Ataxia Telangiectasia-Mutated Is Activated but Not Required for Productive Autographa californica Multiple Nucleopolyhedrovirus Infection

ABSTRACT

Multiplication of the invertebrate DNA baculoviruses activates the host DNA damage response (DDR), which promotes virus DNA replication. DDR signaling is initiated by the host insect’s phosphatidylinositol-3 kinase-related kinases (PIKKs), including ataxia telangiectasia-mutated kinase (ATM). Like other PIKKs, ATM phosphorylates an array of host DDR proteins at serine/threonine glutamine (S/TQ) motifs, the result of which leads to cell cycle arrest, DNA repair, or apoptosis. To define the role of host PIKKs in baculovirus replication, we compared replication levels of the baculovirus prototype species Autographa californica multiple nucleopolyhedrovirus in permissive Spodoptera frugiperda (SF21) cells with and without ATM function. Caffeine, which inhibits multiple DDR kinases, and the ATM-specific inhibitors KU-55933 and KU-60019 each prevented phosphorylation of Spodoptera histone H2AX (SfH2AX), a recognized indicator of ATM activity. However, only caffeine reduced autographa californica multiple nucleopolyhedrovirus (AcMNPV)-induced bulk phosphorylation of S/TQ protein motifs. Furthermore, only caffeine, not KU-55933 or KU-60019, reduced AcMNPV yields, suggesting a limited role for ATM. To investigate further, we identified and edited the Spodoptera ATM gene (sfatm). Consistent with ATM’s known functions, CRISPR/Cas9-mediated knockout of sfatm eliminated DNA damage-induced phosphorylation of DDR marker SfH2AX in SF21 cells. However, loss of sfatm failed to affect the levels of AcMNPV multiplication. These findings suggested that in the absence of the kinase SfATM, another caffeine-sensitive host DDR kinase promotes S/TQ phosphorylation and baculovirus multiplication. Thus, baculoviruses activate and utilize the host insect DDR in an ATM-independent manner.

IMPORTANCE The DDR, while necessary for the maintenance and fidelity of the host genome, represents an important cellular response to viral infection. The prolific DNA baculoviruses activate and manipulate the invertebrate DDR by using mechanisms that positively impact virus multiplication, including virus DNA replication. As the key DDR initiator kinase, ATM was suspected to play a critical role in this host response. However, we show here that baculovirus AcMNPV activates an ATM-independent DDR. By identifying the insect host ATM ortholog (Spodoptera frugiperda SfATM) and evaluating genetic knockouts, we show that SfATM is dispensable for AcMNPV activation of the DDR and for virus replication. Thus, another PIKK, possibly the closely related kinase ATR (ATM- and Rad3-related kinase), is responsible for efficient baculovirus multiplication. These findings better define the host pathways used by invertebrates to engage viral pathogens, including DNA viruses.

INTRODUCTION

Baculoviruses are a family of large DNA viruses which are highly prolific in insects. Like other viruses, baculoviruses modify the host cell to promote multiplication. In turn, the host cell activates antiviral pathways. These intrinsic pathways include the host DNA damage response (DDR). The DDR machinery detects incoming or newly synthesized DNA viral genomes and initiates a complex response involving activation of phosphatidylinositol 3-kinase-like kinases (PIKK) (reviewed in reference 1). PIKKs are activated by phosphorylation as a key step in the response. The three principal DDR PIKKs are ataxia telangiectasia-mutated kinase (ATM), ATM- and Rad3-related kinase (ATR), and DNA-dependent protein kinase (DNA-PK) (2, 3). ATM plays a key role in activating the DDR in response to deleterious double-strand breaks, which are subsequently repaired by homologous recombination (4). ATR responds to DNA lesions that occur due to replicative stress, including that induced by DNA viruses (5). Like ATM, DNA-PK responds to double-stranded breaks. However, DNA-PK promotes DNA-repair via nonhomologous end joining (NHEJ). ATM, ATR, and DNA-PK phosphorylate multiple substrates when activated (6). ATM, and to a lesser degree ATR and DNA-PK, phosphorylate the variant histone H2AX (7–9). Phosphorylated H2AX (γ-H2AX) is a key indicator of ATM-mediated DDR activation. In addition to promoting DNA-repair, DDR activation induces cell cycle arrest, promotes apoptosis, and responds to viral infections.

Diverse viruses trigger the host DDR, including double-stranded DNA (dsDNA) viruses (human cytomegalovirus [HCMV], BK polyomavirus, and adenovirus), single-stranded DNA (ssDNA) viruses (minute virus of mice [MVM]), and RNA viruses (influenza virus) (10–13). Infection often triggers a DDR response that promotes viral multiplication (12, 14, 15). For example, BK polyomavirus requires the virus-induced activation of the ATM and ATR pathways for efficient viral DNA synthesis (16). HCMV activates the ATM pathway, and ATM-phosphorylated host DDR proteins localize to viral replication centers within the nucleus. However, ATM activity is not required for HCMV replication (17). Adenovirus E4 ORF3 reorganizes the Mre11-Rad50-Nbs1 DDR complex to promote the establishment of viral replication centers (18). During MVM infection, viral genomes associate with DNA damage sites on the cellular genome (19). The DNA baculoviruses, including Autographa californica multiple nucleopolyhedrovirus (AcMNPV) (species Autographa californica multiple nucleopolyhedrovirus), also trigger the host insect DDR. AcMNPV-induced DDR activation triggers apoptosis, which is blocked by viral-encoded caspase inhibitors (20).

