On the other hand, introduction of E. coli may also take place away from the areas of seafood harvest such as in the open market by use of contaminated ice and water used for preserving or washing the fish. There is a paucity of data regarding the presence of antibiotic resistance either in clinical or environmental strains of E. coli from India. The presence of cephalothin and penicillin resistance among the seafood strains of E. coli might suggest a hospital or veterinary origin for such strains. E. coli in aquatic environment is exposed to sublethal levels of antibiotics in the aquatic environments brought in by the discharged wastes. The use of antibiotics is also widespread in animal industry and agriculture. It is estimated that the use of antibiotics in animals is 100–1000 times that in human population (Feimen 1998; Levy 1998; Witte 1998). No study so far has been conducted to understand the effects of antibiotic use in animals, agriculture and aquaculture in India. The unregulated use of antibiotics in such systems in India contributes significantly to the antibiotic residues in meat and aquaculture products. The antibiotic resistance patterns of seafood strains of E. coli observed in this study suggests a greater risk in the form of transfer of resistance to other pathogenic bacteria. The frequent exchange of plasmids between E. coli and other coliform bacteria has been previously reported (Grabow et al. 1973, 1976). Our previous study reports the presence of Salmonella sp. in a significant percentage of fish and shellfish (Kumar et al. 2001). The antibiotic resistance in non-pathogenic E. coli reported here cannot be ruled out as insignificant since the transfer of resistance genes can take place between closely related bacteria such as members of Enterobacteriaceae family. The resistant strains in seafood can form commensal flora via the food chain (Van den Bogaard & Stobberingh 1999; Witte 1998). The source and routes of contamination is difficult to predict as the aquatic system receives bacterial population from diverse sources. A comprehensive surveillance is required to determine the presence and distribution of antibiotic resistant E. coli in foods, animals and agriculture in India.

References Bauer, A.W., Kirby, W.M.M., Skerris, J.C. & Turck, M. 1996 Antibiotic susceptibility testing by a standardized single diffusion method. American Journal of Clinical Pathology 45, 494–496. Edberg, S.C., Patterson, J.E. & Smith, D.B. 1994 Differentiation of distribution system, source water and clinical coliforms by DNA analysis. Journal of Clinical Microbiology 32, 139–142.

623 FDA. 1998 Bacteriological Analytical Manual, Food and Drug Administration, 8th ed. Arlington, VA: AOAC International. ISBN 0935584595. Feimen, S.E. 1998 Antibiotics in animal feed: drug resistance revisited. American Society for Microbiology News 64, 24–30. Fratamico, P.M., Sackitey, S.K., Wiedmann, M. & Deng, M.Y. 1995 Detection of Escherichia coli O157:H7 by multiplex PCR. Journal of Clinical Microbiology 33, 2188–2191. Geldreich, E.E. 1997 Coliforms: a new beginning to an old problem. In Coliforms and E. coli: Problem or Solution, eds. Kay, D. & Fricker, C. pp. 3–11. Cambridge: The Royal Society of Chemistry. ISBN 0-85404-771-9. Grabow, W.O.K., Middendorff, I.G. & Prozesky, O.W. 1973 Survival in maturation farms of coliform bacteria with transferable drug-resistance. Water Research 7, 1589–1597. Grabow, W.O.K., Van Zyl, M. & Prozesky, O.W. 1976 Behaviour in conventional sewage purification processes of coliform bacteria with transferable or non-transferrable drug-resistance. Water Research 10, 717–723. Jeyasekaran, G., Karunasagar, I. & Karunasagar, I. 1990 Validity of faecal coliform test in tropical fishery products. Proceedings of the Second Indian fisheries Forum 27–31. Kumar, H.S., Otta, S.K., Karunasagar, I. & Karunasagar, I. 2001 Detection of Shiga-toxigenic Escherichia coli (STEC) in fresh seafood and meat marketed in Mangalore, India by PCR. Letters in Applied Microbiology 33, 334–338. Levy, S.B. 1998 The challenge of antibiotic resistance. Scientific American 278, 46–53. McGeer, A. 1998. Agricultural use of antibiotics and resistance in human pathogens: villain or scapegoat? Canadian Medical Association Journal 159, 1129–1136. Nataro, J.P. & Kaper, J.B. 1998 Diarrheagenic Escherichia coli. Clinical Microbiology Reviews 11, 142–202. Office International des Epizoties (OIE). (1999, March). The use of antibiotics in animals – ensuring the protection of public health. Proceedings of the Scientific Conference, Paris, France. Olsvik, O. & Strockbine, N.A. 1990 PCR detection of heat-stable, heat-labile and shiga-like toxin genes in Escherichia coli. In Diagnostic microbiology: Principles and Applications, eds. Persing, D.H., Smith, T.F., Tenover, F.C. & White, T. J. pp. 271–276. Washington: American Society for Microbiology. ISBN 1-55581-056-X. Pal, A., Ghosh, S., Ramamurthy, T., Yamasaki, S., Tsukamoto, T., Bhattacharya, S.K., Nair, G.B. & Takeda, Y. 1999 Shiga toxinproducing Escherichia coli from healthy cattle in a semi-urban community in Calcutta, India. Indian Journal of Medical Research 110, 83–85. Paton, A.W. & Paton, J.C. 1998 Detection and characterisation of shiga toxigenic Escherichia coli by using multiplex PCR assays for stx1, stx2, eaeA, enterohemorrhagic E. coli hlyA, rfbO111 and rfbO157. Journal of Clinical Microbiology 36, 598–602. Sambrook, J., Fritsch, E.F. & Maniatis, T. 1989 Molecular Cloning: A Laboratory Manual. New York: Cold Spring Harbor Laboratory Press. ISBN 0-87969-309-6. Van den Bogaard, A.E. & Stobberingh, E.S. 1999 Antibiotic usage in animals: impact on bacteria resistance and public health. Drugs 58, 589–609. Williams, R.J. & Heymann, D.L. 1998 Containment of antibiotic resistance. Science 279, 1153–1154. Witte, W. 1998 Medical consequences of antibiotic use in agriculture. Science 279, 996–997.

 Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:625–628 DOI 10.1007/s11274-004-3553-x

Rapid detection of Salmonella typhi in foods by combination of immunomagnetic separation and polymerase chain reaction S. Kumar*, K. Balakrishna, G.P. Singh and H.V. Batra Division of Microbiology, Defence R&D Establishment, Jhansi Road, Gwalior 474 002, India *Author for correspondence: Tel.: 91-751-2340245/354, Fax: 91-751-2341148, E-mail: [email protected]

Keywords: Immunomagnetic separation, IMS-PCR, meat, milk, polymerase chain reaction, Salmonella

Summary A combination of immunomagnetic separation and polymerase chain reaction (IMS-PCR) was used to detect Salmonella typhi in food and water samples. IMS was found to be an effective method for specific capture of S. typhi from artificially inoculated meat rinse samples. The bacteria could be detected within 6 h by IMS-PCR with a sensitivity of 105 cells. However, when tested in milk samples, the method was less effective. In comparison to conventional culture method, IMS-PCR is a rapid and specific method for detection of S. typhi and could be useful in outbreak situations for tracing the source of infection.

Introduction Salmonella typhi is the primary aetiological agent of typhoid fever which is responsible for significant morbidity and mortality, particularly in the developing countries. The incidence of typhoid fever has been estimated to be between 17 and 33 million cases annually with 600,000 associated deaths (Pang 1998; Hoiseth 2000). S. typhi is an obligate human pathogen. It causes infection by the faecal-oral route. Typhoid fever is typically acquired by ingesting food or water that has been contaminated by faeces of typhoid-infected individuals. Many outbreaks, caused either by consumption of contaminated food (Usera et al. 1993; Cote et al. 1995) or water (Nishina et al. 1989; Mermin et al. 1999) have been reported. The standard procedure for isolation of S. typhi from food, water or any other environmental sample usually requires recovery by three steps: pre-enrichment, enrichment and selective plating that take minimum of 3 days and several additional days for confirmation of presumptive positive results. Polymerase chain reaction (PCR) is a sensitive and rapid technique. But it can be inhibited by several factors present in food and other complex environmental samples (Lantz et al. 1994). The removal of these inhibitory substances like food components, humic acid, bile salts and lipids etc. is critical step in the preparation of samples of template DNA for PCR-based detection of food pathogens. Immunomagnetic separation (IMS) is a powerful technique that can specifically capture the target bacteria from food samples for template DNA preparation with little or no non-specific bacteria or interfering factors. The present

work aimed at the concentration of S. typhi bacteria by IMS from artificially inoculated food and water samples and their subsequent detection by PCR.

Materials and methods Preparation of antibodies to flagellin Salmonella typhi (strain SKST), obtained from the Christian Medical College, Vellore, India was grown in nutrient broth or LB broth (Difco) for 18–24 h at 37
C. Flagellin was purified as described by Ibrahim et al. (1985). Antiserum to flagellin was raised in New Zealand white rabbits (Kumar et al. 2003); which gave weak cross-reaction with S. weltevreden, S. infantis and S. paratyphi A. The antiserum was made monospecific by adsorbing 1 ml antiserum with a mixture of cultures obtained by growing each cross-reacting serotype on two nutrient agar plates. Adsorption was carried out at 50
C for 1 h. The antiserum was separated by centrifugation at 10,000 · g for 15 min and checked for cross-reactivity again. Adsorption was repeated against cross-reactive antigens. Immunoglobulins G (IgG) were purified from the adsorbed antiserum by gel exclusion chromatography using Bio-Rad matrix Bio A1.5 (henceforth called polyclonal antibodies). Coating of magnetic beads Epoxy superparamagnetic beads (Dynal, Oslo, 2.8 lm dia.) were coated with anti-flagellar polyclonal antibodies (PoAb) as per manufacturer’s instructions. Briefly,

626 5 mg beads were washed twice in 0.1 M phosphate buffer (PB), pH 7.2, before resuspending in 100 ll of same buffer. Antiflagellar PoAb (5 ll, 20 mg ml)1) was mixed with the beads along with 95 ll PB. This was followed by the addition of 100ll of 3M ammonium sulphate solution. The mixture was mixed in mixer (Dynal, Oslo) overnight at room temperature. The beads were washed 4 times with PBST (0.1 M phosphatebuffered saline, pH 7.2, 0.5% Tween 20) and were stored in 300 ll of PB with 0.1% BSA (w/v) at 4
C.

S. Kumar et al. use. Water samples were collected from laboratory and household supply. 10 ml each of food or water sample were inoculated with 100 ll of S. typhi culture to achieve a concentration of 107–103 c.f.u. ml)1. An uninoculated control was prepared by seeding samples with 100 ll of BPW. After inoculation each sample was incubated for 10 min at room temperature. The inoculated sample was added in 90 ml of BPW and mixed well. One millilitre sample was taken and processed for DNA extraction by boiling after centrifugation or IMS. The DNA was used as template in PCR assay.

Polymerase chain reaction and IMS-PCR A pair of oligonucleotide primers targeting the fliC-d gene was used to detect S. typhi. The primers (Forward 5¢ TGG GCG ACG ATT TCT ATG CC 3¢ and Reverse 5¢ TTT GCG GAA CCT GGT TCG CC 3¢) are reported to pick H=d positive salmonellae specifically (Chaudhry et al. 1997). PCR was carried out in 25 ll reaction containing 30 pmol of each primer, 200 lM of each dNTP, 0.75 units of Taq polymerase, 1.5 mM MgCl2 in 1 · PCR buffer (MBI Fermentas). PCR was taken through 30 cycles of 94
C for 1 min (denaturation), 57
C for 1 min (annealing), and 72
C for 1 min (extension). The DNA was denatured for 4 min in the beginning and finally extended for 5 min at 72
C. PCR products were analysed on 1.0% (w/v) agarose gel. To determine the detection sensitivity of PCR for pure culture, DNA was prepared from serial 10-fold dilutions (ca. 107–102 c.f.u. ml)1) of overnight growth. Each dilution (1 ml) was centrifuged at 8000 · g for 10 min. The cell pellet thus obtained was resuspended in 50 ll of distilled water (DW) and boiled for 10 min. This was immediately placed on ice for 5 min. Particulate material was removed by centrifugation at 12,000 · g for 5 min. The lysate was removed and 2.5 ll was used as the template in PCR assay immediately or following storage at )20
C. Viable count was determined by plate count. S. typhi bacteria were captured with immunomagnetic beads from the second set of dilutions. Coated beads (10 ll, 107 beads) were added to each dilution (1 ml). After mixing for 1 h in Dynal mixer, the beads were washed twice with PBST followed by concentration by magnet (MPL1, Dynal). The beads were suspended in 50 ll of DW and boiled for 10 min. The rest of the procedure was same as described above. Supernatant was separated and used as template DNA. In mixed culture experiments, E. coli MTCC 732 or Salmonella paratyphi A (clinical isolate) were added to the above mentioned dilutions of S. typhi and the process repeated. Artificial inoculation of food and water samples Samples of milk, meat and vegetables were procured from the local market. Meat and green leafy vegetable rinses were prepared by suspending 150 g of each food sample in 150 ml of buffered peptone water (BPW) and rinsing thoroughly with the medium. Aliquots of 10 ml food rinse were made and stored at )20
C for further

Results and discussion Typhoid fever continues to be a major problem of the developing countries. S. typhi, the causative agent is transmitted by consumption of contaminated food or water. In the event of an outbreak, it is essential to locate the source of infection in order to take timely control measures. Owing to the complex nature of food and other environmental matrices, the detection of any bacteria including Salmonella from these matrices, is a difficult task. The procedure of Salmonella isolation and identification generally takes 3–5 days. The sensitivity and rapidity of PCR as a detection method is unquestionable. The only problem it suffers from is that high quality of DNA is required. In this study, PCR generated the expected product of 489 bp. The detection sensitivity of PCR assay was found to be 1.0 · 103 c.f.u. When bacteria were captured by IMS for subsequent detection of extracted DNA by PCR, there was a visible decrease in the intensity of the band at this concentration; however, there was no apparent change in the sensitivity (Figure 1a). The low intensity of the band may be due to the fact that recovery of bacteria with beads is never 100% and there is additional loss of bacteria during washing steps. We observed by plate counts that at least 40% of the bacteria are captured by the beads and another 6–8% are lost during each washing step (unpublished data). It was further observed that the presence of E. coli or S. paratyphi A does not make any difference in detection sensitivity of IMS-PCR (Figure 1b). We also found that two washings can be used during IMS procedure. Three washing were reported to give extremely low level of recoveries and were found unsuitable for detection of Listeria monocytogenes by IMS-PCR (Hudson et al. 2001). The results of IMS-PCR in spiked food samples are shown in Table 1. The meat and milk samples tested in this study had a background bacterial count of about 106 and 108 c.f.u. ml)1 respectively. The background bacterial count in vegetable rinse was however much lower, of the order of 102 c.f.u. ml)1. BPW was added to the spiked samples to dilute the PCR inhibitory substances. PCR performed with DNA samples extracted by direct boiling of spiked meat rinse samples after centrifugation always gave unequivocal results. On most

IMS-PCR for S. typhi

627

Figure 1. PCR amplification of the fliC-d gene of S. typhi bacteria in pure culture (panel a) or mixed culture (panel b). DNA was extracted from the cells ca. 106 (A: lanes 1, 4), 105 (a: lanes 2, 5) or 104 (a: lanes 3, 6) by boiling after centrifugation (a: lanes 1–3,7) or IMS (a: lanes 4–6; b: lanes 1–8). S. typhi bacteria ca. 108 or 105 mixed with (b: lanes 2 & 6, 4 & 8) or without (b: lanes 1 & 5, 3 & 7) equal number of E. coli bacteria were processed during IMS by one (b: lanes 1–4) or two (b: lanes 5–8).washing steps. Positive control and 100 bp ladder are shown in lane 7 and 8 in panel a.

Table 1. Detection of S. typhi in various food and water samples by IMS-PCR. Samples

Meat rinse Milk Vegetable rinse Water

Inoculation level c.f.u. ml)1 number tested positive/ total tested 2.5 · 107

2.5 · 106

2.5 · 105

2.5 · 104

2.5 · 103

5/5 4/5 5/5

4/5 3/5 5/5

0/5 0/5 5/5

0/5 0/5 3/5

0/5 0/5 0/5

5/5

5/5

5/5

4/5

0/5

occasions, a smear was observed. The IMS method was able to capture the S. typhi bacteria in sufficient purity so as to enable their subsequent detection by PCR; however there was a drop in detection sensitivity by two logs when compared to pure culture. But the advantage of IMS-PCR is that it took only about 5 h to complete the assay and could be used directly on meat rinse samples. IMS was partially successful with milk samples. Only three out of five samples were positive ca. 106 c.f.u. ml)1. It is possible that the magnetic beads got entrapped in the fat particles present in the milk. Fats and other substances have been reported to interfere with the process of DNA extraction from milk samples (Ramesh et al. 2002). None of the milk samples tested positive by direct PCR. As calcium ions present in milk may inhibit the PCR reaction, we tried different concentrations of MgCl2 (1.5, 2.0 and 2.5 mM) in PCR reaction without any success. All the tested vegetable rinse and water samples tested positive both by direct PCR and IMS-PCR to same level of detection sensitivity (1.2 · 103c.f.u.). Therefore, there was no added advantage of IMS over direct extraction of DNA in these two samples. This is possibly due to the simple nature of the matrix of water and vegetable rinse samples that may not have much PCR-interfering material.

Superparamagnetic beads coated with specific antibodies specifically capture the target bacteria from complex food samples with the background of inhibitory substances and competing bacteria. The utility of immunomagnetic separation for subsequent culture or PCR has been shown for capture of Listeria monocytogenes, E. coli and other bacteria (Hudson et al. 2001; Kerr et al. 2001). In this study, IMS led to a drop in sensitivity of order of 1–2 log from food samples. We believe that this would still not affect the utility of IMSPCR as a rapid detection method in identifying the source of infection in disease outbreak situation. However, the detection sensitivity of IMS-PCR, most likely can be increased by short pre-enrichment of the captured bacteria.

Acknowledgement Authors are thankful to Director, DRDE for providing the necessary facilities and encouragement.

References Chaudhry, R., Laxmi, B.V., Nisar, N., Ray, K. & Kumar, D. 1997 Standardisation of polymerase chain reaction for detection of Salmonella typhi in typhoid fever. Journal of Clinical Pathology 50, 437–439. Cote, T.R., Convery, H., Robinson, D., Ries, A., Barrett, T., Frank, L., Furlong, W., Horan, J. & Dwyer, D. 1995 Typhoid fever in the park: epidemiology of an outbreak at a cultural interface. Journal of Community Health 20, 451–458. Hoiseth, S.K. 2000 Vaccines, Bacterial. In Encyclopedia of Microbiology, 2nd edn., Vol 4, ed. Lederberg, J. pp 767–778. New York: Academic Press. ISBN 0-12-226800-8. Hudson, J.A., Lake, R.J., Savill, M.G., Scholes, P. & McCormick, R.E. 2001 Rapid detection of Listeria monocytogenes in ham samples using immunomagnetic separation followed by polymerase chain reaction. Journal of Applied Microbiology 90, 614–621.

628 Ibrahim, G.F., Fleet, G.H., Lyons, M.J. & Walker, R.A. 1985 Method for isolation of highly purified Salmonella flagellins. Journal of Clinical Microbiology 22, 1040–1044. Kerr, P., Finlay, D., Thomson-Carter, F. & Ball, H.J. 2001 A comparison of monoclonal antibody-based sandwich ELISA and immunomagnetic bead selective enrichment for the detection of Escherichia coli O157 from bovine faeces. Journal of Applied Microbiology 91, 1–4. Kumar, S., Balakrishna, K., Tuteja, U. & Batra, H.V. 2003 Application of monoclonal antibodies to flagellin of Salmonella typhi for its detection in foods. Indian Journal of Microbiology 43, 193–197. Lantz, P.G., Hahn-Hagerdal, B. & Radstrom, P. 1994 Sample preparation methods in PCR detection of Food pathogens. Trends in Food Science and Technology 5, 384–389. Mermin, J.H., Villar, R., Carpenter, J., Roberts, L., Samaridden, A., Gasanova, L., Lomakina, S., Bopp, C., Hutwagner, L., Mead, P., Ross, B. & Mintz, E.D. 1999 A massive epidemic of

S. Kumar et al. multidrug-resistant typhoid fever in Tajikistan associated with consumption of municipal water. Journal of Infectious Diseases 179, 1416–1422. Nishina, T., Shiozawa, K., Hayashi, M., Akiyama, M., Sahara, K., Miwa, N., Nakatsugawa, S., Murakami, M. & Nakamura, A. 1989 Supply system in Fuji city. Kansenshogaku Zasshi 63, 240–247. Pang, T. 1998 Genetic dynamics of Salmonella typhi – diversity in clonality. Trends in Microbiology 6, 339–342. Ramesh, A., Padmapriya, B.P., Chandrashekar, A. & Varadaraj, M.C. 2002 Application of convenient DNA extraction method and mutiplex PCR for the direct detection of Staphylococcus aureus and Yersinia enterocolitica in milk samples. Molecular and Cellular Probes 16, 307–314. Usera, M.A., Aladuena, A., Echeita, A., Amor, E., Gomez-Garces, J.L., Ibanez, C., Mendez, I., Sanz, J.C. & Lopez-Brea, M. 1993 Investigation of an outbreak of Salmonella typhi in a public school in Madrid. European Journal of Epidemiology 9, 251–54.

Ó Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:629–634 DOI 10.1007/s11274-004-3566-5

DNA amplification fingerprinting as a tool for checking genetic purity of strains in the cyanobacterial inoculum B. Jeberlin Prabina, K. Kumar* and S. Kannaiyan Department of Agricultural Microbiology, Tamil Nadu Agricultural University, Coimbatore 641 003, Tamil Nadu, India *Author for correspondence: Tel.:+91-0422-2431222 ext. 303, E-mail:azollakumar@rediffmail.com

Keywords: Cyanobacteria, cluster analysis, genetic purity, RAPD–PCR profile, strain identification

Summary The electrophoretic patterns for 17 different cyanobacterial cultures derived from 6 different decamer primers were analysed to provide diagnostic fingerprints for each culture and their genetic distances based on RAPD markers.The primer OPB 09 produced a maximum of 24 amplified products. The primers OPB 09, OPG 04 and OPAH 02 generated markers specific for Nostoc cultures. Westiellopsis was found to be distinct from other cyanobacterial cultures in the RAPD profile obtained with the primer OPAH 02. The primer OPF 03 generated specific markers for Tolypothrix tenuis. Fischerella cultures could be identified with the primers OPB 09, OPAG 03 and OPF 05. The study revealed that these RAPD markers could be further used to identify and establish the genetic purity of the strains in the cyanobacterial inoculum. There was a similarity of 60–90% within Westiellopsis cultures. Nostoc cultures shared 50–80% similarity with Westiellopsis cultures. Anabaena cultures were similar to Westiellopsis cultures by 60–70%. The markers produced for each culture were also applied to phylogenetic analysis to infer genetic relatedness in this group of prokaryotes. The dendrogram analysis clearly revealed that free-living cyanobacterial cultures are closely related to each other and are distinct from the symbiotic forms.

Introduction Cyanobacteria are unique among the prokaryotes due to their capacity for oxygenic photosynthesis. An important feature of many cyanobacteria is their ability to fix atmospheric nitrogen both under free-living and symbiotic conditions. The species of cyanobacteria which are known to fix atmospheric nitrogen are classified into three groups viz., heterocystous-aerobic forms, aerobic unicellular forms and non-heterocystous filamentous microaerophilic forms. Nitrogen-fixing cyanobacteria can use sunlight as the sole energy source for the fixation of carbon and nitrogen and therefore have potential as biofertilizers (Kannaiyan 1985). The composite inoculum consisting of cyanobacterial cultures viz., Nostoc, Anabaena, Calothrix, Tolypothrix, Plectonema, Aphanotheca, Gleocapsa, Cylindrospermum, Oscillatoria, Aulosira and Scytonema has been used for inoculation in rice (Kannaiyan 1993). Mass production of high quality inoculum of the desirable strains is an important stage when considering the inoculation of nitrogen-fixing cyanobacteria as biofertilizer for rice crops (Watanabe & Roger 1984). The biofertilizer potential of cyanobacteria is limited, mainly due to the non-availability of good quality inoculants and the difficulty in re-establishing the

inoculated strains varied in the rice field (Roger & Watanabe 1986). Rapid and reliable quality control methods have not yet been developed for cyanobacterial biofertilizers. Current cyanobacterial taxonomy is based primarily on observed morphological characteristics, which is often confusing and does not identify potentially different strains. More importantly, the morphology of cyanobacteria in laboratory cultures is often considerably altered from the original morphology of environmental isolates and the diversity of strains within a culture may be reduced because of selective culturing conditions (Doers & Parker 1998). Analyses of photosynthetic pigment content, isoenzyme variation or differentiated cell culture may also be misleading because of the variable expression of cyanobacterial gene products in culture (Rippka et al. 1979; Kato et al. 1991). Taxonomic characters change so drastically that reliable identification of species becomes difficult or even impossible (Kumar et al. 2000). A number of new valuable molecular biological tools for taxonomic purpose have been developed which could be effectively utilized for checking the genetic purity of strains in the cyanobacterial inoculants. Nelissen et al. (1996) developed a cyanobacterialspecific oligonucleotide which could identify the PCR products from 16S rRNA after separation by agarose

630 gel electrophoresis. DNA/DNA hybridization has been used successfully to resolve taxonomic problems at the generic and species level in cyanobacteria (Kelly & Cowie 1972). Restriction Fragment Length Polymorphisms have been used extensively as an efficient DNA fingerprinting method to identify symbiotic cyanobacteria (Nierzwicki-Bauer & Haselkorn 1986). Polymerase Chain Reaction (PCR) technology had a significant impact in almost all the areas of molecular biology and the modification of this basic procedure has allowed a number of assays for detecting variation at the nucleotide level. Among them, Random Amplified Polymorphic DNA (RAPD) developed by Williams et al. (1990) has been widely used in various fields of molecular biology. The presence of large numbers of decamer sequences in a large genome forms the basis of this technique, which utilizes random primers so that just by chance at several unpredictable locations, two primers will anneal sufficiently close to one another on opposite strands of the template to amplify the intervening regions. The RAPD technique in conjunction with PCR has been employed to identify many organisms to the strain level (Welsh & McClelland 1990). This technique is sensitive and specific because the entire genome of an organism is used as the basis for generating the DNA profile. Eskew et al. (1993) used DNA fingerprinting to differentiate the isolates in the Azolla–Anabaena symbiosis. In this investigation, our objective was to develop an easy and reliable method for checking the genetic purity of strains in the cyanobacterial inoculum using RAPD analysis. We have demonstrated that a PCR fingerprinting method could be used as a genetic tool for identification of species and strains analysis of genetic relatedness of cyanobacteria. The method was found to be accurate in distinguishing and classifying even closely related strains.

Materials and methods Cyanobacterial cultures and culture conditions The cyanobacterial cultures used in this study were collected from different biofertilizer production centres of India. The cultures Anabaena azollae-SK-SL-TNAU1, A. azollae-MPK-SK-AM-24, A. azollae-MPK-SK-AM-25, A. azollae-MPK-SK-AF-38, Anabaena-TR 52-ST 1, Anabaena sp., Nostoc muscorum –DOH, Westiellopsis-4A2, Westiellopsis–C100-TR5-ST3-PA-SK, Anabaena variabilis-SA0, Nostoc sp. and Westiellopsis sp. were selected from the Germplasm collections at the Azolla Laboratory, Department of Agricultural Microbiology, Tamil Nadu Agricultural University (TNAU), Coimbatore, while Westiellopsis–ARM48 was obtained from the Department of Microbiology, PSG College of Arts and Science, Coimbatore and Aulosira pseudoramosa, Tolypothrix tenuis, Westiellopsis prolifica and Fischerella sp. were collected from the National Centre for Conservation and Utilization of BGA (NCCUBGA), Indian Agricul-

B.J. Prabina et al. tural Research Institute (IARI), New Delhi. All the cyanobacterial cultures were grown in BG-11 medium in conical flasks. The flasks were incubated under controlled conditions in culture room with a light intensity of 3000 lux (16 h/8 h day and night cycle) and a temperature of 25 ± 1 °C. Thirty days old cultures were used for the study. Extraction of total genomic DNA for RAPD analysis The total genomic DNA from the cyanobacterial cultures was isolated by the 2% cetyl trimethyl ammonium bromide (CTAB) procedure (Clark 1997) with little modification. Approximately 2 g of blot-dried cyanobacterial culture was ground in an autoclaved and prechilled mortar using liquid nitrogen. The macerated samples along with 2 ml of CTAB buffer were transferred to eppendorf tubes and incubated at 65 °C for 30 min. Protein was removed by treating with equal volumes of a 1:1 chloroform-isoamyl alcohol mixture. DNA was precipitated with the addition of ice-cold isopropanol. The precipitated DNA was hooked out with a bent Pasteur pipette and dissolved in 500 ll of 10 mM Tris–1 mM EDTA buffer. Then the DNA was concentrated by precipitation with ethanol and sodium acetate. The pellet was dissolved in 50 ll of Tris-EDTA buffer and stored at )20 °C for further use. Oligonucleotide primers and PCR amplification The primers were obtained from OPERON Techniques Inc., USA. The DNA amplification was performed in a PTC-100 TM Programmable Thermocycler (MJ Research Inc., Water Town, Mass). The reaction mixture consisted of 3 ll of 25 ng ll)1 DNA, 1 l1 of 10 mM dNTPs, 1.2 ll primer, 2 ll of 10 x Assay buffer, 0.2 ll Taq polymerase (Bangalore Genei Inc.,), 12.6 ll sterile distilled water. The incubation cycle was initial denaturation (95 °C for 2 min), denaturation (94 °C for 1 min), annealing (37 °C for 1 min), extension (72 °C for 2 min) and final extension (72 °C for 5 min). Thirty five incubation cycles were followed for denaturation, annealing and extension steps. After the reaction, 15 ll of the PCR product was separated by 1.5% agarose gels, stained with ethidium bromide, viewed and photographed using alpha imager TM 1200 documentation and analysis system. Dendrogram analysis Each band visualized on a gel was considered a RAPD marker and part of the total RAPD fingerprint generated for a particular genera. Bands bisected by similar perpendicular lines drawn across the gel were considered homologous characteristics. The assumption that these bands contained homologous primer recognition sequences and identical intervening sequence was made. Therefore, the presence or absence of a band at any position on the gel was used to construct a binary

631

Checking genetic purity of strains matrix for cyanobacterial RAPD markers. Each lane of the PCR product for different samples with different primers was scored and cluster analysis was carried out using NTSYS pc version 1.7 (Rohlf 1992).

