PF74 Inhibits HIV-1 Integration by Altering the Composition of the Preintegration Complex

Antiretroviral therapy (ART) that uses various combinations of small molecule inhibitors has been highly effective in controlling HIV. However, the drugs used in the ART regimen are expensive, cause side effects, and face viral resistance. The HIV-1 CA plays critical roles in the virus life cycle and is an attractive therapeutic target. While currently there is no CA-based therapy, highly potent CA-specific inhibitors are being developed as a new class of antivirals. Efforts to develop a CA-targeted therapy can be aided through a clear understanding of the role of CA in HIV-1 infection. CA is well established to coordinate reverse transcription and nuclear entry of the virus. However, the role of CA in post-nuclear entry steps of HIV-1 infection is poorly understood. We show that a CA-specific drug PF74 inhibits HIV-1 integration revealing a novel role of this multifunctional viral protein in a post-nuclear entry step of HIV-1 infection.

ABSTRACT

The HIV-1 capsid protein (CA) facilitates reverse transcription and nuclear entry of the virus. However, CA’s role in post-nuclear entry steps remains speculative. We describe a direct link between CA and integration by employing the capsid inhibitor PF74 as a probe coupled with the biochemical analysis of HIV-1 preintegration complexes (PICs) isolated from acutely infected cells. At a low micromolar concentration, PF74 potently inhibited HIV-1 infection without affecting reverse transcription. Surprisingly, PF74 markedly reduced proviral integration owing to inhibition of nuclear entry and/or integration. However, a 2-fold reduction in nuclear entry by PF74 did not quantitatively correlate with the level of antiviral activity. Titration of PF74 against the integrase inhibitor raltegravir showed an additive antiviral effect that is dependent on a block at the post-nuclear entry step. PF74’s inhibitory effect was not due to the formation of defective viral DNA ends or a delay in integration, suggesting that the compound inhibits PIC-associated integration activity. Unexpectedly, PICs recovered from cells infected in the presence of PF74 exhibited elevated integration activity. PF74’s effect on PIC activity is CA specific since the compound did not increase the integration activity of PICs of a PF74-resistant HIV-1 CA mutant. Sucrose gradient-based fractionation studies revealed that PICs assembled in the presence of PF74 contained lower levels of CA, suggesting a negative association between CA and PIC-associated integration activity. Finally, the addition of a CA-specific antibody or PF74 inhibited PIC-associated integration activity. Collectively, our results demonstrate that PF74’s targeting of PIC-associated CA results in impaired HIV-1 integration.

IMPORTANCE Antiretroviral therapy (ART) that uses various combinations of small molecule inhibitors has been highly effective in controlling HIV. However, the drugs used in the ART regimen are expensive, cause side effects, and face viral resistance. The HIV-1 CA plays critical roles in the virus life cycle and is an attractive therapeutic target. While currently there is no CA-based therapy, highly potent CA-specific inhibitors are being developed as a new class of antivirals. Efforts to develop a CA-targeted therapy can be aided through a clear understanding of the role of CA in HIV-1 infection. CA is well established to coordinate reverse transcription and nuclear entry of the virus. However, the role of CA in post-nuclear entry steps of HIV-1 infection is poorly understood. We show that a CA-specific drug PF74 inhibits HIV-1 integration revealing a novel role of this multifunctional viral protein in a post-nuclear entry step of HIV-1 infection.

INTRODUCTION

The human immunodeficiency virus 1 (HIV-1) infects immune cells expressing the CD4 receptor and the chemokine coreceptors, CCR5 or CXCR4 (1). HIV-1 infection begins with the fusion of the viral membrane to the target cell plasma membrane (2). Subsequently, the fullerene cone-shaped viral core is released into the cytoplasm. Housed inside the viral capsid are two copies of the linear single-stranded viral RNA (vRNA) genome and a number of viral and cellular factors that facilitate postentry steps of infection (3–8). The cytoplasmic release of the core triggers conversion of the vRNA into a double-stranded viral DNA (vDNA) by the reverse transcription complex (RTC) (9). Consequently, the RTC transitions into a poorly characterized preintegration complex (PIC) that contains the vDNA and associated cellular and viral proteins, including the viral integrase (IN) enzyme (10). The PIC transports the vDNA to and across the nuclear pore complex (NPC) and integrates it into the host chromatin in the nucleus (10–13). Although the nuclear import of PICs involves active transport (14), the size limit imposed by the NPC requires shedding of CA from the core in a process defined as uncoating.

Reverse transcription and uncoating are highly coordinated events. Genetic, biochemical, and molecular studies show that the capsid is a major controller of these two early events (3, 4, 15). Perturbations of the capsid, including altering the intrinsic stability of the core via specific substitutions in CA (16, 17), expression of host restriction factors such as TRIM5α (18), and treatment with CA-targeting compounds (19), inhibit reverse transcription. Further, CA-binding cellular cofactors like cyclophilin A (CypA) promote reverse transcription in a CA-dependent manner (20, 21). Although the exact timing, location, and mechanism of uncoating are unclear, several models are currently being investigated. These include rapid uncoating before initiation of reverse transcription (22, 23), gradual uncoating during reverse transcription (22–24), and uncoating following engagement with the NPC (25, 26). While uncoating may facilitate nuclear entry, the prospect of the uncoating process exposing potential nuclear localization signals (NLS) in the viral replication complex lacks experimental evidence. Moreover, disruption of all the reported NLS in the PIC-resident viral elements did not preclude HIV-1 from infecting nondividing cells (27–33). Subsequently, HIV-1 CA as the viral determinant of nuclear entry was demonstrated when a chimeric virus containing MLV CA in the place of the HIV-1 CA, failed to infect nondividing cells (34). Concurrently, the reliance of HIV-1 infection on CA-dependent host factors like CPSF6 (11), transportin 3 (TNPO3) (11, 35), nucleoporin 153 (NUP153) (26, 36, 37), and Ran-binding protein 2 (RanBP2) (38, 39) established a role for CA in nuclear entry of HIV.

Besides the established role of CA in reverse transcription and nuclear entry, indirect evidence suggest that CA is also involved in post-nuclear entry steps of HIV-1 infection. Early biochemical studies implied that CA is completely shed from the PICs (23) and may not be directly involved in post-nuclear entry steps of infection (22, 40). However, subsequent studies detected CA in the viral replication complexes at/near the nuclear pores/envelope (41, 42) and also in the nuclear fraction of infected cells (25). The association of CA with either the replication complex or HIV-1 IN in the nucleus of target cell suggests that CA may directly promote post-nuclear entry steps of infection. Accordingly, an intranuclear role for CA is supported by infection studies of several HIV-1 CA mutants showing alterations in proviral integration and integration site selection. For example, the T54A/N57A and Q63/67A mutations in CA render HIV-1 infection cell cycle dependent (17, 34). While the cell cycle dependence of the T54A/N57A CA mutant maps to post-nuclear entry events (34), the Q63/67A CA mutant (17) is defective in nuclear entry, and the PICs retained more CA and were impaired for integration in vitro (16). In addition, the infectivity defect of the CA mutants A92E and G94D was attributed to impaired chromosomal integration (44, 45). The altered integration efficiency and integration site preferences rendered by the N74D mutation in CA (46, 47) and the retargeted integration by the CypA/RanBP2-independent CA mutant P90A (47, 48) suggest an intranuclear role for CA (38, 46, 47). The recent report that the CA-binding cellular cofactor CPSF6 enables targeting of the vDNA into gene dense regions further supports a role of CA in post-nuclear entry steps of HIV-1 infection (49, 50).

Studies of CA-targeting small molecule inhibitors also support CA’s role in post-nuclear entry steps of HIV-1 (51, 52). The CA-specific inhibitor PF74 inhibits HIV-1 replication and exhibits no inhibitory effect on HIV-1 protease or reverse transcriptase (RT) (53). Selection for drug resistance has identified substitutions in CA that confer resistance to the antiviral effects of PF74 (53–55). PF74 binds at the interface of CA between the N-terminal domain (NTD) and the C-terminal domain (CTD) and inhibits HIV-1 infection by blocking reverse transcription and nuclear entry (53, 54, 56). A recent report shows that PF74 treatment affects distribution of viral DNA integration into the host genome (57). The less potent antiviral compound BI-2 also binds to the same pocket within the CA targeted by PF74 (58). Interestingly, BI-2 does not affect reverse transcription but reduces nuclear entry of HIV-1 (58). Another study attributed that the antiviral effect of coumermycin A1 (C-A1) is linked to integration and the inhibitory mechanism may depend on the binding of the compound to CA (59). Although these studies suggest an intranuclear role for CA, biochemical data demonstrating a direct role of CA in HIV-1 integration are lacking.

We probed a direct link between CA and integration by combining the use of PF74 as a pharmacologic probe with the measurement of integration activity of HIV-1 PICs isolated from acutely infected cells. At 2 μM, PF74 (54) markedly inhibited HIV-1 infection without affecting reverse transcription. Although PF74 reduced nuclear entry, the level of reduction in nuclear entry did not account for the >95% inhibition in infection, indicating that PF74 also blocks a post-nuclear entry step(s). Interestingly, PF74 reduced proviral integration, and this reduction correlated quantitatively with antiviral activity. Unexpectedly, PICs extracted from cells cultured in the presence of PF74 retained higher integration activity in vitro. PF74 treatment neither dramatically affected the viral DNA ends nor resulted in the accumulation of PICs. Partial purification studies revealed that CA cosedimented with functional PICs. Interestingly, PICs extracted from PF74-treated cells contained lower levels of CA compared to untreated PICs. These data suggest that the impaired HIV-1 integration represent a component of the antiviral activity of PF74 and demonstrate a direct role for CA in viral DNA integration.

RESULTS

PF74 at a low concentration inhibits HIV-1 infection without reducing reverse transcription.

PF74 is an attractive small molecule for studying the role of CA in HIV-1 infection (54) and for understanding the antiviral mechanisms of CA inhibitors (57). Earlier studies report that at concentrations of 5 μM and higher, PF74 accelerates core disassembly (53, 54), which leads to abortive reverse transcription and the inhibition of HIV-1 infection (53, 54, 56, 60–62). In contrast, at concentrations of 2 μM and lower, PF74 inhibits HIV-1 infection by reducing nuclear entry of the viral DNA (60, 62). Interestingly, a recent study showed that treatment with a low concentration of PF74 (1.39 μM) resulted in the redistribution of viral DNA integration sites in the host cell genome (57). These studies suggest that PF74 can affect HIV-1 infection both prior to and after reverse transcription, depending on the concentration of the compound.

To better understand the antiviral mechanism of the CA-specific inhibitor, we tested the effects of PF74 on early steps of HIV-1 infection in the cultures of SupT1 cells, a T lymphocytic cell line that supports HIV-1 infection (63). SupT1 cells were inoculated with HIV-1/GFP reporter virions and cultured in the absence or presence of PF74 ranging from concentrations of 0.08 to 20 μM. An inhibitory effect on HIV-1 infection was determined by the reduction in cells with green fluorescent protein (GFP) fluorescence cultured in the presence of PF74 relative to the cells cultured without PF74. Data in Fig. 1A show that HIV-1 infection was inhibited ∼20% at the lowest concentration of PF74 (0.08 μM) used in this assay. The inhibitory effect of PF74 increased in a dose-dependent manner, with ∼90% inhibition observed at 0.63 μM and ∼95% inhibition at 1.25 μM. HIV-1 infection was completely inhibited by PF74 at concentrations of 10 μM and higher in accordance with previously reported data using other cell types (53, 54). The estimated 50% inhibitory concentration (IC₅₀) value of PF74 for HIV-1 infection in SupT1 cells is ∼0.3 μM.

FIG 1.

