2016 年 22 巻 1 号 p. 1-21
Histamine is the main causative agent of scombroid fish poisoning (SFP). To prevent SFP outbreaks, methods for controlling histamine accumulation have been studied. Histamine accumulation in raw fish, including frozen and semi-dried fish products, can be prevented by employing Hazard Analysis and Critical Control Point (HACCP) and Good Agricultural Practices (GAP). However, there is little information about histamine producers and the methods for preventing histamine accumulation in fermented fishery products in Japan. In 2011, CODEX adopted a limit of 400 ppm for histamine content in fish sauce, indicating that many manufacturers of fermented fish products are required to implement appropriate countermeasures. Presented here is recent information about histamine-producing bacteria in fermented fish products.
Histamine Histamine (C5H9N3) is a biogenic amine and one of the important chemical compounds that causes food-borne intoxication; large amounts of histamine can cause hypertension, hypotension, headache, urticaria, nausea, and vomiting, among other symptoms (EFSA, 2011; FAO/WHO, 2012; Hungerford, 2010; Ladero et al., 2010; Taylor, 1986). Chemically, the molecular mass of histamine is 111.14, the aqueous solution is odorless, colorless, and transparent, and histamine is a basic compound since it is a biogenic amine. Histamine is stable at elevated temperatures, as evidenced by its common use in cooking. Decarboxylation of l-histidine by certain bacteria yields histamine (Fig. 1). In addition, in the body, mast cells, basophils, and enterochromaffin-like (ECL) cells are known to produce histamine (FAO/WHO, 2012; Hungerford, 2010). Histamine exerts its pharmacological effects, such as vasodilation, by acting as a mediator of allergic reaction and inflammation. It is usually stored in vivo in these cells or in the granules of basophils, and its transient release into the extracellular space can be triggered by an external stimulus, such as the binding of an antigen to the cell surface antibody.
Conversion of L-histidine to histamine. Decarboxylation of L-histidine is caused by bacterial enzymes in the food.
Histamine accumulation occasionally occurs in food, Scombroid fish (EFSA, 2011; FAO/WHO, 2012), fermented fish products (Harada et al., 2008; Hernandez-Herrero et al., 1999; Mongkolthanaruk et al., 2012; Nakazato et al., 2002; Sato et al., 1995; Satomi et al., 1997, 2008, 2011, 2012; Stute et al., 2002; Yatsunami and Echigo, 1991, 1993), wine (Coton et al., 1998, 2005, 2010; Landete, et al., 2005; Lonvaud-Funel and Joyeux, 1994; Lonvaud-Funel, 2001; Lucas et al., 2005, 2008), cheese (Burdychova et al., 2007; Chang and Snell, 1968; Chang et al., 1985; Joosten and Northlt, 1989), meat products (Landeta et al., 2007; Silla Santos, 1998; Suzzi and Gardini, 2003), and others (Calles-Enríquez et al., 2010; Hamaya et al., 2014; Ibe et al., 2003; Le Jeune et al., 1995; Tsai et al., 2007), and is caused by certain histamine-producing bacteria. Furthermore, a number of helpful reviews have discussed histamine in food (Halasz et al., 1994; Silla-Santos, 1996; Spano et al., 2010). Thus, regulatory limits on histamine have been set in individual countries or by international organizations such as CODEX (Codex Alimentarius Commission)i) (Table 1). Histamine has occasionally been regarded as one of the measures of decomposition, although the correlation between histamine content and decomposition odors is often inconsistent, indicating that histamine content is not suitable for evaluating the freshness of fish products (FAO/WHO, 2012).
| Regulatory bodies (country) | Products | Prescribed limits |
|---|---|---|
| CODEX | Fish and fish products | Histamine levels as indicators of decomposition, and hygiene and handling. A maximum average level of not more than 100 ppm is considered satisfactory in relation to decomposition, while an upper limit of 200 ppm in any one sample is applied for hygiene and handling. |
| Fish sauce | The product shall not contain more than 400 ppm of histamine in any sample unit tested. | |
| FSANZ (Australia and New Zealand) | Fish and fish products | Maximum limit of 200 ppm for histamine |
| FDA (USA) | Tuna and related fish | Guidelines for tuna and related fish establishing a ‘defect action level’ of 50 ppm in any sample (It is said to be indicative of spoilage and may mean that toxic levels are present in other samples). A separate toxicity level of 500 ppm is also given. |
| EU (European Union) | Fish and fish products | Fish species belonging to families known to contain large amounts of histidine (e.g. Scombridae, Clupeidae etc) in their tissues should be tested for the presence of histamine. Nine samples should be tested from each lot and the mean value should be 100 ppm or less. The lot is considered unsatisfactory if more than two samples give results of between 100 and 200 ppm, or if any sample gives a result of 200 ppm or more. |
| Fermented products | Maximum limit of 200 ppm for histamine |
The no observed adverse effect level (NOAEL), the appropriate hazard level for healthy individuals as determined by the Joint FAO/WHO Expert Meeting on the Public Health Risks of Histamine and Other Biogenic Amines from Fish and Fishery Products (2012), was a dose of 50 mg of histamine. Moreover, using the available consumption data combined with expert opinions, the meeting agreed that a serving size of 250 g captured the maximum amount eaten in most countries at a single eating event. Based on the hazard level of 50 mg of histamine and a serving size of 250 g, the maximum concentration of histamine in that serving was consequently calculated to be 200 mg/kg. Most of the histamine that is ingested is rapidly eliminated in the urine. After the remaining histamine in the body is degraded, it is released as carbon dioxide. Therefore, histamine is completely eliminated within a few hours of ingestion and is not accumulated in the body.
Ingestion of foods containing large amounts of histamine can sometimes trigger food poisoning called scombrotoxin fish poisoning (SFP) or allergy-like food poisoning. Foods containing a large amount of l-histidine, a precursor of histamine, require strict quality control. SFP is clearly different from a food allergy that patients develop as an immune reaction after consuming specific causal foods. Individuals do not develop SFP unless they have ingested more than the allowed amount of histamine. In other words, food poisoning does not occur in the absence of a causal agent, such as a large amount of histamine in fish and shellfish (FAO/WHO, 2012). Every year, approximately 20 incidents of SFP occur and about 500 people in Japan are affected. None of the cases of SFP were reported to result in death (Toda et al., 2009).
Since histidine in fish meat is converted to histamine, fish species containing large amounts of free histidine in the muscle are at high risk of histamine accumulation, specifically if the histidine is not released via processes such as muscle protein degradation by fermentation or enzyme preparation. Some fish species, including scombroid fish such as mackerel, sardine, saury, marlin, and tuna are known to be the major cause of SFP in Japan. Outside of Japan, marlin, tuna, and migratory fish, such as dolphin, have been recognized as causal fish species of histamine poisoning. The Joint FAO/WHO Expert Meeting on the Public Health Risks of Histamine and Other Biogenic Amines from Fish and Fishery Products (2012) created a list of the fish species containing high levels of histidine, based on previous reports describing free histidine levels in each fish species, as possible causative species of SFP. Fish species that ranked high on the list are estimated to easily accumulate histamine. Even if fish have low histidine levels, poor handling may cause histamine to accumulate.
