ABSTRACT
In the absence of an effective vaccine against the human immunodeficiency virus (HIV), topical microbicides to prevent the sexual transmission of HIV represent an important strategy to prevent the continued spread of infection. The recent trend in the development of new microbicide candidates includes the utilization of FDA-approved therapeutic drugs that target the early stages of the HIV life cycle, including entry inhibitors and reverse transcriptase inhibitors. We have investigated 12 pyrimidinedione compounds with potent HIV activities and their abilities to inhibit both virus entry and reverse transcription, in an effort to determine a lead microbicide for product development. The candidate compounds were evaluated for efficacy against subtype B, C, and E clinical virus strains in fresh human peripheral blood mononuclear cells and against CCR5-tropic virus strains in both monocyte-macrophages and dendritic cells. Microbicide-specific biological assays and toxicity evaluations were also performed in a variety of established and fresh human cells as well as against Lactobacillus strains common to the vaginal environment. These evaluations resulted in the identification of congeners with cyclopropyl and cyclobutyl substituents at the N-1 of the pyrimidinedione as the most active molecules in the structure-activity relationship series. The pyrimidinediones represent excellent microbicide candidates in light of their significantly high efficacies against HIV-1 (subnanomolar concentration range), potencies (therapeutic index, >1 million), solubility profiles, and dual mechanism of antiviral action that includes two early steps of virus replication prior to the integration of the virus that are considered most important for microbicidal activity.
INTRODUCTION
A recent report issued by UNAIDS on the global AIDS epidemic sighted significant gains in the prevention of new HIV infections in a number of countries most affected by the AIDS epidemic (35). Many of these countries have noted changes in sexual behaviors followed by a decline in the number of new HIV infections. These behavioral changes include increasing condom use among young people with multiple partners and encouraging signs that young people are waiting longer to have sexual intercourse in some of the most heavily affected countries (35). However, the report also stated that despite these declines in new HIV infections, the AIDS epidemic is far from over and that the rates of new HIV infections continue to rise in many other countries. HIV/AIDS continues to be the leading cause of death in Africa (35).
Despite intensive research, a vaccine to prevent HIV infection has not yet been successfully developed. In the absence of a prophylactic vaccine, antiretroviral therapy (ART) has been the primary mechanism utilized to prevent disease progression and prolong the survival of infected individuals. A wide variety of antiretroviral (ARV) agents are currently approved for use, targeting numerous steps in HIV replication, such as attachment, fusion, reverse transcription, integration, and proteolytic processing. With the use of combination therapy strategies, resulting in the evolution of highly active antiretroviral therapies (HAART), profound suppression of plasma viral loads for prolonged periods of time have resulted in significant declines in AIDS-related morbidity and mortality in developed countries, where the therapies can be afforded (24, 30). Although HAART regimens have improved the prognosis for HIV-infected individuals, challenges to effective use of these therapeutic strategies remain, including issues of adherence, side effects and toxicities, drug resistance, and persistent viral replication in latent reservoirs. These challenges necessitate the development of alternative strategies to combat the spread of HIV, including prevention strategies in the form of vaginal and rectal topical microbicides.
With the incidence of HIV infections in women on the rise, especially in underdeveloped countries (35), the development and clinical evaluation of topical microbicides has achieved worldwide focus. It has been suggested that a microbicide with 60% efficacy introduced into 73 low-income countries could prevent 2.5 million HIV infections over 3 years (37). For these reasons, topical microbicides are an ideal HIV prevention strategy that would allow a woman to take responsibility for her own protection from sexually transmitted disease (3, 34), especially in societies where women are unempowered. Minimally, a microbicide product should be safe, acceptable by the user, affordable, stable, and effective without compromising existing physical barriers to sexually transmitted infections (34). Microbicides are currently being developed in gel, cream, film, suppository, sponge, and intravaginal ring formats, providing both a physical barrier that the virus must penetrate as well as specifically targeting critical steps of the virus replication cycle, such as virus entry and reverse transcription (33). Six microbicides have progressed to late-stage clinical development; however, these products, including nonoxynol-9, C31G (Savvy), cellulose sulfate (Ushercell), BufferGel, PRO2000, and carageenan (CarraGuard), failed to demonstrate efficacy in clinical trials (20a, 23a, 26a). Most recently, a successful phase 3 clinical trial of the antiretroviral agent tenofovir formulated as a gel product resulted in protection of women and energized the microbicide field, as it suggested that highly potent ARVs might result in highly successful microbicide products, especially when combined into coformulated microbicide products (1). Additionally, it has recently been reported that systemic prophylactic administration of the highly potent ARV Truvada (a combination product consisting of tenofovir and emtricitabine) to men who have sex with men (MSM) provided protection from HIV transmission (20).
