|Home | About | Journals | Submit | Contact Us | Français|
Nonnucleoside reverse transcriptase inhibitors (NNRTIs) are potent and commonly prescribed antiviral agents used in combination therapy (CART) of human immunodeficiency virus type 1 (HIV-1) infection. The development of drug resistance is a major limitation of CART. Reverse transcriptase (RT) genotypes with the NNRTI resistance mutations K101E+G190S are highly resistant to efavirenz (EFV) and can develop during failure of EFV-containing regimens in patients. We have previously shown that virus with K101E+G190S mutations can replicate more efficiently in the presence of EFV than in its absence. In this study, we evaluated the underlying mechanism for drug-dependent stimulation, using a single-cycle cell culture assay in which EFV was added either during the infection or the virus production step. We determined that EFV stimulates K101E+G190S virus during early infection and does not affect late steps of virus replication, such as increasing the amount of active RT incorporated into virions. Additionally, we showed that another NNRTI, nevirapine (NVP), stimulated K101E+G190S virus replication during the early steps of infection similar to EFV, but that the newest NNRTI, etravirine (ETR), did not. We also showed that EFV stimulates K101E+Y188L and K101E+V106I virus, but not K101E+L100I, K101E+K103N, K101E+Y181C, or K101E+G190A virus, suggesting that the stimulation is mutation specific. Real-time PCR of reverse transcription intermediates showed that although the drug did not stimulate minus-strand transfer, it did stimulate minus-strand strong-stop DNA synthesis. Our results indicate that stimulation most likely occurs through a mechanism whereby NNRTIs stimulate priming or elongation of the tRNA.
The reverse transcriptase (RT) of human immunodeficiency virus type 1 (HIV-1) converts viral genomic RNA into double-stranded DNA through the process of reverse transcription (reviewed in reference 42). The enzyme is crucial for viral replication and has been an important target of antiretroviral therapy since 1987 (15). Efavirenz (EFV) is a nonnucleoside RT inhibitor (NNRTI) that is commonly used in combination with two nucleoside analog RT inhibitors (NRTIs) for the treatment of antiretroviral-agent-naïve patients (36a). NNRTIs inhibit virus replication by preventing DNA polymerization by RT, but EFV has a low barrier to resistance, because a single mutation can cause high-level drug resistance (1, 2). An interesting observation first published by our research group was that viral variants having the NNRTI resistance mutation combination K101E+G190S replicated more efficiently in the presence of EFV than in its absence, suggesting that some variants selected by drug treatment were not only resistant but were actually stimulated in the presence of drug (39). The stimulation observed for the double mutant, K101E+G190S, was seen despite the fact that the single K101E and G190S mutants did not show any stimulation. The K101E+G190S double mutant is seen in 3 to 4% of patients failing EFV and is highly resistant to nevirapine (NVP) and EFV (29, 30, 39). However, the mechanism by which stimulation occurs is as yet unknown.
Others have shown that NNRTIs can affect HIV-1 replication through mechanisms other than DNA replication. RT is a heterodimer enzyme, comprised of the two subunits p51 and p66, and has both polymerization and RNase H activities (reviewed in reference 18). Processing of the Gag-Pol polypeptide by HIV-1 protease results in heterodimer formation of RT during the maturation step of the life cycle and is important for producing infectious HIV-1 virus (33, 40). Studies have shown that NNRTIs, such as EFV, could chemically enhance the binding of p51 with p66 to form the heterodimer in vitro (34). Other studies have shown that EFV also increases Gag-Pol processing in the virion (35). In addition, it was reported that EFV increases intracellular Gag-Pol processing and decreases viral particle release from transfected cells (14). All these studies show that EFV can influence the late stages of the HIV-1 life cycle. In contrast, EFV has also been shown to increase the RNase H activity of the enzyme (24, 27) during early stages of the HIV-1 life cycle.
We previously studied the stimulation of replication by EFV using a multiple-cycle assay, so we were not able to identify the step of the virus life cycle in which stimulation occurred. The drug could stimulate reverse transcription by stimulating RNase H activity, an early step of infection, and/or EFV could stimulate RT maturation and particle production, a late step of infection. In the current study, we achieved the following goals to better elucidate the mechanism of stimulation. (i) We identified other combinations of NNRTI resistance mutations besides K101E+G190S that were stimulated by NNRTIs. (ii) We tested whether stimulation of infection occurred early or late in infection by using a single-cycle assay in which EFV was added either during infection or during virus production. (iii) We determined whether other NNRTIs besides EFV could stimulate K101E+G190S. (iv) Real-time PCR of early reverse transcription intermediates was performed to identify whether early steps of reverse transcription are stimulated by NNRTIs.
