PRO-INFLAMMATORY CYTOKINES IN HIV INFECTION

Pro-inflammatory cytokines, including TNF-α, IL-1β, and IL-6 have been associated to increased levels of HIV replication in vivo and in vitro. At the molecular level, both NF-kB and AP-1 can mediate transcriptional activation of HIV expression either by a direct interaction with DNA binding sites in the HIV LTR or, in the case of AP-1, via binding to an intragenic enhancer [46].

IL-2 is a central regulator of T cell function and survival produced by activated T lymphocytes and, very recently, also by microbial stimulation of DC [51,52]. IL-2 induces proliferation and activation of both CD4⁺ and CD8⁺ T cells [53], potentiates the cytotoxicity of CD8⁺ T lymphocytes and NK cells and stimulates B cell function, therefore playing a major role in the containment of viral infections and in the elimination of intracellular organism [54]. IL-2 expression in LN is barely detectable at all stages of HIV infection in both adults [55] and children [56], although increased expression of this cytokine has been associated with reduction of plasma viremia [57]. In one study, overexpression of IL-2 was observed in HIV infected tonsils [58], whereas a modest inhibition of cytokine expression has been independently reported in the intestinal mucosa [59]. IL-2 mRNA expression in LN of rhesus macaques has been correlated to SIV replication early after infection [41]. PHA-stimulated lymphocytes of SIV infected monkeys, as well as of HIV-infected chimpanzees, showed decreased IL-2 transcription and cell proliferation [42,60].

IL-2 production is deficient in HIV infected individuals as well as after in vitro infection of PBMC, a defect that has been correlated to an increased production of IL-4 and IL-10 [61].

IL-6 is produced by CD4+ T-cells, mast cells, monocytes, macrophages, fibroblasts, and endothelial cells [66], and is a multifunctional cytokine in that it participates in growth and differentiation of B- and T-cell effectors, and hematopoetic precursors, as well as by inducing a pyretic acute phase reaction, and promoting tumor growth and angiogenesis [66].

IL-6 expression has been reported increased in HIV-infected thymocytes [67], thymic epithelial cells [68], and in LN of both HIV-infected individuals [69] and rhesus macaques infected with SIV. Furthermore, IL-6 expression has been correlated to viral replication [41], although other studies did not support this observation [70]. Together with IL-6, IL-1β expression was also found elevated in both HIV and SIV infected LN [41] and thymocytes [67]. HAART has been shown to decrease both IL-1β and IL-6 expression in LN, followed by a decrease of the number of LN-associated T lymphocytes with concomitant increase in peripheral blood [71].

IL-7. Thymic epithelial cells represent the main cellular source of IL-7. As validated by studies in immunodeficient individuals [72], IL-7 plays a master role for T cell development and homeostasis [73]

Related to HIV infection, it has been demonstrated that in the thymus IL-7 is crucial for inducing HIV replication [74] and protecting thymocytes from apoptosis by superinduction of Bcl-2 expression [75]. On the other hand, apoptosis by IL-7 has been observed in both astrocytes and neuronal cell line [76] and primary naive and memory T cells [77].

In both HIV+ individuals [80,81] and uninfected children born from HIV⁺ mothers [82] elevated levels of IL- 7 have been reported, thus suggesting that exposure to HIV or HIV proteins across the placenta may perturb thymopoiesis.

In support of the above findings, a correlation between i) circulating levels of IL-7, ii) CD4⁺ T cells loss, and iii) plasma viremia has been demonstrated [83].

Indeed, it has been reported that IL-7 levels before therapy are predictive of virological but not immunological response to ART in both HIV+ children and adults [80,84]. On the other hand, a correlation between plasma levels of IL-7 and increased CXCR4 expression by peripheral blood mononuclear cells (PBMC) has been observed in children infected with X4 viruses, while reversion of their viral phenotype to CCR5 use induced by highly active antiretroviral therapy (HAART) was associated with reduced plasmatic levels of both viremia and IL-7 [85,86]. Although a positive correlation was observed between CXCR4 levels and the identification of X4 viruses [87], the levels of circulating IL-7 were not found correlated to those of CXCR4 on PBMC of infected adults [85,86]. Independently, a decrease of cells expressing the IL-7 receptor has been observed in PBMC and lymphoid tissues of macaques infected with SIV as a function of disease progression [88], likely explaining why elevated levels of IL-7 are not sufficient to maintain T cell homeostasis in this animal model [89].