The recent failures of the initial microbicides focused critical attention on continued microbicide development and the types of products that must be developed. The recent trend in the development of new microbicide candidates includes the utilization of FDA-approved drugs for HIV therapy that target the early stages of the HIV life cycle, including entry inhibitors and reverse transcriptase (RT) inhibitors (17, 25). As of October 2009 there were 38 ongoing and planned clinical trials for the development of microbicides, and 6 of these included ARV candidates (10). ImQuest has licensed and has begun development of a pyrimidinedione (PYD) series of ARVs which have been shown to have potent activities against the HIV-1 reverse transcriptase and also possess a second mechanism of action that targets virus entry and extends the range of action to HIV-2 (5, 9). The molecules have been evaluated in a wide variety of anti-HIV in vitro efficacy and toxicity assays, and the results have defined 12 microbicide candidates (4, 5). These 12 superior compounds have been further evaluated in more specific microbicide assays in an effort to rationally prioritize the various PYD molecules for advanced preclinical and clinical development. The data generated from these experiments and included in our study indicate that the pyrimidinediones as a class have similar, if not superior, in vitro activity profiles to the ARV microbicide candidates currently being developed (18, 19, 27). These results, combined with the dual mechanism of action that targets virus entry and reverse transcription, make the PYD series of molecules highly attractive candidates as topical microbicides. Our results suggest that PYD analogs with cyclopropyl and 3-cyclopenten-1-yl substitutions at the N-1 of the pyrimidinediones represent highly attractive candidates for further development as topical anti-HIV microbicides.
MATERIALS AND METHODS
Cell lines, virus, and bacteria.The CEM-SS (29), HeLa-CD4-LTR-β-galactosidase (22), GHOST X4/R5 (26), and H9 (32) cell lines, as well as the HIV-1IIIB (CXCR-4 tropic) (32) and CCR5-tropic clinical virus isolates (clade B, US/92/727 and HIV-1BaL; clade C, ZA/97/003, 93/MW/959, and 93/IN/101; clade E, CMU06, 92/TH/020, and 93/TH/07) were obtained from the NIAID AIDS Research and Reference Reagent Program (Rockville, MD). The ME180 (21, 31) cells and Lactobacillus strains (L. crispatus ATCC 33820, L. jensenii ATCC 25258, and L. acidophilus ATCC 11975) were obtained from American Type Culture Collection (Manassas, VA). The HIV-1SK-1 (8, 14, 21, 31) used for developing chronically infected H9 cells (H9/SK-1) was a gift to ImQuest BioSciences from Duke University (Durham, NC). Human peripheral blood mononuclear cells (PBMCs), monocyte-macrophages, and dendritic cells were derived from human blood, which was purchased from Biological Specialty Corporation (Colmar, PA). The cell lines were propagated as recommended and stored in liquid nitrogen; stocks of HIV-1IIIB and clinical virus strains were stored at −80°C prior to being used in the antiviral assays.
Materials.Twelve of 68 active pyrimidinedione compounds (5) were used in these evaluations (Fig. 1). These compounds were defined as the most promising potential microbicides based on in vitro evaluations. Each compound was obtained from Samjin Pharmaceutical Company, Ltd., as a dry white powder. The compounds were solubilized at 1 or 40 mM in 100% dimethyl sulfoxide and stored at 4°C. Zidovudine (AZT) (28), dextran sulfate (molecular weight, 500,000) (2), and Chicago Sky Blue (13) were used as experimental control compounds and were obtained from Sigma-Aldrich Corporation (St. Louis, MO).
Structure of candidate pyrimidinedione microbicides. The table shows the 12 representative pyrimidinediones with the chemical substituents noted at each side chain position (R, R1, R2, R3, and X). Abbreviations: Et, ethyl; Me, methyl; iPr, isopropyl.
CPE inhibition assay.The cytopathic effect (CPE) assay was performed as previously described (6). Briefly, serially diluted compound was added to a 96-well round-bottom microtiter plate in triplicate. CEM-SS cells at a concentration of 2.5 × 103 cells per well and HIV-1IIIB at the appropriate predetermined titer were sequentially added to the microtiter plates. The cultures were incubated at 5% CO2–37°C for 6 days. Following the incubation, the microtiter plates were stained with sodium 2,3-bis(2-methoxy-4-nitro-5-[(phenylamino)-carbonyl]-2H-tetrazolium dye (XTT; Sigma, St. Louis, MO) to evaluate the efficacy and toxicity of the test compounds. AZT was evaluated in parallel as a positive control for the assay.
Virus entry inhibition assay.The virus entry inhibition assay was performed as previously described (36). Twenty-four hours prior to compound and virus addition, HeLa-CD4-LTR-β-galactosidase (MAGI) cells diluted in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and antibiotics were plated in 96-well flat-bottom microtiter plates at 1 × 104 cells per well in a volume of 100 μl and incubated overnight at 37°C–5% CO2. Following the overnight incubation, each test compound was serially diluted and added to the cells in triplicate in a volume of 50 μl per well. HIV-1IIIB was diluted to a predetermined titer in assay medium and added to the microtiter plates. The cultures were incubated for 2 h at 37°C–5% CO2. Following the incubation, the monolayers were washed three times with RPMI 1640 (without additives) to remove residual extracellular compound and unbound virus and incubated at 37°C–5% CO2 for an additional 48 h. Following the incubation, toxicity plates were evaluated by XTT staining, as previously described (6), and efficacy plates were evaluated by chemiluminescence detection using the Gal-Screen system (Applied Biosystems) according to the manufacturer's instructions. Chicago Sky Blue was evaluated in parallel as a positive control for the assay.
