Soluble interleukin-7 (IL-7) receptor α (sCD127) is implicated in the pathogenesis of autoimmune diseases. We show that serum sCD127 concentrations are increased at the onset of type 1 diabetes (T1D; n = 390) as compared with concentrations in age-matched islet autoantibody–negative first-degree relatives of patients (n = 392; P = 0.00001). sCD127 concentration in patients was influenced by islet autoantibody status (P = 0.003) and genotype of the rs6897932 single nucleotide polymorphism within the IL-7RA gene (P = 0.006). Release of sCD127 in vitro was strongly upregulated by activation of T lymphocytes and affected by exposure to cytokines. sCD127 bound IL-7 and was antagonistic to IL-7 signaling and IL-7–mediated T-cell proliferation, suggesting a regulatory feedback mechanism on T-cell expansion. Remarkably, high glucose led to a glycated form of sCD127 that was ineffective as an IL-7 antagonist. The finding of glycated sCD127 in the circulation of patients at onset of T1D suggested that physiological regulation of IL-7–mediated T-cell survival and expansion by sCD127 may be compromised in T1D. The findings indicate that genetic, immunologic, and metabolic factors contribute to a dysregulation of the IL-7/IL-7 receptor pathway in T1D and identify a novel hyperglycemia-mediated interference of immune regulatory networks.
The homeostatic cytokine interleukin-7 (IL-7) interacts with the IL-7 receptor (IL-7R) α chain (CD127) and the gc receptor (CD132) on the T-cell surface, forming a complex crucial to several signaling cascades involved in the survival and proliferation of T cells, including autoreactive T-cell clones. The IL-7/IL-7R pathway has been implicated in the expansion of autoimmunity and alloimmunity. In type 1 diabetes (T1D), increased concentrations of IL-7 are found postislet transplantation and are associated with homeostatic proliferation of T cells (1). Elevated IL-7 concentrations have also been reported in the synovial fluid of patients with rheumatoid arthritis (2), in the serum and cerebrospinal fluid of patients with multiple sclerosis (3), and in the serum of patients with autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (4). Moreover, IL-7 accelerates diabetes onset in the nonobese diabetic mouse (5), while blockade of the IL-7R can reverse diabetes in the same model (6,7).
As with many cytokine receptors, a soluble form of the IL-7R α chain (sCD127) has been identified (8). It is derived both from alternative splicing and by release of membrane-bound sCD127. The production of sCD127 is in part determined by regulation of transcription and mRNA processing, especially splicing, and this is affected by polymorphisms within the CD127 gene. Four common CD127 haplotypes have been described (9). Haplotype 2 is identified by a T allele in exon 6 (rs6897932). This single nucleotide polymorphism (SNP) is associated with reduced exon splicing and lower production of mRNA encoding soluble IL-7Ra in healthy control individuals. In contrast, transcripts that skip exon 6 (C allele of rs6897932) confer susceptibility to T1D and encode sCD127, potentially altering the production of sCD127 and IL-7R signaling in T cells (10).
Little is known with respect to how membrane versus sCD127 affects T-cell responses. Soluble forms of receptors can be antagonistic, such as tumor necrosis factor (TNF) receptor I and II (11) and the IL-6R subunit gp130 (12), or agonistic, such as the soluble IL-15 receptor α, where precomplexed IL-15/soluble IL-15 receptor α increases natural killer and CD8+ T-cell proliferation up to 50 times compared with IL-15 alone (13). In the current study, we asked whether sCD127 functions as an antagonistic or agonistic molecule in IL-7 signaling and whether it is dysregulated in patients with T1D. The findings show that sCD127 acts as an antagonist, is increased by immune activation, and affected by T1D-relevant CD127 polymorphisms. Its concentration is increased at onset of T1D, consistent with immune activation. Remarkably, however, sCD127 is glycated in patients, and the glycated form is ineffective as an antagonist of IL-7–mediated T-cell expansion. These findings show the complexity of cytokine signaling and associations within a disease process and unveil a novel effect of high blood glucose that affects immune function.
