Antiviral Activity and In Vitro Mutation Development Pathways of MK-6186, a Novel Nonnucleoside Reverse Transcriptase Inhibitor

Abstract

MK-6186 is a novel nonnucleoside reverse transcriptase inhibitor (NNRTI) which displays subnanomolar potency against wild-type (WT) virus and the two most prevalent NNRTI-resistant RT mutants (K103N and Y181C) in biochemical assays. In addition, it showed excellent antiviral potency against K103N and Y181C mutant viruses, with fold changes (FCs) of less than 2 and 5, respectively. When a panel of 12 common NNRTI-associated mutant viruses was tested with MK-6186, only 2 relatively rare mutants (Y188L and V106I/Y188L) were highly resistant, with FCs of >100, and the remaining viruses showed FCs of <10. Furthermore, a panel of 96 clinical virus isolates with NNRTI resistance mutations was evaluated for susceptibility to NNRTIs. The majority (70%) of viruses tested displayed resistance to efavirenz (EFV), with FCs of >10, whereas only 29% of the mutant viruses displayed greater than 10-fold resistance to MK-6186. To determine whether MK-6186 selects for novel resistance mutations, in vitro resistance selections were conducted with one isolate each from subtypes A, B, and C under low-multiplicity-of-infection (MOI) conditions. The results showed a unique mutation development pattern in which L234I was the first mutation to emerge in the majority of the experiments. In resistance selection under high-MOI conditions with subtype B virus, V106A was the dominant mutation detected in the breakthrough viruses. More importantly, mutant viruses selected by MK-6186 showed FCs of <10 against EFV or etravirine (ETR), and the mutant viruses containing mutations selected by EFV or ETR were sensitive to MK-6186 (FCs of <10).

INTRODUCTION

Human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT) plays an essential role in the life cycle of HIV-1, converting single-stranded viral RNA into double-stranded proviral DNA via polymerase and RNase H activities (15). Therefore, the inhibition of RT has been one of the primary therapeutic strategies in suppressing the replication of HIV-1 (9, 27). In addition to RT active-site inhibitors, nucleoside reverse transcriptase inhibitors (NRTIs; such as zidovudine [AZT] and lamivudine [3TC]), nonnucleoside reverse transcriptase inhibitors (NNRTIs; such as efavirenz [EFV], nevirapine [NVP], etravirine [ETR], and rilpivirine [RPV]) are non-active-site competitive inhibitors of HIV-1 RT that bind to a hydrophobic pocket in the p66 subunit of the p66/p51 heterodimer at a distance of 10 Å from the polymerase active site (20). NNRTI binding causes conformational changes within p66 that reposition the active-site residues to an inactive conformation, thereby inhibiting the chemical step of polymerization (1). Mutations identified in the viruses from patients who failed an NNRTI-containing regimen mostly are residues around the NNRTI binding pocket (NNRTIBP) that play important roles in NNRTI binding. As a result, viruses harboring mutations in these residues often disrupt the crucial interactions between NNRTIs and RT, thus conferring resistance to the inhibitors. For instance, K103 is located at the entrance of the NNRTIBP, and viruses with the K103N mutation (K103 to N103; present in 40 to 60% of NNRTI-resistant viruses) (22, 42) display significant resistance to older NNRTIs (EFV and NVP). In addition, Y181 plays an important role in the π-π interactions with NNRTIs inside the NNRTIBP, thus viruses with a Y181C mutation (present in 15 to 25% of NNRTI-resistant viruses) exhibit high degrees of resistance to NVP and delavirdine and moderate resistance to ETR (40).

The current standard of treatment for HIV-1-infected patients is highly active antiretroviral therapy (HAART), which is typically composed of three or more drugs with complementary mechanisms of action (29). Patients undergoing HAART have experienced profound and continuous viral suppression, in many cases with substantial immune system recovery and the halt of progression to clinical disease (32). Consensus guidelines for the use of HAART in antiretroviral-naive subjects recommend the use of two NRTIs in combination with an NNRTI, a ritonavir-boosted protease inhibitor, or an integrase inhibitor (13, 17). Although NNRTIs can be key components of effective combination regimens, like all antiretroviral agents their effectiveness can be hampered by the emergence of resistance mutations in viruses. Moreover, a single mutation can lead to significant reductions in susceptibility, often to all available inhibitors within the same class (4, 12). Therefore, extensive efforts have been made to identify novel NNRTIs that are highly active against the prevalent NNRTI-resistant viruses (such as viruses containing the K103N or Y181C mutation) and suitable for a once-daily dosing with excellent safety profiles. These efforts have led to the approval of two NNRTIs (ETR and RPV) and the discovery of several promising NNRTIs that are currently in clinical development (37).

The error-prone nature of the HIV-1 RT enzyme, coupled with the lack of exonuclease proof-reading activity, results in a large genetic diversity among viral subtypes in different geographic regions. In addition, recombination between subtypes further contributes to genetic diversification (8). Currently, nine distinct viral subtypes (A, B, C, D, F, G, H, J, and K) are expanding across the globe (18). Different virus subtypes therefore might display various levels of susceptibility to antiviral agents resulting from minor variations in binding pockets (24). The frequency and pattern of mutations that confer resistance to antiviral agents may differ among HIV-1 subtypes following treatment with antiviral agents, thus virus subtypes from some geographical regions may have a greater propensity to develop a resistance mutation(s) against certain drugs than do other viral variants (36). On a global scale, the most prevalent HIV-1 subtypes are C, A, and B, accounting for 47, 27, and 12% of HIV-1 infected patients, respectively (3). However, despite the fact that HIV-1 subtype B viruses account for only 12% of the global HIV pandemic (39), the antiretroviral drugs used to treat HIV-1 infection have been developed mainly based on in vitro studies with subtype B isolates, and most data on the genetic mechanisms of HIV-1 drug resistance have been obtained from subtype B viruses. Some in vitro and in vivo observations, however, suggest that the various subtypes respond differently to certain antiretroviral drugs (5–7, 16, 23), and these differences could influence therapeutic outcomes (19). In addition, differences in replication capacity or fitness may exist among various HIV subtypes, and these may become magnified under the suppression pressure of drug treatment (36). Therefore, identifying the relevant drug resistance mutations among non-B viral subtypes will be important for monitoring the evolution and transmission of drug resistance, for determining initial treatment strategies for persons infected with non-B viruses, and for interpreting genetic resistance among non-B patients who are unresponsive to antiretroviral therapy.

