Alcohol Production by Fish Spoilage Bacteria

Transcription

1 1055 Journal of Food Protection, Vol. 46, No. 12, Pages (December 1983) Copyright". Intemational Association of Milk, Food and Environmental Sanitarians Alcohol Production by Fish Spoilage Bacteria AEJAZ AHAMED and JACK R. MATCHES'" Institute for Food Science and Technology, College of Ocean and Fishery Sciences. University of Washington, Seattle, Washington (Received for publication August 27, 1982) ABSTRACT Bacterial isolates (244) identified to genera were tested for their ability to produce ethanol, isopropanol and propanol in a fish tissue extract. All of the isolates produced ethanol and 241 and 227 produced isopropanol and propanol, respectively. One high alcohol producing member of each of the groups Moraxellalike, Pseudomonas, Flavobacterium, Micrococcus and corynefonns was selected for utilization of fish components as substrates in production of alcohol. The substrates tested included four sugars, nine amino acids and lactic and pyruvic acids. Although there were some variations in the levels of alcohols produced by the test organisms from the substrates, the organisms appeared to prefer simple 5 and 6 carbon sugars and then utilized the free amino acids. The level of oxygenation greatly affected the levels of alcohols produced. The presence of alcohol in canned fish has been reported (1.2,8,9,11). The measurement of alcohols in raw fish was reported by Holaday (3) for mackerel, salmon and sardines and by Human and Khayat (7) for raw tuna. Only limited data have been reported on the numbers and types of bacteria producing alcohol during decomposition. Human and Khayat (7) measured alcohols and other compounds in decomposing tuna. However, the numbers of bacteria were not measured. This study was designed to determine if the bacteria causing spoilage of salmon and trout could produce alcohol, the possible fish components used for alcohol production, and the effects of oxygenation on alcohol production. MATERIALS ASD METHODS Fish species King salmon (Oncorhynchus tshawytscha) were obtained fresh as they returned to the School of Fisheries hatchery. Rainbow trout (Salmo irrideus) were reared in the School of Fisheries hatehery. Salmon and rainbow trout were held at 5 C for 16 d, and samples were removed at intervals for bacterial isolation. Sample preparation Chunks of fish were blended with distilled water (1:5 dilution) at high speed in a Waring Blendor for 1-1/2 min. For bacterial isolation, O.I-ml quantities of the appropriate serial decimal dilutions were spread on the surface of dried plate count agar (Difco) supplemented with 0.5% :-;lac!. Incubation was at 20-2 I DC for 5 d or until colonies were large enough for isolation. Bacterial isolation and identification Bacterial colonies were picked from the spread plates using a random numbers technique. The plates were placed over a number grid, numbered tags were selected randomly, and a colony was picked from the grid having the number of the tag selected. All isolates were streaked on plate count agar three times to ensure purity. Each isolate was tested for gram reaction, cellular and colony morphology, pigment production, oxidase reaction, catalase production and motility under phase contrast microscopy. Oxidase-positive motile rods were tested for oxidative or fermentative metabolism using the medium of Hugh and Leifson (6). The oxidase test was performed by rubbing the culture on a strip of filter paper impregnated with tetramethyl-p-phenylenediamine 2HCI. Cultures tuming purple within 60 s were considered positive. Test media A fish tissue extract medium was prepared by sterilizing a 1:5 homogenized dilution of fish in distilled water. The liquid portion containing the soluble components was used for alcohol production by the isolates. Dilute media were prepared by adding I g of nutrient broth (Difco) to I L of distilled water and sterilizing. The dilute fish tissue medium was prepared by adding I g of homogenized fish tissue to I L of distilled water and sterilizing. To 100 ml of both the fish tissue and nutrient broth were added 1.0 ml of dilute solutions of the test substrates (0.0 I molar). Each tube of test medium (10 ml) was inoculated with I loop of a 24-h-old culture of the test organism. Organisms used The isolate producing the highest level of alcohol in the fish tissue extract was chosen from each of the groups shown in Table I. These organisms were then used in the substrate studies. Levels of oxygenation To measure levels of oxygenation on alcohol production, the dilute fish medium (I g of fishll of water) was inoculated with 3 test organisms (Moraxella-like, Pseudomonas and Flavobacterium) and 10 ml were dispensed into 16 X 150 mm screw-capped tubes and 50-ml Erlenmeyer flasks. The medium in screw-capped tubes produced the lowest level of oxygenation. The next level was obtained by placing the medium in Erlenmeyer flasks giving a large medium surface area. The highest level of oxygenation was obtained by placing 50-ml Erlenmeyer flasks containing 10 ml of medium on a New Brunswick gyratory water bath shaker at 150 RPM. All samples were incubated at room temperature (20-21 C). Substrates used The substrates used in these studies were chemically pure compounds from several manufacturers. These included four sugars: fructose, ribose, glucose and sucrose; nine amino acids: lysine, methionine, leucine, valine, alanine, glycine, glutamic acid, cysteine and histidine. and lactic and pyruvic acids. JOURNAL OF FOOD PROTECTION, VOL. 46, DECEMBER 1983