Results and discussion RAPD–PCR analysis of cyanobacterial cultures A total of 12 primers were initially chosen to generate RAPD patterns for 17 different cyanobacterial cultures. The criteria for choosing these primers were their availability and the generally accepted bias towards oligonucleotides of high G+C content. All the 12 primers were tried and 6 primers (Table 1) were found to produce informative and reproducible genetic markers for different cyanobacterial cultures. Total genomic DNA extracted from the 17 cyanobacterial cultures were used as templates. The primer OPB 09 produced totally 24 amplified DNA fragments of size ranging from 9000 bp to 125 bp (Figure 1). There was one monomorphic band and 23 polymorphic bands. The percentage of polymorphism observed was 95.8% (Table 2). An amplified fragment of molecular size 1636 bp was unique for Fischerella sp. The primer OPG 04 produced a unique fingerprinting for all the cyanobacterial cultures (Figure 2). The primer amplified a DNA fragment of molecular size 8144 bp exclusively in Westiellopsis-C100-TR5-ST3-PA-SK. Nostoc muscorum had an amplified fragment of size around 3054 bp. A band of molecular size 5090 bp was found in Anabaena sp. and Anabaena TR 52 ST1. The primer OPAH 02 generated distinct pattern for Westiellopsis cultures and it could be used to distinguish Westiellopsis from other cultures (Figure 3). The primer OPF 03 (Figure 4) produced a total of 12 amplified DNA Table 1. Primers used for RAPD–PCR analysis of cyanobacterial cultures. Primer

Sequence

%G+C

OPAH 02 OPAG 03 OPB 09 OPG 04 OPF 05 OPF 03

CACTTCCGCT TGCGGGAGTG TGGGGGACTC AGCGTGTCTG CCGAATTCCC CCTGATCACC

60 70 70 60 60 60

Figure 1. RAPD profile of cyanobacterial cultures for primer OPB 09. See Table 3 for key to the numbers.

fragments of molecular size ranging from 7000 bp to 1000 bp. It produced a single band of molecular size around 6800 bp for Tolypothrix tenuis. Considerable variation in banding pattern was observed between and among the different genera of cyanobacteria investigated. The total number of amplified DNA fragments produced by the primer OPAG 03 was 16 (Figure 5). Among the Westiellopsis cultures, the fragment separation was different for the acid tolerant culture Westiellopsis-4A2. This primer differentiated Fischerella sp. from the other cyanobacterial cultures by the presence of a single fragment of size around 4500 bp. The primer OPF 05 produced a DNA fragment of molecular size around 3052 bp in most of the cultures (Figure 6). For Fischerella sp. it produced two bands of size 7000 bp and 1636 bp, whereas in A. pseudoramosa, it produced a distinct pattern at 6000 bp which was not found in other cultures. RAPD–PCR was used to generate unique and identifying DNA profiles for members of the cyanobacterial genera Anabaena and Microcystis by Neilan (1995). Rasmussen & Svenning (1998) identified the symbiotic and free-living cyanobacterial cultures using RAPD technique. Also it was used for discriminating genotypes of Microcystis (Nishihara et al. 1997). The results of the present study indicated the potential use of RAPD markers as a rapid method to detect genetic variation and the genetic relatedness of the cyanobacterial strains at the level of DNA. The primers OPB 09, OPG 04 and OPAH 02 generated markers specific for Nostoc cultures. Westiellopsis was found to be distinct from other cyanobacterial cultures in the RAPD profile obtained with the primer OPAH 02. The primer OPF 03 generated specific markers for Tolypothrix tenuis. The primers OPB 09, OPAG 03 and OPF 05 could be used for the identification of Fischerella cultures.

Table 2. Number of PCR products generated in RAPD analysis of cyanobacterial cultures with different primers. Primers

No. of monomorphic bands

No. of polymorphic bands

Total no. of bands

Percentage of polymorphism

OPAH 02 OPAG 03 OPB 09 OPG 04 OPF 05 OPF 03

0 0 1 0 0 0

13 16 23 18 13 12

13 16 24 18 13 12

100 100 95.8 100 100 100

632

Figure 2. RAPD profile of cyanobacterial cultures for primer OPG 04. See Table 3 for key to the numbers.

B.J. Prabina et al.

Figure 6. RAPD profile of cyanobacterial cultures for primer OPF 05. See Table 3 for key to the numbers.

required for RFLP-Southern hybridization experiments which makes it more advantageous over traditional methods. The RAPD markers developed may be used as molecular markers for identification of the standard strains used in the preparation of cyanobacterial inoculum which will serve as a valuable tool for checking their genetic purity in order to ensure the supply of quality cyanobacterial inoculants to the farmers at an economic rate. Figure 3. RAPD profile of cyanobacterial cultures for primer OPAH 02. See Table 3 for key to the numbers.

Figure 4. RAPD profile of cyanobacterial cultures for primer OPF 03. See Table 3 for key to the numbers.

Phylogenetic analysis Genetic relatedness among the cyanobacterial genera was determined with banding pattern from the RAPD reaction. The dendrogram analysis of the seventeen cyanobacterial cultures revealed three major clusters (Figure 7). In the first major cluster, all the Westiellopsis cultures are clustered together and show similarity of 71%. The relatedness between different Anabaena sp. ranged between 74.0 and 87.4%. Nostoc muscorum and Nostoc sp. are related by 85.5%. Neilan, 1995 determined the genetic diversity among strains of Anabaena and Microcystis sp. With banding patterns from the multiplex RAPD reaction. The multiplexing of primers CRA 22(CCGCAGCCAA) and CRA 23 (GCGATCCCCA) generated a total of 33 RAPD markers and 29 RAPD markers when analysed on

Figure 5. RAPD profile of cyanobacterial cultures for primer OPAG 03. See Table 3 for key to the numbers.

It should be noted that RAPD profiles may be altered by the presence of transposable elements and plasmids in prokaryotes as well as by the contaminating bacteria in a sample. The RAPD technique does not require previous knowledge of an organism’s gene sequences and requires only 1/1000 of the amount of DNA

Figure 7. Phylogenetic relationship between the cyanobacterial cultures by RAPD analysis.

633

Checking genetic purity of strains Table 3. Relevant characteristics and specific RAPD marker for the cyanobacterial cultures. Sl.No. Cyanobacterial cultures used

Relevant characteristics

Specific RAPD marker

1.

Westiellopsis sp.

The cells are dark green in colour and are larger and compact compared to other cultures. The vegetative cells are cylindrical and true branching is present.

1.2 kb&0.7 kb-OPAH02 4.00 kb-OPB09

2. 4. 6.

Westiellopsis-C100 TR5ST3PA-SK Westiellopsis 4A2 Nostoc muscorum

As 1 Saline-tolerant culture. As 1 Acid-tolerant culture. The trichomes are non motile, wavy and are tightly coiled. Vegetative cells are ovoid and are distinct. Heterocysts intercalary or terminal. Constrictions at the cross - walls are conspicuous.

8.14 kb-OPG04 1.2 kb&0.7 kb-OPAH02 3.05 kb-OPG04

7.

Nostoc sp.

As 6.

–

8.

Anabaena variabilis

9.

A. azollae-SK-SL-TNAU 1

Trichomes are straight. Vegetative cells are spherical, 5.00 kb-OPB09 barrel shaped and separated by conspicuous constrictions at the cross wall. Heterocysts are intercalary or terminal. As 7. 5.00 kb-OPB09

10.

A. azollae MPK-SK-AM-24

As 7.

5.00 kb-OPB09

11.

A. azollae MPK-SK-AM-25

As 7.

5.00 kb-OPB09

12.

A. azollae MPK-SK-AF-38

As 7.

5.00 kb-OPB09

13.

Anabaena sp.

As 7.

5.09 kb-OPGO4

14.

Anabaena-TR52ST1

As 7.

5.09 kb-OPGO4

15.

Tolypothrix tenuis

Basal to apical tapering degree is low. Cells shorter at terminal end. Heterocysts are terminal. The cells are short, cylindrical and filamentous. The cells are enclosed in a sheath.

0.56 kb-OPAH02 6.8 kb-OPF03

16.

Aulosira pseudoramosa

The cells are cylindrical. The heterocysts are elongated and are intercalary or terminal. Trichomes are straight and filamentous.

6.00 kb-OPF05

17.

Fischerella sp.

The cells are spherical and thick walled. True branching is present. Heterocysts are elongate, spherical and are terminal or intercalary.

7.00 kb, 1.63 kb-OPF05 1.63 kb-OPB09

agarose gels. The phenogram clearly delineated the genera Anabaena and Microcystis. Giovannoni et al. (1988) studied the phylogeny of fresh-water cyanobacteria obtained with 16S rRNA gene sequences. It has been shown that as few as three primers used separately provided enough polymorphic information to identify species of the symbiotic genes of Anabaena and to create a phylogenetic tree with a topology similar to that derived with 22 primers (Neilan 1995). The result from the present study also analysed the genetic relatedness of different cyanobacterial genera with 6 primers which revealed that free-living cyanobacteria are closely related to each other and are distinct from symbiotic forms.

Acknowledgement The authors are grateful to the Indian Council of Agricultural Research, New Delhi for the financial support to carry out this study under National Agricultural Technology Project-Team of Excellence on Biofertilizers for rice-based cropping systems.

References Clark, M.S. 1997 Plant Molecular Biology–A laboratory manual. New York: Springer-Verlag, ISBN 3-54058405-6. Doers, M.P. & Paker, D.L. 1988 Properties of Microcystis aeruginosa and M. flosaquae in culture: taxonomic implications. Journal of Phycology 24, 502–508. Eskew, D.L., Caetano-Anolles, G., Bassam, B.J. & Gresshoff, P.M. 1993 DNA amplification fingerprinting of the Azolla–Anabaena symbiosis. Plant Molecular Biology 21, 263–373. Giovannoni, S.J., Turner, S. Olsen, G.J., Barns, S., Lane D.J. & Pace, N.R. 1988 Evolutionary relationship among cyanobacteria and green chloroplasts. Journal of Bacteriology 170, 3584–3592. Kannaiyan, S. 1993 Biofertilizers for Rice. p.60. Coimbatore: Tamil Nadu Agricultural University, Tamil Nadu, India. Kannaiyan, S. 1985 Studies on the Algal Application for Low Land Rice Crop. p. 24. Coimbatore: Tamil Nadu Agricultural University, Tamil Nadu, India. Kato, T., Watanabe, M.F. & Watanabe, M. 1991 Allozyme divergence in Microcystis and its taxonomic inference. Archives of Hydrobiology Suppl. Bd. 64, 129–140. Kelly, M.L. & Cowie, D.B. 1972 DNA–DNA hybridization studies of blue-green algae. Carnegie Institute of Washington Yearbook 71, 276–287. Kumar, K., Annapoorna, N. & Kannaiyan, S. 2000 Determination of genetic purity of the strains in the cyanobacterial inoculants by RAPD-PCR techniques. Paper presented in the Asian Pacific Conference on Plant Tissue Culture and Agribiotechnology

634 (APCPTCA), National University of Singapore, Singapore, 229 pp (Abstr). Neilan, B.A. 1995 Identification and phylogenic analysis of toxigenic cyanobacteria by multiple randomly amplified polymorphic DNAPCR. Applied and Environmental Microbiology 61, 2286–2291. Nelissen, B., De Baere, R., Wilmotte, A. & De Wachter, R. 1996 Phylogenetic relationships of non-axenic filamentous cyanobacterial strains based on 16S rRNA sequence analysis. Journal of Molecular Evolution 42, 194–200. Nierzwicki-Bauer, S.A. & Haselkorn, R. 1986 Differences in mRNA levels in Anabaena living freely or in symbiotic association with Azolla. EMBO Journal 5, 29–35. Nishihara, H., Miwa, H., Watanabe, M., Nagashima,M., Yagi, O. & Takamura, Y. 1997 Random Amplified Polymorphic DNA (RAPD) analysis for discriminating genotypes of Microcystis cyanobacteria. Bioscience Biotechnology and Biochemistry 61, 1067–1072. Rasmussen, U. & Svenning, M.M. 1998 Fingerprinting of cyanobacteria based on PCR with primers derived from short and long tandemly repeated repetitive sequences. Applied and Environmental Microbiology 64, 265–272.

B.J. Prabina et al. Rippka, R., Deruelles, J.,Waterbury, J.B., Herdman, M. & Stanier, R.Y. 1979 Genetic assignments, strain histories and properties of pure culture of cyanobacteria. Journal of General Microbiology 111,1–61. Roger, P.A. & Watanabe, I. 1986 Technologies for utilizing biological nitrogen fixation in Wetland rice. Potentialities, current usage and limiting factors. In Nitrogen Economy in Flooded Rice Soils. Dutta, S.K. & Patrick, N.H Jr. eds. pp. 37–77. Dordrecht: Martinus Nijhoff Publishers, The Netherlands. ISBN 90-2473361-8. Rohlf, F.G. 1992 NTSYS-pc Numerical taxonomic and multivariate analysis system version 1.7 Owner manual. Watanabe, I. & Roger, P.A. 1984 Nitrogen fixation in wetland rice field. In Current Developments in Biological Nitrogen Fixation. Subba Rao, N.S. ed. pp. 237–236 New. Delhi: Oxford and IBH Publishing Co,. India ISBN 0-713112877-1. Welsh, J. & McClelland, M. 1990 Finger printing of genomes using PCR with arbitrary primers. Nucleic Acids Research 18, 7213– 7218. Williams, J.G.K., Kubleik, A.R., Livak, K.J., Raflaski, J.A. & Tingey, S.V. 1990 DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Research 18, 6531–6534.


Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:635–644 DOI 10.1007/s11274-004-3567-4

Utilization of a polyphasic approach in the taxonomic reassessment of antibioticand enzyme-producing Bacillus spp. isolated from the Philippines Marie Antonette Ruth V. Guerra-Cantera1 and Asuncion K. Raymundo2,* 1 National Institute of Molecular Biology and Biotechnology (BIOTECH), University of the Philippines Los Ban˜os, College, Laguna 4031, Philippines 2 Institute of Biological Sciences, UP Los Ban˜os, College, Laguna 4031, Philippines * Author for correspondence: Tel.: +6349-536-2893, Fax: +6349-536-2517, Email: [email protected].

Keywords: Bacillus, chemotaxonomy, molecular fingerprint, phenotype, taxonomy

Summary The taxonomy of 58 locally isolated antibiotic- and enzyme-producing Bacillus isolates deposited at the Philippine National Collection of Microorganisms (PNCM)-BIOTECH was reassessed in this study using a polyphasic approach since they had been only partially identified prior to deposition in the culture collection. The isolates had 41.1–69% G + C, and possessed the characteristic diaminopimelic acid (DAP) and fatty acid methyl ester (FAMEs) properties of Bacillus species. Molecular analysis using specific PCR primers differentiated the isolates into two major groups, the Bacillus cereus group and the Bacillus subtilis group. To further differentiate these isolates, they were subjected to 39 phenotypic tests. Using the dichotomous key constructed for Bacillus, 45 isolates maintained their original identities, five were named at the species level, and 12 were re-identified and renamed. These results showed that the classical phenotypic tests allowed the reclassification of the isolates, while modern techniques of chemotaxonomy and the molecular approach led only to genus and cluster classification and confirmation.

Introduction The genus Bacillus contains a systematically diverse assemblage of Gram-positive, aerobic or facultatively anaerobic spore-forming organisms (Claus & Berkeley 1986). Members of this genus exhibit a wide range of nutritional requirements, growth conditions, DNA base compositions (Slepecky & Hemphill 1991), and major amino acid compositions of the cell wall. In addition to phenotypic heterogeneity, they also appear to be phylogenetically diverse (Ash et al. 1991), with five phylogenetically distinct clusters emerging from a comparative analysis of the small subunit rRNA. For instance, the Bacillus cereus group includes the following species: Bacillus anthracis, Bacillus mycoides, Bacillus thuringiensis and Bacillus cereus (Priest et al. 1988). In contrast, five genotypic groups belonging to the Bacillus subtilis cluster were separated on the basis of Southern blot analyses done by Shida (2001), namely, Bacillus amyloliquefaciens, Bacillus mojavensis, Bacillus licheniformis, Bacillus subtilis subsp. subtilis, Bacillus subtilis spizizenii, Bacillus vallis mortis, Bacillus pumilus and Bacillus atrophaeus. To date, the genus Bacillus has been subjected to numerous taxonomic reclassifications resulting in the proposal of new genera and species (Niimura et al. 1990; Wisotzkey et al. 1992; Ash et al.

1993; Nakamura 1993; Takagi et al. 1993; Shida et al. 1994, 1995, 1996, 1997; Heyndrickx et al. 1996, 1998; Spring et al. 1996; Waino et al. 1999; Fortina et al. 2001; Nazina et al. 2001; Schlesner et al. 2001), using polyphasic identification. The locally isolated antibiotic- and enzyme-producing Bacillus isolates deposited at the Philippine National Collection of Microorganisms (PNCM), National Institute of Molecular Biology and Biotechnology (BIOTECH) had only been partially identified prior to deposition in the culture collection. With the current status of Bacillus taxonomy brought about by these taxonomic rearrangements and a continuous search for novel species, the taxonomy of these industrially important Bacillus isolates needed to be reassessed to give them a more valid description and nomenclature. Due to the high cost and difficulty in obtaining complete rRNA sequences from an adequate number of isolates to infer a reliable outline of phylogenetic relationships, a rapid and less costly technique such as molecular fingerprinting using specific primers must be used in classification and identification. This requirement is an especially important consideration in the developing countries, where taxonomic work must be done under tight budgetary constraints, without the state-of-the art equipment available in Western laboratories. This study

636 established the utility of the different approaches to Bacillus taxonomy and ascertained the appropriate taxonomic position of these local Bacillus isolates using a composite analysis of phenotypic, genotypic and chemotaxonomic characteristics.

Materials and methods Bacterial strains and culture conditions Fifty-eight local Bacillus isolates and seven reference strains of Bacillus were used (Table 1). The cultures were grown in Nutrient Agar (NA), Luria–Bertani (LB) Agar or Tryptone Glucose Yeast Extract Agar (TGYA) at 30 C. Active and glycerol stocks were maintained at 4 and )70 C, respectively. Chemotaxonomic characterization Determination of cellular fatty acid composition Approximately 40 mg of bacterial cells from 24-h old cultures grown in trypticase soy agar (TSA) at 28–30 C were harvested by centrifugation and placed in a sterile screw-capped tube. Fatty-acid methyl esters (FAMEs) were prepared as described by Sasser (2001). The resulting extracts were sprayed with nitrogen gas to avoid oxidation and sent to the Japan Collection of Microorganisms (JCM) in Saitama, Japan for gas chromatography (GC) analysis. Determination of the diaminopimelic acid (DAP) isomer composition of the cell wall DAP isomers were separated by TLC as described by Staneck & Roberts (1974). Standard DL-DAP (Sigma Chemical Co.) was used to confirm the respective Rf values. Genotypic characterization DNA preparation The genomic DNA of the isolates was extracted using the procedure as described by Ausubel et al. (1995). Determination of DNA base composition The DNA G+C content was determined by HPLC (Mesbah et al. 1989) using a Cosmosil 5C18 (4.6 mm · 150 mm) column. HPLC conditions included a flow rate of 1.0 ml/min at 40 C, with detection wavelength of 260 or 270 nm. Peak areas were calibrated using the DNA GC Kit (Seikagaku-Kogyo). 16S rRNA amplification using species-specific primers 16S rRNA was PCR-amplified using either forward primer Bsub-F (5¢-CGG ATG GTT GTT TGA ACC GCA TGG TTG A-3¢) or Bcer-F (5¢-GTT AGG GAA GAA CAA GTG CTA GTT G-3¢) (Shida 2001) and reverse primer 1377R (5¢-GGC ATG CTG ATC CGC GAT TAC TAG C-3¢) (Shida et al. 1996). The PCR

M.A.R.V. Guerra-Cantera and A.K. Raymundo reaction mixture contained 10 ng of total DNA, 50 pmol each of the primers, 1.2 U of Taq DNA polymerase (Roche), 1 · Taq DNA polymerase buffer (Roche), and 0.2 mM dNTPs in a 25-ll reaction mixture. Standard PCR was performed with an Applied Biosystems Gene Amp PCR System 2400 with the following cycle conditions: initial denaturation at 94 C for 5 min, 25 cycles at 94 C for 60 s, 58 C for 90 s and 72 C for 90 s, and final elongation at 72 C for 7 min. The PCR products were photographed using the Gel Documentation System (GDS) after electrophoresis on a 1.2% agarose gel and ethidium bromide staining. Phenotypic characterization The isolates were characterized using different tests as described by Gordon et al. (1973), Raymundo et al. (1995) and Takagi et al. (1993). All tests were performed using cells at the exponential phase of growth, except for the spore test, which required cultures from the early stationary phase.

Results and discussion Microbial researches in the developing countries, including the Philippines, tend to focus on the application of microorganisms for life improvement, with researchers often paying little attention to the identity of those microorganisms. For instance, most of the local isolates with industrial potential (i.e., antibiotic and enzyme producers) have only been partially identified based on a small number of phenotypic tests (i.e., microscopic examination of endospores, Gram reaction, and oxygen requirement, among others) prior to submission and deposition in local culture collections (Raymundo et al. 1985). Hence, a more thorough characterization was done using phenotypic, chemotaxonomic and genotypic approaches. Chemotaxonomic characterization Although the %G + C content of the isolates differed from each other by as much as 10% using HPLC, the figure still fell within the range of values for members of the genus Bacillus, which is 32–69 mol% (Slepecky & Hemphill 1991). To confirm the identity of Bacillus at the genus level, the presence or absence of DAP isomers was determined by TLC. Results showed that all 58 isolates of Bacillus contained DAP isomers characteristic of the Bacillus species. The cellular FAME composition of the isolates was also determined by gas chromatography. The major fatty acid present in the 58 Bacillus isolates was isoC15:0; and either iso-C14:0, iso-C16:0, anteiso-C17:0, and iso-C17:0 as minor components. Other isolates have anteiso-C15:0 as the major fatty acid. Bacillus alcalophilus DSM 485T (B1663) was established to contain C18:1 x7, a rarely encountered fatty acid in Bacillus, which is

Taxonomic reassessment of local Bacillus isolates

637

Table 1. List of Bacillus isolates recharacterized in this study. BIOTECH accession no.

Strain name as indicated in the BIOTECH record

Depositor/ Other strain designationb

Characteristics

1033a 1034 1035 1036 1037a 1038 1039 1040 1041 1042 1043a 1044a 1045 1046a 1060 1100a

B. pumilus B. megaterium B. licheniformis B. licheniformis B. circulans B. licheniformis B. licheniformis B. pumilus B. licheniformis B. licheniformis B. pumilus Bacillus sp. B. circulans B. megaterium B. pumilus Paenibacillus polymyxa Paenibacillus polymyxa Bacillus sp. B. amyloliquefaciens B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. licheniformis B. amyloliquefaciens B. licheniformis B. subtilis B. subtilis B. sphaericus B. sphaericus Bacillus sp. B. subtilis B. subtilis Bacillus sp. B. licheniformis B. cereus B. megaterium B. subtilis B. subtilis B. subtilis B. licheniformis B. subtilis subsp. natto Bacillus sp. B. pumilus B. subtilis subsp. subtilis B. pumilus B. cereus B. mycoides B. megaterium B. subtilis B. alcalophilus B. cereus B. coagulans

AKR (3C2) AKR (8C3) AKR (6C2) AKR (6C5) AKR (6C6) AKR (7C2) AKR (7C4) AKR (8C2) AKR (8C4) AKR (8C5) AKR (3UC6) AKR (4UC1) AKR (4UC2) AKR (8UC1) AKR (7C1) AKR/EYA

Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Antibiotic producer Polymyxin producer

AKR/EYA

ND

NFP (B14) NRA AKR/RHV (Ng 12) AKR/RHV (Ng 18) AKR/RHV (Ng 19) AKR/RHV (Ng 20) AKR/RHV (Ng 22) AKR/RHV (Ng 24) AKR/RHV (Ng 26) AKR/RHV (Ng 1875) AKR/RHV (Ng 0) MAA [LP4(1)O] MAA (LIRI Tap 2) MAA (LIRI Tap 3) MAA BCM BCM BCM BCM (K 160) MPFB (S139) MPFB (S159) MPFB (S137) MPFB (S82) MPFB (SIII) MPFB (S110) MATT MATT MLM LMT MAM MAM MATT (SC 9262) FBE MHDB CTH CTH CTH KDTN KDTN KDTN JCM 2499 RC JCM 2152 JCM 2257

Amylase producer Bam H1 source Mutant, bacitracin producer Mutant, bacitracin producer Mutant, bacitracin producer High-bacitracin producing mutant Mutant, bacitracin producer High-bacitracin producing mutant High-bacitracin producing mutant Low bacitracin- producing mutant Mutant, bacitracin producer From fermented fish paste From tapuy (Rice wine) From tapuy (Rice wine) – Amylase producer Amylase producer Amylase producer Amylase producer Pectinase producer Pectinase producer Pectinase producer Pectinase producer Pectinase producer Pectinase producer Penicillinase producer Penicillin acylase producer Penicillin acylase producer Alkaline protease producer Alkaline protease producer Alkaline protease producer b-lactamase induction For soybean fermentation Gas producer Protease producer Protease producer Protease producer Chitosanase producer Chitosanase producer Chitosanase producer ND Type strain Type strain Type strain

1101a 1124a 1130a 1165 1166 1167 1168 1169 1170 1171 1172 1173 1269a 1277a 1278 1320 1330 1331 1332 1333a 1567 1571 1572a 1573 1574 1575a 1633 1635a 1643 1677 1679a 1680 1810 10099 10101a 10102 10103 10104 10203 10204 10205 1666 1663 1509 1510

638

M.A.R.V. Guerra-Cantera and A.K. Raymundo

Table 1. Continued BIOTECH accession no.

Strain name as indicated in the BIOTECH record

Depositor/ Other strain designationb

Characteristics

1511 1512 1513 1514

B. B. B. B.

JCM JCM JCM JCM

Type Type Type Type

licheniformis megaterium pumilus subtilis

2505 2506 2508 1465T

strain strain strain strain

a

Local Bacillus isolates which were reclassified based on their phenotypic characteristic BIOTECH – National Institute of Molecular Biology and Biotechnology. b Abbreviations: AKR, A.K. Raymundo; BCM, B.C. Mendoza; CTH, C.T. Hedreyda; EYA, E.Y. Ardales; FBE, F.B. Elegado; JCM, Japan Collection of Microorganisms; KDTN, K.D.T. Natividad; LMT, L.M. Tapay; MAA, M.A. Alcala; MAM, M.A. Mercado; MATT, M.A.T. Tavanlar; MHDB, M.H.D. Bayer; MLM, M.L. Mangaban; MPFB, M.P.F. Bite; NFP, N.F. Paje; NRA, N.R. Apuya; RC, R. Calibo; RHV, R.H. Ventura.

usually found in acidophilic–thermophilic bacilli (Komagata & Suzuki 1987). These data confirm that all of the isolates examined are included in the genus Bacillus. However, the chemotaxonomic data obtained were not sufficient to make any reassessment at the species level. Genotypic characterization Based on preliminary morphological data gathered by the depositors, all isolates were labelled as members of the genus Bacillus (data not shown). Members of this genus were previously grouped based on their phenotypic similarities, with several tests used to differentiate at the species level (Gordon et al. 1973). To make differentiation and identification of Bacillus species more rapid and efficient, Shida (2001) successfully developed a method on the basis of PCR amplification of 16S rRNA gene fragments using group-specific primers. Thus, the forward primers Bsub-F and Bcer-F (specific for the members of the B. subtilis and the B. cereus groups, respectively), and the universal reverse primer 1377R, were used to differentiate them from the other isolates, particularly those included in the B. subtilis and B. cereus clusters. PCR amplification using Bsub-F and 1377R resulted in a 1.2 kb PCR fragment from B. coagulans JCM 2257T (B1510), B. licheniformis JCM 2505T (B1511), B. megaterium JCM 2506T (B1512), B. pumilus JCM 2508T (B1513), B.

subtilis JCM 1465T (B1514), and B. subtilis JCM 2499T (B1666), all of which are included in the B. subtilis cluster. The same sized fragment was also amplified from 41 local isolates, suggesting that these belong to the B. subtilis cluster (Figure 1a). On the other hand, B. cereus JCM 2152T (B1509) and all other local isolates produced a 0.925 kb fragment using Bcer-F and 1377R, suggesting that these isolates belong to the B. cereus cluster (Figure 1b). Using this method, the local isolates were differentiated as members of either the B. subtilis or the B. cereus cluster. Phenotypic characterization The isolates were subjected to 39 phenotypic tests to provide more descriptive information that would help in recognizing their taxa (Table 2). All isolates were Grampositive, rod-shaped, catalase-positive, spore-formers, with either an aerobic or facultative anaerobic mode of respiration, characteristics of the Bacillus species. Cells appeared singly and in chains of considerable length. Some of the isolates were good oxidizers, while others were weak fermenters. Most isolates produced acid from glucose, fructose, maltose, sucrose and trehalose. Around 3, 7, 28, 47 and 68% of the total isolates showed a negative reaction for casein, gelatin, starch, Tween 80 and lecithin hydrolysis tests, respectively. Only 10 Bacillus isolates could reduce nitrate (NO3 ) to

Figure 1. (a) Amplification of 16S rRNA gene by PCR with detection primer Bsub-F and universal reverse primer 1377R. Lanes: (1) kb ladder, (2) B. subtilis JCM 1465T (B1514), (3) B. coagulans JCM 2257T (B1510), (4) B. subtilis S82 (B1573), (5) B. subtilis SIII (B1574), (6) B. licheniformis Ng 12 (B1165), (7) B. licheniformis Ng 19 (B1167), (8) B. licheniformis Ng 20 (B1168), (9) B. subtilis (B1332). (b) Amplification of 16S rRNA gene by PCR with detection primer Bcer-F and universal reverse primer 1377R. Lanes: (10) B. pumilus (B10102), (11) B. cereus K-30 (B10203), (12) B. mycoides K-8 (B10204), (13) B. megaterium K-13 (B10205), (14) B. cereus (Bc2) and (15) B. cereus (Bc3).

