Lipid and exopolysaccharide production during hydrocarbon growth of a marine bacterium from the sea surface

The marine bacterium Alcaligenes sp. PHY 9 L.86 was isolated from hydrocarbon-polluted sea-surface waters and grown on 0.1 % tetradecan in batch cultures. Lipid composition of cell pellets and supernatants were examined throughout growth, using thin layer chromatography coupled with flame ionization detection Cellular and extracellular carbohydrate and protein contents were estimated. Bacterial growth on hydrocarbon induced the production of extracellular emulsifying agents (biosurfactants). Formation of foams was observed in the culture medium at early stationary phase; it was related to high emulsifying activity and maximum extracellular lipid production (in particular, free fatty acids and triglycerides). Specific staining for acid polysaccharides revealed the formation of exopolysaccharid fibers associated with vesicles, the size of which depended on the growth phase. Surface-active agents produced by Strain PHY 9 L.86 explain the foam formation. Our results stress the role of biosurfactants in the biodegradation of hydrocarbon in the marine environment. INTRODUCTION carbon substrate; it was therefore selected for detailed investigation of the production of en~ulsifying agents. In coastal areas of the Gulf of Marseille (France), In this study, a qualitative and quantitative analysis which is chronically polluted with hydrocarbons, foams of compounds released into the culture medium by can frequently be observed at the sea surface. The Strain PHY 9 L.86 was performed at various phases of ability of these foams to emulsify crude oil was noted the bacterial culture grown on hydrocarbon substrate. by Rambeloarisoa et al. (1984). Emulsifying agents are responsible for foam formation. They can be either synthetic emulsifiers (detergents) or tensioactive MATERIAL AND METHODS molecules of biological origin (biosurfactants). The fact that detergents have not been detected suggests a Organisms and growth conditions. The bacterial biological origin of the foams. Strain PHY 9 L.86 was identified as Alcaligenes sp. Studies by Rambeloarisoa et al. (1984) showed that according to the classification system proposed by hydrocarbons and hydrocarbonoclastic bacteria Buchanan & Gibbons (1974). accumulated extensively in these foams (lo7 to 108 Bacteria were grown a t 30 "C in 2 1 Fernback flasks bacteria ml-l). The ability of bacteria growing on hycontaining 300 m1 of mineral salt medium with the drocarbon substrates to produce emulsifying agents is following composition (in g 1-' distilled water): Tris well known (Rosenberg 1986). Thus, the production of (hydroxymethyl amino methane) (6.05), NaC1 (23), KC1 surface-active molecules by bacteria could be an im(0.?5), CaClz (1.47), NH,C1 (3.?4), MgS04 .7H20 (12.3), portant factor involved in foam formation. A mixed NaHPO, (0.4 mM). The pH was adjusted to 7.8 with bacterial community (EM4) composed of 8 bacterial Titrosol (Merck) (4 N). strains which degraded crude oil very effectively was Tetradecane (0.1 % , v/v) was provided as sole carbon isolated from foams (Rambeloarisoa e t al. 1984). and energy source. Aeration was provided by agitation Among this community, the marine Strain PHY 9 L.86 on a reciprocal shaker (96 rpm). Growth was monitored exhibited a maximum efficiency for degrading hydroby following absorbance at 450 nm with a Shimadzu 8 Inter-Research/Printed in F. R. Germany 260 Mar Ecol. Prog. Ser UV-visible spectrophotometer, by viable count estimates on agar medium plates (f~ltered seawater enriched with 15 g I-' Bjo-Trypcase [Merieux], 5 g 1-' Phytone [Merieux], 15 g 1-' Agar [Difco]) or by estimating the protein content using the method of Lowry et al. (1951). Whole contents of the flasks were harvested at various phases of growth, either for electron microscopical studies or chemical analyses. Emulsifying activity of the 6 0 0 0 ~ g supernatant was measured as a function of the degree of stability of an emulsion obtained after 5 rnin mechanical agitation of a reaction system or after exposure to sonic oscillation (Zosim et al. 1982) with 2 impulses for 1 rnin at 9 kHz each. The reaction system contained a determined quantity of supernatant corresponding to an amount of 0.09 m g total sugars (estimated according to the method of