Other biogenic amines Although the primary causative agent of SFP is histamine, other biogenic amines derived from amino acids by microbial metabolism can also cause allergy-like symptoms cooperatively with histamine. Tyramine, cadaverine, putrescine, and others sometimes act as antagonists of histamine-metabolizing enzymes, such as diamine oxidase and histamine-N-methyltransferase, though small amounts of these amines cannot affect the body (FAO/WHO, 2012; Hungerford, 2010; Ladero et al., 2010; Taylor, 1986). Therefore, these amines in foods have also been studied. Since tyramine is sometimes accumulated in fermented foods including fish sauce at significant levels (Stute et al., 2002), it is necessary to focus on tyramine-producing bacteria (Coton et al., 2005, 2009; Lucas et al., 2003). Figure 2 shows the successful isolation of tyramine-producing bacteria from fish sauce using a tyrosine agar plate (Satomi et al., 2014). Cells producing tyramine had a ‘halo’ around their colonies because of the conversion of insoluble tyrosine to soluble tyramine. A number of reviews describing biogenic amines are available (FAO/WHO, 2012; Hungerford, 2010; Ladero et al., 2010; Taylor, 1986).
Image of colonies formed by Tetragenococcus halophilus strains, including tyramine-producing strains, on a tyrosine agar plate. Tyramine-producing cells had a ‘halo’ around their colonies because of the conversion of insoluble tyrosine to soluble tyramine.
Histamine-producing bacteria As mentioned above, histamine production is the result of bacterial decarboxylation of l-histidine, and histamine accumulation occurs in many foods other than fish products. Interestingly, the bacterial species that produce histamine are somewhat specific to each food.
In the case of raw fish, such as Scombroid fish, mainly gram-negative bacteria, Enterobacteriaceae, and Photobacterium spp., are known to produce histamine (EFSA, 2011; Emborg et al., 2006; FAO/WHO, 2012; Kanki et al., 2007; Lehane, 2000, Okuzumi et al., 1981; Sato et al., 1994; Takahashi et al., 2003; Torido et al., 2012b). Histamine-producing Enterobacteriaceae include common enteric mesophiles, such as Morganella, Enterobacter, Hafnia, Raoultella, and others. Morganella morganii is particularly well known as a significant histamine producer, owing to its rapid growth and the potent activity of its histamine-producing enzyme. The genus Photobacterium (family Vibrionaceae) encompasses marine bacteria, indicating that they require sodium ions for growth. In addition, it is necessary to pay attention to psychrophilic and psychrotolerant species in chilled fishery products, which are represented by Morganella psychrotropicus, P. phosphoreum, and P. kishitanii (Emborg et al., 2006; Kanki et al., 2004; Torido et al., 2012a, 2012b). According to recent studies (Torido et al., 2012a, 2012b), the psychrophiles can grow and produce histamine at temperatures less than 10°C within 3 days. However, it typically takes 6 days or more to grow and produce histamine at less than 5°C, even in these bacteria, indicating that the risk from histamine is best mitigated by applying basic GAP and, where feasible, a HACCP system (EFSA, 2011; FAO/WHO, 2012).
Gram-positive histamine-producing bacteria were isolated from fermented or aged foods, such as wine, cheese, meat products, fish sauce, and others (references are quoted above). Non-halophilic lactic acid bacteria (LAB), such as Oenococcus oeni, Lactobacillus 30a, Lb. hilgardii, Lb. buchneri, and others, or staphylococci were the major sources of histamine accumulation in dairy products. Additionally, histamine accumulation occasionally occurred in fishery products, such as fish sauce (Fujii et al., 2008; Sato et al., 1995; Satomi et al., 1997, 2008, 2011, 2012; Stute et al., 2002), salted fish (Hernandez-Herrero et al., 1999; Mongkolthanaruk et al., 2012; Yatsunami and Echigo, 1991, 1993), and others (Harada et al., 2008; Hamaya et al., 2014), and certain bacterial strains, belonging to Tetragenococcus, Staphylococcus, and others, were identified as the histamine producers. Generally, tetragenococci are causative agents of histamine accumulation in fermented or aged foods containing large amounts of salt (approximately 15% or more) (Kimura et al., 2001; Satomi et al., 2012). The materials of fermented foods sometimes have high salinity, low pH, or contain alcohol, generally making them an unsuitable environment for bacterial growth. Therefore, histamine producers take longer to produce histamine during fermentation. However, since there is a long fermentation period in the processing of fermented foods, the risk of histamine accumulation may be high in such foods.
Histamine-producing enzymes Histamine-producing bacteria have histidine decarboxylase (HDC: EC.4.1.1.22), which catalyzes the conversion of l-histidine to histamine. Histamine production in bacteria is one of the stress responses to low pH in the bacterial environment (Molenaar et al., 1993). Two HDC isozymes are known as bacterial enzymes: pyridoxal phosphate (PLP)-dependent HDC in gram-negative bacteria and pyruvoyl-type HDC in gram-positive bacteria (Landete et al., 2008). In the many PLP-dependent HDCs, the enzymes are constructed as a 170 kDa homotetramer, which is widely distributed in Enterobacteriaceae, Photobacterium spp, mammal liver, and others (Fujii et al., 1994; Kamath et al., 1991; Kanki et al., 2007; Morii et al., 2003; Tanase et al., 1985). Pyruvoyl-type HDCs are generally constructed as a hexamer (αβ)6, containing alpha and beta subunits, about 200 kDa in the active form (Huynh and Snell, 1985; Recsei and Snell, 1984; van Poelje and Snell, 1990), and the HDC gene is coded on hdcA. A summary of the chemical characteristic of each enzyme in histamine-producing gram-positive bacteria is shown in Table 2. The polypeptide chain translated from hdcA is cleaved to form the alpha and beta subunits, and then the N-terminal residue in the alpha subunit is converted to a pyruvoyl group via pyruvate in the process of maturation. The enzyme is thought to be specifically distributed to gram-positive bacteria, mainly LAB (Calles-Enriquez et al., 2010; Coton et al., 1998; Gallagher et al., 1993; Konagaya et al., 2002; Lucas et al., 2005; Martín et al., 2005; Recsei et al., 1983; Recsei and Snell, 1984; Vanderslice et al., 1986), staphylococci (de las Rivas et al., 2008; Furutani et al., 2014), clostridia (Recsei et al., 1983; Shimizu et al., 2002), and micrococci. (Alekseeva et al., 1976, 1986; Prozorovski and Jörnvall, 1974; van Poelje and Snell, 1990).
