Kaposi's Sarcoma-Associated Herpesvirus Deregulates Host Cellular Replication during Lytic Reactivation by Disrupting the MCM Complex through ORF59

KSHV is the causative agent of various lethal malignancies affecting immunocompromised individuals. Both lytic and latent phases of the viral life cycle contribute to the progression of these cancers. A better understanding of how viral proteins disrupt functions of a normal healthy cell to cause oncogenesis is warranted. One crucial lytic protein produced early during lytic reactivation is the multifunctional ORF59. In this report, we elucidated an important role of ORF59 in manipulating the cellular environment conducive for viral DNA replication by deregulating the normal functions of the host MCM proteins. ORF59 binds to specific MCMs and sequesters them away from replication origins in order to sabotage cellular DNA replication. Blocking cellular DNA replication ensures that cellular resources are utilized for transcription and replication of viral DNA.

ABSTRACT

Minichromosome maintenance proteins (MCMs) play an important role in DNA replication by binding to the origins as helicase and recruiting polymerases for DNA synthesis. During the S phase, MCM complex is loaded to limit DNA replication once per cell cycle. We identified MCMs as ORF59 binding partners in our protein pulldown assays, which led us to hypothesize that this interaction influences DNA replication. ORF59's interactions with MCMs were confirmed in both endogenous and overexpression systems, which showed its association with MCM3, MCM4, MCM5, and MCM6. Interestingly, MCM6 interacted with both the N- and C-terminal domains of ORF59, and its depletion in BCBL-1 and BC3 cells led to an increase in viral genome copies, viral late gene transcripts, and virion production compared to the control cells following reactivation. MCMs perform their function by loading onto the replication competent DNA, and one means of regulating chromatin loading/unloading, in addition to enzymatic activity of the MCM complex, is by posttranslational modifications, including phosphorylation of these factors. Interestingly, a hypophosphorylated form of MCM3, which is associated with reduced loading onto the chromatin, was detected during lytic reactivation and correlated with its inability to associate with histones in reactivated cells. Additionally, chromatin immunoprecipitation showed lower levels of MCM3 and MCM4 association at cellular origins of replication and decreased levels of cellular DNA synthesis in cells undergoing reactivation. Taken together, these findings suggest a mechanism in which KSHV ORF59 disrupts the assembly and functions of MCM complex to stall cellular DNA replication and promote viral replication.

IMPORTANCE KSHV is the causative agent of various lethal malignancies affecting immunocompromised individuals. Both lytic and latent phases of the viral life cycle contribute to the progression of these cancers. A better understanding of how viral proteins disrupt functions of a normal healthy cell to cause oncogenesis is warranted. One crucial lytic protein produced early during lytic reactivation is the multifunctional ORF59. In this report, we elucidated an important role of ORF59 in manipulating the cellular environment conducive for viral DNA replication by deregulating the normal functions of the host MCM proteins. ORF59 binds to specific MCMs and sequesters them away from replication origins in order to sabotage cellular DNA replication. Blocking cellular DNA replication ensures that cellular resources are utilized for transcription and replication of viral DNA.

INTRODUCTION

Kaposi's sarcoma-associated herpesvirus (KSHV) is the etiological agent responsible for Kaposi's sarcoma (1) and other B cell proliferative disorders, including primary effusion lymphoma (PEL) and multicentric Castleman's disease (MCD) (1, 2). Infected immunocompetent individuals harbor the virus in a latent (dormant) state for lifetime; however, immunocompromised patients are at great risk for developing multiple lethal malignancies. As a double-stranded DNA (dsDNA) herpesvirus with a large genome (approximately 165 kb), KSHV encodes nearly 90 open reading frames (ORFs), long noncoding RNAs, and several micro-RNAs.

The biphasic life cycle of KSHV includes two distinct modes, a latent state and a lytic, infectious virion-producing state. During latency, the viral genome is maintained as an extrachromosomal episome tethered to the host chromosomes with only a limited number of viral genes expressed (3). The multifunctional LANA, encoded by ORF73, is responsible for successful replication and segregation of the KSHV genome copies into dividing cells (4, 5). LANA recruits a number of cellular proteins and DNA replication factors to ensure a lifelong persistent infection (6). The low-profile latent infection remains unrecognized by the host's immune surveillance due to its ability to manipulate both the innate and adaptive arms of the immune system (7).

However, when the latently infected cells are perturbed by environmental stimuli, including hypoxia, inflammatory signals, immune suppression, viral coinfection, or treatment with specific compounds, namely, histone deacetylase (HDAC) inhibitors, lytic reactivation may be triggered and KSHV shifts into the lytic phase of its life cycle. During this productive stage, a synchronized cascade of transcriptional activation of viral genes and rapid replication of viral genomes facilitates packaging, assembly, and egress of infections virions. Lytic reactivation plays a crucial role in inducing KSHV-mediated malignancies as well as in producing viral progenies (8).

Among the viral proteins expressed early during lytic reactivation is a DNA processivity factor, ORF59 (9, 10). ORF59 is shown to form a homodimer in the cytoplasm, which interacts with the viral DNA polymerase, translocates it to the nucleus, and assembles on the viral origin of lytic replication (OriLyt) (11–14). We have previously shown that phosphorylation of ORF59 at Ser378 and Ser379 is required for ORF59's role in viral DNA synthesis and production of virions (15). More recently, we determined that ORF59 directly disrupts the activity of a chromatin-modifying enzyme, PRMT5, to promote an epigenetic conformation that favors active transcription of the viral genes (16). This, along with work from other investigators, suggested that aside from its crucial role in DNA processivity, ORF59 exerts additional functions to promote lytic reactivation (17, 18). In each of these instances, ORF59 acts through an association with a particular protein or set of host proteins to implement changes to the cellular conditions that in one way or another support lytic reactivation. In proteomics screening to identify cellular binding partners of ORF59, another fascinating set of interacting proteins identified was the components of the minichromosome maintenance (MCM) complex (16).

The MCM proteins are a component of the DNA replication licensing factors with helicase activity, which function to limit DNA replication to once per cell cycle (19). MCM2, MCM3, MCM4, MCM5, MCM6, and MCM7 (termed MCM2-7) are expressed in all eukaryotes and associate to form a heterohexameric complex at a 1:1:1:1:1:1 ratio (20). The core of this replicative complex, MCM4, MCM6, and MCM7, contains the DNA helicase activity, which utilizes ATP hydrolysis to unwind dsDNA in a toroidal fashion (21, 22).

During G₁ phase, inactive MCM2-7 are recruited to the origins of replication by Cdt1 to join Cdc6 and the origin recognition complex (Orc1 to Orc6, or Orc1-6) in the formation of prereplication complex (pre-RC) (23). Cdc6 and Orc1-6 of the pre-RC actively load MCM2-7 onto DNA, which are required for helicase activity following activation (24). Once the cells enter into S phase, MCM2-7 can no longer be loaded onto the pre-RCs, effectively limiting DNA replication to one round per cell cycle (25). During this phase, cyclin-dependent kinases (CDKs) and Dbf4-dependent kinase (DDK) activate MCM2-7 helicase activity and promote loading of additional replication proteins, including Cdc45 and the GINS complex (26, 27). Replication forks are assembled, and DNA polymerase is loaded before the initiation of bidirectional DNA replication (28, 29). MCM2-7 are part of the traveling replication forks and are necessary for DNA elongation; thus, the disruption, inactivation, or absence of any of the MCM2-7 proteins halts the progression of replisome (30).