The 293 human primary embryonal kidney cell line (American Type Culture Collection) and the 293T cell line, a derivative of 293 cells, that contains the simian virus 40 (SV40) T antigen (American Type Culture Collection) were used to make virus by transient transfection. The following reagents were obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: the cell line, TZM-bl, from John C. Kappes, Xiaoyun Wu, and Tranzyme, Inc. (9, 25, 26, 36, 41); the cell line, PM1, from Marvin Reitz (21); the nonnucleoside reverse transcriptase inhibitors (NNRTIs), efavirenz (EFV), nevirapine (NVP), and etravirine (ETR); and the plasmid pNL4-3-deltaE-EGFP (EGFP stands for enhanced green fluorescent protein) (45). TZM-bl cells are a derivative of the HeLa cell line and were engineered to express CD4 and CCR5 (chemokine [C-C motif] receptor 5). They also contain the firefly luciferase (Luc) and Escherichia coli β-galactosidase genes under the control of the HIV-1 long terminal repeat. The 293, 293T, and TZM-bl cell lines were grown in Dulbecco modified Eagle medium (Cellgro) supplemented with 10% fetal bovine serum (FBS) (Valley Biomedical), penicillin (100 U/ml), and streptomycin (100 U/ml). The PM1 cell line was grown in RPMI 1640 medium (Cellgro) supplemented with 10% FBS (Valley Biomedical), penicillin (100 U/ml), and streptomycin (100 U/ml). EFV, NVP, and ETR were reconstituted in dimethyl sulfoxide (DMSO) to achieve final concentrations of 5 mM, 75 mM, and 4.6 mM, respectively, and stored at −20°C.
The following antibodies were obtained from the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: the HIV-1 RT monoclonal antibody (8C4) from Dag E. Helland, the HIV-1 RT monoclonal antibody (5B2B2) from D. Helland and A. M. Szilvay (32), HIV-1 p24 hybridoma (183-H12-5C) from Bruce Chesebro (8), and the antiserum to HIV-1 integrase (recognizing epitopes mapping to amino acids 1 to 16) from Duane P. Grandgenett (19). Fluorescein isothiocyanate (FITC)-conjugated mouse anti-rat Thy1.1 (HIS51) and R-phycoerythrin (PE)-conjugated rat anti-mouse Thy1.2 (30-H12) monoclonal antibodies were obtained from BD Biosciences.
Mutations in reverse transcriptase (RT) were introduced into pRHAXX (11) using the QuikChange II mutagenesis kit (Stratagene). pRHAXXY181C and pRHAXXG190A (37) were mutagenized with the K101E primer (5′-GCAGGGTTAGAAAAGAAAAAATCAG) (only the sense primers are shown) to create pRHAXXK101E+Y181C and pRHAXXK101E+G190A. pRHAXXV106I, pRHAXXY188L, and pRHAXXG190E were made by mutagenizing pRHAXXWT with the following primers: V106I (5′-CAGAAAAAATCAATAACAGTACTGGATG), Y188L (5′-GGATGATTTGCTTGTAGGATCTGACTTAGAAATAGG), and G190E (5′-GGATGATTTGTATGTAGAATCTGACTTAGAAATAGG). The K101E primer was used to mutagenize pRHAXXV106I, pRHAXXY188L, and pRHAXXG190E to create pRHAXXK101E+V106I, pRHAXXK101E+Y188L, and pRHAXXK101E+G190E. pRHAXXWT was mutagenized with the following primers to create pRHAXXK101E+L100I and -K101E+K103N: K101E+L100I (5′-CCACATCCTGCAGGGATAGAACAGAAAAAATC) and K101E+K103N (5′-CAGGGTTAGAACAGAACAAATCAGTAACAGTAC). The antisense primers used for mutagenesis were the reverse and complement of each sense primer. The resultant clones were sequenced to confirm the presence of the mutation and the absence of spurious mutations. The RT gene containing the K101E mutation in combination with either the L100I, K103N, V106I, Y181C, Y188L, or G190A mutation was subcloned from pRHAXX to pAT2 (12) using the XmaI and XbaI restriction sites. All mutant pAT2 clones were sequenced to verify the presence of the mutation and the integrity of the cloning sites.
pNL4-3XXK101E+G190S (containing the K101E+G190S [ES] mutations), pNL4-3XXL74V+K101E+G190S L74V+K101E+G190S (containing the L74V+K101E+G190S [VES] mutations), pNL4-3XXD10 and pNL4-3XXD10MMTT were constructed as previously described (39). pDAT1 is an env-deleted version of pNL4-3 and has the mouse Thy1.1 gene in place of nef (13). The RT fragments of pNL4-3XXES, pNL4-3XXVES, pNL4-3XXD10, and pNL4-3XXD10MMTT were isolated by digestion by restriction enzymes XmaI and XbaI and were subcloned into the pDAT2 vector. All mutant pDAT2 clones were sequenced to verify the presence of the mutation and the integrity of the cloning sites.
