Searching for a valid animal model of hemolytic uremic syndrome.

Abstract

Shiga toxin–producing Escherichia coli (STEC) is the most common cause of diarrhea-positive hemolytic uremic syndrome (HUS). Producing considerable morbidity and mortality, diarrhea-positive HUS is characterized by hemorrhagic colitis, microangiopathic hemolytic anemia (MHA), thrombocytopenia, and acute renal failure. Although the gastrointestinal tract and the kidney are the organs most commonly affected, central nervous system, skeletal, pancreatic, and myocardial involvement also can occur. Despite the significant progress made during the past few years in understanding the pathogenesis of STEC-associated HUS, the precise sequence of events by which infection causes disease is still unclear. Efforts to address this question in vivo have been hampered greatly by the lack of animal models that fully recapitulate the entire spectrum of HUS as seen in humans. Though STEC strains cause gastrointestinal, neurologic, or systemic symptoms in gnotobiotic piglets, rabbits, and mice, no such models have shown the hallmark histopathologic finding of renal thrombotic microangiopathy (TMA) following oral ingestion of bacteria; a baboon model has shown such changes but only with intravenous infusion of Shiga toxin 1.1 More recently, ferrets infected orally with E coli O157:H7 were shown to develop damage to glomeruli mimicking that seen in HUS by light microscopy.2 In this issue of the Journal, Gunzer and colleagues3 present a gnotobiotic piglet model of renal TMA using oral infection with O157:H7 or O26:H11 wild-type isolates of enterohemorrhagic E coli (EHEC). The authors demonstrate that the morphologic features of renal TMA in the piglets closely resemble those seen in human HUS. In addition, colitis and myelin degeneration, both of which are typical of human disease, were detected. However, the thrombocytopenia, MHA, and renal failure seen in human disease were not produced. HUS has been modeled before with gnotobiotic piglets. Various reports showed neurologic signs and vascular damage in the brains of gnotobiotic piglets infected orally with E coli O157:H7.4,5 However, in the study that looked for renal disease, there was no evidence of renal TMA.5 So, why were Gunzer and coworkers3 able to reproduce the typical renal changes of TMA experimentally while other researchers could not? These seemingly contradictory results may be explained by differences in virulence of the STEC strains used in the respective studies. Gunzer and colleagues3 used E coli isolated from HUS cases in Walla Walla, WA (serotype O157:H7), and Hannover, Germany (serotype O26:H11). Although previous studies also used O157:H7 serotypes, bacteria of the same serotype may have significant genetic differences that could cause variations in virulence. Perhaps the most important factor in virulence is the quantity of Shiga toxin 2 produced by the bacteria. However, qualitative differences in virulence also may be contributory. As the recently completed genomic sequence of EHEC O157:H7 suggests, there may be several additional uncharacterized virulence factors.6 Considering our incomplete understanding of HUS, the opportunity to use genetically modified strains of E coli makes the availability of a suitable animal model all the more valuable. The probable initial step in pathogenesis of STEC-associated HUS, adhesion, is a complex interaction between the bacteria and the intestinal wall. In vitro studies suggest that E coli O157:H7 initially binds small bowel epithelium associated with Peyer patches, followed by large intestine colonization.7