Parvovirus B19 Infection of Human Primary Erythroid Progenitor Cells Triggers ATR-Chk1 Signaling, Which Promotes B19 Virus Replication

Abstract

Human parvovirus B19 (B19V) infection is restricted to erythroid progenitor cells of the human bone marrow. Although the mechanism by which the B19V genome replicates in these cells has not been studied in great detail, accumulating evidence has implicated involvement of the cellular DNA damage machinery in this process. Here, we report that, in ex vivo-expanded human erythroid progenitor cells, B19V infection induces a broad range of DNA damage responses by triggering phosphorylation of all the upstream kinases of each of three repair pathways: ATM (ataxia-telangiectasi mutated), ATR (ATM and Rad3 related), and DNA-PKcs (DNA-dependent protein kinase catalytic subunit). We found that phosphorylated ATM, ATR, and DNA-PKcs, and also their downstream substrates and components (Chk2, Chk1, and Ku70/Ku80 complex, respectively), localized within the B19V replication center. Notably, inhibition of kinase phosphorylation (through treatment with either kinase-specific inhibitors or kinase-specific shRNAs) revealed requirements for signaling of ATR and DNA-PKcs, but not ATM, in virus replication. Inhibition of the ATR substrate Chk1 led to similar levels of decreased virus replication, indicating that signaling via the ATR-Chk1 pathway is critical to B19V replication. Notably, the cell cycle arrest characteristic of B19V infection was not rescued by interference with the activity of any of the three repair pathway kinases.

INTRODUCTION

Human parvovirus B19 (B19V), a member of the genus Erythrovirus within the family Parvoviridae (21), has been associated with a broad spectrum of human diseases. Most commonly, it causes a mild rash in children; this form is known as “fifth disease” (59). However, under some circumstances, B19V infection leads to more severe symptoms. Examples of these diseases include hydrops fetalis in pregnant women infected during the second trimester, chronic pure red cell aplasia in immunocompromised patients, and transient aplastic crisis in sickle cell disease patients (28, 59).

B19V contains a single-stranded DNA (ssDNA) genome; it is approximately 5.7 kb in length and features two identical terminal repeats (ITRs) at both ends (17, 62). Replication of the B19V genome is restricted to nuclei of human erythroid progenitor cells (EPCs) (38, 39, 51). In ex vivo-expanded EPCs, B19V infection induces apoptosis (9, 49) and arrests the cell cycle at the G₂/M transition (55). Analyses in ex vivo-expanded EPCs infected with B19V have revealed that the large nonstructural protein NS1, which is multifunctional and essential to B19V DNA replication (61), causes cell cycle arrest (55) and also plays a role in triggering infection-associated apoptosis (9, 49). In addition, we have demonstrated that cellular signaling pathways triggered by the binding of erythropoietin (Epo) to its receptor (EpoR) are critical to B19V infection of ex vivo-expanded EPCs (7). Beyond the fact that NS1 and EpoR signaling are involved, however, we know little about the mechanisms by which the B19V genome replicates in ex vivo-expanded EPCs.

Several studies have shown that parvovirus infection induces a DNA damage response (DDR) that plays an important role in the replication of parvovirus DNA (1, 33, 44). DDR is mediated by three phophatidylinositol 3-kinase-like kinases (PI3Ks): ATM (ataxia telangiectasia mutated), ATR (ATM and Rad3 related), and DNA-PKcs (DNA-dependent protein kinase catalytic subunit) (14, 20). ATM is activated primarily by double-stranded DNA breaks (DSBs) and is recruited to DSBs by the Mre11-Rad50-Nbs1 (MRN) complex (2, 29, 40). ATR responds to single-stranded DNA breaks (SSBs) and stalled DNA replication forks and is recruited by an ATR-interacting protein (ATRIP) to replication factor A (RPA)-coated ssDNA (3, 47). DNA-PK is activated in response to DSBs and is recruited to damage sites in complex with both Ku70 and Ku80 (27). After any of these kinases is recruited to a site of DNA damage, it phosphorylates a number of substrates (e.g., H2AX, RPA32, Chk1, and Chk2), which in turn target other proteins to silence cyclin-dependent kinases. This leads to the arrest of cell cycle progression, so that the damaged DNA can be repaired or, in the case of irreparable damage, so that the potentially hazardous cells can be eliminated through apoptosis (14, 25).

The kinases that control the DDR can be induced and hijacked by viruses to promote replication of the genomes (56). For example, ATM signaling can be coopted for this purpose by minute virus of canines (MVC) and minute virus of mice (MVM) (1, 33, 44), and DNA-PK is used in the same way by adeno-associated virus 2 (AAV2) during coinfection with adenovirus (16, 46). We reasoned that DDR may likewise contribute to the replication of the B19V genome during B19V infection of EPCs. The viral titer in the plasma of infected patients can be as high as 10¹³ genomic copies per ml (30, 57); this overwhelming replication of the B19V genome may well lead to replication stress and elicit a DDR.

