4Transfusion transmitted Diseases Center, Institute of Blood Transfusion, Chinese Academy of Medical Science, #26 Hua-Cai Road, Long-Tan-Si Industrial Park, Chenghua District, Chengdu, 610052 Sichuan Province, P. R. China
Received Date: October 18, 2011; Accepted Date: December 02, 2011; Published Date: December 05, 2011
Citation: Maeda R, Yoshimura K, Miyamoto F, Kodama E, Harada S, et al. (2011) In vitro and In vivo Resistance to Human Immunodeficiency Virus Type 1 Entry Inhibitors. J AIDS Clinic Res S2:004. doi:10.4172/2155-6113.S2-004
Copyright: © 2011 Maeda R, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of AIDS & Clinical Research
Viral entry is one of the most important targets for the efficient treatment of Human immunodeficiency virus type 1 (HIV-1)-infected patients. The entry process consists of multiple molecular steps: attachment of viral gp120 to CD4, interaction of gp120 with CCR5 or CXCR4 co-receptors, and gp41-mediated fusion of the viral and cellular membranes. Understanding the sequential steps of the entry process has enabled the production of various antiviral drugs to block each of these steps. Currently, the CCR5 inhibitor, maraviroc, and the fusion inhibitor, enfuvirtide, are clinically available. However, the emergence of HIV-1 strains resistant to entry inhibitors, as commonly observed for other classes of antiviral agents, is a serious problem. In this review, we describe a variety of entry inhibitors targeting different steps of viral entry and escape variants that are generated in vitro and in vivo.
CD4-gp120 binding inhibitor; CCR5 antagonist; CXCR4 antagonist; Fusion inhibitor; Resistance; HIV-1
The development of chemotherapy with antiretroviral agents has reduced the morbidity and mortality of Human immunodeficiency virus type 1 (HIV-1)-infected individuals. Successful treatment of HIV- 1-infected patients using chemotherapy is partly due to a combination of different classes of antiviral agents against the viral protease or reverse transcriptase. However, successful eradication of the virus from infected individuals has not been achieved by antiviral treatment, and is often limited by the emergence of drug-resistant HIV-1 strains [1-3]. These problems highlight the need to develop novel anti-HIV-1 drugs that target different steps of the viral replication process. Viral entry is currently one of the most attractive targets for the development of new drugs to control HIV-1 infection. Viral entry proceeds through Env (gp120, gp41)-mediated membrane fusion, and consists of sequential steps: (i) attachment of viral gp120 to the CD4 receptor; (ii) binding of gp120 to CCR5 or CXCR4 co-receptors; and (iii) fusion of the viral and cellular membranes (Figure 1). A large number of inhibitors targeting different steps of the viral entry process have been developed, including peptides/peptide mimics, small molecules, and monoclonal antibodies (MAb).
Enfuvirtide (also known as T-20) was the first of a new class of drugs known as fusion inhibitors, which was approved by the U.S. Food and Drug Administration (FDA) in 2003. Approval was given for the use of this drug in combination with other anti-HIV-1 medications to treat advanced HIV-1 infection in adults and children aged six years and older. The drug is an antiviral peptide that prevents HIV-1 entry by blocking gp41-mediated fusion [4-6]. Small compounds that can bind to the pockets of the extracellular loops of a coreceptor are expected to be potent antiviral agents. Several small-molecule CCR5 inhibitors have progressed through clinical development . Maraviroc [8,9], a CCR5 antagonist, is the second entry inhibitor approved by the FDA in 2007 for treatment-experienced patients infected with a CCR5-tropic (R5-tropic) virus. Extensive research is currently underway to develop the next generation of entry inhibitors, however, the emergence of viral strains resistant to entry inhibitors, as well as other classes of antiviral agents, has been reported In vitro and In vivo [7,10]. In this review, we describe the current status of In vitro and In vivo resistance to HIV-1 entry inhibitors.
Resistance to CD4-gp120 binding inhibitors
Inhibition of CD4-gp120 binding: Entry of HIV-1 into target cells is mediated by the trimeric envelope glycoprotein complex, each monomer consisting of a gp120 exterior envelope glycoprotein and a gp41 transmembrane envelope glycoprotein . Attachment of HIV-1 to the cell is initiated by the binding of gp120 to its primary CD4 receptor, which is expressed on the surface of the target cell. The gp120-CD4 interaction induces conformational changes in gp120 that facilitate binding to additional coreceptors (for example, CCR5 or CXCR4). Attachment inhibitors are a novel class of compounds that bind to gp120 and interfere with its interaction with CD4 . Thus, these agents can prevent HIV-1 from attaching to the CD4+ T cell and block infection at the initial stage of the viral replication cycle (Figure 1). There are two primary types of HIV-1 attachment inhibitors: nonspecific attachment inhibitors and CD4-gp120 binding inhibitor .
