Screening and production of rhamnolipids by ... - Wiley Online Library

3 downloads 3157 Views 182KB Size Report
Pseudomonas aeruginosa GS3 from molasses. Letters in Applied. Microbiology 25, 91–94. Rapp, P., Bock, H., Wray, H. and Wagner, F. (1979) Formation,.
Journal of Applied Microbiology 2000, 88, 379–387

Screening and production of rhamnolipids by Pseudomonas aeruginosa 47T2 NCIB 40044 from waste frying oils E. Haba, M.J. Espuny, M. Busquets1 and A. Manresa Laboratori de Microbiologia, Facultat de Farma`cia, Universitat de Barcelona, and 1Departament de Bioquı´mica i Biologia Molecular, Facultat de Quimica, Universitat de Barcelona, Spain 7140/04/99: received 6 April 1999, revised 2 June 1999 and accepted 28 September 1999

World production of oils and fats is about 2·5 million tonnes, 75% of which are derived from plants. Most of them are used in the food industry for the manufacture of different products, or directly as salad oil. Great quantities of waste are generated by the oil and fat industries: residual oils, tallow, marine oils, soap stock, frying oils. It is well known that the disposal of wastes is a growing problem and new alternatives for the use of fatty wastes should be studied. Used frying oils, due to their composition, have great potential for microbial growth and transformation. The use of economic substrates such as hydrophobic wastes meets one of the requirements for a competitive process for biosurfactant production. In the Mediterranean countries, the most used vegetable oils are sunflower and olive oil. Here we present a screening process is described for the selection of micro-organism strains with the capacity to grow on these frying oils and accumulate surface-active compounds in the culture media. From the 36 strains screened, nine Pseudomonas strains decreased the surface tension of the medium to 34–36 mN/M; the emulsions with kerosene remained stable for three months. Two Bacillus strains accumulated lipopeptide and decreased the surface tension to 32–34 mN/m. Strain Ps. aeruginosa 47T2 was selected for further studies. The effect of nitrogen and a C/N of 8·0 gave a final production of rhamnolipid of 2·7 g l−1 as rhamnose, and a production yield of 0·34 g g−1. E . H AB A , M .J . ES PU N Y, M. B US QU E TS AN D A. MA N RE SA . 2000.

INTRODUCTION

Oils and fats are found in all living cells. World production of oils and fats is about 2·5–3 million tonnes, 75% of which are derived from plants (Mielke 1992). Most of the oils and fats are used in the food industry which generates great quantities of wastes: tallow, lard, marine oils or soap stock, and free fatty acids from the extraction of seed oils. Waste disposal is a growing problem which explains the increasing interest in the use of wastes for microbial transformation. Chemical surfactants are widely used in industrial applications; biosurfactants are biological molecules with similar properties to their chemical counterparts. Probably the most important advantage of biosurfactants over chemical surfactants is their ecological acceptability. Moreover, bioCorrespondence to: Dr A. Manresa, Laboratori de Microbiologia, Facultat de Farma`cia, Universitat de Barcelona c/Joan XXIII s/n, E-08028 Barcelona, Spain, (e-mail: [email protected]). © 2000 The Society for Applied Microbiology

surfactants are non-toxic, natural biodegradable products and thus, essentially compatible with the biogeochemical cycle. Biosurfactants are associated with the uptake of hydrophobic carbon substrates by micro-organisms. However, biosurfactants are not only produced by micro-organisms growing in water-immiscible substrates, although most of the literature is concerned with the conversion of hydrocarbons and vegetable oils as substrates for production (Syldatk et al. 1984; Passeri et al. 1991; Lang and Wullbrandt 1999). Watermiscible substrates such as glucose or other carbohydrates have also been used (Guerra-Santos et al. 1984; Spoecker et al. 1999). Economy is often the bottleneck of biotechnological processes, especially for biosurfactant production. The success of biosurfactant production depends on the development of cheaper processes and the use of low cost raw material, which accounts for the 10–30% of the overall cost (Cameotra and Makkar 1998). Alternative substrates have been suggested

380 E . H AB A ET AL .

(Kosaric et al. 1984; Mercade´ 1994) for biosurfactant production, especially water-miscible wastes: molasses, whey milk or distillery wastes (Babu et al. 1996; Patel and Desai 1997; Daniel et al. 1998, 1999). However, few examples are found of the use of hydrophobic wastes as cheap substrates. Although they have great potential to support microbial growth, most of the reports published deal with contaminant hydrocarbon uptake (Wasko and Bratt 1990) or the use of mineral oil as a substrate for glycolipid production (Mercade´ et al. 1996). Industrial waste-water from olive oil processing, which contains 1–5% of residual oil, is a suitable substrate for glycolipid production (Mercade´ et al. 1993). Here, a study was made of the re-use of olive and sunflower cooking oil as substrates for biosurfactant production by different genera of bacteria and yeast.