AcMNPV both activates and manipulates the DDR to promote viral multiplication (20, 21). This DNA virus, with its 134-kbp circular DNA genome, interferes with the normal ATM-induced DDR response (22). AcMNPV blocks DNA damage-induced accumulation of γ-H2AX at early times after infection (21), through the activity of its early lef-7 gene (22). Previous studies demonstrated that the broad-spectrum PIKK inhibitor caffeine also blocked γ-H2AX accumulation and severely reduced budded virus yields (20, 21). These findings implicated host DDR activity as a key requirement for AcMNPV multiplication.

Here, we further define the connection between AcMNPV and the induction of the invertebrate DDR by investigating the role of the PIKKs in initiating the host response and also contributing to virus multiplication. We report that AcMNPV induces the nuclear accumulation of DDR-specific protein phosphorylations early during infection of permissive Spodoptera frugiperda SF21 cells. Phosphorylation was specific for SQ and TQ (S/TQ) motifs, which is consistent with virus-induced PIKK activity. We confirmed that the PIKK inhibitor caffeine blocked S/TQ phosphorylation. However, other inhibitors known to be specific for ATM did not when used to treat AcMNPV-infected cells at nontoxic levels. Identification and genetic knockout of the Spodoptera frugiperda atm gene demonstrated that SfATM is required for DDR-mediated H2AX phosphorylation but not for AcMNPV productive infection. These findings indicate that a caffeine-sensitive PIKK, other than SfATM, is activated upon AcMNPV infection and contributes to efficient virus multiplication.

RESULTS

ATM-specific inhibitors block DNA damage-induced H2AX phosphorylation in permissive Spodoptera cells.

Upon infection, AcMNPV triggers the host lepidopteran DDR as indicated by the phosphorylation of histone H2AX (20–22), a direct substrate of ATM during DNA repair. These previous studies also suggested a role for ATM in the virus-induced DDR as demonstrated by the reduction of γ-H2AX by treatment of cells with known ATM inhibitors during DNA damage and baculovirus infection. Two ATM inhibitors, caffeine and KU55933, blocked γ-H2AX accumulation and AcMNPV productive multiplication (20, 21). Here, we performed side-by-side comparisons of ATM inhibitors at drug concentrations that did not affect cell viability. To this end, we first conducted dose response assays for cell viability using increasing concentrations of caffeine, KU-55933, and KU-60019, each known ATM inhibitors (23). Using MTT [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide] to monitor SF21 cells 72 h after drug treatment, viability decreased noticeably above concentrations of 8 mM, 54 μM, and 32 μM for caffeine, KU-55933, and KU-60019, respectively (Fig. 1). From these data, we chose 4 mM caffeine, 32 μM KU-55933, and 16 μM KU-60019 as the highest concentrations that exhibited no obvious effects on cell viability.

FIG 1.

SF21 cell viability in the presence of DDR inhibitors. SF21 cells were treated with caffeine, KU-55933, or KU-60019 at the indicated concentrations. Cell viability was determined via MTT assay 72 h later. Data are reported relative to values obtained from cells treated with the DMSO control alone ±SD (n = 3). Arrows indicate the highest drug concentration that exerted limited effects on cell viability.

Because ATM is the principal PIKK responsible for phosphorylation of histone variant H2AX after DNA damage, we tested the independent effect of the three ATM drugs on the accumulation of γ-H2AX after pharmacologically induced DNA damage using the topoisomerase inhibitor etoposide (24). γ-H2AX was monitored by using a green fluorescent protein (GFP)-fusion of SfH2AX that was transfected into SF21 cells prior to treatment with etoposide (20). DDR-mediated phosphorylation of GFP-SfH2AX occurs at a C-terminal SQ motif and is a sensitive marker of the DDR in SF21 cells (20). In dimethyl sulfoxide (DMSO)-treated SF21 cells, etoposide increased GFP-γ-SfH2AX levels compared to DMSO-treated cells alone (Fig. 2A, lanes 1 and 2). In contrast, caffeine, KU-55933, and KU-60019 reduced the level of etoposide-induced GFP-γ-SfH2AX (Fig. 2A, lanes 3, 4, and 5). The immunoblot signal intensities were quantified, and the ratios between GFP-γ-SfH2AX and total GFP-SfH2AX were compared (Fig. 2B). We concluded that caffeine and the ATM-specific inhibitors KU-55933 and KU-60019 blocked DNA damage-induced ATM-mediated phosphorylation of SfH2AX at concentrations that maintained SF21 cell viability. Thus, SfATM is sensitive to each of these inhibitors in SF21 cells.