RT inhibition assay.The RT inhibition assay was performed as previously described (9). A 2× buffer solution was prepared that contained 2 M Tris (pH 8.0), 3 M KCl, 1 M MgCl2, 2 M dithiothreitol, 25 U/ml recombinant rCdG, 1 mM dGTP, and 800 Ci/mM [α-32P]dGTP. The compounds were diluted serially 10-fold in water, and 30 μl was added in triplicate to a 96-well round-bottom tissue culture plate. Fifty microliters (50 μl) of the 2× buffer was added to each sample, including the positive and negative controls. Twenty microliters of HIV-1 wild-type RT enzyme (CHIMERx) diluted in water containing bovine serum albumin (1:100) and Triton-X (1:1,000) was added to all wells except the negative control. The plates were incubated for 50 min at 37°C. Following the incubation, 10 μl of 10-mg/ml DNA (molecular biology-grade fish sperm) was added to each well, followed by the addition of 150 μl of cold 10% trichloroacetic acid (TCA). The plate was incubated at room temperature for approximately 15 min in order to allow for precipitation of the samples. Following the incubation, the contents were transferred to filter plates containing DEAE-filter paper, and contents were filtered using a vacuum manifold. The plate was rinsed two additional times with 200 μl of 10% TCA as described above. The plate was transferred to a reading cassette, and 20 μl of Wallac Supermix scintillant was added to each well. The plate was covered with a plate sealer and read on a Wallac MicroBeta scintillation counter. The nonnucleoside RT inhibitor (NNRTI) UC-38 was evaluated in parallel as a positive control for the assay.
Anti-HIV assay in fresh human peripheral blood mononuclear cells.PBMC-based anti-HIV assays were performed as previously described (9). Briefly, phytohemagglutinin-stimulated PBMCs cultured in the presence of interleukin-2 were suspended at 1 × 106 cells/ml and were added to a 96-well round-bottom plate. Serially diluted test materials were added to the plate in triplicate, followed by the appropriate predetermined titer of the strain of HIV. The culture was incubated for 7 days at 37°C–5% CO2. Following the incubation, supernatants were collected for analysis of virus replication based on supernatant RT activity, and cells were analyzed for viability by XTT dye reduction. AZT was used as an internal assay standard. Clinical subtypes of virus that represent HIV strains found in geographic locales where microbicide products are most likely to be clinically evaluated (clades B, C, and E) were utilized.
Anti-HIV assay in fresh human monocyte-macrophages.Monocyte-macrophage-based anti-HIV assays were performed as previously described (9). Briefly, monocyte-macrophages derived from human PBMCs (4 × 106 cells/well) were plated for 5 to 7 days at 37°C–5% CO2. On the day of assay, the cell monolayer was gently washed with Hanks balanced salt solution (HBSS) or Dulbecco's PBS (DPBS) several times to eliminate residual PBMCs. After the final wash, 50 μl of supplemented RPMI 1640 was placed in each well, and 100 μl of compound prepared at 2 times the high test concentration was transferred to the flat-bottom 96-well plate containing the cells. A predetermined titer of HIV-1BaL was added to the appropriate wells in a 50-μl volume. Following the 24-h incubation, virus was removed by washing with HBSS or DPBS. After 7 days in culture, HIV-1 replication was quantified by the measurement of HIV-1 p24 by using a commercial antigen capture enzyme-linked immunosorbent assay (ELISA) for the HIV p24 antigen (Perkin-Elmer). Compound cytotoxicity was evaluated using the tetrazolium dye XTT. AZT was evaluated in parallel as a control for the assay.
Anti-HIV assay in MO-DCs.Dendritic cell-based anti-HIV assays were performed as previously described (9). Monocyte-derived dendritic cells (MO-DCs) derived from fresh human PBMCs were suspended at 1 × 107 cells/ml in RPMI 1640 with 10% FBS, 2 mM l-glutamine, 25 mM HEPES, 100 U/ml penicillin, 100 μg/ml streptomycin and incubated for 2 h at 37°C–5% CO2 with HIV-1BaL at a multiplicity of infection of 50 to 150 50% tissue culture infective doses. Following the incubation, cells were washed six times and resuspended at 1 × 106 cells/ml in complete medium. Aliquots (100 μl) of MO-DCs were dispensed in a 96-well round-bottom plate. Compounds to be evaluated (100-μl aliquots) prepared in a half-log dilution series in complete medium were added immediately after infection. Following 7 days in culture at 37°C–5% CO2, HIV-1 replication was quantified by the measurement of HIV-1 RT activity. Compound cytotoxicity was determined using the tetrazolium dye XTT. AZT was evaluated in parallel as a positive control for the assay.
MTSA.The microbicide transmission and sterilization assay (MTSA) was performed as previously described (36). Briefly, CEM-SS cells at a concentration of 5 × 105 cells per well and the appropriate predetermined titer of HIV-1IIIB were cocultured for 1 h at 37°C in a 96-well round-bottom microtiter plate. Following the initial infection, the cells and virus suspension were transferred to a T25 tissue culture flask containing 5 ml of tissue culture medium. Five or six concentrations of test or control compound were added to appropriate individual flasks. Three days after the initial infection, cell-free supernatant samples collected from each flask were evaluated for virus content by the RT assay (6), and the cells were subcultured by transferring 20% of the existing culture (1 ml of resuspended cells and supernatant) to 4 ml of fresh cells (at a cell density of 1 × 105 cells/ml) in fresh tissue culture medium containing test compound at the identical fixed compound concentration. Each subculturing, performed at 3-day intervals, was considered a new passage. Subculturing of the infected cultures was performed every third day for a total of 15 passages. Cells in passages 11 through 15 were cultured in the absence of test compound to confirm complete virus sterilization of the culture. The results of each assay are reported as the number of passages determined to be positive for virus replication in the flask (i.e., a value of 0 indicates complete virus sterilization, whereas a value of 15 indicates no inhibition of virus replication).