Bacterial strains, after being grown on TSA (BBL), were maintained at 4 °C. Yeast strains were grown on Saubouraud with 30 mg l−1 of tetracycline and maintained at 4 °C. Microorganisms were also preserved lyophilized on skim milk. Analytical methods

Microbial growth was measured by the protein content of the cultures following the method of Lowry (Lowry et al. 1951) against a standard of bovine serum albumin (Sigma). In liquid culture experiments, protein (g l−1) was measured directly from an aliquot of the culture. The nitrate content of the medium was monitored by the Aquamerck Nitrate test Merck 8032 (Merck). Rhamnolipid production was measured as rhamnose by a specific colorimetric method (Chandrasekaran and Bemiller 1980).

MATERIALS AND METHODS Detection and isolation of surface active compounds Micro-organisms

Several strains isolated in this laboratory from contaminated soil and water samples were selected for their capacity to accumulate surface-active compounds (Robert 1989): Pseudomonas sp. strains 25A3, 3AT, 32T3, 44T1, 55T1, MT22, 42A2 and 47T2; Rhodococcus sp. strains 51T1, 51T7 and 135; Bacillus sp. 2a; Serratia sp. 2T2; Arthrobacter oxydans; Acinetobacter calcoaceticus; Micrococcus luteus; Candida albicans, C. albicans strain 39A2 and C. torulopsis, and unidentified Gram-negative bacilli strains 13A2, 43T2 and 61T3. Other micro-organisms screened were: Ps. aeruginosa ATCC 27823, Ps. aeruginosa ATCC 111, Ps. aeruginosa CECT 378, Ps. fluorescens CECT 844, Ps. fragi CECT 446, Staphylococcus aureus CECT 59, B. pumilus ATCC 14884, B. subtilis ATCC 6633, B. subtilis ATCC 6051, B. cereus CECT 193, B. licheniformis CECT 491, C. tropicalis CECT 1440, C. rugosa IFO 0750, C. lipolytica CECT 13517 and C. albicans ATCC 10231. Growth conditions

Experiments were carried out in 250 ml baffled Erlenmeyer flasks containing 50 ml medium. The composition of the basal medium was (g l−1): NaNO3 14, KH2PO4 2, K2HPO4 4, KCl 0·2, MgSO4.7H2O 1, CaCl2 0·02, FeSO4.7H2O 0·024, yeast extract (YE) 0·02, and 0·05 ml of a trace-element solution containing (g l−1): H3BO3 0·26, CuSO4.5H2O 0·5, MnSO4.H2O 0·5, MoNa2O4.2H2O 0·06, ZnSO4.7H2O 0·7·2%. Finally, 2% or 4% of used frying oil was added. Medium was sterilized at 120 °C at 1 atm for 15 min. The final pH of the medium was adjusted to 7·2. Submerged microbial cultures were incubated at 30 °C on a reciprocal rotary shaker (150 rev min−1).

Surface tension (ST) and interfacial tension (IT) against kerosene from the cell-free culture and the crude extract were determined in a Fisher tensiometer mod 20 (Fisher Scientific, Pittsburg, PA, USA) according to the Du Nou¨y ring method. Critical micelle concentration (CMC−1), a parameter used as an indirect measure of surfactant concentration, was determined by measuring the surface tension of serial dilutions of the crude extract in distilled water at pH 7. Emulsification power was measured by vortexing an equal volume of the previously centrifuged culture with kerosene for 1 min and determining the percentage of volume occupied by the emulsion. The mixture was allowed to settle for 24 h and the height of the emulsion was measured. The stability of the emulsions was also checked for up to 3 months. Surface-active compounds were separated by liquid–liquid extraction from the supernatant fluid (5 ml) with an equal volume of chloroform previously acidified with 1N HCl at pH 2. The separate organic layer was removed and collected. The combined extract was dried with anhydrous Na2SO4 and concentrated in a rotary evaporator at 40 °C (R111 Bu¨cchi, Switzerland) to obtain the crude extract. Crude extracts were analysed by thin layer chromatography (TLC) on silica gel plates (G60 5553; Merck, Germany). Chromatograms were developed with chloroform : methanol : acetic acid (65 : 25 : 4 v/v) and visualized with TLC reagents, i.e. ninhydrin (6705, Merck) for free amino groups, iodine vapours for lipid stains and a-naphthol-H2SO4 for sugar detection (Kates 1972). Molecular weight determination

Individual rhamnolipids were separated by gradient elution high performance liquid chromatography (HPLC) on a Shimadzu LC-9 A gradient system (Shimadzu, Kyoto, Japan)

© 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387

P S. AE R UG IN O SA RH A MN OL I PI D P R OD UC T IO N 381

equipped with a Spherisorb ODS2 column of 250 × 4·5 mm (Teknochroma, San Cugat, Spain). The flow rate was 1 ml min−1 and the acetonitrile gradient programmed as follows: 4 min at 70% acetonitrile–0·1% acetic acid, then from 70% to 100% acetonitrile in 40 min. For detection, a Sedere evaporative light-scattering detector (S.E.D.E.R.E., Orleans, France) was used. Integration of peaks was recorded on a Shimadzu C-R4A Chromatopac Integrator (Shimadzu, Kyoto, Japan). EI-MS was performed with a VG-QUATTRO mass spectrometer with an electrospray electron source. Scans were taken in the positive-ion mode between 300 and 2000 m/z at a resolution of approximately 500. Rhamnolipid samples were dissolved in CH3CN/H2O (1 : 1, v/v) containing 1% formic acid at a final concentration of 20–50 ml. Aliquots were injected via a Rheodyne injector into a constant stream of H2O–acetonitrile (1 : 1, v/v) containing 1% acetic acid. A flow rate of 15 ml min−1 was maintained throughout the analysis via an Isco syringe pump. Mass calibration was carried out with an external horse-heart myoglobin reference standard using Mass Lynx software. Cooking oil analysis