FIG 2.

Comparison of H2AX phosphorylation in the presence of DDR kinase inhibitors. SF21 cells expressing GFP-HA-SfH2AX were treated with DMSO, 10 mM caffeine, 32 μM KU-55933, or 16 μM KU-60019 for 24 h. The cells were then treated with DMSO or 100 μM etoposide for 1 h. (A) Immunoblots. Cell lysates were prepared and subjected to immunoblot analysis using phospho-specific anti-γ-H2AX to detect GFP-γ-SfH2AX (top), anti-HA to detect GFP-HA-SfH2AX (middle), or anti-tubulin (bottom). The results are representative of three independent experiments. (B) Quantitation. The intensities of the anti-γ-H2AX and anti-HA signals were quantified, and the ratio ± standard deviation (SD) is reported relative to that obtained for the untreated cells (lane 1). n = 3. *, P < 0.05; **, P < 0.01.

AcMNPV induces DDR kinase substrate phosphorylation.

During infection, AcMNPV DNA replication triggers DDR activation (20, 21), as judged by host H2AX phosphorylation. Levels of γ-H2AX are highest in the absence of AcMNPV replication factor LEF-7 (22), which functions to reduce or limit γ-H2AX levels early in infection. Because of ATM’s initiator role in the DDR, we hypothesized that pharmacological inhibition of ATM would block these DDR-mediated phosphorylations during infection. To test this hypothesis, we infected control and ATM inhibitor-treated SF21 cells with AcMNPV and assessed DDR kinase-mediated phosphorylations using an antibody specific for the S/TQ phosphorylation motif targeted by ATM/ATR kinases. ATM and ATR preferentially phosphorylate serines and threonines, which are immediately followed by a glutamine (pS/TQ), including that of H2AX (25). S/TQ phosphorylation was monitored by immuno-chemical staining of fixed cells early in infection. Etoposide increased S/TQ phosphorylations relative to control cells (Fig. 3 and 4), confirming DNA damage-induced activation of the DDR. AcMNPV also increased phosphorylations at S/TQ motifs compared to mock-infected cells. The bulk of the increase in S/TQ phosphorylations was localized to the nucleus of AcMNPV-infected cells (Fig. 3), consistent with DDR activation. Moreover, these phosphorylations were distributed throughout the nucleus and were not confined to viral DNA replication centers. Viral replication centers were marked by the accumulation of immediate early transactivator IE1 (Fig. 3), which is an origin of DNA replication-binding protein (26). The presence of IE1 in S/TQ phosphorylated cells also confirmed that all cells were infected.

FIG 3.

Comparison of S/TQ motif phosphorylation in the presence of DDR kinase inhibitors. SF21 cells were treated for 24 h with DMSO alone or treated with the indicated drugs (10 mM caffeine, 32 μM KU-55933, or 16 μM KU-60019) and then infected with AcMNPV recombinant vAcIE1^HA (multiplicity of infection [MOI], 3). Drugs were replaced following removal of the virus inoculum. The cells were fixed 6 h after infection. Mock-infected cells (mock) were treated with DMSO alone (–) or 100 μM etoposide (+) for 1 h and then fixed. Fixed cells were stained with anti-pS/TQ (green), anti-HA to detect IE1^HA (red), and Hoechst (blue). Cells were visualized via confocal microscopy; representative fields are shown from triplicate experiments. Yellow bars, 10 μm.

FIG 4.

Quantitation of AcMNPV-induced S/TQ motif phosphorylation. Fields of immunostained cells, imaged as described in Fig. 3, were quantified by using Fiji. Immunofluorescence values are reported as the mean nuclear intensity of the anti-pS/TQ signal. Lower and upper hinges represent the first and third quartiles, and the whiskers are 1.5 times the interquartile range. The numbers of cells examined for each condition were as follows: mock (−) etoposide, 247; mock (+) etoposide, 201; +vAcIE1^HA (DMSO), 250; +vAcIE1^HA (caffeine), 216; +vAcIE1^HA (KU-55933), 250; +vAcIE1^HA (KU-60019), 193. Statistical analysis was carried out using one-way analysis of variance (ANOVA) followed by Tukey’s honestly significant difference test. ns, P > 0.05; *, P < 0.05; ****, P < 0.0001.

Caffeine treatment significantly reduced (P < 0.05) levels of S/TQ phosphorylation in infected cells (Fig. 4). Furthermore, caffeine eliminated the appearance of IE1-marked viral replication centers in the nucleus (Fig. 3). Thus, a caffeine-sensitive kinase is responsible for S/TQ phosphorylation during infection and contributes directly or indirectly to replication center formation. In contrast, neither of the ATM-specific inhibitors, KU-55933 or KU-60019, prevented AcMNPV-induced phosphorylation at S/TQ sites (Fig. 3 and 4). The KU-55933 and KU-60019 concentrations used here were sufficient to block etoposide-induced phosphorylation of a GFP-SfH2AX fusion protein (Fig. 2). These findings suggested that ATM is not solely responsible for phosphorylation at S/TQ sites during AcMNPV infection and that a caffeine-sensitive kinase other than or in addition to ATM phosphorylates S/TQ sites during AcMNPV infection.