In this report, we describe our investigation of the DDR in ex vivo-expanded EPCs infected with B19V, in particular, with respect to the possibility that the DDR plays a role in virus replication. We discovered that B19V infection leads to a broad spectrum of DDR events, including phosphorylation of all three upstream kinases. We also found that these phosphorylated kinases, together with their downstream substrates or components, localize to the B19V replication center of the infected cell. In addition, disruption of either the ATR or DNA-PKcs signaling pathway significantly reduces the efficiency of B19V DNA replication without disturbing the cell cycle phenotype of infected EPCs.

MATERIALS AND METHODS

Cell line and virus.

Ex vivo-expanded CD36⁺ EPCs were prepared as described previously (7, 58). B19V-containing viremic plasma sample P53 was obtained from ViraCor Laboratories (Lee's Summit, MO). The plasma sample was titrated 10-fold and use to infect CD36⁺ EPCs. At 48 h postinfection (p.i.), the cells were fixed and stained with an anti-B19V NS1 antibody (9). Endpoint titers, in fluorescence focus-forming units (FFU), were determined at the last dilution that gave unequivocal fluorescence. Virus infection was performed at a multiplicity of infection (MOI) of approximately 1 FFU per cell as described previously (7).

Chemicals and treatments.

Hydroxyurea (HU; Calbiochem, Darmstadt, Germany) was dissolved in deionized water to make a 250 mM stock solution. The pharmacological inhibitors of CGK733, KU55933, NU7441, SB218078 (Chk1 inhibitor, abbreviated as Chk1i), and NSC109555 (Chk2 inhibitor, abbreviated as Chk2i) were purchased from Tocris Bioscience (St. Louis, MO) and were dissolved in dimethyl sulfoxide (DMSO) to make a 10 mM stock solution. Caffeine was purchased from Sigma (St. Louis, MO) and dissolved in deionized water as a stock solution at 100 mM.

Treatment with these pharmacological inhibitors was performed 3 h prior to infection with virus. KU55933, CGK733, NU7441, and caffeine were applied to cells at final concentrations of 10 μM, 2.5 μM, 10 μM, and 1 mM, respectively. Chk1i and Chk2i were used at final concentrations of 150 nM and 5 μM, respectively. DMSO at 0.25% was used as a nontreatment control. EPCs treated with HU at a final concentration of 0.25 mM were used as positive controls for DDR.

Lentiviral vectors and lentivirus transduction.

The oligonucleotides used to generate shRNAs were synthesized by Integrated DNA Technologies (IDT; Coralville, IA). shRNA oligos were annealed and ligated to the pLKO-GFP vector (Addgene, Inc.), and lentivirus was produced as described previously (7), following the manufacturer's instructions (http://www.addgene.org/plko). The following validated shRNA sequences (Sigma, St. Louis, MO) were chosen for targeting the genes of interest: shRNA specific to ATM (shATM), 5′-CCGGTGATGGTCTTAAGGAACATCTCTCGAGAGATGTTCCTTAAGAC CATCATTTTTG-3′; shRNA specific to ATR (shATR), 5′-CCGGAAAGAGGCTCCTACCAACGACTCGAGTCGTTGGTAGGAGCCTCTTTCTTTTTG-3′; shRNA specific to DNA-PKcs (shDNA-PKcs), 5′-CCGGCCGGTAAAGATCCTAATTCTACTCGAGTAGAATTAGGATCTTTACCGGTTTTT-3′; shRNA specific to Ku70 (shKu70), 5′-CCGGAAGAGTCTACCCGACATAAGCTCGAGCTTATGTCGGGTAGACTCTTCTTTTTG-3′. The following scrambled shRNA (shScrambled) was used as an shRNA control: 5′-CCGGCCTAAGGTTAAGTCGCCCTCGCTCGAGCGAGGGCGACTTAACCTTAGGTTTTTG-3′. At 48 h postransduction, CD36⁺ EPCs were infected with B19V.

Western blotting and immunofluorescence.

Western blotting and immunofluorescence assays were performed as described previously (33). Confocal images were taken at a magnification of ×40 or ×100 (objective lens), with an Eclipse C1 Plus confocal microscope (Nikon) controlled by Nikon EZ-C1 software.

Antibodies used in this study.