MK-6186 is a novel NNRTI containing a chlorobenzonitrile and two indazole rings (Fig. 1) (14). MK-6186 was identified as a potential new NNRTI that displays excellent activities against not only wild-type (WT) viruses but also a broader panel of NNRTI-resistant viruses, including viruses with K103N and/or Y181C mutation. In addition, MK-6186 also showed good pharmacokinetic profiles in rats and dogs and offers the potential for once-daily dosing in humans (14). Given the prevalence of HIV-1 A, B, and C subtypes, we conducted in vitro resistance selection studies with MK-6186 in these subtypes under low-multiplicity-of-infection (MOI) conditions. The resistance selection with subtype B virus was also performed under high-MOI conditions at fixed concentrations of the inhibitors in the presence of 50% human serum (HS) to account for high viral load and protein content at the beginning of antiretroviral therapy (40).

Fig 1.

Structures of MK-6186 (A), EFV (B), and ETR (C). Letters a, b, and c represent ring moieties in MK-6186 and ETR.

Our results indicate that MK-6186 is highly potent against the majority of NNRTI-associated mutant viruses and different HIV-1 subtype viruses. Moreover, viruses treated with MK-6186 displayed unique mutation development patterns in which L234I was the initial mutation identified in most of experiments under low-MOI conditions. The V106A mutation, on the other hand, was the dominant mutation identified in breakthrough viruses (BTV) from resistance selection under high-MOI conditions. Moreover, mutations selected by MK-6186 were highly susceptible to other NNRTIs and vice versa.

MATERIALS AND METHODS

Materials.

Full-length WT and two mutant RT proteins (K103N and Y181C) derived from strain NL_4-3 were expressed in Escherichia coli BL21(DE3) cells and purified as described previously (34). The t500 RNA template was made by IBA BioTAGnology (Germany), and the biotinylated DNA primer was made by Integrated DNA Technology (IDT; Coralville, IA). The R8 virus (NL_4-3 strain) was a kind gift from Christopher Aiken (Vanderbilt University, Nashville, TN). SupT1 cells were provided by the NIH AIDS Research and Reference Reagent Program. Ruthenylated dUTP (Ru-dUTP) was custom-made by Midland Certified Reagents Company (Midland, TX). Electrochemiluminescence (ECL) detector M-384 and streptavidin-coated magnetic beads were purchased from BioVeris (Gaithersburg, MD). Culture medium (RPMI 1640) and Dulbecco's modified Eagle medium (DMEM)–10% fetal bovine serum (FBS) were from Gibco (Carlsbad, CA). Wallac Microbeta and Victor luminometers were from PerkinElmer (Wellesley, MA). Microplates (384 wells) were purchased from Falcon (Franklin Lake, NJ). The QIAamp kit was from Qiagen (Valencia, CA). The BigDye Terminator kit and ABI 3100 Analyzer were supplied by Applied Biosystems (Foster City, CA). ViroMag R/L magnetic beads were ordered from Boca Scientific (Boca Raton, FL). The Dynal MPC-S magnetic particle concentrator, RPMI 1640 medium, SOC medium, and plasmid pcDNA 3.1(+) were purchased from Invitrogen (Carlsbad, CA). BioStor vials (1.5 ml) were acquired from National Scientific Supply (Claremont, CA). The JM109 Competent Cell Access reverse transcription-PCR (RT-PCR) system and the pGEM-T EasyVector system II were obtained from Promega (Madison, WI). The Guava instrument was from Millipore (Billerica, MA). The QuikChange SDM kit was from Stratagene (La Jolla, CA).

HIV-1 reverse transcriptase biochemical assay.

The ECL RT biochemical assay was performed based on a protocol described previously, with minor modifications (28). Briefly, HIV-1 RT enzyme (10 pM) was combined with an inhibitor or dimethyl sulfoxide (DMSO) (10%) in assay buffer (50 mM Tris-HCl, pH 7.8) followed by incubation at room temperature for 30 min. The polymerization reaction was initiated by the addition of a biotinylated template/primer substrate (5 nM) and deoxynucleoside triphosphates (dNTPs) (0.6 μM each dATP, dGTP, and dCTP and 0 nM Ru-dUTP). The reactions were continued for 90 min at 37°C, followed by the addition of EDTA (10 mM) to terminate the reaction. The resulting solution was incubated at room temperature for an additional 30 min. Streptavidin-coated magnetic beads (80 μg/ml) were added to capture the product and unreacted substrate for detection. The quantification of the product was determined based on the ECL signal as measured by ECL instrument M-384.

HIV-1 single-cycle replication assay.

The single-cycle assay was performed as described previously (24). Briefly, P4/R5 cells containing the ß-galactosidase gene with the long terminal repeat (LTR) as the promoter were detached from the plate and placed in a 384-well plate at a density of 1,000 cells per well. Cells were plated in the presence of medium (40 μl) containing DMEM–10% FBS–1% penicillin and streptomycin. After incubation at 37°C and 5% CO₂ overnight, 20 μl medium was removed from the cells and discarded. Medium with viruses was added to a 384-well plate containing the inhibitor to generate a mixture composed of a 2× concentration of viruses and inhibitor. The resulting mixture (20 μl) was added back to the cells described above, resulting in a 1× concentration of both virus and inhibitor. The mixture then was incubated for an additional 48 h, allowing a single cycle of viral replication. The level of viral replication was assessed by the expression of ß-galactosidase based on the activity assay.

HIV-1 single-cycle replication assay with viruses containing RT sequences obtained from patient isolates.

Monogram Biosciences performed an HIV-1 single-cycle replication PhenoScreen assay with MK-6186, EFV, and ETR against a broad array of clinically relevant NNRTI-resistant viruses using a variation of their clinical diagnostic assays described previously (30).

HIV-1 multiple-cycle replication assay.