RESEARCH DESIGN AND METHODS
Serum samples were collected at disease onset from 390 patients with diabetes aged <20 years (213 males; mean age ± SD, 9.9 ± 4.4 years) participating in the DiMelli study that recruits the majority of new-onset patients with diabetes aged <20 years in Bavaria, Germany (14), and from 392 islet autoantibody–negative first-degree relatives (213 males; mean age ± SD, 9.9 ± 4.4 years) participating in the German TEENDIAB study (15). Among the 390 patients, 349 were positive for at least one islet autoantibody against GAD65, IA-2, or ZnT8. Genotyping for the rs6897932 SNP within the IL-7Ra gene was performed using the MassARRAY system with iPLEX chemistry (Sequenom, San Diego, CA). Serum concentration of sCD127 was also tested in samples from islet autoantibody–negative (n = 317, 169 female; median age 11.2 years; range, 7.3–16.0 years) and –positive first-degree relatives of patients with T1D (n = 317, 158 female; 138 with one, 179 with at least two islet autoantibodies; median age 11.2 years; range, 7.3–16.0 years).
sCD127 binding to IL-7.
ELISA was used to measure sCD127 binding to IL-7. High-binding polystyrene 96-well plates were coated overnight with sCD127 (100 ng/mL; R&D Systems) and/or sCD132 (1 ng/mL; R&D Systems) in PBS, washed, blocked with PBS-2% BSA, and incubated with increasing concentrations (0.125–2,048 pg/mL) of biotinylated IL-7 (R&D Systems) for 2 h. Detection of bound IL-7 was performed by incubation of plates with streptavidin-horseradish peroxidase (HRP; R&D Systems) followed by addition of tetramethylbenzidine substrate solution (R&D Systems) and colorimetric determination of sCD127-bound IL-7.
sCD127 effects on IL-7 signaling and T-cell proliferation.
Peripheral blood mononuclear cells (PBMCs) were incubated for 20 min with IL-7 (10 ng/mL) at 37°C in RPMI 1640 supplemented with 5% FCS in the absence or presence of 100 ng/mL of human CD127 (R&D Systems). PBMCs were fixed with 5% formaldehyde, stained with anti-CD4 Pacific blue (clone RPA-T4; BD Biosciences), permeabilized with methanol, and stained with anti-phosphorylated (p)Y694 signal transducer and activator of transcription 5 (STAT5) Alexa Fluor 647 (clone 47; BD Biosciences) and anti-pT308 Akt phycoerythrin (clone J1-223.371; BD Biosciences). For antigen-specific and polyclonal T-cell proliferation, PBMCs were stained with 5- (and 6-)carboxyfluorescein diacetate succinimidyl ester (CFSE) and incubated with antigens GAD65 and tetanus toxoid (TT) (16), or with anti-CD3/CD28–coated beads at a 1:100 bead/cell ratio, in the absence or presence of IL-7 (1 ng/mL) and varying concentrations of sCD127. After 5 days, PBMCs were stained with anti-CD4, and the percentage of proliferating cells (CFSEdim) was determined on gated CD4+ T cells by flow cytometry.
Regulation of membrane-bound and sCD127.
PBMCs were cultured for 24 h in RPMI 1640 supplemented with 5% FCS with the following stimuli: anti-CD3/CD28–coated beads (1:100 bead/cell ratio; Dynal), IL-2 (100 U/mL; BioSource International), IL-7 (10 ng/mL; R&D Systems), interferon-γ (IFN-γ; 10 ng/mL; BioSource International), IL-4 (10 ng/mL; PeproTech), IL-17 (10 ng/mL; R&D Systems), TNF-α (10 ng/mL; R&D Systems), IL-10 (10 ng/mL; R&D Systems), IL-12 (10 ng/mL; PeproTech), rapamycin (sirolimus, 10 ng/mL; Sigma-Aldrich), FK506 (tacrolimus, 10 ng/mL; Fujizawa Pharmaceutical), mycophenolate mofetil (MMF, 1 μg/mL; Roche), daclizumab (1 μg/mL; Roche), high glucose (33.3 mmol/L; Sigma-Aldrich), and insulin (100 U/mL; Sigma-Aldrich). PBMCs were subsequently stained with anti-CD4 Pacific blue (clone RPA-T4; BD Biosciences), anti-CD8 allophycocyanin-Cy7 (clone SK1; BD Biosciences), and anti-CD127 phycoerythrin (clone M21; BD Biosciences), and surface expression of CD127 in gated CD4+ and CD8+ T cells was analyzed by flow cytometry. Supernatant from PBMC cultures was taken to measure sCD127 concentration.
