Search

To discern if there was a particular genotype associated with clinical enteroaggregative Escherichia coli (EAEC) strains isolated from deployed military personnel (DMP) with travelers' diarrhea (TD), we characterized a collection of EAEC from DMP deployed to Afghanistan, Djibouti, Kenya, or Honduras. Although we did not identify a specific EAEC genotype associated with TD in DMP, we found that EAEC isolated at the first clinic visit were more likely to encode the dispersin gene aap than EAEC collected at follow-up visits. A majority of the EAEC isolates were typical EAEC that adhered to HEp-2 cells, formed biofilms, and harbored genes for aggregative adherence fimbriae (AAF), AggR, and serine protease autotransporters of Enterobacteriaceae (SPATEs). A separate subset of the EAEC had aggR and genes for SPATEs but encoded a gene highly homologous to that for CS22, a fimbriae more commonly found in enterotoxigenic E. coli. None of these CS22-encoding EAEC formed biofilms in vitro or adhered to HEp-2 cells. Whole genome sequence and single nucleotide polymorphism analyses demonstrated that most of the strains were genetically diverse, but that a few were closely related. Isolation of these related strains occurred within days to more than a year apart, a finding that suggests a persistent source and genomic stability. In an ampicillin-treated mouse model we found that an agg4A+ aar- isolate formed a biofilm in the intestine and caused reduced weight gain in mice, whereas a strain that did not form an in vivo biofilm caused no morbidity. Our diverse strain collection from DMP displays the heterogeneity of EAEC strains isolated from human patients, and our mouse model of infection indicated the genotype agg4A+ aar- and/or capacity to form biofilm in vivo may correlate to disease severity.

In this study we compared nine Shiga toxin (Stx)-producing Escherichia coli O157:H7 patient isolates for Stx levels, stx-phage insertion site(s), and pathogenicity in a streptomycin (Str)-treated mouse model. The strains encoded stx2a, stx1a and stx2a, or stx2a and stx2c. All of the strains elaborated 105-106 cytotoxic doses 50% (CD50) into the supernatant after growth in vitro as measured on Vero cells, and showed variable levels of increased toxin production after growth with sub-inhibitory levels of ciprofloxacin (Cip). The stx2a+stx2c+ isolates were 90–100% lethal for Str-treated BALB/c mice, though one isolate, JH2013, had a delayed time-to-death. The stx2a+ isolate was avirulent. Both an stx2a and a recA deletion mutant of one of the stx2a+stx2c+ strains, JH2010, exhibited at least a three-log decrease in cytotoxicity in vitro and both were avirulent in the mice. Stool from Str-treated mice infected with the highly virulent isolates were 10- to 100-fold more cytotoxic than feces from mice infected with the clinical isolate, JH2012, that made only Stx2a. Taken together these findings demonstrate that the stx2a-phage from JH2010 induces to higher levels in vivo than does the phage from JH2012. The stx1a+stx2a+ clinical isolates were avirulent and neutralization of Stx1 in stool from mice infected with those strains indicated that the toxin produced in vivo was primarily Stx1a. Treatment of mice infected with Stx1a+Stx2a+ isolates with Cip resulted in an increase in Stx2a production in vivo and lethality in the mice. Our data suggest that high levels of Stx2a in stool are predictive of virulence in mice.

To discern if there was a particular genotype associated with clinical enteroaggregative Escherichia coli (EAEC) strains isolated from deployed military personnel (DMP) with travelers' diarrhea (TD), we characterized a collection of EAEC from DMP deployed to Afghanistan, Djibouti, Kenya, or Honduras. Although we did not identify a specific EAEC genotype associated with TD in DMP, we found that EAEC isolated at the first clinic visit were more likely to encode the dispersin gene aap than EAEC collected at follow-up visits. A majority of the EAEC isolates were typical EAEC that adhered to HEp-2 cells, formed biofilms, and harbored genes for aggregative adherence fimbriae (AAF), AggR, and serine protease autotransporters of Enterobacteriaceae (SPATEs). A separate subset of the EAEC had aggR and genes for SPATEs but encoded a gene highly homologous to that for CS22, a fimbriae more commonly found in enterotoxigenic E. coli. None of these CS22-encoding EAEC formed biofilms in vitro or adhered to HEp-2 cells. Whole genome sequence and single nucleotide polymorphism analyses demonstrated that most of the strains were genetically diverse, but that a few were closely related. Isolation of these related strains occurred within days to more than a year apart, a finding that suggests a persistent source and genomic stability. In an ampicillin-treated mouse model we found that an agg4A+ aar- isolate formed a biofilm in the intestine and caused reduced weight gain in mice, whereas a strain that did not form an in vivo biofilm caused no morbidity. Our diverse strain collection from DMP displays the heterogeneity of EAEC strains isolated from human patients, and our mouse model of infection indicated the genotype agg4A+ aar– and/or capacity to form biofilm in vivo may correlate to disease severity.

In this study we compared nine Shiga toxin (Stx)-producing Escherichia coli O157:H7 patient isolates for Stx levels, stx-phage insertion site(s), and pathogenicity in a streptomycin (Str)-treated mouse model. The strains encoded stx2a, stx1a and stx2a, or stx2a and stx2c. All of the strains elaborated 105-106 cytotoxic doses 50% (CD50) into the supernatant after growth in vitro as measured on Vero cells, and showed variable levels of increased toxin production after growth with sub-inhibitory levels of ciprofloxacin (Cip). The stx2a+stx2c+ isolates were 90–100% lethal for Str-treated BALB/c mice, though one isolate, JH2013, had a delayed time-to-death. The stx2a+ isolate was avirulent. Both an stx2a and a recA deletion mutant of one of the stx2a+stx2c+ strains, JH2010, exhibited at least a three-log decrease in cytotoxicity in vitro and both were avirulent in the mice. Stool from Str-treated mice infected with the highly virulent isolates were 10- to 100-fold more cytotoxic than feces from mice infected with the clinical isolate, JH2012, that made only Stx2a. Taken together these findings demonstrate that the stx2a-phage from JH2010 induces to higher levels in vivo than does the phage from JH2012. The stx1a+stx2a+ clinical isolates were avirulent and neutralization of Stx1 in stool from mice infected with those strains indicated that the toxin produced in vivo was primarily Stx1a. Treatment of mice infected with Stx1a+Stx2a+ isolates with Cip resulted in an increase in Stx2a production in vivo and lethality in the mice. Our data suggest that high levels of Stx2a in stool are predictive of virulence in mice.