The Connection Domain Mutation N348I in HIV-1 Reverse Transcriptase Enhances Resistance to Etravirine and Rilpivirine but Restricts the Emergence of the E138K Resistance Mutation by Diminishing Viral Replication Capacity

Abstract

Clinical resistance to rilpivirine (RPV), a novel nonnucleoside reverse transcriptase (RT) inhibitor (NNRTI), is associated an E-to-K mutation at position 138 (E138K) in RT together with an M184I/V mutation that confers resistance against emtricitabine (FTC), a nucleoside RT inhibitor (NRTI) that is given together with RPV in therapy. These two mutations can compensate for each other in regard to fitness deficits conferred by each mutation alone, raising the question of why E138K did not arise spontaneously in the clinic following lamivudine (3TC) use, which also selects for the M184I/V mutations. In this context, we have investigated the role of a N348I connection domain mutation that is prevalent in treatment-experienced patients. N348I confers resistance to both the NRTI zidovudine (ZDV) and the NNRTI nevirapine (NVP) and was also found to be associated with M184V and to compensate for deficits associated with the latter mutation. Now, we show that both N348I alone and N348I/M184V can prevent or delay the emergence of E138K under pressure with RPV or a related NNRTI, termed etravirine (ETR). N348I also enhanced levels of resistance conferred by E138K against RPV and ETR by 2.2- and 2.3-fold, respectively. The presence of the N348I or M184V/N348I mutation decreased the replication capacity of E138K virus, and biochemical assays confirmed that N348I, in a background of E138K, impaired RT catalytic efficiency and RNase H activity. These findings help to explain the low viral replication capacity of viruses containing the E138K/N348I mutations and how N348I delayed or prevented the emergence of E138K in patients with M184V-containing viruses.

INTRODUCTION

Since the mid-1990s, combination antiretroviral therapy (ART), also referred to as highly active antiretroviral therapy (HAART), has led to significant declines in HIV/AIDS-associated morbidity and mortality (1, 2). The use of two nucleoside/nucleotide reverse transcriptase (RT) inhibitors [N(t)RTIs] plus one nonnucleoside reverse transcriptase inhibitor (NNRTIs) has been commonly employed in first-line ART, but the rapid replication rate of HIV-1 and the error-prone nature of its RT can drive the development of drug resistance against all drugs currently in use (3, 4).

HIV-1 RT carries out the viral DNA polymerase and RNase H activities (5). Structurally, HIV-1 RT is an asymmetric heterodimer composed of a 560-amino-acid p66 subunit (66 kDa) and a 440-amino-acid p51 subunit (51 kDa) (6). The p66 subunit contains both the DNA polymerase and RNase H active sites and is composed of three domains: the N-terminal polymerase domain (residues 1 to 319), the connection domain (residues 320 to 440), and the C-terminal RNase H domain (residues 441 to 560) (7). The p51 subunit is derived from HIV-1 protease-mediated cleavage of the p66 subunit via removal of the C-terminal RNase H domain. Although the p51 subunit contains the same amino acid sequences that comprise the DNA polymerase domain of the p66 subunit, the polymerase active site in p51 is not functional. The polymerase domain consists of the finger, palm, and thumb subdomains, which are analogous to a right hand in shape (7). Subunit-specific mutational analyses have shown that p51 is involved in the fine-tuning of RT enzymatic activities (8–11).

The N(t)RTIs, which include zidovudine (ZDV), didanosine (ddI), stavudine (d4T), lamivudine (3TC), emtricitabine (FTC), abacavir (ABC), and tenofovir (TFV), compete with deoxynucleoside triphosphate (dNTP) substrates and act as DNA chain terminators; the NNRTIs, including nevirapine (NVP), delavirdine (DLV), efavirenz (EFV), etravirine (ETR), and rilpivirine (RPV), act allosterically and noncompetitively by inducing conformational changes in the RT enzyme through binding to the NNRTI-binding pocket (NNATI-BP) and by inhibiting the chemical step of polymerization. Resistance-associated mutations (RAMs) for NNRTIs are generally located at the NNRTI-BP, whereas RAMs that confer resistance to NRTIs can be scattered throughout the polymerase domain. Mechanisms of resistance to NRTIs are due to either reduced incorporation of nucleotide analogs into DNA or enhanced excision and removal of the incorporated NRTI chain terminator from the 3′ end of the blocked primer (12–16). In contrast, multiple mechanisms of resistance to NNRTIs have been reported (14–18).

Most NNRTI and NRTI resistance mutations are located within the palm and finger subdomains of HIV-1 RT, and thus, standard genotyping for RAMs generally covers the N-terminal ∼300 amino acids in the polymerase domain. During the past 6 years, however, mutations in the C-terminal domain of RT, including the connection and RNase H domains, have been shown to be associated with NNRTI and NRTI resistance (19–24). An N-to-I substitution at position 348 (N348I) in the connection domain is particularly prevalent and can confer resistance to both NRTIs and NNRTIs in ∼10% of treatment-experienced individuals (22, 24). N348I-containing virus is associated with resistance to NVP, ZDV, and ddI (22, 24, 25) and enhances resistance to TFV in tandem with thymidine analog-associated mutations (TAMs) (26, 27). The N348I mutation emerged early in ZDV- and ddI-containing antiretroviral therapy and frequently coappears with TAMs, M184V, and the NNRTI RAMs K103N, Y181C/I and G190A/S (19, 22, 24, 26, 28–30). N348I is especially associated with the simultaneous use of ZDV and NVP (8, 22, 24) and was also five times more prevalent in patients receiving both ZDV and 3TC than in those receiving ZDV alone (26).

The M184V mutation confers high-level resistance to both 3TC and FTC and has a negative impact on enzyme processivity and viral replication capacity (RC) due to the fact that M184V-containing RT is defective in dNTP usage (31, 32). Biochemical and virological data have demonstrated that N348I can complement this deficit of M184V (25, 26).

However, very little is known about the impact of N348I on the newer NNRTIs RPV and ETR, which are di-aryl-pyrimidine (DAPY) compounds that have been approved for use in HIV therapy (33–35). In contrast, the E138K mutation in RT is associated with RPV-related virologic failure (36) and has also been selected by both ETR and RPV in cell culture (37, 38). Although E138K confers modestly reduced susceptibility to ETR and RPV (37–40), it does not result in reduced susceptibility to EFV and NVP (40–42). The mechanism of E138K-mediated RPV resistance is an increase in the dissociation rate of RPV, which overcomes a smaller enhancement in its rate of association (11). E138K and M184I/V have a mutual compensatory effect in regard to RT processivity and viral replication capacity (9, 11, 43, 44). However, unlike N348I, E138K is relatively rare in treatment-experienced patients (40, 45–47), despite the fact that it can mutually compensate for M184I/V (43, 44).

