Human airway mast cells proliferate and acquire distinct inflammation-driven phenotypes during type 2 inflammation

Abstract

Mast cells (MCs) play a pathobiologic role in type 2 (T2) allergic inflammatory diseases of the airway, including asthma and chronic rhinosinusitis with nasal polyposis (CRSwNP). Distinct MC subsets infiltrate the airway mucosa in T2 disease, including subepithelial MCs expressing the proteases tryptase and chymase (MC_TC) and epithelial MCs expressing tryptase without chymase (MC_T). However, mechanisms underlying MC expansion and the transcriptional programs underlying their heterogeneity are poorly understood. Here, we use flow cytometry and single-cell RNA-sequencing (scRNA-seq) to conduct a comprehensive analysis of human MC hyperplasia in CRSwNP, a T2 cytokine-mediated inflammatory disease. We link discrete cell surface phenotypes to the distinct transcriptomes of CRSwNP MC_T and MC_TC, which represent polarized ends of a transcriptional gradient of nasal polyp MCs. We find a subepithelial population of CD38^highCD117^high MCs that is markedly expanded during T2 inflammation. These CD38^highCD117^high MCs exhibit an intermediate phenotype relative to the expanded MC_T and MC_TC subsets. CD38^highCD117^high MCs are distinct from circulating MC progenitors and are enriched for proliferation, which is markedly increased in CRSwNP patients with aspirin-exacerbated respiratory disease, a severe disease subset characterized by increased MC burden and elevated MC activation. We observe that MCs expressing a polyp MC_T-like effector program are also found within the lung during fibrotic diseases and asthma, and further identify marked differences between MC_TC in nasal polyps and skin. These results indicate that MCs display distinct inflammation-associated effector programs and suggest that in situ MC proliferation is a major component of MC hyperplasia in human T2 inflammation.

Figure 1:. Phenotypic characterization of sinus MC hyperplasia

(A) Flow identification of human sinus MCs. (Additional replicates and full gating in Figure S1). (B) MC quantification as percentage of CD45⁺ (left) or total cells (right) in indicated patient groups (n=6–12 donors/group). *, p<0.05; **, p<0.01 (Mann-Whitney). (C) Representative plots showing MC heterogeneity in nasal polyp (left) or CRSsNP tissue (right). (D) MC phenotype in polyp epithelium (left), whole polyp (center) with overlay (right). Red: epithelium; Black: whole polyp. Representative of five donors. (E) Gating distinguishing epithelial MCs (CD117^low SSC^low, red) from subepithelial MCs (CD117^high, blue) in epithelial fraction (left panel) vs unfractionated polyp (right panel) (F) Quantification of epithelial and subepithelial MCs in fractionated epithelium and unfractionated polyp (n=5 donors); *, p<0.05 (paired t-test). (G) FcεR1α expression of epithelial (red) and subepithelial MCs (blue) with quantification (right). Lines denote paired observations (n=24 donors); ****, p<5 × 10⁻⁵ (paired t-test) (H) Toluidine blue (left) and protease immunophenotyping (right) of sorted CD117^low MCs. Two examples provided to show heterogeneity. Green: tryptase; Red: chymase; Yellow: colocalization; Blue: nucleus. (I) Toluidine blue (left) and protease immunophenotyping (right) of CD117^high MCs. Two examples provided to show heterogeneity. Green: tryptase; Red: chymase; Yellow: colocalization; Blue: nucleus.

