Antiviral Therapy 14:123 129 Short communication Genetic barriers for integrase inhibitor drug resistance in HIV type-1 B and CRF2_AG subtypes Almoustapha-Issiaka Maïga 1, Isabelle Malet 1 *, Cathia Soulie 1, Anne Derache 1, Victoria Koita 2, Bahia Amellal 1, Luba Tchertanov 3, Olivier Delelis 3, Laurence Morand-Joubert 4, Jean-François Mouscadet 3, Robert Murphy 5, Mamadou Cissé 2, Christine Katlama 5, Vincent Calvez 1 and Anne-Genevieve Marcelin 1 1 Department of Virology, Pitié-Salpêtrière Hospital, AP-HP, EA 2387, Université Pierre et Marie Curie, Paris, France 2 Centre d Ecoute, de Soins, d Animation et de Conseil, Bamako, Republic of Mali 3 LBPA, CNRS, Ecole Normale Supérieure de Cachan, Cachan, France 4 Department of Bacteriology-Virology, Saint-Antoine Hospital, AP-HP, Université Pierre et Marie Curie, Paris, France 5 Department of Infectious Diseases, Pitié-Salpêtrière Hospital, AP-HP, Université Pierre et Marie Curie, Paris, France *Corresponding author: E-mail: isabelle.malet@psl.aphp.fr These authors made an equal contribution to this work Background: HIV type-1 (HIV-1) integrase (IN) inhibitor resistance is the consequence of mutations that are selected in the viral IN gene targeted by antiretroviral drugs, such as raltegravir (RAL) and elvitegravir (EVG). The genetic barrier, defined as the number of viral mutations required to overcome the drug-selective pressure, is one of the important factors in the development of drug resistance. The genetic barrier for IN inhibitor resistance was compared between HIV-1 subtype B and HIV-1 subtype CRF2_AG, which is highly prevalent in West Africa and becoming more frequent in developed countries. Methods: IN nucleotide sequences from 73 HIV-1 subtype B and 77 HIV-1 subtype CRF2_AG antiretroviral-naive patients were examined at 19 IN amino acid positions implicated in RAL and EVG resistance. Results: The majority (14/19) of the studied positions showed a high degree of conservation of the predominant codon sequences leading to a similar genetic barrier between subtypes B and CRF2_AG. Nevertheless, at positions 14 and 151, the variability between subtypes affected the genetic barrier for the mutations G14C, G14S and V151I with a higher genetic barrier being calculated for subtype CRF2_AG. Conclusions: The major IN mutations E92Q, Q148K/R/H, N155H and E157Q (implicated in the resistance of IN inhibitors RAL and EVG) are highly conserved between subtypes B and CRF2_AG and display a similar genetic barrier. However, subtype CRF2_AG showed a higher genetic barrier to acquire mutations G14S, G14C and V151I as compared with subtype B, which could play a role in the resistance to RAL and/or EVG. Introduction Retroviral integrase (IN) is responsible for an essential step in HIV type-1 (HIV-1) replication involving the integration of retrovirus DNA into the host cell DNA. IN catalyses two reactions: 3 -end processing during which the terminal GpT dinucleotide is cleaved from the 3 -end of each long terminal repeat, producing CpA 3 -hydroxyl ends [1] and strand transfer in which both newly exposed 3 extremities of the viral DNA are covalently linked, by a one-step transesterification, into the host genome [2,3]. Raltegravir (RAL; MK-518) and elvitegravir (EVG; JTK-33/GS-9137) are members of a new class of HIV-1 inhibitors that block HIV-1 IN activity by interfering with the strand transfer stage [4,5]. Recently, IN resistance mutations have been described for RAL and EVG. Concerning RAL failure in vivo, it has been generally associated with two different genetic pathways: N155H associated with the secondary mutations L74M, E92Q, T97A, V151I and G163R or Q148K/R/H. RAL failure in vitro has been associated with the secondary mutations G14S/A and E138K. Other mutations could play a role in the resistance and/or failure, including Y143R/C plus L74A/I, E92Q, T97A, E157Q, I3M and S23R [6 9]. Concerning EVG resistance in vitro, the primary mutations T66I and E92K appearing with several secondary mutations H51Y, F121Y, S147G, S153Y, E157Q and R263K 9 International Medical Press 1359-6535 123
