The preferential selection of K65R in HIV-1 subtype C is attenuated by nucleotide polymorphisms at thymidine analogue mutation sites

Journal of Antimicrobial Chemotherapy Advance Access published June 7, 2013 J Antimicrob Chemother doi:10.1093/jac/dkt204 The preferential selection of K65R in HIV-1 subtype C is attenuated by nucleotide polymorphisms at thymidine analogue mutation sites Cédric F. Invernizzi 1,2, Dimitrios Coutsinos 1 3, Maureen Oliveira 1, Rita S. Schildknecht 1,2, Hongtao Xu 1, Simani Gaseitswe 4, Daniela Moisi 1, Bluma G. Brenner 1,2 and Mark A. Wainberg 1 3 * 1 McGill University AIDS Centre, Lady Davis Institute for Medical Research, Sir Mortimer B. Davis Jewish General Hospital, Montréal, Québec, Canada; 2 Department of Medicine, McGill University, Montréal, Québec, Canada; 3 Department of Microbiology and Immunology, McGill University, Montréal, Québec, Canada; 4 Botswana Harvard AIDS Institute, Gaborone, Botswana *Corresponding author. McGill AIDS Centre, Lady Davis Institute, Jewish General Hospital, 3755 Côte-Ste-Catherine Road, Montréal, Québec H3T 1E2, Canada. Tel: +1-514-340-8260; Fax: +1-514-340-7537; E-mail: mark.wainberg@mcgill.ca Present address: Spiez Laboratory, Spiez, Switzerland. Present address: CSL Behring AG, Bern, Switzerland. Received 22 January 2013; returned 12 March 2013; revised 22 April 2013; accepted 30 April 2013 Objectives: We recently reported the preferential selection of the K65R resistance mutation in subtype C HIV-1 compared with subtype B and showed the underlying mechanism to be dependent on subtype C-specific silent nucleotide polymorphisms, i.e. genomic mutations that change the genotype but not the phenotype. The number of clinical reports demonstrating elevated numbers of K65R nevertheless suggests the existence of factors limiting the increased incidence of K65R mutations. Thus, we investigated the contributions of subtype C-specific silent nucleotide polymorphisms at thymidine analogue mutation (TAM) sites 70, 210 and/or 219 that might reduce the previously described preferential selection of K65R in subtype C HIV-1 associated with subtype C-specific nucleotide polymorphisms at sites 64/65. Methods: Cell culture drug selections were performed with various drugs in MT2 cells. Results: The use of nucleoside/nucleotide reverse transcriptase inhibitors [N(t)RTIs] as single drugs or in combination confirmed the more frequent selection of K65R by multiple N(t)RTIs in a subtype B virus that contained the 64/ 65 nucleotide polymorphisms of subtype C than in a wild-type subtype B virus. This effect was attenuated in the presence of several silent TAM nucleotide polymorphisms, except when stavudine was employed in the selection protocol. Conclusions: These results further demonstrate that stavudine can preferentially select for K65R in subtype C virus and also provide a basis for understanding the importance of silent nucleotide polymorphisms in regard to altered HIV drug resistance profiles. Keywords: subtype differences B/C, nucleotide polymorphisms, K65R resistance mutation, cell culture drug selections, N(t)RTIs, TAMs Introduction We recently described the more rapid tissue culture selection of K65R mutations with tenofovir in clinical isolates of subtype C than subtype B HIV-1. 1 Careful analysis of these drug resistance profiles concluded that there must be an underlying subtypespecific mechanism that accounted for this difference. The reverse transcriptase (RT) enzymes of subtypes B and C behave in similar fashion in biochemical assays, 2 suggesting that differences in nucleotide sequences between subtypes might be involved. Further investigation revealed that two subtype C-specific silent nucleotide polymorphisms that change the genotype but not the phenotype at positions 64 and 65 of RT are involved and that both polymorphisms had to be present in tandem in order for a selection pattern distinct from that of subtype B virus to occur. 