Mechanisms Associated with HIV-1 Resistance to Acyclovir by the V75I Mutation in Reverse Transcriptase*

It has recently been demonstrated that the anti-herpetic drug acyclovir (ACV) also displays antiviral activity against the human immunodeficiency virus type 1 (HIV-1). The triphosphate form of ACV is accepted by HIV-1 reverse transcriptase (RT), and subsequent incorporation leads to classical chain termination. Like all approved nucleoside analogue RT inhibitors (NRTIs), the selective pressure of ACV is associated with the emergence of resistance. The V75I mutation in HIV-1 RT appears to be dominant in this regard. By itself, this mutation is usually not associated with resistance to currently approved NRTIs. Here we studied the underlying biochemical mechanism. We demonstrate that V75I is also selected under the selective pressure of a monophosphorylated prodrug that was designed to bypass the bottleneck in drug activation to the triphosphate form (ACV-TP). Pre-steady-state kinetics reveal that V75I discriminates against the inhibitor at the level of catalysis, whereas binding of the inhibitor remains largely unaffected. The incorporated ACV-monophosphate (ACV-MP) is vulnerable to excision in the presence of the pyrophosphate donor ATP. V75I compromises binding of the next nucleotide that can otherwise provide a certain degree of protection from excision. Collectively, the results of this study suggest that ACV is sensitive to two different resistance pathways, which warrants further investigation regarding the detailed resistance profile of ACV. Such studies will be crucial in assessing the potential clinical utility of ACV and its derivatives in combination with established NRTIs.

clinic to treat infections caused by herpes simplex virus 1 and 2 (HSV-1 and HSV-2) (1-3). As its valyl prodrug, and to a lesser extent as parent ACV, it is also used in treating varicella zoster infections. The drug is an acyclic guanosine analogue that needs to be selectively converted to its triphosphate form (ACV-TP) that is accepted by the viral polymerase and acts as a chain terminator. The herpesviruses provide the kinases that generate the monophosphate (ACV-MP), whereas cellular enzymes are required to synthesize the triphosphate form (4 -6). ACV-TP competes with intracellular dGTP pools for incorporation. Once incorporated, it acts as a chain terminator because of the lack of a structural equivalent of the 3Ј-hydroxyl group of the sugar moiety of a natural nucleotide (7,8). The next complementary nucleotide, immediately downstream of the ACVterminated 3Ј-end of the primer, can still bind to the HSV DNA polymerase and triggers formation of a stable dead-end complex (DEC) (9).
It has recently been demonstrated that under certain conditions ACV also exhibits antiviral activity against the human immunodeficiency virus type 1 (HIV-1) (10,11). ACV was shown to suppress HIV-1 replication in human tissues co-infected with HIV-1 and human herpesviruses (10). The latter provide the viral kinase that facilitates production of ACV-MP. This bottleneck in the production of the active antiviral agent can also be bypassed with a monophosphorylated prodrug (CF2648) (Fig. 1, left), that shows anti-HIV activity in herpesvirus-free cells (4,12,13). Cell-free assays revealed HIV-1 reverse transcriptase (RT) as the target (10,11). ACV-TP binds to the nucleotide binding site, and the incorporated ACV-MP causes DNA chain termination after the release of pyrophosphate (PPi). Like most other nucleoside analogues (14 -16), the incorporated ACV-MP can be excised from the 3Ј-end of the primer in the presence of PPi or the PPi-donor ATP (10). This reaction can reduce the overall inhibitory effect; however, the removal of the chain terminator can be blocked through formation of a DEC (17). DEC formation depends critically on the chemical nature of the inhibitor (18). High concentrations (Ͼ100 M) of the next nucleotide are required to form a DEC with a primer terminated with zidovudine (AZT), whereas submicromolar concentrations are often sufficient to form a DEC with ddNTPs (16). ACV-MP shows a behavior in the middle of the spectrum; ϳ25 M concentrations of the next nucleotide inhibits excision by 50% (10).
