Vol. 4: 27-33. Biological quality of fermented fish offa and chicken by-products T. Mikael Lassén Division ofAnimal Nutrition, Department ofAnimal Science and Animal Health, Royal Veterinary and Agricultural University, Bulowsvej 13, DK-1870 Frederiksberg C„ Denmark The biological quality of fermented animal by-products prepared from fish and chicken offal was evaluated. A quality index (QI) based on analyses of a few important free amino acids and their corresponding biogenic amines was given. A group separation method based on ion exchange chro- matography was found suitable for isolating and purifying amines and amino acids in biological samples. Quality was evaluated in samples fermented with different starter cultures, inoculation siz- es and substrate levels. Slow or incomplete fermentation led to the accumulation ofbiogenic amines, especially tyramine, and resulted in a low QI. Fast initial and continuous stable fermentation for four weeks resulted in low concentrations of biogenic amines and high concentrations of free amino acids and consequently a high QI. The alanine concentration was considered to be a useful tool for estimat- ing proteolysis, and QI related to changes in alanine concentration, lactic acid production and redox potential were considered to give the most useful estimation of the biological quality of fermented fish offal and poultry waste. Key words', biological silage, animal offal, biogenic amines, amino acids, HPLC ntroduction Biogenic amines comprise a large and heteroge- neous groupof natural products of special inter- est owing to their physiological effects. They are produced in enzyme catalysed reactions with amino acids as direct precursors, some ofwhich are common to all living cells. Amino acids are transformed into biogenic amines in sequences of reactions with several intermediates, some of which may be specific for biochemical reactions in many organisms, organs, tissues or cells (Egg- um et al. 1988a). Amino acid decarboxylases (ADC) from microorganisms are important fac- tors for production of biogenic amines (Bprre- sen et al. 1988, Eggum et al. 1988a). Adverse effects on animal growth and health may be a result of excessively high dietary concentrations of both psychoactive and vasoactive biogenic amines (Eggum et al. 1988a). The presence of biogenic amines in animal by-products such as fish and slaughter offal may have toxicological implications. Scombroid fish © Agricultural Science in Finland Manuscript received February 1994 27 AGRICULTURAL SCIENCE IN FINLAND Lassén, T. M.: Biological quality offermentedfish offal and chicken by-products poisoning of humans occurs because of inges- tion of pelagic fish, e.g. tuna and mackerel, which contain unusually high levels ofhistamine (Taylor 1986, 1988).Animal feeds often include whole fish, fish offal or fish meal, and large amounts of fishery products in feed have been implicated in outbreaks ofillness, e.g., diarrhoea, and poor growth performance among animals. High levels of histamine or other biogenic amines have been proposed as causative agents (Skadborg 1985, Eggum et al. 1988b, Klausen 1988). Biogenic amines or their products are con- sidered of interest in relation to the quality of foodstuffs of animal origin and fur-animal feed. Information on types, combinations and concen- trations of the actually harmful compounds caus- ing the problems observed in fermented offal of animal origin (Lassén et al. 1990,Urlings 1992) is scanty. Some of the problems associated with poor quality mink feed and the role of protein decomposition products in feed deterioration have been studied by Eggum et al. (1987). The results indicate that several problems still need to be solved, and that reproducible, simple and efficient methods for feed quality control have to be developed and evaluated. This study fo- cused on improving the analytical