Free fatty acid composition was determined by Gas Chromatography (GC) after methanolysis with NaOH 0·5 N solution of sodium methoxide by standard techniques (Folch et al. 1957). Monomers were analysed on a Hewlett Packard 5890 serie II Chromatograph (Hewlett Packard, Palo Alto, USA) equipped with a splitless injector and flame ionization detector (FID). A Supelco SP 2380, 60 m × 0·25 mm ID fused silica capillary column was used. The helium flow was 1 ml min−1. The oven temperature was kept at 177 °C for 11 min, followed by a ramp of 7 °C min−1 to 235 °C, and held at 235 °C for 10 min. Sample injections were in splitless mode at 250 °C. Detector temperature was set at 300 °C. The split ratio was 1 : 30 and peak integration was carried out on a HPChem integrator (Hewlett Packard, Palo Alto, USA). RESULTS AND D ISCUSSION Substrate characterization

Frying oil is produced in large quantities by the food industry and at domestic scale. After being used, cooking oil changes its composition and contains more than 30% of polar compounds (Kock et al. 1996) depending on the variety of food, the type of frying and the number of cycles used. The fatty acid composition of the olive and sunflower oil used in this work is compared with the standard unused oils in Table 1. The most important differences in the fatty acid composition of the used olive oil compared with the native oil were the presence of 22·52% of fatty acids of low chain length (³C10), and myristic and lauric acid in the used oil. The

concentration of oleic acid decreased from 84·4 to 55·51%, while linoleic acid increased from 4·6 to 7·43%. The presence of linolenic acid (0·67%) and behenic acid (0·1%) is also worth noting. In the case of the sunflower oil, linolenic acid decreases from 3·4 to 0·28% and stearic acid from 4·76 to 2·4%. The content of oleic acid, 24·70%, decreased. In this case, 2·39% of fatty acids had a chain length lower than (³C12). Screening for biosurfactant production from used frying oils

Thirty-six micro-organisms were screened with respect to biosurfactant production when grown in submerged culture with waste olive or sunflower cooking oil as carbon source (2%). The surface tension of the culture media was 57 mN/m; good biosurfactant producers were considered to be those that decreased the surface activity to 40 mN/m. Cell growth, surface activity of the cell-free culture and emulsification power of the biosurfactant against kerosene, are presented in Table 2. After 72 h of incubation, most of the Pseudomonas tested showed satisfactory growth when cultivated on either used olive oil or used sunflower oil. Used olive oil induced biosurfactant production. In most of the Pseudomonas strains, the surface tension was less than 40 mN/m. The lowest values were for: Pseudomonas sp. 44T1 32 mN/m; Ps. aeruginosa ATCC 111, 33 mN/m; Pseudomonas sp. 47T2 and Ps. aeruginosa ATCC 27823, 34 mN/m; Pseudomonas sp. 3AT and Pseudomonas sp. 25A3, 36 mN/m; Pseudomonas sp. MT22 and Pseudomonas sp. 32T3, 35 mN/m; Pseudomonas sp. 55T1 36·5, mN/m; and Ps. fluorescens CECT 844, 39 mN/m. Only three strains showed surface tension values over 40 mN/m: Pseudomonas sp. 42 A and Ps. aeruginosa CECT 378, 44 mN/m; and Ps. fragi CECT 446, 44 mN/m. Stable water in oil emulsion was observed after 24 h (Table 2); values between 55 and 87% were recorded. Biosurfactants accumulated by Pseudomonas sp. 32T3, 35 mN/m (Robert 1989) and Pseudomonas sp. 42A2 NCIB 40005 (ST, 40 mN/m) (Mercade´ et al. 1988) were exceptions. Instead of producing rhamnolipids, these two strains accumulated fatty acids with surface activity. According to the surface tension, CMC, TLC and emulsification power, Ps. aeruginosa CECT 378 and Ps. fragi CECT 446 did not produce surface-active compounds from frying olive oil. Sunflower oil was not as good a substrate as olive oil, neither for cell growth nor for biosurfactant production. The lowest surface tension found for Ps. aeruginosa ATCC 27823, Ps. aeruginosa ATCC 111 and Pseudomonas sp. 551T was 36 mN/m; Pseudomonas 47T2 NCIB 40044 and Pseudomonas sp. MT22 gave 37 mN/m and 38 mN/m, respectively. The value for Pseudomonas 3AT and Ps. aeruginosa 44T1 was 42 mN/m, Pseudomonas sp. 25A3, 47 mN/m and Pseudomonas