CD4-independent virus transmission inhibition assays.The CD4-independent virus transmission inhibition assay was performed as previously described (36). ME180 cells resuspended in RPMI 1640 medium with 10% FBS and antibiotics were plated at a density of 5 × 103 cells per well in a 96-well flat-bottom microtiter plate and incubated overnight at 37°C–5%CO2. Following the incubation, test compound was serially diluted in assay medium and added in triplicate to the cells. Thoroughly washed, mitomycin C-treated chronically HIV-infected H9 cells (H9/SK-1) were added at 2 × 104 cells/well, and the plates were incubated at 37°C–5% CO2 for 4 h. The mitomycin C concentration employed was that which resulted in complete killing of the chronically infected cells within 24 h, to prevent virus production from the H9/SK-1 inoculum from contributing to the end point of the assay. Immediately following the infection period, and again at 24 and 48 h postinfection, the cell monolayers were washed with RPMI 1640 (without additives) to remove residual compound and virus-infected cells. Following a 6-day incubation at 37°C–5% CO2, cell-free supernatant samples from the cell cultures were evaluated for virus content using the HIV-1 p24 antigen ELISA kit (Beckman Coulter) according to the manufacturer's instructions. Duplicate assay plates were evaluated for cellular toxicity by XTT staining. Dextran sulfate was evaluated in parallel as a positive control for the assay.
Cell-free virus and CD4-dependent transmission inhibition assay.The cell-free virus and CD4-dependent transmission inhibition assay was performed as previously described (36). GHOST X4/R5 cells diluted in DMEM supplemented with 10% FBS and antibiotics were incubated in a 96-well flat-bottom microtiter plate (5 × 103 cells per well) overnight prior to assay initiation. Following the incubation the cultures were washed to remove nonadherent cells. Serially diluted test compound was added in triplicate wells, and the cells were infected with a predetermined titer of HIV-1IIIB for 4 h at 37°C. Following the incubation, residual virus and test material were removed by washing with RPMI 1640 (without additives). The cultures were incubated for 6 days, at which time antiviral activity was assessed by evaluating cell-free supernatants for virus content by RT assay as described previously (6). Cytotoxicity was evaluated in parallel using XTT dye reduction. Dextran sulfate was evaluated in parallel as a positive control for the assay.
Cell-associated virus transmission inhibition assay.The cell-associated virus transmission inhibition assay was performed as previously described (36). GHOST X4/R5 cells diluted in DMEM supplemented with 10% FBS and antibiotics were incubated in a 96-well flat-bottom microtiter plate (5 × 103 cells per well) overnight prior to assay initiation. Following the incubation, the cultures were washed to remove nonadherent cells. Serially diluted test compound was added in triplicate wells. Thoroughly washed, mitomycin C-treated chronically HIV-infected H9 cells (H9/SK-1) cells were added at 2 × 104 cells/well, and the plates were incubated at 37°C–5% CO2 for 4 h. The mitomycin C concentration employed was that which resulted in complete killing of the chronically infected cells within 24 h, to prevent virus production from the H9/SK-1 inoculum from contributing to the end point of the assay. Immediately following the infection period, and at 24 and 48 h postinfection, the cell monolayers were washed with RPMI 1640 (without additives) to remove residual compound and virus-infected cells. The cultures were incubated for 6 days, at which time antiviral activity was assessed by evaluating cell-free supernatants for virus content by RT assay as described previously (6). Cytotoxicity was evaluated in parallel based on XTT dye reduction. Dextran sulfate was evaluated in parallel as a positive control for the assay.
Lactobacillus toxicity assay.The Lactobacillus toxicity assay was performed as previously described (23). Briefly, in a 15-ml conical tube 10 ml of MRS medium was inoculated with a stab from a frozen glycerol stock of Lactobacillus crispatus, L. jenseni, or L. acidophilus and incubated for 24 h at 37°C in an anaerobic chamber. The overnight culture was diluted in MRS medium until an absorbance of 0.06 at 670 nm was obtained. Six serial -log dilutions of the compound were performed and were added in a volume of 100 μl to the plate, followed by the addition of 100 μl of the diluted bacteria. The plates were placed in an anaerobic chamber and were incubated at 37°C for 24 h. Following the incubation the plates were read spectrophotometrically at 490 nm on a Spectramax Plus 384 spectrophotometer (Molecular Devices, Sunnyvale, CA). Penicillin-streptomycin was evaluated in parallel as a positive control for the assay.
Statistical software used to generate EC50 and TC50 values.For the 50% efficacy concentration (EC50) and 50% toxic concentration (TC50) assays, raw data were collected using the Softmax Pro 4.6 (or higher) software or generated with the Microbeta Trilux software and were imported into a Microsoft Excel spreadsheet for analysis by 4-parameter or linear curve fit calculations. Both antiviral efficacy and cellular toxicity with a graphical representation of the data were provided in a plate analysis report (PAR) that summarized the activity of the experimental and control compounds; these data are expressed as the EC50 and TC50.
