In this report we uncover three novel findings about the ORF57 protein, an essential KSHV protein involved in viral gene expression. First, these data establish a role for ORF57 in stabilizing nuclear transcripts. Second, our data show that ORF57 binds directly to its target, PAN RNA, in living cells and that binding correlates with function. Third, ORF57 and its homologs can increase the expression of a variety of mRNAs 
, suggesting a relatively nonspecific effect. However, here we demonstrate the existence of an ORF57-responsive element in PAN RNA suggesting that, at least in some cases, specific cis-acting sequences have evolved to recruit ORF57 to its targets. Consistent with previous reports, our data further show that ORF57 primarily enhances transcript accumulation from intron-lacking genes but not from intron-containing genes, even when the ORE is included in the spliced transcript. Together, these data are consistent with a model in which ORF57 binds to intronless viral transcripts and protects them from a cellular RNA quality control pathway.
The data presented here demonstrate ORF57 stabilizes nuclear RNAs that would otherwise be rapidly degraded. Using steady-state analysis () and a transcription pulse assay (), we show that ORF57 increases the stability of an unstable polyadenylated nuclear RNA. Kinetic analysis further supports the conclusion that ORF57 is protecting RNAs, at least in part, from the same rapid RNA decay pathway that the ENE protects transcripts from in cis 
. Even though ORF57 has been implicated in mRNA export 
, its expression does not lead to cytoplasmic accumulation of PAN RNA. Thus, we conclude that the effect of ORF57 on RNA stability is independent of its proposed function in RNA export. This conclusion is consistent with previous reports implying that ORF57 stabilizes its target RNAs 
and with the observations that ORF57 homologs SM and ICP27 can stabilize specific transcripts 
. However, these data are the first direct demonstration that KSHV ORF57 increases the half-lives of nuclear RNAs and that his function is separable from its proposed roles in mRNA export, transcription, and translation.
We observe only a slight increase in polymerase density on PAN RNA genes driven from either the PAN or tetracycline-responsive promoter (). Therefore, we conclude ORF57 increases PAN RNA levels primarily by a post-transcriptional mechanism. We were surprised at the lack of increase in pol II density on the constructs driven by the PAN promoter, which binds the viral transactivator ORF50, an ORF57-interacting protein 
. Assuming these proteins associate in our experimental system, the interaction appears to have little effect on transcription initiation. It should be noted that we used an antibody (8WG16) that preferentially recognizes the initiating hypophosphorylated form of pol II 
. Thus, it remains possible that a hyperphosphorylated elongating form of pol II increases on the PAN RNA gene in response to ORF57. Taken with the observation that PAN RNA is up-regulated by ORF57 from four different promoters (), our data strongly support the conclusion that transcription initiation is unaffected by ORF57.
ORF57 enhances the expression of viral mRNAs, the noncoding nuclear PAN RNA, and heterologous reporter mRNAs, so it appeared that its effects were not strongly influenced by cis-acting sequences. On the contrary, our data demonstrate that a specific cis-acting sequence, the ORE, can enhance the effects of ORF57 on both PAN RNA and on a β-globin reporter mRNA. Using a previously described set of deletions 
, we show that ORF57 has reduced binding to PANΔ1 RNA and that this correlates with loss of activity (, ). Moreover, we placed the ORE into an intronless β-globin mRNA and found that it enhances ORF57-responsiveness by ~5-fold, demonstrating that the ORE affects mRNAs as well as the noncoding PAN RNA (). Tethering ORF57 to the ORE-lacking PANΔ1 transcripts in cells is sufficient to complement the ORE deletion, so it seems likely that ORF57 recruitment resides at the core of ORE activity (). The simplest interpretation of these results is that the ORE is a high-affinity ORF57-binding site, and that RNA binding by ORF57 is necessary for its stabilization activity. Interestingly, intronless β-globin mRNA levels are enhanced by ORF57 ~4-fold in the absence of the ORE. Perhaps ORF57 has enough non-specific RNA-binding activity to account for its general effects on reporter RNAs. Alternatively, this effect may be linked to a separate ORF57 activity that functions independently of RNA binding.
Mechanistically, our data support the model that ORF57 binds to its RNA targets and inhibits the activity of nuclear RNA decay enzymes, but we do not yet know the molecular details of ORF57-mediated RNA stabilization. In one model, ORF57 binds RNA making it inaccessible to RNA decay enzymes. Alternatively, ORF57 could indirectly stabilize transcripts by promoting changes in RNP composition or conformation. It remains formally possible that ORF57 increases PAN RNA stability by retaining the transcripts in the nucleus, thereby protecting them from cytoplasmic decay enzymes. However, given the reported role of ORF57 in mRNA export 
and its ability to shuttle 
, we think this last hypothesis is unlikely.