Specific pharmacologic inhibition of ATM does not affect AcMNPV multiplication.

To determine the role of ATM in viral replication, we next treated SF21 cells with caffeine or the two ATM-specific inhibitors (KU-55933, KU-60019) for 24 h and measured yields of extracellular budded virus. Caffeine reduced the budded virus yields by more than 1,000-fold relative to the DMSO-treated control (Fig. 5), consistent with previous reports (20, 21). In contrast, neither ATM-specific inhibitor reduced budded virus yields. We concluded that caffeine, which broadly inhibits PIKKs, inhibits a target in AcMNPV-infected SF21 cells that is critical for AcMNPV multiplication. Moreover, caffeine’s target(s) of inhibition is likely independent of the target(s) of either ATM inhibitor. These findings suggest that ATM function is not required for AcMNPV multiplication.

FIG 5.

AcMNPV multiplication in the presence of kinase inhibitors. SF21 cells were treated with DMSO, 10 mM caffeine, 32 μM KU-55933, or 16 μM KU-60019 for 24 h and infected with AcMNPV (MOI, 0.5). Extracellular budded virus was collected 72 h later and quantified by using the 50% tissue culture infective dose (TCID₅₀). Yields are reported as the average PFU per mL ± SD (n = 3).

Spodoptera ATM (SfATM) mediates etoposide-induced H2AX phosphorylation.

We next sought an independent strategy to test the impact of ATM kinase activity on AcMNPV multiplication. To this end, we identified the Spodoptera ATM gene (sfatm) and generated SfATM knockout SF21 cells. Using sequence similarity with human ATM, we identified a candidate ATM sequence (accession no. GFAB01000496) via a BLAST search of a sequenced SF21 cell transcriptome (27). Conserved domain analysis (28) of the candidate ATM identified a conserved telomere-length maintenance and DNA damage repair (TAN) motif, a FRAP (FKBP12-rapamycin-associated protein), ATM, and TRRAP (transformation/transcription domain associated protein) (FAT) domain, an ATM kinase catalytic domain, and a FAT C-terminal (FATC) domain. These ATM domains are evolutionarily conserved (Fig. 6A). Importantly, the N-terminal TAN domain within ATM is exclusive to all ATM orthologs and is notably absent in other PIKKs, including the closely related PIKK ATR (29). Alignment of several ATM orthologs, including that of Homo sapiens, Saccharomyces cerevisiae, Drosophila melanogaster, Bombyx mori, and the identified S. frugiperda ATM sequences using Clustal Omega (30) showed high conservation within the TAN domain (Fig. 6B). Based on these data, we concluded that the S. frugiperda ATM ortholog SfATM was identified.

FIG 6.

ATM orthologs. (A) Protein schematics. H. sapiens ATM (3,056 residues), D. melanogaster ATM (2,767 residues), S. cerevisiae ATM (2,787 residues), Bombyx mori ATM (2,817 residues), and S. frugiperda ATM (2,815 residues) are shown. Domains identified from NCBI’s conserved domain search or visual inspection are indicated: telomere-length maintenance and DNA damage repair (TAN, purple), FRAP, ATM, and TRRAP domain (FAT, cyan), catalytic domain of ATM kinase (kinase, magenta), and FRAP, ATM, and TRRAP C-terminal domain (FATC, green). The sgRNAs used to target Cas9 to the genomic location of sfatm are shown; arrows indicate approximate locations. (B) TAN motif conservation. The amino acid sequences of selected ATM orthologs were aligned using Clustal Omega (30), and the N-terminal region containing the TAN motif is shown. The numbering at the top corresponds to the amino acid sequence of human ATM. Shading indicates conservation, with pink indicating similar residues, blue >60% conserved, and purple >90% conserved. Consensus for the TAN motif is shown at the bottom.

To initiate Cas9-mediated genome editing, SfATM was used to query Spodoptera whole-genome shotgun sequences. We identified the sfatm genomic region (accession no. JQCY02002655), which included 35 exons. We next generated two single guide RNAs (sgRNAs) for targeting Cas9 to sfatm using CRISPOR.org (31); sgRNA no. 1 targeted the exon 1 region of sfatm, and sgRNA no. 2 targeted the exon 34 region, which contains the critical FATC domain (Fig. 6A). Following transfection of SF21 cells with the CRISPR/Cas9 plasmids pIE1-Cas9-2A-GFP-U6 and pSfU6-neo, which expressed sgRNA no. 1 and sgRNA no. 2, respectively, DNA from single-cell clones was isolated and screened by PCR, and appropriate portions of sfatm were sequenced (data not shown). We generated multiple single-cell clones that contained lesions at the target sites within sfatm.

