Selection and Characterization of HIV-1 with a Novel S68 Deletion in Reverse Transcriptase Raymond F Schinazi, Ivana Massud, Atlanta Veterans Affairs Medical Center Kimberly L. Rapp, Atlanta Veterans Affairs Medical Center Meta Cristiano, Emory University Mervi Acuzar Detorio, Emory University Richard A. Stanton, Emory University Matthew A. Bennett, Atlanta Veterans Affairs Medical Center Monique Kierlin-Duncan, Atlanta Veterans Affairs Medical Center Johan Lennerstrand, Atlanta Veterans Affairs Medical Center James H Nettles, Emory University

Journal Title: Antimicrobial Agents and Chemotherapy Volume: Volume 55, Number 5 Publisher: American Society for Microbiology | 2011-05, Pages 2054-2060 Type of Work: Article | Final Publisher PDF Publisher DOI: 10.1128/AAC.01700-10 Permanent URL: http://pid.emory.edu/ark:/25593/f0hzv

Final published version: http://aac.asm.org/content/55/5/2054 Copyright information: © 2011, American Society for Microbiology. Accessed October 8, 2021 4:50 AM EDT ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, May 2011, p. 2054–2060 Vol. 55, No. 5 0066-4804/11/$12.00 doi:10.1128/AAC.01700-10 Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Selection and Characterization of HIV-1 with a Novel S68 Deletion in Reverse Transcriptaseᰔ Raymond F. Schinazi,* Ivana Massud, Kimberly L. Rapp, Meta Cristiano, Mervi A. Detorio, Richard A. Stanton, Matthew A. Bennett, Monique Kierlin-Duncan, Johan Lennerstrand,† and James H. Nettles Center for AIDS Research, Laboratory of Biochemical Pharmacology, Department of Pediatrics, Emory University School of Medicine, and Veterans Affairs Medical Center, Decatur, Georgia, 30033

Received 7 December 2010/Returned for modification 3 February 2011/Accepted 16 February 2011

Resistance to human immunodeficiency virus type 1 (HIV-1) represents a significant problem in the design of novel therapeutics and the management of treatment regimens in infected persons. Resistance profiles can be elucidated by defining modifications to the viral genome conferred upon exposure to novel nucleoside

reverse transcriptase (RT) inhibitors (NRTI). In vitro testing of HIV-1LAI-infected primary human lymphocytes treated with ␤-D-2؅,3؅-dideoxy-2؅,3؅-didehydro-5-fluorocytidine (DFC; ; Reverset) produced a novel deletion of AGT at codon 68 (S68⌬) alone and in combination with K65R that differentially affects drug response. Dual-approach clone techniques utilizing TOPO cloning and pyrosequencing confirmed the novel S68⌬ in the HIV-1 genome. The S68⌬ HIV-1 RT was phenotyped against various antiviral agents in a heteropolymeric DNA polymerase assay and in human lymphocytes. Drug susceptibility results indicate that the S68⌬ displayed a 10- to 30-fold increase in resistance to DFC, , , tenofovir, -, and amdoxovir and modest resistance to , ␤-D-2؅,3؅-oxa-5-fluorocytidine, or 9-(␤-D 1,3-dioxolan-4-yl)guanine and remained susceptible to 3؅-azido-3؅-deoxythymidine, 2؅,3؅-dideoxyinosine (ddI), 1-(␤-D-dioxolane)thymine (DOT) and . Modeling revealed a central role for S68 in affect- ing conformation of the ␤3-␤4 finger region and provides a rational for the selective resistance. These data indicate that the novel S68⌬ is a previously unrecognized deletion that may represent an important factor in NRTI multidrug resistance treatment strategies.

There are significant detriments associated with the man- atitis (40, 46). Previous reports indicate that DFC-5Ј-triphos- agement and treatment of human immunodeficiency virus type phate (DFC-TP) has a long intracellular half-life and inhibits 1 (HIV-1) that limit treatment efficacy, including the replication of both wild-type (WT) and mutant strains of drug resistance, adherence, tolerability, and long-term toxicity. HIV-1 observed during treatment with (ZDV), The error-prone HIV-1 reverse transcriptase (RT), along with lamivudine (3TC), and other NRTIs (14). In addition, DFC the high rate of turnover for infected cells, contributes to a can select for I63L, K65R, K70N, K70E, and R172K mutations rapid mutation rate coupled with diverse viral quasispecies in in HIV-1 RT in MT-2 cells (11). vivo (23, 26). Together, these factors promote the emergence We report on a novel deletion of the S68 codon alone or in of drug resistance variants that can lead to antiretroviral fail- combination with K65R in HIV-1 RT that reduces the effec- ure, progression to AIDS, and an increase in opportunistic tiveness of DFC in HIV-1LAI-infected primary human lympho- (21, 4). cytes. To test how this mutation affects the virus response to Amino acid substitutions, insertions, or deletions are asso- different NRTI, drug susceptibility assays were performed ciated with reduced susceptibility to numerous nucleoside RT with an in vitro-selected S68⌬ DFC treated viral pool (HIV- inhibitors (NRTIs) (21, 23, 26, 27). In general, drug resistance 1S68⌬-23) and with a mutant virus constructed by site-directed to multiple NRTI has been associated with an amino acid mutagenesis using a pNL4-3 background. HIV-1 with the S68 substitution at the nucleoside-binding site of the enzyme and deletion produced more than 30-fold-increased resistance to with insertions or deletions in the ␤3-␤4 hairpin loop in the DFC, 3TC, emtricitabine [(Ϫ)-FTC], and finger subdomain of the HIV-1 RT (33). fumarate (TDF) and a 5- to 20-fold increase in resistance to Dexelvucitabine (DFC; D-D4FC; Reverset) is a potent NRTI amdoxovir (DAPD, AMDX), abacavir (ABC), and 9-(␤-D-1,3- of HIV-1 that advanced to phase 2, but its clinical development dioxolan-4-yl)guanine (DXG). These data suggest that the was recently terminated due to adverse events in some patients novel S68⌬ may be an important previously unrecognized de- associated with an increase in amylase associated with pancre- letion in the HIV polymerase that could significantly impact treatment regimens containing nucleoside analogs.

