Molecular Determinants of Multi-nucleoside Analogue Resistance in HIV-1 Reverse Transcriptases Containing a Dipeptide Insertion in the Fingers Subdomain EFFECT OF MUTATIONS D67N AND T215Y ON REMOVAL OF THYMIDINE NUCLEOTIDE ANALOGUES FROM BLOCKED DNA PRIMERS* □ S

Human immunodeficiency virus type 1 isolates having dipeptide insertions in the fingers subdomain of the reverse transcriptase (RT) show high level resistance to 3 (cid:2) -azido-3 (cid:2) -deoxythymidine (AZT) and other nucleoside analogues. Insertions are usually associated with thymidine analogue resistance mutations, such as T215Y. The resistance phenotype correlates with increased ATP-de-pendent phosphorolytic activity, which facilitates removal of thymidine analogues from inhibitor-termi-nated primers. In this report, we show that substituting Thr, Ser, or Asn for Tyr-215 in a multidrug-resistant RT, bearing a Ser-Ser insertion between codons 69 and 70, leads to AZT and stavudine resensitization through the loss of the ATP-mediated removal activity. The mutation D67N, which is rarely found in insertion-containing strains, had no variants a stable ternary complex assessed as described (32, RTs (20 n M were incubated with the labeled D38/25PGA d4T template-primer (0.3 n M ) to form the binary complex under the conditions above and then further incubated at 37 °C in the presence of an excess of dNTP (200 (cid:2) M dATP, in our assay conditions) to obtain a stable ternary complex. The putative ternary complexes formed were challenged with unlabeled D38/25PGA d4T tem- plate-primer at a final concentration of 500 n M . The ternary complex species resistant to the DNA trap were resolved on a 6% native poly- acrylamide gel, subjected to phosphorimaging, and quantified using the Tina software. The apparent dNTP binding affinity (apparent K d [dNTP] ) of the vari- ant RTs was determined as described above, except that the binary complexes were incubated in the presence of increasing concentrations of the complementary nucleotide (dATP, 0–4000 (cid:2) M ) prior to being challenged with the DNA trap. In some experiments, ternary complexes were obtained in the presence of 200 (cid:2) M foscarnet to monitor the effect of the PPi analogue in the apparent K d [dNTP] . The RT-DNA (cid:3) DNA-dNTP complexes resistant to the trap were resolved and quantified as indicated above. To determine the apparent K d [dNTP] , the percentage of total RT-DNA (cid:3) DNA complex converted into stable RT-DNA (cid:3) DNA-dNTP ternary complex as a function of the concentration of dNTP was fitted to the single-site ligand-binding equation.

Human immunodeficiency virus type 1 (HIV-1) 1 reverse transcriptase (RT) replicates the viral genomic RNA to synthesize a double-stranded DNA that integrates into the host genome. The viral enzyme is multifunctional, possessing RNAand DNA-dependent DNA polymerase, RNase H, strand transfer, and strand displacement activities (1). HIV-1 RT is a heterodimeric enzyme composed of two polypeptide chains of 66 and 51 kDa, with subdomains termed fingers, thumb, palm, and connection in both subunits and an RNase H domain in the larger subunit only.
HIV-1 RT is an important target for chemotherapeutic intervention in the control of AIDS. Antiretroviral drugs targeting the viral polymerase include nucleoside analogue inhibitors (NRTIs), acyclic nucleoside phosphonates, and non-nucleoside RT inhibitors (NNRTIs) (reviewed in Ref. 2). Inside the cell, nucleoside derivatives are converted to their active triphosphate forms by host cell kinases and are then incorporated into the HIV-1 genome by the viral RT. Because nucleoside analogues lack a 3Ј-OH group, their incorporation blocks elongation of the growing DNA chain. On the other hand, NNRTIs bind to an allosteric site located 10 -15 Å away from the polymerase active site, distorting the geometry and/or the mobility of the polymerase catalytic site, interfering with the proper positioning of the template-primer in the nucleic acid-binding cleft of the RT or restricting the mobility of the thumb, which results in the impairment of translocation during polymerization (for reviews, see Refs. 3 and 4). Although NNRTIs resistance mutations exert their action by diminishing the ability of the enzyme to bind the inhibitor (i.e. Y181C for nevirapine, K103N for nevirapine, delavirdine, and efavirenz, etc.), NRTI resistance mutations act either by (i) increasing discrimination against the triphosphate forms of the drugs (reviewed in Refs. 5 and 6) or (ii) increasing the ability of RT to unblock inhibitorterminated DNA chains in the presence of physiological concentrations of pyrophosphate (PPi) or ATP (7)(8)(9).
