Main content area

Unveiling hákarl: A study of the microbiota of the traditional Icelandic fermented fish

Osimani, Andrea, Ferrocino, Ilario, Agnolucci, Monica, Cocolin, Luca, Giovannetti, Manuela, Cristani, Caterina, Palla, Michela, Milanović, Vesna, Roncolini, Andrea, Sabbatini, Riccardo, Garofalo, Cristiana, Clementi, Francesca, Cardinali, Federica, Petruzzelli, Annalisa, Gabucci, Claudia, Tonucci, Franco, Aquilanti, Lucia
Food microbiology 2019 v.82 pp. 560-572
Acinetobacter, Alkalibacterium, Alternaria, Candida glabrata, Candida parapsilosis, Candida tropicalis, Candida zeylanoides, Cladosporium, DNA, Debaryomyces, Epicoccum, Erysipelothrix, Lactococcus, Listeria, Ochrobactrum, Phoma, Photobacterium, Pseudomonas aeruginosa, Saccharomyces cerevisiae, Shiga toxin-producing Escherichia coli, Somniosus microcephalus, Sporobolomyces, Staphylococcus, Tissierella creatinophila, Torulaspora, aerobes, community structure, culture media, data collection, denaturing gradient gel electrophoresis, enzymes, fermentation, fermented fish, lactic acid bacteria, meat, pH, plate count, quantitative polymerase chain reaction, ready-to-eat foods, sensory properties, sharks, toxicity, trimethylamine, yeasts
Hákarl is produced by curing of the Greenland shark (Somniosus microcephalus) flesh, which before fermentation is toxic due to the high content of trimethylamine (TMA) or trimethylamine N-oxide (TMAO). Despite its long history of consumption, little knowledge is available on the microbial consortia involved in the fermentation of this fish. In the present study, a polyphasic approach based on both culturing and DNA-based techniques was adopted to gain insight into the microbial species present in ready-to-eat hákarl. To this aim, samples of ready-to-eat hákarl were subjected to viable counting on different selective growth media. The DNA directly extracted from the samples was further subjected to Polymerase Chain Reaction-Denaturing Gradient Gel Electrophoresis (PCR-DGGE) and 16S amplicon-based sequencing. Moreover, the presence of Shiga toxin-producing Escherichia coli (STEC) and Pseudomonas aeruginosa was assessed via qualitative real-time PCR assays. pH values measured in the analyzed samples ranged from between 8.07 ± 0.06 and 8.76 ± 0.00. Viable counts revealed the presence of total mesophilic aerobes, lactic acid bacteria and Pseudomonadaceae. Regarding bacteria, PCR-DGGE analysis highlighted the dominance of close relatives of Tissierella creatinophila. For amplicon sequencing, the main operational taxonomic units (OTUs) shared among the data set were Tissierella, Pseudomonas, Oceanobacillus, Abyssivirga and Lactococcus. The presence of Pseudomonas in the analyzed samples supports the hypothesis of a possible role of this microorganism on the detoxification of shark meat from TMAO or TMA during fermentation. Several minor OTUs (<1%) were also detected, including Alkalibacterium, Staphylococcus, Proteiniclasticum, Acinetobacter, Erysipelothrix, Anaerobacillus, Ochrobactrum, Listeria and Photobacterium. Analysis of the yeast and filamentous fungi community composition by PCR-DGGE revealed the presence of close relatives of Candida tropicalis, Candida glabrata, Candida parapsilosis, Candida zeylanoides, Saccharomyces cerevisiae, Debaryomyces, Torulaspora, Yamadazyma, Sporobolomyces, Alternaria, Cladosporium tenuissimum, Moristroma quercinum and Phoma/Epicoccum, and some of these species probably play key roles in the development of the sensory qualities of the end product. Finally, qualitative real-time PCR assays revealed the absence of STEC and Pseudomonas aeruginosa in all of the analyzed samples.