MK-6186 was tested in cell culture HIV-1 multiple-cycle replication assays using genetically defined wild-type and three laboratory mutant viruses (K103N, Y181C, and K103N/Y181C) as described previously (25). Antiviral activity assays were performed using variants of a laboratory HIV-1 isolate, R8, and MT-4 human T-lymphoid cells in cell culture medium supplemented with 10% FBS or 50% HS.

Resistance selection with subtype B virus under low-MOI conditions.

Resistance selection with NNRTIs under low-MOI conditions was conducted using the same protocol as that described before (24). Briefly, SupT1 cells were grown to log phase, followed by centrifugation, and we resuspended the pellets in growth medium to a density of 2 × 10⁶/ml. The cells were infected with R8 virus at a 50% tissue culture infective dose (TCID₅₀) of 10^5.67/ml. The infected cells were incubated at 37°C in a 5% CO₂ incubator overnight. The cells were subsequently treated with NNRTI at an initial concentration of 1× EC₉₅. Breakthrough viruses (BTV) were harvested and stored at −80°C or immediately subjected to genetic analysis. BTV from the previous round were employed for the subsequent round of selection at a higher concentration of inhibitor (up to 50× EC₉₅).

Resistance selection with subtype A and C viruses under low-MOI conditions.

The resistance selection procedure with subtype A and C viruses was similar to that used for resistance selection with subtype B virus, except that SupT1 cells containing CCR5 receptors were employed for the viral infection. Subtype A (patient RW/92/026) and subtype C (patient 93MW 959) viruses were used for the resistance selection study (24).

Resistance selection under high-MOI conditions.

Under high-MOI conditions, cell infection was determined to be greater than 50% using a procedure described previously (24). Briefly, SupT1 cells were infected with 200 μl of R8 virus, and the resulting infected cells were treated with the NNRTI at three different concentrations (200 nM, 500 nM, and 1.25 μM) in the presence of 50% human serum. When the majority of the cells formed syncytia (>50%), the medium containing the BTV was harvested and stored at −80°C or subjected to immediate genetic analysis.

Identification of resistance mutation(s) in BTV from resistance selection.

DNA sequencing was performed as described previously (24). Briefly, viral RNA of BTV was isolated using a QIAamp kit by following the manufacturer's instructions. The entire RT region of the cDNA sequence was generated via the Access RT-PCR system using the primers AGGTACAGTGTTGGTAGGACCTAC and TGTTCAGCTTGGTCTCTTACCTG for subtype A virus, TTAAAGCCCGGGATGGATGGCCCAAAAGTTAAAC and GTACTTTTCGAATGCCAGCACTGACCAATTTATC for subtype B virus, and ACCTGTCAACATAATTGGAAGAAAT and ATACTTCTCATGCTCTTCTTGAG for subtype C virus. The synthesized cDNA was purified by agarose gel electrophoresis. Population sequencing was performed with the purified cDNA using the BigDye termination kit, and the sequences were resolved on an ABI 3100 Analyzer. Several sequencing primers for each virus subtype were designed and employed to identify mutations in the BTV obtained from resistance selection in cell culture.

Clonal sequencing of the RT DNA sequences of the BTV obtained from resistance selection in cell culture.

Clonal sequencing was conducted as described previously (24). More than 20 white colonies were selected for culturing overnight, and viral DNA was purified and subjected to genetic analysis.

Isolation of HIV-1 subtype A viral particles with ViroMag R/L beads.

BTV resulting from resistance selection with subtype A viruses in the presence of NNRTIs were isolated as described previously (41). Briefly, 100 μl of the medium containing BTV was diluted with 900 μl of serum-free RPMI in a 1.5-ml BioStor vial, followed by the addition of 100 μl of ViroMag R/L beads. The vial was inverted gently back and forth at room temperature for 10 min and placed into a Dynal MPC-S magnetic particle concentrator, allowing all of the beads to attach to the inner wall of the tube. After removing the solution, the beads were resuspended and washed with fresh serum-free RPMI. The beads were finally resuspended in a starting volume of serum-free RPMI and were either used immediately for infection experiments or stored at −80°C for future use.

SDM.

Mutant viruses that contained the desired mutation(s) were generated by introducing the respective mutation into the WT HIV-1 strain HXB2 backbone using the QuikChange site-directed mutagenesis (SDM) kit. The resulting proviral DNA construct was used to transfect 293T cells, and culture supernatants containing the SDM-derived mutant viruses were collected and concentrated and their titers determined prior to the infection experiments (24).

RESULTS

Biochemical inhibitory potency and antiviral activity of MK-6186.

RT containing one of the prevalent NNRTI resistance mutations, K103N or Y181C, was employed to screen all newly synthesized compounds targeting the NNRTIBP. NNRTIs that displayed excellent potencies against the mutant RTs with minimum shifts in potency compared to WT RT were selected for further evaluation (14). As shown in Table 1, MK-6186 displayed subnanomolar potency in the inhibition of RNA-dependent DNA polymerase activity, with 50% inhibitory concentrations (IC₅₀s) of 0.35, 0.60, and 0.58 nM versus WT, K103N, and Y181C RT virus, respectively. These results indicate that MK-6186 is highly active against the two most prevalent RT mutants, with <2-fold shifts in potency compared to WT RT, and K103N and Y181C RTs showed FCs of 2 and 3, respectively, when tested with ETR. The control compound EFV displayed a more than 20-fold shift in potency against the RT containing K103N and no shift against the RT containing the Y181C mutation, as expected (11).

Table 1.

Inhibitory potency of MK-6186, EFV, and ETR against WT, K103N, and Y181C RT viruses

Compound	Inhibitory potency againsta:
	WT		K103N		Y181C
	IC50 (nM)	FC	IC50 (nM)	FC	IC50 (nM)	FC
MK-6186	0.35 ± 0.13 (22)	1.0	0.60 ± 0.17 (21)	1.7	0.58 ± 0.15 (21)	1.7
EFV	0.42 ± 0.06 (5)	1.0	10.2 ± 6.6 (4)	24	0.31 ± 0.17 (5)	0.7
ETR	0.60 ± 0.12 (4)	1.0	1.3 ± 0.17 (4)	2.2	1.96 ± 0.9 (3)	3.3

^aIC₅₀s are means ± standard deviations. FC, fold change compared to WT levels. Numbers in parentheses indicate the number of assays performed.