In this section, we focus on the CD4-gp120 binding inhibitors, the soluble form of CD4 (sCD4), a fusion protein of CD4 with Ig (PRO542), a monoclonal anti-CD4 antibody (Ibalizumab, formerly TNX-355), CD4 binding site (CD4bs) monoclonal antibodies (b12 and VRC01), small-molecule HIV-1 attachment inhibitors (BMS-378806 and BMS- 488043), and a new class of small-molecule CD4 mimics (NBD-556 and NBD-557) and a natural small bioactive molecule (Palmitic acid) (Figure 2). We also describe the resistance profiles against these CD4- gp120 binding inhibitors In vivo and/or In vitro.
Soluble CD4 (sCD4) and PRO542: In the late 1980s, various recombinant, soluble proteins derived from the N-terminal domains of CD4 were shown to be potent inhibitors of laboratory strains of HIV-1 . Based on the potential of sCD4 to inhibit HIV-1 infection In vitro, this protein was tested for clinical efficacy in HIV-1-infected individuals; however, no effect on plasma viral load was observed . Further examination revealed that doses of sCD4 significantly higher than those achieved in the clinical trial were required to neutralize primary clinical isolates of HIV-1, in contrast to the relatively sensitive, laboratoryadapted strains .
The first report of sCD4-resistant variants induced by In vitro selection showed that the resistant variant had a single mutation (M434T) in the C4 region . During selection with sCD4, it was also reported that, seven mutations (E211G, P212L, V255E, N280K, S375N, G380R, and G431E) appeared during In vitro passage . Further, a recombinant clone containing a V255E mutation was found to be highly resistant to sCD4 compared with the wild-type virus (114-fold higher 50% inhibitory concentration [IC50] value). To determine the mutation profiles obtained during In vitro selection with sCD4, the atomic coordinates of the crystal structure of gp120 bound to sCD4 was retrieved from public protein structure database (PDB entry: 1RZJ). From these analyses, it was determined that almost all the described resistance mutations were located the inside the CD4-binding cavity of gp120 .
Recently, a novel recombinant antibody-like fusion protein (CD4- 1gG2; PRO542) was developed in which the Fv portions of both the heavy and light chains of human IgG2 were replaced with the D1D2 domains of human CD4 . PRO542 was shown to broadly and potently neutralize HIV-1 subtype B isolates, and was also able to neutralize strains from non-B isolates with the same breadth and potency as for subtype B strains. PRO542 blocks attachment and entry of the virus into CD4+ target cells and were mainly developed for the prevention and transmission of HIV-1 through external application agents, such as microbicides.
Ibalizumab (TNX-355): Monoclonal anti-CD4 antibodies block the interaction between gp120 and CD4 and, therefore, inhibit viral entry . Ibalizumab (formerly TNX-355) was a first-in-class, monoclonal antibody inhibitor of CD4-mediated HIV-1 entry . By blocking CD4-dependent HIV-1 entry, ibalizumab was shown to be active against a broad spectrum of HIV-1 isolates, including recombinant subtypes, as well as both CCR5-tropic and CXCR4-tropic HIV-1 isolates. Many clinical trials with HIV-1-infected patients have demonstrated the antiviral activity, safety, and tolerability of ibalizumab. A nine-week phase Ib study investigating the addition of ibalizumab monotherapy to failing drug regimens showed transient reductions in HIV-1 viral loads and the evolution of HIV-1 variants with reduced susceptibility to ibalizumab. Further, clones with reduced susceptibility to ibalizumab contained fewer potential N-linked glycosylation sites (PNGSs) within the V5 region of gp120. Reduction in ibalizumab susceptibility due to the loss of V5 PNGSs was confirmed by site-directed mutagenesis .
Monoclonal antibodies, b12 and VRC01: Several broadly neutralizing MAbs isolated from HIV-1-infected individuals define conserved epitopes on the HIV-1 Env. These include the membrane proximal external region of gp41 targeted by MAbs 4E10 and 2F5 ; the carbohydrate- specific outer domain epitope targeted by 2G12 ; a V2- V3-associated epitope targeted by PG9/PG16 ; and the CD4bs  targeted by b12 and VRC01. The CD4bs overlaps with the conserved region on gp120 that is involved in the engagement of CD4. The prototypical CD4bs-directed MAb, b12, neutralizes around 40% of primary isolates, and its structure (in complex with the core of gp120) has been defined . However, Mo et al.  reported the first resistant variant induced by In vitro selection with b12 that showed a P369L mutation in the C3 region of HIV-1JRCSF. Further, several b12-resistant viruses commonly display an intact b12 epitope on the gp120 subunits , suggesting that quaternary packing of Env also confers resistance to b12.