sCD127 was measured in serum and cell-culture supernatant using a sandwich ELISA. High-binding polystyrene 96-well plates were coated overnight in PBS with 50 mg/mL of monoclonal mouse anti-human CD127 antibody (clone 40131; R&D Systems), washed, and the plate blocked with PBS–2% BSA for 3 h. After washing, 50 μL of serum diluted 1:5 in PBS–0.1% BSA or undiluted culture supernatant was added to the wells for 2 h at room temperature. Plates were washed and incubated with 5 μg/mL biotinylated polyclonal goat anti-human CD127 (R&D Systems) for 1 h, washed, streptavidin-HRP added, washed, and developed with tetramethylbenzidine substrate solution and colorimetric measurement.
Bioinformatic analysis for prediction of sCD127 glycation sites and experimental glycation of sCD127.
The mRNA sequence of IL-7Ra/CD127 (Homo sapiens; accession number NP_002176.2) was subjected to computer-assisted analysis to predict the presence of potential glycation sites using an online algorithm (NetGlycate 1.0 Server, http://www.cbs.dtu.dk/services/NetGlycate/).
Recombinant human sCD127 (R&D Systems) was experimentally glycated by incubation for 7 days in the dark at 37°C in the presence of 100 mmol/L d-glucose (Sigma-Aldrich) or PBS as control. sCD127 glycation was evaluated by nontryptophan fluorescence as previously reported (17). Aliquots (200 ng) of native, glucose-treated, and PBS-treated sCD127 were electrophoresed by 12% SDS-PAGE under denaturating conditions and electroblotted onto 0.22-μm nitrocellulose membranes (Bio-Rad).
Detection of glycated sCD127 in serum samples.
Sera were incubated with anti-mouse immunoglobulin G (IgG) 1–coupled beads coated with mouse IgG1 anti-human sCD127 (clone 40131; R&D Systems) to immobilize serum sCD127 on beads, as we previously described (18). Beads were washed and incubated with a recombinant receptor for advanced glycation end products (RAGE)/human IgG protein chimera (R&D Systems) that recognizes and binds to glycated proteins. RAGE detection was performed by incubation with goat anti-human IgG fluorescein isothiocyanate (Sigma-Aldrich).
Expression of CD127, pSTAT5, and pAKT was assessed as median fluorescence intensity (MFI). To evaluate proliferative T-cell responses, the percentage of CD4+CFSEdimcells was calculated. IL-7 binding to sCD127 was assessed as arbitrary units of optical density. Data are presented as mean ± SD, and Student t test was used for comparisons. sCD127 concentration in serum samples was calculated in nanograms per milliliter according to a standard curve. Comparisons of sCD127 and glycated sCD127 concentrations between patients and control subjects and according to CD127 genotype were performed by linear regression analysis and included age and sex as covariates. A two-tailed P value <0.05 was considered significant.
sCD127 binds to IL-7.
