Partially assembled K99 fimbriae are required for protection

Abstract

Antibodies to K99 fimbriae afford protection to F5+ bovine enterotoxigenic Escherichia coli (ETEC). Previous studies show that murine dams immunized with Salmonella vaccine vectors stably expressing K99 fimbriae confer protection to ETEC-challenged neonatal pups. To begin to address adaptation of the K99 scaffold to display heterologous B- and T-cell epitopes, studies were conducted to determine how much of the assembled K99 fimbria is required to maintain protective immunity. Sequential deletions in the K99 gene clusters were made, resulting in diminished localization of the K99 fimbrial subunit in the outer membrane. As placement of the K99 fimbrial subunit became progressively contained within the vaccine vector, diminished immunoglobulin A (IgA) and IgG1 antibody titers, as well as diminished Th2-type cytokine responses, were observed in orally immunized mice. Deletion of fanGH, which greatly reduced the export of the fimbrial subunit to the outer membrane, showed only partial reduction in protective immunity. By contrast, deletion of fanDEFGH, which also reduced the export of the fimbrial subunit to the outer membrane but retained more subunit in the cytoplasm, resulted in protective immunity being dramatically reduced. Thus, these studies showed that retention of K99 fimbrial subunit as native fimbriae or with the deletion of fanGH is sufficient to confer protection.

FIG. 1.

(A) Expression of K99 fimbriae by Salmonella vaccine vectors as examined by IEM at 12,000×. IEM of the AP112 strain showed extracellular expression of K99 fimbrial antigen stained with rabbit anti-K99 fimbriae antibody and gold-anti-rabbit antibody. IEM of the AP114 strain sectioned at 90 nm showed periplasmic expression of K99 subunit (note gold particles as black dots along the circumference of the Salmonella cells). IEM of AP116 strain sectioned at 90 nm showed periplasmic and cytoplasmic expression of K99 subunit (note gold particles localized within the Salmonella cells). (B) Differential expression of the K99 fimbrial subunit as a consequence of deletions in the K99 genes. Depicted are Western blots of K99 fimbrial subunit from strains AP112, AP114, and AP116 and compared to migration of purified K99 subunit. Compared to AP112 by densitometric measurements of Western blots from two experiments, the K99 fimbrial subunit was reduced in AP114 (*, P =0.027) and AP116 (*, P =0.004) outer membrane (OM) fractions, but these strains retained expression in cytoplasmic (C), inner membrane (IM), and periplasmic (P) fractions; however, AP114 C fractions did show significant reduction compared to AP112 (*, P =0.013) and to AP116 (¶, P =0.021). (C) Quantification of K99 antigen in homogenized cells from the different Salmonella strains using a capture-ELISA method (see Materials and Methods). The various Salmonella constructs underwent extraction, and protein concentrations were normalized to equivalence. Depicted are the means (ng of K99 fimbria/ml/mg total protein) ± SD of the extrapolated values obtained from standard curves obtained with purified K99 fimbriae. *, P < 0.001, **, P < 0.024 for amount of K99 fimbria in AP112 versus AP114, AP116, or H647. There was no significant difference in K99 fimbria content between AP114 and AP116.

FIG. 2.

K99 subunit-specific IgA and IgG responses by mice immunized with Salmonella-K99 vaccines. (A) Depicted are the means ± standard deviation of K99 fimbria-specific IgA log₂ endpoint titers in colostrum, serum, and fecal samples that developed 6 weeks after oral immunization of CD-1 mice with 5 × 10⁹ CFU of each Salmonella construct. The greatest IgA antibody titers were found in colostrum followed by serum and fecal samples, and the greatest titers for K99 fimbria were induced by AP112 strain followed by AP114 and AP116 strains. *P < 0.001 for IgA levels by AP112- versus AP114-, AP116-, or H647-vaccinated mice. (B) Depicted are the means ± standard deviation of K99 antigen-specific IgG and IgG subclasses log₂ endpoint titers in serum that developed 6 weeks after oral immunization of same mice with the described Salmonella construct. The greatest antibody titers were induced by the AP112 strain followed by the AP114 and AP116 strains. For mice orally immunized with strain AP112, elevated IgG1, IgG2a, and IgG2b anti-K99 fimbrial titers were induced. As progressive deletions in the transport of the K99 fimbrial subunit were sequentially reduced, progressive decreases in the ratio and magnitude of IgG1 and IgG2b titers were observed. *, P < 0.001 for IgG and IgG subclass responses by AP112- versus AP114-, AP116-, and H647-vaccinated mice.

FIG. 3.

Oral vaccination with AP112 or AP114 induces elevated mucosal Th2 cell responses. Depicted are cytokine responses of isolated lymphocytes following oral immunization of BALB/c mice with each Salmonella-K99 construct, and data are expressed as the means ± standard deviation for (A) Peyer's patches and (B) splenic cytokine-forming cell (CFC) responses. Mice orally immunized with strains AP112 or AP114 showed elevated IL-4, IL-5, and IL-6 CFC in Peyer's patches compared to mice vaccinated with AP116. Different CFC responses were observed in the spleen. Both AP112 and AP116 showed elevations in IL-6 CFC, but IL-4 CFC were significantly greater in AP112-vaccinated mice compared to all other vaccine groups. Minimal IFN-γ responses were detected in either tissue. These results showed that Salmonella strains AP112 and AP114 induced predominantly fimbria-specific mucosal Th2 cell responses. Data are representative of three experiments. *, P < 0.001; **, P = 0.004.

FIG. 4.

Passive immune protection of pups orally challenged with the wild-type ETEC strain. Dams orally immunized with Salmonella strain AP112 provided the best passive immunity to wild-type ETEC-challenged pups, as demonstrated by ∼90% survival (n = 71), while the passive protection diminished in mice orally vaccinated with either the AP114 strain (79% survival, n = 68) or the AP116 strain (30% survival, n = 73). In contrast, pups from dams vaccinated with the Salmonella control strain H647 showed a survival rate of only 6% (n = 69). *, P < 0.001.