RESULTS
Efficacies of lead candidate PYDs.Twelve PYD congeners (Fig. 1) from a structure-activity relationship (SAR) series, which included a total of 68 PYD molecules, were evaluated in relevant primary topical microbicide assays to assess their potential as prevention agents. The 12 molecules in the series were chosen based on their relative antiviral activities against HIV-1 and HIV-2, as well as their relative abilities to inhibit virus entry and reverse transcription using wild-type infectious virus or enzyme, respectively. The most active compounds in CEM-SS cells against HIV-1IIIB included compounds IQP-0406, IQP-0407, IQP-0528, IQP-0558, IQP-0410, and IQP-1187, which exhibited subnanomolar protection against HIV replication with EC50s that ranged from 0.2 to 0.5 nM. The most active compounds against HIV-2ROD were compounds IQP-0406, IQP-0528, IQP-0549, and IQP-0558, which exhibited EC50s ranging from 100 to 500 nM. When evaluated against wild-type HIV-1 RT in a biochemical RT inhibition assay, compounds IQP-0405, IQP-0406, IQP-0529, and IQP-0558 were found to be the most active agents, with EC50s ranging from 12 to 20 nM. As expected based on mechanistic studies performed with the pyrimidinediones (unpublished data), none of the compounds were active in enzymatic assays against purified HIV-2 RT. Compounds IQP-0405, IQP-0407, IQP-0528, IQP-0554, and IQP-0558 were found to be the most active inhibitors against HIV-1IIIB entry, with EC50s ranging from 6 to 12 nM. In addition to evaluating the compounds based on concentrations that inhibited HIV replication, we also ranked the most potent compounds based on their observed in vitro therapeutic index (TI), which takes into account both efficacy against virus replication and toxicity to target cells evaluated in parallel with the same cells over the same period of time. Using TI values to compare the PYDs, the most active HIV-1 inhibitors in CEM-SS cells were IQP-0405, IQP-0407, IQP-0528, IQP-0565, IQP-0410, and IQP-1187. The compounds with the greatest TI values in the entry inhibition assay were the same, except that IQP-0558 outperformed IQP-1187. Against HIV-2 in CEM-SS cells, IQP-0405, IQP-0407, and IQP-0528 were still superior, with the highest TI values, but IQP-0531, IQP-0549, and IQP-0558 also outperformed the other PYDs. These comparative data are presented in Table 1 and provide comprehensive efficacy information to allow the prioritization of the active pyrimidinedione microbicide candidates.
Efficacies of candidate microbicides against HIV-1 and HIV-2a
Efficacies of pyrimidinediones against subtype viruses relevant to microbicide development.In standardized fresh human PBMC assays, the 12 PYDs were evaluated against HIV-1 subtypes A through G and O and were found to be highly efficacious against viruses from each subtype, with EC50s in the low to subnanomolar concentration range for each of the most active compounds (data not shown). A more extensive evaluation was performed with the 12 PYDs against low-passage, clinical HIV-1 subtypes C and E, since these are the predominant strains found in regions of the world where microbicides will most likely be primarily tested. These antiviral data are presented in Fig. 2. The activity of the 12 molecules against clinical subtype B virus (strain US/92/727) ranged from 0.13 to >10 nM, with the most active compounds being IQP-0405, IQP-0406, IQP-0407, IQP-0528, IQP-0410, and IQP-1187 (Fig. 2A). The compounds were also evaluated against three strains of clinical subtype C viruses and exhibited activity ranging from 0.09 nM to 8.9 nM, with compounds IQP-0406, IQP-0407, IQP-0528, IQP-0558, IQP-0410, and IQP-1187 being the most active (Fig. 2B). When tested against three clinical subtype E viruses, the most active compounds were IQP-0405, IQP-0406, IQP-0407, IQP-0528, IQP-0558, IQP-0410, and IQP-1187. EC50s for all 12 PYDs ranged from 0.07 nM to 31.4 nM against clinical subtype E viruses (Fig. 2C).
Efficacies of pyrimidinediones against HIV-1 clade viruses in PBMCs. The results presented were obtained from representative antiviral assays, with appropriate control compounds evaluated in parallel, and were selected from a minimum of three antiviral assays. We found that the standard errors among multiple antiviral assays averaged less than 10% of the respective mean EC50. In each individual assay, mean efficacy values were derived from a minimum of three replicate wells. One clade B strain (A) and three clade C (B) and clade E (C) strains were evaluated.
Efficacy of pyrimidinediones in monocyte-macrophages and monocyte-derived dendritic cells.In addition to evaluation of test compounds in fresh human PBMCs, it is also important that the activity of potential microbicides be evaluated for activity in other relevant target cell types found within the vaginal mucosa, including monocyte-macrophages and dendritic cells. Thus, we evaluated the activity of the PYDs in these two cell types infected with HIV-1BaL. The activity of the PYDs in monocyte-macrophages ranged from 0.20 nM to >10 nM, with the most active compounds being IQP-0528, IQP-0558, IQP-0410, and IQP-1187 (Fig. 3A). The antiviral activities of the test molecules extended into the subpicomolar range in dendritic cells, with EC50s ranging from 0.0003 nM to 0.72 nM. The most active molecules in the series were IQP-0405, IQP-0406, IQP-0528, IQP-0565, IQP-0410, and IQP-1187 (Fig. 3B).
