EFdA, a Reverse Transcriptase Inhibitor, Potently Blocks HIV-1 Ex Vivo Infection of Langerhans Cells within Epithelium

Despite increasing access to antiretroviral drugs, sexual transmission of HIV-1 remains a significant public health threat. A recent clinical trial, CAPRISA 004, of a vaginally administered microbicide using a nucleoside reverse transcriptase inhibitor (NRTI), tenofovir (TDF), has demonstrated that 1% TDF gel reduced HIV-1 acquisition by an estimated 39% overall (Abdool Karim et al., 2010), indicating a potential utility of NRTI-based microbicides. In the VOICE study, however, a once-daily dosing regimen with TDF gel failed to demonstrate protective effects in at-risk women. These studies demonstrate the need to develop additional more potent microbicide candidates to potentially increase the activity to protect women from HIV-1 transmission.

We previously reported that a series of 4′-substituted NRTIs have excellent antiviral properties (Ohrui, 2006), and through optimization of such 4′-substituted NRTIs, 4′-ethynyl-2-fluoro-2′-deoxyadenosine (EFdA) was found to exert extremely potent activity against a wide spectrum of HIV-1 strains including highly multidrug-resistant clinical HIV-1 isolates, with favorable in vitro cell toxicities (Nakata et al., 2007; Ohrui et al., 2007). EFdA inhibited HIV-1 replication in activated peripheral blood mononuclear cells with an EC₅₀ of 0.05 nM, a potency several orders of magnitude greater than any of the current clinically available NRTIs (Michailidis et al., 2009). As the prevalence of new infections with drug-resistant HIV-1 variants could increase in the coming years (Nichols et al., 2011), EFdA may be useful as a topical microbicide.

Langerhans cells (LCs) are dendritic cells located, among other sites, within genital skin and mucosal epithelium (Lederman et al., 2006). In female rhesus macaques exposed intravaginally to simian immunodeficiency virus, up to 90% of initially infected target cells were LCs (Hu et al., 2000). Ex vivo experiments with human foreskin explants show that epidermal LCs in inner foreskin are primary target cells for HIV-1 infection, providing a plausible explanation for why circumcision greatly reduces the probability of acquiring HIV-1 (Ganor et al., 2010; Zhou et al., 2011). LCs also express CD4 and CCR5, but not CXCR4, and demonstrate the distinctive characteristics of emigrating from tissue to draining lymph nodes in order to interact with T cells following contact with pathogens (Lederman et al., 2006). Indeed, epidermal LCs are readily infected ex vivo with R5-HIV-1, but not with X4-HIV-1, and initiate and promote high levels of infection upon interactions with cocultured CD4⁺ T cells (Kawamura et al., 2000; Ogawa et al., 2009, 2013), consistent with previous epidemiologic observations that the majority of HIV-1 strains isolated from newly infected patients are R5-HIV-1 strains (Zhu et al., 1993). Thus, LCs likely have an important role in disseminating HIV-1 soon after exposure to the virus.

To understand how HIV-1 traverses skin and genital mucosa, an ex vivo model was developed in which resident LCs within epithelial tissue explants obtained from suction blisters are exposed to HIV-1 and then allowed to emigrate from the tissue, thus mimicking conditions that occur following mucosal exposure to HIV (Kawamura et al., 2000; Ogawa et al., 2009, 2013). In this model, although relatively few productively infected LCs are identified, these cells induce high levels of HIV-1 infection when cocultured with resting autologous CD4⁺ T cells (Kawamura et al., 2000; Ogawa et al., 2013). As expected, when epidermal tissue explants were pretreated with various concentrations of TDF, EFdA, and CCR5 inhibitor, maraviroc (MVC), prior to R5-tropic HIV-1_Ba-L exposure, HIV-1 infection of resident LCs within epidermis as well as subsequent virus transmission from emigrated LCs to cocultured CD4⁺ T cells was decreased in a dose-dependent manner (Figure 1a and c; for detailed methods, see Supplementary Material). The blocking was confirmed by repeated experiments using skin explants from three additional randomly selected individuals (Figure 1b and d). Strikingly, although the blocking efficiency of TDF or MVC even at 5,000nM was partial, EFdA demonstrated complete blocking of R5-HIV-1 replication in LCs as well as subsequent virus transmission from emigrated LCs to CD4⁺ T cells at doses of 100–5,000nM (Figure 1a–d). Furthermore, EFdA blocked ex vivo virus infection of LCs as well as subsequent virus transmission when two strains of R5-HIV-1, HIV-1_JR-FL and HIV-1_AD8, were utilized in experiments (n=3, Supplementary Figure S1 online).

