5201_a12_p1010-1014 9/16/04 9:20 AM Page 1010
AIDS RESEARCH AND HUMAN RETROVIRUSES Volume 20, Number 9, 2004, pp. 1010–1014 © Mary Ann Liebert, Inc.
Sequence Note
Potential Drug Resistance Polymorphisms in the Integrase Gene of HIV Type 1 Subtype A
BELINDA L. HERRING, 1,* ANTHONY L. CUNNINGHAM, 1 and DOMINIC E. DWYER 1,2
ABSTRACT
Variation in HIV-1 genes within and between subtypes has been best defined in the env gene, however, other more conserved genes vary between subtypes. Integrase (IN) and other regions of the pol gene are highly con- served due to their integral role in HIV replication and therefore are targets for antiviral drugs. In this study 3 individuals, infected heterosexually with HIV-1 subtype A, were examined for IN polymorphisms. Two pa - tients’ sequences clustered phylogenetically with other subtype A sequences and one patient’s sequence was most similar to the circulating recombinant form CRF_02. No polymorphisms were observed in either of the motifs containing residues critical residues for IN activity. Polymorphisms were observed in a residue asso -
ciated with resistance to anti-integrase drugs. In addition, a number of unique polymorphisms were observed in one individual (WM1666). IN can vary significantly within a subtype as well as between subtypes, and mu - tations associated with resistance to anti-integrase compounds can be present in drug naive individuals.
NTEGRATION OF RETROVIRAL DNA into the host cell genome A small number of studies have investigated the presence Iis an essential step in the retroviral life cycle. Integration is and frequency of polymorphisms in IN from wild-type HIV-1 facilitated by integrase (IN), a 51-kDa protein encoded by the isolates, however, they have concentrated predominantly on pol gene. The integrase gene contains three structural/functional subtype B isolates. 14,17,18 Few data are available about natu - domains, two of which contain specific amino acid motifs es - rally occurring sequence polymorphisms among nonsubtype B sential for enzyme activity or viral infectivity, the HHCC and HIV-1 strains, including those that may contribute to resistance the DDE motifs.1–6 Mutagenesis studies have provided critical to new antiintegrase compounds. Studies of HIV-1 protease information on the role of the conserved motifs in IN with re - demonstrate that mutations associated with protease inhibitor spect to infectivity and activity. Consequently, mutation of any resistance occur as natural polymorphisms in protease inhibi - of the amino acids mentioned abolishes integrase activity and tor-naive patients, and may be more frequent in certain sub - therefore viral infectivity and replication. types (e.g., M36I in subtype C).19,20 In this study we examined Integrase is under intense investigation as an antiviral drug eight integrase sequences from three HIV-1 subtype A-infected target since it is essential for HIV-1 replication, is highly con - individuals for the presence of naturally occurring polymor - served, and there is no cellular equivalent. 7 A number of inte - phisms. grase inhibitors have been identified including the diketoacids Cryopreserved peripheral blood mononuclear cells (PBMC) (DKA) L-708,906, L-731,988, and S-1360 and a new class of or primary cocultured cells from three HIV-1 subtype A pa - inhibitors the 5 H-pyrano[2,3-d:-6,5-dЈ]dipyrimidines (PDPs) tients were selected from the Westmead Hospital repository. 21 V-165.8–15 The most promising of these compounds, the DKAs, HIV-1 infection in all cases was acquired by heterosexual con - are in phase I/II clinical trails. 7,16 tact in Africa (Table 1). DNA was extracted and HIV-1 sub -
1Center for Virus Research, Westmead Millennium Institute, Westmead Hospital, Westmead, NSW 2145, Australia. 2Department of Virology, Centre for Infectious Diseases and Microbiology Laboratory Services, ICPMR, Westmead Hospital, Westmead, NSW 2145, Australia. *Present address: Department of Medicine, University of California, San Francisco, San Francisco, California 94118.
