The HIV-1 fusion inhibitor enfuvirtide (fuzeon, or T-20) is a synthetic 36-amino acid peptide derived from the C-terminal region of the HR2 domain . Several reports have shown that enfuvirtide acts through a dominant negative mechanism, packing into the hydrophobic grooves of the HR1 coiled coil and preventing the gp41 wild-type HR2 sequence from also doing so [1,2].
Phase III clinical trials with enfuvirtide (TORO-I and TORO-II) have demonstrated an increase in the CD4 T-cell count and a significant reduction in HIV-1-RNA concentration in enfuvirtide-treated patients [3,4]. However, continued exposure to the drug leads to the emergence of viral resistance when virus replication is not completely suppressed . Sequence analysis of enfuvirtide-resistant viral populations has revealed the acquisition of mutations within the three amino acid motif at position 36–38 (GIV) [5,6], which is highly conserved in the HR1 domain of gp41 [7–9]. Other studies have also found mutations in a wider region of HR1 (amino acids 36–45) [10–13]. In addition, the recent identification of changes associated with enfuvirtide resistance in HR2 positions 126 and 138 [14,15] generates an increasingly complex picture.
The aim of our study was to investigate the molecular evolution of the gp41 protein in HIV-1-infected patients with detectable plasma viraemia during long-term salvage enfuvirtide therapy.
Materials and methods
Sixty-four heavily antiretroviral-experienced HIV-1-infected patients with genotypic and phenotypic resistance to multiple protease and reverse transcriptase (RT) inhibitors who were receiving enfuvirtide (90 mg twice a day) were included in the study. Among these patients, we selected 13 with incomplete virological suppression during follow-up. A median of four (range 2–7) plasma samples were obtained between weeks 1 and 96.
Population-based sequencing of gp41
Plasma samples were collected from the 13 enfuvirtide-treated patients at baseline and during treatment. Viral RNA was extracted and the gp41 coding region was amplified. An RT–polymerase chain reaction (PCR) was performed (SuperScript One-Step RT–PCR kit; Invitrogen, Spain) with primers Egp41F and Egp41R (nucleotides 7728–7749 and 8408–8422, respectively of the HIVHXB2 numbering system). A nested PCR was carried out (Platinum Taq DNA Polymerase High Fidelity; Invitrogen) with primers Ngp41F and Ngp41R (nucleotides 7797–7816 and 8337–8359, respectively). Population-based sequencing was carried out using Big-Dye Terminator Cycle Sequencing and the ABI 3100 sequence analyser (Applied Biosystems, Foster City, California, USA). All sequences were assembled, aligned and edited using the Sequencher and GeneDoc software.
Clonal-based sequencing of gp41
For four patients, at baseline and longitudinally after the initiation of enfuvirtide treatment, three independent RT–PCR for each sample were performed under identical conditions to minimize resampling bias. Nested PCR was then carried out for each RT–PCR and the three PCR products were purified, pooled and cloned into a pBluescript vector. Individual clones were sequenced and analysed as described above.
A Fisher's two-tailed test was used to assess the distribution of resistance-associated mutations in the clonal sequences.
Molecular evolutionary dynamics of gp41 in enfuvirtide-treated patients
Under the selective pressure imposed by the enfuvirtide treatment, resistance mutations to this drug became detectable within the first 4 weeks of treatment (Table 1). Position 43 was the most frequently mutated, with the change from N to D being observed in viruses from eight patients (62%). Viruses from five patients carried mutations at position V38, and the change to A was the most prevalent. Viruses from four patients also developed a change at position 36, viruses from two patients showed a mixture of G36G/S and G36G/D, whereas a G36V mutation was observed in viruses from the other two patients. Although the most prevalent mutations were detected in the HR1 domain, we also found several changes along the gp41 ectodomain, upstream from HR1 (T25S), in the loop (N105N/D; Q110E/K) and in the HR2 domain, where the most prevalent mutation was found at position N126. This change was observed in viruses from three patients: in one patient, N first mutated to S and subsequently to K, and in the other two patients, N changed to N/K and H, respectively. Substitution S138A was observed in only one patient.
