Potent Activity of a Nucleoside Reverse Transcriptase Inhibitor, 4′-Ethynyl-2-Fluoro-2′-Deoxyadenosine, against Human Immunodeficiency Virus Type 1 Infection in a Model Using Human Peripheral Blood Mononuclear Cell-Transplanted NOD/SCID Janus Kinase 3 Knockout Mice

Abstract

4′-Ethynyl-2-fluoro-2′-deoxyadenosine (EFdA), a recently discovered nucleoside reverse transcriptase inhibitor, exhibits activity against a wide spectrum of wild-type and multidrug-resistant clinical human immunodeficiency virus type 1 (HIV-1) isolates (50% effective concentration, 0.0001 to 0.001 μM). In the present study, we used human peripheral blood mononuclear cell-transplanted, HIV-1-infected NOD/SCID/Janus kinase 3 knockout mice for in vivo evaluation of the anti-HIV activity of EFdA. Administration of EFdA decreased the replication and cytopathic effects of HIV-1 without identifiable adverse effects. In phosphate-buffered saline (PBS)-treated mice, the CD4⁺/CD8⁺ cell ratio in the spleen was low (median, 0.04; range, 0.02 to 0.49), while that in mice receiving EFdA was increased (median, 0.65; range, 0.57 to 1.43). EFdA treatment significantly suppressed the amount of HIV-1 RNA (median of 9.0 × 10² copies/ml [range, 8.1 × 10² to 1.1 × 10³ copies/ml] versus median of 9.9 × 10⁴ copies/ml [range, 8.1 × 10² to 1.1 × 10³ copies/ml]; P < 0.001), the p24 level in plasma (2.5 × 10³ pg/ml [range, 8.2 × 10² to 5.6 × 10³ pg/ml] versus 2.8 × 10² pg/ml [range, 8.2 × 10¹ to 6.3 × 10² pg/ml]; P < 0.001), and the percentage of p24-expressing cells in the spleen (median of 1.90% [range, 0.33% to 3.68%] versus median of 0.11% [range, 0.00% to 1.00%]; P = 0.003) in comparison with PBS-treated mice. These data suggest that EFdA is a promising candidate for a new age of HIV-1 chemotherapy and should be developed further as a potential therapy for individuals with multidrug-resistant HIV-1 variants.

Highly active antiretroviral therapy, combining two or more reverse transcriptase inhibitors and/or proteinase inhibitors, has been successful in reducing the morbidity and mortality caused by human immunodeficiency virus type 1 (HIV-1) infection (6, 27). The limitations of antiviral therapy for AIDS are exacerbated by the development of drug-resistant HIV-1 variants, the existence of viral reservoirs (4, 5), and a number of inherent adverse effects (1, 31). Nucleoside reverse transcriptase inhibitors (NRTIs), including zidovudine, didanosine, lamivudine, and stavudine, constitute the most important class of antiretroviral compounds for the treatment of HIV-1 infection (9, 17). However, the application of these compounds is clinically limited due to their cytotoxicity through inhibition of the host DNA polymerase and the rapid emergence of drug-resistant viral strains (2, 16). Therefore, developing new compounds with reduced cytotoxicity and improved antiviral potency, especially against drug-resistant viral strains, has become an urgent therapeutic objective. Recently, a new antiviral agent, 4′-ethynyl-2-fluoro-2′-deoxyadenosine (EFdA), was created (Fig. 1) (21, 23, 24). EFdA shows potent antiviral activity (50% effective concentration = 0.004 μM) and good activity against NRTI-resistant strains (10). Interestingly, EFdA-triphosphate (the active form of EFdA) showed more intracellular stability (21) and generated a more persistent antiviral effect than those of other NRTIs. In addition, EFdA is effective against human polymerases α, β, and γ, suggesting that EFdA might serve as a suitable therapy for treating individuals with HIV-1 infection and AIDS (21).

FIG. 1.

Structure of EFdA.

