Histamine levels in commercially important fresh and processed fish of Oman with reference to international standards

Histamine was investigated in fresh (378), frozen (441), canned (290) and dried (24) fish samples of scombroid and non-scombroid species of Oman using a high performance liquid chromatography with a fluorescence detector. Of the 1133 fish samples tested, histamine was detected in 551 samples with a detection rate of about 41.8% among fresh fish, 61.0% of frozen fish, 78.9% of canned fish and 91.6% of dried fish samples and the mean histamine levels were 2.6, 5.8, 3.1 and 104 mg kg⁻¹, respectively. A total of 3.7% and 0.79% of the total samples exceeded the FDA and EU regulatory limits for histamine. Imported dried anchovies contained high histamine levels. The study confirms that post catching and commercialisation practices of seafood are adequate, warranting good quality fish and may not cause histamine risk to consumer in terms of human diet. While necessary monitoring may be done for imported dried fish products.     

Biogenic amines and potential histamine – Forming bacteria in Rihaakuru (a cooked fish paste)

The biogenic amine concentration in Rihaakuru (a fish paste) (n = 28), obtained from different parts of the Maldives (North, South, and Central), was determined by HPLC. Ten biogenic amines were detected; agmatine, not detected (ND) – 161 ppm; cadaverine, ND – 387 ppm; histamine, ND – 5487 ppm; putrescine, ND – 290 ppm; phenylethylamine, ND – 23 ppm; serotonin, ND – 91 ppm; spermine, ND – 329 ppm; spermidine, ND – 79 ppm; tryptamine, ND – <5 ppm; and tyramine, ND – 50 ppm. Nine biogenic amines were found in 3 samples, 8 in 10 samples, 7 in 6 samples, 6 in 3 samples, 4 in 5 samples, and 1 was found in 1 sample. Histamine was detected at levels that are regarded as a risk to human health. Fourteen isolates were selected from two randomly selected samples out of the 28 samples of Rihaakuru and screened for histamine production. Twelve of the 14 isolates produced histamine, with the highest histamine producers being Bacillus massiliensis Nai5 (6.65 ppm) and Bacillus polyfermenticus (5.58 ppm); while Bacillus malacitensis produced the least (<0.5 ppm).     

Commercial test kits and the determination of histamine in traditional (ethnic) fish products-evaluation against an EU accepted HPLC method

Although commercial test kits are generally used and/or evaluated for determining histamine in fresh and canned fish, and fish meal, there is little information on their performance and the application for traditional fish products (TFPs), which generally differ in product properties. In this study, three quantitative (Food EIA, Veratox, Histaquant) and four qualitative (Histasure, Histameter, Transia qualitative and semi-quantitative) commercial histamine test kits were evaluated against HPLC method for detecting histamine in several traditional fish products. Among the quantitative kits, Histamine Food EIA showed the best correlation with HPLC method for TFPs (R² = 0.9132) as well as good recoveries ranging between 89 (±4.11)% and 117 (±1.50)%. Although good recoveries were also observed with Veratox kit, poor correlation was found with HPLC. Poor correlation and low recoveries were also observed with Histaquant. Histasure and Transia tube histamine kits showed good agreement with HPLC results. However, the detection limit for Transia qualitative kit is 100 ppm, and for Histasure and Transia semi-quantitative kits can be set to 50 ppm. Therefore, Histasure and Transia semi-quantitative kits are found more suitable for either HACCP monitoring histamine in seafood processing plants or regulatory purposes according to Food Drug and Administration (FDA) legislation for TFPs. However, Transia semi-quantitative kit should be used in caution for histamine cut off values <50 ppm. This study shows that each test kit can represent different performance for determining histamine in TFPs according to product type, and therefore new commercial test kits should be evaluated against an approved analytical method before applications in future for these types of products.     