Rat anti-B19V NS1 antibody was produced previously (9). Other anybodies obtained commercially included anti-γH2AX (Millipore, Billerica, MA), anti-Ku70 and anti-phospho(p)-RPA32(Thr21) (Epitomics, Burlingame, CA), anti-p-ATM (Ser1981; Rockland Immunochemicals Inc., Gilbertsville, PA), anti-p-DNA-PK(Ser2056) (Genescript, Piscataway, NJ), and anti-ATM, anti-ATR, anti-DNA-PK, and anti-Ku80 antibodies (Calbiochem). Anti-p-Chk1(Ser345) and anti-p-Chk2(Thr68) antibodies were purchased from Cell Signaling Inc. (Danvers, MA). In addition, two anti-p-ATR (Ser428) antibodies were obtained from Cell Signaling Inc. for immunofluorescence and Santa Cruz Biotechnology (Santa Cruz, CA) for Western blot analysis. Antibody dilutions used for Western blotting and immunofluorescence analysis were those suggested in the manufacturers' instructions.

Southern blot analysis.

Low-molecular-weight DNA (Hirt DNA) was extracted from CD36⁺ EPCs as described previously (24). Southern blotting was performed as described previously using the SalI-digested pM20 probe, which contains the full-length B19V genome (24).

FISH.

A B19V DNA fragment (nucleotides [nt] 802 to 1295; GenBank accession no. AY386330) was PCR amplified and subsequently labeled using the Label IT fluorescence in situ hybridization (FISH) fluorescein kit (Mirus, Madison, WI). In situ hybridization was performed following the manufacturer's instructions. Briefly, CD36⁺ EPCs were cytospun onto slides, fixed in 1% paraformaldehyde for 30 min, and permeabilized with 1% Triton-100 for 10 min. After permeabilization, slides were treated with 5 μg/ml RNase A at 37°C for 1 h and washed with 2× SSC (1× SSC is 0.l5 M NaCl plus 0.015 M sodium citrate) buffer. Slides were sequentially dehydrated with 70%, 85%, and 100% ethanol in order for 2 min at room temperature and then denatured with 70% formamide for 2 min. Sequentially, slides were dehydrated with 70% ethanol (prechilled at −20°C), 85%, and 100% ethanol at room temperature in order for 2 min each, and hybridized with fluorescein-labeled DNA probe at 37°C overnight. After FISH, slides were treated with antibodies for immunofluorescence analysis.

Flow cytometry analysis.

We performed flow cytometry analysis as described previously (8). Briefly, CD36⁺ EPCs were fixed in 1% paraformaldehyde at room temperature for 30 min and permeabilized with phosphate-buffered saline (PBS) containing 0.5% Tween 20. Cells were sequentially incubated with primary and secondary antibodies and then with 4′,6-diamidino-2-phenylindole (DAPI) at 1 μg/ml in PBS containing 0.3% Tween 20. All processed samples were analyzed on a three-laser flow cytometer (LSR II; BD Biosciences) at the Flow Cytometry Core at the University of Kansas Medical Center. All flow cytometry data were analyzed using FACS DIVA software (BD Biosciences).

RESULTS

B19V infection induces a DDR in B19V-infected EPCs.

To determine whether a DDR is induced during B19V infection of ex vivo-expanded EPCs, we located the B19V DNA replication center by using a FISH-based method with a B19V DNA-specific probe. Since the parvovirus large nonstructural protein NS1 is essential to replication of the viral DNA (4, 13, 42), we costained the FISH-processed B19V-infected EPCs with an anti-NS1 antibody (9) to compare their localization. As shown in Fig. 1 A, B19V NS1 (red) signal overlapped with that for the B19V DNA (green), indicating that the NS1 protein was expressed within the viral DNA replication center. In all of the subsequent experiments, we used anti-NS1 immunostaining to mark this center.

Fig. 1.

B19V infection induces a DNA damage response in CD36⁺ EPCs. (A) Combined immunofluorescence- and FISH-based visualization of the B19V replication center. Mock- and B19V-infected CD36⁺ EPCs were analyzed 48 h p.i. Replicated viral genome (green) was detected by hybridization with a B19V-specific DNA probe labeled with fluorescein. B19V NS1 (red) was immunostained with anti-B19V NS1 antibody. Nuclei were stained using DAPI. Confocal images were taken at a magnification of ×100. (B to D) Immunofluorescence analysis of DDR in B19V-infected cells. At the indicated times p.i., mock- and B19V-infected CD36⁺ EPCs were coimmunostained with anti-B19V NS1 (red) and anti-γH2AX (green) (B) or with anti-B19V NS1 (red) and anti-p-RPA32 (green) (C). Confocal images in panels B and C were taken at a magnification of ×40. (D) Confocal images of the same 48-h p.i. samples as shown in panels B and C, but at a magnification of ×100. HU-treated EPCs were used as positive controls for γH2AX and p-RPA32 staining.