Efficacies of pyrimidinediones against HIV-1BaL in monocyte-macrophages (A) and dendritic cells (B). The results presented were obtained from representative antiviral assays, with appropriate control compounds evaluated in parallel, and selected from a minimum of three antiviral assays. The standard errors among multiple antiviral assays averaged less than 10% of the respective mean EC50. In each individual assay, mean efficacy values were derived from a minimum of three replicate wells.
Toxicity of pyrimidinediones to normal vaginal flora Lactobacillus.The presence of the normal hydrogen peroxide-producing Lactobacillus flora of the vagina is critical to a healthy vaginal environment. Each PYD was evaluated for toxicity to three strains of Lactobacillus normally found in the vagina, including Lactobacillus acidophilus, Lactobacillus jensenii, and Lactobacillus crispatus. The PYDs were found to be nontoxic to these normal vaginal microorganisms at concentrations 104 to 106 times higher than those concentrations required for antiviral activity. Concentrations that inhibited the growth of the lactobacilli ranged from 40 μM to >500 μM. IQP-0405 and IQP-0558 were the least toxic compounds, with TC50s of >500 μM against all strains of Lactobacillus evaluated. IQP-0407 and IQP-1187 were found to possess the greatest toxicity; however, their defined average TC50s were 66.1 μM and 48.3 μM, respectively, well above the antiviral activities of the molecules. The remaining compounds had only marginal toxicities, with TC50s ranging from 101.8 μM to >500 μM. Thus, the PYDs are considered to have little to no toxicity to the normal vaginal flora. These toxicity data are presented in Table 2.
Toxicity to normal vaginal flora Lactobacillus spp.
Toxicities of pyrimidinediones to fresh and established human cells.In order to further assess the relative toxicity of the PYDs, each compound was exposed to fresh and established human cells for 24 h and/or 6 days (Fig. 4). Cell growth inhibitory concentrations ranged from 23.9 μM to >500 μM. Compounds IQP-0406, IQP-0549, IQP-0554, and IQP-0558 appeared to have the greatest impact on cellular replication, with most TC50s ranging from 23.9 to 114.7 μM. Compounds IQP-0405, IQP-0407, and IQP-0410 were the least toxic, with all TC50s ranging from 269.5 to >500 μM. The results from the 24-h studies are more likely to mimic typical microbicide use (short-term exposure). The 6-day results are presented to highlight potential effects of chronic microbicide use. Again, it should be noted that the concentrations impacting cell growth were 104 to 106 higher than those that inhibit virus replication, consistent with the high therapeutic index of the PYDs. Also, in nearly all cases the toxicity of the compounds has been found to be related to their overall solubility in aqueous medium as opposed to direct toxic effects on the target cells.
Toxicities of pyrimidinediones to fresh and established human cells. The results presented were obtained from representative antiviral assays, with appropriate control compounds evaluated in parallel, and selected from a minimum of three antiviral assays. We found that the standard errors among multiple toxicity assays averaged less than 10% of the respective mean TC50. In each individual assay, mean toxicity values were derived from a minimum of three replicate wells.
Efficacies in the microbicide transmission and sterilization assay.The 12 PYDs were evaluated in parallel for their potential to completely sterilize a virus-infected culture. This assay quantifies the concentration of the microbicide necessary to completely suppress virus transmission, yielding sterilization of the cell culture. For each test compound a dose-response range of five to six concentrations spanning the TI window of each compound (selected based on the defined EC50 and TC50 of the compound in the CPE inhibition assay) were evaluated. The results are presented as the number of passages which were positive for virus replication at each compound concentration, with a value of 0 representing complete sterilization of the culture and 15 representing no effect on virus replication (36). The concentrations chosen to be evaluated for each compound included 10, 50, 250, 1,250, 6,250, and 31,250 times the EC50 as defined in the CPE assay. Compounds IQP-1187, IQP-0405, and IQP-0528 performed the best, with complete sterilization at 50, 250, and 250 times the EC50, respectively. IQP-0406 and IQP-0531 were slightly less effective, with complete sterilization at concentrations of 1,250 and 6,250 times the EC50. IQP-0407, IQP-0529, IQP-0549, IQP-0554, IQP-0558, IQP-0565, and IQP-0410 were the least effective, with complete sterilization achieved at 31,250 or greater than 31,250 times the EC50 in the CPE assay. These data are graphically presented in Fig. 5.
Sterilization concentration determinations of pyrimidinediones in a microbcide transmission and sterilization assay. The pyrimidinediones were evaluated in the MTSA, and the results are presented as the number of passages which were positive for virus replication at each compound concentration. The results of one representative replicate assay are presented. Each compound was evaluated at concentrations that were 10, 50, 250, 1,250, 6,250, and 31,250 times the EC50, which was defined in the CPE inhibition assay. All of the concentrations tested were significantly below the defined TC50 for CEM-SS cells. Passages that were positive for virus production were defined by detection of virus in the cell-free supernatant in an RT assay. Cells were passaged 10 times and then in the absence of the compound for an additional 5 passages. Values of 0 represent those that were completely sterilized; passages with values of 15 represent concentrations of the test compounds that had no impact on virus replication in the cultures.