In vitro selection experiments revealed that ACV drug pressure is linked to the emergence of mutation V75I in the RT gene (11). A similar change, i.e. V75T, has earlier been associated with resistance to stavudine (19). Mutations M184V and T69N are other previously known resistance-conferring mutations that emerged under the selective pressure of ACV; however, V75I outgrew the culture over protracted periods of time, suggesting that this mutation is strongly associated with ACV resistance. HIV variants containing V75I showed marked increases in 50% effective antiviral concentrations (EC 50 ), which confirms the selection experiments (11).
Here we studied the underlying biochemical mechanism of HIV resistance to ACV associated with V75I. Two major mechanisms of resistance to nucleoside analogue RT inhibitors (NRTIs) have been described (20 -24). The first mechanism is based on substrate discrimination. In this case the mutant enzyme can selectively diminish binding and/or incorporation of the nucleotide analogue, whereas the properties of the natural counterpart remain largely unaffected. M184V that confers high level resistance to 2Ј,3Ј-dideoxy-3Ј-thiacytidine is a prominent example in this regard. The second major resistance mechanism associated with NRTIs is based on excision. In this case, the mutant enzyme can increase the rate of excision of the incorporated inhibitor. Thymidine analogue-associated mutations (TAMs) were shown to be able to recruit ATP as a PPi donor and increase excision of incorporated AZT-MP (17). In this study we demonstrate that V75I discriminates against ACV-TP at the level of incorporation. Excision of the incorporated nucleotide is also increased when compared with wild type RT; however, the effect is less pronounced, as seen with the excision of ACV-MP against a background of TAMs. V75I does not provide further protection from excision through DEC formation. Collectively, the data suggest that ACV is vulnerable to both major resistance mechanisms.

EXPERIMENTAL PROCEDURES
Enzymes and Nucleic Acids-Heterodimeric reverse transcriptase p66/p51 was expressed and purified as described (25). Mutant enzymes were generated through site-directed mutagenesis using the Stratagene QuikChange kit according to the manufacturer's protocol. TAM2 refers to HIV1-RT con-taining the following substitutions: D67N, K70R, T215F, and K219Q. Oligodeoxynucleotides used in this study were chemically synthesized and purchased from Invitrogen and from Integrated DNA Technologies. The following sequence was used as template T50A6: 5Ј-CCAATATTCACCATCAAGGC-TTGATGAAACTTCACTCCACTATACCACTC. The underlined nucleotides are the portion of the templates annealed to the primer. The following primer was used in this study: P1, 5Ј-GAGTGGTATAGTGGAGTGAA.
Synthesis of ACV-TP-ACV (1.5 mmol) was dissolved in 200 l of dry 1,3-dimethyl-2-oxohexahydropyrimidine N,NЈ-dimethylpropylene urea with 12-15 molecular sieves under nitrogen and stirred for 24 h. The mixture was chilled with an icewater bath and stirred for 1 h followed by slow addition of 3 eq of phosphorus oxychloride and stirring for an additional 25 min. A solution of tributylammonium pyrophosphate (4 eq) in 200 l of N,NЈ-dimethylpropylene urea and tributylamine (15 eq) was simultaneously added to the reaction. After 45 min the reaction was quenched with ice-cold water, and then it was slowly brought to the room temperature. The reaction was washed with chloroform, and the aqueous layer was collected and co-evaporated with deionized water three times. The residue was resuspended in 100 l of deionized water and purified on an ion-exchange column by high performance liquid chromatography (max) ϭ 253. To reduce the amount of excess salt, the final product was co-evaporated with water 5 times, giving total yield of ACV-TP (NH 3 ) 4 of 18% with purity Ն95%. The molecular weight of the ACV-TP was confirmed by liquid chromatography-tandem mass spectrometry; m/z (Mϩ1) 4663152 (26). Synthesis of CF2648-The ACV prodrug CF2648 was prepared by the coupling of suitably base-protected ACV with the appropriate phosphorochloridate reagent under anhydrous conditions followed by base deprotection, according to procedures we have previously reported (27).
