Human papillomavirus 16 replication converts SAMHD1 into a homologous recombination factor and promotes its recruitment to replicating viral DNA

ABSTRACT We have demonstrated that SAMHD1 (sterile alpha motif and histidine-aspartic domain HD-containing protein 1) is a restriction factor for the human papillomavirus 16 (HPV16) life cycle. Here, we demonstrate that in HPV-negative cervical cancer C33a cells and human foreskin keratinocytes immortalized by HPV16 (HFK+HPV16), SAMHD1 is recruited to E1-E2 replicating DNA. Homologous recombination (HR) factors are required for HPV16 replication, and viral replication promotes phosphorylation of SAMHD1, which converts it from a dNTPase to an HR factor independent from E6/E7 expression. A SAMHD1 phospho-mimic (SAMHD1 T592D) reduces E1-E2-mediated DNA replication in C33a cells and has enhanced recruitment to the replicating DNA. In HFK+HPV16 cells, SAMHD1 T592D is recruited to the viral DNA and attenuates cellular growth, but does not attenuate growth in isogenic HFK cells immortalized by E6/E7 alone. SAMHD1 T592D also attenuates the development of viral replication foci following keratinocyte differentiation. The results indicated that enhanced SAMHD1 phosphorylation could be therapeutically beneficial in cells with HPV16 replicating genomes. Protein phosphatase 2A (PP2A) can dephosphorylate SAMHD1, and PP2A function can be inhibited by endothall. We demonstrate that endothall reduces E1-E2 replication and promotes SAMHD1 recruitment to E1-E2 replicating DNA, mimicking the SAMHD1 T592D phenotypes. Finally, we demonstrate that in head and neck cancer cell lines with HPV16 episomal genomes, endothall attenuates their growth and promotes recruitment of SAMHD1 to the viral genome. The results suggest that targeting cellular phosphatases has therapeutic potential for the treatment of HPV infections and cancers. IMPORTANCE Human papillomaviruses (HPVs) are causative agents in around 5% of all human cancers. The development of anti-viral therapeutics depends upon an increased understanding of the viral life cycle. Here, we demonstrate that HPV16 replication converts sterile alpha motif and histidine-aspartic domain HD-containing protein 1 (SAMHD1) into a homologous recombination (HR) factor via phosphorylation. This phosphorylation promotes recruitment of SAMHD1 to viral DNA to assist with replication. A SAMHD1 mutant that mimics phosphorylation is hyper-recruited to viral DNA and attenuates viral replication. Expression of this mutant in HPV16-immortalized cells attenuates the growth of these cells, but not cells immortalized by the viral oncogenes E6/E7 alone. Finally, we demonstrate that the phosphatase inhibitor endothall promotes hyper-recruitment of endogenous SAMHD1 to HPV16 replicating DNA and can attenuate the growth of both HPV16-immortalized human foreskin keratinocytes (HFKs) and HPV16-positive head and neck cancer cell lines. We propose that phosphatase inhibitors represent a novel tool for combating HPV infections and disease.

IMPORTANCE Human papillomaviruses (HPVs) are causative agents in around 5% of all human cancers.The development of anti-viral therapeutics depends upon an increased understanding of the viral life cycle.Here, we demonstrate that HPV16 replication converts sterile alpha motif and histidine-aspartic domain HD-containing protein 1 (SAMHD1) into a homologous recombination (HR) factor via phosphorylation.This phosphorylation promotes recruitment of SAMHD1 to viral DNA to assist with replica tion.A SAMHD1 mutant that mimics phosphorylation is hyper-recruited to viral DNA and attenuates viral replication.Expression of this mutant in HPV16-immortalized cells attenuates the growth of these cells, but not cells immortalized by the viral oncogenes E6/E7 alone.Finally, we demonstrate that the phosphatase inhibitor endothall promotes hyper-recruitment of endogenous SAMHD1 to HPV16 replicating DNA and can attenuate the growth of both HPV16-immortalized human foreskin keratinocytes (HFKs) and HPV16-positive head and neck cancer cell lines.We propose that phosphatase inhibitors represent a novel tool for combating HPV infections and disease.replication, and attenuates the growth of HPV16-positive cells.SAMHD1 is dephosphory lated by protein phosphatase 2A (PP2A) during mitotic exit (37), and endothall is an inhibitor of PP2A (38).Endothall treatment attenuated E1/E2 replication in C33a cells and enhanced recruitment of SAMHD1, E1, and E2 to the replicating DNA, mimicking the effects of SAMHD1 T592D.HFK+HPV16 treatment with endothall enhanced the recruitment of SAMHD1 and E2 to the viral genomes and attenuated cellular growth as well as inhibited the development of replication foci following differentiation.All of these results mimicked the results generated with SAMHD1 T592D.Strikingly, the effects of endothall and SAMHD1 T592D on cellular growth were dependent upon the presence of the full-length replicating HPV16 genome; endothall had minimal effects on the growth of isogenic cells immortalized by the HPV16 oncogenes E6 and E7.Finally, we demonstrate that the head and neck cancer cell line, SCC-104, which contains episomal HPV16 genomes, is hypersensitive to endothall, and treatment resulted in increased SAMHD1 and E2 recruitment to the replicating HPV16 genomes.Endothall was less toxic to head and neck cancer cell lines with integrated HPV16 genomes (SCC-47) or an HPV-negative head and neck cancer cell line (HN30).We propose that phosphatase inhibitors represent a novel approach for the treatment of HPV16-positive tumors that have replicating HPV16 genomes.

