Soluble IL7Rα potentiates IL-7 bioactivity and promotes autoimmunity

sIL7Rα Binds IL-7 but Not TSLP.

To study the biology of sIL7Rα, we produced the protein encoded by the Δ6IL7Rα mRNA, as well as a protein comprised solely of the ECD of IL7Rα (IL7Rα-EC), which does not contain the unique 26-aa tail found in sIL7Rα and would be predicted to be identical to shed sIL7Rα. Using surface plasmon resonance, we measured binding affinities of both proteins for IL-7 and TSLP, ligands for cell-associated receptors containing IL7Rα (Fig. 1). The protein encoded by Δ6IL7Rα mRNA displays moderate affinity for rhIL-7 (K_d, 6.3 nM), which was 16-fold stronger than the binding affinity of IL7Rα-EC to rhIL-7 (K_d, 98 nM), primarily because of a faster k₁ rate constant (2.6 × 10 ^{vs. 6.2 × 10 M/s;} Fig. 1 A vs. C). At concentrations up to 0.5 μM, we saw no binding between Δ6IL7Rα and TSLP (Fig. 1B), whereas moderate affinity (K_d, 53 nM) was measured between TSLP to TSLPR-EC (Fig. 1D). Based upon these studies, we conclude that immunobiologic effects of Δ6IL7Rα occur as a result of binding to IL-7 rather than TSLP. We further conclude that binding affinity of sIL7Rα to IL-7 is ∼1 log weaker than the picomolar binding affinity previously measured between IL-7 and the cell-associated IL7R complex (33–35).

Fig. 1.

sIL7Rα binds IL-7 but not TSLP. The binding kinetics for all of the cytokine/receptor interactions fit best to a two-step binding reaction model using two on-rate (k₁ and k₂) and off-rate (k_-1 and k_-2) constants. (A–D) Surface plasmon resonance sensorgrams are shown for each designated protein pair. Black lines represent raw data and red lines represent the global fitting analysis of the sensorgrams to a two-step binding reaction model using ClampXP. (E) Summary of binding affinities measured in A–D. Experiments were performed in triplicate with errors on the order of 5–10% for the rate constants.

sIL7Rα Enhances IL-7–Induced Survival of 2E8 Cells by Diminishing Consumption.

To assess the effects of sIL7Rα on IL-7 bioactivity, we measured IL-7–mediated survival of 2E8, an IL-7–dependent murine pro–B-cell line, in the presence of rhIL7 (2,000 pg/mL) and varying concentrations of sIL7Rα. No effects were observed at early time points, but we observed a trimodal effect on day 12 of culture (Fig. 2A). The fact that effects were not apparent until a late time point led us to postulate that IL-7 consumption could be significantly impacted by the presence of sIL7Rα and that such an effect would become increasingly apparent with longer culture periods. At low molar ratios (≤128), cell survival was substantially diminished from baseline and from earlier time points, and we postulated that this reflected low IL-7 levels attributable to increasing IL-7 consumption over time. At middle molar ratios (256 and 512), we observed enhanced cell survival, and we postulated that this resulted from diminished IL-7 consumption leading to increased IL-7 availability and, thus, increased cell survival. Finally, at high molar ratios (≥1,024), we postulated that the large excess of sIL7Rα diminished free IL-7, thus preventing binding of IL-7 to its cell-associated receptor and leading to cell death.

Fig. 2.

sIL7RΑ potentiates IL-7–mediated survival and diminishes consumption of IL-7. (A) Survival of the IL-7–dependent cell line 2E8 was measured in the presence of rhIL-7 (2,000 pg/mL) plus varying concentrations of sIL7Rα. No effects were seen at any time point using molar ratios of 128:1 and below. However, on day 12, increased survival was observed at sIL7Rα:IL-7 molar ratios of 256–512:1, which was diminished at higher molar ratios. (B) Using lower concentrations of rhIL-7 [250 pg/mL (Left) and 500 pg/mL (Right)], survival of 2E8 from days 0–7 is shown with no cytokine, rhIL-7 alone, or sIL7Rα:rhIL7 molar ratio of 500:1. sIL7RΑ significantly increased survival on days 5 and 7 (*P < 0.05; **P < 0.01; ***P < 0.001). Error bars represent SEM of triplicate experiments. (C) IL-7 levels measured in the cultures described in B. Significantly increased IL-7 levels were measured in the presence of sIL7Rα on days 1, 3, and 5 of culture (*P < 0.05; **P < 0.01; ***P < 0.001). Error bars represent SEM of triplicate experiments. This experiment was performed three times with similar results.