3 6 An increasing numberof reports now indicate that levels of K65R mutations are on the rise, especially in patients infected with subtype C HIV-1 who have failed treatment. 7 9 These reports are in disagreement with multiple studies conducted on subjects infected with non-b subtypes of HIV-1. 10 15 Small sample sizes of virological failures, incomplete treatment regimens, mixtures of different subtypes and non-proportional pooling of non-b subtypes may have biased the results of these studies. 15 Despite the elevated numbers of patients infected harbouring K65R mutations, 7 9 it is nonetheless likely that several factors might limit the proportion of K65R that might otherwise occur. 3 6 # The Author 2013. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com 1of5

Invernizzi et al. First, zidovudine is widely used as part of first-line regimens in developing countries with a high prevalence of subtype C and is known to favour the selection of thymidine analogue mutations (TAMs), which antagonize the development of K65R. 7,16 19 Second, K65R is associated with a loss of viral replicative fitness, 20 which might compromise the ability of this mutation to be generated. Indeed, phenotypic polymorphisms at TAM sites of HIV-2 have been shown to preclude the development of K65R with nucleoside/nucleotide reverse transcriptase inhibitors [N(t)RTIs] used as single drugs or in combination. 21 Therefore, we wished to assess the involvement of known silent nucleotide polymorphisms at three different TAM sites in HIV-1 subtype C that, in contrast to those in HIV-2, do not result in a change in phenotype. Materials and methods Protocols similar to those previously described were employed. 6 The following primers were purchased in desalted purity from Integrated DNA Technologies in order to generate all mutant NL4-3 plasmids NL4-3 (64/65) (forward): 5 -CTCCAGTATTTGCCATAAAAAAGAAAGACAGTACTAAATGGAG-3 ; NL4-3 (64/65) (reverse): 5 -CTCCATTTAGTACTGTCTTTCTTTTTTATGGCAAA TACTGGAG-3 ; NL4-3 (70) (forward): 5 -GAAAGACAGTACTAAGTGGAGAAAAT TAGTAGATTTCAG-3 ; NL4-3 (70) (reverse): 5 -CTGAAATCTACTAATTTTCTCCACT TAGTACTGTCTTTC-3 ; NL4-3 (219) (forward): 5 -CCACACCAGACAAGAAACA TCAGAAAGAACC-3 ; NL4-3 (219) (reverse): 5 -GGTTCTTTCTGATGTTTCTTG TCTGGTGTGG-3 ; NL4-3 (210) (forward): 5 -CTGAGACAACATCTGTTAAGGT GGGGATTTAC-3 ; NL4-3 (210) (reverse): 5 -GTAAATCCCCACCTTAACAGATG TTGTCTCAG-3. Mutated residues are shown in bold. Results 16 19 Due to known antagonistic effects between TAMs and K65R, we compared the consensus sequences of subtypes B and C with a focus on sites known to develop TAMs, since nucleotide polymorphisms at such sites might reduce the probability of selection of K65R by increasing the likelihood of selection of other mutations. Three TAM sites harbour single nucleotide polymorphisms at the third position of the codon, i.e. K70K, L210L and K219K, representing differences between subtypes B and C (Figure 1a). At positions 70 and 219, the change is from A to G, whereas a G to A switch occurs at position 210. To investigate the impact of these silent nucleotide polymorphisms on the preferred development of K65R in subtype C virus, we first generated HIV-1 NL4-3 clones containing single nucleotide mutations at positions 64 and 65 to create viruses behaving as if they were subtype C, i.e. NL4-3 (64/65), as previously described. 