Cell-free and cell-associated virus transmission.Based on performance in the efficacy and toxicity assays, seven PYDs were chosen for further evaluation as microbicide candidates. The lead PYDs were evaluated in parallel with a standard control compound (dextran sulfate) in assays measuring the inhibition of cell-free and cell-associated virus transmission. The activity against the cell-free virus transmission of HIV-1IIIB in GHOST X4/R5 cells ranged from 0.17 to 0.38 μM for all of the tested PYDs. The activity of dextran sulfate was 0.31 μg/ml. The PYDs also effectively inhibited virus transmission from chronically infected H9 cells (HIV-1SK-1 virus) to GHOST X4/R5 cells at concentrations ranging from 0.016 to 0.21 μM, with IQP-0406 having the greatest activity at an EC50 of 0.016 μM. Dextran sulfate was active at 0.78 μg/ml. These data are presented in Table 3.
Cell-free and cell-associated virus transmission
The seven PYDs were also evaluated in assays evaluating CD4-independent and CD4-dependent virus transmission. Consistent with mechanistic data indicating that CD4 was necessary for the activity of the PYDs, all of the compounds tested were unable to inhibit the CD4-independent transmission of virus from chronically infected cells to the CD4-ME180 cervical cell line. The control compound dextran sulfate was active in the CD4-independent transmission inhibition assay at a concentration of 0.63 μg/ml. The PYDs were able to inhibit the transmission of virus from chronically infected cells to CD4+ GHOST X4/R5 cells at concentrations ranging from 0.016 to 0.21 μM. IQP-0406, IQP-0407, IQP-0528, and IQP-1187 were the most active, with EC50s ranging from 0.016 μM to 0.04 μM. IQP-0405 and IQP-0410 were 2- to 5-fold less active, with EC50s of 0.072 μM and 0.08 μM, respectively. IQP-0558 was the least active, with an EC50 of 0.21 μM. Dextran sulfate was active at 0.78 μg/ml. These data are presented in Table 4.
CD4-independent and CD4-dependent virus transmission
DISCUSSION
Twelve PYD molecules were identified from a SAR series as having highly desirable properties as potential vaginal topical microbicide candidates. These properties included high potency, activity against both HIV-1 and HIV-2, and inhibitory activity against two steps of HIV infection and replication that would be beneficial for a microbicide product (inhibition of virus entry and reverse trancsription). In an effort to define and prioritize a lead microbicide candidate for development, the PYDs were further evaluated in a series of in vitro antiviral and toxicity evaluations relevant to the development of microbicides. Based on all the results obtained herein, we can summarize that the PYDs have significant efficacy against laboratory and clinical virus strains (clades B, C, and E) when evaluated in CEM-SS cells, PBMCs, monocyte-macrophages, and dendritic cells at concentrations ranging from 0.1 to 10 nM, with therapeutic indices approaching and exceeding 1 million. In more relevant topical microbicide assays, which evaluated cell-free and cell-associated virus transmission, the compounds were active in the submicromolar concentration range. The molecules were found to be inactive against HIV transmission to cell lines devoid of CD4, consistent with their proposed entry inhibitory mechanism, which demonstrated a requirement for CD4 for the compounds to exert entry inhibition (unpublished data). In target cells that express cell surface CD4, antiviral activity was observed in the high nanomolar concentration range. In the MTSA, a transmission inhibition assay that quantifies the ability of microbicide candidates to completely suppress virus transmission (cell-to-cell and cell-free transmission events), the PYDs had various levels of activity, with the best compound requiring only 50-fold more compound than the experimentally determined EC50 to totally suppress virus infection, while the worst-performing PYD required >35,000-fold more compound to completely “sterilize” the culture. We believe the MTSA may reflect the relative ability of a microbicide candidate to prevent virus transmission events in a microbicide environment and thus may be an important screening tool to prioritize candidate compounds. Thus, our data suggest that compounds IQP-1187, IQP-0405, and IQP-0528 likely represent the best pyrimidinediones for continued development, and these pyrimidinediones were found to be more active than PRO2000, tenofovir, and SPL7013 (observed results). The toxicities of the molecules to established and fresh human cells and to the natural normal flora Lactobacillus were significantly higher (1,000- to 1,000,000-fold greater concentrations) than the defined efficacy concentrations. The experimentally determined toxicity in most cases was attributable to precipitation of the compound, as opposed to overt toxicity to the cells or bacteria.