IL-7 was early investigated in the context of HIV infection, replication and cell depletion in the thymus [75]. This effect at has been recently linked to a skewed enhancement of CXCR4-dependent (X4) virus replication in mature thymocytes concomitantly with an increased expression of CXCR4 [87]. In addition, IL-7 has been shown to upregulate CXCR4 expression in both naive T cells [90], PBMC and thymocytes [87,91], an effect that has been correlated with the emergence of X4 HIV-1 variants in peripheral blood [92]. This hypothesis was supported by the observation of increased plasma levels of both IL-7 and CXCL12 and upregulated expression of CXCR4 and plasma viremia [93].

When ex vivo CD8-depleted PBMC of infected individuals under effective HAART were incubated with various stimuli, IL-7 was identified as the most effective HIV-inductive cytokine in terms of reactivation of latent proviral DNA, while IL-2 alone was ineffective unless combined with the mitogen phytohemagglutinin (PHA) [94]. Furthermore, the analysis of the viral quasispecies induced by IL-7 vs. PHA+IL-2 suggested the possibility that segregated pools of latently infected cells were activated by the two stimuli [94]. However, more recent studies did not observe an inductive effect of IL-7 on latent HIV-infected CD8-depleted cells of infected individuals [95], and synergistic anti-viral effects have been reported by incubating these cells in combination with IFN-α [95].

On the other hand, IL-7 was early documented to increase HIV replication in CD8-depleted PBMC of infected individuals stimulated with anti-CD3 Ab [96,97], to enhance HIV replication in resting naive CD4⁺ T lymphocytes, an effect that was correlated to the autocrine/paracrine release of other cytokines, such as IFN-γ, but also IL-4 and IL-10 [98], and been reported to synergize with TNF-α in sustaining HIV spreading in single CD4⁺ thymocytes by inducing the upregulation of the p75 TNF receptor [74].

Concerning the potential role of IL-7 in the modulation of infection of macrophages, both inhibition [99] and enhancement of virus replication [100] have been reported almost simultaneously. In the latter study, HIV infection or cell stimulation with extracellular Tat upregulated the levels of expression of the IL-7Rα, increasing cellular responsiveness to the related cytokine [100].

In spite of its potentiation of HIV replication, IL-7 remains a potentially important cytokine for the immunologic reconstitution of HIV infected individuals, given its ability to contribute to CTL development [101], and by the fact that in HIV⁺ individuals and children HAART+IL-2 treatment induced significant increase of CD4⁺ cells and plasma levels of IL-7 [102,103], and increased expression of IL-7 receptor (CD127) on CD8⁺ T cells [104].

IL-12 is a heterodimeric cytokine produced by macrophages and DC upon the encounter with the pathogen [105]. It is the most potent inducers of Th1 cell polarization and IFN-γ production, although it also induces proliferation and stimulatory activity on T and NK cells, and CTL maturation [106].

Decreased levels of IL-12 production have been reported from from PBMC of HIV⁺ individuals [107] or SIV_mac251 infected macaques [108,109]. However, even on HIV chronically infected cell lines, cells of HIV infected individuals or SIV infected animals, IL-12 was shown to rescue i) proliferation of T and NK cells, ii) IFN-γ production by T and NK cells, iii) Ag presentation and accessory functions of macrophages and DC, and iv) the cytolytic capacity of CTL and NK cells [107-112]. On the other side, IL-12 has been reported to influence HIV replication, either by enhancing [113], or inhibiting virus replication [114]. The inhibitory effect was correlated to CCR5 down-regulation [115] as consequence of cells stimulation with the cytokine before infection.

Despite the reported decreased production of IL-12 from PBMC of HIV⁺ individuals or SIV_mac251 infected macaques, expression of IL-12 in LN was found to be upregulated during acute SIV infection and this was correlated to the extent of viral replication [41,42].

IL-15 is secreted by macrophages and NK cells and not by activated T cells (like IL-2) and is a Th1 cytokine [116] sharing with IL-2 the βγ chains of the receptor complex produced by MP, thus sharing with IL-2 many activities, such as the ability to induce T cell proliferation and activation [117].

On the other hand, IL-15 is more potent than IL-2 in stimulating both NK cell maturation, differentiation [118] and survival, and NK cells function in HIV+ individuals [117], including the growth and survival of naive rather than memory T cells [118], secretion of IFN-γ and of CCR5-binding chemokines [119]. In fact, in one study, IL-15 stimulation of PBMC enhanced both HIV replication and secretion of CCR5-binding chemokines; however, IL-15 induced levels of CC chemokines (in infected individuals and animals [140]. In fact, it has been proposed that because the lack of Th17 in the gut, thus lacking also recruitment of neutrophils.

