Isolation and Characterization of Human Immunodeficiency Virus Type 1 Resistant to the Small-Molecule CCR5 Antagonist TAK-652

Abstract

TAK-652, a novel small-molecule chemokine receptor antagonist, is a highly potent and selective inhibitor of CCR5-using (R5) human immunodeficiency virus type 1 (HIV-1) replication in vitro. Since TAK-652 is orally bioavailable and has favorable pharmacokinetic profiles in humans, it is considered a promising candidate for an entry inhibitor of HIV-1. To investigate the resistance to TAK-652, peripheral blood mononuclear cells were infected with the R5 HIV-1 primary isolate KK and passaged in the presence of escalating concentrations of the compound for more than 1 year. After 67 weeks of cultivation, the escape virus emerged even in the presence of a high concentration of TAK-652. This virus displayed more than 200,000-fold resistance to TAK-652 compared with the wild type. The escape virus appeared to have cross-resistance to the structurally related compound TAK-779 but retained full susceptibility to TAK-220, which is from a different class of CCR5 antagonists. Furthermore, the escape virus was unable to use CXCR4 as a coreceptor. Analysis for Env amino acid sequences of escape viruses at certain points of passage revealed that amino acid changes accumulated with an increasing number of passages. Several amino acid changes not only in the V3 region but also in other Env regions seemed to be required for R5 HIV-1 to acquire complete resistance to TAK-652.

The introduction of highly active antiretroviral therapy with reverse transcriptase inhibitors and protease inhibitors has achieved significant progress in the treatment of human immunodeficiency virus type 1 (HIV-1) infection (31). In addition, novel inhibitors targeting other essential molecules for viral replication, such as CCR5 and integrase, are now in human clinical trials (8, 22, 25). The chemokine receptors CCR5 and CXCR4 act as major coreceptors of HIV-1 in consort with the primary receptor CD4 (4, 16, 17). It has been reported that HIV-1 using CCR5 as a coreceptor (R5 HIV-1) is isolated predominantly during the asymptomatic stage (5). R5 HIV-1 is also responsible for virus transmission between individuals. On the other hand, HIV-1 using CXCR4 as a coreceptor (X4 HIV-1) generally emerges at the advanced stage of the disease and is related to acceleration of its progression (5, 20). However, several lines of evidence suggest that R5 HIV-1 still plays a major role even in the advanced stage (11, 30). Therefore, suppression of R5 HIV-1 in infected individuals may be more important than that of X4 HIV-1 in terms of blocking viral transmission and delaying disease progression. This hypothesis has been supported by the finding that individuals having homozygous CCR5-Δ32, a truncated and nonfunctional form of CCR5, display profound resistance to HIV-1 infection without obvious health problems (6, 12, 21). These findings have given us the idea that CCR5 antagonists may be effective as anti-HIV-1 agents without serious side effects, even though CCR5 is a host cellular factor.

The first small-molecule CCR5 antagonist, TAK-779, has been reported to be a potent and selective inhibitor of HIV-1 replication by our group (3). This compound inhibits R5 HIV-1 replication at nanomolar concentrations in cell cultures. However, TAK-779 is an anilide derivative with a quaternary ammonium moiety and could not be further developed as an antiretroviral agent because of its poor oral bioavailability. In the meantime, several groups have identified different classes of small-molecule and orally bioavailable CCR5 antagonists, most of which appeared to be promising candidates for further development (8, 13, 25). TAK-220 and TAK-652, novel orally bioavailable CCR5 antagonists, are successors of TAK-779. TAK-220 is one of a novel series of compounds with chemical structures totally different from that of TAK-779 (27). TAK-220 is orally bioavailable and highly inhibitory to HIV-1 replication in vitro. The other compound, TAK-652, is a derivative of TAK-779 with high oral bioavailability and favorable pharmacokinetic profiles in humans (2). This compound is also a highly potent inhibitor of R5 HIV-1 replication in vitro. Thus, both compounds are considered promising candidates for clinical development.