control of fer- mented silage to be used as animal feed by High Performance Liquid Chromatography (HPLC) determination of free amino acids and biogenic amines, and the relationship between them. Material and methods Biogenic amines and free amino acids were ana- lysed at the Chemistry Department, Royal Vet- erinary and Agricultural University, Denmark. To obtain fish silage of different qualities (silos 1 to 3), herring offal was fermented with 107 colony forming units (cfu) Lactobacillus plantarum (L.pl) g and 0,2, and 5% (w/w) dex- trose at 25°C (Lassén 1993a). The effects of dif- ferent starter cultures on the biological quality of different raw materials were studied by fer- menting herring offal (silos 4 to 7) and poultry waste (silos 8-11) with four different starter cul- tures (L.p, L.p :Pediococcus pentosaceus (P.p), L.p -.Pediococcus acidilactici (P.a), and Pelzyme®), and 5 % (w/w) dextrose at 25°C (Lassén 1993b). Analyses for amino acids and biogenic amines were made after 0, 1,2, and 7 days' storage (silos 1-3), and in samples taken weekly during the 4 weeks' storage period (si- los 4-11). Samples were frozen immediately, freeze dried, minced and homogenized. They were analysed for free amino acids and biogen- ic amines using modifications of methods de- scribed by Bjerg et al. (1984). Samples composed of 200 mg of fermented material and 200 pi of internal standard solution (2 pmol 3,4-Dimetoxy Phenylmetylamin and 3 pmol Norvaline/g) were extracted three times in 5 ml of 70% boiling methanol. The raw extract was air dried over- night, dissolved in 2 ml of water and separated into basic amino acids and biogenic amines us- ing CM-Sephadex 25® (Pharmacia, Sweden) ion exchange column (A) and into neutral and acid- ic amino acids using Dowex 50w x 8 200 mesh® (Pharmacia, Sweden) ion exchange column (B). The A column was eluated with 4 M acetic acid:methanol solution (1:1) and the B column with 2 M Pyridine solution. Both eluates were air dried overnight and dissolved in 500 pi of water and purified over Imm Bondapac ClB® (Pharmacia, Sweden). The Bondapac column was eluated with 1.5 ml of water and the eluates were air dried overnight and resolved in 200 pi of water. 10 pi of solution was used for quanti- tative determinationby HPLC. The HPLC meth- od used was pre-column derivationwith OPA (25 mg o-Phatalaldehyde OPA, 2 ml methanole, 250 pi potassium borate, and 25 pi mercapto propi- onicacid) (B-eluates) and NAP (16.5 mg p- Phatalaldehyde, 1.32 ml methanole, 165 pi po- tassium borate, and 29.73 mg N-acetyl-D-peni- cillum, NAP) (A-eluates). A-eluates were eluat- ed for 60 minutes with 25 mM phosphatebuffer (A) and 90% acetonitrile (C) using the follow- ing gradient: 0 min 90% (A): 10% (B); 45 65:35; 50 40:60; 60 90:10; B-eluates were eluated for 28 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 27-33. 60 minutes with 25 mM phosphatebuffer (A) and 50% acetonitrile:phosphatebuffer (B) with the following gradient: 0 min 100% (A):0% (B); 20 80:20; 40 65:35; 50 40:60; 60 100:0. The HPLC column was a SuperPac spherisorb ODS2® 3p.m x 125 mm (Pharmacia, Sweden) and the column temperature was 30°C, flow rate was 1 ml/min, and the absorbance was measured at 340 nm. The raw data were processed by computer with an HPLC manager (Pharmacia, Sweden). Feed quality was presented by a quality in- dex (QI) expressed by the following formula; (Lysine + Arginine + Tyrosine) Ql = (Ornithine +Cadaverine + Putrescine +Tyramine) Samples from the 11 different treatments de- scribed above were analysed for contentsof free amino acids and biogenic amines, and the qual- ity of the silage was evaluated according to QI. To estimate the degree of proteolysis, the alanine content was recorded in addition to QI. Results The study showed that herring offal (silo 1) could be stored anaerobically for two days at 25°C without affecting