1010 5201_a12_p1010-1014 9/16/04 9:20 AM Page 1011
POLYMORPHISMS IN INTEGRASE GENE 1011
TABLE 1. PATIENT CHARACTERISTICS
Subtype Origin of Sample ID Sex Transmission HIV infection Env IN
WM1013 M Heterosexual South Africa A CRF02_AG WM1408 M Heterosexual Uganda A A WM1666a F Heterosexual Africa A A
aSix clones (four clones from PBMC and two from plasma).
type was initially determined using env heteroduplex mobility assay (HMA) as previously described. 21 The entire coding region of IN was amplified from each pa - tient sample using nested polymerase chain reaction (PCR). Briefly, 10 l of extracted DNA was added to a 50 l PCR containing 10 pmol each of primers IN12, 5 Ј GCAG- GATTCGGGATTAGAAG 3Ј (position 4007–4026 on HXB2), and IN13, 5Ј CTTTCTCCTGTATGCAGACC 3Ј (5251–5270). Reaction conditions included three cycles of 94°C 1 min, 55°C 1 min, 72°C 1 min followed by 32 cycles of 94°C 15 sec, 57°C 45 sec, 72 °C 1 min with a final extension for 5 min at 72 °C. Second round reactions were initiated with 2 l from each first- round reaction and contained 10 pmol each of IN1, 5 Ј AAG- GTCTATCTGGCATGGGTA 3 Ј (4137–4157) and BH4, 5 Ј TCCCCTAGTGGGATGTGTACTTC 3Ј (5200–5222). Identi- cal cycling conditions were used for both rounds of amplifica - tion.
Six clones were generated for patient 1666, four from provi- ral DNA and two from plasma RNA. Population sequences were generated for WM1408 and WM1013. PCR products or posi - tive clones were sequenced using overlapping oligonucleotide primers [sense primers BH6, 5 Ј CCTGGTAGCAGTTCATG- TAG 3 Ј (4448–4467), BH7, 5 Ј GCAGTACAAATGGCAGT- ATTCATCC 3 Ј (4752–4776), and antisense primers BH8, 5 Ј CCGTAGCACCGGTGAAATTGC 3 Ј (4585–4605), BH9, 5 Ј GCTGTCCCTGTAATAAACCCG 3Ј (4899–4919)]. Sequenc- ing reactions were performed using the Big Dye Terminator Cycle Sequencing Ready Reaction DNA Sequencing Kit (Ap - plied Biosystems, Foster City, CA) according to the manufac - turer’s instructions. Sequencing reactions were run on a 373 Stretch DNA Sequencer (ABI Foster City, CA) and assembled using Sequencher (GeneCodes, Ann Arbor, MI). Nucleotide and peptide alignments were generated using CLUSTALW.22 Sequences were aligned with consensus se - quences obtained from the Los Alamos HIV Sequence Data - base (http://hiv-web.lanl.gov) representing M group subtypes, FIG. 1. Phylogenetic tree constructed using maximum likeli- including A1, A2, and circulating recombinant forms (CRFs) hood analysis (PAUP) of WM integrase sequences (864 nt) and containing subtype A portions in pol and one subtype O se - representatives from subtypes A1, A2, B, C, F1, F2, CRF01_AE, quence (Fig. 1). Phylogenetic analysis was performed using and CRF02_AG. All consensus sequences were obtained from PAUP 4.0.23 A Maximum Likelihood tree was constructed us - the Los Alamos HIV Sequence Database. Accession numbers ing the HKY model as selected using Modeltest3.06. 24 Base for representative sequences from subtypes A1, A2 and frequencies, gamma distribution, and transition/transversion ra- CRF01_AE and CRF02_AG are as follows: 97CDKTB48, AF286238; 94CY017-41, AF286237; 97CDKS10, AF286241; tio values were determined by Modeltest. One hundred boot - 97CDKFE4, AF286240; ZAM184, U86780; UG455, M62320; strap replicates were performed on the phylogenetic tree. Phy - UG037, U51190; MSA4079, AF457086; SE8538, AF069669; logenetic analysis of nucleotide sequences (867 nt in length) SE8131, AF107771; SE7253, AF069670; NKU3005, AF457089; demonstrated that all WM1666 sequences and WM1408 clus - Q23-17, AF004885; 97TZ02, AF361872; SE7535, AF069671; tered with other subtype A sequences including CRF01_AE CM240, U54771; 95TNIH047, AB032741; IbNG, L39106. (subtype A in pol). WM1013 clustered with CRF02_AG. Phy - Bootstrap values greater than 60% are indicated. 5201_a12_p1010-1014 9/16/04 9:20 AM Page 1012
