The Ataxia Telangiectasia-Mutated and Rad3-Related Protein Is Dispensable for Retroviral Integration

Abstract

Integration into the host cell DNA is an essential part of the retroviral life cycle and is required for the productive replication of a retrovirus. Retroviral integration involves cleavage of the host DNA and insertion of the viral DNA, forming an integration intermediate that contains two gaps, each with a viral 5′ flap. The flaps are then removed, and the gap is filled by as yet unidentified nuclease and polymerase activities. It is thought that repair of these gaps flanking the site of retroviral integration is achieved by host DNA repair machinery. The ATM and Rad3-related protein (ATR) is a member of the phosphatidylinositol 3 kinase-related family of protein kinases that play a major role in sensing and triggering repair of DNA lesions in mammalian cells. In an effort to examine the role of ATR in retroviral integration, we used RNA interference to selectively downregulate ATR and measured integration efficiency. In addition, we examined the possible role that Vpr may play in enhancing integration and, in particular, whether activation of ATR by Vpr (Roshal et al., J. Biol. Chem. 278:25879-25886, 2003) will favor human immunodeficiency virus type 1 integration. We conclude that cells in which ATR has been depleted are competent for retroviral integration. We also conclude that the presence of Vpr as a virion-bound protein does not enhance integration of a lentivirus vector in dividing cells.

Integration into the host cell DNA is an essential part of the retroviral life cycle and is required for the productive replication of a retrovirus (11). Integration consists of several coordinated steps, which are initiated and directed by the viral protein integrase. Following the completion of reverse transcription of the viral DNA, the terminal two or three nucleotides from the 3′ end of the linear viral DNA are cleaved by integrase. This is followed by insertion of the viral DNA into the host chromatin via an integrase-catalyzed, nucleophilic attack by the terminal 3′ recessed and exposed hydroxyl groups on two staggered phosphates of the target DNA. This results in cleavage of the host DNA and insertion of the viral DNA, forming the integration intermediate that contains two gaps, each with a viral 5′ flap. The flaps are then removed, and the gaps are filled by as yet unidentified nuclease and polymerase activities. It is thought that repair of these gaps flanking the site of retroviral integration is achieved by host DNA repair machinery.

In mammalian cells three major proteins are involved in sensing and directing repair of DNA damage, the ataxia telangiectasia-mutated (ATM), the ATM- and Rad3-related protein (ATR), and the DNA-dependent protein kinase (DNA-PK). ATM, ATR, and DNA-PK are all members of the phosphatidylinositol 3 kinase-related family of protein kinases (PIKK). DNA-PK plays a central role in the nonhomologous end-joining pathway. This is the primary repair mechanism of double-strand breaks within the cell and is important for V(D)J recombination during B-cell receptor maturation. ATR and ATM also participate in DNA repair by acting as sensors of various types of genotoxic stress which results in double-strand breaks or replication stress. Upon recognition of such stress, they become activated (for review, see references 1 and 39) and are able to coordinate the cellular response to the damaged DNA.

Recently, Daniel et al. suggested that successful integration of retroviral DNA into the host cell DNA requires ATR but not ATM (13). Our group recently reported that the human immunodeficiency virus type 1 (HIV-1)-encoded protein R (Vpr) activates ATR but not ATM, resulting in Chk1 phosphorylation and G₂ arrest (33, 41). Vpr is an HIV-1-encoded, 14-kDa accessory protein which is packaged into the virion during viral assembly (reviewed in reference 37). It has been proposed that capsid-bound Vpr participates in nuclear transport of the preintegration complex upon infection of growth-arrested cells (17) and naturally nondividing cells (such as macrophages) (12, 36, 38). However, this role of Vpr has been the subject of controversy, as another group reported that Vpr was dispensable for infection of a growth-arrested cell line (5).

In an effort to further examine the role of ATR in retroviral integration, we used RNA interference to selectively downregulate ATR and measured integration efficiency. In addition, we examined the possible role that Vpr may play in enhancing integration and, in particular, whether activation of ATR by Vpr will favor HIV-1 integration.

