4'-Ethynyl-2-fluoro-2'-deoxyadenosine (EFdA) inhibits HIV-1 reverse transcriptase with multiple mechanisms

Abstract

4'-Ethynyl-2-fluoro-2'-deoxyadenosine (EFdA) is a nucleoside analog that, unlike approved anti-human immunodeficiency virus type 1 (HIV-1) nucleoside reverse transcriptase inhibitors, has a 3'-OH and exhibits remarkable potency against wild-type and drug-resistant HIVs. EFdA triphosphate (EFdA-TP) is unique among nucleoside reverse transcriptase inhibitors because it inhibits HIV-1 reverse transcriptase (RT) with multiple mechanisms. (a) EFdA-TP can block RT as a translocation-defective RT inhibitor that dramatically slows DNA synthesis, acting as a de facto immediate chain terminator. Although non-translocated EFdA-MP-terminated primers can be unblocked, they can be efficiently converted back to the EFdA-MP-terminated form. (b) EFdA-TP can function as a delayed chain terminator, allowing incorporation of an additional dNTP before blocking DNA synthesis. In such cases, EFdA-MP-terminated primers are protected from excision. (c) EFdA-MP can be efficiently misincorporated by RT, leading to mismatched primers that are extremely hard to extend and are also protected from excision. The context of template sequence defines the relative contribution of each mechanism and affects the affinity of EFdA-MP for potential incorporation sites, explaining in part the lack of antagonism between EFdA and tenofovir. Changes in the type of nucleotide before EFdA-MP incorporation can alter its mechanism of inhibition from delayed chain terminator to immediate chain terminator. The versatility of EFdA in inhibiting HIV replication by multiple mechanisms may explain why resistance to EFdA is more difficult to emerge.

Keywords: AIDS; Antivirals; EFdA; Enzyme Inhibitor; Human Immunodeficiency Virus (HIV); NRTIs; Nucleoside/Nucleotide Analogue; Reverse Transcriptase; Reverse Transcription.

FIGURE 1.

EFdA-TP inhibits RT-catalyzed DNA synthesis as an ICT or DCT. A, structure of EFdA. B, DNA/DNA T_d31/P_d18 (20 nm) was incubated with 20 nm RT for 15 min in the presence of 10 μm dNTPs, 6 mm MgCl₂, and increasing concentrations of ddATP (lanes 1–4) or EFdA-TP (lanes 5–22). The sequence of the template and the mechanism that EFdA-TP uses to inhibit RT are shown next to different bands of the gel. C, the translocation state of RT after ddAMP (lanes 1–6 in both gels) or EFdA-MP (lanes 7–20 and 7–21) incorporation was determined using site-specific Fe²⁺ footprinting. T_d43/P_d30 or T_d43/P_d30+5 (100 nm) with 5′-Cy3-label on the DNA templates was incubated with 600 nm RT and various concentrations of the next incoming nucleotide (dTTP). The complexes were treated for 5 min with 1 mm ammonium iron sulfate. An excision at position −18 indicates a pre-translocation complex, whereas the one at position −17 represents a post-translocation complex. The sequences of the T/Ps that were used in the experiments are shown under the gels with underlined matched sequences in primer extension and footprinting assays. D, representative reactions that highlight the sites on the template sequences where EFdA-TP acts as an ICT or DCT. The numbers indicate the points (marked in red) of the dATP or dATP analog incorporation (P1, P6, and P10). In green color, we indicate the bases after dATP or dATP analog incorporation (P7 and P11). E, similar to A, the effect of ddATP and EFdA-TP on primer extension was observed using the DNA/DNA T_d31/P_d18+5.

FIGURE 2.

Effect of template sequence on mechanism of inhibition by EFdA-TP. Primer extension assays were performed by incubating various DNA/DNA T/Ps with 20 nm RT for 15 min in the presence of 10 μm dNTPs, 6 mm MgCl₂, and increasing concentrations of EFdA-TP. The sequence of the template is shown next to the gels. Twenty templates with A, T, C, or G at the P0, P2, P5, and P7 positions were used. The nucleotides that vary are shown in blue color and indicated with an asterisk next to the gel bands. In red color, we highlight the T positions where EFdA-TP is expected to be incorporated. Immediate and delayed chain terminations are shown in red and green boxes, respectively. Blue color and asterisks highlight the positions of the nucleotides that vary among different gels.

FIGURE 3.

SPR to determine the kinetic constants of EFdA-TP and dATP binding to HIV-1 RT. Nucleotide binding was performed by using RT covalently cross-linked to T_d37/P_d20, which has a 5′-biotinylated DNA template and a ddGMP incorporated at the primer. The RT-DNA_ddGMP complex was immobilized on a streptavidin sensor chip, and increasing concentrations of dATP or EFdA-TP were flowed to allow nucleotide association and dissociation. A two-state reaction protocol was used to analyze the SPR data, which assume a 1:1 binding of substrate (EFdA-TP or dATP) to an immobilized ligand (RT) followed by a conformational change (closing of fingers subdomain) to form a stable complex. The graph shows the association and dissociation of EFdA-TP over time. This analysis generated the following kinetic values: k_a1, the association rate constant for substrate binding; k_d1, the dissociation rate constant for substrate from the complex; k_a2, the forward rate constant for the conformational change; k_d2, the reverse rate constant for the conformational change; and K_d, the overall equilibrium dissociation constant, which for this type of two-state reaction protocol is defined by K_d = (k_d1/k_a1)·(k_d2/(k_d2 + k_a2)). The -fold change of the various constants is shown in parentheses.