An amperometric biosensor for the rapid assessment of histamine level in tiger prawn (Penaeus monodon) spoilage

Histamine could accumulate in seafood when bacteria spoilage commenced and caused histamine poisoning without altering the fish normal appearance and odor. Therefore, a histamine biosensor using immobilized enzyme diamine oxidase (DAO) has been developed for the rapid monitoring of the histamine levels in tiger prawn (Penaeus monodon). The histamine biosensor had a response time of <1 min and optimum pH of operation was 7.4 with reproducibility and repeatability (n = 5) of 4.87% and 5.26% relative standard deviations (RSD) respectively. Recoveries ranging from 93.11% to 100.58% were obtained for histamine spiked at levels from 5 to 20 ppm. The variation in histamine levels of some tiger prawn samples after a 5-h exposure at temperature of 30 °C ± 2 were studied using the histamine biosensor and the results were comparable to histamine levels determined by an HPLC method. The two methods showed a linear correlation with R² = 0.9612 (Y = 0.9164x + 5.58). The limit of detection was 0.65 ppm of histamine, which is below the indicator level of 50 ppm established by USA FDA. The reusable biosensor is simple and can be used for direct histamine determination without further sample pretreatment, and is suitable for the routine analysis of histamine in tiger prawns to monitor spoilage.     

Histamine formation by histamine-forming bacteria in douchi,a Chinese traditional fermented soybean product

Seven soybean and 19 black bean douchi products sold in the supermarkets in southern Taiwan were purchased and tested to determine the occurrence of histamine and histamine-forming bacteria. The levels of pH, salt content, water content, yeast and mold, and aerobic plate count (APC) in all samples ranged from 4.7 to 5.9, 4.4% to 14.0%, 6.8% to 51.6%, 3.0 to 5.1 log CFU/g, and 5.2 to 9.2 log CFU/g, respectively. None of these samples contained total coliform and Escherichia coli. Although black bean douchi products had an average histamine content of 29.0 mg/100 g, 18 of them had histamine contents greater than 5 mg/100 g, the allowable level set by the US Food and Drug Administration (FDA) for scombroid fish and/or products. In contrast, only four soybean douchi products had histamine levels greater than 5 mg/100 g. Among the black bean samples, four contained histamine at 56.3, 62.1, 80.2 and 80.8 mg/100 g, that are above the 50 mg/100 g hazard action level. Eight histamine-forming bacterial strains, capable of producing 11.7–601 ppm of histamine in trypticase soy broth (TSB) supplemented with 1% l-histidine (TSBH), were identified as Bacillus subtilis (four strains) Staphylococcus pasteuri (one strain) and Staphylocuccus capitis (three strains) by 16S rDNA sequencing with PCR amplification. S. capitis, which was previously reported to be halotolerant, was a potent histamine-former capable of producing more than 500 ppm of histamine in TSBH in the presence of 0.5–10% NaCl.     

Histamine content and histamine-forming bacteria in dried milkfish (Chanos chanos) products

Thirty-two dried milkfish products sold in five retail markets in southern Taiwan were purchased and tested to determine the occurrence of histamine and histamine-forming bacteria. Except for histamine and cadaverine, the average content of various biogenic amines in tested samples was less than 8.5 mg/100 g. Most of the tested dried milkfish products (78.1%) had histamine levels greater than the FDA guideline of 5 mg/100 g for scombroid fish and/or product, while fourteen of them (43.7%) contained > 50 mg/100 g hazard action level. Thirty histamine-producing bacterial strains, capable of producing 5.4 ppm to 562 ppm of histamine in trypticase soy broth supplemented with 1.0% L-histidine (TSBH), were identified as Enterobacter aerogenes (seven strains), Citrobacter sp. (one strain), Staphylococcusxylosus (ten strains), S. sciuri (one strain), Bacillus thuringiensis (two strains), Citrobacter freundii (five strains), Klebsiella pneumoniae (one strain) and E. cloacae (three strains), by 16S rDNA sequencing with PCR amplification.     

