Targeted Disruption of CSL Ligand - Host Cell Receptor Interaction in Treatment of Cryptosporidium parvum Infection

The apicomplexan parasite Cryptosporidium parvum is a leading cause of diarrhea in humans, calves, and other agriculturally important food animals, which can progress to chronic disseminated infection in the immunocompromised host [14,27]. While significant progress has been made [5,9,10,15,23,27,38], there are presently no approved, consistently effective parasite-specific drugs or vaccines for cryptosporidiosis. Specific immune responses prevent or terminate C. parvum infection in immunocompetent hosts [27]. Thus, passive immunization with polyclonal antibodies prepared against whole C. parvum organisms has been investigated for prevention or treatment of cryptosporidiosis in neonatal and immunocompromised hosts, as well as immunocompetent hosts [10,13,21,22,27,28,35]. These studies have met with variable success.

Our long-term hypothesis has been that optimal passive immunization can be achieved by use of neutralizing monoclonal antibodies (MAbs) having high specific activity for selectively targeting antigens known to have an essential role in the parasite life cycle. One such antigen, CSL, is an -1,300 kDa conserved glycoprotein contained in the dense granule and microneme apical complex organelles of the infective sporozoite and merozoite stages of C. parvum [32]. We recently defined the role of CSL in the infection process, demonstrating that it meets biochemical and functional definitions necessary for designation as a ligand- containing parasite glycoprotein [18]. In brief, isolated native CSL was shown to bind specifically and with high affinity, in a dose-dependent, saturable, and self-displaceable manner to intestinal epithelial cells. Isolated CSL was also capable of competitively inhibiting sporozoite attachment to intestinal epithelial cells in vitro via host receptor occupancy. Most recently, initial characterization of the receptor recognized by CSL has been reported, and its expression by cells of epithelial lineage shown to parallel their greater permissiveness to C. parvum than cells of mesenchymal origin [19]. MAb 3E2, originally used to define CSL [32], was shown to neutralize sporozoite infectivity in vitro [18,32]. When evaluated in vivo, prophylactic administration of 3E2 to oocyst-challenged neonatal ICR mice reduced intestinal infection levels by 62–92% and provided complete protection to some mice [34]. Indeed, the neutralizing activity of 3E2 in vitro and its prophylactic efficacy in neonatal mice are profoundly greater than that of all other MAbs we have produced [12,29,34] against multiple, distinct epitopes of defined C. parvum antigens, including P23 [1,6,12,24], GP25–200 [1], and CPS-500 [30]. Based on the optimal prophylactic efficacy of MAb 3E2 observed in neonatal ICR mice, the aim of the present study was to evaluate its therapeutic efficacy against established infection in an adult model of persistent cryptosporidiosis. For these experiments, IFN-γ-depleted SCID mice, in which C. parvum infection had progressed to the chronic disseminated stage prior to initiation of treatment, were used.

MATERIALS AND METHODS

Oocysts of the Iowa C. parvum isolate were obtained from Pleasant Hill Farm (Troy, ID) for use in all experiments following isolation from experimentally infected newborn Cryptosporidium-free calves as previously described [31].

Immunoaffinity chromatography purification of native CSL from C. parvum sporozoites, and use for production of a panel of mouse MAbs from which 3E2 was identified, have been previously described [32,34]. MAb 3E2 and mouse IgM isotype control MAb of irrelevant specificity were produced by growing hybridomas in a hollow fiber bioreactor system using serum free medium. Prior to use, the MAbs were dialyzed against PBS and concentrated by ultrafiltration. MAb concentrations were determined by radial immunodiffusion.

Adult female C.B.17/Icr Tac-SCID mice (Taconic, Germantown, NY) used to evaluate the therapeutic efficacy of 3E2 were housed in sterile microisolator cages with autoclaved food and water throughout the experiment. At four weeks of age, mice were injected intraperitoneally, once daily for two days, with 10⁵ neutralizing units of purified rat antimouse IFN-γ MAb R46A2 [8]. At five weeks of age, each mouse was inoculated with 10⁷ oocysts by gastric intubation using a 22 gauge feeding tube. Four weeks later, fecal pellets were collected from individual mice to confirm persistent infection and to determine the number of oocysts/mg feces using previously described methods [26,28]. Persistently infected mice were then randomly assigned to either the MAb 3E2 or the isotype control MAb treatment group, each group containing seven mice shedding oocysts above the median and seven mice shedding oocysts below the median number/mg feces. Beginning at 39 days after oocyst inoculation, all mice were empirically treated with cimetidine (60 mg/kg per os) at 12 h intervals for a total of three treatments. Twelve h after the third treatment, mice were administered either MAb 3E2 or isotype control MAb by gastric intubation. Each MAb was prepared at a final concentration of 5 mg/ml in PBS containing 1 % (wt/vol) ovalbumin and cimetidine, and administered in a volume of 250 μl per mouse. Mice were treated identically with the MAb preparations every 12 h thereafter for 21 days, and euthanized by CO₂ asphyxiation 10 h after the final treatment. Sections of the gall bladder/common bile duct junction, pylorus, terminal ileum, cecum, and proximal colon were then collected from the same anatomic site in each mouse and processed for histopathology [26]. Slides were coded and examined histologically without investigator knowledge of treatment group to quantitate C. parvum stages in mucosal epithelium. Scores of 0, 1, 2, or 3 (0, no infection; 1, < 33% of mucosa infected; 2, 33 to 66% of mucosa infected; and 3, > 66% of mucosa infected) were assigned to longitudinal sections from each tissue using a standardized method as previously described [28,29,31]. Scores from equivalent lengths of the terminal ileum, cecum, and proximal colon for each mouse were summed to obtain an intestinal infection score [26,28]. Mean tissue infection scores were evaluated for significant differences by analysis of variance using the General Linear Models Program of SAS (SAS Institute Inc., Gary, N.C.).