AI Maïga et al. have been described [1]. In vivo, two profiles have been more frequently described: E92Q with N155H and N155H with E138K, S147G, Q148R, G14S/C or Q148R/H [11]. The genetic barrier, defined as the number of viral mutations required to overcome the drug-selective pressure, is an important factor in the development of resistance. It has been shown that the rate of protease inhibitor (PI), nucleoside reverse transcriptase inhibitor (NRTI) and non-nrti (NNRTI) resistance development on the basis of the calculated genetic barrier might be similar for different HIV-1 subtypes. Nevertheless, several positions involving minor protease substitutions have shown a higher genetic barrier for some individual non-b subtypes, whereas other minor protease substitutions might influence resistance pathways [12]. HIV-1 subtype B is mainly found in the Americas, Europe and Australia, and subtype CRF2_AG, highly prevalent in West Africa, is becoming more frequent in developed countries [13 15]. Because of the lack of information concerning IN inhibitors and the prevalence of subtypes B and CRF2_AG, it is important to explore how the variability between subtypes B and CRF2_AG could affect IN resistance. Indeed, previous results have suggested some differences between subtypes B and CRF2_AG IN in a three-dimensional model [16]. The aim of this study was to explore the IN genetic barrier for evolution of RAL and EVG resistance substitutions. The variability at the nucleotide level between subtypes B and CRF2_AG could have a significant effect on the genetic barrier for RAL and EVG resistance. To explore the effect of this variability, IN nucleotide sequences from 73 HIV-1 subtype B and 77 HIV-1 subtype CRF2_AG antiretroviral (ARV)-naive patients were examined at 19 IN amino acid positions corresponding to 27 substitutions implicated in RAL and EVG resistance. Methods A total of 15 plasma samples were collected from patients who were diagnosed with HIV-1 infection in two clinical centres between 3 and 5 (CESAC, Bamako, the Republic of Mali and Pitié-Salpêtrière Hospital, Paris, France). All of these 15 patients had never received any ARV therapy. Subtype determination was on the basis of the reverse transcriptase and protease coding regions. Among the 15 patients, 73 were infected with HIV-1 subtype CRF2_AG and 77 with HIV-1 subtype B. A 1,86 base pair fragment encompassing the entire IN gene was amplified, sequenced and analysed as previously described [8]. The codons were evaluated for evolution from wild type by comparison with the HxB2 reference sequence in the 15 studied patients to all described RAL and EVG resistance-associated substitutions. The calculations used in this study were performed as described by van de Vijver et al. [12]. In summary, transitions (ts; A G or C T) occurred on average 2.5 more frequently than transversions (tv; A C, A T, C G, G T). The smallest number of ts (scored as 1) and/ or tv (scored as 2.5) required for evolution to RAL or EVG resistance-associated substitutions were used to calculate the genetic barrier. Results A total of 19 IN