Severely immunodeficient mice transplanted with human peripheral blood mononuclear cells (hu-PBMC-SCID mice) represent a useful model for AIDS research, including preclinical evaluation of antiretroviral agents and vaccine development. Although the initial SCID mouse model required many PBMC for engraftment and showed inconsistent efficacy (20), recently introduced NK cell-deficient mice show a markedly improved engraftment efficiency. For this study, we established human PBMC-transplanted, HIV-1_JR-FL-infected nonobese diabetic (NOD)/SCID/Janus kinase 3 (Jak3) knockout (NOJ) mice, in which massive and systemic HIV-1 infection occurs, human CD4⁺/CD8⁺ cell ratios significantly decrease, and high levels of HIV-1 viremia are achieved. In these mice, the novel anti-HIV-1 agent EFdA, an NRTI, exerted potent anti-HIV-1 activity. Thus, our refined hu-PBMC-SCID mouse model is a powerful tool to evaluate antiretroviral activity and the adverse effects of new anti-HIV-1 agents.

MATERIALS AND METHODS

Antiviral agent.

EFdA was synthesized as published elsewhere (21, 23, 24).

Pharmacokinetic analysis of EFdA in BALB/c mice.

Pharmacokinetic analysis of EFdA in BALB/c mice was performed as previously described (22). In brief, plasma samples were collected periodically for 4 h following a single EFdA administration at a dose of 20 mg/kg of body weight dissolved in 250 μl phosphate-buffered saline (PBS). Each plasma sample (50 μl) was centrifuged at 10,000 rpm for 10 min, and the supernatant was injected into a high-performance liquid chromatography system. The eluent was monitored by UV spectroscopy at 262 nm, and the EFdA concentration in plasma was determined.

To examine the adverse effects of high-dose EFdA treatment, EFdA was administered to BALB/c mice twice a day intraperitoneally at a dose of 5 to 50 mg/kg for 14 days, and we observed their status and body weight twice a week.

Transplantation of human PBMC into NOJ mice.

NOJ mice were established and maintained in the Center for Animal Resources and Development, Kumamoto University (Kumamoto, Japan) (26). The mice were 16 to 20 weeks old at the time of transfer of human PBMC. Human PBMC-transplanted NOJ (hu-PBMC-NOJ) mice were generated by previously described methods (22). Briefly, NOJ mice were irradiated (1.8 Gy), and PBMC (1 × 10⁷) were freshly prepared from heparinized blood of a single healthy HIV-1-seronegative donor by Ficoll-Hypaque density gradient centrifugation, resuspended in PBS (0.1 ml), and infused intraperitoneally into each mouse. Peripheral blood was collected from healthy volunteers after informed consent was obtained, according to the institutional guidelines approved by the Faculty of Medical and Pharmaceutical Sciences, Kumamoto University. All animal experiments were performed according to the guidelines of the Kumamoto University Graduate School of Medical Science.

Treatment of HIV-1-infected hu-PBMC-NOJ mice with EFdA.

Five days after PBMC implantation, HIV-1_JR-FL (25,000 50% tissue culture infective doses) (12) was inoculated intraperitoneally into each mouse for which PBMC engraftment was confirmed. Twenty-four hours after HIV-1 inoculation, EFdA (10 μg in 0.1 ml PBS/mouse, twice a day) or PBS was administered for 14 consecutive days (see Fig. 3). On day 15, blood samples were collected from the mouse orbit, and then peritoneal cavity and spleen cells were harvested and resuspended in PBS.

FIG. 3.

Protocol for drug administration. ip, intraperitoneally; bid, twice daily.

Flow cytometric analysis.

Reconstructed human PBMC proliferation in mice was determined by flow cytometric analysis with allophycocyanin (APC)-Cy7-conjugated anti-mouse CD45 (BD Pharmingen, San Diego, CA), Pacific Blue (PB)-conjugated anti-human CD45 (anti-hCD45), APC-conjugated anti-hCD4 (Dako Cytomation, Glostrup, Denmark), phycoerythrin (PE)-Cy7-conjugated anti-hCD3 (e-Bioscience, San Diego, CA), and fluorescein isothiocyanate (FITC)-conjugated anti-hCD8 (Beckman Coulter, Fullerton, CA) monoclonal antibodies. The cells were treated with red cell lysing buffer (155 mM NH₄Cl, 10 mM KHCO₃, and 0.1 mM EDTA) to lyse erythrocytes before staining. Single-cell suspensions were prepared in staining medium (PBS with 3% fetal bovine serum and 0.05% sodium azide) and stained with monoclonal antibodies as described above. After 30 min of incubation on ice, the cells were washed twice with washing medium, fixed in PBS with 0.1% paraformaldehyde for 20 min in the dark, and permeabilized in PBS with 0.01% saponin. After 10 min of incubation on ice, cells were stained with PE-conjugated anti-HIV-1 p24 monoclonal antibody (Beckman Coulter, Fullerton, CA) for 30 min on ice. After being stained, the cells were analyzed on an LSR II flow cytometer (BD Bioscience, San Jose, CA). Data were analyzed with FlowJo (Tree Star, San Carlos, CA) software.