The antiviral activity of MK-6186 was evaluated in a multiple-cycle replication assay in the presence of 10% FBS and 50% HS. As shown in Table 2, in the presence of 10% FBS, the EC₉₅s of MK-6186 were 13, 16, 60, and 109 nM against WT, K103N, Y181C, and K103N/Y181C viruses, respectively. Viruses harboring the K103N or K103N/Y181C mutation showed more than 40-fold resistance to EFV. In contrast, viruses containing the Y181C mutation were highly susceptible to EFV, with almost no shift in potency, which is consistent with results from the biochemical assay. K103N viruses did not exhibit any resistance to ETR, whereas Y181C viruses conferred >5-fold resistance to ETR. Overall, MK-6186 was approximately 2- to 4-fold less potent than ETR in the inhibition of WT and the three mutant viruses.

Table 2.

Antiviral activity of MK-6186, EFV, and ETR against WT, K103N, Y181C, and K103N/Y181C viruses in the presence of 10% FBS and 50% HS

Compound and serum	Antiviral activity againsta:
	WT		K103N		Y181C		K103N/Y181C
	EC95 (nM)	FC	EC95 (nM)	FC	EC95 (nM)	FC	EC95 (nM)	FC
10% FBS
MK-6186	13 ± 2.6 (29)	1.0	16 ± 4.1 (17)	1.2	60 ± 24 (16)	4.6	109 ± 38 (16)	8.4
EFV	6.3 ± 3.5 (190)	1.0	263 ± 55 (153)	42	8.3 ± 3.1 (42)	1.3	289 ± 136 (41)	46
ETR	4.3 ± 2.0 (33)	1.0	4.3 ± 2.3 (26)	1.0	23 ± 12 (11)	5.3	57 ± 21(10)	13
50% HS
MK-6186	74 ± 22 (27)	1.0	112 ± 34 (10)	1.5	342 ± 77 (12)	4.6	697 ± 107 (7)	9.4
EFV	50 ± 29 (176)	1.0	1422 ± 543 (21)	28	79 ± 24 (22)	1.6	3113 ± 693 (12)	62
ETR	35 ± 10 (41)	1.0	37 ± 11(27)	1.1	234 ± 82 (16)	6.7	661 ± 217 (21)	19

^aIC₅₀s are means ± standard deviations. FC, fold change compared to WT levels. Numbers in parentheses indicate the number of assays performed.

In the presence of 50% HS, the concentrations of compounds that were required to suppress 95% of viral replication generally were higher due to protein binding. As shown in Table 2, MK-6186 exhibited an approximately 5- to 7-fold shift in potency as serum was changed from 10% FBS to 50% HS. The antiviral potency of ETR was reduced approximately 10-fold when tested in the presence of 50% HS. As a result, MK-6186 displayed antiviral potency similar to that of ETR when tested against viruses harboring the Y181C and K103N/Y181C mutations, and it was 2- to 3-fold less potent than ETR against WT and K103N viruses in the presence of 50% HS. Under the same conditions, EFV displayed approximately 6- to 10-fold shifts in potency as serum was changed from 10% FBS to 50% HS. EFV showed an EC₉₅ of 50 nM, 1.4 μM, 79 nM, and 3.1 μM in suppressing the replication of WT, K103N, Y181C, and K103N/Y181C viruses, respectively. As a result, EFV was the most potent inhibitor against Y181C viruses. Compared to WT virus, MK-6186 displayed FCs of 1.5, 4.6, and 9.4 in the EC₉₅ against K103N, Y181C, and K103N/Y181C viruses, respectively. The Y181C mutation appeared to have a greater impact on the susceptibility of viruses to ETR, because viruses containing the Y181C and Y181C/K103N mutations exhibited 6.7- and 19-fold shifts in potency, respectively, compared to the WT. K103N viruses, however, were as susceptible as WT virus to inhibition by ETR. In general, compared to the WT, the relative FCs in the potency of the NNRTIs against HIV variants were similar in the presence of either 10% FBS or 50% HS.

Although there was almost no shift in IC₅₀ when MK-6186 was tested with Y181C RT, there was an approximately 5-fold shift in EC₉₅ when the compound was tested with the Y181C virus. The discrepancy in FCs between biochemical- and cell-based assays with Y181C variants has also been observed with another NNRTI (25). The cause of this discordance remains to be elucidated.

Resistance profile of MK-6186.

A panel of 12 prevalent EFV-resistant viruses was employed to assess the mutant profiles of MK-6186, EFV, and ETR in a single-cycle replication assay by Monogram. Among 12 resistant mutant viruses tested, only two relatively rare mutant viruses (Y188L and V106I/Y188L) displayed more than a 100-fold shift in potency when treated with MK-6186 (Fig. 2). On the other hand, eight mutant viruses showed FCs of greater than 100 with EFV. For instance, viruses harboring the K103N/P225H double mutation exhibited greater than 200-fold resistance to EFV but only 2-fold resistance to MK-6186. Furthermore, the other mutant viruses that showed less than 100-fold resistance to EFV also were more susceptible to MK-6186. One example is that viruses containing the K101E mutation had an FC of 23 with EFV but a FC of 7.8 with MK-6186. All of the EFV-resistant mutant viruses were susceptible to ETR, with FCs of <10. Although Y188L mutant viruses exhibited significant resistance to MK-6186 (FC of >100), the mutation was not selected by MK-6186 during in vitro resistance selection as described below. Moreover, Y188L also showed greater than 100-fold resistance to EFV, and it was only observed in 5% of the patients who failed in treatment with an EFV-containing regimen. The lack of a prevalent Y188L mutation may have been because two base changes are required to generate the Y188L mutation (24, 35).

Fig 2.

Susceptibility of mutant viruses containing NNRTI-associated mutations to NNRTIs MK-6186 (A), EFV (B), and ETR (C). An asterisk indicates that the EC₅₀ was not reached at the highest concentration tested. CNDO is a drug-sensitive reference standard that is used to determine the fold changes in drug susceptibility of patient samples. The EC₅₀s against CNDO were 0.65, 2.0, and 5 nM for EFV, ETR, and MK-6186, respectively. MDRC4 is a multidrug-resistant virus control that is used to evaluate and monitor assay performance. The fold change was defined as the ratio of EC₅₀s between mutant virus and CNDO. The EC₅₀ was obtained from a single-compound titration (n = 1).

