Molecular detection and identification of aquatic mycobacteria

Abstract

Abstract Mycobacteriosis (fish tuberculosis) is a progressive disease of a wide range of wild and captive marine and freshwater fish species. While Mycobacterium marinum, M. fortuitum and M. chelonae are the most frequently reported species to be involved in the disease, several new mycobacteria species have also recently been implicated. Conventional detection / identification of fish mycobacteria is based on histopathology, culture and biochemical characteristics. In this study complementary molecular approaches were developed to assist in Mycobacterium identification. First, a highly specific and sensitive multiplex PCR-based assay, targeting two genes (hsp65 and 16S RNA), was established to simultaneously detect the genus Mycobacterium and identify M. marinum, M. fortuitum or M. chelonae from culture or infected fish tissue, based on presence / absence of specific amplicons. In addition, PCR-restriction enzyme analysis (PRA) and DNA sequence analysis of the 16S-23S internal transcribed spacer (ITS) region and a 441 bp fragment of the hsp65 gene demonstrated the limitations of multiplex PCR (and commercial line probe assays) to differentiate among the species of the M. fortuitum complex. However DNA sequence analysis of the hsp65 gene fragment was found to reliably identify M. fortuitum from closely related species, M. conceptionense and M. senegalense. Reliable identification of novel species (or very similar species) of aquatic mycobacteria requires more extensive DNA sequence comparisons. Thus, multigene (polygenetic) analyses, as used here, provide rapid, accurate and reliable species identification of aquatic mycobacteria. Furthermore, a number of novel species of aquatic mycobacteria, M. stomatepiae, ‘M. angelicum’, ‘M. aemonae’ and M. salmoniphilum were discovered using the polygenetic analysis approach. Correct identification of Mycobacterium species by DNA sequence comparisons relies on accurate database information. Difficulties in this study in assigning M. marine and M. gordonae to their correct taxa suggest errors in the current public sequence repositories. The above methods were successfully applied to detect and identify mycobacteria in field samples including formalin-fixed, paraffin-embedded (FFPE) fish tissue, water and frozen fish tissue.