Quantification of murine plasma HIV-1 p24 and viral RNA copy numbers.

The amount of p24 antigen in murine plasma was determined using an HIV-1 p24 antigen enzyme-linked immunosorbent assay kit (ZeptoMetrix Corp., Buffalo, NY). The plasma viral load was quantified with the Amplicor HIV-1 Monitor test, version 1.5 (Roche Diagnostics, Branchburg, NJ).

Statistical analysis.

Nonparametric statistical analyses were performed using the Mann-Whitney U test and StatView software, version 4.51.1 (Abacus Concepts, Berkeley, CA). P values of <0.05 were defined as significant.

RESULTS

Pharmacokinetics of EFdA in BALB/c mice.

We examined the pharmacokinetics of EFdA in BALB/c mice by intraperitoneally administering the compound at a dose of 20 mg/kg. Plasma samples were collected periodically for up to 4 h and subjected to high-performance liquid chromatography analysis. As shown in Fig. 2, the concentration of EFdA reached the maximal concentration 10 to 30 min after intraperitoneal administration and then decreased rapidly. Although the initial blood concentration was highly variable, we found that the areas under the blood concentration-time curve were similar among the four mice (4.18, 2.44, 6.10, and 7.23 mg/liter-h; mean = 4.99 ± 1.68 mg/liter-h). Next, we administered EFdA intraperitoneally to BALB/c mice twice a day at a dose of 5 to 50 mg/kg for 14 days to examine the adverse effects induced by high-dose EFdA treatment. Mice treated with EFdA at doses of 5 to 50 mg/kg did not show any body weight loss (data not shown). No acute and subacute whole-body effects were observed in mice. Mice treated with 50 mg/kg showed ruffled fur, but the main organs of these mice appeared normal. These results suggest that even high doses of EFdA have few adverse effects in mice.

FIG. 2.

Pharmacokinetics of EFdA. Each mouse was administered EFdA intraperitoneally at a dose of 20 mg/kg, and blood samples were taken at 15, 30, 60, 120, and 240 min (n = 4).

Effects of EFdA on CD4⁺ and CD8⁺ cell counts in HIV-1-infected hu-PBMC-NOJ mice.

The in vivo antiviral potency of EFdA was investigated in the hu-PBMC-NOJ mouse model of HIV-1 infection. NOJ mice were intraperitoneally transplanted with human PBMC (1 × 10⁷ cells/mouse) 5 days before inoculation with HIV-1_JR-FL (R5 strain). EFdA (10 μg/mouse; 0.5 mg/kg) was administered intraperitoneally twice a day for 15 days (Fig. 3). PBMC were recovered from murine peripheral blood, the peritoneal cavity, and the spleen on day 16 after HIV-1 inoculation. Samples were stained with anti-mouse CD45-APC-Cy7, anti-hCD45-PB, anti-CD3-PE-Cy7, anti-CD4-APC, and anti-CD8-FITC and subjected to flow cytometric analysis for the determination of CD4⁺/CD8⁺ cell ratios. As shown in Fig. 4A, distinct CD4⁺ cells as well as CD8⁺ cells were seen in PBMC recovered from uninfected PBMC-transplanted mice. There were only a few CD4⁺ cells in PBMC recovered from HIV-1_JR-FL-infected, PBS-treated mice, resulting in a low CD4/CD8 ratio (median, 0.04; range, 0.02 to 0.49); however, the CD4⁺ cell frequency was increased by EFdA treatment (median, 0.65; range, 0.57 to 1.43), up to the level in uninfected mice (median, 0.79; range, 0.73 to 1.43), in the PBMC as well as the spleen and peritoneal cavity (Fig. 4B). The numbers of CD4⁺ cells in PBS-treated mouse peripheral blood, spleens, and peritoneal cavities were significantly lower than those in EFdA-treated (P < 0.001) or uninfected (P < 0.005) mice (Fig. 5), while there were no significant differences in CD8⁺ cell numbers between groups, indicating that EFdA is not toxic to lymphocytes. Thus, EFdA protects CD4⁺ T cells against HIV-1 infection-induced cell death.

