Abstract
In a recent issue of this journal, a study performed in Hong Kong by Chu et al. (2006) concluded that there was no association between the presentation of diarrhoeal symptoms and the faecal isolation of Aeromonas spp. among outpatients from public clinics. Using a multiplex PCR, the authors screened for the presence of three haemolysin genes (ahh1, asa1 and aerA) in order to genotype 253 Aeromonas strains isolated in the investigation of gastroenteritis. No significant association was found between the genotypes, species identification and the amount of growth (on the primary culture), and the presentation of diarrhoeal symptoms defined as bloody, mucous or watery stool specimens.
In our opinion, the authors derived a highly unsubstantiated conclusion from very limited data on a topic that has been under discussion and heavily researched by aeromonad specialists for many years worldwide. In our view, the most important limitations of the study were the criteria on which the authors based their conclusions, which were (1) the presence or absence of the three haemolysin genes, (2) the type of stool as a direct indication of the presence of diarrhoeal symptoms, and (3) the amount of bacteria as an indication of diarrhoea. First, the Canadian PCR genotyping approach (Wang et al., 2003) used by Chu and others was not designed to determine the role that specific strains may play in the development of diarrhoea, but rather was set up simply to screen strains for specific haemolysin genes. The presence or absence of any particular haemolytic virulence gene(s) is only partial evidence of the virulence potential of that strain, and non-haemolytic strains can also be implicated in human infection (Wang et al., 2003). In addition, the authors' results may not even reflect the true presence or absence of the genes studied because, as stated by Wang et al. (2003), PCR may have a low sensitivity. Negative PCR results therefore have to be validated by DNA hybridization with the specific probes. It is well known that the virulence of Aeromonas, as with many other pathogenic micro-organisms, is multifactorial and very complex. Strains may produce different toxins with haemolytic, cytotoxic and enterotoxic activity, as well as many enzymes that are considered virulence factors, and may even possess a type III secretion system (Chopra & Houston, 1999; Chacón et al., 2003; Galindo et al., 2006). No mention is made in the study of Chu and others of the significant body of published work on this subject, nor of the three enterotoxins, Act, Ast and Alt, that have clearly been implicated in cases of Aeromonas-associated diarrhoea (Albert et al., 2000). The considerable array of virulence factors that have been described in Aeromonas has recently been reviewed (Galindo et al., 2006), and many of these have been identified on the recently sequenced genome of the type strain Aeromonas hydrophila ATCC 7966 (Seshadri et al., 2006). However, the specific role that each of these putative virulence factors, alone or in combination, may play in the development of diarrhoea is unknown.