MATERIALS AND METHODS

Cells.

The human cervical cancer cell line HeLa was grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin-l-glutamine solution. Human embryonic kidney (HEK) cell line 293FT (Invitrogen, Carlsbad, Calif.) cells were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 1% penicillin-streptomycin-l-glutamine solution, and 1 mg of G418 per ml. The stably transformed human osteosarcoma cell line U2OS/ATRkd was maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 400 μg of G418 per ml, and 200 μg of hygromycin B per ml.

Plasmids.

pHR-Luc was constructed by removing vpr-ires-gfp from plasmid pHR-vpr-ires-gfp (33) and inserting firefly luciferase cDNA at the restriction site for Acc65I. pCMVΔR8.2ΔvprD116G was constructed by removing a 1,025-bp fragment from within the pol gene of the packaging plasmid pCMVΔR8.2Δvpr (2) with BclI and AflII sites and replacing it with a homologous fragment from an HIV-1_NL4-3 clone containing the D116G integrase mutation (24).

The self-inactivating retrovirus transfer vector SRα-SIN-CMV-LUC was derived from the previously described vector SRαMSVtkNeo (25). SRαMSVtkNeo was digested with ClaI to remove an internal thymidine kinase promoter and the neomycin resistance gene and filled in with Klenow fragment. The resulting DNA fragment was ligated with a DNA fragment containing the cytomegalovirus (CMV) immediate-early enhancer/promoter, followed by the firefly luciferase cDNA. In order to inactivate the long terminal repeat (LTR), a 383-bp sequence containing the murine sarcoma virus enhancer sequence was removed from the U3 region of the 3′ long terminal repeat. This deletion was accomplished by digestion with NheI and SacI, followed by a DNA Klenow fill-in and religation.

Drugs.

Doxycycline was obtained from Sigma and dissolved in water to make a 1 mM solution. Caffeine was also obtained from Sigma and dissolved in water to make a 100 mM stock solution.

Vector production.

We produced the lentivirus vectors pHR-Luc and pHR-LacZ by calcium phosphate-mediated transfection of HEK 293FT cells. Cells were cotransfected with 9 μg of plasmid pHR-luc or pHR'LacZ and 9 μg of pCMVΔR8.2Δvpr, PMM310 (obtained from Mike Miller; expresses the Vpr-β-lactamase fusion protein) and 3.5 μg of human CMV (HCMV)-vesicular stomatitis virus G protein (VSVG) (7). To make pHR-Luc with encapsidated Vpr, the packaging vector pCMVΔR8.2 (27) was used in place of pCMVΔR8.2Δvpr during the transfections. The integrase mutant virus was constructed with pCMVΔR8.2ΔvprD116G in the place of pCMVΔR8.2Δvpr. The supernatants were collected at 48, 60, and 72 h posttransfection and cleared of cell debris by centrifugation at 2,000 rpm in a Sorvall Legend RT. To concentrate the vectors, the cleared supernatants were centrifuged at 25,000 rpm in a Sorvall Discovery 100s for 2 h. The viral pellet was then resuspended in 0.3 ml of culture medium and frozen at −80°C. The murine leukemia virus vector was produced as described above for lentivirus vectors except the following plasmids were used: 12.5 μg of ψ⁻ env⁻ murine leukemia virus (21), 12.5 μg of SRα-SIN-CMV-Luc, and 5 μg of HCMV-VSVG.

