TEXT

Top Abstract Text References

Enterohemorrhagic Escherichia coli (EHEC) strains, including E. coli O157:H7, are emerging pathogens responsible for outbreaks and sporadic cases of diarrhea (11). EHEC isolates often share numerous virulence factors with other pathogenic E. coli strains, but are distinguished by their production of Shiga toxins (Stx). The activity of these A-B-type toxins in the human microvasculature can result in the most severe consequences of EHEC infection, including hemorrhagic colitis and hemolytic-uremic syndrome (11). Two main types of Stx, Stx1 and Stx2, have been described, each consisting of human glycolipid-binding B subunits and enzymatic A subunits that inhibit protein synthesis by cleaving eukaryotic 28S rRNA. More than 60 serotypes of E. coli associated with human disease have been found to encode the Shiga toxins (1).

The stx genes in most, if not all, EHEC strains are carried by lysogenic bacteriophages of the lambdoid family (17, 22, 26, 28). The toxin genes from lysogens of several of these phages were cloned and sequenced (2, 9, 20), and primer extension analyses were used to identify functional promoters immediately 5' of each toxin-coding sequence (5, 27). Although the Stx1 promoter was found to be inducible in low-iron growth media by virtue of an operator for the iron-dependent Fur transcriptional repressor (3, 21), no environmental parameters for transcriptional regulation at the Stx2 promoter have been found (18).

Superimposed on the transcriptional regulation of these toxin gene-associated promoters is the contribution of the phage lytic cycle to Stx production. Whereas most toxin-encoding bacteriophages are thought to serve primarily as vectors for dissemination of toxin genes among bacterial strains, it has become clear that the phage life cycle has a central role in the regulation of Stx production by EHEC. Recent work has shown that phage induction may contribute to EHEC pathogenesis in a number of ways, including (i) increasing toxin gene copy number as a result of phage genome replication, (ii) increasing toxin transcription from phage promoters that are repressed during lysogeny, and (iii) leading to toxin release in the process of phage-mediated bacterial lysis (19, 23, 29). In fact, very little toxin is produced by an EHEC mutant strain which contains a deletion of the late phage promoter, P_R', suggesting that phage induction is critical for toxin expression (P. Wagner, M. Neely, X. Zhang, D. Acheson, M. Waldor, and D. Friedman, submitted for publication).

The molecular event that initiates phage  induction is cleavage of the repressor cI, which holds phage transcription silent during lysogeny (24). cI cleavage is RecA dependent and occurs as a consequence of activating the bacterial SOS response to DNA damage (13). Mitomycin C (18) and fluoroquinolone (16, 31) antibiotics are DNA-damaging agents known to induce the SOS response and are capable of inducing lambdoid phages, as well as Stx production by EHEC strains. However, most patients who develop the severe sequelae of EHEC infection are not exposed to these agents. Therefore, some other, perhaps endogenous, agent in the human intestine may activate the lytic cycle of Stx-encoding bacteriophages and thereby trigger in vivo Stx production by EHEC.

Human neutrophils and other cell types release a variety of antibacterial molecules, some of whichincluding H₂O₂ and NOare known to damage bacterial DNA in a way that induces the SOS response (10, 15). We thus hypothesized that human neutrophils and their products might induce Stx-encoding phages and Stx production by EHEC. To test this hypothesis, we first exposed the O157:H7 EHEC clinical isolate 1:361 to a range of concentrations of H₂O₂ in vitro. A dose-dependent relationship between H₂O₂ concentration and Stx2 production by strain 1:361 organisms was observed, as shown in Fig. 1A. Concomitant with the observed increase in toxin production, a >10-fold increase in phage titer was also observed upon treatment of 1:361 with 0.25 mM H₂O₂ (not shown). The isogenic 1:361 derivative, 1:361Q-P_R', which lacks sequences necessary for late phage transcription and is known to be impaired in toxin production when treated with mitomycin C (Wagner et al., submitted), was similarly impaired in the presence of H₂O₂ (Fig. 1B). These results suggest that the H₂O₂-mediated increase in Stx2 production depends upon late phage transcription.

