| |
Fuzeon Drug Resistance
|
| |
| |
"Impact of HIV-1 gp41 Amino Acid Substitutions Selected during Enfuvirtide Treatment on gp41 Binding and Antiviral Potency of Enfuvirtide In Vitro"
Journal of Virology, October 2005, p. 12447-12454, Vol. 79, No. 19
M. Mink,1 S. M. Mosier,1 S. Janumpalli,1 D. Davison,1 L. Jin,1 T. Melby,1 P. Sista,1 J. Erickson,1 D. Lambert,1 S. A. Stanfield-Oakley,1 M. Salgo,2 N. Cammack,3 T. Matthews,1,{dagger} and M. L. Greenberg1*
Trimeris Inc., Morrisville, North Carolina,1 Roche, Nutley, New Jersey,2 Roche, Palo Alto, California3
ABSTRACT
Enfuvirtide (ENF), a novel human immunodeficiency virus type 1 (HIV-1) fusion inhibitor, has potent antiviral activity against HIV-1 both in vitro and in vivo. Resistance to ENF observed after in vitro passaging was associated with changes in a three-amino-acid (aa) motif, GIV, at positions 36 to 38 of gp41. Patients with ongoing viral replication while receiving ENF during clinical trials acquired substitutions within gp41 aa 36 to 45 in the first heptad repeat (HR-1) of gp41 in both population-based plasma virus sequences and proviral DNA sequences from isolates showing reduced susceptibilities to ENF.
To investigate their impact on ENF susceptibility, substitutions were introduced into a modified pNL4-3 strain by site-directed mutagenesis, and the susceptibilities of mutant viruses and patient-derived isolates to ENF were tested.
In general, susceptibility decreases for single substitutions were lower than those for double substitutions, and the levels of ENF resistance seen for clinical isolates were higher than those observed for the site-directed mutant viruses.
The mechanism of ENF resistance was explored for a subset of the substitutions by expressing them in the context of a maltose binding protein chimera containing a portion of the gp41 ectodomain and measuring their binding affinity to fluorescein-labeled ENF. Changes in binding affinity for the mutant gp41 fusion proteins correlated with the ENF susceptibilities of viruses containing the same substitutions.
The combined results support the key role of gp41 aa 36 to 45 in the development of resistance to ENF and illustrate that additional envelope regions contribute to the ENF susceptibility of fusion inhibitor-naïve viruses and resistance to ENF.
DISCUSSION
This study examined the effect of ENF-selected amino acid substitutions in gp41 on the susceptibility of viruses to inhibition by ENF and the binding of ENF to gp41-derived HR-1 regions containing several of the same substitutions. Such an approach allowed us to evaluate a mechanistic basis for alterations in ENF susceptibility conferred by these substitutions. The effects of substitutions in gp41 amino acids 29 to 45 on ENF sensitivity were assessed in the following three ways: (i) through site-directed mutagenesis of a modified laboratory strain of HIV-1 (NL4-3/G36D), (ii) through an evaluation of primary isolates recovered from patients undergoing chronic treatment with ENF during phase II clinical studies, and (iii) through site-directed mutagenesis of virus envelopes using a pseudotype reporter virus assay. NL4-3-based viruses were engineered to harbor many of the substitutions in the HR-1 region of HIV gp41 that had been observed in the plasma viruses of patients enrolled in ENF phase I/II and phase II clinical trials and in their recovered virus isolates displaying reduced susceptibilities to ENF. We observed that several single amino acid changes in gp41 aa 36 to 45 conferred a >10-fold reduction in NL4-3 susceptibility to ENF (G36E, V38A, V38E, Q40H, and N43D). Reductions in NL4-3 susceptibility conferred by single substitutions ranged as high as 1,100-fold for the V38E mutation. In addition, we found that the N42S polymorphism, which occurs in approximately 16% of fusion inhibitor-naïve viruses, increased NL4-3 sensitivity to ENF inhibition about twofold, consistent with the observations of Melby et al., who found that this polymorphism was associated with lower ENF IC50s for baseline virus envelopes from patients participating in ENF phase III trials (submitted). Substitutions at two positions in aa 36 to 45 generally caused greater reductions in NL4-3 susceptibility to ENF than did single substitutions. Similarly, substitutions at two positions in gp41 were generally associated with larger changes in the ENF IC50 for on-treatment virus isolates than those for on-treatment isolates carrying only one of the changes. The effects of mutations within gp41 aa 36 to 45 on ENF susceptibility and their appearance in the vast majority of viruses from patients experiencing virological failure during ENF therapy confirm the key role of this region in the development of resistance to ENF.