We next examined the expression levels and localization of phosphorylated H2AX (γH2AX) and Thr21-phosphorylated RPA32 (p-RPA32), a hallmark of DDR (5, 34, 52). Both γH2AX and p-RPA32 were detectable as early as 12 h p.i., the time at which B19V NS1 first became visible (Fig. 1B and C). γH2AX either formed a ring at the edge of the nucleus and completely overlapped with NS1 expression or was distributed evenly across the nucleus and thus had a broader expression domain than NS1 (Fig. 1D). In contrast, p-RPA32 consistently colocalized with NS1 (Fig. 1D). Overall, damaged foci were detected only in NS1-expressing cells. Upregulation of the expression of both γH2AX and p-RPA32 in infected cells versus mock-infected cells was confirmed by Western blotting (Fig. 2 B, compare lanes 2 and 1). Thus, our results demonstrate that B19V infection of EPCs induces a DDR.

Fig. 2.

All three upstream kinases of the DNA repair pathways are activated in response to B19V infection of CD36⁺ EPCs. (A) Immunofluorescence analysis. At 48 h p.i., mock- and B19V-infected CD36⁺ EPCs were coimmunostained with anti-B19V NS1 (red) and anti-p-ATM (geen), anti-p-ATR (green), or anti-p-DNA-PKcs (green). Confocal images were taken at a magnification of ×100. (B to D) Western blot analysis. CD36⁺ EPCs were treated with DMSO and the indicated inhibitors for 3 h prior to infection with B19V. Cells were harvested at 48 h p.i., lysed, and immunoblotted with anti-γH2AX and anti-p-RPA32 (B), anti-p-ATM and anti-p-ATR (C), or anti-p-DNA-PKcs (D). In all blots, anti-β-actin was used as a loading control. HU-treated cells were used as a positive control for the DDR.

ATM, ATR, and DNA-PK are activated in B19V-infected EPCs.

We next assessed which kinase pathway is activated in the B19V infection-associated DDR. Strikingly, we found that all three upstream kinases were phosphorylated, and the phosphorylated forms (green) localized to the B19V replication center (overlapping with NS1) (Fig. 2A, red). Specifically, we tested for the following markers: ATM phosphorylated on Ser1981, which typically responds to DSBs (29); ATR phosphorylated on Ser428, which typically associates with SSBs and stalled replication forks (15); DNA-PKcs phosphorylated on Ser2056, a form of the kinase known to be induced by DSBs (14). The phosphorylation status of each protein was confirmed by immunoblotting with their respective antibodies (Fig. 2C, lane 3, and D, lane 2).

To determine which of the three upstream kinases is responsible for H2AX and p-RPA32 phosphorylation in B19V-infected EPCs, we blocked their kinase activities using a panel of pharmacological inhibitors that inhibited their phosphorylation. Use of the inhibitors at the specified concentrations did not result in any obvious cytotoxicity (data not shown). As shown in Fig. 2B, the levels of γH2AX and p-RPA32 were more than 10-fold higher in whole-cell lysates of B19V-infected cells than in lysates of mock-infected counterparts (Fig. 2B, compares lanes 1 and 2). Treatment with the ATM-specific inhibitor KU55933 (at 10 μM) failed to cause a significant decrease in levels of γH2AX and p-RPA32 (Fig. 2B, compare lanes 2 and 4). However, treatment with the DNA-PK-specific inhibitor NU7441 (at 10 μM) reduced the levels of both by approximately 60% (Fig. 2B, compare lanes 2 and 5). Notably, addition of the ATM and ATR pan-inhibitor CGK733 at 2.5 μM significantly decreased the levels of γH2AX and p-RPA32 (by more than 80%) (Fig. 2B, compare lanes 2 and 3). The effectiveness of each inhibitor in preventing kinase phosphorylation was confirmed by Western blotting (Fig. 2C and D); KU55933 and NU7441 inhibited specifically the phosphorylation of ATM and DNA-PK, respectively (Fig. 2C, compare lanes 3 and 5, and D, compare lanes 2 and 3), and CGK733 inhibited the phosphorylation of both ATM and ATR (Fig. 2C, compare lanes 3 and 4). Although an ATR-specific inhibitor is not available, the facts that specifically inhibiting ATM using KU55933 did not affect γH2AX and p-RPA32 expression, whereas inhibiting both ATR and ATM using CGK733 resulted in significantly reduced expression, suggest that ATR signaling is the major determinant of B19V infection-induced DDR and that ATM does not play a major role in spite of its activation within the B19V DNA replication center. Moreover, our analysis suggests that the DNA-PK pathway also contributes to triggering of the DDR.

The ATR and DNA-PK signaling pathways are required for efficient B19V DNA replication in EPCs.

We and others have reported that DDR induced during infection by autonomous parvoviruses facilitates replication of the viral genome (1, 33, 44). To determine whether this is true for B19V, we used flow cytometry to examine the effects of the DDR kinase inhibitors on the percentage of B19V-infected (NS1-expressing) cells. We found that the percentage of NS1-expressing cells was reduced by 60% when cells were treated with CGK733 (inhibitor of ATM and ATR) and by ∼25% when the cells were treated with NU7441 (inhibitor of DNA-PK), but this value was unaffected when the cells were treated with KU55933 (inhibitor of ATM) (Fig. 3 A). In addition, treatment with 1 mM caffeine, another pharmacological inhibitor of ATM and ATR signaling (45), resulted in a 50% reduction in B19V-infected EPCs (Fig. 3A). Considering the importance of NS1 during parvovirus replication, we speculate that B19V DNA replication is impaired when the DDR signal is blocked as a consequence of ATR and DNA-PK pathway signaling.