An expanded panel of 96 NNRTI-associated clinical isolates was also tested for their susceptibility to MK-6186, EFV, and ETR under the same assay conditions. There were 28 (29%) and 19 (20%) of the 96 mutant viruses that displayed greater than 10-fold resistance to MK-6186 and ETR, respectively (Fig. 3A and C). In contrast, the majority (67; 70%) of the mutant viruses showed greater than 10-fold shifts in EC₅₀ compared to the reference strain when tested against EFV (Fig. 3B). Among them, 48% (46 mutant viruses) exhibited greater than 100-fold shifts in potency. Among the 96 viruses, more than 50 viruses displayed <5-fold resistance to MK-6186 (Fig. 3A). The majority (86%) of the mutant viruses displaying reduced susceptibility to MK-6186 (>10-fold shift in EC₅₀) also showed resistance to EFV (FC of >10), whereas only 28% of MK-6186-resistant viruses had reduced susceptibility to ETR.

Fig 3.

Susceptibility of a panel of 96 clinical NNRTI-associated mutant viruses to NNRTIs. (A) MK-6186; (B) EFV; (C) ETR. The fold change was defined as the ratio of EC₅₀s between mutant virus and CNDO. The EC₅₀ was obtained from a single-compound titration (n = 1).

Among 67 mutant viruses that were resistant to EFV (FC of >10), more than 30 of them were sensitive to MK-6186 (FC of <5), and only 8 mutant viruses showed greater than 100-fold resistance to MK-6186 (Fig. 4A). Among 19 mutant viruses that were resistant to ETR, 11 of them (58%) also exhibited greater than 10-fold shifts in EC₅₀ when tested with MK-6186, and 8 of them displayed FCs of <10 (Fig. 4B).

Fig 4.

Susceptibility of EFV and ETR resistance viruses to MK-6186. (A) EFV; (B) ETR. The fold change was defined as the ratio of EC₅₀s between mutant virus and CNDO. The EC₅₀ was obtained from a single-compound titration (n = 1).

Susceptibility of different HIV-1 subtypes to NNRTIs.

A total of 93 WT viruses in 10 different HIV-1 subtypes were evaluated for their susceptibility to MK-6186, EFV, and ETR (Fig. 5). It appears that the WT viruses were slightly more susceptible to MK-6186 than EFV. Interestingly, subtype H was highly sensitive to all three NNRTIs, with 5-fold higher susceptibility. This phenomenon has been reported with another NNRTI (24). The difference in the amino acids that reside in the NNRTI binding pocket between subtype H and other subtypes might contribute to the hypersensitivity. The comparison of RT X-ray structures between subtype H and other subtypes may shed light on understanding the differences in susceptibility and could further aid in elucidating the binding mechanism of NNRTIs.

Fig 5.

Susceptibility of HIV-1 subtypes to NNRTIs. (A) MK-6186; (B) EFV; (C) ETR. The fold change was defined as the ratio of EC₅₀s between each subtype virus and CNDO. The EC₅₀ was obtained from a single-compound titration (n = 1). The number of different samples tested in different subtypes are the following: A (n = 4), A1 (n = 13), AE (n = 5), AG (n = 18), BF (n = 4), C (n = 22), D (n = 9), G (n = 8), H (n = 2), and B (n = 7).

In vitro resistance selection with MK-6186 under low-MOI conditions in HIV-1 subtypes A, B, and C.

Given the diversity of HIV-1 subtypes described above, in vitro resistance selection was performed under low-MOI conditions with 1 viral isolate from each of HIV-1 subtypes A, B, and C. The resistance selection experiments were started at concentrations of 1× EC₉₅ and sequentially escalated up to 50× EC₉₅. The BTV from the prior round of selection were employed for subsequent selection at higher concentrations of each compound. Resistance selection for subtypes A and B was conducted in triplicate and subtype C in duplicate.

Resistance selection with MK-6186 under low-MOI conditions, as shown in Table 3, revealed three pathways for mutation development in subtype A viruses. No mutation was detected in BTV until the MK-6186 concentration was increased to 5× EC₉₅ in the first experiment, and Y181C was the mutation responsible for viral breakthrough at 5× and 10× EC₉₅. There was not enough BTV isolated for DNA sequence analysis in the selection experiment at 50× EC₉₅. A rare NNRTI-associated mutation, L234I, was identified in the second experiment at a concentration of 1× EC₉₅, and the mutation remained the only mutation selected throughout the experiment. For pathway three, F227C was the sole mutation detected in the BTV from the selection with MK-6186 at 2× EC₉₅ and persisted as the only mutation responsible for the viral breakthrough at higher concentrations of MK-6186.

Table 3.

Resistance selection with MK-6186 in different HIV-1 subtypes in the presence of 10% FBS under low-MOI conditions^a

Expt no. and subtype	Resistance selection at:
	1× EC95		2× EC95		5× EC95		10× EC95		50× EC95
	Time (days)	Mutation(s)	Time (days)	Mutation(s)	Time (days)	Mutation(s)	Time (days)	Mutation(s)	Time (days)	Mutation(s)
A
1	19	NM	20	NM	15	Y181C	12	Y181C		NA
2	23	L234(L/I)	10	L234(L/I)	7	L234(L/I)	8	L234(L/I)	8	L234(L/I)
3	19	NM	10	F227(C/F)	9	F227C	8	F227C	7	F227C
B
1	16	E138(K/E), Y181(Y/C)	7	E138(K/E), Y181(Y/C)	9	E138(K/E), Y181(Y/C), V106(A/V)	8	E138(K/E), Y181(Y/C), V106A	8	E138(K/E), Y181(Y/C), V106A, L234(L/I)
2	17	L234I	11	L234I	33	L234I V108(I/V)	13	L234I, V108(I/V), V106(A/V)	8	L234I, V108I V106A
3	17	L234I	9	L234I	13	L234I	11	L234I, Y181(C/Y)	8	L234I, V106A
C
1	10	L234(L/I)	6	L234I	5	L234I, V106(A/V)	5	L234I, V106(A/V)	5	L234I, V106(A/V)
2	10	L234(L/I)	6	L234I	5	L234I, V106(A/V)	5	L234I, V106(A/V)	5	L234I, V106(A/V)

^aThe EC₉₅ of MK-6186 is 13 nM. Three independent experiments were performed with subtype A and B viruses and two independent experiments were performed with subtype C virus in the resistance selection. NM, no mutation detected; NA, not available. Amino acids in parentheses indicate a mixture of wild-type residues and the emerging mutation.

