A STAT3 palmitoylation cycle promotes TH17 differentiation and colitis

Abstract

Cysteine palmitoylation (S-palmitoylation) is a reversible post-translational modification that is installed by the DHHC family of palmitoyltransferases and is reversed by several acyl protein thioesterases^1,2. Although thousands of human proteins are known to undergo S-palmitoylation, how this modification is regulated to modulate specific biological functions is poorly understood. Here we report that the key T helper 17 (T_H17) cell differentiation stimulator, STAT3^3,4, is subject to reversible S-palmitoylation on cysteine 108. DHHC7 palmitoylates STAT3 and promotes its membrane recruitment and phosphorylation. Acyl protein thioesterase 2 (APT2, also known as LYPLA2) depalmitoylates phosphorylated STAT3 (p-STAT3) and enables it to translocate to the nucleus. This palmitoylation-depalmitoylation cycle enhances STAT3 activation and promotes T_H17 cell differentiation; perturbation of either palmitoylation or depalmitoylation negatively affects T_H17 cell differentiation. Overactivation of T_H17 cells is associated with several inflammatory diseases, including inflammatory bowel disease (IBD). In a mouse model, pharmacological inhibition of APT2 or knockout of Zdhhc7-which encodes DHHC7-relieves the symptoms of IBD. Our study reveals not only a potential therapeutic strategy for the treatment of IBD but also a model through which S-palmitoylation regulates cell signalling, which might be broadly applicable for understanding the signalling functions of numerous S-palmitoylation events.

Extended Data Fig. 1 |. DHHC7 and DHHC3 are the palmitoyltransferases for STAT3.

a, b, HEK293T cells were transfected with Flag–STAT3 and HA–DHHC1–23 plasmid to screen for DHHCs that could increase STAT3 palmitoylation. The palmitoylation level was detected using Alk14 metabolic labelling, click chemistry to install a fluorescent tag, and in-gel fluorescence. The DHHC proteins analysed were from mice, and the protein number corresponds to the gene number with the exception of DHHC10/Zdhhc11, DHHC11/Zdhhc23, DHHC13/Zdhhc24, DHHC22/Zdhhc13 and DHHC24/Zdhhc22 (ref. ). b shows multiple replicates. c, Quantification of the relative palmitoylation levels in a and b. The palmitoylation level was normalized by STAT3 protein level and the level of negative control (with Alk14 but without DHHC overexpression) was set to 1. Quantification data are expressed as mean ± s.e.m. **P < 0.01.

Extended Data Fig. 2 |. DHHC7 is the major palmitoyltransferase for STAT3.

a, Endogenous STAT3 palmitoylation in wild-type and DHHC7-knockout HEK293T cells determined using Alk14 labelling. b, Wild-type (WT) and DHHC7-knockout (MUT) HEK293T cells were transfected with DHHC7 or DHHS7 and labelled with Alk14. The palmitoylation level of overexpressed STAT3 was detected by in-gel fluorescence (left). The relative palmitoylation level was quantified (right). c, d, Endogenous STAT3 palmitoylation in wild-type and DHHC7-knockout HEK293T cells (c) and splenocytes from the DSS-induced colitis mice (d) determined using Alk14 labelling. The palmitoylation level of STAT3 was detected by in-gel fluorescence (left) and quantified (right). e, ABE method confirming DHHC7 as the major enzyme that promotes STAT3 palmitoylation. HEK293T cells were transfected with Flag–STAT3 and different HA–DHHC (mouse) plasmids. The pulled-down STAT3 with or without NH₂OH treatment was detected by western blot. f, HEK293T cells were transfected with HA-tagged STAT3 and Flag-tagged human DHHC3/7/19 to confirm that human DHHC7 could increase STAT3 palmitoylation. Palmitoylation level was detected using Alk14 metabolic labelling, click chemistry to install a fluorescent tag and in-gel fluorescence. g, DHHC7 expression does not affect STAT1 palmitoylation. HEK293T cells were transfected with Flag–STAT1 and different HA–DHHC7 plasmids. The palmitoylation level of STAT1 was detected using ABE. Quantification data are mean ± s.e.m. **P < 0.01.