M184V is usually preceded by a M184I mutation that results from a G-to-A hypermutation. M184I is usually outcompeted by M184V due to the higher replication capacity of viruses containing the latter substitution. An NNRTI mutation, Y181C, prevents the emergence of E138K by impairing enzyme and viral fitness without enhancing levels of resistance (38–53).

As stated above, the M184I/V and E138K mutations can mutually compensate for deficits in RT enzyme processivity and viral replicative fitness that are conferred individually by these substitutions. Furthermore, this explains why these mutations appeared very frequently in patients who underwent treatment failure while receiving combination therapy that included both FTC and RPV. However, an obvious question is why the E138K substitution did not spontaneously appear in patients who received therapy with 3TC and developed the M184I/V mutations during the early years of ART, which often included NVP and ZDV together with 3TC. One possible answer is that the N348I RT connection domain mutation that is associated with resistance to both ZDV and NVP may have forestalled the emergence of E138K. Here, we provide experimental data in support of this hypothesis.

MATERIALS AND METHODS

Chemicals, cells and nucleic acids.

Etravirine (ETR) and rilpivirine (RPV) were gifts of Janssen Pharmaceuticals (Titusville, NJ).

Cord blood mononuclear cells (CBMCs) were obtained through the Department of Obstetrics, Jewish General Hospital, Montreal, Canada. The HEK293T cell line was obtained from the American Type Culture Collection (ATCC). The following reagents and cells were obtained through the NIH AIDS Research and Reference Reagent Program: the infectious molecular clone pNL4-3, from Malcolm Martin; the TZM-bl (JC53-bl) cell line, from John C. Kappes, Xiaoyun Wu, and Tranzyme Inc.; NVP; and EFV.

pNL4.3PFB plasmid DNA was a generous gift of Tomozumi Imamichi, National Institutes of Health, Bethesda, MD.

An HIV-1 RNA template, ∼500 nucleotides (nt) in size and spanning the 5′ untranslated region (UTR) to the primer-binding site (PBS), was transcribed in vitro from AccI-linearized pHIV-PBS DNA (54) by using an Ambion T7-MEGAshortscript kit (Invitrogen, Burlington, ON, Canada), as described previously (55). The oligonucleotides used in this study were synthesized by Integrated DNA Technologies Inc. (Coralville, IA) and purified by polyacrylamide-urea gel electrophoresis. The sequences of DNA template D110 and primer D25 used for DNA-dependent DNA polymerase assay are as follows: 5′-GAATTAGATCGATGGGAAAAAATTCGGTTAAGGCCAGGGGGAAAGAAAAAATATAAATTAAAACATATAGTATGGGCAAGCAGGGAGCTAGAACGATTCGCAGTTAATCC-3′ (D110) and 5′-GGATTAACTGCGAATCGTTCTAGCT-3′ (D25). For 5′-end labeling of oligonucleotides with [γ-³²P]ATP, the Ambion KinaseMax kit was used, followed by purification through Ambion NucAway spin columns, according to protocols provided by the supplier (Invitrogen, Burlington, ON, Canada).

Site-directed mutagenesis and preparation of site-directed mutant HIV-1_NL4.3 virus stocks.

To construct HIV-1 RT expression plasmids and recombinant HIV-1_NL4.3 infectious clones harboring the desired mutations in the RT gene, site-directed mutagenesis reactions were carried out by using a QuikChange II XL site-directed mutagenesis kit (Stratagene, La Jolla, CA). This work was performed with HIV-1 RT expression plasmid pbRT6H-PROT DNA (55) and plasmid pNL4.3PFB DNA (56) to generate recombinant RT enzymes and HIV-1_NL4.3 viruses containing the desired RT mutations. DNA sequencing was performed to verify the absence of spurious mutations and the presence of any desired mutation in the RT coding sequences. Recombinant HIV-1 wild-type (WT) and mutant viruses were generated by transfection of the corresponding proviral plasmid DNAs into HEK293T cells using Lipofectamine 2000 (Invitrogen, Burlington, ON, Canada) according to the manufacturer's instructions. Viral supernatants were harvested at 48 h posttransfection, centrifuged for 5 min at 800 × g to remove cellular debris, filtered through a 0.45-μm-pore-size filter, aliquoted, and stored at −80°C. Levels of p24 in viral supernatants were measured by using a PerkinElmer HIV-1 p24 antigen enzyme-linked immunosorbent assay (ELISA) kit according to manufacturer's instructions. Virion-associated RT activity was verified by an in vitro RT assay, as described previously (5).

Selection of HIV-1 mutants in cord blood mononuclear cells under drug selection pressure.

CBMCs stimulated with phytohemagglutinin A (PHA) were cultured in RPMI 1640 medium supplemented with 10% qualified fetal bovine serum (FBS), 20 U of human interleukin-2 (IL-2)/ml, 5 μg of hydrocortisone/ml, 2 mM l-glutamine/ml, 100 U of penicillin/ml, and 100 μg of streptomycin/ml. Cells in 24-well tissue culture plates were infected with recombinant viral clones at a similar multiplicity of infection (MOI). Selection for viral resistance mutations was performed by using increasing concentrations of RT inhibitors at starting concentrations below the 50% effective concentration (EC₅₀), as described previously (37, 38, 57). As controls, all viruses were simultaneously passaged without drugs. Reverse transcriptase assays were performed, as described previously, to monitor viral replication (37, 38, 58, 59). Based on the ratio of the RT value of control wells (100%) to those of wells containing drug at the previous round of replication, drug concentrations were adjusted at subsequent passages. Briefly, if the value of the drug-treated well was between 50 and 100% of the control value, the drug concentration was raised by 2- to 2.5-fold; if this value was between 10% and 49% of the control value, the same concentration was maintained; and if the value dropped to <10% of the control value, drug pressure was released entirely, and the process of selection was reinitiated. Virus-containing culture media were harvested and kept at −80°C for subsequent genotypic analysis. Selections for resistance were performed over a period of 19 weeks.

Measurements of HIV-1 replication kinetics in CBMCs.