| Bacterium | Molecular mass (kDa) | Subunit | Optimum | Vmax | Km | ProHDC (π chain) | References | ||
|---|---|---|---|---|---|---|---|---|---|
| Native | α | β | structure | pH | (µmol·min−1·mg−1) | (mmol·L−1) | detected | ||
| Staphylococcus epidermidis TYH1 | 121 | 26 | 9 | (αβ)3(6?) | 5.0 – 6.0 | 45.5 | 1.1 | + | Furutani et al. (2013) |
| Tetragenococcus muriaticus | 257 | 29 | 13 | (αβ)6 | 4.5 – 7.0 | 16.8 | 0.74 | + | Konagaya et al. (2004) |
| Oenococcus oeni | 190 | 28 | 11 | (αβ)6 | 4.8 | 17.8 | 0.33 | + | Coton et al. (1998) |
| Lactobacillus buchneri | 203 | 25 | 9 | (αβ)6 | 5.5 | 69 | 0.6 | + | Recsei et al. (1983) |
| van Poelje and Snell (1990) | |||||||||
| Lactobacillus30a | 208 | 25 | 9 | (αβ)6 | 4.8 | 80 | 0.4 | + | Vanderslice et al. (1986) |
| van Poelje and Snell (1990) | |||||||||
| Clostridium perfringens | 213 | 25 | 10.5 | (αβ)6 | 4.5 | 25 | 0.2 | + | Recsei et al. (1983) |
| Micrococcus sp. | 100 | 29 | 8 | (αβ)3(6?) | 4.4 – 5.8 | 25 | 0.8 | + | Alekseeva et al. (1976, 1986) |
Gene analysis of pyruvoyl-type HDC Gram-positive histamine-producing bacteria are typified by the presence of pyruvoyl-type HDC, which is encoded by the structural gene hdcA. This gene, along with hdcP, hdcB, and hdcRS, comprise the hdc cluster (Fig. 3). The region related to histamine production is conserved in several LAB. The putative functions of the other genes are as follows: hdcP encodes a histidine/histamine antiporter (Lucas et al., 2005; Martín et al., 2005), hdcB encodes a cleavage factor of immature HDC translated from hdcA (Trip et al., 2011), and hdcRS encodes histidyl-RNA synthetase (Lucas et al., 2005; Martín et al., 2005). Although some genes are deleted in certain species, hdcA and hdcP are likely essential genes for producing HDC (Calles-Enríquez et al., 2010; de las Rivas et al., 2008; Yokoi et al., 2011).
Genetic organization of bacterial histidine decarboxylase loci. The dashed box indicates a set of genes >99% identical. nt, nucleotides; aa, amino acids.
As previously mentioned, gram-positive histamine producers have a single copy of hdcA as the gene encoding the histamine-producing enzyme. The processing of HDC has been studied. In the case of Lactobacillus 30a, HDC was shown to be synthesized in an inactive form of about 310 amino acids (π chain); this form undergoes autoserinolysis, yielding an α chain of about 230 residues, which links to the pyruvoyl group at the N terminus and a β chain of about 80 residues. These subunits presumably associate into an active hexameric (αβ)6 complex (Recsei and Snell, 1984; Gallagher, et al., 1993). Although the numbers of amino acid residues in HDC are different among bacterial species, the cleavage site of α and β chains, the substrate binding and catalytic sites, and the mechanisms of translation and maturation of HDC are conserved. Phylogenetic relationships among hdcA are shown in Fig. 4. The sequence diversity of hdcA is increased among gram-positive bacteria, correspondingly to the taxonomic distance of each bacteria. Thus, staphylococcal and clostridial hdcA are far from the positions occupied by that of LAB. In the case of tetragenococcal hdcA, which is closely involved in histamine accumulation in fermented fishery products, the HDC is composed of 316 amino acids and shares more than 99% sequence similarity with Lb. hilgardii and O. oeni HDCs. Moreover, spacer regions connecting each open reading frame (ORF) in the hdc cluster are also shared, indicating that hdc clusters encoded in tetragenococci, Lb. hilgardii, and O. oeni, transfer as one gene block.
Genetic relationships among the hdcA of different bacteria. Phylogenetic trees were constructed based on the nucleotide sequences of hdcA using the neighbor-joining method of Saitou and Nei (1987). The bar indicates genetic distance (Knuc). The nucleotide accession numbers are as follows: Vibrio anguillarum (AY312585), Morganella morganii (J02577), Klebsiella planticola (M62746), Enterobacter aerogenes (M62745), Lactobacillus sakei (AY800122), Lb. hilgardii (AY651779), Oenococcus oeni (U58865), Tetragenococcus muriaticus (AB040487), T. halophilus (AB076394), strain H (AB362339), Lb. buchneri (AJ749838), Lb. reuteri (CP000705), Lactobacillus 30a (J02613), Clostridium perfringens (NC_003366). Abbreviations are as follows: nt, nucleotide; aa, amino acid.
Fermented fishery products in Japan Japanese fermented seafoods are divided into three major types based on the fermentation procedure: salted, pickled, and molded seafoods (Table 3). The salted fermented foods include fish sauce and salted fish, which are typically made from fish and salt (> 10%). Fermentation of these foods involves the digestion of fish materials by self-digestion and microorganisms (Fujii et al., 2008; Fujii and Sakai, 1984; Fukui et al., 2012; Furutani et al., 2012; Ito et al., 1985a, 1985b; Shozen et al., 2012a; Taira et al., 2007; Yamashita et al., 1991), predominantly gram-positive cocci, including halophilic LAB such as Tetragenococcus spp. The organoleptic features of these foods are a salty taste, preferable with a fishy odor and strong umami quality. Moreover, the features of these fish sauces are similar to the products of other countries (Park et al., 2001, 2002). Pickled fermented foods are made from raw fish materials combined with carbohydrates such as cooked rice or bran. A traditional fermentation barrel used in processing rice bran pickles is shown in Fig. 5. Representative products include funa-zushi (pickled crucian carp with steamed rice), izushi (pickled salmon with steamed rice), and rice bran pickles. In the case of low-salt content products such as funa-zushi, izushi, and so on, they are typically processed by fermentation with LAB. Non-halophilic LAB, such as Lactobacillus spp., Streptococcus, and others, are predominantly used (Fujii, 1992; Fujii et al., 2011). The products are distinctly characterized by their sour taste and flavor, occasionally resembling cheese. The shelf-life and quality stability period are shorter than those of salted products owing to the low salt concentration (except for pickled seafood with rice bran). Since the salt concentration is higher in pickled seafood with rice bran (around 10% salt) than the former products, the transition of microbial flora and chemical compounds is similar to that of fish sauce fermentation (Kobayashi et al., 1995, 2000; Kosaka et al., 2010, 2012; Kosaka and Ooizumi, 2012). Furthermore, boiled, smoke-dried, and molded fish (fushi), referred to as katsuobushi, are made from dried smoked bonito and mold starter (Fujii, 1992). The mold starter functions in the digestion of lipids and dehydration. The unique aromatic flavor, derived from the phenolic compounds produced during the smoking step, is also a distinctive characteristic of katsuobushi. Soup stock extracted from sliced katsuobushi, called dashi, is the essence of Japanese cuisine.