The wild-type (WT) and mutant pAT2 and pNL4-3XX constructs were transiently transfected into either 293 cells using 40 μg of plasmid DNA and Superfect (Qiagen) or 293T cells using 5 μg of DNA and 25 μl of Lipofectamine LTX (Invitrogen) in the absence of drug. Mutant pDAT2 constructs were transiently transfected into 293 or 293T cells using 5 μg of plasmid DNA along with 5 μg of pSV-A-MLV-ENV (expressing the murine leukemia virus [MLV] envelope [ENV]) and 25 μl of Lipofectamine LTX (Invitrogen) as previously described (13). For all transfections, the cells were incubated for 4 h, and the transfection medium was replaced with fresh medium. For the DAT transfections, the medium was replaced without drug or with 400 or 800 nM EFV. Clarified supernatants were harvested after 72 h and stored at −80°C. HIV-1 capsid protein (p24) was quantified for each virus stock using an enzyme-linked immunosorbent assay (ELISA) kit (Perkin Elmer).
The WT, G190S, K101E+G190S (ES), and L74V+K101E+G190S (VES) RT p51 subunit sequences were subcloned into pET-28a(+) expression vector, and the p66 subunits were cloned into pET-21a(+) vectors (Novagen) (37). The WT and mutant His-tagged p51 (His-p51) subunits and untagged p66 subunits were expressed in E. coli. The detection of RT heterodimer in the presence of EFV was performed as previously described (33). Briefly, after induction by the Overnight Express autoinduction system (Novagen), cells were lysed in 1 ml of lysis buffer (50 mM sodium phosphate buffer [pH 7.8], 500 mM NaCl, 0.5% NP-40, 5 mM dithiothreitol [DTT]) containing 1 mg/ml of lysozyme, 500 U Benzonase nuclease (Novagen), and 1× protease inhibitor cocktail (Roche). Lysates of His-p51 and p66 were combined and incubated for 16 h at 4°C in the presence or absence of 0.1, 1, 10, 100, and 1,000 μM EFV. The heterodimer was captured on Talon Ni2+ metal affinity resin beads (BD Biosciences), and unbound subunits were removed by washing with lysis buffer containing 10 nM imidazole (BD Pharmingen). Heterodimer bound to beads was resolved by SDS-PAGE and detected with a mixture of RT monoclonal antibodies 8C4 and 5B2B2.
WT or mutant virus stocks were made in the presence or absence of 400 or 800 nM EFV. Equal amounts of virus as determined by capsid protein quantitation were pelleted by centrifugation at 150,000 × g for 1 h at 4°C. Detection of virion-associated RT content was performed as previously described (37). Briefly, virus pellets were resuspended in 15 μl of NuPAGE 2× sample buffer (Invitrogen), and viral proteins were separated by 4 to 12% Bis-Tris NuPAGE electrophoresis according to the manufacturer's instructions. Anti-RT 8C4, anti-IN, and anti-p24 antibodies were used to detect RT, integrase (IN), and p24 capsid proteins, respectively.
WT or mutant virus stocks were made in the presence or absence of 400 or 800 nM EFV. Equal amounts of virus as determined by capsid protein quantitation were pelleted by centrifugation at 150,000 × g for 1 h at 4°C. Virus pellets were resuspended in 10 μl of phosphate-buffered saline (PBS) and were lysed by adding 10 μl of 0.3% Igepal CA-630 (Sigma-Aldrich). The specific activities of DNA polymerization were measured as previously described (38).
The fitness of the K101E+G190S (ES), D10, and D10MMTT viruses in the absence of drug varies (39). Since the magnitude of the effect of the drugs may vary depending on the number of infectious viruses, the amount of virus used for infection was normalized. In order to optimize the linear range of the flow assay for stimulation, infections were also performed at a low multiplicity of infection (<0.1). To test whether stimulation occurs during the early steps of the virus life cycle, one million PM1 cells were incubated with virus at 37°C for 1 h, washed with PBS, and cultured with medium for 48 h at 37°C, in the presence or absence of EFV (0, 400, and 800 nM), NVP (0, 50 and 200 μM), or ETR (0, 1, 2, 10, and 100 nM). The cells were then stained with the anti-Thy1.1 and anti-Thy1.2 antibodies and analyzed by fluorescence-activated cell sorting (FACS) as previously described (13). Fold stimulation was calculated by dividing the percentage of infected cells in the absence of drug by the percentage of infected cells in the presence of drug. To test whether stimulation occurs during the late steps of the virus life cycle, virus was made in the presence or absence of EFV (0, 400, and 800 nM). PM1 cells were infected as described above but in the absence of drug.
PM1 cells were infected as previously described in the following concentrations of NVP and ETR: 0, 50, 200, and 400 μM for NVP and 0, 2, 4, 8, 16, 32 and 64 nM for ETR (39). After 6 days, clarified culture supernatants were harvested and the capsid p24 concentration was measured by an ELISA kit (Perkin Elmer). The 50% inhibitory concentration (IC50) and stimulation were calculated as previously described (39).