On the other side, other authors were not able to show any correlation between viremia and levels of serum IL-17 [141], and even showed that upon SEB stimulation up to 2% of CD4+ T cells were able to express IL- 17 [141]; however, this was a characteristic mostly of “early-infected individuals (less than 1 year of infection)”, and not revealed in the other populations tested (i.e., chronic infection, LTNP, HAART treated individuals) [141]. These data are in partial agreement with those earlier reported, showing a significant increase in the constitutive production of IL-17 from peripheral CD4⁺ T cells of asymptomatic HIV+ and treatment naive individuals, further inducible by PMA/ionomycin stimulation [142]. However, these two last reports do not agree with the ones reported above.

IL-18 is a pro-inflammatory∕Th1 cytokine produced by activated PBMC and epidermal cells that induces the production of IFN-γ from NK cells [143] and enhances the cytolytic potential of both NK cells and of CD8⁺ cytotoxic T lymphocytes [143]. In virtue of its ability to induce Th1 response, IL-18 has been used as an adjuvant either in mice vaccinated with DNA expressing HIV-1 Nef, Gag/Tat/Nef, or Env [144]. As predicted, IL-18 treated animals showed an enhanced cellular response and decreased Ab production [145,146].

In vitro, both acute HIV infection or incubation of the THP-1 cell line with the accessory viral protein Nef induced expression of IL-18 [147]. As most pro-inflammatory cytokines, IL-18 induced HIV expression in chronically infected monocytic [148] and T cell lines [149] via induction of the release of endogenous TNF-α and IL-6 [148], as previously reported for PMA [150]. In contrast, other studies have indicated that IL-18 inhibited the acute in vitro infection of PBMC infection consequently to its capacity of inducing CD4 downmodulation and release of IFN-γ [151], as previously reported for the otherwise HIV-inductive cytokines TNF-α and IFN-γ [152]. These apparent discrepancies are mostly accounted for by the experimental design in that incubation of cells with high concentrations of these cytokines does result in a downregulation of different cell surface receptors, including CD4 and chemokine receptors, whereas stimulation of already infected cells (or lower concentrations of the same cytokines) results in the opposite effect of increasing virus expression, as previously discussed [153].

In vivo, increased serum levels of IL-18 have been observed during PHI/early infection in association with high levels of IFN-γ and reduced expression of CXCR4 from the surface of CD4+ T cells [154]. Although serum levels of IL-18 were found comparable in LTNP and HIV progressors [155], other studies indicated that they were higher in symptomatic vs. asymptomatic HIV⁺ individuals, including children, or healthy individuals [155-159] and that they were correlated to disease progression [158,160]. Similar results were reported in macaques infected with pathogenic vs. non-pathogenic strains of simian HIV (SHIV) [161]. In partial contrast, a reduced capacity of secreting IL-18 from ex vivo stimulated PBMC was observed in those individuals characterized by high circulating levels of this cytokine [157,162,163]. This apparent discrepancy, in addition to a functional exhaustion of PBMC, could be explained by the fact that IL-18 can be secreted by activated platelets [164] as well as by the adipose tissue[165]. Indeed, higher levels of IL-18 have been observed in HIV+ individuals affected by lipodistrophy compared to those without lipodistrophy [165,166] as well as in HIV+ individuals with hypertriglyceridemia [167]. Successful HAART has been linked to reduction of IL-18 circulating levels [155,156,158,160]. In addition, increased IL-18 concentrations have been observed in the cerebrospinal fluid (CSF) of HIV⁺ individuals with opportunistic infections, although not in those with HIV-associated dementia (HAD) [159].

IL-21 is mostly secreted by activated CD4+ T cells, including Th17 cells on which acts as an amplifier of their function [168], and targets several immune cells belonging to both lymphocytic and myelo monocytic lineages, as reviews elsewhere [169]. In fact, it has recently been reported that serum levels of IL-21 are significantly reduced in AIDS individuals and correlate with CD4+ T cell counts [170]. Concerning HIV infection, IL-21 has been shown to augment NK effector functions, such as perforin expression, in chronically HIV+ individuals [171]. Moreover, IL-21 has been shown to enhance CD8+ T cell function, including secretion of IFN-γ [172,173] and expression of perforin without inducing broad cellular activation or proliferation, in HIV+ individuals [174]. Furthermore, IL-21 has been shown to promote the expansion of HIV-specific CD8+ memory T cells [172,173] as well as the Ab-dependent cellular cytotoxicity (ADCC) and Complement-mediated lysis of antigen-expressing cells in synergy with IL-15 [172,173]. Therefore, IL-21 owns the potential for either immunotherapy or as a vaccine adjuvant. In fact, it has recently reported that, both alone and in combination with IL-15, IL-21 increased the magnitude of the response of mice vaccinated with DNA vaccine expressing the HIV-1 Env glycoprotein providing evidence of resistance to viral transmission [172].

