Human IL-23 is essential for IFN-γ-dependent immunity to mycobacteria

Abstract

Patients with autosomal recessive (AR) IL-12p40 or IL-12Rβ1 deficiency display Mendelian susceptibility to mycobacterial disease (MSMD) due to impaired IFN-γ production and, less commonly, chronic mucocutaneous candidiasis (CMC) due to impaired IL-17A/F production. We report six patients from four kindreds with AR IL-23R deficiency. These patients are homozygous for one of four different loss-of-function IL23R variants. All six patients have a history of MSMD, but only two suffered from CMC. We show that IL-23 induces IL-17A only in MAIT cells, possibly contributing to the incomplete penetrance of CMC in patients unresponsive to IL-23. By contrast, IL-23 is required for both baseline and Mycobacterium-inducible IFN-γ immunity in both Vδ2⁺ γδ T and MAIT cells, probably contributing to the higher penetrance of MSMD in these patients. Human IL-23 appears to contribute to IL-17A/F-dependent immunity to Candida in a single lymphocyte subset but is required for IFN-γ-dependent immunity to Mycobacterium in at least two lymphocyte subsets.

Figure 1:. Six patients from four kindreds with private homozygous IL23R variants

(A) Pedigrees of the four kindreds studied here. The IL23 variants are indicated below the kindred name. Solid black symbols indicate patients with MSMD, and black diagonal stripes indicate CMC. Symbols linked with a double line indicate consanguinity. The genotype is indicated under each symbol, with M corresponding to the variant found in each kindred, and WT indicating wild-type. Arrows indicate the index case in each family. (B) Schematic representation of the WT IL-23R gene/protein, the colorless area represents the extracellular domain, the black area the transmembrane domain and the gray area the intracellular domain of the protein. The positions of the variants found in the patients are indicated by arrows. (C) All homozygous variants found in gnomAD V2.1.1 (in black) for IL23R are plotted according to their combined annotation-dependent depletion (CADD score; y axis) and minor allele frequency (MAF; x axis). The dashed line indicates the mutation significance cutoff (MSC) for IL23R. (D) Gene14 level negative selection. IL23R is not under negative selection, like other genes for which mutations underlie AR inborn errors of immunity (IEI), as determined by CoNeS. (E) Schematic representation of the impact of the different variants of the patients (P1-P6) on IL- 23R.

Figure 2:. Four loss-of-function IL23R alleles, and six patients with AR complete IL23R Deficiency

(A-B) EBV-B^STAT4 cells were either left non-transduced (NT) or were transduced with lentiviruses generated with an empty vector (EV) or with vectors containing the WT or the mutated IL23R cDNA. Cell-surface IL-23R expression was assessed by flow cytometry (A). EBV-B^STAT4 cells were left unstimulated (NS), or were stimulated with IL-23 or IFN-α2b as a positive control. STAT3 phosphorylation was assessed by flow cytometry (B). (C) EBV-B cells from a healthy control (Ctrl.), P1, P3 and an IL-12Rβ1-deficient patient were left unstimulated (NS), or were stimulated with IL-23 or IFN-α2b as a positive control. STAT3 phosphorylation was assessed by flow cytometry. The results shown in A-C are representative of three independent experiments. (D) T-blasts from a healthy control (Ctrl.), P4, P5, P6, and an IL-12Rβ1-deficient patient were left unstimulated (NS), or were stimulated with IL-23 or IFN-α2b as a positive control. STAT3 phosphorylation was assessed by flow cytometry. (E) Whole blood from two healthy controls, P5, and P6 was left unstimulated (NS), or was stimulated with IL-23 or IFN-α2b as a positive control. STAT3 phosphorylation in the various cell subsets was assessed by pSTAT CyTOF and the pSTAT3-positive cells were depicted in a uniform manifold approximation and projection (UMAP) visualization.

Figure 3:. Impaired basal IFN-γ signaling in vivo in patients with inherited IL-23R Deficiency

(A) Frequency of memory CD4⁺ T cells (T_H1, T_H2, T_H17, T_H1*) and absolute counts of NK and γδ T cells in 19 healthy adult controls, 20 healthy pediatric controls, and patients (P4, P5 and P6). (B-F) Single-cell transcriptome analysis. Cryopreserved peripheral blood mononuclear cells (PBMCs) from P3, P4 and P6 were analyzed with cryopreserved PBMCs from healthy adult and pediatric controls, two IL-12Rβ1- and two IL-12Rβ2-deficient patients, and one patient heterozygous for a STAT1 gain-of-function (GOF) variant. (B) Clustering analysis. After batch-correction with Harmony (28), clusters were identified manually with the aid of the SingleR pipeline (78), guided by the MonacoImmuneDataset (79). (C) Pseudobulk differential expression analysis. IL-23R-, IL-12Rβ1-, and IL-12Rβ2-deficient and STAT1-GOF patients were compared with healthy controls (adults and children combined). Gene set enrichment analysis (GSEA) was conducted on the fold-change ranking against the Hallmark gene sets (http://www.gseamsigdb.org/gsea/msigdb/genesets.jsp?collection=H). Only immune-related pathways are shown. (D) Differential gene expression in classical monocytes between adult and pediatric control cells and IL-23R-, IL12Rβ1-, and IL-12Rβ2-deficient and STAT1-GOF cells. Colors reflect Z-transformed normalized pseudobulk read counts. (E) Intercellular communication analysis with CellChat (65). Principal component analysis (PCA) of the CellChat-predicted signaling probability of communication between all cell-to-cell pairs and all receptor-ligand pairs (left panel), and PCA on CellChat-predicted probability of signaling to and from MAIT cells (right panel). (F) GSEA was conducted on the fold-change ranking against all transcription factor target gene sets (http://www.gsea-msigdb.org/gsea/msigdb/genesets.jsp?collection=TFT) of IL-23R-deficient and STAT1-GOF T_H17 cells. NES, normalized enrichment score.