Prolonged use of antiretroviral drugs in the clinical treatment of HIV-infected patients has resulted in the emergence of viral isolates with reduced susceptibilities to multiple RT inhibitors. Among multidrug-resistant viruses, resistance to multiple NRTIs can be achieved by three different major pathways: (i) accumulation of thymidine analogue resistance mutations (TAMs), (i.e. M41L, D67N, K70R, L210W, T215Y/F, and K219Q/E), which have a clinical impact on AZT, d4T, ddI, abacavir, and tenofovir resistance (16 -21); (ii) selection of mutation Q151M, followed by other substitutions (A62V, V75I, F77L, and F116Y) that render viruses with significant resistance to AZT, d4T, ddI, ddC, and abacavir (22)(23)(24); and (iii) emergence of insertions of two residues (often Ser-Ser, Ser-Gly, and Ser-Ala) together with mutation T69S, in virus containing additional amino acid substitutions associated with NRTI resistance, such as M41L, L210W, and/or T215Y. In phenotypic assays, these viruses display high level resistance to AZT and moderate levels of resistance to other NRTIs, such as d4T, ddC, and ddI (25)(26)(27)(28).
Although the biochemical properties of the wild-type HIV-1 BH10 RT were not significantly altered upon introduction of a dipeptide (Ser-Ser) insertion, we have previously shown that eliminating the insertion in the sequence context of an HIV-1 RT bearing the drug-related mutations M41L, A62V, K70R, V118I, M184I, L210W, T215Y, and G333E produced a significant reduction of the ATP-dependent phosphorolytic activity that allows excision of AZT or d4T from inhibitor-terminated primers (13,27). These data correlated with a significant reduction in the viral susceptibility to AZT (27). However, eliminating the insertion was not sufficient to abrogate AZT resistance, because of the presence of TAMs in the insertioncontaining viral isolate. In this work, we have analyzed the impact of TAMs D67N and T215Y on the ATP-and PPi-mediated removal of AZT-and d4T-monophosphate by SS RT. These two mutations were chosen based on available genotypic data showing that T215Y/F is the most frequently found mutation in viral isolates containing insertions in the fingers subdomain, whereas D67N appears very rarely in those viral strains, despite being a commonly selected mutation during treatment with thymidine analogues.

EXPERIMENTAL PROCEDURES
RTs-Expression and purification of wild-type BH10, SS, and mutant SS_2S0S were done as previously described (27). Other mutant RTs were generated by using the QuikChange TM site-directed mutagenesis kit (Stratagene), following the manufacturer's instructions. Individual mutations Y215T, Y215N, Y215S, and D67N were introduced in plasmid pRT6 carrying the nucleotide sequence encoding the 66-kDa subunit of SS RT (27) to generate mutant enzymes designated as SS_Y215T, SS_Y215N, SS_Y215S, and SS_D67N, respectively (see Fig.  1). On the other hand, the mutation T215Y was introduced in a pRT6 construct encoding a mutated p66 subunit of BH10 RT, which contained the T69SSS insertion (27). This double mutant T69SSS/T215Y was designated as BH10_SSSY. The mutagenic primers used were: 5Ј-GG-AGGTGGGGATTTACCACACCAGACAAA-3Ј and 5Ј-TTTGTCTGGTG-TGGTAAATCCCCACCT-3Ј for SS_Y215T; 5Ј-GGAGGTGGGGATTTT-CCACACCAGACAAA-3Ј and 5Ј-TTTGTCTGGTGTGGAAAATCCCCA-CCT-3Ј for SS_Y215S; 5Ј-GGAGGTGGGGATTTAACACACCAGACAA-A-3Ј and 5Ј-TTTGTCTGGTGTGTTAAATCCCCACCT-3Ј for SS_Y215N; 5Ј-GTCATAAAGAAAAAAAATAGTTCTAGTTCTAGATGGAG-3Ј and 5Ј-CTAGAACTAGAACTATTTTTTTTCTTTATG-3Ј for SS_D67N; and 5Ј-GAGGTGGGGACTTTACACACCAGACAAAAAACATC-3Ј and 5Ј-G-ATGTTTTTTGTCTGGTGTGTAAAGTCCCCACCTC-3Ј for BH10_ SSSY. The introduced mutations were confirmed by DNA sequencing, and the appropriate MscI-KpnI inserts were then cloned in plasmids derived from pT51H (27,29), for expression of the p51 subunits of SS or BH10 RTs. All of the RTs were purified as p66/p51 heterodimers, and the mutations were introduced in both subunits of the enzyme. The 51-kDa polypeptide was obtained with an extension of 14 amino acids at its N-terminal end, which includes 6 consecutive histidine residues to facilitate its purification by metal chelate affinity chromatography, as previously described (29).