In resistance selection studies with subtype B virus, E138K and Y181C were the first two mutations detected in experiment one at 1× EC₉₅. Subsequently, another known NNRTI-associated mutation, V106A, emerged at concentrations of 5× EC₉₅ and 10× EC₉₅, followed by the addition of the L234I mutation at 50× EC₉₅. For experiments two and three, L234I was the first mutation selected in the BTV at 1× EC₉₅ and was present in the BTV through the entire experiments. In experiment two, the NNRTI resistance mutation V108I was selected at 5× EC₉₅. An additional mutation, V106A, emerged at 10× EC₉₅, and all three of these mutations were also identified in the BTV from the selection at 50× EC₉₅. In the third experiment, the Y181C mutation was selected at 10× EC₉₅ in addition to L234I selected at 1× EC₉₅. Interestingly, Y181C was replaced by V106A in BTV from selection at an MK-6186 concentration of 50× EC₉₅.

A single mutation development pathway accounted for viral breakthrough in subtype C virus in both experiments, in which L234I again led the mutation pathway at 1× and 2× EC₉₅ of MK-6186 followed by the emergence of V106A at a higher concentration of the inhibitor.

For resistance selection experiments 1 and 3 with MK-6186 in subtype B virus, BTV selected at 10× and 50× EC₉₅ were analyzed by clonal sequencing to determine the linkage of the observed mutations and potentially to identify mutations occurring at lower frequencies within the population. Based on the clonal sequencing results (Table 4), in the first experiment at 10× EC₉₅, 44% of the BTV contained the V106A/E138K mutation, followed by 22% with the triple mutation V106A/E138K/L234I. A minor population (2/18) of the BTV harbored the triple mutation V106A/E138K/Y181C. At 50× EC₉₅, the majority (15/22) of the BTV contained the triple mutation V106A/E138K/L234I, and the rest were equally distributed among V106A/E138K/Y181C, V106A/E138K, and V106A/L234I. In the third experiment at 10× EC₉₅, the L234I/Y181C mutation accounted for more than 65% (15/23) of BTV, and the single mutation L234I represented 26% of the population. At 50× EC₉₅, 90% (17/19) of BTV contained the L234I/V106A mutation, and a minor population harbored the triple mutation L234I/V106A/V108I. These results confirm that Y181C selected at 10× EC₉₅ in experiment three was replaced by V106A at a higher concentration of MK-6186 (50× EC₉₅).

Table 4.

Mutations identified in BTV from resistance selection in subtype B virus with MK-6186 at 10× and 50× EC₉₅ in experiments 1 and 3 based on clonal sequencing

Expt	Mutations identified at:
	10× EC95		50× EC95
	Mutation(s)	No. of colonies	Mutation(s)	No. of colonies
1	V106A/E138K	8	V106A/E138K/L234I	15
	V106A/E138K/L234I	4	V106A/E138K/Y181C	2
	V106A/E138K/L228F	2	V106A/L234I	2
	V106A/E138K/Y181C	2	V106A/E138K	2
	V106A/L234I	1	V106A	1
	V106A/E138K/L234I/Y181C	1
3	L234I/Y181C	15	L234I/V106A	17
	L234I	6	L234I/V106A/V108A	2
	L234I/Y181C/V106I	2

Assessment of the susceptibility of the mutant viruses selected by MK-6186 in subtypes A and B to NNRTIs.

Different approaches were employed to assess the extent of NNRTI resistance conferred by the mutant viruses selected by MK-6186. With subtype A BTV, ViroMag R/L magnetic beads were used to isolate viral particles. Isolated BTV were then employed to infect cells to assess the susceptibility of the BTV to NNRTIs (41). With subtype B virus, conventional SDM was carried out to generate the mutant viruses selected by MK-6186. Attempts to isolate subtype C BTV with the magnetic beads for NNRTI susceptibility assessment were not successful due to the low efficiency of isolation.

The subtype A NNRTI resistance results are summarized in Fig. 6A. Subtype A viruses containing the Y181C mutation displayed FCs of 7.7, 2.3, and 32 against MK-6186, EFV, and ETR, respectively. Surprisingly, L234I-containing BTV exhibited significant resistance to MK-6186, with an FC of >100. However, the mutant viruses were highly susceptible to EFV and ETR, with FCs of 1.2 and 1.0, respectively. In addition, BTV containing F227C also showed a >100-fold shift in potency against MK-6186 compared to that of WT virus and exhibited only 3.1- and 1.6-fold shifts in potency against EFV and ETR, respectively.

Fig 6.

Susceptibility of mutant viruses selected by MK-6186 to ETR, EFV, and MK-6186. Fold change represents the ratio of EC₅₀s between mutant virus and WT virus. (A) Mutant viruses isolated by magnetic beads from subtype A BTV in experiments 1 (Y181C virus), 2 (L234I virus), and 3 (F227C virus). (B) Mutant viruses derived from SDM based on mutations identified in subtype B BTV in experiments 1 to 3. B1, mutant viruses selected from experiment 1; B2, mutant viruses selected from experiments 2; and B3, mutant viruses selected from experiment 3. The level at an FC of 100 indicates that the FC is >100. The values presented in the figure were obtained from means of quadruplicates.