Considering the significance of the MCM complex in regulating a healthy cell cycle, it comes as no surprise that certain viruses influence the normal activities of these proteins to favor their survival. Members of the gammaherpesvirus family, Epstein-Barr virus (EBV) and betaherpesvirus human cytomegalovirus (HCMV), both downregulate cellular chromosomal DNA replication during lytic infection (31, 32). EBV expresses a Ser/Thr protein kinase, BGLF4, early during lytic reactivation, which phosphorylates MCM4 to prevent DNA helicase activity and halts the cell cycle progression (33). Similarly, HCMV sabotages competitive cellular DNA replication by preventing assembly of the prereplication complex and inhibiting MCM loading onto the chromatin (32). Given the association between ORF59 and the MCM proteins in our proteomics assays, we were eager to investigate the implications of this interaction and determine the mechanism by which KSHV restricts cellular DNA replication.

Interestingly, our data showed that depletion of MCM6 (a component of the MCM complex) from KSHV-infected cells enhanced KSHV genome copies and increased the numbers of virions produced following lytic reactivation. We confirmed the binding between ORF59 and members of the MCM complex by coimmunoprecipitation assays in overexpression and endogenous settings. One means by which the MCM's activity is modulated for loading and unloading onto the cellular chromatin is through changes in their phosphorylation status (20). Our results showed a hypophosphorylated form of MCM3 that lacks the ability to load onto the chromatin during lytic reactivation. Changes in the phosphorylation status of MCMs suggested that MCMs dissociate from the chromatin during lytic reactivation, and in fact immunoprecipitation of MCMs (MCM3, MCM4, and MCM6) failed to coprecipitate histone H3, confirming loss of MCM association with chromatin. Further, the levels of MCM3 and MCM4 bound to the cellular origins of replication were depleted upon reactivation. These data strongly support that virus interferes with cellular DNA replication by disrupting the action of the MCM complex through its interactions with ORF59.

RESULTS

ORF59 associated with endogenous MCM proteins.

Since MCM2-7 proteins associate with each other to form the heterohexameric complex, we were interested in determining which components directly bind to ORF59. To this end, endogenous MCM immunoprecipitations (IPs) were performed from two KSHV-infected cell lines, BCBL-1 and iSLK.219, induced for lytic reactivation for 24 h. Immunoprecipitation of MCM3, MCM4, MCM5, and MCM6 were able to coprecipitate ORF59 with various efficiencies (Fig. 1). Anti-MCM3 antibody efficiently immunoprecipitated MCM3 and specifically coprecipitated a detectable level of ORF59 compared to that of the control, IgG antibody (Fig. 1B and H, ORF59 band in lane 3 above the heavy chain). Similarly, anti-MCM4 antibody precipitated MCM4, which showed specific association with endogenous ORF59, as the control antibody (IgG) did not show detectable levels of ORF59 (Fig. 1C and I, lanes 3, ORF59 band above the heavy chain). Remarkably, although the detection and immunoprecipitation of MCM5 with specific antibody was only modest, this protein clearly coprecipitated ORF59 in a specific fashion (Fig. 1D and J, compare lanes 3 with IgG in lanes 2). Additionally, immunoprecipitated MCM6 showed an efficient and specific coprecipitation of ORF59 compared to that of the control, IgG antibody (Fig. 1E and K, compare lanes 3 and 2). Importantly, only two components, MCM2 and MCM7 of the MCM complex, failed to coprecipitate endogenous ORF59 protein from both BCBL-1 (Fig. 1A and F) and iSLK.219 (Fig. 1G and I) cells despite robust expression of ORF59. This confirmed that only certain MCMs (MCM3, MCM4, MCM5, and MCM6) interact with viral ORF59 during lytic reactivation.

FIG 1.

MCM complex associates with KSHV processivity factor. Endogenous MCM proteins were immunoprecipitated from BCBL-1 cells induced for 24 h (A to F) or iSLK.219 cells induced for 24 h (G to L). Inp, input. Each MCM IP (lane 3) was accompanied by an IgG control (lane 2) to ensure specificity. IgG control antibody displayed no affinity for either endogenous MCM proteins or endogenous ORF59 protein, showing that MCM2-7 antibodies specifically precipitate respective MCMs. (A and G) Goat anti-MCM2 IP did not detect coprecipitating ORF59 (band above the heavy chain [hc]) despite efficient MCM2 IP. (F and I) Similar results were obtained for mouse anti-MCM7 IP, which specifically enriched MCM7 but could not coprecipitate ORF59. (B and H) mouse anti-MCM3 IP enriched MCM3 protein and also coprecipitated ORF59. (C and I) Mouse anti-MCM4 IP also successfully immunoprecipitated MCM4 and ORF59. (D and J) Anti-MCM5 antibody efficiently precipitated ORF59. (E and K) Mouse anti-MCM6 IPs in both cell lines efficiently coprecipitated ORF59. MCM3, MCM4, MCM5, and MCM6 proteins in both BCBL-1 cells (B to E) and iSK.219 cells (H to K) associate with ORF59 24 h after lytic reactivation.

MCMs interact with ORF59 without other viral factors.

After confirming the association between ORF59 and MCM3-6 during lytic reactivation from KSHV-infected cells, it was important to determine whether this interaction was direct or mediated by other viral factors. Also, since the MCM proteins and ORF59 are DNA-binding proteins, it was not unreasonable to assume that the association between these proteins could be mediated by a DNA intermediate. To answer this, HEK293T cells were transfected with a plasmid encoding recombinant ORF59-HA, and the endogenous MCM proteins were immunoprecipitated using specific antibodies after treating the lysates with DNase, which was confirmed by an amplification of the β-actin gene (Fig. 2G). Not surprisingly, the same MCMs (MCM3, MCM4, MCM5, and MCM6) showed an immunoprecipitation of ORF59 even after treating the lysates with DNase, confirming the specificity of their interaction. Similar to the KSHV-infected cells, MCM2 and MCM7 did not show any precipitation of ORF59 despite an ample expression of ORF59-HA (Fig. 2A and F, lanes 3). MCM3, MCM4, and MCM5, which were efficiently immunoprecipitated with specific antibodies, showed detectable levels of ORF59 (Fig. 2B, C, and D, lanes 3, above the heavy chain). The strongest association of ORF59 after DNase treatment was observed for MCM6, where endogenous MCM6 efficiently coprecipitated ORF59-HA (Fig. 2E, lane 3).

FIG 2.

MCM6 associates with ORF59 in an overexpression system. (A to F) 293T cells were transfected to express ORF59-HA and then used for immunoprecipitation of endogenous MCM2-7 following DNase treatment. Control IgG IPs demonstrate antibody specificity (lane 2), and goat anti-MCM2 (A) and mouse anti-MCM7 (F) failed to coprecipitate ORF59 despite enrichment of the indicated MCM proteins. (B and C) Mouse anti-MCM3 and mouse anti-MCM4 IPs show significant enrichment and coprecipitation of ORF59. (D and E) Rabbit anti-MCM5 and mouse anti-MCM6 readily coprecipitated ORF59 (lane 3). (G) Treatment of the lysate with DNase, followed by isolation of remaining DNA and subsequent PCR amplification of the β-actin gene, showed loss of DNA after DNase treatment compared to the level for the untreated control cell lysate. (H) Immunoblot to detect the levels of MCM6 in shRNA-depleted (MCM6-dep) BC3 and BCBL-1 cells. GAPDH was used as a loading control. (I) MCM6-depleted BCBL-1 and BC3 cells were subjected to MCM3, MCM4, and MCM5 immunoprecipitations following DNase treatment and subsequent detection of coprecipitating ORF59. MCMs precipitated specifically, and the bound ORF59 was detected with only MCM3 and MCM4. Lack of detectable levels of ORF59 with control antibody (IgG) confirmed specificity of this interaction. (J) Schematic depicting GFP-Myc-tagged ORF59 truncation clones generated for binding assays. (K) MCM6 associates with ORF59 (aa 1 to 132) and ORF59 (aa 264 to 396).