HPV16 replication phosphorylates SAMHD1 to convert it to a homologous recombination factor, and recruits it to E1-E2 replicating DNA
We established that E1-E2 replicating DNA recruits SAMHD1 using our C33a model system (10,39).SAMHD1 levels were knocked down using an shRNA in the presence of E1 and E2 expression along with a plasmid containing the origin of replication (Fig. 1A, compare lane 4 with lane 3).Knockdown of SAMHD1 did not alter the levels of E1 or E2 protein expression.Chromatin immunoprecipitation (ChIP) experiments using chromatin prepared from C33a cells transfected with E1-E2-pOri and shRNA Control (Ctrl) or shRNA SAMHD1 demonstrated that the knockdown of SAMHD1 expression (Fig. 1A) resulted in a reduction of SAMHD1 signal (Fig. 1B, compare lane 6 with lane 3, the results are representative of three independent experiments).In the absence of E1 or E2, there was a minimal ChIP signal detected as reported previously (not shown [7,8,10,[39][40][41][42]). The knockdown of SAMHD1 expression did not significantly alter E1-E2 DNA replication levels in C33a cells (Fig. 1C).In these C33a assays, input DNA is degraded using Dpn1 (which targets 4pb DNA sites methylated in bacteria but not in mammalian cells).This demonstrates that the signals generated come from replicated pOri DNA.Next, we determined that SAMHD1 was recruited to viral genomes in human foreskin keratino cytes immortalized by HPV16 (HFK+HPV16) (Fig. 1E).Chromatin was prepared from two independent donor HFK+HPV16 cell lines (21), and ChIP assays were carried out in triplicate with HA, E2, and SAMHD1 antibodies.The levels of HPV16 long control region (LCR) DNA pulled down were determined relative to input, expressed as percentages, and represent a summary of three independent experiments.The HA antibody pulls down background levels of LCR (lanes 1 and 2) while both E2 (lanes 3 and 4) and SAMHD1 (lanes 5 and 6) pulled down significantly more LCR DNA in both donor lines.The results suggested that SAMHD1 was converted into an HR factor in HFK+HPV16, and this was confirmed by determining the phosphorylation status of SAMHD1 (Fig. 1F).HFKs from two different donors were immortalized by hTERT, the full HPV16 genome (wild-type [WT] HPV16), or E6 and E7 expression via retroviral transduction (E6E7).SAMHD1 levels were similar in all immortalized lines.However, an antibody specific for SAMHD1 phosphorylated on threonine 592 (SAMHD1 p592, middle panel) detected SAMHD1 phosphorylation only in the presence of the entire HPV16 genome (lanes 2 and 5).It is likely that the DDR induced by HPV16 replication drives the phosphorylation of SAMHD1, and this is not due to E6/E7 expression (lanes 3 and 6). Figure 1 demonstrates that E1 and E2 replicating DNA recruits SAMHD1, and that HPV16 full genome converts SAMHD1 into an HR factor.