Nucleotides and Template-Primers-The triphosphate derivatives of NRTIs were obtained from Moravek Biochemicals (AZT-triphosphate) and Sierra Bioresearch (d4T-triphosphate). Stock solutions of dNTPs and rNTPs (100 mM) were from Amersham Biosciences. Before use, nucleoside-triphosphates were treated with inorganic pyrophosphatase to remove traces of PPi, as described (13). As in previous studies, we used the following heteropolymeric oligonucleotides (Invitrogen) as DNA⅐DNA template-primer substrates: 5Ј-GGGTCCTTTCTTACCTG-CAAGAATGTATAGCCCTACCA-3Ј (termed D38) served as DNA template, and 5Ј-TGGTAGGGCTATACATTCTTGCAGG-3Ј (termed 25PGA) was used as the primer strand. Polyacrylamide gel-purified oligonucleotides were labeled at their 5Ј-termini with [␥-32 P]ATP and T4 polynucleotide kinase (Promega). The phosphorylated primers were then annealed to gel-purified templates in a solution containing 150 mM NaCl and 150 mM magnesium acetate, as described (30). The template-primer molar ratio was adjusted to 1:1.
Chain Terminator Removal Assays-RT-catalyzed DNA rescue reactions were performed with D38/25PGA DNA duplexes as previously described (13). Briefly, the phosphorylated template-primer (30 nM) was preincubated at 37°C for 10 min in the presence of the corresponding RT at 15-20 nM active enzyme concentration, in 50 mM Hepes buffer (pH 7.0), containing 15 mM NaCl, 15 mM magnesium acetate, 130 mM potassium acetate, 1 mM dithiothreitol, and 5% (w/v) polyethylene glycol 6000. The reactions were initiated by adding an equal amount of preincubation buffer containing AZT-triphosphate or d4T-triphosphate at a final concentration of 25 M. After incubating the samples at 37°C for 30 min, the rescue reactions were initiated by adding a mixture of all dNTPs at a final concentration of 100 M, in the presence of sodium PPi (200 M) or ATP (3.2 mM) depending on the assay. Because the next complementary dNTP (dATP, in our assay conditions) has an inhibitory effect on the rescue reaction, time courses of the unblocking and extension reactions were carried out in the presence of 1 M dATP. Sodium foscarnet (Sigma) inhibition was monitored under these conditions by adding the inhibitor (at 0 to 800 M), together with the mixture of dNTPs. In experiments designed to assess the inhibitory effect of dATP, the extension reactions were incubated for 0 -15 min in the presence of different concentrations of dATP. In all cases, the incubation times were within the linear range of the corresponding time course. The reactions were stopped by adding an equal amount of 10 mM EDTA in 90% formamide containing 3 mg/ml xylene cyanol FF and 3 mg/ml bromphenol blue. The products were resolved on a denaturing 20% polyacrylamide-urea gel, and primer rescue was quantified by phosphorimaging with a Bas 1500 scanner (Fuji), using the program Tina version 2.09 (Raytest Isotopenmessgerate Gmbh, Staubenhardt, Germany).
Template-Primer Binding Affinity-Dissociation constants (K d ) of RT-DNA⅐DNA complexes were determined in accordance with the procedure described by Astatke et al. (31). For these experiments, primer 25PGA was blocked at its 3Ј end with d4T-triphosphate, using terminal deoxynucleotidyl transferase as previously described (13). Then the gel-purified inhibitor-terminated primer was labeled at its 5Ј terminus with [␥-32 P]ATP and T4 polynucleotide kinase and annealed to template D38 as described above. The 38-mer DNA template primed with 5Ј 32 P-labeled d4T-terminated 25-mer (D38/25PGA d4T ) (0.3 nM) was incubated with varying concentrations of the corresponding RTs in a total volume of 10 l of 50 mM Hepes buffer (pH 7.0), containing 15 mM NaCl, 15 mM magnesium acetate, 130 mM potassium acetate, 1 mM dithiothreitol, and 5% (w/v) polyethylene glycol 6000. Following incubation for 10 min at 37°C, the samples were analyzed on a 6% native polyacrylamide gel in Tris borate buffer (85 mM Tris and 85 mM boric acid, pH 8.0). The amount of template-primer in the RT-bound form was determined by phosphorimaging. The percentage of complexed DNA was plotted against the enzyme concentration, and the K d [DNA] value was determined as the RT concentration at which 50% of the DNA is bound.
Stable Ternary Complex Formation Assays and Determination of dNTP Binding Affinity (Apparent K d [dNTP] ) -The ability of HIV-1 RT variants to form a stable ternary complex was assessed as described (32,33). For this purpose, RTs (20 nM) were incubated with the labeled D38/25PGA d4T template-primer (0.3 nM) to form the binary complex under the conditions described above and then further incubated at 37°C in the presence of an excess of dNTP (200 M dATP, in our assay conditions) to obtain a stable ternary complex. The putative ternary complexes formed were challenged with unlabeled D38/25PGA d4T template-primer at a final concentration of 500 nM. The ternary complex species resistant to the DNA trap were resolved on a 6% native polyacrylamide gel, subjected to phosphorimaging, and quantified using the Tina software.