Fig. 3.

Inhibition of ATR and DNA-PK signaling reduces B19V replication. (A and B) Effects of treatment with pharmacological inhibitors. CD36⁺ EPCs were treated with the indicated inhibitors for 3 h prior to B19V infection. (A) Flow cytometry analysis. At 48 h p.i., mock- or B19V-infected cells were immunostained for intracellular B19V NS1 and analyzed by flow cytometry. Numbers indicate the percentages of cells expressing NS1 in each case. A representative result is shown. The bar graph shows statistical analysis with averages and standard deviations based on at least three independent experiments, with the percentage of NS1-expressing cells in the DMSO-treated group being arbitrarily set as 100. Relative percentages of NS1-expressing cells in the other groups are indicated. (B) Southern blot analysis. At 48 h p.i., half of the cells were collected for Hirt DNA extraction and analyzed by Southern blotting with a B19V DNA-specific probe (upper blot). The other half of each sample was used for Western blotting with the anti-β-actin antibody, which served as a loading control (lower blot). Bands representing replicative form DNA (RF DNA) and ssDNA are indicated. B19V full-length DNA digested from pB19-M20 (62) was used as a marker. (C and D) Effects of treatment with kinase-targeted shRNAs. CD36⁺ EPCs were transduced with shRNA-expressing lentiviruses, as indicated, at 48 h prior to infection. (C) Western blot analysis. At 48 h p.i., cells were collected and analyzed by Western blotting using the indicated antibodies. (D) Flow cytometry analysis. At 48 h p.i., cells were collected and analyzed for NS1 expression by flow cytometry, using anti-B19V NS1. Transduced (shRNA-expressing) cells were selected based on the GFP-positive population, and the percentages of NS1-expressing cells among the GFP-positive population are shown in each histogram. A representative result is shown. Statistical analysis from three independent experiments is also shown; the percentage of NS1-expressing cells in the shScrambled-transduced group was arbitrarily set as 100, and the relative percentages of NS1-expressing cells in other groups are indicated.

We next performed Southern blot analysis to examine whether interference with DDR signaling impairs replication of the B19V genome. As shown in Fig. 3B, the levels of both replicative form DNA (RF DNA) and the ssDNA genome were significantly reduced in B19V-infected EPCs that had been treated with CGK733 and caffeine, but not in those that had been treated with KU55933. Treatment with 2.5 μM CGK733 decreased B19V DNA forms ∼3-fold, and treatment with 10 μM NU7441 reduced the RF DNA and ssDNA by ∼30% (Fig. 3B). We further knocked down ATM, ATR, and DNA-PKcs by transducing cells with lentiviruses expressing specific shRNAs. As shown in Fig. 3C, the efficiency of knockdown using these lentivirus-based shRNAs was high. We selected transduced cells based on the expression of green fluorescent protein (GFP; encoded by the lentiviral vector) and found that the percentage of NS1-expressing cells in the ATR knockdown group was reduced by 50%, and in the DNA-PKcs group it was reduced by 27% (Fig. 3D). Taken together, the kinase inhibitor and knockdown results demonstrate that the ATR and DNA-PK signaling pathways contribute significantly to efficient B19V DNA replication.

The ATR substrate Chk1 and the DNA-PKcs binding complex Ku70/Ku80 are required for B19V DNA replication in EPCs.

Given that ATM, ATR, and DNA-PKcs were activated and also localized to the B19V replication center, we questioned whether their activated substrates or components also localized to this site. To answer this question, we examined the localization of Ser345-phosphorylated Chk1 (p-Chk1), an ATR substrate, Thr68-phosphorylated Chk2 (p-Chk2), an ATM substrate (6, 31, 48, 60), and the Ku70/Ku80 heterodimer, which recruits DNA-PKcs and is part of the active DNA-PK holoenzyme (23, 54). As shown in Fig. 4 A, Chk1 and Chk2 were phosphorylated and localized with NS1 within the nuclei of B19V-infected EPCs. Notably, although both Ku70 and Ku80 were evenly distributed across the nuclei in uninfected cells, they localized to the B19V replication center in infected cells (Fig. 4A, Ku80 and Ku70). These results confirmed that all three signaling pathways (ATM-Chk2, ATR-Chk1, and DNA-PKcs–KU70/80) were activated in the B19V replication center.

Fig. 4.