Taxonomic reassessment of local Bacillus isolates

639

Table 2. Distinctive phenotypic characteristics of local Bacillus isolates used in this study. Characteristics

1033

1037

1043

1044

1046

1101

1124

1130

1330

1509

1510

1511

Gram reaction Cell shape Cell width of >1.0 lm Cell length of >3 lm Spore position

+ Rods )

+ Rods )

+ Rods )

+ Rods )

+ Rods )

+ Rods )

+ Rods )

+ Rods )

+ Rods )

+ Rods +

+ Rods )

+ Rods )

+

+

+

+

+

+

)

)

)

+

)

)

ND

Central

ND

Subterminal ) + + + 6.5

Subterminal ) + + + 6.1

Subterminal ) + + + 6.0

Subterminal ) + + + 6.5

Subterminal + + + W+ 5.5

ND

) + + + 5.8

Subterminal + + + + 5.8

Central

+ VW+ + ) 5.8

Subterminal ) W+ + + 5.8

+ + + + 7.0

) + + + 5.8

+ )

) )

) )

+ ND

+ )

+ )

) )

+ +

) ND

+ )

+ +

+

)

)

)

)

)

)

)

+

)

)

+ + + ) +

+ + + ) +

+ + + ) )

+ ) + + )

+ + + + )

+ + + + +

+ + + ) )

+ + + ) )

+ + + ) )

+ + W+ ) )

+ + + ) )

+ ) )

) ) )

+ ) )

+ ) )

+ + )

+ + +

+ + )

+ W+ )

+ ) )

+ ) )

+ + +

+ + ) + ) )

+ + ) ) + )

+ + ) + + )

+ + ) + + )

+ + ) + ) )

+ + ) + ) )

+ + ) + ) )

+ + ) + + )

+ + ) + ) )

+ + ) + + )

+ + ) + + )

+ ) + ) + ) ) ) + VW+ ) )

+ ) + ) + ) ) ) + VW+ ) )

+ ) + ) + ) ) ) + W+ ) )

+ ) + ) + ) + + + + + )

+ W+ + ) VW+ ) + W+ W+ VW+ W+ )

W+ W+ + ) ) ) + ) VW+ ) ) )

+ VW+ + ) + ) + ) + W+ + )

W+ + + ) + ) + VW+ + ) W+ )

+ ) + ) + ) + ) + ) ) )

VW+ VW+ + ) + ) + ) VW+ ) ) )

+ + + VW+ + ) + + + + + )

Anaerobic growth + Catalase + Nitrate reduction + VP reaction ) pH in Vogues 5.8–6.0 Proskauer broth Utilization of citrate + Utilization of ) propionate Degradation of tyrosine ) Hydrolysis of Casein + Gelatin + Starch + Tween 80 ) Egg yolk lecithin ) Growth in the presence of 5% NaCl + 7% NaCl ) 10% NaCl ) Growth at pH 5.5 + pH 6.8 + 5 C ) 42 C + 50 C ) 65 C ) Acid produced from D )Glucose + L )Arabinose ) D )Fructose + D )Galactose ) Maltose + Lactose ) Sucrose ) D )Xylose ) Trehalose + Glycerol + Mannitol ) Gas produced ) from D )glucose BIOTECH identity B. pumilus

Gram reaction Cell shape Cell width of >0.5 lm Cell length of >3 lm Spore position Anaerobic growth Catalase Nitrate reduction

B. circu- B. pum- Bacillus B. mega- Paeniba- Bacillus B. amylo- B. amylo- B. cereus B. coagulans ilus sp. terium cillus sp. liquefa- liquefa- JCM lans polymyxa ciens ciens 2152 JCM 2257

B.licheniformis JCM 2505

1512

1513

1572

1575

1635

1666

1679

1575

1635

1666

1679

10101

+ Rods +

+ Rods )

+ Rods )

+ Rods +

+ Rods )

+ Rods )

+ Rods )

+ Rods +

+ Rods )

+ Rods )

+ Rods )

+ Rods )

+

)

)

+

)

)

)

+

)

)

)

+

Subterminal ) + )

Subterminal ) + )

Subterminal + + +

ND

Subterminal + + +

Subterminal ) + +

ND

ND + W+ +

Subterminal ) + +

ND

) + )

Subterminal + + +

Subterminal ) + +

+ W+ +

) + )

640

M.A.R.V. Guerra-Cantera and A.K. Raymundo

Table 2. (Continued)

VP reaction pH in Vogues Proskauer broth Utilization of citrate Utilization of propionate Degradation of tyrosine Hydrolysis of Casein Gelatin Starch Tween 80 Egg yolk lecithin Growth in the presence 5% NaCl 7% NaCl 10% NaCl Growth at pH 5.5 pH 6.8 5 C 42 C 50 C 65 C Acid produced from D )Glucose L )Arabinose D )Fructose D )Galactose Maltose Lactose Sucrose D )Xylose Trehalose Glycerol Mannitol Gas produced from D )Glucose BIOTECH identity

1512

1513

1572

1571

1635

1666

1679

1575

1635

1666

1679

1011

) 6.5)7.0

+ 5.0)5.5

) 5.8

+ 5.5

+ 6.1

+ 6.0

+ 6.5

+ 5.5

+ 6.1

+ 6.0

+ 6.5

+ 6.0

+ )

) )

+ )

+ )

+ )

+ )

+ )

+ )

+ )

+ )

+ )

+ )

)

)

)

+

)

)

)

+

)

)

)

)

+ ) + ) ) of + + )

+ + ) ND +

+ + ) ND +

W+ + + ) )

+ + ) + )

+ + + + )

+ + ) ) )

W+ + + ) )

+ + ) + )

+ + + + )

+ + ) ) )

+ + + ND )

) ) )

+ ) )

+ ) )

+ ) )

+ + +

+ ) )

+ ) )

+ ) )

+ + +

+ ) )

+ + +

+ + ) + ) )

+ + ) + + )

+ + ) + ) )

+ + ) + ) )

+ + ) + + )

+ + ) + + )

+ + ) + + )

+ + ) + ) )

+ + ) + + )

+ + ) + + )

+ + ) + + )

+ + ) + + )

+ VW+ + ) + ) + + + ) + )

+ + + ) + ) + W+ + + + )

+ ) + ) + ) + ) + VW+ ) )

+ ) + ) + ) + ) + VW+ ) )

+ + + ) + ) + ) + ) + )

+ VW+ + ) ) ) + VW+ + + + )

+ + + ) VW+ ) + W+ ) ) W+ )

+ ) + ) + ) + ) + VW+ ) )

+ + + ) + ) + ) + ) + )

+ VW+ + ) ) ) + VW+ + + + )

+ + + ) VW+ ) + W+ ) ) W+ )

+ VW+ + ) W+ ) + ) + ) + )

B. megaterium JCM 2506

B. pumilus JCM 2508

Bacillus Bacillus B. cereus B. subtilis B. subti- Bacillus B. cereus B. subtilis B. subti- Bacillus sp. sp. JCM lis sp. JCM lis sp. 2499 2499

Legend: (+), positive; (W+), weak positive; (VW+), very weak positive; ()), negative; (ND), not determined. BIOTECH ) National Institute of Molecular Biology and Biotechnology.

either nitrite (NO2 ) or ammonia (NH3) or gaseous nitrogen (N2). On the other hand, most had either subterminal to terminal spores. Approximately 70% of the total isolates could grow in media with salt concentrations of up to 7%, while 90% could grow in media with pH up to 6.8, and at temperatures up to 42 C but not 50–65 C. Results of the phenotypic characterization further showed the phenotypic diversity of the local Bacillus isolates deposited at the PNCM. Taxonomic reassessment The identity of the local isolates was checked and reassessed by comparing the data gathered from the phenotypic characterization to the identification key established by Gordon et al. (1973) (Figure 2). The use

of phenotypic characteristics, although classical, has been successfully used in the identification of the aerobic spore-forming bacteria (Reva et al. 2001). Analyses showed that some isolates possessed a number of characteristics quite different from those of their type strains, suggesting that those isolates were misclassified. For instance, B1033 (‘B. pumilus’) was able to grow under anaerobic conditions. The type species of B. pumilus JCM 2508T on the other hand, is an aerobe, and therefore could not grow under anaerobic conditions. Moreover, B. pumilus JCM 2508T was unable to hydrolyse starch into a simpler form, but B1043 (‘B. pumilus’) was found to be amylase-positive. Thus, using the modified dichotomous key constructed for Bacillus, the species of five isolates was

Taxonomic reassessment of local Bacillus isolates

641

Figure 2. Schematic diagram for the identification of Bacillus spp. isolated from the Philippines (modified from Gordon et al. 1973; Sneath 1986).

established and 12 were reclassified (Table 3). The identities of B1033 (‘B. pumilus’) and B1043 (‘B. pumilus’) were revised to B. cereus and B. subtilis, respectively, based on their oxygen requirement. B1033, which is unable to grow under strictly aerobic conditions, was misclassified as B. pumilus, which is strictly aerobic. B1037 (‘B. circulans’), on the other

hand, showed the ability to utilize citrate thus was reclassified from B. circulans to B. megaterium. B1046 (‘B. megaterium’) showed a positive reaction to VP test, and so was emended to B. coagulans. Another three local isolates, namely B1100 (‘Paenibacillus polymyxa’), B1277 (‘B. licheniformis’) and B1333 (‘B. subtilis’), were reclassified as P. larvae on the basis of

642

M.A.R.V. Guerra-Cantera and A.K. Raymundo

Table 3. Local Bacillus isolates identified/reclassfied after taxonomic reassessment. BIOTECH accession #

BIOTECH identity

Phenetic identity

Cluster based on 16S rRNA specific primers B. subtilis

B. cereus

FAMEs identity

Final identity

1033 1037

B. pumilus 3C2 B. circulans 6C6

B. cereus B. megaterium

) +

+ +

1043 1044

B. pumilus 3UC6 Bacillus sp. 4UC1

B. subtilis B. subtilis

) )

+ +

1046

B. megaterium 8UCI

B. coagulans

+

+

1100

Paenibacillus polymyxa

Paenibacillus larvae

)

+

1101 1124 1130 1269

B. subtilis B. subtilis B. subtilis Paenibacillus lentimorbus Paenibacillus larvae

+ + + +

) ) ) )

+

)

1330 1572

Paenibacillus polymyxa Bacillus sp. B14 B. amyloliquefaciens B. licheniformis LP4(1)O B. licheniformis LIRI Tap 2 B. amyloliquefaciens Bacillus sp. S137

+ +

) +

1575

Bacillus sp. S110

B. licheniformis Brevibacillus laterosporus B. subtilis

)

+

1635 1679

B. cereus B. subtilis

B. coagulans B. pumilus

+ +

) +

B. subtilis

+

)

B. cereus

)

+

B. coagulans

+

)

B. cereus B. cereus (GC subgroup A) B. coagulans B. coagulans

B. licheniformis

+

)

B. licheniformis

B. megaterium

+

)

B. pumilus B. subtilis

+ +

) )

B. megaterium B. megaterium (GC subgroup A) B. pumilus B. pumilus B. subtilis B. subtilis

1277

10101 Bacillus sp. P4 Reference strains 1509 B. cereus JCM 2152 1510 1511 1512 1513 1666

T

B. coagulans JCM 2257 T B. licheniformis JCM 2505 T B. megaterium JCM 2506 T B. pumilus JCM 2508 T B. subtilis JCM 2499 T

B. thuringiensis B. cereus B. cereus B. megaterium (GC subgroup A) ND B. subtilis B. thuringiensis B. subtilis sv. gallieriae B. cereus B. cereus (GC subgroup A) ND Paenibacillus larvae B. amyloliquefaciens B. subtilis B. subtilis B. subtilis B. subtilis B. subtilis ND Paenibacillus lentimorbus Paenibacillus Paenibacillus lentimorbus lentimorbus ND B. licheniformis ND Brevibacillus laterosporus B. cereus B. cereus (GC subgroup A) B. amyloliquefaciens B. coagulans Paenibacillus B. pumilus lentimorbus B. subtilis B. subtilis

B. licheniformis

Legend: (+), positive; (W+), weak positive; (VW+), very weak positive; ()), negative; (ND), not determined. BIOTECH – National Institute of Molecular Biology and Biotechnology.

the catalase test. Moreover, B1130 (‘B. amyloliquefaciens’) was renamed as B. subtilis, while B1269 (‘B. licheniformis’) was reclassified to Paenibacillus lentimorbus on the basis of catalase test. Still another isolate, B1635 (‘B. cereus’), could not grow at 50 C, and was revised to B. coagulans; B1679 (‘B. subtilis’) failed to hydrolyse starch and was reclassified as B. pumilus. In addition, Bacillus isolates, which had not yet been identified to the species level were also named using the 39 phenotypic tests. For instance, Bacillus spp. under accession numbers B1044, B1124, B1575 and B10101 were named as B. subtilis, while B1572 was identified as Brevibacillus laterosporus. Out of 58 local Bacillus isolates deposited at the PNCM, 45 maintained their original identities, 12 isolates were reclassified into another species of Bacillus, and five were identified up to the species level.

Conclusion The use of modern techniques in molecular biology, genetics, and biochemistry has led to progress in bacterial taxonomy. The systematic study of the aerobic, spore-forming bacteria has resulted in the reclassification of the genus Bacillus into several genera. However, the identification process has become more difficult for the non-specialist in Bacillus taxonomy, especially in culture collections with limited modernized equipment. It is therefore important to try different identification approaches and to determine if the practical and relatively less difficult procedures can still be utilized. The taxonomic reassessment of local Bacillus isolates using chemotaxonomic, molecular fingerprinting and phenotypic analyses resulted in the reclassification of 12 isolates into another species of Bacillus, and the iden-

Taxonomic reassessment of local Bacillus isolates tification of five up to the species level. Only 45 Bacillus isolates maintained their original identities. In this study, the continuing importance of phenotypic characterization in the speciation of bacterial isolates is highlighted. Both chemotaxonomic and genotypic methods were unable to identify the bacteria up to the species level in this particular case, a major limitation of the genotypic method being the necessity of finding primers specific to the species, without which identification of that particular species could not be achieved. Despite the universal application of both genotypic and chemotaxonomic methods, phenotypic characterization remains an important tool in the arsenal of microbial taxonomists. Such results should put the different methods of characterization in perspective and show that classical approaches can still be equally effective, or sometimes even more effective than the modern methods. This should encourage taxonomists in Third World countries, who can ill afford the latest technologies and must do their work with limited equipment. Acknowledgements This work was conducted at the National Institute of Molecular Biology and Biotechnology, UP Los Ban˜os. The authors are indebted to Dr M. Suzuki of the Bioscience Research Center, Japan for analysing the FAMEs, Dr Rosario G. Monsalud of the Philippine National Collection of Microorganims-BIOTECH for the local isolates of Bacillus used in this research, and the Commission on Higher Education (CHED) for funding this study. References Ash, C., Farrow, J.A.E., Wallbanks, S. & Collins, M.D. 1991 Phylogenetic heterogeneity of the genus Bacillus revealed by comparative analysis of small-subunit-ribosomal RNA sequences. Letters in Applied Microbiology 13, 202–206. Ash, C., Priest, F.G. & Collins, M.D. 1993 Molecular identification of rRNA group 3 bacilli (Ash, Farrow, Wallbanks, & Collins) using a PCR probe test. Proposal for the creation of a new genus Paenibacillus. Antonie van Leeuwenhoek 64, 253–260. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. & Struhl, K. 1995 Current Protocols in Molecular Biology, Vol. I. USA: John Wiley and Sons. ISBN 0-471-50338-X. Claus, D. & Berkeley, R.C.W. 1986 Genus Bacillus Cohn 1872. In Bergey’s Manual of Systematic Bacteriology, vol. 2, eds. Sneath, P.H.A., Mair, N.S., Sharpe, M.E. & Holt, J.G. pp. 1105–1139. Baltimore: The Williams & Wilkins Co. ISBN 0-6683-07893-3. Fortina, M.G., Pukall, R., Schumann, P., Mora, D., Parini, C., Luigi Manachini, P. & Stackebrandt, E. 2001 Ureibacillus gen. nov., a new genus to accommodate Bacillus thermosphaericus (Andersson et al. 1995), emendation of Ureibacillus thermosphaericus and description of Ureibacillus terrenus sp. nov. International Journal of Systematic and Evolutionary Microbiology 51, 447–455. Gordon, R.E., Haynes, W.C. & Pang, C.H. 1973 The genus Bacillus. In Agriculture Handbook No. 427. pp. 1–99. Washington, D. C: US Dept. of Agriculture. Heyndrickx, M., Vandemeulebroecke, K., Scheldeman, P., Kersters, K., De Vos, P., Logan, N.A., Aziz, A.M., Ali, N. & Berkeley, R.C.W. 1996 A polyphasic reassessment of the genus Paenibacillus, reclassification of B. lautus (Nakamura 1984) as Paenibacillus

643 lautus comb. nov. and B. peoriae (Montefusco et al. 1993) as P. peoriae comb. nov., and emended description of P. lautus and P. peoriae. International Journal of Systematic Bacteriology 46, 988– 1003. Heyndrickx, M., Lebbe, L., Kersters, K., De Vos, P., Forsyth, G. & Logan, N.A. 1998 Virgibacillus: a new genus to accommodate Bacillus pantothenticus (Proom and Knight 1950). Emended description of Virgibacillus pantothenticus. International Journal of Systematic Bacteriology 48, 99–106. Komagata, K. & Suzuki, K. 1987 Lipid and cell-wall analysis in bacterial systematics. Methods in Microbiology 19, 161–207. Mesbah, M., Premachandran, U. & Whitman, W.B. 1989 Precise measurement of the G + C content of deoxyribonucleic acid by high-performance liquid chromatography. 39, 159–167. Nakamura, L.K. 1993 DNA relatedness of B. brevis Migula 1900 strains and proposal of B. agri sp. nov., nom. rev., and B. centrosporus sp. nov., nom. rev. International Journal of Systematic Bacteriology 43, 20–25. Nazina, T.N., Tourova, T.P., Poltaraus, A.B., Novikova, E.V., Grigoryan, A.A., Ivanova, A.E., Lysenko, A.M., Petrunyaka, V.V., Osipov, G.A., Belyaev, S.S. & Ivanov, M.V. 2001 Taxonomic study of aerobic thermophilic bacilli: descriptions of Geobacillis subterraneus gen. nov., sp. nov. and Geobacillis uzenensis sp. nov. from petroleum reservoirs and transfer of Bacillus stearothermophilus, Bacillus thermocatenulatus, Bacillus thermoleovorans, Bacillus kaustophilus, Bacillus thermoglucosidasius & Bacillus thermodenitrificans to Geobacillus as the new combinations G. stearothermophilus, G. thermocatenulatus, G. thermoleovorans, G. kaustophilus, G. thermoglucosidasius & G. thermodenitrificans. International Journal of Systematic and Evolutionary Microbiology 51, 433–446. Niimura, Y., Koh, E., Yanagida, F., Suzuki, K., Komagata, K. & Kozaki, M. 1990 Amphibacillus xylanus gen. nov., sp. nov., a facultatively anaerobic spore-forming xylan-digesting bacterium which lacks cytochrome, quinone and catalase. International Journal of Systematic Bacteriology 40, 297–301. Priest, F.G., Goodfellow, M. & Todd, C. 1988 A numerical classification of the genus Bacillus. Journal of General Microbiology 134, 1847–1882. Raymundo, A.K., Serrano, E.T., Reyes, G.D. & Zulaybar, T.O. 1985 Isolation and identification of antibiotic-producing Bacillus from the soil. Philippine Agriculturist 68, 393–402. Raymundo, A.K., Zamora, A.F. & Dalmacio, I.F. 1991 Manual in Microbiological Techniques. UPLB: Technology and Livelihood Resource Center. ISBN 971-852-001-8. Reva, O.N., Sorokulova, I.B. & Smirnov V.V. 2001. Simplified technique for the identification of the aerobic spore-forming bacteria by phenotype. International Journal of Systematic and Evolutionary Microbiology 51, 1361–1371. Sasser, M. 2001 Identification of bacteria by gas chromatography of cellular fatty acids. MIDI Technical Note 101, 1–6. Schlesner, H., Lawson, P.A., Collins, M.D., Weiss, N., Wehmeyer, U., Volker, H. & Thomm, M. 2001 Filobacillus milensis gen. nov., sp. nov., a new halophilic spore-forming bacterium with Orn-D-Glutype peptidoglycan. International Journal of Systematic and Evolutionary Microbiology 51, 425–431. Shida, O. 2001 Modern Systematics of Aerobic, Endospore-forming Bacteria. In Proceedings of the 4th Asia-Pacific Biotechnology Congress & 30th Annual Philippine Society for Microbiology, Inc. Convention. pp. 315–322. May 16–18, 2001. Waterfront Cebu City Hotel Lahug, Cebu City Philippines. Shida, O., Takagi, H., Kadowaki, K. & Komagata, K. 1996 Proposal for two new genera, Brevibacillus gen. nov. & Aneurinibacillus gen. nov. International Journal of Systematic Bacteriology 46, 939–946. Shida, O., Takagi, H., Kadowaki, K., Nakamura, L.K. & Komagata, K. 1997 Transfer of B. alginolyticus, B. chondroitinus, B. curdlanolyticus, B. glucanolyticus, B. kobensis and B. thiaminolyticus to the genus Paenibacillus & emended description of the genus Paenibacillus. International Journal of Systematic Bacteriology 47, 289–298.

644 Shida, O., Takagi, H., Kadowaki, K., Nakamura, L.K. & Komagata, K. 1995 Proposal of B. resuzeri sp. nov., B. formosus sp. nov., nom. rev., and B. borstelensis sp. nov., nom. rev. International Journal of Systematic Bacteriology 45, 93–100. Shida, O., Takagi, H., Kadowaki, K., Udaka, S. & Komagata, K. 1994 Bacillus galactophilus is a later subjective synonym of Bacillus agri. International Journal of Systematic Bacteriology 44, 172-173. Slepecky, R.A. & Hemphill, H.E. 1991 The genus Bacillus-nonmedical. In The Prokaryotes, eds. Balows, A., Truper, H. G., Dworkin, M., Harder, W. & Schleifer, K.H. New York: Springer-Verlag. ISBN 3-540-97258-7. spring, S., Ludwig, W., Marquez, M.C., Ventosa, A. & Schleifer, K.H. 1996 Halobacillus gen. nov., with description of Halobacillus litoralis sp. nov. and H. truperi sp. nov., and transfer of Sporosarcina halophilia to H. halophilus comb. nov. International Journal of Systematic Bacteriology 46, 492–496. Staneck, J.L. & Roberts, G.D. 1974 Simplified approach to identification of aerobic actinomycetes by thin-layer chromatography. Applied Microbiology 28, 226–231.

M.A.R.V. Guerra-Cantera and A.K. Raymundo Takagi, H., Shida, O., Kadowaki, K., Komagata, K. & Udaka, S. 1993 Characterization of B. brevis, with descriptions of B. migulanus sp. nov., B. choshinensis sp. nov., B. parabrevis sp. nov., & B. galactophilus sp. nov. International Journal of Systematic Bacteriology 43, 221–231. Waino, M., Tindall, B.J., Schumann, P. & Ingvorsen, K. 1999 Gracilibacillus gen. nov., with description of Gracilibacillus halotolerans gen. nov., sp. nov.; transfer of Bacillus dipsosauri comb. nov., and Bacillus salexigens to the genus Salibacillus gen. nov., as Salibacillus salexigens comb. nov. International Journal of Systematic Bacteriology 49, 821–831. Wisotzkey, J. D., Jurtshuk, Jr., P., Fox, G.E., Deinhard, G. & Poralla, K. 1992 Comparative sequence analysis on the 16S rRNA (rDNA) of B. acidocaldarius, B. acidoterrestris, & B. cycloheptanicus & proposal for creation of a new genus, Alicyclobacillus gen. nov. International Journal of Systematic Bacteriology 42, 263–269.

World Journal of Microbiology & Biotechnology (2005) 21:645–648 DOI 10.1007/s11274-004-3568-3

Ó Springer 2005

Relationship among acidophilic bacteria from diverse environments as determined by randomly amplified polymorphic DNA analysis (RAPD) Tanveer Akbar1, Kalsoom Akhtar1, Muhammad A. Ghauri1, Munir A. Anwar1, Moazur Rehman1, Mehboobur Rehman2, Yusuf Zafar2 and Ahmad M. Khalid1,* 1 Bioprocess Technology Division 2 Plant Biotechnology Division, National Institute for Biotechnology and Genetic Engineering, P.O. Box 577, Faisalabad, Pakistan *Author for correspondence: Tel.: +92-41-651471, Fax: +92-41-651472, E-mail: [email protected]/director@ fsd.comsats.net.pk

Keywords: Acidithiobacillus ferrooxidans, Acidithiobacillus thiooxidans, acidophilic moderate thermophiles, genomic diversity, RAPD analysis

Summary Random amplification of polymorphic DNA (RAPD), a PCR-based technique was applied to evaluate genomic diversity among three strains of Acidithiobacillus thiooxidans, five strains of Acidithiobacillus ferrooxidans and one acidophilic moderate thermophile strain, using 45 random primers of five different series. More than 2200 bands were observed, with an average of 45 bands per primer. Primer OPC-3 produced the maximum number of fragments whereas minimum numbers of fragments were produced with primer OPA-5. A dendrogram was generated using cluster analysis by the unweighted pair group method of arithmetic means (UPGMA). The dendrogram showed three groups with similarity ranging from 29 to 85%. The maximum similarity (85%) was observed between the strains T.t1 and T.t2 of Acidithiobacillus thiooxidans.

Introduction The genus Acidithiobacillus (formerly known as genus Thiobacillus) was described in 1904 by Beijerinck and was recognized as a genus of convenience (Kelly & Harrison 1989). Previously, this genus contained genetically and physiologically heterogeneous bacterial groups. However, Kelly & Wood (2000) have reclassified some species of Thiobacillus to three new genera thus reducing the group heterogeneity to its minimum level. In general, bacteria belonging to genus Acidithiobacillus are obligate acidophiles, aerobic, motile Gram-negative rods and can utilize reduced sulphur compounds to support autotrophic growth. Some species oxidize ferrous iron or use natural and synthetic metal sulphides for energy source, while some can oxidize hydrogen as well. Mesophilic species grow optimally at 30–35 °C and moderately thermophilic species at around 45 °C. The G + C content of the DNA is 52–64 mol% (Kelly & Wood 2000). Acidithiobacillus thiooxidans and Acidithiobacillus ferrooxidans show optimum growth at pH values ranging from 2–5 and temperatures ranging from 25 to 35 °C. At. thiooxidans utilizes inorganic sulphur, whereas At. ferrooxidans can utilize Fe2+ ions in addition to sulphur as a sole electron donor. These acidophiles are present in high numbers in mine drainage and are important in

bioleaching of copper, uranium and gold from low grade ores as well as in microbial removal of pyritic sulphur from coal (Norris & Johnson 1998). They are also intimately involved in the production of acid mine drainage. The most intensively studied species is At. ferrooxidans. It participates in the leaching of iron pyrite (FeS2) by direct oxidation of this mineral, while in the commercial leaching of copper and uranium from ores, this may act indirectly (Tuovinen & Fry 1993). Analysis of randomly amplified polymorphic DNA (RAPD) is widely used to investigate variability among microorganisms (Novo et al. 1996). This technique is fast, cheap and easy to perform, and requires only small amounts of DNA that is available even from dried materials and does not require any previous sequence information (Simpson et al. 2002). In RAPD, agarose gel electrophoresis and ethidium bromide (EtBr) staining are used for DNA detection. RAPD involves the use of arbitrary GC-rich decamers as single primers instead of reverse and forward primers to amplify unknown sites in the target genome. During the last several years, RAPD has been developed and applied for the characterization of bacteria and their classifications, groupings, identification and placement on the level of strains (Novo et al. 1996; Pobell et al. 1997). Moreover, RAPD analysis is based upon PCR,

646

T. Akbar et al.

Table 1. Different isolates of acidophilic bacteria used with their source of isolation. Bacteria Mesophiles Acidithiobacillus thiooxidans ATCC 8085 T.t1 T.t2 Moderate thermophiles Acidithiobacillus ferrooxidans ATCC 13661 NCIMB-13537a MAAC1 MAAC2 TF4 MT13 a

Source

American Type Culture Collection Mach coal samples, Pakistan Khushab coal mines, Pakistan

American Type Culture Collection Saindak copper mine water, Pakistan Harnai coal samples, Pakistan Mach coal samples, Pakistan Uranium mine site, Pakistan Uranium mine site, Pakistan

Isolated in our laboratory and deposited to NCIMB.

which is a very sensitive technique and should give reproducible results. This paper describes the RAPD analysis of five strains of At. ferrooxidans, three strains of At. thiooxidans and one moderate thermophile isolate. The aim of this study was to optimize the RAPD technique and to examine the genomic variability among these bacteria.

Materials and methods Bacterial strains and growth conditions The acidophilic bacteria used in these studies were isolated from samples obtained from coal, copper and uranium mine sites by enrichment followed by solid plate streaking (Ghauri 1991) (Table 1). The single colonies obtained were transferred to liquid medium. Culture purity was ascertained by three successive transfers. All the isolates were grown in basal salt medium containing (g/l) 0.5 MgSo4Æ7H2O, 0.15 (NH4)2SO4, 0.05 KH2PO4, 0.05 KCl, 0.01 Ca(NO3)2, Pyrite (1% ) was added to the medium for At. ferrooxidans and 1% sulphur for At. thiooxidans. For the moderately thermophilic MT13 isolate, the former medium was amended with 0.05% yeast extract. Growth was obtained in shake flasks incubated on a shaker at 30 °C for mesophiles and 45 °C for moderate thermophiles. Cells were harvested by centrifugation at 10,000 rpm, washed with distilled water (pH 2.5 adjusted with H2SO4) and finally suspended in T.E. buffer. DNA isolation and optimization of PCR conditions Total genomic DNA was isolated using the Sambrook et al. (1989) protocol with slight modifications where needed. Forty seven primers from Kit A (OPA-05, OPA07 and OPA-18), Kit B (OPB-05, OPB-08, OPB-10 and OPB-18), Kit C (OPC-01 to OPC-08, OPC-11 to OPC15 and OPC-17), Kit J (OPJ-04 to OPJ-08, OPJ-10, OPJ11, OPJ-13, OPJ-15 and OPJ-18 to OPJ-20) and Kit Z

(OPZ-01, OPZ-04 to OPZ- 07, OPZ-11, OPZ-13 to OPZ-15, OPZ-19 and OPZ-20) were used for the PCR as reported by Williams et al. (1990) with minor modifications for isolation of good quality DNA. The PCR was carried out in a Perkin-Elmer DNA thermal cycler 480. DNA denaturation was done at 94 °C for 5 min followed by 40 cycles amplification (94 °C for 1 min, 36 °C for 1 min and 72 °C for 2 min) and a final extension step at 72 °C for 10 min. Amplification products were separated on 1.2% agarose gel. Statistical analysis Nei and Li’s coefficient (Nei & Li 1979) was used to calculate the relatedness of the studied species among each other. Nei and Li’s co-efficient was calculated by the following statistical equation F ¼ 2Nxy =ðNx þ Ny Þ; where F is the similarity co-efficient, Nx and Ny are the numbers of fragments from population x and y, respectively, whereas Nxy is the number of fragments shared by the two populations. Data were analysed by comparing the RAPD profile on the basis of the presence or absence (1 or 0 respectively) of each reproducible DNA band. A similarity matrix was generated using Nei and Li’s coefficient of similarity and a dendrogram was generated using Unweighted Pair Group Method of Arithmetic Means (UPGMA) (Nei & Li 1979).

Results and discussion The present study was conducted to establish a relationship among five strains of At. ferrooxidans, three strains of At. thiooxidans and a moderate thermophile (isolated from a uranium mine) by employing RAPD analysis. These strains are currently being used in the Bioprocess Technology Division at NIBGE for bioleaching of precious metals and microbial removal of pyritic sulphur from coal. All At. thiooxidans and At. ferrooxidans isolates were Gram-negative mesophiles, motile rods, capable of oxidizing elemental sulphur and reduced sulphur compounds. At. ferrooxidans was also capable of oxidizing Fe2+ ions. Optimum growth was obtained at pH 4.0 and pH 2.0 for At. thiooxidans and At. ferrooxidans respectively . Local isolates T.t1 and T.t2 have greater specific growth rate (0.20 and 0.21 h)1 respectively) than strain ATCC 8085, whose specific growth rate was 0.188 h)1. At. ferrooxidans isolate MAAC2 had the maximum specific growth rate compared to other strains (0.108 h)1). The moderately thermophilic isolate MT13 was mixotrophic and could oxidize both Fe2+ ions and reduced sulphur compounds. The 16S rDNA analysis of this strain revealed it to be a strain of Sulfobacillus thermosulfidooxidans (Ghauri et al. 2003).