PF74 inhibits HIV-1 infection without reducing reverse transcription and partially blocking nuclear entry. (A) Anti-viral activity of PF74. SupT1 cells (5 × 10⁴ cells) were infected with HIV-1 GFP reporter virions (∼15 ng of p24) and cultured in the presence of increasing concentrations of PF74 (0 to 20 μM). At 24 h postinfection, the cells were collected, and the intracellular GFP fluorescence was measured by flow cytometry. The inhibitory effect of PF74 was determined from the percent reduction in GFP fluorescence of infected cells cultured in the presence of PF74 relative to infected cells cultured without the inhibitor. (B to D) Effect of PF74 on HIV-1 reverse transcription. SupT1 cells were inoculated with three different amounts (1,500 ng/ml [B], 150 ng/ml [C], or 15 ng/ml [D] of p24) of HIV-1 virions and cultured in the absence or presence of 2 and 10 μM PF74. At 16 h postinfection, the cells were harvested, and the total DNA isolated from these cells was analyzed by qPCR using primers specific for amplification of late RT products. The levels of viral DNA were quantified by calculating the copies of viral DNA from a standard curve generated in parallel and under same conditions of qPCR using 10-fold serial dilutions of known copy numbers (10⁰ to 10⁸) of an HIV-1 molecular clone. (E and F) Effects of PF74 on HIV-1 nuclear entry. SupT1 cells were inoculated with an equivalent amount of HIV-1 virions (150 ng/ml) and cultured in the absence or presence of 2 and 10 μM PF74. At 16 h postinfection, the cells were harvested, and the total DNA was isolated from these cells. Then, 100 ng of the DNA was subjected to qPCR to amplify the junctions of the HIV-1 2-LTR (E) and 1-LTR circles (F) using specific primer sets. The copy numbers of the 2-LTR and 1-LTR circles were calculated by extrapolating the qPCR data to a standard curve generated using 10-fold serial dilutions of the p2LTR and p1LTR plasmids, respectively. Data shown in panels B to F are mean values from at least three independent experiments, with error bars representing the standard errors of the mean.

Since PF74 can affect both reverse transcription and post-reverse transcription steps, we first tested the effects of PF74 on HIV-1 reverse transcription through the quantification of viral DNA synthesis in the presence or absence of the compound. SupT1 cells were inoculated with three different amounts of DNase I-treated HIV-1 virions and then cultured in the absence or presence of 2 μM (low concentration) and 10 μM (high concentration) PF74. At 24 h postinfection, the total DNA was isolated from the cells and subjected to quantitative real-time PCR (qPCR) to measure the accumulation of reverse transcription products. Viral DNA copy numbers were calculated relative to a standard curve generated using a molecular clone of HIV-1. The data revealed that infection in the presence of 10 μM PF74 resulted in a marked reduction in the accumulation of reverse transcribed HIV-1 DNA across the entire range of virus inoculum (Fig. 1B to D). This is reflective of the extent of PF74’s antiviral activity and is consistent with an early postentry defect previously described for higher concentrations of PF74 (53, 54, 56, 60–62). In contrast, treatment of infected cells with PF74 at 2 μM showed a minimal reduction in the synthesis of reverse transcription products (Fig. 1B to D). Yet, at 1.25 and 2.5 μM, PF74 treatment resulted in an ∼95% reduction in HIV-1 infection (Fig. 1A), indicating that the compound inhibits infection by interfering with a step in infection subsequent to reverse transcription.

PF74 treatment reduces HIV-1 nuclear entry.

To pinpoint the mechanism of inhibition, we tested whether the antiviral effect of PF74 at low concentrations can be attributed to an effect on nuclear entry. We quantified the 1-long terminal repeat (1-LTR) and 2-LTR circles, two commonly used surrogate markers of nuclear import of retroviral DNA (14, 64–67). While the 2-LTR circles are generated by nonhomologous end joining and are exclusively present in the nucleus, more than 90% of the LTR-containing circles are generated by homologous recombination of the 5′ and 3′ LTRs, resulting in the formation of 1-LTR circles in the nucleus (14, 64–67). Although these circles constitute unproductive products, they are useful reporters of HIV-1 nuclear entry (68). First, we quantified copies of 2-LTR circles in infected SupT1 cells cultured in the presence or absence of PF74. The copy numbers of 2-LTR circles were calculated by extrapolating data to a standard curve generated using a plasmid containing the 2-LTR sequence (p2-LTR). We observed a marked reduction in the accumulation of 2-LTR circles upon treatment with either 2 or 10 μM PF74 (Fig. 1E), suggesting that the compound inhibits viral nuclear entry. The essentially complete reduction in the number of 2-LTR circles in cells treated with 10 μM PF74 is likely a manifestation of the severely impaired reverse transcription (Fig. 1B to D). Unexpectedly, the 50% reduction in 2-LTR circles in cells treated with 2 μM PF74 (Fig. 1E) does not quantitatively account for the >95% reduction in infection observed at this concentration of the drug (Fig. 1A).

Besides the nuclear 2-LTR circles formed by end-to-end ligation, 2-LTR circles can also be formed in the cytoplasm of the target cell due to autointegration, an IN-dependent process where the 3′-processed vDNA forms a less than full-length circle (69). Therefore, interpretation of nuclear entry effects through the measurement of 2-LTR circles can be confounded by the presence of autointegration products (20). Therefore, to better understand PF74’s effect on the nuclear entry step, we quantified 1-LTR circles that are formed exclusively in the nucleus of target cells by the nuclear localized homologous recombination machinery (70). We employed a qPCR-based method to quantify 1-LTR circles in infected SupT1 cells cultured in the presence or absence of PF74. The copy numbers of 1-LTR circles were calculated by extrapolating qPCR data to the standard curve generated by employing a plasmid (p1-LTR) containing the 1-LTR sequences of HIV-1. Results from these analyses revealed that the copies of 1-LTR circles in infected cells are markedly higher (∼10-fold) compared to the 2-LTR circles (Fig. 1F), in accordance with the evidence that 1-LTR circles are approximately 10-fold more abundant than 2-LTR circles in infected cells (71). As in the case of 2-LTR circles (Fig. 1E), our data also show that 10 μM PF74 treatment led to a marked and almost complete reduction in the 1-LTR circles, whereas, at 2 μM, the number of 1-LTR circles was reduced by approximately 50% (Fig. 1F). Collectively, the quantification of LTR circles results in Fig. 1E and F establish that at low micromolar concentration PF74 inhibits nuclear entry of HIV-1 without affecting the reverse transcription step.

PF74 potently inhibits proviral DNA integration.

Our results in Fig. 1A to D suggested that the antiviral effect of high concentration of PF74 may be manifested at the reverse transcription step, whereas the nuclear entry step is inhibited at low concentrations of the drug. Interestingly, our results from the 2-LTR and 1-LTR circle measurements (Fig. 1E and F) show that the level of reduction in nuclear entry with 2 μM PF74 did not quantitatively correlate with the level of reduction in infection, indicating that the compound also blocks a step after nuclear entry. To identify the post-nuclear entry step(s) of infection targeted by PF74, we measured integration of HIV-1 DNA into the SupT1 genome by Alu-based nested qPCR (71). SupT1 cells were inoculated with HIV-1 virions at three different concentrations, and genomic DNA from these cells was subjected to qPCR. The copy numbers of the integrated viral DNA (proviral DNA) were determined relative to a standard curve generated using the proviral molecular clone in the second round qPCR. Results from these analysis show that 2 μM PF74 treatment resulted in a marked reduction in proviral integration at all three concentrations of virus inoculum (Fig. 2A to C). Importantly, the 2 μM PF74-induced reduction in integration correlated quantitatively with the >95% reduction in infection (Fig. 2D). These results indicate that the infection block is most likely due to a failure of the virus to integrate into the host genome.

FIG 2.

PF74 treatment markedly reduces HIV-1 integration. (A to C) Effects of PF74 on proviral DNA integration. SupT1 cells were inoculated with three different amounts (1,500 ng/ml [A], 150 ng/ml [B], or 15 ng/ml [C] of p24) of HIV-1 particles and cultured in the absence or presence of 2 and 10 μM PF74. At 16 h postinfection, the cells were harvested, and the total DNA from these cells was subjected to Alu-based qPCR to measure proviral DNA integration into the host genome. The copies of viral DNA integration were calculated by extrapolating the integration data into a standard curve generated using known copy numbers (10⁰ to 10⁸) of an HIV-1 molecular clone. Data in panels A to C represent mean values from three independent experiments, with standard errors of the mean. (D) Cumulative effect of PF74 on HIV-1 reverse transcription, LTR circles, and proviral DNA integration. To analyze PF74’s antiviral effect, the observed data for PF74’s effect on infection (Fig. 1A), reverse transcription (Fig. 1C), 2-LTR circles (Fig. 1E), 1-LTR circles (Fig. 1F), and proviral integration (Fig. 2B) were plotted as the percent versus controls (untreated cells). (E and F) Effects of RAL treatment on PF74-induced LTR circle formation. SupT1 cells were inoculated with an equivalent amount of HIV-1 virions (150 ng/ml) and cultured in the absence or presence of 2 μM PF74, 1 μM RAL, or a combined treatment of 2 μM PF74 and 1 μM RAL. At 16 h postinfection, the total DNA from the infected cells was isolated and analyzed by qPCR to amplify the junctions of the HIV-1 2-LTR and 1-LTR circles. Copies of the 2-LTR circles (E) and 1-LTR circles (F) were quantified using a standard curve, as described for Fig. 1E and F. Data shown in panels A, B, C, E, and F are mean values from three independent experiments, with error bars representing the standard errors of the mean.

Next, we probed whether the marked reduction in viral DNA integration with 2 μM PF74 is due to the 2-fold reduction in nuclear entry block of the virus (Fig. 1E and F). To test this, we measured the effects of the integrase inhibitor raltegravir (RAL) on PF74-induced 2-LTR and 1-LTR circle formation. As a potent inhibitor of the nuclear strand-transfer step in HIV-1 integration, RAL treatment eliminates autointegration and increases the formation of 2-LTR circles in infected cells (71, 74). Hence, we reasoned that if PF74 inhibited viral DNA integration independent of its effect on nuclear entry, then RAL treatment should result in an increase in nuclear LTR circles in PF74-treated cells. SupT1 cells were inoculated with HIV-1 and cultured in the absence or presence of PF74, RAL, or both PF74 and RAL. Total DNA from these cells was isolated and subjected to qPCR to measure accumulation of 2-LTR and 1-LTR circles as described in Fig. 1E and F. The results from these analysis show that RAL treatment alone markedly increased the levels of both 2-LTR (Fig. 2E) and 1-LTR circles (Fig. 2F) in the infected cells compared to the infected cells cultured without RAL. Similar to the data in Fig. 1E and F, 2 μM PF74 treatment resulted in a 2-fold reduction in the number of 2-LTR (Fig. 2E) and 1-LTR (Fig. 2F) circles compared to the untreated control cells. Interestingly, treatment of cells with both RAL and PF74 (2 μM) led to an increase in the number of 2-LTR and 1-LTR circles relative to the respective LTR circles by PF74 treatment alone. For example, the number of 2-LTR circles in cells treated with PF74 (∼1,690 copies) increased ∼45% compared to cells treated with both RAL and PF74 (∼3,160 copies) (Fig. 2E). A similar level (∼45%) of increase in 1-LTR circles was also observed when the number of 1-LTR circles in cells treated with both the inhibitors (∼25,205) was compared to 1-LTR circles in PF74 treated cells (∼13,205) (Fig. 2F). Accordingly, the numbers of 2-LTR and 1-LTR circles in the cells treated with both RAL and PF74 were dramatically reduced compared to the single treatment of RAL (Fig. 2E and F). Interestingly, the observed reduction of circular HIV-1 DNA reflective of the nuclear entry block by PF74 did not quantitatively correlate with the level of inhibition (>95%) of integration (Fig. 2A to C) or infection observed with 2 μM PF74 (Fig. 1A). These data further support an inhibitory effect of PF74 after nuclear entry of the virus and suggest that a component of the antiviral activity of the drug depends on reduced viral DNA integration.