View larger version (51K): [in this window] [in a new window]   FIG. 1.   H2O2 induces Stx2 expression by the E. coli O157:H7 clinical isolate 1:361. (A) An overnight culture of 1:361 was diluted 1/5,000 in L broth containing H2O2 (Sigma, St. Louis, Mo.) at the indicated concentrations. 1:361 cultures did not grow in L broth containing >0.5 mM H2O2. (B) 1:361 and 1:361Q-PR' cultures were diluted 1/5,000 in L broth () or L broth supplemented with 0.25 mM H2O2 (). Stx2 levels (nanograms/milliliter) were measured from sonicated overnight cultures by using a previously described enzyme-linked immunosorbent assay (6). Mean values from three independent cultures are shown along with standard deviations. The addition of 0.25 mM H2O2 resulted in a statistically significant (P < 0.004) increase in Stx2 production by 1:361.

We next cocultured strain 1:361 with human neutrophils isolated from healthy donors. A significant inducing effect of neutrophils on bacterial Stx2 production was observed, relative to growth in medium alone (Fig. 2A). A reasonable explanation for this observation is that neutrophil-derived factors damage bacterial DNA and thereby activate the SOS response and subsequent phage induction. If this hypothesis is correct, then 1:361Q-P_R', which is defective in late phage transcription, should be impaired in Stx2 production when cocultured with neutrophils. Furthermore, inhibitors of neutrophil enzymes required for the generation of H₂O₂ and other factors that activate the SOS response should reduce the effect of neutrophil induction of Stx2 production. Each of these predictions proved correct. 1:361Q-P_R' made very small amounts of Stx2 in the presence of neutrophils (Fig. 2B). Oxygen-derived free radical production by neutrophils can be antagonized by diphenyleneiodonium (DPI), an agent known to inhibit NADPH oxidase (4) but not the production of other microbicidal factors by human neutrophils (7). DPI prevented the neutrophil-associated increase in Stx2 production by 1:361 organisms (Fig. 2B), but not mitomycin C-associated Stx2 induction (not shown). Thus, despite the fact that 1:361 organisms made less Stx2 overall when grown in DPI, these results suggest that DPI inhibited the production of phage-inducing factors by neutrophils rather than the response of strain 1:361 organisms to these factors. Our data are consistent with the hypothesis that the production of NADPH oxidase-dependent reactive oxygen intermediates by neutrophils induces the bacterial SOS response and thereby augments production of Stx2.

View larger version (51K): [in this window] [in a new window]   FIG. 2.   Coculture of 1:361 with human neutrophils results in increased Stx2 production. (A) An overnight culture of 1:361 was diluted 1/5,000 in a solution consisting of one part L broth and one part Hank's buffered saline solution or one part L broth and one part Hank's buffered saline solution containing 106 neutrophils. The addition of neutrophils resulted in a statistically significant (P = 0.01) increase in Stx2 production by 1:361 organisms. (B) 1:361 or 1:361Q-PR' was diluted 1/5,000 from an overnight culture as described above, without () or with (+) 106 neutrophils and with or without DPI (Sigma) at a final concentration of 2 µM. Neutrophils were isolated as previously described (8), and Stx2 levels were measured as described for Fig. 1.

We recently found that Shiga toxin production is regulated as part of the lytic cycle of the toxin-encoding bacteriophages (Wagner et al., submitted). Therefore, an important next step in understanding EHEC pathogenesis is the identification of host factors that induce Stx-encoding prophages and thereby toxin production. Our current data suggest that neutrophils may be a source of such factors. Besides inducing Stx production by EHEC, several recent observations implicate neutrophils in other aspects of EHEC pathogenesis. Epidemiologic studies have revealed that EHEC disease severity correlates with peripheral blood neutrophil counts (25). In vitro data suggest that neutrophil infiltration into the intestinal lumen enhances Stx uptake (B. P. Hurley, C. M. Thorpe, A. J. King, G. T. Keusch, and D. W. K. Acheson, Abstr. 100th Gen. Meet. Am. Soc. Microbiol., abstr. B-101, 2000). Furthermore, Stx2 can induce the respiratory burst of neutrophils (12) and inhibit neutrophil apoptosis (14). Taken together, these observations suggest a model in which neutrophils act to facilitate both Stx production and absorption. Such a model, if correct, has important clinical implications. If host factors such as neutrophils and their products contribute to EHEC pathogenesis, then it is reasonable to consider inhibition of these factors as a novel therapeutic strategy for EHEC treatment. The need for new approaches is underscored by the fact that many antibiotics commonly used to treat diarrhea are known to induce bacteriophages and are associated with increased morbidity in patients with EHEC infections (30).