RAPD analysis in Acidithiobacillus

647

In the present study, the most crucial step was to obtain enough biomass for DNA isolation. Most of the isolates grew slowly on respective media with mean generation times (td) of 6–8 h (data not shown). Genomic DNA from At. thiooxidans ATCC 8085 and At. ferrooxidans ATCC 13661 was used for optimization of PCR amplification conditions, using six different DNA concentrations (3–18 ng) with primers OPA-05 and OPB-05. The RAPD products were separated in a 1.2% agarose gel. The highest intensity of the bands was observed with 12 ng of DNA and this concentration was used for optimizing Taq polymerase and magnesium chloride concentrations. Reactions were conducted using 1.0, 2.0 and 3.0 units of Taq polymerase per 25 ll reaction mixture and 2.0 units were found optimum. Magnesium chloride concentrations of 2.5, 3.0, 3.5 and 4.0 mM were also tested for obtaining high intensity bands, and the best results were obtained with 3.0 mM. Consistent results were obtained with these optimized conditions. Variation in the band profiles was observed for each primer (Figure 1). Most of the primers used in the amplification, grouped the nine isolates into 3 groups (Table 2, Figure 2). Group-I comprised of strains T.t1, T.t2 and At. thiooxidans (ATCC 8085) whereas group-II included At. ferrooxidans strains TF4, MAAC2, ATCC 13661, NCIMB 13537 and MAAC1. MT13 was placed in group-III. Group I strains T.t1 and T.t2 showed the maximum similarity i.e. 85%. Reference strain ATCC 8085 showed 59% similarity to these strains thus indicating group heterogeneity (Figure 2). Among Group-II strains MAAC1 and TF4 showed maximum similarity of 55%. Isolate MAAC1 also shared 52, 36 and 28% similarity with MAAC2, NCIMB 13537 and ATCC 13661 respectively. Isolate MAAC2 showed 36 and 28% similarity with NCIMB 13537 and ATCC 13661 respectively. Strain NCIMB 13537 had maximum similarity of 51% with ATCC 13661. Among strains of group-I and group-II similarity ranged from 33 to 55%. The most distinct isolate MT13 showed similarity ranging from 31–37% with various strains of At. ferrooxidans and At. thiooxidans (Table 2, Figure 2). On the basis of physiological characteristics such high values of dissimilarity were

Figure 1. (a–c) RAPD profile from acidophilic bacteria generated with different primers: (a) Primers OPA-05 & OPJ-04; (b) Primers OPC-03 & 07; (c) Primers OPZ-13 & 14; M: marker (1 kb DNA ladder, GibcoBRL), 1: At. thiooxidans (ATCC 8085), 2: MAAC1, 3: TF4, 4: MAAC2, 5: T.t1, 6: T.t2, 7: At. ferrooxidans (NCIMB 13537), 8: At. ferrooxidans (ATCC 13661) and 9: MT13.

Table 2. Similarity matrix for Nei and Li’s co-efficient of different acidophilic bacteria obtained from RAPD markers.

1 2 3 4 5 6 7 8 9

At. Thiooxidans (ATCC 8085) MAAC1 TF4 MAAC2 T.t1 T.t2 NCIMB 13537 At. ferrooxidans (ATCC 13661) MT13

1

2

3

4

5

6

7

8

0.3564 0.3401 0.3724 0.5792 0.5945 0.3285 0.3276

1 0.5540 0.5175 0.3643 0.3590 0.3555 0.2812

1 0.4702 0.3333 0.3186 0.3477 0.3014

1 0.3538 0.3820 0.4033 0.2870

1 0.8509 0.3965 0.3430

1 0.4046 0.3461

1 0.5094

1

0.3012

0.2859

0.2484

0.2354

0.3682

0.3554

0.3053

0.2899

9

1

648

T. Akbar et al. T.t1

Acknowledgements

T.t2 T.t(ATCC) TF4 MAAC2

This work is a part of Ms Tanveer Akbar’s M.Phil. dissertation and was supported through Pak–Kazakh joint research project funded by Ministry of Science and Technology, Pakistan.

T.f(ATCC) NCIMB 13537 MAAC1 MT13

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Figure 2. Dendrogram of different strains of Acidithiobacillus ferrooxidans, Acidithiobacillus thiooxidans and a moderate thermophile MT13.

quite in line with our anticipations. Biochemical and physiological similarities/differences were also evident in the genetic profile as determined by RAPD analysis among different species and different strains of the same species. Amplification products of the same size from different species represented homologous sequences. In these studies, isolates falling into the same similarity group were screened from different geographic areas. Harrison (1982) constructed homology groups on the basis of DNA–DNA homology and reported that no correlation existed between homology groups and geography. This fact suggests that a similar genotype may prevail in distinct geographic areas due to the presence of similar leaching microenvironments, which may act as a kind of selection pressure for different microbial groups. On the basis of microbiological studies, diversity was expected at both inter-and intra-species levels. Inter-species genomic diversity observed on the basis of RAPD was in accordance to our expectations. However, intra-species genomic diversity was least expected. The existence of genomic variabilities between different strains of the same species of At. ferrooxidans and At. thiooxidans reflected that these groups are heterogeneous. Thus RAPD markers may be used as a quick and reliable alternative for establishing differences among different species as well as among strains of the same species.

References Ghauri, M.A. 1991 A study of the diversity of acidophilic bacteria. Ph.D Thesis, University of Wales, Bangor, UK. Ghauri, M.A., Khalid, A.M., Grant, S., Heaphy, S. & Grant, W.D. 2003 Phylogenetic analysis of different isolates of Sulfobacillus spp. Isolated from uranium-rich environments and recovery of genes using integron-specific primers. Extremophiles 7, 341–345. Harrison, A.P. Jr. 1982 Genomic and physiological diversity amongst strains of Acidithiobacillus ferrooxidans and genomic comparison with Acidithiobacillus thiooxidans. Archives of Microbiology 131, 68–76. Kelly, D.P. & Harrison, A.P. 1989 Aerobic chemolithotrophic bacteria and associated organisms. Genus Thiobacillus, Beijerinck, 1904b, 597 AI. In Bergey’s Manual of Systematic Bacteriology, eds. Staley, J.T. et al. vol.3 pp. 1842–1853. Baltimore: Williams and Wilkins. ISBN 0683079085. Kelly, D.P. & Wood, A.P. 2000 Reclassification of some species of Thiobacillus to the newly designated genera Acidithiobacillus gen. nov., Halothiobacillus gen. nov. and Thermithiobacillus gen. nov. International Journal of Systematic and Evolutionary Microbiology 50, 511–516. Nei, N. & Li, W. 1979 Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences, USA 76, 5269–5273. Novo, M.T.M., Desauza, A.P., Garcia, O. & Ottoboni, L.M.M. 1996 RAPD genomic fingerprinting differentiates Thiobacillus ferrooxidans strains. Systematic and Applied Microbiology 19, 91–95. Norris, P.R. & Johnson D.B. 1998 Acidophilic microorganisms. In Extremophiles: Microbial Life in Extreme Environments eds. Horikoshi, K. & Grant, W.D. pp. 133–153. Wiley-Liss, Inc. ISBN 0-471-02618-2. Pobell, S., Otto, A. & Kutschke, S. 1997 Identification and discrimination of thiobacilli using ARDREA, RAPD and rep-APD. Journal of Applied Microbiology 84, 1085–1091. Sambrook, J. Fritsch, E.F. & Maniatis, T. 1989 Molecular Cloning: A Laboratory Manual, 2nd ed. USA: Cold Spring Harbor Laboratory Press. ISBN 0-87969309-6. Simpson, P.J., Stanton, C., Fitzgerald, G.F. & Ross, R.P. 2002 Genomic diversity within the genus Pedicococcus as revealed by Randomly Amplified Polymorphic DNA, PCR and Pulsed- Field Gel Electrophoresis. Applied and Environmental Microbiology 68, 765–777. Tuovinen, O.H. & Fry, I.J. 1993 Bioleaching and mineral biotechnology. Current Opinion in Biotechnology 4, 344–355. Williams, J.G.K., Kubelik, A.R., Livak, K.J., Rafalski, J.A. & Tingey, S.V. 1989 DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Research 18, 6531–6535.

 Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:649–654 DOI 10.1007/s11274-004-3569-2

Purification and characterization of a thermoalkalophilic xylanase from Bacillus sp. M.P. Sapre, H. Jha and M.B. Patil* Postgraduate Teaching Department of Biochemistry, Nagpur University, Nagpur 440033, India *Author for correspondence: Tel.: +712-223-3904, E-mail: [email protected]

Keywords: alkaliphilic, thermostable, xylanase

Summary An extracellular xylanase was purified to homogeneity from the culture filtrate of a thermophilic Bacillus sp. The molecular weight of the purified xylanase was 44 kDa, as analysed by SDS/PAGE. The enzyme reaction followed app Michaelis–Menten kinetics with Km and Vmax values of 0.025 mg/ml and 450 U/mg protein, respectively, as obtained from a Lineweaver–Burk plot. The xylanase contained no other enzyme activity except for the hydrolysis of xylan substrate. The optimal temperature of the enzyme assay was 50
C. The optimum pH for the xylanase activity was at three peaks 6.5, 8.5 and 10.5, respectively and the enzyme was stable over a broad range of pH from pH 6 to 10.5. Metal ions tested with demetalized enzyme had no effect, with the exception of Hg2+ and Pb2+ (both strong inhibitors). Inhibition of the enzyme activity by N-bromosuccinimide (amino acid modifier) indicated the role of tryptophan residues in the catalytic function of the enzyme. Due to these outstanding properties, the xylanase of Bacillus sp. finds potential applications in biopulping, biobleaching and de-inking of recycled paper and other industrial processes.

Introduction Xylan ranks second only to cellulose in abundance and comprises up to one third of the total weight of higher plants (Lynch et al. 1981; Plus & Poutanen 1989). Knowledge of xylanolytic activity is important to understand biomass recycling in nature and the use of enzymes in controlled lignocellulose conversions. The use of environmentally friendly processes is becoming popular in the pulp and paper industry and therefore, biotechnological processes are coming to the forefront of research. The aim of these processes is to reduce or replace the harmful alkaline extraction of hemicellulose and the need for chlorine in the bleaching process, without affecting the cellulose fibre and thereby the strength of paper products. In addition, the polluting waste effluent can be bio-converted to useful products such as single-cell protein and ethanol. A large number of b-D -xylanases (EC 3.2.1.8) from fungal and bacterial sources have been used to achieve these goals (Paice et al. 1988; Da Silva et al. 1993). The current state of the art of biotechnology in the pulp and paper industry has improved tremendously over the last decade. One such development is the use of xylanase in the bleaching of kraft pulp. The use of xylan- degrading enzymes to aid in the bleaching of chemical pulp was introduced by Viikari et al. (1986). The cellulase-free xylanases with high thermostability are of particular interest for the enzymic hydrolysis or pre-treatment of pulp. The biopulping process operating at higher temperature results in higher reaction

rates and hydrolysis efficiencies. Therefore, xylanases from alkaliphilic bacteria and actinomycetes have been studied widely (Mackenzie et al. 1987, Nakamura et al. 1993; Tasujibo et al. 1997). An indigenously isolated bacterial strain, identified as a Bacillus sp. produced comparable amounts of thermostable xylanases as produced by other Bacillus sp. (Gessesse 1998; Ratankhanokchai et al. 1998). In this paper the purification and biochemical characterization of b-D -xylanase from a Bacillus sp. is reported. Materials and methods Growth conditions Bacillus sp., indigenously isolated in our laboratory was maintained on nutrient agar slants, pH 7.0 at 4
C, by periodic transfer. A loopful of culture was transferred to nutrient broth (2 ml) and incubated at 50
C for 24 h. This was used as inoculum for 1 l fermentation medium in a 5-l flask. The composition of the fermentation medium was (in g l)1): xylan 20, NaNO3 5, KH2PO4 2.5, MgSO4 Æ 7H2O 0.5, serine 5, riboflavin 5, pH 7.0 before autoclaving. The inoculated flasks were incubated for 4 days at 50
C in a laboratory incubator under static conditions. Enzyme purification The purification procedures were carried out by the method of Johnson et al. (1999). For purification

650 purposes the culture filtrate was centrifuged at 10,000 · g for 10 min, at 4
C and the clear supernatant was used. The crude culture filtrate was treated with ammonium sulphate to 80% saturation, the precipitate was collected by centrifugation at 15,000 · g for 15 min. The precipitate was suspended in 10 ml of buffer A (0.02 M sodium phosphate buffer, pH 7.0) and the suspension was dialysed against the same buffer (2 l) for 18 h. The dialysed sample was loaded on to a DEAEcellulose column (1 cm · 50 cm) pH 7.0 and the column was washed until the absorbance at 280 nm remained constant. The proteins were eluted by a linear gradient of 0–1 M NaCl in buffer A. The fractions containing xylanase activity were pooled and dialysed against sucrose for concentraion of the protein, the concentrated pool was further dialysed against buffer A for 2 h to remove unwanted sucrose. The concentrated solution was loaded on to a Sephadex G-75 column (1 cm · 50 cm). The b-D -xylanase was eluted by using buffer A. The purified xylanase was pooled, concentrated and preserved at 4
C until use.

M.P. Sapre et al. To check the stability of xylanase at different pH values the stock enzyme solution was adjusted to different values of pH with buffers having pH values ranging from 2 to 12 and these were than incubated at 50
C for 3 h. Residual enzyme activities were determined in 0.2 M sodium phosphate buffer (pH 6.5) after suitable dilution of the treated enzyme preparations. Also for the determination of the optimum temperature, xylanase activity was assayed over the temperature range of 25, 30, 37, from 40 to 100
C in 0.2 M sodium phosphate buffer (pH 6.5). The stock enzyme was suitably diluted and used in the assay mixture under standard assay conditions. In order to investigate the thermostability of Bacillus b-D -xylanase, aliquots of enzyme were removed for the standard xylanase activity assays after the purified xylanase was incubated at different temperatures for 30 min and the residual enzyme activity was determined under standard assay conditions. Different metal ions and the amino acid modification reagents were preincubated with the standard assay solutions to monitor the effect of all these reagents on xylanase activity.

Enzyme assays The xylanase activity was calculated by measuring the amount of xylose released from 1% (w/v) birchwood xylan according to the method of Mukhopadhyay et al. (1997). The purified or crude xylanase was incubated with a xylan solution (1% w/v) xylan, in 0.2 M sodium phosphate buffer, pH 6.5) at 50
C for 15 min. Dinitrosalicylic acid (2 ml) was added to arrest the reaction. The mixture was boiled for 15 min and allowed to cool naturally. Reducing sugars were determined by measuring the absorbance at 540 nm. One unit of xylanase activity was defined as the amount of enzyme, which produced 1 lmol of reducing sugar, by hydrolysing the xylan substrate, per min at 50
C. The cellulase activity was assayed in a similar manner to the xylanase activity except that 1% (w/v) carboxymethylcellulose was used as the substrate (Denison & Koehn 1977). The b-glucosidase (EC 3.2.1.21), b-xylosidase (EC 3.2.1.37), b-mannosidase (EC 3.2.1.78) and a-L -arabinofuranosidase (EC 3.2.1.55) activities were assayed by incubating for 10 min at 50
C using p-nitrophenyl-b-D -glucopyranoside, p-nitrophenyl-b-D -mannopyranoside, p-nitrophenyl-bD -xylopyranoside and p-nitrophenyl-a-L -arabinofuranoside. Following the addition of two volumes of 1 M sodium carbonate, the product, p-nitrophenol, was quantified at 420 nm using an absorption coefficient of 18.5 ml lmol)1 cm)1 (Herr et al. 1978). The protein content was determined by the Lowry method. For the determination of the optimum pH of the purified enzyme, the activity was checked at different pH values ranging from 2 to 12 using KCl–HCl and universal buffers (citrate/phosphate/borate). Xylanase eluted in the peak fractions of Sephadex G-75 chromatography step during purification of the enzyme was suitably diluted using these buffers. Also the preparation of the substrate solution was carried out using above mentioned buffers.

SDS/polyacrylamide gel electrophoresis SDS/PAGE was carried out as described by Laemmli (1970). Non-denaturing gel electrophoresis was performed using the same method except that SDS was not included in the gel, running buffer and loading buffers, and there was not denaturing procedure. A 7% (w/v) resolving gel was used at all times. Gels were stained using silver nitrate. Determination of the tryptophan content of xylanase with N -bromosuccinimide (NBS) The tryptophan content of xylanase that is believed to be involved in the xylanase activity and/or the substrate binding of the enzyme, was determined by the spectrophotometric method, using NBS as described by Spande & Witkop (1969). The number of tryptophan residues oxidized with NBS per mol of protein was calculated using the following expression: n ¼ ðlmol tryptophan=lmol protein) ¼ ð% decrease A
1:31
MMÞ=ð100
a.f.
5500Þ; ð1Þ where n is the number of tryptophan residues per mole of protein, A the absorbance at 280 nm, a.f. the absorptivity factor to convert absorbance at 280 nm to mg/ml of protein, MM the molecular weight of protein and 5500 is the molar decadic absorptivity for tryptophan at 280 nm. Molecular weight determination The subunit molecular weight of purified b-D -xylanase was determined by SDS/PAGE as described above. The

651

Xylanase from thermoalkalophile native molecular weight was determined by Sephadex G-100 gel filtration chromatography using standards. Blue dextran was used to determine the void volume of the column. Analysis of the birch woodxylan hydrolysis products The products from birchwood xylan hydrolysed by purified b-D -xylanase were analysed by thin layer chromatography (TLC). The hydrolysis was carried out with 200 IU of purified xylanase and 1% birchwood xylan as described above. Aliquots were removed periodically, and 5 ll of each sample was spotted on to the TLC plate (Silica gel G). The end products were separated by the solvent system (ethyl acetate/water/ acetic acid: 10:10:7, by vol.), xylo-oligosaccharides were located using orcinol–sulphuric acid reagent. Xylose and xylobiose were used as standards. Reproducibility of results

Physicochemical properties The subunit molecular weight of the Bacillus sp. xylanase was found to be 44 kDa, based on its mobility in SDS/ PAGE. The native molecular weight was optimized by Sephadex G-100 chromatography to be 44,600 Da. Substrate specificity Bacillus sp. xylanase hydrolysed only xylan and was free from all other enzyme activities examined including those of CM-cellulase, b-glucosidase, b-mannosidase, b-xylosidase, and a-L -arabinofuranosidase. Xylan was hydrolysed into fragments by purified xylanase at the early stage of the hydrolysis process. However, no detectable levels of xylose could be measured during the first 6 h (Figure 1). Significant amounts of low molecular weight compounds, xylobiose and trace amounts of xylose were liberated from xylan after 24 h of incubation. The pattern of hydrolysis appears to be similar to that of an endo-acting xylanase.

Unless otherwise indicated, all values are average values calculated from three independently derived sets of data.

Catalytic properties

Results

The specific activity of xylanase from Bacillus sp. was calculated to be approx. 4.39 units/mg of protein. When

Purification of Bacillus sp. b-D -xylanase The purification of xylanase from a culture filtrate of Bacillus sp. is summarized in Table 1. It was essential to undertake all the purification steps at 4
C, in order to ensure the purity of the final product, b-D -xylanase. About 10% proteins in the crude extract were lost during the ammonium sulphate precipitation step. The specific activity of b-D -xylanase was 44.15 times higher than that after the 80% ammonium sulphate precipitation. From the elution profile of the column chromatography (results not shown) and SDS/PAGE, purified xylanase appeared to be homogeneous. There was only one band detected showing xylanase activity for the purified enzyme in the non-denaturing gel. Purified b-D -xylanase retained its full activity after storage for 3 months at 4
C.

Table 1. Purification chart of xylanase from Bacillus sp. Total Total Total Specific Fold Yield volume activity protein activity purification (%) (ml) (units) (mg) (units/ mg) Crude filtrate 200 Ammonium 40 sulphate fraction DEAE-cellulose 5 fraction Sephadex G-75 2 fraction a

3200 2943

400a 296

2300

15

1800

4.1

8 9.90

1 1.23

100 91.96

153.3

19.16

71.87

439

54.87

56.25

Non-reproducible value due to interfering substances.

Figure 1. SDS/PAGE of b-xylanase purification. Lane 1, Molecular Weight Standards; Lane 2, ammonium sulphate fraction of xylanase; lane 3, DEAE-Cellulose fraction of xylanase enzyme; lane 4, Sephadex G-75 Fraction of xylanase enzyme.

652

M.P. Sapre et al.

Effect of metal ions and chemical modifiers on enzyme activity Most of the inorganic salts, except that of Hg2+ and Pb2+, EDTA, dithiothreitol (D L -DTT), b-mercaptoethanol and common amino acid modification agents such as idoacetate, p-chloromercury benzoate, PMSF, showed no significant effect on xylanase activity. HgCl2 completely inhibited the xylanase activity at a concentration of 4 mM. In both the cases, the addition of 50 mM EDTA did not recover the enzyme activity. The effect of some inducers like glutathione, L-cysteine–HCl, and metal chelators like 1,10-phenanthroline was significant. Glutathione induced the Bacillus sp. xylanase activity by 10% at a concentration of 10 mM. On the other hand 1,10-phenanthroline at a

100

80

60

40

Optimal pH 20

pH Stability of B-D-xylanase from Bacillus sp.

0 2

3

4

4,5

5

6

6,5

7

8

9

10

11

12

pH Max: Activity was - 400 u/mg

Figure 3. Effect of pH on the activity and stability of purified xylanase from Bacillus sp.

concentration of 6 mM showed some inhibitory effect (70% activity retained at 6 mM onwards). NBS (N-bromosuccinimide), capable of oxidizing the tryptophan residues of the protein, was the only amino acid modification reagent which inactivated xylanase. The indole chromophore of tryptophan residue(s) of Bacillus sp. xylanase was oxidized slowly by NBS to oxindole, a much weaker chromophore at 280 nm (Figure 5). The addition of 200 ll of 0.10 M NBS to 2 ml of enzyme at pH 4.0 was enough to inactivate the enzyme completely. The results showed that the modification of 1.64 tryptophan residues, (calculated by using Equation (1)) was sufficient for the inactivation of complete enzyme activity.

120

100

80 Relative Aactivity (%)

120

Relative activity (%)

the concentration of xylan was varied from 0.3 to 5.0 mg/ml, the enzyme reaction followed Michaelis and app Menten kinetics with Km and Vmax values of 0.025 mg/ ml and 450 units/ml/min per mg of protein respectively, using a Lineweaver–Burk plot (results not shown). Bacillus sp. xylanase exhibited maximum activity at 50
C (Figure 2) and at pH 6.5. Purified xylanase was capable of retaining the full activity after incubating at 50
C for more than 23 h (Figure 4) and appeared to be stable when the pH was greater than that of 5.0 (Figure 3). When kept at 30
C Bacillus sp. xylanase retained its complete activity upto 72 h (Figure 4).

60

Discussion

40

Optimal Tempeature

20

Thermostability of B-D-xylanase from Bacillus sp.

0 0

40

50

55

60

70

80

90

100

Temperature (˚C) Max: Activity was - 400 u/mg

Figure 2. Effect of temperature on the activity and thermostability of purified xylanase from Bacillus sp.

Bacillus spp. have been demonstrated to be the most promising potential producers of thermostable, cellulasefree and alkalitolerent xylanases (Shendye & Rao 1993; Gessesse 1998; Ratankonchai et al. 1999). A highly active b-D -xylanase from Bacillus sp. has been successfully purified. Even with a 40–44% loss of total xylanase activity during the purification procedure, more than 4 mg of highly active b-D -xylanase was obtained from 200 ml of crude culture media. Purified xylanase from this local strain showed characteristics similar to that of other Bacillus sp. as studied by Dhillon et al. (2000) and Alexander et al. (1993). The molecular weight of xylanase

653

Xylanase from thermoalkalophile 120

0.6

100 A 0.5

B

Abs.

Relative Aactivity (%)

80

60

0.4

40 50˚C

0.3

60˚C 70˚C 20

0.2 250

260

0 0

1

2

3

4

5

6

7

8

Time (hours) Figure 4. Thermostability of xylanase from Bacillus sp. at pH 10.5. Residual xylanase activity was monitored at various time intervals after incubation at various temperatures.

from Bacillus sp. is similar to those from other Bacillus sp. strains reported in the literature by Akiba & Horikoshi (1988) and Bernier et al. (1983), within the range of 30– 40 kDa. The native molecular weight estimated by gel filtration method was showed little variation from that calculated from SDS/PAGE. Several investigators have reported the abnormal behaviour of b-D -xylanase on gelfiltration columns, (Filho 1993; Belanic et al. 1995). The enzyme probably consists of a single polypeptide chain, since the molecular weight estimated by SDS/PAGE and gel filtration chromatography are not too distinct. Apart from the similarities of the molecular mass values described above, the crude and purified xylanases of all Bacillus sp. strains reported so far and in this study, possess similar range of activity pH (6.0–10.5) and optimum temperature (50
C). Bacillus sp. xylanase exhibited no other activities except b-D -xylanase activity. The results from thin layer chromatography indicated that the b-D -xylanase produced by Bacillus sp. appeared to be an endo-type. The purified xylanase possessed the highest specific activities with birchwood and bagasse xylans compared with those from other strains. These properties make this purified enzyme a potential tool for several industrial applications. NBS inhibited the activity of the purified b-D -xylanase. It is generally believed that one or more tryptophan

270

280

290

300

310

Wavelength (nM)

9 10 11 12 13 24 36

Figure 5. N-Bromosuccinimide-treated xylanase shows decrease in absorbance at 280 nm.

residues are either involved in the binding of the substrate, or directly involved in enzyme xylanase activity (Marui et al. 1985; Keskar et al. 1989; Biswas et al. 1990). Our finding that one or two tryptophan residues oxidized by NBS per enzyme molecule were sufficient to inhibit the enzyme activity completely, indicated the in role in the catalytic process of the enzyme reaction. However, xylan at a concentration of 1 mg/ml gave over 95% protection against this reagent. Clarke (1987) reported that two tryptophan residues in a cellulase from Schizophyllum commune were involved in substrate binding and in the catalytic mechanism, respectively, and furthermore, substrate binding to one of two tryptophan residues exposed the second one to attack by NBS. Acknowledgement Authors thank the Head of the Department, P.G. Department of Biochemistry, Nagpur University, Nagpur, India, for providing laboratory facilities and encouragement. References Akiba, T. & Horikoshi, K. 1988 Xylanases from alkaliphilic thermophilic Bacillus. Methods in Enzymology 160, 655–659.

654 Alexander, K. Iris, A. & Yuval, S. 1993 Purification and characterization of a thermostable xylanase from Bacillus stearothermophilus T-6 Applied and Environmental Microbiology 59(6), 1725–1730. Bernier, R., Desaochers, M., Jurasek, L. & Paice, M.G. 1983 Isolation and characterization of a xylanase from Bacillus subtilis. Applied and Environmental Microbiology 46, 511–514. Belanic, A., Scarpa, J., Peirano, A., Diaz, R., Steiner, J. & Eyzaguirre, J. 1995 Penicillium purpurogenum produces several xylanases: purification and properties of two of the enzymes. Biotechnology 41, 71–79. Biswas, S.R., Jana, S.C., Mishra, A.K. & Nanda, G. 1990 Production, purification and characterization of xylanase from a hyperxylanolytic mutant of Aspergillus ochraceus. Biotechnology and Bioengineering 35, 244–251. Clarke, A.J. 1987 Tryptophan involvement in the function of enzymes and protein harmones as determined by selective oxidation with N-bromosuccinimide. Biochimica et Biophysica Acta 912, 424–431. Da Silva, R., Yim, D.K. & Park, Y.K. 1993 Application af thermostable xylanase from humicala sp. for pulp improvement Journal of Fermentation and Bioengineering 77, 109–111. Denison, D. and Koehn, R.D. 1977 Breakdown of cellulose by yeast species. Mycologia 69, 592. Dhillon, A., Gupta, J.K. & Khanna, S. 2000 Enhanced production, purification and characterization of a novel cellulase-poor thermostable alkalitolerant xylanase from Bacillus circulans AB16. Process Biochemistry 35, 849–856. Filho, F.X., Plus, J. & Coughlan, M.P. 1993 Purification and characterisation of a thermostable xylanase from bacillus sterothermophilus T-6. Journal of industrial Microbiology 2, 171–180. Gessesse, A. 1998 Purification and properties of two thermostable alkaline xylanases from an alkaliphilic Bacillus sp. Applied and Environmental Microbiology 64, 3533–3535. Herr, D., Baumer, F. & Dellweq, H. 1978 Applied Microbiology and Biotechnology 5, 29–36. Johnson, Lin, Larry, M., Singh, S. & Pillay, B. 1999 Purification and biochemical characteristics of b-D -xylanase from a thermophilic fungus, Thermomyces lanuginosus- SSBP. Biotechnology and Applied Biochemistry 30, 73–79. Keskar, S., Srinivasan, C. & Deshpande, V. 1989 Chemical modification of xylanase from thermotolerant Streptomyces. Biochemical Journal 261, 49–55. Laemmli, U.K. 1970 Cleavage of structural protein during the assembly of the head of bacteriophage T4. Nature 227, 680–686.

M.P. Sapre et al. Lynch, J.M., Slater, J.H. & Bennett, J.A. 1981 Cellulase activities of some actinomycetes organisms from soil. Journal of General Microbiology 10, 236–239. Mackenzie, C.R., Bilaus, D., Scheneider, M. & Johnson, K.G. 1987 Induction of cellulolytic and xylanolytic enzyme systems in Streptomyces sp. Applied and Environmental Microbiology 53, 2835–2839. Marui, M., Nacanishi, K. & Yasui, J. 1985 Purification and properties of three types of xylanase induced by methyl-b-D -xyloside from Streptomyces sp. Agricultural and Biological Chemistry 49, 3399– 3407. Mukhopadhyay, S.K., Paul, S., Roy, A. & Chatterjee, S.P. 1997 Indian Journal of Microbiology 37, 77–80. Nakamura, S., Wakabayashi, K., Nakai, R., Anone, R. & Horikoshi, K. 1993 Purification and some properties of an alkaline xylanase from alkaliphilic Bacillus sp. Strain 41M-1. Applied and Environmental Microbiology 59, 2311–2316. Paice, M.G., Bernier, R. & Jurasek, L. 1988 Biosorption of organic liquid in a kraftmill generated lagoon. Biotechnology and Bioengineering 32, 235–239. Plus, J. & Poutanen, K. 1989 Mechanisms of enzymic hydrolysis of hemicelluloses (xylans) and procedures for determination of enzyme activities involved. In Enzyme Systems of Lignocellulosics Degradation, ed. Coughlan, M.P. pp. 151–158. Elsevier. London: ISBN 1-85166411-4 Ratankhanokchai, K., Khin Lay Kyu & Tanticharoen, M. 1998 Purification and properties of a xylan – binding endoxylanase from Alkaliphilic Bacillus sp. strain K-1. Applied and Environmental Microbiology 65, 694–697. Shendye, A. & Rao, M. 1993 Chromosomal gene integration and enhanced xylanase production in an alkaliphilic thermophilic Bacillus sp. (NCIM 59). Biochemical and Biophysical Research Communications 195, 776–784. Spande, T.F. & Witkop, B. 1969 Determination of the tryptophan content of proteins with N-bromosuccinimide. Methods in Enzymology 11, 498–506. Tasujibo, H., Obtsuki, T., Ilo, T., Yamazaki, I., Miyamoto, K., Sugiyama, M. & Inamori, Y. 1997 Cloning & sequence analysis of genes encoding xylanases and acetyl xylan esterases from Streptomyces thermoviolaceus OPC-520. Applied Environmental Microbiology 63, 661–664. Viikari, L., Ranua, M., Kantelinen, A., Sundquist, J. & Linko, M. 1986 Advances in Biochemical Engineering/Biotechnology, Vol. 57, ed. Eriksson, K.-E. p. 261. Germany: Springer-Verlag.


Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:655–659 DOI 10.1007/s11274-004-3570-9

Liquid–liquid extraction of an extracellular alkaline protease from fermentation broth using aqueous two-phase and reversed micelles systems T.S. Porto1,2, T.I.R. Monteiro1,2, K.A. Moreira2, J.L. Lima-Filho1,2, M.P.C. Silva1,2, A.L.F. Porto2,3 and M.G. Carneiro-da-Cunha1,2,* 1 Departamento de Bioquı´mica, Universidade Federal de Pernambuco-UFPE, Av. Prof. Moraes Rego s/n, CEP 50.670420, Recife, PE, Brasil 2 Laborato´rio de Imunopatologia Keizo Asami-LIKA, Universidade Federal de Pernambuco-UFPE, Brasil 3 Departamento de Morfologia e Fisiologia Animal, Universidade Federal Rural de Pernambuco-UFRPE, Brasil *Author for correspondence: Fax: +55-81-32718576, E-mail: [email protected]

Keywords: Aqueous two-phase systems, liquid–liquid extraction, proteases, reversed micelles

Summary The study of recovery of an extracellular alkaline protease from fermentation broth produced by Norcadiopsis sp. was carried out with liquid–liquid extraction through sodium di-(2-ethylhexyl) sulphosuccinate/isooctane reversed micelles systems and aqueous two-phase systems (polyethylene glycol/potassium phosphate). The best conditions for extraction and back-extraction with the reversed micelles system was obtained at pH 9.0 and pH 5.0, respectively, showing a yield of protein of 6.16%, a specific activity of 4.10 U/ml and a purification factor of 1.80. The studies using aqueous two-phase systems of polyethylene glycol/potassium phosphate at pH 10.0 showed purification factors of 2 and 5, and protein yield of 11 and 4%, respectively, for polyethylene glycol 550/potassium phosphate and polyethylene glycol 8000/potassium phosphate. The results indicate that the aqueous two-phase systems are more attractive as a first step in the isolation and purification processes.

Introduction The production of various compounds of industrial interest, such as enzymes, through microbiological processes has greatly increased over the past years. Microbial protease enzymes encompass nearly 60% of the total enzymes used on an industrial scale (Zvidzai & Zvauya 2001; Haki & Rakshit 2003). The development of new recovery and purification techniques for industrial enzymes is very important, thus finding a way to implement them with a certain degree of economy is advisable, as they could turn out to be responsible for up to 70% of the total production cost (Dry & Suttnar 1997). Chromatographic purification processes are expensive. Consequently, alternative techniques such as liquid–liquid extraction with a reversed micelle system or an aqueous two-phases system, have been under study so that some cost reduction and a decrease in processing steps may be achieved (Lima et al. 2002). The reversed micelles are spherical aggregates formed by surfactants in organic solvents. In these micelles the surfactant aliphatic chains are directed towards the exterior (towards the organic solvent) and the polar heads directed towards the interior of the micelles, establishing a polar nucleus that contains an aqueous

microphase. The micellar systems can thus be used to solubilize several highly hydrophilic compounds with large molecules like proteins (Lazarova & Tonova 1999). Liquid–liquid extraction with reversed micelles has basically two steps: extraction and back-extraction. Both steps depend on several factors, which are, pH, ionic strength, surfactant type and addition of cosurfactant (Nascimento et al. 2002). The aqueous two-phase systems are generally formed by an aqueous solution of two polymers or, one polymer and specific salts, above the critical concentration of these components for phase formation (Sebastia˜o et al. 1996). One of the important aqueous two-phase system characteristics is its composition of about 85–99% of water that allows the partition of biomolecules and cellular particles under non-denaturing conditions. The physical properties of aqueous two-phases systems can be changed by manipulation of the concentrations, polymer composition and salts. In this way, the partition of molecules and biological particles can be explored to obtain dissociations that would otherwise be difficult if not altogether impossible to achieve (Venancio et al. 1996). The aim of this work was to study the recovery of an extracellular alkaline protease from its fermentation broth through liquid–liquid extraction, using reversed micellar and aqueous two-phases systems.

656 Materials and methods Materials Reagents Sodium di-(2-ethylhexyl) sulphosuccinate (AOT) with 99% purity, polyethylene glycol (PEG) and azocasein, were from Sigma Chemical Company (USA); bicichoninic acid (BCA) was obtained from Pierce; isooctane, glucose and butanol were from Merck (Germany). All other chemicals were of analytical grade. Methods Microorganism, cell growth and enzyme production The Nocardiopsis sp., isolated from forest soil in Pernambuco, Brazil was maintained at 28 C on ISP-2 agar slants (Pridham et al. 1957) made up of malt extract 1.0% (w/v), yeast extract 0.4% (w/v), agar 2.0% (w/v). The inoculum (105 cells/ml) was prepared in Erlenmeyer flasks (250 ml) containing 50 ml MS-2 medium (Porto et al. 1996), containing the following components: glucose (1.0% w/v), soybean flour (4.0% w/v), MgSO4
7H2O (0.06% w/v), NH4Cl (0.1% w/v), K2HPO4 (0.435%) and 0.1 ml mineral solution (100 mg of the FeSO4, MnCl2
4H2O, ZnSO4, CaCl
H2O distilled water 100 ml, with pH initial 7.0), grown for 48 h at 28 C in an orbital shaker (200 rev/min). After incubation, 12.5 ml of this culture was transferred to an Erlenmeyer flask (500 ml) containing 112.5 ml of MS-2 medium, and incubated again for 48 h at 28 C in an orbital shaker (200 rev/min). The fermentation broth was centrifuged for 15 min at 12,500 · g to remove the biomass and the supernatant was used for the purification process in this study. The alkaline protease is an extracellular enzyme and was produced in the stationary phase of fermentation, after 24 h. Its optimum pH for activity is 10.5 at 50 C (Moreira et al. 2003). Protease liquid–liquid extraction with reversed micelles The reversed micellar system was made by the anionic surfactant, sodium di(2-ethylhexyl) sulphosuccinate, more commonly known as aerosol-OT or AOT, in isooctane. The separation and recovery of the protease by liquid–liquid extraction using a micellar system consisted of two steps. In the first step (extraction), equal volumes (5 ml) of AOT micellar solution (200 mM) and of the fermentation broth (which pH ranged from 5.4 to 9.0 by the addition of 1 M HCl or NaOH) added to this mixture 5% (v/v) butanol (0.5 ml) in a 30 ml beaker under agitation at 700 rev/min, for 10 min at 25 C. The mixture was separated into two phases (residual aqueous phase and micellar phase) by centrifugation for 10 min at 2400 · g. In the second, back-extraction step, the micellar phase containing the solubilized protein was added to an equivalent volume of 200 mM buffered aqueous solution containing 100 mM KCl (sodium citrate pH 5.0, sodium phosphate pH 7.0 and sodium carbonate pH 10.0). The mixture was stirred for 10 min at 700 rev/min at 25 C

T.S. Porto et al. and centrifuged for 10 min at 2400 · g to obtain a distinct phase boundary. The resulting two phases were then carefully separated and analysed for protein concentration. Partition in aqueous two-phase system The aqueous two-phase systems of polyethylene glycol (PEG)/potassium phosphate total (K2HPO4 and KH2PO4) mass system equal to 3 g were formed with the appropriate quantities of PEG (50% w/v), potassium phosphates (40% w/w) and water, in order to obtain the different tie-lines. They were obtained according to data presented by Albertsson (1986) (Table 1). The systems were mechanically stirred for 3 min in a vortex to ensure equilibrium conditions, and then 0.25 ml of fermentation broth, was added and the mixture was centrifuged (1500 · g for 10 min) to separate the phases from each other. Samples of the phases were removed and analysed to quantify the enzymatic activity and total protein concentration. Throughout, the temperature was maintained at ±25 C for all the experiments. Protease assay The protein content was determined by according to the Bradford method for the aqueous two-phases system and according to Smith et al. (1985) for the reversed micelles system, with bovine serum albumin as standard at a range 0–600 lg/ml. Protease activity in the aqueous phases was spectrophotometrically measured according to Leigton et al. (1973). Azocasein (1.0%, w/v, in 200 mM Tris–HCl buffer, pH 7.6) was used as the substrate. One unit of activity is defined as the amount of enzyme that produces an increase in absorbance of 1.0 in 1 h at 440 nm. The specific activity was calculated as the ratio of protease activity to protein mass, which is expressed as U/mg, and the purification factor as the ratio of final specific activity to specific activity of initial aqueous phase (fermentation broth), as follows: Specific activity ¼

Protease activity (U/ml) Protein mass (mg/ml)

Purification factor ¼

Final specific activity Initial specific activity

ð1Þ

ð2Þ

In order to confirm the validity of experimental results, duplicate measurements were done and the mean of these are presented in the results.

Results and discussion Effect of pH on extraction and back-extraction of the protease with reversed micelles The influence of pH on the extracellular protease extraction from the Nocardiopsis sp. fermentation broth

657

Extraction of alkaline protease from broth Table 1. Global composition of the aqueous two-phases systems used for protease extraction from fermentation broth. PEG Concentration (%) (w/w)

550 1000 3350 8000

Tie line 1

Tie line 2

Tie line 3

PEG Potassium phosphate

PEG Potassium phosphate

PEG Potassium phosphate

16.7 16.2 14.0 11.8

17.7 17.7 17.7 14.0

19.7 19.7 19.7 –

14.8 14.3 11.8 9.8

15.7 15.7 15.7 11.8

17.7 17.7 17.7 –

100

10

95

9

90

8

85

7

80

6

75

5

70

4

65

3

60

2

55

1

50

Extracted Protein in Micellar Phase (%)

Activity in Residual Aqueous Phase (%)

(aqueous phase) was evaluated by using the aqueous phase pH, ranging from 5.4 to 9.0 with the addition of 1 M HCl or NaOH. The results presented in Figure 1 show that the quantity of protein extracted to the micellar phase was stable within the pH range of 5.4–7.0 and increased gradually (about 36%) up to pH 9.0 (8.4%). A reduction of 26.9% of protein activity in the residual aqueous phase with the increase of pH up to 9.0 can also be observed, suggesting that this loss of activity was probably due to protease transfer into the micellar phase. Similar cases have been reported for a pure cutinase and from fermentation broth in continuous and discontinuous systems, using micelles of AOT/Isooctane (Carneiro-da-Cunha et al. 1997). In the literature, the pH effect on protein transfer has been reported for several proteins where the pH of the aqueous solution determines the net charge. Electrostatic interactions between the protein and surfactant head groups can favour the transfer of protein into the organic phase (Nascimento et al. 2002; Alves et al. 2003). However, this is not always observed. There have been reports indicating that other forces may also play an important role in the solubilization mechanism, such as hydrophobic interactions (Aires-Barros & Cabral,

0 5

6

8

7

9

10

pH Figure 1. The effect of pH in protease extraction from fermentation broth with reversed micelles of AOT (200 mM)/isooctane. % of Activity in residual aqueous phase (j); % of extracted protein in micellar phase (m).

1991; Pires & Cabral, 1993) and solubilization by an ionpairing mechanism (Hatton et al. 1989). In this work, pH 9.0 was found to be more selective for the transfer of the alkaline protease (8.4%) from fermentation broth to the interior of the micelles. Since the pI of this protease is unknown, it is not possible to affirm what interactions were involved in the transfer mechanism. However it is possible that the extraction occurred by electrostatic interactions. Also, it should be considered that the presence of butanol in the system may change the water properties inside the reversed micelles. Unfortunately, low protein recovery was achieved, although a similar result has been reported in the literature for an alkaline protease from Bacillus sp. fermentation broth (Rahaman et al. 1988) when the same aqueous/organic volume ratios were used for enzyme purification. The protease back-extraction solubilized in the reversed micellar phase from the previous extraction was carried out by using 200 mM of buffered aqueous phase at pH 5.0 (sodium citrate buffer), pH 7.0 (sodium phosphate) and pH 10.0 (sodium carbonate), containing 100 mM KCl (Table 2). The enzyme back-extracted at pH values 5.0, 7.0 and 10.0 presented specific activities of 4.10, 2.43 and 2.62 U/mg, respectively, and the best back-extraction occurred at pH 5.0, with a purification factor of 1.8. At pH values of pH 7.0 and 10.0 it was possible to back-extract 100% of the protein from the micellar phase without any purification, suggesting that the enzyme activity was partially lost, probably due a greater interaction between the protease and micelles. Previous studies in our laboratory have indicated that when the effect of the pH on the extraction and backextraction of the protease was performed by adding 50 mM KCl at initial aqueous phase (fermentation broth) it was not possible to obtain any purification. Recently, this alkaline protease from Nocardiopsis sp., isolated from a soil sample collected from the Northeastern region of Brazil, has been reported to be stable at alkaline pH values between 8.0 and 10.5 at 50 C (Moreira et al. 2003), suggesting that the activity loss is not due to the pH. Similar results were observed in the Bacillus sp. alkaline protease back-extraction (Rahaman et al. 1988) and in the acid phosphatase from Aspergillus niger fermentation broth (Soni & Madamwar 2000), with high values of pH, a decrease of enzymatic activity took place while the total protein recovered increased. This was probably due to interaction of the surfactant with broth constituents (Soni & Madamwar 2000) or to protein instability in presence of organic solvent at high pH values (Rahaman et al. 1988; Krieger et al. 1997). The results were also analysed in terms of protein mass on the back-extraction and, apart from experimental errors, it can be concluded (Table 2) that the protein mass balance can be considered closed. Purification factors of protein similar to the ones obtained in this work (1.8), using micellar systems, are reported in the literature as: 1.5 and 0.15 for a

658

T.S. Porto et al.

Table 2. Effect of pH in the protease back-extraction from the extraction at pH 9.0. Back-extraction pH 5.0

Protein in aqueous phase Protein in micellar phase Protein total

7.0

10.0

mg/ml

%

mg/ml

%

mg/ml

%

0.0440

73.3

0.0657

109.5

0.0609

101.5

0.019

31.6

0.0

0.0

0.0

0.0

–

104.9

–

109.5

–

101.5

U/ml

U/mg

U/ml

U/mg

U/ml

U/mg

0.16

2.43

0.16

2.62

Activity in 0.18 4.09 aqueous phase Purification 1.8 factor

Table 3. Purification factor, partition coefficient and protein yield using aqueous two-phase system at pH 10.0. PEG

Tie line

PFa

KATb

KPTc

Yield (%)

550 550 550 1000 1000 1000 3350 3350 3350 8000 8000

1 2 3 1 2 3 1 2 3 1 2

1 2 2 1 1 2 3 4 5 3 5

2.9 33 44 23 13 30 3.2 5.5 6.5 0.8 1.3

2.3 1.0 33 0.7 0.3 1.0 2.0 0.7 0.5 1.5 2.0

17 10 11 10 15 4 4 2 2 6 4

a b c

1.0

Purification factor. Partition coefficient of total activity. Partition coefficient of total protein.

1.1

Protein concentration in micellar phase: 0.060 mg/ml; Activity in initial aqueous phase: 1.67 U/ml and 2.33 U/mg.

commercial a-amylase and a neutral protease, respectively, (Chang & Chen 1995) and 1.5 for a pre-purified xylanase with ammonium sulphate from fermentation broth (Rodrigues et al. 1999) and 1.8 for an alkaline protease from Bacillus sp. fermentation broth (Rahaman et al. 1988). Although these results are not yet those recommended for enzyme recovery, further efforts should be made to optimize the procedure by changing the other reversed micellar extraction parameters. Protease partition in aqueous two-phases system The protease partition results in aqueous biphasic systems are presented in the Table 3. It can be observed that in a general way, the higher the increase of molecular mass and the length of tie-line, the greater the increase of the coefficient of partition in all the tested systems. These results are corroborated by those obtained by Huddleston et al. (1991), who studied the coefficient of partition of intracellular proteins from Saccharomyces cerevisiae in PEG/phosphate systems, observing for PEG 400 and 1000 preference of proteins for the top phase. Studies of Xu et al. (2002) showed similar behaviour with hexokinase and glucose-6-phosphate dehydrogenase in PEG/PES systems. The importance of the protein molecular weight must also be considered, once the effect of polymer molecular mass is closely connected with the biomaterial molecular mass. In fact, the partition of an amino acid or protein with small molecular mass is not much influenced by variations of polymer molecular mass, which is not the case with proteins with larger molecular mass (Albertsson 1986). Concerning the purification factor for PEG 3350– phosphate salt, an increase can be observed in the top PEG-rich phase with the increase of tie line length. For the PEG 8000–phosphate salt system the purification

factor in the rich phase salt increased with the increase of tie line. The better protease purification factor was detected using the PEG 3350–phosphate salt in tie-line 3 and the PEG 8000–phosphate salt in tie-line 2, both at pH 10.0 with an equal factor of 5.

Conclusion The studies developed in the present work show that it is possible to extract proteases from fermentation broth through the liquid–liquid extraction process by using aqueous two-phases and reversed micelles system. The simplicity of the liquid–liquid extraction process with the aqueous two-phase system, its low cost and effluent generation of low environmental impact and, mainly, the promising results presented in this work when compared with those found with reversed micelles system, makes the two-phase system technique deserving of further development as a first step for the isolation and purification processes of protease enzymes from fermentation broth. Acknowledgements The authors are grateful to CNPq/PIBIC, CAPES, FACEPE and JICA.

References Aires-Barros, M.R. & Cabral, J.M.S. 1991 Selective separation and purification of two lipases from Chromobacterium viscosum using AOT reversed micelles. Biotechnology and Bioengineering 38, 1302– 1307. Albertsson, P.A. 1986 Partition of Cell Particles and Macromolecules, 3rd edn. p. 345, J. Wiley & Sons: New York. ISBN 0471828203. Alves, J.R.S., Fonseca, L.P., Ramalho, M.T. & Cabral, J.M.S. 2003 Optimization of penicillin acylase extraction by AOT/isooctane reversed micellar systems. Biochemical Engineering Journal 15, 81– 86.

Extraction of alkaline protease from broth Carneiro-da-Cunha, M.G., Aires-Barros, M.R., Tambourgi, E.B. & Cabral, J.M.S. 1997 Continuous extraction of a recombinant cutinase with reversed micelles using a perforated rotating disc contactor, Florence, Italy, Proceedings ECCE, 2583–2586. Chang, Q.L. & Chen, J.Y. 1995 Reversed micellar extraction of trypsin effect of solvent on the protein transfer and activity recovery. Biotechnology and Bioengineering 44, 172–174. Dyr, J.E. & Suttnar, J. 1997 Separation used for purification of recombinant proteins. Journal of Chromatography B 699, 383–401. Haki, G.D. & Rakshit, S.K. 2003 Developments in industrially important thermostable enzymes: a review. Bioresource Technology 89, 17–34. Hatton, T.A. Scamehorn, J.F. Horwell, J.H. 1989 Surfactant-based Processes pp. 55–90. Marcel-Dekker Inc: New York. ISBN 0824779290. Huddleston, J.G., Ottomar, K.W., Ngonyani, D.N. & Liddiatt, A. 1991 A influence of system and molecular parameters upon fractionation of intracellular proteins from Saccharomyces by aqueous two-phase partition. Enzyme and Microbial Technology 13, 24–32. Krieger, N., Taipa, M.A., Aires-Barros, M.R., Melo, E.H.M., LimaFilho, J.L. & Cabral, J.M.S. 1997 Purification of Penicillium citrinum lipase using AOT reversed micelles. Journal of Chemical Technology and Biotechnology 69, 77–85. Lazarova, Z. & Tonova, K. 1999 Integrated reversed micellar extraction and stripping of a-amilase. Biotechnology and Bioengineering 63, 583–591. Leigton, T.J., Doi, R.H., Warren, R.A.J. & Kelln, R.A. 1973 The relationship of serine protease activity RNA polymerase modification and sporulation in Bacillus subtilis. Journal of Molecular Biology 76, 103–122. Lima, A.S., Alegre, R.M. & Meirelles, A.J.A. 2002 Partitioning of pectinolytic enzymes in polyethylene glycol/potassium phosphate aqueous two-phase systems. Carboydrate Polymers 50, 63–69. Moreira, K.A.A., Porto, T.S., Teixeira, M.F.S., Porto, A.L.F. & Lima Filho, J.L. 2003 New alkaline protease from Nocardiopsis sp.: partial purification and chracterization. Process Biochemistry 39, 67–72. Nascimento, C.O., Coelho, L.C.B.B., Correia, M.T.S. & Carneiro-daCunha, M.G. 2002 Liquid–liquid extraction of lectin from Cratylia mollis seeds using reversed micelles. Biotechnology Letters 24, 905– 907.

659 Pires, M.J. & Cabral, J.M.S. 1993 Liquid–liquid extraction of a recombinant protein with a reversed micelle phase. Biotechnology Progress 9, 647–650. Porto, A.L.F., Campos-Takaki, G.M. & Lima-Filho, J.L. 1996 Effects of culture conditions on protease production by Streptomyces clavuligerus growing on soy bean flour medium. Applied Biochemistry and Biotechnology 60, 115–122. Pridham, T.G., Anderson, P., Foley, C., Lindenfelser, L.A., Hesseltine, C.W. & Benedict, R.G. 1957 A selection of media for maintenance and taxonomic study of Streptomyces. Antibiotics Annual 947–953. Rahaman, R.S., Chee, J.Y., Cabral, J.M.S. & Hatton, T.A. 1988 Recovery of an extracellular alkaline protease from whole fermentation broth using reversed micelles. Biotechnology Progress 4, 218–242. Rodrigues, E.M.G., Milagres, A.M.F. & Pessoa A.A. Jr. 1999 Xilanase recovery: effects of extraction conditions on the AOTreversed micellar systems using experimental design. Process Biochemistry 34, 121–125. Sebastia˜o, M.J., Cabral, J.M.S. & Aires-Barros, M.R. 1996 Improved purification protocol of a Fusarium solani pisi recombinant cutinase by phase partitioning in aqueous two-phase systems of polyethylene glycol and phosphate. Enzyme and Microbial Technology 18, 251–60. Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gardner, F.H., Provenzano, M.D., Fujimoto, C.K., Goeke, N.M., Olson, B.J. & Klenk, D.C. 1985 Measurement of protein using bicinchoninic acid. Analytical Biochemistry 150, 76–85. Soni, K. & Madamwar, D. 2000 Reversed micellar extraction of an extracellular acid phosphatase from fermentation broth. Process Biochemistry 36, 311–315. Venancio, A., Almeida C. & Teixeira, J.A. 1996 Enzyme purification with aqueous two-phase systems: comparison between systems composed of pure polymers and systems composed of crude polymers. Journal of Chromatography B 680, 131–136. Xu, Y., Vitolo, M., Albuquerque, C.N. & Pessoa A. Jr. 2002 Affinit partitioning of glucose-6-phosphate dehydrogenase and hexokinase in aqueous two-phase systems with free triazine dye ligands. Journal of Chromatography B 780, 53–60. Zvidzai, C.J. & Zvauya, R. 2001 Purification of protease from a alkalophilic Bacillus subtilis CHz1 isolated from a Zimbabwean hot spring Zimbabwe. Journal of Food Biochemistry 25, 1–13.

Ó Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:661–665 DOI 10.1007/s11274-004-3573-6

Bioleaching of nickel from equilibrium fluid catalytic cracking catalysts Oguz Bayraktar Department of Chemical Engineering, Bioreaction Engineering Laboratory, Izmir Institute of Technology, Gu¨lbahc¸e Ko¨yu¨, 35437 Urla-Izmir, Turkey; Tel.: +90 232 750-6287, Fax: +90 232 750-6196, E-mail: [email protected]

Keywords: Aspergillus niger, bioleaching, FCC, fluidized catalytic cracking catalyst, nickel

Summary This study investigates the possibility of reusing metal-contaminated equilibrium fluid catalytic cracking (FCC) catalyst after bioleaching. Leaching with Aspergillus niger culture was found to be more effective in the mobilization of nickel from the catalyst particles compared to chemical leaching with citric acid. Bioleaching achieved 32% nickel removal whereas chemical leaching achieved only 21% nickel removal from catalyst particles. The enhanced nickel removal from the catalysts in the presence of A. niger culture was attributed to the biosorption ability of the fungal mycelium and to the higher local concentration of citric acid on the catalyst surface. It was found that 9% of solubilized nickel in the liquid medium was biosorbed to fungal biomass. After nickel leaching with A. niger culture, the hydrogen-to-methane molar ratio and coke yield, which are the measures of dehydrogenation reactions catalysed by nickel during cracking reactions, decreased significantly.

Introduction The fluid catalytic cracking (FCC) process is used for the conversion of relatively high-temperature boiling hydrocarbons to lower temperature-boiling hydrocarbons in the heating oil and gasoline range. During the cracking process, nickel- and vanadium-containing porphyrins and porphyrin-like complexes from the hydrocarbon feed decompose, leaving metal contaminants on the FCC catalyst surface. These metals increase the production of coke, hydrogen, and light gases formation at the expense of the highly desired gasoline (Cimbalo et al. 1972). After reaction in the fluidized catalytic cracking unit (FCCU) spent catalysts are removed from the unit and replaced by fresh catalysts in order to maintain the desired catalytic activity. These spent catalysts are called equilibrium catalysts (ECat). Spent FCC catalysts constitute a significant portion of the solid wastes generated in the petrochemical industry. Furimsky (1996) pointed out that on a world scale about 400,000 tons of spent catalysts from FCCU are produced annually. Spent catalysts can be used as secondary raw materials, or disposed of in land farming or land filling (Hsu et al. 1996; Su et al. 2001). Since catalysts contain contaminants picked up in the processes for which they were used, there has been an increasing concern about the land disposal of spent catalysts. The ability to restore catalyst activity by using an appropriate demetallization process would greatly reduce toxic waste. Demetallization processes involve the physical

removal of the metal contaminants from the catalysts. The process called ‘Demet’ is claimed to reduce both fresh catalyst addition and spent catalyst disposal when used with any FCC catalyst (Yoo 1998). In this process, first the metals on the catalyst are oxidized, sulphided and chlorinated in the reactor section. Then, the nickel and sodium chlorides are washed from the catalyst. Since chlorination techniques are destructive to zeolitecontaining catalyst, they have received less and less consideration in commercial use. Biotechnological removal of such metal contaminants from used catalysts has not received enough attention. Although the microbial leaching of metals became successful by using chemolithoautotrophic bacteria to solubilize metals from materials containing sulphide and/or ferrous compounds (Sukla et al. 1993; Krebs et al. 1997), only Myerson & Ernst (1985) mentioned the removal of metal impurities from spent catalysts. The use of fungi was thought to be an interesting alternative (Bosecker 1997) since fungi could solubilize metals from ores by the excretion of organic acids. Aspergillus niger and Penicillium simplicissimum are the two widely used species for bioleaching of metals from ores. Bioleaching of valuable elements (Al, Co, Cu, Ni) from ores using micro-fungal species like Penicillia or Aspergilli has been investigated earlier (Sukla et al. 1993). In general a pH range between 2 and 8, a temperature range between 20 and 40 °C, and a high resistance to heavy metals are characteristics of leaching processes with fungi (Burgstaller & Schinner 1993). Fungi have been used

662 in the past for the leaching of carbonaceous low-grade ores and mining wastes (Wenberg et al, 1971), industrial filter dusts (Burgstaller et al. 1992), electronic scrap (Brandl et al. 2001) and coal wastes (Scott et al. 1986). The main objective of this research was the elimination of nickel from equilibrium FCC catalysts so as to recover catalyst activity partially. This study focuses on nickel removal from the metal-contaminated FCC catalysts using A. niger culture. Citric acid constituted the major acid produced from A. niger strain. For this reason citric acid was used in chemical leaching experiments to mimic the chemical action of A. niger culture in order to understand the difference between chemical leaching with citric acid and bioleaching. Besides, the effectiveness of citric acid in solubilizing nickel from ores was reported earlier in the literature (Valix et al. 2001). The equilibrium catalysts being subjected to bioleaching and chemical leaching with citric acid were used in the catalytic cracking of gas oil. During the cracking reactions the hydrogen and coke produced by nickelcatalysed dehydrogenation reactions were measured to compare the effects of both bioleaching and chemical leaching on nickel removal from the FCC catalyst.