The apparent dNTP binding affinity (apparent K d [dNTP] ) of the variant RTs was determined as described above, except that the binary complexes were incubated in the presence of increasing concentrations of the complementary nucleotide (dATP, 0 -4000 M) prior to being challenged with the DNA trap. In some experiments, ternary complexes were obtained in the presence of 200 M foscarnet to monitor the effect of the PPi analogue in the apparent K d [dNTP] . The RT-DNA⅐DNA-dNTP complexes resistant to the trap were resolved and quantified as indicated above. To determine the apparent K d [dNTP] , the percentage of total RT-DNA⅐DNA complex converted into stable RT-DNA⅐DNA-dNTP ternary complex as a function of the concentration of dNTP was fitted to the single-site ligand-binding equation.
Recombinant Virus and Drug Susceptibility Tests-The assays were performed as previously described (27,34). Briefly, full-length RTcoding sequence DNA was amplified from plasmids carrying the different RTs using primers IN5 (5Ј-AATTTTCCCATTAGTCCTATTGAAA-CTGTACCA-3Ј) and IN3 (5Ј-TCTATTCCATCYAAAAATAGTACTTTC-CTGATTCC-3Ј). The PCR products were then cotransfected with an RT-deleted HXB2-D clone in SupT1 cells (35). When the HIV-1 p24 antigen concentration in the cultures surpassed 20 ng/ml, the supernatants were harvested. Progeny virus was propagated and titrated in MT-4 cells. The nucleotide sequence of the RT-coding region of the progeny virus was checked for possible reversions or additional mutations. The SupT1 and MT-4 cells and the deleted HXB2-D clone were obtained from the AIDS Reagent Program (Medical Research Council). HIV-1 drug susceptibility profiles were obtained after infecting 35000 MT-4 cells with 100 50% tissue culture infective doses of virus, at a multiplicity of infection of 0.003, by exposing the HIV-1-infected cultures to various concentrations of each drug (5-fold dilutions). After MT-4 cells were allowed to proliferate for 5 days, the number of viable cells was determined by a tetrazolium-based colorimetric method as described (27,36).

Mutations Associated with Insertions in the Fingers Subdomain of HIV-1 RT-
Although with a relatively low prevalence (0.5-2.7%), several studies have identified heavily treated patients carrying HIV-1 isolates with RTs containing an insertion in the fingers subdomain plus additional resistance mutations, including TAMs (37)(38)(39)(40)(41). We have done an extensive survey of the literature to generate a data base containing 200 sequences of RTs having insertions of one or two amino acids between codons 69 and 70 (supplemental table). The most frequently found insertions were Ser-Gly, Ser-Ser, and Ser-Ala, which represented approximately twothirds of the total number of sequences reported. Interestingly, TAMs frequently associated with insertions include M41L, K70R, L210W, and T215Y/F. An aromatic residue at position 215 was observed in 90% of all isolates and appeared in all strains having a Ser-Ser insertion. On the other hand, the classical AZT resistance mutation D67N was never found in consensus sequences of insertion-containing RTs from clinical samples, although it has been occasionally observed in clones derived from viral populations (42).
The SS RT is a good model to study the effect of those mutations, because this polymerase contains both the insertion and the T215Y mutation while having wild-type Asp at position 67 ( Fig. 1). We constructed mutants having the SS RT background where Tyr-215 was replaced by Thr (SS_Y215T), Ser (SS_Y215S), or Asn (SS_Y215N) or where Asp-67 was replaced by Asn (SS_D67N). The amino acid substitution Y215T involves two nucleotide changes at codon 215, whereas Y215S and Y215N involve just one nucleotide change and are potential intermediates during a possible reversion of Tyr to wildtype Thr. In addition, both the insertion and T215Y were introduced in the sequence context of the wild-type BH10 RT (mutant BH10_SSSY) to determine whether both changes were sufficient to confer resistance to thymidine analogues. The properties of all mutants on primer unblocking, dNTP inhibition, and DNA chain elongation, as well as their ability to form stable ternary complexes were compared with those of reference RTs (i.e. SS and BH10 RTs), as well as of mutant SS_2S0S, which lacks the dipeptide insertion while maintaining the SS RT sequence background.
Removal of AZT-and d4T-Monophosphate from Blocked DNA Primers-The ability of RTs to rescue AZT-and d4Tterminated primers was assessed by using the template-primer shown in Fig. 2A. These experiments were carried out in two steps. First, the HIV-1 RT was incubated with the templateprimer in the presence of the triphosphate form of the inhibitor. Then unblocking and extension reactions leading to the accumulation of a 38-nucleotide product were carried out by adding dNTPs in the presence of physiological concentrations of ATP or PPi (3.2 mM and 200 M, respectively). These reactions can be inhibited by the next complementary dNTP (dATP in our assay conditions), and therefore time courses were carried out in the presence of low dATP concentrations (i.e. 1 M). Noncomplementary dNTPs were included at 100 M to facilitate full extension of primers after removal of chain-terminating nucleotides. At these concentrations, inhibition of the rescue reaction because of potential mispairs between the template T and dGTP, dTTP, or dCTP was found to be negligible.