There may be no exceptions that drug-resistant HIV-1 will emerge under the selective pressure of any single antiretroviral agent. In the case of CCR5 antagonists, there is a serious concern that their long-term use could induce the evolution of X4 HIV-1 in patients (17, 19). In fact, drug-resistant viruses were isolated in long-term cultures of R5 HIV-1-infected cells by the selection pressure of some CCR5 antagonists, such as AD101 and vicriviroc (10, 15, 29). However, the escape viruses were found to retain the R5 phenotype. Therefore, in vitro isolation and analyses of drug-resistant viruses may be able to provide useful information for future clinical development of CCR5 antagonists. In this study, we conducted a long-term culture experiment with R5 HIV-1-infected peripheral blood mononuclear cells (PBMCs) with escalating concentrations of TAK-652. After serial passages of the infected cells for more than 1 year, an escape virus was obtained which displayed complete resistance to TAK-652 but retained full susceptibility to TAK-220.

MATERIALS AND METHODS

Compounds.

The small-molecule CCR5 antagonists TAK-779 (3), TAK-220 (27), and TAK-652 (2) and the CXCR4 antagonist AMD3100 (23) were synthesized by Takeda Pharmaceutical Company, Osaka, Japan. The chemical structures of the CCR5 antagonists are shown in Fig. 1.

FIG. 1.

Structures of TAK-779, TAK-220, and TAK-652.

Cells and virus.

PBMCs were obtained from healthy volunteers after obtaining their informed consent. The cells were isolated with Ficoll-Hypaque gradient density centrifugation and stimulated with 5 μg/ml phytohemagglutinin (PHA) in RPMI 1640 medium supplemented with 20% fetal bovine serum, 100 U/ml recombinant human interleukin-2 (Takeda Pharmaceutical Company), 100 U/ml penicillin G, and 100 μg/ml streptomycin for 3 days. The above medium without PHA was used in viral replication assays. U87 astroglioma cells expressing human CD4 and either CCR5 or CXCR4 (U87.CD4.CCR5 cells or U87.CD4.CXCR4 cells, respectively) were obtained from D. Littman (New York University School of Medicine, New York, NY) and maintained in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum, 300 μg/ml Geneticin, 1 μg/ml puromycin, and antibiotics. The above medium without Geneticin and puromycin was used in viral replication assays. The KK strain of R5 HIV-1 was used for infection of PBMCs. This strain was isolated from a patient in Kagoshima University Hospital who had no treatment history with any antiretrovirals until virus isolation. Its coreceptor usage was previously determined by a replication assay in U87.CD4.CCR5 cells and U87.CD4.CXCR4 cells, as described below.

Long-term culture of infected PBMCs with TAK-652.

The KK strain of HIV-1 (100 ng of p24) was added to 6 ml of PHA-stimulated PBMCs (5 × 10⁶ cells) and incubated at 37°C. To achieve sufficient infection of PBMCs with the clinical isolate, the cells were incubated for 2 days in the absence of compounds. After virus adsorption, the cells were washed three times with culture medium to remove unadsorbed virus particles. The cells were resuspended with culture medium (10 ml) in the presence of 0.2 nM TAK-652. On day 4 after virus infection, the infected cells were subcultured at a ratio of 1:4 with fresh culture medium containing the same concentration of the compound. On day 7, the number of viable cells was counted and adjusted to 1 × 10⁶ cells/ml. Then, 1 ml of the infected cells and 4 ml of freshly prepared and uninfected PBMCs (4 × 10⁶ cells) were suspended with culture medium (10 ml) in the presence of an appropriate concentration of TAK-652 and incubated at 37°C. As control cultures, exactly identical passages of the infected PBMCs in the absence of the compound were carried out in parallel with the cultures exposed to TAK-652. At each passage, p24 antigen levels of culture supernatants were monitored by using an enzyme-linked immunosorbent assay (ELISA) kit (ZeptoMetrix Corp., Buffalo, NY) to confirm virus replication. The concentration of TAK-652 was escalated when the p24 level in the TAK-652-treated culture exceeded 50% of that in the control culture at two consecutive passages. To exclude the effect of different PBMC donors, PBMCs from the same (one) donor were used for passages throughout the experiment. The escape viruses as well as the control viruses were propagated once in PBMCs to remove the compound, their infectivity was determined, and they were used for further experiments.

Susceptibility assay of escape viruses to CCR5 antagonists.