SAMHD1 is in a cellular complex with E1-E2, and a SAMHD1 phospho-mim etic is hyper-recruited to E1-E2 replicating DNA, disrupting viral replication
Phosphatases are integral to the control of mammalian DNA replication (43).Figure 1E demonstrates that HPV16 replication converts SAMHD1 into an HR factor via threonine 592 phosphorylation, promoting recruitment to the viral genome.We hypothesized that disrupting dephosphorylation of SAMHD1 may block viral replication by preventing removal of SAMHD1 from the replication complex.To investigate this, we used SAMHD1 T592 mutants (T592D, threonine to aspartic acid mimicking phosphorylation; T592A, threonine to alanine mimicking non-phosphorylation). Figure 2A demonstrates that the expression of wild-type and SAMHD1 mutants did not significantly alter the expression of the viral proteins E1 and E2 (lanes 7-9, input).The HA-IP (middle panel) immunopreci pitates the HA-tagged E1; E1 interacts with endogenous SAMHD1 (lane 3) and the three SAMHD1 proteins that are overexpressed (lanes 7-9).As expected, E1 interacts with E2, and this is not disrupted by the overexpression of any of the SAMHD1 proteins (lanes 5-9).E2 also interacts with endogenous SAMHD1 (bottom panel, lane 4) and the three overexpressed SAMHD1 proteins (lanes 7-9).E2 also interacted with E1 (lanes 5-9), as expected.Replication assays with the overexpressed SAMHD1 proteins demonstrate that SAMHD1 T592D significantly reduced replication (Fig. 2B, lane 5).E1 plasmid DNA levels detected in the cell extracts were not significantly altered by the overexpression of the SAMHD1 proteins, demonstrating that they were not toxic to the transfected cells (Fig. 2B).ChIP assays (Fig. 2C) demonstrated that the expression of SAMHD1 T592D resulted in a significant increase of E1 (HA), E2, and SAMHD1 T592D (V5) recruitment to replicating DNA when compared with the overexpression of SAMHD1 WT or SAMHD1 T592A (lanes 4, 8, and 12).The overexpression of the latter two proteins did not alter the recruitment of E1 or E2 to replicating DNA when compared with no-SAMHD1 overexpression control (pLX304).These results supported our hypothesis that a failure to dephosphorylate SAMHD1 (T592D mimics "permanent" phosphorylation) can block E1-E2-mediated DNA replication by "freezing" the replication complex on the DNA.
To investigate the relevance of the results in Figure 2 to the viral life cycle, SAMHD1 WT and mutants were stably overexpressed in HFK+HPV16 cells using lentiviral delivery and selection.Figure 3A demonstrates the expression of the SAMHD1 proteins from the delivered lentiviruses in HFK+HPV16 (lanes 6-8 and 14, 15) and HFK+E6/E7 (lanes 2-4 and 10-12) in two different foreskin donor lines.ChIP assays targeting the HPV16 LCR were carried out with the HPV16+HFK cells, and the percent of input chromatin pulled down was determined using E2, V5 (to detect the exogenously overexpressed SAMHD1 proteins), and a control HA antibody (Fig. 3B).The results with the HA antibody (lanes 17-24) detected background levels of LCR pulled down.E2 pulled down significantly more LCR DNA than the HA antibody; there was no significant difference in the E2 signal between any of the samples (lanes 1-8).All of the V5-tagged SAMHD1 proteins were recruited to the viral DNA, although there were significant differences between them.SAMHD1 WT was recruited, and the SAMHD1 T592A mutant was significantly less recruited (compare lanes 10 and 14 with 11 and 15, respectively).SAMHD1 T592D was the most recruited to the viral DNA and was significantly more than SAMHD1 WT and SAMHD1 T592A (compare lanes 12 and 16 with 10, 11 and 14, 15).The results support the hypothesis that phosphorylation of SAMHD1 on T592 promotes recruitment of the protein to replicating viral DNA.The viral DNA levels in the cells were not signifi cantly different between all of the samples (not shown); however, cellular growth was attenuated by the overexpression of the SAMHD1 T592D protein (Fig. 3C).In HFK+E6E7, the expression of the SAMHD1 proteins did not alter cell growth; in HFK+HPV16, cell growth was attenuated only by the SAMHD1 T592D.The results in Figure 3 suggest that overrecruitment of SAMHD1 (T592D) to replicating viral DNA attenuates cellular growth.As viral replication induces a DNA damage response, which would convert SAMHD1 into an HR factor, SAMHD1 T592D could attenuate host cell replication and growth due to the active DNA damage response requiring SAMHD1 recruitment to, and disengagement from, host replication sites to manage cell growth.E1 plasmid (that is not replicated during the assay), and there is no significant difference in the detection of this DNA between any of the samples that were transfected with the E1 plasmid.This demonstrates that the reduction of replication with SAMHD1 T592D is not due to cellular toxicity of the protein killing the transfected cells.(C) Chromatin was prepared from the transfected cells, and ChIP assays were carried out with V5 (detects the SAMHD1 proteins from the co-transfected plasmids), HA (detects E1), and E2 antibodies.(D) For all proteins, there is a significant (*, P value < 0.05) increase in signal in the presence of SAMHD1 T592D.Please note the log scale in this figure .Following differentiation of HFK+HPV16 cells, SAMHD1 T592D blocked the formation of viral replication foci (detected using γ-H2AX, Fig. 4A).Growing, undifferentiated HFK+HPV16 cells had no γ-H2AX foci; HFK+E6E7 cells had no foci irrespective of differentiation status (not shown).The expression of SAMHD1 WT or SAMHD1 T592A had no effect on foci formation (Fig. 4A), and quantitation of foci numbers confirmed only SAMHD1 T592D could block foci formation (Fig. 4B).The quantitation (Fig. 4B) summari zes the data from two independent HFK donor lines; donor 1 is shown in Figure 4A.The expression of the SAMHD1 proteins does not affect differentiation, as involucrin induction was similar in all samples (Fig. 4C). Figure 4D demonstrates that the exogenous levels of SAMHD1 do not change following differentiation, and that SAMHD1 phosphory lation is retained.The phosphorylation signal is enhanced by the exogenous expression of SAMHD1 WT (lanes 2 and 6).To investigate whether the exogenous expression of the SAMHD1 proteins disrupted viral genome integrity, we carried out TV exonuclease (RecBD, NEB) assays combined with real-time PCR, which measure both episomal and integrated DNA levels (21,44,45).Figure 4E demonstrates that there is no consistent pattern of the exogenous SAMHD1 proteins changing viral genome levels in two different donor cells; the observed differences are likely due to the time it takes to generate the cell lines from the parent HFK+HPV16 cells.Next, we investigated HPV16 genome levels in the cells expressing exogenous SAMHD1 following differentiation (Fig. 4F).E6 DNA levels were measured, and the strikingly consistent result is that, in the presence of SAMHD1 T592D, there is a highly significant reduction in genome copy number to almost undetectable levels.Finally, we investigated the genome status using TV assays (Fig. 4G).The only significant result was that the presence of SAMHD1 T592D resulted in enhanced viral genome integration.Combined with the reduction in HPV16 genome levels (Fig. 4F), the results suggest that SAMHD1 T592D binds to the viral genome and blocks replication, resulting in genome fragmentation and the integration of some of the broken genomes into that of the host.The levels of episomal genomes observed in these cells are similar to that we have reported previously, where Southern blots also confirmed the presence of episomal genomes (21).
The results in Fig. 1 to 4 demonstrate that SAMHD1 is recruited to E1-E2 replicating DNA, and that a T592D mutant has increased recruitment to the replicating DNA and can disrupt viral replication.

