Abnormal interleukin-7 function in common variable immunodeficiency

Common variable immunodeficiency (CVID) is a heterogeneous group of immunodeficiency disorders,¹  characterized by deficient B-cell function. However, about half the cases show signs of T-cell deficiency,²  possibly contributing to the defective antibody production.

Interleukin-7 (IL-7) plays a critical role in lymphocyte development and homeostasis.³  High plasma levels of IL-7 have been correlated with low CD4⁺ T-cell counts (eg, in HIV infection), reflecting a homeostatic response to T-cell depletion.^4-6  However, high IL-7 levels may also lead to autoimmunity and ectopic lymphocyte infiltrates, as suggested in animal models examining the effects of exogenous IL-7 administration.⁷  Deficiencies of the IL-7 receptor α chain (IL-7Rα) or the γ common chain (γc) may lead to severe combined immunodeficiency (SCID). Interestingly, a partial IL-7Rα deficiency results in a delayed clinical presentation of immunodeficiency, and may additionally cause hypogammaglobulinemia.⁸ 

Although IL-7 seems to be crucially involved in the function of human lymphocytes and defects in the IL-7 signaling system in humans have been described, the role of IL-7 in CVID has not been studied. Herein, we examine the expression of IL-7 and its role in T-cell function in CVID.

Patients with CVID¹  without clinically apparent infections were included (Table 1). Healthy donors were included as controls. Informed consent was obtained from all subjects before inclusion. The study was conducted according to the ethical guidelines at our hospital and the Helsinki declaration, and approved by the Rikshospitalet University Hospital's authorized representative.

Negative selection of CD3⁺ T cells was performed as previously described.⁹  The CD3⁺ T cells were resuspended in RPMI 1640 with 2 mM L-glutamine and 25 mM HEPES (N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid) buffer (Gibco, Paisley, United Kingdom) supplemented with 10% fetal calf serum (FCS) and incubated in 96-well plates (10⁶ cells/mL). The cells were stimulated with anti-CD3 (SpvT3b; final concentration, 40 ng/mL for optimal stimulation; suboptimal stimulation, 10 ng/mL; Zymed, San Francisco, CA), anti-CD28 (clone 15E8; CLB, Amsterdam, the Netherlands; final concentration 50 ng/mL for optimal stimulation; not added for suboptimal stimulation) with and without IL-7 (R&D Systems, Oxon, United Kingdom) and cross-linked.¹⁰  For cytokine and lactate-dehydrogenase (LDH) analyses, cell-free supernatants were harvested after 24 (interferon γ [IFNγ] and IL-2) and 48 (transforming growth factor β1 [TGFβ1]) hours. Proliferation analyses were performed as previously described.¹¹ 

Absolute numbers of circulating lymphocyte subsets were assessed using TruCount and TriTEST (Becton Dickinson, San Diego, CA). Other staining for flow cytometry was performed using fluorescein isothiocyanate (FITC)-conjugated anti-CD4; phycoerythrin (PE)-conjugated anti-IL-7Rα and anti-Ki67; peridinin chlorophyll-alpha protein (PerCP)-conjugated anti-CD3 and allophycocyanin (APC)-conjugated anti-CD8 (Becton Dickinson); and FITC-conjugated anti-CC chemokine receptor 7 (CCR7; R&D Systems). Apoptosis was estimated using annexin V/propidium-iodide staining (Becton Dickinson). Flow cytometry was performed as previously described.¹⁰ 

Cytokine levels were measured by enzyme-linked immunosorbent assay (ELISA; R&D Systems). LDH activity was measured using Cytotoxicity Detection Kit (LDH; Roche-Applied-Science, Basel, Switzerland).

Data are presented as median [25th-75th percentile]. The 2-tailed Mann-Whitney U test, Wilcoxon signed-rank test, chi-square test, and Spearman signed rank test were used as appropriate. Tests were considered significant when P is less than .05.