Initial studies of the mechanism of ENF action indicated that ENF acts through binding to the HR-1 region of gp41 (4, 24). The binding affinities of ENF for HR-1-containing fusion proteins engineered to carry a subset of ENF resistance-associated substitutions correlated in rank order with the ENF susceptibilities of mutant NL4-3 viruses with the same substitutions. This observation suggests a partial mechanistic basis for ENF resistance. A decreased susceptibility to ENF may in part be due to a decreased binding affinity of ENF for the altered target HR-1 regions of gp41. This inference is consistent with the recent studies of Nameki et al. (21), who found that an I37K mutation reduced the binding of a related fusion inhibitor peptide, C34, to an HR-1 target. In addition to that study, the present study and other recent reports (2, 33) indicate that the mechanism of ENF resistance is not fully accounted for by decreased binding of ENF to the HR-1 target nor restricted to substitutions in gp41 aa 36 to 45.
The levels of resistance seen for NL4-3-based mutants were generally lower than those for clinical isolates bearing the same mutations. There was also a wide variation in the ENF susceptibilities of clinical isolates bearing a given mutation. These observations suggest several alternatives that could influence resistance to ENF. Among them are the following: (i) the envelope background may affect the magnitude of the shift in susceptibility to ENF conferred by substitutions within gp41 aa 36 to 45, (ii) changes occurring in the HR-2 region of gp41 may contribute to ENF resistance, and (iii) changes occurring in gp120 in response to ENF could influence ENF resistance. In this study, gp120 sequences were not obtained from plasma viruses or from the initial baseline or on-treatment clinical isolates, and thus we cannot directly address their potential impact on the results reported here. However, a series of statistical and modeling analyses of entire gp160 sequences and ENF susceptibilities of patients participating in the phase III clinical trials of ENF have been conducted by C. Su and colleagues (unpublished data) to identify changes occurring in the viral envelope during ENF therapy that are significantly associated with alterations in ENF susceptibility. In their study, analysis-of-variance models indicated that approximately 90% of the variations in phenotypic sensitivity were accounted for by changes in gp41 aa 36 to 45. Nevertheless, they did observe that changes at three positions in gp120 and five positions in gp41 outside of aa 36 to 45 were associated with significant reductions in ENF susceptibility, leaving open the possibility that changes in gp120 could influence resistance to ENF.
Derdeyn et al. (5, 6) have suggested that coreceptor specificity defined by the V3 loop may modulate HIV's susceptibility to ENF. However, their findings were not confirmed in a separate study analyzing the ENF susceptibilities of 111 virus isolates using a variety of cell-based assay systems (M. L. Greenberg, C. B. McDanal, S. Stanfield-Oakley, et al., Abstr. 8th Conf. Retrovir. Opportunistic Infect., abstr. 473, 2001). No differences in susceptibility were seen between X4-, R5-, or dual-tropic viruses from patients or between serially obtained R5 and X4 isolates from the same patient. In the phase III studies of ENF, virologic responses were equivalent for patients with baseline isolates exhibiting either R5, X4, or dual tropism (Melby et al., submitted), indicating that the coreceptor tropism at baseline is not a significant determinant of the therapeutic response to ENF.
