Identification of novel mutations responsible for resistance to MK-2048, a second-generation HIV-1 integrase inhibitor

Abstract

MK-2048 represents a prototype second-generation integrase strand transfer inhibitor (INSTI) developed with the goal of retaining activity against viruses containing mutations associated with resistance to first-generation INSTIs, raltegravir (RAL) and elvitegravir (EVG). Here, we report the identification of mutations (G118R and E138K) which confer resistance to MK-2048 and not to RAL or EVG. These mutations were selected in vitro and confirmed by site-specific mutagenesis. G118R, which appeared first in cell culture, conferred low levels of resistance to MK-2048. G118R also reduced viral replication capacity to approximately 1% that of the isogenic wild-type (wt) virus. The subsequent selection of E138K partially restored replication capacity to approximately 13% of wt levels and increased resistance to MK-2048 to approximately 8-fold. Viruses containing G118R and E138K remained largely susceptible to both RAL and EVG, suggesting a unique interaction between this second-generation INSTI and the enzyme may be defined by these residues as a potential basis for the increased intrinsic affinity and longer "off" rate of MK-2048. In silico structural analysis suggests that the introduction of a positively charged arginine at position 118, near the catalytic amino acid 116, might decrease Mg(2+) binding, compromising enzyme function and thus leading to the significant reduction in both integration and viral replication capacity observed with these mutations.

FIG. 1.

Clinical structures of RAL, EVG, and MK-2048.

FIG. 2.

Dose-response curves for MK-2048, RAL, and EVG against resistant viruses. Dose-response curves for MK-2048 (A), RAL (B), and EVG (C) are shown for wild-type, G118R, E138K, and G118R E138K viruses. Data are fitted to sigmoidal dose-response (variable slope) curves, from which EC₅₀ values were calculated. Data are from three experiments, and error bars represent the standard error of the mean (SEM).

FIG. 3.

Fold change in EC₅₀s versus the wild type. The fold change in EC₅₀ values for each drug is shown in terms of fold change versus wild-type virus for that drug (wild type = 1).

FIG. 4.

Replication capacity of wild-type and mutant viruses. Replication capacity was determined by comparing levels of RT activity in culture supernatants between mutated and wild-type viruses at 72 h postinfection. Error bars represent SEM.

FIG. 5.

Reverse transcription and integration evaluated by real-time PCR. (A) Early reverse transcription products at 7 and 24 h postinfection; (B) late reverse transcription products at 7 and 24 h postinfection; (C) 2-LTR circle accumulation at 24 h postinfection; (D) integrated viral DNA at 24 h postinfection.

FIG. 6.

Ribbon representation of the integrase catalytic core. Shown are G118R and E138K. Dark blue, Mg²⁺ cation; red, catalytic amino acids 64 and 116; light blue, amino acids 118 and 138.

FIG. 7.

Electrostatic surface representation of the integrase catalytic core. (A) G118R and E138K structure. On the surface, red represents negatively charged elements, white represents neutral elements, and blue represents positively charged elements. Green, Mg²⁺ cation. Underneath the semitransparent surface, side chains and amino acids 64, 116, 118, and 138 are shown. The arrow indicates amino acids 118 and 138 and the area of charge change as a result of the amino acid substitution.