2 1056 AHAMED AND MATCHES Chromatography Alcohol was measured by injecting a 2 ILl sample into a Carle 211 Analytical Gas Chromatograph fitted with a 6 ft x 1/8 in. stainless steel column packed with mesh Porapak Q. A flame ionization detector operating with H2 and breathing air (29 and 30 PSIG, respectively) and N2 as the carrier gas (30 PSrG) was used. The column temperature was adjusted to 160 C. Injection port and detector temperatures were 180 and 200 C, respectively. During use a range and attenuation of 10 anf 4 were found adequate to give good separation and peak height. The alcohols measured in the studies were ethanol, isopropanol, and propanol. External standards were used and alcohol was calculated from peak area. RESULTS AND DISCUSSION Bacteria were isolated from both king salmon (Oncorhynchus tshawytscha) and rainbow trout (Salmo irrideus) during storage at 5 C for 16 d. A total of 244 isolates were tested and identified. These data are shown in Table 1. The gram-positive flora composed of Micrococcus and coryneforms made up 20% of the total population. The gram-negatives made up the remainder of the organisms isolated. Flavobacterium and Vibrio were present but in relatively low numbers during the storage period. Moraxella-like organisms made up more than half the total population (52%). Although low numbers of bacteria were isolated on several sampling days (3, 5, 6 and 8 d of storage), the number of Moraxella-like organisms made up between 29 and 100% of the populations for each sampling day. The rates (in percentages) of gram-negative to grampositive organisms at days 0 and after 3, 4, 5, 6, 7, 8, 10, 11, and 16 d of storage were 81119, 100/0, 85/15, 56/44, 100/0, 76/24, 89/11, 80/20, and 80120, respectively. The consistent levels of gram-positives and the lower than expected number of Pseudomonas during later storage TABLE 1. Total numbers and Bacterial groups isolated may be due to the water from which the fish were collected. Also, the incubation temperature of 5 C may have selected for the Moraxella-like organisms, which remained high throughout storage. The University of Washington School of Fisheries hatchery receives the major portion of its water from the Lake Washington Ship Canal. Salmon returning to the hatchery must pass through this canal between salt water of Puget Sound at the Chittenden locks and the hatchery pond (approximately 5 miles). The rainbow trout were reared in this water at the hatchery. The kind of bacteria on fish are influenced by the microflora of water. This was shown (5) for Atlantic salmon during upstream migration and for Pacific salmon during out migration, ocean residence and upstream migration (14). Each of the 244 isolates was tested for ability to produce alcohol from sterile fish tissue extract. The organisms and the alcohols produced are shown in Table 2. The concentrations of alcohols produced in the fish tissue extracts are the highest for ethanol followed by isopropanol and propanol. All of the 244 isolates tested produced ethanol at levels as high as 628 ppm. This high level was produced by both Pseudomonas and Moraxella-like groups which also contained strains producing the lowest levels of 64 and 40 ppm, respectively. These data and those for the other genera show some variations within a given genus. Isopropanol was produced by all isolates except three of the Moraxella-like organisms. The levels of isopropanol produced were much lower than the levels of ethanol, and the minimum levels measured were as low as 4 ppm. The pattern for propanol production by the isolates was similar with the exceptions of lower levels produced and fewer isolates showing production. Strains of Vibrio, salmon and rainbow trout stored at 5 C II 16 Moraxella-like Pseudomonas Flavobacterium Micrococcus Coryneforms Vibrio Total TABLE 2. Concentrations tissue extract at 5 C. Number Organism isolates Ethanol Propanol tested max min aver max ave max min ave Pseudomonas 40 (40)b (40) (37) Moraxella-like 132 (132) (129) (120) Flavobacterium 14 (14) (14) (14) Micrococcus 36 (36) (36) (34) Coryneforms 12 (12) (12) (12) Vibrio Total 244 "Maximum, minimum and average levels of alcohol produced. bn umbers in parentheses are the numbers of isolates producing the particular alcohol. JOURNAL OF FOOD PROTECTTON, VOL. 46, DECEMBER 1983

