The Expression and Nuclear Retention Element of Polyadenylated Nuclear RNA Is Not Required for Productive Lytic Replication of Kaposi’s Sarcoma-Associated Herpesvirus

ABSTRACT

Kaposi’s sarcoma-associated herpesvirus (KSHV) is an oncogenic human gammaherpesvirus and the causative agent of Kaposi’s sarcoma (KS), primary effusion lymphoma (PEL), and multicentric Castleman’s disease (MCD). During reactivation, viral genes are expressed in a temporal manner. These lytic genes encode transactivators, core replication proteins, or structural proteins. During reactivation, other viral factors that are required for lytic replication are expressed. The most abundant viral transcript is the long noncoding RNA (lncRNA) known as polyadenylated nuclear (PAN) RNA. lncRNAs have diverse functions, including the regulation of gene expression and the immune response. PAN possesses two main cis-acting elements, the Mta response element (MRE) and the expression and nuclear retention element (ENE). While PAN has been demonstrated to be required for efficient viral replication, the function of these elements within PAN remains unclear. Our goal was to determine if the ENE of PAN is required in the context of infection. A KSHV bacmid containing a deletion of the 79-nucleotide (nt) ENE in PAN was generated to assess the effects of the ENE during viral replication. Our studies demonstrated that the ENE is not required for viral DNA synthesis, lytic gene expression, or the production of infectious virus. Although the ENE is not required for viral replication, we found that the ENE functions to retain PAN in the nucleus, and the absence of the ENE results in an increased accumulation of PAN in the cytoplasm. Furthermore, open reading frame 59 (ORF59), LANA, ORF57, H1.4, and H2A still retain the ability to bind to PAN in the absence of the ENE. Together, our data highlight how the ENE affects the nuclear retention of PAN but ultimately does not play an essential role during lytic replication. Our data suggest that PAN may have other functional domains apart from the ENE.

IMPORTANCE KSHV is an oncogenic herpesvirus that establishes latency and exhibits episodes of reactivation. KSHV disease pathologies are most often associated with the lytic replication of the virus. PAN RNA is the most abundant viral transcript during the reactivation of KSHV and is required for viral replication. Deletion and knockdown of PAN resulted in defects in viral replication and reduced virion production in the absence of PAN RNA. To better understand how the cis elements within PAN may contribute to its function, we investigated if the ENE of PAN was necessary for viral replication. Although the ENE had previously been extensively studied with both biochemical and in vitro approaches, this is the first study to demonstrate the role of the ENE in the context of infection and that the ENE of PAN is not required for the lytic replication of KSHV.

INTRODUCTION

Herpesviruses are large double-stranded DNA viruses. There are eight known human herpesviruses, which include Kaposi’s sarcoma-associated herpesvirus (KSHV), or human herpesvirus 8 (HHV-8). KSHV is an oncogenic herpesvirus and the causative agent of Kaposi’s sarcoma (KS). Other associated malignancies include multicentric Castleman’s disease (MCD) and primary effusion lymphoma (PEL) (1–3). KSHV is in the gammaherpesvirus subfamily, which also includes Epstein-Barr virus (EBV), or human herpesvirus 4 (HHV-4). The distinguishing characteristic of all herpesviruses is to establish a latent infection after the initial primary infection, granting the virus the ability to remain within the host. Latent infection is followed by periodic episodes of lytic reactivation and the subsequent production of infectious virus and viral shedding.

During reactivation, a highly abundant 1.1-kb long noncoding RNA (lncRNA) referred to as polyadenylated nuclear (PAN) RNA (Nut-1 or T1.1) is expressed early during reactivation (4–6). PAN is believed to account for ∼80% of the viral transcriptome (4). Interestingly, PAN RNA is also found in packaged virions (4, 7). The PAN locus is located between K6 and open reading frame 16 (ORF16) and partially overlaps K7 (4). PAN expression is regulated through K-RTA binding to a specific domain within the PAN promoter known as the Rta-responsive element (RRE) (4, 8–10). The PAN promoter is also an important area of genomic looping and has been demonstrated to be required for efficient viral replication even with the exogenous expression of PAN RNA (11). Viral proteins, including ORF57 and ORF59, also interact with PAN RNA. ORF57 plays a direct role in stabilizing the expression of PAN and increasing its nuclear accumulation (4, 12–14). ORF59 binds to PAN during reactivation, recruiting chromatin-modifying factors to the viral genome (15). PAN sequesters LANA from the viral DNA episomes to relieve LANA-mediated repression on lytic promoters during reactivation (4, 16). Previous studies using a PAN deletion virus have shown that the absence of PAN results in a defect in viral transcript accumulation and a decrease in infectious virus production (12, 17, 18). Additionally, PAN may contribute to cell cycle control and suppress the expression of host genes involved in the immune response (17). PAN also plays a role in epigenetic regulation by interacting with the host cellular demethylases UTX and JMJD3 to activate lytic replication (12, 15). Mapping studies have also shown that PAN interacts at multiple sites within the KSHV and host genomes (12). PAN also interacts with cellular histones, including H1.4 and H2A, yet the potential function of this interaction remains unknown (17).

The structural characteristics of PAN RNA contribute to both its stability and its functional properties. There are two main cis-acting elements found within PAN: the expression and nuclear retention element (ENE) and the Mta-responsive element (MRE) (4). The ENE is a 79-nucleotide (nt) U-rich sequence that mainly functions to accumulate PAN in the nucleus (19). Furthermore, the ENE functions as an inhibitor of nuclear RNA decay (20). Interestingly, the crystal structure of the core ENE revealed that it forms a triple-helix structure and binds to the poly(A) tail at the 3′ end of the RNA, preventing the binding of nuclear decay factors (21). The identification of the unique structure of the ENE has led to the discovery of ENE-like structures in other nonhuman herpesviruses such as rhesus rhadinovirus (RRV) and equine herpesvirus 2 (EHV-2) (6, 22). The MRE was demonstrated to promote PAN RNA stability, even in the absence of the ENE (23). The MRE is a 5′ element that forms 3 stem-loops and functions as the binding site for ORF57, preventing PAN RNA decay and promoting PAN RNA accumulation. This is achieved through a 9-nt element at the 5′ end of PAN located within the MRE named the ORE (ORF57-responsive element) (4, 14, 23). Previous studies comparing the effects of ENE and MRE deletions, within the context of a plasmid-based model, suggest that the MRE contributes significantly more to ORF57-mediated stability than the ENE (24). Additionally, other potential cis elements of PAN, including secondary structures and protein interaction sites, have been predicted using computational models (24). The identification of the ENE triple helix has also led to the identification of similar triple helices in cellular RNA, including cellular metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) and multiple endocrine neoplasia β (MENβ) (25, 26). The unique physical structure of the RNA triple helix has been a topic of speculation regarding its biological function and relevance.

