Nontuberculous Mycobacteria in Freshwater Fish and Fish Products Intended for Human Consumption

Introduction

N ontuberculous mycobacteria (NTM) are ubiquitous, potentially pathogenic organisms that have been isolated from a variety of environmental sources, including water, soil, dust, aerosol, and from animals (Kazda et al., 2009). For humans, NTM represent an increasing health risk, especially in immunodeficient individuals. Infections can occur through inhalation, ingestion, or through skin trauma (Falkinham, 1996). NTM have been isolated from various kinds of food, and many studies support the hypothesis that food, especially raw or partially cooked products, plays a role as a source of NTM for humans (von Reyn et al., 1996; Yoder et al., 1999; Argueta et al., 2000).

NTM are associated with various types of aquatic environments, including drinking water (Falkinham, 1996; Primm et al., 2004), and were isolated from many species of marine and freshwater fish, including those intended for human consumption (Brocklebank et al., 2003; Aranaz et al., 2008; Gauthier et al., 2009; Mrlik et al., 2012). Consumption of raw or partially cooked fish/shellfish has been identified as a risk factor for NTM infections in patients with human immunodeficiency virus/acquired immune deficiency syndrome (von Reyn et al., 1996; Ristola et al., 1999).

Our objective was to ascertain whether freshwater fish destined for human consumption and fish products contain NTM and whether they represent a possible risk for exposure and/or infection of humans.

Materials and Methods

A total of 92 samples of freshwater fish and fish products were investigated. One fillet or one gutted fish (muscle tissue) or one package (smoked products) was considered as one sample. Twenty-five samples of freshwater cultured fish from five species intended for human consumption (Table 1) were obtained from fish harvest of four ponds in the Czech Republic. Samples of retail fresh, frozen, and smoked fish and fish products (n=67) from a total of nine fish species (Table 1) were purchased from four supermarkets and one farmers' market.

Twenty-five grams of each sample were homogenized in a Stomacher^® 80 biomaster (Seward Ltd., Worthing, United Kingdom) with 30 mL phosphate buffer solution (pH 7.2). The suspension was centrifuged (3 000 rpm/20 min) and the supernatant was discarded. The pellet was resuspended and decontaminated in 1 N HCl and then neutralized with 2 N NaOH (Matlova et al., 2005). Two hundred microliters of the suspension was inoculated onto two Herrold's egg-yolk media nonsupplemented or supplemented with antibiotic mixture to obtain the following final concentration in the medium: 800 U/L of penicillin, 25 μg/L of chloramphenicol, and 1 mg/L of amphotericin B, and onto Stonebrink-Leslie solid medium for incubation at 25, 30, and 37°C. Culture media were checked after 7, 15, 30, and 60 days of incubation for growth of mycobacteria.

Suspected mycobacterial isolates were stained by Ziehl-Neelsen technique and acid-fast bacilli were analyzed according to a previously described duplex polymerase chain reaction (PCR) to identify the genus Mycobacterium (Moravkova et al., 2008). Corresponding mycobacterial species were identified by their growth characteristics and sequence analysis of the 16S rRNA gene according to a previously described method (Harmsen et al., 2003). Experimental sequences were aligned with the available database entries using the “blastn algorithm” (EZ Taxon: http://www.eztaxon-e.org/). The cut-off value for the identity scores of the 16S rRNA gene was set as 99%. Isolates with identical region of the analyzed 16S rRNA gene were not distinguished in this study (Table 2).

For detection of mycobacterial DNA in investigated samples, DNA was isolated according to Slana et al. (2010) from 50 mg of each pellet (made from 25 g of the sample as described above) before cultivation. Isolated DNA was used as a template for genus-specific conventional PCR assay targeting the rpoB gene, which enables the detection of mycobacteria and differentiation of NTM from members of the M. tuberculosis complex as reported previously (Mokaddas and Ahmad, 2007).

Results

Viable mycobacteria were isolated from 11 (11.9%) out of 92 investigated samples (Table 1). Fifteen mycobacterial isolates were obtained (Table 2) and classified as M. fortuitum (n=5), M. immunogenum (n=2), M. phocaicum/mucogenicum (n=1), M. neoaurum (n=2), M. peregrinum (n=2), M. porcinum (n=1) and M. senegalense/ houstonense/conceptionense (n=2). Obtained mycobacterial species and their growth characteristics are listed in Table 2. Of the 92 fish tissue samples analyzed, 36 (39.1%) were found to contain NTM DNA (Table 1). None of the samples originating from smoked fish products was positive by culture, or by PCR method. The proportion of positive samples originating from heat-untreated samples was the lowest in fresh fish from ponds (4.0% by PCR; 0% by culture), a higher proportion was observed in retail fresh fish (60.9% by PCR; 8.7% by culture), and the highest in retail frozen fish (91.3% by PCR; 39.1% by culture).