FIGURE 4.

Pre-steady state kinetics of correct single nucleotide incorporation of EFdA-TP and dATP by HIV-1 RT. Preincubated HIV-1 RT (50 nm) and T_d31/P_d18 (A) or T_d26/P_d18+5 (50 nm) (B) were mixed by a quench-flow instrument with various concentrations of EFdA-TP or dATP for reaction times varying between 5 ms and 2 s. The amount of product at different reaction times was fit to the burst equation (Equation 1) to obtain burst phase rates of nucleotide incorporation. Observed rate constants were then plotted using a hyperbolic equation (Equation 3) to estimate k_pol and K_d(dNTP) (A and B). The sequences of the T/Ps are shown below the plots, and the calculated constants are shown in C. Error bars represent S.D. of at least two independent experiments.

FIGURE 5.

Incorporation of dNTPs into T/P_EFdA-MP by HIV-1 RT. A, T_d31/P_d18-EFdA-MP (5 nm) was incubated with 20 nm HIV-1 RT, 6 mm MgCl₂, 1 μm dATP, dCTP, dGTP, and increasing concentrations of the next correct dNTP (dTTP, 0–50 μm). The reactions were terminated after 15 min. B, T_d31/P_{d18+5-EFdA-MP} (5 nm) was incubated with 20 nm HIV-1 RT, 6 mm MgCl₂, 1 μm dATP, dCTP, dGTP, and increasing concentrations of the next correct dNTP (dTTP, 0–50 μm). The reactions were terminated after 5 and 30 min. C–E, various T_d31/P_{d18+5-EFdA-MP} (5 nm) where EFdA-MP is incorporated as a mismatch were incubated with 20 nm HIV-1 RT, 6 mm MgCl₂, 1 μm dATP, dCTP, dGTP, and increasing concentrations of the next correct dNTP (dTTP, 0–50 μm). The reactions were terminated after 5 and 30 min. The sequences of the T/Ps are shown below the gels, and the lanes are numbered for clarity.

FIGURE 6.

PP_i- and ATP-dependent unblocking of EFdA-MP-terminated primers. A, purified T_d31/P_d18-EFdA-MP or T_d26/P_{d18+5-EFdA-MP-dTMP} was incubated with HIV-1 RT in the presence of 6 mm MgCl₂ and 150 μm PP_i at 37 °C. Aliquots were removed, and reactions were stopped at the indicated time points (0–30 min). The excision products are shown in braces. B, purified T_d31/P_d18-EFdA-MP or T_d26/P_{d18+5-EFdA-MP-dTMP} was incubated at 37 °C with HIV-1 RT in the presence of 10 mm MgCl₂, 3.5 mm ATP, and d(d)NTPs (100 μm dATP, 0.5 μm dTTP, and 10 μm ddGTP for T_d31/P_d18-EFdA-MP or 100 μm dTTP and 10 μm ddCTP for T_d26/P_{d18+5-EFdA-MP-dTMP}). Aliquots of the reactions were stopped at the indicated time points (0–90 min). C, purified T_d31(6X)/P_{d18+5-EFdA-MP} (where X = T, A, C, or G) was incubated with HIV-1 RT in the presence of 6 mm MgCl₂ and 150 μm PP_i at 37 °C. Aliquots were removed, and reactions were stopped at the indicated time points (0–20 min). The excision products are shown in braces. The lane numbers are shown below each gel.

FIGURE 7.

EFdA-TP and TFV-DP have a different stopping pattern and do not always compete for the same incorporation sites. A, using a long template, T_d100, we investigated the stopping pattern of EFdA-TP and TFV-DP. At some positions, both analogs appear to have similar incorporation efficiency (same intensity of bands). However, at position +6, EFdA-MP is incorporated more efficiently than TFV, or TFV is not incorporated at all (+7; DCT site). The sequence of template is shown next to the gels. B, T_d31(7C)/P_d18 (20 nm) was incubated with 20 nm RT for 15 min in the presence of 1 μm dNTPs, 6 mm MgCl₂, and various concentrations of EFdA-TP and TFV-DP. The sites where the inhibitors act as ICT or DCT are shown in red and green, respectively. In red boxes, we highlight some examples of ICT where both analogs are equally incorporated. In the blue box, we highlight the ICT and DCT sites at P6 and P7, respectively, where TFV and EFdA-MP, respectively, are equally incorporated.

FIGURE 8.

Schematic representation of EFdA-MP-containing complexes. EFdA acts as an ICT by occupying the N-site of RT, whereas it occupies the P-site of RT when it acts as a DCT. In this case, the N-site is free for the next incoming dNMP to be incorporated.

FIGURE 9.

Overview of the multiple mechanisms of RT inhibition by EFdA.