© 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387

382 E . H AB A ET AL .

— –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Free fatty acids Olive oil Sunflower oil Olive/sunflower (g 100 g−1) Native Used Native Used Used oil (1 : 1) — –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 0·1 0·1 0·17 0·12 Lauric – 0·1 – 0·1 0·1 Myristic tr 0·8 – 7·55 8·20 Palmitic 6·9 8·74 5·6 0·2 0·43 Palmitoleic – 0·64 – 4·74 3·49 Stearic 2·3 2·6 2·2 24·7 41·65 Oleic 84·4 55·51 25·1 55·6 27·98 Linoleic 4·6 7·43 66·2 0·28 0·3 Linolenic – 0·67 – 0·8 0·38 Behenic – 0·1 – – – Lignoceric – 0·15 – – – cis-vaccenic – 0·15 – 2·39 14·11 ³C10 – 22·52 – 4·08 3·41 1·8 1·49 1·9 — ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

CECT 844, 47 mN/m. Pseudomonas fragi CECT 446 (ST, 47·5 mN/m) and finally Ps. aeruginosa CECT 378, 57·5 mN/m, did not accumulate surface-active compounds in the culture medium. Stability of water in oil emulsion, after 24 h, was similar to that obtained with olive oil; values ranging from 55 to 87% were found for the producer strains. The biosurfactants extracted from the supernatant fluids of the cultures were developed by thin layer chromatography (TLC) and visualized with specific reagents. Since biosurfactant structure is a characteristic of the producing species they were detected with a-naphthol-H2SO4 indicating, as expected, their glycolipid nature as rhamnolipids. Only strains 42A2 (Guerrero et al. 1997) and 3AT (unpublished results) produced fatty acids with surface activity. Serratia 2T2 produced a lipidic product. Matsuyama and Nakagawa reported a Serratia strain producing a cyclic lipopeptide with surface activity (Matsuyama and Nakagawa 1996). Bacillus species were also screened. Neither used olive oil nor used sunflower oil were satisfactory substrates for the growth of these species. However, the surface tension of the culture media decreased. When grown on used olive oil, the values of the surface activity were: B. subtilis 2a, 32 mN/m; B. pumilus ATCC 14884, 38 mN/m; B. subtilis ATCC 6633, 39 mN/m; B. subtilis ATCC 6051 and B. licheniformis CECT 491, 40 mN/m. It is well known that the genus Bacillus accumulates a variety of lipopeptides (McInerney et al. 1990; Morikawa et al. 1992; Kajimura et al. 1995). The results obtained when sunflower oil was supplied as carbon source were similar to those obtained with olive oil. Values found in the culture media were: the lowest B. pumilus ATCC 14884,

Table 1 Main fatty acid composition of the native and waste oil used as substrate for microbial growth and biosurfactant production

34 mN/m; B. subtilis ATCC 6051 and B. licheniformis CECT 491, 39 mN/m; B. subtilis ATCC 6633 and B. subtilis 2a, 41 mN/m and 42 mN/m, respectively. Spots developed by TLC of the organic extraction of the supernatant fluid were visualized with ninhydrin, indicating that they were lipopeptides. No water-in-oil emulsions were observed because emulsification is a property of some surface-active compounds and lipopeptides lack this property. The strains Rhodococcus sp. 51T1, 51T7 and 135 were isolated in this laboratory. The genus Rhodococcus produces glycolipids, trehalose-lipids when cultivated in hydrocarbon substrates (Rapp et al. 1979; Espuny et al. 1995; Mercade´ et al. 1996; Lang and Philp 1999). Only Rhodococcus sp. 51T1 reduced the surface tension, when incubated with olive oil, to 34 mN/m. When sunflower oil was supplied as carbon source, Rhodoccus 51T7 gave a surface tension of 40 mN/m and Rhodococcus sp. 51T1 gave 37·5 mN/m. Acinetobacter calcoaceticus CECT 441 grew well on both olive oil and sunflower oil; the values of surface tension of the culture media found were 42·5 and 38 mN/m, respectively. Although Staphylococcus aureus, Arthrobacter oxydans and Micrococcus luteus decreased the surface tension of the culture, this was due to the oily substrate remaining. Used olive oil was not a suitable substrate for cell growth, although the surface tension of the supernatant fluid decreased with Candida sp. 39A2 (35 mN/m), C. albicans (39 mN/m), C. rugosa IFO 0750 (39 mN/m), C. tropicalis CECT 1357 (35 mN/m). In this case, C. lipolytica (43 mN/m) and C. torulopsis (45 mN/m) were poor producers. Sunflower oil supported good cellular growth in most cases and surface