CBMCs were isolated and cultured as previously described (57). Recombinant WT viruses and viruses containing the desired mutations were normalized by p24 in order to minimize interinoculum effects, as described previously (58). Briefly, 2 × 10⁶ CBMCs were infected with viruses containing 8 × 10⁶ pg of p24 for 2 h, and the cells were washed with medium, resuspended in 4 ml complete medium after sedimentation, and split into 2 wells for each sample of a 12-well plate. The replication kinetics of mutant and WT viral stocks were assessed on the basis of p24 levels in culture supernatants sampled at variable time points postinfection, as measured by ELISA as described above.

Measurements of HIV-1 replication capacity in TZM-bl cells.

The relative replicative capacities of the following recombinant WT HIV-1_NL4.3 clones containing the E138K, E138K/N348I, and E138K/M184IV/N348I mutations were evaluated in a noncompetitive infectivity assay using TZM-bl cells, as previously described (37, 60). Twenty thousand cells per well were added in triplicate into a 96-well culture plate in 100 μl Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, Burlington, ON, Canada) supplemented with 10% fetal bovine serum (Invitrogen, Burlington, ON, Canada), 1% penicillin-streptomycin, and 1% l-glutamine (Invitrogen, Burlington, ON, Canada). Viral stocks for both wild-type and mutant viruses were normalized by p24, and recombinant viruses were serially diluted 2-fold from viral stock suspensions. After 4 h, 50 μl of DMEM was removed from the wells and replaced by 50 μl of virus dilution; a control well did not contain virus. Virus and cells were cocultured for 48 h, after which 100 μl of Bright-Glo reagent was added and luciferase activity was measured in a 1450 MicroBeta TriLux microplate scintillation and luminescence counter (PerkinElmer), as described above. The viral replication level of each viral variant was expressed as a percentage of relative light units (RLU) with reference to the WT virus.

Analysis of phenotypic drug susceptibility in TZM-bl cells.

Phenotypic susceptibility analyses of RT inhibitors were performed with recombinant HIV-1_NL4.3 clones in a TZM-bl cell-based in vitro assay, as described previously (37, 38, 61). Briefly, RT inhibitors at variable concentrations were added to TZM-bl cells (10⁴ cells/well) in 96-well plates grown in 100 μl supplemented medium. Immediately after drug addition, cells were infected with WT or mutant viruses. At 48 h postinfection, cells were rinsed with 100 μl phosphate-buffered saline (PBS) and lysed with 50 μl/well Promega cell lysis reagent (Fisher Scientific, Ottawa, ON, Canada). Cell lysates were then transferred to a white, opaque 96-well plate (Corning, Tewksbury, MA). Promega luciferase assay reagent (Fisher Scientific, Ottawa, ON, Canada) was added to each well, and RLU/well were measured by using a PerkinElmer 1450 MicroBeta TriLux microplate scintillation and luminescence counter (PerkinElmer, Waltham, MA). The EC₅₀ was calculated by using the GraphPad Prism program (GraphPad Software, San Diego, CA).

Recombinant reverse transcriptase expression and purification.

Recombinant RTs in heterodimeric form were expressed from plasmid pbRT6H-PROT (55) and purified as described previously (62, 63), with minor modifications. In brief, RT expression in Escherichia coli M15(pREP4) (Qiagen, Mississauga, ON, Canada) was induced with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at room temperature. Pelleted bacteria were lysed under native conditions with BugBuster protein extraction reagent containing Benzonase (Novagen, Madison, WI), according to the manufacturer's instructions. After clarification by high-speed centrifugation, the clear supernatant was subjected to the batch method of Ni-nitrilotriacetic acid (NTA) metal affinity chromatography (QIAexpressionist; Qiagen, Mississauga, ON, Canada). All buffers contained Complete protease inhibitor cocktail (Roche, Mississauga, ON, Canada). Hexahistidine-tagged RT was eluted by using an imidazole gradient. RT-containing fractions were pooled, passed through DEAE-Sepharose (GE Healthcare, Mississauga, ON, Canada), and further purified by using SP-Sepharose (GE Healthcare, Mississauga, ON, Canada). Fractions containing purified RT were pooled, dialyzed against storage buffer (50 mM Tris-HCl [pH 7.8], 50 mM NaCl, and 50% glycerol), and concentrated to 4 to 8 mg/ml with Centricon Plus-20 30-kDa-molecular-mass-cutoff filters (Millipore, Etobicoke, ON, Canada). Aliquots of proteins were stored at −80°C. Protein concentration was measured by using a Bradford protein assay kit (Bio-Rad Laboratories, Saint-Laurent, QC, Canada), and the purity of the recombinant RT preparations was verified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The RNA-dependent DNA polymerase activity of each recombinant RT preparation was evaluated as described previously (64), using various concentrations of RT and a synthetic homopolymeric poly(rA)/poly(dT)_12–18 template/primer (T/P) (Midland Certified Reagent Company, Midland, TX).

Susceptibility of RT inhibitors tested by DNA-dependent DNA polymerase activity using recombinant RT enzymes.

Susceptibility to ETR and RPV was assayed by using recombinant RT enzymes and heterodimeric DNA template/primer pair D110/D25, and experiments were performed as described previously (55). Briefly, RT reaction mixture containing 50 mM Tris-HCl (pH 7.8), 6 mM MgCl₂, 60 mM KCl, dNTPs (5 μM each) with 2.5 μCi of [³H]dTTP (70 to 80 mCi/mmol), 30 nM heterogeneous HIV-1 DNA T/P pair D110/D25, the same activity of RT enzymes, and variable amounts of RT inhibitors was included in 50-μl reaction volumes. In the reaction mixtures, the final concentrations of ETR or RPV were 0, 0.01, 0.03, 0.10, 0.30, 1.00, 3.00, and 10.00 μM. Reaction mixtures were incubated at 37°C for 30 min, reactions were terminated by adding 0.2 ml of 10% cold trichloroacetic acid (TCA) and 20 mM sodium pyrophosphate to the mixtures, and the mixtures were incubated for at least 30 min on ice. The precipitated products were filtered through a 96-well MultiScreen HTS FC filter plate (Millipore) and sequentially washed with 200 μl of 10% TCA and 150 μl of 95% ethanol. The radioactivity of incorporated products was analyzed by liquid scintillation spectrometry using a 1450 MicroBeta TriLux microplate scintillation and luminescence counter (PerkinElmer). The 50% inhibitory concentration (IC₅₀) of each RT inhibitor was determined by nonlinear regression analysis using GraphPad Prism5.01 software.

