Increased Bone Marrow Interleukin-7 (IL-7)/IL-7R Levels but Reduced IL-7 Responsiveness in HIV-Positive Patients Lacking CD4+ Gain on Antiviral Therapy

Background The bone marrow (BM) cytokine milieu might substantially affect T-lymphocyte homeostasis in HIV-positive individuals. Interleukin-7 (IL-7) is a bone marrow-derived cytokine regulating T-cell homeostasis through a CD4+-driven feedback loop. CD4+ T-lymphopenia is associated with increased free IL-7 levels and reduced IL-7R expression/function, which are only partially reverted by highly active antiretroviral therapy (HAART). We investigated the BM production, peripheral expression and signaling (pStat5+ and Bcl-2+ CD4+/CD8+ T cells) of IL-7/IL-7Rα in 30 HAART-treated HIV-positive patients who did not experience CD4+ recovery (CD4+ ≤200/µl) and who had different levels of HIV viremia; these patients included 18 immunological nonresponders (INRs; HIV-RNA≤50), 12 complete failures (CFs; HIV-RNA>1000), and 23 HIV-seronegative subjects. Methods We studied plasma IL-7 levels, IL-7Rα+CD4+/CD8+ T-cell proportions, IL-7Rα mRNA expression in PBMCs, spontaneous IL-7 production by BM mononuclear cells (BMMCs), and IL-7 mRNA/IL-7Rα mRNA in BMMC-derived stromal cells (SCs). We also studied T-cell responsiveness to IL-7 by measuring the proportions of pStat5+ and Bcl-2+ CD4+/CD8+ T cells. Results Compared to HIV-seronegative controls, CFs and INRs presented elevated plasma IL-7 levels and lower IL-7Rα CD4+/CD8+ cell-surface expression and peripheral blood production, confirming the most relevant IL-7/IL-7R disruption. Interestingly, BM investigation revealed a trend of higher spontaneous IL-7 production in INRs (p = .09 vs. CFs) with a nonsignificant trend toward higher IL-7-Rα mRNA levels in BMMC-derived stromal cells. However, upon IL-7 stimulation, the proportion of pStat5+CD4+ T cells did not increase in INRs despite higher constitutive levels (p = .06); INRs also displayed lower Bcl-2+CD8+ T-cell proportions than controls (p = .04). Conclusions Despite severe CD4+ T-lymphopenia and a disrupted IL-7/IL-7R profile in the periphery, INRs display elevated BM IL-7/IL-7Rα expression but impaired T-cell responsiveness to IL-7, suggesting the activity of a central compensatory pathway targeted to replenish the CD4+ compartment, which is nevertheless inappropriate to compensate the dysfunctional signaling through IL-7 receptor.

In several lymphopenic conditions, the IL-7/IL-7R axis has been shown to be crucial in sustaining peripheral T-cell homeostasis via a rise in circulating IL-7 levels that acts as a survival signal to both lymphoid progenitors and mature circulating T lymphocytes [7,8].
In line with this model, CD4+ T-lymphopenia in the course of HIV disease is characterized by a substantial increase in IL-7 plasma levels, and yet such an IL-7-enriched milieu fails to preserve the peripheral T-cell pool. Aside from the immunodestructive effects of HIV-1, a major cause of the failure of IL-7 to sustain peripheral Tlymphocyte homeostasis might be the down-regulation of IL-7Rca expression and suppression of IL-7Ra function on peripheral T lymphocytes [9,10,11], which might counteract the positive effect of IL-7 on T-cell homeostasis.
CD4+ T-cell reconstitution following the initiation of highly active antiretroviral therapy (HAART) is associated with a normalization of the IL-7/IL-7R axis that comprises reduced circulating IL-7 and increased IL-7Ra expression on T cells, although the levels seen in healthy HIV-seronegative individuals are not reached. However, the altered responsiveness of T cells to IL-7 has been demonstrated in HAART-treated patients, suggesting the persistence of IL-7/IL-7R dysfunction [9,10,11].
Despite the fact that the majority of patients undergoing HAART undergo full viro-immunologic reconstitution, up to 30% of individuals fail to experience peripheral CD4+ T-cell rescue with either suppressed or unsuppressed viremia and have an increased rate of clinical progression [12,13,14,15]. Most recently, inefficient CD4+ gain on HAART has been associated with reduced recovery of T-cell responsiveness to IL-7 [16,17]. Furthermore, patients with failed CD4+ recovery on HAART display specific alterations in the bone marrow, the primary organ contributing to IL-7 synthesis, which includes altered clonogenic capability, stromal cell dysfunction and imbalanced cytokine milieu [18,19].
Collectively, these findings suggest a crucial influence of the IL-7/IL-7R axis on CD4+ T-lymphocyte reconstitution, and yet the reciprocal interactions of IL-7/IL-7R expression within lymphoid organs and peripheral blood are still poorly understood. The levels of circulating IL-7 might either reflect changes in cytokine production at IL-7-producing sites via a feedback loop with peripheral T-cell pool [1] or follow the dynamics of IL-7Ra cellsurface expression [20,21]. Analogously, the level of IL-7Ra expression is regulated by both the free IL-7 level [11,22] and HIV-mediated effects [23,24].
We aimed to comprehensively investigate the production, peripheral dynamics and function of IL-7/IL-7R in HIV-positive patients failing to recover CD4+ counts following initiation of HAART. In particular, as HIV antigenemia has been shown to affect both bone marrow function and IL-7R peripheral expression [24,25], we reasoned that HAART-treated lymphopenic patients with full viremia suppression (immunological nonresponders, INRs) might feature a different IL-7/IL-7R profile than individuals with complete immune-virological failure (complete failure, CFs).

