Progress 09/01/99 to 08/31/04
Outputs
Enterotoxigenic Escherichia coli (ETEC) is a common cause of diarrhea in neonatal and weaned pigs and bacterial attachment to enterocytes is the first step in the pathogenesis of infection by ETEC. Bacterial attachment is mediated by fimbrial adhesins with K88 fimbriae being the most common. The three serological variants of K88 (K88ab, K88ac, and K88ad) differ from each other in their enterocyte binding specificities and binding specificity appears to affect virulence. Previously we showed that K88ab and K88ac bind an intestinal mucin-type glycoprotein while K88ad binds an intestinal neutral glucosphingolipid. Differences in binding specificity must be accounted for in the major fimbrial subunit proteins as other subunit proteins do not differ between variants. We showed that K88-varient specific monoclonal antibodies (MAbs, or their Fab fragments) block binding of E. coli expressing the variant to brush borders containing a receptor recognized by the variant. By
contrast, MAbs (or Fabs) that recognize all variants failed to block fimbriae binding. This suggests that variant-specific MAbs recognize fimbrial receptor binding domains. In one study, we showed that the minimal peptide recognized by one K88ac variant-specific MAb was a stretch of amino acids from aa64 to aa107. This region contains 8 amino acids that differ between K88 variants and may account for their receptor binding specificity. We have also conducted a number of amino acid substitution studies. Substitutions of Thr-74, Val-94, Ser-172, or Glu-152 of K88ac with Val-74, Glu-94, Pro-172, or Asp-152 of K88ad or visa versa resulted in loss of receptor-binding, suggesting that these amino acids might contribute to the receptor binding domain. In further efforts to identify regions of the K88 binding domain, we constructed chimeric FaeG genes. Complementary regions of the FaeG gene of K88ac and K88ad were PCR amplified, hybridized by overlap PCR to produce K88ac/ad chimeric FaeG
genes. The chimeras were cloned into plasmid 8069 with its FaeG gene removed. This plasmid was derived from pDB88-102p by adding unique restriction sites at the flanking region of the FaeG gene. The constructed plasmids with chimeric K88ac and K88ad FaeG genes were transformed into a non-adherent E. coli strain, which subsequently was tested for binding specificity. Bacterial adherence assays utilizing porcine brush borders adhesive for K88ac and /or K88ad suggested that the region from amino acid 133 to aa156 of the FaeG gene was essential for K88 variant-specific binding. The substitution of the region of the K88ac FaeG gene from amino acid 133 to 156, with that of K88ad of the FaeG gene resulted in the expression of hybrid fimbriae that bound to porcine brush borders that are recognized by K88ad fimbriae. Switch of the same region of K88ad with that of K88ac resulted fimbriae that bound brush borders typically recognized by K88ac fimbrie. Combined, the observations of our studies
suggest that several regions of the FaeG genes are likely involved in receptor recognition, with the region from aa133 to 156 imparting variant receptor specificity.
Impacts
Identification of neutrailzing epitopes of K88 fimbriae will facilitate development of highly effective subunit vaccines for post-weaning E. coli diarrheal disease. A greater understanding the evolution of the K88 adhesin may eventually help in predicting the rise of future pathogens.
Publications
- Zhang, W., Y. Fang and D. Francis. 2004. Characterization of K88ac fimbria binding specificity by using K88ac/ad chimeric FaeG genes. Proceedings, 85th Annual Meeting, Conference of Research Workers in Animal Diseases.
|
Progress 01/01/03 to 12/31/03
Outputs
Enterotoxigenic Escherichia coli (ETEC) is a common cause of diarrhea in neonatal and newly weaned pigs and bacterial attachment to intestinal epithelium is the first step in the pathogenesis of infection by ETEC. Bacterial attachment is mediated by fimbrial adhesins with K88 fimbriae being the most common. The three serological variants of K88 (K88ab, K88ac, and K88ad) differ from each other in their enterocyte binding specificities and binding specificity appears to affect virulence. Previously we showed that K88ab and K88ac bind an intestinal mucin-type glycoprotein (IMTGP) while K88ad binds an intestinal neutral glucosphingolipid (IGLad). Any differences in binding specificity of K88 variants must be accounted for in the major fimbrial subunit proteins as other subunit proteins do not differ between variants. K88-varient specific monoclonal antibodies (MAbs) prepared against K88ab, K88ac and K88ad (or their Fab fragments) were shown to block binding of E. coli
expressing the variant to brush borders containing a receptor recognized by the variant. By contrast, MAbs (or Fabs) that recognize all variants failed to block fimbriae binding. This suggests that variant-specific MAbs recognize fimbrial receptor binding domains. The minimal peptide recognized by one K88ac variant-specific MAb was a stretch of amino acids from no.64 to no.107. This region contains 8 amino acids that differ between K88 variants and may account for their receptor binding specificity. We substituted one of the 8 amino acids, Val-94 in the cloned faeG gene of K88ac with Glu-94 (of K88ad). The switch resulted in loss of the ability of the expressed fimbriae to bind to porcine brush borders and fimbrial recognition by the neutralizing MAb. These results suggest that Val-94 plays a key role in the formation of the receptor-binding epitope of K88ac. However, because the substitution did not switch receptor specificity from that of K88ac to that of K88ad, it alone cannot
account for the receptor recognition differences between the two K88 variants. We also replaced the Thr-74 in the faeG gene of K88ac with Val-74, characteristic of K88ad, with the same outcomes observed with AA-94. In analogous studies with K88ad, Proline-172 was substituted with Serine, characteristic of K88ac, and Alanine 218 with Glutamine also characteristic of K88ac. The first substitution extinguished the receptor binding activity of the fimbria to porcine brush borders, and resulted in the loss of recognition by 1 of 2 K88ad-specific MAbs. Failure to the mutation to facilitate recognition of the modified fimbriae by K88ac MAb suggested that the mutant did not acquire K88ac antigenicity. Further characterization of the Pro172 to Ser172 and the Ala218 to Glu218 mutants of K88ad is in progress.