It was important to determine whether these ORF59-interacting MCMs bound directly or whether the hexameric assembly of MCMs had any role in their association. To this end, we depleted the levels of the strongest associating MCM (MCM6) from two KSHV-infected cell lines, BCBL-1 and BC3 (Fig. 2H), and assayed the binding of ORF59 with other ORF59-interacting MCMs after treating the lysates with DNase. Immunoprecipitations of MCM3 and MCM4 showed efficient coprecipitation of ORF59 despite the lack of MCM6 in both cell lines (Fig. 2I). However, MCM5 was unable to precipitate ORF59 (Fig. 2I), suggesting that MCM6 helps in bridging the association of ORF59 with some of the components of the hexameric complex.

To gain a better understanding of ORF59's association with MCM6 (strongly associating MCMs), a series of GFP-ORF59-Myc truncations (Fig. 2J) were constructed and used for identifying the domain involved in binding to MCM6. HEK293T cells were transfected with control vector or GFP-ORF59-Myc regions, followed by lysing the cells and treating the lysates with DNase prior to an immunoprecipitation with anti-myc antibody. The expression levels of each GFP-ORF59 segment appear comparable between samples, as do the levels of endogenous MCM6 (Fig. 2K, lanes 1 to 4, inputs). The GFP-ORF59-Myc segments immunoprecipitated with somewhat similar efficiencies, but only ORF59 segment 1 (amino acids [aa] 1 to 132) and ORF59 segment 3 (aa 264 to 396) coprecipitated MCM6 (Fig. 2K, lanes 6 and 8). In contrast, the control vector and ORF59 segment 2 (aa 133 to 264) failed to coprecipitate MCM6, confirming that both the N- and C-terminal domains of ORF59 are primarily responsible for binding to MCM6.

MCM6 knockdown enhanced viral genome replication and late gene transcription.

The direct binding between ORF59, an essential component of viral DNA replication, and the MCM complex, a factor necessary for DNA licensing and cellular replication, suggested two possible outcomes for this interaction: ORF59 could recruit the help of MCMs for viral DNA replication, or binding of ORF59 to MCMs could disturb the usual functions of MCM proteins to hinder cellular DNA replication to indirectly promote viral replication. The first step toward unraveling the evidence for either case was to assess the role of MCM proteins during lytic reactivation.

To this end, BC3 and BCBL-1 cells were transduced using lentiviral vector with a doxycycline-inducible shMCM6 virus and selected on puromycin to generate stable cell lines that harbored the short hairpin RNA (shRNA) to deplete MCM6. The doxycycline induction of BC3 and BCBL-1 showed reduced levels of MCM6 (M6-dep) compared to those of the untreated control cells (34) (Fig. 3A, BC3, and E, BCBL-1, row IB: MCM6). The depletion of MCM6 was not complete but was sufficient for assaying its role in subsequent assays. Aside from the reduced levels of MCM6, immunoblotting (IB) for other proteins, including LANA, showed a comparable level in M6-dep and control cells (34) (Fig. 3A and E, IB: LANA).

FIG 3.

MCM6 knockdown assists viral DNA replication. BC3 (A and D) and BCBL-1 (E to H) cells transduced with shMCM6-inducible lentivirus were treated with doxycycline (MCM6 depleted) or left untreated (control), followed by induction of lytic reactivation by NaB and TPA and the detection of protein levels. MCM6 knockdown was accomplished in both cell lines, and lytic proteins were readily detected 72 h after induction. (B and F) The relative genome copies were significantly increased in MCM6-depleted BC3 and BCBL-1 cells compared to control cells. (C and G) The late gene transcripts ORF21 and ORF65 were also enhanced in lytically reactivating MCM6-depleted cells compared to the control (lytic) cells. (D and H) The number of virions released into the supernatant after 96 h of lytic induction in BC3 and BCBL-1 was higher in cells with MCM6 depletion than in the control cells.

To evaluate the potential role of MCMs in lytic genome replication, control and MCM6-depleted BC3s and BCBL-1 cells were lytically reactivated by the addition of sodium butyrate (NaB) and tetradecanoyl phorbol acetate (TPA), and total DNA was harvested after 72 h postinduction for the quantitation of viral genome copies. Induction of lytic reactivation was confirmed by the detection of immediate-early and early genes RTA, ORF57, and ORF59 (Fig. 3A, BC3, and E, BCBL-1). Furthermore, detection of late gene K8.1 showed slightly increased levels of this virion protein in the MCM6-depleted cells compared to levels for controls (Fig. 3A and E). Strikingly, a significantly greater number of viral genome copies were detected in MCM6-depleted BC3 and BCBL-1 cell lines compared to the control cells (Fig. 3B, BC3, and F, BCBL-1). In addition, the expression of late genes detected by two representative genes, ORF21 and ORF65, were also enhanced in MCM6-depleted cells compared to that of the control cells (Fig. 3C, BC3, and G, BCBL-1). While the levels of these late genes were significantly increased, immune detection of immediate-early RTA and early proteins ORF57 and ORF59 revealed the expression of these proteins was only marginally affected, if at all, in both BC3 and BCBL-1 cells depleted for MCM6 (Fig. 3A and E, lanes 3 and 4). Further, we wanted to determine whether higher viral copies and late gene expression have an effect on progeny virion production. We addressed this by inducing the same number of MCM6-depleted and control BC3 and BCBL-1 cells and collecting the supernatant for quantifying cell-free virions. Consistent with the findings of enhanced levels of both viral genome copies and late viral gene transcripts, MCM6-depleted cells produced an increased number of virions (Fig. 3D, BC3, and H, BCBL-1). These findings suggest that despite a depletion in the levels of a replication protein, it nevertheless yielded conditions more favorable to lytic reactivation.

MCM depletion affects the formation of viral replication foci.

To continue the assessment of cellular and viral DNA replication under the condition of MCM6 knockdown, immunofluorescence staining was performed to visualize replication compartments in BCBL-1 depleted for MCM6. Prior to harvesting and fixing on coverslips, cells were incubated with EdU (5-ethynyl-2′-deoxyuridine) for 30 min. EdU is a modified thymidine analog that gets incorporated into the newly synthesized DNA and can be localized following Click-iT reaction with fluorescently labeled azide to visualize the areas of active DNA synthesis. EdU was clicked with Cy3 fluorophore, and PCNA, ORF59, and LANA were detected by specific antibodies. The specificity of EdU labeling and Click-iT reaction was confirmed by the fact that cells incubated with dimethyl sulfoxide (DMSO) instead of EdU showed no specific staining (Fig. 4B, DMSO). In addition, latent BCBL-1 control or shMCM6-depleted cells stained for LANA and EdU displayed the expected punctate dots of LANA expression. The EdU signals were somewhat ubiquitous throughout the nuclei of latent cells, representing DNA replication foci (Fig. 4A). LANA dots, which represent the location of KSHV episomes, also colocalized with EdU dots, indicating active replication of the viral genome during latency, as expected (Fig. 4A).

FIG 4.

LANA colocalized with EdU foci in BCBL-1 control or MCM6-depleted cells. (A) Uninduced BCBL-1 control cells or BCBL-1 MCM6-depleted cells were EdU labeled, fixed on coverslips, and subjected to Click-iT with Cy3 fluorophore azide (63) to localize the actively replicating DNA. LANA was localized by mouse anti-LANA and mouse Alexa Fluor IR647 (Blue) antibodies. In control and MCM6-depleted latent BCBL-1 cells, LANA colocalized with EdU dots. (B) Control or MCM6-depleted BCBL-1 cells were also treated with DMSO as a negative control for EdU and clicked with Cy3 fluorophore azide (63) to determine the specificity of the click reaction. PCNA was detected using rabbit anti-PCNA and anti-rabbit Alexa Fluor IR647 (blue) antibodies. PCNA showed a specific punctate nuclear pattern, while the DMSO control shows virtually no signal confirming the specificity. (C) MCM6 knockdown increases the formation of viral replication foci. IFA images detected subcellular localization of PCNA (stained using rabbit anti-PCNA and anti-rabbit Alexa Fluor IR647, blue), and actively replicating DNA was detected by Click-iT chemistry and visualized using Cy3 azide fluorophore (63) in BCBL-1 Con or MCM6-dep cells induced for lytic reactivation for 24 h. (D) Individual replication foci (represented by EdU dots) from control or MCM6-depleted cells were manually quantified and plotted. (E) ORF59 was detected using rabbit anti-ORF59 antibody and fluorescently labeled with Alexa Fluor IR647 (blue) in BCBL-1 control (Con) or MCM6-dep cells induced for lytic reactivation for 24 h. Actively replicating DNA was clicked with Cy3 azide fluorophore (63) to detect EdU.