Figure 2:. Identification polyp MCs with a progenitor-like cell surface phenotype

(A) Identification of nasal polyps SSC^lowItgβ7 ^high MCs (left). CD117 and SSC expression on Itgβ7^high MC (magenta) relative to all polyp MCs (grey) (right). (B) Toluidine blue (left) and protease immunophenotype (right) of sorted Itgβ7 ^high polyp MCs MCs. Green: tryptase; Red: chymase; Blue: nucleus (C) Representative flow plot of Itgβ7 expression on CRSsNP MCs. (D) Quantification of Itgβ7^high MCs as a percentage of total MCs in CRSsNP, CRSwNP and AERD. **, p<0.01 (Mann-Whitney). (E) CD34 expression on polyp Itgβ7^high MCs (magenta) and circulating MC_P (turquoise) vs isotype (grey), representative of three independent donors. (Magenta and grey are superimposed). (F) Integrin expression on circulating MC_P (left) and polyp Itgβ7^high MCs (right), each representative of three separate donors. (G) Distribution of polyp αE^high (green) and αE^- (purple) Itgβ7^highMCs distribution within the epithelial vs subepithelial gates (left) as in Fig. 1E, with quantification (right). *, p<0.05 (paired t-test).

Figure 3:. scRNA-seq identification of transcriptionally distinct polyp MC subsets

(A) Schematic representation for scRNA-seq analysis of 7,355 cells sorted polyp MCs (Fig. S1) (n=6 donors). (B) Uniform Manifold Approximation and Projection (UMAP) depiction of MC clusters. (Fig. S3 depicts identification of MC vs contaminating clusters). (C) MC expression of canonical MC transcripts. (D) Common (top rows) and cluster-enriched transcripts (columns show donor-averaged row-normalized expression). p_adj < 3.1 × 10⁻²⁹ for cluster enriched genes (Wilcoxon). (E) Polyp MC expression of MC1 and MC3 signatures (% of transcripts/cell). (F) Per-cell expression of MC1 and MC3 signatures across all clusters (top) and MC4 alone (bottom). (correlation r-value). (G) Differentially expressed transcripts between polyp MC_TC (MC1) and MC_T (MC3), (donor averaged, row normalized); p_adj < 6 × 10⁻⁹ (Wilcoxon). (H) Immunomodulatory transcript expression in polyp MC_TC and MC_T, (donor averaged, row normalized); p_adj < 7 × 10⁻¹⁷ (Wilcoxon). (I) FcεR1 signaling pathway expression in polyp MC_TC and MC_TC (donor averaged, row normalized); p_adj < 8 × 10⁻⁴ (Wilcoxon). (J) Expression of transcript scores driven by IgE signaling, IL33 signaling, or shared by both stimuli across the MC1–3 clusters. (lines denote median and quartiles) p***, p<1×10⁻¹⁰, ***, p<1×10⁻¹⁵ (Mann-Whitney); Cohen’s effect size for MC1 vs. MC3: 0.60 (IgE), 0.79 (IL33), and 1.52 (shared).

Figure 4:. Characterization of an IL-4-elicited MC_T phenotype enriched in diseased human lung samples.

(A-C) Re-clustering of MCs from three scRNA-seq datasets accessed through the human IPF atlas: (A) control, IPF and ILD; (B) control, IPF and COPD; and (C) control and IPF. Clustering for each dataset indicated a population (polygon gate) that was statistically enriched for the polyp MC3-associated transcripts IL17RB, CD38, and GPR183 (center panels) and predominantly composed of MCs from diseased tissue relative to healthy (left panels). Circle gate indicates MCs co-expressing the proliferation-associated genes MKI67 and TOP2A. (D) Correlation analysis of polyp scRNA-seq epithelial IL-4/13-induced signature expression, (donor averaged) versus scRNA-seq defined MC percentage for each donor. p<0.001. (E) MC3-enriched transcripts (Supplementary table 4) upregulated in two technical replicate CBMC samples by 96-hour IL-4 stimulus (row normalized expression). Top half: FDR<0.1, bottom half (indicated by purple line): P < 0.05 (DESeq2). (F) MC1-enriched transcripts (supplemental table 4) downregulated in two technical replicate CBMC samples by 96 hours IL-4 stimulus (row normalized expression). Top half: FDR<0.1, bottom half (indicated by purple line): P < 0.05 (DESeq2). (G) CBMCs IL17RB expression (qPCR) following 72-hour stimulus with vehicle or IL-4, (n=4 biologic replicates across three independent experiments), * indicates p<0.05 (paired t-test).