PHA-stimulated PBMCs (4 × 10⁶ cells) were infected with 1,400 50% cell culture infective doses of the virus and incubated at 37°C. After a 4-h incubation, the cells were washed with culture medium to remove unadsorbed viral particles and seeded into a 96-well plate (2 × 10⁵ cells/well) with culture medium containing various concentrations (0.1 to 10,000 nM) of test compounds. On day 4 after virus infection, the cells were subcultured at a ratio of 1:2 with fresh culture medium containing the same concentration of the test compounds. On day 7 after infection, the culture supernatants were collected and p24 antigen levels were determined by using an ELISA kit (ZeptoMetrix Corp.).

Determination of coreceptor usage.

U87.CD4.CCR5 or U87.CD4.CXCR4 cells were seeded into a 48-well plate (1 × 10⁴ cells/well) and incubated overnight at 37°C. The culture supernatants were removed, and the cells were inoculated with the culture supernatant of each passage in a total volume of 400 μl. After an overnight incubation, the cells were washed thoroughly with culture medium to remove unadsorbed viral particles and further incubated. On day 4 after infection, the culture supernatants were removed and incubation continued with fresh culture medium. On day 6, the culture supernatants were collected and p24 antigen levels were determined by using an ELISA kit (ZeptoMetrix Corp.).

Sequence analysis of env genes.

Genomic DNA was extracted from the infected PBMCs with a DNA extraction kit (Wako, Tokyo, Japan). The extracted DNA was subjected to PCR. The PCR consisted of 30 cycles (95°C for 15 s, 55°C for 30 s, and 68°C for 150 s) with the forward and reverse primers EnvS (5′-GAGCAGAAGACAGTGGCAATGAGAGTGAAG-3′) and EnvA (5′-TTTTGACCACTTGCCACCCATCTTATAGCA-3′), respectively, which generated a fragment including nucleotides −18 through 2566 of the env gene corresponding to the JR-FL strain of HIV-1 (GenBank accession number U63632). The amplified products were isolated by gel electrophoresis and purified with a PCR DNA and gel band purification kit (Amersham Pharmacia Biotech, Piscataway, NJ). The purified DNA was sequenced directly with a cycle sequence kit (BigDye Terminator version 3.1; Applied Biosystems Inc., Foster City, CA), using both forward and reverse primers on an automated DNA analyzer (model 3730; Applied Biosystems Inc.). Depending on the sequence result obtained by the analysis, the primer for the next sequence analysis was designed.

Data analysis.

The 50% inhibitory concentrations (IC₅₀) of test compounds were calculated using the SAS system procedure NLIN, which produces least-squares estimates of the parameters of a nonlinear model (logistic model).

RESULTS

Isolation of escape viruses.

To successfully isolate TAK-652-resistant viruses, it seemed important to start an experiment with a genetically heterogeneous primary R5 HIV-1 isolate. In addition, an isolate from a treatment-naïve patient would be preferable. Therefore, we chose the KK strain as the source of R5 HIV-1. In our previous study, TAK-652 was found to inhibit replication of the KK strain with an IC₅₀ and IC₉₀ of 0.043 nM and 0.19 nM, respectively (2). Therefore, PBMC cultures were started in the absence or presence of TAK-652 at a concentration of 0.2 nM (almost identical to its IC₉₀). After three passages, the p24 level of the TAK-652-treated culture rapidly increased and reached more than 50% of the control culture level (Fig. 2). Therefore, the compound concentration was elevated to 0.4 nM at passage 5. At passage 8, viruses were isolated from the TAK-652-treated and control cultures and designated as KK_652-8 and KK_C-8, respectively (Fig. 2).

FIG. 2.

Long-term culture of infected PBMCs with escalating concentrations of TAK-652. PHA-stimulated PBMCs were infected with an R5 HIV-1 clinical isolate (KK strain) and were passaged weekly by replenishing with fresh PBMCs in the presence or absence of TAK-652 at the indicated concentrations. The solid and dotted lines indicate the p24 levels and the concentrations of TAK-652 in culture supernatants, respectively. PBMCs obtained from one healthy volunteer were used throughout the experiment. At passages 8, 43, 56, and 67, viruses were isolated from the supernatants and subjected to phenotypic and genotypic analyses.