During follow-up the pattern of substitutions changed several times in most of the patients. Mutations 36 and 38 emerged rapidly (median emerging time 10 weeks, range 2–24), but they subsequently disappeared (patient nos. 1, 5, 6, 10, 11, and 13). Therefore, contrary to what occurs in the process of the acquisition of resistance to protease and RT inhibitors, changes did not accumulate over time in the HR1 domain, and none of the patients had more than two mutations after 6 months of treatment. Interestingly, at the last follow-up timepoints, the N43D substitution was not observed together with changes at positions 36 or 38 in any patient. Several amino acid changes in the HR2 region were detected during enfuvirtide treatment. In all but two patients these changes emerged subsequent to the selection of mutations in the HR1 domain. Changes at positions 113 (G to G/S) and 126 (N to S) in patient 3 appeared at week 4. Later, at week 24, a change in 126K was selected. These substitutions (G113S and N126K) were maintained until changes in HR1 were selected, with the N126K mutation being present at the end of the follow-up period. In patient 4, viruses harbouring the M24R and K144K/R mutations were detectable as early as one week after initiation of the treatment (data not shown). This mutant genotype became undetectable thereafter and a N43N/D mixture emerged at week 4.
Sequence analysis of gp41 clones
To obtain a more accurate picture of the evolutionary dynamics of the mutations in the gp41 sequence, we performed an analysis of 320 individual clones from four patients (patient nos. 1, 4, 5 and 10). This analysis showed that substitutions at positions 36 and 38 were already detected before week 4. For example, in patient 4, 30% of the clones contained the G36S/D change after only one week of enfuvirtide therapy (Table 2a). Moreover, this clonal analysis confirmed the disappearance of mutations at positions 36 and 38 observed in the population-based sequencing of plasma samples (Table 1). In addition, all of these changes did not coexist in the same viral genome, suggesting an incompatibility (negative pleitropy) between mutations at sites 36, 38 and 43 (Table 2a and b). Analysis of these 320 clones indicated that 61 and 51 clones contained a single mutation at either position 36 or 38, respectively, but none of the sequences contained a double mutation. This highly exclusive relationship was statistically significant (Fisher's exact test, P < 0.001). Likewise, a strong and statistically significant (P < 0.001) exclusion was also observed for the other pair of mutations 36/43 and 38/43 (Table 2b).
Phase III clinical trials (TORO-I and TORO-II) have demonstrated the safety and efficacy of enfuvirtide in treatment-experienced HIV-1-infected patients [3,4]. However, as has happened with all other approved antiretroviral drugs, HIV-1 may develop resistance to this fusion inhibitor [5,6,10–15].
In this study, we analysed the emergence of resistance mutations in 13 heavily pre-treated HIV-1-infected patients failing an enfuvirtide-containing salvage therapy. The selection of several drug resistance-associated mutations was detected along the whole gp41 ectodomain. The most frequent included mutations in the HR1 domain that have already been associated with enfuvirtide resistance. However, we identified other unreported changes in both regions, HR1 and HR2. One of the most interesting results from our study is the transient selection of mutations at positions 36 and 38 during follow-up. Several explanations could be proposed for their transient existence. First, in-vitro experiments showed that each of these mutants displayed reduced fitness compared with the wild type, and that the double mutant 36/38 was even less fit than each single mutant, suggesting that these viruses would probably be eliminated in vivo . Second, and in agreement with this possibility, Kinomoto et al.  showed that changes at position 36 were correlated with decreased syncytium-forming activity and viral infectivity. Third, both mutations cause a delay in fusion kinetics that results in enhanced sensitivity to neutralizing antibodies, suggesting that these viruses may be eliminated in vivo . Finally, the nucleotides encoding some enfuvirtide resistance mutations, such as 36, 37 and 38 apparently modify the structural stability of the Rev-responsive element (RRE), impairing the gp41 protein as well as the RRE functions . These changes seriously affect replication kinetics because the mutations lying in the RRE could impair messenger RNA transport from the nucleus to the cytoplasm. In contrast, other changes (e.g. at position 43) are located at positions where structural changes in RRE may be minimal . Our clonal analysis demonstrated that the three main mutations implicated in resistance to enfuvirtide (mutations 36, 38 and 43) are present in separate clones, thus constituting independent populations. In addition, the low frequency of double or multiple mutants found in our study and in others could be explained by the presence of the mutations in individual genomes. The statistically significant exclusion described here is not unique to gp41, as it has also been described for the K70R and L210W mutations in RT , and for the D30N and L90M mutations in protease .
