Search

A murine pulmonary model was used to study the mucosal immune response to Shigella flexneri serotype 2a infection. Inoculation of BALB/cJ mice with shigellae via the intranasal route resulted in bacterial invasion of bronchial and alveolar epithelia with concomitant development of acute suppurative bronchiolitis and subsequent development of lethal pneumonia. The pathology of pulmonary lesions resembled the colitis that characterizes shigellosis in humans and primates. Significant protection against a lethal dose of S. flexneri 2a was observed in mice previously infected with two sublethal doses of the homologous strain. Immunity against lethal challenge was associated with decreased bacterial invasion of the mucosal epithelium. Over the course of two sublethal challenges, which constituted primary and secondary immunizations, mice developed pulmonary and serum immunoglobulin G and A antibody recognizing both lipopolysaccharide and invasion plasmid antigens IpaB and IpaC. Immune mice and naive control mice differed in lung lavage cytokine levels following lethal challenge. Immune mice developed significantly elevated levels of pulmonary gamma interferon within 6 h of challenge, while naive control mice developed elevated levels of this cytokine later during the initial 24-h period. Both groups had elevated levels of gamma interferon during the 24- to 48-h period of infection. Both groups also had elevated levels of tumor necrosis factor alpha within 6 h of challenge, but the control mice had significantly higher levels at the 48- and 72-h time points. Elevated levels of interleukin-4 were observed only in immunized mice. This cytokine appeared within 24 h and receded between 48 and 72 h. Fluorescence-activated cell sorter analysis of lung parenchymal cells showed that both groups experienced an initial influx of monocytes, but the proportion of this cell type began to recede in immunized mice after 48 h of infection, while peak levels were maintained in the control animals. These studies suggest that elements of local B lymphocyte activity, as well as Th1 and Th2 lymphocyte activity, may contribute to the survival of immune mice after intranasal challenge with shigellae.

This study used the guinea pig keratoconjunctivitis model to examine the importance of route of administration (mucosal versus parenteral), frequency and timing of immunization (primary versus boosting immunization), and form of antigen given (live attenuated vaccine strain versus O-antigen-protein conjugate) on the production of protective immunity against Shigella infection. Since local immune response to the lipopolysaccharide (LPS) O-antigen of Shigella spp. is thought to be important for protection against disease, O-antigen-specific antibody-secreting cells (ASC) in the spleen and regional lymph nodes of immunized animals were measured by using an ELISPOT assay. Results indicated that protective efficacy was associated with a strong O-antigen-specific ASC response, particularly in the superficial ventral cervical lymph nodes draining the conjunctivae. In naive animals, a strong ASC response in the cervical lymph nodes and protection against challenge were detected only in animals that received a mucosal immunization. Protection in these animals was increased by a boosting mucosal immunization. While parenteral immunization alone with an O-antigen-protein conjugate vaccine did not protect naive animals against challenge, a combined parenteral-mucosal regimen elicited enhanced protection without the addition of a boosting immunization. Although O-antigen-specific serum immunoglobulin A titers were significantly higher in animals receiving a mucosal immunization, there was no apparent correlation between levels of serum antibody and protection against disease.

A 140-megadalton plasmid (pWR110), which has previously been associated with virulence in Shigella flexneri, was transferred to Escherichia coli K-12. Segments of S. flexneri chromosomal material were then transferred to the plamid-bearing K-12 strains. The virulence of these transconjugant hybrids was assessed in the HeLa cell model, in ligated rabbit ileal loops, or in the Sereny test. A K-12 strain which harbored only pWR110 invaded HeLa cells, produced minimal lesions in the rabbit ileal mucosa, and was negative in the Sereny test. Plasmid-containing K-12 hybrids which had incorporated various shigella chromosomal regions gave differential reactions in the rabbit ileal loops and in the Sereny test. Analysis of these transconjugants indicated that three regions were linked with virulent phenotypes. These included the his region (when the genes responsible for O-antigen synthesis were cotransferred) and the kcp locus (linked to the lac-gal region). Either of these chromosomal regions was sufficient to allow invasion of the rabbit ileal mucosa. In addition to both of these regions, another shigella chromosomal segment linked to the arg and mtl loci was necessary for fluid production in the rabbit ileal loop and for a positive Sereny reaction. Thus, derivatives of an E. coli K-12 strain, constructed by the stepwise conjugal transfer of a large plasmid and three chromosomal segments from S. flexneri, appeared to contain the necessary determinants for full pathogenicity in a variety of laboratory models.