RESULTS AND DISCUSSION

Passive immunotherapy with otCSL MAb 3E2 in chronically infected SCID mice significantly reduced C. parvum levels in the intestinal tract (Table 1), indicating that pharmacologically active levels of orally administered MAb can be maintained in the adult intestinal lumen and mediate neutralization of C. parvum. This finding parallels the previously observed ability of MAb 3E2 to passively protect neonatal mice against oocyst challenge [34]; however, the reduction in parasite burden achieved was substantially lower in adult SCID mice. While infection levels in the pylorus and biliary tract in 3E2-treated mice were also lower than isotype control MAb-treated mice, the differences were not statistically significant (Table 1).

These observations most likely reflect the fundamental differences in evaluating therapeutic efficacy of MAb administered after infection becomes persistent and disseminated in an immunodeficient adult model vs. prophylactic efficacy of MAb against primary infection by oocyst challenge in a neonatal model. Adult SCID mice, depleted of IFN-γ to augment infection, provide a stringent model of persistent cryptosporidiosis for therapeutic evaluation [8]. Unlike neonatal mice, adult SCID mice have a mature gastrointestinal system and develop severe suppurative enterocolitis during chronic infection [28]. The observed lack of efficacy against gastric infection is most likely related to the adult gastric environment, where acidic conditions could effectively reduce or negate the affinity of 3E2-CSL binding. Finally, chronic intestinal infection in SCID mice may also spread to the hepatobiliary tract. This site can then serve as a reservoir for maintenance of C. parvum infection that is sheltered from orally administered MAb [26,28,36].

Passive immunization with locally delivered antibody has been shown to be efficacious in preventing or treating infection by a variety of enteropathogens [39,40]. Its feasibility was extended for use in cryptosporidiosis by results of early studies to evaluate antibodies produced against uncharacterized whole C. parvum organism preparations [3,13,21,26,28,35]. In the present study, we reasoned that passive immunotherapy against chronic C. parvum infection could be improved by using neutralizing MAb prepared against CSL, an apical glycoprotein known to function as a zoite ligand during infection [18,32]. Because surface pellicle and exocytosed apical organelle antigens such as CSL are essential for infection, and their exposure also allows binding of neutralizing antibody, they offer opportune targets for intervention against C. parvum [2,4,7,16,17,20,25,29,30,33,34] as well as other apicomplexan parasites [11]. However, limited knowledge on the specific mechanisms of antibody-mediated neutralization of C. parvum and the pathogenesis of attachment and invasion has hampered development of immunologic control strategies and drug discovery [10,27,37]. The mechanism of neutralization by 3E2 involves binding to a repetitive, carbohydrate-dependent epitope on CSL, occurrence of the circumsporozoite precipitate-like reaction, and release of CSL-antibody precipitates from the sporozoite surface [32]. Sporozoites having undergone this reaction are no longer capable of attaching to host cells [18]. It is possible that C. parvum has redundant mechanisms for infection of different host cell types, and probable that parasite stages use more than one ligand in a multistep attachment and invasion process [18,32,37]. In any case, MAb 3E2-mediated disruption of CSL ligand-host cell receptor binding could explain the biological basis for the therapeutic efficacy observed in the present study, and the prophylactic efficacy reported previously [34]. Therefore, the rationale for further characterizing the active ligand moiety(ies) contained in CSL is evident. To this end, preliminary studies have been performed to determine the glycosyl composition of purified native CSL, and the linkages and anomeric configurations of residues comprising the epitope recognized by MAb 3E2. Preliminary experiments to determine the molecular weight of the CSL protein backbone following N- and O- deglycosylation have also been performed (Riggs, M. W. & McNeil, M. R., unpublished data). Continuing studies to determine the complete molecular structure of functional glycan and/or peptide domains in CSL are expected to lead to new immunologic avenues for disrupting ligand-receptor interactions in the pathogenesis of cryptosporidiosis, as well as host receptor-based therapeutic approaches using chemically synthesized ligand antagonists.

ACKNOWLEDGMENTS

This work was supported by Public Health Service Grant AI 30223 from the National Institutes of Health, Bethesda, MD, and United States Department of Agriculture NRICGP Grants 94–37204–0496, 99– 35204–8533, and 2001–35204–09960.

We thank Deborah Dalton, Beth Auerbach-Dixon, Kathryn Huey Tubman, Michael S. Scherman, and Jessica Chodos for excellent technical assistance, and E. Havell for providing R46A2 hybridoma cells as a source of rat anti-mouse IFN-γ antibody.