Materials and methods Catalyst samples Commercial equilibrium FCC catalysts supplied by Marathon-Ashland Oil Inc., Ashland (USA) were used in bioleaching and cracking experiments. They are named ECat-LOW, ECat-INT and ECat-HIGH based on their nickel concentration. The concentration of nickel in Ecat-LOW, ECat-INT and ECat-HIGH were 300 nickel 900 and 2600 mg kg)1 respectively. Bioleaching and chemical leaching A lyophilized culture of A. niger from Presque Isle Cultures (Erie, USA) was used, in all bioleaching experiments. First, the lyophilized culture was revived properly and growth of the fungus was observed and monitored by measuring the pH for 21 days. The organism was cultivated on a sucrose medium in baffled 250-ml Erlenmeyer flasks Wid incubated at 30 °C on an orbital shaker at 150 rev min)1. The growth medium contained (in g l)1) sucrose (100), NaNO3 (1.5), KH2PO4 (0.5), MgSO4 Æ 7H2O (0.025), KCl (0.025), and yeast extract (1.6). The medium was sterilized by autoclaving at 120 °C to reduce the interference from other microorganisms. The fungal culture was grown in the presence of different amounts of ECat-HIGH catalyst. The pulp densities used in the inhibition experiments were 5 l)1and 10 g l)1. Inoculation was conducted by adding the fungal spores to the mixture of catalyst and growth medium. Fungal growth in the presence of equilibrium catalysts sample was monitored by measuring the pH of the medium for 21 days. In all

O. Bayraktar et al. cases the initial pH of the growth medium was adjusted to a value of 6.5. All of the bioleaching experiments were conducted at a pulp density of 5 g l)1. The inhibition and control (only culture medium without fungi) experiments were conducted using only the highly nickelcontaminated (ECat-HIGH) catalyst sample. However, both bioleaching and chemical leaching with citric acid experiments were carried out with three catalyst samples named ECat-LOW, ECat-INT and ECat-HIGH based on their nickel concentration. Controlled chemical leaching experiments with citric acid were also conducted. Citric acid was used because of its ability for solubilizing nickel. This acid also constitutes the major acid produced from the fungus strains. Chemical leaching of the spent FCC catalyst was achieved by adding a solution 39 g citric acid l)1 to equilibrium catalyst particles. This concentration value was selected since this was the highest citric acid concentration in the A. niger culture at the end of the 21 day incubation period. The slurry mixture was kept at 30 °C on an orbital shaker (150 rev min)1) and leaching was carried out for a period of 21 days. In both chemical and bioleaching experiments, catalyst samples were taken at the end of the incubation period, and were filtered and then washed with distilled water. The amount of nickel on the catalyst particles was determined using inductively coupled plasma-atomic absorption spectroscopy (ICP-AES) with standard procedures. 0.5 g catalyst sample was weighed into a platinum dish and approximately 20 ml of deionized water was added. After slowly adding 3 ml of concentrated hydrofluoric acid, 5 ml concentrated sulphuric acid and l ml concentrated nitric acid were subsequently added to the platinum dish. Volatile compounds were fumed off until the residue was slightly moist with sulphuric acid. After cooling down, approximately 40 ml of deionized water and l5 ml of hydrochloric acid were added to the platinum dish, wich was warmed until the sample was completely dissolved. The solution was tranferred into a 100 ml volumetric flask and diluted with appropriate amount of deionized water. In a second platinum dish an analogous mixture of deionized water, hydrofluoric acid, sulphuric acid and nitric acid was fumed off and taken up with hydrochloric acid. The solution was transferred into a 100 ml volumetric flask. After appropriate dilution with deionized water it was used as a blank sample. Similarly, acid digestion was used using nitric acid in combination with perchloric acid to determine nickel in the biomass sample. The amount of citric acid produced by A. niger cultures at the end of 21 days was measured by a spectrophotometric method (Marier & Boulet 1958). Microactivity test (MAT) evaluation MAT measurements of catalyst activity were performed according to the procedures of ASTM D-3907 using sour imported heavy gas oil (Davison Chemical Co.) feed. The MAT measurements for catalyst samples

663

Bioleaching of FCC catalysts named ECat-LOW, ECat-INT and ECat-HIGH based on their nickel concentration were carried out before and after bioleaching experiments. The MAT unit was manufactured by Industrial Automated Systems (Parlin, New Jersey). The microactivity test unit consists of a furnace for maintaining constant cracking temperature (500 °C), a syringe pump whose speed can be varied to inject the cracking feed, a liquid product receiver and a gas collection vessel. Liquid products were condensed at the exit of the reactor by using an ice-water mixture. Product gases were collected by downward displacement of water. To achieve different conversions the reaction severity was varied by changing the catalyst-to-oil ratio (C/O) where C/O ratio is defined as the amount of catalyst divided by the total amount of feed. This ratio was changed by changing the amount of feed while keeping the catalyst amount constant at 5 g. C/O ratios of 3, 5, and 8 were used in the cracking experiments. The distribution of gaseous products was analysed by using a Varian 3600 gas chromatograph. The H2, C1, C2, C3, C4, and Cþ 5 lumps were determined quantitatively. The liquid products, after weighing, were dissolved in CS2 solvent and analysed by GC simulated distillation. The D2887 simulated distillation method was used to determine the percentage of the liquid products boiling in the range of gasoline (IBP – 421 F), light cycle oil, LCO (421–640 F), and heavy cycle oil, HCO (640 F and above). A Supelco Petrocol B packed column installed in a Varain 3400 GC was used. Using this method, the conversion as defined by ASTM 3907 was calculated. The yields were calculated as weight percent of reactant. The fractions of gasoline, LCO and HCO were determined by the cut-off points at 421 and 640 F. Conversion was defined as 100 wt% ff (LCO, wt% ff + HCO, wt% ff). The amount of coke on the spent catalyst was determined by temperature-programmed oxidation (Bayraktar & Kugler 2003). In this study, all experiments were performed at least three times and average values were taken as data points.

Results and discussion Depostion of metal contaminants (e.g., nickel, vanadium and iron) on the catalyst causes a decrease in catalyst activity. Both coke and hydrogen productions increase at the expense of gasoline. The negative effect of metal contaminants on the yields can be seen from Table 1. Data presented in Table 1 were obtained at a constant MAT conversion of 50%. MAT conversion can be defined as percentage of the feed converted to gas, gasoline and coke during catalytic cracking reactions. The yields of valuable products, gasoline and LCO, decreased whereas coke yield increased when the nickel content on the cracking catalyst increased from 300 mg kg)1 (ECat-LOW) to 2600 mg kg)1 (ECatHIGH). With increasing nickel content on the catalyst, gasoline yield decreased from 38.43% (untreated ECatLOW) to 34.47% (untreated ECat-HIGH). Similarly, LCO yield decreased from 30.87 to 28.74%. On the other hand, coke yield increased significantly from 2.73 to 7.15%. Studies for deposition of these metals on the catalyst showed that nickel was in general homogeneously distributed throughout the catalyst surface (Kugler & Leta 1988). Nickel was found to be primarily responsible for promoting the dehydrogenation activity which produced coke and gaseous products, in particular, hydrogen and methane at the expense of desired liquid products (Venuto & Habib 1979). H2 yields from MAT experiments have usually been reported by using hydrogen-to-methane (H2/CH4) molar ratio. Hydrogen production was found to be very sensitive to catalyst metal contamination (Venuto & Habib 1979). At low catalyst metals level (3.5%) is still essential for cell viability and activity. Maximum yield coefficient Y(X/N)max of 32.8 g of dried biomass per g nitrogen identified for C. oleophila was comparable with the 28.2 g biomass/g NH3 reported by Klasson et al. (1989). Under excess of ammonia this value was only 9.2 g biomass/g NH3. The model used in our work was unable to identify changes in nitrogen content during the entire range of RT. At low RT nitrogen concentrations above 50% came out, although maximum nitrogen content in biomass does not exceed 8– 10% (Dellweg 1987). The Monod-model could not be used for determination of maximum growth rate, because of total exhaustion of nitrogen over almost the entire range of RT. Furthermore, the variables glucose and biomass concentration used in the model depend both on each other and on the nitrogen supply also. The BSPs achieved during present work were three times higher compared with literature values of between

S. Anastassiadis et al. 0.0249 and 0.045 g/(g h), reported for other yeast strains in batch, fed-batch, semi- or continuous cultures (Aiba & Matsuoka 1979; Briffaud & Engasser 1979; Maddox & Kingston 1983; Enzminger & Asenjo 1986; Rane & Sims 1995). Even higher specific and volumetric productivities and product concentrations were obtained after process optimization and process development (see Anastassiadis 1994a, b; Anastassiadis et al. 1993, 1999) using yeast strains. More than 230 g of citric acid/l were produced in repeated fed batch fermentations (unpublished data obtained by Anastassiadis & Stephanopoulos at M.I.T., USA, 1995) and 220–250 g/l in a continuous mode using an ATCC strain within 50 to 130 h, running for several months at very high yields (unpublished data obtained by Anastassiadis at Research in Biotechnology Co, Greece, 2004).

Conclusion Concluding from the above results, continuous fermentation in the chemostat seems to be a very good tool for studying and understanding complex microbial systems and for sophisticated process optimization and development as well. Continuous citric acid production by yeasts is feasible and very promising for future industrial applications, even in today’s highly competitive industry. Biomass-specific nitrogen feed rate is the most important influencing factor on continuous citric acid production in yeasts. The concentration of trace elements and other important nutrients has to be adapted to nitrogen concentration in order to fulfil the microorganism’s requirements for optimum production. Further medium optimization and process development in CSTR experiments increased productivity in continuous citric acid fermentation by C. oleophila, substantially (paper in preparation). Glucose concentrations higher than 450 g/l were utilized in continuous fed-batch fermentations under optimized fermentation conditions. Final citric acid concentrations of 120–150 g/l that were achieved in chemostat cultures by C. oleophila (Anastassiadis et al. 1993, 1994, 2001; Anastassiadis 1994a, b), have now been surpassed by the more than 230 g/l achieved in continuous repeated fed-batch operations and 200 g/l in the chemostat (unpublished data of S. Anastassiadis & G. Stephanopoulos; Chemical Engineering Department, M.I.T., Boston, USA, 1995) and 230–250 g/l produced within 50–130 h for several months (S. Anastassiadis 2004, Research in Biotechnology Co. Greece 2004) show the importance of these very time-consuming preliminary experimental studies in the chemostat. In today’s industry, 150–180 g of citric acid/l are produced in traditional batch or fed-batch fermentations using A. niger within more than 6 days in addition to sporulation and precultivation times, and preparation efforts for next fermentation run, including energy and time consuming cleaning and sterilization of fermenter.

Continuous citric acid fermentation by C. oleophila Acknowledgement We thank Professor Dr U. Stottmeister (UFZ Ctr. Envtl. Res. Leipzig-Halle, Germany) for his helpful advices and support. The experiments of the present manuscript comply with the currant laws of the country Germany, where the experiments were performed (Institute of Biotechnology 2 of Research Center Ju¨lich 2, RCJ; formerly known as KFA). References Aiba, S. & Matsuoka, M. 1978 Citrate production from n-alkanes by Candida lipolytica in reference to carbon fluxes in vivo. European Journal of Applied Microbiology and Biotechnology 5, 247–261. Aiba, S. & Matsuoka, M. 1979 Identification of metabolic model: citrate production from glucose by Candida lipolytica. Biotechnology and Bioengineering 21, 1373–1386. Akiyama, S.-I., Suzuki, T., Sumino, Y., Nakao, Y. & Fukuda, H. 1973a Induction and citric acid productivity of fluoroacetatesensitive mutant strains of Candida lipolytica. Agricultural and Biological Chemistry 37, 879–884. Akiyama, S.-I., Suzuki, T., Sumino, Y., Nakao, Y. & Fukuda, H. 1973b Relationship between aconitase hydratase activity and citric acid productivity in fluoroacetate-sensitive mutant strain of Candida lipolytica. Agricultural and Biological Chemistry 37, 885–888. Anastassiadis, S. 1993 Determination of organic acids, especially citric acid and isocitric acid, in fermentation solutions and fruit juices. In: HPLC Applications (1993). p. 4. Du¨ren, Germany: MACHEREY-NAGEL GmbH & Co. KG, Application No. 8. Anastassiadis, S. 1994a Kontinuierliche Fermentation von Glucon- und Citronensa¨ure mit hefea¨hnlichen Pilzen und Hefen. PhD thesis, Westfa¨lische Wilhelms-Universita¨t Mu¨nster, Germany. Anastassiadis, S. 1994b Zymotiki methodos gia tin sinechi paragogi tou kitrikou oxeos; Process for the continuous production of citric acid by fermentation. Greek Patent No. 940100098 (24.02.1994). Anastassiadis, S., Aivasidis, A. & Wandrey, C. 1993 Fermentationsverfahren zur kontinuierlichen Citronensa¨uregewinnung. German Patent P 43 17 488.4–09. Anastassiadis, S., Aivasidis, A. & Wandrey, C. 1994 Fermentationsverfahren zur kontinuierlichen Citronensa¨uregewinnung. (Process for the continuous production of citric acid by fermentation.) Austrian Patent No. 473/94 (03/07/1994). Anastassiadis, S., Aivasidis, A. & Wandrey, C. 2001 Process for the continuous production of citric acid by fermentation. US Patent No. 08/208,123 (03/08/1994). Anastassiadis, S., Aivasidis, A. & Wandrey, C. 2002 Citric acid production by Candida strains under intracellular nitrogen limitation. Applied Microbiology and Biotechnology 60, 81–87. Antonucci, S., Bravi, M., Bubbico, R., Di Michele, A. & Verdone, N. 2001 Selectivity in citric acid production by Yarrowia lipolytica. Enzyme and Microbial Technology 28, 189–195. Behrens, U., Weissbrodt, E. & Lehmann, W. 1978 Zur Kinetik der Citronensa¨urebildung bei Candida lipolytica. Zeitschrift fu¨r Allgemeine Mikrobiologie 18, 549–558. Behrens, U., Thiersch, A., Weissbrodt, E. & Stottmeister, U. 1987 Particularities in the kinetics of growth and citric acid accumulation by Saccharomycopsis lipolytica. Acta Biotechnologica 7, 179– 183. Briffaud, J. & Engasser, J.M. 1979 Citric acid production from glucose, I. Growth and excretion kinetics in a stirred fermentor. Biotechnology and Bioengineering 21, 2083–2092.

705 Crolla, A. & Kennedy, K.J. 2001 Optimization of citric acid production from Candida lipolytica Y-1095 using n-paraffin. Journal of Biotechnology 89, 27–40. Dellweg, H.-D. 1987 Biotechnologie – Grundlagen und Verfahren. Verlag Chemie, Weinheim. ISBN 3527265333. Einsele, A. & Fiechter, A. 1971 Novel energy and carbon sources. B. Liquid and solid hydrocarbons. Advances in Biochemical Engineering 1, 168–194. Enzminger, J.D. & Asenjo, J.A. 1986 Use of cell recycle in the aerobic fermentative production of citric acid by yeast. Biotechnology Letters 8, 7–12. Kim, E.K., Ambriano, J.R. & Roberts, R.S. 1987 Vigorous stationary phase fermentation. Biotechnology and Bioengineering 30, 805–808. Kim, E.K. & Roberts, R.S. 1991 Rate equations for the vigorous stationary phase fermentation of citric acid by Saccharomycopsis lipolytica. Biotechnology and Bioengineering 37, 985–988. Klasson, T.K., Clausen, E.C. & Gaddy, J.C. 1989 Continuous fermentation for the production of citric acid from glucose. Applied Biochemistry and Biotechnology 20/21, 491–505. Kristiansen, B. & Sinclair C.G. 1979 Production of citric acid in continuous culture. Biotechnology and Bioengineering 21, 296–315. Kubicek, C.P. & Ro¨hr, M. 1980 Regulation of citrate synthase from the citric acid accumulating fungus, Aspergillus niger. Biochimica et Biophysica Acta 615, 449–457. Lockwood, L.B. 1979 Production of organic acids by fermentation. In: Microbial Technology vol. 1 (Chapter 11), eds. Peppler H.J. and Perlman D. Academic Press, New York, pp. 355–387. ISBN 0125515014. Lozinov, A.B. & Finogenova, T.V. 1982 Einfluß der Limitation des Wachstums von Hefen auf den oxidativen Stoffwechsel und die Produktsynthese. Acta Biotechnologica 2, 317–324. Maddox, I. & Kingston, P. 1983 Use of immobilized cells of the yeast, Saccharomycopsis lipolytica, for the production of citric acid. Biotechnology Letters 5, 795–798. Mayilvahanan, D., Annadurai, G., Raju V., Chellapandian M., Krishnan M.R.V. & Jayaraman K. 1996 Citric acid production, Part 1: Strategies for reduction in cycle time for targeted yields. Bioprocess Engineering 15, 323–326. Milson, P.E. 1987 Organic acid fermentation, especially citric acid. Food Biotechnology 1, 273–307. Rane, K.D. & Sims, K.A. 1995 Citric acid production by Candida lipolytica Y1095 in cell recycle and fed-batch fermentors. Biotechnology and Bioengineering 46, 325–332. Saha, M.L. & Takahashi, F. 1997 Continuous citric acid fermentation by magnetic rotating biological contactors using Aspergillus niger AJ 117173. Journal of Fermentation and Bioengineering 84, 244– 248. Stottmeister, U., Behrens, U., Weissbrodt, E., Weizenbeck, E., Du¨resch, R., Kaiser, M., No¨lte, D., Richter, H.-P., Schmidt, J., Kochmann, W., May, U., Kreibich, G. & Scho¨ppe, G. 1986. Patentschrift DD 239 610 A1, ISSN 0433-6461. Stottmeister, U. & Hoppe, K. 1991 Organische Genußsa¨uren. In Lebensmittelbiotechnologie, Entwicklungen und Aspekte, 1st edn. Ruttloff, H. pp. 516–547 Akademie Verlag. Stottmeister, U. & Weissbrodt, E. 1991 Product formation by Yarrowia lipolytica – some generalizing aspects. In Metabolismus von Alkanen und ihre mikrobiellen Syntheseprodukte, eds. Finogenova TV and Scharyschev, A.A. Puschtschino: Akademie-Verlag UDSSR. Vassilev, N., Baca, M.T., Vassileva, M., Franco, I. & Azcon, R. 1995 Rock phosphate solubilization by Aspergillus niger grown on sugar-beet waste medium. Applied Microbiology and Biotechnology 44, 546–549. Vassileva, M., Azcon, R., Barea, J.-M., Vassilev, N. 2000 Rock phosphate solubilization by free and encapsulated cells of Yarrowia lipolytica. Process Biochemistry 35, 693–697.

 Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:707–715 DOI 10.1007/s11274-004-4490-4

The influence of different yeasts on the fermentation, composition and sensory quality of cachac¸a Evelyn Souza Oliveira1,*, Helena Maria Andre´ Bolini Cardello2, Elisangela Marques Jeronimo3, Elson Luiz Rocha Souza4 and Gil Eduardo Serra3 1 Departamento de Alimentos, Faculdade de Farma´cia, Universidade Federal de Minas Gerais, Brasil 2 Departamento de Nutric¸a˜o, Faculdade de Engenharia de Alimentos, Universidade Estadual de Campinas, Brasil 3 Departamento de Tecnologia de Alimentos, Faculdade de Engenharia de Alimentos, Universidade Estadual de Campinas, Brasil 4 Laborato´rio de Ana´lises de Bebidas e Vinagres de Andradas, Ministe´rio da Agricultura e do Abastecimento, Brasil *Author for correspondence: Av. Antoˆnio Carlos, 6627, Cidade Universita´ria – Pampulha, Belo Horizonte, Minas Gerais-Brasil, CEP 31.270-901, Brasil, Tel.: +31-3499-6915, Fax: +31-3499-6730, E-mail: [email protected]

Keywords: Alcoholic fermentation, cachac¸a, sensorial analysis, volatile compounds, yeasts

Summary Cachac¸a (sugarcane wine) was produced using different yeast strains, six being strains of Saccharomyces cerevisiae and one each of Candida apicola, Hanseniaspora occidentalis, Pichia subpelliculosa and Schizosaccharomyces pombe. The ethanol yields (%) of the non-Saccharomyces strains were similar to those of the Saccharomyces strains. The following determinations were carried out on the cachac¸a: acetaldehyde, ethyl acetate, propanol, isobutyl alcohol, isoamyl alcohol, volatile acidity. The cachac¸as showed variations in the levels of secondary compounds, but these variations did not result in differences (P £ 0.05) in the sensory attributes of aroma and flavour and overall impression. Of the volatile compounds quantified in the cachac¸as, only propanol showed a positive correlation (P £ 0.05) with the flavour attributes and overall impression. The S. pombe strain was considered inadequate for the production of cachac¸a. The cachac¸as were classified into five groups in an exploratory Hierarchical Cluster Analysis as a function of the volatile compounds. Principal Component Analysis showed that 93% of the variation (PC 1) occurred among the samples, and was explained by the individual volatile compounds and the total secondary compounds, with the exception of isoamyl alcohol only 7% (PC 2) was associated with the volatile acidity. The negative correlations shown between the volatile compounds of the cachac¸as and the ethanol content of the sugarcane wine, with the exception of acetaldehyde, showed that the variation in ethanol content of the sugarcane wine is an important factor for standardization of the ethanol/volatiles ratio and the beverage quality.

Introduction Cachac¸a is a typical and exclusive type of beverage with peculiar sensory characteristics produced in Brazil. It has an alcohol concentration of 38–48% (v/v) at 20
C (Brasil 2003). It is obtained by distillation of the fermented must from sugarcane juice. Brazilian production of cachac¸a is estimated at 1.3 · 109 l per year. Various factors interfere in the quality of distilled alcoholic beverages, such as the raw material and the fermentation, distillation and ageing processes etc. However, the yeasts and the fermentation conditions have been indicated as the factors which most influence the flavour of alcoholic beverages, since the majority of the flavour compounds are formed during fermentation (Suomalainen & Lehtonen 1979; Lehtonen & JounelaEriksson 1983).

The flavour of alcoholic beverages is due to numerous volatile and non-volatile compounds which confer the typical taste and odour of the beverage. These compounds can be divided into various groups according to their chemical nature. Higher alcohols (fusel oil), fatty acids and esters form the largest groups in alcoholic beverages, both quantitatively and qualitatively, the higher alcohols being the most abundant (Lehtonen & Jounela-Eriksson 1983; Berry 1995). According to Cole & Noble (1995), with few exceptions, the flavour in complex systems such as alcoholic beverages cannot be inferred from analyses, since the perceived flavour is the result of a combination of various components, instead of that of an impact compound. Despite the fact that different beverages are easily distinguished from one another by their sensory characteristics, no great differences in their chemical compositions are observed. The

708 most important difference appears to be in the quantities of the aroma compounds (Suomalainen & Lehtonen 1979). Studies in foods and beverages aimed at relating the compounds responsible for their flavour to product quality are normally monitored by the sensory quality, carried out directly according to human perception. This method is still the only way of evaluating the acceptability of these products (Stone & Sidel 1998). The literature encountered with respect to the sensory quality of cachac¸a, mostly refer to aged cachac¸as (Cardello & Faria 1998). With respect to the sensory quality of recently distilled cachac¸as, there are few published papers. Oliveira (2001) evaluated 24 yeast strains belonging to the species Saccharomyces cerevisiae and six from other species, isolated from fermentations in cachac¸a distilleries. The Hierarchical Cluster Analysis showed that these strains could be separated into different groups according to their similarity with respect to the fermentation parameters (the yield was the parameter of greatest impact in differentiating the strains) and in relation to the formation of volatile compounds (acetic acid was the component of greatest impact). In continuing this study, fermentations using 10 Saccharomyces cerevisiae and non-Saccharomyces strains belonging to different groups were performed with the objective of determining the influence of these strains on the composition and sensory quality of recently distilled cachac¸a produced on a pilot plant scale.

Material and methods Microorganisms The cachac¸as were produced from different yeast strains, six being Saccharomyces cerevisiae and four non-Saccharomyces. All the strains were isolated from small cachac¸a distilleries in the state of Minas Gerais and selected from previous studies as a function of their fermentative characteristics (Oliveira et al. 2004) and their capacities to form volatile compounds in synthetic media (Oliveira 2001). The majority of the strains came from the culture collection of the Federal University of Minas Gerais (UFMG: Sc3 (UFMG-A1605) S. cerevisiae; Sc4 (UFMG-A1667) S. cerevisiae; Sc6 (UFMG-A1676) S. cerevisiae; Sc8 (UFMG-A905) S. cerevisiae; Sc13 (UFMG-A1492) S. cerevisiae; Sc21 (UNICAMP-AV3) S. cerevisiae; Ca (UFMG-A972) Candida apicola; Ho (UFMG-A893) Hanseniaspora occidentalis; Ps (UFMG-A960) Pichia subpelliculosa; Sp (UFMG-A1113) Schizosacharomyces pombe. Inoculum preparation To prepare the inoculum, each of the strains maintained in slant culture in GYMP agar (2% glucose, 0.5% yeast extract, 1% malt extract, 0.2% NaH2PO4 and 2% agar)

E. S. Oliveira et al. was aseptically transferred to six flasks containing 100 ml of sterile synthetic medium composed of 40 g/l glucose; 5.0 g/l potassium diphosphate; 5.0 g/l ammonium chloride; 1.0 g/l magnesium sulphate heptahydrate; 1.0 g/l potassium chloride; 6.0 g/l yeast extract, pH 6.0. The flasks were incubated at 30
C, and agitated at 150 rev/ min for 24 h. The contents of each flask were then aseptically transferred to six 2-l flasks containing 1000 ml of the same sterile culture medium with the addition of 60 g glucose/l, and inoculated as before. The cell mass was separated by centrifugation, re-suspended in distilled water, washed twice and then used as the inoculum for the fermentation of cachac¸a on a pilot plant scale. Fermentation experiments The sugarcane juice was extracted immediately after cutting in a previously desinfected industrial grinder. The roots and tips of the cane were excluded before extraction and the surface of the cane was washed with potable water. The juice was sieved through a strainer to remove bagasse, diluted with drinking water to 15
Brix, and inoculated with a suspension containing sufficient cells to give between 107 and 108 cells/ml of juice, with a cell viability of about 99%. After inoculation, the juice was homogenized and a sample was collected for control analyses, such as the soluble solids content (
Brix), total reducing sugars (TRS), cell count and determination of cell viability. The fermentations were conducted in a simple batch system in the pilot plant of the Department of Food Technology of the Faculty of Food Engineering/UNICAMP. A cylindrical stainless steel tank (diameter 50 cm and height 54 cm) with the 40 l must at an average temperature of 32
C was used. The termination of fermentation was determined by the stabilization of the
Brix reading, that is, when two consecutive readings gave the same value, with an interval of 1 h between samplings. At the end of the fermentation, a sample of the sugarcane wine was collected for the determination of ethanol. Distillation The distillation of each fermentation broth was carried out in a copper still with a working capacity of 30 l, equipped with a condenser and gas heater. The temperature of the sugarcane wine was controlled between about 91 and 97
C to maintain the distillation rate at about 1 l/h. The head fraction was collected separately and standardized at a fixed volume of 250 ml (corresponding to about 5% of the total volume of cachac¸a), thus equalizing the distillation time of the head components. The heart fraction was then collected until an ethanol concentration of about 39% v/v was reached. The cachac¸as were stored in glass bottles with caps and plastic coverings and maintained at room temperature in a cool place without extreme environmental changes for later sampling and sensory analysis.

709

Sugarcane wine fermentation Analytical determinations The TRS were quantified using DNS (dinitrosalicylic acid) (Miller 1959) and ethanol in the sugarcane wine by the modified potassium dichromate spectrophotometric method (Zimmermann 1963), based on the oxidation of ethanol to acetic acid using potassium dichromate and sulphuric acid; these reagents turn the solution into a greenish colour directly proportional to the ethanol concentration, and at 600 nm wave-length the Cr+3 can be read without interference. The cell count and cell viability were determined using an optical microscope and a Neubauer chamber and the dye erythrosin, according to Bonneu et al. (1991). The secondary compounds in the cachac¸a were determined by GLC. For the determination of volatile acidity, the samples of cachac¸a were first steam distilled in an Enochimico Gibertini distilling apparatus and the distillate was titrated with 0.1 M NaOH using phenolphthalein as indicator. The ethanol concentration was determined in an A. PAAR model DMA 48 densitometer. Acetaldehyde, ethyl acetate, propanol, isobutyl alcohol and isoamyl alcohol were determined in the distillate by GLC using a Shimadzu model GC-17-A gas chromatograph with automatic injection, a flame ionization detector and a DB-WAX capillary chromatographic column (30 m · 0.25 mm · 0.25 lm). The analyses were conducted under the following conditions: injector temperature, 180
C; detector temperature, 190
C; temperature programme for the column as follows, 40
C (5 min), 20
C/min up to 120
C (1 min), 30
C/ min to 180
C (1 min), giving a total run time of 13 min; the carrier gas was nitrogen; split mode injection of 1:15 and sample volume of 1.0 ll. The compounds were quantified using the external standard technique, the standards being prepared in a 40% v/v ethanol solution. Sensory analysis The 10 cachac¸a samples were submitted to sensory acceptance tests using a 9-point hedonic scale (Stone & Sidel 1993). A panel of 41 judges evaluated the samples with respect to aroma, flavour and overall impression. The evaluations were carried out in individual booths in a sensory analysis laboratory. The samples were served in transparent, colourless glass containers covered with a watch glass and codified with three digits. All the samples were presented in a monadic manner in a total of 10 sessions per judge.

(SAS Institute 1993) were applied to the data obtained in the sensory analysis, and the normal distribution of the scores given by the judges was verified with respect to the hedonic scale used.

Results and discussion Ethanol yield (%)of fermentation The Saccharomyces cerevisiae and Pichia subpelliculosa strains presented lower ethanol yields (percentage of theoretical maximum ethanol yield) in the pilot scale sugarcane fermentations than in synthetic medium, results also obtained by Oliveira et al. 2004. In contrast, the non-Saccharomyces (Candida apicola, Hanseniaspora occidentalis and Schizosaccharomyces pombe) strains presented higher ethanol yields in fermentations conducted in sugarcane juice, certainly because of the presence of nutrients which favoured the activity of these yeasts or because of the lower content of dissolved oxygen (no shaking). A positive correlation (r ¼ +0.81, P £ 0.1) was observed between the ethanol yield on a pilot scale using sugarcane and that obtained using a synthetic medium for Saccharomyces cerevisiae. This observation validated the use of synthetic medium for the characterization and screening of Saccharomyces cerevisiae strains. This fact is especially important, since it is difficult to maintain standardized sugarcane compositions. The ethanol yields using sugarcane juice varied between 58.0% (Sc21) and 84.8% (Sp). Of the strains studied, S. pombe presented the highest yield (84.8%) with the sugarcane juice. A similar result was obtained by Fahrasmane et al. (1986), who obtained ethanol yields of 79% and 85% from sugarcane juice with S. cerevisiae and S. pombe, respectively. The yield of ethanol in the majority of cachac¸a distilleries varied from 60 to 75% with respect to the sugars in the must. Differences in ethanol yield (%) between those obtained using sugarcane juice and those using synthetic medium could also be due to the manner in which the fermentation is conducted, and not just the composition of the medium. The fermentations in synthetic medium were conducted in 250 ml flasks shaken under standardized conditions of temperature and pH, whereas the fermentations in sugarcane juice was performed on a pilot scale, with no shaking or temperature control. Correlation between the ethanol content of the sugarcane wine and the secondary compounds in the cachac¸as

Statistical analysis The exploratory Hierarchical Cluster Analysis (HCA) and the Principal Component Analysis (PCA) were applied to check for similarities between the cachac¸as with respect to the levels of the various volatile components which characterised each cachac¸a (PIROUETTE computer program). The analysis of variance (ANOVA) and Tukey´s means comparison test

Various examples of correlation were observed (Figure 1) between the ethanol content of the sugarcane wines in the sugarcane fermentations and the concentration of secondary compounds (mg/100 ml anhydrous alcohol) in the respective cachac¸as. The ethanol content presented negative correlations with acidity, ethyl acetate, propanol, isobutyl alcohol, isoamyl alcohol, and with the sum of the higher alcohols and the total volatile

140

Ethyl acetate (mg/100 mL aa)

E. S. Oliveira et al. Acidity (mg/100 mL aa)

710 y = -29.042x + 273.01 r = - 0.83***

120 100 80 60 40 20 0 6

5

8

7

9

10

40

y = -8.195x + 77.803 r = - 0.83***

35 30 25 20 15 10 5 0 5

Isobutanol (mg/100 mL aa)

Propanol (mg/100 mL aa)

y = -2.8533x + 33.331 r = - 0.56*

15 10 5 0 6

5

8

7

9

y = -15.503x + 156.84 r = - 0.81***

70 60 50 40 30 20 10 0 6

5

10

9

10

y = -65.522x + 751.14 r = - 0.74**

400

200 150

(mg/100 mL aa)

350

Higher alcohols

Isoamyl (mg/100 mL aa)

y = -47.141x + 560.8 r = - 0.68**

250

100

300 250 200 150 100 50

6

5

8

7

9

5

10

6

7

8

9

10

Ethanol (% v/v)

Ethanol (% v/v) 200

600

y = -103.21x + 1118.5 r = - 0.87****

500

non - alcohols (mg/100 mL aa)

(mg/100 mL aa)

8

7

Ethanol (% v/v)

50

Total volatile compounds

10

9

80

Ethanol (% v/v) 300

8

Ethanol (% v/v)

25 20

7

6

Ethanol (% v/v)

400 300

175

y = -37.683x + 367.38 r = - 0.82***

150 125 100 75 50 25

200

0 5

100 5

6

7

8

9

10

6

7

8

9

10

Ethanol (% v/v)

Ethanol (% v/v)

Figure 1. Equations and correlation coefficients (r) for the ethanol content of the sugarcane wine and the various secondary compound in the cachac¸a obtained by fermentation of sugar cane juice, with a significance of 0.1%(****), 1%(***), 5%(**) and 10%(*).

compounds. These correlations demonstrated the importance of the ethanol content of the sugarcane wine on the concentration of secondary compounds in the cachac¸as. That is, the greater the ethanol content of the sugarcane wine, the greater the volume of cachac¸a obtained with a determined concentration of ethanol, and the secondary compounds present in the sugarcane wine will suffer a dilution effect. This fact shows that a variation in the ethanol content of the sugarcane wine is an important factor for the standardization of the ethanol/volatiles ratio and the quality of the commercialized beverage in the industrial process. Volatile compounds in cachac¸as The results for the analyses of volatile compounds in cachac¸as are presented in Table 1. The ethanol concentration (% v/v) of the cachac¸as varied from 38.4 to 39.9;

volatile acidity from 28.4 to 121.4; acetaldehyde from 7.4 to 18.2; ethyl acetate from 9.4 to 36.5; propanol from 5.6 to 20.5; isobutyl alcohol from 16.2 to 76.2; isoamyl alcohol from 79.4 to 283.8, higher alcohols from 101.1 to 351.6; and total volatile compounds from 164.0 to 527.7 mg/100 ml of anhydrous alcohol (aa). The quality of cachac¸a in Brazil is defined by the Federal Decree no. 2314 of 04/09/97, Decree no. 4851 of 02/10/03 and by law no. 3712 of the Ministry of Agriculture of 18/09/1974, which established the following standards of identity and quality: the alcoholic content should be from 38 to 48% v/v at 20
C and the total secondary compounds (aldehydes, acids, esters, furfural and higher alcohols) should not be less than 200 mg/100 ml of anhydrous alcohol. The maximum values (mg/100 ml anhydrous alcohol) were 150 for volatile acidity as acetic acid, 200 for esters as ethyl acetate, 30 for aldehydes as acetaldehyde, 5 for furfural

Ethan1

14.39± 0.33 16.31± 0.15 14.11± 0.11 7.36± 0.50 12.49± 0.56 18.16± 1.70

12.04± 0.15 9.33± 0.14 11.10± 0.17 17.39± 0.08

87.32± 1.34 36.90± 0.25 70.01± 4.09 36.07± 0.22

Acetald3

38.52± 4.41 50.82± 0.30 57.99± 0.21 28.39± 0.50 62.30± 0.35 121.42± 4.42

VolAc2

25.24± 0.20 15.66± 0.19 13.69± 1.08 9.35± 0.25

14.53± 0.92 16.28± 0.19 17.64± 0.62 9.77± 1.21 15.41± 0.02 36.49± 1.25

EthylAc4

9.84± 0.29 12.27± 0.09 11.24± 0.43 5.61± 0.02

11.36± 0.79 8.64± 0.13 14.39± 0.15 16.05± 0.30 13.10± 0.17 20.52± 0.42

Prop5

43.88± 0.55 44.33± 0.25 38.91± 0.18 16.24± 0.11

43.44± 0.32 28.43 0.20 52.03± 0.35 35.74± 0.55 46.70± 0.27 76.19± 1.11

Isob6

196.53± 2.29 199.38± 1.10 220.85± 0.84 79.39± 0.70

261.83± 2.67 206.01± 1.56 283.75± 1.84 189.79± 2.76 241.20± 2.34 254.92± 3.53

Isoam7

250.25± 3.10 255.98 1.44 271.00 1.10 101.14 0.82

316.63± 3.72 243.08± 1.87 350.16± 2.04 241.58± 3.55 301.00± 1.78 351.62± 4.98

HigAl8

374.85± 2.59 317.87 1.44 365.80 5.41 163.95 0.67

384.07± 6.89 326.49± 2.33 439.91± 2.57 287.10± 4.68 391.20± 2.87 527.69± 2.96

TVC9

0.29 0.35

1.61

0.27

2.00

2.86

0.28

3.34

0.28

0.45

5.30

4.14

0.28

3.90

0.22

0.30

2.91

2.01

0.26

4.89

5.68

4.50

4.48

3.35

5.17

5.31

5.45

7.25

6.03

Isoam12 Isob

Higher alcohols/non-

10

Prop11 Isob

4.69

HigAl10NAlc

Ethanol (%, v/v); 2Volatile acidity as acetic acid; 3Acetaldehyde; 4Ethyl acetate; 5Propanol; 6Isobutyl alcohol; 7Isoamyl alcohol; 8Higher alcohols; 9Total volatile compounds, Alcohols(S 1,2 e 3); 11Propanol/Isobutyl alcohol; 12 Isoamyl alcohol/Isobutyl alcohol.