Dubois et al. 1956) (the quantity of liquid used could vary between culture media, but the sugar concentration remained constant), and 0.1 m1 of tetradecane and mineral salt medium without NH4C1 to a final volume of 10 ml. The stability of the emulsion over time was measured at 610 nm (Roy et al. 1979) at 20 "C. A similar reaction system prepared from synthetic seawater (i.e. without biosurfactants) served as control. Electron microscopy. Ahquots of cultures were sampled under the tetradecane upper layer and prepared for electron microscopy analysis. The techniques used were adapted to allow for the high fragility of the surface materials (exopolysaccharides, vesicules, etc). Since the classical methods of preparation for electron microscopic observation of thin sections turned out to be inappropriate, we used either negative staining with 1 ?'o potassium phosphotungstate (PTA) and 1 O/O uranyl acetate (UA), with or without previous stabilization of the bactenal surface structures with ruthenium red/ glutaraldehyde (G/RR/UA) which specifically reveals acidic polysaccharides (Mutaftshiev et al. 1982). A carbon-coated Formvar grid was deposited on top of a drop of a bactenal cell suspension culture, and left (2 min) for adsorption of the bacteria and their exostructures on the carbon surface. Free-standing material was removed by blotting with filter paper. Grids were treated with a mixture (v/v) of 1 % glutaraldehyde and 1 % ruthenium red, and then contrasted with 0.1 O/O uranyl acetate. Grids were stained for 5 rnin and examined with a Hitachi 600 electron microscope at 75 kV. Analytical methods. Centrifugation ( 6 0 0 0 ~ g for 15 min, twice, at 5 "C) of cultures ylelded 2 fractions desi.gnated cell pellets and supernatant. The 2 freeze-dried fractions were analysed for protein, carbohydrate and lipid content. Protein was estimated according to the method of Lowry'et al. (1951) with bovine serum albumin as standard. Carbohydrate was estimated using the phenol-sulfuric acid method of Dubois et al. (1956). h p i d was extracted according to the method of Bligh & Dyer (1959). Aliquots of the lyophilized fractions were suspended in 0.8 volumes of distilled water, 1 volume of chloroform and 2 volumes of methanol. The mixture was stirred overnight under nitrogen. After addition of 1 volume chloroform and 1 volume distilled water and stirring for 10 min, the mixture was filtered through GF/C glass fiber filters. Filtrates were allowed to separate into 2 phases. The bottom phase was collected as the Lipid extract. The filter was again extracted using the same procedure and the extracts combined. Lipids were separated by thn-layer chromatography coupled with flame ionization detection (TLC/FID) using a Iatroscan apparatus TH 10 (Iatron Laboratories, Tokyo). The Iatroscan was operated wlth a hydrogen flow of 75 kg cm-2, a n air flow of 2000 m1 min-' and a scanning speed of 0.32 cm S-' (30 tooth gear). The Iatroscan was connected to a Shimadzu-chromatopac CRlB Integrator. Lipid classes were separated on chromarods S I1 using several developments and partial scans, as proposed by Pamsh & Ackman (1983). In the first development (40 rnin hexane-diethyl etherformic acid 99: 1 : 0.1), the less polar lipids (hydrocarbons [HC], wax esters [WE] and free fatty acids [FFA]) moved away from the point of application. There were scanned and the scan was stopped manually. The remaining neutral Lipids (triacyglycerides [TG]; fatty alcohols [Alc]; diacylglycerides [DG]; sterols [ST]) were separated in the second development (40 rnin in hexane-diethyl ether-formic acid 80 : 20 : 0.1). Again, a partial scan was performed. This was followed by a short development in acetone (3 cm above the origin). This development separates monoglycerides, glycolipids and pigments from total phospholipids which do not move in acetone. In this set of samples, the pigments were visibly separated from monoglycerides and glycolipides. Lipids were identified on the basis of their ability to CO-chromatograph with authentic standards purchased from Sigma LTP Corp. Lipids were quantified with reference to calibration curves performed for each rod and each class of compound. Each rod was considered as an isolated analytical unit as recommended by Delmas et al. (1984). Calibrations were performed at 5 levels with loads in the range 0.5