Extended Data Fig. 3 |. STAT3 is palmitoylated at Cys108.

a, HEK293T cells were transfected with HA–DHHC7 and different Flag–STAT3 mutants and labelled with Alk14. The palmitoylation levels of immunoprecipitated STAT3 were visualized by in-gel fluorescence. Co-immunoprecipitated DHHC7-HA was blotted. b, The mutant STAT3(C108S) could not be palmitoylated by DHHC7. Wild-type and DHHC7-knockout HEK293T cells were transfected with the indicated plasmids and treated with Alk14. The palmitoylation level of immunoprecipitated STAT3(C108S) was visualized by in-gel fluorescence. c, Wild-type and DHHC7-knockout mouse splenocytes were transfected with different Flag–STAT3 mutants and labelled with Alk14. The palmitoylation levels of immunoprecipitated STAT3 were visualized by in-gel fluorescence. d, Palmitoylation of wild-type STAT3 and mutants detected using the ABE method. HEK293T cells were transfected with HA–DHHC7 and different Flag–STAT3 mutants. e, DHHC3 expression can increase STAT3 C108 palmitoylation. HEK293T cells were transfected with Flag–STAT3 and HA–DHHC3 plasmids and labelled with Alk14. The fatty acylation level of STAT3 was detected by in-gel fluorescence.

Extended Data Fig. 4 |. STAT3 C108 palmitoylation promotes STAT3 membrane localization.

a, The subcellular localization of EGFP-tagged STAT3 in wild-type and DHHC7-knockout HEK293T cells (left) was examined by confocal imaging. The percentage of cells with STAT3 plasma membrane and endomembrane was quantified (right). Scale bars, 50 μm. b, DHHC7 knockout decreases STAT3 membrane localization and increases its nuclear localization. Subcellular fractionation was carried out using wild-type and DHHC7-knockout splenocytes. Equivalent amounts of the nuclear and membrane fractions were then analysed by western blots (left). The relative STAT3 levels were quantified (right). Mem, membrane fraction; cyto, cytoplasmic fraction; nuc, nuclear fraction. c, STAT3(C108S) mutation decreases STAT3 membrane localization and increases nuclear localization. DHHC7-knockout HEK293T cells were transfected with HA–DHHC7 and the indicated Flag–STAT3 constructs and subcellular fractionation was performed. Equivalent amounts of the nuclear and membrane fractions were then analysed by western blot (left). The relative STAT3 levels were quantified (right). d, Confocal imaging showing the subcellular localization of EGFP-STAT3 (wild type) and EGFP-STAT3(C108S) in DHHC7-knockout HEK293T cells ectopically expressing DHHC7. Scale bars, 100 μm. e, STAT3 palmitoylation in different subcellular fractions of wild-type and DHHC7-knockout HEK293T cells. The cells were transfected with Flag–STAT3 and labelled with Alk14. Cell fractionation was performed and the protein level of STAT3 was readjusted using Coomassie Brilliant Blue to ensure the equal loading of STAT3 in wild-type and knockout cell fractions for the gel analysis. The palmitoylation levels of immunoprecipitated STAT3 in the membrane, cytoplasmic and nuclear fractions were visualized by in-gel fluorescence. Quantification data are expressed as mean ± s.e.m. **P < 0.01.

Extended Data Fig. 5 |. DHHC7 and DHHC3 promotes STAT3 phosphorylation.

a, Phosphorylation levels of Flag–STAT3 in HEK293T cells expressing HA-tagged mouse DHHC1–23. STAT3 was pulled down with Flag beads and subjected to western blot analyses. b, Quantification of the relative phosphorylation levels in a. The phosphorylation level was normalized to the STAT3 protein level and the level of the negative control was set to 1. Quantification data are expressed as mean ± s.e.m. **P < 0.01.

Extended Data Fig. 6 |. Palmitoylation promotes STAT3 phosphorylation.

a, DHHC3 and DHHC7 expression increases endogenous STAT3 phosphorylation in HEK293T cells. HEK293T cells were transfected with HA–DHHC3 and HA–DHHC7. The cells were lysed and subjected to western blot analyses as indicated. b, DHHC7 and DHHC3 expression in mouse splenocytes increased endogenous STAT3 phosphorylation. Mouse splenocytes were transfected with different HA–DHHCs and catalytically inactive DHHS7. The cells were lysed and subjected to western blot analyses as indicated. c, DHHC7 knockout in HEK293T cells decreased endogenous STAT3 phosphorylation. Cells were treated with Alk14, lysed, and STAT3 was immunoprecipitated with STAT3 antibody and protein A/G magnetic beads. Immunoprecipitated samples were analysed by western blot. d, The phosphorylation of STAT3(C108S) could not be promoted by DHHC7. Wild-type and DHHC7-knockout cells were transfected with indicated HA–DHHC7 and Flag–STAT3 plasmids. STAT3 was immunoprecipitated with Flag beads and analysed by western blot. e, Distribution of STAT3 and p-STAT3 in subcellular fractions of DHHC7-knockout HEK293T cells reintroduced with DHHC7 or catalytically inactive DHHS7 (left). The relative p-STAT3 levels were quantified (right). f, The subcellular localization of STAT3 and p-STAT3 was analysed using confocal imaging after EGFP–STAT3 constructs and HA-tagged DHHC7 were transfected into DHHC7-knockout HEK293T cells. Co-localization of STAT3 and DHHC7 was analysed using Pearson’s coefficient. The level of p-STAT3 was quantified. Scale bars, 100 μm. Quantification data are mean ± s.e.m. *P < 0.05; **P < 0.01.