| Fermentation type | Name of products | Raw materials | Salt concentration | Fermentation period | Risk of histamine during fermentation | Contribution during fermentation | ||
|---|---|---|---|---|---|---|---|---|
| General name | Japanese product name | Self-digestion | Microorganisms | |||||
| Salted type | Fish sauce | Gyosho | Squid, sand fish, sardine, etc. | >20% | >1 year | + | ◎ | △ |
| Gyosho (using mold starter) | approx. 20% | >6 months | + | ◎ | △ | |||
| Shiokara (Salted fish) | Ika-shiokara | Squid | approx. 10% | 2 – 3 weeks | w | ◎ | △ | |
| Shutou | Organ of skipjack tuna | >20% | 1 – 2 months | w | ◎ | △ | ||
| Konowata | Intestine of sea cucumber | <5% | 2 – 3 days | w | ◎ | △ | ||
| Kusaya | Kusaya | Amberstripe scad, flying fish, etc. | 3 – 8% | >100 years | – | ╳ | ○ | |
| Pickled type | Nare-zushi (fermented sushi) | Funazushi | Crucian carp | 2 – 4% | 2 years | w | △ | ◎ |
| Izushi | Salmon, etc. | <4% | 1 – 2 months | w | △ | ◎ | ||
| Nuka-zuke (pickled seafood with rice bran) | Heshiko | Mackerel | >7% | >7 months | + | ○ | ○ | |
| Fugunoko-nukazuke | Puffer fish ovaries | 10 – 15% | 2 years | + | ○ | ◎ | ||
| Molded type | Fushi (boiled, smoke-dried, and molded fish) | Katsuobushi | Skipjack tuna | <1% | > 3 months | – | ╳ | △ |
| Sababushi | Mackerel and its relatives | <1% | > 3 months | – | ╳ | △ | ||
w; weak
Image of a wooden barrel used for making rice bran pickle.
In the fermented products described above, histamine accumulation was observed mainly in the salted-type fermented products, though rice bran pickles (nuka-zuke) occasionally undergo histamine accumulation. Both products have a high salt concentration (>7% in nukazuke, and >15% in fish sauce), indicating that the predominant microorganisms are halophilic LAB. A summary of the dominant bacteria and histamine-producing bacteria in Japanese fermented fishery products is given in Table 4.
| Category of products | Name of product | Substrate | Nature and use | Major microorganisms | Histamine-producing bacteria | References |
|---|---|---|---|---|---|---|
| Fish sauce | Fish sauce | Fish or shellfish, salt, bacterial starter (if applicable) | Salty, fish odor, umami, condiment | Tetragenococcus halophilus, T. muriaticus, Staphylococcus spp., Chromohalobacter spp. | T. halophilus, T. muriaticus | Fujii (1992) Satomi et al. (1997, 2012) Fukui et al. (2012) |
| Fish sauce (using koji starter) | Fish or shellfish, salt, koji: mould starter, bacterial starter (if applicable) | Salty, fish odor, umami, similar to soy sauce taste, condiment | T. halophilus, Staphylococcus spp., Chromohalobacter spp., Aspergillus sp. | T. halophilus | Satomi et al. (2008, 2011) | |
| Pickled seafood with rice bran | Heshiko, Konka-iwashi, etc. | Mackerel, sardine, herring, etc. | Salty, fish and rice bran odor, sour, side dish, snack for drinking | T. halophilus, T. muriaticus, Staphylococcus spp., Chromohalobacter spp., Haloanaerobium spp. | T. halophilus, T. muriaticus, Staphylococcus spp. | Yatsunami and Echigo (1991) Sato et al. (1995) Kosaka et al. (2011) Satomi et al. (2012) Kobayashi et al. (1995, 2000) |
| Fugunoko-nukazuke (puffer fish ovaries fermented with rice bran) | Puffer fish ovaries | |||||
| Pickled seafood with steamed rice | Nare-zushi, Izushi, Funazushi, etc. | Crucian carp, mackerel, salmon, sand fish, etc. | Picked fish, sour, sometimes cheese-like flavor, side dish and snack for drinking | Lactobacillus plantarum, Lb. alimentarius, Lb. coryniformis, Lb. sakei, Lb. sanfrancisco, Lb. kefir, Lb. fermentum, Lactococcus lactis, Pediococcus sp., Leuconostoc spp. | Not determined at present | Fujii (1992) Fujii et al. (2011) |
| Katsuobushi (boiled, smoke-dried and molded skipjack tuna) | Katsuobushi, Fushi | Boiled, smoke-dried and molded skipjack tuna (mackerel and its relatives are applicable) | Umami, aromatic (phenolic compounds), sliced before using soup stock (dashi) | Aspergillus spp., A. glaucus | Not determined at present | Fujii (1992) |
Histamine-producing bacteria
Tetragenococcus The primary causative agents of histamine accumulation in Japanese fermented fishery products are Tetragenococcus spp. These bacteria are gram-positive tetrad cocci (Fig. 6), halophilic, and lactic acid producers (Collins et al., 1990; Dicks et al., 2009; Garvie, 1986; Satomi et al., 1997; Weiss, 1992). While these bacteria are not typically histamine producers, certain Tetragenococcus strains cause histamine accumulation in fermented fishery products (Kimura et al., 2001; Kobayashi et al., 2000; Konagaya et al., 2002; Mongkolthanaruk et al., 2012; Sato et al., 1995; Satomi et al., 1997, 2008, 2011, 2012; Udomsil et al., 2010). Their physiological and biochemical properties correspond closely with those of typical non-halophilic LAB, except for being halophiles, meaning they are facultative anaerobic bacteria without a respiratory pathway. The genus encompasses five species, and two of those species, T. halophilus and T. muriaticus, are commonly isolated from Japanese fermented fishery foods, such as fish sauce, pickled seafood with rice bran, and others. Both species proliferate in foods with salt concentrations >20%. T. halophilus is commonly isolated from salted-type fermented products including soy sauce, miso paste, and fishery products; moreover, the species is widely distributed in high osmotic pressure environments, such as salted foods, sugar thick juice, and others (Fujii et al., 2008; Fukui et al., 2012; Ito et al., 1985a, 1985b; Justé et al, 2008, 2012; Kobayashi et al., 1995, 2000; Kosaka et al., 2012; Kuda et al., 2001; Röling and van Verseveld, 1996; Sato et al., 1995; Taira et al., 2007). T. halophilus strains are predominant bacteria during the processing of fermented fishery products and soy sauce (Ito et al., 1985a, 1985b; Fukui et al., 2012; Röling et al., 1996; Udomsil et al., 2010). They play an important role in lactic acid production, which reduces the pH of products during fermentation. Therefore, T. halophilus is also used as a fermentation starter in these foods (Kimura et al., 2015; Shozen et al., 2012a; Udomsil et al., 2010; Zaman et al., 2011). This bacterium has an optimal growth temperature of around 30°C; it does not grow at temperatures higher than 40°C and hardly grows below 10°C. Despite being LAB, they prefer an alkaline growth environment, and their growth is significantly inhibited at pH 5.0 or less. T. muriaticus has also been isolated from fermented fishery products (Satomi et al., 1997, 2012), and was originally isolated from squid liver sauce (Satomi et al., 1997); however, there is little information about its distribution.