Our study focused on determining the functional role of the PAN ENE in the context of KSHV infection. Our findings demonstrated that the ENE is not required for viral replication and does not play a role in regulating viral transcription or protein expression. We demonstrated that although PAN RNA abundance is not affected, the absence of the ENE disrupts PAN nuclear retention and localization patterns. These data indicate that the ENE is not required for efficient viral replication and suggest that other regions may contribute to the function of PAN during lytic replication.

RESULTS

cis elements within PAN RNA and generation of a PAN ΔENE KSHV bacmid.

Previous studies on the function of PAN indicated that the MRE and ENE of PAN might have a regulatory function during the KSHV lytic cycle (6, 19). Major structures and viral protein binding sites of PAN that have been previously described are summarized in Fig. 1A (17, 19, 23, 24). As previous studies using deletion and knockdown of PAN had shown it to be required for efficient viral replication, we sought to determine if the ENE plays a role in the function of PAN during lytic replication.

FIG 1.

Major structural features and viral protein binding regions for PAN. (A) PAN cis-acting MRE (blue) and ENE (green) regions. (Top) Major viral protein interaction domains within PAN. (Bottom) PAN RNA secondary structure model of the 3 major stem-loops of the MRE, including the ORE, MRE core, and ENE triple helix showing a poly(A) tail. (B) Whole-genome sequence mapping reads showing the PAN sequences in WT and ΔENE viruses compared to the KSHV reference genome as presented in Qiagen CLC Genomics Workbench. The ENE-specific deletion is shown.

First, a KSHV bacmid containing a deletion of the 79-nt ENE of PAN was generated. Wild-type (WT) KSHV bacterial artificial chromosome 16 (BAC16) used in these experiments contained an in-frame hemagglutinin (HA) tag of ORF59 (BAC16 ORF59-HA) to facilitate downstream experiments using HA reagents and was previously characterized (27). KSHV BAC16 ORF59-HA ΔENE was constructed using homologous recombination with the insertion of a kanamycin (Kan) I-SceI cassette, followed by the seamless removal of the kanamycin cassette. The insertion and removal of the kanamycin cassette were confirmed by whole-genome DNA sequencing, and the construct was verified to contain only the specific deletion (Fig. 1B). No other indels were detected within the viral genome for the BAC mutants used in the following experiments. BAC sequences were deposited in the NCBI Sequence Read Archive (SRA), and accession numbers are listed in Materials and Methods. To generate the viral cell lines, iSLK cells were transfected with either BAC16 ORF59-HA WT, ΔPAN, or ΔENE and selected with 1.2 mg/ml hygromycin in addition to 1 μg/ml puromycin and 250 μg/ml G418 used to maintain the iSLK cells. In all subsequent experiments within this study, iSLK BAC16 ORF59-HA WT, ΔPAN, or ΔENE cells were reactivated with 0.25 μg/ml doxycycline (Dox), 0.25 mM sodium butyrate (NaB), and 10 ng/ml 12-O-tetradecanoyl-phorbol-13-acetate (TPA).

The ENE is not required for expression of IE, E, and L viral transcripts but facilitates nuclear retention of PAN during reactivation.

As the ENE has previously been reported to contribute to the accumulation of PAN RNA, we sought to determine if the absence of the ENE affected the accumulation of PAN RNA during lytic replication. Along with PAN RNA, we also measured other viral transcripts at the immediate early (IE), early (E), and late (L) stages during reactivation. iSLK BAC16 ORF59-HA WT, ΔENE, and ΔPAN cells were induced and harvested at 6, 24, 48, and 72 h postinduction (hpi). mRNA accumulation was measured by quantitative PCR (qPCR) using primers/probes specific for ORF50, PAN, ORF58/59, ORF9, ORF57, ORF K2 (viral interleukin-6 [vIL-6]), ORF K13 (vFLIP), and ORF26. No significant differences were observed in the accumulation of IE, E, and L viral mRNAs measured in ΔENE and ΔPAN compared to WT cells. As expected, ΔPAN cells showed no PAN RNA (Fig. 2). Interestingly, the amount of PAN RNA itself was not affected by the ENE deletion. Surprisingly, we observed that transcript accumulation in ΔPAN cells was not affected. A deletion of PAN had been previously characterized to have a disruption in transcript accumulation in BAC36CR-infected 293L cells (5). It is important to note that although we did not observe similar results in reactivated BAC16 iSLK cells, this could be due to the overexpression of the inducible K-RTA that is present within the iSLK cells. Thus, it is likely that the overexpression of K-RTA is sufficient for transcript expression even in the absence of PAN. The other difference is the BAC templates, BAC16 and BAC36, which may also account for the discrepancy between our data and those from previous reports.

FIG 2.

The ENE does not affect the expression of IE, E, and L viral transcripts. RNA was harvested and purified from iSLK BAC16 ORF59-HA WT, ΔENE, and ΔPAN cells at 6, 24, 48, and 72 hpi. Purified RNA was DNase treated and used for cDNA synthesis followed by qPCR using primers specific for ORF50, PAN, ORF58/59, ORF9, ORF57, ORF K2, ORF K13, and ORF26. Results represent data from three independent experiments. Bar graphs represent means ± standard deviations (SD). A one-way ANOVA (analysis of variance) nonparametric Freidman test was performed (P value of <0.05).