1012 HERRING ET AL.
logenetic relationships were supported by significant bootstrap quence (2.15%) and showed greater similarity to subtype A and values (Ͼ75%). A consensus nucleotide sequence was gener - subtype A1 sequences than to the WM1666 clones. Interest - ated for WM1666 for further analysis. Recombination analysis ingly, WM1013 was most similar to CRF02_AG (2.93%) than was performed on all three sequences using SimPlot. 25 Re- it was to other subtype A sequences. combination was not observed in WM1408, and WM1013 was The deduced amino acid sequences (288 amino acids) of IN most similar to the CRF_02 sequence, which is itself a recom - genes were aligned using ClustalW 22 to A1 and A2 represen - binant of subtypes G, A, and U in IN. Similarity plots indicated tatives, UG455 and CDKS10, respectively, a subtype A/A1/A2 that WM1666 was most similar to subtype A consensus and a subtype B consensus sequence (Los Alamos HIV Se - throughout the majority of IN although a small portion of the quence Database), and CRFs containing A-like sequences in gene was more similar to subtype D (530–650 nt). Neverthe - env and pol. No mutations were observed in the residues of ei - less, this relationship was not supported by bootscanning or vi- ther the HHCC or the DDE motifs in the subtype A sequences sual analysis of alignments (data not shown). examined (Fig. 2). However, mutations were observed in Genetic distances were calculated between the subtype A se- WM1666 at residues flanking the HHCC motif (E11D and quences; a consensus sequence derived from the six WM1666 R14K). Amino acid sequences from isolate WM1666 also con- clones and a population sequence from WM1013 and WM1408, tained a number of unique amino acid substitutions (n ϭ 6) that and subtype A reference sequences and a consensus A sequence. were not consistent with consensus sequences from subtype A, The average intrapatient genetic distance for 1666 clones was A1, A2, other HIV-1 subtypes, or CRF sequences (Fig. 2). 1.95% (range 0.47–2.83%). One WM1666 clone contained an WM1013 contained amino acid substitutions consistent with in-frame 9-bp deletion, which increased the distance of this those observed in the CRF02 representative IbNG. Clustering clone to all others to 2.83%. Clones from WM1666 were on of mutations, in particular functional domains, was not ob - average more similar to the subtype A consensus (4.04%) than served, although a greater number of mutations were observed to representatives from subtype A1 (7.16%) and A2 (7.93%). in the catalytic core than the N-terminal or C-terminal domains. WM1408 was most similar to the subtype A consensus se - The number of nonsynonymous mutations observed for all
FIG. 2. Amino acid alignment of WM integrase sequences with a consensus A/A1/A2 sequence and a representative from sub - types A1, A2, B, and CRFs containing subtypes A and/or U in IN. Accession numbers for sequences are as follows: UG455; M62320, CDKS10; AF286241, CM240; U54771, IbNG; L39106, CY032; AF049337, VI1310; AF193253, BFP90; AF064699, GR17; AF179368, MU2079; AF516184. Consensus sequences were obtained from the Los Alamos HIV Sequence Database. Conserved motifs are shaded, dark and light gray for the HHCC and DDE motifs, respectively. The residues boxed indicate the position of the three amino acid deletion in one WM1666 sequence. Positions that agree with the A/A1/A2 consensus are indi - cated by “.” Arrows indicate positions where resistance mutations to the compounds currently in clinical trials have been de - scribed. 5201_a12_p1010-1014 9/16/04 9:20 AM Page 1013
POLYMORPHISMS IN INTEGRASE GENE 1013
WM1666 clones was greater than the number of synonymous SEQUENCE DATA mutations when each cloned sequence was compared to the A/A1/A2 consensus sequence (Los Alamos HIV-1 Sequence Integrase sequences described in the manuscript were Database); the average ds/dn ratio was 3.3. In comparison, the deposited in GenBank, accession numbers AF407637 to ds/dn ratios for WM1013 and WM1408 when compared to the AF407642, AY580066, and AY580067. same consensus were 10.68 and 2.32, respectively. However, when WM1013 was compared to the consensus CRF02_AG se- quence the ds/dn ratio was 7.54. ACKNOWLEDGMENTS HIV-1 survives in part by changing its genetic makeup through mutation to avoid survival challenges such as the host’s This project was supported by Commonwealth AIDS Re - immune mechanisms and antiretroviral drugs. However, viral search Grant 956071. change must be tempered with viral replicative integrity so that functionality of viral proteins is maintained. Integrase, along with the other pol genes, is essential for HIV-1 replication and REFERENCES is in general highly conserved. Sequence variation in RT and protease, resulting in the presence of resistance mutations in 1. Wiskerchen M and Muesing MA: Human immunodeficiency virus treatment-naive individuals, both within and between subtypes, type 1 integrase: Effects of mutations on viral ability to integrate, has serious implications for the successful use of antiretroviral direct viral gene expression from unintegrated viral DNA tem - drugs.19,20 Although the focus to date has been on sequence plates, and sustain viral propagation in primary cells. J Virol variation in RT and protease, a small number of studies have 1995;69(1):376–386. examined the dynamics of sequence variation in IN both within 2. Gerton JL, Ohgi S, Olsen M, DeRisi J, and Brown PO: Effects of and between clades. 