The conformations of DNA replication foci in lytically reactivated control or MCM6-depleted BCBL-1 cells were examined next. Notably, the number of EdU dots in control cells appeared to be less than that in MCM6-depleted cells (Fig. 4A, EdU, and B). PCNA was used for the detection of replication foci (Fig. 4C). Multiple cells from 10 different frames were examined, and the replication foci from those control or MCM6 cells were counted. Interestingly, induction of these cells led to the formation of more discrete replication foci, presumably the viral replication foci, compared to that of the latent cells (compare Fig. 4A and C). Overall, the average number of distinct replication foci was higher in the MCM6-depleted cells than in the control cells (Fig. 4D). To ensure that these replication foci (EdU) corresponded with viral DNA replication, subcellular localization of ORF59 was assessed in combination with actively replicating DNA. ORF59 in both control and MCM6-depleted cells formed a somewhat punctate pattern throughout the nucleus, and this pattern overlapped in places with EdU in those lytic cells (Fig. 4E, merge). This suggested that the majority of the EdU signals detected in lytically reactivated cells correspond to viral DNA replication compartments. These results suggest that the knockdown of MCM6 resulted in the formation of a greater number of replication foci and yielded conditions more favorable for robust lytic DNA replication.

Hypophosphorylated MCM3 is detected upon lytic reactivation.

Since the knockdown of MCM6 was favorable for lytic DNA replication, we were eager to test for any signs of aberrant regulation of the MCM proteins during the course of lytic reactivation. Various members of the MCM proteins have been shown to display a variety of posttranslational modifications to control their activity, including acetylation, sumoylation, and most commonly, phosphorylation (20). The selective phosphorylation of these proteins serves to carefully regulate helicase activity as well as association and loading onto the chromatin (20). To examine potential changes in phosphorylation levels of the MCM proteins, the expression levels of MCM3, MCM4, and MCM6 were examined from equal numbers of latent and lytic BCBL-1 and iSLK.219 cells (Fig. 5A and B, respectively). In both cell lines, levels of MCM proteins were not found to be drastically altered upon lytic reactivation (Fig. 5A and B, compare lane 1 with 2), suggesting they are not targeted for degradation during reactivation. One observation was obvious in regard to the expression of MCM3, which was detected in two different isoforms in lytic cells of both lines (BCBL-1 and iSLK.219). It was hypothesized that the lower band corresponds to a hypophosphorylated form of MCM3, but to confirm that the change in migration pattern of MCM3 observed was due to differences in phosphorylation levels of the protein, further testing was performed.

FIG 5.

Phosphorylation status of MCM3, MCM4, and MCM6 before and after lytic reactivation. Lysates from equal numbers of BCBL-1 cells (A) or iSLK.219 cells (B) left uninduced or induced for 24 h were resolved by denaturing SDS-PAGE (lanes 1 and 2) and using the Phos-tag chemical in conjunction with SDS-PAGE (lanes 3 and 4) to determine the levels of endogenous MCM3, MCM4, and MCM6. No obvious difference could be observed in the banding pattern of either MCM4 or MCM6 protein in either the denaturing SDS-PAGE gel or the Phos-tag gel between uninduced (latent) cells and those induced for 24 h. However, a significant mobility shift was observed with regard to MCM3 after lytic reactivation in both cell lines (MCM3 panel, lanes 2 and 4) (hypophosphorylated MCM3 is demarcated by red arrowheads, and hypophosphorylated MCM4 and MCM6 are demarcated by black arrowheads). (C) HEK293L cells containing BAC36 WT or ORF59 deleted (Δ59) were used for the detection of phosphorylation status on MCMs following induction using denaturing SDS-PAGE gel (lanes 1 to 4) or the Phos-tag gel (lanes 5 to 8). No detectable differences were observed in the levels of MCMs on denaturing SDS-PAGE gel in uninduced (Und) or induced (24hInd) cells. ORF59 was detected in WT BAC36 following induction, as expected (IB: ORF59). GAPDH was used as a loading control. The hypophosphorylated form of MCM3 was detected on MCM3 from WT BAC36 (red arrowhead) but not in ORF59-deleted BAC36 (Δ59) cells. Black arrowheads represent hypophosphorylated MCM4 and MCM6.

The same lysate used for assessing MCM expression levels was additionally run on a Phos-tag gel (Fig. 5A and B, lanes 3 and 4). The Phos-tag chemical is specifically designed to capture phosphorylated Ser/Thr/Tyr and His/Asp/Lys residues and is mixed in combination with divalent metal ions and incorporated in the SDS-PAGE gel. This compound slows the migration of phosphoproteins, and the separation of bands directly correlates with the phosphorylation levels of the protein (35, 36). In our hands, the addition of Phos-tag and MnCl₂ altered the cosmetic appearance of the MCM protein bands, as they appeared in denaturing SDS-PAGE (Fig. 5A) but distinct bands were nevertheless easily distinguished (Fig. 5A, lane 4, red arrow). Consistent with what was observed in conventional denaturing SDS-PAGE, the Phos-tag gel clearly showed two separated bands of MCM3 in the induced cells (Fig. 5B, compare lane 4 with 3 in the IB: MCM3 row). Also, although multiple bands could be observed for MCM4 and MCM6, there did not appear to be a significant difference in the pattern of those bands before or after lytic reactivation (Fig. 5A and B, lanes 3 and 4, MCM4 and MCM6). The apparent dephosphorylation of MCM3, as indicated by a hypophosphorylated form appearing more prominent after lytic reactivation, certainly appeared indicative of a mechanism by which the MCM proteins are disrupted from their normal role in cellular DNA replication after lytic reactivation.

In order to determine a direct role of ORF59 on the phosphorylation status of MCMs, we utilized ORF59-deleted bacterial artificial chromosome 36 (BAC36Δ59) (15) and compared the levels of MCM phosphorylation with wild-type (WT) BAC36. The levels of MCMs were comparable in both WT and ORF59-deleted BAC36 (Δ59) under uninduced as well as induced (for 24 h) cells (Fig. 5C). Expression of ORF59 was detected in WT BAC36 but not in Δ59 BAC36, as expected (Fig. 5C, lane 3, IB: ORF59). Importantly, the lysates from cells induced for 24 h showed a hypophosphorylated MCM3 band only in WT BAC36 but not in ORF59-deleted cells on Phos-tag gel (Fig. 5C, lanes 5 to 8), confirming the role of ORF59 in modulating the phosphorylation status of MCM3 (Fig. 5C, IB: MCM3, lane 7, red arrow). MCM4 and MCM6 did not show any detectable change in their phosphorylation levels in cells lacking ORF59 (Fig. 5C).

MCM3, MCM4, and MCM6 binding with histones is disrupted during lytic reactivation.

MCM proteins are located within the nucleus and regulated according to their association or disassociation with chromatin throughout the cell cycle (20). The interactions between MCM complex and histone H3 was first characterized over 20 years ago, and since then various work studying the nuances of eukaryotic DNA replication has found that phosphorylated MCM3 is more commonly associated with chromatin-rich subcellular compartments (37, 38). By manipulating the phosphorylation status of MCM3, cells have the ability to regulate MCM loading onto chromatin, DNA replication, or checkpoint activation (39). Thus, following the detection of a hypophosphorylated form of MCM3 during lytic reactivation, we hypothesized that this alteration affects the downstream organization of the MCM proteins.