The ENE was demonstrated to allow for the nuclear retention of PAN RNA. Thus, we were interested in determining whether the same nuclear retention of PAN would be observed in the context of KSHV infection. To assess whether the deletion of the ENE disrupts the accumulation of PAN in the nucleus, we quantified PAN in nuclear and cytoplasmic compartments. iSLK BAC16 ORF59-HA WT and ΔENE cells were induced at 24 and 72 hpi, nuclear and cytoplasmic fractions were separated using nuclear extraction buffer (NEB), and RNA was extracted and used to measure the accumulation of PAN in these fractions by qPCR. Primers and probes specific for PAN, 7SK as a cellular nuclear noncoding RNA (ncRNA), and ORF26 as a cytoplasmic viral mRNA were used in the qPCRs. We observed a significantly higher accumulation of PAN in the cytoplasmic fraction with the ENE deletion than with the WT at 72 hpi (Fig. 3A). To confirm that there was no contamination between fractions, we tested the fractions by Western blotting. iSLK BAC16 ORF59-HA WT and ΔENE cells were induced at 24 hpi, nuclear and cytoplasmic fractions were separated using NEB, and protein was extracted to verify correct nuclear and cytoplasmic fraction separation using antibodies specific for nuclear H3 and cellular glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Fig. 3B). These observations indicated that the ENE plays an important role in retaining PAN at the nucleus, which is consistent with previous reports showing the nuclear retention of a β-globin–ENE construct (19).

FIG 3.

The PAN ENE facilitates nuclear retention. (A) RNA was extracted from nuclear and cytoplasmic fractions at 24 and 72 hpi. RNA was treated with DNase and used for cDNA synthesis and qPCR using primers/probes specific for 7SK, PAN, and ORF26. Bar graphs represent means ± SD. Two-way ANOVA Sidak’s multiple-comparison test was used (P value of <0.05). (B) Western blot showing protein expression in nuclear and cytoplasmic fractions.

PAN localizes and aggregates as nuclear punctate spots in the absence of ENE.

qPCR data of nuclear and cytoplasmic fractions demonstrated that in the absence of the ENE, there was higher enrichment of PAN in the cytoplasmic fractions than in the nuclear fraction. We sought to visualize WT PAN and ENE deletion through fluorescent in situ hybridization (FISH). We used specific biotin probes to detect PAN followed by streptavidin coupled to a fluorophore. LacZ was used as a negative control. Surprisingly, we observed that PAN still localizes almost exclusively to the nucleus as punctate spots in the absence of the ENE (Fig. 4). Furthermore, although some PAN was punctate in the WT, the majority appeared more diffused than the aggregated spots in the cells that contained the ΔENE virus. Our observations are consistent with previous studies looking at the localization of PAN with an ENE deletion in a plasmid transfection system (14). Although, based on qPCR of nuclear and cytoplasmic fractions, we detected higher levels of PAN RNA in the cytoplasm for the ΔENE virus, this was not visually apparent by FISH. We hypothesize that this is due to a lack of concentration of PAN in a specific area within the cytoplasm and is below the level of detection for FISH.

FIG 4.

PAN localizes as nuclear punctate spots in the absence of the ENE. (Top) FISH was performed using biotin oligonucleotides specific for either PAN or LacZ as a negative control in iSLK BAC16 ORF59-HA WT and ΔENE cell lines at 12, 24, and 48 hpi. PAN is shown (red), nuclei are stained with DAPI (blue), and merged images are shown. (Bottom) LacZ as a negative control.

Viral and cellular proteins retain the ability to bind to PAN in the absence of the ENE during reactivation.

Since viral and cellular proteins that interact with PAN have been identified, we sought to determine if these proteins retained the ability to bind to PAN in the absence of the ENE. To assess this, RNA immunoprecipitations (RNA-IPs) were performed in iSLK BAC16 ORF59-HA WT and ΔENE cells. RNA-IPs were performed at 48 hpi using HA-, LANA-, ORF57-, H1.4-, H2A-, or IgG-specific antibodies. RNA from the input lysate and immunoprecipitated complexes was isolated and used for reverse transcriptase PCR (RT-PCR) with PCR-specific primers for PAN, cellular U1, and GAPDH. U1-specific primers for ORF57-IP were used as a positive control, and GAPDH was used as a negative control, because ORF57 was previously reported to interact with U1 and other spliceosome-associated factors (28). The absence of reverse transcriptase (no RT) was included in RT-PCRs as an additional control (Fig. 5). The RNA-IPs showed that all assessed proteins were able to bind to PAN in the absence of the ENE. These data suggested that the ENE in PAN is not the required binding site of viral proteins during viral reactivation. Previous studies that identified potential sites for binding of viral proteins to specific regions of PAN had been performed in the absence of infection or using in vitro assays, which may account for the outcome differences (24). It is possible that the viral proteins assessed have multiple binding sites that are required for PAN binding, similarly to what has been reported for the MRE and ORF57 (24).

FIG 5.

ORF59, ORF57, LANA, H1.4, and H2A retain the ability to interact with PAN in the absence of the ENE. RNA-IPs were performed on iSLK BAC16 ORF59-HA ΔENE cells treated with NaB, Dox, and TPA for 48 h. Protein-RNA complexes were immunoprecipitated using antibodies specific for ORF59-HA, ORF57, LANA, H1.4, or H2A. RNA was isolated and used as the template for RT-PCR using primers specific for either PAN, U1, or GAPDH.

Production of infectious virus in BAC16ΔENE.

To determine if the ΔENE virus has defects in the expression of viral proteins, iSLK BAC16 ORF59-HA WT and ΔENE cells were reactivated, and protein was harvested at 24 and 72 hpi. Protein lysates were separated on SDS-PAGE gels, and the expression of LANA, K-RTA, ORF59-HA, ORF57, and GAPDH was determined using specific antibodies (Fig. 6A). As shown in Fig. 6A, compared to the WT, there were no appreciable defects in IE or E protein expression in the ENE deletion virus.

FIG 6.