Further passages of the infected PBMCs were carried out weekly with an increasing concentration of TAK-652 from 0.8 to 1.6 nM. During this period, HIV-1 replication in the TAK-652-treated culture appeared to be suppressed well, compared to that in the control culture. However, the p24 levels gradually increased after 37 passages, and viruses from the TAK-652-treated and control cultures were isolated at passage 43 (KK_652-43 and KK_C-43, respectively) (Fig. 2). At passage 52, the p24 levels of the TAK-652-treated and control cultures were comparable. After this point, suppression of HIV-1 replication was not observed for the TAK-652-treated culture, even when its concentration was elevated to 100 nM. Viruses were obtained from the TAK-652-treated and control cultures at passage 56 (KK_652-56 and KK_C-56, respectively) and passage 67 (KK_652-67 and KK_C-67, respectively) (Fig. 2). The concentrations of TAK-652 were 10 and 1,000 nM at passages 56 and 67, respectively.

Susceptibility of escape viruses to CCR5 antagonists.

When TAK-652 was examined for its inhibitory effect on KK_652-67 replication in PBMCs obtained from two different donors, it did not show any significant inhibition at concentrations up to 10,000 nM (Fig. 3). TAK-779 exhibited a dose-dependent but only partial antiviral activity against KK_652-67. Interestingly, TAK-220, from a different class of CCR5 antagonists, was highly inhibitory to the replication of KK_652-67, irrespective of PBMC donor. As shown in Table 1, the IC₅₀ of TAK-652 for KK_652-67 was more than 10,000 nM. Considering that the IC₅₀ of TAK-652 for the wild-type virus (KK_WT) was 0.043 nM in PBMCs, KK_652-67 had strong (more than 200,000-fold) resistance to this compound. The IC₅₀s of TAK-779 for KK_652-67 were 77 and 2,000 nM in experiments 1 and 2, respectively. In another experiment under identical assay conditions, an IC₅₀ of 2.1 nM was obtained with TAK-779 for the wild type (Table 1), suggesting that KK_652-67 displayed cross-resistance to TAK-779. In contrast, KK_652-67 appeared to retain complete susceptibility to TAK-220 because little, if any, increase in its IC₅₀ was observed in comparison with those for KK_WT (Table 1). Surprisingly, TAK-652 did not inhibit the replication of the control virus isolated at passage 67 (KK_C-67) at concentrations up to 10,000 nM (data not shown), suggesting that the virus is an X4 HIV-1, a dual-tropic (R5X4) HIV-1, or a mixture of R5 HIV-1 and X4 HIV-1.

FIG. 3.

Antiviral activities of TAK-652, TAK-220, and TAK-779 against the escape virus at passage 67 (KK_652-67). PHA-stimulated PBMCs were infected with the virus and incubated for 4 h. The cells were washed to remove unadsorbed viral particles and seeded into a 96-well plate with culture medium containing various concentrations of test compounds. On day 4 after virus infection, the cells were subcultured at a ratio of 1:2 with fresh culture medium containing the same concentration of the test compounds. On day 7 after infection, the culture supernatants were collected and p24 antigen levels were determined by ELISA. The results of two different experiments are shown. In each experiment, different donors were used for the antiviral assays of the compounds.

TABLE 1.

Anti-HIV-1 activities of TAK-652, -220, and -779 against KK_652-67 in PBMCs^a

Compound	KKWTb	IC50 (nM) (fold increase)
		KK652-67
		Expt 1	Expt 2
TAK-652	0.043	>10,000 (>230,000)	>10,000 (>230,000)
TAK-220	1.2	1.9 (1.6)	1.3 (1.1)
TAK-779	2.1	77 (37)	2,000 (940)

^aCells from different donors were used in each experiment. The IC₅₀ is the 50% inhibitory concentration based on the reduction of p24 antigen levels in culture supernatants on day 7 after virus infection. Assays were performed in triplicate wells. Values in parentheses represent the fold increase (ratio of the IC₅₀ for the wild type to the IC₅₀ for KK_652-67).

^bThe IC₅₀s of TAK-652 and TAK-220 for the wild type were taken from references 2 and 27, respectively. The IC₅₀ of TAK-779 for the wild type is based on an experiment performed under identical assay conditions.