Using two of our SF21 deletion clones, we tested for SfATM kinase function by screening for GFP-Sf H2AX phosphorylation in response to etoposide-mediated DNA damage. In parental cells, etoposide induced accumulation of GFP-γ-SfH2AX relative to untreated cells (Fig. 7A, lanes 1 and 2), as expected. In contrast, neither of two chosen CRISPR-Cas9-edited clones showed increased levels of GFP-γ-SfH2AX in response to etoposide (Fig. 7A, lanes 3 to 6). When quantified, both of the knockout clones showed significantly reduced accumulations of GFP-γ-SfH2AX compared to parental SF21 cells (Fig. 7B). We concluded that the CRISPR-Cas9-edited cells are SfATM deficient and that SfATM is responsible for DDR-activated SfH2AX phosphorylation.

FIG 7.

Comparison of H2AX phosphorylation in SfATM KO cell lines. SF21 cells, including the parental cell line, and two SfATM KO clones (KO no. 1 and no. 2) were transfected with a plasmid expressing GFP-HA-SfH2AX. After 48 h, cells were treated with either DMSO (–), or 100 μM etoposide (+) for 1 h. (A) Immunoblots. Cell lysates were prepared and subjected to immunoblot analysis using phospho-specific anti-γ-H2AX (top), anti-HA (middle), or anti-tubulin (bottom). The results are representative of three independent experiments. (B) Quantitation. The intensities of the anti-γ-H2AX and anti-HA signal were quantified, and the normalized ratio is reported relative to that obtained for the untreated cells (lane 1). n = 3. *, P < 0.05; **, P < 0.01.

AcMNPV yields are not reduced in SfATM knockout cells.

To further investigate the potential role of ATM in AcMNPV multiplication, we compared the yields of AcMNPV budded virus from parental and SfATM knockout (KO) cells at intervals after infection. Wild-type AcMNPV multiplied in both SfATM KO cell clones to levels at or above levels in parental SF21 cells (Fig. 8). Thus, SfATM is not required for AcMNPV multiplication. This finding indicated that SfATM does not account for the caffeine-induced inhibition of AcMNPV replication (Fig. 5). Consequently, we hypothesized that caffeine, but not ATM inhibitor KU-55933 or KU-60019, would restrict productive AcMNPV multiplication in SfATM KO cells. To test this hypothesis, we first determined the relative toxicity of the ATM-specific inhibitors in SfATM KO cells since the absence of ATM may have affected cell sensitivity to each drug. Caffeine, KU-55933, and KU-60019 affected the viability of the SfATM KO cells (Fig. 9) at concentrations comparable to those for parental SF21 cells (Fig. 1). We next infected cells pretreated with caffeine, KU-55933, or KU-60019 with AcMNPV and measured budded virus yields. As expected, caffeine reduced yields of budded virus relative to DMSO control cells (Fig. 10). In contrast, neither KU-55933 nor KU-60019 affected virus yields from SfATM KO cells. (Fig. 10). We concluded that a caffeine-sensitive kinase other than SfATM is required for efficient AcMNPV multiplication.

FIG 8.

Comparison of AcMNPV multiplication in SfATM KO cells. SF21 parental cells and two SfATM KO clones (ATM KO no. 1 and no. 2) were infected with AcMNPV (MOI, 0.5). Extracellular budded virus was collected at the indicated times and quantified by using the TCID₅₀. BV yields are reported as average PFU ± SD per mL (n = 3).

FIG 9.

SfATM KO cell viability in the presence of DDR inhibitors. Cell lines SfATM KO no. 1 and no. 2 were treated with caffeine, KU-55933, or KU-60019 at the indicated concentrations. Cell viability was determined via MTT assay 72 h later. Data are reported relative to values ± SD obtained from cells treated with DMSO alone (n = 3). Arrows indicate the highest drug concentration that exerted limited effects on cell viability.

FIG 10.

AcMNPV multiplication in SfATM KO cells in the presence of DDR inhibitors. The two cell lines SfATM KO no. 1 and no. 2 were treated with DMSO, 10 mM caffeine, 32 μM KU-55933, or 16 μM KU-60019 for 24 h and then infected with AcMNPV (MOI, 0.5). Extracellular budded virus was collected 72 h later and quantified using the TCID₅₀. Budded virus yields are reported as average PFU ± SD per mL (n = 3).

DISCUSSION

The host DDR plays an important role during infection by diverse viruses. The invertebrate DDR is activated upon initiation of viral DNA synthesis during baculovirus infection, including that by AcMNPV (20–22). Whether the DDR positively or negatively impacts baculovirus multiplication, including virus DNA replication, is of interest because of the important role these viruses play as efficient and popular vectors for foreign gene expression. Our findings here that SfATM contributes to the baculovirus-induced host DDR but is not solely responsible for baculovirus replicative productivity provides insight into invertebrate virus-host interactions by suggesting that additional host PIKKs are involved.

Pharmacological inhibition of SfATM.