Competition between ACV-TP and dGTP-DNA synthesis was monitored with 5Ј-end-labeled primers unless otherwise indicated. 150 nM DNA/DNA (T50A6/P1) was incubated with 30 nM HIV-1 RT in a buffer containing 50 mM Tris-HCl, pH 7.8, 50 mM NaCl, constant concentrations of dGTP (1 M) and ddTTP (2 M) or dNTP mix (1 M) and increasing concentrations of ACV-TP. Nucleotide incorporation was initiated by the addition of MgCl 2 to a final concentration of 10 mM, and the reactions were allowed to proceed for 5 min. The reactions were stopped by the addition of 3 reaction volumes of formamide containing traces of bromphenol blue and xylene cyanol. The samples were then subjected to 15% denaturing PAGE followed by phosphorimaging. Inhibitory concentrations of ACT-TP required to inhibit DNA synthesis at position ϩ2 by 50% (IC 50 ) was determined by normalizing the product fraction formed at position ϩ2 in the presence of ACV-TP to the corresponding value in the absence of inhibitor. Data points were fit to a sigmoidal dose response (variable slope) function using GraphPad Prism (Version 5.0).
ATP-dependent Excision-The T50A6/P1 DNA/DNA substrate was extended by a single nucleotide to generate an oligonucleotide with ACV-MP at the 3Ј-end. The extended primer was gel-purified and annealed with template T50A6. 50 nM sub- strate was then incubated with 500 nM HIV-1 RT in a buffer containing 50 mM Tris-HCl, pH 7.8, 50 mM NaCl, 10 mM MgCl 2 , and 3.5 mM ATP (pyrophosphatase-treated). Aliquots were taken at different time points and analyzed as described.
Site-specific Footprinting-In preparation of the footprinting experiments, template T50A6 was 5Ј-end-labeled and heat-annealed with primer P1. 50 nM DNA/DNA hybrid was incubated with 750 nM HIV-RT for 10 min in a reaction mixture containing sodium cacodylate, pH 7 (120 mM), NaCl (20 mM), DTT (0.5 mM), MgCl 2 (10 mM), and 25 M ddGTP or 5 M ACV-TP in a final volume of 15 l. In control reactions ddGTP and ACV-TP were either omitted or substituted with 100 l phosphonoformic acid (PFA or foscarnet). After complete complex formation and/or nucleotide incorporation, increasing concentrations of dTTP were added to the reactions followed by an incubation of 5 min at 37°C. For the actual footprinting, complexes were treated with 0.1 mM ammonium iron (II) sulfate hexahydrate (28). The reactions were allowed to proceed for 5 min and were processed and analyzed as described.
DEC Formation-DNA synthesis at template position ϩ1 was conducted in a similar fashion as described above except that a chain terminator (ddGTP, 25 M or ACV-TP, 2.5 M) was incorporated at this position. A time course of incorporation of the chain terminator at position ϩ1 in the absence or in the presence of increasing concentrations of the nucleotide at the following position, ϩ2, was monitored. The reactions were processed and analyzed as described. Slopes of the linear portion of product formation illustrate the velocities of the reaction, which when normalized to the enzyme concentration, determine the turnover number (k cat ). Inhibition of DNA synthesis by DEC formation is illustrated by a decrease in k cat .
