HIV-1 Tropism

New antiviral options that have become available through the development of antiretrovirals directed toward novel targets but also of antiretrovirals being developed in the existing classes, provide opportunities for new antiviral strategies involving agents directed concomitantly at a multitude of viral life-cycle targets, including HIV-1 entry into host cells. Entry inhibitors in development aim at blocking HIV-1 attachment, a prerequisite for the virus to accomplish fusion and entry into host cells, by binding either to the viral surface receptors (e.g., gp120 and gp41) or the host human cell receptors expressing CD4 or the chemokine coreceptors (CCR)5 and CXCR4 [1–11]. This is a particularly valuable method of hindering HIV infection as by blocking entry into human host cells the genetic material of the assaulting HIV particles is not incorporated into host cells and viral replication is prevented. In addition to entry inhibitors, the maturation inhibitors currently in development, disrupt the capsid protein and render virions noninfectious [12].

In addition to binding the CD4 receptor, a second adjacent coreceptor is required for HIV to interact with so as to enter target cells. CCR5 and CXCR4 are the major CCRs used by HIV to enter into human host cells [13–15]. Based on coreceptor use, HIV-1 strains have been classified according to their tropism: CCR5 tropic (R5); CXCR4 tropic (X4); or dual tropic (R5/X4) [16]. This corresponded to previous observations where nonsyncytium-inducing viral phenotype was consistent with replication in monocyte/macrophages (M tropic), linked with less virulent strains, whereas syncytium-inducing viral phenotype was consistent with replication in T lymphocytes (T tropic) and linked with more virulent strains [17–20], suggesting that tropism may be related with virulence and disease progression or stage of disease.

HIV-1 strains present in early infection preferentially use the CCR5 receptor, while CXCR4- using strains emerge in approximately 50% of patients over time and are associated with accelerated immunologic decline and progression to AIDS [19,21–24]. As HIV-1 evolves during the course of infection, coreceptor usage may also change. It is unclear what factors suppress the emergence of X4-tropic viruses during the early phase of the infection or contribute to their emergence in advanced disease. Furthermore, it is unknown whether the appearance of X4-tropic strains is a consequence or cause of accelerated immunologic decline and disease progression to AIDS. However, CXCR4 viruses can infect immature thymocytes, precursor cells of mature CD4 lymphocytes, which express high levels of X4 but lower levels of R5 coreceptors, with resultant rapid depletion of thymus cells; this may explain the rapid CD4 T-cell depletion, faster progression to AIDS and higher virulence observed in patients infected with CXCR4-tropic virus, in contrast with CCR5-tropic viruses, which reach the thymus via dendritic cells or mature CD4 T cells but do not affect thymopoiesis [25–29].

CCR tropism of HIV is associated with CD4-cell counts, HIV-1 RNA levels and natural killer cell counts. The presence of mixed/dual-tropic CCR5/CXCR4 populations or CXCR4-using virus may be seen at all CD4-cell counts and viral loads but is more common at lower CD4-cell counts and higher viral loads. Consequently, these laboratory markers can not be used to predict the presence of CCR5-using virus in the clinical setting and screening for chemokine receptor use would be necessary before the initiation of treatment with CCR antagonists and routinely during their administration [30].

Entry inhibitor studies, with proof-of-concept in terms of primary and secondary end points, warrant further development to determine the optimal agents, dosing, long-term safety and efficacy; the exception is maraviroc (UK-427, 857), a potent, orally bioavailable and selective small molecule inhibitor of CCR5, with broad spectrum anti-HIV-1 activity, while vicriviroc, a similar agent, is lagging behind in terms of development. Maraviroc studies published recently demonstrated efficacy and safety in treating triple class, experienced patients infected with R5-tropic HIV-1 [31,32]. Maraviroc (twice daily [b.i.d.] or once daily [q.d.]) plus optimized background treatment (OBT) provided significantly greater virologic suppression rates and increases in CD4 count compared with placebo plus OBT at 48 weeks in the treatment-experienced population studied. There were no clinically relevant differences in safety between maraviroc (b.i.d. or q.d.) and placebo treatment arms [33]. In lieu of this, maraviroc had been granted accelerated review by both US and EU regulatory authorities and was licensed in August, 2007. Much attention has been turned to this agent since drug characteristics may be class or coreceptor determined, pertaining to therapeutic efficacy, mechanism of action, antiviral activity, viral resistance, drug interactions and reasons for failure.

