Are corneal cells susceptible to antibody-mediated killing in corneal allograft rejection?

Purpose: The precise role of antibodies in corneal transplantation is controversial. Clinical and experimental evidence both supports and refutes the contribution of donor-derived alloantibody in corneal allograft rejection. Accordingly, we prospectively evaluated the presence of donor-derived alloantibody in two high-risk donor-host combinations. We also evaluated the ability of this alloantibody to kill corneal epithelial, keratocytes, and endothelial cells in complement-dependent and complement-independent fashions. Methods: C3H/Hej (H-2^k) and Balb/c (H-2^d) corneal grafts were transplanted orthotopically to CB6F1 (H-2^b/d) and C57BL/6 (H-2^b) recipients, respectively. These two donor-host combinations represent disparity at the entire MHC and multiple minor histocompatibility loci. Four objectives were addressed. First, we wished to determine if there was a correlation between the appearance of donor-specific serum IgG antibody and corneal graft rejection. Second, we evaluated the effect of passive transfer of hyperimmune donor-specific antibody on corneal allograft rejection. Third, we examined the capacity of donor-specific alloantibody to mediate complement-dependent cytolysis of corneal cells. Finally, we determined the capability of donor-specific alloantibody to mediate apoptosis of corneal cells. Results: The presence of donor-specific serum IgG alloantibodies did not correlate with corneal graft rejection. One hundred percent of CB6F1 and C57BL/6 hosts rejected their C3H and Balb/c orthotopic corneal allografts, respectively. However, two of these seven CB6F1 hosts and one C57BL/6 host did not produce donor-specific IgG alloantibody that was significantly different from naive donors. Passive transfer of hyperimmune allo-antiserum prior to corneal transplantation did not increase the incidence, severity, or tempo of corneal allograft rejection in either donor-host combination. Hyperimmune allo-antiserum produced complement-mediated lysis of C3H corneal endothelial but not C3H corneal epithelial cells in the C3H-CB6F1 donor-host combination. Interestingly, all three corneal cell layers were vulnerable to complement-mediated cytolysis in the Balb/c-C57BL/6 donor-host combination. Additionally, Balb/c corneal epithelial, keratocytes, and endothelial cells were vulnerable to complement-independent, antibody induced apoptosis. Conclusions: Corneal graft rejection does not appear to correlate with the production of IgG alloantibody and can occur in the absence of donor-specific IgG alloantibody. Antibody-mediated killing of the corneal endothelium can occur in a complement-dependent or complement-independent fashion.