Histamine contents of salted seafood products in Taiwan and isolation of halotolerant histamine-forming bacteria

Fifty-seven salted seafood products sold in the fishing village stores in Taiwan, including salted fish product, salted mollusc product and salted shrimp product, were tested to determine the occurrence of histamine and histamine-forming bacteria. Although the average content of each of nine different biogenic amines in all samples was less than 5.0 mg/100 g, 10.5% (6/57) of tested samples had the histamine content greater than the 5.0 mg/100 g allowable limit suggested by the US Food and Drug Administration. One histamine-producing bacterial strain, capable of producing 78.5 ppm of histamine in trypticase soy broth supplemented with 1.0% l-histidine (TSBH), was identified as Bacillus megaterium. The B. megaterium isolate was a halotolerant bacterium which grew well to an elevated NaCl concentration of 15% in TSBH medium. Besides, it had a consistent ability to produce >300 ppm of histamine at 10% NaCl concentration in TSBH medium after 72 h.     

Histamine contents of fermented fish products in Taiwan and isolation of histamine-forming bacteria

Twenty-seven imported fermented fish products from Southeast Asian countries and sold in the supermarkets in Taiwan, including fish sauce, fish paste and shrimp paste, were tested to determine the occurrence of histamine and histamine-forming bacteria. The levels of pH, salt content, total volatile basic nitrogen, trimethylamine, and aerobic plate count in all samples ranged from 4.8% to 6.5%, 16.2% to 45.3%, 51 to 275 mg/100 g, 5.4 to 53.9 mg/100 g and 1.0 to 4.2 log CFU/g, respectively. The average content for each of eight different biogenic amines in all samples was less than 90 ppm, except for histamine which has an average content of 394 ppm in fish sauce, 263 ppm in fish paste, and 382 ppm in shrimp paste. Most of the tested fermented fish products (92.6%) had histamine levels greater than the FDA guideline of 50 ppm, while seven of them (25.9%) contained >500 ppm of histamine. Although Bacillus coagulans and Bacillus megaterium were identified as the two histamine-producing bacteria capable of producing 13.7 and 8.1 ppm of histamine, respectively, in trypticase soy broth broth supplemented with 1.0% l-histidine, they were not determined to be the main contributors to histamine accumulation in these fermented fish products.     

Prediction of the amount and rate of histamine degradation by diamine oxidase (DAO)

Histamine is a biogenic amine that forms in a variety of foods and can cause food poisoning at high concentrations (>500 ppm). In situations where the formation of histamine in food cannot be prevented through refrigeration, diamine oxidase (DAO) enzyme may be used to degrade histamine to safe levels. The aims of this work were to apply DAO in model (buffer) and real (cooked tuna soup used in the manufacture of a fish paste product, Rihaakuru) systems, in order to obtain predictions for the rates and amounts of histamine degradation. The two systems were set up with a constant concentration of histamine (500 mg/L) and the DAO enzyme (2534 units/L) at a temperature of 37 °C, agitation at 100 rpm and an incubation time of 10 h with variable pH (5–7) and salt concentrations (1–5%). A total of 15 experiments were designed for each system using central composite design (CCD). The data from these experiments were fitted into regression models; initially the data were used to generate an exponential decline model and then the data from this were fitted into a secondary response surface model (RSM) to predict the rate and amount of histamine degradation by DAO. The model system results indicated that DAO activity was not significantly affected by salt (p > 0.05), and that activity reached a maximum within the pH range of 6–6.5 with an optimum at pH 6.3. However, the results obtained with the tuna soup model showed that the optimum oxidation of histamine using DAO occurred between pH 6–7 and salt 1–3%. This study defined the conditions for the use of DAO to degrade 500 mg/L of histamine in tuna soup used to manufacture Rihaakuru. The models generated could also be used to predict the rate and amount of histamine degradation in other foods that have similar characteristics to tuna soup.     