quality as measured by QI; however, it putrefied after one week of storage. Offal in silo 2 also putrefied after one week of storage, but fermentation with 5% dextrose (silo 3) resulted in stable silage with a high QI (0.75) after one week of storage (Fig. 1). The alanine content increased from an initial concentration of 18.7 pmol/g to 212.2 pmol/g after one week in silo 2, and from 16.5 pmol/g to 148.4 pmol/g in silo 3. If no dextrose was added (silo 1), alanine decreased from an initial concentration of 24 pmol/g to 7.9 pmol/g after two days and then increased once more to 14.0 pmol/g after one week (Table 1). In herring offal (silos 4-7) QI increased dur- ing the first week of storage and then decreased Table 1. Changes in alanine content (pmol/g) during storage for silage of different organoleptic qualities (silos 1-3) and for herring offal (silos 4-7) and poultry waste (silos 8-11) fermented with different starter cultures. Storage time Days Storage time Weeks S* 0 1 2 7 S* 0 I 2 3 4 1 24.0 13.7 7.9 14.0 2 18.7 43.3 58.6 212.2 3 16.5 26.6 50.6 148.5 4 14.5142.2 145,0 218.8214.3 7.856.5 89.2103.3 120.5 11.355.6 97.1123.3 149.6 25.059.8 80.8184.1 104.1 20.197.4 120.1144.7 183.8 44.098.9 112.5134.6 144.0 34.485.94 113.0135.7 156.4 35.098.2 104.3139.2 203.1 5 6 7 8 9 10 II * S = silo number Fig. 1. Quality index (QI) for herring offal (silos 1-3) fermented with 107 colony forming units/g Lactobacillus plantarum and 0,2, and 5% dextrose at 25°C. 29 AGRICULTURAL SCIENCE IN FINLAND Lassén, T M.: Biological quality offermentedfish offal and chicken by-products and stabilized at about 0.75 for good quality si- lage and 0.25 for poor quality silage (Figure 2.). In this type of raw material alanine increased during the first three weeks from an initial con- centrationof 14.5 pmol/g to 218.8 pmol/g (L.p), and from 25.0 pmol/g to 184.1 pmol/g (Pelzyme®), and then decreased to 214.3 pmol/ g, and 104.1 pmol/g, respectively. For L.p:P.p and L.p:P.a alanine increased throughout the period from initial concentrations of 7.8 pmol/g and 11.3 pmol/g to 120.5 pmol/g and 149.6 pmol/g, respectively (Table 1), but increasing concentrations of biogenic amines, especially tyramine, resulted in a low QI. The quality of the fermented silage was acceptable for all four types of silage, Pelzyme® yielding the most ac- ceptable odourand texture. There was no major difference in QI between poultry wastes (silo 8-11), except slightly in- creased QI for silage fermented withL.p:P.p and L.p:P.a after two weeks of storage, but after four weeks of storage the final QI was almost the same (Figure 3). Alanine content (Table 1) in- creased throughout the period of storage from an initial concentration of 20.1 pmol/g to 183.8 pmol/g (L.p), 44.0 pmol/g to 134.6 pmol/g (L.p:P.p), 34.4 pmol/g to 156.4pmol/g (L.p:P.a), and 35 pmol/g to 203.1 pmol/g (Pelzyme®). The quality of the fermented silage was acceptable for all silages with a primarily sour odour after four weeks of storage. The offal liquefied after one week of fermentation, and separated into oily, aqueous and solid fractions due to hydrol- ysis of the material; herring offal was the most liquefied. Discussion QI is a new way of presenting analytical infor- mation obtained from analyses of the amino acids released from protein and accumulated bio- genic amines. The amino acids lysine, arginine, and tyrosine were used in the index for the esti- mation of quality, because they are readily re- leased from protein and metabolised to the cor- responding biogenic amines (cadaverine, orni- thine and putrescine, and tyramine). Histidine and histamine were not chosen, because in- creased concentrations of histamine in ferment- ed products (herring offal and chicken offal) were only detected when the product was already organoleptically unacceptable. The alanine con- tent was chosen as a tool to compare the amounts