Protection against lethal viral infection by neutralizing and nonneutralizing monoclonal antibodies: distinct mechanisms of action in vivo.

Monoclonal antibodies (MAb) reactive with the glycoprotein of vesicular stomatitis virus (VSV) serotypes Indiana (VSV-Ind) and New Jersey (VSV-NJ) were used to protect mice against lethal infection. MAb which reacted with a number of distinct epitopes and which could neutralize the virus in vitro could also protect against infection in vivo. MAb which could not neutralize the virus in vitro but which were specific for the glycoprotein of a single serotype were also able to protect mice against lethal VSV challenge. Interestingly, a group of MAb which cross-reacted with the glycoproteins of VSV-Ind and VSV-NJ could passively protect against challenge with either serotype. It was shown that as early as 2 h after infection, neither neutralizing nor nonneutralizing MAb could protect. Nonneutralizing MAb were found to be less effective at in vivo protection than neutralizing MAb. Furthermore, nonneutralizing MAb demonstrated a much lower binding efficiency to intact virions than did neutralizing MAb. These observations, plus the fact that the nonneutralizing MAb could lyse virus-infected cells in the presence of complement, suggested that in vivo protection by these antibodies may involve cell-associated viral determinants. To compare the mechanisms by which neutralizing and nonneutralizing MAb protected in vivo, F(ab')2 fragments were used in protection experiments. Although the F(ab')2 of a neutralizing MAb was still able to protect animals lethal virus challenge, the F(ab')2 of a cross-reactive nonneutralizing MAb was unable to do so. The reactivity of nonneutralizing MAb with virions and the apparent necessity of an intact Fc portion for protection further distinguish these antibodies from those MAb that are able to neutralize VSV solely by binding to the glycoprotein.