RT-catalyzed RNase H activity.

RNase H activity was assayed by using a 41-mer 5′-³²P-labeled heteropolymeric RNA template, kim40R, that was annealed to a complementary 32-nucleotide DNA oligomer, termed kim32D, at a 1:4 molar ratio, as described previously (44). Reactions were conducted at 37°C in mixtures containing an RNA-DNA duplex substrate (20 nM) with RT enzymes in assay buffer, 50 mM Tris-HCl (pH 7.8), 60 mM KCl, and 5 mM MgCl₂. Aliquots were removed at different time points after initiation of reactions and quenched by using an equal volume of formamide sample loading buffer (96% formamide, 0.1% each bromophenol blue and xylene cyanol FF, and 20 mM EDTA). The samples were heated to 90°C for 3 min, cooled on ice, and electrophoresed through 6% polyacrylamide–7 M urea gels. The gels were analyzed by phosphorimaging.

RNA-dependent DNA polymerase activity.

The same 497-nt RNA and 5′-end-³²P-labeled D25 primers described previously (44) were used to assess the polymerization efficiency of recombinant RT enzymes in time course experiments. Final reaction mixtures contained 20 nM T/P, 400 nM RT enzyme, 50 mM Tris-HCl (pH 7.8), and 50 mM NaCl. Reactions were initiated by adding 6 mM MgCl₂ and dNTPs at 200 μM to the mixtures, and the mixtures were sampled at 30 s, 60 s, and 5 min, respectively, and mixed with 2 volumes of stop solution. Reaction products were separated by 6% denaturing polyacrylamide gel electrophoresis and analyzed as described previously (44).

Processivity assays.

The processivity of recombinant RT proteins was analyzed as described previously, using a heteropolymeric RNA template in the presence of a heparin enzyme trap to ensure a single processive cycle, i.e., a single round of binding and of primer extension and dissociation (44). The T/P was prepared by annealing the 497-nt HIV PBS RNA with 5′-end-³²P-labeled 25-nt DNA primer D25 at a molar ratio of 1:1, denatured at 85°C for 5 min, and then slowly cooled to 55°C for 8 min and to 37°C for 5 min to allow for specific annealing of the primer to the template. RT enzymes with equal amounts of activity and 40 nM T/P were preincubated for 5 min at 37°C in a buffer containing 50 mM Tris-HCl (pH 7.8), 50 mM NaCl, and 6 mM MgCl₂. Reactions were initiated by the addition of dNTPs at 0.5 μM and a heparin trap (final concentration, 3.2 mg/ml), and the mixtures were incubated at 37°C for 30 min; 2 volumes of stop solution (90% formamide, 10 mM EDTA, and 0.1% each xylene cyanol and bromophenol blue) were then added to stop the reaction. Reaction products were denatured by heating at 95°C and analyzed by using 6% denaturing polyacrylamide gel electrophoresis and phosphorimaging. The effectiveness of the heparin trap was verified in control reactions in which the trap was preincubated with the substrate before the addition of RT enzymes and dNTP.

Steady-state kinetics analysis.

Kinetics studies were carried out with a modification of a previously described method using homopolymeric poly(rA)/poly(dT)_12–18 and complementary dTTP as the nucleotide substrate (44). The reaction mixture (10 μl) contained 50 mM Tris-HCl (pH 7.8), 60 mM KCl, 6 mM MgCl₂, 5 mM DTT, 0.5 U/ml poly(rA)/poly(dT)_12–18, RT enzymes, variable concentrations of the tracer [³H]dTTP, and cold dTTP (0.2 to 200 μM). Reactions were run at 37°C and quenched by adding 0.2 ml of 10% cold trichloroacetic acid and 20 mM sodium pyrophosphate to the mixtures, and products were filtered onto Millipore 96-well MultiScreen HTS FC filter plates (MSFCN6B) and sequentially washed with 200 μl of 10% TCA and 150 μl of 95% ethanol. The radioactivity of incorporated products was analyzed by liquid scintillation spectrometry using a PerkinElmer 1450 MicroBeta TriLux microplate scintillation and luminescence counter. The steady-state kinetic parameters V_max and K_m for nucleotide substrates were determined by nonlinear regression analysis of the substrate concentration and initial velocity data using the Michaelis-Menten equation with the program GraphPad Prism 5.01 according to the manufacturer's instructions. The k_cat was calculated by using the equation k_cat = V_max/[E], where E represents the enzyme.

RESULTS

The N348I mutation delays or prevents the emergence of E138K in CBMCs under drug pressure.

To determine whether the N348I mutation, alone or in combination with M184V, might prevent the emergence of E138K under ETR or RPV pressure, we introduced the M184V, N348I, and M184V/N348I mutations into the HIV-1 NL_4.3 proviral clone and performed viral passage experiments in CBMCs using increasing drug concentrations over a period of 19 weeks. Figure 1 shows the weekly increment of drug concentrations employed. In the selection experiments with ETR and RPV, we were unable to increase drug concentrations at the same rate as with the WT virus when the M184V/N348I double mutant virus was employed. With RPV, there was also a delay in the escalation of drug concentrations with the N348I single mutant virus compared to the WT virus or virus containing M184V.

FIG 1.

Stepwise increment of drug concentrations in cell culture selection experiments. The graph shows the weekly increment of ETR or RPV concentrations for selection of resistant variants in cord blood mononuclear cells (CBMCs) using wild-type (WT) virus and viruses containing the N348I and M184V/N348I substitutions at baseline. CBMCs were infected with viruses at the same multiplicity of infection and cultured in the presence of progressively increasing concentrations of drugs.

Genotyping was performed to verify the emergence of mutations (Table 1). As demonstrated previously (9, 37, 38), E138K was rapidly selected by both ETR and RPV in the case of WT viruses and viruses containing M184V. However, selections of the M184V/N348I double mutant virus by ETR or RPV did not lead to the emergence of resistance substitutions after 19 weeks. The results in Fig. 1 also show that we were unable to increase drug concentrations at the same rate as when WT viruses were employed. These results indicate that M184V/N348I prevented or delayed the emergence of E138K. In addition, our results show that the presence of N348I alone under RPV pressure led to the emergence of substitutions at positions V179 and Y181 but not E138K, thus altering the mutational pathway seen with the WT virus. In addition, we were unable to increase drug concentrations at the same rate as with the WT virus. These selections were performed on three different occasions, with similar results being obtained each time.