PF74 and RAL additively inhibit HIV-1 infection.

To better understand the inhibitory effect of PF74 at a post-nuclear entry step, we titrated PF74 and RAL against each other and measured HIV-1 infection. We reasoned that if PF74’s inhibitory effect on HIV-1 integration is a consequence of the block at nuclear entry, then RAL-inhibition will be insensitive to PF74 treatment. We used a luciferase reporter assay by inoculating TZM-bl cells with pseudotyped HIV-1 particles. Infected cells were cultured in the absence or presence of various concentrations of PF74, RAL, or both PF74 and RAL. Cellular lysates prepared at 48 h postinfection were analyzed for luciferase activity as an indicator of infection. Results from these analysis show that RAL treatment inhibits HIV-1 infection in a dose-dependent manner (Fig. 3A). RAL at 5 nM reduced infectivity by ∼20%, whereas treatment with RAL at 50 nM inhibited infectivity by ∼80% (Fig. 3A). Accordingly, infectivity was reduced to >90% with 100 nM RAL, followed by >95% inhibition with a 200 nM concentration of the inhibitor (data not shown). Similarly, PF74 treatment inhibited HIV-1 infection in a concentration-dependent manner in the TZM-bl cells (Fig. 3B). Treatment with 0.1 μM PF74 reduced infection up to 50%, whereas >95% inhibition was achieved with concentrations above 0.5 μM (Fig. 3B), consistent with the antiviral activity of the CA inhibitor observed with SupT1 cells (Fig. 1A). Then, based on the extent of inhibition, we selected 25 and 50 nM RAL for the titration studies against a range of PF74 concentrations (0.2 to 2.0 μM). TZM-bl cells inoculated with HIV-1 particles were treated with PF74 in the absence or presence of either 25 nM (Fig. 3C) or 100 nM (Fig. 3D) RAL. Luciferase activity measurements revealed that the combined treatment of PF74 and RAL resulted in an additive antiviral effect compared to the inhibitory effect conferred by either of the inhibitors (Fig. 3C and D). For example, treatment with 0.2 μM PF74 alone reduced infection by ∼45%, whereas 25 nM RAL showed an ∼70% reduction in infection (Fig. 3C). Interestingly, the combined treatment of 0.2 μM PF74 and 25 nM RAL inhibited infection by ∼85%. The additive antiviral effect of PF74 with RAL was consistently observed in studies with 0.5 and 2.0 μM PF74 (Fig. 3C). Accordingly, titration studies of PF74 against 100 nM RAL also revealed that the presence of both drugs reduced HIV-1 infection to a greater degree than did a singular treatment of PF74 or RAL (Fig. 3D). Finally, titration of 2.0 μM PF74 with a range of concentrations of RAL (5 to 100 nM) confirmed the additive antiviral effects of these two inhibitors (Fig. 3E). For instance, with 5 nM RAL, treatment infectivity was reduced by ∼15%. However, in the presence of 5 nM RAL and 2.0 μM PF74, a >95% reduction in infection was observed. Considering a 2-fold reduction in nuclear entry with 2.0 μM PF74 treatment (Fig. 1E and F), this level of reduction in infectivity (from ∼85% to ∼5%) with both the inhibitors strongly support that the antiviral effect of PF74 in part depends on the inhibition of a post-nuclear entry step of HIV-1 infection.

FIG 3.

Effects of RAL treatment on the antiviral activity of PF74. TZM-bl cells were inoculated with 5 ng/ml of pseudotyped HIV-1 particles and cultured in the absence or presence of RAL (5 to 100 nM) (A) or PF74 (0.1 to 2 μM) (B). At 48 h postinfection, cellular lysates were prepared and subjected to luciferase activity measurements as an indicator of infection. Data shown are mean values from three independent experiments, with error bars representing the standard errors of the mean. (C and D) Titration of PF74 against RAL. TZM-bl cells inoculated with pseudotyped HIV-1 were treated with either 25 nM RAL (C) or 100 nM RAL (D) in the absence or presence of a range of concentrations of PF74 (0.2 to 2 μM). At 48 h postinfection, the luciferase activity was measured in the cellular lysates, and the data are plotted as the percent infection versus the controls. (E) Titration of RAL against PF74. TZM-bl cells were inoculated with pseudotyped HIV-1 and treated with PF74 (2 μM) in the absence or presence of RAL (5 to 100 nM). At 48 h postinfection, the luciferase activity was measured and plotted as the percent infection versus the controls. Data presented in panels C to E are representative of three independent experiments.

PF74’s inhibitory effect on HIV-1 integration is not due to a delay in the integration pathway.

To better understand the antiviral mechanism of the CA inhibitor, we probed whether the inhibitory effect of PF74 on HIV-1 integration is due to slowing down of the viral replication complexes (PICs) proceeding to integration. To test this, we performed a time-of-addition infectivity assay using RAL as a probe. TZM-bl cells were inoculated with HIV-1 particles in the presence or absence of 2.0 μM PF74. RAL (25 nM) was then added to the cells at 0, 1, 2, 3, 5, 8, 12, and 16 h postinfection. At 48 h postinfection, the luciferase activity was measured in the cellular lysates as an indicator of infectivity (Fig. 4). As expected, addition of RAL during infection (i.e., at 0 h) markedly inhibited HIV-1 infection (Fig. 4A). This level of inhibition (∼70 to 75% relative to untreated control cells) remained consistent in cells exposed to RAL up to 5 h postinfection (Fig. 4A). However, when RAL was added at or after 8 h postinfection, the inhibitory effect was diminished relative to the drug treatment at 0 h. For example, the ∼75% reduction in infection with RAL treatment during infection was decreased to ∼55% in cells treated with RAL at 12 h and to ∼50% with the inhibitor treatment at 18 h. Analysis of the infection data relative to the RAL treatment at 0 h show that RAL treatment at 12 and 18 h rescued infection up to 2-fold (Fig. 4A). Interestingly, in the presence of PF74, infectivity never recovered irrespective of the time of RAL addition (Fig. 4B). These data indicate that the rescue of infectivity from RAL inhibition is countered by PF74 in accordance with an additive antiviral effect by both the drugs.

FIG 4.

Time-of-addition assay of PF74 and/or RAL and measurement of HIV-1 infectivity. TZM-bl cells inoculated with pseudotyped HIV-1 particles (5 ng/ml) were subjected to a time-of-addition assay in the absence or presence of RAL, PF74, or RAL plus PF74. At 48 h postinfection, cellular lysates were prepared and subjected to luciferase activity measurements as an indicator of infection. (A and B) Time of RAL addition assay. Cells were cultured in the absence (A) or presence (B) of PF74 (2 μM). To these cells, RAL (25 nM) was added at 0, 1, 2, 5, 8, 12, and 16 h relative to infection. At 48 h postinfection, the luciferase activity was measured in the cellular lysates and plotted as the percent infection versus the controls. (C and D) Time-of-PF74-addition assay. Cells were cultured in the absence (C) or presence (D) of RAL (25 nM). PF74 (2 μM) was added at different time points (0, 1, 2, 5, 8, 12, and 16 h) relative to the infection to these cells. At 48 h postinfection, cellular lysates were prepared, and the luciferase activity was measured. The data are plotted as the percent infection versus the controls. Data presented are representative from three independent experiments, with error bars representing the standard errors of the mean.

Next, we measured infectivity by adding PF74 at increasing time points relative to infection in the absence or presence of RAL (Fig. 4C and D). Similar to the RAL inhibition data (Fig. 4A), PF74 treatment during infection dramatically reduced HIV-1 infection (Fig. 4C). The level of PF74’s inhibition (>95%) remained consistent until the addition of the drug 5 h postinfection. Interestingly, when PF74 was added at or after 8 h, the antiviral effect was reduced relative to the drug treatment at 0 h (see Fig. 6C), which is consistent with the rescue of RAL inhibition (see Fig. 6A). Accordingly, when PF74 was added at 12 and 16 h postinfection, a marked rescue of inhibition was observed (see Fig. 6C). Interestingly, the timing of rescue of PF74 inhibition (see Fig. 6C) is quite similar to that of the RAL addition data (see Fig. 6A), indicating that the inhibitor did not delay the integration pathway. Subsequently, in the presence of RAL, the inhibitory effect of PF74 treatment after 8 h was not rescued (Fig. 4D). The lack of rescue by RAL imply that the inhibitory effects of PF74 on infection is not due to the slowing down of the viral replication complexes proceeding to integration.

FIG 6.

Analysis of PIC composition. (A) Fractionation of Cy-PICs and integration activity measurements. Portions (1 ml) of Cy-PICs were overlaid on a freshly prepared, 10 to 50% linear gradient of sucrose and subjected to ultracentrifugation for 2 h at 4°C. Fractions of 1 ml were collected starting from the top and analyzed for integration activity using 200-μl aliquots of each fraction. (B) PIC activity. Integration assays were performed in the presence of the integrase inhibitor RAL using fraction 14 as the source of PIC activity. (C) Viral DNA content in the fractionated PICs. Aliquots (200 μl) of each fraction were subjected to qPCR to quantitate copies of vDNA using a standard curve. (D and E) Quantification of CA in the fractionated PICs. Aliquots (100 μl) of each fraction were subjected to ELISA to quantitate the amount of CA by extrapolating the data to a standard curve generated using purified recombinant CA. The data show the CA levels in the fractioned PICs isolated from cells inoculated with enveloped (D) and VSV-G-pseudotyped (E) HIV-1. The data shown are averages of three independent experiments. (F and G) Detection of PIC-associated CA by immunoprecipitation. To detect PIC-associated CA by pulldown assays, we used magnetic bead-coated GST-CypA (A) and anti-CA antibodies (B). The GST-CypA or antibody-coated beads or equivalent amounts of protein A/G-magnetic beads (negative control) were incubated with cPICs or CA-A14C/E45C tubes (positive control) or recombinant monomeric CA. The bound complexes were subjected to centrifugation and washed, and the pelleted CA was subjected to immunoblot analysis. Lanes: S, supernatant after centrifugation; W, wash; E, eluate. The data are representative of three independent experiments.

PF74 does not dramatically alter formation of viral DNA ends.

Proviral integration of HIV-1 is dependent on the formation of proper viral DNA ends that are necessary for the strand transfer step (10, 12). Therefore, next we tested whether PF74’s inhibitory effect on HIV-1 integration is due to alterations in the viral DNA ends. To test this, we analyzed the nucleotide sequences of the 2-LTR junctions that are formed by ligation of the unintegrated vDNA. SupT1 cells were inoculated with HIV-1 and cultured in the absence or presence of 2.0 μM PF74. Total DNA from these cells was isolated and the 2-LTR junction DNA was amplified by PCR. The amplified DNA was cloned and sequenced, and the junctions were analyzed by multiple alignment. Results from this analysis reveal that in untreated cells ∼35% of the sequences of the 2-LTR junctions represent the canonical GT-CA nucleotides (Table 1 ). In PF74-treated cells, the percentage of canonical sequences was slightly reduced to ∼25% (Table 1). As expected, there were panoply of other sequences detected at the 2-LTR junctions both in the untreated and PF74-treated cells. For example, ∼15% of the sequences in the untreated cells represented unprocessed nucleotides of both U5 and U3 regions. This number was not changed in PF74-treated cells, given that ∼15% of unprocessed sequences were also detected. Similarly, there were sequences containing both deletions and insertions of nucleotides of varied lengths in both untreated and PF74-treated cells. Interestingly, the percentages of insertions of one or more nucleotides into the junctions were modestly higher (35%) in PF74-treated cells than in untreated cells (20%). The levels of deletions were not dramatically altered in the inhibitor-treated cells. Collectively, the analysis of the 2-LTR junctions show that PF74 treatment did not dramatically alter the formation of viral DNA ends that are necessary for proviral integration.