To directly test the hypothesis that sIL7Rα modulates IL-7 consumption, we used the same model system but lowered the IL-7 concentration to 250–500 pg/mL and then assessed sIL7Rα-mediated effects on 2E8 survival and on IL-7 levels with a focus on molar ratios that mediated potentiation. Under these conditions, sIL7Rα plus IL-7 significantly increased 2E8 survival beginning on day 5 (Fig. 2B), and significantly increased IL-7 levels were present in cultures containing sIL7Rα on days 1 through 5 (Fig. 2C). We observed no effect of sIL7Rα on IL-7 levels maintained at 37 °C in a serum-containing, but cell-free, system, ruling out the possibility that sIL7Rα diminished spontaneous or proteolytic degradation of IL-7. Furthermore, at the molar ratios used in this system, sIL7Rα had no effect on the accuracy of the IL-7 ELISA in measuring IL-7 levels. Therefore, sIL7Rα diminishes IL-7 consumption by IL7R expressing target cells, leading to enhanced IL-7 bioactivity in settings where IL-7 is limited.

sIL7Rα Modulates IL-7 Signaling in Human T Cells.

The cell line 2E8 is an IL-7–dependent murine cell line, raising the possibility that these results could reflect a competitive disadvantage of cell-associated mouse IL7R complex for rhIL-7 binding. We, therefore, explored the effects of sIL7Rα on IL-7 signaling using human T cells. Using STAT5 phosphorylation as a readout of early IL7R signaling, 50:1 molar ratio of sIL7Rα:rhIL-7 had no effect, incomplete blockade was noted with 500:1 ratios, and at 5,000:1, sIL7Rα completely inhibited IL-7 signaling in both CD4 ^{and CD8 human T cells (}Fig. 3A). Consistent with decreased early signaling, as evidenced by diminished STAT-5 phosphorylation, the 500:1 ratio of sIL7Rα:rhIL-7 also diminished IL-7–induced CD127 down-regulation and CD95 up-regulation on day 1 (Fig. 3B). However, at later time points, cultures containing sIL7Rα predominantly showed augmented rhIL-7–induced effects, as evidenced by more profound and persistent CD127 down-regulation and CXCR4 up-regulation. Furthermore, IL-7 levels were higher when sIL7Rα was present (Fig. 3C), similar to the results observed in the 2E8 model system. These results confirm that sIL7Rα diminishes IL-7 consumption by human T cells, resulting in more potent and prolonged biologic effects.

Fig. 3.

sIL7Rα-mediated modulation of IL-7 signaling on human T cells leads to diminished IL-7 consumption and diminished SOCS1 and CD95 Induction. Human PBMCs were coincubated with rhIL-7 with or without sIL7Rα and then analyzed at the times shown. Molar ratios were as noted in A, whereas 500:1 sIL7Rα:rhIL7 molar ratio was used in B–D. (A) IL-7–induced STAT-5 phosphorylation was measured in gated CD4 ^{and CD8 T cells after 15 min. Cells were serum-starved before IL-7 was added. No effect was seen at 50:1 molar ratio, incomplete inhibition was seen at 500:1 ratio (similar effects at 1 and 10 ng/mL rhIL-7), and complete inhibition was seen at 5,000:1 ratio. Controls using human serum albumin and rat IgG2a at similar concentrations showed no significant impact on STAT-5 signaling. (}B) Modulation of IL-7–induced changes in CD127, CXCR4, and CD95 expression by sIL7Rα. Representative flow-cytometric histograms of CD127 expression at day 5 on CD4 ^{and CD8 T cells are shown on the right. (}C) IL-7 consumption by human PBMCs is diminished in the presence of sIL7Rα, as measured by ELISA. (D) SOCS1 levels are reduced in the presence of sIL7Rα after 24 h of incubation with IL-7. Relative quantity (RQ) of SOCS1/GAPDH mRNA expression compared with untreated cells was determined. Error bars represent 95% CI of triplicate experiments. A total of three independent experiments were performed using PBMCs from different donors with comparable results. Statistical significance shown (*P < 0.05) reflects comparisons between rhIL-7 alone and rhIL-7 plus sIL7Rα using a two-tailed t test.