Studies by Heil et al. and Stanfield-Oakley et al. suggested that both the HR-1 and HR-2 regions of the viral envelope may contribute to ENF susceptibility (M. Heil, J. M. Decker, J. Chen, et al., Abstr. 2nd Int. AIDS Soc. Conf. HIV Pathogenesis Treatment, abstr. 801, 2003; S. A. Stanfield-Oakley, J. Jeffrey, C. B. McDanal, et al., abstr. 12th Int. HIV Drug Resistance Workshop, abstr. 56, 2003). Those findings were extended by Xu et al. (33), who found that an S138A HR-2 mutation could enhance resistance to ENF, and by Baldwin et al., who reported that an N-to-K mutation in HR-2 (equivalent to N126K in HXB2) contributed to ENF resistance and conferred enhanced fusogenicity on the virus (2). Nameki et al. (21) found that the N126K HR-2 mutation led to tighter binding between an HR-2 peptide and an HR-1 target. Taken together, these studies suggest another mechanism of resistance to ENF involving HR-2. In this case, alterations in the HR-2 region may provide a competitive advantage over ENF for binding to the mutant HR-1 coiled coil, either through increases in binding affinities or by shortening the kinetic window for six-helix bundle formation, as posited by Reeves et al. (23).
Finally, our studies also demonstrate that the envelope background in which an ENF-selected mutation arises can influence the degree of resistance to ENF that is conferred. This finding adds to our previous observation that the envelope background of fusion inhibitor-naïve viruses influences the sensitivity to ENF (27). Previous reports have documented that fusion inhibitor-naïve isolates display a wide range of sensitivities to ENF, despite the fact that the vast majority have the consensus sequence GIVQQQNNLL for gp41 aa 36 to 45. Those findings indicate that regions outside of this motif can influence the sensitivity to ENF. The current study extends the influence of the envelope context to the effects of HR-1 mutations on ENF resistance. The data presented here clearly demonstrate that substitutions in the region of aa 36 to 45 of gp41 are primary determinants for the development of viral resistance to ENF. Overall, these results lend strong support to the mechanism of ENF action targeting the HR-1 region as a competitive inhibitor that acts to block the interactions of the viral HR-1 and HR-2 regions required for the fusion of viral and target cell membranes. These studies also indicate that an important mechanism of resistance to ENF involves decreases in ENF binding to the HR-1 target sequences that are altered during exposure to ENF. In addition, there are other mechanisms contributing to ENF resistance. These include alterations in HR-2 that may increase binding to HR-1 or shorten the kinetic window of six-helix bundle formation, changes in gp120 that act in ways that are unclear to affect ENF susceptibility, and influences of the envelope background on resistance to ENF. Further studies will be required to more fully understand these mechanisms and the influence of the envelope's genetic context.
INTRODUCTION
The emergence of drug-resistant viruses remains one of the most serious impediments to successful antiretroviral therapy for human immunodeficiency virus (HIV)-infected individuals. All currently available antiretrovirals belong to one of four mechanistic classes: nucleoside/nucleotide reverse transcriptase inhibitors, nonnucleoside reverse transcriptase inhibitors, protease inhibitors, and the newest class, fusion inhibitors. Cross-resistance within the nucleoside/nucleotide reverse transcriptase inhibitor, nonnucleoside reverse transcriptase inhibitor, and protease inhibitor classes is extensive, and treatment options for multidrug-experienced patients are often severely limited. Consequently, there is a need for new antiretrovirals that act against novel targets and will not be affected by viral resistance to earlier classes of drugs; the process of viral entry represents one such target.