Production of infectious virus in BAC16ΔENE. (A) Expression of viral proteins in iSLK BAC16 ORF59-HA WT and ΔENE cells at 24 and 72 hpi. (B) Total DNA was harvested at 6, 24, 48, and 72 hpi. DNA was purified and used in qPCRs with KSHV-specific primers/probes to determine the relative levels of viral DNA accumulation compared to normalized cellular DNA and untreated samples. (C) The ENE does not affect infectious virus production. Quantification of infectious virus was performed as described in Materials and Methods. Statistical significance was determined by an unpaired t test (P value of <0.05). Results represent data from three independent experiments. NS, not significant. (D) Rate of replication of WT compared to ΔENE virus. Induced cells at the indicated time points were harvested, and viral genomes were quantified by qPCR using KSHV-specific primers/probes. Experiments were performed in triplicate.

As we found that the deletion of the ENE does not affect the expression of viral transcripts and protein production, we measured viral DNA synthesis in the absence of the ENE. We performed qPCR on DNA harvested from iSLK BAC16 ORF59-HA WT, ΔENE, and ΔPAN cells at 6, 24, 48, and 72 hpi and compared the results to those for DNA harvested from noninduced cells. Using primers and probes specific for ORF26 to detect viral DNA and cellular 7SK as a normalization control, qPCR analysis demonstrated that there are no significant differences in the accumulation of viral DNA (Fig. 6B). This was further tested by quantifying infectious virus production. iSLK BAC16 ORF59-HA WT, ΔENE, and ΔPAN cells were induced for 5 days, and the virus was then harvested through freeze-thawing of infected cells, followed by the removal of cellular debris in the lysate and concentration of the virus by high-speed centrifugation. Purified virus was used to infect 293L cells, and green fluorescent protein-positive (GFP⁺) cells were counted at 72 hpi and plotted as GFP⁺ cells per milliliter (Fig. 6C). There were no significant differences between WT and ΔENE infectious virus production. However, we observed a significant defect in the production of infectious virus in the ΔPAN compared to the WT and ΔENE viruses. Together, these results indicated that the ENE is not required for viral DNA synthesis or the production of infectious virus.

Finally, because of the interaction between PAN RNA and ORF59, the polymerase processivity factor, we were interested in determining if the rate at which viral DNA is synthesized is affected by the absence of the ENE. We performed qPCR on DNA harvested from iSLK BAC16 ORF59-HA WT and ΔENE cells at 6, 24, 48, and 72 hpi. The rate of WT and ΔENE synthesis is reported as base pairs per minute at the indicated times postinduction (Fig. 6D). The numbers of genomes and rates of DNA synthesis in WT and ΔENE viruses are comparable and not significantly different, confirming that the ENE in PAN is not involved in or required for viral DNA synthesis during reactivation.

DISCUSSION

In previous studies investigating the function of PAN during lytic replication, two strategies were used to disrupt PAN, through creating a virus with either a deletion of PAN or antisense oligonucleotide depletion of PAN. These reports concluded that with the absence or reduction of PAN, there was a severe defect in lytic gene expression and a subsequent decrease in virion production (5, 18). To explore this phenotype, we further characterized the function of the ENE of PAN RNA during KSHV replication. Previous work suggested that the ENE plays a role in PAN RNA stability and nuclear retention (4). Part of the interest in the ENE is due to its unique triple-helix structure. ENE-like elements, and the associated triple-helix structure, have been found in cellular noncoding RNAs such as cellular MALAT1 and MENβ (25). Both PAN and MALAT1 ENE-like elements were found to increase the expression of intronless β-globin RNA when ENE-like sequences were inserted into reporter plasmids (19, 25). The triple-helix structure of the ENE in PAN was resolved by crystallography, and later computational analyses identified putative ENEs in four other gammaherpesviruses (21, 22). Interestingly, while the suspected ENE was identified in RRV, retroperitoneal fibromatosis-associated herpesvirus (RFHV), EHV-2, and EHV-5, not all PAN homologues contain an ENE, as can be noted for bovine herpesvirus 4 (BHV-4), which lacks an ENE (22, 29). These findings suggested that either these lncRNAs are not comparable homologues or some PAN equivalents can function without the structure of the triple-helix ENE. There has been much effort put toward understanding the role of the ENE, and many studies have used biochemical and in vitro analyses to conclude that the ENE is an important and functional cis element. There are inherent limitations to in vitro assays, although these experiments are incredibly useful and essential to fundamental knowledge. Despite previous reports, the function of the ENE during the viral replication cycle had yet to be determined.

Here, we report that the ENE is not required for efficient viral replication or the production of infectious virions. Instead, the deletion of the ENE affects the localization of PAN to be distributed in both the cytoplasm and nucleus rather than being retained in the nucleus as observed in the WT. Additionally, in our attempt to visualize PAN localization through FISH, we observed ΔENE PAN localization as nuclear punctate spots, contrary to the diffused localization within the nucleus in WT PAN. The punctate nuclear localization has been noted for other lncRNAs. In fact, cellular MALAT1 has also been observed to localize to the nucleus as speckles (30). Previous studies identified that one of the contributing factors for the nuclear speckles was pre-mRNA splicing factors, which modulate distribution into nuclear speckles (30). The poly(A) tail of PAN has been reported to be sequestered and wrapped around the ENE structure, likely to protect from RNA decay and retention at the nucleus (21). Since the ENE is no longer present, it is possible that the poly(A) tail is exposed and available for RNA processing factors to bind, which may explain the punctate aggregation observed in the PAN ΔENE.

During viral replication, PAN RNA accumulates in the nucleus through a process proposed to involve the ENE, which protects the poly(A) from being degraded by exonucleases (20). Additionally, ORF57 binds to a 9-nt core within PAN to stabilize both WT and ENE deletion forms of the RNA (23, 31). ORF57 binding to PAN prevents PABPN1 and the PAPα/γ-mediated RNA decay (PPD) pathway by preventing the binding of RNA helicase protein MTR4 (hMTR4) (31). We did not observe a significant difference in the amount of PAN RNA in the absence of the ENE. This is likely due to the activity of ORF57 binding to PAN even in the absence of the ENE to protect from cellular RNA decay pathways, which is consistent with what was previously reported by Ruiz et al. (31). Further studies of the ORF57 binding sites within PAN, including the MRE, are complicated by the overlapping reading frame of K7 with the 5′ regions of PAN, which hinders straightforward deletions or nucleotide mutations of the locus. As such, it remains to be seen what the functional role of the MRE is during viral replication.