14,17–20,26 mutations in residues near the active site of human immunodefi - Here we have described three integrase sequences, two sub - ciency virus type 1 integrase on specific enzyme-substrate inter - actions. J Virol 1998;72(6):5046–5055. type A and one CRF02_AG sequence, each containing IN 3. Cannon PM, Wilson W, Byles E, Kingsman SM, and Kingsman polymorphisms. Two recent studies examined HIV-1 integrase AJ: Human immunodeficiency virus type 1 integrase: Effect on vi- sequences from other clades and demonstrated that IN differs ral replication of mutations at highly conserved residues. J Virol 15,27,28 between subtypes. Recombination has been described 1994;68(8):4768–4775. within the integrase gene, 27 however, it was not observed 4. Engelman A and Craigie R: Identification of conserved amino acid among the sequences examined. residues critical for human immunodeficiency virus type 1 inte - grase function in vitro. J Virol 1992;66(11):6361–6369. A number of antiintegrase compounds, predominantly the DKAs, have been developed and at least one (S1360) is in Phase 5. Engelman A, Englund G, Orenstein JM, Martin MA, and Craigie II clinical trials. 7,16 A number of in vitro-induced resistance mu- R: Multiple effects of mutations in human immunodeficiency virus tations to DKAs affect amino acid positions in both the N-termi - type 1 integrase on viral replication. J Virol 1995;69(5):2729–2736. 6. Asante-Appiah E and Skalka AM: Molecular mechanisms in retro- nal domain and the catalytic core of IN. 12,29 However, the pres - virus DNA integration. Antiviral Res 1997;36(3):139–156. ence of naturally occurring resistance mutations to these drugs has 7. Dayam R and Neamati N: Small-molecule HIV-1 integrase inhib - only briefly been addressed. A recent study examined in vitro sus- itors: The 2001–2002 update. Curr Pharm Des 2003;9(22): ceptibility of a number of HIV-1 isolates from different clades to 1789–1802. the DKA, L-713,988. These isolates were susceptible to L- 8. Espeseth AS, Felock P, Wolfe A, et al.: HIV-1 integrase inhibitors 713,988, albeit at a lower level than the subtype B isolates ex - that compete with the target DNA substrate define a unique strand amined and one isolate, an A/C recombinant, exhibited a six-fold transfer conformation for integrase. Proc Natl Acad Sci USA decrease in susceptibility to the compound, with no evidence of 2000;97(21):11244–11249. the described resistance mutations being present. 15 9. Hazuda D, Blau CU, Felock P, et al.: Isolation and characteriza - Naturally occurring DKA resistance mutations have been tion of novel human immunodeficiency virus integrase inhibitors from fungal metabolites. Antiviral Chem Chemother 1999; identified in subtype B IN sequences (unpublished data), and 10(2):63–70. although previously described resistance mutations were not ob- 10. Pannecouque C, Pluymers W, Van Maele B, et al.: New class of served in this study, resistance of an A/C integrase recombi - HIV integrase inhibitors that block viral replication in cell culture. nant to L-731,988 suggests that mutations, in addition to those Curr Biol 2002;12(14):1169–1177. described, may be responsible for IN drug resistance and they 11. Goldgur Y, Craigie R, Cohen GH, et al.: Structure of the HIV-1 may exist naturally in non-subtype B isolates. integrase catalytic domain complexed with an inhibitor: A platform The data presented here suggest that IN can tolerate poly - for antiviral drug design. Proc Natl Acad Sci USA 1999; morphisms and yet retain functionality, however, if placed un - 96(23):13040–13043. der selective pressure by antiintegrase drugs, additional muta - 12. Hazuda DJ, Felock P, Witmer M, et al.: Inhibitors of strand trans - tions that reinstate IN functionality may evolve. fer that prevent integration and inhibit HIV-1 replication in cells. Science 2000;287(5453):646–650. Since more diverse HIV-1 strains are being identified in 13. Pluymers W, Pais G, Van Maele B, et al.: Inhibition of human im- countries with established health care systems and resources to munodeficiency virus type 1 integration by diketo derivatives. An- provide therapy, anti-IN resistance through natural polymor - timicrob Agents Chemother 2002;46(10):3292–3297. phisms may be an obstacle in the effective use of anti-IN drugs. 14. Reinke R, Steffen NR, and Robinson WE Jr: Natural selection re - More extensive studies on larger panels of isolates are required sults in conservation of HIV-1 integrase activity despite sequence to ensure the efficacy of anti-IN drugs globally and to avoid variability. AIDS 2001;15(7):823–830. potential pitfalls in their use. 15. Reinke R, Lee DJ, and Robinson WE Jr: Inhibition of human im - 5201_a12_p1010-1014 9/16/04 9:20 AM Page 1014