We found significantly higher plasma levels of IL-7 in CVID patients (n = 72) compared with healthy controls (n = 23), with particularly high levels in 13 patients (> 10 pg/mL; hereafter termed CVID^IL-7high; Figure 1A). Except for a higher proportion of males in the CVID^IL-7high group (P = .026), there was no association between IL-7 levels and clinical characteristics of the study group (Table 2). In contrast to previous reports in T-cell-depleted individuals, we did not find any correlation between IL-7 levels and CD3⁺ or CD4⁺CD3⁺ T-cell counts in CVID (Figure 1B). In fact, the 13 CVID^IL-7high patients had significantly higher CD8⁺ T-cell counts than the remaining patients (1043 [299-2040] vs 531 [353-722] CD3⁺CD8⁺ cells/mm³; P = .02). There were no different proportions of CD45RA⁺ T cells, but the predominance of CCR7^- effector-memory T cells^11,12  was significantly stronger in the CVID^IL-7high patients (88% [75%-90%] vs 68% [45%-90%] CCR7^- of CD3⁺ cells; P = .01). There was no difference in the expression of the proliferation marker Ki67 between CVID and controls or between the CVID^IL-7high and the remaining patients, indicating that the increased numbers of CD8⁺ T cells were not due to increased proliferation in vivo. However, the proportion of apoptotic cells after in vitro isolation correlated inversely to IL-7 levels in the CVID patients (r = -0.080; P = .055), suggesting a protective effect of IL-7. There was no difference between CVID patients and controls in the expression of IL-7Rα on T cells, and there was no association between in vitro responses of IL-7 in proliferation or cytokine assays and the expression of IL-7Rα.

In further in vitro analyses of 16 randomly selected CVID patients, we found an inverse correlation between IL-7 levels and in vitro T-cell proliferation in the CVID^IL-7high patients (r = -0.55, P = .034). While IL-7 had no effect on T-cell proliferation when given alone or in combination with optimal stimulation in patients or controls, it strongly enhanced proliferation in suboptimally stimulated T cells in healthy controls (Figure 1C). In contrast, while there was a heterogeneous response among CVID patients with normal IL-7 levels, none of the CVID^IL-7high patients reached similar levels (Figure 1C).

While the addition of IL-7 in vitro increased IFNγ secretion in optimally stimulated T cells (fold change, ∼ 2), there was a particularly strong response to IL-7 in suboptimally stimulated T cells in both CVID patients and controls (fold change, ∼ 12; Figure 1D). Notably, in the CVID group there was a significant correlation between IL-7-stimulated proliferation and IFNγ response (fold change in suboptimally stimulated T cells; r = 0.71, P = .01), and all 4 individuals with poor IL-7 response in IFNγ secretion also had a poor IL-7 response in the proliferation assay at suboptimal stimulation.

We found that T cells from CVID patients spontaneously secreted less TGFβ1 than cells from healthy controls. However, while adding IL-7 significantly increased TGFβ1 secretion in T cells from healthy controls in all conditions, no consistent effect of IL-7 was seen in CVID, although again, the effect varied within the CVID group. Interestingly, TGFβ1 has been demonstrated to inhibit secretion of IL-7 by bone marrow stromal cells,¹³  indicating a possible feedback mechanism from T cells for regulation of IL-7.¹⁴  While there was a positive correlation between plasma levels of IL-7 and in vitro secretion of TGFβ1 in patients with normal levels of IL-7, this correlation was disrupted in the CVID^IL-7high patients (Figure 1E), suggesting a deficient feedback mechanism for IL-7 secretion in these patients.

Finally, the impaired responses to IL-7 in T-cell proliferation and IFNγ secretion were not due to increased in vitro cell death as assessed by LDH levels in cell culture supernatants. In fact the LDH concentration was lower in CVID^IL-7high patients suggesting protective effects of IL-7.

We describe a subgroup of CVID patients with high plasma levels of IL-7, associated with elevated numbers of CD8⁺ T cells and increased proportion of terminally differentiated effector-memory T cells^11,12  that do not show signs of increased proliferation, but appear to have reduced apoptosis. Some of these patients also have reduced response to IL-7 in vitro. Our observations suggest abnormalities in IL-7-mediated lymphocyte homeostasis in some CVID patients, possibly involving mechanisms such as abnormal signaling pathways (similar to certain forms of SCID),⁸  or impaired feedback to IL-7-producing stroma cells due to primary T-cell defects.^11,15-18  IL-7 also plays a role in APC-mediated T-cell stimulation.³  Thus, the IL-7 dysfunction may also be involved in the defective dendritic cell function recently described in CVID.^19,20  Considering the role of IL-7 in both development and peripheral homeostasis of lymphocytes,³  it is conceivable that the abnormal IL-7 function as demonstrated in this study may be involved in the skewing of lymphocyte subsets and in the development of an impaired T- and B-cell function in some CVID patients.