It was important to examine the susceptibility of the viruses isolated at certain points during the long-term passage of infected PBMCs. For accurate evaluation and comparison of these viruses, all escape and control viruses as well as the wild type needed to be examined simultaneously. In the first experiment, the IC₅₀ of TAK-652 for the wild type was 0.14 nM (Table 2). This value was 3.3-fold higher than that obtained previously (2). When the IC₅₀ for the escape virus at passage 8 (KK_652-8) was compared to that for its control virus (KK_C-8), no reduction in its susceptibility to TAK-652 was observed in either experiment 1 or 2. However, slight (2.0- to 3.3-fold) and considerable (110- to 380-fold) increases in the IC₅₀s were identified for KK_652-43 and KK_652-56, respectively, compared with those of the corresponding control viruses. Thus, viruses with a different degree (low, middle, or high) of resistance to TAK-652, namely, KK_652-43, KK_652-56, and KK_652-67, were obtained in the long-term culture experiment.

TABLE 2.

Anti-HIV-1 activity of TAK-652 against escape and control viruses obtained at passages 8, 43, and 56 in PBMCs^a

Virus	IC50 (nM) (fold increase)
	Expt 1	Expt 2
KKWT	0.14	NDb
KKC-8	0.24 (1)	0.27 (1)
KK652-8	0.13 (0.55)	0.26 (0.97)
KKC-43	0.20 (1)	0.34 (1)
KK652-43	0.66 (3.3)	0.69 (2.0)
KKC-56	0.63 (1)	0.64 (1)
KK652-56	240 (380)	67 (110)

^aCells from different donors were used in each experiment. The IC₅₀ is the 50% inhibitory concentration based on the reduction of p24 antigen levels in culture supernatants on day 7 after virus infection. Assays were performed in triplicate wells. Values in parentheses represent the fold increase (ratio of the IC₅₀ for virus from the control passage to the IC₅₀ for virus from the corresponding TAK-652-treated passage).

^bND, not determined.

Coreceptor usage of escape viruses.

To determine whether the escape viruses and their control viruses acquired the ability to use CXCR4 as an alternative coreceptor, the replication of these viruses were examined in U87.CD4.CCR5 or U87.CD4.CXCR4 cells. All viruses except for KK_C-67 replicated well in U87.CD4.CCR5 cells but not in U87.CD4.CXCR4 cells (Table 3), indicating that they could not use CXCR4 as a coreceptor for infection. Although we did not examine their replication in U87 cells expressing other chemokine receptors, it would be unlikely that these viruses could use a chemokine receptor other than CCR5 and CXCR4. KK_652-67 was found to be highly susceptible to TAK-220 (Fig. 2 and Table 1), which is an antagonist highly specific to CCR5 (27).

TABLE 3.

Replication of escape and control viruses obtained at passages 8, 43, 56, and 67 in U87.CD4.CCR5 and U87.CD4.CXCR4 cells^a

Virus	p24 (ng/ml)b
	U87.CD4.CCR5	U87.CD4.CXCR4
KKC-8	+	−
KKC-43	+	−
KKC-56	+	−
KKC-67	+	+
KK652-8	+	−
KK652-43	+	−
KK652-56	+	−
KK652-67	+	−

^aAssay procedures are described in Materials and Methods.

^b+, the p24 level in the culture supernatant was above 1 ng/ml or the virus replicated in the indicated cells; −, the p24 level was below 0.1 ng/ml or the virus did not replicate in the indicated cells.

KK_C-67 could replicate in both U87.CD4.CCR5 and U87.CD4.CXCR4 cells (Table 3). Therefore, to determine whether the virus is a dual-tropic (R5X4) HIV-1 or a heterogeneous mixture of R5 HIV-1 and X4 HIV-1, PBMCs were infected with KK_C-67 and cultured in the presence of either 5 μM TAK-652, 5 μM AMD3100 (a specific inhibitor of CXCR4), or both. These concentrations of TAK-652 and AMD3100 are enough to suppress the replication of R5 HIV-1 and X4 HIV-1, respectively. After a 4-day incubation, the cells were washed thoroughly and further incubated for 3 days in the absence of any compounds. Active replication of KK_C-67 was observed for the cells initially exposed to TAK-652 alone or AMD3100 alone. However, no replication was observed for the cells initially exposed to both compounds (data not shown). Then, PBMCs were infected with the virus derived from the TAK-652-exposed culture and incubated in the presence of 5 μM AMD3100, and no virus replication was identified (data not shown). An identical result was obtained when PBMCs were infected with the virus derived from the AMD3100-exposed culture and incubated in the presence of 5 μM TAK-652 (data not shown). These results indicate that KK_C-67 is likely to be a heterogeneous mixture of R5 HIV-1 and X4 HIV-1 rather than R5X4 HIV-1.