TABLE 1.

Effects of PF74 on the 2-LTR junction sequences^a

2-LTR junction type(s)	No. of clones/total no. of clones tested (%)
	(–) PF74	(+) PF74
Consensus	7/20 (35)	5/20 (25)
U5 unprocessed and U3 processed	1/20 (5)	2/20 (10)
U5 processed and U3 unprocessed	2/20 (10)	1/20 (5)
Partially processed	1/20 (5)	1/20 (5)
Insertions	4/20 (20)	7/20 (35)
Deletions	4/20 (20)	3/20 (15)
Other	1/20 (5)	1/20 (5)

^aSupT1 cells were inoculated with 150 ng/ml p24 equivalent of HIV-1 virions and cultured in the absence (–) or presence (+) of 2 μM PF74. At 16 h postinfection, the cells were harvested, and the total DNA was isolated from these cells. Portions (100 ng) of the DNA were subjected to PCR to amplify the junctions of the HIV-1 2-LTR circles. The amplified DNA was gel purified, cloned, and analyzed by DNA sequencing. The DNA sequences from 20 representative clones from untreated and treated samples were aligned by multiple alignment to determine effects of PF74 on viral DNA ends. Sequences are calculated as the percentages of clones over the total clones analyzed. The data are presented as the percentages of various 2-LTR junction sequences in untreated and PF74-treated cells.

HIV-1 PICs assembled in the presence of PF74 show enhanced integration activity.

To understand the mechanism by which PF74 inhibits HIV-1 integration, we biochemically studied PICs, the viral replication complex that carries out viral DNA integration and retains integration activity in vitro (72, 75). We predicted that PICs recovered from infected cells treated with 2 μM PF74 would be impaired for integration activity since the inhibitor did not affect reverse transcription (Fig. 1B to D), reduced nuclear entry to an extent that is insufficient to account for the marked inhibition in infection (Fig. 2D to F), did not delay the integration pathway (Fig. 4), or resulted in a dramatic alteration in the formation of viral DNA ends (Tables 1 and 2). To test this hypothesis, we prepared cytoplasmic extracts containing PICs (Cy-PICs) from SupT1 cells acutely infected with a high concentration (∼1500 ng of p24 equivalent) of DNase I-treated HIV-1 particles (72, 73, 76). These PICs were subjected to integration assay in vitro, and viral DNA integration was quantified by nested qPCR (76, 77). The data in Fig. 5A show that the Cy-PICs efficiently integrate the viral DNA into the target DNA and that addition of the integrase inhibitor RAL significantly impairs such integration activity. Importantly, the preparation of cytoplasmic extracts from uninfected cells and the absence of Cy-PICs exhibited no integration activity, showing the specificity of viral DNA integration in the assay. Moreover, dilution of the Cy-PICs resulted in a proportional reduction in integration activity, indicating that the assay measurements are conducted mostly within the linear range of PIC assay (Fig. 5B and C).

TABLE 2.

Nucleotide sequence of HIV-1 2-LTR junction DNA^a

Junction type(s)	Right LTR sequence	Terminal nucleotides	Left LTR sequence
Consensus	GGAAAATCTCTAGCA	GTAC	TGGAAGGGCTAATTC
U5 unprocessed and U3 processed	GGAAAATCTCTAGCA	GT	TGGAAGGGCTAATTC
U5 processed and U3 unprocessed	GGAAAATCTCTAGCA	AC	TGGAAGGGCTAATTC
Partially processed	GGAAAATCTCTAGCA	G	TGGAAGGGCTAATTC
Insertions	GGAAAATCTCTAGCA	GT(X)AC	TGGAAGGGCTAATTC
Deletions	GGAAAATCTCTA		------ AAGGGCTAATTC

^aSequences of various types of HIV-1 2-LTR junctions with nucleotides at the right LTR, left LTR, and termini are shown.

FIG 5.

Effect of PF74 on the integration activity of HIV-1 PICs. (A) Measurement of PIC-associated integration activity in vitro. SupT1 cells were inoculated with high-titer HIV-1 particles (1,500 ng/ml of p24), and PIC-containing cytoplasmic extracts (Cy-PICs) were isolated from these cells. In vitro integration assays using the Cy-PICs as the source of integration activity and quantification of viral DNA integration by nested qPCR were carried out as described previously (76, 77). Several controls were included in parallel to ascertain the specificity of the PIC-associated integration activity. (B) Integration activity with different amounts of PICs. Various amounts of Cy-PICs were used as the source of integration activity in the assay, and viral DNA integration was calculated as described for panel A. (C) Relative integration activity. Copies of integrated viral DNA in diluted Cy-PICs were plotted as the percentage versus the copies of viral DNA integration relative to undiluted (200 μl) Cy-PICs. (D to G) Integration activity of wild-type HIV-1 PICs assembled in the presence of PF74. (D) Integration activity of Cy-PICs. Acutely infected SupT1 cells were cultured in the absence or presence of PF74. The cells were harvested, and cytoplasmic extracts containing PICs were isolated from these cells. The integration activities of these PICs were measured in vitro, and the viral DNA copy numbers were determined. Mean values from three independent experiments are shown, with error bars representing the standard errors of the mean. (E) Viral DNA copy numbers in the Cy-PICs. Copies of viral DNA in Cy-PICs assembled in the absence or presence of PF74 were quantified by qPCR using a standard curve generated according to the method described for Fig. 1B. (F) Integration activity of Nu-PICs. To generate Nu-PICs, nuclear pellets of acutely infected SupT1 cells were homogenized, and the PIC-containing nuclear extracts (Nu-PICs) were isolated as described in Materials and Methods. In vitro integration assays of Nu-PICs and quantification of integrated viral DNA were performed using a standard curve as described for Fig. 5A. (G) Viral DNA copy numbers in the Nu-PICs. Copies of Nu-PIC-associated viral DNA were quantified by qPCR. The data shown are mean values from three independent experiments, with error bars representing the standard errors of the mean. (H to K) Integration activities of PF74-resistant HIV-1 5Mut PICs assembled in the presence of PF74. SupT1 cells were inoculated with a high titer (∼1,500 ng/ml of p24) of HIV-1 5Mut particles, and Cy-PICs and Nu-PICs were prepared from these cells. The integration activities of Cy-PICs (H) and Nu-PICs (I) were measured, and the copy numbers of viral DNA Cy-PICs (J) and Nu-PICs (K) were determined by qPCR as described in the text. Mean values from three independent experiments are shown, with error bars representing the standard errors of the mean.

Having established the specific conditions to quantify viral DNA integration in vitro, we analyzed the effects of PF74 on PIC-associated integration activity. Cy-PICs were extracted from SupT1 cells that were inoculated with HIV-1 and cultured in the presence or absence of 2 and 10 μM PF74. The activity of these Cy-PICs was assayed in vitro, and viral DNA integration was quantified as described in Fig. 5A. Surprisingly, the Cy-PICs from the 2 μM PF74-treated cells displayed higher integration activity relative to the Cy-PICs from untreated cells (Fig. 5D). The amount of viral DNA in the Cy-PICs from untreated and 2 μM PF74-treated cells was comparable (Fig. 5E), suggesting that PF74 specifically reduced the activity of PICs. In addition, PIC preparations from 10 μM PF74-treated cells showed markedly reduced integration activity (Fig. 5D), owing to reduced viral DNA in these PICs (Fig. 5E).

Although the integration activity and the viral DNA content of Cy-PICs were unaffected by 2 μM PF74 treatment, the LTR circle data in Fig. 1E and F suggest that only ∼40 to 50% of the PICs are able to enter the nuclei of infected cells. To analyze whether the PICs in the nuclei (Nu-PICs) of cells remain competent for integrating the viral DNA into the host genome, we prepared nuclear extracts (Nu-PICs) from acutely infected cells cultured in the absence or presence of PF74 and measured their integration activity in vitro. The activity data show that Nu-PICs assembled in the presence of 2 μM PF74 are competent for integrating the viral DNA into target DNA (Fig. 5F). Interestingly, as in the case of the Cy-PICs in Fig. 5D, the Nu-PICs from the 2 μM PF74-treated cells displayed elevated integration activity relative to the Nu-PICs from untreated cells (Fig. 5F). Accordingly, qPCR analysis also showed that the copies of viral DNA in the Nu-PICs were unaffected by 2 μM PF74 treatment (Fig. 5G). These results indicate that in 2 μM PF74-treated cells, the Cy-PICs entering the nucleus remain functional and retain robust integration activity. Collectively, the functional analysis of Cy-PICs and Nu-PICs suggest that 2 μM PF74 treatment does not negatively affect the PIC-associated integration activity.

PF74 does not affect the integration activity of PF74-resistant HIV-1 (5Mut) PICs.

The antiviral action of PF74 has been genetically linked to the binding of the compound at a CA-CA interface between the NTD and the CTD of adjacent subunits within CA hexamers (53, 54, 78). Five substitutions (5Mut) in CA collectively confer resistance to the antiviral effects of PF74 by markedly reducing binding of this inhibitor to capsid (53, 54). To test whether the effects of PF74 on PIC function is mediated via the compound specifically acting on HIV-1 CA, we studied the Cy-PICs and Nu-PICs of 5Mut virus. SupT1 cells were inoculated with HIV-1 5Mut virions and cultured in the presence of 2 and 10 μM PF74. Cytoplasmic and nuclear fractions were prepared as the sources of Cy-PICs and Nu-PICs. Integration activity measurements showed no significant effect of 2 or 10 μM PF74 on viral DNA integration of 5Mut Cy-PICs compared to untreated controls (Fig. 5H). Accordingly, PF74 treatment did not alter the levels of associated viral DNA of the 5Mut PICs (Fig. 5I). Similar results were also obtained with Nu-PICs from cells cultured in the presence of PF74 (Fig. 5J and K). These data indicate that the effect of PF74 on PIC activity is a result of the compound specifically acting on HIV-1 CA and suggest that PIC-associated CA can affect viral DNA integration.

Analysis of PIC composition show cosedimentation of CA with active PICs.

PICs assembled in the presence of PF74 showed elevated integration activity (Fig. 5D and F), whereas such an increase was not evident with drug-resistant 5Mut PICs (Fig. 5H and J). Since PF74 specifically binds to HIV-1 CA and has been reported to alter capsid stability (54, 79), we hypothesized that PF74 affects integration activity by altering CA levels in the PICs. To test this, we fractionated PICs on a linear gradient of sucrose. A portion (1 ml) of Cy-PICs was overlaid onto a freshly prepared sucrose gradient and subjected to ultracentrifugation (16). Thereafter, 1-ml fractions were collected from the top and analyzed for integration activity and viral DNA by qPCR in parallel. Integration activity measurements show that fractions 1 through 6 lacked PIC activity (Fig. 6A). However, integration activity was detected in fractions 7 through 15, with peak integration activity eluting in fractions 12 through 14 (Fig. 6A). Quantification of viral DNA also showed a peak in fractions 12 through 14 (Fig. 6C) corresponding to the integration activity in these fractions (Fig. 6A). The elution of overlapping peaks of integration activity and viral DNA in fractions 10 through 15 indicated the presence of functional PICs.

