Shigella and Escherichia Coli at the Crossroads: Machiavellian Masqueraders Or Taxonomic Treachery?
Total Page:16
File Type:pdf, Size:1020Kb
J. Med. Microbiol. Ð Vol. 49 2000), 583±585 # 2000 The Pathological Society of Great Britain and Ireland ISSN 0022-2615 EDITORIAL Shigella and Escherichia coli at the crossroads: machiavellian masqueraders or taxonomic treachery? Shigellae cause an estimated 150 million cases and genera. One authority has even proposed that entero- 600 000 deaths annually, and can cause disease after haemorrhagic E. coli EHEC) such as E. coli O157:H7 ingestion of as few as 10 bacterial cells [1]. They are are essentially `Shigella in a cloak of E. coli antigens' spread by the faecal±oral route, with food, water, [7]. fomites, insects especially ¯ies) and direct person-to- person contact. S. dysenteriae causes brisk and deadly Shigella-like strains of E. coli that cause an invasive, epidemics, particularly in the developing world; S. dysenteric diarrhoeal illness were ®rst described in ¯exneri and S. sonnei account for the endemic form of 1971, over a decade before the appearance in 1982 of the disease, particularly in industrialised nations; S. the new EHEC strains that launched the current wave boydii is rarely encountered [1, 2]. of interest in the E. coli±Shigella connection [8]. Termed `enteroinvasive E. coli' EIEC), these strains, Shigellosis is a locally invasive colitis in which bacteria like shigellae, were able to invade and proliferate invade and proliferate within colonocytes and mucosal within intestinal epithelial cells, eventually causing cell macrophages, trigger apoptosis of macrophages and death [4, 8]. EIEC share with shigellae a c. 140-MDa spread through the mucosa from cell to cell [1]. plasmid pINV) that encodes several outer-membrane Cytokines produced by epithelial cells and macro- proteins involved in invasion of host cells [4, 8]. Like phages in response to invading bacteria stimulate an shigellae, EIEC are usually non-motile and lactose- intense acute in¯ammatory response. This limits the negative. They are also usually lysine decarboxylase- extent of bacterial invasion, but also contributes to negative, a characteristic which, in both EIEC and mucosal injury [1]. Despite their local invasiveness, shigellae, is due to large deletions `black holes') in the shigellae rarely cause bacteraemia. However, shigello- genome that correspond to chromosomal regions sis occasionally precipitates the haemolytic-uraemic present in E. coli K-12. The deleted genes encode syndrome HUS), which is characterised by vaso- enzymes involved in metabolic reactions that produce occlusive renal failure, consumptive thrombocytopenia, compounds such as cadaverine that interfere with micro-angiopathic haemolytic anaemia and neuro- uptake of shigellae and EIEC by host cells. Hence, logical dysfunction. These manifestations result in part the deletions may enhance the ability to invade host from the actions of Shiga toxin, the distinctive cells [4]. EIEC and shigellae also exhibit considerable cytotoxin produced by S. dysenteriae type 1, on antigenic cross-reactivity [4, 8]. However, EIEC do not neurons and on renal and cerebral endothelial cells produce Shiga toxins and are not known to cause HUS. [1, 2]. Perhaps because of this, or because they are not known to have caused large-scale outbreaks, EIEC have Ever since the initial discovery of the dysentery received considerably less public health and media bacillus now S. dysenteriae) by Kiyoshi Shiga in attention than the now-notorious EHEC [7]. 1898 [3], Shigella spp. have generally been regarded on both microbiological and clinical grounds as distinct EHEC and shigellae exhibit a number of clinical and from Escherichia coli. Whereas most E. coli are pathogenetic differences. The prototypical EHEC- motile, ferment lactose, are lysine decarboxylase- and associated syndrome is haemorrhagic colitisÐpassage indole-positive, and are mostly avirulent, shigellae are of large-volume, grossly bloody stools without obvious non-motile, are typically lactose-, lysine decarboxylase- pus or mucus, accompanied by abdominal pain, but not and indole-negative, and are obligate pathogens [1, 4]. tenesmus or fever [5, 6]. This differs from classical Consequently, the emergence of E. coli O157:H7 and dysentery, and probably re¯ects the pancolitis of other Shiga toxin-producing E. coli strains that, like EHEC, in contrast to the more focal proctocolitis of shigellae, cause food- and water-borne outbreaks of shigellosis and EIEC infection, and the toxigenic but haemorrhagic colitis, a dysentery-like illness some- non-invasive nature of EHEC disease, in contrast to the times complicated by HUS [5, 6], has prompted a mucosal invasion of shigellosis. EHEC share with S. reconsideration of the relationship between these dysenteriae type 1 the