* Corresponding author. Mailing address: Veterans Affairs Medical Center, 1670 Clairmont Rd., Medical Research 151H, Decatur, GA MATERIALS AND METHODS 30033. Phone: (404) 728-7711. Fax: (404) 728-7726. E-mail: rschina Cells culture. Primary human peripheral blood mononuclear (PBM) cells were @emory.edu. separated by Ficoll-Hypaque (Histopaque 1077: Sigma, St. Louis, MO) density † Present address: Uppsala University, Department of Medical Sci- gradient centrifugation from buffy coats obtained from the American Red Cross ences, 75185 Uppsala, Sweden. (Atlanta, GA). Buffy coats were derived from healthy seronegative donors. Cells ᰔ Published ahead of print on 28 February 2011. were activated with 3 ␮g of phytohemagglutinin A (PHA; Associated of Cape

2054 VOL. 55, 2011 NOVEL S68 DELETION IN HIV-1 RT 2055

Cod, Inc., East Falmouth, MA) in 500 ml of RPMI 1640 (Mediatech, Inc., to physiological concentration (3.2 mM). In addition, ATP was only used in assay Herndon, VA) containing 20% heat-inactivated fetal bovine serum (HyClone, with virus pellets as a sample, not only to obtain primer unblocking but also to Logan, UT), 83.3 IU of penicillin/ml, 83.3 ␮g of streptomycin/ml, 1.6 mM L-glu- protect degradation of substrate in the crude sample. Furthermore, the deoxy-

tamine (Mediatech), for 2 to 3 days prior to use. HIV-1LAI obtained from the nucleoside triphosphate (dNTP) level (including Brd-UTP) was increased from Center for Disease Control and Prevention (Atlanta, GA) was used as the virus 1to4␮M for assay with the virus pellet samples. Subsequently, in an assay with for selecting a resistant pool. A multiplicity of infection of 0.1 was selected to recombinant purified RT enzyme, only 1 ␮M dNTP (no ATP) was used. begin the infected pool, as determined by a limiting dilution method in human The level of resistance to NRTI-triphosphate (NRTI-TP) by the S68 deletion PBM cells. mutants compared to WT RT was determined as the 50% inhibitory concentra-

Selection of resistant virus. To select resistant virus, naive activated human tion (IC50) based on RT activity. Fold increases in resistance values were deter- ␮ PBM cells were treated with DFC at 0.1 M for 1 h prior to inoculation with mined by dividing the IC50 for the mutant by the IC50 for the respective WT

HIV-1LAI (WT) as previously described (35). Briefly, the vehicle (control) and strain. The RT activity was linear during the assay time within the substrate range DFC-treated human PBM cell groups were infected for 1 h. Supplemented used, and thus steady-state kinetics were assumed. The NRTI-TPs used were RPMI 1640 containing interleukin-2 (26 IU/ml) was then added for a final ZDV-TP, DFC-TP, and (Ϫ)-FTC-TP, all prepared in our laboratory with purity concentration of 106 cells/ml. Virus was passaged every 6 days with a fresh Ͼ96% as determined by liquid chromatography-tandem mass spectrometry. treatment of DFC, ranging from 0.0 to 6.0 ␮M over 52 weeks. The RT activity in Molecular modeling. All models of drug interaction were extrapolated from supernatant was measured weekly and used to determine the percent inhibition X-ray crystallographic atomic coordinates, available to the public at the RCSB by DFC (35). Protein Data Bank (PDB; www.pdb.org) (1). HIV-1 genome isolation and amplification. Total RNA was isolated from Ligand model selection. Four experimental data sets were selected from the culture supernatants by using a commercial QIAamp Viral RNA minikit PDB for reference modeling of the S68⌬ mutation. All selected coordinate sets (Qiagen, Inc., Valencia, CA). RT-PCR was performed using SuperScript III RT contained HIV-RT with terminated and cross-linked DNA (template-primer (Invitrogen, Carlsbad, CA) and DECAprime II primers (Ambion, Austin, TX) to [T/P]) and small molecule nucleotide ligands (4, 18, 43). 1RTD and 1T05 contain generate cDNA from viral RNA. PCR was performed using Platinum Taq DNA the WT RT/DNA complex with the ligands thymidine-5Ј-triphosphate (dTTP) Polymerase High Fidelity (Invitrogen). A 1,346-bp fragment of the HIV-1-RT and tenofovir-diphosphate (TDP), respectively (18, 43). Files 3JYT and 3JSM genome was amplified using the forward primer 5Ј-TTGACTCAGATTGGTTG contain the K65R mutant form of HIV-RT with T/P and 2Ј-deoxyadenosine-5Ј- CACTTTAA-3Ј and the reverse primer 5Ј-AAGAACCCATAGTAGGAGCAG triphosphate (dATP) and TDP solved by the same lab as 1T05 (4). AAAC-3Ј. The PCR product was purified by using a QIAquick PCR purification Bioinformatics: data assembly and sorting. Custom scripting was used to kit (Qiagen), and the fragment was cloned using TOPO cloning (Invitrogen). download and process more than 150 available structures of HIV-RT. The Sequencing of HIV-1 RT amino acids 1 to 300 was performed in parallel between 66-kDa A-chain of the ternary complex Ϫ1RTD was chosen as a structural the control virus and DFC-treated virus to uncover mutations selected from the reference (18). Every other structure was fit to the reference coordinate space applied drug pressure. through a command scripted Chimera interface to the Needleman-Wunsch al- Pyrosequencing. Amplicons generated from supernatant samples collected at gorithm (29). All were fit to the single reference coordinate space by using a weeks 14, 20, 23, and 52 were sent to Research and Testing Laboratory (RTL; BLOSUM-62 matrix, a gap extension penalty of 1, and a secondary structure Lubbock, TX) for pyrosequencing using a 454 Genome Sequences FLX system score of 20% (13). Pairwise RMSD of C-alpha carbons were calculated for (Roche Diagnostic Corp., Indianapolis, IN). Pyrosequencing was performed chains. The data were plotted and used to select relevant related clusters of using standard amplicon protocols according to the manufacturer’s instructions. structures for detailed protein and ligand modeling as described above. Each library was sequenced in both forward and reverse directions. Totals of Fitting and analysis of protein-ligand interactions. All structures were fit to 79,259, 80,594, 95,818 and 74,424 raw sequences were derived from each library, the 1RTD coordinate space as described in Bioinformatics using UCSF Chimera respectively. Sequences were quality screened using a custom script (RTL) and for visualization and protein preparation (30). Ramachandran plots were gen- exported to the FASTA format. erated by using the SwissPDB viewer (31). Ligands were prepared and modeled Drug susceptibility assay. The procedures for the antiviral assays in human in the Molecular Operating Environment (MOE; Chemical Computer Group, PBM cells have been previously published (34). Briefly, uninfected PHA-stimu- Montreal, Quebec, Canada) using the Protonate 3D and ligand interaction

