Abnormal Expression Pattern of the IL-2 Receptor β-Chain on CD4+ T Cells in ANCA-Associated Vasculitis

Background/Aim. ANCA-associated vasculitis (AAV) is a small-vessel vasculitis of autoimmune origin. In addition to autoantibodies, T cells have a pivotal pathophysiological role in this disease. T-cell homeostasis and immune tolerance critically depend on IL-2 and its receptor expressed by T cells. In this study, we investigated the IL-2 receptor (IL-2r) expression on CD4+ T cells in AAV. Methods. Thirty patients with AAV and 15 age-matched healthy controls (HC) were enrolled. T cells from peripheral blood were analysed by flow cytometry for expression of the IL-2r α- and β-chain. Results. The IL-2r α-chain was overexpressed in AAV as compared to HC (36 ± 16% versus 20 ± 9%, P < 0.005). The IL-2r-β-chain expression was significantly reduced on CD25+ CD4+ T-cells and CD4+CD25+FoxP3pos regulatory T-cells (Tregs; AAV versus HC: 48 ± 14% versus 62 ± 9%, P = 0.002 and 38 ± 18% versus 68 ± 5%, P = 0.002). Low β-chain expression in AAV was associated with relapsing disease and systemic vasculitis with renal involvement. Conclusion. The IL-2r expression pattern is abnormal in AAV. To our knowledge, we are the first to show that the β-chain expression is drastically diminished on T cells in AAV and related to a less favorable disease course. Given the indispensable function of the β-chain in IL-2 signaling of T cells, diminished expression may contribute to disturbed immune homeostasis in AAV.


Introduction
ANCA-associated vasculitis (AAV) is a necrotizing smallvessel vasculitis of autoimmune origin, which is characterized by the presence of antineutrophil-cytoplasmic-antibodies (ANCA) [1]. ANCA have a dominant pathogenic role in AAV and are either directed against proteinase-3 (PR3) or myeloperoxidase (MPO) [1]. Recent data suggest that, in addition to autoantibodies, T cells are pathogenic factors in AAV [2]. The isotype of the autoantibodies indicates that a T-cell dependent class switch has taken place [3]. Furthermore, T-cell infiltrates are present in organ lesions commonly observed in AAV [2,4,5]. In addition, granuloma formation-which is regarded as a T-cell driven processis a histological feature of specific AAV subtypes [6]. In line with this, several phenotypical and functional T-cell abnormalities have been reported in AAV patients [2,[7][8][9].
IL-2 is essential for the homeostasis Tregs and effector T cells (Teff) [15,16]. There are three different subunits of the IL-2r: -chain (CD25), -chain (CD122), and -chain (CD132, constitutively expressed by all lymphoid cells) [15][16][17]. Both the -chain and the -chain have intracellular, signal transducing domains which are indispensable for proper function of the IL-2r [16]. Interestingly, CD122 is also the -subunit of the IL-15 receptor and mediates signal transduction of this cytokine. Moreover, both the IL-2rchain and the -chain have IL-2 binding domains. The IL-2r exists in two forms: as a heterodimer consisting of the 2 Disease Markers -chain and the -chain forming the low-affinity IL-2r and secondly as a heterotrimer consisting of all three chains forming the high-affinity IL-2r [16]. Teff usually express the low-affinity IL-2r and upregulate the -chain to form the high-affinity receptor only transiently upon activation [16][17][18]. Tregs, which control Teff responses, usually express the high-affinity IL-2 receptor [16][17][18].
In contrast to Teff, the survival and functionality of Tregs-and thus immune tolerance-is critically dependent on IL-2 [15,[19][20][21]. The lack of a functional IL-2r leads to detrimental autoimmunity due to breakdown of immune tolerance [16]. In murine models, the knockdown of the IL-2r -chain results in an expansion of activated Teff and reduced development of functional Tregs followed by lethal, autoimmune organ inflammation [16,22]. This underscores the importance of the IL-2r and especially of the IL-2rchain for T-cell homeostasis and immune tolerance.
In AAV, T-cell homeostasis is persistently disturbed and regulatory mechanisms seem to be impaired [2]. Considering the important role of the IL-2r and in light of the fact that the -chain CD122 has never been studied in AAV, it was the aim of this study to investigate the IL-2r expression on CD4 + T cells in AAV and its implications for disease pathogenesis.