Inhibition of Chk1 and Ku70, but not Chk2, reduces B19V replication. (A) Immunofluorescence analysis of localization of the substrates of ATR and ATM and the DNA-PK components. At 48 h p.i., mock- or B19V-infected CD36⁺ EPCs were coimmunostained with anti-B19V NS1 (red) and one of the following: anti-p-Chk1 (green), anti-p-Chk2 (green), anti-Ku70 (green), or anti-Ku80 (green). Confocal images were taken at a magnification of ×100. Nuclei were stained with DAPI. (B to D) Treatment with pharmacological inhibitors of Chk1 and Chk2. CD36⁺ EPCs were treated with DMSO, a Chk1 inhibitor or a Chk2 inhibitor for 3 h prior to infection. (B) Flow cytometry analysis of p-Chk1 and p-Chk2 expression. At 48 h p.i., cells were stained for intracellular p-Chk1 and p-Chk2 and analyzed by flow cytometry. The MFI is indicated. The background sample was treated with second antibody only. (C) Flow cytometry analysis of NS1 expression. At 48 h p.i., cells were stained for intracellular NS1 and analyzed by flow cytometry. Numbers show percentages of NS1-expressing cells in the population. A representative result is shown. Statistical analysis from three independent experiments is also shown. The percentage of NS1-expressing cells in the DMSO-treated group was arbitrarily set as 100; the values shown for the other cells are relative to the DMSO control. (D) Southern blot analysis. At 48 h p.i., Hirt DNA was extracted from one-half of the cells and analyzed by Southern blotting (upper panel); the remaining cells were used for Western blot analysis with anti-β-actin, which served as a loading control (lower panel). (E) Effect of treatment with Ku70-specific shRNA. CD36⁺ EPCs were transduced with lentiviruses expressing scrambled or Ku70-specifc shRNA at 48 h prior to infection. At 48 h p.i., cells were collected and analyzed for NS1 expression by flow cytometry. Transduced (shRNA-expressing) cells were selected based on the GF-positive population, and the percentages of NS1-expressing cells among the GFP-positive population are shown in each histogram. A representative result is shown. Western blot analysis at the bottom indicates knockdown efficiency of Ku70 in treated cells at 48 h p.i.

We used Chk1- and Chk2-specific pharmacological inhibitors to further explore the roles of the ATR and ATM pathways in B19V DNA replication. The efficiency of each inhibitor was confirmed by flow cytometry, following staining for the phosphorylated, active forms of these proteins (Fig. 4B). Compared to cells in the mock group, those infected with B19V expressed ∼4-fold more p-Chk1 and p-Chk2, as determined based on the mean immunofluorescence intensity (MFI). When Chk1 and Chk2 inhibitors were applied, cell proliferation was not affected (data not shown), although the phosphorylation of Chk1 and Chk2 was inhibited by ∼60 to 70%. Notably, only the inhibition of Chk1 led to reduced NS1 expression (∼40% reduction) (Fig. 4C). Consistent with this finding, application of the Chk1 inhibitor resulted in an ∼70% decrease in replication of the B19V genome, whereas application of the Chk2 inhibitor did not result in a significant change relative to that in cells treated with DMSO (Fig. 4D). Taken together, these results demonstrate that Chk1, a direct substrate of ATR signaling, plays an important role in B19V DNA replication.

To further examine the role of DNA-PK, we knocked down Ku70 by lentivirus-mediated shRNA. As shown in Fig. 4E, more than 70% of Ku70 was knocked down in shKu70-transduced EPCs compared with the scrambled shRNA control. Consequently, interference with Ku70 resulted in more than 50% reduction of NS1 expression, again suggesting that DNA-PK signaling plays a role in B19V replication.

Inhibition of DDR signaling does not rescue B19V infection-induced G₂/M cell cycle arrest.

DNA damage triggers biochemical pathways that arrest cell cycle progression (14, 45). Given that B19V infection of EPCs induces cell cycle arrest at the G₂/M phase (Fig. 5A, DMSO) (55), we monitored cell cycle changes in response to infection-induced DDR in the presence of kinase inhibitors and kinase-specific shRNAs. We found that none of the inhibitors tested (KU55933, CGK733, NU7441, Chk1i, or Chk2i) reduced the cell cycle arrest observed in B19V-infected cells (Fig. 5A, NS1+). Consistent with this finding, individual knockdown of ATM, ATR, or DNA-PKcs failed to significantly change this phenotype (Fig. 5B, NS1+); in NS1-positive cells of both the treated and control groups, more than 90% of the cells were arrested at G₂/M.

Fig. 5.