Scanning electron microscope image of Tetragenococcus halophilus IAM 1676T.
Genetic information regarding tetragenococcal HDC has been recently reported (Mongkolthanaruk et al., 2012; Satomi et al, 2008, 2011, 2012), though some points remain to be clarified. At present, it is likely that histamine-producing tetragenococci, including T. halophilus and T. muriaticus, harbor a hdcA-encoding plasmid. Generally, bacterial amino acid decarboxylase gene systems are chromosomally located, excepting an aspartate decarboxylase operon detected on a 25-kb plasmid of T. halophila (Abe et al., 2002) and hdc cluster on an 80-kb plasmid of Lb. hilgardii (Lucas et al., 2005). From the sequencing results of the representative plasmids encoding hdcA, several pieces of information have been clarified. First, the strains harbor an approximately 21 – 37 kb plasmid encoding a single copy of the hdc cluster (Table 5), which consists of four genes, hdcP, hdcA, hdcB, and hdcRS (Fig. 3). Second, the nucleotide sequence of the hdc cluster shares >99% sequence similarity with that of LAB (Fig. 3), including non-halophilic LAB that are phylogenetically distinct from Tetragenococcus spp. (Fig. 4). Finally, the structure of most tetragenococcal plasmids are identical, and two putative mobile genetic elements, ISLP1 (Nicoloff and Bringel, 2003)-like and IS200 (Beuzón et al., 2004)-like, were identified in the up- and down-stream regions of the hdc cluster (Fig. 7). Additionally, the order of genes within the hdc cluster was identical to the sequences of LAB reported as histamine producers. The sequence types of the putative plasmid replication processes are divided into two groups: theta-type replication and rolling circle replication (RCR), though almost all parts related to the hdc cluster including the putative transposons are conserved in both types of plasmids. The theta-type plasmid encoded the putative oriT and mob gene clusters related to plasmid transfer, but a putative oriT region and its related genes were not determined in rolling circle replication. Since theta-type plasmids have been previously isolated from LAB, such as Pediococcus pentosaceus (Giacomini et al., 2000; Kantor et al., 1997), T. halophilus (Benachour et al., 1997), and others, it should be distributed in LAB including tetragenococcal species. However, at present, it is not clear if theta-type plasmids can transfer to other hosts by using the mob gene. The RCR related region in both tetragenococcal species showed 100% sequence similarity to that of Staphylococcus haemolyticus plasmid pLNU4 (Lüthje et al., 2007), indicating that the replication region of the plasmids isolated from T. halophilus and T. muriaticus are the same. Therefore, the RCR plasmids encoding hdcA should be able to replicate in both species. Remarkably, there are putative insertion sequences (ISs) positioned in adjacent regions to the hdc cluster. Two putative ISs were found to be conserved sequences, although many transposase genes, including truncated genes, were identified around the hdc cluster from sequenced plasmids. They shared high sequence similarity with known ISs, including ISLP1-like and putative truncated IS200-like. Moreover, ISLP1-like sequences are determined from the upper region of the tyrosine decarboxylase gene encoded on tetragenococcal plasmids (Satomi et al., 2012). Although IS1216V (Heatona et al., 1996) was also found upstream of ISLP1-like on most plasmids, it was not identified on some plasmids. Therefore, the hdc cluster and other genes encoding biogenic amine-producing enzymes are transferred by this IS. However, it is difficult to conclude that the spread of the hdc cluster among LAB is caused by the transposon. In fact, many ISs and transposase-like ORFs, which can explain simple gene transfer, were present in the plasmids; however, most of the transposase-like sequences and IS elements were truncated or incomplete. Further study is needed to elucidate the mechanisms of hdc transfer among Tetragenococcus strains.
| Strainsa | Plasmid designation | Sourceb | Identificationc | RFLPd typing | Amine-producing gene | Length (bp)e | Putative replication type (closest sequence)f | oriT | Tng | Reference | Accession number | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| I-1 | pHDC-I-1 | Flyingfish cured in rice bran | Tc. muriaticus | I | hdc | 21,359* | RCR | St. haemolyticus pLNU4H | − | + | Satomi et al., 2012 | AB710473 |
| I-8 | pHDC-I-8 | Mackerel fish sauce A | Tc. muriaticus | I | hdc | 21,514* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2012 | - |
| I-9 | pHDC-I-9 | Mackerel fish sauce A | Tc. muriaticus | II | hdc | 23,196* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2012 | AB710474 |
| I-13 | pHDC-I-13 | Mackerel fish sauce B | Tc. muriaticus | I | hdc | 21,514* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2012 | - |
| H | pHDC-H | fish sauce A | Tc. halophilus | III | hdc | 29,924* | theta | Ped. pentosaceus pMD136I | + | + | Satomi et al., 2008 | AB362339 |
| A | pHDC-A | fish sauce B | Tc. halophilus | IV | hdc | 32,501* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2011 | AB588176 |
| HO | pHDC-HO | fish sauce C | Tc. halophilus | V | hdc | 36,638* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2011 | AB588177 |
| RI | pHDC-RI | fish sauce D | Tc. halophilus | VI | hdc | 21,231* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2011 | AB588180 |
| Tyr A | pTDC-A | fish sauce E | Tc. halophilus | - | tdc | 27,115* | RCR | St. haemolyticus pLNU4 | - | + | Satomi et al., 2014 | AB914479 |
| Tyr B | pTDC-B | fish sauce E | Tc. halophilus | - | tdc | 28,689* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2014 | AB914743 |
| JCM10006T | pHDC-Tm | fish sauce F | Tc. muriaticus | I | hdc | 21,516* | RCR | St. haemolyticus pLNU4 | − | + | Satomi et al., 2011 | AB588178 |
Physical and genetic maps of plasmids encoding hdc in Tetragenococcus strains. A box indicates putative function predicted from sequence similarity. Arrows indicate transcriptional direction of gene clusters. The highest degree of similarity of each ORF is shown in Table 3. The regions encoding ISLP1, HDC, and IS200 were conserved among all plasmids with 99% sequence similarity
