Targeting a host element as a strategy to block HIV replication: is it nice to fool with Mother Nature?

There is increasing recognition that numerous host elements interact with HIV as it makes its way through the cells it must infect in order to propagate ^[1,2]. With the recent approval of maraviroc as a treatment for HIV infection, the Food and Drug Administration (FDA) licensed the first antiretroviral agent that instead of targeting a viral element directly, targets a host element that the virus uses for its replication cycle. Targeting CCR5 was a logical first target as a dysfunctional allele - one encoding a 32 base pair deletion in the CCR5 open reading frame - has a frequency as high as 10-14% in the white population ^[3,4] and about 1% are homozygous rendering these persons almost completely resistant to HIV infection ^[3,5,6]. Moreover, despite such a high prevalence in select populations today, this allele exists in apparent Hardy Weinberg equilibrium implying that on a large scale, there is neither gross advantage nor disadvantage to any of the three possible genotypes ^[7]. Targeting a host element as an antiretroviral therapy has a number of theoretical (but unproven) advantages such as, for example, a possible more arduous pathway for the emergence of resistance ^[8]. In this issue, pathways for the emergence of resistance are described and they are quite interesting and effective. These pathways include binding to a different region of the coreceptor that is not occupied by the inhibitor, binding to the complex of receptor that is bound and altered by the inhibitor and the acquisition of the ability to also utilize CXCR4 as a coreceptor for entry.

Yet targeting a host element is not without risk, especially if this element plays a role in host survival or happiness. This situation might be the case for CCR5, even though there is not a robust effect of this knockout on survival in the general population ^[7]. As reviewed in this issue, there is mounting evidence that persons who otherwise seem to do just fine without CCR5 appear to be at increased risk for morbidity and mortality if they encounter certain flaviruses such as West Nile virus ^[7,9] and tick-borne encephalitis virus ^[10]. The recognized promiscuity of receptor ligand interactions among chemokines and their receptors ^[11] may account for the general tolerance of the CCR5 knockout; perhaps, this functional redundancy (at least for CCR5) is just at the margin of tolerability only in certain anatomic sites such as the central nervous system. Will this fact be a problem for HIV-infected persons receiving therapy with small molecule CCR5 inhibitors? Right now, it is just not clear. On the one hand, there may be profound differences in the adaptation of host defenses of persons who have lived without CCR5 since birth and persons who only begin to experience drug-induced CCR5 blockade during adulthood. Likewise, blocking CCR5 in a person with preexistent HIV-related immune deficiency might be not as well tolerated than the congenital absence of CCR5 might be. Alternatively, small molecule CCR5 blockade might not be as complete as the genetic knockout, especially in sites such as the central nervous system, wherein defenses against certain flaviruses might be most important ^[12,13]. So far there are no data indicating that administration of these small molecule CCR5 inhibitors places persons at greater risk for these infections, but this should be monitored carefully in ongoing trials and through postmarketing surveillance.

Could targeting host CCR5 provide benefits beyond antiviral effects in persons with HIV infection? There is reason to suspect that the interaction between CCR5 and its ligands (the so-called 'axis of evil' ^[14]) plays a role in HIV disease pathogenesis. Genetic studies ^[15,16] from several cohorts implicate polymorphisms in the genes for CCR5 and one of its key ligands, CCR3L1, in differential outcomes of HIV infection that is in part independent of virologic effects. Moreover, among nonhuman primates that are natural hosts for simian immunodeficiency virus (SIV) infection, decreased expression of CCR5 on CD4⁺ T cells is a concomitant of the low-level immune activation that accompanies the benign clinical course and circulating CD4⁺ T cell preservation that is characteristic of SIV infection in these species ^[17]. Persons with HIV infection treated with the CCR5 antagonists, maraviroc and vicriviroc, tend to experience CD4 T cell gains with treatment that may not be attributed only to antiviral activity ^[18,19]. Signalling through CCR5 may play a role in the systemic immune activation that is thought to help drive the pathogenesis of immune deficiency and CD4⁺ T cell depletion in chronic HIV infection ^[20], and blocking CCR5 may help to reverse this.

There are other potential clinical outcomes of CCR5 blockade that merit consideration in the setting of HIV infection. As cellular activation and chemotaxis mediated in part through CCR5 interactions with its ligands may promote inflammation at tissue sites of microbial invasion, blockade of CCR5 may help to limit inflammatory complications of HIV infection such as the immune response inflammatory syndrome (IRIS) that is seen in a proportion of persons with advanced disease who begin antiretroviral therapy ^[21,22]. The CADIRIS trial (CCR5 antagonism to Decrease the Incidence of IRIS), which will begin in early 2009 is designed to ascertain whether adding maraviroc to a standard antiretroviral therapy regimen will decrease the occurrence or severity of IRIS. As the population surviving with HIV infection ages, increased risks for cardiovascular disease are predictable. Will CCR5 blockade decrease cardiovascular risk as suggested by animal studies ^[23,24] or will these risks be enhanced as suggested by early clinical surveillance? What will be the effects of CCR5 blockade on vaccine responsiveness? One might predict some level of impairment but to date, its occurrence or its significance if it occurs is not known.

Why do viruses that utilize CCR5 for entry predominate in early HIV infection ^[25] and why do not viruses utilizing CXCR4 emerge more often and earlier in the course of HIV infection? Work presented in this issue of Current Opinion in HIV and AIDS indicates that the fitness cost to R5 viruses of evolving to utilize CXCR4 may provide a substantial barrier to this emergence ^[26]. Yet both the frequency of cells coexpressing CXCR4 and CD4 and the density of CXCR4 on target CD4⁺ cells in lymphoid tissue are substantially greater than the frequencies and densities of CCR5 on target CD4⁺ cells at these sites of HIV replication ^[27]. In addition, X4 viruses tend to outgrow R5 viruses in in-vitro competitive fitness assays ^[28,29]. Although viruses that utilize only CXCR4 for entry are uncommon ^[30], viruses capable of utilizing CXCR4 are commonly found in plasma and in genital secretions of persons with HIV infection ^[31,32], yet these viruses do not appear to be transmitted or to predominate in the newly infected host. There are a number of host factors that could limit the ability of HIV to utilize CXCR4 and these factors might be at least in part responsible for both the selective acquisition and predominance of R5 using viruses in HIV-infected persons ^[33]. It is likely that some of these factors are linked to host defenses as with time and with progression of immune deficiency, the frequency with which X4 utilizing viruses can be found in plasma tends to increase ^[34,35]. Although a number of potential sites of restriction for X4 virus acquisition have been proposed ^[33], it is not clear which (if any) of these continues to limit the emergence of X4 viruses throughout the course of most infections. On the other hand, the observation that blockade of CCR5 is sufficient to induce the emergence of minor, preexisting X4 virus could suggest that no other immunologic factor besides CCR5 availability is limiting X4 dominance over R5 virus. Resistance to CCR5 inhibitors in cell culture can occur in the absence of coreceptor switch. However, virus may be more prone to coreceptor switch under conditions in which low but undetectable CCR5 and high CXCR4 expression are preestablished ^[36].

These observations underscore the interdependency among HIV, its human host and the pathogenesis of disease. A better understanding of these relationships should lead to the design of better strategies to manage HIV infection and its complications and also may identify targets for treatment of other disorders that may be driven by host immune or inflammatory intermediates.