IL-7 receptor blockade blunts antigen-specific memory T cell responses and chronic inflammation in primates

Abstract

Targeting the expansion of pathogenic memory immune cells is a promising therapeutic strategy to prevent chronic autoimmune attacks. Here we investigate the therapeutic efficacy and mechanism of new anti-human IL-7Rα monoclonal antibodies (mAb) in non-human primates and show that, depending on the target epitope, a single injection of antagonistic anti-IL-7Rα mAbs induces a long-term control of skin inflammation despite repeated antigen challenges in presensitized monkeys. No modification in T cell numbers, phenotype, function or metabolism is observed in the peripheral blood or in response to polyclonal stimulation ex vivo. However, long-term in vivo hyporesponsiveness is associated with a significant decrease in the frequency of antigen-specific T cells producing IFN-γ upon antigen restimulation ex vivo. These findings indicate that chronic antigen-specific memory T cell responses can be controlled by anti-IL-7Rα mAbs, promoting and maintaining remission in T-cell mediated chronic inflammatory diseases.

Fig. 1

Schematic representation of cytokine-induced receptor heterodimerization signaling mechanisms as previously proposed. During the initiation step, IL-7 interacts with the extracellular domain 1 (D1) of IL-7Rα, generating the site-1 interface. This leads to the intermediate step where a 1:1 complex can associate with the shared common gamma-chain (γ_c) receptor. The binding of γ_c receptor involves an interface between IL-7 and γ_c called site-2a and an interface between D2 regions of the IL-7Rα and γ_c receptor called site-2b. The stabilized heterodimer complex activates the JAK/STAT and possible other signaling pathways

Fig. 2

Anti-IL-7Rα mAbs induce long-term control of skin inflammation. a Cutaneous erythema diameters measured daily after tuberculin intradermal injection in baboons treated with a single intravenous injection with 10 mg/kg of the site-1/2b IgG1 mAb (red, n = 3), site-1/2b IgG4 mAb (blue, n = 3) or site-1 IgG4 mAb (green, n = 3). Control erythema (closed symbols) was performed 4 weeks before treatment. Baboons were then rechallenged 24 h after mAb treatment (open symbols). b Representative skin immunostaining (green: CD3, red: CD68, blue: nucleus) 4 days after tuberculin challenge performed a month before or 24 h after mAb injection as described in (a). c Area under curve (AUC) of cutaneous erythema diameters measured after tuberculin intradermal challenges performed monthly or at the indicated time in baboons treated with site-1/2b mAb (IgG1: red and n = 3, IgG4: blue and n = 3) as in (a), or in historical placebo-treated control animals (black histograms) following the same protocol. n.d. not determined in control animals. Data are mean ± SEM and erythema diameters were measured by at least two observers. *p < 0.05; **p < 0.01 one-way ANOVA within site1/2b treated animal group across all timepoints as compared to pre-treatment responses. # p < 0.05 Mann–Whitney between pre-re-vaccination and post-re-vaccination

Fig. 3

Anti-human IL-7Rα mAbs, epitopes, and antagonist/agonist signals. a Epitope determination by Hydrogen Deuterium Exchange and Mass Spectrometry (HDX-MS). Heatmaps of the relative fractional uptake differences between the bound and unbound antigen (recombinant human IL-7Rα) with the site-1/2b IgG4, site-1 IgG4 #1, and site-1 IgG1 #2 anti-human IL-7Rα mAbs. The relative fractional uptake of each amino acid is inferred from the value obtained for the smallest overlapping peptide, or the most C-terminal if two or more peptides are of equal length. Amino acids essential for IL-7 binding on IL-7Rα (site 1) are circuled green, the predicted site-2b on IL-7Rα interaction with γ-chain is circuled red and the interaction site of IL-7Rα with TSLPR circuled in dotted black as previously described^,,. b Representation of the IL-7 (brown)/IL-7Rα (grey) structure of each epitope recognized by site-1/2b or site-1 anti-human IL-7Rα mAbs as determined by HDX-MS in (a). Epitopes (red, blue, orange or green colors) were determined on the most intense relative fractional uptake differences between the bound and unbound antigen. c Representative phospho-STAT5, phospho-PI3k-p55, phospho-ERK1/2, and GAPDH western blot of one out of seven representative human donor cells. PBMCs were pretreated with 10 µg/mL of one anti-IL-7Rα mAb and then incubated for 10 min at 37 °C with or without 5 ng/ml of human IL-7. d Same as (c) with quantification of pSTAT5, pPI3K, and pERK signals corrected to GAPDH expression and normalized to medium control conditions (n = 7 different donors). *p < 0.05; **p < 0.01 between indicated groups