1

Saccharomyces cerevisiae Sc3 39.2± 0.30 Sc4 39.0± 0.23 Sc6 39.0± 0.14 Sc8 39.9± 0.71 Sc13 38.4± 0.15 Sc21 38.8± 0.11 Non-Saccharomyces C. apı´cola 38.9± 0.04 H. occidentalis 38.4± 0.24 P. subpelliculosa 39.1± 0.05 S. pombe 39.2± 0.25

Strains

Table 1. Ethanol content (% v/v) and volatile compounds concentrations (mg/100 ml anhydrous alcohol) in the cachac¸as obtained with the different strains of yeast.

Sugarcane wine fermentation 711

712 and 300 for higher alcohols.Vargas (1995) analyzed 683 reports of analyses issued by the Vegetable Laboratory of the Ministry of Agriculture and Agrarian ReformMG in the period from 1989 to 1994 regarding cachac¸a bottled and/or sold in the state of Minas Gerais. She showed that the parameters for identity and quality had high coefficients of variation (CV), indicating that these parameters are unstable variables, with the exception of the ethanol concentration. The higher alcohols showed an average value of 187 mg/100 ml of anhydrous alcohol (aa), with minimum and maximum values of 6 and 490 mg/100 ml, respectively. The average value found for volatile acidity was 118 mg/100 ml aa, varying from not detected to 587 mg/100 ml. The average value for esters was 87 mg/100 ml aa with values varying from 2 to 484 mg/100 ml. The average value for aldehydes was 13.3 mg/ml aa, with a wide range of variation from 1 to 282 mg/100 ml. The average value found for total volatile compounds was 391 mg/100 ml aa, varying from 66 to 1060 mg/100 ml. The cachac¸a fermented with the Sc21 strain showed the highest concentrations of all volatile compounds analysed, with the exception of isoamyl alcohol, which was highest in the cachac¸a produced with the strain Sc6 (283.8 mg/100 ml aa). The cachac¸as obtained with the strains Candida apicola, Hanseniaspora occidentalis, and Pichia subpelliculosa contained volatile compounds within the limits required by Brazilian law. The cachac¸a produced with the S. pombe strain showed the smallest amounts of propanol, isobutyl alcohol and isoamyl alcohol and, consequently, of higher alcohols. It also presented very low values for total volatile compounds (164.0 mg/100 ml aa), this value being below the minimum value stipulated by Brazilian law, which is 200 mg/100 ml aa. Fahrasmane et al. (1986), comparing two strains of yeast (S. cerevisiae and S. pombe) with respect to their productions of higher alcohols, also showed that the strain of S. pombe produced much less higher alcohols than the S. cerevisiae strain. Parfait & Jouret (1975) also reported that S. pombe produced relatively small quantities of higher alcohols. The low levels of acetaldehyde and ethyl acetate found in the cachac¸a can be explained by the separation of the head fraction, containing the more volatile compounds, during distillation. The components of all the cachac¸as were within the legal limits, with the exception of the higher alcohols in the cachac¸as obtained with the strains Sc3, Sc6, Sc13 and Sc21. The concentration of these alcohols exceeded the maximum limits stipulated by law (300 mg/100 ml anhydrous alcohol). According to Lehtonen & Jounelaiksson (1983), the distillation system exerts a pronounced effect on the flavour of alcoholic beverages, especially on the quantitative composition of the volatile components. The higher alcohol/non-alcohol ratio in the cachac¸as obtained (Table 1) with the 10 strains of yeast studied, varied from 1.6 to 5.3, with an average of 3.3. The S. pombe (Sp) strain presented the lowest value and the

E. S. Oliveira et al. S. cerevisiae (Sc8) strain the highest value. This ratio can be used to monitor the distillation in the separation of the head fraction, so as to obtain an adequate ratio. For the Saccharomyces yeasts, the average value was 3.69. The sensory analysis (Table 2) attributed the lowest degree of acceptance to the cachac¸a fermented with the S. pombe strain. This cachac¸a had the lowest value for the higher alcohol/non-alcohol ratio. The ratio propanol/isobutyl alcohol in the cachac¸as varied from 0.22 to 0.45, with an average of 0.30. This ratio is also indicative of differentiation; the Sc8 strain produced a cachac¸a with the highest ratio (0.45) and considerably different from the other strains. This index was used by Singer (1966) to differentiate different alcoholic beverages, the ratio showing an average of 0.40 for cognacs and 0.42 for whiskeys. The ratio isoamyl alcohol/isobutyl alcohol varied from 3.35 to 7.25, with an average of 5.21. Singer (1966) found average values of 2.86 for this ratio in cognacs and of 1.09 in whiskeys, considerably lower than the values found in the cachac¸as in this study. The similarities between the samples of cachac¸a for the variables acetaldehyde, ethyl acetate, volatile acidity, propanol, isobutyl alcohol, isoamyl alcohol, higher alcohols and total volatile compounds can be seen in the dendrogram (Figure 2) obtained with the results of the HCA. The cachac¸as formed five distinct groups, with a similarity of 0.76. The scale, varying from 0.0 to 1.0, represents the indices of similarity. The cachac¸as obtained from the Saccharomyces cerevisiae strains were divided into three groups with low similarities among them, demonstrating that the S. cerevisiae strains could produce cachac¸as with differentiated compositions of volatile compounds. The PCA (Figures 3 and 4) demonstrated that 92.9% of the variation occurring among the samples could be explained by the first axis (Principal Component 1) and only 6.7% by the second axis (Principal Component 2). Together, Principal Components 1 and 2 explained

Table 2. Average scores obtained for the cachac¸as in the acceptance test for the attributes aroma, flavour and overall impression. Cachac¸as

Sensory attributes Aroma

Saccharomyces cerevisiae Sc3 6.54ab Sc4 6.76ab Sc6 6.66ab Sc8 6.93 a Sc13 6.27ab Sc21 6.56ab Non-Saccharomyces C. apicola 6.71ab H. occidentalis 6.46ab P. subpelliculosa 6.59ab S. pombe 5.66b MSD

1.13

Flavour

Overall I.

5.78a 5.90a 5.95a 6.56a 6.00a 6.24a

6.10a 6.27a 6.29a 6.71a 6.12a 6.59a

6.00a 6.00a 6.02a 5.56a

6.24a 6.27a 6.29a 5.76a

1.23

1.13

MSD = Minimum significant difference between the means according to Tukey’s test (P = 0.05).

713

group 3

group 2

group 1

Sugarcane wine fermentation

Ca

located on the opposite side from that produced by the Sp strain, since it presented elevated contents of volatile compounds. Oliveira (2001), working with the same yeasts in synthetic media, showed that Sc21 and Sp also formed high and low contents of volatile compounds, respectively, in synthetic medium.

Ho

Sensory analysis

Sc4

The average scores obtained by the samples for the sensory attributes in the sensory evaluation can be found in Table 2. The cachac¸as did not vary from one another with respect to aroma, flavour and overall impression (P £ 0.05). Only the sample fermented by the S. pombe strain (Sp) differed statistically with respect to aroma from the sample Sc8 (P £ 0.05). This result is especially expressive, considering the analytical results which indicated that S. pombe was not adequate for the production of cachac¸a. Despite the fact that the cachac¸as did not differ significantly from one another, that produced by the S. cerevisiae strain Sc8 received higher scores for all the attributes, whereas that fermented by the S. pombe strain (Sp) received the lowest scores. The cachac¸as fermented with strains of C. apicola, H. occidentalis and P. subpelliculosa did not differ statistically (P £ 0.05) with respect to aroma, flavour and overall impression from the samples fermented with S. cerevisiae strains. Thus, the cachac¸as produced with these yeasts had an acceptance comparable to those obtained from S. cerevisiae strains. All the cachac¸as were well accepted, always being evaluated above ‘liked slightly’ by the taste panel, with the exception of sample Sp, whose scores were between ‘liked slightly’ (6) and ‘neither liked nor disliked’ (5). These scores were certainly due to the lower contents of volatile compounds formed by this yeast, as was mentioned above. With respect to the other cachac¸as,

1.0

0.8

0.6

0.4

0.2

0.0

Sp Ps

group 4

Sc8 Sc21

grroup 5

Sc13 Sc3 Sc6

Figure 2. Dendrogram obtained for the 10 cachac¸as obtained using S. cerevisiae and non-S. cerevisiae strains.

99.6% of the variation between the samples. When Figures 3 and 4 are compared to each other, the variability of the samples associated with the first axis (PC1) was mainly due to the individual volatile compounds, and the sum of total secondary compounds with the exception of isoamyl alcohol. In the case of PC2, volatile acidity was the most important variable. The samples were clearly distinguished from one another, being marked by clearly defined locations on Figure 3. The S. pombe strain produced a cachac¸a clearly differentiated from the others, mainly because of its lower content of higher alcohols and volatile components. The cachac¸a produced by the Sc21 strain was

group 4 Sc21

60 group 2 group 1

Ca

Sp

PC2

20

Ps Sc4

group 3 -20

Sc13

group 5

Ho Sc8

Sc6 Sc3

-300

-200

-100

0

100

200

PC1

Figure 3. Graph of the Scores (PC1 versus PC2) for the cachac¸as obtained using S. cerevisiae and non-S. cerevisiae strain.

714

E. S. Oliveira et al. volatile acidity

60

secondary components

ethyl acetate

20

PC2

isobutanol acetaldehyde propanol

-20 higher alcohols isoamyl alcohol -300

-200

-100

0

100

200

PC1

Figure 4. Graph of the loadings (PC1 versus PC2) of the volatile components of the cachac¸as obtained using S. cerevisiae and non-S cerevisiae strains.

despite the variations in quantities and proportions of the volatile contents, these variations did not influence their acceptance. Since non-Saccharomyces strains are frequently found in cachac¸a fermentations, it is apparent that these do not have a large influence on the quality of the cachac¸as, or on the compounds which affect their sensory qualities. Of the compounds quantified in the cachac¸as, only propanol showed a statistically significant (P £ 0.05) correlation with the flavour attributes (r ¼ +0.76) and overall impression (r ¼ +0.81). These results are contrary to those of Boza & Horii (1998) who reported that the propanol content showed a negative influence on the sensory quality of cachac¸a. Ribeiro (1997) also evaluated the physico-chemical and sensory characteristics of three cachac¸as fermented by different strains of Saccharomyces cerevisiae and concluded that it was difficult to correlate the constitution of cachac¸a with its sensory quality, suggesting the need for an amplification of the studies in this direction.

Conclusions The ethanol yields of the non-Saccharomyces strains were similar to those of the Saccharomyces cerevisiae strains. The Exploratory Hierarchical Cluster Analysis, exhibited a separation of the cachac¸as into five groups, as a function of the volatile compounds determined, and the Principal Component Analysis demonstrated that the variation between the samples was mainly due to the the individual volatile compounds, and the sum of total secondary compounds; with the exception of the isoamyl alcohol, volatile acidity was the most important variable in the case of PC2. The cachac¸as produced by the different strains of yeast presented variations in the contents and propor-

tions of the main volatile compounds, but these variations did not result in perceptible differences in aroma, flavour or overall impression. The cachac¸a produced by S. pombe was the exception, with an aroma considered worse than that produced by the strain of Saccharomyces which tended to show the best attributes. This observation indicated that the non-Saccharomyces yeasts studied present in small scale sugarcane fermentations, did not exert a negative influence on the sensory quality of the cachac¸a.

Acknowledgements This study was financed by Coordenac¸a˜o de Aperfeic¸oamento de Pessoal de Nı´ vel Superior (CAPES). We thank Professor David Lee Nelson at Faculdade de Farma´cia of Universidade Federal de Minas Gerais for the revision of the manuscript.

References Berry, D.R. 1995 Alcoholic beverage fermentations. In Fermented Beverage Production, 1st edn., eds. Lea, A.G.H., Piggott, J.R. pp. 32–44. London: Blackie Academic & Professional. ISBN 0751400270. Bonneu, M., Crouzet, M., Urdaci, M. & Aigle, M. 1991 Direct selection of yeast mutants with reduced viability on plates by erythrosine B. staining. Analytical Biochemistry 193, 225–230. Boza, Y. & Horii, J. 1998 Influeˆncia da destilac¸a˜o sobre a composic¸a˜o e a qualidade sensorial da aguardente de cana-de-ac¸u´car. Cieˆncia e Tecnologia de Alimentos, 18, 391–396. Brasil Portaria no 371 de 18 de setembro de 1974 do Ministe´rio da Agricultura Pecua´ria e Abastecimento. Complementac¸a˜o dos padro˜es de identidade e qualidade para destilados alcoo´licos. Dia´rio Oficial da Unia˜o, Sec¸a˜o I, Parte I, Suplemento, Brası´ lia, 19 set. 1974, pp. 53–62.

Sugarcane wine fermentation Brasil Decreto no 2.314 do Ministe´rio da Agricultura Pecua´ria e Abastecimento de 04 de setembro de 1997. Dispo˜e sobre o registro, classificac¸a˜o, padronizac¸a˜o, produc¸a˜o e fiscalizac¸a˜o das bebidas. Dia´rio Oficial da Unia˜o, Brası´ lia, 05 de set. 1997. Brasil. Decreto no. 4.851 do Ministe´rio da Agricultura Pecua´ria e Abastecimento de 2 de outubro de 2003. Altera dispositivos do Regulamento aprovado pelo Decreto n. 2.314, de 4 de setembro de 1997, que dispo˜e sobre a padronizac¸a˜o, a classificac¸a˜o, o registro, a inspec¸a˜o, a produc¸a˜o e a fiscalizac¸a˜o. Dia´rio Oficial da Unia˜o, Brası´ lia, 3 out. 2003. Cardello, H.M.A.B. & Faria, J.B.1998 Ana´lise descritiva quantitativa da aguardente de cana durante o envelhecimento em tonel de carvalho (Quercus alba L.). Cieˆncia e Tecnologia de Alimentos 18, 169–175. Cole, V.C. & Noble, A.C. 1995 Flavor chemistry and assessment. In Fermented Beverage Production, 1st edn., eds. Lea, A.G.H., Piggott, J.R. pp. 361–385. London: Blackie Academic & Professional. ISBN 0751400270. Fahrasmane, L., Parfait, A. & Galzy, P. 1986 Proprietes fermentaires des levures de rhumerie. Industries Alimentaires Agricoles 103, 125– 127. Lehtonen, M. & Jounela-Eriksson, P. 1983 Volatile and non-volatile compounds in the flavour of alcoholic beverages. In Flavour of Distilled Beverages: Origin and Development, ed. Piggott, J.R. pp. 64–78. Florida: Verlag Chemie International Inc. ISBN 0–89573– 131–2 Miller, G.L. 1959 Use of dinitrosalicylic acid reagent for determination of reducing sugar. Analytical Chemistry 31, 426–428. Oliveira, E.S. 2001 Caracterı´ sticas fermentativas, formac¸a˜o de compostos vola´teis; e qualidade da aguardente de cana obtida por linhagens de leveduras isoladas de destilarias artesanais. Tese de

715 doutorado, UNICAMP, Faculdade de Engenharia de Alimentos, Campinas, Brasil. Oliveira, E.S., Rosa, C.A., Morgano, M.A. & Serra, G.E. 2004 Fermentation characteristics as criteria for selection of cachac¸a yeast. World Journal of Microbiology and Biotechnology 20, 19–24. Parfait, A. & Jouret, C. 1975 Formation of higher alcohols in rum. Annales de Technologie Agricole 24, 421–36. Pirouette 1997 Multivariate Data Analysis for IBM PS Systems, Version 2.1, Infometrix, Seattle, WA, USA. Ribeiro, C. A. F. 1997 Potencialidades de diferentes linhagens de levedura da espe´cie Saccharomyces cerevisiae na tecnologia de aguardente de cana. Dissertac¸a˜o de mestrado, USP, Escola Superior de Agricultura ‘Luiz de Queiroz’, Piracicaba, Brasil. SAS Institute 1993 SAS User’s Guide: Statistics. Cary, USA ISBN 1555445543. Singer, D.D. 1966 The analysis and composition of potable spirits: Determination of C3, C4 and C5 alcohols in whisky and brandy by direct gas chromatography. Analyst, 91, 127–134. Stone, H. & Sidel, J.L. 1993 Sensory Evaluation Pratices, 2nd edn. San Diego: Academic Press Inc., Serie (Food Science and Technology). ISBN 0-12-672482-2. Suomalainen, H. & Lehtonen, M. 1979 The production of aroma compounds by yeast. Journal of the Institute of Brewing 85, 149– 156. Vargas, E.A. 1995 Qualidade da aguardente de cana produzida, engarrafada e/ou comercializada em Minas Gerais. Dissertac¸a˜o de mestrado, UFMG, Faculdade de Farma´cia, Belo Horizonte, Brasil. Zimmermann, H.W. 1963 Studies on the dichromate method of alcohol determination. American Journal of Enology and Viticulture 14, 205–213.


Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:717–722 DOI 10.1007/s11274-004-4045-8

Biological control of sheep parasite nematodes by nematode-trapping fungi: in vitro activity and after passage through the gastrointestinal tract E´rika B. N. Graminha1, Alvimar J. Costa2, Gilson P. Oliveira3, Antonio C. Monteiro1,* and Solange B. S. Palmeira3 1 Laborato´rio de Microbiologia, Departamento de Produc¸a˜o Vegetal, Faculdade de Cieˆncias Agra´rias e Veterina´rias (FCAV), Unesp, SP, Brasil 2 Departamento de Patologia Veterina´ria e Centro de Pesquisas em Sanidade Animal (CPPAR), FCAV, Unesp, SP, Brasil 3 CPPAR, FCAV, Unesp, SP, Brasil *Author for correspondence: Tel.: +55-16-3209-2652, Fax: +55-16-3202-4275, E-mail: [email protected]

Keywords: Arthrobotrys conoides, Arthrobotrys musiformis, nematodes, nematophagous fungi, sheep

Summary The main method used for the control of gastrointestinal nematodes in sheep production is the application of chemotherapeutic agents, which often lead to the selection of parasites resistant to given active principles. Biological control can be considered a promising alternative, contributing to an increase in the efficacy of verminous control. We determined the in vitro activity and in situ survival of the predatory fungi Arthrobotrys musiformis and Arthrobotrys conoides during passage through the gastrointestinal tract of sheep after oral administration of conidia in microencapsulated form and as a liquid in natura. Initial in vitro tests showed that both fungi were efficient in the predation of trichostrongylid L3 larvae present in the faeces of sheep naturally infected with gastrointestinal nematodes. The fungi presented high nematophagous activity, which was 99.3% for A. conoides and 73.7% for A. musiformis. A. conoides did not survive passage through the gastrointestinal tract under the conditions of the present experiment. On the other hand, A. musiformis was reisolated after administration in either microencapsulated or liquid form, suggesting that this species is a promising alternative for the control of nematodes in sheep since it survives without any protection (in natura).

Introduction In view of the high susceptibility of animals to nematode infection, the control of these gastrointestinal parasites on sheep farms becomes a challenge in the sanitary management of the herd. The economic losses associated with the acquisition of vermifugal agents and the constant deterioration of the animal’s health due to nematophagous spoliation by some worm species are regarded as an obstacle to productivity. However, the drug treatment used does not always lead to the desired effect, because of the process of resistance developed by the parasites as a result of the long-term use of these drugs, underdosage and an inadequate control strategy, facts favouring the selection of resistant parasite populations (Prichard 1990; Shoop 1993; Pen˜a et al. 2002). Biological control is an option in which a living organism is used to control another target organism (Thamsborg et al. 1999). Predatory fungi, a group of microorganisms that produce a nematode-trapping net of mycelium, because of their structure are able to capture and destroy nematodes in the pre-parasitic phase (Barron 1977; Nansen et al. 1988). Laboratory

interaction tests are necessary to demonstrate the action of predatory fungi on infectious larvae of domestic animals. The predatory activity of species of the genus Arthrobotrys, as well as of species of other genera, in culture, faecal bolus and pasture has been demonstrated in various experiments (Naves & Campos 1991; Mendoza-de Gives et al. 1992; Arau´jo et al. 1994; Padilha 1996), indicating variations in the predatory capacity of different isolates from the same species (Arau´jo & Patarroyo 1995). The ability of these fungi to survive passage through the gastrointestinal tract is an essential prerequisite that needs to be studied for the strategic use of these organisms as biological control agents of gastrointestinal nematodes (Waller et al. 2001). The latter authors obtained important results showing that fungi which produce conidia with a thin wall do not seem to survive gastrointestinal passage. Probably because of this, Grønvold et al. (1993a) obtained successful results with the fungus Duddingtonia flagrans, which showed an elevated activity after passage through the gastrointestinal tract due to its ability to produce thick-wall chlamydospores.

718 Tests with isolates of the fungi A. robusta and A. conoides have been performed by Arau´jo et al. (2001), who reported efficacy against the infectious larvae of Oesophagostomum radiatum after passage of conidia through the gastrointestinal tract compared to the control group (p < 0.05). However, although all fungi passed through the gastrointestinal tract, these species were unable to predate a higher number of infectious larvae when compared to the untreated group. The authors concluded that efficacy was lost after passage, probably due to the intense stress to which the fungi were submitted, or because they had been transferred for many years under laboratory conditions. The objective of the present study was to determine the in vitro predatory activity and the in situ survival of the nematode predatory fungi A. musiformis and A. conoides during passage through the gastrointestinal tract of sheep after oral administration of conidia in natura (liquid medium) and in microencapsulated form.

E.B.N. Graminha et al. and incubated in the dark at 25 C for 48 h. Next, 1 ml of a concentrated suspension of L3 larvae of the nematode species Haemonchus (47%), Trichostrongylus (45%), Cooperia (6%) and Strongyloides (2%) were transferred to the Petri dishes and the plates were again incubated. After 24 h, the material was examined under a stereomicroscope for the observation of nematodes captured by the fungi. Parasitized and/or predated specimens were fixed in 3% glutaraldehyde in 0.05 M potassium phosphate buffer, pH 7.4, for 72 h. The specimens were then washed six times at 15 min intervals in pure buffer solution for fixation in 2% osmium tetroxide in the same buffer. The material was again washed as described before, dehydrated in ethyl alcohol, dried in a desiccator to the critical point using CO2, mounted, sputtered with a 35 nm gold layer, and photographed under a JEOL JSM 5410 scanning electron microscope operating at 15 kV (Maia & Santos 1997). Micropellet preparation

Material and methods Arthrobotrys musiformis and A. conoides, isolated from pasture soil (Graminha et al. 2002) in the municipality of Jaboticabal, State of Sa˜o Paulo, Brazil, were cultured at ambient temperature in Petri dishes containing 2% agar– water. To stimulate predation, parasitism and, consequently, the production of conidia, 1 ml Panagrellus sp. suspensions containing about 1 · 103 larvae/ml were frequently added. These free-living nematodes were cultured on Petri dishes containing moistened and crushed oat flakes as described by Heintz & Pramer (1972). The conidia were removed from the plates with a Drigalsky spatula by washing with 1 ml sterile distilled water. After quantification, the conidia were appropriately diluted to obtain the concentration to be tested in each assay. In vitro evaluation Twenty-four 2 g ovine faecal samples positive for trichostrongylids were individually weighed and divided into the following three groups: group 1: eight samples without the addition of conidia (control), group 2: eight samples containing 105 A. musiformis conidia/g faeces, and group 3: eight samples containing 105 A. conoides conidia/g faeces. After addition of the conidia, the faecal samples were homogenized individually, vermiculite was added (2:1), and the samples were incubated at 25 C for 7 days. After this period, infectious larvae were extracted by the method of Baermann and quantified (Chandrawathani et al. 1998). The larvae were identified based on the morphological criteria proposed by Keith (1953).

Sodium alginate mucilage containing oat flour Sodium alginate was gradually added to a container with distilled water under shaking until complete hydration, followed by the addition of sodium benzoate as a bacteriostatic agent. In another container with distilled water, polysorbate 80 was mixed with oat flour and this mixture was added to sodium alginate plus benzoate. To the mucilage formed after addition was then added the conidial suspension. Microgranulation The mucilage was transferred under constant shaking to a separating funnel and dripped into a calcium chloride solution which was also shaken throughout the procedure (adapted from Fravel et al. 1985). The microgranules obtained were collected on a sieve, washed with distilled water and transferred to a plane sieve, where they were dried by forced ventilation at ambient temperature. Nematophagous activity of fungi resistant to passage through the gastrointestinal tract of sheep Ten mongrel sheep aged 8–10 months were randomly divided into five groups of two animals, with each animal in its respective group receiving one of the following treatments:

Determination of fungal action by scanning electron microscopy

G1: control (untreated animals); G2: 1 ml suspension containing 2 · 106 A. musiformis conidia in natura; G3: 1 ml suspension containing 2 · 106 A. conoides conidia in natura; G4: 2 · 106 microencapsulated A. musiformis conidia; G5: 2 · 106 microencapsulated A. conoides conidia.

Pure cultures of the fungi A. musiformis and A. conoides were replated on Petri dishes containing 2% agar–water

Each animal was fed daily 30 g soy bran plus mineral salt and 3 kg autoclaved corn silage for a period of 5 days before and after administration of the conidia.

719

Sheep parasite nematodes by predatory fungi Faecal samples were collected from each sheep 12, 16, 20, 24, 28, 40, 62 and 74 h after conidia administration. Concentrated suspensions of free-living nematodes (see above) were added to the samples collected in plastic bags and the samples were incubated at 25 C for 3 days in the dark to stimulate fungal growth. Next, aliquots of these samples were placed on Petri dishes containing 2% agar–water and suspensions with the same free-living nematodes were again added. All plates were kept at 25 C in the dark and examined daily under a stereomicroscope for a period of 15 days for the determination of fungal growth, the formation of traps and nematode predation, followed by reisolation of the fungi. Statistical analysis In the in vitro assay was used the fully randomized design, consisting of three groups and eight replicates each. Log transformation (x + 5) was used for analysis of variance of the data as proposed by Little & Hills (1978). The analyses were performed by the F-test and measures were compared by the Tukey test at a 5% probability of error. All analyses were performed using the SAS program (1996).

Results and discussion The results of the in vitro test showed that A. conoides preyed on 99.3% of the larvae, while A. musiformis preyed on 73.7%. Both species differed significantly from the control group, demonstrating their high in vitro predatory capacity (Table 1). Scanning electron microscopy (Figure 1) illustrates the predation by the two microorganisms that were characterized by broad expansions of the mycelium forming the three-dimensional traps that are characteristic of these predatory fungi. These results are consistent with those obtained by some investigators but exceed those reported by others. Table 1. Parasitic nematodes of sheep recovered from faecal cultures after treatment of samples with conidia of the fungi Arthrobotrys musiformis and A. conoides. Repetition

A. musiformis

A. conoides

Control

1 2 3 4 5 6 7 8

110 360 180 800 750 190 190 160

20 10 10 10 0 0 10 10

1040 550 2270 510 2110 460 1230 2260

Total Mean SD Efficacy (%)

2740 342.5B 276.70 73.70

70 8.75C 6.41 99.30

10430 1303.75A 800.28 –

a

Original values expressed in arithmetic media but statistical analysis performed with log transformation (x + 5).