of amino acids released from protein over time, because alanine is readily released but hardly metabolised at all during storage. Therefore, the reliability of QI improves with simultaneous re- cording of alanine content in the product. The finding of a low QI in spoiled fish offal Fig. 2. Quality index (QI) for herring offal (silos 4-7) fermented with different lactic acid bacteria cultures ( 10* colony forming units/g, 5% dextrose at 25°C). Fig. 3. Quality index (QI) for chicken offal (silos 8-11) fermented with different lactic acid bacteria cultures (1 O*col- ony forming units/g, 5% dextrose and 12% extruded wheatmeal:feathers (2:1) at 25°C). 30 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 27-33. (stored at 25°C for one week, silo 1) was as ex- pected, due to the high concentrations of bio- genic amines. Spoiled herring offal was, how- ever, characterized by an increase in QI during the first two days of storage, primarily due to low concentrations of biogenic amines. This is probably an outcome of less hydrolase activity and slower release of amino acids from protein in unfermented herring offal, which was con- firmed by lower concentrations of free alanine. After one week at 25°C almost all the released amino acids had decomposed to biogenic amines (cadaverine and tyramine) and ornithine, where- as the more stable silage (silo 3) had a higher QI due to lower concentrations of biogenic amines. High concentrations of biogenic amines were also found in poor quality silage (silo 2) after one week of storage (low QI). When herring of- fal was fermented with different lactic acid bac- teria (LAB) (silos 4-7), a difference in QI de- velopment between the cultures was observed, the silage culture Pelzyme® generally having a higher QI than the others. For chicken offal fer- mented under similar conditions (silos 8-11) no majordifference between the LAB cultures test- ed was observed. Fermented silage was general- ly characterized by an increased alanine concen- tration over time, which describes the hydroly- sation and proteolytic activity in the silage. Using QI not only provides a simple way of expressing feed quality, but also a way of com- paring analyses based on wet weight (ww) and dry matter. Data from a study by Klausen (1988), recalculated to QI, give the following informa- tion. At the beginning of the storage, herring vis- cera had a very high QI (10°C:10.2 and 20°C;6.9), due to the high content of free amino acids and low concentrations of biogenic amines. After 24 h QI decreased to 1.15 and 0.98, re- spectively, and finally decreased to 0.29 and 0.20, at which point the herring viscera were considered putrefied and spoiled. Both in the present study and in that of Klausen (1988), low QI could be explained by a high content of biogenic amines (tyramine and cadaverine) and low concentrations of free ami- no acids in the herring viscera. An initial increase in QI can be attributetd to release of amino acids from protein with little or no transformation into biogenic amines. This was often observed dur- ing the first week of storage after successful in- itial fermentation (silos 4,6, and 7). However, after one week of storage, QI decreased, and high concentrations of biogenic amines and ornithine were observed after four weeks of storage, even if the pH and redox potential in the silage were still low enough (Lassén 1993b) to inhibit en- zyme activity. The lactic acid producing bacte- ria itself might be responsible for ADC activity at low pH, because most of the spoilage bacteria are inhibited at pH < 4.5 or lack cellular activity (Lindgren 1985). The biochemistry of fish vis- cera is obviously quite different from that of the fillet (Bprresen et al. 1988), but studies on the biochemistry offish viscera are scanty. Stede and Stockemer (1981) found for whole herring stored at 6°C that the viscera