Staphylococcus The genus Staphylococcus is comprised of gram-positive cocci, facultative anaerobes with a respiratory pathway, and encompasses more than 40 species, including S. aureus, a well known causative agent of food-poisoning and occasional clinical bacterium (Kloos and Shileifer, 1986; Götz et al., 2006). In addition, staphylococci have been frequently reported as histamine-producers in salted fishes (Hernandez-Herrero et al., 1999; Yatsunami and Echigo, 1991, 1993), fish paste (Harada et al., 2008), and fermented meat (Silla Santos, 1998; Suzzi and Gardini, 2003; Landeta et al., 2007) and soy bean (Tsai et al., 2007) products. There are only two histamine-producing staphylococcal species reliably identified, S. capitis (de las Rivas et al., 2008) and S. epidermidis (Yokoi et al., 2011). The remainder of the histamine-producing staphylococci remains unidentified. Therefore, there is little information about the distribution and significance of staphylococci as histamine producers in fishery products. Although histamine-producing staphylococci are problematic in dairy or meat products, they cannot be ignored in salted fish products because of their tolerance to high salt concentrations. Staphylococci are infrequently isolated from fermented fish products with a long-term fermentation period. Since most fermented fishery products, like fish sauce, contain >15% salt concentration, staphylococci are rarely isolated as dominant bacteria over the entire fermentation period in these foods. According to previous reports (Kloos and Shileifer, 1986; Götz et al., 2006), staphylococci typically grow at <15% salt concentration, indicating that these bacteria are halotolerant. Actually, staphylococci are commonly isolated from foods containing <10% salt, such as salted dry fish, dry-cured sausage, and others (Hernandez-Herrero et al., 1999; Landeta et al., 2007; Silla Santos, 1998; Suzzi and Gardini, 2003; Yatsunami and Echigo, 1991, 1993). In a study of the changes in bacterial flora during fish sauce fermentation, it was reported that staphylococci, such as S. xylosus and its relatives, are the dominant bacteria only during the initial stage of fermentation, thereafter declining (Fig. 8; Fukui et al., 2012), indicating that a high salt concentration is not optimal for its growth. The bacterial flora of the fish sauce mash was initially dominated by Staphylococcus spp., and then the predominant species changed over time to Tetragenococcus.
Changes in microflora during fish sauce fermentation. Phylogenetic classifications were carried out using the clone library method. Bacterial 16S rRNA genes were amplified using PCR with DNA directly extracted from fish sauce mash. Staphylococcus spp., 
: Tetragenococcus halophilus, 
: Lentibacillus-like, 
.
Staphylococcal HDCs are also pyruvoyl-type HDCs, as in other gram-positive histamine producers. As is shown in the phylogenetic tree based on HDC amino acid sequences, the phylogenetic position of its HDC is far from that of LAB but rather close to clostridial HDC (Furutani et al., 2014; Yokoi et al., 2011; de las Rivas et al., 2008). Moreover, since hdcB and hdcRS are lacking in the hdc cluster (de las Rivas et al., 2008; Yokoi et al., 2011), the genetic characteristics are similar to those of Clostridium. Previous studies strongly indicate that the HDC gene (hdcA) is encoded in genomic DNA and is located in mobile genetic elements (de las Rivas et al., 2008; Yokoi et al., 2011). In the case of S. epidermidis TYH1 isolated from fish-miso, the hdc cluster resides in the staphylococcal cassette chromosome elements (Yokoi et al., 2011), which was discovered as a mobile genetic element and is composed of the mec gene complex, encoding methicillin resistance (SCCmec) (Ito et al., 1999). Since the SCC gene complex encodes recombinases of the invertase/resolvase family (Ito et al., 2003; International Working Group on the Classification of Staphylococcal Cassette Chromosome Elements, IWG-SCC, 2009), which mediate the site-specific integration of SCC into the chromosome of staphylococcal strains, the hdc cluster may transfer to other strains by using this mobile genetic element.
Non-halophilic histamine-producing bacteria In agricultural products, such as wine, cheese, sausage, and other dairy products, O. oeni, Lactobacillus spp., and Streptococcus thermophilus are isolated as problematic histamine-producing bacteria (references are quoted above). These bacterial species are typical LAB known since ancient times, and both the physiological and biochemical properties are consistent with those of typical LAB. Certain Lactobacillus strains have long been known as representative histamine-producing bacteria. According to the previous literature (Hammes et al., 1992; Kandler and Weiss, 1986), Lactobacillus strains are bacilli or cocci, gram positive, and are facultative anaerobes without a respiratory pathway. These strains do not require salt to grow; in fact, growth is inhibited by salt concentrations of 6% or more. The lactic acid-producing ability of these species is great, and they can grow in acidic environments as low as pH 4.0. Their optimal temperature is around 30°C. Histamine-producing LAB are rarely isolated from marine products except for fermented products; however, some lactobacilli or relatives have been isolated from kamaboko as slime-producing bacteria or rancid bacteria. Recently, Lactobacillus otakiensis-like strains have been isolated as histamine producers from pickled mackerel in Japan (Hamaya et al., 2014). Originally, this species was isolated from Japanese vegetable pickles as a non-histamine-producing dominant bacterium (Watanabe et al., 2009). The physiological characteristics of this bacterium are shown in Table 6. Since it is generally accepted that the major histamine producers in pickled mackerel production are gram-negative bacteria resulting from the contamination of materials (Furutani et al., 2013), this requires further study. Notably, pickled fish products are not fermented foods. However, as histamine-producing Lactobacillus strains are present in the field where fishery products are manufactured, this should be taken into consideration when investigating histamine accumulation in pickled-type fermented fishery products.
| Characteristics | Lactobacillus otakiensis JCM15040T | Isolates (n = 19) |
|---|---|---|
| Gram stain | + | + |
| Cell shape | Rod | Rod |
| Motility | − | − |
| Oxygen requirement | − | − |
| Oxidase | − | − |
| Catalase | − | − |
| O/F test | F | F |
| Gas production from glucose | + | d |
| Type of lactate | D + L | D + L or D |
| Histamine production | − | + |
| Growth at 4°C within 4 weeks | + | + |
| at 45°C | − | − |
| Growth at pH 3.6 | − | − |
| at pH 3.8 | − | −(d) |
| at pH 4.0 | + | + |
| Growth in 10% NaCl | − | −(d) |
| in 15% NaCl | − | − |
| Sequence similarity (%) with Lactobacillus otakiensis JCM15040T (AB366386) | 100 | 100 |
Symbol: d, differs among strains.