To determine whether lytic reactivation could disrupt the binding of MCM proteins to chromatin, endogenous MCM proteins were immunoprecipitated to evaluate the association between MCM3, MCM4, or MCM6 with histone H3. Equal numbers of latent BCBL-1 or iSLK.219 cells and cells lytically induced for 24 h were subjected to immunoprecipitation with MCM-specific antibodies or control IgG and resolved on a higher-percentage gel to enable the detection of histone H3. Similar to the previous experiments, immunoprecipitation of MCM3 from both BCBL-1 and iSLK.219 cells showed an efficient immunoprecipitation, including the presence of different hypophosphorylated isoforms in the lysates from the lytic reactivated cells (Fig. 6A, BCBL-1, and B, iSLK.219, compare lanes 3 and 6). MCM3 coprecipitated histone H3 from the latent cells, as expected, but surprisingly no coprecipitated H3 was detected in the cells undergoing lytic reactivation despite comparable levels of histone H3 in the input lanes (Fig. 6A and B, compare lanes 3 and 6, IB: H3).

FIG 6.

Association of MCM3, MCM4, and MCM6 with histone proteins is reduced upon lytic reactivation. The lysates from two cell lines, BCBL-1 and iSLK.219, were used for the immunoprecipitation of endogenous MCM proteins and to test their association with chromatin by detecting coprecipitated histone H3. (A and B) Mouse anti-MCM3 IP. (C and D) Mouse anti-MCM4 IP. (E and F) Mouse anti-MCM6 IP. In each instance, the MCM proteins reliably associate with histone H3 in latent cells (lane 3). In lytic cells, despite unchanged expression of histone H3 (lane 4) and a comparable IP efficiency in lytically reactivating cells (MCM panels, lane 6), histone H3 no longer coprecipitated with MCM3, MCM4, or MCM6. IgG control antibody IP failed to enrich either histone H3 or MCM proteins, indicating the interaction with MCMs and H3 during latency is highly specific. (G) HEK293 cells with BAC36WT (36wt) or BAC36 with ORF59 deleted (36Δ59) were induced for 24 h following immunoprecipitation of MCM3 to detect any coprecipitating histone H3. BAC36WT did not show coprecipitating histone H3. ORF59-deleted BAC36 showed detectable levels of histone H3.

These results were also corroborated with MCM4- and MCM6-immunoprecipitated samples. Even though endogenous detection in the input lanes was somewhat faint (Fig. 6C, BCBL-1, and D, iSLK.219, lanes 1 and 4), immunoprecipitation of MCM4 in both BCBL-1 and iLSK.219 displayed a highly efficient and specific enrichment (Fig. 6C and D, lanes 3 and 6, IB: MCM4) from both latent and lytic cells. Importantly, histone H3 coprecipitated with MCM4 in the latent cells (Fig. 6C and D, lane 3, IB: H3) but could not be detected coprecipitating with MCM4 in lytic reactivated cells (Fig. 6C and D, lane 6, IB: H3). Lastly, the immunoprecipitation of MCM6 from both BCBL-1 and iSLK.219 efficiently precipitated MCM6 from the lysate of latent and lytic cells (Fig. 6E and F, lanes 3 and 6, IB: MCM6), along with coprecipitation of H3 from uninduced cells but not from the induced cells (Fig. 6E and F, compare lanes 3 with lanes 6, IB: H3). For each MCM IP, the appropriate control IgG antibody failed to precipitate either MCM bait or H3, indicating that binding between MCM3, MCM4, and MCM6 with H3 from these cells was highly specific, and this binding was disrupted in cells undergoing lytic reactivation.

We confirmed the role of ORF59 on MCM association with histone H3 during reactivation by utilizing BAC36 deleted for the ORF59 gene (BAC36Δ59). HEK293L cells containing WT BAC36 or BAC36Δ59 were induced for 24 h, followed by immunoprecipitation of MCM3 to detect coprecipitating histone H3 in WT BAC36, as expected (Fig. 6G). However, the cells lacking ORF59 showed detectable levels of histone H3, confirming a direct role of ORF59 in MCMs loading onto the chromatin (Fig. 6G).

Presence of MCM3 and MCM4 is reduced at cellular replication origins.

The hypophosphorylation of MCM3 along with the disrupted binding of MCM3, MCM4, and MCM6 with histone H3 suggested that MCM complex was disrupted from chromatin during viral lytic reactivation. As the binding of MCM complex is important for the initiation of DNA replication at cellular origins, it was necessary to determine whether lytic reactivation facilitated the detachment of MCMs from chromatin and more specifically from cellular origins of replication. Although the specific recognition sequences and locations of eukaryotic replication origins are still largely a mystery and the topic of ongoing research, a few origins have been characterized (40). These include the lamin B2 origin and the DHFR (dihydrofolate reductase) origin (34, 41–44). As these specific sites are identified to be the origins of replication and the prereplication complex (pre-RC) is necessary for replication initiation, past studies confirmed the association between the MCM complex at the lamin B2 and DHFR origins using chromatin immunoprecipitation (ChIP) assays (45–48).

ChIPs were performed from uninduced BCBL-1 and iSLK.219 cells or cells lytically reactivated for 24 h. Due to the changes observed in phosphorylation status of MCM3 after lytic reactivation and the fact that MCM4 is a part of the crucial MCM4, MCM6, and MCM7 subcomplex responsible for DNA helicase activity (49), both MCM3 and MCM4 ChIPs were performed to assess their binding to the lamin B2 and DHFR promoters. To test the relative enrichment of the MCM proteins at these cellular origins, previously characterized primer sequences were used (50, 51) (see Table 2). MCM3-bound chromatin DNA was isolated from equal numbers of uninduced or induced (for 24 h) BCBL-1 and iSLK.219 cells, and the enrichment was determined by relative quantitative PCR (qPCR) analysis. In both cell lines, relative binding of MCM3 to both lamin B2 and DHFR promoter regions was decreased following lytic reactivation (Fig. 7A, BCBL-1, and B, iSLK.219). Similarly, the MCM4 ChIP from uninduced and induced (24 h) cells showed a decreased level of MCM4 bound to both the cellular replication origins (lamin B2 and DHFR promoters) in lytic cells compared to that of latent cells (Fig. 7C, BCBL-1, and D, iSLK.219). Collectively, these results show that the binding of MCMs with chromatin at cellular replication origins is disrupted during lytic reactivation.

TABLE 2.

qPCR primers

Primer	Sequence (5′–3′; sense, antisense)
LANA (KSHV genome copies)	TTGCCTATACCAGGAAGTCCCACA
	GGAGGAAGACGTGGTTACGGG
OriLyt L	GCTTCCAGCGAACGGAAATAAG
	ACGGAGAATAGACCAGCCTGAA
ORF21	ACGACGAATCAAGCACCTCCACAA
	CGTAGAGGGAGTTGTCGTGCATTA
ORF65	TGGATCATGACTACGCTCACCA
	CCATCCTCCTCAGATAGGCCTCATAA
TR region (KSHV virion genomes)	GGGGGACCCCGGGCAGCGAG
	GGCTCCCCCAAACAGGCTCA
Lamin B2 promoter	TGCATGCCTAGCGTGTTCTTTTTTTTTTCCAATGA
	GGCAGAACCTAAAATCAAAATGTTTATTGGAGTG
DHFR promoter	TCGCCTGCACAAATAGGGAC
	AGAACGCGCGGTCAAGTTT

FIG 7.

