Highly active antiretroviral therapy and beta-chemokines

Chemokines (chemotactic cytokines) are a large family of immune factors that regulate leucocyte trafficking during physiological and pathological processes. Based on the positioning of cysteines, chemokines are divided into four subfamilies, CXC (also called alpha-chemokines), CC (beta-chemokines), C (gamma-chemokines) and CX3C (delta-chemokines) [1]. Research in the field of chemokines accelerated significantly after the report by Cocchi et al. that a subset of these factors, namely CC chemokines macrophage inflammatory protein (MIP)-1α, MIP-1β and regulated upon activation normal T cell-expressed and secreted (RANTES), can function as inhibitors of HIV-1 [2]. This finding led to identification of CCR5, a receptor for these chemokines, as a main coreceptor for primary HIV-1 strains [3–7]. An alternative co-receptor used by T cell-line tropic viruses and by HIV-1 strains appearing mainly at later stages of disease is CXCR4, a receptor for a CXC chemokine SDF-1α [8]. The discovery of involvement of chemokine receptors in HIV-1 binding and entry opened the door for new approaches to controlling HIV-1 infection.

The feasibility of chemokine receptors as targets for anti-HIV therapy was supported by a finding that a genetic mutation in CCR5, delta32, which occurs naturally in a small percentage of individuals, makes them highly resistant to HIV-1 infection [9]. High levels of CC chemokines that utilize CCR5 would be expected to achieve a similar effect, both by down-regulating CCR5 on target cells and by competing with HIV-1 for CCR5 binding. This assumption was confirmed in a report that resistance of CD4⁺ T cells of HIV-1-exposed but uninfected individuals to in vitro infection by CCR5-tropic HIV-1 isolates correlates with high levels of CC chemokines produced by these cells [10]. Another report demonstrated that PBMCs of uninfected haemophiliacs exposed repeatedly to HIV-contaminated blood products were able to produce higher levels of CC chemokines than PBMCs from a control group of haemophiliacs [11].

However, attempts to establish correlation between the levels of MIP-1α, MIP-1β and RANTES and the progression of HIV disease or success of therapy brought contradictory results. When production of CC chemokines from stimulated PBMCs or cultured T cell clones was examined, higher levels of all or some of MIP-1α, MIP-1β and RANTES chemokines were associated with slow progression or asymptomatic state [12–15]. In contrast, another similar study did not find such a correlation [16]. In yet another study high levels of RANTES in plasma were even correlated with rapid progression to AIDS [17].

As a result, the protective or predictive value of CC chemokine levels for the success of the anti-HIV-1 therapy is not accepted uniformly. In some studies, a correlation was established between increasing levels of MIP-1α, MIP-1β and RANTES in plasma and decreasing levels of HIV-1 RNA during indinavir-based therapy [18,19], whereas a study of patients on ritonavir- and saquinavir-based therapy found that only the ability of PHA-stimulated PBMCs to secrete CC chemokines had a predictive value, but not the levels of MIP-1α, MIP-1β and RANTES in serum or plasma [20]. In patients on highly active antiretroviral therapy (HAART), high levels of RANTES production by CD4⁺ and CD8⁺ T cells before initiation of therapy were indicative of a positive outcome of HAART [21]. Also an increased production of inducible MIP-1β [22] or ΜΙP-1α [23] or both MIP-1β and RANTES [24] was demonstrated during HAART. However, others reported that high levels of MIP-1α, MIP-1β and RANTES at the start of HAART decreased to the levels of uninfected controls during therapy [25] or no changes in the levels of CC chemokines during therapy were found [26]. In a different cohort, patients exhibited increase of some CC chemokines and yet a decrease of other chemokines as a consequence of HAART [27]. A potential problem with the interpretation of the results of these studies is that chemokines are probably only one of many factors determining the prognosis of HIV-1 infection and success of therapy. Some of the discrepancies between the individual studies could be explained by differences between the cohorts of patients in the state of HIV disease, mode of treatment and drugs used, methods of preparation of the cell cultures and means of activation, virus load and virus tropism (R5 versus X4). To demonstrate clearly the role of CC chemokines in HIV disease a well-defined population of subjects would be needed, in whom at least some of the contributing factors could be standardized to reveal the specific effects of chemokines.

The main issue that needs to be addressed before using chemokines as therapeutic agents is whether chemokine levels could be manipulated to achieve anti-HIV protection without causing pathological side effects. Indeed, high levels of CC chemokines are associated with severe inflammatory conditions and could have pathological effects on their own [28]. It appears that chemokines should be considered as a supportive rather than the main anti-HIV factor in therapy. When used together with anti-HIV drugs, moderate chemokine levels will reduce the efficiency of HIV-1 entry, thus providing a better control of infection. In addition, such levels might be sufficient to control rebound of the virus during structured therapy interruptions, which are used to reduce side effects of anti-HIV drugs. While these assumptions seem to be supported firmly by results of in vitro experiments, they still await confirmation from clinical data.

The paper by Burton et al. in this issue of Clinical and Experimental Immunology [29] introduces an unexpected spin to a problem of chemokines in anti-HIV therapy. It suggests that chemokine levels can be manipulated by choosing specific components of the HAART treatment. If this result is confirmed, it may contribute to our ability to design the optimal composition of the anti-HIV drug cocktail used in HAART. However, several concerns need to be considered before therapy decisions can be made (see Table 1). The most important concern is the possibility that CC chemokines might favour emergence of X4 (CXCR4-utilizing) HIV-1 strains [30], which are associated with faster progression of HIV disease [31,32]. Such a possibility is illustrated by HHV-6, which may suppress replication of CCR5-tropic viruses in ex vivo experiments by inducing expression of RANTES [33], but is still a confounding factor in HIV-1 infection [34]. Another potential problem is that high chemokine concentration may attract additional CD4⁺ T cells to the sites of infection, thus providing new targets for the virus to spread. Chemokines may also activate the cells through chemokine receptors, thus contributing to hyperactivation of the immune system, which is one of the pathogenic mechanisms in AIDS.

Table 1.

CC chemokines and HIV-1 infection

Favourable	Potentially dangerous
Inhibition of R5 strains	R5 to X4 switch
Attraction and activation of CD4+ T cells: stimulation of anti-HIV response	Attraction and activation of CD4+ T cells: new targets for the virus
Attraction of CD8+ T cells to the site of infection	Activation of latent provirus in HIV-infected cells

It could be a delicate balance between various chemokines and other soluble factors, as well as their concerted interaction with several receptors that is actually determining the course of HIV-1 infection and the success of therapy. A caution has to be exercised before this balance is manipulated by future vaccination or therapeutic approaches.