Histamine and other biogenic amines are present in various amounts in many foods. Fresh fish at harvest are virtually free of histamine, but post-harvest conditions that allow for growth of certain spoilage bacteria can result in histamine formation. This article describes histamine fish poisoning, the methods used for detection of histamine-producing bacteria in fish and the potential for the use of molecular-based analytical methods in future developments in the field.
by Dr Kristin Björnsdóttir-Butler and Professor David P. Green

Histamine fish poisoning
Histamine (or scombrotoxin) fish poisoning occurs worldwide through ingestion of fish containing hazardous levels of histamine and other biogenic amines (putrescine and cadaverine). In the United States, this type of posoning accounted for 7.5 % of all foodborne outbreaks, and 38 % of all seafood-related illnesses reported from 1990 to 2003 [1]. Human illness occurs rapidly after ingestion of fish with elevated histamine levels and lasts from several minutes to a few hours. Symptoms include allergic-like responses, such as headache, dizziness, swelling of the tongue, nausea, vomiting, diarrhoea and stomach pain [2,3]. Histamine fish poisoning is usually self-limiting and recovery is usually complete.

Histamine is produced by certain spoilage organisms through action of the enzyme histidine decarboxylase (hdc) which converts the amino acid histidine to histamine. There are two classes of hdc enzyme: one class is found in eukaryotic cells and Gram-negative bacteria while the second group is found in Gram-positive bacteria. Both groups have different coenzymes associated with them. Gram-positive histamine-producing bacteria are more commonly associated with fermented products like salami, cheese, sauerkraut and wine while Gram-negative histamine-producing bacteria are more common in fish.

A wide range of Gram-negative bacteria can produce histamine in fish but the major types are mesophilic enteric and marine bacteria [4]. Bacteria capable of producing large quantities of histamine (>1000 ppm after 24-48 h incubation at temperatures above 15°C in tuna fish infusion broth,TSB, supplemented with 2%) include Morganella morganii, Enterobacter aerogenes, Raoultella planticola, R. ornithinolytica and Photobacterium damselae. Species capable of producing low quantities of histamine (250-500 ppm after 48 h incubation) include Hafnia alvei, Citrobacter freundii, and C. brakii [5].

The primary method used by industry for the control of histamine fish poisoning is  rapid chilling of fish at harvest and continuous chilling during transportation aimed at reducing growth of bacteria and histamine formation. However, some bacteria such as Morganella psychrotolerans grow at low temperature (2.5 °C) and produce large quantities of histamine [6-8]. The US Food and Drug Administration (FDA) considers fish with more than 50 ppm histamine to be decomposed and fish containing more than 500 ppm histamine to be a human health hazard.
Hence, high-histamine-producing bacteria pose a greater risk for human health than low-histamine producers.
Methods for the analysis of histamine in fish include qualitative and quantitative immuno-based test kits as well as high performance liquid chromatographic (HPLC) methods. A better understanding of the relationship between the presence of histamine-producing bacteria and formation of histamine in fish is needed in order to develop effective mitigation strategies. Prompt analysis for histamine-producing bacteria can help identify conditions leading to histamine production and histamine fish poisoning. Therefore, a need remains for reliable, rapid and accurate methods for detection of histamine-producing bacteria in fish.

Detection of histamine-producing bacteria
There are several methods for the detection of histamine-producing bacteria [Table 1] The traditional method is by differential culture-based methods using agar or broth.

Culture-based methods
Niven [9] first developed a differential agar-based medium for detection of histamine-producing bacteria that contains the pH indicator bromocresol purple, tryptone, L-histidine hydrochloride and a few other components. Histamine production occurs during growth of the bacteria and the increased pH of the surrounding medium results in a colour change from green/brown to purple [Figure 1, Left Panel]. Yamani and Unterman [10] adapted Niven’s agar to a broth-based medium for quantification by MPN methods. Others manipulated the pH and the incubation time and temperature to produce a modified Niven’s medium with increased selectivity [11].

Although traditional culture-based methods are easy to use and inexpensive, they have several disadvantages. Low pH (5.3) of the medium can inhibit the growth of some histamine-producing bacteria. In fact, Niven et al. reported a one-log reduction in bacterial growth on Niven’s agar compared to standard plate count agar [9]. In addition, some bacterial strains such as P. damselae and H. alvei do not grow well at low pH [12, 13]. Yoshinaga and Frank adjusted the pH of the medium to 6.5 to support the growth of acid-sensitive Clostridia [14]. However, increasing pH can increase the risk of false-positive reactions. Other workers reported 15 to 72 % false-positive results due to other bacterial metabolic processes that increase the pH of the medium and lead to a colour change similar to that caused by histamine production [15-17]. For these reasons, differential culture-based methods are considered best suited in screening of fish for the presence of histamine-producing bacteria with subsequent confirmation of the production of histamine by chemical methods.

Physicochemical methods
Another detection technique employs physicochemical analysis. Klausen and Huss developed a potentiometric-based method for detection of histamine-producing bacteria in fish; the method measures increased conductance in culture media and can be used to screen bacterial isolates [18]. The method is highly effective for detection of high-histamine producing bacteria, but is not suitable for detection of low-histamine producers [15].

Molecular-based methods
Recent research has focused on the development of molecular-based techniques to detect the hdc gene from Gram-negative histamine-producing bacteria [19]. Several investigators used nucleic acid amplification methods targeting 724, 709 and 534 bp fragments of the hdc gene [20-22]. A PCR method has been developed for the detection of the high-histamine producer, M. morganii, based on the amplification of the variable region of the 16S rDNA gene [23]. Methods such as these can be used to confirm bacterial cultures that screen positive on Niven’s medium, but they do not eliminate the need to verify histamine production by chemical analysis.

