Mutations in the HIV-1 3′-Polypurine Tract and Integrase Strand Transfer Inhibitor Resistance

LETTER

Integrase strand transfer inhibitors (INSTIs), in combination with other antivirals, are widely used for the treatment of HIV-1-infected individuals. HIV-1 resistance to INSTIs is typically associated with mutations in the viral integrase gene (1). However, recent studies suggested that mutations in the HIV-1 3′-polypurine tract (3′-PPT) may also confer viral resistance to INSTIs. Specifically, Malet et al. (2) reported in an in vitro selection experiment that a high concentration of dolutegravir (DTG) selected for mutations in the 3′-PPT that conferred INSTI resistance. Wijting et al. (3) reported on the selection of mutations in the 3′-PPT, but not integrase gene, in the virus from an infected individual who failed DTG maintenance monotherapy. However, no follow-up studies have ascertained whether the 3′-PPT mutations reported by Wijting et al. reduce INSTI sensitivity in phenotypic analyses of drug resistance. The primary objective of this study was to address this critical knowledge gap.

Using site-directed mutagenesis, we constructed subtype B HIV-1^LAI infectious viruses in which the 3′-PPT sequence was changed from 5′-TTAAAAGAAAAGGGGGG-3′ (HIV-1^WT) to 5′-TTAAAAGAAAAGCAGTΔACTGGAA-3′ (HIV-1^PPT(1)) or 5′-TTAAAAGAAAAGGGAGC-3′ (HIV-1^PPT(2)), in line with the mutations reported by Malet et al. (2) and Wijting et al. (3), respectively. We first assessed whether the 3′-PPT mutations affected the replication capacity of HIV-1 using a relatively low (27 ng p24) or high (81 ng p24) input of virus (Fig. 1). We found that HIV-1^PPT(2) replicated as efficiently as HIV-1^WT. In contrast, we could barely detect replication of HIV-1^PPT(1), suggesting that the mutations dramatically reduce replication efficiency. In light of this diminished replication efficiency, HIV-1^PPT(1) was excluded from further analysis. HIV-1 susceptibility to the INSTIs DTG, cabotegravir, elvitegravir, and raltegravir was assessed in a single-cycle assay, as described previously (4). We found that HIV-1^WT and HIV-1^PPT(2) exhibited similar susceptibility to each of the INSTIs (Table 1). Previously published studies have suggested that HIV-1 reverse transcriptase (RT)-mediated plus-strand DNA synthesis from the 3′-PPT is exquisitely sensitive to inhibition by nonnucleoside RT inhibitors (NNRTIs) (5–7). Therefore, we also assessed HIV-1^WT and HIV-1^PPT(2) susceptibility to efavirenz and rilpivirine and found that the 3′-PPT mutations conferred neither resistance nor hypersusceptibility to these drugs (Table 1). Collectively, these data underscore that the 3′-PPT mutations reported by Wijting et al. do not confer INSTI resistance, and thus it is unlikely that their selection can be attributed to the observed DTG failure in the HIV-1-infected individual.

FIG 1.

Replication capacity of HIV-1^WT, HIV-1^PPT(1), and HIV-1^PPT(2). TZM-bl cells were infected with 27 ng p24 (A) or 81 ng p24 (B) of each virus. A single-copy assay for extracellular virion-associated HIV-1 RNA was used to quantify virus production, as described previously (9). Data are reported as the average of two independent experiments.

TABLE 1.

Susceptibility of HIV-1 containing mutations in the 3′-PPT to INSTIs, efavirenz, and rilpivirine

Inhibitor by drug class	EC50 (nM)a of:		Fold changeb
	HIV-1WT	HIV-1PPT(2)
INSTI
DTG	0.66 ± 0.23	0.85 ± 0.42	1.3
Cabotegravir	0.32 ± 0.06	0.37 ± 0.02	1.1
Elvitegravir	0.75 ± 0.51	0.76 ± 0.50	1.0
Raltegravir	4.75 ± 0.68	5.16 ± 0.82	1.1
NNRTI
Efavirenz	0.91 ± 0.05	1.08 ± 0.50	1.2
Rilpivirine	0.42 ± 0.08	0.52 ± 0.18	1.2

^aConcentrations of drug required to inhibit viral replication by 50% (EC₅₀). Data reported as a mean ± standard deviation from at least 3 independent experiments.

^bMean fold change in EC₅₀ of mutant versus wild-type (WT) virus. EC₅₀ values between HIV-1^WT and HIV-1^PPT(2) were compared for statistically significant differences (P < 0.05) using a t test. None were found to be statistically significant.

Clinical management of HIV-1 infection relies on genotypic analyses of drug resistance that inform on the utility of including a specific drug class in a combination regimen. In this regard, it is important to know whether mutations outside the HIV-1 integrase gene contribute to INSTI resistance. Malet et al. (8) recently reported that the selection of mutations in the 3′-PPT in individuals failing INSTI resistance was rare, and our study highlights that if they do arise, they do not necessarily contribute to INSTI resistance. As such, genotypic analysis of the 3′-PPT in HIV-1 resistance to INSTIs is not warranted.