Endothall, a PP2A inhibitor, promotes recruitment of endogenous SAMHD1 to replicating viral DNA
PP2A dephosphorylates SAMHD1 during transition from mitosis (37), and endothall is an inhibitor of PP2A (38).As SAMHD1 T592D (mimicking constant phosphoryla tion of SAMHD1) disrupts E1-E2-mediated DNA replication and viral genome amplifica tion during mitosis, the ability of endothall to "mimic" SAMHD1 T592D by inducing increased endogenous SAMHD1 phosphorylation was tested.Figure 5A demonstrates that endothall significantly reduces the levels of E1-E2 DNA replication in transient C33a transfection experiments.Endothall did not induce death of the transfected cells as the transfected E1 plasmid levels were the same in the vehicle-and endothall-treated cells (Fig. 5B).Endothall induced increased levels of E2, E1, and endogenous SAMHD1 recruited to the E1-E2 replicating DNA (Fig. 5C).The attenuation of replication and hyper-recruitment to the replicating DNA following endothall treatment is similar to the results generated with SAMHD1 T592D (Fig. 2).

Endothall promotes recruitment of SAMHD1 to the viral genome in HFK+HPV16 cells, attenuates their growth, and blocks viral replication foci formation during differentiation
As endothall is attenuating E1-E2-mediated DNA replication in C33a cells and promoting recruitment of endogenous SAMHD1 onto the replicating DNA, the ability of endothall to disrupt the growth of HFK+HPV16 and HFK+E6E7 cells from two independent donors was determined.Figure 6A demonstrates that endothall significantly attenuates the growth of HFK+HPV16 cells (left panel).It also attenuates the growth of HFK+E6E7 cells but less than HFK+HPV16 (right panel).Chromatin was prepared from the treated HFK+HPV16 cells, and the interaction of E2 and SAMHD1 with the replicating viral DNA was determined using ChIP assays (Fig. 6B).For E2 and SAMHD1, there was a significant increase in recruitment to the viral DNA with donor 1 (compare lanes 2 to 1 and 6 to percentage of all cells.There was a significant reduction in the percentage of cells with γH2AX foci in the SAMHD1 T592D samples (*P value < 0.05) when compared with all other samples.(C) RNA was prepared from the differentiated cells, and involucrin levels were determined.The results are expressed relative to no calcium treatment = 1.In all cases, there is a significant increase in involucrin RNA (P value < 0.05), confirming that differentiation has not been affected by the expression of any of the SAMHD1 proteins.
(D) Following differentiation, the expression of the exogenous SAMHD1 proteins does not change, and SAMHD1 remains phosphorylated on T592.The exogenous SAMHD1 proteins do not affect differentia tion as keratin 10 expression was induced following differentiation irrespective of SAMHD1 expression (compare lanes 5-8 with 1-4).Combined with the involucrin RNA results, we conclude that differentiation is not affected by the exogenous SAMHD1 proteins.(E) The expression of exogenous SAMHD1 proteins did not reproducibly or significantly affect HPV16 genome DNA levels in two independent HFK+HPV16 cell lines.(F) Following differentiation, there is a consistent and significant reduction in HPV16 genome levels in the SAMHD1 T592D expressing HFK+HPV16 cells, P value < 0.05*.There was no consistent change in genome levels with the expression of any of the other exogenous SAMHD1 proteins.(G) To investigate HPV16 genome status (episomal vs integrated), TV exonuclease assays were carried out.
The results are expressed relative to mitochondrial DNA (which is episomal) equaling 100%.The only consistent and significant change in HPV16 genome status was observed following differentiation of HFK+HPV16 cells expressing SAMHD1 T592D, where integration levels increased, P value < 0.05*.IVL, involucrin.
5, respectively).For donor 2, there was a significant increase in SAMHD1 recruitment (compare lane 8 with 7), and although there was an increase in E2 recruitment, it did not reach significance as the P value was greater than 0.05.HA was used as a control antibody (lanes 9-12), and there was significantly less viral DNA pulled down when compared with E2 and SAMHD1, and there was no significant difference in HA signal following endothall treatment.In Figure 6C, we confirmed that the endothall treatment increased SAMHD1 phosphorylation in the two donor HFK+HPV16 cell lines (compare lanes 2 and 4 with 1 and 3), whereas there was a very low SAMHD1 phosphorylation in the HFK+E6E7 cells that was not altered by endothall treatment (lanes 5-8).
The addition of endothall prevented replication foci formation following differentia tion of HFK+HPV16 cells (Fig. 7A).γH2AX is used as a marker for the viral replication foci, and there are no foci detectable in HFK+E6E7 cells (the top two sets of panels) irrespec tive of endothall treatment.As expected, in HFK+HPV16, replication foci developed 3 days following the addition of calcium (third panel set down).Endothall blocked the formation of these replication foci (bottom panel set).The results shown are for donor 1 HFK cells, identical results were obtained in both donors, and quantitation is shown in Figure 7B.Treatment with endothall did not affect differentiation, as involucrin induction was similar in all samples (Fig. 7C).