Impacts
Identification of neutrailzing epitopes of K88 fimbriae will facilitate development of highly effective subunit vaccines for post-weaning E. coli diarrheal disease. A greater understanding the evolution of the K88 adhesin may eventually help in predicting the rise of future pathogens.
Publications
- No publications reported this period
|
Progress 01/01/02 to 12/31/02
Outputs
Four, one-day-old gnotobiotic piglets of the A phenotype were alternatively inoculated with strain 3030-2 (O157:K88ac, LT, STb) or the Morris strain (O8:K88ad; LT, STb). Another 3 pigs served as negative controls and were inoculated with non-pathogenic E. coli G58-1. Pigs were observed for signs of disease for 18 hrs, then euthanized, after which specimens were collected to assess piglet phenotype and extent of bacterial colonization. Piglets inoculated with strain 3030-2 were severely affected by challenge, with one pig dying before the 18 hr experiment termination time, and two other pigs becoming markedly dehydrated. A fourth pig from that group exhibited diarrhea. Two of four pigs receiving the Morris strain developed diarrhea, one of which exhibited some dehydration. The other two pigs in the group exhibited fluid-filled colons, but no diarrhea. None of the pigs receiving G58-1 developed diarrhea or increased accumulation of colonic fluid. Epithelial colonization
of E. coli in sections of duodenum, jejunum (2 locations) and ileum of pigs challenged with 3030-2 ranged from complete coverage of all sections to 25% coverage in the ileum. Only rare attached bacteria were observed on any section of intestine taken from pigs challenged with the Morris strain. Bacterial attachment was not observed in tissue sections from pigs inoculated with the negative control strain. Bacteria grew to approximately 2 logs greater concentration in the ilea of pigs infected with 3030-2 when compared with those infected with the Morris strain. This lower level of colonization was also exhibited in impression smears of jejunal and ileal tissue subjected to immunofluorescence staining with anti-K88 antibodies. We subjectively tested for the translocation of bacteria by culturing liver tissue. No-strain-to-strain difference was observed. Differences in virulence exhibited between the 3030-2 and the Morris strain may be accounted for by differences in receptors targeted
by the two K88 variants. K88ac binds to a large mucin-type sialoglycoprotein (IMTGP), while K88ad binds to a small glycosphingolipid (IGLad). The large IMTGP likely projects through the glycocalyx, thus is relatively accessible to the fimbriated E. coli. By contrast the IGLad is sufficiently small that it likely is hidden below the glycocalyx, perhaps only being exposed on cells not exhibiting much of a glycocalyx, such as the dome epithelial cells covering Peyer's patches.
Impacts
The results of recent studies suggest that the size and structure of the receptor targeted by the enteric bacteria is critical to its function in bacterial colonization. The projection of large molecules like IMTGP through the relatively thick glycocalyx covering of intestinal epithelial cells may make them a preferred target for enteric bacteria. An understanding of this phenomenon may make predicting the characteristics of, and identifying the receptors for other enteric pathogens easier.