IL-27. As for IL-12 and IL-23, IL-27 is also primarily produced by macrophages, monocytes, and DCs following their activation by pathogen recognized through Toll-like receptors (TLRs) [175,176]. In fact, IL- 27 belongs to the IL-6/IL-12 family and can exert both pro- and anti-inflammatory effects as well as affecting T helper cell commitment, T cell proliferation and cytotoxic activity [177,178]. In fact, as for IL-12 and IL- 23, IL-27 induces IFN-γ production by naive T cells and NK cells [175,176,179] and promotes Th commitment toward Th1 cell differentiation and proliferation [175,180,181]. In addition, it has been shown to affect the maturation of the Ab response by inducing isotype switching in B cells [182].

The potential role of IL-27 in HIV infection has been thus far investigated in the context of DC and antigen presentation. When DC derived from IFN-α stimulation were compared to those generated in mice immunized with CD40L and IL-4 they were found to be superior in inducing in vitro cross-priming of HIV- specific CD8+ T cells. Of interest is the fact that this effect was correlated to an enhanced expression of both IL-23 and IL-27 [183]. Furthermore, binding of human papilloma virus (HPV)-like particles (VLPs) to DC and induce the expression of IFN-α, IFN-γ and IL-10 and suppress the replication of both X4 and R5 HIV-1 without affecting the expression of HIV receptors and co-receptors. In addition, these VLPs induced the expression of IL-27 that inhibited HIV-1 replication in activated PBMCs, CD4+ T cells, and macrophages. However, very recently it has been shown that IL-27 inhibits in vitro HIV infection mainly of monocyte- derived macrophages, independently of the expression of IFNs [184].

TNF-a is released in its mature form as a trimer, and is produced by a wide variety of cells including T cells, macrophages [185] and dendritic cells [186]. In macrophages and DC, TNF synthesis can be induced by various pathogens including viruses, parasites, bacteria, LPS and as well as by cytokines (IL-1, IL-2, IFN-γ, GM-CSF, M-CSF, and TNF itself). On the other hand, TNF inhibits interferon-γ priming for production of high levels of IL-12 by macrophages [187]. Thus, by blocking TH1 cytokine production, TNF might limit the extent and duration of inflammatory response in vivo. Thus, chronic TNF exposure suppresses the response of both TH1 and TH2 subsets and attenuates T-cell receptor signalling [188]. TNF is produced in small quantities in quiescent cells, but becomes one of the major factors secreted in activated cells [185]. Thus chronic TNF stimulation suppresses T-cell function in vivo and might have important implications for understanding pathogenesis of chronic inflammatory diseases [188], such as represented by the HIV infection. Anti-inflammatory cytokines, such as TGF-β, IL-4 and IL-10 inhibits TNF-α production in macrophages

The multiple activities of TNF are mediated through two receptors; type 1 TNFR and type 2 TNFR with molecular masses of 60 and 80 kDa, respectively. Both TNFR are type I transmembrane glycoproteins and are present virtually on all cell but red blood cells, although TNFR1 is more ubiquitous, and TNFR2 is often more abundant on endothelial cells and cells of hematopoietic lineage.

TNFR1 is the major mediator of TNF biological functions including the effects on apoptosis and cytotoxicity. On the other hand, TNFR2 signaling appears to be mainly confined to cells of the immune system, and is involved in the proliferation of thymocytes [189], proliferative response of human mononuclear cells [190], in

the induction of GM-CSF secretion, in the inhibition of early hematopoiesis [191], and in downregulating activated T cells by inducing apoptosis [192]. Indeed, the biological response to TNF is believed to be a result of the balance of multiple signals delivered via both TNFR1 and TNFR2. In fact, in latently HIV-infected lymphocytic cells (ACH-2), the TNFR1 plays a major role in stimulation of HIV production [193]. In contrast, when both TNFR are activated simultaneously by agonistic antibodies or co-culture with cells expressing a noncleavable membrane form of TNF, HIV production is downregulated, and cell death is enhanced [193].