One day before infection, TZM-bl cells were seeded into a 24-well plate at a density of 5 × 104 cells per well. One hour before infection, the cells were treated for 1 h at 37°C with medium containing either 0, 200, 400, 800, 1,600, 3,200, or 6,400 nM EFV. The cells were infected with 10 ng of p24 ES, D10, or D10MMTT virus per well. After incubation for 2 h at 37°C, the cells were washed twice with PBS and cultured for 48 h in medium with or without EFV. Cells were lysed by passive lysis buffer (Promega), and the luciferase activity of the cell lysate was measured using the luciferase assay system (Promega).
The ApaI/EcoRI fragment from pNL4-3-deltaE-EGFP was replaced with the same fragment from pAT2D10MMTT to create pNL4-3-deltaE-EGFP-D10MMTT, which has the D10MMTT RT sequence instead of WT. pNL4-3-delta-EGFP-WT and pNL4-3-delta-EGFP-D10MMTT virus stocks were made by transient transfection of 293T cells using 5 μg of plasmid DNA along with 5 μg of pSV-A-MLV-ENV and 25 μl of Lipofectamine LTX (Invitrogen).
In order to remove plasmid DNA present in the virus stock during transfection, virus stocks were incubated for 1 h at 37°C in the presence of 40 U/ml of Turbo DNase (Ambion) and 4 mM MgCl2. A portion of the DNase-treated virus was heat inactivated by incubating at 65°C for 1 h. Three million PM1 cells were infected with 900 ng of p24 of DNase-treated WT or D10MMTT virus. Cells and virus were incubated for 1 h at 37°C, washed to remove unbound virus, and resuspended in medium with no drug or with 400 or 800 nM EFV. The cells were incubated at 37°C and at 0, 2, 4, and 8 h after incubation, 1 million cells were washed twice with PBS, and the pellets were frozen at −80°C. Genomic DNA was prepared from each pellet using the Qiagen DNA blood kit (Qiagen). Genomic DNA was incubated for 4 h at 37°C in the presence of DpnI. DpnI was heat inactivated, and the genomic DNA was then used for real-time PCR. Real-time PCR was performed by the protocol of Mbisa et al. (22) using TaqMan PCR master mix (Applied Biosystems) and the primer/probe sets (Integrated DNA Technologies) used to amplify minus-strand strong-stop DNA synthesis (RU5) and minus-strand transfer (U3RU5). Reactions were performed using 25-μl reaction mixtures, the Bio-Rad CFX96, a final primer concentration of 0.6 μM, 0.08 μM probe, and 5 μl of genomic DNA. The following cycling conditions were used: (i) 10 min at 95°C; (ii) 40 cycles (1 cycle consisting of 15 s at 95°C, 30 s at 55°C, and 30 s at 60°C); (iii) 5 min at 60°C; and (iv) 10 min at 95°C. Cycling data were analyzed using the Bio-Rad CTX Manager software.
Our group has previously shown that efavirenz (EFV) can stimulate virus replication for reverse transcriptase (RT) genotypes containing the nonnucleoside reverse transcriptase inhibitor (NNRTI) resistance mutations K101E+G190S (ES) (39). Our results were obtained by measuring the amount of viral p24 capsid released into the culture supernatant. We also wanted to know whether the number of infected cells increased during infection. In the current study, we used a multiple-cycle assay, which identified infected cells using a reporter gene expressed on the cell surface and measured the percentage of infected cells by flow cytometry (12). We used a panel of three RT genotypes cloned into pAT2 (12): (i) the NL4-3RT containing the site-directed mutations K101E and G190S (ES); (ii) a RT cloned from a plasma sample obtained from a patient failing EFV, which had the nucleoside drug-resistant mutations M41L and T215Y and the nonnucleoside drug-resistant mutations K101E and G190S and 28 other polymorphisms compared to the NL4-3 RT backbone (D10); and (iii) the same patient (D10) RT with the M41L and T215Y nucleoside resistance mutations back-mutated to the wild type but still retaining the K101E and G190S mutations (D10MMTT) (39). We infected PM1 T cells with ES, D10, or D10MMTT in the presence or absence of EFV. Figure 1A shows the fold increases in the percentages of infected cells compared to the percentages of infected cells in the no-drug control. The data show a significant increase in infection rates for ES at 200 and 400 nM EFV (P < 0.001), D10 at 400 and 800 nM EFV (P < 0.0001), and D10MMTT at 400, 800, 1,600, and 3,200 nM EFV (P < 0.001). The results shown here are consistent with previous results obtained by measuring p24 capsid protein production.