Change in coreceptor usage by HIV-1 after exposure to maraviroc, while virologic suppression is achieved but also on viral rebound or post-treatment failure [31,32,34], remains an issue of paramount importance to be answered and resolved in view of X4 viruses being linked with the infection of immature thymocytes, resulting in thymus cell and subsequent CD4 T-cell depletion, faster progression to AIDS and higher virulence [25–29,35,36]. Viral tropism seems to independently influence HIV disease progression [30,37].

In patients with virologic failure, a change in coreceptor use from R5 to dual or mixed R5/R4 was recorded in the maraviroc studies: 65% (n = 32 of 49) of patients in the b.i.d. group, 63% (n = 31 of 49) in the q.d. group and 5% (n = 4 of 84) in the placebo group, respectively [31,32,34].

Very accurate recombinant virus tropism assays, utilized before the initiation of entry inhibitor treatment and on follow up, can make a huge difference in the concomitant determination of HIV tropism and viral susceptibility to HIV entry inhibitors, whilst the threshold for detection of minority quasispecies in dual/mixed viral populations and X4 viruses will be markedly lowered; the current limit of detection for minor populations is 10% for R5 viruses and 20% for X4 viruses by the Phenoscript assay (Viralliance, Paris, France) [38] and 5% for R5 viruses and 10% for X4 viruses by the Phenosense assay (Monogram Biosciences, San Francisco, CA, USA) [34,39]. This limitation, pertaining to threshold ranges, may have important repercussions in the implementation of treatment involving CCR5 blockage, as minor dual/mixed or X4 viral populations not detected at baseline may subsequently emerge and expand [40–42]. Also, these assays are not able to discriminate between mixed and dual tropic R5/X4 populations. The plasma viral load above which samples can be reliably amplified in the Trofile assay is 1000 copies/ml of plasma.

Trofile amplifies the envelope gene from a patient‘s HIV genome (obtained from their blood sample) and then uses it to make HIV particles containing the patient‘s virus envelope protein. The resultant HIV particles are then used to infect cells that contain the CCR5 coreceptor or the CXCR4 coreceptor on the cell surface. Once the virus infects the cell, it undergoes a single round of replication. Virus replication results in the production of luciferase from a luciferase gene that is carried into the cell by the virus. The production of luciferase in either CCR5 cells, CXCR4 cells or both cell types, defines the coreceptor tropism of the patient virus [43]. It is a validated commercial assay used in all the current clinical trials, which aims to accelerate the development of entry inhibitors and assist clinicians in the selection of appropriate treatment regimens. It is utilized to select patients based on their coreceptor usage prior to treatment at the trial enrolment stage and detect coreceptor switching on viral rebound after treatment initiation. This method offers the most clinically validated data with currently approximately 25,000–30,000 analyzed samples.

Recent enhancements to the Trofile assay enable reliable detection of CXCR4-using HIV-1 subpopulations at less than 1% without sacrificing the ability to reliably detect CCR5 variants [44]. Attempts to improve coreceptor tropism predictions based on V3 loop sequence in antiretroviral-experienced patients was not successful for the detection of CXCR4-using variants because envelope sequencing is hampered by diversity in variable-loop length amongst subpopulations of viruses in approximately 25% of patients [45]. Nevertheless, clinical applications and orthologic utilization of tropism determinations will depend on the development of accurate and easily accessed tropism assays.