HIV infection is characterized by a progressive depletion of CD4⁺ T cells resulting in cellular immunodeficiency [194] and chronic status of inflammation, as highlighted by the increases in both cellular and soluble markers of immune activation, such as cellular markers (CD38, HLA-DR, CD44) and levels of neopterin, β2 microglobulin, soluble CD30, TNF and the soluble form of TNFR2 [195].

In vitro, it has been shown that TNF modulates the viral cycle of HIV-1, in both T cells and macrophages, by targeting two main steps of the viral life cycle such as viral entry and transcription.

TNF has been shown to inhibit entry (either membrane fusion or viral uncoating, but preceeding reverse transcription) of R5 HIV-1 strain into macrophages [152]. On the other hand, both CD4 and CCR5 are downregulated on the cell surface by TNF [196,197], that might result from increased CCR5-binding chemokine production following TNF treatment [198]. Another explanation is that TNF triggers the release of GM-CSF, that has been shown to downregulate CCR5 and subsequently block entry of R5 viruses into macrophages [199].

The second viral step influence by TNF is the viral transactivation. In fact, recombinant TNF stimulates HIV- 1 replication in chronically infected U1 and ACH-2 cell lines through activation of NF-, B and subsequent transactivation of the proviral LTR [150]. Stimulation with TNF of HIV-1 replication in U1 cell lines is mediated by engagement of TNFR1 [200] and activation of NF-kB. Moreover, even HIV expression by the phorbol esters such as phorbol-myristate-acetate (PMA), a broadly used chemical agent that activates protein kinase C, also induced viral transcription and expression in U1 and ACH2 cell lines at least in part by triggering an autocrine pathway mediated by endogenous TNF-α [150]. In addition to cell lines, TNF-α sustains HIV replication also in primary MDM and PBMC [32]. Of note, even TNF-β (also known as lymphotoxin-α, LT-a) a molecule that utilizes the same receptors of TNF-α, showed a similar effects than TNF-α, at least in the chronically infected cells lines [201].

The relevance of TNF-a/b inducing HIV expression via NF-kB is stressed by the fact that clade B HIV-1 possesses two binding sites for NF-kB in close proximity of the transcription start site, whereas clade C virus, spreading in Sub-Saharan Africa, displays three binding regions, and thus express higher levels of virions upon TNF-α stimulation, than clade B or A (1 NF-kB binding site) [202]. Of interest, HIV-2, a related AIDS causing virus characterized by slower disease progression and inefficient vertical transmission from mother to child as well as its closely related SIV have only one NF-kB binding site [203].

Moreover, TNF-α increases HIV replication by inducing NF-kB activation [204]. In tonsil and lymph node, HIV does not influence TNF-α expression [205], although enhances virion trapping [206]. In SIV infection of rhesus macaques strongly correlates with TNF-α expression in lymph nodes [41], whereasTNF-a, induced SIV production from simian alveolar macrophages [207].

Monocytes and tissue macrophages are amongst the major cell sources of this cytokine and their stimulation by bacterial products, such as lipopolysaccharide (LPS), or other microbial agents such as P.falciparum [208], may increase both cytokine expression and HIV replication [209]. In vivo, TNF-α may increase HIV replication at both mucosal [210,211] and systemic [212-214] levels, as well as in HIV spreading in the central nervous system (CNS) [215,216]. However, because the ability of TNF also to inhibit HIV entry, into macrophages, LPS/bacterial infections might result in a controlled viral production within infected macrophages that could be critical to avoid macrophage death and to optimize viral production.

Conversely, at least some HIV strains and/or their gp120 Env [10,31] can upregulate TNF-α secretion from macrophages as well as from T and B lymphocytes [217].

In vivo, a positive correlation between viral replication and the increased levels of TNF-α in the sera [218] LN [41,219], PBMC [220,221], bronchoalveolar lavage mononuclear cells [219], and microglial cells [221] has been shown in macaques infected with SIV. Moreover, high levels of TNF-α[219,222,223] has also been reported in vaginal mucosa, favoring infection of HIV clade C [224-227], possibly in virtue of the transcription binding sites (3 sites for NF-kB and 1 site for AP1) forming its LTR promoter [202]. Converserly, no substantial up- or down modulation of TNF-αexpression was observed in tonsils and LN of infected individuals [205], although an enhanced virion trapping has been correlated with the expression of this cytokine during primary infection [206].