Role of In Vitro Stimulation with Lipopolysaccharide on T-Cell Activation in HIV-Infected Antiretroviral-Treated Patients

We investigated the effect of LPS in vitro stimulation on T-cell activation in HIV-infected patients with different CD4+ recovery on HAART. PBMCs from 30 HIV-positive, HAART-treated, aviremic individuals with different CD4+ reconstitution (Low Responders: CD4+ < 350/μL; Intermediate Responders: CD4+ 350–599/μL; High Responders: CD4+ ≥ 600/μL) were cultured with LPS and the proportion of HLA-DR/CD38- and Ki67-expressing CD4+/CD8+ T-cells was measured (flow cytometry). Upon LPS stimulation, significantly higher CD4+ and CD8+HLA-DR+ cells were shown in LR and IR versus HIV-negative controls. While no differences in the proportion of LPS-stimulated CD4+CD38+ cells were recorded amongst HIV-positive subgroups, CD8+CD38+ cells were more elevated in patients with lower CD4+ recovery on HAART (i.e., LR and IR). Upon in vitro LPS stimulation, HLA-DR and CD38 expression on T-cells are differentially regulated. While HLA-DR induction reflects impaired CD4+ reconstitution on HAART, cell-surface CD38 expression is increased only on CD8+ T-cells, allowing to speculate that the sole induction of CD38 on CD4+ cells may not be sufficient to depict LPS-driven immune activation in HIV.


Introduction
Untreated HIV disease is characterized by high levels of Tcell activation which account for the progressive depletion of CD4+ T-cells [1,2].
Microbial translocation, which occurs following the breakdown of the gastrointestinal barrier, has been put forward as a possible mechanism underlying immune activation in HIV disease [2,3]. Most interestingly, microbial translocation-induced T-cell hyperactivation may also represent a cause of impaired immune restoration on virologically suppressive HAART [4][5][6], as suggested by high levels of circulating lipopolysaccharide (LPS) in patients with poor CD4+ T-cell recovery in course of effective treatment [4].
In keeping with these observations, in vitro exposure of PBMCs to LPS and other Toll-Like Receptor (TLR) agonists has been shown to induce the expression of CD38, indicator of T-cell activation, on CD4+ and CD8+ T-cells in healthy individuals [7], thus suggesting a model for HIV pathogenesis. Indeed, CD38 upregulation on CD8+ cells has been consistently described in HIV infection and correlates with disease progression better than HIV RNA load [8][9][10][11]; conversely, the issue of whether high CD38 expression on CD4+ T-cells is a marker of poor prognosis [8,9,11,12] has been debated for some time [13]. Taken together, these findings imply that the actual biological significance of CD38 in the context of HIV/AIDS remains elusive, given the multiple functions it retains [14]. Indeed, along with its association with T-cell activation, recent data have demonstrated that CD38 on CD4+ cells may also identify a hypoproliferating cell subset, thus offsetting the paradigm of CD38 as T-cell activation marker [13,15].
In the light of these premises, we investigated the dynamics of HLA-DR and CD38 expression on peripheral Tcells following selective in vitro LPS stimulation of PBMCs from HIV-infected HAART-treated individuals. We hypothesized that LPS may account for diverse levels of immune activation in HIV-infected patients according to the degree of immunological recovery in course of suppressive HAART. We also hypothesized that HLA-DR and CD38 may not be equivalent in reflecting the extent of T-cell activation following in vitro stimulation with LPS in the setting of HIV disease.
A graphical time-course representation of the effect of LPS stimulation on CD38 in HIV-infected individuals with different response to HAART and in HIV-negative controls is summarized in Figure 3.        12.6-18.9) compared to LR (5.5%, IQR 1.4-15.5; P = 0.07; Figure 4(b)).
No differences were detected in terms of CD8+ Tcell proliferation among study groups following 48-hour stimulation with LPS (Figures 4(k), and 4(l)).

