564 COMMUNICATION Production of Eicosapentaenoic Acid by Freshwater Vibrio Einar Ringei*, Jens Petter Jelstensen and Rolf Erik Olsen 1 The Norwegian College of Fishep] Science, University of Tromsd, N-9001 Tromsd, Norway Arctic chart, Salvelinus alpinus (L.), were fed a moist pellet diet based on casein, dextrin and coconut oil in fresh water. The diet contained substantial amounts of satu- rated fatty acids, especially 12:0. Polyunsaturated fatty acids in the n:3 series were not detected in the diet, but analysis of the intestinal liquor stripped from the chart showed that eicosapentaenoic acid (EPA, 20:5n-3) ao counted for 3.5% of fatty acids in total lipid. Colony form- ing units (CFU) of aerobic microorganisms were approx- imately 104 per mL intestinal liqilor. Seventeen bacterial strains isolated from the intestinal liquor were screened for EPA production. Four isolates contained a high pro portion of EPA in their total lipid. These isolates belong to the species Vibrio. Lipids 27, 564-566 (1992). TABLE 1 Ingredients (g/kg dry weight) of the Diet a Ingredients Casein 706.5 Dextrin 170.0 Gelatin 17.0 Premix I 30.0 II 1.7 III 1.8 IV 20.0 Coconut oil 53.0 aA detaileddescriptionof the added pre- mixes and vitamins isgiven by Olsen et al. (11). Bacteria are generally considered not to contain polyun- saturated fatty acids (PUFA); instead such fatty acids are only believed to be formed de novo by photosynthetic organisms (1}. However, the presence of linoleic acid {18:2n-6) in cellular lipids has been demonstrated in Vibrio cholera (2), several Vibrio species {3) and in three bacterial species of the human pathogen Yersinia {4,5}.Eicosapenta- enoic acid (20:5n-3) has been reported as a membrane con- stituent in several marine bacterial species: deep-sea bacteria {6), Alteromonas sp. (7) Shewanella Alteromonas putrefaciens {8,9}and Flexibacter (10). These observations may indicate that microorganisms contain PUFA more frequently than earlier suggested. No investigations to date have reported that freshwater bacteria contain PUFA. The focus of our study was therefore to investigate whether bacteria isolated from freshwater fish contain PUFA or not. MATERIALS AND METHODS Fish. Alevins of Arctic chart, Salvelinus alpinus (L.), were reared on commercial feed (Tess Elite Pluss, Skretting LTD, Stavanger, Norway) from the initial feeding stage to an average weight of approximately 10 g. Thereafter the fish were fed the artificial diet. Diet. The diet consisted of a moist pellet based on ca- sein, dextrin and coconut off. The composition of the diet is given in Table 1. A detailed description of the different premixes and vitamins added to the diet is given by Olsen et aL {11).The diet components were thoroughly mixed, pelleted, and stored at -80~ prior to use *To whom correspondence should be addressed at the University of Troms~, Instituteof Biology and Geology, Dramsveien 201, N-9001 Troms~, Norway. ICurrent address: Finnmark College, Follumsvei, N-9500 Alta, Norway. Abbreviations: CFU, colonyforming units; EPA, eicosapentaenoic acid; GC, chromatography; MSD, mass selectivedetector; PUFA, polyunsaturated fatty acids; TSA, tryptic soy agar; TSB, tryptic soy broth. Experimental conditions. We examined 30 Arctic charr in our study, following the experimental design of Ring~ and Nflsen (12}. This study was undertaken under a natural photoperiod at 70~ from October to April with an average water temperature of 6.0~ _ 1.0~ Sampling of intestinal liquor. Ten fish, each weighing about 40 g, were stripped after having been anesthetized in 0.3% benzocaine. The belly of the fish was pressed and intestinal liquor was extruded through the rectum. The intestinal liquor was separated from the fecal pellet after standing for 2 min. Lipid analyses of diet and intestinal liquor. Analyses of diet and intestinal liquor were carried out in triplicate. Lipids were extracted by the method of Bligh and Dyer (13). Methyl esters of fatty acids were prepared by H2SO4 catalyzed transesterification of total lipid in methanol (14), and were analyzed by gas chromatography (Hewlett- Packard, Model 5890 A, Waldbronn 2, Germany) using a SP 2330 capillary column (30 • 0.25 mm i.d.) and helium as carrier gas. The temperature program employed was 60~ for 1 rain, followed by an increase of 30~ to 180~ for 7 rain and thereafter 5~ to 240~ Indi- vidual fatty acids were identified by comparison with known standards (Supelco 4-7019, 4-7042, 4-7033, 4-5589, Supelca Bellefont~ PA) and quantitated using a Hewlettt Packard 3393A integrator. Eicosapentaenoic acid (20:5n-3) was identified by gas chromatography-mass selective de- tector (GC-MSD) using a 5970 MSD (Hewlett-Packard) connected to a Hewlett-Packard work station 9000-300, and further confirmed by gas chromatography-mass spectrometry (GC-MS) using conditions described else- where (15). Isolation of microorganisms. Total viable counts were performed on TSAg plates containing TSA (tryptic soy agar) 40 g/L and 5 g/L glucose. Intestinal liquor was diluted in sterile 0.9% saline; 0.1 mL vol of appropriate dilutions were spread on the surface of the TSAg plates. The plates were incubated at 12~ and inspected daily for up to 4 wk. After enumeration, a representative selec- tion of colonies was subcultured on TSAg plates. After LIPIDS, Vol. 27, no. 7 (1992)