TABLE 1.

Mutations emerging in HIV-1 mutants selected with RT inhibitors in CBMCs^a

Virus	Mutationsb
	ETR	RPV
WT	E138K, L100I/L	E138K, L100I
M184V	E138K, M230M/L	A62A/V, K101E/K, E138K
N348I	V179I, Y181C, N348I	L100I, E138K, N348I, E399G
M184V/N348I	V179I/V, N348I	N348I, E399G

^aSelection experiments in CBMCs were performed over a period of 19 weeks.

^bBaseline mutations that persisted at the time of genotypic analysis are underlined.

The N348I substitution together with E138K diminishes viral replication capacity.

The N348I substitution was shown to impair viral replication capacity in a variety of cell types, including HEK293T cells, Jurkat cells, Sup T1 cells (25), MT2 cells, and peripheral blood mononuclear cells (PBMCs) (22). In addition, the E138K mutation impaired HIV-1 replication capacity in a short-term viral replication assay in TZM-bl cells and in a multiple-cycle viral replication assay in CBMCs (37, 38, 44).

We wished to investigate the impact of interactions between N348I or M184V/N348I and E138K on viral replication, and therefore, we performed multiple-cycle viral replication assays in CBMCs. We infected CBMCs with recombinant HIV-1 clones that were normalized in terms of inoculum on the basis of p24 antigen (65). Quantification of virus production at various time points was carried out by measurement of p24 antigen. As shown in Fig. 2A, the relative replication ability of viruses containing E138K alone was impaired compared to that of the WT, while the replication capacity of the E138K/N348I double mutant virus was further decreased compared to that of E138K alone. Interestingly, the addition of M184V to the E138K/N348I double mutant did not further diminish viral replication, and in contrast, a slight compensatory effect on viral replication capacity was observed with E138K/M184V/N348I triply mutated viruses. The order of replication capacity in CBMCs was WT > E138K > E138K/M184V/N348I > E138K/N348I. Thus, N348I further enhanced the replication deficit of viruses containing E138K.

FIG 2.

(A) Comparative analysis of viral replication capacity in CBMCs. Viral stocks of WT HIV-1_NL4.3 and the E138K, E138K/N348I, and E138K/M184V/N348I mutants were harvested by transfection of HEK293T cells and used to infect CBMCs through normalization of the p24 antigen level of the inoculum. Virus growth was monitored at the indicated time points by measuring p24 antigen levels in cell-free culture media by ELISA. Values are means of three independent experiments. Error bars represent standard deviations. (B) Comparative analysis of viral replication capacity in TZM-bl cells. Viral stocks of WT HIV-1_NL4.3 and the E138K, E138K/N348I, and E138K/M184V/N348I mutants were normalized for p24 antigen and used to infect TZM-bl cells. Luciferase activity was measured at 48 h postinfection as an indication of viral replication. The relative infectivity of WT HIV-1_NL4.3 compared to mutant viruses is shown on the y axis, while the x axis denotes the input of p24. Values are means of three independent experiments. Error bars represent the standard deviations. At p24 inputs of 8 × 10⁵ pg/ml and 2.4 × 10⁵ pg/ml, the relative replication of mutant viruses was diminished compared to that of the WT virus. These values translate to a 3-fold virus replication disadvantage for E138K virus and 7- and 5-fold disadvantages for the E138K/N348I and E138K/M184V/N348I viruses, respectively, compared with the WT virus.

Previous results had shown that the replication capacity (RC) of HIV-1 containing E138K was impaired by 3-fold compared to that of the WT virus in TZM-bl cells (37, 38, 44). Thus, we now also employed this cell type and determined the impact of the E138K, E138K/N348I, and E138K/M184V/N348I substitutions on viral RC by infecting TZM-bl cells with serially diluted viral stocks normalized on the basis of p24 antigen. The infectiousness of the WT and mutant viruses was determined by measuring luciferase activity at 48 h postinfection. The results show that the relative replication of all mutant viruses tested was diminished compared to that of the WT virus (Fig. 2B). At p24 inputs of 8 × 10⁵ pg/ml and 2.4 × 10⁵ pg/ml, the relative RC of viruses containing E138K was decreased by ∼3-fold compared to that of the WT, consistent with previous observations (37, 38, 44), while the replication capacities of viruses containing either E138K/N348I or E138K/M184V/N348I were decreased by ∼7-fold and ∼5-fold, respectively. The order of replication capacity in TZM-bl cells was WT > E138K > E138K/M184V/N348I > E138K/N348I, in agreement with the above-described findings in CBMCs. N348I, alone or in combination with M184V, further impaired the viral replication capacity of viruses containing the E138K substitution.

N348I enhances E138K-mediated resistance to ETR and RPV.

It has been shown that HIV-1 containing E138K displays modest phenotypic resistance to the newer NNRTIs ETR and RPV (37, 38, 44, 66), as determined by a high-throughput short-term viral replication assay in TZM-bl cells. Next, we determined the impact of interactions between N348I or M184V/N348I and E138K on susceptibility to the NNRTIs ETR, RPV, EFV, and NVP using this same system. Table 2 shows that E138K conferred modest levels of resistance to ETR (2.2-fold), RPV (2.6-fold), and EFV (1.8-fold) without impacting NVP susceptibility (0.7-fold). Levels of E138K-mediated resistance (fold changes of EC₅₀) were increased by the addition of N348I for ETR (5-fold), RPV (5.6-fold), and EFV (4.1-fold). In addition, susceptibility to NVP was also changed from no resistance to low-level resistance (2.9-fold). These data indicate that the addition of N348I to E138K enhanced resistance by 2.3-fold for ETR and EFV, 2.2-fold for RPV, and 3.8-fold for NVP. The enhancement of E138K resistance against ETR and RPV by N348I was further verified by studies using an in vitro recombinant RT assay, as described below (Table 3). However, the addition of M184V to the E138K/N348I double mutant did not further enhance levels of resistance but rather diminished resistance against ETR (2.9-fold), RPV (2.4-fold), and EFV (2.5-fold) to levels obtained with the E138K single mutant alone while not impacting susceptibility to NVP (2.4-fold).

TABLE 2.