Endothall preferentially attenuates the growth of an HPV16-positive head and neck cancer cell line with episomal viral genomes
To determine whether the endothall growth attenuation results with HFK+HPV16 (Fig. 6) could be extended to HPV16-positive head and neck cancer cell lines, we treated three cell lines with endothall: HN30, which are HPV negative and p53 wild type; SCC-104, which are HPV16 positive and contain episomal viral genomes; and SCC-47, which are HPV16 positive and contain integrated viral genomes (46).The three cell lines were also treated with cisplatin.The IC50 for the treatments is shown in Figure 8A.All cell lines were equally sensitive to cisplatin, whereas UMSCC104 was statistically more sensitive to endothall than the other two cell lines.Chromatin was prepared from the cells, and the recruitment of SAMHD1 and E2 to the viral genome in UMSCC104 was determined with and without endothall treatment (Fig. 8B).Endothall statistically increased the recruit ment of E2 to the viral genome in UMSCC104 (right-hand panel) and increased SAMHD1 recruitment, although it marginally failed to reach statistical significance (left-had panel, P value = 0.0505).The HA antibody was used as a control and resulted in background signals that were not varied by endothall.(B) Chromatin was prepared from both HFK+HPV16 donor lines, and ChIP assays were carried out with E2 and SAMHD1 (HA was used as a negative control).
There is a significant increase in SAMHD1 recruitment to the viral DNA in the presence of endothall for both donors (*, P value < 0.05).Donor 1 has a statistically significant increase in recruitment of E2 to the viral genome in the presence of endothall (*, P value < 0.05).In donor 2, there was an increase in E2 recruitment to the viral genome in the presence of endothall, but this did not reach significance.(C) Treatment with endothall increased the phosphorylation of SAMHD1 on T592 (pSAMHD1 signal) in both HFK+HPV16 donor lines, but did not do so in either HFK+E6E7 cell lines.foci was quantitated and expressed as a percentage of all cells.There was a significant reduction in the percentage of cells with γH2AX foci in the SAMHD1 T592D samples (*, P value < 0.05) when compared with all other samples.(C) RNA was prepared from the differentiated cells, and involucrin levels were determined.
The results are expressed relative to no calcium treatment = 1.In all cases, there is a significant increase in involucrin RNA (P value < 0.05), confirming that differentiation has not been affected by endothall treatment.