amino acid positions (51, 66, 74, 92, 97, 121, 138, 14, 143, 147, 148, 151, 153, 155, 157, 163, 3, 23 and 263) related to 27 IN resistance mutations were studied to explore the genetic barrier for RAL and EVG resistance. The global analyses of the 15 nucleotide sequences (73 from subtype B and 77 from subtype CRF2_AG) at the 19 studied positions showed 42 wild-type codons encoding for the expected wild-type amino acids determined on the basis of the HxB2 reference sequence and 12 codons reflecting a polymorphism at the amino acid level corresponding with L/I/M74, T/A97, V/I151, E/Q157, G/E/Q/N163, I/M/Q3 and S/N23 (the expected wild-type amino acid is represented in bold; Table 1). At each position, we always found a predominant codon that represented 74% of the population except at position 92 for subtype B, where the codons and accounted for 63% and 37%, respectively. At four positions, the predominant codons were different depending on the subtype: /GGC for subtype CRF2_AG/ subtype B for G14, AGC/AGT for S147, GTG/GTA for V151 and GGG/ for G163 (Table 1). The 42 wild-type codons present at the 19 positions described above were analysed. For each codon, we evaluated the number of ts and/or tv required for a drug resistance-associated substitution. A total of 27 amino acid substitutions implicated in RAL and/or EVG resistance were analysed and all the codons corresponding to a drug resistance substitution were evaluated (Table 1). Then, to calculate the genetic barrier from each wildtype codon, the mutated codon that allows the smallest number of ts and/or tv was retained (minimal score column in Table 1). Most of the predominant codons were remarkably conserved among subtypes B and CRF2_AG as the same prevalent codon was found in 14/19 positions (51, 66, 74, 97, 121, 138, 143, 148, 153, 155, 157, 3, 23 and 263); thus, variable scores of 1, 2, 2.5 or 3.5 were calculated according to the amino acid substitution considered, independently of the subtype (Figure 1). Different predominant codons between subtypes B and CRF2_AG were observed in 5/19 positions (92, 124 9 International Medical Press
The genetic barrier of HIV-1 integrase Table 1. Evaluation of the number of transitions and transversions required to obtain codons encoding for mutated amino acids and the calculated minimal score for genetic barrier analyses IN codon Codon, % position Subsitution a Codon Subtype B Subtype CRF2_AG Mutational resistance codon b Minimal score c 51 H51Y CAT 99 TAT 1ts TAC 2ts 1 CAC 1 2ts 1ts 1 66 T66I ACA 97 96 ATT 1ts, 1tv ATC 1ts, 1tv 1ts 1 ACC 3 1 2ts 1ts 1ts, 1tv 1 ACT 3 1ts 2ts 1ts, 1tv 1 74 L74A CTG 87 74 GCT 1ts, 2tv GCC 1ts, 2tv GCA 2ts, 1tv GCG 1ts, 1tv 3.5 CTA 4 5 1ts, 2tv 1ts, 2tv 1ts, 1tv 2ts, 1tv 3.5 TTG 4 3 1ts, 2tv 1ts, 2tv 2ts, 1tv 1ts, 1tv 3.5 TTA 6 1ts, 2tv 1ts, 2tv 1ts, 1tv 2ts, 1tv 3.5 (I74A) 4 8 3ts 2ts, 1tv 2ts 3ts 2 (M74A) 1 4 2ts, 1tv 2ts, 1tv 3ts 2ts 2 L74I CTG 87 74 ATT 2tv ATC 2tv 1ts, 1tv 3.5 CTA 4 5 2tv 2tv 1tv 2.5 TTG 4 3 2tv 2tv 2tv 5 TTA 6 2tv 2tv 1tv 2.5 (I74) 4 8 1tv 1tv (M74I) 1 4 1tv 1tv 1ts 1 L74M CTG 87 74 1tv 2.5 CTA 4 5 1ts, 1tv 3.5 TTG 4 3 1tv 2.5 TTA 6 1ts, 1tv 3.5 (I74M) 4 8 1ts 1 (M74) 1 4 92 E92Q 37 99 1tv CAG 1ts, 1tv 2.5 63 1 1ts, 1tv 1tv 2.5 E92K 37 99 AAA 1ts AAG 2ts 1 63 1 2ts 1ts 1 97 T97A ACA 96 GCT 1ts, 1tv GCC 1ts, 1tv GCA 1ts GCG 2ts 1 ACG 1 1ts, 1tv 1ts, 1tv 2ts 1ts 1 (A97) GCA 3 1tv 1tv 1ts 121 F121Y TTC 88 TAT 1ts, 1tv TAC 1tv 2.5 TTT 12 1tv 1ts, 1tv 2.5 138 E138K 99 AAA 1ts AAG 2ts 1 1 2ts 1ts 1 14 G14S 9 99 AGT 1ts, 1tv AGC 1ts, 1tv TCT 3tv TCC 3tv TCG 1ts, 2tv TCA 2tv 3.5 GGC 77 2ts 1ts 1ts, 2tv 2tv 