FIG. 4.

Effects of EFdA on the CD4⁺/CD8⁺ cell ratio in HIV-1-infected hu-PBMC-NOJ mice. (A) PBMC recovered on day 16 after R5 HIV-1_JRFL inoculation were subjected to flow cytometry. Representative flow cytometric analysis profiles are shown. (B) PBMC, spleen cells, and peritoneal cavity cells recovered on day 16 after HIV-1 inoculation were subjected to flow cytometry. CD4/CD8 ratios are shown for each mouse (n = 12).

FIG. 5.

Effects of EFdA on numbers of CD4⁺ and CD8⁺ cells. PBMC (n = 9), spleen cells (n = 12), and peritoneal cavity cells (n = 12) recovered on day 16 after HIV-1 inoculation were counted and subjected to flow cytometry. Short bars indicate the medians. N.S., not significant.

EFdA suppresses HIV-1 viremia in hu-PBMC-NOJ mice.

The amount of HIV-1 p24 in plasma was also found to be very high in PBS-treated mice (median, 1.9 × 10³ pg/ml; range, 8.3 × 10² to 5.6 × 10³ pg/ml). EFdA was found to significantly suppress the amount of plasma p24 on day 15 (median, 2.1 × 10² pg/ml; range, 8.3 × 10¹ to 6.3 × 10² pg/ml; P < 0.001) (Fig. 6A). We also determined the HIV-1 RNA copy number in infected, PBS-treated mice and found that the median copy number was 9.9 × 10⁴ (range, 1.3 × 10⁴ to 5.4 × 10⁵) copies/ml on day 15 after HIV-1 inoculation; however, EFdA significantly suppressed viremia (median, 9.0 × 10² copies/ml; range, 8.1 × 10² to 1.1 × 10³ pg/ml; P < 0.001) on day 15.

FIG. 6.

Effects of EFdA on the amounts of plasma p24 and HIV-1 RNA. Blood samples were collected from the mouse orbit on day 16 after HIV-1 inoculation. (A) Amounts of plasma p24 antigen (n = 12). (B) HIV-1 RNA copy numbers (n = 11). Short bars indicate the medians.

Effects of EFdA on intracellular p24 levels in HIV-1-infected hu-PBMC-NOJ mice.

The number of p24-expressing (p24⁺) cells in human CD3⁺ cells in the spleen, peripheral blood, and peritoneal cavity was analyzed by flow cytometric analysis. The frequency of p24⁺ cells in the spleen was found to be high for PBS-treated mice (median, 1.90%; range, 0.33% to 3.68%). EFdA was found to significantly suppress the level of p24⁺ cells (median, 0.11%; range, 0.00% to 1.00%; P = 0.003) (Fig. 7A and B). The frequency of p24⁺ cells in peripheral blood and the peritoneal cavity was also found to be high for PBS-treated mice and significantly suppressed after EFdA treatment. No apparent EFdA-associated adverse effects were seen throughout the study period.

FIG. 7.

Effects of EFdA on HIV-1-infected cells. (A) PBMC recovered on day 16 after HIV-1_JRFL inoculation were stained with anti-p24-PE, anti-mouse CD45-APC-Cy7, anti-hCD45-PB, anti-hCD4-PB, anti-hCD4-APC, anti-hCD3-PE-Cy7, and anti-hCD8-FITC and subjected to flow cytometry. Representative flow cytometric analysis profiles of the mouse CD45⁻ hCD45⁺ hCD3⁺ hCD8⁻ gated fraction are shown. (B) PBMC, spleen cells, and peritoneal cavity cells recovered on day 16 after HIV-1 inoculation were subjected to flow cytometry. The percentage of p24⁺ cells among CD4 T cells (CD45⁻ hCD45⁺ hCD3⁺ hCD8⁻ gated) is shown (n = 8).