Several changes were also selected over time in the HR2 domain. Similar to Xu et al. , we found the S138A substitution, although the frequency of this mutation was low in agreement with other previous reports [10,13]. Changes in HR2 at codon 126 were found in samples from three out of 13 patients, but only in two cases was the previously reported N126K mutation detected . It has been suggested that changes in the HR2 domain could represent secondary or compensatory mutations [14,15]. However, in some cases we detected changes in the HR2 region even before the emergence of HR1 mutations (G113G/S, N126S-126K, 144K/R). It thus appears unlikely that all of the changes in HR2 could represent compensatory mutations. Further studies are required to determine whether these mutations are involved in resistance to enfuvirtide.
In the current study we have not observed a pattern of molecular convergence in the HIV-1 gp41 quasispecies isolated from different patients. These evolutionary differences between patients could be caused by the different env genetic backgrounds in each patient. In addition, as the env gene is the main target of the immune system, the selective pressure of the immune system could also play a role in the viral evolution in the presence of enfuvirtide.
In conclusion, the evolutionary dynamics of enfuvirtide resistance-associated mutations in HIV-1 gp41 are complex. Our data show that the initial mutant virus population, selected very rapidly under the pressure of an enfuvirtide-containing regimen, is continuously replaced by other new mutant quasispecies during long-term salvage treatment. This dynamic shift in the HIV population might indicate that the initial population containing mutations 36 or 38 confer a fitness disadvantage to the virus.
The authors would like to thank E. Grau from the IrsiCaixa Foundation for her laboratory support and N. Pérez from the Universitat Politècnica de Catalunya for her statistical analysis assistance. C. Gutiérrez from the Hospital Ramon y Cajal and M. Fuster from Hospital de la Santa Creu i Sant Pau are gratefully acknowledged for providing clinical samples. Special thanks go to the patients attending the HIV Clinical Unit at the Hospital Germans Trias i Pujol. The authors would also like to thank D.R. Kuritzkes and J. Blanco for their critical reading of the manuscript.
Sponsorship: This study was partly supported by the Fundació IRSICaixa (Institut de Recerca de la SIDA- Caixa), the Fondo de Investigación Sanitaria, project 04/00271 and by the Spanish AIDS network Red Temática Cooperativa de Investigación en SIDA (Red G03/173). C.C. (FIS 04/00271) and J.M-P. were supported by the Institute for Research on Health Sciences Germans Trias i Pujol in collaboration with the Spanish Health Department and the Catalan Health Department, respectively.
1. Wild CT, Shugars DC, Greenwell TK, McDanal CB, Matthews TJ. Peptides corresponding to a predictive alpha-helical domain of human immunodeficiency virus type 1 gp41 are potent inhibitors of virus infection. Proc Natl Acad Sci U S A 1994; 91:9770–9774.
2. Kilgore NR, Salzwedel K, Reddick M, Allaway GP, Wild CT. Direct evidence that C-peptide inhibitors of human immunodeficiency virus type 1 entry bind to the gp41 N-helical domain in receptor-activated viral envelope. J Virol 2003; 77:7669–7672.
3. Lazzarin A, Clotet B, Cooper D, Reynes J, Arasteh K, Nelson M, et al
. Efficacy of enfuvirtide in patients infected with drug-resistant HIV-1 in Europe and Australia. N Engl J Med 2003; 348:2186–2195.
4. Lalezari JP, Henry K, O'Hearn M, Montaner JS, Piliero PJ, Trottier B, et al
. Enfuvirtide, an HIV-1 fusion inhibitor, for drug-resistant HIV infection in North and South America. N Engl J Med 2003; 348:2175–2185.
5. Wei X, Decker JM, Liu H, Zhang Z, Arani RB, Kilby JM, et al
. Emergence of resistant human immunodeficiency virus type 1 in patients receiving fusion inhibitor (T-20) monotherapy. Antimicrob Agents Chemother 2002; 46:1896–1905.
6. Poveda E, Rodes B, Toro C, Martin-Carbonero L, Gonzalez-Lahoz J, Soriano V. Evolution of the gp41 env region in HIV-infected patients receiving T-20, a fusion inhibitor. AIDS 2002; 16:1959–1961.