© 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387

P S. AE R UG IN O SA RH A MN OL I PI D P R OD UC T IO N 383

Table 2 Cell growth, surface tension (ST), emulsifying power after 24 h (E-24) and type of biosurfactant of the microorganisms screened after grown on olive or sunflower used oil –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Cell growth ST (mN/m) E–24 (%) Oil Oil Oil Microorganism Olive Sunflower Olive Sunflower Olive Sunflower Chemical type –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Pseudomonas sp.3AT ¦¦¦ ¦ 36 42 55 56·4 glycolipid Pseudomonas sp.25A3 ¦¦¦ ¦ 36 47 31·5 64 fatty acids Pseudomonas sp.32T3 ¦¦¦ ¦¦ 35 44 50·6 87·5 glycolipid P. aeruginosa 42A2 NCIB 40045 ¦¦¦ ¦¦¦ 41 43 57·5 12·5 glycolipid Pseudomonas sp.44T1 ¦¦¦ ¦¦¦ 32 42 60 59·2 glycolipid P. aeruginosa 47T2 NCIB 40044 ¦¦¦ ¦¦¦ 34 37 53·4 56·4 glycolipid Pseudomonas sp.55T1 ¦¦¦ ¦ 36·5 36 61·3 57·6 glycolipid Pseudomonas sp.MT22 ¦¦¦ ¦¦¦ 35 38 nd 55 glycolipid P. aeruginosa ATCC 27823 ¦¦¦ ¦¦¦ 34 36 0 0 glycolipid P. aeruginosa ATCC 111 ¦¦¦ ¦¦¦ 33 36 0 0 glycolipid P. aeruginosa CECT 378 ¦ ¦ 44 57·5 0 0 glycolipid P. fluorescens CECT 844 ¦ ¦¦ 39 47 0 0 lipopeptide P. fragi CECT446 ¦ ¦ 44 47·5 0 0 lipopeptide Bacillus sp. 2a ¦¦ ¦ 32 42 0 0 lipopeptide B. subtilis ATCC 6633 ¦ ¦ 39 41 1·9 0 lipopeptide B. subtilis ATCC 6051 ¦ ¦ 40 39 5·2 0 lipopeptide B. pumillus ATCC 14884 ¦ ¦ 38 34 0 0 lipopeptide B. cereus CECT 193 ¦ ¦ 45 49 0 0 lipoprotein B. licheniformis CECT 491 ¦ ¦¦ 40 39 0 0 lipoprotein Candida albicans 39A2 ¦ ¦¦¦ 35 35 0 0 lipoprotein C. albicans ¦ ¦ 39 43 0 0 lipoprotein C. rugosa1970 IFO 0750 ¦ ¦¦¦ 39 39 0 0 lipoprotein C. tropicalis CECT 1440 ¦¦ ¦ 35 43 1·5 0 lipoprotein C. lipolytica CECT 1357 ¦¦ ¦ 43 40 0 0 glycolipid C. torulopsis ¦ ¦¦ 45 40 0 0 glycolipid Rhodococcus sp. 51T7 ¦ ¦¦¦ 56 40 9·7 0 glycolipid Rhodococcus sp. 51T1 ¦¦ ¦ 34 37·5 0 0 nd Rhodococcus sp. 135 ¦¦¦ ¦¦ 60 47 0 0 fatty acids S. aureus CECT 59 ¦ ¦ 34 47 45·3 18·7 nd Serratia sp. 2T2 ¦ ¦ 34·5 42 45·3 18·7 nd A. calcoaceticus CECT441 ¦¦¦ ¦¦ 42·5 38 0 0 lipoprotein A. calcoaceticus CECT441 ¦¦¦ ¦¦ 42·5 38 0 0 nd Arthrobacter oxydans ¦ ¦ 42 34 1·9 0 nd strain 13A2 ¦ ¦ 47 33 0 0 nd strain 43T2 ¦ ¦ 51 51·5 1·9 0 nd strain 61T3 ¦ ¦ 36 36 0 0 nd –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– —

tension decreased as follows: Candida sp. 39A2, 35 mN/m; C. rugosa, 39 mN/m; C. lipolytica and C. torulopsis, 40 mN/m. There are numerous reports on the isolation and surfactant production of different species of the genus Candida. The surfactants produced by this genus can differ widely from one species to another. It has been reported to produce sophorose lipids (Shepherd et al. 1995; Brakemeier et al. 1998), a lipidcarbohydrate complex (Desai and Desai 1993) and long-chain fatty acids (Ka¨ppeli et al. 1978) when grown either on hydrophobic or water-miscible substrates. The surfactants pro-

duced from used frying oils did not have emulsifying properties. By TLC, visualized with a specific reagent, they were presumptively characterized as lipoproteins. Stability of rhamnolipids emulsions

In a further step in the transformation of waste frying oil into biosurfactants, several Pseudomonas strains were incubated in submerged culture with single used oil or with a mixture of used olive and sunflower oil (1 : 1). Good growth was observed

© 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387

384 E . H AB A ET AL .

and surface tension was between 32 mN7m and 36 mN/m. Emulsion of kerosene with the supernatant fluid of cultures with mixed oils or single oils had similar values, between 53 and 60%, except for Pseudomonas sp. 42A2, when incubated with olive oil (Table 3). The stability of the emulsions was relatively good; only emulsions with the supernatant fluid from Pseudomonas M22 collapsed from 53·5 to 37·7% after 3 months. In the case of the cultures from sunflower oil, the only emulsion which collapsed was that from Ps. aeruginosa 44T1, from 87·5 to 56·4% Water-in-oil emulsions of the supernatant fluid from the cultures with the oil mixture were very stable and compact (Table 3). Emulsions were measured after 24 h, kept, and measured every 3 months at room temperature (stable after 10 months). Rhamnolipid production by Pseudomonas 47T2