In parallel, to determine the association of CA in functional PICs, we quantified CA levels in the fractions by p24 enzyme-linked immunosorbent assay (ELISA) (16). We detected highest levels of CA in the top fractions of the gradient, likely corresponding to soluble CA in the Cy-PIC preparation (Fig. 6D). The levels of CA in the subsequent fractions decreased markedly with the lowest levels of CA detected in fractions 8 and 9. Interestingly, from fraction 10, the amount of CA started to increase relative to fraction 9 and a peak of CA elution was observed in fractions 12 through 14. The increased levels of CA in the last fraction (fraction 15) represented the pelleted protein due to centrifugation. Comparative analysis of the levels of CA to the integration activity and viral DNA in fractions 11 through 14 (Fig. 6D) show that CA is associated with functional PICs. Accordingly, immunoprecipitation assays with a monoclonal CA-specific antibody also detected CA in the cytoplasmic PICs (Fig. 6G) but not with a GST (glutathione S-transferase)-CypA fusion protein (Fig. 6F). We also fractionated PICs extracted from cells treated with the HIV-1 reverse transcriptase inhibitor efavirenz. Surprisingly, extraction of cytoplasmic PICs from these cells (even with 5 μM efavirenz) retained detectable levels of integration activity concurrent with the elution of very small amount of CA in the PIC fractions (data not shown). We predict that the incomplete arrest of reverse transcription is due to the use of high titer virus that is necessary to extract sufficient PICs. Unfortunately, fractionation studies with a smaller amount of virus were not feasible due to the detection limit of the p24 ELISA.

In an effort to determine whether the presence of CA in the functional PIC fractions is dependent on the viral entry pathway, we analyzed PICs of vesicular stomatitis virus G (VSV-G)-pseudotyped HIV-1 that utilizes the endocytic pathway to enter the target cell (80, 81). Similar to the elution profile of the enveloped HIV-1, the integration activity and viral DNA eluted in fractions 11 through 14 (data not shown). Accordingly, a CA peak was observed in fractions 11 through 14 (Fig. 6E) identical to the elution of CA containing fractions of PICs recovered from cells inoculated with nonpseudotyped HIV-1 particles (Fig. 6D). The levels of CA in these fractions were higher than the CA levels of PICs recovered from cells inoculated with nonpseudotyped HIV-1 (Fig. 6C). These data suggest that the presence of CA in the PIC fractions was not dependent on the viral entry pathway and that the CA was part of the functional PIC complex.

In vitro treatment of PICs with PF74 or CA antibody inhibits integration activity.

Biochemical analysis of PICs established that CA cosedimented with functional PICs (Fig. 6). To further probe that the CA is functionally associated with PIC, we measured integration activity in the presence of PF74. A range of concentrations of PF74 (0 to 0 μM) were added to integration reaction mixtures containing PICs of wild-type and drug-resistant (5Mut) HIV-1 and target DNA. Quantification of viral DNA integration illustrate that PF74 treatment markedly inhibits PIC-associated integration activity (Fig. 7A). The reduction in PIC activity is CA specific since PF74 treatment had no effect on viral DNA integration by the 5Mut PICs (Fig. 7B). These studies suggested that the associated CA is directly linked to the integration activity of PICs. Concurrently, to further probe CA’s role in PIC function, we tested whether a CA-specific antibody would interfere with PIC-associated integration activity (Fig. 7C). The presence of the anti-CA antibody (183-H12-5C) reduced viral DNA integration in vitro, whereas a control antibody against HIV-1 gp120 had little effect (Fig. 7C). Collectively, these integration activity measurements reveal that CA is functionally associated with active HIV-1 PICs recovered from cells.

FIG 7.

Effects of PF74 and CA antibody treatment on PIC-associated integration activity in vitro. (A and B) Integration activity measurement in the presence of PF74. Various concentrations of PF74 (0 to 10 μM) were added to an in vitro integration assay mixture containing wild type Cy-PICs (A) and 5Mut Cy-PICs (B). The integration activity was measured by qPCR, and copies of viral DNA integration were plotted relative to the respective control PICs. (C) Integration activity measurements in the presence of anti-CA antibody. Various amounts of an HIV-1 CA monoclonal antibody or the nonspecific control antibody (anti-gp120 antibody) were added to the in vitro integration assay mixture containing wild-type Cy-PICs, and the integration activities were quantified by qPCR. Viral DNA integration was plotted relative to the respective control PICs. Mean values from three independent experiments are shown, with error bars representing the standard errors of the mean.

PICs assembled in the presence of PF74 contain a reduced quantity of CA.

Our results (Fig. 6 and 7) demonstrate that the enhanced activity of PICs recovered from cells cultured in the presence of PF74 is dependent on the binding of the inhibitor to CA. PF74 has been reported to alter HIV-1 capsid stability (79, 82). Therefore, to test whether PF74 affects the PIC-associated integration activity by altering the levels of associated CA, we fractionated PICs assembled in the presence of 2 μM PF74 on a 10 to 50% linear sucrose gradient. Gradient fractions were analyzed for integration activity, viral DNA, and levels of CA. Integration activity and viral DNA elution profiles were similar to the Cy-PICs of untreated cells (data not shown). Quantification of the CA levels in the fractions showed that CA is associated with active PICs (Fig. 8A). Similar to the fractions in the untreated PICs, the levels of CA decreased markedly, with the lowest levels of CA detected in fraction 9. Interestingly, a peak of CA elution was observed in fractions 10 through 12 in contrast to the peak in fractions 12 through 14 of untreated PICs (Fig. 8A and B). In addition, the amounts of CA eluted in the fractions (11 through 14) were lower compared to the untreated PICs (Fig. 8B). Calculation of a relative ratio of CA levels revealed that PICs assembled with PF74 retained ∼40% and ∼50% smaller amounts of CA in fractions 12 and 13, respectively (Fig. 8C). These results indicate that the level of associated CA was substantially reduced in PICs extracted from cells cultured in the presence of PF74.

FIG 8.

PF74 treatment reduces CA levels in the PICs. Cy-PICs isolated from acutely infected SupT1 cells cultured in the presence of 2 μM PF74 were fractionated through a linear gradient of sucrose as described for Fig. 6. Fractions of 1 ml were collected and analyzed for PIC-associated CA by p24 ELISA. (B) CA levels in fractions 10 through 14. (C) Relative levels of CA in PICs of untreated cells versus PF74-treated cells. Representative data from three independent experiments are presented. The data in panels B and C are plotted as mean values from three independent experiments, with error bars representing the standard errors of the mean. *, P < 0.05 (comparison of untreated PICs to PF74-treated PICs).

DISCUSSION

HIV-1 CA plays critical roles in viral replication and is an emerging target for antiretroviral therapy (4, 13, 15, 16, 83–85). While it is well established that CA coordinates reverse transcription and nuclear entry of HIV-1 (15–17, 34, 43, 61), a functional link between CA and post-nuclear entry steps of infection has not been clearly defined. To span the knowledge gap of CA’s role in integration, we used the CA-specific antiviral compound PF74 as a probe. PF74 binds to HIV-1 CA at an interface that is currently being targeted to develop inhibitors for clinical use (52–54, 56, 60, 78, 79, 86). In accordance with previous studies (53, 54, 56, 60–62), a high concentration (10 μM) of PF74 dramatically reduced reverse transcription, whereas viral DNA synthesis was minimally affected at a low concentration (2 μM) of the inhibitor (Fig. 1B to D). Through the quantification of 2-LTR circles, an assay often used for studying HIV-1 nuclear entry (60, 62), we observed an estimated 2-fold reduction in the number of 2-LTR circles at a low concentration of PF74 (Fig. 1E). However, 2-LTR circles can be generated through autointegration of the viral DNA in the cytoplasm (69) and can confound interpretation of PF74’s specific effect on nuclear entry. Therefore, we also quantified 1-LTR circles that are exclusively formed in the nuclei of infected cells (71, 87, 88) to better probe PF74’s effect on nuclear entry. Reduction in the levels of 1-LTR circles with low concentrations of PF74 (Fig. 1F) confirmed the specific inhibitory effect of the compound on the nuclear entry of the virus. Surprisingly, the observed reduction in nuclear entry by PF74 (Fig. 1E and F) did not quantitatively correlate with infectivity (Fig. 1A), indicating that an additional block after the nuclear entry of the virus contributes to the antiviral activity of the CA inhibitor.

To examine the specific effects of PF74 on the post-nuclear entry steps of HIV-1 infection, we measured proviral integration. PF74 treatment resulted in a marked reduction in HIV-1 integration (Fig. 2A to C). Interestingly, the level of inhibition in proviral DNA integration quantitatively correlated with the inhibition of infectivity at low concentrations of PF74 (Fig. 2F), suggesting that impaired integration could be an important component of the antiviral activity of the CA inhibitor. To address that the dramatic reduction in proviral integration could be a consequence of the 2-fold decrease in nuclear entry (Fig. 1E and F), we used the integrase inhibitor RAL as a probe. RAL treatment increases the number of 2-LTR circles in the nucleus by inhibiting proviral integration (71, 74). Therefore, we predicted that if the inhibition of nuclear entry by PF74 is sufficient for the marked reduction in integration and infectivity, then a combined treatment of PF74 and RAL will not alter the number of LTR circles compared to PF74-treated cells. Surprisingly, the combined treatment of RAL and PF74 resulted in an increase in the number of 2-LTR and 1-LTR circles compared to circles formed with PF74 treatment alone (Fig. 2E and F). Therefore, PF74’s inhibitory effect on proviral integration is distinct from the reduction in nuclear entry and involves a block subsequent to the nuclear entry of the virus. Even though this is the first demonstration of the inhibitory effect of PF74 on HIV-1 integration, a recent study showing that PF74 treatment alters HIV-1 integration targeting into transcriptional units and gene-dense regions of the host genome (57) supports our results on a post-nuclear entry effect of PF74.

Accumulating evidence suggests that HIV-1 nuclear entry and post-nuclear entry steps are facilitated by the engagement of CA-dependent host factors, including CPSF6, Nup153, and TNPO3 (49, 50). Interestingly, PF74 binds to the CA binding pocket that is targeted by these host factors (54, 78). Several reports support the prediction that the competition of binding to CA between PF74 and host factors such as Nup153 is responsible for the nuclear entry defect by the drug at low concentrations (36, 37, 57). Therefore, in infections carried out in the presence of a high concentration of PF74, the drug likely occupies most of the available binding sites on the viral capsid. This would exclude Nup153 from engaging with the capsid, causing premature disassembly of the capsid and a severe defect upon reverse transcription. However, at low concentrations, we predict that PF74 occupies fewer binding sites on capsid, resulting in minimal alterations to the structural integrity. Thus, Nup153 can still engage with the capsid at PF74-unoccupied binding sites. This would enable the nuclear entry of a reduced number of PICs, reflective of the 2-fold decrease in nuclear entry in PF74-treated cells.

Even though PF74’s effect on reverse transcription and nuclear entry has been reported by several groups (55, 57, 79, 82), the inhibitory effects of PF74 on post-nuclear entry steps such as viral DNA integration are unknown. We propose that PF74 inhibits HIV-1 integration by disrupting engagement of CPSF6 with the PIC in the nucleus. Recent studies suggest that HIV-1 integration into gene-dense regions is coordinated by CPSF6 (50). Given that PF74 and CPSF6 share the same binding pocket in the capsid (56, 89), competition between PF74 and CPSF6 to bind the viral replication complex-associated CA in the nucleus would disrupt viral DNA integration. This model also predicts that the absence of Nup153 in the NPC (e.g., the knockdown or knockout of Nup153) will preclude PF74 from imposing its inhibitory effect on proviral integration. For PF74 to manifest its inhibitory effect inside the nucleus, the PICs must enter the nucleus. Accordingly, titration of PF74 and RAL showed that the combined treatment resulted in a greater antiviral activity compared to either of the drugs alone (Fig. 3A to E). The additive antiviral effect was independent of the 2-fold reduction in nuclear entry, providing further evidence that at low concentrations PF74 inhibits the integration activity of PICs that have entered the nucleus.