In the presence of ATP, AZT removal reactions catalyzed by SS and mutant SS_D67N RTs showed the highest efficiency, whereas SS RT mutants lacking an aromatic side chain at position 215 (i.e. mutants SS_Y215T, SS_Y215N, and SS_Y215S) were devoid of ATP-dependent phosphorolytic activity (Fig. 2B). Substituting Tyr-215 had a larger effect than Residues that are identical to those found in the wild-type BH10 RT are indicated by dots. Underlined amino acids are those related to drug resistance.
deleting the two serines forming the insertion. However, the presence of the insertion together with an aromatic residue at position 215, as in mutant BH10_SSSY, was not sufficient to confer the high ATP-dependent phosphorolytic activity displayed by the SS_RT. The large differences observed between mutant enzymes in ATP-dependent phosphorolysis using AZTterminated primers were not detected in phosphorolytic reactions when PPi was used as the donor (Fig. 2C). All of the enzymes showed AZT-monophosphate removal rates that were within a 2-fold range of each other.
The time courses for d4T removal in the presence of ATP were similar to those obtained with AZT-terminated primers for all tested RTs (Fig. 3A). D67N had no effect on the removal reaction, whereas eliminating the aromatic side chain of Tyr-215 had a deleterious effect of the unblocking activity. Interestingly, some differences were also detected between the enzymes when analyzing PPi-mediated removal of d4T monophosphate from chain-terminated primers. Excision rates were approximately three times higher for SS RT than for mutants SS_Y215T, SS_Y215S, and SS_Y215N (Fig. 3B).
Primer Unblocking Inhibition by the Next Complementary dNTP and Formation of Stable Ternary Complexes-As previously shown, removal reactions can be inhibited by the next complementary nucleotide (13,27). AZT removal reactions catalyzed by BH10_SSSY, SS, SS_2S0S, and SS_D67N were highly resistant to dNTP inhibition (IC 50 Ͼ 0.4 mM), but unblocking of d4T-terminated primers was Ͼ20-fold more sensitive to inhibition by the next complementary dNTP (Table I). Under our assay conditions, the pyrophosphate donor (i.e. ATP or PPi) or the specific mutation introduced in the recombinant RT had no significant influence on the IC 50 values for the next complementary dNTP. However, RTs having a dipeptide insertion between codons 69 and 70 and the T215Y mutation were shown to be Ͼ5-fold less sensitive to inhibition by the next complementary dNTP in ATP-dependent removal reactions, using dideoxyadenosine-terminated primers (43). In addition, it has been reported that for AZT-and ddT-terminated primers, the concentration of the next complementary dNTP required to shift the RT from the pretranslocational to the post-translocational stage was remarkably larger for mutant M41L/T69SSS/ T215Y than for the wild-type HIV-1 RT (44). In an attempt to clarify why our assay failed to detect differences in RT sensitivity to dNTP inhibition, we analyzed how mutations could impact dNTP binding by looking at ternary complex formation.
Ternary complexes formed by the RT, a d4T-terminated template-primer, and the next complementary dNTP, using wildtype and mutant RTs, were analyzed by electrophoretic mobility retardation assays. For this purpose, each mutant enzyme was first allowed to form binary complexes with 32 P-labeled d4T-terminated template-primers and then supplemented with the next complementary dNTP to form the ternary complex. All of the tested RTs showed normal DNA polymerase activity (data not shown), and the equilibrium dissociation constants (K d ) of the corresponding binary complexes ranged between 1.6 and 7.4 nM (Fig. 4).
Binding of dNTP to the enzyme is an ordered mechanism that occurs only after DNA binding, and the amount of labeled template-primer that remains bound to the RT in the presence of high dNTP concentration and an excess of DNA trap represents the level of stable ternary complex that is formed. This is All of the dNTPs in these assays were supplied at 100 M, except for dATP whose concentration was 1 M. Active enzyme concentration in the assays was 10 nM. The represented values were obtained from two to three independent experiments. The standard deviations were below 20% in these assays. illustrated in Fig. 5A, which shows the higher stability of the ternary complex versus the binary complex using wild-type BH10 RT as a model. In the presence of over 1500-fold molar excess of DNA trap, the RT-DNA⅐DNA complex was competed out, whereas a significant amount of ternary complex was resistant to competition with DNA trap. The results shown in Fig. 5B, reveal that the RT-DNA⅐DNA complexes formed by each of the mutant enzymes (lane 1) were fully competed out by the DNA trap (lane 2), whereas a portion of d4T-terminated template-primer bound in RT-DNA⅐DNA-dNTP ternary complexes remains stable and resistant to the trap (lane 4). This portion was roughly similar for all tested RTs.