1014 HERRING ET AL.
munodeficiency virus type 1 isolates by the integrase inhibitor 24. Posada D and Crandall KA: MODELTEST: Testing the model of L-731,988, a diketo acid. Antimicrob Agents Chemother 2002; DNA substitution. Bioinformatics 1998;14(9):817–818. 46(10):3301–3303. 25. Lole KS, Bollinger RC, Paranjape RS, et al .: Full-length human 16. Billich A: S-1360 Shionogi-GlaxoSmithKline. Curr Opin Invest immunodeficiency virus type 1 genomes from subtype C-infected Drugs 2003;4(2):206–209. seroconverters in India, with evidence of intersubtype recombina - 17. Huang Y, Zhang L, and Ho DD: Characterization of gag and pol tion. J Virol 1999;73:152–160. sequences from long-term survivors of human immunodeficiency 26. Gunthard HF, Leigh-Brown AJ, D’Aquila RT, et al .: Higher se - virus type 1 infection. Virology 1998;240(1):36–49. lection pressure from antiretroviral drugs in vivo results in 18. Skinner LM, Lamers SL, Sanders JC, Eyster ME, Goodenow MM, increased evolutionary distance in HIV-1 pol. Virology 1999; and Katzman M: Analysis of a large collection of natural HIV-1 259(1):154–165 [published erratum appears in Virology 1999; integrase sequences, including those from long-term nonprogres - 261(2):367]. sors. J Acquir Immune Defic Syndr Hum Retrovirol 1998;19(2): 27. Burns CC, Gleason LM, Mozaffarian A, Giachetti C, Carr JK, and 99–110. Overbaugh J: Sequence variability of the integrase protein from a 19. Cornelissen M, van den Burg R, Zorgdrager F, Lukashov V, and diverse collection of HIV type 1 isolates representing several sub - Goudsmit J: pol gene diversity of five human immunodeficiency types. AIDS Res Hum Retroviruses 2002;18(14):1031–1041. virus type 1 subtypes: Evidence for naturally occurring mutations 28. Swanson P, Devare SG, and Hackett J Jr: Molecular characteriza - that contribute to drug resistance, limited recombination patterns, tion of 39 HIV isolates representing group M (subtypes A–G) and and common ancestry for subtypes B and D. J Virol 1997;71(9): group O: Sequence analysis of gag p24, pol integrase, and env 6348–6358. gp41. AIDS Res Hum Retroviruses 2003;19(7):625–629. 20. Lech WJ, Wang G, Yang YL, et al.: In vivo sequence diversity of 29. Fikkert V, Van Maele B, Vercammen J, et al.: Development of re- the protease of human immunodeficiency virus type 1: Presence of sistance against diketo derivatives of human immunodeficiency protease inhibitor-resistant variants in untreated subjects. J Virol virus type 1 by progressive accumulation of integrase mutations. J 1996;70(3):2038–2043. Virol 2003;77(21):11459–11470. 21. Herring BL, Ge YC, Wang B, et al .: Segregation of human im - munodeficiency virus type 1 subtypes by risk factor in Australia. Address reprint requests to: J Clin Microbiol 2003;41(10):4600–4604. Belinda L. Herring 22. Thompson JD, Higgins DG, and Gibson TJ: CLUSTAL W: Im - Department of Medicine proving the sensitivity of progressive multiple sequence alignment University of California, San Francisco through sequence weighting, position-specific gap penalities and weight matrix choice. Nucleic Acids Res 1994;22:4673–4680. 270 Masonic Avenue 23. PAUP: Phylogenetic analysis using parsimony, version 3.1.1 [com- San Francisco, California 94118 puter program]. Version. Illinois Natural History Survey, Cham -
paign, IL, 1991. E-mail: [email protected]