We also wanted to know whether the stimulation of RT by EFV is unique to the K101E+G190S genotype. To test this possibility, we made a series of K101E double mutants, in which K101E was paired with several NNRTI drug-resistant mutations shown to occur during EFV failure in patients and tested their replication efficiency in the presence and absence of 1, 10, 100, and 1,000 nM EFV (Fig. 1B). K101E in combination with L100I, K103N, G190A (data not shown), and Y181C (Fig. 1B) did not show significant stimulation of replication. However, K101E+V106I showed stimulation at 1 nM EFV (P = 0.0315) and K101E+Y188L showed the highest level of stimulation, reaching levels similar to those of D10 and D10MMTT at 100 and 1000 nM EFV (P < 0.02).
The assays used to test stimulation were multiple-cycle assays, which include all steps of the virus life cycle, both early steps (entry, reverse transcription, and integration) and late steps (RNA production, protein expression, and viral particle production). In order to determine whether stimulation occurred during early or late steps, we modified a single-cycle assay previously designed by our group (13). We used the envelope-deleted vector, pDAT2, and with a murine leukemia virus (MLV) envelope-expressing construct, made MLV-pseudotyped virus stocks in the presence or absence of EFV. The pDAT2 vector is identical to the pAT2 vector, except for the deletion in env. We then used these stocks to infect PM1 T cells in the presence or absence of drug (Fig. 2). We compared the infectivity of viruses made in the absence of drug with the infectivity in the presence of 400 nM or 800 nM EFV, but with no additional drug added during the infection step. Using the proportion of infected cells, we calculated the fold stimulation for ES, D10, and D10MMTT made in 400 and 800 nM EFV compared to virus made in 0 nM EFV (Fig. 3). We found that no stimulation occurred for these three genotypes when EFV was present during virus production (P > 0.2).
In order to confirm the single-cycle assay results, we wanted to test whether the dimerization of mutant RTs was impaired compared to the wild type and whether EFV had an effect on dimerization efficiency. Mutants with reduced dimerization can have reduced amounts of RT in virions. Therefore, we hypothesized that our mutants may have reduced dimerization and that EFV could correct this defect and thereby increase RT content and virus replication. In order to test whether EFV could increase the heterodimerization of NNRTI-resistant mutants that showed EFV-dependent replication stimulation, we compared the ratio of p66 to p51 using recombinant RT protein in a His-tagged p51 (His-p51) pulldown assay. We incubated p66 with His-tagged p51 in the presence and absence of increasing concentrations of EFV (Fig. 4A and B). EFV, in concentrations from 0.1 to 100 μM, increased the amount of p66 bound to p51 for wild-type RT by 40 to 70% (Fig. 4C), compatible with previous reports demonstrating that EFV stabilizes the RT heterodimer (34). ES, the G190S single mutant, and the L74V+K101E+G190S (VES) triple mutant did not show a decrease in p66/p51 dimerization in the absence of drug or an enhancement of dimerization in the presence of 0.1 to 1,000 μM EFV. We have previously shown that G190S and VES are not stimulated by EFV (32). These results show that regardless of whether or not a mutant has dimerization defects or is stimulated by EFV, EFV does not increase dimerization, and therefore, the mechanism of stimulation does not occur through an enhancement of dimerization.
We have previously shown that some NNRTI-resistant mutants have decreased amounts of RT in their virions (37). We wanted to know whether ES, D10, and D10MMTT virions had a reduced amount of RT and whether this reduction could be corrected in the presence of EFV. We assessed the amount of RT in virions made in the presence or absence of 400 and 800 nM EFV by Western blotting. We had previously shown that these concentrations give maximum levels of replication stimulation. Virus with VES NNRTI resistance mutations has a level of resistance to EFV similar to that of virus with ES NNRTI resistance mutations but is not stimulated by the drug. No significant difference was found in the amount of p24 capsid protein between stocks in the absence of drug, suggesting that these mutants produce similar amounts of virions (data not shown). In addition, the presence of EFV did not increase the amount of capsid protein, indicating that EFV does not stimulate virion production.
We also assessed the relative amounts of Gag, integrase, and RT compared to capsid protein of each virus stock by Western blotting (Fig. 5A). Consistent with the p24 ELISA, there was no significant difference in the relative amount of Gag intermediates (p55 and p41) for the drug-resistant mutations L74V+K101E+G190S (LES), D10, and D10MMTT compared to the WT. Virus with ES mutations had a slight increase in the amount of p41, indicating that this mutant may have a defect in Gag processing. However, the presence of EFV did not alter the amount of Gag intermediates for any of the mutant virions (Fig. 5A). Next we measured the relative amounts of integrase protein for each mutant compared to the wild type and found all mutant virions, except ES, had similar amounts of integrase (Fig. 5A). The reduced amount of integrase in virus with ES mutations may indicate a reduced amount of Gag-Pol incorporation into virions or a reduced amount of Gag-Pol processing. The presence of EFV did not alter the amount of integrase in virions compared to the control without drug.
