Human immunodeficiency virus type 1 (HIV-1) integrase: resistance to diketo acid integrase inhibitors impairs HIV-1 replication and integration and confers cross-resistance to L-chicoric acid

Abstract

The diketo acids are potent inhibitors of human immunodeficiency virus (HIV) integrase (IN). Mutations in IN, T66I, S153Y, and M154I, as well as T66I-S153Y and T66I-M154I double mutations, confer resistance to diketo acids (D. J. Hazuda et al., Science 287:646-650, 2000). The effects of these IN mutations on viral replication, enzymatic activity, and susceptibility to other HIV inhibitors are reported herein. By immunofluorescence assay and real-time PCR, all mutant viruses demonstrated a modest delay in viral spread compared to that of reference HIV. These viruses also showed a statistically significant defect in integration without defects in reverse transcription. Recombinant IN containing S153Y, T66I, and M154I-T66I mutations had an approximately twofold decrease in both disintegration and 3'-end-processing-strand transfer activities in vitro. In contrast, IN containing M154I demonstrated a greater than twofold increase in specific activity in both reactions. All mutant HIVs were resistant to l-chicoric acid, a dicaffeoyltartaric acid IN inhibitor, both in tissue culture and in biochemical assays, yet remained susceptible to the reverse transcriptase inhibitors zidovudine and nevirapine. Thus, IN mutations conferring resistance to the diketo acids can yield integration defects, attenuated catalysis in vitro, and cross-resistance to l-chicoric acid.

FIG. 1.

Replication kinetics of mutant HIV determined by IFA and with RT. Equal amounts of HIV as determined by RT activity were inoculated onto H9 cells in triplicate reactions. (A) HIV antigen synthesis was monitored by IFA, and the percentage of HIV antigen-positive cells was measured. (B) Viral supernatant from each culture was cleared of cells, and RT activity was measured from lysed virions by quantifying [³H]dTTP incorporation into a poly(rA)-oligo(dT) template. Points are the means of triplicate infections. Error bars are ±1 standard deviation.

FIG. 2.

Replication kinetics of mutant HIV determined by real-time PCR. Equal amounts of HIV_NL4-3 (solid circles), NL4-3_IN:M154I (open inverted triangles), and NL4-3_{IN:T66I-M154I} (solid squares) were inoculated onto H9 cells in triplicate reactions. Next, 10⁶ cells were lysed and real-time PCR was performed at each time point. (A) Infected-cell equivalents of minus-strand strong-stop DNA (AA55-M667) with standard curve (inset); (B) completely synthesized HIV cDNA (M661-M667) with standard curve (inset); (C) two-LTR-circle DNA (MH535-MH536) with standard curve (inset); (D) mean ratio of two-LTR-circle DNA to cDNA. Each point is the mean of triplicate infections, and error bars are ±1 standard deviation. The number of replicates for each standard curve is indicated in each insert.

FIG. 3.

HIV containing T66I-S153Y is defective for integration. Integrated HIV cDNA was amplified from cellular lysates of triplicate infections with Alu and M661 primers followed by real-time PCR with the AA55 and M667 primers. (A) Nested PCR. In the first round of amplification, Alu primers bind to human Alu sequences and M661 binds within HIV gag. Twenty cycles of PCR in primer excess results in linear amplification of integrated HIV. The PCR products are of differing sizes dependent upon how close to Alu a virus integrates (N_n). In the second round, performed under real-time conditions, internal primers that amplify HIV DNA, AA55 and M667, generate products of a single size, 140 bp, which can be detected using SYBR Green I. (B) Infected-cell equivalents of integrated HIV DNA from HIV_NL4-3 (closed circles) and NL4-3_{IN:T66I-S153Y} (open circles) with standard curve (inset). Each point in panel B is the mean of triplicate infections. The Alu amplification to generate the standard curve was performed three separate times; n is the number of AA55-M667 replicates, and dashed lines are 95% confidence intervals. Error bars are ±1 standard deviation.

FIG. 4.

SDS-PAGE analysis of recombinant HIV IN. Recombinant IN was transformed into BL21 pLysS cells, induced with 0.3 mM IPTG (isopropyl-β-d-thiogalactopyranoside) for 4 h, and purified over a Ni²⁺ affinity column. Lanes: M, markers; 1, IN from HIV_NL4-3; 2 to 6, IN containing mutation(s) T66I (lane 2), S153Y (lane 3), M154I (lane 4), T66I-M154I (lane 5), and C280S (lane 6). An approximately 32-kDa band corresponding to purified IN was recovered for each protein (arrow). Numbers at left are molecular masses in kilodaltons.

FIG. 5.

Enzymatic activities of recombinant IN proteins. Increasing amounts of recombinant IN were incubated in the presence of oligonucleotide substrates for 1 h at 37°C in triplicate reactions. Products were separated by denaturing PAGE. (A) dBY-1 dumbbell disintegration substrate is comprised of viral DNA and target DNA components which, upon addition of IN, are resolved into their respective parts; (B and C) representative gels from disintegration assays are shown for reference IN (B) and IN containing T66I-M154I (C); (D) V1-V2, the 21-mer corresponding to the viral LTR, undergoes 3′-end processing and strand transfer in the presence of IN; (E) reference IN; (F) IN containing T66I-M154I. Numbers at the top of each lane indicate nanomolar enzyme concentrations. S, substrate control; *, substrate; →, disintegration product; −2→, 3′-end processing product; S.T.P., strand transfer products.

FIG. 6.

Specific activities of recombinant IN proteins. Panels show linear regression analysis of representative enzyme activity assays used to determine specific activity (picomoles of product per hour per picomole of IN). Solid circles, reference IN; open triangles, IN_M154I; open squares, IN_T66I-M154I. (A) Disintegration; (B) 3′-end-processing-strand transfer assays. Points are the means of triplicate reactions, and error bars are ±1 standard deviation.

FIG. 7.

Activities of recombinant IN proteins at 100 nM. A 100 nM concentration of each recombinant IN was incubated in the presence of: dBY-1 (A) or V1-V2 (B) substrates for 1 h at 37°C in triplicate reactions. Products were separated by denaturing PAGE. Lanes: 1, IN_C280S; 2, IN_T66I-M154I; 3, IN_M154I; 4, IN_S153Y; 5, IN_T66I; 6, reference IN. S, substrate control; *, substrate; →, disintegration product; −2→, 3′-end-processing product; S.T.P., strand transfer products.

FIG. 8.

IN containing T66I-M154I mutations is resistant to l-731,988 and l-CA. Recombinant IN protein was incubated for 1 h at 37°C in triplicate reactions in the presence of oligonucleotide substrate and an 0.03 to 10.0 μM concentration of either l-731,988 or l-CA. Representative results used to determine each IC₅₀ (Table 4) are shown. (A and B) 3′-end processing-strand transfer for IN from HIV_NL4-3 (A) or IN_T66I-M154I (B) in the presence of l-731,988. (C) Quantification of data from triplicate experiments shown in panels A and B. Inset shows semilog plots used to derive IC₅₀s. (D and E) 3′-end processing-strand transfer for reference IN (D) or IN_T66I-M154I (E) in the presence of l-CA. (F) Quantification of data from triplicate experiments shown in panels D and E. Inset shows semilog plots used to derive IC₅₀s. +, 25 μM l-CA; −, 25 μM l-tartaric acid; E, substrate with enzyme; S, substrate only. Products were resolved by denaturing PAGE and quantified by phosphorimager analysis. *, substrate; →, −2 products; S.T.P., strand transfer products. Numbers above the lanes are the concentrations (micromolar) of each inhibitor.