In a large cohort of 663 treatments, naive patients (18%) were identified with X4 tropism, all as dual/mixed R5/X4 tropisms, and 12% had non-B subtypes, although geographic origins were not described; X4 tropism was found more commonly among patients with lower CD4-cell counts and CD4/CD8 ratios. The phenotype assay utilized failed to identify 8.6% (57 samples) and the resolution of tropism 1 month later in 26% (15 of 57 samples) might indicate continuing evolution, rapid dynamics and changing tropism in the absence of external selective pressures like antiretroviral therapy, while it remains unknown whether continuous switching between R5 and X4 tropism in response to fitness, immune pressures and receptor expression levels on lymphocytes take place [42]. HIV-1 coreceptor switch was induced by antagonism to CCR5 in cell cultures; in the absence of drug pressure, three of six strains used were able to switch from the R5 to X4 phenotype and demonstrated an increased replication rate. Coreceptor switch could be delayed by zidovudine or TAK-779, whilst under CCR5 drug pressure faster emergence of CXCR4-using HIV was induced [46].

The impact of highly active antiretroviral therapy (HAART) on HIV tropism is unpredictable. Even with longstanding complete suppression of viremia, HAART may influence the selection of X4 viruses; a switch from R5 to X4 variants in 11 of the 23 patients who harbored a majority virus population of R5 variants at baseline out of a total of 32 patients receiving HAART with viral load below detection levels for over 5 years, whereas X4 variants persisted from baseline in nine patients, in a longitudinal follow-up study. This could be attributed to peripheral expansion of long-lived (naive CD45RA⁺) T cells in which archival X4 variants are present [47].

Another study demonstrated that HAART might delay the selection of X4 viruses; 24 of the 32 patients analyzed harbored exclusively R5 virus in plasma before the initiation of treatment, whereas eight had mixed R5/X4 virus populations. In four of these eight patients, all of whom initially responded to treatment, the persistence of R5 virus in plasma was observed, whereas the X4 component of the virus population became undetectable. The suppression of the X4 virus component was a transient phenomenon and variants able to use CXCR4 re-emerged after a variable delay [48]. However, in another report, the initiation of a lopinavir–ritonavir-based salvage regimen in 20 HIV-positive children showed that virological responses and antiretroviral efficacy were associated with X4 virus suppression [49].

A high frequency of syncytium-inducing and CXCR4-tropic viruses among HIV-1 subtype C-infected patients receiving antiretroviral therapy was also observed in a 28 patient cross-sectional study [50].

Coreceptor tropism has been established as a determinant of disease progression; individuals with virus using the CCR5 coreceptor generally have a slower rate of progression and lower viral load than those with virus using CXCR4 [51–53].

Several studies show that baseline HIV-1 coreceptor tropism predicts disease progression. In one study, detectable CXCR4-using virus at baseline was associated with a lower baseline CD4⁺ T-cell count and a higher plasma HIV-1 RNA level, while it independently predicted a greater decrease in CD4⁺ T-cell count over time and was associated with a 3.8-fold increased risk of progression to clinical AIDS [37].

In another study, antiretroviral naive subjects with CXCR4/mixed-tropic virus were associated with faster CD4 cell decline compared with individuals that had CCR5 tropism [54].

Although HIV-1 subtypes are, in principle, CCR5 and/or CXCR4 tropic, variation in coreceptor usage that is determined by viral clade has also been proposed. HIV-1 subtype A has been found to also utilize alternative coreceptors such as CXCR6 (Bonzo) and BOB/GPR15 [55–59], even in the early stages of infection [55,57]. It was shown that 50% of isolates efficiently utilized BOB, CXCR6 (Bonzo) or BOB and CXCR6 (Bonzo) in addition to CCR5 to achieve host cell entry and establish infection [55]. It was also demonstrated that viruses utilizing BOB and/or CXCR6 (Bonzo) can enter cells and establish infection when CCR5 and CXCR4 are blocked [60]. HIV-1 subtype D studies demonstrated a considerable predominance of X4 and R5/X4-tropic viruses [56,61]. This calls for careful surveillance and intermittent determination of coreceptor tropism in geographic areas where these subtypes are predominant, especially when coreceptor antagonists are administered.

