Introduction.

After infection with human immunodeficiency virus type 1 (HIV-1), CD4+ T lymphocytes decline in number at a variable rate unless treatment is started with antiretroviral therapy (ART) [1]. One of the strongest predictors of disease progression is T cell activation. For reasons that remain to be elucidated, the frequency of circulating cells that express CD38 and HLA-DR (both of which are thought be markers of cell activation) predict rate of CD4+ T cell decline and disease progression in both untreated and treated disease; importantly, this effect appears to be independent of viral load [2]–[5]. Patients with advanced HIV-1 disease as measured by low CD4+ T cell blood count have the highest levels of T cell activation, concomitant with rapid proliferation and apoptosis [6], [7], enhanced CD4+ T cell susceptibility to infection [7].

Chronic cellular immune activation is accompanied by expression of apoptosis markers such as CD95 and PD-1, and by proinflammatory cytokines [8]–[13]. T lymphocytes expressing CD38 show an abnormal cytokine expression [14], [15] suggesting that HIV-1 perturbs cell function as well as cell number. HIV-1-associated inflammation invariably declines during effective antiretroviral therapy, but often remains higher than that observed in uninfected controls [16]. This residual activation has been associated with cardiovascular disease, malignancy, renal disease and accelerated aging [17].

The mechanisms by which HIV-1 infection leads to immune activation have not been completely defined, although several possible pathways have been postulated. Recent attention has focused on the translocation of microbial products from the intestine to the systemic circulation due to a disrupted mucosal barrier [18], [19]. Another potential contributor to T cell activation is reactivation of CD8+ responses specific for latent infections such as cytomegalovirus (CMV) and Epstein-Barr virus (EBV) [20]–[22]. Consistent with this hypothesis is the observation that most activated CD8+ T cells are antigen-specific [23].

In this work we examined the possibility that human endogenous retroviruses (HERVs) could also be contributing to T cell immune activation.

HERVs are proviruses of ancestral infections of germ cells that have been incorporated into the genome. About 8% of the genome is comprised of sequences thought to derive originally from these retroviral infections [24], and about 1.1% of the genome is comprised of recognizable open reading frames (ORFs) of retroviral origin [25]. There are several families of HERVs, usually named with one amino acid letter by the host's tRNA assumed to be required for priming reverse transcription at identification time, and encompassing similarities with all retroviral genera except lentiviruses (including HIV-1) [26]. HERV expression has been associated with several autoimmune disorders such as multiple sclerosis [27] and rheumatoid arthritis [28]. Also, HERV transcripts have been found in biopsies from lymphomas, breast cancer, melanoma and teratocarcinomas [29]. Some evidence points to an association of HERV with schizophrenia [30]. With regard to infectious diseases, HERV expression has been shown to be increased in HHV-6 A and B, Epstein-Barr virus (EBV), herpes simplex virus and HIV-1 [31], [32]. The exact role that these associations have on disease mechanisms, if any, remains unclear and in some cases controversial [33].

The HERV-K family of endogenous retroviruses are probably the most recently incorporated, with some inserts having occurred after the rise of hominids [34]. The first association between HERV-K products and HIV-1 infection was reported in 1996 by Löwer and colleagues [35], who reported an observation where there was increase in HERV-K-specific antibodies in HIV-1 infected individuals. Ten years later Contreras-Galindo et al. [31] found that HERV-K RNA plasma expression was higher in HIV-1 infected patients than healthy controls, and that this expression could be induced in vitro by HIV-1 infection [36]. Recently, they reported specific HERV-Ks being differentially expressed in HIV-1 infected patients, but not in breast cancer patients, and the presence of molecules that bind HERV-K env antibodies [37]. Increased HERV-K expression was confirmed by our group in a study in which we also identified T cell responses to HERV peptides in subjects with early HIV-1 infection [38]. We have also recently demonstrated that HERV-specific T cells responses persist in chronic HIV-1 infection and are associated with improved virologic control [39]. Interestingly, in addition to HERV-specific responses being highly increased in HIV-1-infected patients compared to healthy donors, within the HIV-1 infected cohort, the responses were much higher in individuals who naturally control HIV replication (“controllers”) than in those who failed to control virus replication and were almost completely absent in patients with advanced disease (as defined by high viral load and low CD4+ T cell count). This suggests that anti-HERV-K T cell immunity could play a role in HIV-1 control (reviewed in [40]).

Since immune responses against chronic latent viral infections such as EBV and cytomegalovirus (CMV) can be reactivated during HIV-1 infection and induce immune activation of their cognate T cells [20], and virus-induced activation can be seen in other viral infections such as CMV [41] we hypothesized that HERVs could act in a similar manner and thereby cause an increase in generalized immune activation in HIV-1 infected subjects We focused on HERV-K env and gag RNA transcripts, since these proviruses have been found to be transcribed in multiple reports, and given their relative new incorporation, are likely to have intact open ready frames. In addition, HERV RNA could interact directly with innate viral recognition through such mechanisms as TLRs in dendritic cells [42]. Therefore, we hypothesized that HERV expression would be a driver of chronic immune activation in individuals with untreated HIV-1 infection.

We found that this hypothesis was refuted by our data, since the relationship was in the opposite direction, and that HERV-K RNA expression was inversely correlated with immune activation in HIV-1 disease.

Abstract.

Human endogenous retroviruses (HERV) are remnants of ancestral retroviral infections integrated into the germ line, and constitute approximately 8% of the genome. Several autoimmune disorders, malignancies, and infectious diseases such as HIV-1 are associated with higher HERV expression. The degree to which HERV expression in vivo results in persistent inflammation is not known. We studied the association of immune activation and HERV-K expression in 20 subjects with chronic, untreated progressive HIV-1 infection and 10 HIV-1 negative controls. The mean HERV-K gag and env RNA expression level in the HIV-1 infected cohort was higher than in the control group (p = 0.0003), and was negatively correlated with the frequency of activated CD38+HLA-DR+CD4+ T cells (Rho = −0.61; p = 0.01) and activated CD38+HLA-DR+CD8+ T cells (Rho = −0.51; p = 0.03). Although HIV-infected persons had higher levels of HERV-K RNA expression (as expected), the level of RNA expression was negatively associated with level of T cell activation. The mechanism for this unexpected association remains to be defined.