Discussion

The occurrence of NTM in fish can be the result of mycobacterial infection or alternatively, mycobacteria can passively contaminate fish or fish products (Gauthier et al., 2009). In our study, the presence of mycobacterial DNA was found in 36 (39.1%) out of a total of 92 investigated freshwater fish and fish products. However, culture examination of samples verified the presence of viable mycobacteria in 11 (11.9%) samples. A possible explanation could be the presence of only low numbers of viable mycobacteria or the presence of dead or noncultivable, cell-wall-deficient cells (Beran et al., 2006). Moreover, the long-term cultivation of mycobacteria requires the use of a decontamination procedure to inhibit the growth of rapidly growing microbes. However, decontamination also reduces the number of viable, colony-forming mycobacteria (Falkinham, 1996).

Viable mycobacteria were predominantly isolated from frozen fish samples. As has been previously demonstrated, freezing did not devitalize mycobacteria during storage of samples including frozen fish (Livanainen et al., 1995; Mediel et al., 2000).

M. fortuitum isolated in our study is one of the most common causes of fish mycobacteriosis (Gauthier et al., 2009), and may cause various pulmonary, skin, or soft-tissue infections in humans (Falkinham, 1996; Amorim et al., 2010). Other species closely related to M. fortuitum were also isolated although not distinguished to the species level in most cases (i.e., M. phocaicum/mucogenicum, M. senegalense/houstonense/ conceptionense, M. peregrinum, and M. porcinum). These mycobacterial species have also been described as important facultative pathogens in humans (Brown-Elliott and Wallace, 2002; Amorim et al., 2010; Brown-Elliott et al., 2011; Kim et al., 2012).

M. immunogenum, isolated from one tissue sample of tilapia, has been identified as the etiologic agent of a variety of infections in humans including hypersensitive pneumonitis, (Wallace et al., 2002). M. neoaurum, isolated from two different pangasius fillets, has been associated especially with catheter-related sepsis in humans (Brown-Elliott et al., 2010).

No viable mycobacteria were isolated from fresh fish from ponds, and mycobacterial DNA was detected in only one (4.0%) sample. The low prevalence (1.7%) of mycobacteria in tissue samples of freshwater fish from water reservoirs, ponds, and farms in the Czech Republic was also ascertained by Mrlik et al. (2012).

In our study, viable mycobacteria and/or their DNA were frequently detected in the retail fresh and even more frequently in the frozen fish samples, but in none of the smoked fish products. Retail fresh fish are in contact with drinking water during the distribution and subsequent retail when they are chilled on ice. The frozen fish products contain drinking water in the amount of 20–25% (according to labels) mainly as a glaze on the surface of fillets. This could influence the presence of mycobacteria in these samples, because water, including drinking water, is considered to be an important reservoir of NTM (Falkinham, 1996; Primm et al., 2004). In the study of Mediel et al. (2000), various NTM species (M. fortuitum, M. gordonae, M. chelonae, M. nonchromogenicum, M. peregrinum, and M. terrae complex members) were isolated from the tissues and defrosted water of frozen marine fish.

Eight out of 11 samples, in which we detected viable mycobacteria, were imported frozen fillets of pangasius and tilapia. These tropical freshwater fish species are significant sources of white fish aquaculture products on the global market and require warm water generally above 21°C for their breeding. Such temperature conditions are favorable also for the growth of NTM (Kazda et al., 2009), and may have contributed to the frequent presence of mycobacteria in the tissues of these fish species.

Conclusions

The results of our study suggest that freshwater fish might be a reservoir of various species of NTM for humans. Mycobacteria can be found especially in retail fresh and frozen fish products, probably due to secondary contamination during subsequent processing. However, such fish products are usually cooked prior to consumption and the risk of human exposure to viable NTM through improperly cooked fish and fish products is likely to be low. Nevertheless, proper and safe handling and treatment of freshwater fish before consumption is recommended.