3tv 3tv 1 GGT 14 1 1ts 2ts 2tv 1ts, 2tv 3tv 3tv 1 G14A 9 99 GCT 2tv GCC 2tv GCA 1tv GCG 1ts, 1tv 2.5 GGC 77 1ts, 1tv 1tv 2tv 2tv 2.5 Evaluation of the number of transitions (ts) and transversions (tv) required to obtain codons encoding for mutated amino acids implicated in raltegravir and elvitegravir resistance, and the calculation of the minimal score to evaluate the genetic barrier for each codon. a The substitution is specified in brackets when the amino acid is different from the HxB2 wild-type amino acid and is underlined when it already corresponds to the expected mutant amino acid. b The number of ts and tv selected to calculate the minimal score are in italics and bold. c The calculated score is dependent on the number of ts and tv; 1 ts is scored as 1 and 1 tv is scored as 2.5. IN, integrase. Antiviral Therapy 14.1 125
AI Maïga et al. Table 1. Continued IN codon Codon, % position Subsitution a Codon Subtype B Subtype CRF2_AG Mutational resistance codon b Minimal score c GGT 14 1 1tv 1ts, 1tv 2tv 2tv 2.5 G14C 9 99 TGT 2tv TGC 2tv 5 GGC 77 1ts, 1tv 1tv 2.5 GGT 14 1 1tv 1ts, 1tv 2.5 143 Y143C TAC 99 TGT 2ts TGC 1ts 1 TAT 1 1ts 2ts 1 Y143R TAC 99 CGT 3ts CGC 2ts CGA 2ts, 1tv CGG 2ts, 1tv AGA 1ts, 2tv AGG 1ts, 2tv 2 TAT 1 2ts 3ts 2ts, 1tv 2ts, 1tv 1ts, 2tv 1ts, 2tv 2 147 S147G AGT 97 22 GGT 1ts GGG 1ts, 1tv 1ts, 1tv GGC 2ts 1 AGC 3 78 2ts 1ts, 1tv 1ts, 1tv 1ts 1 148 Q148H 96 99 CAT 1tv CAC 1tv 2.5 CAG 4 1 1tv 1tv 2.5 Q148K 96 99 AAA 1tv AAG 1ts, 1tv 2.5 CAG 4 1 1ts, 1tv 1tv 2.5 Q148R 96 99 CGT 1ts, 1tv CGC 1ts, 1tv CGA 1ts CGG 2ts AGA 1ts, 1tv AGG 2ts, 1tv 1 CAG 4 1 1ts, 1tv 1ts, 1tv 2ts 1ts 2ts, 1tv 1ts, 1tv 1 151 V151I GTA 98 12 ATT 1ts, 1tv ATC 1ts, 1tv 1ts 1 GTG 88 1ts, 1tv 1ts, 1tv 2ts 2 GTC 1 2ts 1ts 1ts, 1tv 1 (I151) 1 1tv 1tv 153 S153Y TCT 92 94 TAT 1tv TAC 1ts, 1tv 2.5 TCC 7 6 1ts, 1tv 1tv 2.5 TCA 1 2tv 2tv 5 155 N155H AAT 99 96 CAT 1tv CAC 1ts, 1tv 2.5 AAC 1 4 1ts, 1tv 1tv 2.5 157 E157Q 96 94 1tv CAG 1ts, 1tv 2.5 3 1 1ts, 1tv 1tv 2.5 (Q157) 1 5 1ts 163 G163R GGG 11 88 CGT 2tv CGC 2tv CGA 1ts, 1tv CGG 1tv AGA 2ts AGG 1ts 1 83 1 1ts, 1tv 2tv 1tv 1ts, 1tv 1ts 2ts 1 (E163R) 4 1ts, 2tv 1ts, 2tv 1ts, 1tv 2ts, 1tv 2ts 3ts 2 (Q163R) 1 1 1ts, 1tv 1ts, 1tv 1ts 2ts 1ts, 1tv 2ts, 1tv 1 (E163R) 1 1ts, 2tv 1ts, 2tv 2ts, 1tv 1ts, 1tv 3ts 2ts 2 (N163R) AAT 1 1ts, 1tv 2ts, 1tv 1ts, 2tv 1ts, 2tv 1ts, 1tv 1ts, 1tv 3.5 3 I3M 98 96 1ts 1 (Q3M) 1 1ts, 2tv 6 (M3) 1 4 23 S23R AGC 9 CGT 1ts, 1tv CGC 1tv CGA 2tv CGG 2tv AGA 1tv AGG 1tv 2.5 (N23R) AAC 1 2ts, 1tv 1ts, 1tv 1ts, 2tv 1ts, 2tv 1ts, 1tv 1ts, 1tv 3.5 263 R263K AGA 96 97 AAA 1ts AAG 2ts 1 AGG 4 3 2ts 1ts 1 126 9 International Medical Press
The genetic barrier of HIV-1 integrase Figure 1. Comparison of the integrase genetic barriers for the evolution of RAL and EVG resistance substitutions between HIV-1 subtypes B and CRF2_AG Subtype B 6 4 Predominant codons, % Subtype CRF2_AG Subtype B 4 6 H51Y 6 4 CAT CAC ACA ACC ACT T66I CTG CTA TTG TTA L74A CTG CTA TTG TTA L74I CTG CTA TTG TTA L74M ACA GCA ACG TTC TTT GGC GGT GGC GGT GGC GGT E92Q E92K T97A F121Y E138K G14S G14A G14C Subtype CRF2_AG 4 6 TAC TAT TAC TAT AGT AGC CAG CAG CAG GTA GTG GTC Y143C Y143R S147G Q148H Q148K Q148R V15II TCT TCC TCA AAT AAC S153Y N155H E157Q GGG AAT G163R RAL and EVG resistance-associated substitutions AGC AAC AGA AGG 13M S23R R263K Calculated score 1 2 2.5 3.5 5 6 Comparison of the genetic barrier in the integrase (in relation to raltegravir [RAL] and elvitegravir [EVG] resistance mutations) between HIV type-1 (HIV-1) subtypes B and CRF2_AG. The frequency of each codon is represented by bars (up for subtype B and down for subtype CRF2_AG) at positions related to RAL and EVG resistance mutations. The minimal calculated score for each corresponding codon is represented by a colour code. Mutations with different calculated genetic barriers between subytpes B and CRF2_AG are underlined. Antiviral Therapy 14.1 127