7. Hanna SL, Yang C, Owen SM, Lal RB. Variability of critical epitopes within HIV-1 heptad repeat domains for selected entry inhibitors in HIV-infected populations worldwide. AIDS 2002; 16:1603–1608.
8. Zollner B, Feucht HH, Schroter M, Schafer P, Plettenberg A, Stoehr A, Laufs R. Primary genotypic resistance of HIV-1 to the fusion inhibitor T-20 in long-term infected patients. AIDS 2001; 15:935–936.
9. Xu L, Hue S, Taylor S, Ratcliffe D, Workman JA, Jackson S, et al
. Minimal variation in T-20 binding domain of different HIV-1 subtypes from antiretroviral-naive and -experienced patients. AIDS 2002; 16:1684–1686.
10. Poveda E, Rodes B, Labernardiere JL, Benito JM, Toro C, Gonzalez-Lahoz J, et al
. Evolution of genotypic and phenotypic resistance to Enfuvirtide in HIV-infected patients experiencing prolonged virologic failure. J Med Virol 2004; 74:21–28.
11. Marcelin AG, Reynes J, Yerly S, Ktorza N, Segondy M, Piot JC, et al
. Characterization of genotypic determinants in HR-1 and HR-2 gp41 domains in individuals with persistent HIV viraemia under T-20. AIDS 2004; 18:1340–1342.
12. Sista PR, Melby T, Davison D, Jin L, Mosier S, Mink M, et al
. Characterization of determinants of genotypic and phenotypic resistance to enfuvirtide in baseline and on-treatment HIV-1 isolates. AIDS 2004; 18:1787–1794.
13. Menzo S, Castagna A, Monachetti A, Hasson H, Danise A, Carini E, et al
. Genotype and phenotype patterns of human immunodeficiency virus type 1 resistance to enfuvirtide during long-term treatment. Antimicrob Agents Chemother 2004; 48:3253–3259.
14. Baldwin CE, Sanders RW, Deng Y, Jurriaans S, Lange JM, Lu M, Berkhout B. Emergence of a drug-dependent human immunodeficiency virus type 1 variant during therapy with the T20 fusion inhibitor. J Virol 2004; 78:12428–12437.
15. Xu L, Pozniak A, Wildfire A, Stanfield-Oakley SA, Mosier SM, Ratcliffe D, et al
. Emergence and evolution of enfuvirtide resistance following long-term therapy involves heptad repeat 2 mutations within gp41. Antimicrob Agents Chemother 2005; 49:1113–1119.
16. Lu J, Sista P, Giguel F, Greenberg M, Kuritzkes DR. Relative replicative fitness of human immunodeficiency virus type 1 mutants resistant to enfuvirtide (T-20). J Virol 2004; 78:4628–4637.
17. Kinomoto M, Yokoyama M, Sato H, Kojima A, Kurata T, Ikuta K, et al
. Amino acid 36 in the human immunodeficiency virus type 1 gp41 ectodomain controls fusogenic activity: implications for the molecular mechanism of viral escape from a fusion inhibitor. J Virol 2005; 79:5996–6004.
18. Reeves JD, Lee FH, Miamidian JL, Jabara CB, Juntilla MM, Doms RW. Enfuvirtide resistance mutations: impact on human immunodeficiency virus envelope function, entry inhibitor sensitivity, and virus neutralization. J Virol 2005; 79:4991–4999.
19. Nameki D, Kodama E, Ikeuchi M, Mabuchi N, Otaka A, Tamamura H, et al
. Mutations conferring resistance to human immunodeficiency virus type 1 fusion inhibitors are restricted by gp41 and Rev-responsive element functions. J Virol 2005; 79:764–770.
20. Yahi N, Tamalet C, Tourres C, Tivoli N, Ariasi F, Volot F, et al
. Mutation patterns of the reverse transcriptase and protease genes in human immunodeficiency virus type 1-infected patients undergoing combination therapy: survey of 787 sequences. J Clin Microbiol 1999; 37:4099–4106.
21. Sugiura W, Matsuda Z, Yokomaku Y, Hertogs K, Larder B, Oishi T, et al
. Interference between D30N and L90M in selection and development of protease inhibitor-resistant human immunodeficiency virus type 1. Antimicrob Agents Chemother 2002; 46:708–715.