Fig. 4

Dual agonist/antagonist (site-1) anti-IL-7Rα mAbs induce transcriptional modification and activation of human leukocytes. a RNA-Seq analysis of human PBMCs (n = 7) incubated without IL-7 for 3.5 h with different anti-human IL-7Rα mAbs (10 µg/ml, blue: site-1/2b IgG4, red: site-1 IgG4#1, green: site-1 IgG1#2). Upper left: Venn diagram of the 481 differentially expressed genes (FDR 5%, FC > 1.5) comparing anti-IL-7Rα mAbs and medium control conditions. Upper right: Principal component analysis (PCA) of anti-IL-7Rα mAbs versus control and IL-7 stimulation (5 ng/ml) on the 93 most differentially expressed genes (FDR 5%, FC > 2) comparing IL-7 stimulation and control conditions. Bottom: Relative expression (fold-change as compare to control medium condition) of selected genes with each anti-IL-7Rα mAb (blue: site-1/2b IgG4, red: site-1 IgG4#1, green: site-1 IgG1#2). *p < 0.05; **p < 0.01; ***p < 0.001 as compare to control medium condition. b RNA-Seq analysis of human PBMCs (n = 7) incubated with IL-7 (5 ng/ml) for 3.5 h with the different anti-human IL-7Rα mAbs (10 µg/ml). Upper: Heatmap of the expression of the 93 most differentially expressed genes (FDR 5%, FC > 2) between IL-7 stimulation and control condition. Bottom: Quantification of the median profile of the three IL-7 induced clusters in IL-7 stimulated, control and IL-7 + anti-human IL-7Rα mAbs conditions. Same colors as in (a). *p < 0.05; **p < 0.01; ***p < 0.001 between indicated groups

Fig. 5

Antagonist anti-IL-7Rα mAb induces antigen-specific memory T cell hyporesponsiveness. a Epidermal thickness measured histologically on skin biopsies performed 4 days after tuberculin intradermal injection in baboons treated with a single intravenous injection of 10 mg/kg of a humanized site-1/2b IgG4 mAb (n = 3). Pre-T: 4 weeks pre-treatment with the mAb; W5, W8, W11 are the number of weeks (W) after mAb administration. W26 post-BCG is a new tuberculin challenge after re-vaccination of animals with BCG vaccine. b Cutaneous erythema response at the indicated time-point as in (a), represented with area under the curve (AUC) of daily erythema diameters. c IFN-γ-secreting cell frequencies in baboon PBMCs after ex vivo tuberculin restimulation at the indicated time-point as in (a). d At indicated time-points, week 21 (before BCG re-vaccination) and week 26 (post-BCG re-vaccination), 600 IU/ml of recombinant human IL-2 or 10 µg/mL of the humanized site-1/2b IgG4 mAb were added in some wells during ex vivo tuberculin restimulation. In some conditions, CD25⁺ depletion was performed on PBMCs before tuberculin restimulation to assess a potential role of Tregs. Data are mean ± SEM and erythema diameters were measured by at least two observers. *p < 0.05; **p < 0.01 one-way ANOVA with Dunn’s multiple comparison test

Fig. 6

Antagonist anti-IL-7Rα mAb inhibits immune cells skin infiltration. Representative hematoxylin & eosin (upper), CD68 (middle), and CD3 (bottom) staining of skin biopsies performed 4 days after tuberculin challenge at the indicated time-points in animals treated with a single intravenous injection of 10 mg/kg of a humanized site-1/2b IgG4 mAb. Pre-T: 4 weeks pre-treatment with the mAb; W5, W8, W11 are numbers of weeks (W) after mAb administration. Post-BCG: new tuberculin challenge after re-vaccination of animals with BCG

Fig. 7

Antagonist anti-IL-7Rα mAb does not disturb polyclonal T-cell activation or modify metabolism. a Peripheral blood T-cell subset frequencies determined by flow cytometry for baboons treated with a single intravenous injection of 10 mg/kg of a humanized site-1/2b IgG4 mAb (n = 4). T cell sub-populations were defined using the following gating strategy CD3⁺ cells for Tem: effector memory T cells (CCR7⁻ CD45RA⁻), Tcm: central memory T cells (CCR7⁺ CD45RA⁻), Tn: naive T cells (CCR7⁺ CD45RA⁺), TEMRA: CD45RA expressing effector memory T cells (CCR7⁻ CD45RA⁺). b CD25⁺ (upper) and CD69⁺ (lower) activated T-cell frequencies at the indicated time-point in the same animals as in (a), after ex vivo culture (48 h), without stimulation or with anti-CD3 polyclonal stimulation, and supplemented with human IL-2 (300 UI/mL) or human IL-7 (10 ng/mL) as mentioned. c Oxygen consumption rate (OCR) and extracellular acidification rate (EACR) of T cells purified one week before and after anti-IL-7Rα treatment in the same animals as in (a). OCR and EACR was determined on freshly isolated and unstimulated T cells or after over-night polyclonal activation using Seahorse XF24 Analyzer. d ATP concentration (µg/mL) and mean fluorescence intensity (MFI) of 2-NBD glucose uptake in the same condition as in (c). Data are mean ± SEM

Fig. 8

Antagonist anti-IL-7Rα mAb inhibits antigen-specific human memory T cells persistence after antigen rechallenge. Human CPD-labeled PBMCs (n = 4) were stimulated with MHCI or MHCII -restricted pool of peptides (H1N1 flu; CEFT: CMV, EBV, Flu, Tetanos; CEF: CMV, EBV, Flu) or with anti-CD3 + anti-CD28 mAbs for 3 days (a), 8 days (b) or 10 days (c). Cells were cultured in medium alone (white histogram) or supplemented with 5 ng/ml of recombinant human IL-7 (black histogram) or 5 ng/mL of human IL-7 plus 10 µg/ml of the site-1/2b humanized anti-IL-7Rα mAb (grey histogram). The histograms show the percentage of proliferated cells (% of CPD^low cells) and viable cells (% of Annexin-V⁻ PI⁻ cells) among total cells, or within proliferated (CPDlow, Fig. 7c middle) versus quiescent (CPD high, Fig. 7c right) cells. Horizontal bars mean ± SEM. *p < 0.05 between indicated groups