Late transcripts are protected from the PPD RNA decay pathway albeit mediated through the host nuclear RNA interference (RNAi)-defective 2 (NRDE2) proteins (32). During late viral replication, NRDE2 localizes as nuclear punctate spots with replication compartments where it interacts with MTR4 to prevent PPD complexes from degrading viral late transcripts (32). This observation leads to speculation regarding whether in the absence of the ENE, NRDE2 binds to PAN in addition to ORF57 to protect from RNA decay. Furthermore, because PAN interacts with mRNA export proteins such as ALYREF, it would be interesting to determine if a connection exists between the interaction of PAN, mRNA export proteins, and the export of late transcripts (33). Finally, exploring the possibilities of why PAN localizes as nuclear speckles in the absence of the ENE could provide valuable information on the function of PAN during viral replication as well as help determine if PAN RNA undergoes further processing and regulation in addition to a 5′ cap and poly(A) tail.

An alternate interpretation of the observed data is that without the ENE, PAN RNA punctate spots may be the result of a change in the ability of PAN to interact with certain proteins, resulting in a shift to predominantly bind to viral factors such as LANA, which has previously been reported to also localize as punctate spots in the nucleus (6, 16). LANA is reported to interact with PAN at a region spanning nt 600 to 1062, which overlaps the ENE regions located at nt 894 to 972 (16). Our data suggest that LANA is binding to a region other than the ENE at the 3′ end of PAN, which could lead to the observed punctate aggregations. Since the interaction between LANA and PAN regulates the lytic-latent cycle by PAN preventing LANA from reassociating with the viral genome, it is possible that a disruption of the PAN ENE hinders the regulation of the viral cycle. This could result in a virus that may have a discrete phenotype within the context of other infection models such as in vivo animal studies that could test the reestablishment of latency following lytic reactivation.

PAN has been widely studied, and it has been repeatedly demonstrated to be essential for replication (5, 18). As we focused on our studies involving the ENE of PAN, we compared the BAC16ΔENE virus to the BAC16ΔPAN virus, which has the majority of PAN deleted except for the overlapping region that is shared with K7. This deletion has been previously characterized in BAC36 (5). Unexpectedly, we observed that in iSLK cells, BAC16ΔPAN retained the ability to accumulate transcripts and DNA, contrary to what was previously reported for BAC36ΔPAN and antisense oligonucleotide knockdown of PAN in BCBL1 cells (5, 18). It is possible that this observation could be due to the presence of the inducible K-RTA within the iSLK cells, which promotes reactivation and viral replication. However, we observed that BAC16ΔPAN still had a significant defect in infectious virus production, which is consistent with what was previously reported for mutants or knockdown of PAN. Additionally, other studies using BAC16 with a mutation of the interaction domain of K-RTA (RRE) within the PAN promoter showed a decrease in PAN RNA production and a defect in chromatin looping, which resulted in a decrease in infectious virus production (11). These results show that PAN plays an important role in replication, especially in infectious virus production. Additionally, the inducible K-RTA expressed in iSLK cells may be sufficient to support transcription and DNA accumulation without the presence of PAN. The defect in the production of infectious virus with BAC16ΔPAN, while not displaying a significant reduction in viral gene expression or DNA synthesis, suggests that it may be a defect in assembly and late protein localization. It is also important to note that the 5′-most sequence, which overlaps K7, remains within the KSHV BAC16 genome. This leads to speculation that the 5′ region present within the PAN deletions may be sufficient to also support transcript and DNA accumulation. Further characterization studies of this region, which includes the MRE cis element, may reveal that the domains of PAN that grant its function in replication may be clustered within the 5′ region.

MATERIALS AND METHODS

Cell lines.

iSLK cells were maintained in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Corning) in 1 μg/ml of puromycin and 250 μg/ml of G418 (InvivoGen). iSLK cells containing a KSHV bacterial artificial chromosome (BAC) were maintained in the medium described above with the addition of 1.2 mg/ml of hygromycin B (U.S. Biologicals). For the induction of lytic reactivation in the above-described cell lines, 0.25 mM sodium butyrate (NaB) (Sigma-Aldrich), 1 μg/ml doxycycline (U.S. Biologicals), and 10 ng/ml of phorbol 12-myristate 13-acetate (TPA) (Sigma-Aldrich) were used at the indicated time points. 293L cells (obtained from the NIH AIDS reagent program) were maintained in DMEM with 10% FBS. All cell lines were grown at 37°C in a humidified incubator supplemented with 5% CO₂.

Recombinant bacmid with ENE deletion.