Pseudomonas 47T2 NCIB 400044 was selected for rhamnolipid production. This strain produced 1·4 g l−1 rhamnolipid based on rhamnose content from olive oil mill wastewater (Mercade´ et al. 1993). It is well known that there are different structural variants of rhamnose lipids; the type produced depends on the Pseudomonas strain, the carbon source used and the strategy of production as well. When cultivated with used vegetable oils, this strain produced a mixture of two rhamnolipids with an Rf of 0·45 and 0·7, respectively. These rhamnolipids, isolated by HPLC, gave a retention time of 5 and 9 min, respectively. The molecular mass found was 488 Da and 633 Da, respectively, and, corresponded to (R1): L-a-rhamnopyranosyl-b-hydroxydecanoyl-b-hydroxydecanoate and (R2): 2-O-a-L-rhamnopyranosyl-a-L-rhamnopyranosyl-b-hydroxydecanoyl-bhydroxydecanoate (Lang and Wullbrandt 1999).

Effect of nitrogen source concentration on production

A critical effect was found on rhamnolipid production when sodium nitrate supplied was 3, 4, 5 and 7 g l−1. As expected, an increase in biomass was observed of 3·7, 8·3, 8·6 and 10·56 g l−1, respectively (Fig. 1a). A major effect was found for rhamnolipid production. When nitrate content was 3 g l−1, production increased steadily during the incubation time. In this case, biosurfactant accumulation was observed during both the exponential and the stationary phase of cell growth. Meanwhile, when nitrogen content was increased at 4, 5 or 7 g l−1, product accumulation ran in parallel with cell growth and nearly ceased in the stationary phase. Production yield (Yp/x), calculated at 48 h, was 0·27 when initial nitrate was 7 g l−1, 0·34 at 5 g l−1 of nitrate supply, 0·23 with 4 g l−1 of initial nitrate content and 0·29 at 80 h when nitrate supply was 3 g l−1. Effect of C/N concentration on production

The C/N ratio affects biomass and biosurfactant formation (Fig. 2). The carbon substrate content supplied was 40 g l−1 and different nitrogen sources were added (3, 4, 5 and 7 g l−1). A C/N ratio of 8 gave the highest production yield (Yp/x) of 0·34. These results are similar to those found earlier with Pseudomonas 44T1 (Manresa et al. 1991). Kinetics of product accumulation

After 2 h of inoculation, biomass increased slowly for 24 h to reach the highest biomass concentration (9·63 g l−1) observed; then the stationary phase of growth was established (Fig. 3).

Table 3 Stability of the w/o emulsions formed with the supernatant fluids of the cultures with kerosene from different Pseudomonas

strains with time –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Emulsion (%) Olive oil Sunflower oil Olive/Sunflower ST (1 : 1) Micro-organisms Cell growth (mN/m) 24 h 3 months 24 h 3 months 24 h 3 months –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — Pseudomonas sp. 3AT +++ 36 55 50·6 56·4 52·6 47·1 45 Pseudomonas sp. 42A2 NCIB 40045 +++ 39 31·5 30·6 64 — 54·7 — Ps. aeruginosa 44T1 +++ 30 50·6 50 87·5 56·4 54·4 50 Ps. aeruginosa 47T2 NCIB 40044 +++ 32 57·7 57·7 12·2 12·5 54·5 54·5 Pseudomonas sp. 55T1 +++ 36 60 56 59·2 56·5 59 56 Ps. aeruginosa MT22 +++ 35 53·5 37·7 56·4 56·1 56 52 Ps. aeruginosa ATCC 27823 +++ 33 61·3 60 57·6 57·1 55·7 55 Ps. aeruginosa ATCC 111 +++ 35 57·7 ND 55 ND 75·6 63·1 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– — © 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387

P S. AE R UG IN O SA RH A MN OL I PI D P R OD UC T IO N 385

10 9 8 7 6 5 4 3 2 1 0

(a) 3·5

Biomass, Nitrate, –1 Rhamnose (g l )

3

–1

Rhamnose (g l )

2·5 2 1·5 1

0

10

0·5

20

30

40 50 60 Time (h)

70

80

90 100

Fig. 3 Time course of cell growth and biosurfactant production. Biomass (); rhamnose (); residual nitrate (r)

0 (b) 12

–1

Biomass (g l )

10 8 6 4 2 0

0

10

20

30

40 50 Time (h)

60

70

Fig. 1 Effect of the sodium nitrate content on biomass and biosurfactant production (measured as rhamnose). NaNO3: 7 g l−1 (ž); 5 g l−1 (); 4 g l−1 (r); 3 g l−1 ()