Patient Cohort.
Thirty-one consecutive patients with AAV visiting the outpatient clinic of the Department of Nephrology were enrolled (mean age 58 ± 14 years; 17 males, and 14 females). Four of these 31 patients were sampled during active disease. Three of the four were sampled again during follow-up in remission. Three of the four patients with active AAV were untreated at the time of sampling; one had already received one cycle of intravenous cyclophosphamide. The four active patients presented with a Birmingham vasculitis activity score (BVAS) of 16, 16, 21, and 6, respectively. Twentyseven patients were in remission at the time of sampling and nine were sampled a second time during follow-up. Remission was defined according to Hellmich et al. as the complete absence of active clinical disease reflected by a BVAS of 0 [23]. PR3-ANCA was detectable in 29 patients at the time of diagnosis; two patients had ANCA with specificity for MPO at the time of diagnosis. Sixteen of the patients sampled during remission were treated with mycophenolate mofetil (MMF) at the time of sampling; seven received azathioprine (AZA), four were treated with cyclophosphamide (CYC), one patient was treated with cotrimoxazole, and another one with methotrexate (MTX). Low-dose steroids <10 mg/day were administered to 21 patients with quiescent disease in addition to MMF, AZA, or CYC. Steroids alone without MMF, AZA, or CYC were given to one patient. The diagnosis of AAV was made in accordance with the criteria of the American College of Rheumatology and Chapel Hill consensus [24][25][26]. According to the definitions published by Hellmich et al., AAV was classified as localized disease without renal involvement in eight patients, whereas the remaining patients had systemic AAV with biopsy-proven renal involvement (Table 1) [23]. The mean disease duration at the time of sampling was 99 ± 80 months. Clinical data was obtained retrospectively based on patient file records. According to Hellmich et al., a relapse was defined as reactivation of disease attributable to active inflammation requiring intensified prednisone and/or therapy with AZA, CYC, MTX, or MMF. Applying this definition, 26 relapses in 14 patients were found; the median Birmingham vasculitis activity score was 10.
Fifteen age-matched healthy controls (HC, mean age 53 ± 7 years; 6 males and 9 females) with no history of chronic infections, cancer, or autoimmune diseases were enrolled as control cohort. Patients with IgA nephropathy ( = 18, mean glomerular filtration rate (GFR) = 45 ± 14 mL/min) and six patients with unilateral nephrectomy ( = 5 due to living-related kidney donation, and = 1 due to renal cell cancer 10 years ago; mean GFR = 45 ± 8 mL/min) served as additional control cohorts. Informed consent and approval by the local ethics committee of the University Hospital Essen were obtained.

Flow Cytometry: Surface and Intracellular Staining.
Expression levels of the receptors were measured by multicolour surface staining on unstimulated lymphocytes from whole blood. Phycoerythrin (PE), fluorescein isothiocyanate (FITC), peridin chlorophyll protein (PerCP), and allophycocyanin-(APC-) labeled antibodies with different specificity were used: CD4 (mouse IgG1, PerCP), CD25 (mouse IgG1, FITC), and CD122 (mouse IgG1, PE). Appropriate isotype controls (Becton Dickinson, Heidelberg) were used. Peripheral whole blood was stained with labeled monoclonal antibodies for 20 min at room temperature followed by red blood cell lysis. Intracellular staining for FoxP3 (clone 259D/C7, APC, Becton Dickinson) was performed on ficoll separated PBMC of patients and HC by using a fixation/permeabilization kit according to the manufacturer's instructions. A surface staining was performed with anti-CD4, anti-CD25, anti-CD122, or appropriate isotype controls followed by fixation and permeabilization (FoxP3 Fixation/permeabilization kit, Becton Dickinson, Heidelberg, Germany). Afterwards, intracellular staining with anti-FoxP3 was performed. Measurements were performed with a fluorescence activated cell sorter (FACS) Calibur from Becton Dickinson. The FACS data was analyzed by using the software Flow Jo Version 7.6.5 (Treestar Inc., Ashland, USA). Regulatory T cells were defined as CD4 + CD25 + FoxP3 pos and activated T-helper cells were defined as CD4 + CD25 + FoxP3 neg . The gating strategy for FACS analysis of CD122 expression on CD25 + CD4 + and CD25 neg CD4 + T cells is given in Figures 1(a) and 1(b). The gating strategy to determine expression of CD122 on CD4 + CD25 + FoxP3 + Tregs and CD4 + CD25 + FoxP3 neg activated T-helper cells is given in Figures 2(a) and 2(b).

Statistics.
All values are expressed as mean ± standard deviation. Significance for the differences between groups was determined using the Mann-Whitney test. Matched pair analysis was performed using Wilcoxon signed-rank test. Spearman's rank correlation was applied for detecting correlations between different study parameters.

IL-2 Receptor -Chain Expression on CD4 + CD25 + T Cells Correlates with Relapse Rate and Systemic, Renal AAV.
To further study the clinical implications of the diminished presence of CD122, a correlation analysis was performed. Patients with lower expression of CD122 had experienced more relapses as indicated by the negative correlation of CD122 expression on CD25 + CD4 + T cells and relapse rate ( = −0.55, = 0.004, Figure 5(c)).