Biogenic amine survey and organoleptic changes in fresh, stored, and temperature-abused bluefish (Pomatomus saltatrix).

Changes in histamine, putrescine, and cadaverine concentrations in bluefish filets (Pomatomus saltatrix) stored at 5, 10, and 15 degrees C were determined using high-performance liquid chromatography. An organoleptic assessment was conducted simultaneously with the biogenic amine analyses. The histamine levels found in fresh bluefish obtained from wholesale seafood distributors ranged between <1 ppm and 99 with an average of 39 ppm. Putrescine and cadaverine were not found in fresh bluefish. Fish fillets stored at each of the three temperatures developed histamine. The greatest accumulation of histamine was observed in fish stored at 15 degrees C, which developed histamine levels as high as 2,200 ppm. Putrescine levels increased at each temperature during storage. Cadaverine was present only in uninoculated bluefish stored at 15 degrees C. Histamine achieved higher levels in bluefish pieces inoculated with Morganella morganii, which demonstrates that bluefish support bacterial histamine formation. Histamine levels at each temperature exceeded the 50-ppm advisory level established by the Food and Drug Administration before 100% sensory rejection. Standard plate counts increased during storage of fish at all temperatures, but the correlation between histamine levels and standard plate count was not significant.     

Biogenic amine formation and bacterial contribution in fish,squid and shellfish

Forty-one species of fish, squid and shellfish were analyzed for biogenic amine (BA) contents. Most of the fish samples showed lower BA contents, whereas some samples showed higher contents than the allowable levels. Shellfish and squid samples had negligible BA levels. Four fish species containing high BA levels were analyzed for changes in histamine contents during storage. In the most samples, the histamine contents remarkably increased up to 36.6–2123.9 mg/kg after 24 h of storage at 25 °C, while the contents began to gradually increase after 2–3 days of storage at 4–10 °C. The dominant microbial group was enterobacteria throughout the storage period. Meanwhile, out of total 119 strains isolated from different fish species showing high BA levels, 23 strains identified as Enterobacter aerogenes produced large amounts of histamine, putrescine and cadaverine, and 33 strains identified as two different Enterobacter spp. produced less histamine but large amounts of putrescine and cadaverine.     

Histamine formation by histamine-forming bacteria and yeast in mustard pickle products in Taiwan

The occurrence of histamine, histamine-forming bacteria and yeast were tested in 37 mustard pickle products sold in both retail markets and supermarkets in southern Taiwan. Aerobic plate count (APC), total coliform, and Escherichia coli were also tested for microbiological quality. Salt content, pH value, titratable acidity and sulphite content were determined for quality of mustard pickle products. Only one retail market sample and one supermarket sample had 8.9 and 7.4 mg histamine per 100 g products, although the average content for each of the nine biogenic amines was less than 2 mg/100 g. Ten histamine-forming bacterial strains and 6 histamine-producing yeast strains capable of producing 8.7 to 1260 ppm of histamine in trypticase soy broth (TSB) supplemented with 1% l-histidine (TSBH) were identified as Staphylococcus capitis (four strains),Staphylococcus pasteuri (two strains), Enterobacter cloacae (four strains), Candida glabrata (two strains) and Candida rugosa (four strains). S. capitis, which was previously reported to be halotolerant, was a potent histamine-former, capable of producing more than 1000 ppm of histamine in TSBH in the presence of 0.5–10% NaCl. The numbers of the aerobic plate count (APC) in all samples were below the Taiwanese regulatory level of 5 log CFU/g. None of the samples contained total coliform or E. coli. The values of pH, salt content, titratable acidity and sulphite content in all samples ranged from 3.8% to 5.0%, 2.0% to 10.0%, 0.21% to 1.18% and <2.0–1876 ppm, respectively.     