A recently described CD4bs-directed MAb, VRC01, had been shown to be able to neutralize over 90% of diverse HIV-1 primary isolates . The structure of VRC01 in complex with the gp120 core reveals that the VRC01 heavy chain binds to the gp120 CD4bs in a manner similar to that of CD4 . The gp120 loop D and V5 regions contain substitutions uniquely affecting VRC01 binding, but not b12 or CD4-Ig binding. In contrast to the interaction of CD4 or b12 with the HIV-1 Env, occlusion of the VRC01 epitope by quaternary constraints was not a major factor limiting neutralization. Interestingly, many Ala substitutions at non-contact residues increased the potency of CD4- or b12-mediated neutralization; however, few of these substitutions enhanced VRC01-mediated neutralization . This study suggests that VRC01 approaches its cognate epitope on the functional spike with less steric hindrance than b12 and, surprisingly, with less hindrance than the soluble form of CD4 itself. These differences might be related to the distinctly different angle of approach to the CD4bs employed by VRC01, in contrast to the more loop-proximal approach employed by CD4 and b12.
BMS-378806 and BMS-488043: BMS-378806 (Figure 2) is a recently identified small-molecule HIV-1 attachment inhibitor with good antiviral activity and pharmacokinetic properties . BMS-378806 binds directly to gp120 with a stoichiometry of approximately 1:1 and with a binding affinity similar to that of soluble CD4. The potential BMS- 378806 target site was localized to a specific region within the CD4 binding pocket of gp120 using HIV-1 gp120 variants carrying either compound-selected resistant substitutions or gp120-CD4 contact site mutations . M426L (C4) and M475I (V5) substitutions located at or near gp120/CD4 contact sites were shown to confer high levels of resistance to the In vitro mutated HIV-1 variants, suggesting that the CD4 binding pocket of gp120 was the antiviral target. M434I and other secondary changes (V68A and I595F) also affect the drug susceptibility of recombinant viruses, presumably by influencing the gp120 conformation . BMS-378806 (Figure 2) exhibited decreased, but still significant activity against subtype C viruses, low activity against viruses from subtypes A and D, and poor or no activity against subtypes E, F, G, and Group O viruses .
BMS-488043 (Figure 2) is a novel and unique small-molecule that inhibits the attachment of HIV-1 to CD4+ lymphocytes. BMS-488043 exhibits potent antiviral activity against macrophage-, T-cell-, and dual-tropic HIV-1 laboratory strains (subtype B) and potent antiviral activity against a majority of subtype B and C clinical isolates . Data from a limited number of clinical isolates showed that BMS-488043 exhibited a wide range of activity against the A, D, F, and G subtypes, with no activity observed against three subtype AE isolates . The antiviral activity, pharmacokinetics, viral susceptibility, and safety of BMS-488043 were evaluated in an eight-day monotherapy trial that demonstrated significant reductions in viral load. To examine the effects of BMS-488043 monotherapy on HIV-1 sensitivity, phenotypic sensitivity assessment of baseline and post-dosing (day 8) samples were performed. The analyses revealed that four subjects showed emergent phenotypic resistance. Population sequencing and sequence determination of the cloned envelope genes revealed five gp120 mutations at four loci (V68A, L116I, S375I/N, and M426L) associated with BMS- 488043 resistance; the most common (substitution at the 375 locus) located near the CD4 binding pocket .
NBD-556 and NBD-557: Targeting the functionally important and conserved CD4bs on HIV-1 gp120 represents an attractive potential approach to HIV-1 therapy or prophylaxis. Recently, a new class of small-molecule CD4 mimics was identified [36-38]. These compounds, which include the prototypic compound, NBD-556, and its derivatives, mimic the effects of CD4 by inducing the exposure of the coreceptor- binding site on gp120 [17,39]. NBD-556 and -557 (Figure 2) show potent cell fusion and virus-cell fusion inhibitory activity at low (micromolar) concentrations. A mechanistic study showed that both compounds target viral entry by inhibiting the binding of gp120 to its cellular receptor, CD4. A surface plasmon resonance study showed that these compounds bind to unliganded HIV-1 gp120, but not to CD4 . Another recent study identified NBD-analogs as CD4 mimetics that were used for the prophylaxis and treatment of HIV-1 infection . These compounds inhibited HIV-1 transmission by inhibiting the binding of the natural ligand, CD4, and prematurely triggering the envelope glycoprotein to undergo irreversible conformational changes. NBD-556 binds to the F43 cavity, which is formed by binding of gp120 to the CD4 receptor in a highly conserved manner [17,39].
Recently, our group reported that NBD-556 has potent neutralizing antibody-enhancing activity toward plasma antibodies that cannot access neutralizing epitopes hidden within the trimeric Env, such as gp120-CD4 induced epitope (CD4i) and anti-V3 antibodies . Therefore, to investigate the binding site of NBD-556 on gp120, we in duced HIV-1 variants that were resistant to NBD-556 In vitro. Two amino acid substitutions (S375N in C3 and A433T in C4) were identified at passage 21 in the presence of 50 μM NBD-556. The profiles of the resistance mutations after selection with NBD-556 and sCD4 were very similar with regard to their three-dimensional positions.