IL-7 interacts with the IL-7Ra chain (CD127) and the γc (CD132) for signaling on the T-cell surface, and both CD127 and CD132 have a circulating soluble form. We therefore asked whether sCD127 is able to bind to IL-7 and whether CD132 is required for IL-7 binding. We used ELISA (Fig. 1A), in which plastic is coated with sCD127, sCD132, and in combination, and IL-7 binding is detected by streptavidin-biotin HRP. Typical in vivo concentrations of sCD127 (100 ng/mL) and sCD132 (1 ng/mL) were used. IL-7 binding to sCD127 was detectable from an IL-7 concentration of 1 pg/mL, concentrations that are found in the circulation of immunocompetent subjects (Fig. 1B). Binding increased to 32 pg/mL, which is an IL-7 concentration seen in immunosuppressed subjects, and was saturated at 256 pg/mL. IL-7 did not bind to sCD132, and the presence of sCD132 did not influence the binding of IL-7 to sCD127.
sCD127 inhibits IL-7 signaling and T-cell proliferation.
We next sought to determine whether sCD127 is agonistic or antagonistic with respect to the biological effect of IL-7 on T-cell signaling and proliferation (Fig. 2). Stimulation with IL-7 induced high amounts of pSTAT5 in CD4+ and CD8+ T cells (MFI mean ± SD, CD4+: 2,265 ± 506; CD8+: 3,095 ± 647) and pAKT (MFI mean ± SD, CD4+: 880 ± 149; CD8+: 1,086 ± 181), and this was markedly reduced in T cells when sCD127 was added to the culture (soluble STAT5 MFI mean ± SD, CD4+: 519 ± 174, P < 0.0001; CD8+: 1,112 ± 309, P < 0.0001; pAKT mean ± SD, CD4+: 585 ± 131, P < 0.0001; CD8+: 775 ± 108, P < 0.0001). The inclusion of IL-7 (1 ng/mL) to PBMCs cultured with GAD65, TT, or anti-CD3/CD28 beads increased CD4+ T-cell proliferation, and this was reversed by the addition of sCD127 for both healthy control subjects and patients with T1D (percent CD4+CFSEdim mean ± SD; medium + IL-7: 0.51 ± 0.25 vs. medium + IL-7 + sCD127: 0.34 ± 0.21, P = 0.45; GAD65 + IL-7: 6.83 ± 2.15 vs. GAD65 + IL-7 + sCD127: 4.15 ± 1.32, P < 0.0001; TT + IL-7: 8.02 ± 3.45 vs. TT + IL-7 + sCD127: 5.05 ± 2.23, P < 0.0001; anti-CD3/CD28 + IL-7: 62.69 ± 12.45 vs. anti-CD3/CD28 + IL-7 + sCD127: 46.18 ± 8.02, P < 0.0001; Fig. 2C and D). These data indicate that sCD127 is an antagonist soluble cytokine receptor.
Membrane-bound and sCD127 are regulated by immune stimuli.
The primary source(s) of sCD127 has not been determined but may be the result of alternative splicing, receptor shedding, and receptor cleavage. Antigenic stimulation of PBMCs induces downregulation of membrane-bound CD127 in the proliferating subset and a correspondent release of sCD127 in the culture supernatant proportional to the percentage of proliferating cells (Fig. 3A and B). We aimed to clarify whether active immune processes and immunomodulation can influence both the expression of CD127 on the surface of T cells and shedding/cleavage of sCD127. PBMCs from healthy control subjects (n = 12) and patients with T1D (n = 6) were cultured in the presence of immune stimuli: anti-CD3/CD28–coated beads (T-cell receptor [TCR]/costimulation triggering), IL-2 (T-cell proliferation), IL-7 (T-cell survival/homeostatic proliferation), IFN-γ (Th1), IL-4 (Th2), IL-17 (Th17), TNF-α (inflammation), IL-12 (dendritic cells), IL-10 (immune regulation); immunosuppressive drugs: rapamycin, FK506, anti-CD25 daclizumab; or metabolic stimuli: high glucose and insulin (Fig. 3C and D). CD127 was highly expressed in both CD4+ (MFI; control subjects, 5,550 ± 1,029; patients, 5,453 ± 1,039; P = 0.89) and CD8+ (MFI; control subjects, 5,741 ± 881; patients, 6,266 ± 867; P = 0.29) T cells, and only low concentrations of sCD127 were present in the culture supernatant in unstimulated conditions (ng/mL; control subjects, 1.7 ± 0.5; patients, 1.9 ± 0.7; P = 0.45). T-cell activation with anti-CD3/CD28 downregulated membrane-bound CD127 (CD4+ MFI: 3,995 ± 840, P = 0.0005; CD8+ MFI: 3,468 ± 1,179, P = 0.0003) and increased the concentration of sCD127 in the supernatant (ng/mL; 19.1 ± 8.9; P = 0.0008). A similar effect was observed with IL-2 (40 IU/mL) on surface expression of CD127 (CD4+ MFI: 3,847 ± 1,076, P = 0.0003; CD8+ MFI: 3,468 ± 1,179, P = 0.0004) and sCD127 in the culture supernatant (ng/mL: 10.6 ± 5.7; P = 0.014). The addition of IL-7 (1 ng/mL) strongly downregulated CD127 on the cell surface of T cells (CD4+ MFI: 845 ± 389, P < 0.0001; CD8+ MFI: 1,722 ± 736, P < 0.0001), but sCD127 concentration in the culture supernatant remained unchanged (ng/mL; 3.0 ± 1.87; P = 0.47), suggesting a mechanism of receptor internalization rather that shedding, as we have previously described (18). Other cytokines had minor or no effects on CD127 surface expression, but generally increased sCD127 concentration in the supernatant, the strongest effect seen with the noncommon γ chain using cytokine IL-10 (ng/mL; IFN-γ: 1.7 ± 0.6, P = 0.88; IL-4: 5.1 ± 2.2, P = 0.009; IL-17: 3.4 ± 2.1, P = 0.038; TNF-α: 3.1 ± 2, P = 0.041; IL-12: 4.3 ± 2.6, P = 0.07; and IL-10: 8.4 ± 4.3, P = 0.002).
No effect on surface CD127 and sCD127 was observed by the addition of the immunosuppressive drugs rapamycin (10 ng/mL), FK506 (5 ng/mL), mycophenolate mofetil (MMF) (1 mg/mL) and daclizumab (1 μg/mL) or by the addition of high glucose (33.3 mmol/L) or insulin (40 IU/mL).
Circulating sCD127 is increased in patients with autoimmune diabetes.
Since sCD127 was observed in culture supernatants after TCR stimulation, we asked whether sCD127 concentration was increased in autoimmunity (Fig. 4). The mean concentration of sCD127 was significantly higher in patients with diabetes diagnosed before age 20 years (85.1 ± 48.6 ng/mL; n = 390) as compared with age- and sex-matched antibody-negative relatives (70 ± 46.6 ng/mL; P = 0.00001; Fig. 4A). Moreover, patients who were islet autoantibody–positive had significantly higher sCD127 (90.0 ± 49.7) as compared with patients who were islet autoantibody–negative (69.7 ± 40.0; P = 0.031; Fig. 4B). No difference in sCD127 concentration was observed in patients with one autoantibody as compared with patients with multiple islet autoantibodies. Circulating sCD127 concentration was not associated with sex or age, and among the patients, there was no association between sCD127 concentrations and HLA genotype, C-peptide concentration, and HbA1c (data not shown). The release of sCD127 as a spliced variant is determined by polymorphisms of the IL-7R gene (10), and accordingly, sCD127 concentrations in patients with diabetes were associated with the genotype of the rs6897932 SNP (Fig. 4C). Patients homozygous for the T1D risk allele C had higher circulating sCD127 (CC genotype, 99.6 ± 55.8; n = 149) as compared with heterozygous patients (CT, 71.6 ± 32.5; n = 98; P = 10−6) and patients without the risk allele (TT, 45.7 ± 22.4; n = 13; P = 10−7). The data suggest that autoimmunity and IL-7R genotype are likely to influence the serum concentration of sCD127.