Pre-steady-state Kinetics for Nucleotide Incorporation-Nucleotide incorporation under single-turnover conditions was monitored using a rapid quench-flow instrument (KinTek RQF-3). Reactions involved rapid mixing of a solution containing preincubated 100 nM DNA/DNA template/primer hybrid with 500 nM HIV-1 RT in a buffer consisting of 50 mM Tris-HCl, pH 7.8, 50 mM NaCl, and 10 mM MgCl 2 at 37°C with a solution of the same buffer composition except that template/primer and RT were substituted with a given concentration of dGTP or ACV-TP. Nucleotide incorporation was monitored at time points of 0.015, 0.025, 0.05, 0.075, 0.1, 0.2, 0.3, 0.5, and 1 s. The reactions were processed and analyzed as described. Data points from time courses were fit by nonlinear regression (GraphPad Prism (Version 5.0)) to a single exponential equation, [product] ϭ A(1Ϫexp(Ϫk observed t)), where A represents the amplitude, and k observed is the first order rate constant for dGTP or ACV-TP incorporation. k observed were replotted versus increasing concentrations of dGTP or ACV-TP to determine the respective k observed . Data points were fit to a hyperbolic function, where k pol is the maximum first-order rate constant for dGMP or ACV-MP incorporation, and K d,dNTP is the equilibrium dissociation constant for the interaction of dGTP or ACV-TP with the RT-template/primer complex.
Selection of Resistance-MT-4 cells were obtained from the NIH AIDS Research and Reference Reagent Program and infected with HIV-1 LAI.04 in the absence or presence of CF2648. Every 3-5 days, 3% of the culture supernatant was used to infect fresh cells. Cultures were maintained in the absence or presence of gradually escalating concentrations of CF2648. HIV-1 replication in infected cultures was assessed by measuring p24 gag from culture supernatants, as previously described (10). Phenotypic resistance of viruses serially passed in the absence or in the presence of the CF2648 was evaluated in MT-4 cell cultures using drug concentrations ranging from 0 to 150 M. The effective concentration that inhibits 50% replication was calculated by fitting the data points to a sigmoidal dose-response curve using GraphPad Prism (Version 4.0).
Genotyping of the Drug-exposed HIV-1 Strains-Viral RNA was extracted from plasma using the QIAamp Viral RNA Mini Kit (Qiagen, Hilden, Germany). HIV-1 RNA was reverse-transcribed into cDNA, and a 2878-bp nucleotide fragment encompassing protease and reverse transcriptase was amplified in an outer PCR using SuperScript III One-Step RT-PCR System with Platinum Taq High Fidelity (Invitrogen) and outer primers AV190-1 and CR1 (29). A 2853-bp nucleotide fragment was amplified using Expand High Fidelity PCR System (Roche Diagnostics) and the inner primers AV190-2 and CR2. Amplification products were separated on a 1% agarose gel and visualized by ethidium bromide staining. PCR products were purified with Microspin S-400 (GE Healthcare). Sequencing was performed using the ABI PRISM BigDye Terminator v3.1 Ready Reaction Cycle Sequencing kit as described before (29). The reactions were run on an ABI3100 Genetic Analyzer, and analysis was performed using Sequence Analysis Version 3.7 and SeqScape version 2.0 (Applied Biosystems, Nieuwerkerk a/d Ijssel, The Netherlands).

RESULTS
Selection of Resistance with CF2648-The V75I mutation in HIV-1 RT was shown to emerge rapidly under the selective pressure of ACV (11). The poor production of ACV-MP and, in turn, the poor production of ACV-TP is a possible factor that diminishes the efficacy of ACV and facilitates the selection of resistance. It is currently unknown whether monophosphorylated prodrugs can prevent or significantly delay the development of resistance. To address this problem, we attempted to select for resistance with the monophosphorylated prodrug CF2648. HIV-1 was propagated in MT-4 cells in the presence of increasing concentrations of CF2648. The antiviral activity of CF2648 after 39 serial passages (ϳ120 days) in the presence of the compound was reduced relative to the control propagated in the absence of the drug. Sequencing of HIV-1 RT revealed the emergence of V75I, and phenotypic resistance measurements revealed a 15-fold increase in the EC 50 value (Table 