Elucidation of the detailed molecular mechanisms governing the interaction between gp120 and NBD compounds will enable the optimization and evaluation of this strategy in more complex biological models of HIV-1 infection. Consequently, we will continue to synthesize NBD analogs and search for drugs with greater potency to change the tertiary structure of the envelope glycoproteins and reduce host cytotoxicity [40,41].
Palmitic acid : Previous studies with whole Sargassum fusiforme (S. fusiforme) extract and with the bioactive SP4-2 fraction demonstrated inhibition of HIV-1 infection in several primary and transformed cell lines . Palmitic acid (PA), which was isolated from the SP4-2 bioactive fraction, specifically block productive X4 and R5-tropic HIV-1 infection . PA occupies a novel hydrophobic cavity on the CD4 receptor that is constrained by amino acids F52-to-L70 , which encompass residues that have been previously identified as a region critical for gp120 binding. PA is mainly developed as microbicides .
Resistance to CCR5 antagonists
CCR5 antagonists: The binding of HIV-1 to CD4 molecules induces conformational change in gp120, resulting in the recognition of either CCR5 or CXCR4 as a coreceptor for HIV-1 (Figure 1). It has been shown that CCR5-utilizing HIV-1 (R5 virus) is associated with human- to-human transmission that predominate throughout the infection, while CXCR4-utilizing HIV-1 (X4 virus) emerges during the late stage of infection in approximately half of HIV-1-infected individuals and is associated with disease progression . Most strikingly, it had been shown that homozygous individuals having a 32-bp deletion in the CCR5 coding region (CCR5Δ32) were found to be resistant to R5 HIV-1 and remained apparently healthy [47,48]. These findings suggested that CCR5 would be an attractive therapeutic target for treating HIV-1 infection, although it is a host factor. Several small molecule compounds have been developed and were found to bind CCR5 and inhibit R5 virus replication [49-53]. Molecular studies using CCR5 mutants indicated that these compounds bind to a cavity formed by transmembrane helices of CCR5, and thereby inducing the conformational change in an allosteric manner that is not recognized by gp120 of HIV-1 [54-58]. Among these, TAK-779 (Figure 3) was the first compound developed  that could inhibit not only HIV-1 infection, but also binding of RANTES (CCR5 ligand) to CCR5-expressing cells at nanomolar concentrations, but was terminated due to poor oral bioavailability. Maraviroc (MVC, UK427, 857) (Figure 3), however, has been approved and used in the clinic for the treatment of HIV-1 infection . Another promising drug, vicriviroc (VCV, SCH-D, SCH- 417690) (Figure 3), recently completed phase III trials but has not yet been approved .
Figure 3: Profile of CCR5 antagonist-resistant mutants. The CCR5 antagonist-resistant mutants were isolated in vitro and in vivo across different subtypes of HIV-1. Resistance-related mutations were found in the V3 and non-V3 regions including the C2, V4, C4, and gp41. Chemical structures of representative CCR5 antagonists are shown.
Resistance to CCR5 antagonists: Although CCR5 antagonists target a host cell receptor, the In vitro [59-64] and In vivo [65-67] emergence of viruses resistant to CCR5 antagonists in different subtypes has been reported, as shown in Figure 3. The most intuitive mechanism of resistance to CCR5 antagonists is likely to be the acquisition of CXCR4 use or selection of minority variants of CXCR4- or dual/mixed-tropic viruses [61,68-70]. Numerous studies showed that coreceptor selectivity of HIV-1 is primarily dependent on the third hypervariable region (V3 loop) of gp120 [71-74]. Furthermore, there is a simple rule to predict HIV-1 coreceptor usage called the 11/25 rule: if either the 11th or 25th amino acid position of V3 is positively charged, the virus will use CXCR4 as the coreceptor, otherwise it will use CCR5 . Thus, a single amino acid substitution in the V3 loop is sufficient to acquire usage of CXCR4. However, these are rare cases when the viruses exclusively use CCR5.
Indeed, escape variants from selective pressure by natural ligand for CCR5, such as MIP-1α (CCL3) , or CCR5 antagonists , still use CCR5 and do not involve acquisition of CXCR4 usage. These studies indicate that acquisition of CXCR4 usage conferred by mutations in the V3 loop of gp120 results in the loss of replication fitness, as previously described . However, the escape variants from CCR5 antagonists usually retain CCR5 usage [60,61,69,78], and recognize the antagonist-bound form of CCR5 as well as the free CCR5 form for entry by the accumulation of multiple amino acid mutations, called noncompetitive resistance [61,79]. In non-competitive resistance, once saturating concentrations of antagonists were achieved, further inhibition was not observed, resulting in the plateau of inhibition, while competitive resistance can achieve inhibition of viral replication by a sufficient inhibitor concentration, resulting in a shift in the IC50 value (Figure 4). A principal determinant for the reduced sensitivity to CCR5 antagonists has been shown to be the V3 loop of gp120 although the mutations appear to be isolate-specific and antagonist-dependent .