MCM3 and MCM4 ChIP on cellular origins before and after lytic reactivation. BCBL-1 (A and C) or iSLK.219 (B and D) cells left uninduced (gray bars) or induced for 24 h (black bars) were harvested, cross-linked, and utilized for ChIP assay to determine relative binding of both MCM3 (A and B) and MCM4 (C and D) at cellular origins of DNA replication of lamin B2 and DHFR promoter. (A and B) Mouse-anti-MCM3 ChIP did not show a strong association of MCM3 with either the cellular lamin B2 promoter origin or the DHFR cellular origin promoter after lytic reactivation. (C and D) Mouse anti-MCM4 ChIP also showed reduced association of MCM4 at the cellular origins at 24 h of lytic reactivation.

Cellular DNA replication is restricted during lytic reactivation.

To directly test the effects of the apparent loss of MCM proteins at cellular origins of DNA replication, nascent DNA production from these loci was measured. This was accomplished by pulse labeling of newly synthesized DNA by treating both uninduced and induced (for 48 h) BCBL-1 cells with the thymidine analog EdU for 30 min prior to harvesting and extracting total DNA. Total DNA was then sonicated to fragments with an average length of 500 bp before Click-iT reaction was performed. The nascent DNA complexes were captured and purified, and relative quantities were assessed by relative qPCR analysis. The efficiency of this assay was confirmed by quantifying the level of active DNA synthesis at the KSHV lytic origin (OriLyt), which was expected to show an active incorporation of EdU during lytic reactivation. As expected, the OriLyt region of the KSHV genome showed levels of enrichment several hundredfold higher than those of the uninduced latent cells, confirming active viral DNA synthesis (Fig. 8B). Interestingly, when the exact same DNA was used to quantify the levels of actively replicated DNA at the lamin B2 promoter, a significant decrease in the levels of EdU-labeled DNA was observed in cells undergoing lytic reactivation (Fig. 8A).

FIG 8.

Cellular DNA replication stalls to favor viral DNA replication during lytic reactivation. (A) BCBL-1 cells left uninduced or induced for 48 h were labeled with EdU, total DNA was extracted, and the Click-iT procedure was used to isolate DNA from the active sites of replication. Relative qPCR analysis at the cellular lamin B2 origin revealed significantly less nascent DNA replication after 48 h of lytic induction. (B) BCBL-1 cells were robustly replicating viral DNA, as shown by relative qPCR analysis of the viral OriLyt region. (C) TRExBCBL-1-Rta cells left uninduced or induced with doxycycline for 24 h were used for Click-iT isolation of actively replicating DNA, as mentioned above, revealing reduced cellular DNA replication. (D) Viral genome actively replicated during reactivation, as measured by replicated DNA at the OriLyt region. (E) 293LBac36WT or 293LBac36Δ59 cells were induced for 24 h, followed by Click-iT isolation of nascently replicating DNA as detailed above. Uninduced cells were used as a control. 293L cells with WT BAC36 displayed a stronger inhibition of cellular DNA replication than cells lacking ORF59 (293LBac36Δ59) during lytic replication. (F) Viral DNA replication was inhibited in cells with deleted ORF59 (293LBac36Δ59) but not in the 293LBac36WT cells. (G) Schematic of a potential mechanism by which KSHV ORF59 associates with MCM complex and deregulates MCM loading onto the cellular chromatin by an aberrant phosphorylation of MCM3. As a result, cellular DNA replication is reduced to favor the replication of viral DNA.

Detection of active DNA replication during lytic replication was also repeated using TRExBCBL-1-Rta cells by using only doxycycline as a means of lytic induction. Twenty-four hours postinduction, cells were EdU labeled and subjected to detection of newly replicated, EdU-labeled DNA as described above. The results were consistent with those observed from the above-described BCBL-1 cells, which was a significant decrease in nascent cellular DNA replication (Fig. 8C) and an actively synthesized viral DNA (Fig. 8D). To further test the role of ORF59 in curbing cellular DNA replication during lytic reactivation, 293LBac36WT and 293LBac36Δ59 cells were induced for lytic reactivation by the addition of sodium butyrate for 24 h. The levels of actively synthesizing DNA were assayed by EdU labeling and precipitation as described above. Not surprisingly, 293LBac36Δ59 cells lacking ORF59 were unable to limit cellular DNA replication to the same degree as 293LBac36WT cells, suggesting ORF59 is a crucial protein in regulating cellular DNA replication (Fig. 8E, compare black bar with dark gray bar). Uninduced cells were used as a reference for both WT and 36Δ59-deleted cells. This confirmed that cellular replication origins are suppressed to favor the replication of viral genome during lytic reactivation.

DISCUSSION

KSHV-encoded early protein ORF59, classically the DNA processivity factor requisite for successful lytic replication of the viral genomes (9–11, 14, 15, 52), has in recent years been implicated in additional alternative functions, demonstrating the multifunctional purpose of ORF59 (16–18). We recently demonstrated ORF59's role in facilitating chromatin remodeling of the viral genome favorable for robust transcription of the lytic genes (16). Thus, aside from the task of shuttling viral DNA polymerase from the cytoplasm to the viral OriLyt inside the nucleus of infected cells, ORF59 associates with various host proteins to modify cellular conditions conducive to lytic reactivation and the production of infectious virions. In this study, we report a novel mechanism by which ORF59 modulates cellular DNA replication licensing through binding with MCM proteins to sabotage host DNA replication and give direct advantage to viral DNA replication.

As the MCM proteins bound together in a hexameric ring structure, it was necessary to test the extent of ORF59's association with these factors, which identified 4 of the 6 components of the MCM complex, namely, MCM3, MCM4, MCM5, and MCM6, to be associated with ORF59. Interestingly, depletion of MCM6 eliminated the association of MCM5 with ORF59, suggesting MCM6's role in bridging ORF59 with some components of the MCM complex. The fact that ORF59 was detected binding with half of the proteins of this complex suggested this is a significant interaction, although consideration still has to be given to an involvement of additional viral proteins or perhaps simply interlinking DNA (both MCMs and ORF59 bind DNA) in facilitating this association. However, when the binding between ORF59 and MCM3 to -6 proteins was confirmed in an overexpression system even after treating the lysates with DNase, it was clear that ORF59 is capable of interacting with MCM proteins without the cooperation of additional viral factors, and further that this interaction was not mediated by any interlinking DNA. This deliberate affiliation between an essential and multitasking viral protein and the MCM complex brought on the question of what role MCMs could be playing during lytic reactivation.

Importantly, the role of MCM in the context of KSHV lytic replication was not understood earlier. During latency, on the other hand, it has been shown that cellular origin replication licensing factors, including MCM3, were required for successful latent DNA replication (53). EBV latent replication is also functionally assisted by cellular replication proteins, including MCM2, which binds to the latent origin (54). However, the involvement of MCMs in latent DNA replication does not suggest anything about their activity during lytic reactivation aside from the fact that they have the potential to be regulated by viral factors. Importantly, the MCM6 knockdown in BC3 and BCBL-1 cells provided an important tool to determine the significance of MCM complex when cells were undergoing lytic reactivation. In both cell lines, depletion of MCM6 exhibited more robust lytic DNA replication, late gene transcription, and virion production than the control cells with undisturbed MCM6 protein expression. The subcellular localization of the replication foci in MCM6-depleted BCBL-1 appeared favorably modified toward lytic DNA replication, as multiple images of these cells revealed an increased number of active viral replication foci. Since a reduction of MCM6 expression benefited reactivation, it was hypothesized that these proteins are not necessary to assist lytic viral DNA replication but instead interfere with their activity so the virus attains an advantage toward its replication.