The antiviral activity of PF74 is dependent on binding of the compound to specific residues in CA (5, 53, 56, 79). Thus, PF74’s inhibitory effect on integration (Fig. 2) implies that CA acts on post-nuclear entry steps of HIV-1 infection. Associations of CA with the viral replication complexes at or near the nuclear pores/envelope (41, 42) and in the nuclear fraction of infected cells (25) also support an intranuclear role for the viral protein. Although CA mutants show altered integration targeting (47, 50), a direct link between CA and post-nuclear entry steps is lacking. To probe CA’s putative role in viral DNA integration, we studied the biochemical properties of HIV-1 PICs. The PIC is a large viral nucleoprotein complex that contains linear vDNA and several cellular and viral proteins, including IN that possesses the enzymatic activity for integration (10). Biochemical analysis of PICs isolated from acutely infected cells has been instrumental in elucidating the mechanistic details of viral DNA integration (23, 70). Given that HIV-1 integration in vivo is carried out in the context of PICs (23, 70), this viral replication complex represents a suitable reagent for studying a direct role of CA in HIV-1 integration. Integration activity measurements of PICs extracted from cells cultured in the presence of low concentrations of PF74 revealed that the PICs in the cytoplasm and the nucleus exhibited enhanced integration activity (Fig. 5). Given that PF74 treatment dramatically reduced proviral DNA integration in infected cells (Fig. 2), these surprising results indicated that the PICs assembled in the presence of PF74 are competent for viral DNA integration. Even though a 2-fold decrease in nuclear entry by PF74 (Fig. 2F) can reduce the number of PICs in the nucleus by half, activity measurements showed that the Nu-PICs from inhibitor-treated cells remain functional and carry out robust integration of the viral DNA (Fig. 5C). Importantly, integration activity of PICs from cells infected with the PF74-resistant mutant (5Mut) were unaffected (Fig. 5H to K), indicating that the effect of PF74 on viral integration is directly linked to CA. These results are also consistent with a recent report showing that the inhibitory effect of the DNA gyrase inhibitor C-A1 on HIV-1 integration is dependent on binding to CA (90).

The increased integration activity exhibited by PICs assembled in the presence of PF74 could be a consequence of a delay in the integration pathway of the replication cycle. This will predict that in the presence of the CA inhibitor a lower proportion of the PICs proceed to integration. Studies of the rescue of RAL inhibition in the absence or presence of PF74 demonstrate that PF74 does not rescue the virus from inhibition by RAL (Fig. 4). Accordingly, time-of-addition assays did not support a delay in the integration pathway by the CA inhibitor (Fig. 4). If PF74 treatment resulted in PIC accumulation, then the number of PICs is expected to be higher than for untreated cells. However, quantification of the PIC-associated viral DNA of both cytoplasmic and nuclear PICs clearly show that 2 μM PF74 does not increase the number of PICs, as determined by the amount of viral DNA (Fig. 5E and G) or the accumulation of late reverse transcription products (Fig. 1B to D). These studies suggest that the increase in PIC activity by PF74 is not due to a temporal effect of the drug on the integration pathway; rather, it is a specific structural effect on the composition of the PIC.

To test whether CA is directly involved in PIC function, we carried out partial purification of PICs through sucrose gradient fractionation. Integration activity measurements concurrent with the presence of viral DNA enabled us to identify functional PICs that carry out integration of the viral DNA (Fig. 6A to C). We also detected CA associated with the functional PICs (Fig. 6D and G), in accordance with several recent reports on the association of CA with the IN or PICs in infected cells (61, 62, 91, 92). Given the suggestion that PF74 treatment results in premature uncoating of the viral capsid (53, 54, 89), we fractionated PICs assembled in the presence of PF74 to understand whether the inhibitor affected PIC-associated CA. Our results revealed that the levels of CA in these PICs were lower than in the PICs from untreated cells (Fig. 8). Most importantly, the low levels of CA in these PICs, along with the higher integration activity (Fig. 5), imply a negative correlation between the PIC-associated CA level and the integration activity. Sequencing of 2-LTR junctions revealed that PF74 does not dramatically alter viral DNA end sequences (Tables 1 and 2). In this context, PICs recovered from PF74-treated cells with lower associated CA levels should be competent to carry out integration even though there was a dramatic reduction in proviral integration in target cells (Fig. 2). Recent studies suggest that the insertion of viral DNA into host genome may depend on specific interaction of CA-dependent host factors such as CPSF6, Nup153, and TNPO3 (49, 50). Interestingly, PF74 binds to the CA-binding pocket that is targeted by these host factors (54, 78). Therefore, the addition of PF74 may prevent the engagement of these host factors with the PIC, resulting in reduced proviral integration in cells. This speculation does not explain why PICs assembled with PF74 efficiently integrate the viral DNA and retain higher activity in vitro (Fig. 5). It is plausible that the insertion of PIC-associated vDNA into a naked target in vitro may not depend on the required macromolecular interactions between the viral replication complex and the host chromatin.

To further probe the relationship between CA in PICs and HIV-1 integration, we quantified PIC-associated integration activity in the presence of PF74 and a CA-specific antibody. We envisioned that if CA is functionally associated with the PIC, then disrupting the associated CA would affect PIC-associated integration activity. The addition of PF74 markedly inhibited viral DNA integration (Fig. 7A), whereas the drug showed no such effect on the activity of 5Mut PICs (Fig. 7B). Similarly, the integration activity was reduced in the presence of a CA-specific monoclonal antibody (Fig. 7C). These results strongly support that CA is biochemically linked to PIC-associated integration activity. Even though our results do not define the mechanism underlying CA’s role in PIC activity, we speculate that HIV-1 integration is dependent on the presence of low levels of PIC-associated CA. These observations also predict the need for removal of excess CA from the wild-type nuclear PICs for successful proviral DNA integration. This model is supported by data showing selective shedding of CA after nuclear entry could be pertinent or essential for PIC function (90). In addition, excess CA in the nuclear PIC may hinder the engagement of host factors such as LEDGF that are necessary for viral DNA integration. Conversely, a late stage of uncoating after nuclear entry of the PIC may facilitate host factors such as CPSF6 for targeting the viral DNA into gene dense regions of the host genome. In summary, our results reveal a novel mechanistic aspect of the antiviral activity of PF74 and provide the first biochemical evidence for a direct role of CA in HIV-1 integration.

MATERIALS AND METHODS

Chemicals, plasmids, and cell culture.

PF74 was synthesized and purified in the Chemical Synthesis Core, Vanderbilt Institute for Chemical Biology (78). Raltegravir (RAL), efavirenz, and azidothymidine (AZT) were obtained from the NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH. Stock solutions (10 mM) were prepared by dissolving PF74, efavirenz, and AZT in dimethyl sulfoxide (DMSO) and RAL in sterile distilled water. The viruses used in this study were generated from the full-length HIV-1 molecular clone R9 (93); its mutant derivative 5Mut containing the codon changes Q67H, K70R, H87P, T107N, and L111I in CA (53, 54); and HIV-1 reporter plasmid HIV-1 GFP that does not express the envelope but encodes the GFP in place of the nef gene (94). The plasmid pHCMV-G encoding VSV-G was used in cotransfection experiments for generating VSV-G-pseudotyped HIV-1 particles (81). Purification of recombinant wild-type and mutant CA proteins was carried out according to the published protocol (95).

HEK293T and SupT1 cell lines were obtained from the American Type Culture Collection (Manassas, VA). The TZM-bl reporter cell line was obtained from John C. Kappes, Xiaoyun Wu, and Tranzyme, Inc., through the NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH. HEK293T and TZM-bl cells were cultured in Dulbecco modified Eagle medium supplemented with 10% heat-inactivated fetal bovine serum (FBS), 2 mM glutamine, 1,000 U/ml penicillin, and 100 mg/ml streptomycin. SupT1 cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 2 mM glutamine, 1,000 U/ml penicillin, and 100 mg/ml streptomycin. Cells were cultured at 37°C with 5% CO₂.

Virus stocks.

High titer virus stocks were generated by calcium phosphate-mediated transient transfection of HEK293T cells with HIV-1 plasmid constructs (96). Briefly, 2 × 10⁶ cells were seeded per 10-cm culture dish and cultured overnight. Next day, cells in each culture dish were transfected, using calcium phosphate, with 20 μg of plasmid DNA. At 12 h posttransfection, the cells were washed once with phosphate-buffered saline (PBS), replenished with 6 ml of growth medium, and cultured further for 48 h. The virus-containing culture supernatants of transfected cells were harvested, cleared of cell debris by low-speed centrifugation, filtered by passage through 0.45-μm-pore-size syringe filters, and treated with DNase I (Calbiochem; 20 μg/ml of supernatant) in the presence of 10 mM magnesium chloride for 1 h at 37°C. The HIV-1 capsid (p24) concentration in the virus stocks was quantified by a p24-specific ELISA according to standard methods (96). Virus infectivity was determined by using TZM-bl indicator cells, as described previously (97).

Single-cycle infection assay with pseudotyped HIV-1 particles.

For the single-cycle HIV-1 infection assay, SupT1 cells seeded in 96-well plates (5 × 10⁴ cells per well) were inoculated with the HIV-1 GFP reporter virus (equivalent to 15 ng of p24 per well) in the presence of a range of PF74 concentrations (0 to 20 μM) and then cultured for 48 h. The cells were then fixed in a 4% final concentration of formaldehyde, and flow cytometry (BD FACSCalibur Platform) was used to quantify the GFP fluorescence (480-nm laser for excitation) as a measure of viral infectivity. The data were analyzed using the FlowJo software (Tree Star, Inc.), and the half-maximal inhibitory concentration (IC₅₀) values were computed by fitting the data to a least-squares model using GraphPad Prism software (GraphPad Software, Inc.).

HIV-1 infection and total DNA isolation from SupT1 cells.

SupT1 cells (4 × 10⁶ cells per well) in 12-well plates were spinoculated at 480 × g with DNase I-treated virus stocks for 2 h at 25°C and then cultured for 16 h at 37°C. The cells were then harvested by transferring the cell cultures to 2-ml microcentrifuge tubes, followed by centrifugation (1,500 × g) for 5 min at 25°C. The cell pellets were washed once with PBS and then resuspended in 200 μl of PBS. Total DNA was isolated using the QIAmp DNA blood minikit (Qiagen) according to the manufacturer’s instructions.

qPCR for measuring the reverse transcription products and 2-LTR and 1-LTR circles.

Quantitative PCR (qPCR) was used to measure the reverse transcription products and 2-LTR and 1-LTR circles. The late reverse transcription products and 1-LTR circles were quantified by using SYBR green-based qPCR, and the 2-LTR circles were quantified by using TaqMan probe-based qPCR. The SYBR green-based qPCR mix contained 1× iTaq Universal SYBR Green Supermix (Bio-Rad), 300 nM concentrations (each) of forward primer (late RT product, 5′-TGTGTGCCCGTCTGTTGTGT-3′; 1-LTR circle, LA1 5′-GCGCTTCAGCAAGCCGAGTCCT-3′) and reverse primer (late RT product, 5′-GAGTCCTGCGTCGAGAGAGC-3′; 1-LTR circle, LA16 5′-GTCACACCTCAGGTACCTTTAAGACCAATGAC-3′), and 100 ng of the total DNA from infected cells. The TaqMan probe-based qPCR mix contained 1× iTaq Universal Probe Supermix (Bio-Rad), 300 nM concentrations (each) of forward primer (5′-AACTAGGGAACCCACTGCTTAAG-3′) and reverse primer (5′-TCCACAGATCAAGGATATCTTGTC-3′), 100 nM the TaqMan probe (5′-[FAM] ACACTACTTGAAGCACTCAAGGCAAGCTTT-[TAMRA]-3′), and 100 ng of total DNA from infected cells. The qPCR cycling conditions for quantifying late RT products included an initial incubation at 95°C for 3 min, followed by 39 cycles of amplification and acquisition at 94°C for 15 s, 58°C for 30 s, and 72°C for 30 s. The qPCR cycling conditions for quantifying the 1-LTR circles included an initial incubation at 95°C for 3 min, followed by 39 cycles of amplification and acquisition at 94°C for 15 s, 58°C for 30 s, and 72°C for 1 min. For the SYBR green-based qPCR, the thermal profile for melting-curve analysis was obtained by holding the sample at 65°C for 31 s, followed by a linear ramp in temperature from 65 to 95°C with a ramp rate of 0.5°C/s and acquisition at 0.5°C intervals. The TaqMan probe-based qPCR included an initial incubation at 95°C for 3 min, followed by 39 cycles of amplification and acquisition at 94°C for 15 s, 58°C for 30 s, and 72°C for 30 s. During qPCR of the samples, a standard curve was generated in parallel and under same conditions using 10-fold serial dilutions of known copy numbers (10⁰ to 10⁸) of the HIV-1 molecular clone (for late RT products) (93) or the p2LTR plasmid (71) containing the 2-LTR junction sequence (for 2-LTR circles) or the p1LTR plasmid (for 1-LTR circles) that was constructed in-house. Briefly, an amplicon spanning the full-length LTR (634 nucleotides) flanked by Gag-specific (80 nucleotides) and Nef-specific (82 nucleotides) sequences was obtained by PCR using the total DNA of HIV-1-infected SupT1 cells as the template and the primer pair LA1 and LA16. The amplicon was then cloned into the pMiniT2 plasmid (NEB), and the resulting plasmid (p1LTR) was verified by DNA sequencing. CFX Manager software (Bio-Rad) was used to analyze the data and determine the copy numbers of the late RT products and the LTR circles by plotting the data against the respective standard curves.