Figure 4: Typical competitive and non-competitive resistance profiles. Competitive resistance can achieve inhibition of viral replication by a sufficient inhibitor concentration, resulting in a shift in the IC50 value (left panel). In non-competitive inhibition, increasing concentrations of inhibitors have no effect, resulting in no increase in the inhibitory effect (right panel).
In general, primary R5 viruses or laboratory-adapted R5 infectious clones cultured in stimulated peripheral mononuclear cells (PBMCs) have been used for the selection of CCR5 antagonist-resistant variants. However, the use of PBMCs for virus passage is donor-dependent and labor-intensive. Additionally, the use of a single clone for selection would need long-term passage to induce resistant viruses. To overcome these problems, we constructed R5-tropic infectious clones containing a V3 loop library, HIV-1V3Lib. To construct replication competent HIV- 1V3Lib, we chose 10 amino acid positions in the V3 loop and incorporated random combinations of the amino acid substitutions derived from 31 subtype B R5 viruses into the V3 loop library (Figure 5). This novel In vitro system enabled the selection of escape variants from CCR5 antagonists over a relatively short time period.
Figure 5: Schematic structure of HIV-1 V3 loop library showing introduced mutations in V3 for the analysis of escape mutants. Residues in boldface indicate the substitutions that were randomly incorporated in the V3 loop, possible >2 x 104 combinations. The amino acid substitutions were detected in 31 R5 clinical isolates.
In addition to the V3 library, we are currently using PM1/CCR5 cells for virus passages. The PM1/CCR5 cell line was generated by standard retrovirus-mediated transduction of parental PM cell line with the CCR5 gene, as previously described [63,76], and is highly sensitive to the R5 viruses compared to the parental PM1 cell line. Remarkably, the infection of PM1/CCR5 cells with R5 viruses induces prominent cell fusion, which is clear sign of virus proliferation. Thus, the use of PM1/ CCR5 cells with the HIV-1V3Lib allows us to focus on the contribution of the V3 loop in gp120 in CCR5 antagonist-resistance with a shortened selection period compared to the use of PBMCs with wild-type virus. As expected, we were able to isolate TAK-779-  and MVCresistant  variants using replication competent HIV-1V3Lib. Indeed, TAK-779- and MVC-resistant variants were determined to contain several amino acid substitutions within the V3 loop sequence. However, MVC-resistant variants also contained several amino acid substitutions in non-V3 regions (T199K and T275M), such as elsewhere in the gp120 to retain infectivity [80,81]. However, these mutations could not confer non-competitive resistance, indicating the importance of the V3 loop for non-competitive resistance.
Mechanisms of resistance: It is thought that docking of gp120 to CCR5 without CCR5 antagonists involves interactions of both the V3 tip with the second extracellular loop of CCR5 (ECL2) and the V3 stem-C4 region (bridging sheet) with the CCR5 N-terminus (NT) . Since small molecule inhibitors interact with the pocket formed by transmembrane helices, thereby inducing allosteric conformational change in the ECL2, the wild-type virus can no longer interact with the ECL2. It is assumed that binding of small molecule inhibitors alters orientation between the ECL2 and NT regions, disrupting multipoint binding sites for gp120, thereby impeding gp120-CCR5 interaction . Indeed, studies using CCR5 mutants showed that the escape variants were more dependent on tyrosine-sulfated CCR5 NT than wild-type viruses [65,66,84]. Furthermore, these escape variants were more sensitive to monoclonal antibodies recognizing the NT portion of CCR5 . These studies indicated that the escape variants from CCR5 antagonists showed enhanced interactions with the NT that may be a consequence of a weakened interaction with the ECL2 (Figure 6).
Figure 6: Resistant HIV-1 viruses can enter host cells in the presence of the CCR5 antagonist. The successful viral fusion requires the interaction of the V3 loop in gp120 with the ECL2 and NT of CCR5. CCR5 antagonists bind to the pocket formed by TM helices and induce allosteric conformational changes in the ECL2, thereby disrupting the interaction of gp120 with CCR5. The CCR5 antagonists-resistant viruses containing multiple amino acid substitutions in the V3 loop can recognize antagonist-bound forms of CCR5 by enhanced interaction with the NT.
Another genetic pathway is independent of V3 mutations. Vicriviroc- resistant mutants have been developed with multiple amino acid substitutions throughout the gp120 spanning the C2-V5 region without any changes in the V3 loop . Recently, three amino acid changes in the fusion peptide domain of gp41 have been shown to be responsible for resistance although the effect of these mutations was context-dependent [84,85]. Thus, the mechanisms by which changes in the fusion peptide alter the gp120-CCR5 interaction still remain to be determined.