A note on the relationships between cured-cooked and dry-cured ham processing yields

Right and left hams from 353 pigs slaughtered at around 100 kg body weight were processed into cured-cooked hams and dry-cured hams, respectively. Weights and yields at various stages of each process, carcass lean content and fresh meat quality traits were registered. Technological yield of cured-cooked processing (saleable cooked ham weight/defatted–deboned fresh ham weight) was more closely correlated to ultimate pH (r = 0.51, p < 0.001) than to carcass leanness (r = −0.13, p < 0.05) whereas the reverse situation – r = 0.15 (p < 0.01) and r = −0.62 (p < 0.001), respectively – was found for technological yield of dry-cured processing (saleable dry ham weight/trimmed fresh ham weight). The correlation between the two technological yields was significantly positive but of fairly moderate magnitude (r = 0.36). The correlation between the overall yields (saleable processed ham weight/entire fresh ham weight) of the two processes revealed to be very close to zero (r = −0.01).     

Determination of histamine and histamine-forming bacteria in tuna dumpling implicated in a food-borne poisoning

An incident of food-borne poisoning causing illness in seven victims, due to ingestion of tuna dumpling, occurred in March 2006, in Chiayi Prefecture, southern Taiwan. The leftovers of the victims’ tuna dumpling and the five other tuna dumpling samples from five other retail stores were collected and tested to determine the occurrence of histamine and histamine-forming bacteria. The levels of pH, salt content, aerobic plate count (APC), total volatile basic nitrogen (TVBN), total coliform (TC) and Escherichia coli in all samples ranged from 6.08 to 6.43, 0.46% to 0.81%, 5.90 to 8.95 log CFU/g, 6.38 to 21.29 mg/100 g, 750 to 8000 most probable number (MPN)/g, and <3 to 1000 MPN/g, respectively. The suspected tuna dumpling contained 160.8 mg/100 g of histamine greater than the hazard action level of 50 mg/100 g set by the US Food and Drug Administration (FDA) for tuna fish. Given the allergy-like symptoms of the victims and the high histamine content in the suspected tuna dumpling, this food-borne poisoning was strongly suspected to be due to histamine intoxication. In addition, although thirteen histamine-producing bacteria strains capable of producing 8.1–19.7 ppm of histamine in trypticase soy broth (TSB) supplemented with 1.0% l-histidine (TSBH), were identified as Enterobacter sp. (three strains), Pantoea agglomerans (two strains), Klebsiella variicola (four strains) and Serratia marcescens (four strains), by 16S rDNA sequencing with PCR amplification, they were not determined to be the main contributors to histamine accumulation in suspected tuna dumpling.     

Chemical characterisation and histamine-forming bacteria in salted mullet roe products

Sixteen salted mullet roe products sold in the retail markets in Taiwan were purchased and tested to determine the occurrence of histamine and histamine-forming bacteria. The levels of pH, salt content, water content, total volatile basic nitrogen (TVBN) and aerobic plate count (APC) in all samples ranged from 5.4 to 5.8, 5.1% to 7.2%, 15.4% to 27.3%, 32.0 to 69.6 mg/100 g and <1.0 to 7.1 log CFU/g, respectively. None of these samples contained total coliform and Escherichia coli. The average content of each of the nine biogenic amines in all samples was less than 4 mg/100 g, and only one mullet roe sample had the histamine content (8.18 mg/100 g) greater than the 5.0 mg/100 g allowable limit suggested by the US Food and Drug Administration. Two histamine-producing bacterial strains capable of producing 10.7 ppm and 9.6 ppm of histamine in trypticase soy broth (TSB) supplemented with 1.0% l-histidine (TSBH) were identified as Staphylococcus carnosus by 16S rDNA sequencing with PCR amplification, and they were isolated from the sample with higher histamine content (8.18 mg/100 g).     