Extended Data Fig. 7 |. APT2 depalmitoylates STAT3 and promotes the nuclear translation of p-STAT3.

a, Flag–APT2 pulled down HA–STAT3 when co-expressed in HEK293T cells. The interaction of HA–STAT3 with Flag–APT2(C2S) was much weaker. b, Wild-type APT2 could remove the DHHC7-introduced palmitoylation on wild-type STAT3 better than that on the Y705F mutant. DHHC7-knockout HEK293T cells were transfected with the indicated plasmids. The cells were labelled with Alk14 and the palmitoylation of STAT3 was determined by in-gel fluorescence (left). The relative palmitoylation levels were quantified (right). c, In HEK293T cells, overexpression of wild-type APT2—but not the C2S or S122A mutant—decreased the palmitoylation level of Flag–STAT3 as determined by ABE. ML349 treatment (20 μM for 36 h) also increased the palmitoylation level. d, APT2 inhibition or knockdown increases Flag–STAT3 palmitoylation. DHHC7-knockout HEK293T cells were reintroduced with HA–DHHC7 and Flag–STAT3. Cells were then treated with siLYPLA2 or 20 μM of ML349 for 36 h before Alk14 labelling and in-gel fluorescence detection. e, LYPLA2 siRNA knockdown in HEK293T cells decreased the mRNA levels of RORC and CCND1, two STAT3 target genes. f, Overexpression of Flag–APT2 (wild type), but not mutants, increased RORC mRNA levels in HEK293T cells. g, Co-expression of wild-type DHHC7 and APT2 strongly activated STAT3 signalling. DHHC7-knockout HEK293T cells were transfected with Flag–STAT3, different HA–DHHC7, and Flag–APT2 plasmids as indicated. The total mRNA was extracted and subjected to qPCR analysis of the indicated mRNA. h, STAT3 palmitoylation does not affect its dimerization. HEK293T cells were transfected with HA-tagged DHHC7, HA-tagged STAT3, and Flag-tagged STAT3. The level of Flag and HA-tagged STAT3 was determined after Flag-tag immunoprecipitation. i, Flag–APT2 interacted with HA–STAT3 (wild-type) better than it did with the Y705F mutant in HEK293T cells. Quantification data are mean ± s.e.m. *P < 0.05; **P < 0.01.

Extended Data Fig. 8 |. Expression of wild-type DHHC7 and APT2 in mouse splenocytes promotes T_H17 cell differentiation.

a, Co-expression of DHHC7 and STAT3 in splenocytes promotes T_H17 cell differentiation. Mouse splenocytes were transfected with different DHHC7 and STAT3 plasmids as indicated and then treated with a cytokine cocktail for 4 days to initiate differentiation. The Rorc and Il17a mRNA levels, markers for T_H17 differentiation, were analysed using qPCR. b, Splenocytes were transfected with EGFP-STAT3 plasmids and observed with confocal imaging. Scale bars, 100 μm. c, mRNA levels in splenocytes that were treated with a cytokine cocktail for 4 days to initiate differentiation. d, ZDHHC7 mRNA levels were measured in splenocytes transfected with different DHHC7 and DHHS7 plasmids as indicated. e, IL17A mRNA levels were measured in splenocytes transfected with different plasmids as indicated. f, Mouse splenocytes were treated with a cytokine cocktail and transfected with different plasmids as indicated for 4 days, then the cells were collected and analysed by flow cytometry to detect CD4 and IL-17 positive cells (T_H17 cells). The cytokine cocktail contains 3 ng ml⁻¹ TGF-β, 40 ng ml⁻¹ IL-6, 30 ng ml⁻¹ IL-23, 20 ng ml⁻¹ TNF and 10 ng ml⁻¹ IL-1β. g, RORC and CCND1 mRNA levels in HEK293T cells treated with the indicated concentrations of ML349 for 36 h. Quantification data are mean ± s.e.m. *P < 0.05; **P < 0.01.