production of bacteriophage- 584 EDITORIAL encoded Shiga toxin s) and expression of a type III assigned to two adjacent lineages within the distant) secretion apparatus [1, 9]. However, in contrast to phylogenetic group D. These data con®rmed the shigellae, EHEC produce a unique enterohaemolysin genetic diversity within Shigella spp., the intermingling and exhibit `attaching and effacing' adherence to of Shigella spp. and E. coli, and discrepancies between enterocytes similar to that exhibited by enteropatho- conventional Shigella `species' designations and ob- genic E. coli EPEC) [9]. Furthermore, whereas in served genetic relationships. They also localised shigellae the type III secretion apparatus is encoded on Shigella strains speci®cally to E. coli phylogenetic the 140-MDa virulence plasmid [1], in both EHEC and groups A and D [15]. Sequence analysis of the EPEC it is encoded on a chromosomal pathogenicity- housekeeping gene mdh and ribotyping yielded largely associated island termed `locus of enterocyte efface- similar results, but showed somewhat different dis- ment'. Moreover, the virulence plasmid of EHEC is tributions of the Shigella spp. among the E. coli distinct from that of Shigella spp. and EIEC as well as phylogenetic groups [15, 19]. from that of EPEC) [8±10]. Thus, EHEC, EIEC and EPEC all share with one another, and with shigellae, The data of Pupo et al. [15] showed that EIEC, EPEC one or more important clinical or pathogenetic features. and ETEC were distributed across multiple E. coli These perplexing similarities and differences are best lineages within phylogenetic groups A and B1, with understood in the context of the population structure of little clustering according to pathogenic type. No E. coli and Shigella spp. diarrhoeal strains were assigned to phylogenetic group B2, which is the source of most E. coli strains that The ®rst direct genetic insights into the evolutionary cause urinary tract infections, bacteraemia and neonatal relationships between Shigella spp. and E. coli were meningitis [20]. Of the diarrhoeagenic types, only provided by DNA±DNA re-association studies which EHEC of which all representatives studied were E. showed that shigellae exhibit .75% nucleotide simi- coli O157:H7) appeared among the non-aligned larity to E. coliÐwithin the range observed among lineages of reference strains, where they constituted a different E. coli strains [11]. Although the authors of single cluster [15]. These data contradicted the this report proposed only that shigellae could be part of hypothesis that O157:H7 EHEC are derived from the genus Escherichia, the data actually justify placing either shigellae or EIEC, and revealed that EIEC, Shigella spp. within E. coli proper. EPEC and ETEC have evolved independently several times in different E. coli lineages, presumably through Comparative DNA sequence analysis and multilocus horizontal acquisition of the requisite virulence traits enzyme electrophoresis MLEE) subsequently showed [15]. unequivocally that, from a population genetics stand- point, shigellae were better regarded as pathotypes or If O157:H7 EHEC did not evolve from shigellae or clones of E. coli [12±16]. These data showed that the EIEC, they may have arisen instead from a pre-existing four `Shigella' species constituted two related clusters lineage of O157 antigen-positive non-EHEC through within the larger E. coli population, one comprising horizontal acquisition of EHEC-speci®c virulence representatives of three of the four putative Shigella genes. However, this hypothesis was, in turn, contra- species, the other comprising representatives of all four dicted by analysis of the population structure of O157 species [12, 16]. Within each cluster, different Shigella antigen-positive E. coli strains representing diverse H species intermingled in an overlapping fashion in types, clinical sources and host species, which showed adjacent lineages, but in some instances were actually that O157:H7 EHEC occupy their own lineage, far placed together in the same lineage [16]. This indicated removed from non-EHEC, non-H7 O157 strains [21]. that the conventional nomenclature for Shigella spp. Thus, O157:H7 EHEC appeared to have acquired the could be justi®ed only on historical and clinical O157 antigen by horizontal transfer from an unknown grounds [13]. distant donor rather than by vertical inheritance from a clonal ancestor [21]. Over the past decade, understanding of the population structure of E. coli has been re®ned [17, 18]. The The ancestral clonal group within which O157:H7 precise position of Shigella strains within the E. coli EHEC did evolve was revealed by MLEE data showing phylogenetic tree was clari®ed by MLEE analysis of that the nearest neighbours of O157:H7 EHEC are diverse strains of shigellae, EHEC, EIEC, EPEC and EPEC strains of serotype