TGFβ receptor mutations impose a strong predisposition for human allergic disease

Abstract

Transforming growth factor-β (TGFβ) is a multifunctional cytokine that plays diverse roles in physiologic processes as well as human disease, including cancer, heart disease, and fibrotic disorders. In the immune system, TGFβ regulates regulatory T cell (Treg) maturation and immune homeostasis. Although genetic manipulation of the TGFβ pathway modulates immune tolerance in mouse models, the contribution of this pathway to human allergic phenotypes is not well understood. We demonstrate that patients with Loeys-Dietz syndrome (LDS), an autosomal dominant disorder caused by mutations in the genes encoding receptor subunits for TGFβ, TGFBR1 and TGFBR2, are strongly predisposed to develop allergic disease, including asthma, food allergy, eczema, allergic rhinitis, and eosinophilic gastrointestinal disease. LDS patients exhibited elevated immunoglobulin E levels, eosinophil counts, and T helper 2 (TH2) cytokines in their plasma. They had an increased frequency of CD4(+) T cells that expressed both Foxp3 and interleukin-13, but retained the ability to suppress effector T cell proliferation. TH2 cytokine-producing cells accumulated in cultures of naïve CD4(+) T cells from LDS subjects, but not controls, after stimulation with TGFβ, suggesting that LDS mutations support TH2 skewing in naïve lymphocytes in a cell-autonomous manner. The monogenic nature of LDS demonstrates that altered TGFβ signaling can predispose to allergic phenotypes in humans and underscores a prominent role for TGFβ in directing immune responses to antigens present in the environment and foods. This paradigm may be relevant to nonsyndromic presentations of allergic disease and highlights the potential therapeutic benefit of strategies that inhibit TGFβ signaling.

Figures

Fig. 1. Evidence for allergic predilection in LDS patients.

(A to C) Biopsies stained with hematoxylin and eosin demonstrating an eosinophilic infiltrate in the esophagus (A), stomach (B), and colon (C) of a child with LDS. Magnification, ×40. (D) Percentage of eosinophils in the peripheral blood of LDS patients (n = 50). Levels were significantly increased (P = 0.009) compared to the norm (shaded box) by Wilcoxon test. Line and whiskers indicate mean and SD, respectively. (E) Total serum levels of IgE (kU/liter) from LDS patients versus age (n = 41). Levels were elevated (P = 0.016; Student’s t test, two-tailed) above the 95% confidence interval for age as indicated by the solid line. Each point represents an individual patient in (D) and (E). (F) Levels of IL-5, IL-13, CCL2 (MCP-1), and CCL5 (RANTES; pg/ml) in plasma from patients with LDS (n = 24) and age-matched nonallergic controls (n = 16). Significant P values are indicated; comparisons were done by Wilcoxon test.

Fig. 2. Frequency and function of Foxp3+ cells in patients with LDS and nonsyndromic allergic disease.

(A) Number of total Tregs (CD4+CD25+CD127lo) in individuals with LDS (n = 16) and nonsyndromic allergic disease (n = 26) compared to age-matched nonallergic controls (n = 8). Further analysis revealed increases in rTregs and CD45RA−Foxp3inter cells, but no difference in the frequency of aTregs, in LDS (n = 16) and nonsyndromic allergic subjects (n = 26) compared to nonallergic controls (n = 8). (B to D) Percentage of CD4+ lymphocytes as well as each Treg subset that expresses the cytokines IL-13 (B), IFN-γ (C), and IL-17 (D) in LDS patients (n = 7), children with non-syndromic allergic disease (n = 5), and nonallergic controls (n = 5). Significant P values are indicated; all comparisons were done by Wilcoxon test.

Fig. 3. Suppressive activity of LDS Tregs.

(A and B) The three subsets of Tregs (rTregs, CD45RA−Foxp3inter Tregs, and aTregs) were purified and cocultured at various ratios (1:1, 2:1, and 5:1) with CD4+ effector T cells that had been labeled with CellTrace Violet. Cultures were stimulated with anti-CD3 and irradiated antigen-presenting cells, and dilution of the dye, a marker of proliferation, was assessed 4 days later. (A) Results from a representative experiment (n = 3). (B) Percent suppression of effector T cell proliferation at the different ratios (median and range are indicated). LDS (n = 3) and nonallergic (NA) control (n = 3) Tregs effectively suppress effector T cell proliferation.

Fig. 4. Expression of Treg markers by Foxp3+ cells from patients with LDS, nonsyndromic allergic disease, and nonallergic controls.

(A) Gating scheme used to identify the three distinct subsets of Foxp3+ cells. Peripheral blood mononuclear cells (PBMCs) from LDS patients (n = 8), age-matched subjects with nonsyndromic allergic disease (n = 5), and nonallergic controls (n = 5) were stained with CD4, CD25, and CD127 antibodies. Tregs were defined as CD25+CD127lo cells after first gating on lymphocytes [based on forward scatter (FSC) and side scatter (SSC)] that were CD4+. CD4+CD25+CD127lo cells were then divided into rTregs, CD45RA−Foxp3inter Tregs, and aTregs based on their expression of CD45RA and Foxp3. (B to E) Percentage of each Treg subset that expresses the indicated Treg marker (CTLA-4, ICOS, and GITR). Boxes define the 25 and 75% quartiles, divisions within the boxes the medians, and whiskers the range. No difference in expression of any Treg marker was found except that a significantly greater percentage of rTregs and CD45RA−Foxp3inter Tregs from LDS patients expressed intracellular CTLA-4 compared to individuals with nonsyndromic allergic disease and nonallergic controls. Significant P values are indicated; comparison by Wilcoxon test.

Fig. 5. Skewing potential of naïve CD4+ lymphocytes from LDS subjects and nonallergic controls after TGFβ stimulation.

(A) Representative dot plot depicting expression of Foxp3 and IL-13 in LDS (n = 7) and nonallergic (NA; n = 4) CD4+ lymphocytes 4 days after naïve CD4+ T cells were cultured with increasing doses of recombinant TGFβ1 as indicated. (B) Percentage of IL-13–expressing Foxp3+ and Foxp3− cells in cultures from LDS patients (n = 7) compared to nonallergic controls (n = 4) after treatment with the indicated doses of recombinant TGFβ1. Significant P values are indicated; comparisons by Mann-Whitney test.

Fig. 6. Status of TGFβ signaling in the immune system of LDS patients.

(A) Immunostaining for pSmad2 in thymic tissue from three patients with LDS and age-matched controls (n = 3) with increased intensity of nuclear pSmad2 in LDS patients. (B) Percentage of nuclei that stained positively for pSmad2 in LDS versus control thymi from (A) as evaluated by a blinded observer. P < 0.001, Fisher’s exact test. (C) Expression of pSmad2/3 before (No TGFβ1) and at various time points after (+ TGFβ1) treatment of whole blood with recombinant TGFβ1 from patients with LDS (n = 4) and unaffected relatives (n = 4; Controls), and LDS patients receiving therapeutic doses of losartan (n = 4; LDS on losartan). Mean fluorescence intensities (MFI) of pSmad2/3 staining in gated CD4+ lymphocytes are plotted. Levels of pSmad2/3 were increased (P = 0.050; comparison by longitudinal analysis) after stimulation in LDS versus controls and LDS on losartan. Medians and IQRs (whiskers) are shown.