In order to titer pHR-Luc/pCMVΔR8.2Δvpr, pHR-LacZ/pCMVΔR8.2Δvpr, pHR-Luc/pCMVΔR8.2, pHR-Luc/pCMVΔR8.2Δvpr-D116G, and pHR-LacZ/pCMVΔR8.2Δvpr-D116G, we adapted a method designed to study HIV-1 virion fusion. These viruses were made in the presence of a vpr-β-lactamase fusion protein (PMM310), which subsequently is incorporated into the virion. HeLa cells were infected for 2 to 4 h and then they were washed and loaded with the lipophilic fluorogenic substrate CCF2-AM (Invitrogen, Carlsbad, Calif.). CCF2-AM is composed of 7-hydoxycoumarin and fluorescein linked by a cephalosporin core. Excitation of the 7-hydoxycoumarin at 409 nm results in fluorescence resonance energy transfer to the fluorescein, causing it to emit green fluorescence. However, if the cell is infected, the cephalosporin core is cleaved by the β-lactamase of the fusion protein, inhibiting fluorescence resonance energy transfer and resulting in a blue fluorescence signal. This method allows us to accurately determine the titer of integration-incompetent viruses as well as our other lentivirus vectors based on entry. Murine leukemia virus was titered according to its luciferase activity.

All infections were performed at a multiplicity of infection of 2.5 in the presence of 10 μg of Polybrene per ml. The virus was left on the cells for 4 h, at which point the virus was removed and the cells were washed with phosphate-buffered saline and fresh complete growth medium was replaced.

RNA interference.

ATR was knocked down by transfecting HeLa cells with RNA duplexes from Dharmacon (Lafayette, Colo.) as previously described (33, 41). Forty-eight hours posttransfection the cells were infected with the appropriate vector as described above. Forty-eight hours postinfection, the cells were assayed for luciferase activity.

U2OS/ATRkd cells were treated with 2 μM doxycycline for 2 days prior to infection and then infected with virus at a multiplicity of infection of 2.5. Doxycycline was maintained in the culture medium during the infection and until cells were assayed. Forty-eight hours postinfection, cells were assayed for luciferase activity and normalized to milligrams of protein.

Cells infected with pHR-LacZ were fixed at room temperature for 5 min with 0.2% gluteraldehyde. The cells were washed three times with phosphate-buffered saline, and the developing solution was added. The cells were then incubated for 8 h to overnight, and the number of blue cells/field of view was counted.

Luciferase assays.

Cells were washed twice with phosphate-buffered saline, at which point they were lysed and assayed for luciferase activity as directed by the luciferase assay system (Promega, Madison, Wis.). The protein concentration of cell lysates was determined with the Pierce (Rockford, Ill.) BCA protein assay kit, and luciferase activity was normalized to milligrams of protein.

Western blot analyses.

Cells were detached and counted, and 10⁶ cells were lysed in 200 μl of Laemmli sodium dodecyl sulfate sample buffer, 300 μl of water, and 10 μl of a 1,000-fold solution of the Complete cocktail of protease inhibitors from Roche; 30 μl of cell lysate was loaded onto a sodium dodecyl sulfate-6% polyacrylamide gel and transferred to a polyvinylidene difluoride membrane (Amersham). The membrane was blocked with phosphate-buffered saline-5% milk and probed with an antibody to ATR (Amersham) (28) (1:1,000 dilution), or an affinity-purified polyclonal actin antibody (Santa Cruz, Santa Cruz, Calif.) (1:500 dilution) followed by a horseradish peroxidase-conjugated secondary antibody. Proteins were then detected with an enhanced chemiluminescence reagent (Amersham).

Real-time quantitative PCR.

A standard for the Alu-LTR real-time PCR integration assay was performed after infecting HeLa cells with pHR-GFP at an multiplicity of infection of 0.5. The infected cells were cultured for 5 days to allow any unintegrated viral DNA to be degraded (30). The cells were harvested and treated with Turbo DNase (Ambion), following which genomic DNA was isolated with the DNeasy tissue kit (Qiagen). Quantitative Alu-LTR PCR for integrated provirus in samples was performed at 48 h posttransduction, as previously described (6, 8, 29).

RESULTS

Presence of ATR is not required for efficient integration of HIV-1-based lentivirus vectors.