Publications
- No publications reported this period
|
Progress 01/01/01 to 12/31/01
Outputs
Previously, we established that an intestinal mucin-type glycoprotein (IMTGP) is the biologically relevant receptor for K88ac and K88ab fimbriae of enterotoxigenic E. coli (ETEC) and that the K88ad variant of the fimbria binds to a porcine enterocyte neutral glycosphingolipid (IGLad). Here we report that in contrast to K88ab and K88ac, the K88ad fimbrial variant predominantly recognizes membrane glycolipids. The receptor preference of the three K88 variants may explain other observations made by us that wild-type K88ab+ and K88ac+ ETEC caused severe diarrhea in gnotobiotic pigs, while wild-type K88ad+ strains did not. While less than half of the pigs inoculated with K88ad+ ETEC become severely ill, those that did, died with little or no evidence of diarrhea, suggesting that they had succumbed to septicemia. Large glycoproteins like IMTGP project out from enterocyte apical membranes, perhaps allowing for a loose tethering of fimbriated bacteria. The much smaller
membrane-imbedded glycolipids such as IGLad would be less accessible or inaccessible for bacterial attachment on an enterocyte do to the cells' relatively thick glycocalyx. However, they may be available on M-cells, which have a poorly developed glycocalyx and are a frequent portal for bacterial invasion.
Impacts
Observations made in this research help clarify how bacteria identify and attack host cells. Greater understanding of the strategies employed by bacteria in gaining access to host cells will enable us to develop countermeasures to prevent pathogenic bacteria form colonizing host animals.
Publications
- Grange, P.A., Mouricout, M.A., Levery, S.B., Francis, D.H. and Erickson, A.K. 2002. Evaluation of receptor binding specificity of Escherichia coli K88 (F4) fimbrial adhesin variants using porcine serum transferrin and glycosphingolipids as model receptors. Infect. Immun. Pending.
|
Progress 01/01/00 to 12/31/00
Outputs
Binding specificity of E. coli K88 fimbriae to lipid receptors on pig intestinal epithelial cells was carried out using thin-layer chromatography (TLC) analysis. K88ab+ and K88ac+ bacteria and purified fimbriae bound to monohexosylceramide, whereas K88ad+ E. coli bound lactosylceramide. In addition, K88ad specifically bound to glycolipids in the 4-5 sugar regions of the non-acid fraction. No such binding was shown after overlay assay with the other two adhesins. All variants bound to glycolipids in the 5-6 sugar region. No binding activity was detected in the acid fraction of the glycolipid. All bacterial adhesins bound phosphatidylethanolamine, phosphatidylcholine and phosphatidylglycerol, but not phosphatidylserine.
Impacts
This work contributes to our understanding of the binding activity of K88 fimbriae on glycoconjugates associated with enterocyte brush border membranes. While we have made much progress in understanding the mechanism of the binding of K88+ E. coli to porcine epithelial cells, the full sequence of events in this pathogenic event remains to be determined.
Publications
- Sun, R., T. J. Anderson, A. K. Erickson, E. A. Nelson, and D. H. Francis. 2000. Inhibition of the adhesion of E. coli K88ac fimbria to its receptor, IMTGP, by a monoclonal antibody directed against a variable domain of the fimbria. Infect. Immun. 68: 3509-3515.
|
Progress 01/01/99 to 12/31/99
Outputs
Three antigenic variants of the fimbrial adhesin K88 of enterotoxigenic Escherichia coli have been identified, K88ab, K88ac and K88ad. Each variant exhibits uniqueness of specificity with regard to hemagglutination of erythrocytes from various species, and enterocyte binding among pigs from different genetic backgrounds. This uniqueness in specificity appears to be entirely attributable to the major fimbrial subunit, as it is the only structural protein in the fimbria that differs between variants. Using monoclonal antibodies to map the major fimbrial subunit of K88ac, we identified the amino acid stretch from No. 64 to No. 129 as a significant contributor to receptor recognition. A monoclonal antibody that maps to this region is non-crossreactive with K88ab or K88ad, indicating that the region is antigenically unique, and may be responsible for the receptor specificity difference exhibited between K88 variants. The region contains eight amino acids that differ among
the three K88 variants, two of which (Thr-74 and Val-94) are unique to K88ac. To determine whether either of these are responsible for the switch in specificity exhibited between K88ac and the other variants, we constructed K88ac fimbriae containing either of the two amino acids corresponding to K88ad in the two positions in question (Glu-94 and Thr-74). Both of the constructs were expressed by a bacterial host strain, but neither altered fimbria was recognized by the K88ac-specific monoclonal antibody whose epitope is within the amino acid stretch from No. 64 to No. 129. In addition, neither of the modified fimbria bound to receptors typically recognized by K88ac or receptors typically recognized by K88ad. These results suggest a critical role of Thr-74 and Val-94 in K88ac receptor recognition. However, they provide no evidence that either of these amino acids is responsible for the unique receptor binding specificity of K88ac.
Impacts
This project improves general knowledge on how E. coli infection occurs, thus potentially provides better strategies for prevention and treatment. This work and that to follow on this project will show the evolutionary strategy of the pathogen in selection of a host, and may enable better prediction of future challenges to livestock and human beings that this organism may pose in the future.
Publications
- Francis D.H., A.K. Erickson, and P. A. Grange. 1999. K88 adhesins of Escherichia coli and their porcine enterocyte receptors. Adv. Exp. Med. Biol. 473: 147-154
|
|