Inhibition of DDR signaling does not rescue the G₂/M cell-cycle arrest associated with B19V infection. (A) Effects of pharmacological inhibitors of DDR kinases on cell cycle arrest. CD36⁺ EPCs were treated with DMSO or the indicated pharmacological inhibitor for 3 h prior to B19V infection, and the effects on the cell cycle were assessed by flow cytometry. (B) Effects of shRNAs targeting DDR kinases on cell cycle arrest. Cells were transduced with the indicated shRNA-expressing lentiviruses 48 h prior to infection, and the effects on cell cycle were assessed by flow cytometry. At 48 h p.i., GFP-expressing cells were selectively gated for anti-B19V NS1 staining, followed by DAPI staining for cell cycle analysis. Cell cycle in mock-infected cells with (NS1+) or without NS1 gating (total) are shown. In each panel, the percentages of cells in the G1, S and G2/M phases, respectively, are indicated (top left).

When we looked at infection-associated cell cycle changes across the entire cell population, we observed that inhibiting either ATR signaling (whether by applying CGK733, caffeine, or Chk1 inhibitor, or an ATR-targeted shRNA) or DNA-PK signaling (by applying a DNA-PKcs or Ku70 shRNA) reduced the number of G₂/M-arrested cells across the whole population of B19V-inoculated EPCs (both NS1⁺ and NS1⁻) by ∼10 to 15% (Fig. 5A, total). Since B19V NS1 induces G₂/M cell cycle arrest in B19V-infected EPCs (55), we believe that this reduction is due to a general decrease in NS1 expression in all of the cells (Fig. 3B and 4D).

Collectively, these results suggest that during B19V infection of EPCs, a mechanism that is independent of DNA damage checkpoints regulates G₂/M cell cycle arrest.

DISCUSSION

In this study, we have demonstrated, for the first time, that infection of human primary EPCs by B19V leads to the phosphorylation of H2AX and RPA32, a hallmark of DDR, and that three DDR-mediating kinase pathways are activated. Nevertheless, only ATR-Chk1 signaling appears to have a strong influence on B19V DNA replication, with DNA-PKcs activation contributing to a lesser extent. We note that failure of ATM signaling to significantly influence B19V DNA replication and the induced DDR was surprising, given the importance of this pathway in MVC- and MVM-induced DDR (1, 33, 44). Finally, our results indicate that a DDR-independent checkpoint is responsible for the arrest of B19V-infected cells at the G₂/M transition of the cell cycle.

Several members of the family Parvoviridae have recently been shown to induce DDR. During infection by the autonomous parvoviruses MVC and MVM, ATM signaling is activated and required for replication of the viral genome (1, 33, 44); ATR signaling is also activated in MVC-infected cells but is not essential for MVC DNA replication (33). In the case of infection by AAV2, a member of the genus Dependovirus of the family Parvoviridae, it is predominantly the DNA-PK pathway that is activated in the presence of adenovirus, although ATM is also activated to some extent (16, 46). Notably, in cells inoculated with UV-inactivated AAV2, only ATR-Chk1 signaling is activated (26).

The DDR induced during B19V infection of primary human EPCs is unique in that all three kinase pathways are activated. More importantly, B19V takes advantage of both the ATR-Chk1 and DNA-PK signaling pathways to promote replication of its genome. B19V contains an ssDNA genome with a long ITR of 383 nt at both ends, and the gap between the two ends is quite large. This structure is a perfect trigger for the DDR, as the cellular DNA damage machinery recognizes the primed DNA as an SSB and activates ATR-mediated signaling (22). In addition, parvovirus DNA replicates according to a strand displacement model (18), producing new ssDNA ends and mimicking SSBs. Thus, B19V DNA replication also contributes to ATR activation. Similarly, parvovirus DNA replication produces nicked DNA intermediates that could be recognized as a DSB (33) and may activate ATM and DNA-PK (14).

We observed that the regular cell cycle of EPCs was moderately disturbed by knockdown of DNA-PKcs and Ku70 (Fig. 5B), indicating the importance of the DNA-PK complex in sustaining EPC proliferation. Indeed, it has been shown previously that functional inactivation of either Ku70 or Ku80 in human somatic cell lines is lethal, and inactivation of DNA-PK causes a proliferation deficit (43). However, this notion does not affect the results obtained using a DNA-PK-specific inhibitor, which did not affect the cell cycle of EPCs. Thus, we believe that inhibition of B19V replication by knocking down DNA-PKcs and Ku70 is a combined effect of inhibiting the DNA-PK signaling pathway and partial blockage of the G₁/S transition.

DNA-PK, an enzyme that contributes to the repair of DSBs through nonhomologous end joining (NHEJ), plays an important role in the persistence of recombinant AAV2 (rAAV2) in rAAV2-transduced tissues. It may exert its effects on promoting rAAV2 DNA circularization by recruiting Ku70 and Ku80, which bind directly to the ITR structure of rAAV2 (11, 50). However, the role of DNA-PK in wild-type AAV2 DNA replication is to recruit the complex heterodimer of Ku70 and Ku80 to the AAV2 ITR, which serves as an origin of DNA replication and facilitates AAV2 DNA replication in vitro (12). Ku70 and Ku80 act in this context by executing their helicase activities, and as such function much like the MCM2-7 complex to promote replication of the AAV2 DNA (12, 37). In the current study, we have shown that during B19V infection, Ku70 and Ku80 are recruited to the B19V DNA replication center, as that likely also includes phosphorylated DNA-PKcs. Since the ITRs of the B19V genome are identical, as is the case for those of the Dependovirus AAV2 (19), we speculate that the mechanism underlying DNA replication may be similar for these viruses, with Ku70 and Ku80 playing the same role.

