Role of the Adenovirus Type 5 L4-100K Protein During Lytic Infection

Abstract

With the onset of the late phase, one of the first adenovirus type 5 (Ad5) late proteins translated, L4-100K (100K), starts to perform a number of functions that are essential for efficient completion of lytic virus infection. This adenovirus non-structural late protein alters the cellular machinery in favor of translating large amounts of virus products, leading to their subsequent nuclear accumulation for capsid assembly. L4-100K achieves this by contributing to the transport and selective translation of late viral mRNAs, acting as a chaperone for hexon trimerization and being involved in its transport. L4-100K also plays a role in preventing apoptosis of the infected cell by interacting with granzyme-B (GrB) and inhibiting its activity. However, most of the mechanisms underlying these processes, and how 100K is regulated to accomplish these, remain unclear.

In this work the nucleocytoplasmic shuttling properties of L4-100K required to efficiently accomplish cytoplasmic and nuclear tasks and the posttranslational modifications related to this process, as well as its interaction with viral and cellular factors were investigated by generating mutant viruses. These recombinant viruses carry amino acid exchanges at the following locations: (1) C-terminal arginine-glycine-glycine (RGG) boxes; (2) Nuclear export motif; (3) Sumo conjugation motif and (4) N-terminal tyrosine residue of the L4-100K coding sequence. These mutant viruses were not only tested for DNA replication, late protein synthesis and virus yield properties, but also their L4-100K proteins were analyzed for subcellular distribution and protein-protein/RNA interaction patterns.

Arginine to glycine exchanges in the RGG boxes significantly diminished 100K methylation in the course of infection, and substantially reduced virus growth. This demonstrates that 100K methylation in RGG motifs is an important host-cell function required for efficient Ad replication. Further investigation of this mutant virus indicated that PRMT1-catalyzed arginine methylation in the RGG boxes regulates the binding of 100K to hexon and promotes the nuclear localization and capsid assembly of the structural protein, as well as modulating 100K’s tripartite leader-containing mRNA (TL-mRNA) interaction. Furthermore, substitutions in the following glycine-arginine-rich domain (GAR), but not RGG regions, affected 100K nuclear import, implying that the nuclear localization signal of 100K is located within the GAR sequence.

Significantly, phenotypic analyses in different human tumor cell lines revealed that the L4-100K-NES mutant virus is severely defective in both late viral gene expression and virus production compared to wild-type virus, suggesting that nuclear export of 100K is required for maximal virus growth. Consistent with a previous publication, localization of the 100K NES mutant was restricted to the nuclei of infected cells. A similar result was obtained when CRM1 was inactivated by leptomycin B (LMB), showing that CRM1 is the major export receptor for 100K in Ad5 infection. These results demonstrate that CRM1 plays a critical role in the late phase of lytic Ad5 infection, particularly in L4-100K shuttling.

Interestingly, the sumoylation of L4-100K could not be confirmed in the tested human tumor cell lines, although this motif is highly conserved among different Ad serotypes and shows great sequence similarity to the defined Sumo consensus motif. Consistently, the Sumo mutant virus was as efficient as the wt virus in the late phase, yielding high amounts of infectious virus particles. In contrast, the N-terminal tyrosine mutant was found to be defective in late protein synthesis and virus yield. Nonetheless, the function of this tyrosine residue in the highly conserved motif hypothesized to be modified by sulfation still remains to be clarified, since such a modification could not be detected in L4-100K.

Furthermore, in this study the interaction of the late non-structural L4-100K protein with early regulator E1B-55K was determined for the first time by immunoprecipitation assays. This interaction was preserved in all of the tested L4-100K and E1B-55K mutant viruses. Interaction sites in both proteins were investigated by GST-pull-down assays using GST-fused-protein-fragments. According to the results, the interaction between these two multifunctional regulatory proteins apparently plays a critical role in the virus-specific translation mechanism, which requires further investigation.