To examine the requirement for ATR during integration, we used several different methods to downregulate the amounts and/or activity of ATR in the cells. We then infected these cells with an HIV-1-based lentivirus vector, pHR-Luc, which expresses firefly luciferase. The lentivirus vector pHR-Luc was produced by transient transfection of pHR-Luc plasmid DNA along with pCMVΔR8.2Δvpr and HCMV-VSVG (7). Previous studies showed that expression of a reporter gene, such as luciferase or chloramphenicol acetyltransferase, from HIV-1-derived constructs can only occur in the presence of integration (24, 34), and therefore reporter gene activity can be used as an indirect but reliable measure of integration. A recent report by Poon et al. (31) suggested that unintegrated viral DNA from a luciferase-encoding lentivirus vector could produce small but significant amounts of luciferase activity. To account for potential expression from unintegrated DNA, we performed a control in which the same transfer vector, pHR-Luc, was packaged with an integration-defective packaging construct, pCMVΔR8.2ΔvprD116G, in which residue D116 within the catalytic site of integrase was mutated to a glycine. This mutation was previously shown to eliminate integrase activity without affecting reverse transcription (24).

We began to examine the role of ATR in integration with a U2OS cell line which had been stably transformed with a doxycycline-inducible form of ATR that is catalytically inactive (ATRkd) and acts as a dominant-negative construct (28). U2OS/ATRkd cells were plated and maintained in the presence or absence of 2 μM doxycycline for 2 days to allow induction of ATRkd. The cells were then infected with pHR-Luc/pCVMΔR8.2Δvpr at a multiplicity of infection of 2.5. Doxycycline was maintained in the culture throughout the infection; 2 days postinfection, the cells were lysed and assayed for luciferase activity. The levels of transduction in the presence and absence of ATRkd were not significantly different (Fig. 1A). Overexpression of ATRkd was evidenced by the increased intensity of the corresponding band on a Western blot (Fig. 1B). This result was inconsistent with a role for ATR function in retroviral integration.

FIG.1.

ATR is not needed for transduction of an HIV-1-based lentivirus vector. (A) U2OS/ATRkd cells were incubated with or without doxycycline to induce expression of ATRkd, following which they were mock infected or infected with pHR-luc/pCMVΔR8.2Δvpr and assayed for luciferase activity as a measure of transduction. (B) Western blot analysis monitoring the expression of ATRkd upon treatment of cells with doxycycline. (C) Alu-LTR PCR-based method for detection of integrated provirus. The results are representative of four independent experiments. (D) ATR was knocked down in HeLa cells by RNA interference and then the cells were mock infected or infected with pHR-Luc/pCMVΔR8.2Δvpr and assayed for luciferase activity. (E) Western blot analysis monitoring knockdown of ATR.

In order to directly quantitate integrated proviruses, a real-time Alu-LTR PCR assay was used (6, 8, 29). U2OS/ATRkd cells were induced or not induced to express ATRkd and then infected with pHR-Luc as described above. The genomic DNA was isolated from these cells 48 h postinfection, and the number of integrated proviruses was measured via the quantitative Alu-LTR PCR assay. In agreement with our luciferase assays, we found that ATR was not required for efficient lentiviral integration (Fig. 1C).

To test an alternative method of targeting ATR, we used RNA interference as described previously (33, 41). HeLa cells were transfected with small interfering RNA specific to ATR, scrambled small interfering RNA, or mock transfected, incubated for 2 days, and then infected with pHR-Luc/pCMVΔR8.2Δvpr at a multiplicity of infection of 2.5. The cells were harvested 48 h postinfection, lysed, and assayed for luciferase activity. In agreement with our previous results, transduction efficiency in the presence of ATR knockdown was not inhibited (Fig. 1D). To verify the knockdown of ATR, we performed a Western blot analysis on cell lysates, which showed a 90% knockdown of ATR at the protein level (Fig. 1E). To test whether our knockdown was sufficient to eliminate ATR activity, we transfected ATR knockdown cells with a plasmid expressing Vpr and analyzed the cell cycle by flow cytometry. If ATR remained active after knockdown, then the transfected cells should arrest in G₂ in response to Vpr expression (33, 41). We calculated the percentages of cells in G₁, S, and G₂ as well as the G₂/G₁ coefficients (Fig. 2). Whereas normal, untransduced cells displayed a normal cell cycle profile (G₂ = 8.16%; G₂/G₁ = 0.12%), cells transfected with a scrambled small interfering RNA and transduced with pHR-VPR effectively arrested in G₂ (G₂ = 92.66%; G₂/G₁ = 12.1%). When cells were transfected with ATR-specific small interfering RNA, pHR-VPR transduction led to a minor degree of G₂ arrest (G₂ = 24.86%; G₂/G₁ = 0.58%), consistent with loss of ATR activity due to the knockdown.