6 Next, we introduced polymorphisms at TAM positions 70, 219 and/or 210 (Figure 1b). These three TAM polymorphism-containing viruses, i.e. NL4-3 (64/65/70), NL4-3 (64/65/70/219) and NL4-3 (64/65/70/219/210), were compared with wild-type NL4-3 and NL4-3 (64/65) in cell culture drug selections with N(t)RTIs under single or combination drug pressure. Based on known TAM antagonisms toward K65R, 16 19 all three viruses containing TAM polymorphisms were expected to show an attenuated frequency of K65R selection compared with NL4-3 (64/ 65) and to behave in an manner more similar to that of wild-type virus, i.e. NL4-3 (wt). Such a result would explain, at a nucleotide level, discrepancies between previous cell culture results with NL4-3 (64/65) in which K65R had been selected at high frequency, 6 compared with clinical data available from patients infected with subtype C virus who often did not possess K65R at such a high incidence. 7 9 In a first round of selections, we included tenofovir, didanosine, stavudine, apricitabine and abacavir as single drugs for 20 weeks or until a first resistance mutation had appeared (Figure 2). We did not include lamivudine, emtricitabine and zidovudine, since no subtype-specific differences are expected with these drugs; i.e. lamivudine and emtricitabine preferentially select for M184I/V mutations, while zidovudine favours the development of TAMs, notably K70R. 6 Selections with tenofovir resulted in the appearance of K65R in all viruses tested, whereas the four other drugs, i.e. didanosine, stavudine, apricitabine and abacavir, failed to select for K65R with NL4-3 (wt) but did select for K65R when NL4-3 (64/65) was employed, consistent with earlier results. 6 In contrast, the preferential selection of K65R that was observed with NL4-3 (64/65) was attenuated in more than one-third of the TAM polymorphism-containing viruses that were exposed to didanosine, apricitabine or abacavir but not to stavudine. Interestingly, none of the acquired resistance mutations were TAMs. These single drug selections demonstrate that silent TAM nucleotide polymorphisms may be able to shift the resistance mutation selection pattern away from NL4-3 (64/65), which favours K65R. In drug combination selections, we subjected the same viruses to tenofovir/emtricitabine, tenofovir/lamivudine, stavudine/didanosine, abacavir/emtricitabine and abacavir/lamivudine for 20 weeks or until the first resistance mutation had appeared (Figure 2). As expected, abacavir/lamivudine selected for the M184I mutation in all cases, and no subtype differences could be noted. 6 The selection of K65R in the case of NL4-3 (64/65) was attenuated in many of the TAM polymorphism-containing viruses that were exposed to tenofovir/emtricitabine, tenofovir/lamivudine or abacavir/emtricitabine. In contrast, the combination of stavudine/didanosine continued to select for K65R in all cases with the exception of subtype B virus, i.e. NL4-3 (wt). These results demonstrate that silent TAM nucleotide polymorphisms may be able to diminish the likelihood of selection of K65R when various N(t)RTI combinations are studied, although not in the case of stavudine/didanosine. Discussion These results show that silent nucleotide polymorphisms found at TAM sites can attenuate the tendency of viruses containing the 64/ 65 subtype C sequence to favour the K65R resistance mutation pathway. The recombinant viruses that were tested developed K65R under drug pressure at high rates, but below the rates of viruses that did not contain subtype C-specific TAM polymorphisms, as previously reported. 6 These data are consistent with those that might be associated with K65R development in patients infected with subtype C HIV previously treated with TAM-selecting drugs such as zidovudine. 7 9 In contrast, the results obtained with stavudine indicated that K65R selection occurred in all circumstances tested, consistent with the results of clinical studies in areas in which subtype C virus is pandemic that have pointed to high rates of K65R in stavudine-treated individuals. 7 9 Other studies have shown that RNA template sequences that differ between subtype C viruses 2of5