Interestingly, we observed two exceptions to the pattern of sIL7Rα-mediated potentiation of IL-7 bioactivity. Both CD95 up-regulation and suppressor of cytokine signaling 1 (SOCS1) up-regulation were consistently diminished in the presence of sIL7Rα compared with IL-7 alone, even at late time points (Fig. 3 B and D). Thus, sIL7Rα also modulates the character of IL-7 signaling by diminishing CD95 and SOCS1, both negative regulators of immune responses, mimicking reduced IL-7 concentration at early time points (Fig. S2). Reduced CD95 expression in the presence of sIL7Rα showed functional significance, as evidenced by reduced anti-Fas–mediated killing compared with IL-7 alone (Fig. S3). Screening for other cytokines in culture supernatants did not reveal any statistically significant differences in IL-1b, IL-2, IL-6, IL-8, IL-10, IL-12p70, GM-CSF, IFNγ, or TNFα production between the culture conditions, suggesting these effects are directly linked to alterations in the IL-7 signal itself. Together, the ability for sIL7Rα to diminish IL-7 consumption and diminish the induction of negative regulators of immune responses led to the prediction that sIL7Rα would potentiate biologic effects of IL-7 in vivo.

sIL7Rα Potentiates IL-7 Bioactivity in Vivo.

We next explored the effect of sIL7Rα on IL-7–induced homeostatic expansion in vivo. IL7 ^{mice were chosen for these studies because they provide a lymphopenic milieu within which effects on homeostatic expansion can be readily measured, and they eliminated the possibility of confounded results attributable to signaling by murine IL7. In concordance with our in vitro findings, higher levels of IL-7 were measured 24 h following administration of rhIL-7 with sIL7Rα compared with that measured following administration of rhIL-7 alone (}P < 0.05 at 24 h; Fig. 4A), and coadministration of sIL7Rα with rhIL-7 enhanced rhIL-7–driven homeostatic peripheral expansion of congenic, adoptively transferred lymph node cells, with the most potent effects observed for IL-7–induced CD4 ^{T-cell expansion (}Fig. 4B). Thus, similar to the results obtained in vitro, sIL7Rα diminishes clearance of IL-7, resulting in more prolonged exposure and more potent biologic effects.

Fig. 4.

sIL7Rα diminishes IL-7 clearance in vivo and increases IL-7–mediated homeostatic peripheral expansion. (A) On day 0, IL-7 ^{mice received one dose of rhIL-7 (5 μg) with or without sIL7Rα (100 μg) coinjected in the same syringe. A 10:1 molar ratio was chosen because of limited availability of recombinant sIL7Rα. Plasma levels were measured at 24 and 96 h. Mice injected with rhIL-7 plus sIL7Rα have higher plasma IL-7 levels after 24 h than those receiving rhIL-7 alone. (}B) On day 0, mice received IL-7 with or without sIL7Rα, as in A above, plus 2 × 10 ^{congenic lymph node cells. On day 8, mice injected with rhIL-7 plus sIL7Rα showed higher numbers of adoptively transferred CD4 and CD8 splenocytes compared with mice receiving rhIL-7 alone. Asterisks denote significant differences. These experiments were repeated once for two independent experiments with similar results (}n = 5 per group). (C) C57BL/6 mice (n = 10/group) were immunized with MOG and then scored for EAE symptoms. PBS, rhIL-7, sIL7Rα, or rhIL-7 plus sIL7Rα was administered i.p. on days 9 and 15 postimmunization (arrows). Mice injected with rhIL-7 plus sIL7Rα showed increased EAE scores (Left), more rapid progression to EAE score 3 (Center), and higher overall disability as measured by area under the curve of EAE score over time per individual mouse (Right). *P < 0.05 as determined by Mann–Whitney test. No significant differences were seen when injecting IL-7 with a control protein (αhIL6Rα) compared with IL-7 alone, using the models shown in A, B, or C. This experiment was carried out three times with similar results.