The potential of histamine formation, even in these LAB, is strain-dependent; therefore, it is thought that the histamine-forming enzyme gene is mobile. The homology of histamine-producing enzymes and the amino acids of the aforementioned halophilic LAB are highly conserved, indicating that the origin of this enzyme. Little is known about the genetic information of histamine-producing L. otakiensis. At present, it has been reported that histamine-producing strains have hdcA, which shares > 99% sequence similarity with that of Lb. hilgardii and Tetragenococcus spp. (Hamaya et al., 2014).
Other histamine-producing bacteria In addition to the aforementioned bacteria, Bacillus spp., Enterococcus spp., and Micrococcus sp. are also known histamine producers (FAO/WHO, 2012), though their behavior in fishery products requires further elucidation. Some Bacillus strains have been isolated from fishery products including fishmeal (Tsai et al., 2006); however, it is unclear if these bacteria can produce significant levels of histamine in fermented fishery products with high salt concentrations. They are thought to produce spores under high salt conditions. In the case of soy sauce fermentation, Enterococcus strains have been isolated as histamine producers, which is contrary to expectations (Ibe et al., 2003). However, further information, such as genetic analysis of hdcA and taxonomy, has not yet been reported. In addition, besides the N-terminal sequence of HDC, little information is available on the histamine-producing Micrococcus sp. (Alekseeva et al., 1976, 1986; Prozorovski and Jörnvall, 1974; van Poelje and Snell, 1990). Since these bacteria have similar growth features to the previously discussed histamine producers in fermented products, future research should include their genetic and biochemical analyses.
Since histamine accumulation was observed mainly in the salted-type fermented products, fish sauce and salted pickled seafood with rice bran, we present here information about the behavior of histamine-producing bacteria in both products.
Fish sauce fermentation The history of fish sauce production in Japan is older than that of soy sauce, as it was a popular condiment until the distribution of soy sauce (Ishige and Raddle, 1990). Recently, to improve its fishy smell and to shorten the fermentation period, mold starter or koji has been added to the fish materials during fish sauce processing (Funatsu et al., 2000). The fermentation mechanism of fish sauce involves the digestion of fish materials by self-digestion enzymes and microorganisms, resulting in the accumulation of chemical compounds that affect the taste and flavor. The accumulation of amino acids and peptides, which are associated with the typical fish sauce taste, results from a self-digestion enzyme secreted by the fish materials. As shown in Fig. 9, during fermentation, bacterial counts, total nitrogen, amino acids, and peptides increase; hence, the pH decreases due to lactic acid accumulation. Halophilic LAB, mainly T. halophilus, are dominant during the fermentation of a variety of Japanese fish sauces. Changes in the bacterial flora of Japanese fish sauce during fermentation have been studied previously (Fujii, 1992; Fujii et al., 2008; Fukui et al., 2012; Shozen et al., 2012a). Halophilic LAB play an important role in the accumulation of lactic acid during fish sauce fermentation, though they have little effect on the digestion of fish materials or the enzymatic production of amino acids. The basic quality of fish sauce is its stability, so long as an appropriate salt concentration is used. However, the accumulation of biogenic amines, including histamine and tyramine, can occur during fish sauce production. In 2011, CODEX adopted a limit of 400 ppm for histamine content in fish sauce. The behavior of histamine-producing tetragenococci matches that of the predominant bacteria also occupied by tetragenococci (Fig. 10), since most of the histamine producers and dominant bacteria belong the same genus, suggesting that the growth characteristics of both are the same in fish sauce mash. In the case of serious histamine accumulation in fish sauce, viable counts are almost the same for histamine producers and total halophilic bacteria, due to the significant proportion of total viable counts for histamine-producing bacteria (Kimura et al., 2015). Simply, the amount of histamine-producing bacteria reflects the histamine content in fish sauce mash for the entire fermentation period (Kimura et al., 2015). The initial counts of histamine producers in fish sauce fermentation also affect histamine accumulation, meaning that even minor contamination with histamine producers may cause serious accumulation.
Changes in the viable counts of microorganisms and chemical compounds during Japanese fish sauce fermentation. In the upper graph, ●; total viable halophilic bacterial count, ■; standard plate count, ▴; viable mold count (mainly Aspergillus sp.). Lower graph, ■; total nitrogen compounds, ▴; pH.
Changes in the viable counts of total halophilic and histamine-producing bacteria and histamine contents in Japanese fish sauce showing high histamine accumulation. In the upper graph, ●; total viable halophilic bacterial count (mainly Tetragenococcus sp.), ▴; viable histamine-producing bacterial count (also mainly Tetragenococcus sp.). Lower graph, ●; histamine content in the sample accumulating histamine.
Rice bran pickles Although histamine-producing bacteria have been isolated from pickled seafood with rice bran, little is known of their behavior during fermentation (Satomi et al., 2011; Yatsunami and Echigo, 1991). Changes in the bacterial flora of mackerel pickles with rice bran were studied (Kosaka et al., 2012), and showed that staphylococci are predominant in the initial stages of fermentation. The predominant organisms in samples changed to tetragenococci, suggesting that the transition of bacterial flora was similar to that observed in fish sauce. In addition, since T. halophilus was isolated as a histamine producer from sardine pickles with rice (Satomi et al., 2012), it is likely to be the primary histamine-producing bacterium. As a secondary causative agent, staphylococci are proposed as histamine producers in these products (Yatsunami and Echigo, 1991), since there are a number of cases where this bacterium was isolated as a histamine producer. As previously mentioned, under high salinity, Tetragenococcus spp. may be problematic due to their salt tolerance; hence, in the comparatively low salt environments of dried fish, sausages, and fish pastes, staphylococci are also likely to be isolated.
Starter cultures In Japan, starter cultures have not typically been used in the production of traditional fermented fishery products. However, starter cultures have obvious utility in producing fermented fishery products, increasing fermentation stability and product quality as well as inhibiting histamine accumulation (Kimura et al., 2015; Shozen et al., 2012a; Zaman et al., 2011). For example, it has been shown that using a bacterial starter can inhibit histamine accumulation during fish sauce fermentation (Fig. 11; Kimura et al., 2015). Two lots of fish sauce mashes were experimentally prepared, samples A and B, in which the number of histamine-producing bacteria was lower than the detection limit (0.3 MPN/g) at the initial stage. A starter culture of T. halophilus was added to sample A (final concentration of approximately 105 cfu/g). No starter culture was added to sample B. Raw materials in both fish sauce mashes were identical, containing the whole bodies of sardines, salt (final concentration of about 15%), and glucose (final concentration of about 1%), except for the addition of the starter culture. After mixing of the ingredients, both fish sauce mashes were fermented at 30°C for 28 days, and then viable counts of halophilic and histamine-producing bacteria, and the amount of histamine were measured. The number of histamine-producing bacteria in sample A was approximately 102 MPN/g after 7 days of fermentation and did not increase afterwards. On the other hand, the number of histamine-producing bacteria in sample B increased considerably from the initial stage, reaching a high level of 108 MPN/g after 7 days. In addition, histamine accumulation was not found in sample A (added starter culture), whereas sample B accumulated high levels of histamine during the fermentation period, reaching approximately 100 mg/100 g within 28 days, with accompanying viable counts of histamine producers. Interestingly, though changes in the number of total halophilic bacteria in samples A and B were almost identical during fish sauce fermentation, the majority of halophilic bacteria in sample B were histamine-producing bacteria, which is the same bacterial type as in the fish sauce mash containing the useful bacteria, T. halophilus. This indicates that the addition of a starter can inhibit the growth of histamine-producing bacteria during fermentation. Of course, it is important to provide a suitable environment for the starter culture in fish sauce mash, ensuring optimal temperature, nutritional and other factors. Actually, additions of sub-material as nutritional help to starter cultures are effective in preventing histamine accumulation during fish sauce fermentation. Figure 12 shows the effect of starter inoculation with or without sucrose addition on the accumulation of histamine in fish sauce mashes during fermentation; sucrose supplementation of the starter culture was obviously effective. It is likely that the carbohydrate compounds that are essential nutrients for the growth of LAB are not sufficient in the fish sauce mashes (Shozen et al., 2012a).