Figure 1. Preincubation of skin explants with EFdA blocks R5-HIV-1 infection in LCs and subsequent virus transmission to cocultured CD4⁺ T cells.

LCs within skin explants were preincubated with no drug (○) or the indicated concentrations of EFdA (●), TDF (▲), and MVC (■) for 30 minutes, exposed to HIV-1_Ba-L for 2 hours, and then floated on culture medium to allow migration of LCs from the explants. Emigrating cells from the epidermal sheets were collected 3 days following HIV-1 exposure. HIV-1-infected LCs were assessed by HIV-1 p24 intracellular staining in langerin⁺ CD11c⁺ LCs (a, b), or further cocultured with autologous CD4⁺ T cells and culture supernatants were assessed for p24 content by ELISA on the indicated days (c, d). Summary of percent inhibition of LC infection (b) and virus transmission to CD4⁺ T cells (d) of 12 experiments using skin explants from 12 individuals with the indicated each concentration of EFdA (●), TDF (▲), and MVC (■) are shown. Mean values obtained from different donors are shown as horizontal marks (b, d). EFdA, 4′-ethynyl-2-fluoro-2′-deoxyadenosine; LCs, Langerhans cells; MVC, maraviroc; TDF, tenofovir.

Similar to the results in epidermal LCs, preincubation of monocyte-derived LCs (mLCs) with 100–5,000nM of EFdA completely blocked HIV-1 replication in mLCs as well as subsequent virus transmission from mLCs to cocultured CD4⁺ T cells, whereas both TDF and MVC at the same doses only partially inhibited the transmission (Figure 2a and b; for detailed methods, see Supplementary Material online).

Figure 2. Preincubation of skin explants with EFdA blocks subsequent R5-HIV-1 infection in LC in a dose-dependent manner.

mLCs were preincubated with no drug (○) or the indicated concentrations of EFdA (●), TDF (▲) and MVC (■) for 30 minutes, and then immediately exposed to HIV-1Ba-L for 2 hours (a, b), or thoroughly washed to remove the extracellular drug and further cultured for 1, 2, or 3 days prior to exposure to HIV-1Ba-L for 2 hours (c–f). After 7 days of HIV-1 exposure, HIV-1-infected mLCs were assessed by HIV-1 p24 intracellular staining in langerin⁺ CD11c⁺ mLCs (a, c, e), or further cocultured with autologous CD4⁺ T cells and culture supernatants were assessed for p24 content by ELISA on the indicated days (b, d, f). Summary of percent inhibition of mLC infection (a, e) and virus transmission to CD4⁺ T cells (b, f) of three independent experiments are shown. Mean values are shown as horizontal marks (a, b, e, f). EFdA, 4′-ethynyl-2-fluoro-2′-deoxyadenosine; LCs, Langerhans cells; mLCs, monocyte-derived LCs; MVC, maraviroc; TDF, tenofovir.

Intriguingly, even in 1–3 days following the removal of EFdA (1,000nM), EFdA completely blocked HIV-1 infection of mLCs as well as subsequent virus transmission from mLCs to cocultured CD4⁺ T cells, whereas TDF and MVC rapidly lost their anti-HIV-1 activity within days (Figure 2c–f). No cellular toxicity was noted for any of these drugs at the doses used in these experiments (Supplementary Figure S2 online). When similar experiments were conducted using peripheral blood mononuclear cell as target cells, virtually identical favorable persistency of EFdA in antiviral activity compared with that of TDF was observed (data not shown).

In the present work, we demonstrated that EFdA exerted extremely more potent anti-HIV-1 activity in LCs than did TDF and MVC, and the potent anti-HIV-1 activity of EFdA persisted for at least 3 days. Of note, the efficacy of TDF gel in CAPRISA 004 has been linked to its long intracellular half-life (Abdool Karim et al., 2010; Rohan et al., 2010). Our data strongly indicate that EFdA may serve as a promising microbicide to block sexual transmission of HIV-1 because of its potent anti-HIV-1 activity, low cytotoxicity, and superior persistence of antiviral activity against HIV-1 in LCs.

Abbreviations:

EFdA

4′-ethynyl-2-fluoro-2′-deoxyadenosine

LC

Langerhans cell

mLC

monocyte-derived LC

MVC

maraviroc

NRTI

nucleoside reverse transcriptase inhibitor

TDF

tenofovir