3 ALCOHOL PRODUCED BY FISH BACfERIA 1057 cornyneforms and Flavobacterium all produced propanol, but 12, 3 and 2 strains of Moraxella-like, Pseudomonas and Micrococcus strains, respectively, failed to produce this alcohol. Few data are available in the literature on production of these alcohols. Human and Khayat (7) reported ethanol, propanol and butanol in tuna but not isopropanol. The presence of ethanol was reported in canned fish (1,2,4,8,9) and raw fish (3,7). Data in this report show the production of alcohol by the bacteria causing spoilage of both salmon and trout. The substrates used by the bacteria for alcohol production were speculative. Lerke and Huck (9) suggested that ethanol is a common metabolite of a variety of microbial genera and could be derived from carbohydrates via glycolysis and also from the deamination and decarboxylation of alanine. The levels of glycogen in fish have been reported to be as high as 5% in teleosts (12), and this glycogen could be utilized by bacteria. Generally glycolysis proceeds soon after death when the tissues become anaerobic. By the time bacterial activity starts following rigor much of the glycogen could have been converted to lactic acid. Therefore lactic acid was also considered a possible substrate. Other components of fish tissue considered as possible substrates included ribose from ATP degradation, glucose and amino acids. Also included were non-fish components, sucrose and fructose. In preliminary experiments, the fish tissue extract was prepared using I part of fish and 4 parts (w/v) of water. The effects of added substrates on alcohol production could not be measured because the nutrient level of the fish was too high. Therefore, both a dilute nutrient broth (l gil) and dilute fish tissue homogenate (1 gil) were prepared. To these media were added 1 ml of each substrate (0.01 molar solution)/1 00 ml. These data are shown in Table 3 for the 5 test organisms. In the controls without added substrate, the a1cohollevels were below the minimum 3 ppm level of detection with our system. The data in Table 3 show that there is some variation in levels of alcohol produced by the 5 test organisms. Two of the common fish spoilage bacteria, Moraxella-Iike organisms and Pseudomonas, produced alcohol in nearly all the substrates. These 2 organisms utilized all the sugars tested: glucose, ribose, fructose and sucrose, two of which are fish components, for production of alcohol. In addition, Moraxella-like organisms utilized all of the amino acids tested (except cysteine) as well as lactic and pyruvic acids. The highest levels of ethanol were produced by Moraxella-like organisms in alanine, glycene, pyruvic acid and lysine; by Flavobacterium in glucose and fructose; by Micrococcus in ribose; coryneforms in fructose, glucose and sucrose and by Pseudomonas in glucose and ribose. this indicates that amino acids are preferred by the Moraxella-like organisms and the sugars are preferred by the other four organisms. Also, the amount of ethanol produced in the free amino acids (non-protein nitrogen compounds) is comparable to the levels of ethanol produced in the sugars. It can be hypothesized from these data that the organisms prefer simple 5 and 6 carbon sugars and then utilize the free amino acids (NPN compounds) when the levels of sugars decrease. This pattern would probably be followed by proteolysis and the production of additional NPN compounds. With some microorganisms, and especially yeast, production of alcohol is related to the level of oxygenation. Although the organisms normally causing fish decomposition are aerobic or facultative anaerobes, the effects of oxygenation on alcohol production were tested. The data obtained with glucose, ribose, lysine, cysteine, methionine and alanine are shown in Table 4. No alcohol was produced by the 3 test organisms in the dilute fish tissue medium control without supplementation. At the lowest level of TABLE 3. Production of alcohol (ppm) in dilute nutrient broth and dilute fish tissue extract (1.0 g per liter) supplemented with various substrates.oml molar 100 ml bacteria. Moraxella-like Pseudonwnas Flavobacterium Micrococcus Coryneforms Substrate NB' Fishb NB Fish NB Fish NB Fish NB Fish Control Fructose Lysine Methionine Leucine Valine Alanine Glycine Glutamic acid Ribose Cysteine Glucose Lactic acid Sucrose Histidine acid anb: Nutrient broth. bfish: Fish tissue extract. JOURNAL OF FOOD PROTECTION, VOL. 46, DECEMBER 1983