We next examined the prediabetes period. Serum concentration of sCD127 was similar in islet autoantibody–positive first-degree relatives of patients with T1D (75.2 ± 47.9 ng/mL; n = 318) and age-matched islet autoantibody–negative relatives (70.5 ± 51.5 ng/mL; n = 318; P = 0.15; Fig. 4D).
Glycation of sCD127.
Unlike other autoimmune disorders, T1D is characterized by high blood glucose. This is relevant to a number of biochemical processes since it leads to nonenzymatic glycation of circulating proteins, and this can influence protein structure and function. We therefore asked whether sCD127 can undergo glycation and whether this could affect its binding capacity to IL-7(19). The sequence of the IL-7Ra/CD127 (Homo sapiens; accession number NP_002176.2) was analyzed for potential glycation sites. Lysine residues in the extracellular domain of CD127 potentially involved in sCD127 protein glycation included K81, K97, K119, K157, K161, K187, K204, and K214 (Fig. 5A and B). Recombinant sCD127 was therefore experimentally glycated by incubation with d-glucose and the presence of glycation products assessed by spectrofluorimetric analysis to measure the nontryptophan fluorescence as a known glycation marker (Fig. 5C). A significant increase of glycation fluorescence was recorded for sCD127 incubated with glucose (GsCD127) as compared with sCD127 incubated with PBS, indicating that sCD127 was sugar-modified under the experimental conditions. Western blot analysis suggested that the experimental glycation did not cause significant protein alterations (Fig. 5D). As an iso-osmotic control, sCD127 was also incubated with the nonreducing sugar mannitol (20), which did not produce glycation products nor structural or functional modifications (not shown).
sCD127 binding to IL-7 is reduced by experimental glycation.
To determine whether sCD127 glycation influences its binding capacity to IL-7, experimentally glycated (from 1 to 100 mmol/L of d-glucose) sCD127 (GsCD127) was used in the IL-7–binding ELISA. Experimental glycation of sCD127 progressively reduced its binding capacity to IL-7 with a significant decrease observed at a d-glucose concentration of 20 mmol/L and further decrease at 50 and 100 mmol/L (Fig. 6A). In contrast to sCD127, addition of GsCD127 poorly inhibited IL-7–induced STAT5 and AKT phosphorylation, suggesting an impaired function of the glycated receptor (pSTAT5 MFI CD4+; IL-7 + sCD127 = 519 ± 174 vs. IL-7 + GsCD127 = 1,790 ± 536, P = 0.003; CD8+; IL-7 + sCD127 = 1,112 ± 309 vs. IL-7 + GsCD127 = 1,864 ± 559; P = 0.003; pAKT MFI CD4+, IL-7 + sCD127 = 585 ± 131 vs. IL-7 + GsCD127 = 830 ± 163, P < 0.0001; CD8+; IL-7 + sCD127 = 775 ± 108 vs. IL-7 + GsCD127 = 957 ± 139, P = 0.03; Fig. 6B). Finally, addition of sCD127 significantly reduced the proliferative effect of IL-7, whereas the addition of GsCD127 only minimally affected IL-7–assisted proliferation (CD4+CFSEdim: GAD65 + IL-7 + sCD127 = 3.43 ± 0.67 vs. GAD65 + IL-7 + GsCD127 = 4.99 ± 1.5; n = 9, P = 0.037; TT responsiveness, TT + IL-7 + sCD127 = 5.69 ± 2.65 vs. TT + IL-7 + GsCD127 = 8.01 ± 2.85; n = 9, P < 0.0001; Fig. 6C). These data suggest that glycation of sCD127 impairs its function of IL-7 binding.
Glycated sCD127 and membrane-bound CD127 in patients with T1D.