Amino acid changes of escape viruses.

To determine what amino acid changes are associated with resistance to TAK-652, full-length env genes of four escape and four control viruses as well as the wild type were sequenced. Several amino acid changes were observed not only in the gp120 subunit but also in the gp41 subunit (Fig. 4). Each amino acid change could be classified into one of three categories. The first category includes changes that were identified for both escape and control viruses. The second one is changes that were identified only for the control viruses. The last one, which is the most important, includes the changes that were identified only for the escape viruses under the selection pressure of TAK-652. The amino acid changes in the last category have been summarized in Table 4. KK_652-8 had no amino acid changes in this category, which corresponds to the observation that the susceptibility of KK_652-8 to TAK-652 was comparable that of KK_C-8 and KK_WT (Table 2). Six amino acid changes, including one heterogeneous change, were observed for the modestly resistant virus KK_652-43. The number of amino acid changes increased with increasing passage number and resistance. The highly resistant virus KK_652-67 had 12 amino acid changes: one in C2, two in V3, two in V4, two in C4, and five in gp41 (Table 4).

FIG.4.

Env amino acid sequences of isolated viruses. Blue letters indicate the amino acid changes identified for both drug escape and control viruses. Green letters indicate the amino acid changes identified only for the control viruses but not for the escape viruses. Red letters indicate the amino acid changes identified only for the escape viruses but not for the control viruses. There was heterogeneity for four amino acids, which are indicated numerically (1, P or A; 2, P or T; 3, M or I; 4, S or L).

TABLE 4.

Env amino acid changes considered to be involved in TAK-652 resistance^a

Virus	Amino acid at position:
	221 (C2)	306 (V3)	309 (V3)	401 (V4)	403 (V4)	422 (C4)	424 (C4)	506 (gp41)	637 (gp41)	690 (gp41)	766 (gp41)	769 (gp41)
KKWT	K	T	Q	T	T	M	A	M	S	L	V	I
KK652-8	—b	—	—	—	—	—	—	—	—	—	—	—
KK652-43	—	—	—	I	I	M/I	—	V	A	I	—	—
KK652-56	—	K	—	I	I	I	T	V	A	I	A	—
KK652-67	N	K	E	I	I	I	N	V	A	I	A	S

^aBased on the amino acid sequence data shown in Fig. 4.

^b—, identical to the amino acid of the wild type.

DISCUSSION

In this study, we isolated an R5 HIV-1 virus highly resistant to the novel CCR5 antagonist TAK-652 through long-term culture of infected PBMCs. The phenotypic analysis revealed that the virus was highly resistant to TAK-652 and had partial cross-resistance to TAK-779, probably due to their structural similarities (Fig. 1 and Table 1). In contrast to TAK-779, TAK-220, a structurally different CCR5 antagonist, was equally inhibitory to the replication of the TAK-652-resistant virus and the wild type. A similar finding has been reported for SCH-C and vicriviroc (SCH-D) (25, 26). Although both compounds are structurally related, the subtype G Russian clinical isolate RU570, which was weakly susceptible to inhibition by SCH-C (IC₅₀ > 1 μM), retained high susceptibility to vicriviroc. Furthermore, four amino acid changes in the V3 region of gp120 were necessary and sufficient to confer resistance to SCH-C (10), whereas vicriviroc-resistant viruses had no amino acid changes in the V3 region (1). Thus, Env amino acid changes responsible for resistance to CCR5 antagonists differ from one compound to another. It would be of special interest to determine whether TAK-652 has sufficient antiviral activity against TAK-220-resistant R5 HIV-1.