AI Maïga et al. 14, 147, 151 and 163). Only at positions 14 and 151 did these differences have an effect on the calculated genetic barrier (1 versus 3.5 for G14S, 2.5 versus 5 for G14C and 1 versus 2 for V151I, respectively, for subtype B versus subtype CRF2_AG; Figure 1). Discussion The genetic barrier is an important determinant for the development of resistance and could be influenced by the existence of HIV-1 variability at the nucleotide level among the different HIV-1 subtypes. In this study, HIV-1 IN sequences isolated from subtypes B and CRF2_AG ARV-naive patients were studied at 19 amino acid positions related to RAL and EVG resistance-associated substitutions and used to calculate the genetic barrier. The majority of the studied positions (14/19) responsible for RAL and/or EVG resistance, particularly concerning the primary substitutions E92Q, Q148H/K/R, N155H and E157Q, showed a high degree of conservation of the predominant codon sequences, thus suggesting a similar genetic barrier between subtypes B and CRF2_AG. It seemed that the IN inhibitors mostly target sites that are functionally important as they are similar across the two studied subtypes B and CRF2_AG. Concordant results were described in a study concerning PI, NRTI and NNRTI resistance-associated positions where it has been shown that the calculated genetic barrier for nine studied HIV-1 subtypes were the same with just a few positions concerning minor protease substitutions described [12]. Among the five positions with predominant codons that differed between subtypes B and CRF2_AG (5/19), substitutions related to positions 92, 147 and 163 did not lead to a different genetic barrier, whereas polymorphisms at positions 14 and 151 led to higher scores for subtype CRF2_AG for the three mutations G14C, G14S and V151I. In the same way, in the study concerning the calculated genetic barrier for PI, NRTI and NNRTI described above, few positions that differentiated subtypes have shown a higher genetic barrier for non-b subtypes than for subtype B [12]. Concerning the resistance to RAL, most of the studies have shown that Q148K/R/H is always associated with G14S during RAL failure [6,8,1]. The fact that the genetic barrier calculated for G14S is higher for subtype CRF2_AG than B could possibly suggest that it would be more difficult for subtype CRF2_AG to become resistant to RAL or perhaps it will preferentially fail via the N155H pathway. In some patients, natural codons encoding for amino acids implicated in the RAL primary resistance (L74I/M, T97A, V151I, E157Q and I3M) were described, suggesting that resistance to IN inhibitors could be natural or could occur faster in these patients. In conclusion, the major IN mutations E92Q, Q148K/R/H, N155H and E157Q, implicated in the resistance of IN inhibitors RAL and EVG are very well conserved between the two more frequent subtypes in France, B and CRF2_AG, and they display a similar genetic barrier. However, subtype CRF2_AG showed a higher genetic barrier to acquire mutations G14S, G14C and V151I as compared with subtype B and this should be further investigated in clinical practice. Acknowledgements This work was presented at the 15th ICASA, 3 7 December 8, Dakar, Sénégal as an oral presentation. This work was supported by Sidaction and the ANRS (the French National Agency for AIDS research). Disclosure statement The authors declare no competing interests. References 1. Engelman A, Mizuuchi K, Craigie R. HIV-1 DNA integration: mechanism of viral DNA cleavage and DNA strand transfer. Cell 1991; 67:1211 1221. 