Fold changes in NNRTI susceptibilities for recombinant HIV-1_NL4.3 mutant viruses containing specific mutations, as assessed in TZM-bl cells^a

Virus	Mean EC50 (nM) ± SD (FC)
	ETR	RPV	EFV	NVP
WT	1.7 ± 0.2	0.9 ± 0.1	2.1 ± 0.3	321.6 ± 22
E138K	3.8 ± 0.2 (2.2)	2.3 ± 0.2 (2.6)	3.8 ± 0.7 (1.8)	216 ± 13 (0.7)
E138K/N348I	8.5 ± 1.0 (5.0)b	5.1 ± 0.7 (5.6)b	8.4 ± 0.3 (4.1)b	922.9 ± 51 (2.9)b
E138K/M184V/N348I	5.0 ± 0.7 (2.9)	2.2 ± 0.5 (2.4)	5.2 ± 0.7 (2.5)	778.2 ± 39 (2.4)

^aData represent the mean EC₅₀ (50% drug effective concentration) values (nM) ± standard deviations from 3 independent experiments. Fold changes (FC) in EC₅₀ compared to that of the HIV-1 WT NL_4.3 clone are shown in parentheses. Values in boldface type differ significantly from the values for the WT (P < 0.05, as shown by analysis of variance using Tukey's multiple-comparison test).

^bEC₅₀ values differ significantly from those for E138K (P < 0.05).

TABLE 3.

Susceptibilities to ETR and RPV of recombinant WT and mutated RT enzymes

RT enzyme	Mean IC50 (nM) ± SD (fold change)a
	ETR	RPV
WT	56 ± 13	49 ± 12
E138K	129 ± 16 (2.3)	98 ± 14 (2.0)
E138K/N348I	207 ± 24 (3.7)b	152 ± 22 (3.1)b
E138K/M184V/N348I	157 ± 17 (2.8)	103 ± 15 (2.1)

^aIC₅₀s (50% drug inhibitory concentrations) were determined in recombinant RT assays using a heteropolymeric DNA template and DNA primer. Data represent the means ± standard deviations of 3 independent experiments. Values in parentheses represent the fold changes in IC₅₀ for mutated RTs compared to that of WT RT. Values in boldface type differ significantly from values for the wild type (P < 0.05, as shown by analysis of variance using Tukey's multiple-comparison test).

^bIC₅₀ values differ significantly from those for E138K (P < 0.05).

The 50% inhibitory concentrations (IC₅₀s) of ETR and RPV were determined by in vitro DNA-dependent DNA polymerase assays using purified recombinant RT enzymes and a heteropolymeric DNA template/primer. The results in Table 3 show that E138K alone conferred modest resistance to ETR (2.3-fold) and RPV (2.0-fold). In contrast, the E138K/N348I double mutant conferred a 3.7-fold change in the IC₅₀ for ETR and a 3.1-fold change for RPV, indicating that the addition of N348I to E138K enhanced levels of resistance to ETR and RPV by about 1.6-fold. The addition of M184V to the E138K/N348I double mutant decreased resistance levels against ETR (2.8-fold) and RPV (2.1-fold) to that of the E138K single mutant, consistent with phenotyping data discussed above that were obtained with TZM-bl cells. Similar results were obtained in each of three separate experiments.

N348I diminishes the efficiency of DNA synthesis regardless of the presence of E138K.

Although E138K is known to cause a decrease in RT polymerization efficiency at high dNTP concentrations, the addition of M184I/V can restore the efficiency of processive DNA synthesis (44). To determine whether N348I in RT can compensate for the diminished efficiency in polymerization associated with E138K, we performed RNA-dependent DNA polymerase reactions at high dNTP concentrations (200 μM) in time course experiments for 30 s, 60 s, and 5 min using the WT, E138K, E138K/N348I, and E138K/M184V/N348I recombinant RT enzymes. RT molecules were used at a ∼20-fold excess over the substrate so that any RTs that dissociated from the primer terminus during synthesis would be rapidly replaced and the rate-limiting step would be the addition of nucleotides (44, 67). The efficiency of polymerization was calculated as the number of nucleotide additions divided by the reaction time, and the longest extension products at 60 s (indicated by arrows in Fig. 3) were used to compare the polymerization rates of WT and mutant enzymes.

FIG 3.

Time course experiments showing the effect of the E138K, E138K/N348I, and E138K/M184V/N348I mutations on the efficiency of RT processive DNA polymerization. The ³²P-labeled D25 primer (³²P-D25) was annealed to the 497-nt RNA template, and extension assays were performed with an excess of recombinant RT enzymes at dNTP concentrations of 200 μM. Reactions were stopped at 30 s, 60 s, and 5 min, respectively. All reaction products were resolved by denaturing 6% polyacrylamide gel electrophoresis and visualized by phosphorimaging. The sizes of some fragments of the ³²P-labeled 25-bp DNA ladder (Invitrogen, Burlington, ON, Canada) in nucleotide (nt) bases are indicated on the left side. Positions of the ³²P-labeled D25 primer are indicated on the right. The longest extension products generated at 60 s are identified by arrows and indicate differences in polymerization efficiencies. Experiments were repeated at least twice, with similar results being obtained each time. The figure shows a gel from a representative experiment.

The results show that the WT RT enzyme showed a rate of processive DNA synthesis of ∼3.2 nt/s, consistent with previously reported maximum polymerization rates for WT HIV-1 RT (44, 67, 68). The E138K/N348I double variant had a similarly diminished efficiency of polymerization compared to that of E138K RT, i.e., ∼2.0 nt/s, while the addition of M184V, i.e., E138K/M184V, increased this value to ∼2.6 nt/s. These data confirm that E138K/N348I RT is impaired in processive DNA synthesis in a manner similar to that of RT containing a single E138K substitution. In addition, M184V was able to play a compensatory role in the efficiency of processive DNA polymerization in regard to the E138K/N348I double substitutions. This helps to explain how the E138K/M184V/N348I triple substitution results in higher replication capacity than when only E138K and N348I are present.

N348I-containing enzymes display diminished catalytic efficiency, as shown by steady-state kinetics assays.