We next compared the effects of sIL7Rα on IL-7–induced potentiation of experimental autoimmune encephalitis because previous work had demonstrated that coadministration of rhIL7 exacerbates the severity of this disease (12). Myelin oligodendrocyte glycoprotein (MOG)-injected C57BL/6 mice receiving rhIL-7 (5 μg) plus sIL7Rα (100μg) showed significantly worsened experimental autoimmune encephalomyelitis (EAE) symptoms than animals receiving rhIL-7 alone (Fig. 4C), as evidenced by an increased mean EAE score for the group as a whole (Fig. 4C, Left), diminished time to progression to an EAE score of 3 (Fig. 4C, Center; log-rank P = 0.03), and overall disability of individual mice (Fig. 4C, Right). Thus, sIL7Rα potentiates IL-7–induced exacerbation of autoimmune disease. Interestingly, this effect was observed in nonlymphopenic mice with normal levels of murine IL-7, thus demonstrating that the in vitro-potentiating effects of sIL7Rα on IL7 bioactivity occurred even in the presence of normal levels of murine IL-7, and, thus, the observation is generalizable across several in vitro and in vivo model systems.

IL7R Genotype Modulates sIL7Rα and IL-7 Levels.

To validate previous reports of associations between IL7R genotype and IL7Rα mRNA splicing (23), we measured Δ6IL7Rα:full-length isoform ratios in resting peripheral blood mononuclear cells (PBMCs) from individuals with IL7R*CC (autoimmune-predisposing) vs. IL7R*TT (autoimmune protective) genotypes. The Δ6IL7Rα:full-length ratio was significantly higher in IL7R*CC vs. IL7R*TT individuals (mean ± SEM, 0.028 ± 0.001 vs. 0.011 ± 0.001; Fig. 5A, Left), attributable to an increase in the Δ6IL7Rα isoform (mean ± SEM, 0.78 ± 0.13 vs. 0.33 ± 0.08; Fig. 5A, Right). We observed no significant difference in the amount of the more plentiful full-length isoform between genotypes. Both MS patients and controls with other neurologic diseases (OND controls) showed a similar pattern, consistent with previous studies in healthy controls (23).

Fig. 5.

Individuals with the autoimmunity-associated IL7R*CC genotype have increased Δ6IL7Rα mRNA and increased levels of circulating sIL7Rα. (A) Quantitative, isoform-specific real-time PCR measured Δ6IL7Rα mRNA and full-length IL7Rα mRNA in PBMCs obtained from IL7R*CC and IL7R*TT MS patients and controls with ONDs. Each shape represents one patient’s sample. IL7R*CC individuals had increased Δ6IL7Rα/full-length mRNA ratios (Left) and increased Δ6IL7Rα mRNA levels (Right) compared with IL7R*TT MS patients and OND controls. (B) IL7R genotype modulates plasma protein levels of sIL7Rα in healthy controls (n = 41 of each genotype) and MS patients (n = 35–41 of each genotype) as measured using ELISA. The patient and control cohorts analyzed in B are distinct from those analyzed in A. (C) sIL7Rα was not detected in the CSF of MS patients, regardless of genotype. The dotted line indicates detection limit of the assay. P values were calculated using an unpaired two-tailed t test.