1). Overall, the data suggest that the selection of V75I in the RT enzyme may not be prevented with prodrugs that bypass the first phosphorylation step. Given the potential relevance of this mutation in clinical settings, we studied the underlying biochemical mechanisms. Discrimination against ACV-TP-We initially studied substrate discrimination as a possible mechanism for resistance. Classic protocols that address this problem include steady-state kinetics that measure efficiencies of incorporation of the inhibitor and its natural counterpart in separate experiments. However, with this approach we were unable to detect significant differences between purified wild type (WT) RT and the mutant enzyme containing V75I (data not shown). We considered two potential problems associated with this assay. First, incorporation of the inhibitor was measured in the absence of the natural counterpart dGTP. Therefore, we devised an assay that monitors incorporation of ACV-MP in the presence of both ACV-TP and dGTP ( Fig. 2A). The primer/template was designed to monitor incorporation of ACV-MP at the next template position (n ϩ 1) after the 3Ј-end of the primer (n). Incorporation of ACV-MP causes chain termination, whereas incorporation of dGMP allows the addition of a second nucleotide that is provided as a "stop nucleotide," which limits DNA synthesis and reduces complexity of the assay. The natural dGTP was provided at a constant concentration of 1 M, and ACV-TP was provided at varying concentrations (Fig. 2B). We found that the concentrations required to diminish dinucleotide extensions by 50% were significantly higher with the mutant enzyme, which suggests that the V75I mutant increases discrimination against the inhibitor (Fig. 2C). We obtained essentially the same result with a different primer/template sequence that allowed us to monitor ACV-MP incorporation at template position ϩ7 (Fig. 3).  A second common problem associated with classic steadystate kinetics is that the multiple turnovers can mask differences in substrate binding or catalysis (30). To address this problem we measured kinetic parameters under pre-steadystate conditions, which provides at the same time mechanistic insight ( Table 2). WT RT binds ACV-TP and dGTP with similar affinity as evidenced by K d values of 2.2 and 1.3 M, respectively. However, the rate of incorporation is significantly higher for dGMP (k pol ϭ 66 s Ϫ1 ) as compared with ACV-TP (k pol ϭ 14 s Ϫ1 ), which translates to an 8-fold selective advantage of the natural nucleotide over the inhibitor with regard to the efficiency of nucleotide incorporation (k pol /K d ).
The efficiency of incorporation of the natural nucleotide is not largely affected by the V75I mutation. The subtle increase in the K d value is neutralized by a subtle increase in the k pol value, and these variations are not considered to be relevant. However, we measured a marked decline in the rate of incorporation for ACV-MP when V75I is compared with WT RT (k pol ϭ 1.5 s Ϫ1 versus k pol ϭ 14 s Ϫ1 ), whereas the K d value is not significantly increased. The selective effect on the rate constant suggests that V75I affects the catalytic step. Nucleotide binding of ACV-TP is not significantly changed. Overall, the data point to a 10-fold increase in discrimination against the inhibitor at single nucleotide resolution.
Excision of ACV-MP-An increased rate of excision is yet another possible mechanism that could help to explain the resistance phenotype associated with V75I. We devised a primer that was terminated with ACV-MP and studied the efficiency of the ATP-dependent excision reaction with WT RT, V75I, and a mutant enzyme that contains four TAMs (D67N, K70R, T215F, and K219Q). This particular combination represents the TAM2 pathway. Like other combinations of TAMs, this cluster facilitates the recruitment of ATP as a PPi donor. In agreement with previous reports, excision with WT RT is inefficient (Fig. 4). However, V75I appears to increase the efficiency of ACV-MP excision over time, and the TAM-containing enzyme shows further significant increases. Thus, increases in excision of the incorporated ACV-MP may be considered as another possible factor that contributes to ACV resistance.
