TSLP production by dendritic cells is modulated by IL‐1β and components of the endoplasmic reticulum stress response

Thymic stromal lymphopoietin (TSLP) produced by epithelial cells acts on dendritic cells (DCs) to drive differentiation of TH2‐cells, and is therefore important in allergic disease pathogenesis. However, DCs themselves make significant amounts of TSLP in response to microbial products, but little is known about the key downstream signals that induce and modulate this TSLP secretion from human DCs. We show that human monocyte derived DC (mDC) secretion of TSLP in response to Candida albicans and β‐glucans requires dectin‐1, Syk, NF‐κB, and p38 MAPK signaling. In addition, TSLP production by mDCs is greatly enhanced by IL‐1β, but not TNF‐α, in contrast to epithelial cells. Furthermore, TSLP secretion is significantly increased by signals emanating from the endoplasmic reticulum (ER) stress response, specifically the unfolded protein response sensors, inositol‐requiring transmembrane kinase/endonuclease 1 and protein kinase R‐like ER kinase, which are activated by dectin‐1 stimulation. Thus, TSLP production by mDCs requires the integration of signals from dectin‐1, the IL‐1 receptor, and ER stress signaling pathways. Autocrine TSLP production is likely to play a role in mDC‐controlled immune responses at sites removed from epithelial cell production of the cytokine, such as lymphoid tissue.


TSLP production by dendritic cells is modulated by
. β-glucan stimulated mDC induce Syk phosphorylation, IκBα degradation and p38 MAPK phosphorylation. (A) mDC were pre-incubated for one hour with or without the Syk inhibitor and were then stimulated with CUR for 4 hours. (B) mDC were pre-incubated for one hour with or without a dectin-1 neutralising antibody or isotype control and were then stimulated with peptidoglycan (PGN) for 24 hours (n=2). (C) mDC were pre-incubated for one hour with or without NFκB inhibitor and were then stimulated with CUR for 15 minutes and analysed for IκBα degradation.
(D) mDC were stimulated with CUR for 4 hours and analysed for p38 MAPK phosphorylation. (E) mDC were pre-incubated for one hour with or without a p38 MAPK inhibitor and were then stimulated with CUR for 24 hours (n=3). IκBα degradation and Syk and p38 MAPK phosphorylation were measured by immunoblot. IL-1β and IL-23 were measured in 24 hour cell culture supernatants by ELISA. Cumulative data are shown as mean +SEM.

IκBα β-actin
Supplemental Figure 2. Functionality of the anti-TNFα antibody utilised in this study is demonstrated by its ability to inhibit TNFα induced IκBα degradation. mDC were pre-incubated for one hour with or without anti-TNFα and were then stimulated with TNFα for 15 minutes. IκBα degradation was measured by immunoblot.

A B
Supplemental Figure 4. Syk, NFκB and p38 MAPK are required for IL-1β and TSLP production. (A-C) mDC were pre-incubated for one hour with or without Syk inhibitor (n=5), NFκB inhibitor (n=5) or p38 MAPK inhibitor (n=7) and were then stimulated with CUR, MP or C. albicans for 24 hours. (D-F) mDC were pre-incubated for one hour with or without the Syk, NFκB or p38 MAPK inhibitors and were then stimulated with CUR, MP or C. albicans with or without recombinant IL-1β for 24 hours (n=3). TSLP and IL-1β were measured in 24 hour cell culture supernatants by ELISA. Cumulative data are shown as mean +SEM. (A-C) Statistical significance calculated using t test. p values for IL-1β production by CUR stimulated mDC with Syk-i, NFκB-i and p38 MAPK-i respectively (* p=0.0127, ** p=0.0072 and ** p=0.0016), for MP stimulated mDC (* p=0.0491, * p=0.0488 and ** p=0.0036) and for C. albicans stimulated mDC (* p=0.0197, ** p=0.0073 and * p=0.0146). (D-F) Statistical significance was calculated using two-way ANOVA with Bonferroni post-tests. (*** p=0.001, ** p=0.01). Supplemental Figure 5. Inhibitors of IRE1α and PERK inhibit XBP1 splicing and eIF2α phosphorylation respectively, but do not affect DC activation markers. (A-B) mDC were pre-incubated for one hour with or without IRE1α or PERK inhibitors and were then stimulated with CUR or TM for 4 hours, and analysed for XBP1 splicing (n=3, representative experiment shown) and phospho eIF2α (n=2, representative experiment shown) analysed as functional readouts for the efficacy of inhibitors.
(C-D) mDC were pre-incubated for one hour with or without PKR inhibitor and were then stimulated with CUR for (C) 4 hours for PKR phosphorylation 4 hours or (D) 24 hours for TSLP secretion (n=2).
(E-G) mDC were pre-incubated for 1 hour with or without IRE1α or PERK inhibitors and were then stimulated with CUR for 4 hours analysed for OPG and IL-1β mRNA expression and for CD83, CD86 and HLA-DR surface expression by flow cytometry. Unstimulated control (grey filled profile), CUR stimulated (black line profile), CUR plus IRE1α or PERK inhibitors (grey line profile). XBP1, OPG and IL-1β mRNA expression were measured by qRT-PCR. PKR phosphorylation was measured by immunoblot. TSLP was measured in 24 hour cell culture supernatants by ELISA. Cumulative data are shown as mean +SEM. Statistical significance calculated using one-way ANOVA with Bonferroni posttests (ns = not significant, *** p=0.001, ** p=0.01, * p=0.05).  Figure 6. The absence of CHOP does not affect dectin-1 induced TSLP expression. (A-B) BMDC from C57/B6 wild-type and CHOP KO mice were stimulated with CUR and MP for 4 hours (n=4). TSLP and IL-23p19 mRNA expression were measured by qRT-PCR. Cumulative data are shown as mean +SEM. Statistical significance was calculated using t test. p values for IL-23p19 expression by CUR stimulated BMDC (** p=0.0029), and for MP stimulated BMDC (* p=0.0143).