lated human PBM cells were infected with HIV-1LAI. The drugs were then added diagramming algorithms (3, 24). to duplicate or triplicate cultures. Uninfected and untreated human PBM cells Modeling WT and single and double mutations. To reduce bias from manual were grown in parallel at equivalent cell concentrations to control for cell via- loop model building, random-number seeds were applied through the UCSF bility. The cell viability was determined by using a trypan blue exclusion assay. Modeler application to generate five different HIV-1, p66 chain, three-dimen-

The cultures were maintained in a humidified 5% CO2 to 95% air incubator at sional models given identical primary sequence (7). Sequences input for WT, 37°C. The RT activity in supernatant was assessed 6 days after infection when RT S68⌬, K65R, and S68⌬/K65R double mutants were fit against 1RTD, 1T05, and levels are at maximum levels (34, 39). The supernatant was clarified, and the viral 3JSM as templates, yielding 60 theoretical models that were back-fit against the particles were pelleted at 12,000 rpm for2hinaJouan centrifuge (model GR experimental structures for calibration and statistical analysis. Rotamers were 4.1) and suspended in virus-disrupting buffer (1 M Tris-chloride [pH 7.8], 5 M mapped and scored using the Dunbrack rotamer library accessed through Chi- NaCl, 100% glycerol, 100 mM phenylmethylsulfonyl fluoride, 10% Triton X-100, mera and MOE rotamer explorer as described previously (12). and distilled water to produce 1 liter of solution). The RT assay was performed Nucleotide sequence accession number. The sequences obtained in the present

in 96-well microdilution plates using poly-rA-p(T)12-18 as the template primer study have been deposited in GenBank under accession number GQ245682. (39). The RT results were expressed in disintegrations per minute per milliliter of originally clarified supernatant (36). Site-directed mutagenesis for protein expression. Site-directed mutagenesis RESULTS was performed on a hybrid plasmid (designated MC002) created by digesting self-ligated pCR2.1 (Invitrogen) and pNL4-3 (AF3244930) with restriction en- In vitro selection of S68⌬. In the presence of DFC, virus zymes EcoRI and SpeI. The 4.2-kb band from pNL4-3 containing RT was ligated containing the S68 deletion was first detected in supernatant of into pCR2.1 using the cutting sites EcoRI/SpeI. The primers used had the S68 codon AGT removed, thus introducing the deletion into pNL4-3 RT. The dele- HIV-1LAI-infected human PBM cells at week 14 as a mixture tion was confirmed by sequencing in both directions. To ligate the RT of pNL4-3 of S68⌬ and WT viruses (Table 1). Population sequencing, with the deletion at codon 68 into protein expression vector pE60 (Invitrogen), from amino acids 1 to 300 in the HIV-1 RT region, revealed MC002 was amplified and cloned into the NcoI/BglII site by using T4 DNA ligase S68⌬ as the dominant mutation by week 19. It then became a (New England Biolabs). The digested products were gel extracted by using QIAEX II (Qiagen). The ligation was transformed into Alpha-Select competent mixture containing K65R and WT sequences by week 25, and bacteria (Bioline, Taunton, MA). Plasmid DNA was extracted by using a the K65R mutation became dominant by week 28. No other HiSpeed plasmid maxi kit (Qiagen) and sequenced in multiple directions to confirm mutations in the RT region were detected. Adjustment of the the mutation. DFC concentration was at times necessary in order to increase Heteropolymeric DNA polymerase assay. The principle and performance of virus replication. this nonradioactive RT assay has previously been described by Lennerstrand et al. (25). The following modifications were introduced to obtain ATP primer A search in the Stanford University HIV Drug Resistance unblocking reaction in the assay, the ATP (Amersham/GE Healthcare) was set Database (http://hivdb.stanford.edu) did not reveal matching 2056 SCHINAZI ET AL. ANTIMICROB.AGENTS CHEMOTHER.