Attenuated selection of K65R in HIV-1 subtype C JAC (a) (b) Figure 1. (a) Location of the K65R mutation in HIV-1 subtype B and C sequences. Representation of the sites relevant to this study and the surrounding nucleotide sequences, and comparison of the subtype B and C consensus sequences. All subtype C nucleotides that are different from subtype B sequences are shown in bold. Frequencies of subtype-specific silent nucleotide polymorphisms relevant to this study are shown in percentages that were calculated for each subtype from alignments of complete sequences available from the HIV Sequence Database of Los Alamos National Laboratory. The nucleotide change leading to the K65R mutation is shaded in black and indicated by an arrow. Of note, the underscored poly(a) sequence is shifted in subtype C. (b) Sequences of wild-type and mutated NL4-3 viruses. The sequence of NL4-3 (wt) corresponds to the subtype B consensus sequence in (a). Single nucleotide changes to the sequence in the four mutants are shown in bold. All the mutations introduced have no effect on the phenotype. and those of other subtypes are responsible for these differences in propensity for selection of K65R. None of the TAM polymorphismcontaining viruses developed TAMs in the absence of K65R but rather seemed to favour other mutations such as M184I. A possible explanation may lie in the known antagonism between K65R and TAMs, 16 19 and ultrasensitive assays for the detection of individual mutations that are present at levels below limits of detection may help to resolve this issue. In summary, the data presented here help to explain the fact that patients who have been treated with zidovudine as part of first-line regimens in developing countries may possess TAMs that antagonize the later development of 16 19 K65R mutations following subsequent therapy. Other nucleotide polymorphisms may, of course, also militate against the development of K65R, and this should be investigated. 3of5

Invernizzi et al. Virus TFV ddi d4t ATC ABC wt K65R L74V D76G M184I M184I 64/65 K65R K65R K65R K65R K65R 64/65/70 K65R K65R K65R M184T K65R 64/65/70/219 K65R # K65R K65R K65R M184I 64/65/70/210/219 K65R V75I K65R K65R M184I Virus TFV/FTC TFV/3TC d4t/ddi ABC/FTC ABC/3TC Funding This research was supported by grants from the Canadian Institutes of Health Research. Transparency declarations None to declare. wt K65R M184I* V75I M184I M184I 64/65 K65R K65R K65R K65R M184I 64/65/70 M184I* K65R K65R K65R M184I 64/65/70/219 M184I M184I K65R M184I M184I* 64/65/70/210/219 K65R K65R K65R M184V M184I Figure 2. Mutational preferences of wild-type and mutated NL4-3 viruses in MT2 cells under single and combination drug pressure with N(t)RTIs. Viruses: wt denotes wild-type subtype B NL4-3 virus; 64/65, 64/65/70, 64/65/70/219 and 64/65/70/210/219 denote subtype B NL4-3 viruses with subtype C sequences at positions 64, 65, 70, 210 and/or 219. Drug abbreviations: TFV, tenofovir; ddi, didanosine; d4t, stavudine; ATC, apricitabine; ABC, abacavir; FTC, emtricitabine; 3TC, lamivudine. The duration of the study was 20 weeks. The mutations listed were fully developed and unique at 20 weeks or earlier. *Can vary between experiments (M184I or K65R). # The mutation developed later than week 20. All selections for resistance were performed at least three times. Selection of K65R is shown in bold. References 1 Brenner BG, Oliveira M, Doualla-Bell Fet al. HIV-1 subtype Cviruses rapidly develop K65R resistance to tenofovir in cell culture. AIDS 2006; 20: F9 13. 2 Xu H, Quan Y, Asahchop E et al. Comparative biochemical analysis of recombinant reverse transcriptase enzymes of HIV-1 subtype B and subtype C. Retrovirology 2010; 7: 80. 