Patient characteristics
Thirty HIV-positive patients were recruited: 18 INRs and 12 CFs. Subjects were comparable with respect to age, sex, risk factors for HIV, AIDS diagnosis, and HAART length and regimen at the time of analysis (Table 1).
As per inclusion criteria, CFs displayed significantly higher mean HIV RNA levels (p = .0001) and comparable mean absolute and percent current CD4+ counts (p = .4) and nadir (p = .2) ( Table 1). No differences were shown with respect to HAART regimen and duration between the groups; 83% of the patients (25/30) were on a protease inhibitor (PI)-based regimen (Table 1).
INRs had high circulating IL-7 levels and low IL-7Ra expression on T cells

INRs had reduced IL-7Ra expression by PBMCs
According to cell-surface expression, HIV-positive patients presented significantly lower IL-7Ra mRNA levels than HIVseronegative subjects (HIV positive: 138.76109.9; HIV seroneg-  Figure 1D).

INRs displayed augmented IL-7/IL-7Ra expression in bone marrow mononuclear cell (BMMC) cultures
Having shown differences in IL-7/IL-7Ra expression and production in peripheral blood, we investigated the expression at the bone marrow level. According to data for the periphery, HIV-positive patients as a whole presented a trend toward higher IL-7 levels in BMMC supernatants compared to HIV-seronegative subjects, though the difference was not statistically significant (HIV positive: 2.867.7 pg/ml, HIV seronegative: 0.560.4 pg/ml; p = .7) (

INRs had diminished levels of pStat5+CD4+ and Bcl-2+CD8+ T cells following IL-7 stimulation
Given the higher IL-7 BM production despite low CD4+ counts in INRs, we investigated the ex vivo response to IL-7.
Compared to HIV-seronegative subjects, INRs had higher constitutive pStat5+CD4+ levels, whereas no differences were shown between constitutive pStat5+CD8+ T cell levels  Figure 3D). No major changes in the proportions of Bcl-2+ and CD8+ T cells were observed following IL-7 stimulation in both patient groups (p = .2 and p = .9 for INRs and HIV-seronegative subjects, respectively) ( Figure 3D).