Biomass

12

Rhamnose

10

gl

–1

8 6

80

Nitrogen content was exhausted during the first 24 h of incubation. Biosurfactant production started soon after inoculation and increased progressively to 2·7 g l−1. During the exponential phase of cell growth, biosurfactant accumulation was 0·63 g l−1, calculated on the basis of the rhamnose content; in this phase, the volumetric productivity (P g l−1 h−1) was 0·026 g l−1 h−1. Production during the stationary phase reached 2·7 g l−1; productivity (P) in this late phase of production was higher at 0·26 g l−1 h−1. Final production was 2·7 g l−1. Few reports have been published on the use of waste as substrates for rhamnolipid production. From distillery waste and whey milk, a production of 1·85 g l−1 and 1·78 g l−1, respectively, was reported (Koch et al. 1988; Babu et al. 1996); Patel and Desai produced 0·24 g l−1 from molasses (Patel and Desai 1997). This paper aims to contribute to the use of industrial wastes as substrates for the production of biosurfactants. Microbial surfactants are not yet competitive with those produced by the chemical industry, but efforts should be made on the different aspects of production to find a suitable and economic substrate and to develop new strategies to increase the volumetric productivity. Having shown that waste cooking oils are suitable substrates for biosurfactant production, we are now working on the second stage, the optimization of culture media and the production strategy to increase the productivity of the process with Pseudomonas aeruginosa 47T2 NCIB 40044.

4

ACKNOWLEDGEMENTS

2 0

5·7

10·0

8·0

13·3

C/N

Fig. 2 Effect of C/N ratio on cell growth and biosurfactant production (measured as rhamnose)

This work was financially supported by the Comisio´n Interministerial de Ciencia y Tecnologı´a (CICYT), Project AMB96–1429 and the Fundacio´n Bosch i Gimpera (University of Barcelona). The authors thank Robin Rycrofft for the revision of the manuscript.

© 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387

386 E . H AB A ET AL .

REFERENCES Babu, P.S., Vaidya, A.N., Bal, A.S., Kapur, R., Juwarkar, A. and Khanna, P. (1996) Kinetics of biosurfactant production by Pseudomonas aeruginosa strain BS2 from industrial wastes. Biotechnology Letters 18, 263–268. Brakemeier, A., Wullbrandt, D. and Lang, S. (1998) Candida bombicola: production of novel alkyl glycosides based on glucose/2dodecanol. Applied Microbiology and Biotechnology 50, 161–166. Cameotra, S.S. and Makkar, R.S. (1998) Synthesis of biosurfactants in extreme conditions. Applied Microbiology and Biotechnology 50, 520–529. Chandrasekaran, E.V. and Bemiller, J.N. (1980) Constituent analysis of glucosamonoglucans. In Methods in Carbohydrate Chemistry, Volume, Vol III ed. Wrhiste, L. and Wolfrom, M.L. pp. 89–97. N.Y.: Academic Press. Daniel, H.J., Otto, R.T., Binder, M., Reuss, M. and Syldatk, C. (1999) Production of sophorolipids from whey: development of a two stage process with Cryptococcus curvatus ATCC 20509 and Candida bombicola ATCC 22214 using deproteinized concentrates as substrates. Applied Microbiology and Biotechnology 51, 40–45. Daniel, H.J., Otto, R.T., Reuss, M. and Syldatk, C. (1998) Sophorolipid production with high yields on whey concentrated and rapeseed oil without consumption of lactose. Biotechnology Letters 20, 805–807. Desai, J.D. and Desai, A.J. (1993) Production of biosurfactants. In Biosurfactants, Vol 48, Surfactant Science Series ed. Kosaric, N. pp. 65–98. N. Y.: M. Dekker. Espuny, M.J., Ejido, S., Mercade´, M.E. and Manresa, A. (1995) Characterization of trehalose tetraester produced by a waste lube oil degrader Rhodococcus sp. 51T7. Toxicological and Environmental Chemistry 48, 83–89. Folch, J., Lees, M. and Sloane Stanley, G.H. (1957) A simple method for the isolation and purification of total lipids from animal tissues. Journal of Biological Chemistry 226, 497–509. Guerra-Santos, L., Ka¨ppeli, O. and Fiechter, A. (1984) Pseudomonas aeruginosa biosurfactant production in continuous culture with glucose as carbon source. Applied and Environmental Microbiology 48, 301–305. Guerrero, A., Casals, I., Busquets, M., Leo´n, Y. and Manresa, A. (1997) Oxidation of oleic acid to (E) -10-hydroperoxy-8-octadecenoic acid (E)-10-hydroxy-8-ocatadecenoic acids by Pseudomonas sp. 42A2. Biochimica et Biophysica Acta 1347, 75–81. Kajimura, Y., Sugiyama, M. and Kaneda, M. (1995) Bacillopeptins, new cyclic lipopeptide antibiotics from Bacillus subtilis FR-2. Journal of Antibiotics 48, 1095–1103. Ka¨ppeli, O., Mu¨ller, M. and Fiechter, A. (1978) Chemical and structural alterations at the cell surface of Candida tropicalis, induced by hydrocarbon substrate. Journal of Bacteriology 133, 952–958. Kates, M. (1972) Techniques of lipidology. In Biochemistry and Biochemical Biology ed. Work, T.S.W.E. pp. 393–444. Amsterdam: North-Holland Publishing Co. Kock, J.L.F., Botha, A., Blerh, J. and Nigam, S. (1996) Used cooking oil: science tackles a potential health hazard. South African Journal of Science 50, 513–514. Koch, A.K., Reiser, J., Ka¨ppeli, O. and Fiechter, A. (1988) Genetic construction of lactose-utilizing stains of Pseudomonas aeruginosa

