Inhibition of Human Immunodeficiency Virus Type 1 Isolates by the Integrase Inhibitor L-731,988, a Diketo Acid

Abstract

L-731,988 inhibits human immunodeficiency virus (HIV) replication through integrase. In this study, approximately 600 nM L-731,988 inhibited the replication of 12 HIV type 1 isolates from multiple clades, including primary isolates and cloned viruses. These data suggest that diketo acids or their derivatives may prove useful on a worldwide basis in treating HIV infection.

The human immunodeficiency virus (HIV) genome encodes three enzymes: protease (PR), reverse transcriptase (RT), and integrase (IN). Agents targeted at RT and PR inhibit viral replication within infected individuals (1, 2, 4). Despite the success of PR and RT inhibitors, recent evidence suggests that the transmission of drug-resistant HIV is increasing and that its rate of transmission can approach 25% (13), making it clear that new therapeutic targets must be identified. IN is an enzyme unique to retroviruses and is required for productive retroviral infection. Given the latter, as well as the clinical effectiveness of other retroviral enzyme inhibitors, IN is an attractive therapeutic target. Recently, the diketo acids were found to curb viral replication through the inhibition of IN (6). Amino acid changes that lead to resistance to diketo acids are located within IN and have been documented previously for HIV type 1 (HIV-1) (6). Analysis of the HIV sequence database shows that changes at 2 of the 3 amino acids conferring resistance to diketo acids are present within the infected population. To investigate the role that natural variation has in the susceptibility of HIV to diketo acids, a series of primary, cloned, and laboratory HIV-1 isolates were examined for susceptibility to L-731,988 (a diketo acid) (Fig. 1) in both cultured cell lines and peripheral blood lymphocytes (PBLs).

FIG. 1.

Chemical structure of L-731,988.

Laboratory-adapted strains of HIV and the infectious molecular clone HIV_NL4-3 can replicate in cultured cell lines, unlike most primary isolates of HIV. Therefore, the susceptibilities to L-731,988 of laboratory-adapted strains (HIV_LAI and HIV_MVP5180) and HIV_NL4-3 clones containing different IN genes (HIV_NL4-3:IN7-3, HIV_NL4-3:INR104, HIV_NL4-3:INR106, and HIV_{NL4-3:IN92NG003}) were determined on cultured MT2 cells by using a vital dye assay (8, 10, 12).

HIV_LAI and HIV_NL4-3:IN7-3 are clade B, laboratory-adapted HIV strains that are susceptible to diketo acids (6; R. Reinke et al., unpublished data). Cloned isolates were constructed with HIV_NL4-3, which contains two unique restriction sites that allow the native IN to be excised and the IN gene of interest to be cloned in frame into the plasmid vector (7). HIV_NL4-3:IN7-3 contains the HIV_NL4-3 IN gene and was subjected to the same tissue culture and cloning procedures used for the other viruses (7) to control for changes due to culturing of virus and cloning. HIV_NL4-3:INR104 and HIV _NL4-3:INR106 contain IN genes derived from the North American clinical isolates described previously (9). HIV_{NL4-3:IN92NG003} contains an IN gene derived from a near-full-length clone of HIV originally isolated in Nigeria (3), and HIV_MVP5180 is a group O isolate from Cameroon (5). Nomenclature used throughout indicates the specific IN gene cloned into HIV_NL4-3: HIV_NL4-3:INR106 contains the IN gene from clinical isolate R106.

HIV-1 primary isolates (HIV_92RW016, HIV_92RW009, HIV_98IN026, HIV_93IN101, HIV_92THA005, HIV_98CN006) were obtained through the National Institutes of Health AIDS Research and Reference Reagent Program and passaged on phytohemagglutinin-stimulated PBLs with interleukin-2; each virus supernatant was collected when cultures exhibited peak staining with polyclonal anti-HIV serum by an immunofluorescence assay. Supernatants were filtered through 0.45-μm-pore-size filters and assayed for exogenous RT activity. Viral supernatants (36,000 cpm of RT activity) were incubated with 0, 0.5, 1.0, and 5.0 μM L-731,988 for 30 min at 37°C. Following incubation, viral supernatants were added to 1.5 × 10⁶ PBLs containing either 0, 0.5, 1.0, or 5.0 μM L-731,988 in a total volume of 2 ml. All infections were performed in triplicate in 24-well plates. Culture supernatants were monitored by indirect immunofluorescence assay and RT assay (11) every 2 days until RT activity within control infections without L-731,988 peaked. The percent inhibition at each drug concentration was calculated on the basis of the control infections. The 50% effective doses (ED₅₀) were calculated for the viruses when RT values for the control infections were at maximum. ED₅₀ were calculated using the CalcuSyn for Windows software package.

All six viruses tested in the MT2 cell assay were susceptible to the inhibitor, exhibiting only minor differences (two- to threefold) in ED₅₀ (Table 1). ED₅₀ for these isolates ranged from 0.5 to 1.6 μM. Similarly, the ED₅₀ of L-731,988 against the primary isolates ranged from 0.6 to 0.9 μM, with the exception of that for HIV_92RW009, which increased about sixfold compared to that for HIV_LAI (Table 1). Two assays were utilized in this study to examine the susceptibilities of natural variant IN sequences to L-731,988. Despite the differences between the two assays, the data for a control virus, HIV_LAI, correlate between the assays (Table 1) (0.51 to 0.65 μM), demonstrating that the antiviral data from both assays are directly comparable.