Analysis of protein levels in the lysates of BCBL-1 and iLSK.219 cells before and after lytic reactivation did not suggest any significant changes in the expression levels of MCM3, MCM4, or MCM6 (MCM4 and MCM6 both being members of the core helicase activity subunit). This was not surprising, as it has been shown that licensing and helicase activity of the MCM complex is primarily regulated by the association of specific cofactors and posttranslational modifications (20, 55). Kudoh et al. found that EBV-encoded protein kinase (EBV-PK) could hyperphosphorylate MCM4 to suppress DNA unwinding activity and limit cellular DNA replication during EBV lytic reactivation (33). Interestingly, we did not detect any significant changes in phosphorylation status of MCM4 but instead repeatedly detected a hypophosphorylated form of MCM3. The phosphorylation of MCM3 is required for the proper assembly of the MCM2-7 complex and the stable recruitment of MCMs at the chromatin, thus proper phosphorylation of MCM3 is essential for cell cycle progression (56). To investigate how the hypophosphorylation of MCM3 during lytic reactivation could affect the function of MCM complex, we tested whether MCM proteins were effectively loaded onto the chromatin by examining their interactions with histone H3 in lytic cells and comparing their association in latent cells. Notably, the interaction between MCM3 and histone H3 diminished after lytic reactivation, and neither MCM4 nor MCM6 coprecipitated H3 from the cells undergoing lytic reactivation. The loading of MCM3 and MCM4 onto the chromatin also confirmed a specific decrease in MCM3 and MCM4 binding at cellular origins of replication in cells undergoing lytic reactivation. This suggests a mechanism whereby KSHV prevents the loading of the MCM proteins onto the cellular replication origins by blocking the efficient phosphorylation of MCM3.

This strategy is similar to what has been documented during HCMV lytic replication. Biswas et al. determined that early lytic gene expression altered the assembly of pre-RC, and as a result the loading of MCMs on chromatin was reduced (57). By prohibiting the loading of MCMs onto chromatin, HCMV stops replication licensing and stalls active DNA synthesis (32). It was later determined that HCMV pUL117 has a role in preventing the MCM complex from associating with chromatin (58). Additional reports involving the interplay between MCM proteins and HCMV lytic reactivation indicated that while the virus has no need to use the MCM complex to license viral DNA replication initiation, it conveys a clear benefit to viral DNA replication (32, 57–59). It has been speculated that deregulating the activity of host MCM proteins and limiting cellular DNA replication that competes with viral DNA replication in nuclei of infected cells provides an optimal environment for HCMV viral DNA synthesis (60, 61).

Indeed, our experiments suggest that KSHV employs a similar strategy to promote its own lytic viral DNA synthesis, as we detected significantly lower levels of active DNA replication at the cellular origin during lytic reactivation. In contrast, DNA synthesis at the viral origin of DNA replication was enhanced significantly, as expected. Altogether, this suggests a mechanism where MCMs fulfill normal duties in host DNA replication and also assist in latent replication of the KSHV episome in the latent cells. The latent tumor cells follow the cell cycle pattern where MCMs are loaded onto the origin once per cell cycle to unwind DNA during the S phase. When cells are stimulated for lytic reactivation, however, ORF59 associates with the MCM complex and ensures the hypophosphorylated state of MCM3, which in turn prevents efficient loading of the MCM hexamers onto the cellular replication origins and prohibits cellular DNA replication to stall host DNA synthesis (Fig. 8C). In summary, this study presents evidence for a novel process that takes place in KSHV-infected cells undergoing lytic reactivation where the early lytic protein ORF59 assists viral DNA replication indirectly by disrupting cellular MCM proteins to stall host DNA replication.

MATERIALS AND METHODS

Cell culture.

293T (ATCC, Manassas, VA) cells were grown in high-glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 8% bovine growth serum (HyClone, Logan, UT), 2 mM l-glutamine, 25 U/ml penicillin, and 25 μg/ml streptomycin. iSLK.219 (generous gift from Don Ganem) cells were maintained in DMEM supplemented with 10% tetracycline-free fetal bovine serum (FBS) with an additional 600 μg/ml hygromycin B, 400 μg/ml G418, and 1 μg/ml puromycin. iSLK.219 cells with recombinant KSHV BACs were induced by doxycycline and sodium butyrate. KSHV-positive PEL cells, BCBL-1 and BC-3, were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 2 mM l-glutamine, and 5 U/ml penicillin–5 μg/ml streptomycin. All cultures were incubated at 37°C in a humidified environment supplemented with 5% CO₂. HEK293BAC36WT and HEK293BAC36Δ59 were cultured in DMEM with 10% FBS supplemented with 600 μg/ml hygromycin B.

Antibodies.

The following antibodies were used: mouse anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH; US Biological), mouse anti-Myc 9E10 (Sigma-Aldrich, St. Louis, MO, USA), rabbit anti-HA (6908; Sigma-Aldrich, St. Louis, MO), rabbit anti-control IgG (ChIP grade; 2729; Cell Signaling Technology), mouse anti-control IgG (sc-2025; Santa Cruz Biotechnology), anti-KSHV ORF57 rabbit polyclonal (Sigma-Aldrich, St. Louis, MO, USA), rabbit anti-PCNA (sc-7907; Santa Cruz Biotechnology), goat anti-MCM2 (sc-9839; Santa Cruz Biotechnology), mouse anti-MCM3 (sc-365616; Santa Cruz Biotechnology), mouse anti-MCM4 (sc-28317; Santa Cruz Biotechnology), rabbit anti-MCM5 (sc-22780; Santa Cruz Biotechnology), mouse anti-MCM6 (sc-393618; Santa Cruz Biotechnology), and mouse anti-MCM7 (sc-56324; Santa Cruz Biotechnology). Mouse monoclonal anti-LANA antibody and rabbit polyclonal anti-ORF59 were generated at GenScript. Anti-RTA antibody was generously provided by Yoshihiro Izumiya, UC Davis.

Plasmids.

pxi-ORF59-HA plasmid was cloned using PCR amplification and restriction enzyme digests as previously described (15). ORF59-GFP-Myc segments were cloned into the pEGFP-C1-NLS-Myc plasmid vector (62) by PCR amplification and inserted at the BamHI and EcoRI restriction enzyme digest sites. Cloning primer sequences are available in Table 1.

TABLE 1.

ORF59-GFP-Myc cloning primers

Primer	Sequence (5′–3′)
S-ORF59-GFP-Myc-1-132aa	CGAGCTCAAGCTTCGAATTCAATGCCTGTGGATTTT
AS-ORF59-GFP-Myc-1-132aa	TTATCTAGATCCGGTGGATCCGTAGGAAATGGTGGT
S-ORF59-GFP-Myc-133-264aa	TCTCGAGCTCAAGCTTCGAATTCAGGGGACAACCTCACC
AS-ORF59-GFP-Myc-133-264aa	TTATCTAGATCCGGTGGATCCCCAACCCGGGACTTT
S-ORF59-GFP-Myc-264-396aa	CGAGCTCAAGCTTCGAATTCATTTACCCCCGGGCTG
AS-ORF59-GFP-Myc-264-396aa	TTATCTAGATCCGGTGGATCCAATCAGGGGGTTAAA

Inducible MCM6 knockdown in BCBL-1 and BC3.

Lentiviral vectors containing doxycycline-inducible shRNA for MCM6 were obtained from Dharmacon (GE Life Sciences), cotransfected with packaging vectors pCMV-Δ8.2 and pCMV-VSVG (Addgene, Inc.) into 293T cells using PEI (polyethylenimine), and stimulated by addition of 10 mM NaB to produce lentiviral particles. Supernatants were collected for 4 days, and the lentivirus particles were concentrated by ultracentrifugation at 25,000 rpm for 2 h at 10°C. Concentrated lentivirus was used to transduce BCBL-1 cells, and cells were selected on 1 μg/ml puromycin to obtain a pure population of cells. MCM6 knockdown was then accomplished by addition of doxycycline for at least 1 week before knockdown efficiency was assessed by Western blotting using specific antibodies for MCM6 protein.

Coimmunoprecipitations and Western blotting.

