IFIT (interferon-induced with tetratricopeptide repeats) proteins are critical mediators of mammalian innate antiviral immunity. Mouse IFIT1 selectively inhibits viruses that lack 2'O-methylation of their mRNA 5' caps. Surprisingly, human IFIT1 does not share this antiviral specificity. Here, we resolve this discrepancy by demonstrating that human and mouse IFIT1 have evolved distinct functions using a combination of evolutionary, genetic and virological analyses. First, we show that human IFIT1 and mouse IFIT1 (renamed IFIT1B) are not orthologs, but are paralogs that diverged >100 mya. Second, using a yeast genetic assay, we show that IFIT1 and IFIT1B proteins differ in their ability to be suppressed by a cap 2'O-methyltransferase. Finally, we demonstrate that IFIT1 and IFIT1B have divergent antiviral specificities, including the discovery that only IFIT1 proteins inhibit a virus encoding a cap 2'O-methyltransferase. These functional data, combined with widespread turnover of mammalian IFIT genes, reveal dramatic species-specific differences in IFIT-mediated antiviral repertoires.DOI: http://dx.doi.org/10.7554/eLife.14228.001

Transfer RNAs (tRNAs) are transcribed by RNA polymerase III (RNAPIII) and play a central role in decoding our genome, yet their expression and noncanonical function remain understudied. Many DNA tumor viruses enhance the activity of RNAPIII, yet whether infection alters tRNA expression is largely unknown. Here, we present the first genome-wide analysis of how viral infection alters the tRNAome. Using a tRNA-specific sequencing method (DM-tRNA-seq), we find that the murine gammaherpesvirus MHV68 induces global changes in premature tRNA (pre-tRNA) expression, with 14% of tRNA genes upregulated more than 3-fold, indicating that differential tRNA gene induction is a characteristic of DNA virus infection. Elevated pre-tRNA expression corresponds to increased RNAPIII occupancy for the subset of tRNA genes tested; additionally, posttranscriptional mechanisms contribute to the accumulation of pre-tRNA species. We find increased abundance of tRNA fragments derived from pre-tRNAs upregulated by viral infection, suggesting that noncanonical tRNA cleavage is also affected. Furthermore, pre-tRNA accumulation, but not RNAPIII recruitment, requires gammaherpesvirus-induced degradation of host mRNAs by the virally encoded mRNA endonuclease muSOX. We hypothesize that depletion of pre-tRNA maturation or turnover machinery contributes to robust accumulation of full-length pre-tRNAs in infected cells. Collectively, these findings reveal pervasive changes to tRNA expression during DNA virus infection and highlight the potential of using viruses to explore tRNA biology.IMPORTANCE Viral infection can dramatically change the gene expression landscape of the host cell, yet little is known regarding changes in noncoding gene transcription by RNA polymerase III (RNAPIII). Among these are transfer RNAs (tRNAs), which are fundamental in protein translation, yet whose gene regulatory features remain largely undefined in mammalian cells. Here, we perform the first genome-wide analysis of tRNA expression changes during viral infection. We show that premature tRNAs accumulate during infection with the model gammaherpesvirus MHV68 as a consequence of increased transcription, but that transcripts do not undergo canonical maturation into mature tRNAs. These findings underscore how tRNA expression is a highly regulated process, especially during conditions of elevated RNAPIII activity.

Short interspersed nuclear elements (SINEs) are RNA polymerase III (RNAPIII)-transcribed, retrotransposable noncoding RNA (ncRNA) elements ubiquitously spread throughout mammalian genomes. While normally silenced in healthy somatic tissue, SINEs can be induced during infection with DNA viruses, including the model murine gammaherpesvirus 68 (MHV68). Here, we explored the mechanisms underlying MHV68 activation of SINE ncRNAs. We demonstrate that lytic MHV68 infection of B cells, macrophages, and fibroblasts leads to robust activation of the B2 family of SINEs in a cell-autonomous manner. B2 ncRNA induction requires neither host innate immune signaling factors nor involvement of the RNAPIII master regulator Maf1. However, we identified MHV68 ORF36, the conserved herpesviral kinase, as playing a key role in B2 induction during lytic infection. SINE activation is linked to ORF36 kinase activity and can also be induced by inhibition of histone deacetylases 1 and 2 (HCAC 1/2), which is one of the known ORF36 functions. Collectively, our data suggest that ORF36-mediated changes in chromatin modification contribute to B2 activation during MHV68 infection and that this activity is conserved in other herpesviral protein kinase homologs. IMPORTANCE Viral infection dramatically changes the levels of many types of RNA in a cell. In particular, certain oncogenic viruses activate expression of repetitive genes called retrotransposons, which are normally silenced due to their ability to copy and spread throughout the genome. Here, we established that infection with the gammaherpesvirus MHV68 leads to a dramatic induction of a class of noncoding retrotransposons called B2 SINEs in multiple cell types. We then explored how MHV68 activates B2 SINEs, revealing a role for the conserved herpesviral protein kinase ORF36. Both ORF36 kinase-dependent and kinase-independent functions contribute to B2 induction, perhaps through ORF36 targeting of proteins involved in controlling the accessibility of chromatin surrounding SINE loci. Understanding the features underlying induction of these elements following MHV68 infection should provide insight into core elements of SINE regulation, as well as disregulation of SINE elements associated with disease.