Previous results of steady-state kinetic experiments showed that E138K RT has enhanced dNTP-binding affinity (lower K_m value) and can restore the deficit in dNTP usage of RT containing the M184I/V mutations (9, 44). To assess the impact of N348I or M184V/N348I on E138K-containing RT on dNTP-binding affinity and catalytic efficiency, steady-state kinetics studies were performed by using WT RT and the E138K, E138K/M184V, and E138K/M184V/N348I RT variants (9, 44). The results in Table 4 show that RT enzymes containing E138K/N348I and E138K/M184V/N348I variants had K_m values similar to those of WT RT for dTTP but displayed diminished catalytic efficiency (k_cat/K_m), i.e., 34% and 39% of the WT value, respectively. As reported previously, E138K has enhanced dTTP-binding affinity, as demonstrated by a lower K_m value (0.6-fold of the WT value). The order of enzyme catalytic efficiency is WT > E138K > E138K/M184V > E138K/M184V/N348I. These findings showed that the N348I-containing E138K/N348I and E138K/M184V/N348I enzymes display impaired catalytic efficiency, in agreement with data presented above on viral replication capacity and gel-based polymerization efficiency.

TABLE 4.

Kinetics parameters of recombinant RT enzymes, as determined by steady-state kinetics analysis^a

Parameter	Value for RT enzyme
	WT	E138K	E138K/N348I	E138K/N348I/M184V
Mean kcat (min−1) ± SD	22.4.±3.2	11.7 ± 3.0	8.3 ± 1.9	10.1 ± 2.8
Mean Km (μM) ± SD	5.4 ± 1.2	3.2 ± 1.3	6.0 ± 1.7	6.5 ± 2.0
kcat/Km (min−1 μM−1)	4.1	3.7	1.4	1.6

^aThe steady-state kinetics parameters k_cat and K_m for dTTP of recombinant WT HIV-1 RT and its mutant derivatives were determined by using poly(rA)/poly(dT)_12–18 template/primers. The recombinant RT enzymes were purified in a heterodimeric form by immobilized metal affinity chromatography. Values are means ± standard deviations from 3 experiments.

N348I impairs RNase H activity in a background of E138K and E138K/M184V.

Substitutions at certain residues in HIV-1 RT can impair RNase H activity and contribute to reductions in HIV-1 replication fitness (69, 70). In addition, some NNRTI resistance mutations are associated with impaired RNase H activity (69, 71–75), and N348I has been shown to reduce the rate of RNA template degradation (8, 10, 24, 76). Furthermore, reduced RNase H activity may be associated with enhanced NNRTI resistance (77). To investigate the impact of N348I or M184V/N348I on the RNase H activity of E138K-containing RT, we monitored multicycle RNase H-mediated RNA cleavage in time course experiments using WT RT and the E138K, E138K/M184V, and E138K/M184V/N348I RT variants. All of the E138K, E138K/N348I, and E138K/M184V/N348I mutant enzymes yielded the same cleavage profiles as did the WT enzyme but at lower efficiencies, as demonstrated by the relative band densities of the uncleaved RNA substrate and the cleaved products (Fig. 4). These results show that the N348I and M184V/N348I substitutions, when introduced into E138K-containing RT, did not compensate for the diminished RNase H activity of E138K alone and may also contribute to diminished viral replication capacity (Fig. 5).

FIG 4.

RNase H activity of WT and mutant recombinant RT enzymes. (A) Graphic representation of the RNA/DNA (kim40R/kim32D) substrate duplex used to monitor the cleavage efficiency of mutant and WT RTs. Positions of cleavage sites relative to the 3′ end of the primer are shown at the top. The 40-mer RNA kim40R was labeled at its 5′ terminus by ³²P and annealed to 32-mer DNA oligonucleotide kim32D. (B) RNase H activity was analyzed by monitoring substrate cleavage in time course experiments. The time points were 0.2, 0.5, 1, 3, 5, and 10 min (lanes 1 to 6, respectively). The uncleaved substrate and cleaved products relative to the 3′ terminus of the DNA primer are indicated on the left. All reactions were resolved by denaturing 6% polyacrylamide gel electrophoresis. Experiments were repeated at least twice, with similar results being obtained each time. The figure shows a gel from a representative experiment.

FIG 5.

Comparative analysis of enzyme processivity of WT RT and RT enzymes containing specific mutations. The processivity of purified recombinant RT enzymes was analyzed by using a 5′-end-labeled DNA primer (D25) annealed to a 497-nt HIV-1 PBS RNA template as the substrate; the resulting full-length DNA (FL DNA) is 471 nt in length. Processivities were determined by the size distribution of DNA products in fixed-time experiments with 0.5 μM dNTPs in the presence of a heparin trap. All reaction products were resolved by denaturing 6% polyacrylamide gel electrophoresis and visualized by phosphorimaging. Positions of ³²P-labeled primer D25 (³²P-D25) and the 471-nt FL DNA extension products are indicated on the right. Lanes: 1, WT; 2, M184V; 3, E138K/M184V/N348I; 4, E138K/N348I; 5, E138K. Experiments were repeated at least twice, with similar results being obtained each time. The figure shows a gel from a representative experiment.

N348I does not impair enzyme processivity in a background of E138K and E138K/M184V.

Diminished HIV-1 RT processivity is a determinant of impaired viral replication capacity, and both processivity and viral replication can be influenced by the dNTP concentration (78–81). For example, the M184V mutation diminishes dNTP usage and possesses a deficit in enzyme processivity, which can be correlated with lower replication fitness in cell types that have small dNTP pools (78, 82). However, enzyme processivity is not always directly proportional to viral replication capacity, and both E138K and N348I have been demonstrated to compensate for the deficit in processivity associated with M184V (8, 26, 44). To investigate whether interactions between N348I or M184V/N348I and E138K might have an impact on enzyme processivity, we performed gel-based single-cycle processivity assays using recombinant WT RT or RT containing the E138K, M184V, E138K/N348I, and E138K/M184V/N348I substitutions at low dNTP concentrations. Figure 5 shows that only M184V possessed diminished processivity and that all N348I-containing RTs had processivity similar to that of WT RT, as demonstrated by the band density of full-length (FL) DNA products. The E138K mutant enzyme had higher processivity than did WT RT, consistent with previous results (9, 44). These findings show that RTs that contain N348I, i.e., E138K/N348I and E138K/M184V/N348I, do not have diminished enzyme processivity. Therefore, the lower replication capacity of N348I-containing viruses is not apparently due to a defect in RT enzyme processivity.