Discussion
The failure to gain CD4+ T cells and the time spent with nonprotective CD4+ counts following HAART initiation are associated with increased morbidity and mortality [26,27], making the identification of alternative therapeutic approaches to promote immune reconstitution essential. Recently, adjuvant recombinant human IL-7 (rhIL-7) was demonstrated to be effective in sustaining CD4+ recovery in HIV-positive patients with different degrees of immune deficiency [28,29]. Theoretically, due to its role as a major regulator of T-lymphocyte homeostasis, IL-7 might be particularly helpful in patients on HAART with persistent CD4+ lymphopenia, but the actual clinical value of IL-7 treatment in these patients is questioned by the evidence of already elevated endogenous IL-7 levels and reduced IL-7Ra expression [16,17,28,29]. Therefore, we studied the bone marrow expression and peripheral levels of IL-7 and IL-7Ra in HIV-positive patients failing to recover CD4+ counts following initiation of HAART with and without ongoing HIV viral replication.
In addition to confirming the elevated plasma IL-7 levels and reduced IL-7Ra expression in circulating T cells in HIV-positive individuals [10,20,30,31], we found that HIV-positive patients had levels of bone marrow IL-7 production comparable to that of healthy HIV-seronegative controls, with a trend for higher IL-7Ra bone marrow expression compared to HIV-seronegative individuals. As these findings were unexpected given the altered bone marrow architecture and function resulting from HIV infection [18,32,33,34,35], we separately analyzed HIV-positive individuals to test the hypothesis that the bone marrow functions of subjects differed according to the immune-virological response to HAART.
Despite a similar peripheral profile characterized by severe CD4+ T-lymphopenia, heightened plasma IL-7 levels and reduced IL-7Ra+CD4+/CD8+ T cell levels, the investigation of the bone marrow compartment revealed divergent IL-7/IL-7R production between INRs and CFs.
Patients with complete viro-immunological failure to HAART displayed totally hampered production of IL-7 and IL-7Ra, likely due to the complete loss of function of both bone marrow stromal cells and peripheral lymphocytes. This finding is in line with evidence suggesting that during the course of HIV infection, bone marrow stromal auxiliary cells are persistently infected and dysfunctional, impairing the marrow's hematopoietic functions [18,35,36]. Thus, elevated levels of circulating IL-7 in CFs likely result from release by down-regulated cell-surface IL-7Ra levels and not from increased production.
Our findings indicate the activity of opposing pathways in INRs, in which stromal cell production of both IL-7 and IL-7Ra seems to be elevated. Newly produced IL-7 might be released into the peripheral blood and rapidly bind to cell-surface receptors, resulting in the down-regulation of IL-7R in both CD4+ and CD8+ T cells [37].
With this perspective, equally reduced peripheral IL-7Ra production and expression reflect dissimilar immune pathways. In CFs, reduced peripheral IL-7Ra production and expression might be due to IL-7Ra down-regulation driven by both elevated circulating IL-7 levels, a direct viral effect [24,25], and HIVinduced immune activation [22,23], whereas in INRs, reduced peripheral IL-7Ra production and expression might result from down-regulation of IL-7 receptor expression on CD4+ and CD8+ T cells following IL-7/IL-7R engagement [37].
Aiming to reconcile increased IL-7/IL-7R production with persistent CD4+ T-lymphopenia in spite of removing the viral challenge in INRs, we investigated the effect of reduced IL-7mediated signaling on T-lymphocyte homeostasis.
When compared to HIV-seronegative subjects, INRs displayed higher constitutive pStat5 levels exclusively within CD4+ cells, suggesting the persistence of HIV-induced up-regulation of Statdependent signaling pathways despite viremia suppression [38,39]. However, IL-7 stimulation did not result in increased pStat5+ Tcell proportions, confirming the reduced IL-7 responsiveness of the peripheral T cells [16,17].
Our data suggest a model whereby CD4+ breakdown in the absence of viral challenge is associated with compensatory pathway(s) at the level of bone marrow stromal cells, resulting in increased IL-7 production and activated Stat5 signaling, specifically in CD4+ T cells. It is tempting to speculate that this is a compensatory pathway targeted to maintain peripheral Tlymphocyte homeostasis. However, this supposition may be inappropriate for continuous CD4+ exhaustion given the failure of the IL-7-mediated peripheral signaling. One possibility is that with the presence of such compensatory constitutive bone marrow activation, the IL-7/IL-7R system might experience a ''functional exhaustion'' with respect to the continuous demand provided by peripheral T-lymphopenia, resulting in an overall desensitization to further IL-7 stimulation.
Further studies on the regulatory pathways behind bone marrow IL-7/IL-7R production and IL-7Ra-mediated signaling in HIV/AIDS patients are needed to identify the most efficacious clinical use of IL-7 in the (adjuvant) treatment of HIV-positive patients.

Patients
Between October 2006 and June 2009, we performed a crosssectional study including 30 HIV-positive patients who were followed at the Clinics of Infectious Diseases, ''Luigi Sacco'' and ''San Paolo'' Hospitals, University of Milan. Inclusion criteria were stable HAART for at least six months and CD4+ #200/mL over the last six months. Based upon virological response to sixmonth HAART, patients were divided into two groups: INRs (HIV-RNA#50 cp/ml) and CFs (HIV-RNA.1000 cp/ml).
As controls, we included 23 HIV-seronegative patients undergoing heart surgery.
All enrolled patients provided written informed consent according to the Ethics Committees of our Institutions (Comitato Etico, Ospedale ''San Paolo'' and Comitato Etico, Ospedale ''Luigi Sacco'', Milan, Italy). Both ethics committees specifically approved this study.