Finally, we compared the relative amounts of RT (p66 and p51) for each protein and showed that all mutants had reduced amounts of RT compared to the wild type, a finding consistent with our previous report that NNRTI-resistant mutants have reduced RT. However, the addition of EFV during virus production did not correct this defect, indicating that the mechanism of stimulation does not depend on increasing the amount of RT in virions. We also tested the activity of the RT present in virions by measuring DNA polymerization using a homopolymeric template (Fig. 5B) and found that the amount of active RT in virions correlated with the amount of RT as measured by Western blotting. The activity of wild-type RT was inhibited by EFV incorporated into the virions during virus production, but EFV did not stimulate ES, D10, or D10MMTT polymerase activity.
Taken together, there was no stimulation of virus infectivity when EFV was present during the virus production step, no increase in dimerization for mutant RT by EFV, and no increase in the amount of RT in mutant virions made in the presence of EFV. We conclude from the above results that stimulation by EFV does not occur during late steps of infection.
Since EFV does not stimulate HIV-1 replication by enhancing the infectivity of virions made in the presence of drug, we next asked whether EFV enhances the infectivity of virions when present during the early stages of viral replication. We tested this hypothesis using the single-cycle assay (Fig. 2), but added EFV during the infection step rather than during virus production.
The percentages of infected cells were determined for the control with no drug and the cultures with 400 and 800 nM EFV. The ratio of drug to no drug is shown in Fig. 6. ES showed no stimulation of infection in the presence of EFV (P > 0.6). D10 showed a 29% (P = 0.0003) increase in infection at 400 nM EFV and a 16% (P = 0.0466) increase in infection at 800 nM EFV. D10MMTT showed a 214% (<0.0001) increase in infectivity at 400 nM EFV and a 98% (P < 0.0001) increase in infectivity at 800 nM EFV (Fig. 3). The relative hierarchy of stimulation was D10MMTT > D10 > ES which was the same hierarchy found using the multiple-cycle assay. We also measured the inhibition of the wild-type virus by EFV using the single-cycle assay and found that the 50% inhibitory concentration (IC50) was similar to previously published results (data not shown). These results show that EFV stimulates the early stages of virus replication.
The single-cycle assay was performed using a vector that has a deletion in nef and was pseudotyped with MLV envelope. In order to determine whether nef or env impacted the level of stimulation see in the single-cycle assay, we used a complete intact HIV-1 genome (NL4-3XX) and measured stimulation after 40 h of infection. This short time course simulates a single cycle of replication. We infected TZM-bl cells, which express the firefly luciferase (Luc) under the control of the HIV-1 long terminal repeat (LTR) promoter with NL4-3XXES, NL4-3XXD10, and NL4-3XXD10MMTT (Fig. 7). The luciferase activity was measured in the presence and absence of EFV. The results mirrored those of the single-cycle assay and showed that the replication rate of NL4-3ES increased 20% in the presence of 1,600 nM EFV (P < 0.005 by t test) compared to the control with no drug. D10 infection was stimulated 30 to 40% in 200 to 1,600 nM EFV (P < 0.001 by t test), and D10MMTT infection was stimulated 80 to 280% in 200 to 3,200 nM EFV (P < 0.001 by t test; Fig. 7). The hierarchy of stimulation was D10MMTT > D10 > ES, consistent with the hierarchy of the single-cycle assay. These data suggest that the presence of EFV specifically enhanced the early stages of the HIV-1 life cycle.
We also wanted to determine whether other NNRTIs could stimulate the replication of HIV-1 with the K101E+G190S mutations. Therefore, we measured the IC50 and stimulation levels of ES, D10, and D10MMTT in the presence of nevirapine (NVP) and etravirine (ETR) using a multiple-cycle assay. The IC50 of wild-type to NVP was 12.9 ± 0.53 nM. Due to cellular toxicity, we could not achieve a NVP concentration high enough to calculate an IC50 for ES, D10, and D10MMTT. NVP concentrations above 200 μM decreased the viability of the PM1 cells. Since NVP is reconstituted in DMSO, we believe the cellular toxicity was due to the higher final concentrations of DMSO in the culture medium. ES showed a trend toward stimulation of 1.9- and 1.3-fold stimulation in 50 (P = 0.119) and 100 (P = 0.106) μM NVP, respectively, compared to the absence of drug (Fig. 8A). D10 showed a significant level of stimulation of 2.9- and 2.6-fold stimulation in 50 and 100 μM NVP, respectively (P < 0.02), and D10MMTT showed 8.0-, 7.9-, and 5.0-fold stimulation in 50, 100, and 200 μM NVP, respectively (P < 0.004; Fig. 8A). We also wanted to know whether the stimulation that occurred during the multiple-cycle assay occurred during the infection step, so we used the single-cycle assay described above to measure the infection rates of D10MMTT in the presence of 0, 50, and 200 μM NVP (Fig. 8B). There was a significant 2- to 3-fold increase in infectivity when NVP was added during the infection step (P < 0.02).