TABLE 1. Population sequencing from amino acids 1 to 300 in the weeks 14, 20, 23, and 52, detected S68⌬, K65R, and other 64 HIV-1 RT region demonstrates selection of the S68⌬ virus in polymorphisms. Of the detected polymorphisms, an overage of supernatant from DFC-treated human PBM cells 26 occurred at a frequency higher than 3%, and all are non- DFC selection pressure Culture Mutation by population synonymous mutations (Table 3). The populations occurring at ͓␮ ͔ a (concn M ) wk sequencing a frequency greater than 3% or higher are unlikely to represent 0.0 4 WT sequence artifact making this a conservative cutoff for identi- 0.1 10 WT fication of authentic variants (19, 28, 45). Direct comparison of 1.0 11 WT the FLX system versus TOPO cloning revealed a 2- to 10-fold 1.0 12 WT 0.1 14 S68⌬/WT difference in the frequency of virus polymorphisms in samples 0.1 15 WT obtained at weeks 14 and 20 (Table 4). The two sequencing 0.1 16 WT methods showed less difference in the results from the samples ⌬ 1.0 17 S68 /WT obtained at weeks 23 and 52. A higher number of emerging 1.0 18 S68⌬/WT 1.0 19 S68⌬ polymorphisms during early time points could be responsible 0.0* 20 S68⌬ for these finding. The polymorphisms were scored using the 1.0 22 S68⌬ Stanford University HIV Drug Resistance Database, revealing 0.0* 23 S68⌬ that none were associated with known NRTI and NNRTI drug 2.0 24 S68⌬ ⌬ resistance. The FLX system results support the existence of the 3.0 25 S68 /K65R/WT ⌬ 3.0 26 S68⌬/K65R/WT S68 with or without the K65R mutation on the same genome. 3.0 27 S68⌬/K65R/WT In addition, these results provide information related to poly- 6.0 28 K65R morphisms that evolved under DFC treatment and were not 6.0 29 K65R detected by conventional DNA sequencing. 6.0 30 K65R 1.5 36 K65R Susceptibility to RT inhibitors. To determine the suscepti- 6.0 49 K65R bility of the S68⌬ virus against various NRTI and NNRTI, 6.0 52 S68⌬/K65R HIVS68⌬-23 was expanded and tested. HIVS68⌬-23 was sequenced a *, Adjustment of the DFC concentration was necessary in order to increase to ensure the dominant population was the deletion at virus replication, thus providing material for sequence analysis. codon 68. Drug susceptibility assays indicated that virus with the S68 deletion produced a Ͼ30-fold increased resistance to the sequences for the S68⌬ indicating that it is a novel mutation NRTIs DFC, TDF, 3TC, and (Ϫ)-FTC, a 5- to 20-fold in- (Table 2) (2, 10, 20, 21, 42, 44, 22). To confirm the mutant creased resistance to DXG, DAPD, and ABC, and a modest population, the virus obtained from supernatants at weeks 11, resistance (Ͻ5-fold) to ddC, d4T, D-FDOC, or DOT but re- 12, 14, 16, 17, 18, 19, 20, 23, 26, 27, 29, 30, 37, and 52 was mained susceptible to ZDV (Fig. 2). The correlation coeffi- TOPO cloned and sequence analysis was conducted on at least cient for the dose-response curve was higher than 0.9 for all of 10 clones per week (Fig. 1). Similar results were found using the assays (data not shown). As expected, the HIVS68⌬-23 sus- population sequencing (Table 1) and TOPO cloning (Fig. 1). ceptibility to an NNRTI (Sustiva) and protease inhibitor (PI; The mutation K65R appeared at week 14 in 10% of the total lopinavir) was not markedly changed. population and became 100% at week 29. On the other hand, Enzymatic characterization of the recombinant S68 deletion S68⌬ was apparent at week 14 as 70% of the total population, RT. Enzymatic studies with the HIV-1 RT obtained from the and it continued to be detected until week 52, where it ap- recombinant and particle-derived S68⌬ virus in the presence or peared as a mixture in the same genome with K65R (63%). absence of 3.2 mM ATP resulted in a 2.5-fold, 6- to 8-fold, and Interestedly, two clones of 10 at week 23 contained the muta- 8- to 10-fold increases in resistance to ZDV-TP, DFC-TP, and tions S68⌬ in combination with T69A or T69S on the same (Ϫ)-FTC-TP (Fig. 3). As expected, the control M184V enzyme genome. These results demonstrate that S68⌬ virus obtained had a significant Ͼ10-fold increase in resistance to (Ϫ)- from the DFC selection process can occur independently or as FTC-TP but was not resistant to DFC-TP. Neither the S68⌬ a mixture with WT, K65R, T69A, or T69S viruses. nor the M184V recombinant RTs showed significant resistance Pyrosequencing analysis (FLX system) of samples from to ZDV-TP, confirming the results of the cell-based suscepti-

TABLE 2. Alignment of S68⌬ with previously published deletions and/or mutations in codons 63 to 73 of the HIV-1 ␤3-␤4 loop

Deletion and/or Amino acid at position: Virus (reference[s]) mutation(s) 63 64 65 66 67 68 69 70 71 72 73 S68⌬ S68⌬ I KKKD TKWRK AF271766 (2, 10, 21) D67⌬, T69G, K70R I K K K SGRWRK AF311203 (42) K70⌬, S68N IKKKDNT WRK DQ394304 (16) K70⌬, S68G IKKKDGT WRK AF311157 (42) T69⌬, D67S, S68G I K K K S G KWRK EF154395 (44) T69⌬, S68G, K70G I K K K D G GWRK LAI IKKKDSTKWRK pNL4-3 IKKKDSTKWRK VOL. 55, 2011 NOVEL S68 DELETION IN HIV-1 RT 2057

