Analysis of rabies virus glycoprotein sequences in relation to the proposed use of monoclonal antibodies for post-exposure prophylaxis

Abstract

The demand for rabies immune globulin (RIG) for post-exposure prophylaxis (PEP) is significant. Unfortunately, the cost of RIG is prohibitive for many patients in developing countries. Several monoclonal antibodies (MAbs) which neutralize rabies virus (RABV) have been proposed as a replacement for conventional RIG due to the ability of their large-scale production at a reduced cost. In the present study, we generated 487 RABV glycoprotein (G) sequences from a variety of viral lineages, and supplemented the dataset with 154 complete and 115 partial G sequences available in GenBank. The objective was to evaluate variability of known MAb-binding epitopes on the G, which may preclude virus neutralization. The analysis demonstrated that binding site of MAb CR57 (amino acids 226-231 of the G ectodomain) is very conservative. The substitutions detected (such as K226R in several raccoon and African dog RABV isolates; L231P/S in several skunk, raccoon, and various bat RABV lineages; etc.) did not preclude virus neutralization from previously published studies. No substitutions that abolished binding of MAb CR57 in escape mutant studies were detected in naturally occurring field RABV isolates. In contrast, numerous substitutions were detected in the binding site of MAb CR4098 (AA 330-338 of the G ectodomain). Examples include a K330N substitution in a bat isolate from Brazil; V332I/F substitutions in several RABV lineages, associated with big brown bats; N336D in several viruses associated with big brown bats in North America, in South-African mongoose RABV, in one African and one Korean dog RABV isolate; N336G/S in several raccoon isolates; E337D in several canine RABV from Serbia and in the southcentral skunk RABV isolates; I338T in the canyon bat and Arctic RABV isolates. Substitutions in position 336, particularly the N336D, were detected earlier in escape virus studies and precluded neutralization of such viruses by MAb CR4098. Nevertheless, no isolates had substitutions in binding sites for MAbs CR57 and CR4098 simultaneously. There is no reason to expect that any of the viruses from our study would escape neutralization by a combination of these MAbs in vivo. The situation is different for HuMAb RAB1 (also referred to as 17C7). We confirmed numerous substitutions, particularly in position 336, which may abolish binding of MAb RAB1 as was shown previously by escape virus generation. The RAB1 was proposed as a single MAb for use in human rabies PEP, claiming that there are no natural RABV isolates which harbor critical substitutions in its binding site (the combination 336D-346K in the G ectodomain). We encountered this combination in the majority of viruses from one of the lineages associated with big brown bats distributed broadly in North America. Our findings clearly demonstrate that the proposed use of a single MAb for rabies PEP is inappropriate, in line with international recommendations.