Extended Data Fig. 9 |. Human IBD patient gene expression data.

a, b, Human peripheral blood mononuclear cells (PBMCs) from 26 healthy participants, 24 patients with Crohn’s disease (CD; 7 remission stage and 17 active stage) and 10 patients with ulcerative colitis (UC; 1 remission stage and 9 active stage) were extracted. The levels of ZDHHC3 (we analysed ZDHHC3 because DHHC3 also showed some activity towards STAT3), RORC and IL17A mRNA were analysed using qPCR. c, The correlation between IL17A and indicated mRNA levels in 26 healthy participants were analysed. d, The correlation between p-STAT3 and indicated mRNA levels in 26 healthy participants were analysed. e, The correlation between ZDHHC3 and IL17A mRNA levels in 26 healthy participants and 34 patients with IBD (24 CD and 10 UC) were analysed. f, The correlation between RORC and IL17A mRNA levels in 26 healthy participants and 34 patients with IBD (24 CD and 10 UC) were analysed. g, The correlation between LYPLA2 and RORC mRNA levels in 26 healthy participants and 34 patients with IBD (24 CD and 10 UC) were analysed. h, The correlation between ZDHHC7 and RORC mRNA levels in 26 healthy participants and 34 patients with IBD (24 CD and 10 UC) were analysed. Quantification data are mean ± s.e.m. *P < 0.05; **P < 0.01.

Extended Data Fig. 10 |. Mouse DSS model data.

a, ML349 toxicity study. C57BL/6J mice were intraperitoneally injected with ML349 for 7 days at the indicated doses and changes in body weight were evaluated. b, c, C57BL/6J mice were pretreated with daily ML349 intraperitoneal injections as indicated, then the ML349 treatment was combined with 2.5% DSS treatment in their drinking water ad libitum. Survival rate (b) and body weight changes (c) were evaluated. d, C57BL/6J mice were treated with 2.5% DSS treatment in their drinking water ad libitum and ML349 was injected daily intraperitoneally on the day that DSS treatment started. Colon morphology was evaluated at the end of the treatment. e, Quantitation of colon length shown in d. f, Wild-type and DHHC7-knockout mice were treated with 2.5% DSS in their drinking water ad libitum. Colon length was evaluated. Quantification data are mean ± s.e.m. **P < 0.01.

Fig. 1 |. DHHC7-induced palmitoylation promotes STAT3 membrane translocation.

a, HEK293T cells were transfected with Flag–STAT3 and HA–DHHC7. Palmitoylation levels of STAT3 with or without hydroxylamine (NH₂OH) treatment were detected using Alk14 labelling. b, Left, the subcellular localization of endogenous STAT3 and JAK2 was analysed using confocal imaging in wild-type and DHHC7-knockout HEK293T cells. Scale bars, 50 μm. Right, quantification of the colocalization of STAT3 and JAK2 using Pearson’s correlation coefficients. c, Left, the subcellular localization of EGFP–STAT3 and EGFP–STAT3(C108S) in DHHC7 knockout HEK293T cells ectopically expressing DHHC7. Scale bars, 100 μm. Right, the percentage of DHHC7-positive cells in which STAT3 is translocated from the nucleus to the plasma membranes and endomembranes. d, Wild-type and DHHC7-knockout HEK293T cells were transfected with Flag–STAT3 and labelled with Alk14. Subcellular fractionation was performed and STAT3 protein levels were adjusted to ensure that there were equal amounts of STAT3 in the wild-type and knockout cell fractions used for the gel. The palmitoylation levels of immunoprecipitated STAT3 in the membrane (mem.), cytoplasmic (cyto.) and nuclear (nuc.) fractions were visualized by in-gel fluorescence. Data are mean ± s.e.m. **P < 0.01. Raw data gels can be found in Supplementary Fig. 1 and statistical and reproducibility information can be found in the Supplementary Information.

Fig. 2 |. APT2 is a depalmitoylase of STAT3 and palmitoylation/depalmitoylation promotes p-STAT3 nuclear localization.

