Emergence of Exhausted B Cells in Asymptomatic HIV-1-Infected Patients Naïve for HAART is Related to Reduced Immune Surveillance

Alterations of B cell subpopulations have been described up to date as characterizing advanced stage of HIV-1 infection. However, whether such defects are relevant in subjects with a preserved number of CD4+ T cells (>350 cells/μL) is unclear. In a cross-sectional study, we investigated if signs of B cells exhaustion and impaired viral immune surveillance are present in a cohort of 43 asymptomatic HIV-1-infected patients with preserved CD4+ T cell counts (>350 cells/μL) and highly active antiretroviral therapy (HAART) untreated. A dramatic expansion of exhausted tissue-like memory B cells (CD10−CD21lowCD27−) was observed. B cells alteration was related to an increase in Torque teno virus (TTV) load, used as surrogate marker of immune function. Successfully HAART-treated patients showed normalization of B cell subpopulations frequency and TTV load. These results provide new insights on B cell in HIV-1 infection and show that development of B cell abnormalities precedes CD4+ T cell decline.


Introduction
Through direct or indirect effects, HIV-1 infection leads to several perturbations on most cells of the immune system. The persistent viral replication induces loss of CD4 + T cells and a general chronic cellular activation that affect viability, subsets distribution, phenotype, and function of all the major immune cell populations [1]. B cells exhibit numerous effects of HIV-1-induced immune activation. Increased expression of activation and proliferation markers, polyclonal B cell hyperactivation, and hypergammaglobulinemia are some of the B cell defects detectable in HIV-1-infected patients [2,3]. Moreover, B cells derived from HIV-1-infected patients show several alterations in vitro, as high spontaneous antibodies secretion, defective responses to B cell stimuli, and impaired costimulatory functions [4][5][6][7][8]. These functional defects have been recently linked to perturbation in the frequency of several B cell subpopulations. Naïve B cells (CD21 high CD27 − ) constitute the largest B cell population in the blood of healthy individuals, followed by memory B cells (CD21 high CD27 + ). The frequency of memory B cells is strongly decreased in HIV-1-infected individuals, and this occurrence is accompanied by an increased levels of naïve B cells, activated mature B cells (CD20 + CD21 low CD27 + ), and plasma cells (CD20 − CD21 low CD27 + ) [9][10][11]. Moreover, a relevant expansion of B cell subpopulations characterized by a CD21 low CD27 − phenotype has been recently described in chronically infected patients in advanced stage of HIV-1 infection with a pronounced CD4 + T cell lymphopenia and ongoing HIV-1 replication [12]. This B cell population, which is normally present at low frequency in the peripheral blood of healthy individuals, can be further divided, depending on the expression of CD10, into immature/transitional B cells (CD10 + ) or exhausted tissue-like memory B cells (CD10 − ) [13][14][15].
Torque teno virus (TTV) is a single-strand DNA virus harmlessly carried by the majority of healthy individuals [16]. TTV is considered a surrogate marker of the immunological status of the host [17]. In fact, TTV viral load has been found significantly higher in HIV-1-infected patients presenting AIDS and decreased survival than healthy individuals [18].
Here we demonstrate that asymptomatic patients naïve for HAART with preserved CD4 + T cell counts (>350 cells/μL) show an aberrant B cell subsets distribution and a concurrent increase in TTV viral load compared to HIV-1uninfected individuals as well as HAART successfully treated patients.

Study
Population. Blood samples were obtained from a cohort of 43 chronic HIV-1-infected patients naïve for HAART and without any history of opportunistic infections or malignancies. Clinical features of these patients are described in Table 1. Blood samples were also obtained from a cohort of 20 HAART-treated patients, aviremic for at least 12 months (Table 2), and from 34 healthy laboratory workers. The HIV-1-infected patients enrolled in this study were recruited at the Institute of Infectious and Tropical Diseases of the University of Brescia, provided their written informed consent. The study was approved by the Ethical Committee of the University Hospital of Brescia according to the declaration of Helsinki.

IL-7 Serum Levels Detection.
Detection of IL-7 serum concentration was performed through Quantikine HS IL-7 Immunoassay, according to the protocol provided by the supplier (R&D Systems, Minneapolis, USA).

Virus Detection and Quantification.
HIV-1 plasma viral load was detected using Roche Amplicor PCR kit (Roche-Diagnostic, Milan, Italy). DNA extracted from 200 μL of plasma samples was examined for TTV genome using a single step universal TaqMan real-time PCR assay as previously described [17]. The sensitivity of the assay was of 3.0 log 10 DNA copies/mL of plasma. All samples were tested simultaneously in triplicate.