DISCUSSION
HPV activates the DDR to promote the viral life cycle, and inhibition of the DDR can block viral genome amplification (16).One reason the virus activates the DDR is to promote HR replication of the viral genome, allowing amplification in the presence of an active DDR (22).A host of HR factors are recruited to the replicating viral genome and are critical for the viral life cycle (7, 8, 10, 12-14, 19, 47).Here, we demonstrate that viral replication converts SAMHD1 into an HR factor via phosphorylation on threonine 592; the viral oncogenes E6 and E7 are unable to do this.It may not necessarily be viral replication per se that converts SAMHD1 into an HR factor, but rather the expression of the viral helicase E1, a known activator of the DDR when overexpressed (48)(49)(50).The role of SAMHD1 in HR is to promote DNA end resection that generates substrates required for HR via recruitment of CtIP protein to sites of DNA damage (35).In addition to regulation by phosphorylation, the role of SAMHD1 in HR can be controlled by SIRT1 deacetylation as this facilitates SAMHD1 binding to double-strand DNA breaks (36).Moreover, SAMHD1 SUMOylation by PIAS1 also regulates its DNA binding and anti-viral activity (51).SIRT1 is a known component of the HPV16 E1-E2 DNA replication complex that can deacetylate E2 and HR host factors, such as WRN, to promote high fidelity E1-E2 DNA replication and control the viral life cycle (7-9, 52).Overall, the results suggest that HPV16 replication (or replication factors) converts SAMHD1 into an HR factor that assists with replication of the viral genome.We demonstrate here that SAMHD1 is recruited to E1-E2 replicating DNA in C33a cells, and also to the HPV16 genome in HFK cells immortalized by HPV16.CRISPR/ Cas9 removal of SAMHD1 in N/Tert-1 cells containing the HPV16 genome resulted in increased cell proliferation, viral replication, and DNA damage, demonstrating that SAMHD1 is a restriction factor controlling the virally infected cell (33).The phenotypes observed could also be related to the effects on host DNA replication in the absence of SAMHD1.If the virus converts SAMHD1 into an HR factor, this could be important in managing host DNA replication in the presence of an active DDR generated by viral replication.
Although HPV genomes integrate in the majority of cervical cancers, in oropharyng eal cancers, the majority retain an episomal HPV16 genome (53,54).The presence of viral replication in these cancers presents a potential therapeutic target, as reducing viral replication will ultimately reduce the levels of E6 and E7 proteins potentially increasing the expression of their tumor suppressor targets p53 and pRb, respectively.Such an increase would promote the response of the tumors to chemotherapeutic agents such as radiation and cisplatin.The conversion of SAMHD1 into an HR factor only in the HFK cells containing replicating viral genomes creates a difference with control cells that we sought to exploit.Phosphatases are an important component of DNA replication complexes, and we investigated whether a SAMHD1 phospho-mimetic could disrupt HPV16 replication.The results demonstrate that SAMHD1 T592D (the aspartic acid providing a negative charge mimicking phosphorylation) is hyper-recruited to E1-E2 replicating DNA in C33a cells and reduces the levels of replication.This result supports the idea that dephosphorylation of SAMHD1 in the replication complex may cycle it off the replicating DNA, allowing continuous replication.SAMHD1 T592D also promoted recruitment of E1 and E2 to the replicating DNA in C33a cells, suggesting that the replication complex has become stalled, as there is a significant reduction in replication even though more replication factors are being recruited to the replicating DNA.In HFK+HPV16 cells, SAMHD T592D reduced the growth rates of the cells but could not do this in HFK+E6E7 cells.SAMHD1 T592D is also hyper-recruited to the viral genome in HFK cells and promotes an increase in the recruitment of E2 to the viral genome.The presence of SAMHD1 T592D also prevented the formation of viral replication centers following differentiation of HFK+HPV16 cells, as demonstrated by a significant reduction in γH2AX foci development in these cells following calcium treatment when compared with control cells.Strikingly, SAMHD1 T592D resulted in a significant reduction of viral DNA following differentiation, indicating that viral replication was blocked, resulting in viral genome degradation.This is supported by the observation that residual viral DNA becomes integrated in the HFK+HPV16 cells expressing SAMHD1 T592D.This suggests that the mode of HPV16 replication switches upon replication, as in nondifferentiated cells, the viral genome integrity is not disrupted by the expression of SAMHD1 T592D.This demonstrates that SAMHD1 T592D can disrupt viral replication during differentia tion.Although the HFK+HPV16 SAMHD1 T592D cells grew slower, we did not see a significant reduction in the viral DNA in these cells.This is perhaps due to cells dying due to reduced viral DNA over the longer term of these growth assays, or it could be due to SAMHD1 T592D regulating host DNA replication in the HFK+HPV16 cells and attenuating growth via slowing down host replication.It could be a combination of several factors that we will investigate in future studies.
To progress our results in a therapeutic direction, we next investigated the ability of the phosphatase PP2A inhibitor endothall to attenuate HPV16 replication and cell growth.PP2A dephosphorylates SAMHD1 (37) and, therefore, we anticipated that PP2A inhibition would mimic the phenotypes of SAMHD T592D.This is indeed what we saw as endothall treatment increased SAMHD1, E1, and E2 recruitment to replicating DNA in our C33a model while attenuating replication levels; endothall attenuated the growth of HFK+HPV16 cells preferentially when compared with control cells; endothall increased the recruitment of SAMHD1 and E2 to viral DNA in HFK+HPV16 cells, and prevented the development of replication foci following calcium-induced differentiation.Finally, we demonstrate that the growth of UMSCC104 cells, an HPV16-positive head and neck cancer cell line containing episomal viral genomes, is attenuated more than other head and neck cancer cell lines following endothall treatment, and SAMHD1 and E2 recruit ment to the viral genome is increased by endothall treatment of UMSCC104 cells.
Overall, our results suggest that phosphatase inhibitors are therapeutic candidates for the management of HPV16-positive cancers that contain episomal genomes.Although we have demonstrated that endothall treatment increases the recruitment of SAMHD1 to E1-E2 replicating DNA, endothall treatment also likely increases the levels of recruitment of additional DDR/HR factors to the viral DNA.In addition, the toxicity of endothall in the HPV16 full genome cells could be related to the targeting of host replication, which is operating in an environment of an active DDR that would ordinarily arrest DNA replication.Future work will continue investigating the potential of targeting phosphata ses for the treatment of HPV malignancies that contain replicating DNA.
To measure proliferation, cells were seeded in triplicate into 6-well plates at a density of 1 × 10 5 cells per well and were grown to 80% confluency (typically 3 days).The cells were then harvested by trypsinization and were stained with trypan blue, and viable cells were counted.The cells (1 × 10 5 per dish) were re-plated, and this was repeated every 3-4 days for 3 weeks.To measure the effect of endothall on proliferation, the cells were cultured in either 5 µM endothall or vehicle in media throughout.
To measure the sensitivity of cells to endothall, cells were seeded into 6-well plates at a density of 1 × 10 4 cells per dish and were cultured in either vehicle, 1, 5, 10, 25, or 50 μM endothall.Once vehicle wells reached 70%-80% confluency, media were removed and the cells were washed twice with phosphatebuffered saline (PBS).One milliliter of 0.5% crystal violet was added per well, and the plates were incubated by shaking at room temperature for 30 minutes.The wells were washed five times with PBS, and the plates were left to dry.Crystal violet images were scanned using the Odyssey CLx Imaging System, and ImageJ was used for quantification.

