Computational prediction of intracellular targets of wild-type or mutant vesicular stomatitis matrix protein

by Matthew C. Morris, Thomas M. Russell, Cole A. Lyman, Wesley K. Wong, Gordon Broderick, Maureen C. Ferran

The matrix (M) protein of vesicular stomatitis virus (VSV) has a complex role in infection and immune evasion, particularly with respect to suppression of Type I interferon (IFN). Viral strains bearing the wild-type (wt) M protein are able to suppress Type I IFN responses. We recently reported that the 22–25 strain of VSV encodes a wt M protein, however its sister plaque isolate, strain 22–20, carries a M[MD52G] mutation that perturbs the ability of the M protein to block NFκB, but not M-mediated inhibition of host transcription. Therefore, although NFκB is activated in 22–20 infected murine L929 cells infected, no IFN mRNA or protein is produced. To investigate the impact of the M[D52G] mutation on immune evasion by VSV, we used transcriptomic data from L929 cells infected with wt, 22–25, or 22–20 to define parameters in a family of executable logical models with the aim of discovering direct targets of viruses encoding a wt or mutant M protein. After several generations of pruning or fixing hypothetical regulatory interactions, we identified specific predicted targets of each strain. We predict that wt and 22–25 VSV both have direct inhibitory actions on key elements of the NFκB signaling pathway, while 22–20 fails to inhibit this pathway.