The issue of ‘early R5 viruses‘ versus ‘late R5 viruses‘ for patients developing AIDS in the absence of X4 viruses remains to be elucidated further so that the impact on HIV disease progression following the administration of entry inhibitors can be put into perspective; enhanced cytopathic effects owing to decreased sensitivity to inhibition by the β-chemokine RANTES [62–64] or increased levels of CD4 T-cell death [65], characterize late R5 viruses. Observations that primary R5 viruses appear to generate two different phenotypes, where early R5 strains are highly sensitive to inhibition by T-20 and TAK-779 and require higher by comparison numbers of CD4/CCR5 receptors for cell entry whereas late R5 strains have a reduced dependence on CD4/CCR5 levels to achieve entry into target cells and sensitivity to entry inhibitors T-20 and TAK-779, may indicate the ability of viruses to evolve/develop CCR5 utilization in time and concomitantly modify their virulence and susceptibility to entry inhibitors [66].

Humanized CCR5 monoclonal antibody PRO 140 and small molecule CCR5 antagonists maraviroc (UK-427,857), vicriviroc (SCH-D) and TAK-779 were found to be more potently, statistically significantly synergistic when used together than in combination with inhibitors that block other stages of HIV-1 entry; although additive effects were observed for both small molecule CCR5 antagonists and PRO 140 in combination with other classes of HIV-1 inhibitors. The combination of PRO 140 and enfuvirtide (the gp41 fusion inhibitor) trended towards significance and could be further explored [67]. Some studies reported broad synergy between CCR5 inhibitors and other HIV-1 treatment classes [68–70]. Moderate synergy was observed in single experiments with atazanavir, indinavir and enfuvirtide, with additive effects observed in repeat experiments [71].

Synergy between HIV-1 attachment, coreceptors and fusion inhibitors have been demonstrated in culture with simultaneous use of PRO 542 (a CD4-attachment inhibitor), PRO 140 (a CCR5 inhibitor) and the fusion inhibitors T-20 and T-1249; when used in combination, lower concentrations of drug were required to achieve similar levels of inhibition than when used alone [72]. This may indicate that in the future, combinations of antiretrovirals targeting a multitude of HIV cell-cycle sites can be used at lower doses thereby reducing toxic and side/adverse effects while at the same time providing a more efficient inhibition of HIV infection.

Resistance of HIV-1 variants to CCR5 antagonists is not yet fully understood. It has been proposed that HIV could develop resistance by:

• Switching to use another coreceptor (X4 to R5 or R5 to X4);

• Developing the capacity to enter the cell by a nonreceptor pathway;

• Using the CCR despite the fact that it is occupied by the drug by somehow avoiding the inhibitor.

This last mechanism was demonstrated in an in vitro study and suggested that maraviroc-resistant primary isolates (CC1-85 and RU570 derived) had the ability to utilize the CCR5 coreceptor in the presence of bound maraviroc; on serial passaging with maraviroc, resistance developed as mutations in the viral envelope accumulated. The majority of mutations were seen within the V3 loop, with additional mutations seen in the V2, C3 and V4 regions, as well as in gp41 with CC1-85-derived maraviroc- resistant viruses, and in the V1, C4, V4 and C5 regions with RU570-derived maraviroc-resistant viruses. These maraviroc-resistant viruses remained susceptible to other CCR5 antagonists and enfuvirtide [73,74]. However, HIV-1 variants resistant to the CCR5 antagonists SCH 351125 and SCH 417690 demonstrated cross-resistance to other CCR5 inhibitors but were sensitive to other drug classes. Mutational patterns differed among individual resistant viruses but typically clustered in the V3 loop [75].