We have also conducted a long-term culture experiment with R5 HIV-1-infected PBMCs under the selection pressure of TAK-220. However, no escape virus could be obtained with escalating concentrations of TAK-220, even after 132 passages (data not shown). At present, the reason for such difficulty in inducing TAK-220-resistant viruses is unclear. Nishikawa and colleagues recently analyzed the binding sites for TAK-220 on human CCR5 and found that TAK-220 shares some interacting amino acid residues with TAK-779 but also requires distinct amino acid residues for its inhibitory effect on HIV-1 (18). It is possible that the conformation of CCR5 might be more extensively altered by binding of TAK-220 to CCR5 than by binding of TAK-652. Nevertheless, TAK-652 has unique properties with which TAK-220 and other CCR5 antagonists are not endowed. TAK-652 has good oral bioavailability and a long plasma half-life in humans (2). Therefore, it is assumed that TAK-652 is able to retain a plasma concentration sufficiently higher than that required for virus inhibition by once-daily administration at a reasonable dose. TAK-652 is a potent inhibitor of ligand binding not only to CCR5 but also to CCR2b, which has been observed for neither TAK-220, maraviroc, vicriviroc, nor aplaviroc (8, 13, 25, 27).

Amino acid changes in the Env region accumulated with an increasing period of cultivation (Fig. 4). Among the amino acid changes, several changes did not seem to be attributable to the selection pressure by TAK-652 but were the consequences of in vitro passage of infected cells, since these changes could be identified not only for the escape viruses but also for the corresponding control viruses. In addition, there were some amino acid changes that were found only for the control viruses. Selection of viruses with certain amino acid changes despite the absence of compounds has been commonly observed after in vitro passages of primary isolates and may be attributed to better replication fitness of these viruses. Apart from the amino acid changes unrelated to TAK-652 resistance, there were amino acid changes identified only for the escape viruses, and they accumulated with an increased period of cultivation (Table 4). In particular, a considerable gap in drug susceptibility was found between KK_652-43 and KK_652-56 (Table 2). In addition to the amino acid changes observed for KK_652-43, the three changes T306K (V3), M424T (C4), and V766A (gp41) were identified for KK_652-56. Furthermore, three amino acid changes, K221N (C2), Q309E (V3), and I769S (gp41), occurred in the Env region of the highly resistant virus KK_652-67. It has been reported that, unlike the resistance to reverse transcriptase and protease inhibitors, one amino acid change of the Env region does not bring about the resistance to CCR5 antagonists (10, 14). Trkola and colleagues have proposed two possible mechanisms that confer resistance to CCR5 antagonists on HIV-1 (29). The first one is the increase of gp120 binding affinity to CCR5. In this case, the virus can compete more strongly with a CCR5 antagonist and infect target cells. The second one is the creation of a substantially different binding site on gp120 for CCR5. In this case, the virus is still able to infect the target cells, even when the binding site of gp120 is already occupied with a CCR5 antagonist. Since the two mechanisms may not be mutually exclusive and can act sequentially, further studies, including the introduction of site-directed mutations, are required to elucidate the amino acids responsible for the resistance to TAK-652.

In accordance with previous experiments by others (29), no HIV-1 coreceptor switch from CCR5 occurred in the escape viruses in this study (Table 3). Instead, the control virus, KK_C-67, could use both CCR5 and CXCR4 for infection. It is known that only a few amino acid changes, especially in the V3 region of gp120, can convert R5 HIV-1 to X4 HIV-1 (7, 9, 24). Indeed, seven amino acid changes, including two heterogeneous changes, were found in the V3 region of KK_C-67, suggesting that the virus might be a heterogeneous mixture of R5 HIV-1 and X4 HIV-1 rather than a dual-tropic (R5X4) virus. This assumption was confirmed by the drug-swapping experiment with TAK-652 and AMD3100 (see Results for details). Since the original strain, KK_WT, is a clinical isolate from a treatment-naïve patient, it is not surprising that a small population of X4 HIV-1 existed in the infected cells and expanded during their long-term culture.

In conclusion, this study provides important information on TAK-652-resistant viruses, such as no cross-resistance to TAK-220 and no coreceptor switch to X4 HIV-1. While our experiments using a clinical isolate and a single PBMC donor may reflect the in vivo scenario of drug resistance better than those using a laboratory strain and multiple donors, it is possible that different mutants will be selected in individual experiments. Furthermore, in a clinical setting, CCR5 antagonists must be used in combination with existing antiretrovirals (28), which may alter the pattern for TAK-652 resistance. Thus, the emergence of drug resistance should be further investigated and confirmed in clinical trials.