Without an approved and/or highly successful microbicide product, there currently does not exist a “gold standard” for use as a model for the development of a superior microbicide. In choosing inhibitors to prioritize for continued development, algorithms exist which assist with the definition of the microbicidal properties of the compounds in comparison with other compounds that have progressed to human trials or are currently in development. PRO2000 has been in development since the mid-1990s, and until recently was the most advanced compound in human clinical trials. PRO2000 is an HIV inhibitor that targets entry, specifically, the CD4-gp120 interaction, possesses submicromolar antiviral activity, and was robust and potent enough in the in vitro assay systems used to advance compounds in order to move forward to human clinical trials. Unfortunately, results from a phase III clinical trial indicated that PRO2000 was not protective against the vaginal transmission of HIV (15). This was an unexpected result and focused attention on what properties a successful microbicide candidate will need to possess for success in the clinic. For the antiretroviral agent tenofovir, a nucleotide analog that inhibits viral reverse transcriptase, a phase III human clinical trial (CAPRISA 004) was recently completed, and it showed marginal levels of protection. In this trial, tenofovir vaginal gel provided moderate protection when administered before and after sexual intercourse, exhibiting a 39% lower risk of HIV acquisition overall and a 54% reduction among women with the highest level of adherence to the trial protocol (12). The results from our evaluations suggest that the most potent PYDs are significantly more active inhibitors of HIV than tenofovir and that the PYDs would likely work potently in combination with tenofovir, since they exert antiviral activity by complementary and nonoverlapping mechanisms of action. The NNRTI dapivirine is also approved for human therapeutic use and is in development for use as a microbicide (19). The PYDs have similar potency and antiviral characteristics as dapivirine but may have the advantage of being less toxic and more soluble in formulated form. Additionally, the experimental NNRTI UC781 (thiocarboxanilide) has progressed to advanced stages of microbicide development. UC781 possesses activity at nanomolar concentrations and a TI of approximately 62,000 (7), but it has a yellow color and sulfur odor, which are not ideal characteristics for a microbicide. Additionally, unlike the lead pyrimidinediones, UC781 appears to be more unstable in aqueous solutions (16). NNRTI-resistant virus strains which preexist in the population and the relative ease of selection of highly NNRTI-resistant viruses to the compound render UC781 a problematic microbicide for use unless it were to be a component of a microbicide combination product. Thus, the PYDs represent a class of compounds that meet all of the advantageous in vitro characteristics of a microbicide and might be superior to each of these two NNRTIs. The PYDs are colorless molecules that have nanomolar to subnanomolar activities and little to no cellular toxicity, yielding TIs of >100,000 and in some cases >1,000,000. The FDA-approved CCR5 antagonist maraviroc (Selzentry) exhibits activity against CCR5-tropic strains of virus in the low nanomolar concentration range. Although possessing similar in vitro activity to the PYDs, maraviroc is only active against CCR5-tropic strains of virus. Although it is known that a large percentage of the sexually transmitted viruses have CCR5 specificity, it has not been conclusively shown that CXCR4-tropic viruses are not sexually transmitted. The PYDs are highly active against CXCR4-, CCR5-, and dual-tropic (CXCR4/CCR5) viruses and also possess an entry inhibitory mechanism, yielding antiviral properties that might make them equivalent or superior to maraviroc. As noted above, the combination of maraviroc and a PYD would represent an effective combination product with complementary and nonoverlapping antiviral mechanisms of action.
The dogma surrounding potential microbicide candidates has previously involved a compound exhibiting anti-HIV activity at submicromolar concentrations and having a mechanism of action that targeted the early preintegration steps in virus replication, including attachment, entry, and reverse transcription (34). That viewpoint and focus, however, is changing with the introduction of integrase and protease inhibitors into the microbicide pipeline (11). Not only do many of the compounds in development specifically target virus replication as traditional antiretroviral agents, but also, some of the compounds now in development are already approved for use as therapeutic agents (10, 17, 25). The data presented herein summarize the initial in vitro efficacy and toxicity evaluations needed to define and prioritize the PYD congeners for continued microbicide development. Although the activity of the compounds is relatively similar, there are seven compounds from this subset of highly active pyrimidinedione analogs that we believe have properties that provide a rationale for additional evaluations as a means to define the optimal lead candidate. These highly active and potent microbicide candidates include IQP-0406, IQP-0407, IQP-0528, IQP-0558, IQP-0410, and IQP-1187. Of these seven molecules, the first four are structurally related (Fig. 1) with a cyclopropyl substituent at the N-1 of the pyrimidinedione. In preliminary experiments, these molecules have been determined to have much higher chemical and metabolic stabilities than the molecules with the 3-cyclopenten-1-yl substitution (IQP-0410 and IQP-1187), although they have similar efficacies against HIV. IQP-0410 has progressed to an Investigational New Drug submission as a potential therapeutic agent for the treatment of HIV infection. Additional studies that are more relevant to the vaginal microenvironment during virus transmission are in progress with these seven molecules, including antiviral evaluations in the presence of additives, such as vaginal fluids, seminal plasma, and mucin, as well as a pH transition assay. Additionally, the lead PYD compounds are being evaluated in combination with other potential microbicide products, such as tenofovir, and are being comparatively evaluated in human cervical explant efficacy and toxicity assays. Stability and preformulation of the molecules is also being evaluated. From these additional data we will define a lead PYD candidate microbicide for late-stage development.
ACKNOWLEDGMENTS
The studies presented were made possible through support to ImQuest BioSciences obtained from a Small Business Innovative Research phase 1 grant (R21 AI067047) and a Microbicide Innovation Program grant (R21 AI076967) from the NIAID, NIH.
FOOTNOTES
- Received 5 July 2011.
- Returned for modification 2 August 2011.
- Accepted 25 August 2011.
- Accepted manuscript posted online 6 September 2011.
- Copyright © 2011, American Society for Microbiology. All Rights Reserved.