As previously mentioned, the patterns of mutations in escape variants against CCR5 antagonists were hypervariable and context-dependent, due in part to extensive sequence heterogeneity of HIV-1 env. Resistance to CCR5 antagonists was also found to be dependent upon cellular conditions such as cell tropism and the availability of CCR5. The differential staining of CCR5-expressing cells by various CCR5 monoclonal antibodies suggested that CCR5 exists in heterogeneous forms  and compositions of these multiple forms differed in cell type . These findings suggested that different conformations of CCR5 with CCR5 antagonists might induce different substitutions in gp120. Moreover, the development of cross-resistance to other CCR5 antagonists is inconsistent, where some studies suggest that it may occur [69,78,79] and some suggest that it may not occur . Additional data from In vitro and In vivo studies will be needed to elucidate the meaning of these studies.
Resistance to CXCR4 antagonists
CXCR4 as a target: CXCR4 is a coreceptor that is used for entry by X4-tropic viruses ; however, it is not always regarded as a suitable therapeutic target molecule for HIV-1 infection (Figure 1). R5 and X4 HIV-1 variants are both present in transmissible body fluids; however, R5-tropic HIV-1 transmits infection and dominates the early stages of HIV-1 pathogenesis , whereas X4-tropic HIV-1 evolves during the later stages and leads to acceleration of disease progression due to faster decline in CD4+ T lymphocytes [90,91]. Coreceptor switching from CCR5 to CXCR4 occurs in approximately 40–50% of infected individuals ; in addition, the R5 virus is still present as a minor viral population even after emergence of the X4 virus. Furthermore, CXCR4 deletion in mice was shown to induce a variety of severe disorders and resulted in embryonic lethality , suggesting that CXCR4-targeting drugs may be less well tolerated than CCR5 inhibitors. These studies indicate that administration of CXCR4 inhibitors is relatively restricted to the later stage of infection after coreceptor switching. Therefore, the development of CXCR4 antagonists has proceeded at a deliberate pace when compared with that of other types of entry inhibitors.
Escape from CXCR4 antagonists: Based on the manner of escape of R5-tropic HIV-1 from CCR5 antagonists, four main resistance pathways may be intuitively possible for X4 HIV-1 escape from CXCR4 antagonists: (i) coreceptor switching from CXCR4 to CCR5; (ii) outgrowth of the pre-existing R5 virus; (iii) decrease in CXCR4 susceptibility by mutation(s) in Env; and (iv) utilization of the drug-bound form of CXCR4. The first mechanism comprises a shift in coreceptor usage from CXCR4 to CCR5, which is induced by selective pressure from CXCR4 antagonists. However, this is unlikely to occur frequently because coreceptor switching from CCR5 to CXCR4, and vice versa, requires multiple mutations throughout gp160 via transitional intermediates with poor replication fitness .
There is an evolutionary gap in viral fitness between viruses using CXCR4 and those using CCR5. However, an R5X4 dual-tropic virus can shift from X4-dominated tropism to R5-dominated tropism . The R5X4 dual-tropic 89.6 mainly uses CXCR4 as a coreceptor, but after selection with the CXCR4 antagonist T140, coreceptor usage shifted from a phenotype that mainly used CXCR4 to one mainly using CCR5 due to a single amino acid substitution (R308S) in the V3 loop In vitro. These results indicated that the R5X4 virus could shift its main coreceptor usage due to a low genetic barrier to the development of resistance. In contrast, an outgrowth of the pre-existing minority of the R5 virus caused by CXCR4 antagonists, is expected to lead to virologic failure. AMD3100 is a small molecule compound called a bicyclam that has potent antiviral activity against a variety of X4-tropic strains [94-99]. However, it is not clinically available because of low oral bioavailability . After treatment of clinical isolates In vitro with AM3100 for 28 days, the major population of viruses using CXCR4 was promptly replaced by the pre-existing minor population using CCR5 with multiple mutations in the V3 loop In vitro .
The third possible pathway results from accumulation of mutations in the viral envelope that allow interaction between gp120 and the coreceptor in the presence of the inhibitor. AMD3100-resistant viruses selected In vitro from NL4-3 strain still used CXCR4 as a coreceptor and contained several mutations in the V3 loop and showed poor fitness . In contrast, other viruses resistant to POL3026, a specific β-hairpin mimetic CXCR4 antagonist, did not show any fitness cost and contained four mutations (Q310H, I320T, N325D, and A329T) in the gp120 V3 loop . These four mutations were shared by viral strains resistant to SDF-1α  and T134 , indicating that the V3 loop is a crucial region for the acquisition of CXCR4 antagonist resistance.