Figure 5:. Identification of MC_TC heterogeneity across tissues

(A) Transcripts shared between diseased and control skin MCs and polyp MC_TC (top) vs. enriched in skin MCs (upper middle), polyp MC_T (lower middle), or polyp MC_TC (bottom) (donor averaged, row normalized). (B) Representative analysis of MRGPRX2 expression (red) on nasal poly MCs (left) and skin MCs (right) relative to isotype (grey). Staining representative of 3 donors per group. (C) Identification of MCs based on TPSAB1 expression within a scRNAseq dataset of healthy human lung tissue containing samples from distal, medial and proximal lung generated by Travaglini et al(47). (D) Violin plots showing expression of the MC_TC-associated transcripts CMA1, CTSG, MRGPRX2, and C5AR1 in healthy human lung. No displayed transcripts were significantly differentially expressed between distal and proximal lung, MRGPRX2 expression was only observed in proximal lung.

Figure 6:. CD38 expression marks nasal polyp MC_T and an unpolarized intermediary subset

(A) Representative plot showing nasal polyp MC (gated as in Fig. S1) expression of CD38 and CD117. Arrow indicates sequential gating. (B) SSC vs CD117 profile of CD38^high epithelial MCs (red), CD38^high subepithelial MCs (orange) and CD38^low subepithelial MCs (blue). (C) Representative plot showing CD38 and CD117 expression in Itgβ7^high MCs. (D) Representative plot of MC CD38 and CD117 expression in CRSsNP control tissue. (E) Quantification of CD38^highCD117^low MC_T (red), CD38^highCD117^high intermediate MCs (orange) and CD38^lowCD117^high MC_TC (blue) in sinus tissue of patients with CRSsNP (closed circles), CRSwNP (open squares), or AERD (closed triangles). ** indicates p<0.01 (Mann-Whitney) (F) Intracellular chymase expression in epithelial MCs (red), CD38^high subepithelial MCs (orange), and CD38^low subepithelial MCs (blue) versus isotype (grey). (G) Quantification of chymase expression in indicated MC subsets (n=5 donors); *, p<0.05, ***, p<0.001 (t-test). (H) Intracellular FITC-avidin staining in epithelial MCs (red), CD38^high subepithelial MCs (orange), and CD38^low subepithelial MCs (blue). (I) Quantification of FITC-avidin fluorescence in indicated MC subsets (n=9 donors); *, p<0.05 (t-test).

Figure 7:. Unpolarized MC proliferation underlies MC expansion in human nasal polyps.

(A) Schematic approach for Cellular Indexing of Transcriptomes and Epitopes by Sequencing (CITE-seq) analysis of sorted polyp-derived MCs (B) Identification of four clusters across 2,902 MCs from 2 donors with expression of the MC1 and MC3 gene signatures (Supplementary table 2) and MKI67. (C) Cell surface expression of CD117 and CD38 within each MC cluster (line denotes median and quartiles); **, p<1×10⁻⁴; ***, p<1×10⁻¹²; ****, p<1×10⁻¹⁵; NS, not significant (Mann-Whitney). (D) UMAP plot showing cell surface expression of Itgβ7 across nasal polyp MCs. (E) Representative plot of nuclear Ki67 in nasal polyp MCs. (F) Overlay showing Ki67+ MCs (purple) and all MCs (grey) with flow gates defined in Fig. 6A (left). Quantification of Ki67+ MCs by subset (right); ***, p<0.001 (t-test). (G) Quantification of Ki67+ MCs in indicated disease endotypes. ***, p<0.00, (t-test). (H) Correlation of MC proliferation (Ki67 staining) and peripheral blood eosinophil counts across polyp patients. Pearson R = 0.481, p < 0.05