a, C108 is the major palmitoylation site of STAT3. HEK293T cells were transfected with HA–DHHC7 and Flag–STAT3 (wild-type or mutants) and labelled with Alk14. STAT3 was pulled down and subjected to Alk14 labelling and western blot analyses. b, Left, distribution of wild-type or C108S mutants of STAT3 and p-STAT3 in the subcellular fractions of DHHC7-knockout HEK293T cells into which DHHC7 or DHHS7 were reintroduced. Right, quantification of the relative p-STAT3 levels. c, The subcellular localization of STAT3 and p-STAT3 was analysed using confocal imaging after different EGFP–STAT3 constructs and HA-tagged DHHC7 were transfected into DHHC7-knockout HEK293T cells. Scale bars, 50 μm. d, In HEK293T cells, overexpression of wild-type APT2—but not the C2S or S122A mutants—decreased the palmitoylation level of Flag–STAT3 as determined by Alk14 labelling. e, APT2 preferentially depalmitoylates wild-type STAT3 over the Y705F mutant. DHHC7-knockout HEK293T cells were transfected with the indicated plasmids and labelled with Alk14. The palmitoylation of STAT3 was determined by in-gel fluorescence (left) and quantified (right). f, APT2 inhibition or knockdown increases the palmitoylation of Flag–STAT3. DHHC7-knockout HEK293T cells were reintroduced with HA–DHHC7 and Flag–STAT3, and treated with LYPLA2 siRNA or 20 μM of ML349 for 36 h before Alk14 labelling and in-gel fluorescence detection. g, Left, distribution of STAT3 and p-STAT3 in different subcellular fractions of APT2-knockdown HEK293T cells. Right, quantification of the relative STAT3 and p-STAT3 levels in these fractions. h, Left, distribution of STAT3 and p-STAT3 in subcellular fractions of DHHC7-overexpressing HEK293T cells, with or without treatment with 1 μM of the JAK2 inhibitor fedratinib. Right, quantification of the relative STAT3 levels in these fractions. Data are mean ± s.e.m. *P < 0.05; **P < 0.01.

Fig. 3 |. The STAT3 palmitoylation–depalmitoylation cycle promotes T_H17 cell differentiation.

a, DHHC7 promotes phosphorylation of wild-type STAT3, but not of the C108S mutant, in mouse splenocytes. Left, STAT3 and p-STAT3 blots; right, quantification of the relative p-STAT3 levels. b, T_H17 cell differentiation was quantified in the splenocyte samples in a by flow cytometry. c, APT2 inhibition decreases T_H17 cell differentiation in a dose-dependent manner. Mouse splenocytes were treated with a cytokine cocktail (3 ng ml⁻¹ TGF-β, 40 ng ml⁻¹ IL-6, 30 ng ml⁻¹ IL-23, 20 ng ml⁻¹ TNF and 10 ng ml⁻¹ IL-1β) and different concentrations of ML349 for 4 days, then collected and analysed by flow cytometry to detect CD4- and IL-17-positive cells. d, DHHC7 knockout in splenocytes inhibits T_H17 cell differentiation. Wild-type and DHHC7-knockout mouse splenocytes were treated with a cytokine cocktail for 4 days to initiate differentiation, then the cells were collected and analyse by flow cytometry to detect CD4- and IL-17-positive cells. e, Left, STAT3 and p-STAT3 blots of wild-type and DHHC7-knockout splenocytes. Right, quantification of the relative p-STAT3 levels. Data are mean ± s.e.m. **P < 0.01.

Fig. 4 |. The STAT3 palmitoylation–depalmitoylation cycle aggravates colitis.

a, b, Human PBMCs from 26 healthy participants (Ctrl), 24 patients with Crohn’s disease (CD; 7 remission stage (rem.) and 17 active stage (act.)) and 10 patients with ulcerative colitis (UC; 1 remission stage and 9 active stage) were extracted. LYPLA2 and ZDHHC7 mRNA levels were analysed using qPCR. The relative p-STAT3 levels were quantified by western blots. c, d, The correlations between IL17A (c) or p-STAT3 (d) and the mRNA levels of the indicated genes in 34 patients with IBD. e, f, C57BL/6J mice were treated with 2.5% DSS in drinking water ad libitum and ML349 was intraperitoneally injected daily on the day that DSS treatment started. Body weight changes (e) and T_H17 cell levels in the spleen (f) were evaluated. g, h, Wild-type and DHHC7-knockout mice were treated with 2.5% DSS in drinking water ad libitum. Body weight changes (g) and T_H17 cell levels in the spleen (h) were evaluated. i, Model of the regulation of STAT3 by the palmitoylation–depalmitoylation cycle. Palmitoylation of STAT3 by DHHC7 promotes the membrane recruitment and phosphorylation of STAT3. APT2 promotes the nuclear translocation of p-STAT3 by selectively depalmitoylating p-STAT3 over STAT3. The palmitoylation–depalmitoylation cycle drives the membrane localization and phosphorylation of STAT3, and the nuclear translocation of p-STAT3. The direction of the cycle is ensured by the preference of APT2 for p-STAT3 over STAT3. Data are expressed as the mean ± s.e.m. *P < 0.05; **P < 0.01.