Positioning of RT on ACV-terminated Primers-The precise positioning of RT on its primer/template and, in turn, enzymatic function can be influenced by the chemical nature of the 3Ј-end of the primer (31,32). Before nucleotide incorporation, the enzyme needs to translocate a single position farther downstream, which liberates the nucleotide binding site from the 3Ј-end of the primer (28,33,34). In contrast, excision can only take place when the 3Ј-end of the primer occupies the nucleo-tide binding site (15,21,33). The two conformations are referred to as post-and pre-translocational states, and the RT enzyme is able to shuttle between the two conformations. We have recently developed site-specific footprinting techniques that allow us to distinguish between the two complexes (28,32). Incubation of RT-DNA/DNA primer/template complexes with Fe 2ϩ generates specific cuts on the template strand. These cuts are mediated through the RT-associated ribonuclease H active site at the C-terminal domain of the enzyme. Cleavage at position Ϫ17 is indicative for post-translocated complexes, whereas cleavage at position Ϫ18 is indicative for pre-translocated complexes. Here we compared footprints obtained with WT RT and V75I in complex with either ddGMP-or ACV-terminated primers (Fig. 5). Complexes with the bound nucleotide can only exist in the post-translocational state, whereas complexes with the PPi-analogue PFA exist pre-translocation (see controls). In an attempt to gradually stabilize the complex (DEC formation), we increased concentrations of the next complementary nucleotide, i.e. dTTP, which correlated with an increase in signal intensity. Complexes with WT RT and ddGMP-terminated primers exist predominantly in the post-translocated state even in the absence of nucleotide substrate. The signal is increased in the presence of increasing concentrations of the next nucleotide, which confirms formation of a DEC in the post-translocational state. In contrast, corresponding complexes with ACVterminated primers are difficult to identify. Much higher concentrations of the next nucleotide are required to trap the complex in the post-translocational state. We obtained very similar results with the V75I mutant. The trend is the same, although the overall band intensity appears slightly diminished when compared with WT RT. Overall, these findings suggest that DEC formation is compromised with ACV-MP, and the  V75I mutation may further contribute to this effect. The diminished signal points to increased complex dissociation. DEC Formation-To compare and quantify the ability of WT RT and V75I to form a DEC, we measured the turnover after incorporation of ACV-MP and ddGMP, respectively. The simultaneous inclusion of the next nucleotide at different concentrations allows formation of the ternary DEC (Fig. 6). In this set-up, increases in the turnover number (k cat ) correlate inversely with DEC formation (Table 3). WT RT shows a decline in the turnover with ddGMP-terminated primers and increasing concentration of the next nucleotide. DEC formation is ϳ8-fold enhanced at concentrations of 100 M dTTP when compared with the control in the absence of the nucleotide substrate. Under the same conditions DEC formation is compromised with ACV-terminated primers (only 2-fold increases with 100 M dTTP), and V75I further enhances this effect. These findings suggest that DEC formation is likely to be insignificant under physiologically relevant conditions with dNTP concentrations below 100 M. Thus, the ACV-terminated primer is unlikely to be protected from excision in the mutational context of V75I.

DISCUSSION
The potential clinical benefit of ACV or the prodrug version valacyclovir in treating HIV/HSV-2 co-infection has been tested in large clinical trials (35)(36)(37). HIV co-infection with HSV-2 can exacerbate disease progression (38,39). In support of this notion, several trials have been shown that ACV-mediated suppression of HSV-2 is associated with reductions in HIV viral load (36,40). In contrast, other trials failed to show that anti-herpetic therapy prevents infection with HIV-1 (41,42). Two subsequent studies demonstrated that ACV can also directly inhibit HIV-1 replication by targeting the RT enzyme (10,11). However, the development of resistance can compromise the antiviral activity in clinical settings. Three different mutations in HIV-1 RT were shown to emerge under the selective pressure of ACV in vitro, M184V, T69N, and V75I (11). Experiments with constructs that were generated by site-directed mutagenesis confirmed the selection experiments and revealed that each of the three mutations can reduce susceptibility to ACV. V75I shows by far the strongest effect in this regard. Moreover, here we demonstrate that V75I is likewise selected in the presence of a monophosphorylated prodrug derivative of ACV that was developed to bypass the bottleneck in the metabolic conversion of ACV to ACV-TP. Although other mutations independent of or in conjunction with V75I may also contribute to HIV-1 resistance to ACV or its prodrugs, here we focused on the characterization of the effect of V75I on the function of RT.