FIG. 1. Frequency of HIV-RT mutations arising from DFC-treated human PBM cells. TOPO cloning and sequence analysis were conducted on at least 10 clones per week. The bar at the top of the figure indicates DFC selection pressure (␮M) by week.

bility assays described above. Both enzymatic assay and recom- S68⌬ and S68⌬/K65R gave lowest overall RMSD from the binant RT assay demonstrated similar fold increases in resis- experimental references based upon the K65R geometry. tance with or without ATP. Thus, these data demonstrated that ⌬ S68 resulted in resistance to several clinically important DISCUSSION nucleoside analogs but remained susceptible to ZDV, thus supporting the previously published cell-based studies (6, 9, Deletions occur Ͻ0.2% within the ␤3-␤4 hairpin loop-cod- 15, 40). ing region of HIV-1-infected subjects treated with NRTI, and Computational analysis of S68 and K65 structural rela- they appear most likely to emerge in viruses from subjects with tions. Comparative analyses of WT and K65R RT experi- extensive histories of antiretroviral therapy (27, 33, 42). These mental complexes suggest a potential mechanism of drug deletions in the HIV-1 RT have been associated with resis- interaction and resistance that is structurally mediated tance to NRTI when they appear in combination with other through residues 68 and 69. Ramachandran plotting of peptide mutations in the RT-coding region. For instance, deletions backbone angles for the WT/TTP and WT/TDP enzyme/ligand that occur in the RT region at codons 67 and 70 have been complexes (1RTD and 1T05) highlight S68 as the most known to occur in combination with mutation T69G or strained residue in phi/psi space, followed by M184 and T69 in Q151M (16, 21). both (data not shown). The same type of plot comparing WT/ In the present study, S68⌬ was found to be independent of TDP (1T05) to the K65R/TDP complex, published recently by any other reported mutations and/or deletions in the HIV-1 the same lab (3JSM) (4), reveals a marked movement of S68 to ␤3-␤4 loop (Table 2). Sequences used for comparison to S68⌬ the Ramachandran favored region accompanying the K65R were generated from clinical samples that have undergone mutation (Fig. 4a and c). Molecular representations of the multiple drug treatments for HIV-1 (2, 10, 20, 21, 42, 44). The experimental structures illustrate the S68 residue in an ex- presence of a novel deletion in position 68 of the HIV-1 RT tended solvent exposed geometry in the WT to one less genome was selected when the infected cells were exposed to strained in the K65 mutant (Fig. 4b and d). DFC. The presence of the S68⌬ was confirmed in vitro using While theoretical modeling of side-chain mutations is often TOPO cloning and pyrosequencing. extended from existing protein structural backbones and dis- These data establish, for the first time, that S68⌬ can be cussed in terms of ligand interactions, deletion or insertion detected in vitro under single drug pressure. In contrast to modeling requires independent analysis of flexible protein previous reports where an S68G mutation appeared in subjects space. The UCSF Modeler software was used to test possible loop conformations available to an S68⌬ mutation with or without an accompanying K65R mutation (7). The models for TABLE 4. Comparison of the population frequencies of S68⌬ and K65R obtained with TOPO cloning and pyrosequencinga

Population frequency (%) at: Analysis TABLE 3. Frequency of polymorphisms detected in supernatant Wk 14 Wk 20 Wk 23 Wk 52 from DFC-treated human PBM cells using S68⌬ TOPO cloning 70.0 86.0 94.0 63.0 ultradeep pyrosequencing S68⌬ pyrosequencing 6.0 44.9 96.0 92.5 Total no. of polymorphisms K65R TOPO cloning 5.0 7.0 0 88.0 Wk K65R pyrosequencing 0.1 0.1 1.6 99.5 Ͼ0.5% Ͼ3% Ͻ3% Ͻ0.5% a Direct comparison revealed 2- to 10-fold differences in the population fre- 14 72 35 37 47 quencies for virus obtained on weeks 14 and 20 and slight difference by weeks 23 20 60 22 38 41 and 52. These differences may be due to an increase in the percentage of new 23 64 20 44 37 polymorphisms in the early weeks and more error-prone (5 to 10 errors/kb) than 52 62 26 36 44 those obtained from Sanger sequencing (0.01 errors/kb). The K65R and S68⌬ haplotypes were as previously reported (19, 46). 2058 SCHINAZI ET AL. ANTIMICROB.AGENTS CHEMOTHER.

FIG. 2. Fold increase in drug resistance of HIV-1LAI week 23 against various nucleosides in human PBM cells. All experiments were performed in replicates. Viral inhibition was measured by using a FIG. 3. Effect of NTP analogs on recombinant and particle-derived [3H]TTP RT incorporation assay from cell supernatants. High resis- RT containing either S68⌬ or M184V. Black bar, virally derived S68⌬; tance levels were determined by a fold increase of Ͼ5. The S68⌬ dark gray bar, recombinant S68⌬; light gray bar, virally derived developed a Ͼ30-fold increase resistance to DFC, 3TC, (Ϫ)-FTC, M184V. Note that M184V has a Ͼ10-fold increase in resistance to TDF, ABC, and DAPD. The fold increase is calculated as EC HIV- 50 (Ϫ)-FTC-TP. The EC values are averages from three independent 1S68⌬/EC HIV-1 WT (FI ) and EC HIV-1S68⌬/EC HIV-1 WT 50 50 50 90 90 experiments. (FI90). Light gray bar, FI90; dark gray bar, FI50.