FIG. 2.

Cell cycle analysis of ATR knockdown cells. ATR was knocked down in HeLa cells, which were then infected with pHR-vpr. Following the infection, the cells were stained with propidium iodide, and the cell cycle was analyzed by flow cytometry with the Modfit software package (Verity Software House, Inc., Topsham, Maine).

ATM is not required for integration.

The ataxia telangectasia-mutated (ATM) gene is a close relative of ATR and is involved in the cellular response to double-stranded DNA breaks (16). ATR and ATM are both targets for inhibition with caffeine, the drug used in the studies by Daniel et al. (13). To test the possibility that ATM may be involved in the integration process, we used RNA interference to knock down ATM and assayed transduction efficiency as described for ATR. We found that ATM was not required for transduction with a lentivirus vector expressing luciferase as a reporter (Fig. 3A). Downregulation of ATM protein was evident by Western blot analysis (Fig. 3B).

FIG. 3.

Lentiviral transduction of ATM knockdown cells. ATM was knocked down in HeLa cells, which were then infected with the lentivirus vector pHR-Luc/pCMVΔR8.2Δvpr. Two days postinfection, the cells were assayed for luciferase activity as a measure of transduction.

Integration of a lentivirus vector is unaffected by the nature of the reporter gene.

The above knockdown experiments conclusively show that the presence of ATR or its function is not required for transduction with a lentivirus vector expressing luciferase as a reporter gene. These experiments therefore are in contradiction with the report by Daniel et al. (13), in which a different reporter, lacZ, was used.

To rule out the possibility that the results obtained were reporter specific, we conducted a similar experiment in U2OS/ATRkd cells in which the vector pHR-LacZ/pCMVΔR8.2Δvpr (26, 27), a lentivirus vector similar to pHR-Luc/pCMVΔR8.2Δvpr, expresses LacZ instead of luciferase. Upon assaying for β-galactosidase activity, we found no significant change in transduction (Fig. 4A) whether ATR was active or inhibited by overexpression of the dominant negative mutant. Western blot analysis was performed on cell lysates in order to monitor the induction of ATRkd, as shown in Fig. 1B (data not shown).

FIG. 4.

Transduction assay is not reporter specific and is not affected by unintegrated virus. (A) β-Galactosidase activity of U2OS/ATRkd cells incubated with or without doxycycline in order to induce expression of ATRkd infected with pHR-LacZ/pCMVΔR8.2Δvpr. (B) Luciferase activity of HeLa cells that were mock infected or infected with pHR-Luc/pCMVΔR8.2Δvpr or pHR-Luc/pCMVΔR8.2Δvpr-D116G.

Expression of luciferase or LacZ from unintegrated viral DNA does not account for the high level of luciferase activity in the absence of ATR.

It is not clear under what conditions and to what extent unintegrated retroviral DNA can be used as a template for transcription. A recent report by Poon et al. (31) suggested that unintegrated viral DNA from a luciferase-encoding lentivirus vector could produce small but significant amounts of luciferase activity. In addition Bell et al. reported that under certain cellular conditions unintegrated DNA could serve as a template to achieve expression equivalent to that of an integrated provirus (4).