The IC50s of wild-type, ES, D10, and D10MMTT to ETR were 2.8 ± 0.24 nM, 20.6 ± 5.94 nM, 5.3 ± 0.68 nM, and 31.9 ± 4.9 nM, respectively. ES, D10, and D10MMTT did not show any significant levels of stimulation in ETR concentrations ranging from 2 to 64 nM (Fig. 8C), and we did not see stimulation of D10MMTT using the single-cycle assay when ETR was added during the infection step (data not shown).
Intermediate products of reverse transcription, such as minus-strand strong-stop DNA synthesis (−SSDNA) and minus-strand transfer, can be measured using real-time PCR at early time points after infection of cells with single-cycle virus (22). We used this method to measure the rate of −SSDNA synthesis and the efficiency of minus-strand transfer in cells infected with NL4-3-deltaE-EGFP WT or D10MMTT virus in the absence of drug (Fig. 9A). We also measured whether EFV could stimulate these steps by infecting cells with D10MMTT in the presence of 400 and 800 nM EFV. By comparing the average slopes of −SSDNA product over time, the rate of synthesis for D10MMTT was reduced 70% ± 5% (P = 0.003) compared to the WT in the absence of drug (data not shown). Inclusion of 800 nM EFV increased the rate of synthesis of D10MMTT 48% ± 18% (P = 0.0005) compared to the absence of drug (Fig. 9B). At 8 h, the number of RU5 copies at 400 nM EFV increased 5-fold compared to the number at time zero (data not shown). We also evaluated the efficiency of strand transfer by comparing the relative amounts of the RU5 and U3U5 products (Fig. 9B). In the absence of drug, the mutant appears to be more efficient at minus-strand transfer than the wild type (data not shown), indicating a possible change in the ratio of polymerization to RNase H activity. In the presence of 800 nM EFV, there was no significant difference in the strand transfer efficiency for D10MMTT at 4 h (P = 0.27) or 8 h (P = 0.67; Fig. 9C). At 400 nM EFV, there was no significant difference in strand transfer efficiency at 4 h (P = 0.11), but there was a 50% increase at 8 h (P = 0.001), indicating that with 400 nM EFV, there may be a small amount of stimulation (Fig. 9C).
In order to determine in which step of the virus life cycle stimulation occurred, we used a single-cycle assay. Since we replaced the nef gene with the mouse Thy 1 gene in our single-cycle vector, the Thy 1 protein was expressed early in infection, and because the env gene was also deleted, no progeny viruses were produced. Since we detected stimulation only when drug was added after the virus was adsorbed to the cells, and not when virus stocks were made during the transfection process, stimulation could only occur during the steps of the life cycle from viral capsid uncoating to protein (i.e., Thy1.2) expression. Therefore, these studies provide evidence that stimulation by NNRTIs occurs during early steps of virus replication. This conclusion is derived from the finding that the relative hierarchy of stimulation measured in the multiple-cycle assay matches that of the single-cycle assay, when drug was present during the infection step. In the single-cycle assay, no stimulation of infectivity was observed when drug was added during the virus production step. We also did not observe an increase in functional RT in virions in the presence of drug, and we did not observe an enhancement of recombinant RT heterodimerization in the presence of drug. We also showed using real-time PCR of reverse transcription intermediates that minus-strand strong-stop DNA synthesis, but not minus-strand transfer efficiency, is stimulated by efavirenz (EFV).
We observed that in the absence of drug, ES, VES, D10, and D10MMTT virions all had reduced RT content compared to the wild type. This result is consistent with our previously published results showing that NNRTI mutants have reduced RT content (37). The addition of the nucleoside resistance mutation L74V to K101E+G190S (VES) increased the virion RT content compared to K101E+G190S alone. However, the presence of M41L and T215Y in D10 did not increase the RT content compared to D10MMTT, where the M41L and T215Y mutations were back-mutated to the wild type. The finding that L74V and M41L+T215Y eliminate stimulation yet have discordant effects on RT content further supports the conclusion that the mechanism of stimulation does not occur through RT content. However, the presence of L74V may improve the virus fitness cost of K101E+G190S in the absence of drug.
We showed that K101E+G190S has reduced Gag-Pol processing and reduced amount of integrase compared to the wild type. We have previously shown that G190S alone also has reduced Gag-Pol processing (37). Therefore, the reduction effect of K101E+G190S on Gag-Pol processing is probably through the G190S mutation. However, G190S was not observed to have reduced amounts of integrase. The reduction in the amount of integrase in K101E+G190S virions indicates that this double mutation may have a defect in Gag-Pol incorporation. These defects most likely contribute to the reduced fitness of K101E+G190S. However, since EFV did not correct the processing defect or integrase amounts, these steps of the virus life cycle are not involved in the stimulatory effect of NNRTIs on virus replication.