that already had K65R (32), the present study demonstrates cating that the S68⌬ in the HIV-1 genome has potential clinical that the S68⌬ can appear independent of and before K65R relevance and should be monitored in HIV-infected persons. (Table 1). The mechanism by which the S68⌬ virus confers a Early structural studies of resistance mutations effecting nu- decrease in DFC efficiency is largely unknown. According to cleoside analogs relied upon atomic coordinates of HIV-RT the enzymatic studies, S68⌬ RT showed similar resistance to derived from one of two relevant experimental structures, NRTI-TP (ZDV-TP, FTC-TP, and DFC-TP) with or without 1RTD and 1T05 (18, 43). These contain the ternary WT pro- ATP. These data suggest that the mechanism of S68⌬ resis- tein complexed with template-primer and phosphorylated 2Ј- tance is ATP independent and most likely occurs by enhancing deoxynucleotide dNTP and TDP, respectively (18, 43). Our substrate discrimination. study benefits from two recent crystallographic sets of the The serine deletion in position 68 is located in the fingers resistant mutant K65R-RT ternary complex with TDP and subdomain of the HIV-1 RT, between residues that make dATP as ligands (4). Evaluation of protein geometries across important contacts with the incoming dNTP. This region helps these four experimental structures reveals a structural anomaly position, bind, and activate the incoming dNTP for catalytic at S68/D69 (Fig. 4). The S68/D69 protein backbone is highly incorporation into the growing DNA chain during polymeriza- strained in the WT structures, which appears pinned in that tion (18). Our models suggest S68⌬ alone or in combination conformation by the tight pairing of charge-localized K65 to with K65R could impair the dNTP incorporation through re- beta/gamma phosphate oxygens (18, 43). positioning of charge groups above the plane of catalysis de- To minimize bias from manual model building, the UCSF fined by the bound triphosphate. As a consequence, virus rep- Modeler algorithm was used to build more than 60 models of lication rate could decrease overall, with selective resistance to WT, S68⌬, K65R, and the S68⌬/K65R double-mutant models certain chemically related nucleoside analogs. for further analysis (see Materials and Methods for details). Mutations at position 65 in HIV RT are known to influence According to our models, resistance to S68⌬/K65R may the nucleotide biding specificity of the enzyme (4, 8, 47). Fur- be explained through a mechanism repositioning catalytic thermore, a new resistance profile, which includes K65R with residues through loop dynamics. The least-strained loop con- S68G, was detected in persons failing ABC, ddI, and d4T formation formed by deletion of S68 was found to be closer to treatment. The authors of those studies suggested that the K65R than WT structures. These findings suggest that S68⌬ presence of the S68G mutation might increase the replicative and K65R may exist compatibly on the same genome, but we capacity of the mutant virus K65R (32, 41). Thus, these find- expected that the double mutation may be even less fit than ings underscore the potential relevance of the S68⌬ on virus single mutations. Since both S68⌬ and K65R effect binding replication, which developed resistance not only against DFC “above” the catalytic plane defined by the incoming triphos- but also against other important NRTIs involved in the HIV-1 phate, substrates with substitutions below the plane, such as therapy. natural 3Ј-OH, azido would be preferred for incorporation According to previous reports, mutations that occur in the over ddNTP, nonsubstituted L-nucleoside, or acyclic analogs. HIV-1 RT region between amino acids 62 and 78 significantly The spring-like strain placed in loop ␤3-␤4 by residue S68 in increase NRTI resistance and restore the replication defect relation to placement of K65 may increase processivity and associated with the K65R mutation (16, 41). After evaluation promiscuity of substrates in WT HIV-1RT versus either of HIVS68⌬-23 against 13 NRTIs, it was found that this virus mutant. Conversely, the rearranged backbone conformation Ͼ ⌬ developed 5-fold increased resistance (FI50) in the presence needed to accommodate either S68 or K65R substitutions of DFC, DAPD, 3TC, (Ϫ)-FTC, TDF, DXG, and ABC, indi- diminish loop dynamicity and promote side chain placements VOL. 55, 2011 NOVEL S68 DELETION IN HIV-1 RT 2059