In this study, we observed that early in infection AcMNPV dramatically increases the accumulation of phosphorylated PIKK substrates possessing S/TQ motifs, including Spodoptera histone H2AX (Fig. 2 and 3). Due to ATM’s role as an apical DDR kinase and its contribution to the replication of multiple viruses (16, 32, 33), we focused on ATM functions during AcMNPV multiplication. Caffeine, a known PIKK inhibitor, reduced PIKK-mediated S/TQ motif phosphorylation and dramatically reduced AcMNPV BV yields (Fig. 4 and 5), which confirmed previous findings (20, 21). To our surprise, ATM-specific KU-55933 failed to reduce DDR phosphorylation or AcMNPV multiplication when used at nontoxic levels. Our viability assays (Fig. 1) suggested that the previously reported KU-55933 (50 μM) inhibition of AcMNPV (20) was possibly due to unknown off-target effects that compromised cell viability. Similarly, when used at nontoxic levels, ATM-specific KU-60019 reduced DDR phosphorylations but failed to affect AcMNPV yields (Fig. 4 and 5). Thus, the suggestion that ATM plays a central role in AcMNPV multiplication required further investigation.

Spodoptera SfATM is distinct from ATR.

To directly assess ATM’s functions, we used CRISPR/Cas9-mediated genome editing of cultured SF21 cells, a model system for baculovirus studies. The availability of ATM knockout cells obviated the dependence upon PIKK inhibitory drugs with their potential off-target effects. Using available databases, we identified a Spodoptera ATM candidate (2,815 amino acid residues) with striking sequence similarity to ATM from humans, yeast, and other insects (Fig. 6). In particular, SfATM possesses multiple highly conserved ATM domains, including the kinase catalytic domain near the C terminus that functions in ATP binding and phosphate transfer to S/TQ motif-containing substrates. Importantly, SfATM possesses a TAN (Tel1/ATM N-terminal) motif (Fig. 6B), which distinguishes it from other PIKKs, including ATR, DNA-PK, or mTOR, that lack a TAN motif (29). Because the principal PIKKs share significant sequence similarity, it was necessary to differentiate SfATM from other Spodoptera PIKKs. Indeed, using a Spodoptera transcriptome database, we identified a putative ATR with high sequence similarity to the human and Drosophila ATRs (data not shown). The Spodoptera ATR lacks a distinct TAN motif located near the N terminus of SfATM. Unlike ATR, ATM and DNA-PK are not essential genes (reviewed in reference 34). Our confirmed SfATM knockout cell lines were viable. Finally, we demonstrated that ATM-specific DDR functions, including DNA damage-induced H2AX phosphorylation (Fig. 7), were absent in the SfATM knockout cells. Taken together, we concluded that SfATM is the Spodoptera frugiperda ATM ortholog. To our knowledge, this is the first report of a lepidopteran ATM.

AcMNPV multiplication in the absence of ATM function.

Our SfATM deletions and pharmacological inhibition studies indicated that although SfATM is responsible for DNA damage-induced phosphorylation of SfH2AX, SfATM is not required for AcMNPV multiplication. Thus, in SF21 cells, ATM-generated γ-H2AX is not needed for AcMNPV replication, despite its contribution to replication of other DNA viruses (16, 32, 33). Indeed, AcMNPV early protein LEF-7 reduces DDR-mediated phosphorylation of SfH2AX during viral DNA replication (22). AcMNPV may therefore modulate the insect DDR to avoid ATM-mediated antiviral effects or to subvert an arm of the host DDR to promote viral multiplication. In mammals, ATM is the central kinase contributing to DNA repair by homologous recombination; it contributes to multiple stages of the homologous recombination process (30). During infection, ATM substrates involved in homologous recombination relocate to sites of viral replication, including Rad51, which traffics to replication centers of human papillomavirus (HPV), herpes simplex virus 1 (HSV-1), Epstein-Barr virus (EBV), and Simian virus 40 (SV40) (35–38). Like many large DNA viruses, the mechanisms of baculovirus DNA replication are poorly understood. However, it is likely that recombination-dependent DNA replication contributes to the generation of baculovirus genomic DNA (39). Multiple virus-encoded factors that are either required for or stimulate AcMNPV DNA synthesis promote homologous recombination (40), indicating that virus DNA replication stimulates homologous recombination. Thus, our findings here raise the possibility that AcMNPV promotes homologous recombination by a mechanism independent of ATM.

Which PIKK contributes to baculovirus replication?

Our findings indicate that AcMNPV replication events, including AcMNPV-induced phosphorylation of S/TQ motifs, is mediated by a caffeine-sensitive kinase other than SfATM. The most likely candidate is one of the other PIKK kinases, ATR or DNA-PK. Both enzymes phosphorylate S/TQ motif-containing substrates (25, 41) and are activated by other DNA viruses. ATR localizes to HSV-1 DNA replication centers and is required for efficient HSV-1 multiplication (33, 42). HPV also activates the ATR pathway (43). Kaposi’s sarcoma-associated herpesvirus (KSHV) lytic reactivation causes DNA-PK activation (44), while Adeno-associated virus (AAV) also recruits DNA-PK to AAV replication centers when coinfected with adenovirus (45). ATR is an DDR regulatory component that is essential for genome integrity and cell survival. It monitors replication stress, often detected as aberrant replication forks with exposed single-stranded DNA regions (5–7, 34). ATR activation leads to phosphorylation of downstream substrates, including CHK1, which are involved in cell cycle arrest and slowing replication fork processing. Considering that baculoviruses use multiple origins of viral DNA replication to expedite an enormous production of viral DNA, often exceeding the host cell’s own DNA content (39), it is anticipated that these prolific viruses have evolved mechanisms to handle this potentially catastrophic stress. Activation and manipulation of the host ATR pathway may be one such mechanism.