To determine whether genotype influenced sIL7Rα protein levels, we measured circulating IL7Rα levels in IL7R*CC vs. IL7R*CT vs. IL7R*TT individuals (21). Circulating IL7Rα levels in healthy controls (HCs) and MS patients show a wide range [mean (5–95 percentile): HCs, 39.8 ng/mL (7.9–97.4); MS, 42 ng/mL (8.3–111.8)], with an ∼threefold increase in mean sIL7Rα levels in IL7R*CC vs. IL7R*TT in HCs [mean (5–95 percentile): IL7R*CC, 70.3ng/mL (28.6–146.6); vs. IL7R*TT, 19.7 ng/mL (3.7–42.2)] (Fig. 5B). We observed a similar ∼threefold increase in mean sIL7Rα in IL7R*CC vs. IL7R*TT MS patients (55.8 vs. 19.8 ng/mL), similar to the magnitude of the mRNA increase for the Δ6IL7Rα isoform and the Δ6IL7Rα/IL7Rα full-length ratio (Fig. 5A). Heterozygotes (IL7R*CT) showed intermediate levels of circulating sIL7Rα, implicating an allele-dose effect. sIL7Rα was not measurable in the cerebrospinal fluid (CSF), implying that it does not cross the blood–brain barrier (Fig. 5C). This confirms the observations of Hoe et al. (23), who demonstrated previously that IL7R*CC vs. IL7R*TT individuals experience increased IL7R mRNA splicing, which leads to increased circulating sIL7Rα. Interestingly, both quantitative PCR and the measured protein levels demonstrate that the autoimmunity predisposing IL7R*CC genotype induced an approximately threefold increase in Δ6IL7Rα mRNA levels and in circulating sIL7Rα protein levels.

Given that we clearly observed effects of sIL7Rα on IL-7 consumption in vitro, that previous work has concluded that receptor mediated consumption is a primary mechanism by which IL-7 levels are regulated (36), and that under normal circumstances, IL-7 is considered to be a limited resource (3), we postulated that genotype-induced modulation of sIL7Rα levels could significantly impact IL-7 availability in humans in vivo. To test this, we compared IL-7 levels in plasma and CSF obtained from a cohort of patients with MS (n = 42) classified according to genotype. As shown in Fig. 6A (Left), we observed significant increases in plasma IL-7 levels among IL7R*CC MS patients compared with IL7R*TT patients, consistent with a model wherein increased sIL7Rα leads to diminished consumption and a secondary increase in IL-7 availability. Importantly, we saw no significant effect on CSF IL-7 levels (Fig. 6A, Right), which is consistent with this model because sIL7Rα is not present in the CNS (Fig. 6C). To validate these findings, we screened the larger cohort of MS patients (n = 123) and healthy controls (n = 119) enriched for the IL7R*TT genotype previously used to determine sIL7Rα levels (Fig. 5B). This cohort confirmed a significant effect of IL7R genotype on circulating IL-7 levels in patients with MS, such that IL7R*CC MS patients have, on average, double the mean IL-7 levels (∼8 pg/mL) compared with MS patients with IL7R*CT or IL7R*TT genotypes or healthy controls (∼4 pg/mL; Fig. 6B). Interestingly, we observed no effect of IL7R genotype on IL-7 levels in healthy controls. Serum from both MS cohorts was obtained at the time of disease onset but before initiation of therapy, and, therefore, these findings cannot be attributed to immunosuppressive therapies. What remains to be determined is whether the onset of the autoimmune disease triggers increased IL-7 levels among IL7R*CC in MS patients or whether elevations of IL-7 levels in IL7R*CC MS patients predate the onset of the disease and, thus, potentially predispose to the development of autoimmunity.

Fig. 6.

IL7R*CC MS patients have increased plasma IL-7 levels. (A) IL-7 levels were measured in plasma and CSF from patients with MS according to genotype. IL7R*CC patients had significantly higher mean plasma IL-7 levels than IL7R*TT patients, whereas no difference was observed between IL7R*CC and IL7R*CT heterozygotes. No correlation between CSF IL-7 levels and IL7R genotype was observed. The dotted line denotes limit of detection. (B) Plasma IL-7 levels were measured in plasma from a second independent cohort of patients with MS, as well as a separate cohort of healthy controls. IL7R*CC MS patients have significantly higher mean IL-7 levels than IL7R*TT MS patients, whereas no difference was observed according to genotype in healthy controls. Each shape represents a plasma or CSF sample analyzed from one patient or healthy control.

Real-Time PCR.

Plasmid cDNA standards for each IL7Rα isoform β-actin were synthesized (Mr. Gene). TaqMan gene expression assays specific for the full-length (Hs00904814_m1) and Δ6 (Hs00902337_m1) isoforms were purchased from Applied Biosystems. mRNA was extracted from PBMCs using the RNeasy kit (Qiagen) and transformed into cDNA using the Super script III First Strand Synthesis System (Invitrogen). Genotyping of rs6897932 was carried out using allelic discrimination (Assay ID: C_2025977_10; Applied Biosystems) on an ABI TaqMan 7000 Real-Time PCR machine. SOCS1-relative expression levels were determined by semiquantitative real-time PCR using GAPDH as a housekeeping gene (Assay ID: SOCS1, Hs00705164_s1; GAPDH, Hs02758991_g1; Applied Biosystems).