Statistical
Analysis. Statistical analysis were performed using GraphPad Prism v4.0c for Mac OS X. The distributions of each immune parameter between healthy donors and HIV-1-infected patient cohorts were compared using the nonparametric Mann-Whitney U test. All P values were 2 sided and unadjusted. Correlation between variables was tested using Spearman's rank correlation coefficient test. For all statistical analysis, the 0.05 level of significance was used (P < 0.05). Data in the text were expressed as median.

Several B Cell Subset Alterations in Asymptomatic and
HAART-Naïve HIV-1-Infected Patients. In the present study we examined the frequency and the function of B cell populations in 43 HIV-1-infected patients characterized by being clinically asymptomatic, with CD4 + T cell count >350 cell/μL (median: 510 cell/μL), median viral load of 6779 RNA copies/mL and HAART-naïve. As control, B cell populations were examined in 34 healthy individuals.
We observed that the median of CD19 + B cell counts of the asymptomatic HIV-1-infected patients was superimposable to the median of healthy individuals (129.5 versus 143 cell/μL, resp.) (Figure 1(a)). Moreover, the frequency of B cell subsets, such as mature/activated, immature/transitional or tissue-like memory, resting/memory, and naïve B cells was analyzed in all tested individuals. Although the normal number of total B cells, several perturbations were detected in the B cell subpopulations of asymptomatic HIV-1infected patients. As shown in Figures 1(b) and 1(c) among the CD21 high B cells we observed that the frequencies of both resting/memory (CD27 + ), and naïve (CD27 − ) were significantly lower in the peripheral blood of the asymptomatic HIV-1-infected patients compared to healthy donors (median percentage: 23.5 versus 29.4, P = 0.026, and 54.2 versus 64.4, P < 0, 005, resp.). However, we observed that activated/mature CD21 low CD27 + and immature/transitional or tissue-like memory CD21 low CD27 − B cells were significantly higher in the asymptomatic HIV-1-infected patients compared to uninfected donors (median percentage: 9.6 versus 2.4, P < 0.0001, and 10.4 versus 2.1, P < 0.0001, resp.). A previous study from Malaspina et al. [13] showed that the frequency of CD21 low CD27 − B cells was increased in HIV-1-infected patients in advanced stage of disease (CD4 + T cell count <200 cell/μL). This cell subset included both CD10 + elements, accounting for an immature/transitional B cell phenotype, and CD10 − cells defining exhausted tissuelike memory B cells. In our cohort of asymptomatic HIV-1-infected patients the percentage of CD10 + B cells was particularly low and similar to those of healthy donors. The CD21 low CD27 − B cells subtype was essentially negative for CD10, attesting for an exhausted tissue-like memory B cells phenotype (Figure 1(b)). The percentage of CD21 low CD27 − B cells positively correlated to the HIV-1 viral load (r = 0.39, P = 0.012). A higher percentage of CD21 low CD27 − B cells was detected in patients with high plasma viremia (>10000 RNA copies/mL), and, vice versa, a lower percentage of circulating CD21 low CD27 − B cells was found in patients with low plasma viremia (<10000 RNA copies/mL) (P < 0.01) ( Figure 2). No correlation was observed between the percentage of CD21 low CD27 − B cells and CD4 + T cell count (r = −0.27, P = 0.085, not shown).

Increased Levels of IL-7 in Sera of Asymptomatic HIV-1-Infected
Patients. IL-7 is a pleiotropic cytokine controlling T lymphopoiesis and T cell peripheral expansion through interaction with its surface receptor (IL-7Rα or CD127) [19]. Several investigators observed increased serum levels of IL-7 in HIV-1-infected patients characterized by severe CD4 + lymphopenia, and showed direct correlation between IL-7 serum concentration and viral load, loss of CD4 + T cells, and alteration of B cell subsets [13,20]. Moreover, the frequency of immature/transitional CD10 + CD21 low CD27 − B cells is positively correlated to the IL-7 concentration detected in lymphopenic HIV-1-infected patients [13]. For these reasons, high levels of IL-7 are considered a hallmark of advanced HIV-1 disease. However, we observed a dramatic increase of IL-7 serum levels also in asymptomatic HIV-1-infected patients compared to healthy individuals (median concentration: 12.9 pg/mL versus 1.7 pg/mL, P < 0.01) despite a relatively preserved number of CD4 + T cells (median number count: 510 versus 831 cell/μL) (Figure 3(a)). No correlation between IL-7 serum levels and CD4 + T cell counts or exhausted tissue-like memory B cell subset frequency were detected (Figure 3(b)).