Nested qPCR for measuring HIV-1 proviral integration in SupT1 cells.

To measure integration of HIV-1 proviral DNA in infected cells, a nested qPCR method (73), involving a first round of standard PCR with primers designed to amplify only the junctions of the chromosomal-integrated viral DNA but not any unintegrated viral DNA (73, 77), followed by a second round of qPCR with primers that amplify viral LTR-specific sequences present in the first round PCR amplicons, was used with some modifications. Briefly, the first-round PCR was performed in a final volume of 50 μl containing 100 ng of total DNA from infected SupT1 cells, 1× GoTaq reaction buffer (Promega), deoxynucleoside triphosphate (dNTP) nucleotide mix containing 200 μM concentrations of each nucleotide (Promega), 500 nM concentrations (each) of primers that target the host chromosomal Alu repeat sequence (5′-GCCTCCCAAAGTGCTGGGATTACAG-3) and HIV-1 Gag sequence (5′-GTTCCTGCTATGTCACTTCC-3′), and 1.25 U of GoTaq DNA polymerase (Promega). The thermocycling conditions included an initial incubation at 95°C for 5 min; followed by 23 cycles of amplification at 94°C for 30 s, 50°C for 30 s, 72°C for 4 min; and a final incubation at 72°C for 10 min. The second-round qPCR mix contained a 1/10 volume of the first-round PCR products as the template DNA, 1× iTaq Universal Probe Supermix (Bio-Rad), 300 nM concentrations (each) of the viral LTR-specific primers that target the R region (5′-TCTGGCTAACTAGGGAACCCA-3′) and the U5 region (5′-CTGACTAAAAGGGTCTGAGG-3′), and a 100 nM concentration of TaqMan probe (5′-[6-FAM]-TTAAGCCTCAATAAAGCTTGCCTTGAGTGC-[TAMRA]-3′). The qPCR cycling conditions included an initial incubation at 95°C for 3 min, followed by 39 cycles of amplification and acquisition at 94°C for 15 s and 58°C for 30 s, and a final incubation at 72°C for 30 s. During qPCR of the samples, a standard curve was generated in parallel and under same conditions using 10-fold serial dilutions of known copy numbers (10⁰ to 10⁸) of the HIV-1 molecular clone plasmid. Data were analyzed using CFX Manager software (Bio-Rad), and integrated viral DNA copy numbers were determined by plotting the qPCR data against the standard curve.

Measurement of HIV-1 infectivity through titration of PF74 and RAL.

For titration studies, we used TZM-bl cell based luciferase reporter assay (97). A total of 4 × 10⁴ TZM-bl cells were plated into each well of a 24-well plate and cultured for 16 to 24 h. After the cells were washed once with fresh culture medium, they were layered with 100 μl of culture medium or viral stocks prepared in culture medium containing Polybrene (6 μg/ml) and DMSO or various concentrations of PF74 and/or RAL. After culture for 2 h, the culture medium was removed, the cells were washed once with 1× PBS, 1 ml of fresh cell culture medium supplemented with vehicle or appropriate drug was added to each well, and the cells were cultured further. At 48 h postinfection, the cell culture medium was removed, the cells were washed once with 1× PBS, and the cells in each well were lysed in 200 μl of 1× Glo lysis buffer (Promega). The firefly luciferase activity was measured in triplicates of 50-μl aliquots of each cell lysate with firefly assay reagent (Promega), and the values were averaged. For titration studies, a range of RAL (0 to 1 μM) and PF74 (0 to 2 μM) concentrations were tested for their inhibitory effect on HIV-1 infection. Then, selected concentrations of PF74 and RAL were used to measure the effects of combined treatment of the inhibitors on HIV-1 infection.

Time-of-addition infectivity assay.

To probe post-nuclear entry effects of PF74, we carried out rescue-of-inhibition studies by employing TZM-bl cell-based luciferase reporter assay (97). As described in the previous section, TZM-bl cells were inoculated with viral stocks and then cultured in the presence or absence of PF74 (2 μM) and/or RAL (25 nM). In this assay, one drug was added at various time points (0, 1, 2, 5, 8, 12, and 16 h) relative to the time of infection in the absence or presence of the other inhibitor. At 48 h postinfection, the luciferase activity was measured in the cellular lysates as described above.

Sequence analysis of 2-LTR circle junctions.

SupT1 cells (4 × 10⁶) were inoculated with HIV-1 in the presence of DMSO or 2 μM PF74, and total DNA was isolated from the infected cells 24 h postinfection, as described in “HIV-1 infection and total DNA isolation from SupT1 cells.” To obtain a sufficient amount of 2-LTR circle junction DNA for cloning, a nested PCR strategy involving two different sets of primers was used. Briefly, the first-round PCR was performed in a final volume of 50 μl containing 500 ng of total DNA from infected SupT1 cells, 5× Phusion HF buffer (NEB), a dNTP nucleotide mix containing 200 μM concentrations of each nucleotide (Promega), 500 nM concentrations (each) of the primers 2-LTR Forward-M (5′-AGCCTGGGAGCTCTCTGGCTAAC-3′) and 2-LTR Reverse-M (5′-AGCCTTGTGTGTGGTAGATCCAC-3′), and 1 U of Phusion Hot Start Flex polymerase (NEB). The thermocycling conditions included an initial incubation at 98°C for 2 min; followed by 30 cycles of amplification at 98°C for 10 s, 71°C for 1 min, and 72°C for 15 s; and a final incubation at 72°C for 10 min. The second-round PCR mix contained a 1/10 volume of the first-round PCR products as the template DNA, 5× Colorless GoTaq reaction buffer, 500 nM concentrations (each) of the primers MH535 (5′-AACTAGGGAACCCACTGCTTAAG-3′) and 2-LTR Reverse (5′-TCCACAGATCAAGGATATCTTGTC-3′), and 1.25 U of GoTaq DNA polymerase (Promega). The thermocycling conditions included an initial incubation at 95°C for 5 min; followed by 25 cycles of amplification at 95°C for 15 s, 58°C for 1 min, and 72°C for 30 s; and a final incubation at 72°C for 10 min. The second-round PCR product was resolved on 2% agarose gel at 80 V, and the DNA band corresponding to the 2-LTR junction amplicon was excised from the gel and purified using a Zymoclean gel DNA recovery kit according to the manufacturer-recommended protocol (Zymo Research). A portion (10 ng) of the purified DNA was ligated to 50 ng of pGEMT-Easy vector according to the manufacturer-recommended protocol (Promega), and DH5α competent cells were transformed with the ligation reaction products. Plasmid DNA was isolated from white bacterial colonies selected on X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside)/IPTG (isopropyl-β-d-thiogalactopyranoside) indicator plates and screened for the presence of 2-LTR junction DNA insert by diagnostic restriction digestion analysis. The sequences of the 2-LTR junction inserts in the recombinant clones were determined by Sanger DNA sequencing using primers flanking the cloning site. The sequences of the viral DNA ends were analyzed by multiple alignment.

Isolation of PIC from HIV-1-infected SupT1 cells.

HIV-1 PICs were isolated from HIV-1-infected T cells using a published method (72). Briefly, SupT1 cells (8 × 10⁷) were spinoculated (480 × g) with DNase I-treated virions (wild type or VSV-G pseudotyped) for 2 h at 25°C, and the cells were then cultured for 5 h at 37°C. The cells were then harvested by centrifugation (300 × g) for 10 min at 25°C. The cell pellet was washed twice with 2 ml of buffer K^−/− (20 mM HEPES [pH 7.6], 150 mM potassium chloride, 5 mM magnesium chloride) at 25°C. Subsequently, the cell pellet was thoroughly resuspended in 2 ml of ice-cold buffer K^+/+ (20 mM HEPES pH 7.6, 150 mM potassium chloride, 5 mM magnesium chloride, 1 mM dithiothreitol [DTT], 20 μg/ml aprotinin, 0.025% [wt/vol] digitonin). The cell suspension was transferred to a 2-ml microcentrifuge tube and rocked on a rocking platform shaker (60 to 80 rocking motions/min) for 10 min at room temperature. The cell lysate was centrifuged (1,500 × g) for 4 min at 4°C to separate the cytoplasmic fraction from the nuclear fraction. The supernatant (cytoplasmic fraction) was transferred to a new 2-ml microcentrifuge tube, and the pellet (nuclear fraction) was saved for the isolation of nuclear PICs. The cytoplasmic fraction was centrifuged (16,000 × g) for 1 min at 4°C, and the resulting supernatant was transferred to a new microcentrifuge tube. Subsequently, RNase A (Invitrogen) was added to a final concentration of 20 μg/ml, and the contents were incubated for 30 min at room temperature. Finally, sucrose (60% [wt/vol]) was added to a final concentration of 7%, and the contents were mixed well. Aliquots of the cytoplasmic fraction were used as the source of cytoplasmic PICs in assays or were flash-frozen in liquid nitrogen and placed in a –80°C freezer for long-term storage.

To recover PICs from cell nuclei, the nuclear pellet was resuspended in 2 ml of ice-cold buffer K^+/− (20 mM HEPES [pH 7.6], 150 mM potassium chloride, 5 mM magnesium chloride, 1 mM DTT, 20 μg/ml aprotinin, 20 μg/ml RNase A), transferred to 5-ml polypropylene round-bottom tube, and placed on ice. The nuclei were lysed by using a mounted homogenizer (VWR 200) at 10 cycles of 30 s of homogenization interrupted by 30 s of cooling on ice. The lysate was transferred to a new 2-ml microcentrifuge tube and centrifuged (16,000 × g) for 3 min at 4°C. The supernatant was then transferred to a new microcentrifuge tube, and aliquots of this nuclear fraction were used as the source of nuclear PICs in assays or were flash-frozen in liquid nitrogen and placed in a –80°C freezer for long-term storage.

Assay for PIC-associated integration activity in vitro.