Ternary complex formation was tested on the labeled D38/ 25PGA d4T template-primer as a function of dATP, the next complementary nucleotide (Fig. 6). The apparent K d [dNTP] values for SS, SS_2S0S, and SS_Y215N were Ͼ5-fold higher than for the wild-type BH10 RT and somewhat higher than for mutants BH10_SSSY, SS_Y215T, and SS_Y215S (Table II). For most enzymes, the apparent K d [dNTP] values were remarkably higher than the IC 50 values obtained in rescue assays carried out in the presence of ATP or PPi (Table I). This dis-crepancy could be due to the presence of the PPi donor in the reaction. Analyzing ternary complex formation in the presence of ATP or PPi is difficult, because these metabolites are substrates of the phosphorolytic reaction. To circumvent this limitation, we used sodium foscarnet, a PPi analogue. Foscarnet is a relatively weak inhibitor of the replication of recombinant viruses containing wild-type BH10 and SS RTs (27). The foscarnet IC 50 values for the recombinant BH10 and SS RTs in primer rescue assays using d4T-terminated primers were 99.8 Ϯ 7.2 and 46.7 Ϯ 5.2 M, respectively. Foscarnet is expected to bind at the dNTP-binding site, but unlike in the presence of dNTPs, ternary complexes of RT-DNA⅐DNA-foscarnet are rather unstable even at concentrations of the inhibitor above 200 M (data not shown) (45). Interestingly, when RT-DNA⅐DNA-dNTP complexes were formed in the presence of 200 M foscarnet, all of the studied RTs displayed similar apparent K d [dNTP] values ( Fig. 6 and Table II).
Viability of Recombinant HIV-1 Variants and Resistance to Nucleoside Analogue Inhibitors-All enzymes described were found to be viable after transfecting cells with a recombinant infectious HIV-1 clone harboring the corresponding RT. All All of the dNTPs in the assays were supplied at 100 M, except for dATP whose concentration was 1 M. Active enzyme concentration in these assays was 10 nM. The represented values were obtained from two to three independent experiments. The standard deviations were below 20% in these assays. The assays were carried out in the presence of 3.2 mM ATP or 200 M PPi. All of the dNTPs in these assays were supplied at 100 M, except for dATP whose concentration ranged from 1 to 800 M, depending on the assay. Active enzyme concentrations in these assays were in the range of 10 -20 nM, and the concentration of the template-primer was 30 nM. The samples were incubated for 0 -15 min depending on the assay. In all cases, the incubation times were within the linear range of the corresponding time course. The percentage of inhibition was plotted against the concentration of dATP, and the data were fitted to a hyperbola to obtain the IC 50 for each enzyme. Reported values were obtained from two to three experiments and are given as the means Ϯ standard deviations. b N.D., not determined due to the low efficiency of the excision reaction. viruses recovered from transfections were able to infect and replicate in SupT1 cells. Recombinant HIV-1 clones were assayed to measure their susceptibility to nucleoside analogues, including AZT and d4T (Table III). SS RT showed high level resistance to AZT and 3TC and moderate resistance to d4T, ddI, and ddC. Deleting the two-serine insertion or introducing mutations at codons 67 and 215 did not produce large effects on resistance to 3TC, ddC, or ddI. The mutant SS_D67N showed a resistance profile that was very similar to the one displayed by SS RT. However, replacing Tyr-215 with Thr, Ser, or Asn in the multidrug-resistant SS RT led to virus sensitive to both thymidine analogues. On the other hand, introducing the T215Y mutation together with the insertion T69SSS within the wildtype BH10 RT sequence background produce a significant increase in resistance to AZT and particularly to d4T. These data are in good agreement with the results of the biochemical assays, indicating that Tyr-215 is a very important residue in the acquisition of resistance to thymidine analogues displayed by multidrug-resistant insertion-containing RTs.

DISCUSSION
The presence of a dipeptide insertion at codons 69 -70 in combination with amino acid substitutions related to drug resistance (including TAMs) in the HIV-1 RT has been reported by several groups. These viral isolates display high level resistance to AZT and moderate to low level resistance to other NRTIs in phenotypic assays. The insertion by itself does not have a major impact on resistance in the absence of drugresistant mutations but contributes to AZT resistance in the presence of TAMs and other mutations (27,46). A good correlation between the results of phenotypic assays and the ATPdependent phosphorolytic activity displayed by the recombinant RTs has been observed. This activity is higher with AZTterminated primers, followed by d4T-terminated primers and very low (albeit detectable) for primers terminated with other NRTIs, including cytidine analogues (13) and tenofovir (47).