Our data are consistent with the model that ORF57 counteracts a nuclear RNA quality control pathway that rapidly degrades transcripts that derived from intronless genes 
. It is important to point out that this model does not depend on cells specifically recognizing intronless RNAs. Rather, because of the extensive coupling between the steps of cellular RNA biogenesis including pre-mRNA splicing 
, we believe it more likely that intronless RNAs are inefficiently processed. If RNA surveillance and RNA maturation are in kinetic competition as formally proposed by Doma and Parker 
, these inefficiently processed intronless transcripts would be predicted to be subject to degradation by RNA quality control pathways. Because the majority of KSHV genes are intronless 
, we propose that ORF57 functions to counteract this RNA decay pathway to promote the robust expression of viral genes. ORF57 has been reported to promote the export of intronless viral mRNAs 
, so it is easy to imagine that ORF57 allows mRNAs to bypass nuclear decay systems by enhancing the efficiency of their export. However, in the case of PAN RNA, we can uncouple ORF57 mRNA export activities from its function in RNA stabilization. Therefore, our data suggest a more active role for ORF57 in protecting transcripts from degradation in the nucleus. Current work focuses on further testing this model by identifying the cellular decay machinery involved as well as the viral RNAs bound and protected by ORF57.
Because PAN RNA accumulates to such high levels in KSHV lytically reactivated cells 
, it likely performs an important function for the virus. Therefore, studying the biogenesis of this unusual transcript is essential to understanding KSHV biology. In addition, PAN RNA provides a useful a tool to separate ORF57 functions in RNA stability from its role in RNA export. We used an unstable ENE-lacking PAN RNA for our decay studies (), but several observations suggest an important role for ORF57 in PAN RNA biogenesis in infected cells. First, steady-state levels of PAN RNA containing the ENE are up-regulated in the presence of ORF57 ( and ). Second, published reports have shown that ORF57-deleted bacmids produce reduced levels of PAN RNA during lytic infection 
. Finally, we have observed that insertion of multiple copies of the ENE leads to higher levels of PAN RNA (unpublished observations) or βΔ1,2 mRNA 
than insertion of one ENE, suggesting that a single ENE does not completely block RNA degradation. Thus, the proposed overlapping activities of the ENE and ORF57 may both be essential to fully stabilize PAN RNA during lytic replication. Perhaps more importantly, our studies provide insights into the possible mechanism of ORF57 activity on the accumulation of intronless viral transcripts that lack ENE-like elements. However, further experimentation is necessary to test the role that ORF57 plays on PAN RNA stability in the context of viral infection and on the stability of intronless viral mRNAs.
A particularly interesting component of ORF57-mediated RNA stabilization is the role of the poly(A) tail in regulation of transcript stability. After transcription shut-off in the presence of ORF57, we found that some transcripts become hyperadenylated, while others are partially deadenylated (, data not shown). Because both the hypo- and hyperadenylated PAN transcripts are present 8 hrs subsequent to transcription shut-off, it seems that ORF57 stabilizes both forms. Indeed, when we over-expose our northern blots, we observe transcripts resembling these hyper- and hypoadenylated forms in the samples from cells lacking ORF57, but at significantly reduced levels (data not shown). Several different non-exclusive roles for the poly(A) tail in nuclear RNA stability can be imagined that are consistent with our observations. In the first model, PAN RNA is recognized by the cell as an aberrant transcript, presumably due to its lack of export. In manner analogous to yeast and bacterial systems, hyperadenylation of the transcripts is linked to quality control 
. Interestingly, the host-shutoff mechanism employed by KSHV appears to involve destabilization of hyperadenylated cellular mRNAs 
. Moreover, recent work implicates mRNA export factors as regulators of poly(A) length in cells and in polyadenylation assays performed in vitro 
. In the second model, ORF57 promotes polyadenylation, which then leads to greater transcript stability. In a third model, ORF57 stabilizes nuclear PAN RNA, and the hyperadenylation results from promiscuous polyadenylation of the stabilized nuclear transcripts. Distinguishing among these models will shed light into host-virus interactions between KSHV ORF57, cellular poly(A) machinery, and cellular RNA decay pathways.