We subsequently asked whether glycated sCD127 is present in the circulation of patients with T1D. Glycated proteins are engaged by RAGE, a transmembrane receptor of the Ig superfamily belonging to the pattern recognition receptors. We developed a flow cytometric assay in which the glycated form of the sCD127 protein is detected by its capacity to bind to RAGE (Fig. 6D) and searched for GsCD127 in a subset of the patients with diabetes (n = 26; age, 9.7 ± 4.5 years; all islet autoantibody–positive) and antibody-negative relatives (n = 35; age, 9.1 ± 4.5 years). Serum amounts of GsCD127 were increased in patients with diabetes (MFI, 219.3 ± 62.3; n = 26) as compared with those in islet autoantibody–negative relative subjects (MFI, 170.5 ± 49.2; P = 0.0011; Fig. 6E). GsCD127 concentration was not associated with age (P = 0.92) or sCD127 concentration (P = 0.29) and remained significantly associated with T1D after adjustment for these covariates (P = 0.002). Within the patients, GsCD127 concentration was higher in those with the rs6897932 SNP CC genotype (P = 0.04), but there was no relationship with HbA1c (P = 0.5) or autoantibody status (P = 0.6). In order to evaluate glycation of the membrane-bound CD127, we induced sCD127 release by stimulation of T cells from healthy subjects and patients with anti-CD3/CD28 and measured RAGE binding (Fig. 6G). Glycation was increased in sCD127 released from T cells from patients with T1D as compared with T cells from healthy subjects after 24 h (GsCD127 MFI: healthy subjects, 97.5 ± 17.4 vs. T1D, 129.3 ± 24.3; P = 0.027). sCD127 released at later time points in culture after 3 and 7 days displayed similar glycation levels.
Alterations of the soluble form of IL-7Ra has been associated with the pathogenesis of autoimmune diseases (3,21). In this study, we report that sCD127 is antagonistic to IL-7 signaling and that its concentration is affected by immune activation and genetic polymorphisms in the IL-7Ra gene. Of particular interest, the concentration of sCD127 was increased in patients with T1D. However, the potential beneficial effects of downregulating IL-7–mediated T-cell pathogenesis by sCD127 may be diminished in patients by the presence of a glycosylated low-activity form of the soluble receptor.
Relevant to T1D was the increased concentration of sCD127 in the circulation at disease onset. The findings were generated in a relatively large cohort of patients with young age (<20 years) and an age-matched cohort of islet autoantibody–negative first-degree relatives. Moreover, within patients, sCD127 concentrations were only increased in the islet autoantibody–positive subset consistent with a relationship to autoimmune diabetes. There was a clear relationship between sCD127 concentration and CD127 genotype of the patients similar to what was reported for control subjects and patients with multiple sclerosis (22). However, it is unlikely that the increased concentration observed in patients was solely due to this relationship since the association of CD127 genotype and T1D is modest. Although genotyping data were unavailable in the control subjects to confirm this, the minor allele frequency within the patient cohort was 24%, which is similar to that in 2,298 German islet autoantibody–negative first-degree relatives (25%, data not shown) and what is reported in Caucasian control subjects (21).
We predict that the increased sCD127 concentration in patients represents T-cell activation and inflammation at diabetes onset, since we did not observe increases in islet autoantibody–positive relatives. T-cell activation and expansion via TCR engagement, costimulation, and IL-2 downregulated membrane expression of CD127 induced substantial release of sCD127 in the culture medium. These findings are consistent with shedding of membrane-bound CD127 in these settings. In comparison, the homeostatic cytokine IL-7 markedly downregulated the expression of CD127 on the cell surface with only modest release of sCD127, consistent with our previous findings that IL-7 leads to internalization of CD127 (18). It is perhaps noteworthy that among the non–γ-chain cytokines, the suppressive cytokine IL-10 was the most powerful inducer of sCD127, whereas the Th1 cytokine IFN-γ had no influence. Altogether, the in vitro data indicate that T-cell activation and inflammation are likely to augment sCD127 concentrations.