The protocol for BAC mutagenesis was performed as previously described. Briefly, a KSHV BAC16 ENE mutant was generated by homologous recombination using gBlocks (IDT) that were designed to contain the kanamycin cassette and the I-SceI restriction enzyme site. The cassette was flanked by a sequence homologous to that of the PAN ENE region for removal. The following gBlocks for BAC16 ORF59-HA ΔENE were used: GCTAGGTTGACTAACGATGTTTTCTTGTAGGTGAAAGCGTTGTGTAACAATGATAACGGGGAAAAACATGTTATACTTTcgatttattcaacaaagccacgttgtgtctcaaaatctctgatgttacattgcacaagataaaaatatatcatcatgaacaataaaactgtctgcttacataaacagtaatacaaggggtgttatgagccatattcaacgggaaacgtcttgctcgaggccgcgattaaattccaacatggatgctgatttatatgggtataaatgggctcgcgataatgtcgggcaatcaggtgcgacaatctatcgattgatgggaacccgatgcgccagagttgtttctgaaacatggcaaaggtagcgttgccaatgatgttacagatgagatggtcagactaaactggctgacggaatttatgcctcttccgaccatcaagcattttatccgtactcctgatgatgcatggttactcaccactgcgatccccgggaaaacagcattccaggtattagaagaatatctgatcggtgaaaatattgttgatgcgctggcagtgttcctgcgccggttgcattcgattcctgtttgtaattgtccttttaacagcgatcgcgtatttcgtctcgctcagcgcaatcacgaatgaataacggtttggttgatgcgagtgattttgatgacgagcgtaatggctggcctgttgaacaagtctggaaagaaatgcataagcttttgccattctcaccggattcagtcgtcactcatggtgatttctcacttgataaccttatttttgacgaggggaaattaataggttgtattgatgttggacgagtcggaatcgcagaccgataccaggatcttgccatcctatggaactgcctcggtgagttttctccttcattacagaaacggctttttcaaaaatatggtattgataatcctgatatgaataaattgcagtttcatttgatgctcgatgagtttttctaatcagaattggttaattggttgtaacactggcattaccctgttatccctagatcgatgtcgggccagatatacgcgTTGTGTAACAATGATAACGGGGAAAAACATGTTATACTTTTGACAATTTAACGTGCCTAGAGCTCAAATT for BAC16 ORF59-HA ΔENE and GCGGGTTATTGCATTGGATTCAATCTCCAGGCCAGTTGTAGCCCCCTTTTATGATATGCGTACTTTTGACAATTTAACGTcgatttattcaacaaagccacgttgtgtctcaaaatctctgatgttacattgcacaagataaaaatatatcatcatgaacaataaaactgtctgcttacataaacagtaatacaaggggtgttatgagccatattcaacgggaaacgtcttgctcgaggccgcgattaaattccaacatggatgctgatttatatgggtataaatgggctcgcgataatgtcgggcaatcaggtgcgacaatctatcgattgtatgggaagccgatgcgccagagttgtttctgacatggcaaaggtagcgttgccaatgagttacagatgagatggtagactaaactggctgacggaatttatgcctcttccgaccatcaagcattttatccgtactcctgatgatgcatggttactcaccactgcgatccccgggaaaacagcattccaggtattagaagaatatcctgattcaggtgaaaatattgttgatgcgctggcagtgttcctgcgccggttgcattcgattcctgtttgtaattgtccttttaacagcgatcgcgtatttcgtctcgctcaggcgcaatcacgaatgaataacggtttggttgatgcgagtgattttgatgacgagcgtaatggctggcctgttgaacaagtctggaaagaaatgcataagcttttgccattctcaccggattcagtcgtcactcatggtgatttctcacttgataaccttatttttgacgaggggaaattaataggttgtattgatgttggacgagtcggaatccagaccgataccaggatcttgccatcctatggaactgcctcggtgagttttctccttcattacagaaacggctttttcaaaaatatggtattgataatcctgatatgaataaattgcagtttcatttgatgctcgatgagtttttctaatcagaattggttaattggttgtaacactggcattaccctgttatccctagatcgatgtacgggccagatatacgcgGCCCCCTTTTATGATATGCGTACTTTTGACAATTTAACGTGCCTAGAGCTCAAATTAAACTAATACCATAACGTAATGCA for ΔPAN. Underlining indicates the homologous regions used to target specific viral sequences (boldface type indicates the sequence used for initial insertion) and for the generation of the ENE and PAN deletions. The remaining lowercase sequences indicate the Kan cassette. Thirty nanograms of the gBlock was electroporated (0.1-cm cuvette, 1.8 kV, 200 Ω, and 25 μF) into competent Escherichia coli GS1783 cells harboring BAC16. The Kan^r/I-SceI-containing mutants were selected on chloramphenicol and kanamycin plates. Individual bacterial colonies were selected; the correct clones were confirmed by PCR, restriction enzyme digestion, and Southern blotting; and the ENE deletion mutant was confirmed by whole-genome sequencing. BAC DNA was purified using the NucleoBond Xtra BAC kit (Clontech) according to the manufacturer’s protocol. Twenty microliters to 30 μl of BAC DNA from the BAC16 ENE mutant was transfected into iSLK cells with FuGene HD reagent at a 1:4 ratio according to the manufacturer’s instructions (Promega). iSLK cells containing the ENE mutant bacmid were selected with 1.2 mg/ml of hygromycin to obtain a population of cells containing the bacmid.

Sequencing.

WT BAC16 ORF59-HA and the BAC16 ORF59-HA ΔENE and BAC16 ORF59-HA ΔPAN mutants were sequenced at the UNR Genomics Center using an Illumina NextSeq500 platform. Fasta files were aligned to the reference genome using CLC Genomics Workbench (Qiagen).

Western blotting.

iSLK BAC16 ORF59-HA WT and ΔENE mutant cells were induced for 24 and 72 h with 0.25 mM NaB, 0.5 μg/ml of doxycycline, and 10 ng/ml of TPA. Protein extracts were prepared using 200 μl IP lysis buffer (Pierce, Thermo Fisher) and protease inhibitors (catalogue number P8340; Sigma-Aldrich). Lysates were sonicated using two to three 10-s pulses with a probe disruptor (Misonix 200; Thermo Fisher). Lysates were centrifuged at 13,000 rpm for 10 min at 4°C to remove cellular debris. One hundred microliters of Laemmli sample buffer (Bio-Rad) with beta-mercaptoethanol was added to 100 μl of the protein lysate, and the mixture was boiled for 5 min. Forty-five microliters of protein was separated through an SDS-PAGE gel and subsequently transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad). Blots were blocked for 10 min with 5% skim milk in Tris-buffered saline with 0.1% Tween 20 (TBST) (10 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.05% Tween 20) and then incubated with the following specific antibodies: rabbit anti-HA at 1:2,000 (catalogue number H6908; Sigma-Aldrich), rabbit anti-RTA at 1:5,000 (provided by Don Ganem, UCSF), mouse anti-ORF57 at 1:400 (catalogue number SC-135746; Santa Cruz Biotechnology), rat anti-LANA at 1:500 (catalogue number 13-210-100; Advanced Biotechnologies), and rabbit anti-GAPDH at 1:2,500 (catalogue number 128915; Abcam). All primary antibodies were incubated at 4°C overnight, followed by washing with TBST and incubation with either IR-dye 680 and 800 anti-mouse or anti-rabbit (Li-Cor) secondary antibody conjugated for 30 min, and proteins were detected using a ChemiDoc MP imaging system (Bio-Rad).