Discussion
We hereby investigated the induction of CD38, HLA-DR and Ki67 on CD4+ and CD8+ T-cells following selective in vitro LPS stimulation in HIV-infected individuals with different immunological recovery upon virologically suppressive HAART.
We enrolled patients on stable antiretroviral therapy and with comparable demographic and viro-immunological parameters. Of note, patients with less efficient immunereconstitution were significantly older and presented decreased T CD4+ cell nadir compared to subjects displaying sustained CD4+ T-cell recovery; this finding is in keeping with literature data demonstrating that age and CD4+ cell Clinical and Developmental Immunology    . T-cell proliferation levels did not vary according to the degree of immune-reconstitution in course of HAART ((b) and(d)). At T1, LPS stimulation did not account for significant increases in cell proliferation ((e)-(h)). At T2, a trend to increased Ki67 levels on CD4+ T-cells upon stimulation was detected in LR compared to HR and negative controls (j); no differences were detected in terms of Ki67+CD8+ T-cells proliferation among study groups following 48-hour stimulation with LPS ((k) and (l)). counts at HAART initiation are determinants of incomplete immunological reconstitution in course of suppressive HAART [16].
Our data show a differential effect of LPS in vitro stimulation on the cell-surface expression of two T-cell activation markers, that is, HLA-DR and CD38: (i) despite an overall reduction of HLA-DR+ T-cells, patients with an inefficient response to HAART display a higher proportion of activated HLA-DR+CD4+/CD8+ compared to individuals with better CD4+ recovery; (ii) CD38 expression is induced on T-cells, and yet only the CD38+CD8+ pool is significantly expanded according to the degree of immunological impairment.
HIV infection is characterized by high levels of immune activation [1,2] which is a major cause of the progressive CD4+ T-cell loss in untreated disease [1,2,[8][9][10] and impaired immunological recovery in course of HAART [5,17]. A possible driver of immune activation is the translocation of bacterial bioproducts, mainly LPS, through the gastrointestinal barrier to the systemic circulation [2,3]. Indeed, literature findings show that non-progressive disease in SIV natural host primates associates with no microbial translocation and/or immune activation [18,19]; conversely, microbial translocation has been indirectly implicated in driving immune activation in chronically HIVinfected humans and SIV-infected Rhesus macaques [19].
In keeping with this model, in HIV-negative individuals, PBMCs stimulated in vitro with LPS have been demonstrated to increase the expression of CD38, and not HLA-DR [7]. 8 Clinical and Developmental Immunology Consistent with these findings and with recent data showing a dramatic decline in the percentages of DR+ subsets following PBMCs in vitro cultures [20], we hereby show a reduction of HLA-DR+CD4+/CD8+ cells following LPS in vitro stimulation in both healthy donors and HIVinfected individuals. Interestingly enough however, when evaluating HIV-patients according to the degree of immune recovery on HAART, while HLA-DR expression was reduced in patients with good CD4+ recovery, subjects with lower CD4+ reconstitution, (i.e., LR and IR) maintained stable HLA-DR-expressing T-cells. The net outcome of such diverse effect of LPS stimulation according to the extent of CD4+ recovery, is that subjects with poor immunological recovery display significantly higher HLA-DR+ T-cell proportions. Of note, LPS accounted for a higher proportion of proliferating Ki67+ T-cells in the CD4+ subset alone, thus suggesting that the combined effect of microbial stimulation on T-cell proliferation and activation seems to be restricted to the CD4+ cell compartment.
By demonstrating highest LPS-dependant HLA-DR+ Tcells exclusively in patients with lowest CD4+ recovery, our findings add up to the consolidated bulk of evidence on the inverse correlation between HLA-DR expression and CD4+ lymphopenia [10,17]. Thus, HLA-DR might be a faithful marker of T-cell activation secondary to microbial translocation in the setting of severe immune impairment.
A different trend was shown in CD38-expressing T-cells upon LPS stimulation.