4 1058 AHAMED AND MATCHES oxygenation the maximum alcohol content was 40 ppm produced by Flavobacterium from glucose. In the flasks without agitation, 125 ppm ethanol were produced from cysteine and methionine by Pseudomonas. This alcohol level was greatly increased to 425 ppm ethanol in flasks with agitation by Moraxella-Iike organisms from ribose. In these studies, the incubation times were increased with the lower levels of oxygenation to allow for oxygen to dissolve and diffuse into the medium and permit the organism to increase in numbers. All of the media were sampled when the total populations reached 10 7 or higher per ml so that aeration or rate of oxygenation was the major variable. The counts obtained for the Pseudomonas test organism in the six substrates and also the levels of ethanol produced (calculated in ppm) by 10 6 cells are shown in Table 5. Although there are some inconsistencies, the data show that the highest levels of ethanol were produced from cysteine, glucose and methionine. Also, the lowest counts were obtained in the dilute medium containing cysteine. This was also true in two of three levels of oxygenation for glucose and methionine. The data presented here show that the bacteria causing decomposition of fish (salmon and trout) can produce alcohols. The organisms isolated from decomposing fish and shown to produce alcohol in a fish medium were also shown to be able to utilize fish components (sugars and NPN compounds) in the production of ethanol. Another measurement not done in this study is the possible use of TABLE 4. Production of ethanol by 3 test organisms in a dilute fish tissue extract supplemented with substrates at 3 levels of oxygenation. Substrate Control Pseudomonas Moraxella-like Flavobacterium ethanol Tubes Flasks without Flasks with Pseudomonas Glucose M oraxe lta-like Flavobacterium Pseudomonas Ribose Moraxella-like Flavobacterium Pseudomonas 16 SO 325 Lysine Moraxella-like Flavobacterium Pseudomonas Cysteine Moraxella-like Flavobacterium Pseudomonas Methionine Moraxella-Iike Flavobacterium Pseudomonas Alanine Moraxella-like ISO Flavobacterium "1.0 ml of 0.01 molar substrate per 100 ml of medium. TABLE 5. of Pseudomonas cells present in each substrate at the time of ethanol analyses and parts per million ethanol produc Jd' cells. Tubes with Flask without Flask with Substrate Glucose Ribose Lysine Cysteine Methionine Alanine 2.85 X 10 7 LI2 3.0x ,2 X 10 9, X 10 9, X X 10 9,OS 1.6 X X ,78 3,0 X 10 8, X 10 8, X 10 8, X 10 9, X X X x X x D ,07 JOURNAL OF FOOD PROTECTION. VOL 46. DECEMBER 1983

5 ALCOHOL PRODUCED BY FISH BACTERIA 1059 alcohol by the spoilage bacteria in the fish tissue. These studies were not accompanied by the use of labelled compounds which would be desirable. Since the fish spoilage bacteria can produce alcohol during decomposition and can utilize fish components as substrates, the use of ethanol as a chemical index of decomposition appears to be more promising, especially in products such as salmon and rainbow trout. Production of ethanol has not been tested in other fish species in our laboratory; however, the bacterial populations on fish from temperate waters are similar and we feel alcohol production would take place. ACKNOWLEDGMEST This research was supported by the Washington Sea Grant Program. Grant No. NA 79-D from the National Oceanic and Atmospheric Administration. Contribution No. 631, School of Fisheries. University of Washington. Seattle, Washington. REFERESCF.8 1. Crosgrove, D. M A rapid method for estimating ethanol in canned salmon. J. Food Sci. 43: Hillig. F Determination of alcohol in fish and egg products. J. Assoc. Off. Agr. Chern. 41: Holaday, D. A The alcohols as a measure of spoilage in canned fish. J. Assoc. Off. Agr. Chern. 41: Hollingworth, T. A., Jr., and H. R. Throm Correlation of ethanol concentration with sensory classification of decomposition in canned salmon. J. Food Sci. 47: Horsley, R. W The bacteria of Atlantic salmon (Salrna salar L.) in relation to its environment. J. Appl. Bacteriol. 36: Hugh, R., and E. Leifson The taxonomic significance of fermentative vs. oxidative metabolism of carbohydrates by various gram-negative bacteria. J. Bacteriol. 66: Human, J., and A. Khayat Quality evaluation of raw tuna by gas chromatography and sensory methods. J. Food Sci. 46: , Khayat, A Correlation of off-odor scores of canned tuna with gas chromatographic data. J. Food Sci. 44: Lerke, P. A., and R. W. Huck Objective determination of canned tuna quality: identification of ethanol as a potentially useful index. J. Food Sci. 42: Liston, J Microbiology in fishery science. Symposium on Fish Science and Technology, Torry Research Station, Aberdeen. Scotland. II. Miller. A.. and H. Throm Ethanol determination an index of decomposition in canned salmon. Private communication. 12. Portman. W Changes in proteins. nueleotides and carbohydrates during rigor mortis. pp In R. Kreuzer (cd) The technology of fish utilization. FAO Fishing News Ltd. London. England. 13. Shewan. J. M The microbiology of fish and fishery products- a progress report. J. App!. Bacteriol. 34: Yoshimizu, M.. and T. Kimura Studies on the intestinal microflora of salmonids. Fish Patho!. 10: lou RNAL OF FOOD PROTECTION, VOL. 46. DECEMBER 1983