IFIT (interferon-induced with tetratricopeptide repeats) proteins are critical mediators of mammalian innate antiviral immunity. Mouse IFIT1 selectively inhibits viruses that lack 2'Omethylation of their mRNA 5' caps. Surprisingly, human IFIT1 does not share this antiviral specificity. Here, we resolve this discrepancy by demonstrating that human and mouse IFIT1 have evolved distinct functions using a combination of evolutionary, genetic and virological analyses. First, we show that human IFIT1 and mouse IFIT1 (renamed IFIT1B) are not orthologs, but are paralogs that diverged >100 mya. Second, using a yeast genetic assay, we show that IFIT1 and IFIT1B proteins differ in their ability to be suppressed by a cap 2'O-methyltransferase. Finally, we demonstrate that IFIT1 and IFIT1B have divergent antiviral specificities, including the discovery that only IFIT1 proteins inhibit a virus encoding a cap 2'O-methyltransferase. These functional data, combined with widespread turnover of mammalian IFIT genes, reveal dramatic species-specific differences in IFIT-mediated antiviral repertoires.

12Transfer RNAs (tRNAs) are transcribed by RNA polymerase III (RNAPIII) and play a 13 central role in decoding our genome. Many DNA tumor viruses enhance the activity of RNAPIII, 14 yet whether infection alters tRNA expression is largely unknown. Here, we present the first 15 genome wide analysis of how viral infection alters the tRNAome. Using a tRNA-specific 16 sequencing method (DM-tRNA-seq), we find that the murine gammaherpesvirus MHV68 17 induces global changes in pre-tRNA expression, indicating that differential tRNA gene induction 18 is a characteristic of DNA virus infection. Elevated pre-tRNA expression corresponds to 19 increased RNAPIII occupancy, but post-transcriptional mechanisms also contribute. Pre-tRNA 20 accumulation, but not RNAPIII recruitment, requires the virally encoded mRNA endonuclease 21 muSOX, suggesting interplay between the different facets of virus-induced gene regulation. 22Collectively, these findings reveal pervasive changes to tRNA expression during DNA virus 23 infection and highlight the potential of using viruses to explore tRNA biology. 24 25 Impact 26 The first genome-wide analysis of virus-induced tRNA expression changes reveals that 27 tRNA transcription and processing are perturbed by the gammaherpesvirus MHV68. 28 29 30The elegant design of the tRNA decoder, an adaptor molecule which converts genetic 31 information into protein, was proposed over 60 years ago (1). Although tRNAs were the first 32 non-coding RNA described, our understanding of tRNA biology, especially their functions 33 outside of protein translation and what dictates their expression, remains limited. This is partly 34 due to the difficulties in applying RNA sequencing and analysis platforms to tRNAs, but recent 35 methodological advances have resulted in a surge of new studies that help to resolve these 36 complexities (2-4). Non-canonical functions of tRNA transcripts have been described, namely 37 the discovery that both premature and mature tRNAs undergo fragmentation into shorter RNAs 38 with numerous regulatory functions in response to stress, cancer, and viral infection (reviewed 39 in (5, 6)). Given the evidence that the ~400 predicted tRNA genes in mammalian genomes are 40 differentially expressed across cell lines (7-9) and that viral infection can lead to post-41 transcriptional modulation of tRNAs (10, 11), defining the tRNAome under varying conditions will 42 be central to our understanding of tRNA function, canonical and otherwise. 43RNA polymerase III (RNAPIII) transcribes tRNA genes and exhibits enhanced activity 44 during infection with DNA viruses (12-17), many of which encode their own RNAPIII genes. 45Though RNAPIII activation stimulates expression of viral RNAPIII genes, concomitant 46 accumulation of host RNAPIII transcripts can trigger antiviral immune responses. For example, 47 the induction or misprocessing of RNAPIII-generated host 5S RNA pseudogene and vault RNA 48 transcripts during Herpes Simplex Virus-1 (HSV-1) and Kaposi's sarcoma-associated 49 herpesvirus (KSHV) infection...