It is noteworthy that caffeine directly or indirectly inhibits phosphorylation of the AcMNPV immediate early 1 (IE1) transactivator (46), which contributes to virus DNA replication as an origin binding protein and stimulates recombination (40). Phosphorylation of the N-terminal replication domain containing a consensus cyclin-dependent kinase site is required for IE1-mediated DNA replication events but not transcriptional activation. Thus, caffeine’s inhibition of AcMNPV multiplication may also be due in part to its inhibition of a cellular or viral kinase responsible for IE1 phosphorylation. Thus, an important question that remains to be answered is how the host DDR contributes to AcMNPV multiplication. Given that caffeine reduces the synthesis of AcMNPV DNA during infection (20), caffeine-sensitive factors play an important but poorly understood role in viral DNA synthesis. Further study on the role of these host DDR factors in AcMNPV multiplication will provide mechanistic insight of insect virus DNA synthesis and likely uncover novel yet evolutionarily conserved host antiviral responses.

MATERIALS AND METHODS

Cells and infections.

S. frugiperda IPLB-SF21 (SF21) cells (47) were maintained at 27°C in TC100 medium (Invitrogen) supplemented with 10% fetal bovine serum (FBS) (HyClone). Wild-type AcMNPV (L-1 strain) (48) was previously described. The AcMNPV recombinant wt/vAcIE1^HA-H6 that encodes epitope hemagglutinin (HA)-tagged immediate early IE1 (AcIE1^HA-H6) under the control of the AcMNPV IE1 promoter was generated by David J. Taggart using the Bac-to-Bac system (Invitrogen) (49). In all infections, virus inoculum was allowed to adsorb for 1 h at room temperature with rocking. The inoculum was then replaced with fresh medium at time zero. When indicated, the inoculum was replaced with fresh medium containing drug or DMSO alone. Extracellular budded virus (BV) was quantified by TCID₅₀ assays as previously described (50).

Plasmids and transfections.

CRISPR-Cas9 constructs were designed for genome editing of SF21 cells. In brief, the S. frugiperda U6 snRNA (sfU6) promoter and terminator sequences (accession no. NJHR01000755) were identified using the Drosophila melanogaster U6 sequence to query S. frugiperda whole-genome shotgun contigs. While our work was ongoing, a similar strategy was published (51). To create pSfU6-neo, the sfU6 gene (accession no. NJHR01001538) and 400 bp upstream and 100 bp downstream, along with a Kan/neoR cassette under the control of an AcMNPV IE1 promoter, were cloned into the pBluescript KS+ (Invitrogen) backbone using NEBuilder HiFi (New England Biolabs). To create pIE1-Cas9-2A-GFP-U6, the sfU6 sequence was cloned along with Cas9-2A-GFP from pSpCas9(BB)-2A-GFP (PX458) into the pBluescript KS+ backbone using NEBuilder HiFi. pSpCas9(BB)-2A-GFP (PX458) was a gift from Feng Zhang (Addgene plasmid no. 48138) (52). Transfections were carried out using TransIT-Insect according to the manufacturer’s instructions (Mirus Bio). In each transfection, 10 μg of plasmid DNA was dissolved in unsupplemented TC100 and combined with 20 μL of TransIT-Insect. The transfection mixture was added dropwise to individual wells of 6-well plates containing 10⁶ SF21 cells in TC100. Unless otherwise noted, experiments were conducted 24 h posttransfection.

Inhibitors.

ATM-specific inhibitors KU-55933 and KU-60019, and the topoisomerase II inhibitor etoposide (Selleckchem), were dissolved in DMSO and diluted in TC100 medium supplemented with 10% FBS to the indicated concentrations. Caffeine (Alexis Biochemicals) was dissolved directly in TC100 medium supplemented with 10% FBS. When indicated, SF21 cells were overlaid with TC100 medium containing inhibitor at the indicated concentrations and incubated at 27°C. For infection of drug-treated cells, all drugs were replaced immediately following inoculation.

Immunoblots and antisera.