DISCUSSION

The recent introduction of the newer NNRTIs ETR and RPV into clinical practice and the identification of novel NNRTI-associated mutations have resulted in new patterns of drug resistance (42, 83). ETR and RPV were designed with intrinsic flexibility that allows efficient binding to a high-plasticity NNRTI-BP (84, 85), and they are active against viruses that contain mutations associated with resistance to NVP and EFV (86, 87). Structural studies have revealed similar binding modes within the NNRTI-BP for ETR and RPV (84, 88, 89), indicating cross-resistance. Although three or more mutations are often required to confer high-level resistance to ETR and RPV (36, 39, 90, 91), the E138K substitution can confer cross-resistance to both of these drugs on its own (42). This study is the first to show that the connection domain mutation N348I in HIV-1 RT can prevent or delay the emergence of E138K under ETR or RPV pressure while enhancing the levels of resistance conferred by E138K to these NNRTIs. We have also provided biochemical and virological evidence to show that this antagonism between N348I and E138K is due largely to an exacerbation of the impairment in enzyme and viral fitness associated with E138K and that this deficit cannot be rescued by the further addition of M184V.

The E138K mutation emerges both in vitro and in vivo and confers cross-resistance to ETR and RPV by reducing susceptibility 2- to 3-fold (38–40, 92), but E138K was shown to have no cross-resistance to NVP or EFV (40). In the ECHO and THRIVE clinical trials that led to the approval of RPV for use in treatment-naive patients, E138K was the most frequent NNRTI mutation at treatment failure (36, 87), indicating that E138K is a signature mutation for this drug. In contrast, E138K is much rarer in patients who have failed therapy with ETR, even though this mutation confers resistance to ETR in tissue culture. One reason for this may be that the patients who received ETR were almost all treatment experienced and harbored the N348I substitution that, either alone or in combination with M184V, can prevent or delay the emergence of E138K under ETR and RPV drug pressure, as shown here. It is notable that Y181C is also antagonistic with E138K and can prevent the emergence of E138K in cell culture under ETR or RPV pressure (38, 48, 61). Y181C was present in the DUET studies conducted on the efficiency of ETR in treatment-experienced subjects in 32% of patients at baseline (90); this explains the low prevalence of E138K in these trials. The present study makes clear that N348I can also delay the emergence of E138K.

M184V confers high-level resistance to 3TC and FTC and can impair viral fitness (31, 32) while delaying the development of resistance to EFV (93) or hypersensitizing to NNRTIs (94, 95). M184I/V was also common, together with E138K, in patients who failed combination antiretroviral FTC/TDF/RPV therapy in the ECHO and THRIVE clinical studies (36), due to mutual compensatory effects between the E138K and M184I/V substitutions in regard to enzyme processivity, polymerization efficiency, and viral replication capacity (9, 43, 44, 96). A strong association between N348I and M184V has also been observed, and the selection of N348I is associated primarily with the use of ZDV and NVP (8, 22, 24). Biochemical and virological data have demonstrated that N348I has a compensatory effect on M184V in regard to enzyme processivity and viral replication capacity (25, 26). Previous studies showed that M184I/V at baseline did not impact the emergence of E138K in CBMCs under ETR or RPV selection pressure (9, 38). However, we now show that N348I can restrict the emergence of E138K, even though the addition of M184V to N348I does not reverse this antagonism. These findings help to explain why E138K did not emerge spontaneously in patients who possessed M184V in order to compensate for the fitness deficits associated with the latter mutation.

We have recently shown that E138K consistently emerges in CBMCs under ETR selection pressure but not in MT2 cells that have larger dNTP pools (37, 48). We now also demonstrate that the addition of N348I to E138K further decreased viral replication in both cell types and that the addition of M184V to the E138K/N348I double mutant slightly increased viral replication capacity compared to E138K/N348I alone. These data help to explain the impact of N348I, alone or in combination with M184V, on the emergence of E138K.

The presence of E138K alone, and the combination of E138K plus N348I, resulted in diminished efficiency of polymerization in a gel-based RNA-dependent DNA polymerase assay. The negative effect of N348I on polymerization was further confirmed by steady-state kinetic analyses, which showed that N348I further diminished the catalytic efficiency of E138K without significantly impacting dNTP-binding affinity. The addition of N348I or M184V/N348I to E138K did not impact enzyme processivity. Also, neither N348I nor E138K alone impaired enzyme processivity, and both were able to restore the diminished processivity of M184V (26, 44). Although we used steady-state kinetics rather than a pre-steady-state kinetic analysis, our data are in agreement with the results of gel-based RNA-dependent DNA polymerase assays, which show that the lower replication capacity associated with E138K/N348I and E138K/M184V/N348I is due mainly to a deficit in enzyme catalytic efficiency.

Our data also indicate that N348I can enhance resistance to ETR and RPV in a background of E138K. Previous studies showed that N348I can increase resistance to ETR (29, 97, 98). Now, we also show that N348I can enhance resistance to both ETR and RPV if E138K is also present.

In addition to the nonemergence of E138K in patients who first possessed the M184IV mutation, a related issue is why E138K did not appear in tissue culture selections performed exclusively with 3TC or FTC as a means of increasing viral fitness. Part of the answer may lie in the fact that the M184I/V substitutions result in diminished viral replication capacity. A second reason may be that M184V is known to result in increased RT fidelity (31, 32), also making it less likely that a subsequent spontaneous mutation might emerge. In order to provide further information on this topic, we are assessing the possible presence of E138K in 3TC-selected cultures using an allele-specific PCR assay, which is more sensitive than the population-based genotyping carried out to date.

As N348I is not located at the NNRTI-BP, it is not clear how N348I enhances the resistance levels conferred by E138K to ETR and RPV. Transit-kinetic analysis has shown that E138K resistance to RPV is due mainly to an enhanced dissociation rate of RPV (11). N348I also confers resistance to NVP through decreased binding affinity (8) while inhibiting RNase H activity (10, 77, 99). Further biochemical and structural analyses may lead to the design of better NNRTIs with improved resistance profiles and potency. Our study also supports the idea that the connection domain should be monitored in routine HIV resistance genotyping, because of its potential to enhance levels of drug resistance.

In summary, we have shown that the N348I mutation alone or in combination with M184V can prevent or delay the emergence of the E138K substitution by diminishing viral replication capacity. We have also provided mechanistic insights into the reasons for this decreased replication capacity. The copresence of N348I and E138K impairs RT catalytic efficiency and also diminishes RNase H activity, while the addition of N348I to E138K enhances resistance to both ETR and RPV. Ours is also the first study to show that a connection domain mutation can have an antagonistic interaction with a NNRTI RAM in terms of enzymatic function and viral replication.