Transient DNA replication assay
C33a cells were plated out at 5 × 10 5 in 10-cm dishes.The following day, plasmid DNA was transfected using the calcium phosphate method.Three days post-transfection, a low molecular weight DNA was extracted using the Hirt method as previously described by Boner et al. (47).The digested sample was extracted twice with phenol:chloroform:iso amyl alcohol (25:24:1) and was precipitated with ethanol.Following centrifugation, the DNA pellet was washed with 70% ethanol, dried, and resuspended in a total of 150 µL water.Forty-two microliters of sample were digested with DpnI (New England Biolabs, Ipswich, MA, USA) overnight to remove unreplicated pOri16LacZ; the sample was then digested with ExoIII (New England Biolabs) for 1 hour.Replication was determined by real-time PCR, as described previously (39).For endothall treatment, cells were cultured in 5 µM endothall 24 hours post-transfection, which was also 48 hours before harvesting.The fidelity of E1-E2 DNA replication was determined as previously described (42,55).

Immunoprecipitation assay
Cell lysate was prepared as described above.The lysate (250 µg) was incubated with lysis buffer (0.5% Nonidet P-40, 50 mM Tris [pH 7.8], and 150 mM NaCl), supplemented with protease inhibitor (Roche Molecular Biochemicals) and phosphatase inhibitor cocktail (MilliporeSigma) to a total volume of 500 µL.A primary antibody of interest or an HA-tag antibody (used as a negative control) was added to this prepared lysate and was rotated at 4°C overnight.The following day, 40 µL of protein A beads per sample (MilliporeSigma; equilibrated to lysis buffer as mentioned in the manufacturer's protocol) was added to the above mixture and was rotated for another 4 hours at 4°C.The samples were gently washed with 500 µL lysis buffer by centrifugation at 1,000 × g for 2-3 minutes.This wash was repeated three times.The bead pellet was resuspended in 4× Laemmli sample buffer (Bio-Rad), heat denatured, and centrifuged at 1,000 rcf for 2-3 minutes.The supernatant was applied to an SDS-PAGE system to separate and resolve proteins, and was then transferred onto a nitrocellulose membrane using wet-blot transfer method.The membrane was probed for the presence of specific proteins as mentioned in the description of Western blotting above.
For endothall treatment, cells were cultured in 5 μM endothall for 24 hours before fixation and chromatin harvest.

Immunofluorescence
For fluorescent staining, cells were cultured on coverslips coated with poly-L-lysine (MilliporeSigma) and were differentiated by culturing in 1.5 mM calcium-containing media (MCF-154, supplemented with CaCl 2 , both Invitrogen) for 3 days.For endothall treatment, cells were cultured in 5 μM endothall for the 3 days of differentiation.Coverslips were then washed in PBS and were fixed by incubation with ice-cold methanol for 10 minutes.These were permeabilized by incubation in 0.2% Triton X-100/PBS at room temperature for 15 minutes.Background staining was reduced by blocking the coverslips in 10% normal goat serum (LifeTechnologies).Primary antibodies were used at the dilutions as follows: SAMHD1 (1/500, Cell Signaling Technology, 49158), phospho-S139 Histone yH2AX 1/200 (Cell Signaling Technology, 9718).Immune complexes were visualized using Alexa 488-or Alexa 595-labeled anti-species specific antibody conju gates (Molecular Probes).Cellular DNA was stained with 4' ,6-diamidino-2-phenylindole (DAPI, Santa Cruz, sc-3598).The coverslips were mounted onto slides using Vectashield medium (Thermo Fisher) and were visualized, and fluorescence was quantified using the Keyence imaging system.

Exonuclease V assay
PCR-based analysis of viral genome status was performed similarly to Myers et al. (45).Briefly, 20 ng genomic DNA was either treated with exonuclease V (RecBCD, NEB), or no enzyme control, for 1 hour at 37°C, followed by heat inactivation at 95°C for 10 minutes.Reactions containing linearized HPV16 DNA or circular plasmid were prepared and treated alongside genomic DNA as controls for enzyme function.Each reaction (2 ng) was then subject to PCR quantification using HPV16 E6 primers (to quantify viral DNA) or MITO and GAPDH primers (to quantify human circular and linear DNA, respectively).PCR was performed using a 7500 FAST Applied Biosystems thermocycler with SYBR Green PCR Master Mix (Applied Biosystems) and 100 µM of the following primer pairs: HPV16 E6 F: 5′-TTGCTTTTCGGGATTTATGC-3′ R: 5′-CAGGACACAGTGGCTTT TGA-3′; human mitochondrial DNA F: 5′-CAGGAGTAGGAGAGAGGGAGGTAAG-3′ R: 5′-TA CCCATCATAATCGGAGGCTTTGG-3′; and human GAPDH DNA F: 5′-GGAGCGAGATCCCTCC AAAAT-3′ R: 5′-GGCTGTTGTCATACTTCTCATGG-3′.The following cycling conditions were used: 50°C for 2 minutes, 95°C for 10 minutes, 40 cycles at 95°C for 15 seconds, and a dissociation stage of 95°C for 15 seconds, 60°C for 1 minute, 95°C for 15 seconds, and 60°C for 15 seconds.Analysis was performed in Microsoft Excel, with "ExoV-treated Ct" -"No-enzyme control Ct" giving an indication of DNA digestion and, therefore, the amount of linear DNA present.