T215Y is the most frequent mutation associated with dipeptide insertions in the RT-coding region. HIV clones containing the insertion together with mutation T215Y in the sequence context of wild-type BH10 RT had reduced susceptibility to AZT and d4T, although AZT resistance was not as high as with the SS RT that contains additional TAMs such as M41L or L210W. These observations are broadly in agreement with phenotypic data reported by others (26,28) and correlate with the ATP-dependent phosphorolytic activities displayed by the mutant RTs on AZT-and d4T-terminated primers. Although the presence of a Ser-Ser insertion together with an aromatic side chain at position 215 is sufficient to confer some resistance, it is clear that further changes are needed to achieve the high level resistance displayed by the SS RT. Previously reported evidence revealed that T215Y alone produced a very small increase (Ͻ3-fold) in the ATP-dependent phosphorolytic activity on primers terminated with AZT (48,49). Therefore, the question of whether Tyr-215 is a requirement for maintaining ATP-dependent phosphorolytic activity on primers blocked with thymidine analogues was addressed by substituting Thr, Ser, and Asn for Tyr-215 in the SS RT. The mutant enzymes obtained were unable to remove AZT-monophosphate and d4T- monophosphate from inhibitor-terminated primers, and the corresponding mutant viruses were all susceptible to inhibition by thymidine analogues. Interestingly, the PPi-mediated removal activity of primers terminated with d4T, conferred by Tyr-215 mutants was also reduced in comparison with the parental SS RT. Although the physiological relevance of PPi is not clear, all of these data suggest that these mutations have a significant effect in the optimal alignment of the chain terminated primer relative to the PPi donor.
HIV clones having mutations Y215T, Y215S, or Y215N within the SS RT retained low level resistance to ddI and ddC and high level resistance to 3TC in phenotypic assays. High level resistance to 3TC is expected because of the presence of M184I in all SS-derived clones, but SS RT lacks drug resistance-specific mutations for ddI and ddC (i.e. K65R, L74V). It has been shown that primers terminated with ddC and ddA (the active metabolite of ddI) are substrates of the ATP-dependent phosphorolytic activity catalyzed by AZT-resistant RTs carrying the T215Y/F mutation. However, the efficiency of the removal reaction was very low in comparison with thymidine analogues (11)(12)(13)(14)(15). For ddA-terminated primers, it has been shown that ATP-dependent removal reactions catalyzed by RTs having mutations M41L/T69SSS/T215Y and M41L/T69SSS/ L210W/R211K/L214F/T215Y were 10-and 18-fold more efficient than the wild-type enzyme, whereas the insertion alone had no effect on the unblocking reaction (43). Despite these large differences that point to M41L and T215Y as key determinants for ATP-dependent phosphorolytic activity, none of the recombinant viruses generated with those enzymes showed significant resistance to ddI in phenotypic assays (26). Taken together, these data suggest that the contribution of the ATPdependent phosphorolytic activity to the low level resistance to ddI observed with SS RT is likely to be small and is probably due to the complex arrangement of drug resistance mutations found in the multidrug-resistant RT.
The efficiency of nucleoside analogue removal under physiological conditions depends on the rate of the reaction but also on the ability of dNTPs to inhibit it. HIV-1 RT bound to a chain-terminated template-primer and an incoming dNTP complementary to the next nucleotide on the template strand forms a stable ternary complex. A crystal structure of a ternary complex containing a dideoxyguanosine-terminated DNA primer complexed with a DNA template and dTTP is available (50) (Fig. 7). Under these conditions, the excision reaction is blocked. The concentrations of dNTP required to inhibit the unblocking reaction have been determined for a number of RTs displaying ATP-dependent phosphorolytic activity (i.e. D67N/ K70R/T215F/K219Q, M41L/T215Y, M41L/T69SSS/T215Y, etc.) (9,11,13,43). Although the dNTP concentration required to inhibit AZT-monophosphate removal was rather high (IC 50 40 to Ͼ250 M), the IC 50 values obtained with d4T-, ddT-, and ddA-terminated primers were usually below 25 M. These differences have been attributed to the presence of a bulky substituent at the 3Ј position of the ribose in the AZT-terminated primers that sterically hinders the dNTP-binding site (termed the N site), reducing its accessibility by dNTPs (52). Under our assay conditions none of the mutations tested had a significant effect on the IC 50 values for AZT and d4T removal. Furthermore, the presence of ATP or PPi had no effect on the inhibition by the next complementary dNTP.
The formation of ternary complexes has been monitored for wild-type BH10 RT and mutant M41L/T69SSS/T215Y with chemical footprinting techniques in the presence of increasing concentrations of the next complementary nucleotide and DNA⅐DNA complexes having AZT-terminated primers (44). The IC 50 values obtained were 22 M for the wild-type RT and 570 M for the mutant containing the insertion. Higher IC 50 s were also observed with the mutant using DNA⅐DNA complexes having a ddT-terminated primer. However, large differences were not detected in our assays measuring RT suscepti-  bility to dNTP inhibition in AZT and d4T removal reactions. This could be attributed to the RT sequence context or the different assay conditions. However, monitoring ternary complex formation using a gel mobility shift assay has also revealed significant differences in the apparent K d for dNTP binding with the enzymes described in this work. SS RT was about 6-fold more resistant to dNTP inhibition than the wildtype enzyme. Furthermore, Tyr-215 was partially responsible for the observed differences, because those RTs having Ser or Thr at position 215 were somewhat more sensitive to inhibition than those with Tyr at this position. The reason for the discrepancy between determinations based on excision followed by extension reactions and those obtained by monitoring ternary complex formation relates to the presence of the PPi donor.