DISCUSSION

In the present study, we demonstrated the potent activity of EFdA as an agent against HIV in hu-PMBC-NOJ mice. As demonstrated, this particular model is well suited to the study of therapeutic interventions in the HIV arena, providing information on the treatment effects on CD4⁺ T-cell counts as well as on viral markers, such as plasma p24, HIV-1 RNA, and intracellular p24, which are important parameters in determining the overall effectiveness of a treatment in HIV-1-positive patients.

SCID mice implanted with human PBMC, which are known as hu-PBMC-SCID mice, have been used as an animal model for investigating the pathogenesis of HIV infection (15, 18, 19); however, PBMC reconstitution of the SCID mouse varies considerably among transplantation methods, laboratories, experiments, graft sources, and even individual mice (20). PBMC transplantation into NOD/SCID animals resulted in a significant increase in the positive transplantation rate compared to that obtained by identical treatment of SCID animals (7, 13). More recently, the introduction of mice with a complete loss of NK cells, such as NOD/SCID/common γ^−/− mice (8, 32), BALB/c Rag-2^−/− γ^−/− mice (30), and NOJ mice (26), markedly improved the engraftment of PBMC as well as human hematopoietic stem cells and has enabled more stable and precise analysis (14, 22, 29). HIV-1 was challenged 2 weeks after peripheral blood lymphocyte (PBL) transplantation in the previous work (22, 28), since an HIV-1 R5 virus is not adequately infective soon after transplantation (3). We optimized the time of viral infection and found that HIV_JR-FL could successfully infect cells and replicate during virus challenge as early as 5 days after PBL transplantation. Since the HIV-1-infected hu-PBL-NOJ mouse model needed a relatively smaller amount of human PBL and a shorter duration of HIV-1 infection and replication than those in previous studies (7, 13, 22, 28), it could be a more useful instrument for analyzing the pathogenesis of HIV-1 infection and testing the efficacy of antiviral agents.

A number of 4′-ethynyl (4′-E)-2′-deoxynucleosides and their analogs have been synthesized, and a series of potent anti-HIV-1 compounds have been identified to block the replication of a wide spectrum of laboratory and clinical HIV-1 strains in vitro (11, 23, 25). By optimization of such 4′-E nucleoside analogs, EFdA was found to have potent anti-HIV activity, including activity against highly multidrug-resistant variants, with favorable in vitro cell toxicities (21, 24). EFdA shows unique anti-HIV-1 function and characteristics. EFdA-triphosphate shows greater intracellular stability and generates a more persistent antiviral effect than those of other NRTIs, such as zidovudine or tenofovir. EFdA acts as a chain terminator upon incorporation at the primer end; however, it showed no inhibition of cellular polymerases (21). In addition, unlike other adenosine-based NRTIs, EFdA shows adenosine deaminase resistance (10), and moreover, it has a very high selectivity index, and high-dose EFdA is not toxic to BALB/c mice. In the present study, hu-PBMC-NOJ AIDS model mice treated with EFdA maintained high levels of human CD4⁺ lymphocytes (Fig. 4 and 5), suppressed plasma levels of p24 and HIV-1 RNA (Fig. 6), and reduced the number of infected (p24⁺) cells without apparent adverse effects. Although we cannot directly compare EFdA with previously studied anti-HIV-1 agents, our study suggests that EFdA is expected to be effective for clinical use and is a favorable anti-HIV-1 therapeutic agent. It is notable that determination of the precise pharmacokinetics and pharmacodynamics is awaited in clinical trials when EFdA is assessed in humans.

In summary, the data presented here provide strong evidence that the hu-PBMC-NOJ mouse is a valuable model for preclinical testing of new antiretroviral agents. Using this HIV-1 infection mouse model system, we have demonstrated that a new antiretroviral agent, EFdA, has potent anti-HIV-1 activity in vivo, without apparent adverse effects. Since EFdA has unique functional properties, low cytotoxicity, and superior persistence of antiviral activity, it is a promising candidate for a new age of HIV-1 chemotherapy.