The autonomous parvoviruses MVC and MVM hijack ATM signaling for DNA replication (1, 33, 44). Notably, infection by the autonomous parvovirus B19V activates ATM-Chk2 signaling. Both ATM and Chk2 were also recruited to the B19V DNA replication center; however, this activation had no significant effects on replication of the B19V genome. The structures of the MVC and MVM genomes are very similar; each contains a T-shaped palindromic repeat at the left end and a U-shaped palindromic repeat at the right hand (19, 53). This could explain why they would use a similar DDR-based strategy to replicate their DNA. ATM activation during MVC infection activated p53, which is responsible for MVC infection-induced apoptosis (33). Notably, p53 was phosphorylated during B19V infection of EPCs (55). Thus, we speculate that the ATM-Chk2 activation may contribute to apoptosis of B19V-infected EPCs.

ATR signaling is responsible for the repair of SSBs and that of associated stalled replication forks. Activation of its direct substrate, Chk1, results in slowed firing at the replication origin and controls cell cycle arrest, replication fork stability, and replication fork restart (15). During parvovirus replication, NS1 creates a nick site at the terminal resolution site on the ITR (24), and the viral DNA undergoes replication according to a strand displacement model (18), producing additional ssDNA ends that mimic SSBs structurally. Such events occur on multiple copies of the replicating B19V genome and possibly trigger robust ATR activation. This leads to recruitment of ATR-dependent substrates (e.g., RFC, RPA32, MCM2-7, MCM10, PCNA, and several DNA polymerases) to, or possibly their stabilization at, the replication fork (15). Notably, the ATR substrates also are important for AAV2 replication. A study involving an in vitro reconstitution assay showed that RFC, PCNA, MCM2-7, polymerase δ, and Rep78 are the minimum proteins required for efficient DNA replication of AAV2 (35, 36). In addition, one study revealed that infection with UV-inactivated AAV2 activates ATR-Chk1 signaling and that activated ATR-Chk1 complexes then recruited polymerase δ to the AAV2 ITR by the ATR-dependent substrates (26). Unfortunately, it was not feasible to examine the role of this recruitment in AAV2 DNA replication in that study, since the genome of UV-inactivated AAV2 was cross-linked (26). We have shown that ATR can phosphorylate both RPA32, a single-stranded DNA binding protein that is directly involved in replication of eukaryotic cells, and MCM2 (data not shown), a helicase that initiates formation of the prereplication complex. However, additional investigation will be required to understand the mechanism by which ATR-Chk1 signaling promotes B19V DNA replication.

DDR can lead to three distinct outcomes. When damage is severe and irreversible, the host cell may undergo apoptosis (10, 41). In the context of mild damage, it may recruit the repair machinery, through either NHEJ or a mechanism involved in homologous recombination, and cell cycle arrest (14, 32). Our analysis of the effects of DNA repair pathway inhibitors that significantly reduce B19V DNA replication failed to detect a change in the cell cycle pattern in infected cells. Moreover, although caffeine treatment caused a slight G₁ arrest of the cell cycle in EPCs, it also failed to rescue the G₂/M arrest in NS1⁺ cells. In addition, inhibiting Chk1 and Chk2 failed to interfere with G₂/M cell cycle arrest in NS1⁺ cells, although these proteins normally function as checkpoint controls, regulating the cell cycle under stressful conditions (31, 48). Consistent with these findings, knockdown of ATM, ATR, and DNA-PKcs did not affect B19V infection-induced G₂/M arrest. All these lines of evidence strongly suggest that B19V infection-induced G₂/M arrest is dependent on a DDR-independent cell cycle checkpoint. Indeed, it has been shown that B19V NS1 per se is able to arrest NS1-expressing cells at G₂/M by deregulating E2F family transcription factors (55). However, we cannot rule out the possibility that the ATR-Chk1 checkpoint, as well as the ATM-Chk2 checkpoint, may be redundant with the E2F-mediated checkpoint in arresting cells at G₂/M during B19V infection of EPCs. We are currently investigating this possibility.

In conclusion, we have identified a de novo role for ATR-Chk1 and DNA-PK signaling in B19V DNA replication in primary human EPCs. Components of these pathways can be explored as candidate drug targets for inhibiting B19V replication and for the treatment of B19V-asociated diseases in patients infected with B19V.