Residue Val-75 is located in close proximity to template position n ϩ 1 opposite the incoming nucleotide (Fig. 7) (43). Thus, a direct effect of a mutation at this position on binding and/or incorporation of nucleoside analogues is not evident. V75I is part of the "Q151M cluster" that is associated with multiple resistance to NRTIs (44,45). However, in this context, V75I does not appear to contribute to the resistance phenotype; the mutation rather compensates for enzymatic deficits that are introduced by Q151M (46). A related mutation, i.e. V75T, has been associated with low level (3-4-fold) resistance to stavudine, which is a 2Ј,3Ј-unsaturated thymidine analogue and, therefore, structurally distinct from ACV (19,46). The increase of approximately 1-2 orders of magnitude in resistance to ACV is by far the strongest effect that has been reported for changes at this position (Ref. 11 and this study). We consider three complementary mechanisms that help to explain the resistant phenotype.
Discrimination against ACV-TP-Competition experiments under steady-state conditions and pre-steady-state kinetics suggest that V75I further discriminates against the inhibitor. The selective advantage for the natural nucleotide over the inhibitor that is seen with WT RT is 10-fold increased with the V75I mutant. The difference can be assigned to equivalent changes in k pol values. Thus, the V75I mutation appears to compromise the chemical step, whereas affinities to substrate and inhibitor remain largely unchanged. This result is unexpected given the considerable distance between Val-75 and the active site or the incoming nucleotide (Fig. 7). The V75I change may, therefore, indirectly affect the catalytic step. Specific decreases in k pol values have been reported for other NRTI resistanceassociated mutations (47,48). K65R and Q151M are important examples in this regard (47)(48)(49)(50). Lys-65 is located in close proximity to the ␥-phosphate of the incoming nucleotide, which indicates that changes at this position may affect its proper alignment with the attacking 3Ј-end of the primer. The side chains of Gln-151 and also Arg-72 are seen in the vicinity of the ␣-phosphate of the incoming nucleotide, which suggests a direct role in the catalytic step (51,52). By extension, changes at this position can directly compromise catalysis at the center of the reaction. It is, therefore, conceivable that changes at position Val-75 can affect the precise positioning of one of the two residues or both, which could help explain the observed decreases in k pol values. The backbone of Val-75 is in contact distance with the backbone of Gln-151, and subtle structural alterations at Val-75, i.e. V75I, can impact on the positioning of Arg-72, both part of the flexible ␤3-␤4 hairpin loop that traps the incoming nucleotide. Regardless of the precise mechanism, the effect of V75I on ACV-MP incorporation appears to be indirectly mediated through other amino acids. In contrast, classic NRTI-associated mutations, such as M184V, K65R, and Q151M, discriminate against the inhibitor directly at the active site.
Previous other studies have shown that changes at Val-75 can also affect the affinity to the substrate. V75T showed increases in K d values for stavudine without affected k pol (46). Moreover, mispair extension experiments with the V75I mutant enzyme showed likewise significant increases in K d values without affecting k pol (53). Together the results suggest that changes at position Val-75 can affect nucleotide incorporation through   a DEC formation is calculated as a ratio of k cat in the absence of dTTP over k cat in the presence of dTTP. b -Fold change reflects differences in DEC formation by WT RT over V75I mutant RT. c The 3Ј-end of the primer was terminated in the presence of ddGTP or ACV-TP to prevent incorporation of the next nucleotide during DEC formation. d Next complementary nucleotide. e S.D. were determined on the basis of at least three independent experiments. different mechanisms. Mutations at this position can affect substrate binding or the catalytic step. Whether the effects of V75I on the catalytic step of ACV-MP incorporation can be ascribed to the acyclic nature of the inhibitor remains to be seen. However, in this context it is interesting to note that the acyclic phosphonate tenofovir, which is an important component in currently used drug regimens, is sensitive against the Q151M cluster that includes V75I (54).