RNA immunoprecipitation.

RNA immunoprecipitation (RNA-IP) was performed as previously described (5). Briefly, one 10-cm dish was induced with 0.25 mM NaB, 0.5 μg/ml doxycycline, and 10 ng/ml TPA for 48 h. Cells were fixed with 1% formaldehyde for 15 min and quenched with 250 mM glycine. Cells were washed once with cold phosphate-buffered saline (PBS), and cells were harvested in 1 ml IP lysis buffer (Pierce, Thermo Fisher) with protease inhibitors (Sigma-Aldrich), RNase Out (Invitrogen), and 1 mM phenylmethylsulfonyl fluoride (PMSF) (Sigma-Aldrich). Samples were sonicated using 10 20-s pulses with a probe disruptor (Misonix 200; Thermo Fisher). Following sonication, samples were spun down for 5 min at 800 × g at 4°C to remove debris. Ten percent of the lysate was saved as the input. Protein G magnetic beads (Thermo Fisher, Life Technologies) were preblocked with 1% bovine serum albumin (BSA) and 10 μg/ml of yeast RNA for 30 min.

Protein-RNA complexes were immunoprecipitated with anti-IgG, anti-HA (1:200), anti-LANA (1:200), anti-ORF57 (1:40), anti-H1.4 (1:200) (catalogue number CD4J50; Cell Signaling), or anti-H2A (1:200) (catalogue number 18255; Abcam) antibody and RNaseOUT. Fifty microliters of preblocked protein G magnetic beads was added to each sample, and immunoprecipitates were rotated overnight at 4°C. The beads were washed once with low-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 7], and 150 mM NaCl), high-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 7], and 500 mM NaCl), and LiCl wash buffer (0.25 M LiCl, 1% NP-40, 1% Na-deoxycholate, 1 mM EDTA, and 10 mM Tris-HCl [pH 7)] and twice with 1× Tris-EDTA (TE). The input and beads were resuspended with 100 μl and 200 μl of reverse buffer (10 mM Tris-HCl [pH 7], 5 mM EDTA, 10 mM dithiothreitol [DTT], and 1% SDS), respectively. One hundred microliters of 2× proteinase K buffer (10 mM Tris-HCl [pH 7], 1 mM EDTA, 0.5% SDS, 100 mM NaCl, and 5% proteinase K) was added to the input and beads. The cross-link was reversed for 1 h at 42°C followed by 1.5 h at 65°C. RNA was extracted by using TRIzol LS (Thermo Fisher, Life Technologies) according to the manufacturer’s instructions. DNA was removed from extracted RNA using a Turbo DNA-free kit (Thermo Fisher). cDNA was generated using iScript reverse transcription supermix for RT-qPCR (Bio-Rad). PCR was performed using Prime Star Max (TaKaRa) with primers specific for PAN (forward, 5′-TTGGCCTGAGAGCTGTAGTA-3′; reverse, 5′-CACAGAACCGAAACAACGAATG-3′), U1 (forward, 5′-ATACTTACCTGGCAGGGGAG-3′; reverse, 5′-CAGGGGAAAGCGCGAACGCA-3′), and GAPDH (forward, 5′-GAAATCCCATCACCATCTTCCA-3′; reverse, 5′-GAGTCCTTCCACGATACCAAAG-3′).

Real-time qPCR.

For quantitating viral transcripts, iSLK BAC16 ORF59-HA WT, ΔENE, and ΔPAN cells were induced and harvested at 6, 24, 48, and 72 hpi, and uninduced cells were used as controls. Total RNA was isolated using a PureLink RNA minikit (Thermo Fisher, Invitrogen). DNA was removed from extracted RNA using a Turbo DNA-free kit (Thermo Fisher, Invitrogen). cDNA was generated using an iScript cDNA synthesis kit (Bio-Rad) and amplified on a T100 thermal cycler (Bio-Rad). Transcripts were measured by qPCR using SsoAdvanced universal probe supermix (Bio-Rad) according to the manufacturer’s instructions. KSHV-specific primers were normalized to cellular 7SK. Each qPCR was performed in triplicate. KSHV primers/probes are described in Table 1.

TABLE 1.

Primer pairs used for qPCR^a

Gene	Sequence
	Primer 1	Primer 2	Probe
7SK	5′-TGA CTA CCC TAC GTT CTC CTA C-3′	5′-GTC AAG GGT ATA CGA GTA GCT G-3′	5′–56-FAM–CCC TGC TAG–ZEN–AAC CTC CAA ACA AGC T–3IABkFQ–3′
ORF9	5′-AGG TGG GAA ACT ACT GTG TTA TT-3′	5′-TGG CCA GCT TGG CTA TTT-3′	5′–56-FAM–TGC TAC GGT–ZEN–TTC AGA CCC ATG TTG AG–3IABkFQ–3′
ORF26	5′-ATG GCA CTC GAC AAG AGT ATA G-3′	5′-GGC AGT ACG CTC CCT ATT T-3′	5′–56-FAM–AGA CTC TTC–ZEN–GCT GAT
ORF50	5′-CTG GTA CAG TCC TTG CAG AAT A-3′	5′-CAG AGT CTA TTC GCC CTG TTA G-3′	5′–56-FAM–CCC TGA AAC–ZEN–ATG GGA TGT CGG GTC–3IABkFQ–3′
ORF57	5′-GAT GGG AAA CCG CTT AGT AGA G-3′	5′-GCC TGG GAT AGT TAG GAC AAA G-3′	5′–56-FAM–TAA CCT TCT–ZEN–TGG CGA GGT CAA GCT–3IABkFQ–3′
ORF58/59	5′-GTG GCT AGC GGA TAA GGT AAC-3′	5′-CAG TGT TTG CTG CTG TAG TTT G-3′	5′–56-FAM–CCG GTA TGA–ZEN–AGG GCA CAC GAG AAA–3IABkFQ–3′
ORF K2	5′-TCC CTG AAG CCT CCC TAA TA-3′	5′-GAA GAC CTT AGG ATG GGA CAT AC-3′	5′–56-FAM–TTT GGG TGG–ZEN–ACT GTA GTG CGT CTT–3IABkFQ–3′
PAN	5′-CGATTTACACTCAATCCGCTTTC-3′	5′-GTGCACTACCTATCTGCTCATT-3′	5′–56-FAM–CGTTGTTTC–ZEN–GGTTCTGTGTTTGTCTGA–3IABkFQ–3′
ORF K13	5′-TAACCTGCCCTCCTCCTTTA-3′	5′-GGATGACAGGGAAGTGGTATTG-3′	5′–56-FAM–TTCATACCT–ZEN–CAACCCACACTGGCC–3IABkFQ–3′