Stimulation of virus replication was not restricted to EFV, since NVP also shows stimulation for the D10MMTT genotype. However, ETR did not demonstrate stimulation. Etravirine (ETR) is unlike EFV and nevirapine (NVP) in that it has a elastic structure that allows more flexibility in binding to the NNRTI binding pocket (17, 28, 31). We show here that the ETR IC50 of K101E+G190S is low, indicating that the flexibility of ETR allows it to bind to the NNRTI binding pocket despite the presence of the K101E and G190S mutations.
We show that other NNRTI drug-resistant mutations, such as K101E+V106I and K101E+Y188L, also demonstrate stimulation. However, stimulation was not seen for other combinations of K101E with a second NNRTI resistance mutation. K101E+L100I, K101E+K103N, K101E+G190A, and K101E+Y181C were inhibited to different degrees by EFV but were not stimulated. These results support the conclusion that stimulation is mutation dependent.
Drug-dependent replication of HIV-1 mutants has been demonstrated previously for protease inhibitors and the T20 fusion inhibitor (3, 4, 23). Protease inhibitors (PIs) prevent the enzyme from cleaving the Gag and Gag-Pol polypeptides (reviewed in reference 16), but the mechanism by which protease inhibitors can enhance replication of PI-resistant variants is unknown. Fusion inhibitors act by preventing the binding of heptad region 1 (HR1) with heptad region 2 (HR2) in the gp41 protein of the env gene during virus entry (43, 44). Fusion inhibitors stimulate replication by acting as a “safety pin,” which prevents the premature binding of HR1 to HR2 of fusion inhibitor-resistant genotypes during entry (reviewed in reference 3). Since the mechanisms by which NNRTIs, PIs, and fusion inhibitors prevent virus replication are different, it is likely that the mechanisms by which NNRTIs, PIs, and fusion inhibitors stimulate replication are also different.
The specific mechanism by which stimulation occurs for NNRTIs is not yet understood. However, to gain some information about the steps of viral replication involved, real-time PCR measurements of reverse transcription intermediates were performed. These studies show that minus-strand strong-stop DNA synthesis is stimulated by EFV. Therefore, tRNA priming or elongation may be stimulated by the drug. Minus-strand transfer efficiency had variable results with no significant stimulation at 4 or 8 h for 800 nM EFV and 50% stimulation at 8 h only for 400 nM EFV. The magnitude of the stimulation of minus-strand transfer at 400 nM EFV was 10 fold less than the stimulation seen for minus-strand strong-stop DNA synthesis. Since it is known that the RNase H and strand transfer activities of RT are important for efficient minus-strand transfer (5–7, 10, 20), we conclude that these activities are not stimulated by NNRTIs. However, we cannot rule out the possibility that the effects of the drug on RNase H activity are more complicated and that specific RNase H cleavages, such as cleavage of the polypurine tract and tRNA primers from the ends of the provirus could be stimulated. Stimulation of receptor binding is likely not involved, since drug is not added until the virus has had time to bind to the cell in the single-cycle assay used here. However, even in light of the real-time PCR results, we cannot rule out the possibility that in addition to reverse transcription, stimulation by the drug could occur during uncoating, integration, or protein expression.
We believe it most likely that the effect of the drug is mediated through a direct interaction with RT, rather than through another viral or cellular protein. This concept is supported by the facts that EFV and NVP, but not ETR, stimulate the relevant drug-resistant genotypes and that the addition of nucleoside resistance mutations in RT reduces or eliminates the stimulation. We postulate that a direct interaction could occur through the creation of a new binding site in a different location than the NNRTI binding pocket, functional only in the presence of the K101E+G190S mutations, and therefore, it would not be present in WT RT. Alternatively, the mutations could alter the binding of the drug to the NNRTI binding pocket, creating new hydrogen bonds and other interactions. This alteration could allow the drug to bind in a different conformation than occurs to inhibit the enzyme, instead causing a stimulatory effect. These mechanisms are allowable even though the drug cannot bind to the NNRTI binding pocket in the same manner as the drug would bind WT RT. Although outside the scope of this work, future studies assessing the structural properties of the K101E+G190S RT and where NNRTIs bind will shed light on the mechanism of stimulation.
Since NNRTIs are widely used to treat HIV-1 infection, virologic failure with mutants containing K101E, V106I, Y188L, and G190S, is likely to develop in significant numbers of patients, despite the low frequency of these mutants relative to K103N. Resistance and stimulation may both play an important role in the development of NNRTI-resistant mutants in patients, particularly when K101E is present. Our results suggest that any newly developed NNRTI should be tested to ensure that it does not stimulate known NNRTI-resistant mutants of HIV-1. Understanding the mechanism by which stimulation by NNRTIs occurs will be necessary to develop drugs that do not select for this phenomenon. Therefore, NNRTI stimulation is clinically relevant, impacts future drug development, and clearly warrants further investigation.
This work was supported in part by NIH R01-AI-065217 and R01-AI-041387 and the University of Rochester Developmental Center for AIDS Research (P30-AI-078498).
Published ahead of print on 10 August 2011.