To evaluate to what extent, if any, expression from unintegrated DNA might account for the observed reporter activity under conditions of ATR knockdown, we performed parallel experiments with lentivirus vectors pHR-Luc and pHR-LacZ produced with the packaging construct pCMVΔR8.2ΔvprD116G. Although integration-defective vectors cannot be titrated via expression of the reporter gene, they are competent for viral fusion and entry (9, 20). Therefore, integration-defective vectors can be titrated by the β-lactamase activity that results from the packaging of a Vpr-β-lactamase fusion protein into the virion during viral production (see Materials and Methods). Infection of HeLa cells with these integration-defective vectors showed that there was a low level of expression of luciferase, which accounts for 5% of the activity seen from transduction of an integrase-positive vector (Fig. 4B). Therefore, reporter gene expression from unintegrated viral DNA does not explain the high luciferase activity in the absence of ATR. Similar results were obtained with a LacZ reporter vector (data not shown).

Caffeine inhibition of ATR does not affect transduction by a lentivirus vector but does interfere with expression from the CMV immediate-early promoter.

Caffeine is known to be an inhibitor of ATR and ATM and was recently shown by Daniel et al. to inhibit retroviral integration (13). Because inhibition of ATR by ATRkd or via RNA interference in our experiments did not have a negative impact on lentiviral transduction, we reasoned that the apparent inhibition of transduction in the studies by Daniel et al. might be due to an unknown activity of caffeine other than ATR inhibition.

We tested the role of caffeine by adding it at the time of infection (4 mM) and maintaining it in the culture up to 24 h postinfection, as was described by Daniel et al. (13). The cells were then assayed for luciferase activity at 5 days postinfection. No change in the level of transduction was observed whether or not ATR had been inhibited by caffeine (Fig. 5A). To ensure that incubation with caffeine had sufficiently inhibited ATR, we repeated the above experiment, adding caffeine to the cells at the time of plating and maintaining caffeine treatment until the cells were assayed for luciferase activity. We observed a fourfold decrease in apparent transduction with extended incubation in 4 mM caffeine, suggesting that caffeine treatment may be affecting some late event during infection, such as expression of the luciferase reporter (Fig. 5B). To further examine this possibility, HeLa cells were electroporated with the pHR-Luc plasmid used to make the pHR-Luc vector. Following the electroporation, the cells were plated into two flasks, one containing 4 mM caffeine and the other containing no caffeine. The cells were incubated for 2 days, at which point they were assayed for luciferase activity. We observed a fourfold decrease in luciferase activity from the cells that had been treated with caffeine. The previous observation suggests that caffeine inhibits expression from the plasmids employed, not integration (Fig. 5C).

FIG. 5.

Effects of caffeine inhibition of ATR on lentivirus transduction and transcription. (A) Luciferase activity of HeLa cells treated with caffeine or not and mock infected or infected with pHR-Luc/pCMVDR8.2Dvpr. Caffeine was removed 24 h postinfection, and the cells were assayed for luciferase activity 5 days postinfection. (B) Caffeine was added to HeLa cells 24 h prior to infection with pHR-Luc/pCMVΔR8.2Δvpr or mock infection, and the cells were maintained in culture until luciferase activity was assayed, 5 days postinfection. (C) HeLa cells were electroporated with plasmid pHR-luc, following which the cells were plated with or without caffeine and assayed for luciferase activity 2 days postelectroporation.

Murine leukemia virus does not require ATR for successful integration.

We demonstrated that ATR is not required for HIV-1-derived lentivirus vectors to achieve efficient integration. Due to the remarkable differences between lentiviruses and oncoviruses, we felt it was compelling to examine whether ATR might play a role in facilitating or enhancing integration of an oncovirus vector. Thus, we performed experiments similar to those described in Fig. 1A with a murine leukemia virus vector, SRα-SIN-CMV-LUC, expressing firefly luciferase as a reporter. U2OS/ATRkd cells were treated with doxycycline in order to induce expression of ATRkd and then infected with SRα-SIN-CMV-LUC. Transduction efficiency was monitored via luciferase activity. Similar to the results found previously with pHR-Luc/pCMVΔR8.2Δvpr and pHR-LacZ/pCMVΔR8.2Δvpr, murine leukemia virus does not require ATR in order to integrate efficiently into the host DNA (Fig. 6).

FIG. 6.