Increase of TTV Load in Asymptomatic HIV-1-Infected
Patients and Correlation with the CD21 low CD27 − B Cells Frequency. So far, our findings showed profound phenotypic alterations in the B cell compartment of patients in the asymptomatic stage of HIV-1 infection. Previous studies showed that B cell subsets, abnormally expressed in the peripheral blood of HIV-1-infected patients in advanced stages of disease, and scarcely detectable in healthy individuals, displayed several functional defects [13,21]. The aim of our study was trying to understand the impact of these uncommon B cell subsets on immunological surveillance in vivo.
TTV is an ubiquitous virus which infects up to 80% of the healthy population [22], and that is present at high titer in the blood of patients in advanced stages of HIV-1 infection compared to healthy subjects. In HIV-1-infected patients TTV load inversely correlated with the CD4 + T cell count [18]. TTV plasma viremia was measured in all individuals enrolled in our study. Asymptomatic HIV-1-infected patients showed a significantly higher amount of circulating TTV compared to healthy individuals (titer median: 5.7 versus 4.6 log DNA copies/mL, P < 0.005, Table 3). A direct correlation was observed between TTV load and both HIV-1 viremia and percentage of CD21 low CD27 − B lymphocytes (r = 0.49 P < 0.01, and r = 0.40, P < 0.01, resp.) ( Figure 4). These data suggest that in asymptomatic HIV-1-infected patients the increased percentage of CD21 low CD27 − B cells may be related to the lack of in vivo immunological control of TTV replication.  P < 0.005, and versus 143 cell/μL, P < 0.01, resp.) ( Table 3). In the cohort of HAART-treated patients we observed no presence of uncommon B cell populations with percentages of CD21 low CD27 − and of CD21 low CD27 + B cells superimposable to those of healthy individuals ( Figure 5). At the same time, the percentage of CD21 high CD27 − B cell subpopulation was higher in HAART-treated than asymp-tomatic HIV-1-infected cohort patients (P < 0.0005), and the IL-7 levels of treated-patients were superimposable to those of healthy donors. Finally, a successful immunological surveillance was observed in HAART-treated patients as they displayed a TTV plasma viremia within ranges commonly observed in healthy subjects (median: 4.1 versus 4.6 log DNA copies/mL).