The IL-7/IL-7R axis plays an important role in the survival and expansion of T cells, including autoreactive clones. IL-7 accelerates diabetes onset in the nonobese diabetic mouse model (5), and modulation of IL-7 concentration or biological function in NOD mice by blockade of the IL-7R with monoclonal antibodies was shown to reverse diabetes (6,7). IL-7 is constitutively produced by stromal cells, and its circulating concentration is apparently determined by consumption by IL-7R–positive T cells (23). In line with a previous report in the context of HIV infection (24), we show that circulating sCD127 can bind to IL-7 and strongly inhibit its bioactivity. Thus, sCD127 represents an endogenous regulator of IL-7 bioactivity and a candidate molecule for therapeutically targeting IL-7. The affinity of CD127 for IL-7 (Kd = 10−8 M) is considerably lower than the CD127/CD132 complex (Kd = 10−11 M) on the T-cell surface (25). However, we found serum concentrations of sCD127 (70–80 ng/mL) >10,000-fold higher than its ligand IL-7 (5 pg/mL) in steady-state conditions (26) and >1,000-fold higher when IL-7 is elevated in lymphopenic conditions (1). We therefore hypothesize that endogenous sCD127 may bind a significant fraction of circulating IL-7. In our hands, sCD132, which was reported to exist in a soluble circulating form at high concentration (1 ng/mL) (24), did not bind to IL-7 and did not contribute to sCD127/IL-7 interaction.
The increased sCD127 concentration observed at onset of T1D should assist the regulation of immune activation and T-cell expansion. We remained intrigued, however, that T1D onset is notionally a highly inflammatory state in which immune regulation appears to be ineffective. Hyperglycemia is the hallmark of T1D and is itself a mediator of inflammation (27). One mechanism by which it mediates inflammation is the production of advanced glycation end products that have proinflammatory properties through engagement of pattern recognition receptors such as RAGE (28). Nonenzymatic glycation occurs when proteins are exposed to high glucose. In patients with diabetes, glycation has been reported for proteins with long half-lives such as hemoglobin, but also for proteins with short half-lives such as insulin (29). Bioinformatic analysis showed that sCD127 is rich in lysine residues as potential sites of glycation. Consistent with this, we were able to show that sCD127 can be glycated in vitro and found that serum from patients at onset of T1D contained increased levels of a glycated form of sCD127. Importantly, we found that glycation of sCD127 almost completely abolished its ability to bind IL-7 and prevent IL-7 biological activity. These novel findings indicate that hyperglycemia may have previously undiscovered effects on physiological immune-regulatory mechanisms and that potential glycation of proteins in patients with diabetes should be considered in pathogenesis and therapy.
We and others have hypothesized that altered T-cell homeostasis can contribute to the expansion of autoreactive T cells in different autoimmune diseases. We now show that IL-7 biological activity can be regulated by sCD127, but this regulation may be compromised in patients with hyperglycemia. In addition to demonstrating the complexity of cytokine signaling, the findings unveil a novel interplay between metabolic and immunological processes in T1D.
This work was supported with funds from the Juvenile Diabetes Research Foundation (grants 17-2012-16 and 17-2011-601), the Center for Regenerative Therapies Dresden-DFG Research Center and Cluster of Excellence (FZ 111), and the German Ministry for Education and Research to the German Centre for Diabetes Research and the Competence Network of Diabetes (grants 01 GI 0805 and 01 GI 1188). P.M. is supported by an advanced postdoctoral fellowship from the Juvenile Diabetes Research Foundation (10-2009-284).
No potential conflicts of interest relevant to this article were reported.
P.M. conceived the study, designed and performed experiments, and wrote the manuscript. C.B. designed and performed molecular analysis of glycation. M.K. executed the clinical studies providing samples. A.G.Z. conceived and oversaw the clinical studies on which the study was based. E.B. provided input to experimental plan and design of the study and reviewed and wrote the manuscript. P.M. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
The authors thank Christine Milz, Frank Hofmann, and Manfred Keller from the Kassenärztliche Vereinigung Bayern and Ramona Puff, Katharina Warncke, and Peter Achenbach from the Institute of Diabetes Research Munich for contributions to patient recruitment, immune monitoring, and data management.