Using an excess of foscarnet (a PPi analogue), all of the RTs displayed similar susceptibility to inhibition by the next complementary dNTP. The apparent K d [dNTP] values were similar to the IC 50 values reported for the excision reactions. These results suggest that despite being outside the dNTP-binding site and away from the catalytic pocket, Tyr-215 exerts a significant influence on the conformation of the dNTP-binding pocket. This effect can be counteracted in the presence of a PPi analogue.
Modeling studies have led authors to propose that Tyr-215 (and other TAMs) could participate in ATP binding to facilitate nucleotide analogue excision of chain-terminated primers when the primer terminus sits at the N site (52,53). Tyr-215 is exposed to the solvent in the crystal structures of RTs containing the mutation T215Y (and other TAMs) (54,55), and a direct participation of this residue in binding the adenine base of ATP through van der Waals' interactions could occur. However, ATP cannot contact simultaneously the mutated side chains of residues 41 and 215. Available biochemical data are also not consistent with the hypothesis that ATP is bound more tightly by AZT-resistant RTs (15). The available crystal structures of AZT-resistant RTs are complexes of RT bound to a non-nucleoside RT inhibitor and show a conformation where the distance between fingers and thumb subdomains is even larger than in the "open" conformation of the binary complex. In this scenario, the possibility of Tyr-215 (and other TAMs) going through large conformational changes during polymerization that could explain their effects on phosphorolysis cannot be excluded. Another factor that adds further complexity to the study of the mechanism of ATP-based excision is that the removal reaction requires millimolar concentrations of ATP, which are high enough to compete with dNTPs for binding at the N site (56). Available crystal structures of binary and ternary complexes containing the wild-type RT show significant conformational changes at the fingers and thumb subdomains (Fig. 7) that result in a large movement of the ␤3-␤4 hairpin loop that contains the insertion in the SS RT. In the closed configuration of the ternary complex, Lys-65 interacts directly with the ␥ phosphate of the incoming dNTP (50), whereas in the binary complex, this residue is at the tip of the fingers domain, far from the dNTP-binding site (51). In contrast, conformational changes affecting ␤ strand 11a, which contains Thr-215, are less important. Our data suggest that under physiological conditions (i.e. high concentrations of ATP), the resistance mechanism is dependent on the orientation of the PPi moiety of the FIG. 7. Structural analysis of HIV-1 RT binary and ternary complexes showing conformational changes of the ␤3-␤4 hairpin loop and the location of residue 215. A, superimposition of the ␣-carbon backbones of the RT forming an RT-DNA⅐DNA binary (dark blue ribbon) and an RT-DNA⅐DNA-dNTP ternary complex (light blue ribbon). The template (red) and primer (white) strands shown with a stick representation are those found in the ternary complex. The incoming dNTP is shown in yellow using a CPK representation. B, detailed view of the location of Asp-67, Thr-69, and Lys-70 in the ␤3-␤4 hairpin loop and Thr-215 in ␤-strand 11a to show the conformational changes involved in ternary complex formation. C, shows side views of the ternary (left panel) and binary (right panel) complexes illustrating the large conformational changes involving the tip of the ␤3-␤4 hairpin loop (yellow asterisk). The DNA⅐DNA complex and the incoming dNTP are represented with the same color codes as above but using CPK models. The 66-and 51-kDa subunits are shown using a cyan or a green ribbon representation, respectively. The coordinates have been obtained from Protein Data Bank codes 1RTD (50) and 3HMI (51), and the models have been prepared using the program Insight II (Accelrys Inc.).
donor, which appears to be largely affected by the side chain of Tyr-215 and to a lesser extent by the conformation of the ␤3-␤4 hairpin loop.
Previous studies have shown that interrupting nucleoside analogue therapy leads to replacement of the multidrug-resistant virus by wild-type strains (42,57). Replacing Tyr with Thr involves two nucleotide substitutions. However, one mutation (i.e. Y215N or Y215S) would be sufficient to abrogate thymidine analogue resistance, thereby facilitating an alternative sensitization pathway. This reversion pathway has not been described in patients infected with strains having insertion-containing RTs, but it is relatively common in patients having the classical AZT resistance mutations (58,59), where mutations T215D, T215C, and T215S have a relatively low fitness cost in the absence of drugs and are often observed in transmitted virus (60,61). Mutation D67N is very rare in insertion-containing RTs. Virus population dynamics studies have shown that in some isolates this mutation is detected very early together with T215Y and the insertion but is quickly replaced by D67S, suggesting that D67N confers a fitness loss to the multidrugresistant virus. Our biochemical studies did not reveal any significant differences between the SS RT and mutant SS_D67N, whereas recombinant virus containing those enzymes were viable and replicated efficiently in SupT1 cells. Detailed studies are currently in progress to measure any possible fitness differences between both enzymes.