and their application in biosurfactant production. Bio/Technology 51, 1335–1339. Kosaric, N., Cairns, W.L., Gray, N.C.C., Stechey, D. and Wood, J. (1984) The role of nitrogen in multiorganism strategies for biosurfactant production. JAOCS 61, 1735–1743. Lang, S. and Philp J. (1999) Surface active lipids in Rhodococci. Antonie van Leeuwenhoek 74, 59–70. Lang, S. and Wullbrandt, D. (1999) Rhamnose lipids– biosynthesis, microbial production and application potential. Applied Microbiology and Biotechnology 51, 22–32. Lowry, O.H., Rosebrought, N.J., Farr, A. and Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry 139, 265–274. Manresa, A., Bastida, J., Mercade´, M.E. et al. (1991) Kinetic studies on surfactant production by Pseudomonas aeruginosa 44T1. Journal of Industrial Microbiology 8, 133–136. Matsuyama, T. and Nakagawa, Y. (1996) Surface-active exolipids: analysis of absolute chemical structures and biological functions. Journal of Microbiological Methods 25, 165–175. McInerney, M., Javaheri, M. and Nagle, D. (1990) Properties of the biosurfactant produced by Bacillus licheniformis strain JF-2. Journal of Industrial Microbiology 5, 95–102. Mercade´, E., Manresa, A., Robert, M., Espuny, M.J., Bosch, M.P., Parra, J.L. and Guinea, J. (1988). New surfactant isolated from Pseudomonas sp 42A2. JAOCS 65, 1915–1916. Mercade´, M.E., Manresa, A., Robert, M., Espuny, M.J., de Andre´s, C. and Guinea, J. (1993) Olive oil mill effluent (OOME). New substrate for biosurfactant production. Bioresource Technology 43, 1–6. Mercade´, M.E. (1994) The use of agroindustrial by-products for biosurfactant production. JAOCS 71, 61–64. Mercade´, M.E., Monleo´n, L., de Andre´s, C., Rodo´n, I.E.M., Espuny, M.J. and Manresa, A. (1996) Screening and selection of surfactant-producing bacteria from waste lube oil. Journal of Applied Bacteriology 81, 161–166. Mielke, S. (1992) Trends in supply, consumption and prices. In Oils and Fats in the Nineties ed. Shukla, F.V.K.S. and Gunstone. Lystrup, Denmark: International Food Science Centre. Morikawa, M., Ito, M. and Imanaka, T. (1992) Isolation of a new surfactin producer Bacillus pumillus A-1, and cloning and nucleotide sequence of the regulator gene,psf-1. Journal of Fermentation and Bioengineering 74, 255–261. Passeri, A., Lang, S., Wagner, F. and Wray, V. (1991) Marine biosurfactants, II. Production and characterisation of an anionic trehalose tetraester from the marine bacterium Arthrobacter sp. EK 1. Zeitscrift fu¨r Naturfosch 46c, 204–209. Patel, R.M. and Desai, A.J. (1997) Biosurfactant production by Pseudomonas aeruginosa GS3 from molasses. Letters in Applied Microbiology 25, 91–94. Rapp, P., Bock, H., Wray, H. and Wagner, F. (1979) Formation, isolation and characterization of trehalose dimycolates from Rhodococcus erythropolis grown on n-alkanes. Journal of General Microbiology 115, 491–503. Robert, M. (1989) Aislamiento y seleccio´n de microorganismos productores de biotensioactivos. Produccio´n y caracterizacio´n de biotensioactivos por Pseudomonas aeruginosa 44T1. PhD Thesis. Barcelona: University of Barcelona. Shepherd, R., Rockey, J., Sutherland, I.W. and Roller, S. (1995)

© 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387

P S. AE R UG IN O SA RH A MN OL I PI D P R OD UC T IO N 387

Novel bioemulsifiers from microorganisms for use in foods. Journal of Biotechnology 40, 207–217. Spoecker, S., Wray, V., Nimtz, M. and Lang, S. (1999) Glycolipids of the smut fungus Ustilago maydis from cultivation on renewable resources. Applied Microbiology and Biotechnology 51, 33–39. Syldatk, C., Lang, S. and Wagner, F. (1984) Chemical and physical characterization of four interfacial-active rhamnolipids from Pseudomonas sp. DSM 2874 grown on n-alkanes. Zeitscrift fu¨r Naturforsch 4c, 51–60.

Wasko, M.P. and Bratt, R.P. (1990) Properties of a biosurfactant produced by the fuel contaminant Ochrobactrum anthropii. International Biodeterioration 27, 265–273. Yakimov, M., Timmis, K., Wray, V. and Fredrickson, H. (1995) Characterization of a new lipopeptide surfactant produced by thermotolerant and halotolerant subsurface Bacillus licheniformis BAS50. Applied and Environmental Microbiology 61, 1706– 1713.

© 2000 The Society for Applied Microbiology, Journal of Applied Microbiology 88, 379–387