FIG. 4. Experimental structures reveal S68 loop dynamics. (a) Ramachandran plot of three-dimensional backbone phi/psi angles of the WT K65 HIV-1 RT 66-kDa subunit (554 amino acids) bound to TDP (PDB-1t05). Green regions are most favored, blue regions are tolerated, and white space is predicted as strained. S68 on loop ␤3-␤4 is the most strained amino acid in this experimental data set (43). Five key loop residues involved in enzyme function, including S68, are depicted in panel b. (b) Active sites from PDB-1t05. Five ␤3-␤4 loop residues show side chains of strained S68. (c) Plot of the mutant R65 HIV-1 RT 66-kDa subunit (554 amino acids) bound to TDP (PDB-3jsm) shows contrasting less-strained phi/psi conformations of S68 and T69. (d) Active site of R65 mutant from PDB-3jsm (4). Rotation of the peptides position the S68 side chain in the pocket and brings K70 into contact with the gamma phosphate of the ligand. that increase substrate selectivity while reducing overall en- REFERENCES zyme efficiency. 1. Berman, H. M., et al. 2000. The protein data bank. Nucleic Acids Res. In conclusion, the present study reveals the significance of a 28:235–242. ⌬ 2. Boyer, P. L., T. Imamichi, S. G. Sarafianos, E. Arnold, and S. H. Hughes. novel deletion (S68 ) in the hairpin loop of HIV-1 RT that 2004. Effects of the Delta67 complex of mutations in human immunodefi- may have potential clinical relevance when it occurs alone or in ciency virus type 1 reverse transcriptase on nucleoside analog excision. J. Vi- combination with the K65R mutation. rol. 78:9987–9997. 3. Clark, A. M., and P. Labute. 2007. 2D depiction of protein-ligand complexes. J. Chem. Infect. Model 47:1933–1944. 4. Das, K., et al. 2009. Structural basis for the role of the K65R mutation in ACKNOWLEDGMENTS HIV-1 reverse transcriptase polymerization, excision antagonism, and teno- fovir resistance. J. Biol. Chem. 284:35092–35100. This study was supported in part by Public Health Service grants 5. Deval, J., et al. 2004. Mechanistic basis for reduced viral and enzymatic 5R37-AI-0419801 and 5P30-AI-50409 (CFAR) and by the Department fitness of HIV-1 reverse transcriptase containing both K65R and M184V of Veterans Affairs. Computational modeling was performed with as- mutations. J. Biol. Chem. 279:509–516. sistance of the Emory School of Medicine Biomolecular Computing 6. Erickson-Viitanen, S., et al. 2003. Cellular pharmacology of D-d4FC, a Resource (BIMCORE). Hardware and software was supported in part active against drug-resistant HIV. Antivir. Chem. Che- through an Academic Excellence Grant from Sun Microsystems and an mother. 14:39–47. Academic Research Award from Accelrys Corp. 7. Eswar, N., et al. 2006. Comparative protein structure modeling using Mod- eller. Curr. Protoc. Bioinformatics Chapter 5:Unit 5.6. R.F.S. is the inventor of DFC, (Ϫ)-FTC, and 3TC. The royalties for Ϫ 8. Garforth, S. J., T. W. Kim, M. A. Parniak, E. T. Kool, and V. R. Prasad. 2007. ( )-FTC were sold to Gilead Sciences and Royalty Pharma in 2005. Site-directed mutagenesis in the fingers subdomain of HIV-1 reverse tran- However, R.F.S. continues to receive royalties from the sales of 3TC scriptase reveals a specific role for the ␤3-␤4 hairpin loop in dNTP selection. until 2011. R.F.S. is also a major shareholder in RFS Pharma, LLC, a J. Mol. Biol. 365:38–49. company that currently develops Amdoxovir for HIV-1 infections. 9. Geleziunas, R., et al. 2003. HIV-1 resistance profile of the novel nucleoside 2060 SCHINAZI ET AL. ANTIMICROB.AGENTS CHEMOTHER.