ATR is not required for murine leukemia virus integration. U2OS/ATRkd cells were induced to express ATRkd and then mock infected or infected with the murine leukemia virus vector SRα-SIN-CMV-Luc containing luciferase as a reporter. At 2 days postinfection, the cells were assayed for luciferase activity.

Vpr does not enhance integration of HIV-1 in cycling cells.

In view of our earlier observation that Vpr has the ability to activate ATR (33, 41), we wished to test the effect, if any, that encapsidated Vpr had on integration efficiency. We packaged pHR-Luc with either pCMVΔR8.2 (which expresses wild-type Vpr) or pCMVΔR8.2Δvpr and HCMV-VSVG to produce virions that are identical except for the Vpr protein content in the virions. Neither of these vectors is capable of expressing Vpr once integrated, and therefore, any differences observed upon integration must be a result of encapsidated Vpr. Induced and uninduced U2OS/ATRkd cells were infected with either pHR-Luc/pCMVRΔ8.2Δvpr or pHR-Luc/pCMVΔR8.2, and luciferase activity was assayed. We did not observe any difference in luciferase activity whether or not Vpr was present (Fig. 7).

FIG. 7.

Vpr carried in the virion does not enhance integration of lentivirus vectors in dividing cells. U2OS/ATRkd cells were induced to express ATRkd and then infected with pHR-Luc/pCMVΔR8.2Δvpr or pHR-Luc/pCMVΔR8.2, which has Vpr packaged in the virion. At 2 days postinfection, the cells were assayed for luciferase activity as a measure of transduction.

DISCUSSION

Host DNA repair enzymes have become strong candidates for retroviral integration cofactors. One such repair enzyme, DNA-dependent protein kinase (DNA-PK), is a member of the phosphatidylinositol 3 kinase-related family of kinases (15). DNA-PK plays a central role in the nonhomologous end-joining pathway, which is responsible for repair of most double-strand breaks within the cell (23). Subsequent studies probing the role of DNA-PK in retroviral integration (3, 19, 22), however, have failed to reproduce the results of Daniel et al. (15). Here we have shown by several methods that ATR, another member of the PIKK family that is responsible for sensing double-strand breaks and replication stress (1), is not required for retroviral transduction of cycling cells. In addition we have shown that another PIKK, ATM, does not participate in retroviral integration. These data, taken together, suggest that the integration intermediate formed by retroviral integration may not be detected as a double-strand break or as replication stress.

It is possible that detection and repair of the integration intermediate is accomplished by other host DNA repair enzymes. For example, Yoder et al. found that various combinations of cellular enzymes with redundant activities (DNA polymerases beta and delta; DNA ligases I, III, and IV; and flap endonuclease) were competent for repair of synthetic viral-cellular DNA junctions in vitro (40).

The results that we have presented here are in disagreement with those reported by Daniel et al. (13, 14). While it is possible that some of the differences may be attributed to differences in cell lines (Daniel et al. used GM847 [10], whereas we used GK41 [28]), this would not explain why RNA interference-mediated knockdown of ATR has no effect on retroviral infection. Inhibition of the CMV promoter by caffeine may partly explain the discrepancies, but even in light of this, the data cannot be entirely reconciled.

Vpr was previously identified in the preintegration complex (for review, see reference 35) and has been shown to have a role in nuclear import of the preintegration complex into the nucleus in growth-arrested (17) and naturally nondividing cells (such as macrophages) (12, 36, 38). Virion-bound Vpr was also shown to play active roles during HIV-1 infection of cycling cells with regard to transactivation, induction of cell cycle disruption, and apoptosis (18, 31, 32). It is therefore possible that Vpr may play a role in HIV-1 integration in addition to its role in nuclear import, especially in cycling cells, where active nuclear import is not necessary. In addition to examining the role of ATR during integration in cycling cells, we were also able to explore the ability of Vpr packaged in the virion to facilitate or enhance HIV-1 infection. We have shown that encapsidated Vpr does not increase the efficiency of transduction by an HIV-1-derived vector in cycling cells. Our results do not rule out the possibility that encapsidated Vpr may play an unknown role during the early phases of infection of cycling cells.