Subtype B viruses containing mutations selected by MK-6186 all displayed higher levels of resistance to MK-6186 than to EFV or ETR (Fig. 6B). More importantly, viruses containing mutations that evolved at early stages conferred lower degrees of resistance to MK-6186 than the mutations that developed at later stages. For instance, in experiment one, E138K/V106A mutant viruses selected at 10× EC₉₅ exhibited 65-fold resistance to MK-6186, and the dominant E138K/V106A/L234I mutant viruses in BTV from selection at 50× EC₉₅ showed >170-fold resistance (Fig. 6B, graph 1). Both mutant viruses were sensitive to EFV and ETR, with FCs of <2. In experiment two, viruses containing the L234I mutation displayed moderate resistance to MK-6186, with an FC of 5.1, as opposed to the FC of >100 observed in subtype A viruses with the same mutation. The L234I mutant virus had FCs of 1.2 and 0.6 against EFV and ETR, respectively (Fig. 6B, graph 2). The emergence of additional V106A/V108I mutations at higher concentrations of MK-6186 dramatically reduced the susceptibility of the mutant virus to MK-6186, with the FC increasing from 5.1 to >200 (Fig. 6B, graph 2). However, the triple-mutant virus was still highly susceptible to EFV and ETR, with FCs of 1.7 and 0.3, respectively. Mutant viruses with L234I/V106A mutations selected in experiment three drastically increased the resistance to MK-6186, with an FC of 180, but they were quite sensitive to EFV and ETR suppression, with FCs of 0.9 and 0.3, respectively (Fig. 6B, graph 3).

In vitro resistance selection studies were also conducted with EFV and ETR, where L100I, V179D, and K103N were selected by EFV and Y181C, V179I, and L100I were selected by ETR (24). Subtype B mutant viruses were generated by SDM to contain the mutations selected by EFV or ETR. The mutant viruses were evaluated for their susceptibility to MK-6186. As shown in Fig. 7A, viruses harboring the mutations selected by EFV were highly susceptible to MK-6186, with FCs of <2. Moreover, MK-6186 was also highly active against mutant viruses selected by ETR, with FCs of <5 (Fig. 7B).

Fig 7.

Susceptibility of mutant viruses containing mutations selected by EFV or ETR to MK-6186. Fold change represents the ratio of EC₅₀s between mutant virus and WT virus. (A) Mutant viruses selected by EFV in experiments 1 (L100I and L100I/N348I), experiment 2 (V179D and V179D/L100I), and experiment 3 (K103N and K103N/Y188C). (B) Mutant viruses selected by ETR in experiment 1 (Y181C/A272T and Y181C/A272T/L100I) and experiments 2 and 3 (Y181C/V179I and Y181C/V179I/L100I). The values presented in the figure were obtained from means of quadruplicates.

In vitro resistance selection with MK-6186 under high-MOI conditions.

Resistance selection was also conducted under high-MOI conditions in the presence of 50% HS at fixed concentrations of NNRTIs to better reflect higher protein concentrations in blood plasma. The inhibitor concentrations used for the study were 200 nM, 500 nM, and 1.25 μM. Viral breakthroughs were observed with all three concentrations of MK-6186. As shown in Table 5, V106A was identified in BTV from five of six experiments. A triple mutation of V106A/V108I/F214F was found in one of the experiments at 1.25 μM. To assess the resistance conferred by the mutations selected under high-MOI conditions in the presence of 50% HS, V106A mutant viruses derived from SDM were tested with NNRTIs. The mutant viruses displayed >20-fold resistance to MK-6186 (data not shown), but the mutant viruses were highly susceptible to EFV and ETR, with FCs of 1.5 and 0.7, respectively (data not shown).

Table 5.

Resistance selection with MK-6186 on subtype B viruses under high-MOI conditions in the presence of 50% human serum

MK-6186 concn (nM)	Resistance selection in expt.:
	1		2
	Duration (days)	Mutation(s)	Duration (days)	Mutation(s)
200	9	V106A	9	V106A
500	10	V106A	13	V106A
1,250	13	V106A/V108I/F214L	13	V106A

The mutation development pathways selected by MK-6186 among the different subtypes of viruses, after discounting non-NNRTI-associated mutations, are summarized in Table 6. Consistently with previous findings in resistance selection experiments with MK-4965 (24), subtype C viruses showed a more convergent mutation development pathway: only a single pathway was identified from two independent experiments. In contrast, subtypes A and B showed more diverse pathways of mutation development: three different pathways were selected from the three different experiments. Among the total of seven mutation development pathways with A, B, and C subtypes, L234I was the initial mutation selected in the BTV in four mutation pathways. Therefore, the L234I mutation plays the most important role in mutation development in resistance selection with MK-6186 under low-MOI conditions. In contrast, V106A was the dominant mutation selected in BTV under high-MOI conditions.

Table 6.

Mutation development pathways evolved from resistance selection with MK-6186 in HIV-1 A, B, and C subtypes

Subtype and expt no.	Mutation pathway
A
1	Y181C
2	L234I
3	F227C
B
1	E138K/Y181C→E138K/Y181C/V106A
2	L234I→L234I/V108I→L234I/V108I/V106A
3	L234I→L234I/Y181C→L234I/V106A
C
1	L234I→L234I/V106A

DISCUSSION

The L234I mutation was the major contributor to viral breakthrough in resistance selection with MK-6186 under low-MOI conditions, especially with subtype B and C viruses. Subtype B mutant viruses harboring the L234I mutation exhibited approximately 5-fold resistance to MK-6186. Although the X-ray structure of the MK-6186-RT complex is available, the resolution of the C-ring is poor. Therefore, a model of the structure of the MK-6186-RT complex was generated based on a close analog of MK-6186 with minor modification on the a and b rings. As shown in Fig. 8, L234 is in close proximity to the a ring of MK-6186, resulting in van der Waals interactions between the γ methyl group of Leu and a ring of MK-6186. The mutation of Leu to Ile, one of the methyl groups at the γ carbon of Leu, is relocated to the β carbon of Ile. The hydrophobic methyl group at the β carbon of Ile is in close proximity to the hydrophilic indazole moiety of the b ring. These unfavorable interactions result in repulsive interactions between the Ile and MK-6186, thus reducing the susceptibility of the L234I mutant virus to MK-6186. L234I appears to be a unique resistance mutation with MK-6186, because it is a mutation that is rarely associated with older NNRTIs (33). It is also not selected by ETR and RPV and the NNRTIs currently under development, but it was found to be associated with capravirine (31). Interestingly, E138K/V106A mutant viruses exhibited ∼2-fold resistance to ETR, whereas, with the addition of L234I, the FC was reduced to 0.6. In addition, viruses with the V106A mutation have an FC of 0.7, and the FC was further decreased to 0.3 for viruses containing L234I/V106A double mutations. Therefore, the L234I mutation may provide advantageous interactions between ETR and L234I RT in the presence of other mutations, resulting in the hypersensitivity of L234I-containing viruses to ETR. These hypersensitivity phenomena were also observed in a resistance selection study with ETR (40), in which the emergence of the L234I mutation with Y181C always enhances the susceptibility of the mutant virus to ETR. WT and L234I-containing viruses displayed similar sensitivities to EFV and ETR. Consequently, this MK-6186-specific mutation would not cause cross-resistance to EFV and ETR. Moreover, since the L234I mutation is rarely associated with other NNRTIs (33), mutant viruses containing this mutation may not display resistance to other NNRTIs that possess structures distinct from that of MK-6186. In contrast, the emergence of L234I along with other mutation(s) may further enhance the susceptibility of the viruses to other NNRTIs, as is the case with ETR.