^a56FAM, fluorescein; ZEN, Integrated DNA Technologies (IDT) proprietary quencher; 3IABkFQ, 3’ Iowa Black FQ quencher.

For quantitating transcripts in nuclear and cytoplasmic fractions, cells were washed with cold PBS followed by the addition of 300 μl of nuclear extraction buffer (NEB) (20 mM HEPES [pH 7.2], 50 mM NaCl, 3 mM MgCl₂, 300 mM sucrose, and 0.5% NP-40) for 15 min at 4°C. Cells were collected, and nuclei were pelleted at 800 × g for 10 min at 4°C. The cytoplasmic supernatant was collected, the nuclear pellet was washed once with 0.5 ml of cold PBS and spun down at 800 × g for 10 min at 4°C, and nuclei were resuspended in 300 μl of NEB. Nuclei were sonicated twice for 10 s. RNA was extracted using TRIzol LS (Thermo Fisher, Life Technologies) according to the manufacturer’s instructions. Any DNA carryover was removed using a Turbo DNA-free kit (Thermo Fisher, Invitrogen). cDNA synthesis and qPCR were performed as described above.

For quantitating viral DNA synthesis, iSLK BAC16 ORF59-HA WT, ΔENE, and ΔPAN cells were induced and harvested at 6, 24, 48, and 72 hpi, and uninduced controls were used. Total genomic DNA was harvested by adding DNA extraction buffer (2% SDS, 100 mM Tris-HCl [pH 8.0], 10 mM EDTA, and 50 mg/ml of proteinase K). Cell lysates were incubated at 65°C for 1 to 2 h, followed by phenol-chloroform extraction and then chloroform extraction. DNA was precipitated with 100% ethanol and resuspended in 1× TE. DNA was analyzed by qPCR in triplicate using KSHV-specific primers (Table 1) normalized to cellular 7SK.

Fluorescent in situ hybridization.

The fluorescent in situ hybridization (FISH) protocol was performed as previously described, with some modifications (34). iSLK BAC16 ORF59-HA WT and ΔENE cells were grown on coverslips and induced at 12, 24, and 48 h. Cells were washed with PBS, fixed in 4% paraformaldehyde for 30 min, washed with PBS three times, and permeabilized with 0.5% Triton X-100 for 10 min. Cells were washed with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and prehybridized with hybridization buffer (50% formamide, 10% dextran sulfate, 2× SSC, 0.1% BSA, 500 μg/ml of salmon sperm DNA, 125 μg/ml yeast RNA, and 1 mM vanadyl ribonucleoside complexes) for 1 h at 37°C. All probes were biotinylated at the 3′ end with an 18-carbon spacer arm as previously described (12). Biotin probes were pooled at 10 pmol each in hybridization buffer and incubated overnight at 37°C with the samples. Cells were washed twice for 10 min with 1× SSC. Cells were fixed a second time with 4% paraformaldehyde for 15 min. Cells were washed three times with 1× PBS and permeabilized with 0.5% Triton X-100 for 10 min. Cells were incubated with 25 μg/ml of streptavidin 594 in 0.1% BSA–PBS for 1 h at room temperature with rocking and washed three times with 1× PBS. Cells were mounted onto glass slides using ProLong gold antifade reagent with 4′,6-diamidino-2-phenylindole (DAPI) (Thermo Fisher, Life Technologies) and visualized using a fluorescence microscope (Carl Zeiss, Inc.).

Infectious virus assay.

Confluent 10-cm dishes (approximately 5 × 10⁶ cells) of iSLK BAC16 WT, ΔENE, and ΔPAN cells were induced for 5 days. Cells were harvested and freeze-thawed three times. Debris was spun down three times for 10 min at 4,000 rpm. Virus was pelleted by centrifugation at 28,000 rpm for 90 min in an SW28 swinging-bucket rotor. The pellet was resuspended in 0.5 ml of medium. Semiconfluent 6-cm dishes of 293L cells were infected at different serial dilutions, and GFP foci were counted at 72 hpi. GFP⁺ cells per milliliter were plotted.

Quantification of viral rate of replication.

A total of 250,000 iSLK BAC16 WT and ΔENE cells were plated onto 3.5-cm dishes and induced for 0, 6, 24, and 72 h. Total genomic DNA was harvested as described above. DNA was analyzed by qPCR with SsoAdvanced universal probe supermix (Bio-Rad) according to the manufacturer’s instructions. KSHV-specific primers were used to quantitate viral DNA synthesis and were normalized to 7SK. Each qPCR was performed in triplicates. KSHV primers/probes used are described in Table 1. Standard curves were generated using purified KSHV BAC16 DNA. The rate of viral DNA replication was calculated as base pairs synthesized per minute using the following formula, as previously reported (27):

$$\frac{\text{136,000 bp/genome}\times \text{[log2(final number of genomes)}\hspace{0.17em}-\hspace{0.17em}\text{[log2(initial number of genomes)]}}{\text{\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}Duration}}$$

Data availability.

All sequences have been deposited in the NCBI Sequence Read Archive (SRA) and are accessible through the following SRA series accession numbers: SRX7856829 for KSHV WT BAC16 ORF59-HA, SRR13285304 for KSHV BAC16 ORF59-HA ΔENE, and SRR13748339 for KSHV BAC16 ORF59-HA ΔPAN.