The fourth possible mechanism involves acquisition of the ability to utilize the inhibitor-bound form as well as the drug-free form of CXCR4 for viral entry. Several clinical isolates demonstrate infection through the AMD3100-bound form of CXCR4, indicating a noncompetitive mode of drug resistance . The V1/V2 region of one of the isolates is responsible for this property, suggesting that baseline resistance to this kind of CXCR4 antagonist should be considered while developing CXCR4 antagonists. Recent advances have led to the development of orally-active CXCR4 antagonists, including AMD11070 , KRH-3955 , and GSK81297 . Therefore, to prevent the possible emergence of pre-existing forms of the CCR5 virus, it is likely that CXCR4 antagonists will be effective only in combination with a CCR5 antagonist or other antiviral drugs.
Fusion inhibitory peptides and their mechanisms of action
Fusion inhibitors: Enfuvirtide (T-20) was approved by the FDA in 2003 as the first fusion inhibitor that efficiently suppresses the replication of HIV-1 resistant to available classes of anti-HIV-1 drugs (Figure 1), such as reverse transcriptase inhibitors (RTIs) and protease inhibitors (PIs). Hence, it has been widely used for treatment of HIV-1 infected patients where treatment with other antiretroviral drugs has failed . T-20 comprises a 36 amino acid peptide derived from the gp41 HIV-1 C-terminal heptad repeat (C-HR), as shown in Figure 7.
Figure 7: Schematic view of HIV-1 gp41 functional domains and mutation map for T-20. Putative hydrophobic pocket region of the N-HR is shown (green) and may form a leucine-zipper-like domain. In the C-HR, two tryptophan-rich domains (TRD; pink) are located at the N- and C-terminal regions (N-TRD and C-TRD, respectively). The N-TRD binds to the hydrophobic pocket in the N-HR, whereas the C-TRD plays a key role in membrane association. FP; fusion peptide domain, which penetrates into the target cell membrane. TM; transmembrane region. The amino acid sequence of the HXB2 clone is shown as a representative HIV-1 sequence. Only mutations located in the extracellular domain of gp41 are shown. Mutations observed in in vitro and in vivo selections are indicated by an asterisk (*). I37T was only selected in vitro. Primary and secondary mutations were most frequently associated with T-20 resistance (red and blue, respectively). In addition, T25S/A, S35A/T, R46K, L55F, Q56R/K, V72L, A101I/T/V/G, L108Q, N109D, D113G/N, E119Q, L130V, I135L, N140I, and L158W were selected in patients under T-20 containing regimens, but observed in some drug-naïve HIV-1 strains (Los Alamos HIV Sequence Data Bank, https://www.hiv.lanl.gov/content/index (natural polymorphisms). Corresponding regions of T-20, SC34EK, and T2635 are shown. T-20 is comprised of the original sequence but others are extensively modified.
During HIV-1 entry, binding of gp120 to CD4 and either CCR5 or CXCR4 initiates penetration of the hydrophobic fusion peptide domain at the N-terminal heptad repeat (N-HR) of gp41 into the target cell membrane . In the gp41 extra-cellular domain, the α-helical region at the C-HR begins to fold and interact with a trimeric form of the N-HR in an anti-parallel manner. This intramolecular folding forms a stable six-helix bundle and facilitates the fusion of the virus envelope and cellular membranes. During the fusion step of HIV-1 replication, T-20 can interfere with the formation of the six-helix bundle consisting of a trimeric N-HR/C-HR complex.
In the C-HR, two tryptophan-rich domains (TRDs) are located in close proximity to the connection loop (N-TRD) and the membranespanning or transmembrane region (C-TRD). Both TRDs resemble a leucine zipper structure and are believed to be important for interactions of the N-HR and the C-HR. T-20 contains the amino acid sequence of the C-TRD, whereas C34-based peptides, such as SC34EK and T2635, contain the N-TRD. T-20 is believed to bind to the N-HR as a decoy and prevents the formation of the six-helix bundle , resulting in the inhibition of HIV-1 entry. This mode of action has been well documented with another fusion inhibitory peptide, C34, and remains controversial whether the mechanisms of action of T-20 and C34 are in fact the same.