Fig 8.

Modeled structure of the residues surrounding the NNRTIBP in the presence of MK-6186 (based on a close structural analog of MK-6186).

V106A was the dominant mutation identified in BTV under high-MOI conditions in the presence of 50% HS. Viruses containing the V106A mutation displayed ∼20-fold resistance to MK-6186, supporting the view that the V106A mutation was responsible for the viral breakthrough. Based on the modeled structure of the RT-MK-6186 analog complex, the dimethyl group of V106 appears to interact with the hydrophobic portion of b and c rings via van der Waals interactions (Fig. 8). The mutation of Val to Ala results in the weakening of the hydrophobic interactions between this residue and MK-6186 due to the loss of two methyl groups in the Ala mutant. As a result, the affinity of the compound to the NNRTIBP is significantly diminished. In contrast, V106A mutant viruses are very sensitive to ETR, with an FC of 0.7. This sensitivity may be ascribed to the additional space gained by the replacement of the bulky isopropyl group in Val with the methyl group in Ala, thus providing extra volume to accommodate alternative conformations that may be adapted by ETR, which is a highly flexible molecule that can fit into the NNRTIBP with various conformations (10).

Despite higher levels of resistance to MK-6186 imparted by the V106A mutation compared to the L234I mutation, L234I was the major mutation selected at 1× EC₉₅, whereas V106A was selected at later stages (>5× EC₉₅) in the resistance selection studies under low-MOI conditions. Under high-MOI conditions, however, V106A was the only mutation selected in five of six experiments. Viruses harboring the V106A mutation have lower replication capacities (2, 26), thus limiting their ability to compete with other mutant viruses that have higher replication capacity. These results are consistent with the contention that mutant viruses with higher levels of resistance yet lower replication capacities would be selected only under higher selective pressure (24). In other words, mutant viruses with better fitness and lower levels of resistance tend to be selected at lower concentrations of the antiviral agent during resistance selection. Although a primer grip mutation of L234A was found to inhibit the dimerization of p66/p51 RT subunits (38), information regarding the fitness of L234I mutant viruses is not available, and it is predicted that L234I-containing virus would have a better replication capacity than V106A-containing virus. In conclusion, the interplay between the level of resistance and the fitness of the mutant virus plays an important role in mutation development pathways (24).

Surprisingly, subtype A viruses with the L234I mutation displayed significantly higher resistance to MK-6186 than subtype B virus with the same mutation (Fig. 6A versus B, graph 2). These results suggest that the L234I mutation interferes with the binding of MK-6186 to the NNRTIBP of subtype A virus to a greater extent than that of subtype B virus. Therefore, the conformations of the residues within the NNRTIBP may be different between subtype A and B viruses. In addition, although subtype B F227C mutant virus is reportedly nonviable, (21) F227C was the single mutation identified in BTV during resistance selection with subtype A virus in experiment three (Table 3). Therefore, the F227C mutation in subtype A virus may have less impact on the replication capacity of the virus. On the other hand, the RT polymerase active site in subtype B virus may be markedly perturbed by the F227C mutation, thus debilitating the replication capacity of the virus. Taken together, these observations suggest that the relative conformations of the side chains of the residues inside the NNRTIBP are different in subtype A and B viruses (24).

In resistance selection studies in subtype B viruses, the Y181C mutation along with L234I was detected in the BTV from experiment 3 at an MK-6186 concentration of 10× EC₉₅. However, the Y181C mutation was replaced by V106A in the resistance selection at 50× EC₉₅. One possible explanation for the replacement is that the L234I/Y181C mutant virus does not confer high enough resistance to overcome the selection pressure at 50× EC₉₅. As a result, the emergence of another mutation with higher resistance power is necessary for the virus to escape the inhibition pressure. In this case, the emergence of V106A with L234I gives an FC of >180. However, mutations that arise during in vitro resistance selection generally are developed by the sequential emergence of mutations as inhibitor concentrations are increased to overcome the selection pressure. Thus, it is expected that the V106A mutation would emerge with a double mutation (L234I/Y181C) to give a triple mutation (L234I/Y181C/V106A) instead of replacing Y181C at 50× EC₉₅. The fact that Y181C was replaced by V106A suggests that L234I/Y181C/V106A mutant viruses are not viable, as the triple-mutant virus was not detected even by clonal sequencing. The contention that the replication capacity of the mutant virus containing the triple mutation is limited is also supported by the clonal sequencing results of the BTV from resistance selection at 50× EC₉₅ in experiment one (Table 4). Four mutations (V106A, L234I, E138K, and Y181C) were identified in the BTV by population sequencing. However, based on clonal sequencing results, no combination of the triple mutation (L234I/Y181C/V106A) was detected in any of the clones despite mutant viruses with triple mutations of L234I/V106A/E138K and Y181C/V106A/E138K being identified in the BTV (Table 4).

Although ETR appears to show a better resistance profile based on the susceptibility test of 96 NNRTI-resistant viruses, MK-6186 possesses a unique mutation development pathway where L234I and V106A mutations dominate in resistance selection in vitro under low- and high-MOI conditions, respectively. Interestingly, these V106A and L234I mutant viruses are highly susceptible to ETR and EFV. More importantly, the resistance mutations selected by ETR and EFV under the same conditions are sensitive to MK-6186. Taken together, MK-6186 possesses desired antiviral activity and resistance profiles as a new NNRTI.