Statistics
Standard error was calculated from three independent experiments, and significance was determined using a Student's t-test.

FIG 1 (
FIG 1 (A) C33a cells were transfected with the indicated plasmids, and the expression levels of E2, E1 (HA tagged), and SAMHD1 were determined.(B) Chromatin was prepared from the indicated transfected cells, and ChIP assays were carried out to detect the levels of E1 (HA), E2, and SAMHD1 protein recruited to the replicating DNA.There was a significant (*, P < 0.05) reduction in SAMHD1 recruitment when the protein was knocked down, demonstrating that the signal detected is due to SAMHD1 recruitment to the E1-E2 replicating DNA.(C) Quantitative replication assays demonstrate that the knockdown of SAMHD1 does not significantly change E1-E2 DNA replication levels.(D) The E1-E2 replicated DNAs in C were investigated for their fidelity, and removal of SAMHD1 significantly increased the mutational levels as determined by mutations in the lacz gene (42) (*P value < 0.05).(E) Chromatin was prepared from HFK+HPV16 (human foreskin keratinocytes immortalized by HPV16) cells, and the presence of E2 and SAMHD1 on the viral genome was confirmed using ChIP assays.The HA antibody serves as a negative control and represents background levels; both the E2 and SAMHD1 antibodies significantly increased the signal over background.(F) Western blotting of protein extracts from the indicated cell lines demonstrates that SAMHD1 is phosphorylated (pSAMHD1) only in the presence of the full HPV16 genome.

FIG 2 (
FIG 2 (A) C33a cells were transfected with the indicated plasmids, and protein extracts were prepared.Input levels of the proteins under study are shown in the top panels (input).HA (E1) immunoprecipitation demonstrates that E1 can interact with E2 and endogenous SAMHD1 (lanes 3 and 5), as well as with the three co-transfected SAMHD1 expression vector proteins (lanes 7-9).Similarly, E2 can interact with E1, endogenous SAMHD1 (lanes 4 and 5), and the three co-transfected SAMHD1 expression vector proteins (lanes 7-9).The exogenous SAMHD1 is tagged with a V5 epitope.(B) The expression of SAMHD1 T592D significantly (**, P value < 0.05) reduced E1-E2 DNA replication levels in C33a cells (top panel, compare lanes 2 and 5).The bottom panel monitors the levels of

FIG 3 ( 7 FIG 4
FIG 3 (A) Western blot demonstrating the expression of exogenous SAMHD1 proteins (detected with V5 antibody) in the indicated cell lines.(B) Chromatin immunoprecipitation assays demonstrate that SAMHD1 T592D had increased recruitment to viral DNA than exogenous wild-type SAMHD1 (P value < 0.05).(C) Growth curve of the indicated cell lines.Exogenous SAMHD1 T592D significantly (*, P value < 0.05) reduced the growth rate in HFK+HPV16 cell lines, but not in HFK+E6E7 cells.

FIG 5 (
FIG 5 (A) Replication assays were carried out in C33a cells transfected with the indicated plasmids; substantial replication is only detected in the presence of E1 and E2 expression (lanes 7 and 8).Treatment with endothall significantly reduced the replication signal (**, compare lane 8 with lane 7, P value < 0.05).(B) The levels of E1 input plasmid were determined in the control and endothall-treated cells, demonstrating no reduction in E1 levels following drug treatment (lanes 7 and 8).This demonstrates that the combination of transfection and endothall treatment is not killing the transfected cells.(C) Chromatin was prepared from C33a cells transfected with the indicated plasmids, and ChIP assays were carried out with HA (E1), E2, and endogenous SAMHD1 antibodies.The treatment with endothall significantly increased the recruitment of all three proteins to the replicating DNA (*, P value < 0.05).

FIG 6 (
FIG6 (A) The indicated cell lines were treated with endothall (or vector control) over 3 weeks, and cell growth was monitored.In both HFK+HPV16 donors (left-hand panel), there was a significant reduction in cellular growth at most time points tested in the endothall-treated samples (*, P value < 0.05).For HFK+E6E7 cells (right-hand panel), there was only a significant reduction in cell growth at the 21-day time point with endothall treatment (*, P value < 0.05).

Full 11 FIG 7
FIG 7 (A) Calcium was added to the indicated cell lines to differentiate them in the presence of endothall or vehicle (Veh), and cells were fixed 3 days later and were stained for γH2AX as a marker of viral replication foci.Donor 1 is shown, and donor 2 cells behaved in a similar manner.(B) The number of cells with γH2AX

Full 12 FIG 8 Full
FIG 8 (A) The indicated cells were treated with endothall and cisplatin, and the IC50 was determined.All three cell lines were equally sensitive to cisplatin, whereas UMSCC104 (which contains episomal HPV16 genomes) was significantly more sensitive to endothall than the other two lines (*, P value < 0.05).(B) Chromatin was prepared from UMSC104 cells treated with and without endothall, and ChIP assays were carried out for SAMHD1 (left panel) and E2 (right panel).E2 recruitment was significantly increased in the presence of endothall (*, P value < 0.05), and although SAMHD1 recruitment was also increased, it failed to reach significance.