Discussion
Our results show that the different subunits of the IL-2 receptor have an abnormal expression pattern in AAV. The -chain of the IL-2r is overexpressed on CD4 + T cells in AAV, whereas the expression of the IL-2r -chain CD122 is significantly diminished on activated T-helper cells and Tregs when compared to healthy controls. Interestingly, low expression of CD122 was associated with increased relapse rate, worse renal function, and renal involvement in AAV.
Our results confirm previous observations which described an overexpression of the IL-2r -chain CD25 on CD4 + T cells in patients with AAV [11,27]. CD25 as part of the high-affinity IL-2r is usually expressed constitutively by Tregs or transiently by activated T-helper cells [16,17]. The enhanced expression of CD25 in AAV is most likely due to persistent activation of Teff as the fraction of Tregs was comparable between AAV patients and HC in line with other studies [7,27,28]. Chronic challenge with antigen may contribute to persistent activation [1,2]. A similar phenomenon has been observed in other human autoimmune diseases [29,30].
We are the first to describe the expression pattern of the IL-2r -chain on CD4 + T cells in AAV. Surprisingly, the expression of the IL-2r -chain CD122 was sharply diminished on CD25 + CD4 + T cells, activated T-helper cells 8 Disease Markers and Tregs in AAV. Under physiological conditions, activated T-helper cells and Tregs should express the heterotrimeric high-affinity IL-2r which consists of all three subunits [15,16]. Accordingly, in HC, more than half of the CD25 + CD4 + T cells and about 70% of the Tregs coexpressed the IL-2r -chain CD122 along with the -chain CD25. CD122 is indispensable for proper function of the IL-2r as it is critically involved in signal transduction with its cytoplasmic domain [16]. Therefore, CD25 + CD4 + T cells and Tregs lacking CD122 may show decreased responsiveness to IL-2. IL-2 is considered as a Treg "growth factor" promoting Treg differentiation and survival [15]. Deprivation of IL-2 due to a dysfunctional receptor may lead to diminished Treg development, Treg apoptosis, reduced suppressive function, or even loss of lineage commitment with permanent downregulation of FoxP3 [15,16,31,32]. Indeed, it has been demonstrated before that Tregs are functionally impaired in AAV and fail to suppress effector T cells [13,14,28]. Moreover, it has recently been reported that decreased IL-10 production is associated with a higher risk for relapse [33]. Reduced sensitivity of Tregs to IL-2 may be one of the mechanisms which contributes to Treg dysfunction and decreased IL-10 production [34]. In line with this, defective IL-2 signaling in type 1 diabetic patients results in loss of the essential transcription factor FoxP3 in Tregs leading to dysfunctional suppression [34,35].
Interestingly, CD122 is also expressed on a regulatory subset of CD8 + T cells [36]. These CD8 + Tregs promote immune tolerance, prevent autoimmunity, and inhibit effector CD8 + T cells which lack CD122 [36]. It is tempting to speculate that the alterations of the -chain expression in AAV may not only affect CD4 + T cells but also CD8 + T cells. However, the focus of our study was CD4 + T cells and thus the impact on CD8 + CD122 + T cells in AAV remains unclear. The reason for the diminished presence of CD122 remains elusive. A genetic basis for AAV was recently demonstrated by a large genome wide association study [37]. Although not proven in the aforementioned GWAS, genetic alterations may play a role in the aberrant expression pattern of the IL-2r in AAV. A previous study by Carr et al. reported an association of a specific gene polymorphism coding for the -chain of the IL-2r and AAV [38]. However, an association of AAV and a specific polymorphism of the gene coding for the -chain of the IL-2r has not been reported so far. We cannot exclude that immunosuppressive treatment influenced the expression of CD122 because the majority of patients were treated with immunosuppressants in our study. Due to the relapsing nature of AAV, patients have to be treated with immunosuppression even if remission is achieved. Longitudinal data covering start of treatment and cessation of immunosuppressants will help to clarify the impact of medication on CD122 expression.
Lastly, CD122 expression on CD25 + CD4 + T cells was stable over time during remission and correlated with the relapse rate. This indicates that the course of the disease is related to CD122 expression. It is not clear from our study if low CD122 expression is the cause or consequence of an increased relapse rate. The value of CD122 as a biomarker for risk of relapse remains to be studied in a prospective study approach.

Conclusion
In conclusion, there is an aberrant expression pattern of the IL-2r in AAV. Whereas the -chain is overexpressed, the expression of the -chain is sharply diminished on activated T-helper cells and Treg. The reduced expression of thechain CD122 was associated with a less favorable disease course. Given the critical role of the IL-2/IL-2r pathway in immune homeostasis, one may speculate that the abnormal expression pattern of the IL-2r may be a contributing factor in the pathogenesis of AAV.