CD4+ Group 1 Innate Lymphoid Cells (ILC) Form a Functionally Distinct ILC Subset That Is Increased in Systemic Sclerosis

Abstract

Innate lymphoid cells (ILC) are a heterogeneous group of cellular subsets that produce large amounts of T cell-associated cytokines in response to innate stimulation in the absence of Ag. In this study, we define distinct patterns of surface marker and cytokine expression among the ILC subsets that may further delineate their migration and function. Most notably, we found that the subset previously defined as group 1 ILC (ILC1) contains CD4(+) CD8(-), CD4(-) CD8(+), and CD4(-) CD8(-) populations. Although all ILC1 subsets shared characteristics with Th1 cells, CD4(+) ILC1 also demonstrated significant phenotypic and functional heterogeneity. We also show that the frequencies of CD4(+) ILC1 and NKp44(+) group 3 ILC, but not CD4(-) ILC1 or group 2 ILC, are increased in the peripheral blood of individuals with systemic sclerosis (SSc), a disease characterized by fibrotic and vascular pathology, as well as immune dysregulation. Furthermore, we demonstrate that CD4(+) and CD4(-) ILC1 are functionally divergent based on their IL-6Rα expression and that the frequency of IL-6Rα expression on ILC is altered in SSc. The distinct phenotypic and functional features of CD4(+) and CD4(-) ILC1 suggest that they may have differing roles in the pathogenesis of immune-mediated diseases, such as SSc.

Figure 1

CD4+, CD8+ and DN populations in the ILC1 subset. (A) Gating for peripheral blood ILC subsets: After gating on lymphocytes (FSC^lowSSC^low), singlets, and live CD45+ cells, total ILC were defined as Lin - CD127+. Lineage markers included CD3ε, CD19, CD14, CD123, CD11c, FcεRIα, CD34, CD94, ± CD16. Within total ILC, ILC2 were defined as CRTH2+, ILC1 as CRTH2− c-kit− CD56− and ILC3 as CRTH2− c-kit+. A small fraction of ILC3 were NKp44+. (B) ILC subsets in fresh and frozen PBMC from control subjects were assessed for expression of CD4 and CD8α. Flow plots are from a single representative individual (n > 20). (C) ILC subsets from fresh and frozen PBMC were assessed for surface (s) and intracellular (ic) TCRαβ and CD3ε. CD3ε was also included in the lineage markers in these analyses. TCRαβ and CD3ε MFI in ILC subsets (n = 10), with ** p < 0.0001 using two-tailed unpaired Mann-Whitney tests. (E) Sorted CD4+ and CD4− ILC1 populations were cultured with irradiated feeder cells, IL-2 and IL-7 for 7 days. Live cells were then analyzed for surface expression of CD3ε, TCRαβ and CD56 by flow cytometry. Flow plots are from a single individual (n = 2).

Figure 2

Gradient of Th1 phenotype in ILC1 subsets. ILC subsets in fresh or frozen PBMC from control subjects were analyzed for transcription factor and CXCR3 expression. (A) T-bet and GATA-3 expression in ILC subsets (gray line = isotype control; black line = transcription factor). Histograms show representative staining from n = 5 - 9 subjects. (B) T-bet, Eomes and CXCR3 expression in CD4+, CD8+ and DN ILC1 by flow cytometry. Flow plots are from a single representative individual from a total of n = 5 - 9 subjects for each stain set. (C) Frequency of T-bet, EOMES and CXCR3 positive cells in CD4+, CD8+ and DN ILC1 (n = 4 - 5). In (A) and (C), * p < 0.05 and ** p < 0.0001 using two-tailed unpaired Mann-Whitney tests.

Figure 3

CD4+ and CD4− ILC1 cytokine production. Fresh PBMC were prepared from peripheral blood, lineage-depleted using a MACS column, stained for ILC surface markers, and then flow sorted for (A) CD4+ and CD4− ILC1 subsets (n = 3) or (B) CXCR3+ and CXCR3− populations within the CD4+ and CD4− ILC1 subsets (n = 3). In both (A) and (B), the sorted populations were stimulated with PMA/ionomycin/brefeldin A for 6 hours, and then examined for intracellular cytokine expression by flow cytometry.

Figure 4

Distinct chemokine receptor and costimulatory marker expression on ILC subsets. (A) Fresh PBMC from control subjects were isolated from peripheral blood, rested for one hour at 37°C/5% CO₂, and then chemokine receptor expression was analyzed on ILC subsets by flow cytometry. Flow plots are from a single representative individual (n = 5 - 7). (B) ILC subsets from fresh or frozen PBMC were examined for their activation marker and CD45RO expression by flow cytometry. Flow plots are from a single representative individual (n > 10).

Figure 5

High frequency of IL-6Rα expression on CD4+ ILC1. Fresh or frozen PBMC were analyzed for (A) IL-6Rα and gp130 expression on ILC subsets and (B) phosphorylation of STAT-1 and STAT-3 following a 20-minute stimulation with 5 ng/mL of IL-6. Flow plots are representative stains from n > 5 subjects for each stain set. (C) gp130+/IL-6Rα+ frequencies, IL-6Rα MFI and post-stimulation p-STAT-1 and pSTAT-3 frequencies in ILC subsets from fresh PBMC (n = 5 per stain set). * p < 0.05 and ** p < 0.01 using two-tailed unpaired Mann-Whitney tests.

Figure 6

A small number of genes are differentially expressed in CD4+ versus CD4− ILC1. RNAseq was performed on peripheral blood CD4+ and CD4− ILC1 from 3 healthy control subjects. The fold-change (FC), expressed as Log2(FC), of genes with an FDR < 0.05 using DESeq2 were considered significant.

Figure 7

Altered ILC subset frequencies in SSc. (A) Frozen PBMC from gender- and age-matched controls and SSc subjects were thawed, stained for ILC subsets, and then analyzed by flow cytometry using the gating strategy outlined in Figure 1A (n = 38 per cohort). (B) A subgroup of the subjects examined in (A) designated “1st cohort,” included analyses of CD4 expression, and these data on CD4+ and CD4− ILC1 frequencies were replicated in a second cohort of age- and gender-matched control and SSc subjects (n= 19 - 20 per cohort). (C) ILC subset expression of CD2, CD27 and CD28 was examined on previously frozen PBMC from SSc subjects. Histograms are representative staining from n = 5 subjects. (D) ILC subset expression of IL-6Rα and gp130 was examined on previously frozen PBMC samples from control and SSc subjects (n = 20 per cohort). P-values were determined using two-tailed unpaired Mann-Whitney tests.