Graminha et al. (2001) also confirmed the activity of the fungi A. musiformis and A. conoides by scanning electron microscopy, showing that A. musiformis preyed on 66.3 and 94.4% of infectious Ancylostoma spp. and Haemonchus contortus larvae, respectively, while A. conoides preyed on 51.7 and 89.3% of the same larvae, indicating effective control. According to Arau´jo et al. (2001), A. conoides preyed on 24% of Oesophagostomum radiatum and 76% of Cooperia and Haemonchus larvae, but fungal action was not convincing since the number of non-preyed larvae of the respective nematodes upon fungal treatment did not differ significantly from that obtained for the control group (p > 0.05). The results obtained by these authors were possibly due to nematode-related factors such as size, agility and physicochemical characteristics, together with the particularities inherent to the fungal isolate itself. The predatory action of fungi promoting a reduction in the number of infectious larvae of parasitic nematodes of sheep has been reported by various investigators. The most promising results were obtained with D. flagrans. Chlamydospores of this fungus were effective in reducing the number of infectious larvae in sheep faeces (Faedo et al. 1998; Knox & Faedo 2001) and a reduction in L3 larvae of H. contortus using D. flagrans spores has been reported by Pen˜a et al. (2002). According to Faedo et al. (2000), if deposited at the same time as the parasite eggs, D. flagrans prevents the transmission of third-stage larvae from the faecal bolus to pasture but has no effect on these larvae in surrounding soil. Although being more effective than A. musiformis in the in vitro test (Table 1), A. conoides did not survive gastrointestinal passage under the experimental conditions of the present study. Conidia of this fungus could not be reisolated after administration in either microencapsulated form or in natura at any of the time points of faeces collection (Table 2). It is possible that the conidia were inactivated by adverse factors in the gastrointestinal tract such as acids, pH, competition with the microbiota or even temperature. According to Grønvold et al. (1993b) who studied the faecal bolus of calves infected with Ostertagia ostertagi and fed one Arthrobotrys oligospora isolate and two Duddingtonia flagrans isolates, the reduction in L3 larvae of O. ostertagi in pastures was only significant with the D. flagrans isolates. The A. oligospora isolate was unable to pass through the gastrointestinal tract and the authors could not explain why the fungus had lost this ability. When analysing the ability of nine D. flagrans isolates and eight Arthrobotrys spp. isolates to survive passage through the gastrointestinal tract of sheep, Faedo et al. (1997) considered the results regarding the survival of Arthrobotrys to be inconsistent, whereas D. flagrans showed excellent survival ability which resulted in marked reductions in the number of Trichostrongylus colubriformis larvae in culture.

720

E.B.N. Graminha et al.

Figure 1. Scanning electron photomicrographs of the fungi Arthrobotrys musiformis and A. conoides and predation of trichostrongylid L3 larvae by the respective fungi. Left panel: (a) A. musiformis conidia; (b) L3 larvae captured in the net; (c) detail of the three-dimensional trap of A. musiformis. Right panel: (a) A. conoides conidium; (b) L3 larvae captured by the fungus; (c) detail of the three-dimensional net of A. conoides.

Table 2. Recovery of fungi at different times of faeces collection after administration to sheep. Time after treatment

12 16 20 24 28 40 62 74

A. musiformis

A. conoides

In natura

Microencapsulated

In natura

Microencapsulated

+ + + ) ) ) ) )

+ ) + ) ) ) ) )

) ) ) ) ) ) ) )

) ) ) ) ) ) ) )

+, presence of fungi accompanied by trap formation and nematode predation; ), absence of fungi, traps and predation.

A. musiformis was detected at the first three collection times after treatment (12, 16 and 20 h) when administered in the form of conidia in natura. Samples collected from the group of animals treated with microencapsulated conidia showed growth at the first and third collection times (12 and 20 h after treatment) (Table 2). It is possible that the presence of the fungus would have

been detected between 20 and 24 h, but no samples were collected during this period. The results of the present study are similar to those reported by other investigators for D. flagrans, but differ from those obtained with isolates of the genus Arthrobotrys. The differences reported by different investigators indicate that species and predatory abilities may vary according to the regions where the strains were isolated and factors such as the environmental conditions used. The presence of the fungus Monacrosporium thaumasium in goat faecal samples 21 and 24 h after the oral administration of conidia in natura has been shown by Melo et al. (2003), while Arau´jo (1996) observed a larger number of A. robusta conidiophores 24 h after the administration of conidia. Arau´jo et al. (1999) tested fungal species of the genera Arthrobotrys and Monacrosporium, including A. conoides, and obtained successful predation after passage through the gastrointestinal tract of cattle. Furthermore, this species was the most effective in the production of mycelium. The authors emphasized the importance of the study of nematophagous fungi in each specific habitat, since their predatory abilities may vary according to different ecological niches. Manueli et al. (1999) isolated 12

Sheep parasite nematodes by predatory fungi strains of the genus Arthrobotrys from 23 samples containing nematophagous species selected based on the sampling of 2500 faecal boluses from sheep and goats. Periodic transfers are necessary for maintenance of the fungi but may trigger contamination and favour the occurrence of mutations, a fact compromising the predatory capacity of these microorganisms (Campos et al. 2004). However, a biological control agent needs to remain viable throughout the storage period in order to be used for research and eventual commercialization. The microencapsulation technique used in the present study might contribute not only to protection during gastrointestinal passage but also to the maintenance of the nematophagous characteristics of the microorganism even during the storage process. The present results, especially those referring to microencapsulation, are preliminary but encouraging. A. musiformis could be reisolated after administration in two forms, i.e., microencapsulation and liquid form, suggesting that this species is a promising alternative for the control of nematodes in sheep since it survives in the absence of any protection (in natura). However, the results are not conclusive and further studies are needful to identify fungal species and/or strains that survive passage through the gastrointestinal tract of sheep, as well as to establish the conditions for successful passage.

Acknowledgments We thank to Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo (FAPESP: 00/10322-5) for the fellowship granted to EBNG, Bellman Nutric¸a˜o Animal Ltda. for kindly providing the mineral salt, and Dr. Johanna Martha Kopte for technical collaboration.

References Arau´jo, J.V. 1996 Interac¸a˜o entre as larvas infectantes de Cooperia punctata e os fungos predadores do geˆnero Arthrobotrys, caracterizac¸a˜o dos isolados de Arthrobotrys e seu uso no controle biolo´gico de nemato´ides gastrintestinais de bovinos. PhD thesis, ICB, UFMG, Belo Horizonte, MG, Brasil. Arau´jo, J.V., Campos, A.K., Paiva, F. & Bressan, M.C.R.V. 2001 Efeito antagonista de fungos predadores do geˆnero Arthrobotrys sobre larvas infectantes de Oesophagostomum radiatum, Cooperia punctata e Haemonchus placei. Revista Brasileira de Cieˆncia Veterina´ria, 8(2), 81–84. Arau´jo, J.V. & Patarroyo, J.H. 1995 Initial interaction between Haemonchus placei infective larvae and different Arthrobotrys isolates. Arquivo Brasileiro de Medicina Veterina´ria e Zootecnia 47, 733–738. Arau´jo, J.V., Santos, M.A., Ferraz, S. & Maia, A.S. 1994 Biological control in vitro of infective Haemonchus placei larvae by predacious fungi Arthrobotrys musiformis. Arquivo Brasileiro de Medicina Veterina´ria e Zootecnia 46(3), 197–204. Arau´jo, J.V., Stephano, M.A. & Sampaio, W.M. 1999 Passage of nematode-trapping fungi through the gastrintestinal tract of calves. Veterinarski Arhin 69(2), 69–78.

721 Barron, G.L. 1977 The Nematode-Destroying Fungi. 140p. Ghelph, Ontario: Canadian Biological Publications Ltda. ISBN: 0-92037000-4. Campos, A.K., Mota, M.A., Arau´jo, J.V. & Cecon, P.R. 2004 Atividade predato´ria, crescimento radial e esporulac¸a˜o de fungos predadores de nemato´ides Monacrosporium spp, submetidos a` criopreservac¸a˜o. Cieˆncia Rural 34(2), 465–469. Chandrawathani, P., Omar, J. & Waller, P.J. 1998 The control of the free-living stages of Strongyloides papillosus by the nematophagous fungus, Arthrobotrys oligospora. Veterinary Parasitology 76, 321– 325. Faedo, M., Barnes, E.H., Dobson, R.J. & Waller, P.J. 1998 The potential of nematophagous fungi to control the free-living stages of nematode parasites of sheep: Pasture plot study with Duddingtonia flagrans. Veterinary Parasitology 76, 129–135. Faedo, M., Larsen, M. & Thamsborg, S. 2000 Effect of different times of administration of the nematophagous fungus Duddingtonia flagrans on the transmission of ovine parasitic nematodes on pasture – a plot study. Veterinary Parasitology 94, 55–65. Faedo, M., Larsen, M. & Waller, P.J. 1997 The potential of nematophagous fungi to control the free-living stages of nematode parasites of sheep: Comparison between Australian isolates of Arthrobotrys spp. and Duddingtonia flagrans. Veterinary Parasitology 72, 149–155. Fravel, D.R., Marois, J.J., Lumsden, R.D. & Connick Jr., W.J. 1985 Encapsulation of potential biocontrol agents in an alginate-clay matrix. Phytopathology 75(7), 774–777. Gordon, H.M. & Whitlock, H.V. 1939 A new technique for counting nematode eggs in sheep faeces. Australia Commonwealth Scientific and Industrial Research Organization 12, 50–52. Graminha, E.B.N., Maia, A.S., Monteiro, A.C. & Costa, A.J. 2002 Biological control of sheep parasite nematodes. In The Role of Genetics and Evolution in Biological Control; Abstracts of the Global IOBC International Symposium co-organized with C.I.L.B.A., Montpellier, France. p 31. ISBN: 2-9818864-0-3. Graminha, E.B.N., Maia, A.S., Santos, J.M., Candido, R.C. & Silva, G.S. 2001 Avaliac¸a˜o in vitro da patogenicidade de fungos predadores de nemato´ides parasitos de animais dome´sticos. Semina: Cieˆncias Agra´rias 22(1), 11–16. Grønvold, J., Wolstrup, J., Nansen, P., Henriksen, S.A., Larsen, M. & Bresciani, J. 1993a Biological control of nematode parasites in cattle with nematode-trapping fungi: a survey of Danish studies. Veterinary Parasitology 48, 311–325. Grønvold, J., Wolstrup, J., Larsen, M., Henriksen, S.A. & Nansen, P. 1993b Biological control of Ostertagia ostertagi by feeding selected nematode-trapping fungi to calves. Journal of Helminthology 67, 31–36. Heintz, C.E. & Pramer, D. 1972 Ultrastructure of nematode-trapping fungi. Journal of Bacteriology 110, 1163–1170. Keith, R.K. 1953 The differentiation on the infective larvae of some common nematode parasites of cattle. Australian Journal Zoology 1(2), 223–235. Knox, M.R. & Faedo, M. 2001 Biological control of field infections of nematode parasites of young sheep with Duddingtonia flagrans and effects of spore intake on efficacy. Veterinary Parasitology 101, 155–160. Little, T.M. & Hills, F.J. 1978 Agricultural experimentation designs and analysis. 350p. New York: Wiley. ISBN: 0471023523. Maia, A.S. & Santos, J.M. dos. 1997 A SEM technique for preparing biological control agents of nematodes in action. Acta Microscopica 6 (suppl. B), 550–551. Manueli, P.R., Waller, P.J., Faedo, M. & Mahommed, F. 1999 Biological control of nematode parasites of livestock in Fiji: screening of fresh dung of small ruminants for the presence of nematophagous fungi. Veterinary Parasitology 81, 39–45. Melo, L.M., Bevilaqua, C.M.L., Arau´jo, J.V. & Melo, A.C.F.L. 2003 Atividade predato´ria do fungo Monacrosporium thaumasium contra o nemato´ide Haemonchus contortus, apo´s passagem pelo trato gastrintestinal de caprinos. Cieˆncia Rural 33(1), 169– 171.

722 Mendoza-de Gives, P., Zavaleta-Mejia, E., Quiroz-Romero, H., Herrera-Rodriguez, D. & Perdomo-Roldan, F. 1992 Interaction between the nematode-destroying fungus Arthrobotrys robusta (Hyphomycetales) and Haemonchus contortus infective larvae in vitro. Veterinary Parasitology 41, 101–107. Nansen, P., Grønvold, J., Henriksen, S.A. & Wolstrup, J. 1988 Interaction between the predacious fungus Arthrobotrys oligospora and third-stage larvae of a series of animal-parasitic nematodes. Veterinary Parasitology 26, 329–337. Naves, R.L. & Campos, V.P. 1991 Ocorreˆncia de fungos predadores de nemato´ides no Sul de Minas Gerais e estudo da capacidade predato´ria e crescimento in vitro de alguns de seus isolados. Nematologia Brasileira 15(2), 152–162. Padilha, T.P. 1996 Atividade de fungos nemato´fagos nos esta´gios pre´parasita´rios de nemato´deos trichostrongilı´ deos. Cieˆncia Rural, 26(2), 333–341. Pen˜a, M.T., Miller, J.E., Fontenot, M.E., Gillespie, A. & Larsen, M. 2002 Evaluation of Duddingtonia flagrans in reducing infective

E.B.N. Graminha et al. larvae of Haemonchus contortus in feces of sheep. Veterinary Parasitology 103, 259–265. Prichard, R.K. 1990 Anthelmintic resistance in nematodes: Extent, recent understanding and future directions for control and research. International Journal for Parasitology 20, 515–523. SAS Institute, 1989–1996. SAS  User’s Guide: Estatistics. SAS Institute, Inc. Cary, NC, USA. Shoop, W.L. 1993 Ivermectin resistance. Parasitology Today 9(5), 154– 159. Thamsborg, S.M., Roepstorff, A. & Larsen, M. 1999 Integrated and biological control of parasites in organic and conventional production systems. Veterinary Parasitology 84, 169–186. Waller, P.J., Knox, M.R. & Faedo, M. 2001 The potential of nematophagous fungi to control the free-living stages of nematode parasites of sheep: feeding and block studies with Duddingtonia flagrans. Veterinary Parasitology 102, 321–330.

 Springer 2005

World Journal of Microbiology & Biotechnology (2005) 21:723–728 DOI 10.1007/s11274-004-4798-0

Micro-encapsulation of Bifidobacterium lactis for incorporation into soft foods L.D. McMaster*, S.A. Kokott1 and P. Slatter2 1 Department of Food and Agricultural Sciences, Cape Technikon (1), Flow Process Centre, Cape Technikon, P. O. Box 652, 8001 Cape Town, South Africa 2 Department of Civil Engineering, Cape Technikon, P. O. Box 652, 8001 Cape Town, South Africa *Author for correspondence: Tel.: +27-21-4603507, Fax: +27-21-4603424, E-mail: [email protected]

Keywords: Beverages, Bifidobacterium lactis, microencapsulation, probiotics, rheology

Summary Micro-encapsulation of the probiotic micro-organism Bifidobacterium lactis isolated from a bio-yoghurt starter culture, was carried out using a mixture of hydrated gellan and xanthan gums. Rheological studies showed that the gum mix was suitable for encapsulation of B. lactis, for incorporation into soft foods/beverages. The gel behaved as a non-Newtonian material, and the flow curve fitted well to the Herschel–Bulkley model. The average yield stress of the gum was 1.515 Pa, indicating gum stability, and the yield stress range was 1 Pa over a temperature range of 35–50
C. Almost constant minimum gum viscosity occurred between 46 and 61
C. Oval/round capsules were synthesized manually using a monoaxial gum flow through a 27.5 G bevelled needle, together with a superposed air stream (air knife technique). The diameter of the capsules, measured using laser diffractometry, varied from 20 to 2200lm, with 50% of the capsules having a diameter of £637 lm. Numbers of viable B. lactis in the capsules were estimated using high power ultrasound (20 kHz). By using a concentrated inoculum of B. lactis, microcapsules containing log10 11–12 c.f.u. g)1 were synthesized. Apart from the anaerobic culturing of B. lactis, all other work was done in the presence of atmospheric oxygen. The organism exhibited a high degree of oxygen tolerance. A 21-day survival study of immobilized cells in 1 M sodium phosphate buffer (pH 7) stored at either 4 or 22
C indicated that B. lactis survived in excess of log10 11 c.f.u. g)1 microcapsule. This technique represents a suitable means of supplying viable probiotics to the food and/or pharmaceutical industries. Mathematical symbols Herschel–Bulkley model: s0 ¼ sy
j
n y s0 yield stress sy shear stress jy fluid consistency index n flow behaviour index

Introduction The consumption of foods containing probiotic organisms is becoming increasingly popular. Probiotic micro-organisms, generally bacteria, are defined as micro-organisms which when consumed in certain numbers, exert health benefits beyond nutrition (Lactic Acid Bacteria Industrial Platform, LABIP, cited by Guarner & Schaafsma 1998). One group of organisms gaining credibility as probiotics is the genus Bifidobacterium. The significance and development of probiotic foods, and their use as alternatives to antibiotics, is predicted to increase in the future (Doyle 2003; Kroger 2003). In the European Union, consumption of probiotic foods increased 40% during the first quarter of 2004 (Ross

2004). However, the provision and supply of foods containing viable Bifidobacterium in the numbers required for a probiotic organism to be beneficial (RDA log10 8 c.f.u.), has long been recognized to be problematic (Hughes & Hoover 1991; Klaver et al. 1993; Temmerman et al. 2003). Two technologies used to supply Bifidobacterium in foods are freeze- and spraydrying. Although improvements in these methods have been reported (Siuta-Cruce & Goulet 2001), the use of spray- and freeze-drying of bifidobacterium for incorporation into foods for human consumption has shown that both the Bifidobacterium spp. and the carrier medium used in process influence survival of the organism (Lian et al. 2002). Bifidobacterium cells undergoing these treatments lose viability (Klaver et al. 1993; Lian et al. 2002). In addition, microbial injury during

724 both of these processes is common, and injured cells require time for repair in a suitable recovery medium (Ray 2001). Bacteria supplied in foods as both freezeand spray-dried cells require time to rehydrate and reconstitute in the human host. The human gastrointestinal tract (GIT) is not an ideal environment for resuscitation of injured bacteria; hence it is likely that Bifidobacterium present as spray- or freeze-dried cells in probiotic foods do not all respond optimally under conditions associated with the gastro-intestinal tract. An alternate means of delivering living bifidobacteria directly to the GIT is through the use of micro-capsules. Unlike the freeze- and spray-drying processes where the dormant micro-organism is delivered directly into the environment, the capsule matrix surrounding the bacteria protects the viable organisms from the gastric juices of the upper gastro-intestinal tract, and once in the colon, capsules break down to release the living organisms over a period of time (Krasaekoopt et al. 2003). Micro-encapsulation of probiotic bifidobacteria is reported in some cases to improve survival of the organism in various foods. Edible gums such as alginate, locust bean gum, j-carrageenan, xanthan and gellan are often used to immobilize bacterial cells (Sun & Griffiths 2000; Siuta-Crus & Goulet 2001; Kailasapathy 2002; Truelstrup-Hansen et al. 2002; Krasaekoopt et al. 2003). These gums form a soft microcapsule, suitable for inclusion into foods. The aim of this investigation was to provide microcapsules of suitable dimensions synthesized from edible gums, in which a high concentration of living Bifidobacterium was present, for incorporation into soft foods and beverages. To achieve this, we undertook a preliminary rheological study of the gum/s of choice before proceeding with micro-encapsulation. We also wished to establish the best method for synthesising micro-capsules with a view to upgrading production to an industrial scale. For technological reasons, it was important to establish oxygen tolerance of the selected Bifidobacterium spp. A further objective was to establish the viability of immobilized cells when stored in a buffer, as a possible means of supplying the microcapsules to industry for incorporation into foods.

Materials and methods Bacterial strains and culture conditions Bacterial cultures were propagated anaerobically at 37
C in a Bactron 1.5 anaerobic chamber (Shel Lab) at 37
C in an environment consisting of hydrogen (4%), carbon dioxide (10%) and nitrogen (86%). Cells were grown in tryptone yeast glucose (TYG) media which contained the following, in g l)1: tryptone 10 (Biolab); yeast extract, 5 (Biolab); glucose, 5 (Saarchem); Tween 80, 1.0 (Sigma); NaCl, 4.5; KCl, 0.25; MgCl2Æ6H2O, 0.15; KH2PO4, 0.4; K2HPO4, 0.2; NH4Cl, 0.4 (Saarchem); cystein HCl, 0.5 (Sigma); rezasurin, 0.001% (Sigma).

L.D. McMaster et al. Agar (1.5% (w/v) (Biolab) was added for solid media. Bifidobacterium lactis was isolated from a yoghurt starter culture as described by Trindade et al. (2003). Rheology of the gum mix Based on work done elsewhere, gellan and xanthan were selected as the gums of choice, and were used as a mixture of 0.75% w/v gellan and 1% w/v xanthan (Sun & Griffiths 2000). Immediately prior to the rheological studies, the gums were prepared and re-hydrated as described below. A Paar Physica MCR 300 rotational rheometer with a cone-plate measuring system was used for the gellan:xanthan gum mix to determine a flow curve in controlled shear rate mode, at shear rates of 0.1–100 s)1, and a temperature sweep at a constant shear rate of 5 s)1 between 30 and 60
C. In addition, a time dependency test, at a constant shear rate of 5 s)1 at temperatures of 30, 40 and 45
C was done. Micro-encapsulation of B. lactis A variation of the method proposed by Sun & Griffiths (2000) was used. Gellan (0.1 g), (Sigma) and xanthan (0.2 g) (Sigma) were added to 20 ml distilled water. The solution was mixed using a magnetic stirrer (Ikamag) and heated to 80
C for 1 h. The gel mix was then autoclaved at 121
C for 15 min. B. lactis was grown in a Bactron 1.5 anaerobic chamber at 37
C overnight in 250 ml TYG broth to an OD600 0.9–1.1. All subsequent procedures were done under aerobic conditions. Cells were harvested by centrifugation at 8000 · g for 10 min in a Beckman J21 centrifuge at 4
C. The pellet of cells was washed three times by resuspension in 20 ml aliquots of sterile distilled water, followed by centrifugation. Finally, the cells were suspended in sterile distilled water, to give a volume of 2.5 ml. One ml of this concentrate was mixed into 20 ml sterile gellan:xanthan gum at 55
C. Microcapsules were generated by a bead entrapment method, using a superposed airflow together with mono-axial extrusion of the gum/bacteria mixture (Huebner & Buccholz 1999; Park & Chang 2000). The gum/cell mix was manually extruded through a 27.5 G bevelled needle (Teruma) fitted on to a sterile 20 ml syringe (Promex). Gel and sterile air flow rates were 10 ml min)1 and 250 ml s)1 respectively. Microdroplets formed by this method were hardened into spheres by free fall into a sterile 0.1 M CaCl2 solution (100 ml). After 1 h, beads were separated from the solution by aseptic filtration through sterile Whatman filter paper No 1. Microcapsules remaining on the filter paper were washed thoroughly with sterile 0.1 M CaCl2. All procedures were carried out in a laminar flow hood. Estimation of microcapsule size The diameter and size distribution of microcapsules containing B. lactis were estimated by laser

Micro-encapsulation of B. lactis diffractometry using a Malvern Mastersizer S, version 2.19. The analysis model used was Polydisperse, and the presentation – 30HD – ‘‘standard wet’’ for use with the Mie theory. Enumeration of B. lactis To estimate numbers of B. lactis in microcapsules, 0.1– 0.2 g of capsules were weighed into 10 ml of 1 M sodium phosphate buffer (NaH2PO4 46.3 g l)1, NaH2PO4 53.7 g l)1), pH 6.8. At predetermined time intervals, the mix was subjected to sonication to release B. lactis cells from the gel matrix. A Vibracell Ultrasonic Processor VCX750 (20 kHz) (Materials and Sonics) was used with a standard probe, with a tip diameter of 13 mm generating a sound wave amplitude of 105 lm for 15 s. All samples were treated in 20 ml sterile cylindrical Polytop glass cylinders immersed in an ice water bath. After completion of the treatment, to estimate the number of released B. lactis, standard serial dilution and plate counts were done. Cells were diluted in 2% buffered peptone water (Difco). Representative 0.1 ml volumes from dilutions were spread in triplicate/dilution onto TYG agar plates. These were incubated anaerobically at 37
C for 48 h, and colonies were counted. Values obtained were expressed as either log10 c.f.u. g)1 or ml)1 or as log surviving fraction. The log surviving fraction was calculated as Nt/N0, where N0 and Nt are the average number (N) of c.f.u. ml )1 (free cells) or g)1 (microcapsules) at the start of the experiment (0), and at time (t) respectively. Survival of immobilized B. lactis in 1 M sodium phosphate buffer (pH 6.8) Sodium phosphate buffer (1 M) pH 6.8 was prepared as described. Known weights (0.2 g) of entrapped B. lactis were added to 10 ml of sterile buffer. The solution was stored aerobically at either 4 or 22
C for 21 days. Viable numbers of entrapped B. lactis were determined using the Vibracell Ultrasonic Processor VCX 750 followed by serial plate dilution.

725 Rheology of the gellan–xanthan gum mix We elected to use gellan (0.75%) and xanthan (1%) for immobilizing materials as a combination of these gums has been reported to have better technical properties for micro-encapsulation than do alginate, j-carrageenan and locust bean gums (Sun & Griffiths 2000). Although alginate is frequently used to microencapsulate probiotics, it has undesirable attributes, such as susceptibility to degradation by acids (Sun & Griffiths 2000; Truelstrup-Hansen et al. 2002; Krasaekoopt et al. 2003). Hence alginate micro-capsules are likely to erode in the hydrochloric acid of the stomach and not reach the colon intact. Rheological studies used to determine the shear stress data of the gellan:xanthan gum indicated that the gel behaved as a non-Newtonian material, and the flow curve fitted well to the Herschel–Bulkley model s0 ¼ sy)k)n y , which is used to describe the flow behaviour of a yield pseudoplastic material (Figure 1). For microencapsulation, Newtonian liquids such as water or honey are unsuitable as encapsulating materials, as liquid droplets, not solid capsules, are formed during extrusion. In order for successful encapsulation of micro-organisms to occur, the selected material should exhibit non-Newtonian properties i.e. be a relatively viscous material with solid properties. The data reported here indicated that the gum mix of gellan and xanthan satisfied this requirement. The apparent yield stress calculated for the gellan– xanthan mix indicated that the gums would only flow or deform plastically when external forces were greater than the internal structural forces inherent in the gum. The average yield stress of the gum mix was 1.515 Pa, a value similar to that of raw meat batter (Steffe 1996). This value indicated that the gum mix was stable and that at lower stresses, the gel would behave as a solid. The yield–stress range was approximately 1 Pa over a temperature range of 35–50
C (Figure 2). Any material selected for micro-encapsulation should be able to traverse the upper gastro-intestinal tract (GIT) intact to protect the bacteria, but on arrival in the colon, the yield stress of the gum should be such that the immobilized bacteria are released into the human host.

Results and discussion 30

Using an AB yoghurt starter culture, Bifidobacterium lactis was isolated and used for all subsequent studies (Trindade et al. 2003). B. lactis was chosen as the probiotic organism for encapsulation, as it is commonly found in yoghurt starter cultures. The organism possesses attributes important in selecting a probiotic for commercial use. These include oxygen tolerance reported for up to 24 h exposure, fermentation of a wide variety of carbohydrates (Trindade et al. 2003) and survival at pH 2–3 (Truelstrup-Hansen et al. 2002).

Shear stress (Pa)

Bacterial strain isolated

25 20 15 10 5 0 0

20

40

60

80

100

120

Shear rate (s-1) Figure 1. Flow curves of the gum mix at different temperatures, showing non-Newtonian shear thinning behaviour. Gum at 30
C (s); 40
C ()); 40
C (*); 45
C (h); 50
C (d).

726

L.D. McMaster et al. such as 1–3
C. The latter values would require critical temperature control to maintain a constant shear viscosity immediately prior to the extrusion process during which microdroplets are formed. Maintenance of a constant minimum shear viscosity would enhance control over microcapsule size and shape, as well as permit optimization/standardization of mixing procedures necessary for even dispersal of B. lactis in the viscous gum held at these temperatures, prior to extrusion.

7

Viscosity (Pa)

6 5 4 3 2 1 0 0

50

100

150

200

250

300

350

Micro-encapsulation

Time (s) Figure 2. Time sweep of shear viscosity at 5 s)1 at different temperatures for the gum. Temperature range: 30
C (s); 40
C ()); 45
C (*).

The immobilizing material should not be comprised of materials with a high yield stress, as probiotics encapsulated in such a material would not be released. The value obtained in this study for the gellan–xanthan gum mix indicated that stresses associated with GIT movements/peristalsis, could release the bacteria in the colon. The average yield stress of the gum indicated that gellan:xanthan would behave as a solid under low mechanical stress conditions, but would easily be broken by shear stresses such as those associated with chewing. Therefore B. lactis enclosed in a gellan:xanthan microcapsule would have to be supplied in soft foods not requiring mastication, to prevent breaking of the capsule in the mouth. The yield stress value is also important when considering the means of addition of the capsules to foods such that the capsules remain intact and are not disrupted during process. The shear viscosity of the gum mix maintained an almost constant minimum value at temperatures between 46 and 61
C (Figures 2 and 3). These temperatures are within the range for survival of Bifidobacterium. It is recommended that encapsulation using gellan:xanthan as the support matrix should be done between 46 and 61
C. For industrial application of micro-encapsulation it is of importance that the shear viscosity is constant over a broad 15
C range, as opposed to a limited range

Using the methods described above, micro-encapsulation of B. lactis in the gellan–xanthan matrix was successful. Estimation of viable c.f.u. in the capsules over periods of time indicated that oxygen present in the superposed airflow during microencapsulation was not lethal to B. lactis (data not shown). Visual examination of microcapsules using a light microscope indicated that capsules were generally oval/ round in shape, with a dense core comprised of B. lactis cells (Figure 4). When the capsules were broken open and stained by the Gram method, bacterial cells isolated from the core stained gram positive and demonstrated a typical bifidus morphology. The shape of the capsules was variable. Using the manual air knife technique described, we were unable to generate a constant capsule shape. Although the irregularity of microcapsule shape often reported for probiotics immobilized in edible gums is related to process, this problem requires further elucidation, as microcapsule shape could influence both viability and numbers of entrapped bacteria. Shape will also determine flow properties of the capsules, important for industrial process. The use of a constant temperature and gum

4.5 4

Viscosity (Pa)

3.5 3 2.5 2 1.5 1 0.5 0 26

31

36

41

46

51

56

61

Temperature (°C) Figure 3. Shear viscosity of the aqueous gum mix at 5 s)1, as a function of temperature.

Figure 4. Phase contrast view of a single capsule ( · 500), with a densely packed core of B. lactis surrounded by the capsule support material.

Micro-encapsulation of B. lactis

727

flow rate during extrusion, not possible in our encapsulation technique, should improve uniformity of shape. Despite having a constant rate of air flow, a wide variation in the size distribution of the capsules was noted. This ranged from a maximum diameter of 2200 lm to a minimum of 20 lm, with 50% of the capsules having a diameter of