Human Samples.

All patient and control samples were acquired after written consent was obtained and the use of samples was approved by the regional ethical review board at Karolinska Institutet. Human plasma samples were obtained from two different Swedish cohorts: the first is from the Epidemiological Investigation of Multiple Sclerosis (52) (Figs. 5B and ​and6B),6B), in which MS patients’ plasma was collected at diagnosis (according to McDonald criteria), in patients between 16–70 y of age from different Swedish clinics. Healthy controls were selected from an age- and geographically matched control cohort based on IL7R genotype. The second was from the Stockholm Prospective Assessment Study of MS (Figs. 5 A and C and ​and6A),6A), in which plasma and CSF and PBMC samples were collected from MS patients at diagnosis, but before any MS-directed therapy at the Karolinska University Hospital, Stockholm, Sweden. Controls with ONDs were visitors to the same hospital but confirmed by clinic neurologists as not having MS. PBMCs from healthy controls were obtained from healthy blood donors via the National Institutes of Health (NIH) Department of Transfusion Medicine, who underwent informed consent according to NIH Clinical Center Institutional Review Board approved protocols. Plasma, cDNA, and CSF samples were stored at −80 °C and shipped on dry ice before analyses.

ELISAs.

Human IL-7 levels in human plasma, mouse plasma, and culture supernatants were measured by high-sensitivity IL-7 ELISA kit according to the manufacturer’s instructions (Quantikine; R&amp;D Systems). Plasma IL7Rα levels were measured by a custom-made validated ELISA as described previously (21).

Protein Production-Affinity Measurements.

Human IL-7 was expressed from Escherichia coli and purified as described previously (53). The human IL7R ECD was expressed from Drosophila Schneider 2 (S2) insect cells and purified as described previously (53). The sIL7Rα gene (residues 23–254; UniProt accession no. {"type":"entrez-protein","attrs":{"text":"P31785","term_id":"400048","term_text":"P31785"}}P31785) was cloned into the BglII and EcoRI restriction sites of the transfer vector pMT-BipA (Invitrogen) and subsequently confirmed by DNA sequencing. The sIL7Rα construct contains an extra Arg-Ser and an eight-residue His tag at the N-terminal end. A stable S2 insect cell line secreting sIL7Rα was generated, and the protein was expressed and purified using similar methods described previously for the IL7Rα-ECD (53). The human TSLP cDNA was purchased from Origene, cloned into the NcoI and BamHI restriction sites of the pET-15b expression vector, confirmed by DNA sequencing, and expressed in E. coli. The TSLP construct contains an extra M-G at the N-terminal end. TSLP was expressed and purified using similar procedures described for IL-7 (53). Human TSLPR fused as an F_c chimera expressed from NS0 cells was purchased from R&amp;D Systems and was used without further purification. Molar absorption coefficients at 280 nm of 6.9, 16.6, 31.9, and 38.8 mM/cm were used to determine protein concentrations of IL-7, TSLP, IL7Rα-EC, and sIL7Rα, respectively (54).

Protein Production-Functional Studies.

Δ6IL7Rα cDNA was synthesized to order based on the GenBank accession no. {"type":"entrez-nucleotide","attrs":{"text":"AK301220","term_id":"194375360","term_text":"AK301220"}}AK301220 sequence by Mr. Gene. The Δ6IL7Rα protein was purified from supernatant obtained from HEK293E cells transiently expressed with a 7.1-kb mammalian expression clone containing the Δ6IL7Rα cDNA sequence flanked by the CMV promoter and His6. Δ6IL7Rα protein was purified from culture supernatant by immobilized metal ion-affinity chromatography.

Surface Plasmon Resonance.