contained 1.00 (imol his- tamine/g ww after 2 days at 6°C. Corresponding values for herring fillet were 0.41 pmol hista- mine/g ww. Klausen (1988) studied the content of free amino acids and biogenic amines in her- ring fillet and herring viscera stored at O°C for 10 days, and found the contents of tyramine, cadaverine and agmatine in the viscera to be approximately 1000, 200 and 125 times, respec- tively, the contents found in the fillet. Bprresen et al. (1988) found the concentration of most biogenic amines in freshly caught herring to be lower than 15 nmol/g ww in the fillet, and 1.5 pmol/g ww in the viscera, which was supported by the results obtained here. The biochemistry of fermentedpoultry offal is largely unknown, but Urlings (1992) studied the concentrations of biogenic amines in poul- try offal fermented with 10% beet pulp, 2% dex- trose and Lactobacillus plantarum and stored at 15°C for 21 days, and found lower concentra- tions than those reported by Eggum et al. (1987, 1988b). Eggum et al. (1987) found very low concen- trations of biogenic amines in offal from freshly slaughted poultry, but cadaverine and tyramine increased heavily after 24 h at 20°C. This find- ing was supported by the present study, in which 31 AGRICULTURAL SCIENCE IN FINLAND 1 Lassén, T. M.: Biological quality offermentedfish offal and chicken by-products cadaverine and tyramine increased during the first week of fermentation from almost undetec- table concentrations, (0.6, and 0.3 (tmol/g DM, respectively) to 12.7 pmol/g DM, and 13.3 (imol/ g DM, respectively. There are various possible mechanisms for explaining the accumulation of free amino acids and biogenic amines in fermented silage (Egg- um et al. 1988a, Klausen 1988). Undoubtedly the pattern observed in the present study resulted from the combined effect of autolytic and mi- crobial enzyme activities in which amino acids were released from protein. Biogenic amines were formed from free amino acids by autolytic decarboxylation or decarboxylation by microbi- al enzymes and removed by oxidative deamina- tion or other unidentified pathways. The de- creased alanine concentration in putrefied silage might be explained by the Stickland reaction, in which alanines react with glycine and H2O to form acetic acid, ammonia and carbon dioxide (Schlegel 1986). From the present and earlier studies (Lassén et al. 1990, Urlings 1992), it can be concluded thataccumulation ofbiogenic amines in ferment- ed animal by-products is a problem due to the decarboxylase activity caused by the LAB cul- tures added. QI might be an informative way of presenting and comparing data on biological feed quality in fermented animal by-products, and its informative value increases if combined with concentration of free alanine and parameters such as lactic acid production, pH and changes in redoxpotential. Changes in concentrations of both biogenic amines and free amino acids dur- ing storage ought, however, to be presented in pmol active molecules/g material regardless of whether the analyses are performed on dry or wet material. Acknowledgements. Financial support for this study was provided by the Academy of Finland. Sincere gratitude is due to Mrs Birthe lessen of Chr. Hansen’s Lab, A/S, Hors- holm, Denmark for providing the starter cultures, to Mr Peter Möllerof the Chemistry Department, for helpful and skilful assistance with HPLC analyses, and to Associate Professors Niels Enggaard Hansen, Anne-Helene Tauson and Hilmer Sorensen for valuable suggestions and com- ments regarding the manuscript. References Bjerg, 8., Olsen, 0., Rasmussen, K. W. & Sorensen, H. 1984. New principles of ion exchange techniques suit- able to sample preparation and group separation of nat- ural products prior to liquid chromatography. Journal of Liquid Chromatography 7: 691-707. Borresen, T., Klausen, N. K, Larsen, L. M. & Sorensen, H. 