Glycolipid nature of the complement-fixing host cell antigen of vesicular stomatitis virus.

Serological evidence for the presence of a host-cell component of vesicular stomatitis virus particles was provided by Cartwright & Pearce 0968). They showed by complement fixation reactions that purified virus particles prepared from virus grown in BHK zI cells reacted with antiserum to BHK cell but not with antiserum to pig kidney cells. Conversely, virus grown in pig kidney cells fixed complement with antiserum to pig kidney cells but not with antiserum to BHK 2I cells. The amount of complement fixed with the cell antisera was as great as that fixed with antiserum to virus. It could be argued that the presence of host cell constituents in the purified virus preparations, possibly adhering to the surface of the particles, would account for these reactions. However, two observations suggest that the reactions are not due to the adventitious presence of cellular constituents. In the first place, treatment of the virus with Tween 8o, which has a mild detergent action but does not affect the infectivity of the virus, has little effect on the complement-fixing activity of the virus with cell antiserum. Secondly, trypsin, which removes the glycoprotein projections from the surface of the virus with the consequent loss of complement-fixing activity with virus antiserum, has no effect on the complementfixing activity with cell antiserum (Cartwright, Smale & Brown, I969). We expected that a host cell component which fixed complement would contain protein but found, in agreement with Wagner, Schnaitman & Snyder (I969) and Kang & Prevec 0969), that virus grown in cells pre-labelled with [14C]-amino acids was not radioactive. Since this suggested that the host cell component was not a protein, we directed our attention to non-protein constituents of the virus. We had shown previously that virus grown in BHK 21 cells pre-labelled with [32PO4] was radioactive (Cartwright & Pearce, 1968). None of the [3~p] of virus labelled under these conditions was associated with the RNA, and polyacrylamide gel electrophoresis of the virus following disruption with sodium dodecyl sulphate under conditions we have described previously (Cartwright, Talbot & Brown, I97O ) showed that the [a2p] was not associated with any of the virus proteins. Similarly, virus grown in cells pre-labelled with [laC]-choline was also radioactive and the [14C] migrated to a position in polyacrylamide gels well ahead of the area to which the virus proteins migrated. These observations suggested that phospholipid and lipid constituents of the host cell were closely associated with the virus and confinned the analytical results of McSharry & Wagner (I97I) which showed that some of the virus lipid is derived from the host cell. Evidence that the complement-fixing activity with host cell antiserum was likely to be associated with the lipid fraction was obtained by using virus which was grown in the presence of [3H]-amino acids in BHK 2I ceils pre-labelled with [14C]-choline. The virus was disrupted with 3 % sodium dodecyl sulphate and passed through a column of Sephadex G-25. There was no separation of the [~4C] and [3H] but the procedure served to remove the detergent. The peak fraction of radioactivity fixed complement with BHK 2I cell antiserum but only to a small extent with virus antiserum. Hydrolysis of the peak fraction with trypsin (25o/zg./ml.) eliminated all the complement-fixing activity with virus antiserum but did not reduce the activity with cell antiserum (Table 0.