In vitro integration assays were performed using a modified version of published protocol (72). The target DNA used was a PCR-amplified 2-kb region of the ΦX174 genome. The PCR was performed in a final volume of 50 μl containing 50 ng of the ΦX174 RF plasmid (Promega or NEB), 500 nM concentrations (each) of the forward primer (5′-CGCTTCCATGACGCAGAAGTT-3′) and the reverse primer (5′-CACTGACCCTCAGCAATCTTA-3′), 1× GoTaq reaction buffer (Promega), dNTP nucleotide mix containing 200 μM concentrations of each nucleotide (Promega), and 1.25 U of GoTaq DNA Polymerase (Promega) under the following thermocycling conditions: initial incubation at 95°C for 2 min; followed by 34 cycles at 95°C for 30 s, 53°C for 30 s, and 72°C for 2 min; and a final incubation at 72°C for 10 min. PCR products were resolved by standard agarose gel electrophoresis, and the ΦX174-specific PCR amplicon was gel purified by using a QIAquick gel extraction kit (Qiagen). The in vitro integration reaction was carried out by mixing 200 μl of PICs and 600 ng of target DNA and then incubating the mixture at 37°C for 45 min. The integration reaction was stopped and deproteinized by adding SDS, EDTA, and proteinase K to final concentrations of 0.5%, 8 mM, and 0.5 mg/ml, respectively, followed by incubation overnight at 56°C. The deproteinized sample was mixed with an equal volume of phenol (equilibrated with 10 mM Tris HCl [pH 8]), mixed thoroughly by vortexing, and centrifuged (16,000 × g) for 1 to 2 min at room temperature. The aqueous phase is extracted once with an equal volume of phenol-chloroform (1:1) mixture, followed by an equal volume of chloroform. The DNA was precipitated by adding 2.5 volumes of 100% ice-cold ethanol in the presence of sodium acetate (0.3 M, final concentration) and the coprecipitant glycogen (25 to 100 μg, final concentration), followed by incubation overnight at –80°C. The sample was centrifuged (16,000 × g) for 30 min at 4°C, and the resultant DNA pellet was washed once with 80% ethanol using centrifugation (16,000 × g) for 10 min at 4°C. The precipitated DNA was air dried at room temperature, resuspended in 50 μl of nuclease-free water, and used as the template DNA in the nested qPCR. A first-round standard PCR, designed to amplify only the integrated virus-target DNA junctions, was carried out in a final volume of 50 μl containing 5 μl of purified DNA product from the integration reaction, 500 nM concentrations (each) of primers targeting the target DNA (5′-CACTGACCCTCAGCAATCTTA-3′) and the viral LTR (5′-GTGCGCGCTTCAGCAAG-3′), 1× GoTaq reaction buffer (Promega), dNTP nucleotide mix containing 200 μM concentrations of each nucleotide (Promega), and 1.25 U of GoTaq DNA polymerase (Promega) under the following thermocycling conditions: initial incubation at 95°C for 5 min; followed by 23 cycles at 94°C for 30 s, 55°C for 30 s, and 72 °C for 2 min; and a final incubation at 72°C for 10 min. The second-round qPCR designed to amplify only the viral LTR-specific region contained a 1/10 volume of the first-round PCR products as the template DNA, 1× iTaq Universal Probe Supermix (Bio-Rad), 300 nM concentrations (each) of the viral LTR-specific primers that target the R region (5′-TCTGGCTAACTAGGGAACCCA-3′) and the U5 region (5′-CTGACTAAAAGGGTCTGAGG-3′), and a 100 nM concentration of TaqMan probe (5′-[6-FAM]-TTAAGCCTCAATAAAGCTTGCCTTGAGTGC-[TAMRA]-3′). The qPCR run included an initial incubation at 95°C for 3 min, followed by 39 cycles of amplification and acquisition at 94°C for 15 s and 58°C for 30 s, and a final incubation at 72°C for 30 s. During qPCR of the samples, a standard curve was generated in parallel and under the same conditions using 10-fold serial dilutions of known copy numbers (10⁰ to 10⁸) of the HIV-1 molecular clone plasmid. Data were analyzed using CFX Manager software (Bio-Rad), and integrated viral DNA copy numbers were determined by plotting the qPCR data against the standard curve. To determine the integration efficiency (i.e., ratio of chromosome-integrated viral DNA copy numbers to corresponding PIC-associated viral DNA copy numbers) of the in vitro integration reactions, the PIC-associated viral DNA copy numbers were determined. Briefly, using methodologies described above, 0.2 ml of the corresponding PICs was deproteinized, extracted with phenol-chloroform, and subjected to ethanol precipitation, and the resulting purified DNA was directly used in qPCR designed to amplify the viral LTR-specific regions. Data were analyzed uisng CFX Manager software (Bio-Rad), and the viral DNA copy numbers were determined by plotting the qPCR data against the standard curve generated, as described above.

Velocity gradient fractionation of PICs.

The PICs were partially purified by velocity gradient ultracentrifugation through 10 to 50% continuous sucrose gradient. Fresh 10 and 50% sucrose solutions were prepared in buffer K (20 mM HEPES [pH 7.4], 5 mM magnesium chloride, 150 mM potassium chloride, 1 mM DTT, and 20 μg/ml aprotinin) and filter sterilized using 0.22-μm-pore-size syringe filters. Equal volumes (7.1 ml) of the 50 and 10% sucrose solutions were added to the mixing and reservoir chambers, respectively, of the gradient maker connected to an Auto Densi-Flow density gradient fractionator (Labconco). According to the manufacturer’s instructions, a 10 to 50% continuous sucrose gradient was poured in a 15-ml round-bottom polyallomer ultracentrifuge tube (Beckman Coulter), followed by incubation on ice for 1 h. The PICs (1 ml) were then gently overlaid on top of the gradient, without disturbing the layers of sucrose solution, and centrifuged (137,000 × g) in an SW 32.1 Ti Beckman rotor at 4°C for 2 h. Fractions (1 ml each) were collected in microcentrifuge tubes, starting from the top of the gradient using the gradient fractionator in accordance with the manufacturer’s instructions. The gradient fractions in the tubes were mixed thoroughly by inverting the tubes (10 times). Aliquots of each fraction were used in assays or were flash frozen in liquid nitrogen and stored at –80°C.

Quantification of CA by p24 ELISA.

The p24 ELISA protocol was adapted from published reports (16). Unless indicated otherwise, antibodies were obtained from the NIH AIDS Reagent Program. Briefly, anti-p24 antibody (183-H12-5C; 4 μg/ml in PBS) was added to the wells, followed by incubation overnight. Blocking solution (5% dialyzed calf serum in PBS) was then added at 250 μl/well, followed by incubation for 1 h. Standards or samples diluted in ELISA sample diluent (10% dialyzed calf serum and 0.5% Triton X-100 in Dulbecco's PBS) were then added, followed by incubation for 2 h. After incubation for 1 h with HIV IgG diluted in ELISA sample diluent, horseradish peroxidase (HRP)-conjugated ImmnunoPure goat anti-human IgG(H+L) cross-adsorbed secondary antibody (Pierce) diluted in ELISA sample diluent was added, followed by incubation for 1 h. Finally, substrate solution (TMB microwell peroxidase substrate system; KPL, Inc.) was added, followed by incubation at room temperature until color development. The reaction was terminated by adding equal volume of 4 N H₂SO₄ solution, and the absorbance of each well was read at 450 nm with a 650-nm reference in an ELISA microplate reader. The standard curve was generated by plotting absorbance against p24 concentration of the standards, and the unknown concentrations were calculated by interpolation.

PIC-associated integration activity measurements in the presence of PF74 and anti-CA antibody.

To probe the effects of PF74 on the integration activity of PICs in vitro, cytoplasmic PICs were extracted from SupT1 cells inoculated with wild-type or drug-resistant (5Mut) virions according to the method described in “Isolation of PIC from HIV-1-infected SupT1 cells.” A range of PF74 concentrations (0 to 10 μM) were added to the integration reaction mixture containing cytoplasmic PICs (200 μl). In parallel experiments, anti-CA antibody (183-H12-5C; dissolved in PBS) and a nonspecific control antibody (anti-gp120) were added to integration reaction mixture containing wild-type cytoplasmic PICs (200 μl). Integration activity measurements and quantification of viral DNA integration was carried out as per the protocol described in “Assay for PIC-associated integration activity in vitro.”

Pulldown of PIC-associated CA by immunoprecipitation.

To detect PIC-associated CA, we carried out pulldown assays of PIC preparation using GST-CypA fusion protein and anti-CA antibody. As a positive control, CA tubes were used in the pulldown assays. CA-A14C/E45C tubes were assembled by mixing 160 μM CA-A14C/E45C protein with equal volumes of 2× reaction buffer (100 mM Tris-HCl [pH 8.0], 2 M NaCl), followed by incubation at 4°C overnight. For the GST-CypA pulldown assays, glutathione-agarose beads (20-μl suspension equivalent; Pierce) were first equilibrated with binding/wash buffer (50 mM Tris [pH 8], 150 mM NaCl, 0.1 mg/ml BSA). GST-CypA (1.5 μg) was then immobilized onto the glutathione-agarose beads in binding/wash buffer in a final volume of 100 μl at room temperature for 1 h on a rotary shaker. The beads were then washed three times in 500 μl of binding/wash buffer at room temperature with incubation times of 15, 10, and 5 min, respectively. The GST-CypA-coated beads or equivalent amounts of glutathione-beads (negative control) were incubated with cPICs, CA-A14C/E45C tubes, or monomeric CA protein in binding/wash buffer in a final volume of 100 μl at room temperature for 1 h on rotary shaker. The beads were washed three times in 500 μl of binding/wash buffer at room temperature, with incubation times of 15, 10, and 5 min, respectively. The beads were then incubated in 15 μl of β-mercaptoethanol-supplemented 2× Laemmli buffer at 95°C, and the proteins were resolved using 4 to 12% SDS-PAGE and then electrophoretically transferred to a nitrocellulose membrane. Western blotting was performed with mouse anti-CA monoclonal antibody (1:1,000 dilution, 183-H12-5C; NIH AIDS Reagent Program) and secondary HRP-conjugated goat anti-mouse IgG(H+L) (1:5,000 dilution; Bio-Rad), and the immunocomplexes were detected by enhanced chemiluminescence (Pierce).

The coimmunoprecipitation assays with anti-CA monoclonal antibody were performed using the protein A/G-magnetic beads (25 μl of suspension equivalent; Pierce) and a magnetic separator (DYNA-L magnetic separator; Invitrogen). The protein A/G-magnetic beads were first equilibrated with binding/wash buffer (25 mM Tris [pH 7.5], 150 mM NaCl, 0.05% Tween 20). The mouse anti-CA monoclonal antibody (5 μg, 183-H12-5C) was then immobilized onto the protein A/G-magnetic beads in binding/wash buffer in a final volume of 500 μl at 4°C overnight on a rotary shaker. The beads were washed three times in 500 μl of binding buffer at room temperature with incubation times of 15, 10, and 5 min, respectively. The antibody-coated beads or equivalent amount of protein A/G-magnetic beads (negative control) were incubated with cPICs, CA-A14C/E45C tubes, or monomeric CA protein in binding/wash buffer in a final volume of 500 μl at room temperature for 1 h on rotary shaker. The beads were washed three times in 500 μl of binding/wash buffer at room temperature, with incubation times of 15, 10, and 5 min, respectively. The beads were then washed once in 500 μl of ultrapure water and magnetically separated, followed by resuspension with 30 μl of ultrapure water. For immunochemical analysis, an aliquot of the eluate was mixed with an equal volume of β-mercaptoethanol-supplemented 2× Laemmli buffer, followed by incubation at 95°C, and the proteins were resolved by 4 to 12% SDS-PAGE and then electrophoretically transferred to a nitrocellulose membrane. Western blotting was performed with mouse anti-CA monoclonal antibody (1:1,000 dilution, 183-H12-5C; NIH AIDS Reagent Program) and secondary TrueBlot Ultra HRP-conjugated rat anti-mouse IgG (1:1,000 dilution; Rockland Immunochemicals, Inc.). The immunocomplexes were detected by enhanced chemiluminescence (Pierce).

Statistical analysis.

Data were expressed as means ± the standard errors of the mean obtained from three independent experiments. Significant differences between control and treated samples were determined by a Student t test. P values of <0.05 were considered statistically significant.