Experiments were performed using a Biacore 3000 SPR instrument at 25 °C. IL-7Rα-ECD, sIL7Rα, and TSLPR coupling and binding kinetics were measured using a CM5 sensor chip. The receptors were amine-coupled to the sensor chip using previously described methods (55). Experiments were performed in 10 mM Hepes (pH 7.4), 150 mM NaCl, 3 mM EDTA, and 0.005% Tween-20 at 25 °C. Binding kinetics of the association between the cytokines and receptors were performed with a flow rate of 50 μL/min. Twofold serial dilutions of differing cytokine concentrations were assayed. Each 250-μL protein or buffer injection was followed by a 40-s dissociation period. Surfaces were regenerated for subsequent runs with a 5-μL injection of 4 M MgCl₂.

Sensorgrams were trimmed and double-referenced (56) using BIAevaluation Version 4.1 (BiaCore). Global-fitting analysis of the sensorgrams used the program ClampXP (57). Experiments were performed in triplicate, and errors were propagated using Taylor series expansion (58). All of the binding interactions fit best to a two-step (three-state) reaction model (55) originally described for SPR analysis (56). Apparent equilibrium dissociation constants (K_d) were calculated using the following equation: K_d = k_-1k_-2/k₁(k₂ + k_-2).

In Vitro Culture.

The 2E8 cells (IL-7–dependent murine pro–B-cell line) and human PBMCs were cultured in media supplemented with a single dose of IL-7 and sIL7Rα at the concentrations indicated. Cell counts were carried out using a Z2 cell counter (Beckman Coulter), and viability was determined by 7AAD gating via flow cytometry. Flow-cytometric analysis was carried out on a FACS LSRFortessa (Becton-Dickinson) using fluorochrome-labeled staining antibodies (listed below). For Stat5 phosphorylation assays, human PBMCs were serum-starved overnight and given a single dose of IL-7 with or without sIL7Ra, and 15 min later, cells were fixed (Fix Buffer I; BD Biosciences) and permeabilized (Perm buffer III), and then anti-Stat5 phosphoantibodies were added according to the manufacturer’s instructions.

Mouse Experiments.

All studies were conducted according to National Cancer Institute Animal Care and Use Committee-approved protocols. IL7^{, CD45.2 female mice were injected with 2 × 10 CD45.1 lymph node cells plus 5 μg of rhIL-7 (CYT107; Cytheris SA) with or without 100 μg of sIL7Rα. Mice were humanely euthanized on day 8, and then total splenocyte counts were determined by homogenizing mouse spleen using a GentleMACS Dissociator (Miltenyi Biotech), lysing red cells (ACK lysing buffer, Lonza), and counting using a Coulter counter. In parallel, IL7 male mice were injected with the same amounts of rhIL-7 with or without sIL7Rα without lymphocytes. Plasma IL-7 levels were determined at days 1 and 4 following retroorbital blood collection.}

For the EAE studies, female C57BL/6 mice were immunized by two s.c. injections of 200 μg of MOG_35-55 (American Peptide Company) emulsified in Complete Freund’s Adjuvant containing 5 mg/mL H37Ra (Chondrex) in both flanks near the tail base; 400 ng of pertussis toxin (Sigma-Aldrich) was given i.p. on days 0 and 2 post MOG injection. On days 9 and 15, PBS, 5 μg of IL-7, 100 μg of sIL7Rα, 5 μg of IL-7 plus 100 μg of sIL7Rα, or 5 μg of IL-7 plus 100 μg of hαIL6Rα (control protein) was injected i.p. (n = 10 per group). Mice were scored daily by blinded observers starting on day 8 using the following system: 0, normal; 1, flaccid tail, no paraparesis; 2, hind-limb weakness evidenced by inability to right itself when placed on the back, or inability to grasp with its hind limbs; 3, partial hind limb paralysis evidenced by inability to move one hind limb (e.g., to withdraw one limb when pinched but able to bear weight on one limb); 4, complete hind-limb paralysis evidenced by inability to move or withdraw limb; and 5, quadriplegia evidenced by inability to move front and hind limbs (humane end point).

Statistical Analyses.

Graphs represent mean values ± SEM. P values were calculated as indicated in each respective figure using Student’s t test or the Mann–Whitney U test. P < 0.05 was considered statistically significant and is illustrated with an asterisk in the figures.

Supplementary Material