For overexpression experiments, HEK293T cells were plated to 70 to 80% confluence, and expression vectors were transfected by combining transfection reagent, PEI, with 150 mM NaCl and mixing, followed by incubation at room temperature for 15 min. Transfection mixtures were added to 293T cells and incubated for 4 h at 37°C with 5% CO₂, and then medium was changed to remove PEI. At 24 h after transfection, cells were harvested for immunoprecipitation assays by washing with ice-cold phosphate-buffered saline (PBS) and then lysing in 0.75 ml of cold radioimmunoprecipitation assay (RIPA) buffer (1% Nonidet P-40 [NP-40], 50 mM Tris [pH 7.5], 1 mM EDTA [pH 8.0], 150 mM NaCl) supplemented with protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 1 μg/ml aprotinin, 1 μg/ml pepstatin, 1 μg/ml sodium fluoride, and 1 μg/ml leupeptin). Cell pellet was sonicated briefly with a probe sonicator to ensure complete lysis, and cell debris was removed by centrifugation at 13,000 × g for 10 min. For coimmunoprecipitation samples shown in Fig. 2A to F, samples were treated with DNase and incubated at 37°C for 30 min before proceeding with the preclearing step. Efficacy of DNase treatment was assessed by extracting a small aliquot from the DNase-treated lysate, isolating any remaining DNA, and performing PCR to determine the level of DNA left after treatment. Figure 2G showed that DNase treatment of lysate was highly efficient.

Samples were precleared for 30 min by addition of protein A-protein G-conjugated Sepharose beads and rotated at 4°C. After this step, 5% of lysate was saved for input controls, and the remaining lysate was incubated with 1.0 μg of specified antibody to capture proteins by overnight rotation at 4°C. To capture immune complexes, approximately 30 μl of protein A-protein G-conjugated Sepharose beads were added to lysates and rotated at 4°C for 2 h. Complexes were then pelleted and washed three times with RIPA buffer before IPs and input lysates were boiled for 5 min in Laemmli buffer and resolved by SDS-PAGE. The gel was then transferred onto nitrocellulose membrane (Bio-Rad Laboratories), blocked, and incubated with the indicated antibodies and detected by infrared dye-tagged secondary antibodies imaged using an Odyssey imager (LI-COR Inc., Lincoln, NE)

Phos-tag gel.

To differentiate phosphorylated isoforms of the MCM proteins, the Phos-tag chemical (Wako Chemicals USA, Inc.) was added into 8% SDS-PAGE gels according to the manufacturer's instructions. Specifically, 50 μM Phos-tag chemical and 100 μM MnCl₂ were added to resolving gel mixture, and the gel was solidified at room temperature before a stacking layer was added (which had no Phos-tag or MnCl₂). Cell lysates were incubated with Laemmli buffer for 1 min at 95°C before being resolved by SDS-PAGE. Once sufficiently resolved, the gel was rocked in 1× transfer buffer containing 10 mmol/liter EDTA for 20 min, and after that the transfer buffer was replaced for a total of two 20-min EDTA washes. Lastly, the gel was washed in transfer buffer without EDTA, and then the gel was transferred to nitrocellulose membrane. This membrane was blocked and incubated with appropriate antibodies similar to the other immunoblots in this study.

ChIP.

ChIP assay was performed similarly to previously described methods (16). Briefly, approximately 5 million cells were cross-linked by addition of 1% formaldehyde and rocked at room temperature for 10 min. The reaction was quenched by addition of glycine to a final concentration of 125 mM, and cells were washed with cold PBS supplemented with protease inhibitors (1 μg/ml leupeptin, 1 μg/ml aprotinin, 1 μg/ml sodium fluoride, 1 μg/ml pepstatin, and 1 mM phenylmethylsulfonyl fluoride). Cells were resuspended in buffer D for chromatin shearing (Diagenode, Inc.) supplemented with protease inhibitors, and DNA was sonicated to an average fragment length of 300 to 500 bp. Cell debris was removed by centrifugation at 9,000 × g for 1 min, and the sonicated supernatant was diluted with ChIP buffer (0.01% SDS–1.0% Triton X-100–1.2 mM EDTA–16.7 mM Tris [pH 8.1]–167 mM NaCl, including protease inhibitor) before preclearing by constant rotation at 4°C with A and G Sepharose bead slurry for 30 min. Precleared supernatant was collected by brief centrifugation, and 10% of the total volume was saved for the input fraction. The remaining fraction was used to capture chromatin by addition of specific antibodies and overnight rotation at 4°C. The next day, DNA-protein complexes were precipitated using protein Sepharose A/G slurry with constant rotation for 2 h at 4°C. Complexes were washed three times using low-salt wash buffer (0.1% SDS–1.0% Triton X-100–2 mM EDTA–20 mM Tris [pH 8.1]–150 mM NaCl) and once in Tris-EDTA. Elution buffer (1% SDS–0.1 M NaHCO₃) was added to elute chromatins, and then reverse cross-linking was performed by adding 0.3 M NaCl (final concentration) and incubating samples at 65°C for 4 h. DNA was then precipitated and subjected to proteinase K treatment for 2 h at 55°C and lastly purified with Qiagen Min-Elute PCR purification columns according to the manufacturer's instructions. Purified DNA served as a template for qPCR assays using primers listed in Table 2.

Click-iT EdU immunofluorescence assay (IFA).

Control (no doxycycline) or MCM6-depleted (doxycycline-treated) BCBL-1 cells, uninduced or induced for 24 h, were incubated with 30 μM the thymidine analogue EdU (5-ethynyl-2′-deoxyuridine) 30 min prior to harvesting, washed with PBS, evenly spread on coverslips, and air dried before fixation with 4% paraformaldehyde at room temperature for 20 min. The cells were then permeabilized for 20 min by addition of PBS containing 0.15% Triton X-100. With coverslips being kept in the dark from this point forward, Click-iT reaction mixture containing Cy3 fluorophore was added to coverslips and incubated at room temperature for 30 min. The cells were then blocked in a solution of PBS containing 0.4% fish skin gelatin (FSG) for 1 h at room temperature. Following a PBS washing, coverslips were incubated with primary antibody (ORF59, PCNA, or LANA) overnight at room temperature. The next day, samples were PBS washed before Alexa Fluor 647 secondary antibody was applied for 1 h at room temperature and after a final PBS wash was fixed to slides using ProLong diamond antifade (Thermo Scientific). A laser scanning confocal microscope (Carl Zeiss, Inc.) was used to obtain images.

Replication foci were manually quantified from 24 h in control or MCM6-depleted BCBL-1 cells from 10 different frames each, and statistical analysis by Student's t test (P value of <0.05) was conducted using Prism (GraphPad Software).

Click-iT-EdU IP.

BCBL-1 cells, uninduced or induced for 24 h, were labeled with 30 μM the thymidine analogue EdU for 40 min and then harvested for total DNA extraction using Hirt's method. DNA was then sonicated using BioRuptor Pico (Diagenode, Inc.) fragments about of 500 to 750 bp in length. Click-iT reaction was then performed by addition of Click-iT reaction components, including 20 μM biotin azide, and incubation for 2 h at room temperature. The sample volume was then increased to 500 μl, streptavidin beads were added, and samples were rotated overnight at room temperature. The next day, samples were washed 2 times with low-salt wash buffer (0.1% SDS–1.0% Triton X-100–2 mM EDTA–20 mM Tris [pH 8.1]–150 mM NaCl) and 2 times in Tris-EDTA (TE). Samples were then eluted by resuspending beads in elution buffer (1% SDS–0.1 M NaHCO₃) and incubating at room temperature with mild agitation, after which beads were collected by centrifugation and eluted DNA supernatant was kept in fresh tubes. The elution step was repeated once more, and after mild agitation samples were incubated at 70°C for 10 min, followed by centrifugation and transfer of supernatants to tubes containing the product of the first elution step. Samples were next purified by PCNA, DNA was precipitated, and pellets were resuspended in 50 μl TE. This DNA was next used in qPCR analysis to assess levels of actively replicating DNA at both cellular and viral origins.