Resistance to enfuvirtide does not affect susceptibility to other classes of entry inhibitors [76].

Indirect effects following the inhibition of R5 and X4 coreceptors are not known. CCR5 was found to be a critical antiviral and survival determinant in West Nile virus (WNV) infection of mice that acts by regulating trafficking of leukocytes to the infected brain [77]. Genetic analyses in humans and mice provided evidence for the CCR5 control of infection by WNV, a re-emerging pathogen able to cause fatal encephalitis [78], and a link between CCR5Δ32 and fatal WNV infection [79]. In a cohort of WNV-infected patients from Arizona (USA), CCR5Δ32 homozygotes were significantly associated with fatal outcome, with a 6.6-fold increased risk of death from WNV infection and 4.7-times higher liklihood than others (nonhomozygotic) to have WNV infection. Similarly, among patients from Colorado (USA), CCR5Δ32 homozygotes were 4.2-times more likely to have WNV infection [80].

The CCR5Δ32 mutation may also have an impact on disease outcome in individuals with HCV. Some studies [81,82], but not all [83,84], have reported an association between the frequency of CCR5Δ32 polymorphisms and better outcome of HCV infection due to spontaneous viral clearance or reduced portal inflammation and milder fibrosis, respectively. Others reported that increased prevalence of CCR5Δ32 homozygosity was associated with increased viral loads in patients with chronic HCV, suggesting that the CCR5Δ32 mutation may be an adverse host factor in HCV [85]. In a number of studies, the CCR5Δ32 allele frequency, as determined by PCR, did not show statistically significant difference between HCV patients and controls [86–88]. Moreover, the response rates to IFN-α monotherapy were reported to be reduced in HCV- infected patients carrying the CCR5Δ32 mutation, which was overcome by administering interferon/ribavirin combination therapy [89].

Investigation to determine whether CCR5 inhibition, through the administration of vicriviroc, had any effect on plasma titers of Epstein–Barr virus, which is associated with many AIDS-related lymphomas, following the occurrence of two Hodgkin‘s and two non-Hodgkin‘s lymphomas, one squamous cell carcinoma of the skin and one gastric adenocarcinoma in the AIDS Clinical Trials Group protocol A5211, did not show upregulation of Epstein–Barr virus replication [90]. These findings highlight the possible risks and benefits in targeting this chemokine coreceptor either for the prevention or treatment of infectious diseases.

Latent reservoirs and tissue compartments express high levels of CCR5 [91]; virus reservoirs such as resting memory (CD45RO) CD4⁺ T lymphocytes and resting naive (CD45RA) CD4⁺ T lymphocytes harbor replication competent but latent HIV-1 proviral DNA [92] and provide persistent, low-level viral replication [92–94]. In addition, tissue compartments (as in the CNS), with the blood–brain barrier reducing permeability to drugs like protease inhibitors, have residual viral replication.

Conditions for serious curtailment of viral activity or even possible cure of HIV infection would aim at safe and effective means in overcoming viral latency concomitantly with efficacious antiretroviral therapy targeted at all virus–host cell interactions [91–94]. Major new antiviral options are now provided with new classes of antiretrovirals being made available; the entry and integrase inhibitors. Their prudent utilization can supplement and augment current therapeutic regimens, providing a new platform for efficacious antiretroviral therapy in multi-experienced failing patients.

Antiretroviral therapy may be revolutionized if, entry inhibitors combined with intensified HAART (EI-HAART) containing T-20 and integrase inhibitors (all available antiretrovirals acting on the maximum of HIV lifecycle targets), can be shown to deplete latent HIV-1 infection in vivo. It remains to be seen when valproic acid, a histone deacetylase 1 inhibitor, is coadministered with EI-HAART if it can be shown to markedly improve on the frequency of decrease of resting cell infection as previously achieved [95].

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.