HIV type 1 (HIV-1) entry occurs via a multistep process involving the interaction of trimeric envelope (Env) glycoprotein complexes with cellular receptors and subsequently with the target cell membrane. The process ultimately brings the viral and target cell membranes into contact, leading to the formation of a fusion pore and viral entry. In the first part of the process, HIV-1 glycoprotein 120 (gp120) attaches to the cellular CD4 receptor, resulting in conformational changes that expose the coreceptor-binding domain and the trimeric coiled coil formed by the heptad repeat 1 (HR-1) regions of gp41 (26). Binding to cellular coreceptors (CCR5 or CXCR4) then triggers the release of gp120 from the viral envelope glycoprotein complex. In the second phase of entry, a putative hydrophobic fusion peptide at the N terminus of gp41 exposed by the release of gp120 inserts into the target cell membrane. At this stage, the HR-1 regions of the three gp41 molecules are thought to exist as a trimeric coiled-coil structure, and three cognate HR-2 domains bind along the grooves in the HR-1 coiled coil to form a six-helix bundle, thereby facilitating contact between the viral and target cell membranes. Finally, a fusion pore between the viral and target cell membranes that is permissive for viral entry is thought to be created by the aggregation of multiple gp41 trimers that have converted to the six-helix bundle conformation (9, 30).
Several entry inhibitors have been described in the literature and have entered clinical evaluation, including the CD4 binding antagonists PRO 542 (11) and BMS 806 (19), the CXCR4 binding inhibitor AMD3100 (7), the CCR5 binding inhibitors SCH-C and -D (28), UK-427,857 (25), and AK602/GW873140 (20), and the peptide fusion inhibitors enfuvirtide (ENF, formerly T-20) and T-1249 (8, 13). ENF, which has been approved by a number of health authorities worldwide, is a 36-amino-acid (aa) synthetic peptide whose primary sequence was derived from the gp41 HR-2 region of HIV-1LAI (32). ENF has potent antiviral activity against HIV-1 in vitro (32) and during clinical trials has demonstrated significant virological and immunological benefits when used in combination with other antiretrovirals in treatment-experienced patients (15-18).
An understanding of mechanisms of resistance and the potential for intra- and interclass cross-resistance is an important aspect of antiviral drug development. In vitro passaging of HIV-1 IIIB and NL4-3 viruses in the presence of increasing concentrations of ENF resulted in the selection of ENF-resistant viruses carrying mutations in the N-terminal HR-1 region of HIV-1 gp41 (24). Clones derived from these viral populations revealed that aa substitutions at positions 36 to 38 of gp41 were responsible for the reduction in susceptibility to ENF. The involvement of this three-aa motif in conferring viral resistance to ENF was also confirmed by site-directed mutagenesis experiments. These studies suggested that the 36S, 36D, 37T, and 38M mutations could all contribute to ENF resistance but that these mutations had to be present in pairs in order to generate a resistant phenotype. These findings were later confirmed by Derdeyn et al. (5, 6). They constructed mutant viruses based on HIV-1 NL4-3 with mutations to aa 36 to 38 and showed that viruses carrying 36D or 36S plus 38M had significant resistance to ENF. The selection of resistance mutations in aa 36 to 38 of the HR-1 region of gp41 is consistent with the hypothesis that ENF exerts its anti-HIV activity by binding to this region (4). This hypothesis is further supported by the fact that radiolabeled ENF could bind a peptide representing residues 29 to 75 of wild-type HR-1 but not to the equivalent peptide carrying the SIM sequence at positions 36 to 38. In addition, the 29- to 75-residue wild-type peptide could competitively block the antiviral activity of ENF (24), consistent with earlier studies (4).
Studies of viral resistance to ENF carried out during early clinical trials have suggested that changes in the gp41 genotype conferring resistance to ENF may cover a wider aa region than the aa-36-to-38 motif identified in the initial in vitro selection experiments (24, 27, 29; M. Mink, M. L. Greenberg, S. Mosier, et al., Abstr. 11th Int. Drug Resistance Workshop, abstr. 22, 2002). The aim of the present study was to further investigate this possibility by examining the effects of mutations arising in vivo in response to therapy with ENF on ENF susceptibility in vitro and on the binding of ENF to HR-1.
|
|
| |
| |
|
|
|