Oxygen-Sensor-Based Simple Assay of Histamine in Fish Using Purified Amine Oxidase

Oxygen consumption was measured by an oxygen sensor after addition of purified fungal amine oxidase to fish extracts. The oxidation of histamine to imidazoleacetaldehyde proceeded stoichiometrically. Based on the equimolar relationship between histamine and oxygen consumption, histamine was determined selectively by the oxygen sensor. Neither sample pretreatment removing interfering materials nor daily calibration by histamine standard was required. Histamine contents in scombroid fish were determined rapidly with good accuracy. AOAC and oxygen sensor methods showed a very high correlation (r=0.999, n=6).     

Fingerprint of enological tannins by multiple techniques approach

In this study, six enological tannins (T1–T6) from different source were analysed by selected analytical techniques, including UV–Vis, FTIR, NMR, SEC, LC-MS/MS and electronic nose. Moreover, the samples were tested for their antioxidant activity. Spectra and chromatograms are considered as a fingerprint with a pattern that is distinctive for a given type of sample. In particular, tannins showed the maximum UV–Vis absorbance (λ_max) at 273–280 nm, the inflection point (λ_min) at 247–264 nm, and the ratio λ_max/λ_min was 1.03, 1.35, 1.67, 1.36, 2.15, 2.02 for tannin T1–T6, respectively. The FTIR analysis of tannins showed the presence of most important absorption bands at 1522, 1453, 1380, 1350, 1328, 1286, 1210, 1148 and 1044 cm⁻¹. Besides, SEC and NMR analysis allowed to sort the tannins according to their average DP as follows: T1 > T2 = T4 > T3 > T6 = T5. Based on this multiple techniques approach three condensed tannins (sample T3, T5 and T6) were clearly identified, together with two mixtures of hydrolysable tannins (sample T1 and T2), whereas one sample (T4) showed a peculiar pattern different from the commercial tannin commonly used in enology.     

In vitro binding of bile acids by bananas,peaches, pineapple,grapes, pears,apricots and nectarines

The in vitro binding of bile acids by bananas (Musa × paradisiaca), peaches (Prunus persica), pineapple (Ananus comosus), grapes (Vitis spp.), pears (Pyrus communis), apricots (Prunus armeniaca) and nectarines (Prunus persica, nectarina) was determined using a mixture of bile acids secreted in human bile at a duodenal physiological pH of 6.3. Six treatments and two blank incubations were conducted, testing various fresh fruits on an equal dry matter basis. Considering cholestyramine (bile acid binding, cholesterol lowering drug) as 100% bound, the relative in vitro bile acid binding percentages on dry matter (DM), total dietary fibre and total polysaccharides basis were 2–9%, 15–101% and 10–101%, respectively. Bile acid binding, on a DM basis, for bananas was significantly (P ? 0.05) higher and that for nectarines significantly lower than those for peaches, pineapple, grapes, pears and apricots. The bile acid bindings for peaches and pineapple were similar and significantly higher than those for grapes, pears and apricots. Binding values for grapes and pears were significantly higher than apricots. These results point to the relative health promoting potential of bananas > peaches = pineapple > grapes = pears > apricots > nectarines, as indicated by their bile acid binding on a DM basis. The variability in bile acid binding between the fruits tested maybe related to their phytonutrients, antioxidants, polyphenols, flavonoids (anthocyanins, flavonols, and proanthocyanidins), structure, hydrophobicity of undigested fractions, anionic or cationic nature of the metabolites produced during digestion or their interaction with active binding sites. Animal studies are planned to validate in vitro bile acid binding of fruits, observed herein, to their healthful potential, atherosclerosis amelioration and cancer prevention.     