Figure 4:. Impaired ex vivo IL-23-mediated production of IFN-γ in patients with inherited IL-23R deficiency

(A-D) Single-cell RNA sequencing. We analyzed PBMCs from two IL-23R-deficient (P4 and P6), one IL-12RbR1-deficient patient, and one STAT1-GOF patient, together with six healthy controls (including P4’s sister). Cells were either left unstimulated or were stimulated with IL-23 for 6 hours. (A) Clustering analysis. (B) Pseudobulk differential expression (DE) analysis. Log₂ fold-changes in expression were estimated with DESeq2 for the interaction between stimulation (non-stim. vs. IL-23) and genotype (WT vs. IL-23R-deficient, IL-12Rβ1-deficient or STAT1 GOF). GSEA for the Hallmark gene sets was performed on the basis of log₂ fold-change ranking. Gene sets with FDR-adjusted P values below 0.05 for at least one cell subset are shown. (C) Percentage of cells expressing IFNG among each cell subset, without (left panel), or with (right panel) IL-23 stimulation. (D) Two-dimensional plot, log₂ fold-change (FC) for the interaction between IL-23R deficiency and stimulation (x axis) and the interaction between IL-12Rβ1 deficiency and stimulation (y axis) in T_H1, T_H2/T_H17, T_H1*, CD8 EM, MAIT, Vδ2⁺ γδ T, and NK cells (red, IFNG. blue, TNF). (E) Percentages of IFN-γ⁺ cells, on intracellular flow cytometry, for the cell subsets indicated, following stimulation with and without live M. bovis-BCG in the presence and absence of IL-12 and IL-23, or PMA/ionomycin (P/I). Nonparametric Mann-Whitney tests were used for analysis (E), with *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 5:. Normal development of BCG- and C. albicans specific memory CD4⁺ T cells in patients with inherited IL-23R deficiency

(A-B) Naïve (A) or memory (B) CD4⁺ T cells from healthy controls (Ctls), P4, and P5 were either kept under T_H0 (T-cell activation/expansion [TAE] beads) conditions, or were polarized under T_H1 (TAE beads + IL-12), T_H2 (TAE beads + IL-4), T_H17 (TAE beads + IL-1/IL-6/IL-21/IL-23/TGF-β), or T_H9 (TAE beads + IL-9) conditions. After five days of culture, the percentages of IFN-γ⁺, IL-17A⁺, and IL-17F⁺ cells were assessed intracellularly by flow cytometry. (C-F) Multiple memory CD4⁺ T cell lines were generated by stimulation with PHA, IL-2 and irradiated allogeneic PBMCs from the sorted memory CD4⁺ T cells of three healthy controls, P2, P3, P4, and P6. Lines were screened for reactivity with peptide pools covering antigens from BCG, Mycobacterium tuberculosis (MTB), or C. albicans (CA). (C) The proliferation of memory CD4⁺ T-cell lines after stimulation with autologous B cells pulsed with BCG, MTB, or CA peptide pools was measured by determining ³H-thymidine incorporation, and the frequencies (mean number per million CD4⁺ memory T cells) of BCG-, MTB-, and CA-specific memory CD4⁺ T cells were estimated assuming a Poisson distribution. (D) Proliferation, measured by ³H-thymidine incorporation, of memory CD4⁺ CCR6⁻ or CD4⁺ CCR6⁺ T-cell lines, stimulated with autologous B cells pulsed with BCG or CA peptide pools. (E-F) BCG, MTB, or CA reactive memory CD4+ T-cell lines from each individual were selected, and the concentrations of IFN-γ (E) and IL-17A (F) in the supernatant were determined with a Luminex assay. Each dot on the graph corresponds to the value for a single antigen-reactive T-cell line.

Figure 6:. Impaired ex vivo IL-23-mediated production of IL-17 cytokines in patients with inherited IL-23R deficiency

(A) Percentages of IL-17A⁺ cells on intracellular flow cytometry for the indicated cell subsets, from the indicated individuals (healthy controls, Ctrls; patients, P3, P4, P6; one IL-12Rβ1 deficient patient, and one STAT1-GOF patient) unstimulated (–), or stimulated with IL-23 or PMA/ionomycin (P/I). (B) IL-17A and IL-17F secretion by PBMCs, assessed by Legendplex assays on the supernatants, for the indicated individuals (healthy controls, Ctrls; patients P5, P6; and one IL-12Rβ1-deficient patient). The cells were cultured for five days, and were then either left unstimulated (–), or stimulated with IL-1β (+), and/or IL-23 (+), in the presence (+) or absence (–) of heat-killed C. albicans (HKCA), or anti-CD2/CD3/CD28 mAb-coated beads (CD2/3/28). Nonparametric Mann-Whitney tests were used for analysis in panels A-B (*p<0.05, **p<0.01, ****p<0.0001).