Although molecular-based methods result in fewer false-positive results than the culture-based methods, they appear to detect only the high-histamine producing bacteria. We have compared culture-based, potentiometric and traditional PCR methods for detection of histamine-producing bacteria and found that while the Niven’s method detected both high- and low-histamine producing strains, it gave a high number (38 %) of false-positive results [15]. The potentiometric and PCR-based methods produced no false-positive results for high-histamine-producing strains, but failed to detect low-histamine producers.

The failure to detect low-histamine-producing bacteria by molecular-based methods may have several explanations. Firstly, the specific gene in the low-histamine producers is not yet known. It is possible that these bacteria do not carry the hdc gene and produce histamine by some other pathway (indeed, other amino acid decarboxylases are known to decarboxylate histidine in addition to their natural substrates). Secondly, the hdc gene in these bacteria may be plasmid-associated and might not be detected by the molecular method or it could be lost during subculturing of the strains.

Colony lift hybridisation
In order to overcome some limitations inherent to molecular techniques, we combined culture-based methods with nucleic acid hybridisation. Colony lift hybridsation is one technique that is uniquely suited to situations where performance of selective or differential media is less than perfect, as it provides more accurate quantitative results because the target organism can be confirmed without the need for sub-culturing. Hence, we developed digoxigenin-labelled (DIG) probes for the detection and quantification of histamine-producing bacteria using colony hybridisation [Figure 1 Right Panel]. This new method targets high-histamine-producing strains of bacteria based on the specificity of probes targeting the hdc gene of these organisms.

A probe mix created by PCR amplification and labelling of the hdc gene of four high-histamine-producing strains (M. morgannii, R. planticola, E. aerogenes and P. damselae) performed extremely well when applied to strains producing >1000 ppm histamine. The study showed that there was 100 % specificity for high-histamine-producing strains using the probe mix against a culture library of 152 Gram-negative histamine- and non-histamine-producing bacteria. The method did not detect or discriminate between low- and non-histamine-producing bacteria. However, the study provided proof-of-concept evidence that colony lift hybridisation can be used to detect and quantify high-histamine-producing bacteria.

More recently, a real-time PCR method for detection of histamine-producing bacteria was developed by researchers at the US Food and Drug administration [12]. Real-time PCR is considerably less time-consuming than traditional PCR, with results normally obtained within an hour. In addition to template amplification, an internal nucleic acid control can be incorporated into the assay to signal false-negative results.

Future prospects
Development of reliable, rapid and accurate methods for detection of histamine-producing bacteria in fish remains a goal of our research. Important steps have been taken to develop faster and more reliable molecular-based methods, but additional research is needed for the detection of low-histamine-producing bacteria. Prompt analysis of histamine-producing bacteria can be a useful tool for identifying conditions leading to histamine production and histamine fish poisoning. Once a more comprehensive detection method is developed, it can be validated against known histamine-producing organisms to determine the risk of histamine fish poisoning.
 
References
1. Dewaal CS et al. Food Protection Trends 2006; 26:466-473.
2. Jansen SC et al. Annals of Allergy Asthma & Immunology 2003; 91:233-241.
3. Maintz L et al. American Journal of Clinical Nutrition 2007; 85:1185-1196.
4. Kim SH et al. Food Science and Biotechnology 2003; 12:451-460.
5. Bjornsdottir-Butler K et al. Journal of Food Microbiology 2010; 139:161-167.
6. Emborg J et al. International Journal of Systematic and Evolutionary Microbiology 2006; 56:2473-2479.
7. Emborg J et al. International Journal of Food Microbiology 2005; 101:263-279.
8. Haaland H et al. Journal of Food Science and Technology 1990; 25:82-87.
9. Niven CF et al. Applied and Environmental Microbiology 1981; 41:321-322.
10. Yamani MI et al. International Journal of Food Microbiology 1985; 2:273-278.
11. Mavromatis P et al. Journal of Food Protection 2002; 65:546-551.
12. Bjornsdottir-Butler K et al. Food Microbiology 2010. In review.
13. Chen CM et al. Journal of Food Protection 1989; 52:808-813.
14. Yoshinaga DH et al. Applied and Environmental Microbiology 1982; 44:447-452
15. Bjornsdottir K et al. Journal of Food Protection 2009; 72:1987-91.
16. Fletcher GC et al. Journal of Food Protection 1998; 61:1064-1070.
17. Lopez-Sabater EI et al. International Journal of Food Microbiology 1996; 28:411-8.
18. Klausen NK et al. International Journal of Food Microbiology 1987; 5:137-146.
19. Landete JM et al. Critical Reviews in Food Science and Nutrition 2008; 48:697-714.
20. De las Rivas B et al. FEMS Microbiology Letters 2005; 244:367-372.
21. Kanki M et al. Applied and Environmental Microbiology 2002; 68:3462-3466.
22. Takahashi H et al. Applied and Environmental Microbiology 2003; 69:2568-2579.
23. Kim SH et al. Journal of Food Protection 2003; 66:1385-1392.

The authors
Dr Kristin Björnsdóttir-Butler
Post Doctoral Fellow,
US FDA Gulf Coast Seafood Laboratory,
1 Iberville Dr.,
Dauphin Island, AL., USA
e-mail: Kristin.Butler@fda.hhs.gov
&
Dr David P. Green
Professor and Extension Leader
Dept. Food,
Bioprocessing and Nutrition Sciences
North Carolina State University Center for Marine Sciences and Technology
Seafood Laboratory 303 College Circle
Morehead City, NC 28557, USA
Tel: +1 252.222.6304
e-mail: dpg@ncsu.edu