1988. Aminosyredekarboxylaser; Egenskaber og re- lation til biogene aminer og kvalitet af pelsdyrfoder. Faglig Årsberetning 1987. Dansk Pelsdyravlerforening. p. 174- 184. Eggum, B. 0,, Hansen, N. E., Henriksen, P. & So- rensen, H. 1987. Biogene aminer i pelsdyrfoder. Faglig Årsberetning 1986. Dansk Pelsdyravlerforening. p. 233- 245. -, Hansen, N. E., Henriksen, P. & Sorensen, H. 1988b. Biogene aminer i relation til kvalitet af pelsdyrfoder. Faglig Årsberetning 1987. Dansk Pelsdyravlerforening. p. 185- 207. -, Hansen, N. E. & Sorensen, H. 1988a. Amino acid pre- cursors of biogenic amines. In: Friedman, M. (ed.). Ab- sorption and utilization of amino acids. CRC Press, Boca Raton, Florida. 42 p. Klausen, N. K. 1988. Decarboxylation of tyrosine in re- lation to metabolism of amino acids and biogenic amines in fish during storage: Enzymes, Kinetics, and impor- tance. Ph.D Thesis, Chemistry Department, Royal Vet- erinary and Agricultural University, Denmark. 185 p. Lassén, T. M. 1995a, Evaluation of conditions for fer- mentation of fish offal. Agricultural Science in Finland 4: 11-17. -1995b. Lactic acid fermentation of fish offal and chick- en by-product with different starter cultures. Agricultural Science in Finland 4: 19-26. Hildén, A., Hildén, B. H. & Laitinen, M. J. 1990. Prak- tisk tillämpning av erfarenheter från försök med biolo- gisk konserverad ensilage i foder till mink och räv. NJF- Utredning/Rapport Nr. 60. 20 p. Lindgren, S. E. 1985. Användning av mjölksyrajäsande bakterier för konservering av animaliska råvaror. NJF- Seminarium Nr. 85. Aalborg, Denmark. 6 p. Schlegel, H. G. 1986. General Microbiology. Sixth Edi- tion Cambridge University Press. Cambridge, UK. p. 293- 302. Skadborg, J. 1985. Sundhedsskadelige staffer dannet ved mikrobiologisk, enzymatisk og oxidativ aktivitet i foder till pelsdyr. Hovedopgave i Pelsdyrproduktion, Royal Vet- 32 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 27-33. erinary and Agricultural University, Denmark. 64 p. Stede, M. & Stockemer, J. 1981. Bildung von Histamin in trischen Heringen und Makrelen. Fleischwirtschaft 61: 1746-1749. Taylor, S. L. 1986. Histamine Food Poisoning: Toxicolo- gy and Clinical Aspects. CRC Critical Reviews in Toxi- cology 17: 91-128. - 1988. Marine toxins of microbial origin. Food Technol- ogy 42: 94-98. Urlings, H. A. P. 1992. Fermentation of animal by-prod- ucts. Microbiological aspects of processing, epidemiolo- gy and animal nutrition. Diss. Utrecht University, Faculty of Veterinary Medicine, Department of the Science of Food of Animal Origin, Utrecht, The Netherlands. 135 p. SELOSTUS Fermentoitujen kala- ja kanajätteiden biologinen laatu T. Mikael Lassén Royal Veterinary and Agricultural University, Tanska Tutkimuksessa selvitettiin fermentoitujen kala- ja kanajätteiden biologista laatua. Tutkituille jätteille määritettiin laatuindeksi, jonka määrittämiseksi ana- lysoitiin tärkeät vapaat aminohapot ja niitä vastaavat biogeeniset aminit. Näytteet analysoitiin ryhmäerot- telu-menetelmällä, joka perustuu ioninvaihtokroma- tografiaan. Biologinen laatu tutkittiin näytteistä, jot- ka olivat eri bakteeriviljelmistä, erikokoisista siirros- tuksista ja erilaisilta kasvualustoilta. Hidas tai epätäydellinen fermentaatio aiheutti biogeenisten aminien, erityisesti tyramiinin, määrän kasvun ja lopputuloksena oli alhainen laatuindeksi. Nopean alkufermentaation ja jatkuvan vakaan neljän viikon fermentaation seurauksena oli alhainen bio- geenisten aminien pitoisuus, korkea vapaiden amino- happojen pitoisuus ja korkea laatuindeksi. Alaniinipitoisuuden todettiin olevan käytännölli- nen proteolyysin arviointiväline. Lisäksi muutokset alaniinipitoisuudessa vaikuttivat laatuindeksiin. Mai- tohappopitoisuus ja redox-potentiaali soveltuivat kaikkein parhaiten fermentoitujen kala- ja kanajättei- den biologisen laadun arviointiin. 33 AGRICULTURAL SCIENCE IN FINLAND