Host factors that interact with the pestivirus N-terminal protease, Npro, are components of the ribonucleoprotein complex.

UNLABELLED: The viral N-terminal protease N(pro) of pestiviruses counteracts cellular antiviral defenses through inhibition of IRF3. Here we used mass spectrometry to identify a new role for N(pro) through its interaction with over 55 associated proteins, mainly ribosomal proteins and ribonucleoproteins, including RNA helicase A (DHX9), Y-box binding protein (YBX1), DDX3, DDX5, eIF3, IGF2BP1, multiple myeloma tumor protein 2, interleukin enhancer binding factor 3 (IEBP3), guanine nucleotide binding protein 3, and polyadenylate-binding protein 1 (PABP-1). These are components of the translation machinery, ribonucleoprotein particles (RNPs), and stress granules. Significantly, we found that stress granule formation was inhibited in MDBK cells infected with a noncytopathic bovine viral diarrhea virus (BVDV) strain, Kyle. However, ribonucleoproteins binding to N(pro) did not inhibit these proteins from aggregating into stress granules. N(pro) interacted with YBX1 though its TRASH domain, since the mutant C112R protein with an inactive TRASH domain no longer redistributed to stress granules. Interestingly, RNA helicase A and La autoantigen relocated from a nuclear location to form cytoplasmic granules with N(pro). To address a proviral role for N(pro) in RNP granules, we investigated whether N(pro) affected RNA interference (RNAi), since interacting proteins are involved in RISC function during RNA silencing. Using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) silencing with small interfering RNAs (siRNAs) followed by Northern blotting of GAPDH, expression of N(pro) had no effect on RNAi silencing activity, contrasting with other viral suppressors of interferon. We propose that N(pro) is involved with virus RNA translation in the cytoplasm for virus particle production, and when translation is inhibited following stress, it redistributes to the replication complex. IMPORTANCE: Although the pestivirus N-terminal protease, N(pro), has been shown to have an important role in degrading IRF3 to prevent apoptosis and interferon production during infection, the function of this unique viral protease in the pestivirus life cycle remains to be elucidated. We used proteomic mass spectrometry to identify novel interacting proteins and have shown that N(pro) is present in ribosomal and ribonucleoprotein particles (RNPs), indicating a translational role in virus particle production. The virus itself can prevent stress granule assembly from these complexes, but this inhibition is not due to N(pro). A proviral role to subvert RNA silencing through binding of these host RNP proteins was not identified for this viral suppressor of interferon.

Small RNA analysis in Sindbis virus infected human HEK293 cells.

INTRODUCTION: In contrast to the defence mechanism of RNA interference (RNAi) in plants and invertebrates, its role in the innate response to virus infection of mammals is a matter of debate. Since RNAi has a well-established role in controlling infection of the alphavirus Sindbis virus (SINV) in insects, we have used this virus to investigate the role of RNAi in SINV infection of human cells. RESULTS: SINV AR339 and TR339-GFP were adapted to grow in HEK293 cells. Deep sequencing of small RNAs (sRNAs) early in SINV infection (4 and 6 hpi) showed low abundance (0.8%) of viral sRNAs (vsRNAs), with no size, sequence or location specific patterns characteristic of Dicer products nor did they possess any discernible pattern to ascribe to a specific RNAi biogenesis pathway. This was supported by multiple variants for each sequence, and lack of hot spots along the viral genome sequence. The abundance of the best defined vsRNAs was below the limit of Northern blot detection. The adaptation of the virus to HEK293 cells showed little sequence changes compared to the reference; however, a SNP in E1 gene with a preference from G to C was found. Deep sequencing results showed little variation of expression of cellular microRNAs (miRNAs) at 4 and 6 hpi compared to uninfected cells. Twelve miRNAs exhibiting some minor differential expression by sequencing, showed no difference in expression by Northern blot analysis. CONCLUSIONS: We show that, unlike SINV infection of invertebrates, generation of Dicer-dependent svRNAs and change in expression of cellular miRNAs were not detected as part of the Human response to SINV.