Excision of ACV-MP-Excision of incorporated NRTIs is a second major mechanism for NRTI resistance. TAMs increase the excision of AZT-MP when compared with WT RT (15,16); however, the reaction is not restricted to thymidine analogues. The efficiency of the ATP-dependent excision reaction is relatively high with AZT-MP at the 3Ј-end of the primer. In contrast, ddNMPs and 2Ј,3Ј-dideoxy-3Ј-thiacytidine-MP terminated primer strands are poor substrates for the reaction, although these nucleotides analogues are not completely resistant to excision (55). Here we demonstrate that a mutant enzyme containing four TAMs can also markedly increase excision of ACV-MP. Of note, purine analogues are usually poorly excised when compared with their pyrimidine counterparts, which indicates that the base moiety is an important determinant for the reaction (56). However, the acyclic purine analogues tenofovir (32) and ACV-MP are both efficiently removed from the primer terminus, suggesting that the increased flexibility of the acyclic linker between the base and the phosphate or phosphonate of these compounds can facilitate the reaction.
The V75I mutation shows an intermediate phenotype when the excision of ACV-MP is compared with WT RT and TAMs. However, the mechanism associated with such an increase in efficiency of the excision reaction is likely to be different as described for TAMs. TAMs were shown to facilitate binding of ATP in a catalytically competent fashion (15). The aromatic side chains of T215Y or T215F, i.e. the hallmark for AZT resistance, are implicated in stacking interaction with the base moiety of ATP, which facilitates its binding in an orientation that allows its use as a PPi donor (Fig. 7). In contrast, the position of Val-75 in close proximity to template position n ϩ 1 and the nature of the amino acid substitution do not support such interaction. The mutation may indirectly affect the excision reaction in a similar manner as proposed for the incorporation of ACV-MP; however, such an interaction would be counterproductive. Thus, we also consider an alternative interpretation. Excision of ACV-MP with ATP generates the ACV-(5Ј)-tetraphospho-(5Ј)-A (ACVp4A) that is eventually released from the complex. The tetraphosphate can be used as a substrate (16,57), resulting in re-incorporation of ACV-MP. At this level, the V75I mutant has a disadvantage over WT RT, which leads to an accumulation of the excised product. In contrast, TAM-containing enzymes do not significantly affect the efficiency of incorporation of nucleotide analogues (58 -60).
DEC Formation-The formation of a DEC with the next complementary nucleotide after the chain terminator at the 3Ј-end of the primer has two major consequences. First, the bound nucleotide stabilizes the complex, which translates in a diminished turnover of the reaction under steady-state conditions (61). This experimental set-up allowed us to compare and to quantify DEC formation of WT RT with mutant V75I RT. ACV-MP generally diminishes DEC formation when compared with ddGTP, and V75I increased this effect. A similar effect has been observed when comparing tenofovir with ddATP, suggesting that the acyclic linker can negatively influence nucleotide binding (32). V75I may further affect DEC formation through structural alterations at template position n ϩ 1 that is complementary to the next nucleotide. Of note, our data show at the same time that neither the ACV-terminated primer nor the V75I mutation affect the distribution of pre-and posttranslocated complexes.
The second major consequence of a stable DEC is that the 3Ј-end of the primer is protected from excision (16). The nucleotide traps the complex in its post-translocational state, which prevents excision to occur (33). This observation is of potential biological relevance considering that WT RT can excise ACV-MP, although the reaction is inefficient.
Conclusion-The results of this study suggest that discrimination against the inhibitor at the level of incorporation is the dominant mechanism associated with ACV resistance conferred by V75I. Protection against excision is compromised through diminished formation of a ternary DEC, and TAMs can directly increase efficiency of excision of ACV-MP. Thus, ACV appears to be vulnerable to both major NRTI-associated resistance pathways. It is, therefore, essential to characterize the detailed resistance profile of this compound to better assess its potential clinical utility in combination with established antiretrovirals. The use of ACV might be compromised in persons who were recently infected with resistant HIV variants. In addition, the clinical use of ACV may cause the emergence of resistance mutations that can decrease susceptibility to established antiretroviral agents.