Cultures of bone marrow mononuclear cells (BMMCs)
Bone marrow aspirates were collected in EDTA. Ficollseparated BMMCs were cultured in 12-well plates in IMDM medium containing 20% FCS, 100 UI/ml L-glutamine and 100 UI/ml penicillin-streptomycin at a concentration of 1610 6 cells/ml in a total volume of 3 ml/well.

Spontaneous IL-7 production by cultured BMMCs
BMMC cultures were incubated at 37uC in humidified air with 5% CO 2 in the absence of stimuli to verify spontaneous production of IL-7. Briefly, 24 hours after BMMC confluence, supernatants were collected and IL-7 levels were measured by ELISA (Quantikine HS human IL-7; R&D, Milan, Italy) according to the manufacturer's instructions.

IL-7 mRNA quantification from BMMC-derived stromal cells (SCs)
Following BMMC culture as described above, nonadherent cells were removed from cultures at specific intervals and were replaced by 1000 ml of fresh supplemented IMDM. The cultures were subsequently maintained until stromal confluence (3-4 weeks). BMMC-derived SCs were collected by trypsinization and were used to quantify IL-7 mRNA by real-time PCR. (i) RNA extraction and reverse transcription. Total RNA was isolated from BMMC-derived SCs (TRIzolH, Invitrogen, Carlsbad, CA, USA). Potential genomic DNA contamination was removed by incubation with RNase-free DNase I (Invitrogen, Carlsbad CA, USA). From each sample, 4 to 8 ml of total RNA (corresponding to about 250-500 ng) was used for cDNA synthesis using the SuperScript TM III First-Strand Synthesis System for RT-PCR according to the manufacturer's instructions (Invitrogen, Carlsbad CA, USA). (ii) IL-7 mRNA quantification. A portion of the cDNA mixture, corresponding to 1/3, 1/6, 1/21, or 1/210, was subjected to SYBR green-based IL-7 and GAPDH mRNA real-time PCR [40]. All amplifications were carried out on a 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) in a final volume of 50 ml with 100 nM of primers using the Hot-Rescue Real-Time PCR Kit-SG (Diatheva s.r.l., Fano, Italy). For IL-7 amplification, after one cycle at 95uC for 10 min, a twostep PCR procedure was used consisting of 15 sec at 95uC and 45 sec at 68uC for 45 cycles. For GAPDH amplification, the cycling conditions were 10 min at 95uC followed by 45 cycles of 15 sec at 95uC, 15 sec at 60uC and 35 sec at 72uC. To verify the specificity of the PCR products, a dissociation curve analysis was performed. He IL-7 and GAPDH primers amplified 157-bp and 176-bp cDNA fragments, respectively [41]. An external RNA standard curve was used to quantify IL-7 mRNA levels. Ten-fold serial dilutions from 10 7 to 10 3 of synthetic IL-7 RNA molecules synthesized using the T7 RiboMAX TM Express Large-Scale RNA Production System (Promega, Milan, Italy) were reverse transcribed and amplified in the same run with unknown samples. For data normalization, variable amounts (500-100-50-10-5 ng) of mRNA (extracted from the PBMCs of healthy donors) were used to generate a ''gold standard'' for GAPDH quantification. The standard curves were created automatically by the Applied Biosystems software based upon threshold (Ct) values. A curve was accepted when the slope was between -3.50 and -3.32 (93-100% efficiency) and when the minimum value of the correlation coefficient (R 2 ) was 0.98. Data were expressed as IL-7 mRNA copy number/10 ng GAPDH mRNA.

IL-7Ra mRNA quantification from PBMCs and BMMCderived SCs
Following RNA extraction/reverse transcription as described above, detection of IL-7Ra mRNA in PBMCs and BMMCderived SCs was performed using the Assay on Demand Kit (ID: Hs00233682_m1 Applied Biosystems). The ''gold standard'' described above was used for IL-7Ra and GAPDH mRNA quantification. Data were expressed as ng IL-7Ra mRNA/10 ng GAPDH mRNA.

Statistics
Baseline categorical and continuous parameters were compared using Fisher's exact test, the Pearson chi-square test and Student's t test for independent samples, respectively. Variables for HIVpositive patients and healthy controls were compared with Student's t test and the Mann-Whitney nonparametric U test. The same variables were also compared between HIV-positive subjects, INRs and CFs using one-way ANOVA followed by the Games-Howell post hoc test for multiple comparisons. For comparisons between the parameters of stimulated and unstimulated cells, the Wilcoxon matched-pairs signed rank test was used. Statistics were calculated using SPSS software (version 18.0).