2. Brown PO, Bowerman B, Varmus HE, Bishop JM. Retroviral integration: structure of the initial covalent product and its precursor, and a role for the viral IN protein. Proc Natl Acad Sci U S A 1989; 86:2525 2529. 3. Ellison V, Abrams H, Roe T, Brown P. Human immunodeficiency virus integration in a cell-free system. J Virol 199; 64:2711 2715. 4. Lataillade M, Kozal MJ. The hunt for HIV-1 integrase inhibitors. AIDS Patient Care STDS 6; :489 51. 5. Pommier Y, Johnson AA, Marchand C. Integrase inhibitors to treat HIV/AIDS. Nat Rev Drug Discov 5; 4:236 248. 6. Cooper D, Gatell J, Rockstroh C, et al. Results of BENCHMRK-1, a phase III study evaluating the efficacy and safety of MK-518, a novel HIV-1 integrase inhibitor, in patients with triple-class resistant virus. 14th Conference of Retroviruses and Opportunistic Infections. 25 28 February 7, Los Angeles, CA, USA. Abstract 15aLB. 7. Goethals O, Clayton R, Wagemans E, et al. Resistance mutations in HIV-1 integrase selected with raltegravir or elvitegravir confer reduced susceptibility to a diverse panel of integrase inhibitors. XVII International HIV Drug Resistance Workshop. 1 14 June 8, Sitges, Spain. Abstract 9. 8. Malet I, Delelis O, Valantin MA, et al. Mutations associated with failure of raltegravir treatment affect integrase sensitivity to the inhibitor in vitro. Antimicrob Agents Chemother 8; 52:1351 1358. 9. Steigbigel R, Kumar P, Eron J, et al. Results of BENCHMRK-2, a phase III study evaluating the efficacy and safety of MK-518, a novel HIV-1 integrase inhibitor, in patients with triple-class resistant virus. 14th Conference of Retroviruses and Opportunistic Infections. 25 28 February 7, Los Angeles, CA, USA. 1. Jones G, Ledford R, Yu F, et al. Resistance profile of HIV-1 mutants in vitro selected by the HIV-1 integrase inhibitor, GS-9137 (JTK-33). 14th Conference of Retroviruses and Opportunistic Infections. 25 28 February 7, Los Angeles, CA, USA. 128 9 International Medical Press
The genetic barrier of HIV-1 integrase 11. McColl D, Fransen E, Gupta S, et al. Resistance and crossresistance to first generation integrase inhibitors: insights from a Phase II study of elvitegravir (GS-9137). XVI international HIV Drug Resistance Workshop. 12 16 June 7, Barbados, West Indies. 12. van de Vijver DA, Wensing AM, Angarano G, et al. The calculated genetic barrier for antiretroviral drug resistance substitutions is largely similar for different HIV-1 subtypes. J Acquir Immune Defic Syndr 6; 41:352 36. 13. Derache A, Traore O, Koita V, et al. Genetic diversity and drug resistance mutations in HIV type 1 from untreated patients in Bamako, Mali. Antivir Ther 7; 12:123 129. 14. Descamps D, Chaix ML, Andre P, et al. French national sentinel survey of antiretroviral drug resistance in patients with HIV-1 primary infection and in antiretroviral-naive chronically infected patients in 1 2. J Acquir Immune Defic Syndr 5; 38:545 552. 15. Wensing AM, van de Vijver DA, Angarano G, et al. Prevalence of drug-resistant HIV-1 variants in untreated individuals in Europe: implications for clinical management. J Infect Dis 5; 192:958 966. 16. Malet I, Soulie C, Tchertanov L, et al. Structural effects of amino acid variations between B and CRF2_AG HIV-1 integrases. J Med Virol 8; :754 761. Accepted for publication 13 October 8 Antiviral Therapy 14.1 129