reverse transcriptase inhibitor ␤-D-2Ј,3Ј-dideoxy-2Ј,3Ј-didehydro-5-fluoro- variants in antiretroviral-naive persons with reverse transcriptase codon 215 cytidine (Reverset). Antivir. Chem. Chemother. 14:49–59. revertant mutations. J. Virol. 82:10747–10755. 10. Giri, J., H. J. Rueda, A. Monticelli, and N. Planes. 2000. Case report of a 29. Needlema, S. B., and C. D. Wunsch. 1970. A general method applicable to novel amino acid deletion in codon 67 and T69G substitution in the reverse search for similarities in amino acid sequence of two proteins. J. Mol. Biol. transcriptase of HIV-1. Antivir. Ther. 5:227–228. 48:443–453. 11. Hammond, J. L., et al. 2005. In vitro selection and analysis of human 30. Pettersen, E. F., et al. 2004. UCSF chimera: a visualization system for immunodeficiency virus type 1 resistant to derivatives of ␤-2Ј,3Ј-di- exploratory research and analysis. J. Comput. Chem. 25:1605–1612. dehydro-2Ј,3Ј-dideoxy-5-fluorocytidine. Antimicrob. Agents Chemother. 31. Ramachandran, G. N., C. Ramakrishnan, and V. Sasisekharan. 1963. 49:3930–3932. Stereochemistry of polypeptide chain configurations. J. Mol. Biol. 7:95–99. 12. Hari, M., et al. 2006. Paclitaxel-resistant cells have a mutation in the pacli- 32. Roge, B. T., et al. 2003. K65R with or without S68: a new resistance profile taxel-binding region of beta-tubulin (Asp26Glu) and less stable microtu- in vivo detected in most patients failing abacavir, , and stavudine. bules. Mol. Cancer Ther. 5:270–278. Antivir. Ther. 8:173–182. 13. Henikoff, S., and J. G. Henikoff. 1992. Amino acid substitution matrices from 33. Ross, L., et al. 2000. Deletions in the ␤3-␤4 hairpin loop of HIV-1 reverse protein blocks. Proc. Natl. Acad. Sci. U. S. A. 89:10915–10919. transcriptase are observed in HIV-1 isolated from subjects during long-term 14. Hernandez-Santiago, B. I., J. S. Mathew, K. L. Rapp, J. P. Grier, and R. F. antiretroviral therapy. J. Hum. Virol. 3:144–149. Schinazi. 2007. Antiviral and cellular metabolism interactions between dex- 34. Schinazi, R. F., D. L. Cannon, B. H. Arnold, and D. Martino-Saltzman. 1988. and lamivudine. Antimicrob. Agents Chemother. 51:2130–2135. Combinations of isoprinosine and 3Ј-azido-3Ј-deoxythymidine in lympho- 15. Herzmann, C., et al. 2005. Safety, pharmacokinetics, and efficacy of (Ϯ)-␤- cytes infected with human immunodeficiency virus type 1. Antimicrob. 2Ј,3Ј-dideoxy-5-fluoro-3Ј-thiacytidine with and stavudine in anti- Agents Chemother. 32:1784–1787. retroviral-naive human immunodeficiency virus-infected patients. Antimi- 35. Schinazi, R. F., et al. 1993. Characterization of human immunodeficiency crob. Agents Chemother. 49:2828–2833. viruses resistant to oxathiolane-cytosine nucleosides. Antimicrob. Agents 16. Hu, Z., et al. 2007. Virologic characterization of HIV type 1 with a codon 70 Chemother. 37:875–881. deletion in reverse transcriptase. J. Acquir. Immune Defic. Syndr. 45:494– 36. Schinazi, R. F., et al. 1990. Activities of 3Ј-azido-3Ј-deoxythymidine nucle- 500. otide dimers in primary lymphocytes infected with human immunodeficiency 17. Huang, G., et al. 2004. Molecular basis for a dominant inactivation of virus type 1. Antimicrob. Agents Chemother. 34:1061–1067. RUNX1/AML1 by the leukemogenic inversion 16 chimera. Blood 103:3200– 37. Sharma, P. L., J. H. Nettles, A. Feldman, K. Rapp, and R. F. Schinazi. 2009. 3207. Comparative analysis of in vitro processivity of HIV-1 reverse transcriptases 18. Huang, H., R. Chopra, G. L. Verdine, and S. C. Harrison. 1998. Structure of containing mutations 65R, 74V, 184V and 65Rϩ74V. Antivir. Res. 83:317– a covalently trapped catalytic complex of HIV-1 reverse transcriptase: im- 323. plications for drug resistance. Science 282:1669–1675. 38. Sluis-Cremer, N., D. Arion, N. Kaushik, H. Lim, and M. A. Parniak. 2000. 19. Huse, S. M., J. A. Huber, H. G. Morrison, M. L. Sogin, and D. M. Welch. Mutational analysis of Lys65 of HIV-1 reverse transcriptase. Biochem. J. 2007. Accuracy and quality of massively parallel DNA pyrosequencing. Ge- 348(Pt. 1):77–82. nome Biol. 8:R143. 39. Spira, T. J., et al. 1987. Micromethod for assaying reverse transcriptase of 20. Imamichi, T., et al. 2000. Relative replication fitness of a high-level 3Ј-azido- human T-cell lymphotropic virus type III/lymphadenopathy-associated virus. 3Ј-deoxythymidine-resistant variant of human immunodeficiency virus type 1 J. Clin. Microbiol. 25:97–99. possessing an amino acid deletion at codon 67 and a novel substitution 40. Stuyver, L. J., et al. 2004. Potent antiviral effect of reverset in HIV-1-infected (Thr3Gly) at codon 69. J. Virol. 74:10958–10964. adults following a single oral dose. Antivir. Ther. 9:529–536. 21. Imamichi, T., M. A. Murphy, H. Imamichi, and H. C. Lane. 2001. Amino 41. Svarovskaia, E. S., et al. 2008. The A62V and S68G mutations in HIV-1 acid deletion at codon 67 and Thr-to-Gly change at codon 69 of human reverse transcriptase partially restore the replication defect associated with immunodeficiency virus type 1 reverse transcriptase confer novel drug resis- the K65R mutation. J. Acquir. Immune Defic. Syndr. 48:428–436. tance profiles. J. Virol. 75:3988–3992. 42. Tamalet, C., et al. 2000. Multidrug resistance genotypes (insertions in the 22. Imamichi, T., et al. 2000. High-level resistance to 3Ј-azido-3Ј-deoxythymi- ␤3-␤4 finger subdomain and MDR mutations) of HIV-1 reverse transcrip- dine due to a deletion in the reverse transcriptase gene of human immuno- tase from extensively treated patients: incidence and association with other deficiency virus type 1. J. Virol. 74:1023–1028. resistance mutations. Virology 270:310–316. 23. Johnson, V. A., et al. 2008. Update of the drug resistance mutations in 43. Tuske, S., et al. 2004. Structures of HIV-1 RT-DNA complexes before and HIV-1. Top. HIV Med. 16:138–145. after incorporation of the anti-AIDS drug tenofovir. Nat. Struct. Mol. Biol. 24. Labute, P. 2008. The generalized Born/volume integral implicit solvent 11:469–474. model: estimation of the free energy of hydration using London dispersion 44. Villena, C., et al. 2007. Relative fitness and replication capacity of a multi- instead of atomic surface area. J. Comput. Chem. 29:1693–1698. nucleoside analogue-resistant clinical human immunodeficiency virus type 1 25. Lennerstrand, J., C. K. Chu, and R. F. Schinazi. 2007. Biochemical studies isolate with a deletion of codon 69 in the reverse transcriptase coding region. on the mechanism of human immunodeficiency virus type 1 reverse tran- J. Virol. 81:4713–4721. scriptase resistance to 1-(␤-D-dioxolane)thymine triphosphate. Antimicrob. 45. Wang, C., Y. Mitsuya, B. Gharizadeh, M. Ronaghi, and R. W. Shafer. 2007. Agents Chemother. 51:2078–2084. Characterization of mutation spectra with ultra-deep pyrosequencing: appli- 26. Menendez-Arias, L. 2008. Mechanisms of resistance to nucleoside analogue cation to HIV-1 drug resistance. Genome Res. 17:1195–1201. inhibitors of HIV-1 reverse transcriptase. Virus Res. 134:124–146. 46. Waters, L., and M. R. Nelson. 2005. New drugs. J. HIV Ther. 10:16–18. 27. Menendez-Arias, L., T. Matamoros, and C. E. Cases-Gonzalez. 2006. Inser- 47. White, K. L., et al. 2002. Molecular mechanisms of resistance to human tions and deletions in HIV-1 reverse transcriptase: consequences for drug immunodeficiency virus type 1 with reverse transcriptase mutations K65R resistance and viral fitness. Curr. Pharm. Des. 12:1811–1825. and K65RϩM184V and their effects on enzyme function and viral replica- 28. Mitsuya, Y., et al. 2008. Minority human immunodeficiency virus type 1 tion capacity. Antimicrob. Agents Chemother. 46:3437–3446.