Inactivation of Escherichia coli O157:H7 and Salmonella spp. on alfalfa seeds by caprylic acid and monocaprylin

Alfalfa and other seed sprouts have been implicated in several Escherichia coli O157:H7 and Salmonella spp. human illness outbreaks in the U.S. Continuing food safety issues with alfalfa seeds necessitate the need for discovery and use of novel and effective antimicrobials. The potential use of caprylic acid (CA) and monocaprylin (MC) for reducing E. coli O157:H7 and Salmonella spp. populations on alfalfa seeds was evaluated. The effectiveness of three concentrations of CA and MC (25, 50, and 75 mM) to reduce E. coli O157:H7 and Salmonella spp. populations in 0.1% peptone water and on alfalfa seeds was evaluated. Surviving populations of E. coli O157:H7 and Salmonella spp. were enumerated by direct plating on tryptic soy agar (TSA). Non-inoculated alfalfa seeds were soaked for up to 120 min to evaluate the effect of CA and MC solutions on seed germination rate. For planktonic cells, the efficacy of the treatments was: 75 MC > 50 MC > 25 MC > 75 CA > 50 CA > 25 CA. Both E. coli O157:H7 and Salmonella spp. were reduced to below the detection limit (0.6 log CFU/ml) within 10 min of exposure to 75 MC from initial populations of 7.65 ± 0.10 log CFU/ml and 7.71 ± 0.11 log CFU/ml, respectively. Maximum reductions of 1.56 ± 0.25 and 2.56 ± 0.17 log CFU/g for E. coli O157:H7 and Salmonella spp., respectively, were achieved on inoculated alfalfa seeds (from initial populations of 4.74 ± 0.62 log CFU/g and 5.27 ± 0.20 log CFU/g, respectively) when treated with 75 MC for 90 min. Germination rates of CA or MC treated seeds ranged from 84% to 99%. The germination rates of CA or MC soaked seeds and water soaked seeds (control) were similar (P > 0.05) for soaking times of ≤ 90 min. Monocaprylin (75 mM) can be used to reduce E. coli O157:H7 and Salmonella spp. on alfalfa seeds without compromising seed viability.     

Novel starter cultures to inhibit biogenic amines accumulation during fish sauce fermentation

Bacteria with amine oxidase activity have become a particular interest to reduce biogenic amines concentration in food products such as meat and fish sausages. However, little information is available regarding the application of these bacteria in fish sauce. Hence, our study was aimed to investigate the effect of such starter cultures in reducing biogenic amines accumulation during fish sauce fermentation. Staphylococcus carnosus FS19 and Bacillus amyloliquefaciens FS05 isolated from fish sauce which possess amine oxidase activity were used as starter cultures in this study. Fermentation was held for 120 days at 35 °C. The pH value increased in all samples, while salt concentration remained constant throughout fermentation. Aerobic bacteria count was significantly lower (p < 0.05) in the control than in inoculated samples as a result of starter cultures addition. However, it decreased during fermentation due to the growth inhibition by high salt concentration. Proteolytic bacterial count decreased during fermentation with no significant difference (p > 0.05) among samples. These bacteria hydrolyzed protein in anchovy to produce free amino acid precursors for amines formation by decarboxylase bacteria. The presence of biogenic amines producing bacteria in this study was considered to be indigenous from raw material or contamination during fermentation, since our cultures were negative histamine producers. Amino acid histidine, arginine, lysine and tyrosine concentration decreased at different rates during fermentation as they were converted into their respective amines. In general, biogenic amines concentration namely histamine, putrescine, cadaverine and tyramine increased throughout fermentation. However, their concentrations were markedly higher (p < 0.05) in the control (without starter cultures) as compared to the samples treated with starter cultures. Histamine concentration was reduced by 27.7% and 15.4% by Staphylococcus carnosus FS19 and Bacillus amyloliquefaciens FS05, respectively. Both cultures could also reduce other amines during fermentation. After 120 days of fermentation, the overall biogenic amines concentration was 15.9% and 12.5% less in samples inoculated with Staphylococcus carnosus FS19 and Bacillus amyloliquefaciens FS05, respectively, as compared to control samples. These findings emphasized that application of starter cultures with amines oxidase activity in fish sauce fermentation was found to be effective in reducing biogenic amines accumulation.