Effect of starter inoculation on suppressed histamine accumulation in fish sauce mashes during fermentation. The addition of a starter (sample A) inhibited the growth of histamine-producing bacteria during fermentation, while sample B (not inoculated with starter) showed histamine accumulation. Vertical bars indicate standard deviation (n = 3) of means. (a) Halophilic bacterial count. (b) Histamine producing bacterial count. (c) Histamine content. Symbols:, NBRC12172 (105 MPN/g) as Sample A; □, No inoculation as Sample B; ◊.
Effect of starter inoculation with or without sucrose addition on histamine accumulation in fish sauce mashes during fermentation. The addition of a starter with sucrose (lots 1 and 2) inhibited histamine accumulation during fermentation, while the sample inoculated with only starter culture showed histamine accumulation. Symbols: ◆, addition of only starter culture; □, addition of starter culture with sucrose (lot 1: 2% final concentration); and △, addition of starter culture with sucrose (lot 2: 2% final concentration).
Reduction of histamine-producing bacteria at the initial fermentation stage As mentioned above, starter cultures are useful in preventing histamine accumulation during fish sauce fermentation. However, the initial contamination levels by histamine-producing bacteria in fish sauce mash are considerable. Here, the sample case is as follows. Two lots of fish sauce mashes were experimentally produced, samples A and B, in which the number of histamine-producing bacteria was lower than the detection limit (0.3 MPN/g) and approximately 9.3 × 102 MPN/g at the initial stage, respectively. Raw materials in both fish sauce mashes were the same as indicated above. After mixing together all of the ingredients including the starter culture (final viable counts, approximately 105 MPN/g), both fish sauce mashes were fermented at 30°C for 28 days, and then viable counts of halophilic and histamine-producing bacteria, and the amount of histamine were measured (Fig. 13). The number of histamine-producing bacteria in sample A was approximately 102 MPN/g after 7 days of fermentation and did not increase thereafter. On the other hand, the number of histamine-producing bacteria in sample B increased considerably from the initial stage, reaching the high level of 108 MPN/g after 7 days. In each sample, the amount of accumulated histamine in the fish sauce mash reflected the number of histamine-producing bacteria. Histamine was not generated in sample A during the entire fermentation period, while sample B accumulated high levels of histamine, which reached approximately 100 mg/100 g within 28 days, despite the addition of a starter culture. Although changes in the number of halophilic bacteria in samples A and B were almost identical, the majority of halophilic bacteria in sample B were histamine-producing bacteria, suggesting that the dominant bacterium in sample B was a histamine-producing halophilic LAB. This study demonstrates the limit in effectiveness of starter culture addition, and indicates the necessity of reducing the number of histamine-producing bacteria at the initial fermentation stage during fish sauce processing. Moreover, it may also reflect the risk of histamine accumulation in products in which the initial counts of histamine-producing bacteria are able to increase during the fermentation process, as the case above introduced. Thus, using a starter culture in combination with a reduction in the initial counts of histamine producers is essential to the processing of fermented fishery products. The effects of sanitation and fermentation control on suppressing histamine accumulation in the manufacture of fish sauce are shown in Fig. 14 as a successive example. Before sanitation and fermentation control, the products contain large amounts of amines. After controlling fermentation (by addition of sucrose to the fish sauce mash), tyramine content was dramatically decreased and subsequent sanitation effects decreased histamine accumulation.
Effect of initial counts of histamine-producing bacteria on histamine accumulation in fish sauce mashes during fermentation. The initial counts of histamine-producing bacteria reflect histamine accumulation in fish sauce mashes during fermentation. (a) Halophilic bacterial count. (b) Histamine-producing bacterial count. (c) Histamine content. Symbols: △, fish sauce mash sample A containing <0.3 MPN/g of histamine-producing bacteria; □, fish sauce mash sample B containing approximately 9.3 × 102 MPN/g of histamine-producing bacteria.
Effects of sanitation and fermentation control on suppressed histamine accumulation during fish sauce manufacturing. Before sanitation and fermentation control, the products contained large amounts of amines. After fermentation control (addition of sucrose to the fish sauce mash), the tyramine content was dramatically decreased and sanitation resulted in decreased histamine accumulation.
Other methods Recently, bioreactors have been developed to eliminate histamine from liquid fermented products such as fish sauce. The principal two methods employed are as follows. The first is an absorption method using bentonite and similar matrices (Shozen et al., 2012b), while the second involves the use of histamine-degrading enzymes (Tapingkae et al., 2010). The histamine absorption procedures of the former method and an example of histamine absorption of fish sauces are shown in Fig. 15 and Fig. 16, respectively. Both methods are applicable to reduce the histamine content of liquid products; however, their absorption or degrading capacity is limited, indicating that these methods alone cannot fully resolve the issue of histamine accumulation in foods.
A flow chart of bentonite utilization.
Effect of bentonite treatment on histamine adsorption in various fish sauces. Bentonite was added to each fish sauce at 10% (W/V). Each of the fish sauce models was incubated at 20°C for 2 hours.
Acknowledgements The author thanks Y. Shibata, T. Takano, and S. Shinya for supplying samples. The technical assistance of N. Hatano and N. Sakai is gratefully acknowledged. Drs. T. Fujii, Y. Funatsu, Y. Yano, T. Ooizumi, A. Furutani, M. Kimura, Y. Fukui, K. Yokoi, Y. Kosaka, H. Oikawa, MSc M. Mori, MSc K. Shozen, Y. Harada, Y. Hamaya, and T. Takewa are acknowledged for their helpful discussions. A number of findings in the text are obtained from joint research with the Rakuno Gakuen University, Ishikawa Prefectural Fisheries Research Center, and the Toyama Prefectural Agricultural, Forestry & Fisheries Research Center.