Salmonella typhi 5076-IC, which contains a plasmid that encodes the form I antigen of Shigella sonnei and which expresses S. typhi 9 and 12 and S. sonnei form I antigens, was used to immunize rabbits via chronically isolated ileal loops. Intestinal immunoglobulin A activity was detected against S. typhi, S. sonnei form I, and S. typhi strain 5076-IC. Thus S. typhi 5076-IC can effectively elicit mucosal immunoglobulin A to both S. typhi and S. sonnei.

Several variables inherent in chronically isolated ileal (Thiry-Vella) loops in rabbits were studied for their effect on the local immune response of the intestine to live, locally invasive bacteria ( Shigella X16). A much more vigorous local immunoglobulin A response to Shigella X16 was elicited when rabbits were immunized in their Thiry-Vella loops shortly after surgical creation of the loop than if a week were allowed to pass before they were immunized. Three major differences existed in Thiry-Vella loops on the day after surgery and a week later: (i) their microbial flora, (ii) nonspecific acute inflammation due to the surgery itself, and (iii) the histological appearance of the intestine. On day 1 after surgical creation of the Thiry-Vella loop, there were few bacteria in the loop, and the histology was that of normal small bowel except for mild acute inflammation due to the surgery. By day 6 after surgery, all loops contained large numbers of Pseudomonas aeruginosa and other aerobes, an atrophy of intestinal epithelium occurred, and the acute inflammation due to surgical trauma had subsided. By artificially colonizing Thiry-Vella loops with 10 8 or 10 10 live P. aeruginosa on the day of surgery, we found that the presence of these bacteria alone did not greatly diminish local immune responses to live Shigella . Furthermore, when the acute inflammation due to surgical trauma was recreated in loops 6 days old, no enhancement of the immune response was seen as compared to nontraumatized 6-day-old Thiry-Vella loops. The difference between immunization soon after surgery and a week later related to changes that occur in the loop itself with increased isolation. Finally, multiple immunizations of Thiry-Vella loops resulted in a more vigorous local immunoglobulin A response than a single immunization. These studies demonstrated that Thiry-Vella loop models can be useful in studying the kinetics of local immune responses by the intestine only if careful attention is paid to key variables inherent in the Thiry-Vella loop models themselves.

One major stumbling block in the development of an effective means to immunize against shigellosis and other enteric diseases has been the lack of a means to assess sequential mucosal immune responses to different potential immunogens. In the present study, we compared the abilities of live invasive organisms, noninvasive organisms, and nonviable antigen preparations of shigella to elicit mucosal immune responses. Whereas previous studies have found that effective immunity was produced best by vaccination with live invasive strains of shigella, in the present study, live noninvasive strains that did not produce any histopathological damage were consistently able to produce local (immunoglobulin A) immune responses as vigorous as those of the invasive strains. Further, acetone-killed shigella antigen was also an effective mucosal immunogen, whereas hot phenol-water-extracted shigella lipopolysaccharide was ineffective, possibly due to the method of preparation. A single oral or parenteral priming was ineffective in enhancing the mucosal immune response when restimulated 1 month later with the same antigen. However, a mucosal memory response was found to be present several months after a triple mucosal stimulation with a locally invasive vaccine strain.

In spite of several macroscopic criteria for predicting the presence of histological abnormalities in rabbit ileum, microscopic ileal abnormalities still can escape detection. The effect of histologically abnormal rabbit ileum was evaluated on basal intestinal absorption, on basal absorption, on basal adenylate cyclase activity, and on cholera toxin-induced secretion and cholera toxin-induced stimulation of adenylate cyclase activity. Compared to histologically normal rabbit ileum, the presence of histological abnormalities was associated with decreased basal intestinal water, Na, Cl, and glucose absorption, absent glucose-dependent water absorption, and elevated basal adenylate cyclase activities. However, histologically abnormal rabbit ileum responded to inoculation of purified cholera toxin with stimulation of intestinal water secretion and adenylate cyclase activity similar to that in histologically normal ileum. These data have implications concerning the design of experiments that attempt to study the pathogenesis of diarrheal diseases by correlating changes in ileal transport with changes in ileal mucosal adenylate cyclase activity. In spite of abnormal ileal histology, studies of intestinal secretory states which attempt to define the role of adenylate cyclase in secretory processes can be performed provided animals are used as their own controls. However, when groups of animals are compared, the presence of an histologically abnormal ileum can cause changes in basal and intestinal secretagogue-stimulated ileal water and electrolyte transport and in basal and intestinal secretagogue-stimulated mucosal adenylate cyclase activity which can lead to erroneous conclusions if the presence of the abnormal ileal histology is not considered.