3 Coutsinos D, Invernizzi CF, Moisi D et al. A template-dependent dislocation mechanism potentiates K65R reverse transcriptase mutation development in subtype C variants of HIV-1. PLoS One 2011; 6: e20208. 4 Coutsinos D, Invernizzi CF, Xu H et al. Factors affecting template usage in the development of K65R resistance in subtype C variants of HIV type-1. Antivir Chem Chemother 2010; 20: 117 31. 5 Coutsinos D, Invernizzi CF, Xu H et al. Template usage is responsible for the preferential acquisition of the K65R reverse transcriptase mutation in subtype C variants of human immunodeficiency virus type 1. J Virol 2009; 83: 2029 33. 6 Invernizzi CF, Coutsinos D, Oliveira M et al. Signature nucleotide polymorphisms at positions 64 and 65 in reverse transcriptase favor the selection of the K65R resistance mutation in HIV-1 subtype C. J Infect Dis 2009; 200: 1202 6. 7 Doualla-Bell F, Avalos A, Brenner B et al. High prevalence of the K65R mutation in human immunodeficiency virus type 1 subtype C isolates from infected patients in Botswana treated with didanosine-based regimens. Antimicrob Agents Chemother 2006; 50: 4182 5. 8 Lyagoba F, Dunn DT, Pillay D et al. Evolution of drug resistance during 48 weeks of zidovudine/lamivudine/tenofovir in the absence of real-time viral load monitoring. J Acquir Immune Defic Syndr 2010; 55: 277 83. 9 Wallis CL, Erasmus L, Varughese S et al. Emergence of drug resistance in HIV-1 subtype C infected children failing the South African national antiretroviral roll-out program. Pediatr Infect Dis J 2009; 28: 1123 5. 10 Geretti AM. HIV-1 subtypes: epidemiology and significance for HIV management. Curr Opin Infect Dis 2006; 19:1 7. 11 Kantor R, Katzenstein DA, Efron B et al. Impact of HIV-1 subtype and antiretroviral therapy on protease and reverse transcriptase genotype: results of a global collaboration. PLoS Med 2005; 2: e112. 12 Kantor R. Impact of HIV-1 pol diversity on drug resistance and its clinical implications. Curr Opin Infect Dis 2006; 19: 594 606. 13 DART Virology Group and Trial Team. Virological response to a triple nucleoside/nucleotide analogue regimen over 48 weeks in HIV-1-infected adults in Africa. AIDS 2006; 20: 1391 9. 14 Miller MD, Margot N, McColl D et al. K65R development among subtype C HIV-1-infected patients in tenofovir DF clinical trials. AIDS 2007; 21: 265 6. 15 Brenner BG. Resistance and viral subtypes: how important are the differences and why do they occur? Curr Opin HIVAIDS 2007; 2: 94 102. 16 Kagan RM, Lee TS, Ross L et al. Molecular basis of antagonism between K70E and K65R tenofovir-associated mutations in HIV-1 reverse transcriptase. Antiviral Res 2007; 75: 210 8. 17 Parikh UM, Zelina S, Sluis-Cremer N et al. Molecular mechanisms of bidirectional antagonism between K65R and thymidine analog mutations in HIV-1 reverse transcriptase. AIDS 2007; 21: 1405 14. 18 Martin-Carbonero L, Gil P, Garcia-Benayas Tet al. Rate of virologic failure and selection of drug resistance mutations using different triple 4of5

Attenuated selection of K65R in HIV-1 subtype C JAC nucleos(t)ide analogue combinations in HIV-infected patients. AIDS Res Hum Retroviruses 2006; 22: 1231 5. 19 Wainberg MA. Generic HIV drugs enlightened policy for global health. N Engl J Med 2005; 352: 747 50. 20 Cong ME, Heneine W, Garcia-Lerma JG. The fitness cost of mutations associated with human immunodeficiency virus type 1 drug resistance is modulated by mutational interactions. J Virol 2007; 81: 3037 41. 21 Ntemgwa ML, Toni TD, Brenner BG et al. Nucleoside and nucleotide analogs select in culture for different patterns of drug resistance in human immunodeficiency virus types 1 and 2. Antimicrob Agents Chemother 2009; 53: 708 15. 5of5