Discussion
Disturbances in differentiation and function of B cells characterize HIV-1 infection, and mostly, these impairments are correlated to the loss of CD4 + T cells and the increase of HIV-1 load [12]. In the present study, we demonstrate that asymptomatic HIV-1-infected patients naïve for HAART are characterized by normal B cell numbers but impaired B cell subsets frequencies. In particular, we observed that CD21 low CD27 − and CD21 low CD27 + B cells were overexpressed, whereas the frequencies of CD21 high CD27 − and CD21 high CD27 + B cells, which respectively account for naïve and resting memory B cells, were lower compared to healthy donors. The increased uncommon CD21 low CD27 − B cell population in the asymptomatic HIV-1-infected patients consisted exclusively of exhausted tissue-like memory B cells since no CD10 molecule surface was detected in this cellular subset. It is likely that the appearance of CD21 low CD27 + B cells, which have undergone HIV-1-induced activation and differentiation to plasma blasts, and are usually referred as activated mature elements, as well as the appearance of CD21 low CD27 − B cells, which derive from exhausted B cell subpopulation, can be ascribed to the B cell hyperactivation driven by the chronic yet asymptomatic presence of HIV-1   in patients. The presence of exhausted B cells in asymptomatic patients is strongly correlated to the HIV-1 load but not to the CD4 + T cell counts. Immature/transitional CD10 + CD21 low CD27 − B cells, previously described in the peripheral blood of HIV-1 positive patients in advanced stages of disease, were not present in the cohort of asymptomatic patients probably because they appear consequently to a strong reduction of CD4 + T cell counts [13,15]. In fact, immature/transitional B cells were also observed in patients with non-HIV-1-related idiopathic CD4 + lymphopenia suggesting that HIV-1-induced CD4 + lymphopenia (and not HIV-1 viremia itself) drives the expansion of these elements in HIV-1-infected patients [23]. Therefore, our data show that exhausted tissue-like memory B cells emerge independently from immature/transitional B cells in HIV-1-infected patients, and precede the appearance of the latter possibly arising from homeostatic pressures induced by lymphopenia, and/or from conditions that increase the expression levels of early B cell stimulatory factors such as IL-7 [13,23].
Several studies report that CD4 + lymphopenia driven by HIV-1 infection or bone marrow transplantation and idiopathic CD4 + T lymphopenia are often associated with increased serum levels of IL-7 [21,[23][24][25]. In order to explain the increased levels of IL-7 in CD4 + T lymphopenic individuals, it has been suggested that IL-7 augmentation occurs as part of a compensatory effect to enhance differentiation, survival, and expansion of T cells in order to respond to the HIV-1-mediated T cell depletion [24]. Surprisingly, we detected increased levels of IL-7 in sera of asymptomatic HIV-1-infected patients, but no correlations with CD4 + T cells count or HIV-1 load were detected. This result may find a possible explanation in the recent discovery by Guimond et al. [26] that systemic IL-7 concentrations increase solely because of diminished use of the lymphokine by target cells.
So, it is possible to hypothesize that the presence of exhausted memory B cells in asymptomatic HIV-1-infected patients is likely to prejudice the proper functioning of immune cells, and, therefore, a suitable immune surveillance.
So far, no pathological effect can be ascribable to TTV [22,27]. TTV viral load seems to depend on the immunological status of the host, being patients with tumours or HIV-1 infection characterized by a higher TTV load compared to healthy individuals [16,28]. Moreover, recent findings obtained by studying TTV in bone marrow transplanted patients show that the increase of viral loads is correlated with an increase in the percentages of dysfunctional CD8 + CD57 + T lymphocytes circulating in blood [17,29]. In the study we show that asymptomatic HIV-1infected individuals have TTV viral load higher than healthy controls and that its titre is directly correlated to the HIV-1 viral load but not to the CD4 + T cells count. Moreover, TTV viral load directly correlated to the percentage of CD21 low CD27 − B cells. These correlations may attest for a scarce immune control of TTV replication by B cells and for an already compromised immune surveillance in untreated asymptomatic HIV-1-infected patients.
Several studies have shown that HAART-mediated suppression of HIV-1 plasma viremia is followed by normalization of B cell counts and hyperactivation. In addition, reduction of HIV-1 viremia by HAART is associated with decreased frequency of exhausted tissue-like memory B cells [30]. Here we have described that patients who experienced HAART, leading to an undetectable HIV-1 viremia for the last 12 months of therapy, display a normal B cell subsets distribution. The role of HAART in the normalization of IL-7 serum levels is controversial: some investigators showed only a partial decrease in the IL-7 levels, never declining toward normal values whereas others documented a complete normalization of this immunological parameter [19,31]. In our study, successfully HAART-treated patients showed IL-7 values superimposable to those of healthy HIV-1-uninfected individuals.
Interestingly, we reported that all HAART-treated patients experienced TTV load to levels usually detected in healthy individuals. This finding shows that antiretroviral therapies aimed at controlling HIV-1 load and normalizing B cell subpopulations can help to recover immune functions to levels capable of controlling the replication of this endogenous virus. Normalization of B cell counts and subpopulations by HAART is followed by improved B cell responses to both T cell-independent and T cell-dependent immunogens [29,[32][33][34]. Of particular interest is the recent observation that an early control of viral replication through HAART preserves the longevity of B cell responses in vaccinated HIV-1-infected children thus underscoring the direct role of HIV-1 viremia in B cell terminal differentiation [35]. These observations have important implications not only for preserving immune responses against secondary pathogens but also for maintaining or boosting HIV-1 specific immune responses, including antibody responses.

Conclusions
We have provided evidences that asymptomatic HIV-1infected patients with a preserved CD4 + T cell number, and naïve for HAART, display alterations in B cell subset phenotype and impaired immune responses, as manifested by their inability to control TTV replication. On the other hand, normal B cell phenotypes and counts were found in patients who initiated HAART during chronic disease and displayed a negative HIV-1 viral load in the last 12 months of therapy. Normal B cell subpopulations in HAART-treated patients were mirrored by a normal TTV viral load, attesting for a recovery of adequate immunological surveillance. From this point of view, HAART as well as therapies aimed at decreasing immune cell deregulation can help to better control the loss of immune function if administered earlier during HIV-1 infection.