549e Defective Virus Genomes: toward Mechanisms of Emergence and Growth

Kristen A. Thompson, Chemical and Biological Engineering, University of Wisconsin - Madison, 1415 Engineering Drive, Madison, WI 53706 and John Yin, Chemical and Biological Engineering, University of Wisconsin-Madison, 3633 Engineering Hall, 1415 Engineering Drive, Madison, WI 53706.

Virus-like particles with defective genomes are formed as a byproduct of infections by diverse viruses including influenza virus, poliovirus and bacteriophage. Genomes of these particles typically carry large deletions in essential functions, making their growth dependent on complementation by a co-infecting standard virus. Moreover, by competing for limited intracellular resources such virus-like particles can themselves interfere with and parasitize infections, reducing growth yields of standard virus. Over multiple generations such defective-interfering (DI) particles can drive the growth of standard virus to extinction, a property that could have therapeutic implications.

It is well established that DI particles arise most readily in culture under conditions of excessive virus concentration, characterized by the multiplicity of infection (MOI) or average number of virus particles per cell. However, it is not known what values of MOI are necessary or sufficient for DI particle formation and propagation. Further, it is not known what MOIs are sufficiently low to avoid DI particle formation and accumulation in culture. To better understand the effects of MOI on DI particle emergence we have employed a genetically well-characterized rabies-like RNA virus, vesicular stomatitis virus (VSV) grown on baby hamster kidney cells.

VSV gene regulation is correlated with the order of its genes on its linear single-stranded RNA genome. The gene order of the wild type virus is 3'-N-P-M-G-L-5', where the N gene, encoding the viral nucleocapsid, is expressed at the highest level, and the L gene, encoding the large subunit of the viral polymerase, is expressed at the lowest level. VSV-N1 has the wild-type VSV gene order while VSV-N4 is a recombinant virus with the N gene in the 4th position from the 3' end and overall gene order 3'- P-M-G-N-L-5'. VSV-N4 does not replicate as well as VSV-N1, likely owing to an inability of VSV-N4 to meet the high demand for nucleocapsid protein.

We have performed serial-passage infections of VSV-N1 and VSV-N4 with average MOIs spanning from 100 down to 10-4 virus particles per cell. Serial passage of VSV-N1 at high MOIs led to significantly reduced yields over four passages. At MOIs of 100 yields of infectious virus dropped by two orders of magnitude from ~109 to ~107 plaque forming units/ml, while passages with MOIs of 5 and 10 caused yields to drop by one order of magnitude. Passages performed at MOIs of 1 to 10-4 did not cause significant changes in yields of infectious particles. By contrast, when VSV-N4 strain was serially passaged at an MOI of 1, infectious virus yields dropped by one order of magnitude in four passages. These results indicate higher passage MOIs either produce more DI particles or create a higher susceptibility of wild-type infection to interference by DIs. Further, they suggest that the magnitude of the interference effect is influenced by imbalances in the distribution of viral protein expression. Current work aims to isolate DI particles and quantify their effects on the production of DI and viable virus particles.