The time-course analysis of the effect of LPS stimulation on CD38 expression showed an overall rise in CD38expressing T-cells that was, however, delayed in HIV-infected patients compared to healthy controls (i.e., 48 versus 24 hours). Most interestingly, when investigating HIV-positive subjects with different immune recovery separately, only patients with inefficient CD4+ reconstitution (i.e., LR and IR) displayed an expansion of the CD38+ T-cell pool, whereas individuals with good immunological response (HR) maintained a stable expression of CD38 on T-cells overtime. Of note, although CD38 appeared selectively induced on CD4+/CD8+ cells upon longer stimulation, only CD38+CD8+ cells displayed a linear relationship with immune impairment on HAART, with LR/IR patients showing higher LPS-stimulated CD38+CD8+. Conversely, no differences in CD38+CD4+ were shown according to the degree of immune reconstitution.
These findings are consistent with in vivo data showing that CD38+CD8+ T-cells are an independent predictor of disease progression [8][9][10][11]17]; although CD38 has been described to have a similar prognostic value when expressed on T CD4+ cells, this seems less consistent [8,11,12].
By showing higher LPS-induced expansion of CD38+CD8+ cells, our findings suggest a diverse regulation of CD38 in CD4+ and CD8+ T-cell compartments and allow to speculate that the sole expression of CD38+ on CD4+ T-cells may not be sufficient to portray the state of immune activation in HIV disease. Indeed, CD38 has been shown to be constitutively expressed on phenotypically naïve T-cells [13,15,21], thus offsetting its role as an activation marker in the CD4 pool. Indeed, in patients with impaired immunological recovery in course of HAART, Massanella et al. recently reported lower proportion of CD38+CD45RA+CD4+ and higher percentage of CD38+CD45RA-CD4+ cells, probably reflecting a reduced naive compartment and a higher level of activation in CD45RA-cells [21]. This finding is in keeping with data from Benito et al. [12] who showed that CD38 expression on CD4+ cells correlated with disease progression when associated with the expression of phenotypic markers of mermory/activated (CD45R0 or HLA-DR).
In our study, the co-induction of HLA-DR and CD38 on CD4+ and CD8+ T-cells followed similar kinetics upon longer (48 hours) stimulation with LPS, with increased levels of both molecules in subjects with severe immune impairment on HAART. This finding, together with the differential expression of single cell surface molecules, once again suggests that CD38 on CD4+ cells may not faithfully mirror activation in HIV disease.
The present study has several limitations. Our finding of an in vitro effect of LPS stimulation in reducing the expression of activation markers cannot exclude the possibility of increased cellular death following LPS challenge, as previously shown in HIV-negative patients [7].
Our research was not designed to define the mechanisms underlying variable LPS effects and is therefore descriptive in nature. Functional data on stimulated PBMCs, such as the production of type I interferons and other cytokines, would help give a qualitative answer to what activated cells produce upon microbial challenge. In addition, data on APC (macrophages, dendritic cells) activation and signalling following TLR recognition and on the cooperation between APCs and T-cells would shed light on the precise mechanisms underlying T-lymphocyte activation upon microbial stimulation.
As a further limitation, we chose to investigate the effect of LPS alone on T-cell activation. However, given that the composition of translocating microflora may influence the immunological response in course of treatment [22], it would be interesting to study the effect of other stimuli (e.g., PHA, anti-CD3, PMA) as well as other TLR ligands, representing microbial components from Gram-positive bacteria (e.g., Pam3CysK4, TLR1/2 agonist; FSL-1, TLR6/2 agonist) or viral agents (e.g., ssRNA40, TLR8 agonist; CpG oligonucleotide, TLR9 agonist) on T-cell activation and response to HAART.
Our results thus advocate further functional studies to gain deeper insight into the regulation of T-cell activation by TLR agonists in course of HIV disease and response to therapy. These data might provide the scientific background to investigate/explore alternative/adjuvant therapeutical approaches in HIV infection aimed at manipulating the negative effects of LPS in inducing immune activation [23].