Cells were collected and lysed in 50 mM Tris, pH 7.4, 500 mM NaCl, 0.4% SDS, 5 mM EDTA, 1 mM dithiothreitol (DTT), 1× PhosSTOP (Roche) phosphatase inhibitor, and 1× cOmplete (Roche) protease inhibitor for 10 min on ice. Lysates were subjected to SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes (Osmonics, Inc.). Membranes were blocked in 0.2% (wt/vol) I-Block (Applied Biosystems) dissolved in phosphate-buffered saline (PBS) containing 0.1% Tween 20. The indicated primary antibodies were used, followed by IRDye 680 or 800 secondary antibodies (LI-COR) and imaged on a LI-COR Odyssey FC system. The following commercial antisera were used: monoclonal mouse anti-HA (Covance, MMS-101P), polyclonal rabbit anti-γ-H2A.X pS139 (Abcam, ab11174), mouse anti-tubulin (Sigma, DM 1A), and rabbit anti-pS/TQ (Cell Signaling, 2851S).

Immunofluorescence microscopy.

SF21 cells on glass coverslips were fixed with 1% formaldehyde in PBS, washed 3 times with PBS containing 0.1% Triton X-100 and 0.05% Tween 20 (PBST), and then blocked with 5% bovine serum albumin and 5% goat serum in PBST. Coverslips were incubated with the indicated primary antibodies and then washed 3 times in PBST and stained with Alexa Fluor 594-conjugated goat anti-rabbit IgG (Invitrogen, A-11017) and Alex Fluor 488-conjugated goat anti-mouse IgG (Invitrogen, A-11020). Fixed SF21 cells were then washed, counterstained with Hoechst 33342, and mounted using Fluoromount-G (Southern Biotech, 0100-01). Coverslips were imaged on a Leica Stellaris 5 confocal microscope using a 63× lens objective. Images were analyzed using Fiji (53), and figures were prepared by using Adobe Illustrator.

Sequence identification.

The gene (sfatm) encoding the Spodoptera frugiperda ataxia telangiectasia-mutated (SfATM) kinase was identified by using a BLAST search of a whole-cell transcriptome of SF21 cells (accession no. PRJNA271593). The human ATM amino acid sequence was used as the query. The BLAST search identified a transcript containing the potential SfATM coding sequence (accession no. GCTM01013287). Subsequently, the SfATM transcript and predicted protein sequence were used to identify domains matching those within human ATM by using NCBI’s Conserved Domain Database (28). The putative sfatm sequence was used to query an S. frugiperda whole-genome shotgun sequence (accession no. PRJNA380964) to identify the sfatm genomic context (accession no. NJHR01000076). The species and accession numbers of the ATM sequences used for comparison are Homo sapiens (EAW67115), Saccharomyces cerevisiae (CAA84909), Drosophila melanogaster (NP_001036712), Bombyx mori (XP_037877845), Spodoptera frugiperda (XP_035446109), Pan troglodytes (XP_016777413), Gorilla gorilla gorilla (XP_030871796), Pongo abelli (XP_024110822), Macaca mulatta (XP_014971094), Cercocebus atys (XP_011921885), Mus musculus (AAC52673), Rattus norvegicus (XP_038936996), Bos taurus (NP_001192864), Meles meles (XP_045873314), Gallus gallus (NP_001155872), Daphnia pulex (XP_046449745), Musca domestica (XP_005181014), Culex pipiens pallens (XP_039444669), Aedes aegypti (XP_021700538), Danaus plexippus plexippus (XP_032529702), Helicoverpa zea (XP_047041350), Trichoplusia ni (XP_026737947), Manduca sexta (XP_030029907), Spodoptera litura (XP_022833925).

CRISPR-Cas9 cell lines.

Cas9-cleavage sites and corresponding single guide RNAs (sgRNAs) were identified using crispor.tefor.net (31) targeting the sfatm genomic region. Oligo DNAs corresponding to the sgRNAs 5′-CACCGATAAAAGAAAACAGGCGATA-3′ and 5′-GCCGACCGTTGGGATAAACAGCCAC-3′ were cloned into pIE1-Cas9-2A-GFP-U6 and pSfU6-neo, respectively. SF21 cells were transfected with TransIT-Insect (Mirus Bio) and placed under selection in TC100 medium containing 500 μg/mL G418 (Gibco). Seven days later, the G418 was removed and cells were diluted into 96-well plates to obtain single-cell clones. Colonies were expanded, and genomic DNA was screened via PCR using primer pairs that amplified 500 bp upstream (5′-CTGCGATTATTCTGAAATTTTCATT-3′ and 5′-GTTCGGATGCAGCGGTACAGAAG-3′) and downstream (5′-TAAACCCTGCTTTACATACACAATTGGTAGAATG-3′ and 5′-ACCGGGATTATTGCCAGTTTCGATGA-3′) of both sgRNA sites. DNA isolated from selected colonies was sequenced. The selected colonies had two distinct edited copies of SfATM from the two alleles present in each cell. ATM KO no. 1 had two edited sequences: a 17,613-bp deletion that removed a section from exon 1 through exon 35, leaving only the C-terminal 5 amino acids, and a 1-bp deletion in the second codon causing a premature stop codon after 6 amino acids. ATM KO no. 2 had two edited sequences: a 17,722-bp deletion that removed all exons and a 360-bp deletion in exon 1 and 2 resulting in a premature stop codon after 6 amino acids and a 2-bp deletion in exon 35.