TABLE 1.

Susceptibilities of HIV-1 isolates to L-731,988

HIV isolate	Virus cladea	Origin	IN amino acid diver- gence (%)b	Anti- HIV assayc	ED50d (μM)
LAI	B	United States		MT2	0.505
NL4-3:IN7-3	B	United States		MT2	0.6
MVP5180	Group O	Cameroon	15	MT2	1.59
NL4-3:INR104	B	United States	3.5	MT2	0.637
NL4-3:INR106	B	United States	2	MT2	1.27
NL4-3:IN92NG003	A/G	Nigeria	5.2	MT2	0.685
LAI	B	United States		PBL	0.654
92RW016	A	Rwanda	6.6	PBL	0.590
92RW009	C/A	Rwanda	5.0	PBL	3.14
98IN026	C	India	6.3	PBL	0.544
93IN101	C	India	5.2	PBL	0.917
92THA005	A/E	Thailand	5.2	PBL	0.628
98CN006	C	Central African Republic	5.9	PBL	0.923

^aClades based on gag/env sequences.

^bReference sequence is HIV_NL4-3. Divergence was calculated based on the number of substitutions out of 288 amino acids.

^cMT-2 assays were performed at least twice in triplicate (six replicate infections). PBL assays were performed using lymphocytes from one donor in triplicate infections.

^dED₅₀ were calculated from the means of values from triplicate infections by using the CalcuSyn for Windows software package.

The ability of compounds to inhibit primary isolates of HIV is of paramount importance in their development as drugs. The data presented here show that a group O isolate, group M primary viral isolates representing clades A, B, and C, and a clade A/E recombinant were all susceptible to L-731,988. Compared to the control HIV_LAI, most of these viruses exhibited a less-than-twofold difference in susceptibility to L-731,988. In contrast, the isolate HIV_92RW009, a C/A recombinant, showed an approximately sixfold decrease in susceptibility to L-731,988. While complete resistance was not observed, this partial-resistance phenotype was detected at all drug concentrations tested, as evidenced by less growth inhibition than with other viruses (Fig. 2). IN sequence analysis of HIV_92RW009 failed to reveal amino acid changes matching previously identified resistance mutations for either the diketo acid or the l-chicoric acid class of IN inhibitors (C65, T66, G140, S153, or M154) (6, 7, 14). These data suggest that other amino acid changes within the HIV_92RW009 IN may contribute to resistance to diketo acids. Amino acid changes unique to HIV_92RW009 IN were a valine-to-alanine change at position 79, a threonine-to-asparagine change at position 124, and a lysine-to-arginine change at position 173. However, the peak viral RT counts for the HIV_92RW009 cultures were greater than for other viruses tested. Thus, an increased burst size may contribute to the resistance phenotype.

FIG. 2.

Representative susceptibilities of primary HIV isolates to L-731,988 in PBLs. PBLs were inoculated with HIV in the presence of 0.5, 1.0, or 5 μM L-731,988, and RT activity was measured. (A) Isolate HIV_98CN006; (B) isolate HIV_92RW009. Abbreviations: VC, virus control; DKA, diketo acid. Each point represents a mean of values from three independent infections, with error bars corresponding to 1 standard deviation.

To date, three mutations have been associated with resistance to the diketo acids: T66I, S153Y, and M154I (6). Two of the viruses tested contain amino acid changes at positions 153 and 154 of IN. Each virus had other mutations that could account for the resistance phenotype, but in particular, HIV_MVP5180 contains an S153A change and HIV_NL4-3:INR104 has an M154L change. HIV_MVP5180 does show an increase in ED₅₀ of roughly threefold, nearly identical to that observed by Hazuda et al. (6) for the S153Y mutation. Although S153 and M154 have been associated with resistance to diketo acids, the changes contained within HIV_MVP5180 and HIV_NL4-3:INR104 do not dramatically affect their susceptibility to L-731,988, suggesting that mutations occurring at these positions require additional changes, perhaps at position 66 (6), or specific amino acid substitutions for a high-resistance phenotype. Interestingly, S153Y conferred resistance to another diketo acid, L-708,906, but not to L-731,988 (6). It should be noted that if resistance occurs through the synergistic effect of mutations at positions 66, 153, and 154, as suggested by Hazuda et al. (6), the variability at positions 153 and 154 within the natural population may result in a more rapid development of resistance phenotypes. Nevertheless, 15% amino acid divergence for HIV_MVP5180 results in only a threefold decrease in susceptibility to L-731,988. Overall, these data suggest that naturally occurring genetic variation within the IN gene is not a major obstacle in the development of the diketo acids for therapeutic use.

Nucleotide sequence accession numbers

The sequences of all IN genes reported herein have been deposited with GenBank. The accession numbers are R104-AF203330, R106-AF203331, 92NG003-U88825, 93IN101-AF500777, 98CN006-AF500778, 98IN026-AF500779, 92RW009-AF500780, 92RW016-AF500781, and 92THA005-AF500782.