Binding and activation of complement is a key effector mechanism by which therapeutic IgG antibodies mediate their anti-tumor effects and, in some cases, their systemic toxicity. Previously, it was unclear how cell-bound IgG activates complement as the Fc tail of the immunoglobulin molecule has a low binding affinity for C1, the first component of the classical pathway cascade. In this study, an international group of investigators led by Dr. Paul Parren of Genmab in the Netherlands has elucidated the molecular basis for this crucial antibody function, and their findings suggest an approach to enhancing the efficiency of complement activation by modifying the structure of IgG.
Building upon their prior crystallographic finding that anti-HIV IgG1 can pack as a hexameric ring, the investigators hypothesized that such a structure might match the six antibody-binding headpieces of C1q (the binding component of the multimeric C1 complex), thereby overcoming the low affinity of individual IgG molecules for C1q. They used a small peptide to block the key site in the Fc region that mediates interactions between multimeric IgG molecules and showed that this blockade caused a marked inhibition of complement-dependent cytotoxicity (CDC) produced by anti-CD20 and anti-CD38 IgG1 antibodies. Additional experiments used a genetic approach, introducing a variety of mutations in the multimerization interface in the Fc region of IgG to show that, at key points, single amino acid substitutions could substantially impair CDC, while antigen binding and C1q binding to IgG1 immobilized on a solid surface was unaffected. Most mutations outside the critical region were neutral in their effects on CDC, while a change at one site (E345R) enhanced CDC by 10-fold. This effect was observed for each of the four subtypes of IgG and was demonstrated for different antigen specificities including CD20 and CD38. Introducing three different enhancing mutations into IgG also resulted in hexamer formation in solution in equilibrium with the monomeric form. This molecule was capable of directly activating complement in human serum as demonstrated by C4d generation and was significantly more potent in CDC assays when bound to cells. The presence of hexameric IgG on a membrane was modeled on liposomes using cryo-electron tomography, which showed an electron-density cluster consistent with a six-molecule structure only when the gain-of-function E345R mutant IgG was used. The tomography studies also showed the presence of associated multimeric C1q, and crystal packing experiments in silico confirmed the potential docking of the two into a large membrane-bound complex with a ratio of IgG to C1q of 6:4. The modeling suggested that one arm of the IgG would bind antigen, while the other protruded from the membrane. This monovalent binding was confirmed by comparing the CDC potency of a bispecific antibody in which one Fab region had anti-CD20 specificity and the other Fab region had irrelevant specificity. This bispecific antibody that was functionally monovalent (i.e., only one of the two Fab arms bound to cell-surface CD20) was found to be a more potent mediator of CDC than a bispecific antibody in which both Fab regions had anti-CD20 specificity.
In Brief
The conclusion from this sophisticated and highly technical study is that when monomeric IgG binds to antigens on the cell surface, a hexameric aggregate can form, thereby generating high-affinity bindings sites for C1q resulting in activation of the classical pathway of complement. This model may help to explain why complement activation varies between antigens and between antibodies of similar specificity, in that the epitope geometry of the combinations will accommodate hexamerization much more readily for some than others, leading to differences in complement activation.
These findings have important implications for the design of antibodies that are aimed at activating complement. CDC remains one of several mechanisms of action for anti-lymphoma monoclonal antibodies such as rituximab and ofatumumab. While the exact contribution of CDC to the therapeutic efficacy of these and other antibodies is still incompletely understood, and probably varies widely, the work of Dr. Parren and colleagues suggests a genetic engineering approach by which complement activation can be modulated. In some instances, enhancing CDC may be beneficial, while in other instances, reducing complement activation could ameliorate treatment-associated toxicity.
Competing Interests
Dr. Johnson indicated no relevant conflicts of interest.