Primary and secondary mutations for fusion inhibitors: Although some fusion peptides, such as N36  and IQN17 , are designed using the N-HR sequence, most have been designed using the C-HR sequence. Primary mutations for a representative C-HR derived peptide, T-20, are generally introduced within the N-HR, a putative binding site of T-20 [112,113]. Mutations frequently reported In vivo are located at amino acid positions 36–45 of the gp41, including G36D/S/E/V, V38A/M/E, Q40H, N42T, and N43D/K (Figure 7) . Using circular dichroism analysis, others and we clearly demonstrated that these primary mutations reduce the binding affinity of C-peptides with the N-HR [112,115]. This mutation also impairs physiological intra-molecular binding of the C-HR with the N-HR, providing a replication cost . Therefore, HIV-1 develops secondary or compensatory mutations in the C-HR to restore the reduced stabilities of the six-helix bundle by the introduction of primary mutations. N126K, E137K/Q, and S138A [115,117] have been reported In vivo, usually in combination with N-HR mutations. Mutations in the C-HR restore the intra-molecular folding/interaction of the C-HR with the N-HR. The enhanced binding affinity by the secondary mutations can be applied to peptide design, such as C34 with N126K and T-20 with S138A, which maintain anti-HIV-1 activity, even to drug-resistant HIV-1 .
Secondary mutations of the N-HR are not only non-synonymous, but also synonymous. A part of the RNA coding region for the env gene, including gp41, also encodes the Rev-responsible element (RRE), which is an RNA secondary structure important for unspliced RNA export from the nucleus that is required for efficient viral protein synthesis and packaging of genomic RNA [118,119]. Primary mutations at positions 36 and 38 for stem II and at 43 for stem III affect the RRE structure. Synonymous and non-synonymous mutations introduced into the gp41 compensate for RRE structure stability, such as T18A for V38A  and A30V for G36D , and Q41 (CAG to CAA) and L44 (UUG to CUG) for N43D . This association between the gp41 and RRE results in some genetic restrictions.
Impact of mutations on clinical potency: Only one or two amino acid substitutions in gp41 appear to be sufficient for clinical treatment failure, where after the emergence of mutations, viral load gradually increases . For example, G36E, V38A, Q40H, and N43D were shown to confer 39.3-, 16-, 21-, and 18-fold reductions in susceptibility to T-20, respectively . Double or triple substitutions have also been identified in clinical isolates from patients undergoing therapy with T-20. Mutations such as N42T+N43S, V38A+N42D, and Q40H+L45M confer 61-, 140-, and 67-fold reductions in susceptibility to T-20, respectively . Mutations at codons 36 (G36E/D/S) and 38 (V38A/G/M) seem to emerge relatively rapidly In vivo, whereas Q40H and N43D emerge more slowly . After prolonged therapy, HIV-1 has been shown to develop secondary mutations and may confer more apparent resistance with improved replication kinetics. Therefore, combination regimens with other inhibitors, such as RTIs and PIs, are indispensable for sufficient positive viral responses.
T-20 appears to inhibit replication of HIV-1 subtype independently [124-126], since T-20 has mainly been used for subtype B HIV-1 infected patients. Based on the mechanism of action of T-20, interference of N-and C-HR interactions may be expected, where amino acid sequences are highly conserved across all subtypes. However, in non-B subtype HIV-1, N42S predominantly emerged as a resistance-related mutation [124,125].
Resistance to the next generation inhibitors: Next generation inhibitors have been designed using several strategies, such as the introduction of specific amino acid motifs and secondary mutations into the sequence of the original peptide inhibitors  to enhance the stability of the α-helical structure between inhibitors and fusion domain at the N-HR. In contrast to T-20, primary mutations to third generation inhibitors were not selected In vitro [127,128]; therefore, the accumulation of multiple mutations is likely necessary for the development of resistance. In the case of SC34EK, 13 amino acid substitutions (D36G, Q41R, N43K, A96D, N126K, E151K, H132Y, V182I, P203S, L204I, S241F, H258Q, and A312T) were introduced and single amino acid substitutions only conferred weak resistance (<6-fold) . For another peptide, T-2635, 12 amino acids in 10 positions (A6V, L33S, Q66R/L, K77E/N, T94N, N100D, N126K, H132Q, E136G, and E151G) were selected, and single mutations did not confer resistance to T-2635 . Interestingly, some of these mutations were located outside the N-HR and C-HR. Cross-resistance between SC34EK and T-2635 was only examined for the SC34EK-resistant virus and revealed little crossresistance . Further studies of resistance profiles might be helpful in defining new strategies for the design of fusion inhibitors that can suppress the replication of resistant variants of HIV-1.
The emergence of viruses resistant to entry inhibitors, as well as other classes of antiviral agents (reverse transcriptase or protease inhibitors), has been reported In vitro and In vivo. Resistance to entry inhibitors, including attachment inhibitors and coreceptor antagonists, is mainly conferred as a result of missense mutations within the gp120 subunit of the env gene, which differ from one inhibitor to another. Alternatively, treatment failure can occur through the expansion of pre-existing CXCR4-using virus for CCR5 antagonists, and vice versa. Agents that target gp41-dependent fusion select for HIV-1 variants with mutationswithin the gp41 envelope gene. These results indicate the incredible flexibility of the HIV-1 genome to escape from a variety of entry inhibitors. Therefore, the development of novel entry inhibitors for clinical use is needed to limit escape mutants by effective combination therapy.
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals