Alternative splicing contributes to transcriptomic and proteomic diversity. This process is tightly regulated in different tissues, cell types and differentiation stages, but its dysregulation can drive tumorigenesis1,2 or immune-related diseases3. Atopic dermatitis (AD) and psoriasis are clinically independent inflammatory skin diseases4. AD is strongly driven by helper type 2 T cells (TH2 cells) and is associated with IL-4 and IL-13 overproduction, whereas psoriasis is largely driven by TH17 cells and associated with IL-17 activation5. Although an abnormal keratinocyte response to T cell-derived cytokines is intrinsic to the pathology of AD and psoriasis, whether they share common mechanisms that regulate keratinocyte inflammation remains unclear.
Deep RNA-sequencing (RNA-seq) and high-throughput, genome-wide transcriptome analyses of skin from healthy controls and patients with AD or psoriasis show that AD and psoriasis have a similar profile of alternatively spliced transcripts in skin lesions6,7,8. Analysis of splicing signatures indicates that multiple candidate-splicing factors, including HNRNPA1, U2AF1 and DDX5, might be responsible for RNA-splicing changes in AD and psoriasis6,7. Among these splicing factors, the DEAD (AspGluAlaAsp) box (DDX), RNA helicase DDX5 regulates diverse aspects of RNA biogenesis by unwinding or destabilizing the stem–loop structure of genes to facilitate splicing factors binding to splice sites9 and has been associated with disease in humans10. Downregulation or depletion of DDX5 leads to male infertility by changing the alternative splicing pattern of multiple genes during spermatogenesis11, whereas high expression of DDX5 drives tumor development by modulating the alternative splicing of H-Ras or mH2A1 in breast cancer12. DDX5 also facilitates vesicular stomatitis virus propagation by promoting RNA decay of antiviral transcripts and nuclear export of transcripts DHX58, p65 and IKKγ in mouse embryo fibroblasts13. DDX5 has been predicted to be a shared biological factor against viral replication in inverse, erythrodermic and chronic plaque psoriasis14 and genetic variants at the DDX5 locus predispose individuals to asthma, an atopic disease associated with AD15,16.
The response of keratinocytes to the IL-17 family of cytokines, such as IL-17A, IL-17C, IL-17E (also named IL-25) and IL-17F, contributes to the cycle of inflammation and cellular proliferation that results in epidermal hyperproliferation and lesion formation in AD and psoriasis17,18,19. IL-17A activates the IL-17RA–IL-17RC complex or IL-17RC–IL-17RD complex to induce keratinocyte proliferation and the expression of chemokines, such as CXCL1 and CCL20, that recruit neutrophils, TH17 cells or γδ T cells into skin lesions in psoriasis17,20. IL-17E activates IL-17RB–STAT3 signaling in keratinocytes to drive skin inflammation in psoriasis18 or the IL-17RA–IL-17RB complex to induce keratinocyte differentiation and the expression of IL-4 and IL-13 in AD21. IL-17C and IL-17F also induce keratinocyte proliferation and inflammatory cytokine production in psoriasis22. IL-17D, the least understood member of IL-17 family, regulates intestinal homeostasis through its receptor, CD93 (ref. 23), but the role of IL-17D in skin inflammation is completely unknown.
To address whether dysregulation of DDX5 function or expression had a role in propagating inflammation in the skin, we analyzed the expression profile of DDX5 in AD and psoriasis and generated mice with conditional knockout of Ddx5 in keratinocytes. We show that inhibition of DDX5 expression in keratinocytes downstream of IL-17D signaling amplified skin inflammation by controlling the pre-mRNA splicing of IL-36R, which in turn increased IL-36R expression and amplified the IL-36R-mediated skin inflammation.
DDX5 is reduced in keratinocytes and drives skin inflammation
Gene ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of overlapping differentially expressed genes (DEGs) from a publicly available RNA-seq dataset (Gene Expression Omnibus (GEO) database: accession no. GSE121212) from healthy controls and patients with AD or psoriasis6 showed a similar RNA-splicing pattern, based on the presence of shared spliceosome complexes in the skin lesions of patients with AD and psoriasis (Extended Data Fig. 1a). Based on quantitative reverse transcription PCR (RT–qPCR), the mRNA for DDX5, a component of spliceosome complexes9, but not all other tested splicing factors, including CDC5L, PRPF18, DHX8, DHX15, HNRNPA1 and SRSF1, was significantly decreased in whole skin lesions from patients with AD and psoriasis compared with healthy controls (Fig. 1a and Extended Data Fig. 1b). Based on immunoblotting and immunofluorescence staining, expression of DDX5 protein was reduced in whole skin lesions from patients with AD and psoriasis (Fig. 1b,c) and slightly increased in epidermal tissues from skin carcinomas (Extended Data Fig. 1c), compared with healthy skin. The mRNA and protein expression of Ddx5 were progressively reduced in whole lesional skin from wild-type mice treated with MC903 (calcipotriol), a vitamin D3 analog, by painting the ears for 15 d consecutively or placing ovalbumin (OVA) patches on dorsal skin for 7 d consecutively after two intraperitoneal injections of OVA at 1-week intervals to induce AD, or imiquimod (IMQ), a toll-like receptor 7 agonist, painting on dorsal skin for 5 d consecutively to induce psoriasis, compared with skin from untreated wild-type mice (Fig. 1d–f and Extended Data Fig. 1d,e). Reanalysis of single-cell RNA-seq (scRNA-seq) datasets of skin from healthy adults and patients with AD and psoriasis24 indicated that DDX5 mRNA was downregulated in keratinocytes, fibroblasts and lymphocytes from lesions or nonlesional skin in patients with AD and psoriasis (Fig. 1g). Immunofluorescence staining indicated that DDX5 protein was highly expressed in epidermal keratinocytes from healthy skin of humans and mice (Fig. 1h–k), but was almost undetectable in whole lesional skin from patients with AD and psoriasis and MC903- or IMQ-treated mice (Fig. 1h–k), indicating a reduction in keratinocyte DDX5 in inflamed skin in humans and mice.
To determine the role of keratinocyte DDX5 in skin inflammation, we generated mice with a specific ablation of Ddx5 in keratinocytes by crossing Ddx5fl/fl mice with K14Cre transgenic mice, hereafter referred to as Ddx5∆KC mice. Compared with Ddx5fl/fl littermates, Ddx5∆KC mice had greatly thickened scaling and more patches and earlier (day 16 compared with day 8 postinitiation of MC903) emergence of dryness and spongiosis on the ear skin (Extended Data Fig. 2a,b), which resembled clinical features in patients with AD25. The expression of Il4, Il13, Ccl11, Ccl17 and Ccl22 mRNA was increased two- to fourfold (Fig. 1l), whereas the expression of Tslp mRNA was not changed (Fig. 1l), in the skin lesions of Ddx5∆KC mice compared with Ddx5fl/fl mice at day 16 post-MC903 administration. The percentages and absolute numbers of CD45+ immune cells, including CD11b+SiglecF+ eosinophils, CD49b+IgE+ basophils and CD3+CD4+ T cells, were increased in the ear skin of MC903-treated Ddx5∆KC mice compared with Ddx5fl/fl littermates (Fig. 1m), indicating more severe inflammation. Ddx5∆KC mice also had more thickened plaques and increased acanthosis on the dorsal skin on day 5 post-IMQ administration compared with Ddx5fl/fl littermates (Extended Data Fig. 2c,d). Moreover, the expression of Il23, Il17a, Ccl20, Cxcl1, Cxcl2 and S100a7 (Extended Data Fig. 2e), which is linked to the pathogenesis of psoriasis26, and the infiltration of CD45+ immune cells, including CD11b+Ly6G+ neutrophils, CD45+MHCII+ antigen-presenting cells (APCs) and γδTCR+ T cells (Extended Data Fig. 2f), were increased in whole skin lesions in IMQ-treated Ddx5∆KC mice compared with Ddx5fl/fl littermates. Thus, DDX5 deficiency in keratinocytes drove cutaneous inflammation.
DDX5 expression is inhibited by IL-17D in keratinocytes
To investigate the factors that inhibited DDX5 expression in keratinocytes during inflammation we treated neonatal human epidermal keratinocytes (NHEKs) with 11 inflammatory cytokines, including IL-17A, IL-17F, tumor necrosis factor (TNF), IL-1β, IL-4 and IL-36γ. Among all those tested, only IL-17D reduced the expression of DDX5 mRNA and protein in a time- and dose-dependent manner (Fig. 2a–c and Extended Data Fig. 3a,b). Production of IL-17D steadily increased in the skin of MC903- or IMQ-treated wild-type mice compared with untreated mice (Fig. 2d and Extended Data Fig. 4a). Expression of Ddx5 in whole lesional skin from MC903- or IMQ-treated Il17d–/– mice was higher compared with wild-type counterparts and was similar to that in healthy skin from wild-type mice (Fig. 2e,f and Extended Data Fig. 4b,c). MC903- or IMQ-treated Il17d–/– mice exhibited fewer plaques in the ear skin (Extended Data Fig. 4d), had reduced expression of Il4, Il5, Il13, Ccl11, Ccl17 and Ccl22 (Fig. 2g) and decreased infiltration of CD45+ immune cells, including CD11b+SiglecF+ eosinophils, CD49b+IgE+ basophils and CD3+CD4+ T cells (Fig. 2h and Extended Data Fig. 4e), in lesions compared with wild-type counterparts. These data suggested that IL-17D inhibited the expression of DDX5 in keratinocytes during AD and psoriasis.
IL-17D activates CD93-mediated signaling to inhibit DDX5
CD93, a type I transmembrane protein with an amino-terminal C-type lectin-like domain has been identified as a receptor for IL-17D23. Ddx5 was increased in the whole skin lesions of MC903- or IMQ-treated Cd93–/– mice compared with their wild-type counterparts (Fig. 3a and Extended Data Fig. 5a), along with fewer plaques on the ears or dorsal skin (Extended Data Fig. 5b) and reduced expression of Il4, Il13, Ccl11, Ccl17 and Ccl22 (Fig. 3b) or Cxcl1, Cxcl2, Ccl20, Il23, Il17a and Il36γ (Extended Data Fig. 5c). Silencing of CD93 in NHEKs by small interfering RNAs (siRNAs) targeting CD93 (Fig. 3c and Extended Data Fig. 5d), and inhibition of p38 MAPK (mitogen-activated protein kinase), AKT1/2 and SMAD2/3 with SB202190, AKT1/2 kinase inhibitor and SB431542, respectively (Fig. 3d,e), abrogated the reduction in DDX5 expression in IL-17D-stimulated NHEKs compared with dimethylsulfoxide-treated NHEK cells. Treatment with IL-17D induced the phosphorylation of p38 MAPK, AKT and SMAD2/3 in NHEKs or whole lesional skin from MC903- or IMQ-treated wild-type mice, but this effect was lost in CD93-silenced NHEKs or whole lesional skin from MC903- or IMQ-treated Cd93–/–mice (Fig. 3f,g and Extended Data Fig. 5e). Moreover, the p38 MAPK inhibitor SB202190 dampened IL-17D-induced phosphorylation of AKT and SMAD2/3, whereas the AKT1/2 kinase inhibitor and SMAD2/3 inhibitor SB431542 did not block the IL-17D-induced phosphorylation of p38 MAPK in NHEKs (Fig. 3h–j). The AKT1/2 kinase inhibitor prevented IL-17D-induced phosphorylation of SMAD2/3, but the SMAD2/3 inhibitor had no effect on the IL-17D-induced phosphorylation of AKT1/2 (Fig. 3i, j). These data indicated that IL-17D activated a p38 MAPK–AKT–SMAD2/3 signaling pathway downstream of CD93 to inhibit expression of DDX5 in keratinocytes.
DDX5 deficiency amplifies IL-36–IL-36R-mediated inflammation
AD and psoriasis share multiple drivers of epidermal hyperplasia and inflammation, such as IL-17A, TNF, IL-17E, IL-36 and IL-4 (refs. 5,18,27,28,29). To test whether DDX5 regulated the keratinocyte response to these cytokines, we treated wild-type and DDX5–/– immortalized human keratinocytes (HaCaT cell line) with IL-17A, IL-36γ, TNF, IL-25 and IL-4. The expression of multiple chemokines, including CCL20, CXCL1, CXCL2, CXCL6, CCL3, CCL17 and CCL22, was increased in DDX5–/– HaCaT cells 2 h after stimulation with IL-36γ, whereas the expression of some chemokines, such as CCL20 and CXCL2, was slightly increased in DDX5–/– HaCaT cells treated with TNF or IL-25 (Extended Data Fig. 6a–e), suggesting that DDX5 deficiency amplified an IL-36γ-mediated inflammatory response in keratinocytes.
Next, we performed high-throughput RNA-seq in wild-type and DDX5–/– HaCaT cells 4 h post-IL-36γ stimulation or in lesional skin from MC903- or IMQ-treated Ddx5∆KC and Ddx5fl/fl mice. Global transcriptional profiling showed that the expression of multiple inflammatory mediators such as IL-1β, CXCL1 and CXCL2 was increased in DDX5–/– HaCaT cells or in lesions of MC903- or IMQ-treated Ddx5∆KC mice compared with IL-36γ-stimulated, wild-type HaCaT cells or lesions from MC903- or IMQ-treated Ddx5fl/fl mice (Fig. 4a and Extended Data Fig. 6f,g). GO enrichment analysis of three RNA-seq results from IL-36γ-stimulated wild-type and DDX5–/– HaCaT cells or skin lesions from MC903- or IMQ-treated Ddx5∆KC and Ddx5fl/fl mice indicated that the DDX5-dependent changes in gene expression mostly affected leukocyte migration, positive regulation of cytokine production and cell chemotaxis (Fig. 4b). RT–qPCR confirmed that the expression of mRNA for CCL11 (3.6-fold), CCL17 (2.6-fold) and CCL22 (3.1-fold), which attract eosinophils and TH2 cells to AD skin lesions30, and the expression of mRNA for CCL20 (40.5-fold), CXCL1 (8.5-fold) and CXCL2 (51.2-fold), which recruit γδT cells and neutrophils to psoriatic lesions31, were significantly increased in DDX5–/– HaCaT cells compared with wild-type HaCaT cells in response to IL-36γ (Fig. 4c,d). The siRNAs targeting IL1RL2, which encodes IL-36R, in IL-36γ-stimulated DDX5–/– HaCaT cells decreased the expression of IL-36γ-induced CCL11, CCL17, CCL22, CCL20, CXCL1 and CXCL2 compared with IL-36γ-stimulated DDX5–/– HaCaT cells without IL1RL2 silencing (Fig. 4c,d).
IL-36γ-stimulated DDX5–/– HaCaT cells also had increased expression of the IL36R mRNA, but not the IL-36R antagonists IL36RN or IL38 (Fig. 4a,e,f and Extended Data Fig. 6h,i). An increase in IL-36R protein was also observed in NHEKs transduced with siRNA targeting DDX5 and keratinocytes isolated from Ddx5∆KC newborn mice (Fig. 4g,h). Expression of IL-36R protein was increased 4.8-fold or 1.8-fold in whole skin lesions from MC903- or IMQ-treated Ddx5∆KC mice (Fig. 4i and Extended Data Fig. 6j), but reduced by, on average, 68% in whole skin lesions from MC903- or IMQ-treated Il17d–/– mice or Cd93–/– mice (Fig. 4j,k and Extended Data Fig. 6k,l), compared with their wild-type counterparts. The expression of IL-36R was increased by 190% or 80% in the skin lesions of patients with AD or psoriasis, respectively, compared with healthy skin (Fig. 4l and Extended Data Fig. 6m). As such, decreased expression of DDX5 amplified the keratinocyte response to IL-36γ by upregulating the expression of IL-36R.
DDX5 and SF2 regulate pre-mRNA splicing to produce sIL-36R
Alternative pre-mRNA splicing is a critical step in the posttranscriptional regulation of gene expression32. Quantification of splicing events in the RNA-seq datasets from IL-36γ-stimulated wild-type and DDX5–/– HaCaT cells or skin lesions from MC903- or IMQ-treated Ddx5∆KC and Ddx5fl/fl mice indicated that DDX5 deficiency in keratinocytes changed multiple splicing events (Extended Data Fig. 7a), with a profound effect on exon skipping and the splicing pattern of the genes with skipped exons (Extended Data Fig. 7a,b). To test whether DDX5 regulated the splicing of the IL1RL2 pre-mRNA, we examined the presence of IL-36R splicing transcripts in primary human and mouse keratinocytes. Two IL36R transcripts were amplified by PCR in NHEKs and primary keratinocytes isolated from wild-type newborn mice. Sequencing of these transcripts identified a human IL36R transcript that skipped exon3 and a mouse Il36R transcript that skipped exon6 (Extended Data Fig. 7c–f). Both transcripts encoded the ectodomain of IL-36R (Extended Data Fig. 7g,h), a previously unknown soluble sIL-36R.
We next tested whether DDX5 influenced IL36R mRNA splicing to generate alternative transcripts. IL36R transcripts were increased, whereas sIL36R transcripts were decreased in DDX5–/– HaCaT cells compared with wild-type HaCaT cells, from both the endogenous gene and an IL36R reporter minigene transduced in these cells (Fig. 5a,b and Extended Data Fig. 7i). Silencing of the other nine DDX proteins in NHEKs had no effect on the expression of IL36R and sIL36R (Fig. 5b and Extended Data Fig. 7j). Murine Il36R was increased and sIl36R was reduced in keratinocytes isolated from Ddx5∆KC newborn mice compared with Ddx5fl/fl keratinocytes (Extended Data Fig. 7k). Immunoblotting with an antibody specific for sIL-36R indicated that expression of sIL-36R protein was reduced in DDX5-silenced NHEKs or Ddx5∆KC keratinocytes (Fig. 5c,d). In contrast to IL-36R, sIL-36R was gradually reduced in whole skin lesions from wild-type mice at day 8 post-MC903 administration or at day 2 post-IMQ-administration (Fig. 5e,f) and was almost undetectable in skin lesions from patients with AD or psoriasis or from MC903- and IMQ-treated Ddx5∆KC mice (Fig. 5g–j), whereas its expression in skin lesions from MC903- and IMQ-treated Il17d–/– and Cd93–/– mice was similar to that in healthy skin from wild-type mice (Fig. 5k–n). These observations indicated that DDX5 deficiency led to IL36R pre-mRNA splicing that favored the production of IL-36R.
DDX5 cannot recognize sequence specificity in its RNA substrates9,33. Immunoprecipitation and mass spectrometry (MS) indicated that the splicing factor SF2 (also known as SRSF1), which recognizes specific splice sites and defines splicing patterns34, was the second most abundant protein, after DDX17, that interacted with DDX5 (Extended Data Fig. 8a). The interaction was confirmed by immunoprecipitation of SF2 and DDX5 on the same Coomassie Blue-stained gel (Fig. 6a) and by coimmunoprecipitation (Extended Data Fig. 8b,c). Expression of SF2 protein was reduced in the skin lesions of patients with AD and psoriasis or MC903- or IMQ-treated wild-type mice (Fig. 6b–e). Less sIL36R transcript from an IL36R reporter minigene was detected in SF2–/– compared with wild-type HeLa cells, and expression was restored by overexpression of SF2, but not DDX5 (Fig. 6f). The siRNA targeting of SF2 in IL-36γ-stimulated NHEKs reduced the expression of sIL-36R, increased the expression of IL-36R (Fig. 6g) and increased the expression of CCL20, CXCL1, CXCL6, CCL17, CCL3, CCL11 and CCL27 compared with IL-36γ-stimulated wild-type NHEKs (Extended Data Fig. 8d). An exonic splicing enhancer (ESE) finder program (http://exon.cshl.edu/ESE)35 identified multiple putative SF2-binding ESEs in exon3, exon2 and exon4 (Extended Data Fig. 8e) and native RNA immunoprecipitation indicated that SF2 bound predominantly in exon4 of IL36R mRNA (Fig. 6h and Extended Data Fig. 8c,e), suggesting that this interaction might mediate exon3 skipping (Extended Data Fig. 8f). SF2 bound to exon3 and exon4 of IL36R mRNA in DDX5–/– HeLa cells, whereas DDX5 did not interact with exon3 and exon4 of IL36R mRNA in SF2–/– HeLa cells (Fig. 6h). Deletion of the ESE in exon4 abrogated SF2 binding (Fig. 6i and Extended Data Fig. 8g) and inhibited the generation of sIL36R in HaCaT cells transduced with IL36R reporter genes (Fig. 6j). These data indicated that DDX5 cooperated with SF2 to regulate IL-36R splicing for sIL-36R production in keratinocytes.
Soluble IL-36R antagonizes IL-36R signaling to control skin inflammation
Next, we determined the role of sIL-36R in skin inflammation. Immunoprecipitation assays showed that sIL-36R bound to IL-36γ and competed with IL-36R for IL-36γ binding in a dose-dependent manner (Fig. 7a), whereas flow cytometry indicated that sIL-36R inhibited the interaction between IL-36R and IL-36γ (Extended Data Fig. 9a). Moreover, overexpression of Flag-tagged sIL-36R decreased the IL-36γ-induced phosphorylation of p65 and p38 MAPK and expression of CCL20, CXCL1, CXCL2, CCL3, CCL17 and CCL27 in NHEKs (Fig. 7b,c), and CCL20, CXCL1, CXCL2, CXCL3, CCL3, CCL11, CCL17 and CCL27 in DDX5-silenced NHEKs (Extended Data Fig. 9b), compared with their nontransduced counterparts. Next, we injected recombinant murine sIL-36R into the ears of MC903- or IMQ-treated wild-type mice. Compared with phosphate-buffered saline (PBS) controls, MC903-treated wild-type mice that received sIL-36R had a reduced number of ear skin patches (Extended Data Fig. 9c), decreased ear thickness (Fig. 7d), less expression of Il4, Il13, Tslp, Ccl11 and Ccl17 (Fig. 7e) and reduced infiltration of CD45+ immune cells, including CD11b+SiglecF+ eosinophils, CD49b+IgE+ basophils and CD3+CD4+ T cells (Fig. 7f). Administration of sIL-36R also reduced ear-skin plaques, decreased epidermal acanthosis, suppressed the production of Ccl20, Cxcl1, Il23 and Il17a and inhibited the infiltration of CD45+ immune cells, including Ly6G+ neutrophils, CD45+MHCII+ APCs and γδTCR+ T cells in the skin lesions of IMQ-treated wild-type mice compared with mice that received a PBS injection (Extended Data Fig. 9c–f). These data indicated that sIL-36R antagonized IL-36R signaling to control cutaneous inflammation.
Soluble IL-36R inhibits cutaneous inflammation in Ddx5 ∆KC mice
Next, we investigated whether the decrease in cutaneous sIL-36R was a causative factor for the immunopathology of AD and psoriasis. MC903- or IMQ-treated Ddx5∆KC mice intradermally injected with sIL-36R had fewer skin patches or plaques (Extended Data Fig. 10a,b), a significant decrease in ear thickness (Fig. 8a,b) and lower expression of IL-4, IL-13, thymic stromal lymphopoietin (TSLP) proteins and Ccl11, Ccl17 and Ccl22 mRNA in skin lesions from MC903-treated Ddx5∆KC mice or CCL20, CXCL1, IL-23, IL-17A, IL-17F and TNF proteins in skin lesions from IMQ-treated Ddx5∆KC mice (Fig. 8c–e) compared with PBS controls. When Ddx5∆KC mice were crossed with sIL-36RTg/KC mice, which express CAG promoter-driven human sIL-36R in keratinocytes, MC903-treated Ddx5∆KCsIL36RTg/KC mice did not develop skin plaques, and had thinner ear and epidermis (Fig. 8f and Extended Data Fig. 10c,d) and reduced expression of IL-4, IL-13, TSLP proteins and Ccl11, Ccl17 and Ccl22 mRNA in lesions (Fig. 8g,h) compared with MC903-treated Ddx5∆KC mice. In addition, IMQ-treated Ddx5∆KCsIL36RTg/KC mice did not develop plaques or acanthosis in the skin (Extended Data Fig. 10e,f), and had decreased ear thickness and fewer CCL20, CXCL1, IL-23 and IL-17A proteins (Fig. 8i,j) compared with IMQ-treated Ddx5∆KC mice. These data indicated that a reduction in cutaneous sIL-36R was a causative factor for the exacerbated inflammation observed in MC903- or IMQ-treated Ddx5∆KC mice.
In the present study, we observed that IL-17D signaling modulated the expression of DDX5 in keratinocytes, which in turn controlled the extent of inflammation in the skin by regulating the alternative splicing of IL-36R. DDX5 expression in keratinocytes was inhibited by IL-17D through the activation of a CD93–p38 MAPK–AKT–SMAD2/3 signaling pathway and selectively amplified the inflammatory response to IL-36 and aggravated cutaneous inflammation. DDX5 required SF2, which recognized ESEs in exon4 of the IL1RL2 gene, to alternatively splice a soluble form of IL-36R, which antagonized membrane-bound IL-36R to limit cutaneous inflammation.
AD and psoriasis are mediated by distinct T cell polarity and immune responses to T cell-derived cytokines. High-throughput sequencing analyses showed that AD and psoriasis are shaped by dysregulated immune responses to different cytokines and chemokines when DEGs from AD and psoriatic lesions, compared with nonlesions in their corresponding individuals, were analyzed6,7,8. However, when shared DEGs from nonlesional and lesional skin in patients with AD and psoriasis compared with healthy skin were used for GO and the KEGG pathway enrichment analyses, RNA splicing regulated by shared spliceosome complexes was uncovered in both AD and psoriasis. Among all identified shared spliceosomes from GO enrichment analysis, DDX5 is a critical component of multiple spliceosome complexes, such as DDX5/DDX17, DDX5/HNRNPA1 and DDX5/SF2. Our high-throughput RNA-seq results of genetic deletion of DDX5 in HaCaT cells or keratinocyte Ddx5 in mice confirmed DDX5 as a key mediator responsible for the change of alternative splicing patterns observed in AD and psoriasis
Cytokines such as IL-17A, IL-17E, IL-4 and IL-36γ regulate the inflammatory responses in keratinocytes in both AD and psoriasis. In addition to those, we showed that IL-17D signals through CD93–p38 MAPK–AKT–SMAD2/3 to inhibit DDX5 expression in keratinocytes. IL-17D is known to regulate the function of group 3 innate lymphoid cells (ILC3 cells) against intestinal inflammation23, promote tumor rejection through recruitment of natural killer cells36 or promote Listeria infection by suppressing CD8+ T cell activity37. In the present study, we showed that IL-17D regulated IL-36R-mediated skin inflammation through inhibition of DDX5 expression. IL-17D is expressed by endothelial cells, adipocytes or epithelial cells, not by T and B cells23,38. We did not observe IL-17D expression in skin adipocytes and keratinocytes during AD and psoriasis. Therefore, the cells that express IL-17D in AD and psoriasis remain to be determined. Moreover, CD93 is expressed in multiple cells, including keratinocytes, ILC3 cells, endothelial cells and macrophages. Although we observed that IL-17D regulated keratinocyte inflammation by activation of CD93-mediated signaling, whether IL-17D would regulate immune responses of other CD93+ cells to drive skin inflammation needs further investigation.
DDX5 acts as a component of the spliceosome and cannot recognize sequence specificity in its RNA substrates39, suggesting that other cofactors might specifically recognize the alternative splice site in IL36R pre-mRNA. Immunoprecipitation and MS identified multiple DDX5-interacting splicing factors, such as serine/arginine-rich splicing factors (SRSFs) and heterogeneous nuclear ribonucleoproteins (HNRNPs). Among these, SF2 determines splice sites by binding to ESEs in flanking exons of an alternative exon40,41, whereas HNRNPs block the access of spliceosome elements and inhibit splice-site selection by binding to exonic or intronic splicing silencers42. Silencing or deletion of SF2 changed the IL-36R splicing pattern in keratinocytes, whereas silencing of DDX17 or overexpression of HNRNPs did not, excluding their involvement in DDX5-mediated IL-36R splicing. How DDX5 interacts with SF2 to drive IL-36R splicing requires further investigation.
IL-36 cytokines, including IL-36α, IL-36β and IL-36γ, bind and signal through a heterodimeric receptor composed of IL-36R and the IL-1R accessory protein (IL-1RAcP)43. IL-36 is elevated in inflamed skin and associates with the pathogenesis of multiple inflammatory skin diseases, such as AD and psoriasis44,45. Keratinocyte-released IL-36 cytokines increase IL-4-mediated immunoglobulin (Ig)E production in B cells in AD, and treatment with an IL-36R-blocking antibody decreases IgE production and alleviates the disease phenotype46. In psoriasis, IL-36 induces keratinocyte proliferation and dendritic cell activation47,48. Deletion of IL-36R in mice reduces the number of dermal IL-17-producing γδ T cells and protects mice from psoriasiform dermatitis48. IL-36–IL-36R signaling is suppressed by the IL-36R antagonist (IL-36Ra), which is encoded by IL36RN and can bind IL-36R49. Missense mutations in IL36RN that lead to loss of IL-36Ra expression associate with the development of generalized pustular psoriasis50. Our functional and genetic analysis indicated that alternative splicing generated sIL-36R as an additional antagonist of IL-36R that suppressed IL-36R signaling by competing for binding to IL-36 in keratinocytes. Whether IL-36R splicing regulated by the DDX5–SF2 complex and sIL-36R inhibition of IL-36–IL-36R signaling are keratinocyte specific requires further investigation.
In conclusion, our observations indicate that the IL-17D–CD93-mediated decrease in DDX5 expression amplifies cutaneous inflammation and suggest a potential for IL-17D and DDX5 as therapeutic targets in inflammatory skin diseases, whereas the identification of sIL-36R provides insights into the contribution of aberrant RNA splicing to skin inflammation, and may ultimately lead to the development of alternative therapeutic approaches in AD and psoriasis.
Skin samples were obtained from seven patients with AD (moderate to severe), ten patients with psoriasis (mild plaque-type psoriasis), three patients with basal cell carcinoma, three patients with squamous cell carcinoma and eight healthy patients with a 2-mm punch biopsy. The information for all the patients is shown in Supplementary Table 1. All these samples were used for protein extraction or paraffin section. Sample acquisitions, including skin biopsies, were approved by the ethics review committees of Huashan Hospital and Shanghai Tenth People’s Hospital and performed in accordance with the declaration of Helsinki principles. Informed consent was obtained for all procedures. All patients are volunteers and did not receive compensation.
The mice used for OVA induction are BABL/c background and those used for MC903 or IMQ induction are C57BL/6 background. All mice were bred in the specific pathogen-free animal facility at either East China Normal University or Tsinghua University. DDX5fl/fl and sIL36Rfl/fl mice were generated by Shanghai Model Organisms Center, Inc. The K14Cre transgenic mice were obtained from Shanghai Model Organisms Center, Inc., whereas the K5Cre transgenic mice were obtained from Xiao Yang lab in the Academy of Military Medical Sciences in China. DDX5fl/fl mice were crossed with the K14Cre transgenic mice to generate Ddx5∆KC mice, sIL36Rfl/fl mice were crossed with the K5Cre mice to generate sIL36RTg/KC mice and Ddx5∆KC mice were crossed with sIL36RTg/KC mice to generate Ddx5∆KCsIL36RTg/KC mice. All the animal experiments were performed with the use of the protocols (protocol nos.: m20210233 for AD, m20200316 for psoriasis) approved by the Animal Care and Use Committee at East China Normal University. All surgery was conducted under anesthesia and all efforts were made to minimize suffering. For all animal studies, we performed preliminary experiments to determine the requirements for sample size. Mice were grouped according to genotypes but not performed in a blinded manner. Unless otherwise stated, 8-week-old littermates were used for each animal experiment.
Primary murine keratinocytes were isolated from newborn pups (1 or 2 d after birth) and cultured in Medium 154CF (Gibco) supplemented with 0.05 mM Ca2+ and human keratinocyte growth supplement (HKGS, Gibco). NHEKs (Lifeline Cell Technology, catalog no. FC-0007) stored in liquid nitrogen were defrosted and cultured in EpiLife medium (Gibco) containing EpiLife Defined Growth Supplement (EDGS, Gibco) and 0.06 mM Ca2+ (Gibco). The medium was refreshed every 2 d and cells were subcultured according to the cell fusion. Cells at passages 3–5 were used for subsequent experiments. Immortalized human keratinocyte cell line HaCaT (Cobioer, catalog no. CBP6033) was cultured in RPMI 1640 medium (Gibco) containing 10% fetal bovine serum (FBS; Gibco), 50 U ml−1 of penicillin and 50 μg ml−1 of streptomycin (Shanghai Yuan Pei) under standard culture conditions. HeLa and HEK293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) containing 10% FBS, 50 U ml−1 of penicillin and 50 μg ml−1 of streptomycin under standard culture conditions. All cells were cultured at 37 °C and 5% CO2.
AD mouse model
For the calcipotriol (MC903)-induced AD mouse model, 1 nmol of MC903 (Sigma-Aldrich) dissolved in 10 μl of ethanol was painted on the ears of 8-week-old mice (C57BL/6 background) for 15 d consecutively. Photographs of the ears were taken every day and the ear thickness was measured with vernier calipers (Meinaite) every day. For the OVA-induced AD mouse model, 8-week-old mice were intraperitoneally inoculated with 10 μg of chicken OVA mixed with 4 mg of aluminum hydroxide (ImjectAlum; Thermo Fisher Scientific) in a volume of 200 μl at 1-week intervals (that is, at days 0, 7 and 14). At day 13, the dorsal skin of mice was shaved and tape stripped six times with 3M tape. At day 14, 100 μg of OVA in 100 μl of PBS was placed on a 1-cm2 patch of sterile gauze to make an OVA patch, which was attached on the shaved dorsal skin with a transparent dressing (Tegaderm, 3M) for 7 d (that is, from day 14 to day 20). OVA patches were changed daily. Each mouse had a total of three 1-week exposures to the patch at the same site of separation from each other at 2-week intervals. Dorsal skin or ears were collected for flow cytometry, immunoblotting, RT–qPCR, ELISA or immunofluorescence and hematoxylin and eosin (H&E) staining.
Psoriasis mouse model
Mice aged 8 weeks (on a C57BL/6 background) were subjected to a daily topical dose of 62.5 mg of IMQ cream (5%) (Shichuan MedShine Pharmaceuticals Co.) on the shaved back for 5 d consecutively or 25 mg per ear for 7 d consecutively. The ear thickness was measured with vernier calipers every day. The dorsal skin or ears were collected for flow cytometry, immunoblotting, RT–qPCR, ELISA or immunofluorescence and H&E staining.
Intradermal injection of sIL-36R
Each mouse was injected intradermally with 1 μg of purified recombinant murine sIL-36R (dissolved in 20 μl of PBS) in the right ear or 20 μl of PBS in the left ear 1 d before MC903 or IMQ application. On day 16 for MC903-induced AD mice or day 8 for IMQ-induced psoriasis mice, the ears were collected for histology, immunoblotting, ELISA, qPCR or flow cytometric analysis.
Primary keratinocyte culture and stimulation
Primary murine keratinocytes were isolated from DDX5fl/fl and DDX5∆/KC neonates by using 10 mg ml−1 of dispase II (Sigma-Aldrich) digestion overnight at 4 °C, followed by 0.05% trypsin–EDTA (Shanghai Yuan Pei) for 10 min at 37 °C. The cells were cultured with Medium 154CF supplemented with 0.05 mM CaCl2 and HKGS. NHEKs were cultured in EpiLife medium containing EDGS and 0.06 mM Ca2+. To test keratinocytes in response to cytokines, cells were stimulated by the indicated concentrations of cytokines or inhibitors for the indicated time.
Gene deletion or silencing
For the depletion of DDX5 and SF2 in HaCaT and HeLa cells, lentiCRISPRv2 plasmids containing hSpCas9 and the guide (g)RNA targeting DDX5 or SF2 in Supplementary Table 2 were constructed. These plasmids were cotransfected HEK293T cells with the packaging plasmids pMD2G (AddGene) and psPAX2 (AddGene) for lentiviral production. HaCaT or HeLa cells were transduced in suspension with 500 μl of viral supernatant in the wells of a 24-well plate. Cells with DDX5 or SF2 depletion were screened out by using 1 μg ml−1 of puromycin. PCR and Immunoblot analyses were used to confirm that DDX5 or SF2 was deleted in these cells.
For CD93, IL-36R, DDXs or SF2 silencing in primary human keratinocytes, siRNAs targeting human CD93, IL-36R, DDXs and SF2 were synthesized by GenePharm. Using lipo3000 transfection reagent (Invitrogen), siRNAs targeting CD93, IL-36R, DDXs or SF2 (Supplementary Table 2) were transfected into primary human keratinocytes at a final concentration of 20 nM. The silencing efficient was analyzed by RT–qPCR.
Mouse ear tissues were homogenized in TRIzol (Takara) using Beadbeater (Biospec), whereas cells were directly resuspended in TRIzol (Takara). Total RNAs were isolated and reverse transcribed into complementary DNA by using RevertAid First Strand cDNA Synthesis Kit (Roche) according to the manufacturer’s instructions. RT–qPCR was performed in triplicate using SYBR green master mix (Roche) on a StepOnePlus Real-Time PCR System (Applied Biosystem). Samples with a low yield of RNA were predetermined and excluded. Results were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and the comparative ∆∆CT method was used to determine the quantification of gene expression. Primers used in this paper are listed in Supplementary Table 3.
ScRNA-seq data analysis
The processed scRNA-seq data were downloaded from ref. 24 and converted to a Seurat object using SeuratDisk (https://mojaveazure.github.io/seurat-disk). Followed by the typical Seurat workflow (http://satijalab.org/seurat), 2,000 highly variable genes were normalized, scaled and identified using the NormalizeData, ScaleData and FindVariableGenes function from Seurat. The number of positive cells was determined by visual inspection of the ElbowPlot. Uniform Manifold Approximation and Projection with a resolution of 0.5 was used to determine cell clusters. The annotation information was reserved for defining clusters. To evaluate the expression of DDX5 in keratinocytes, fibroblasts and lymphocytes, cell subsets were combined and the expression level was visualized by VlnPlot from Seurat.
WT and DDX5–/– HaCaT cells treated with 100 ng ml−1 of IL-36γ (Novoprotein) for 4 h or lesional skin from MC903-treated or IMQ-treated Ddx5∆KC and Ddx5fl/fl mice were collected. Total RNAs were isolated with TRIzol. RNA-seq was performed using an Illumina system following Illumina-provided protocols for 2 × 150 paired-end sequencing in WuXi NextCODE at Shanghai, China. To obtain clean reads, FastQC (v.0.11.9) was used to assess the overall quality of raw reads and Trimmomatic (v.0.39)51 was applied for raw read quality control to cut adapters and remove low-quality reads. Then all the clean reads were mapped on to the human hg38 genome using Hisat2 (v.2.1.0)52. Samtools (v.1.7)53 was used to convert SAM format files into BAM format files and sorted the BAM files. Gene expression levels were quantified by FeatureCounts (v.1.6.3)54. Differential expression analysis was performed by R package DESeq2 (v.1.34.0)55 and the false recovery rate <0.05 was considered to be significantly differentially expressed. GO and KEGG pathway enrichment analyses were performed in ClusterProfiler (v.4.2.0)56. Differential alternative splicing events were detected by rMATS (v.3.1.0)57 and events with P < 0.05 were identified as significantly differentially expressed, alternative splicing events.
Skin from MC903- or IMQ-treated mice was homogenized in pre-cooled PBS, pH 7.4, by using Beadbeater (Biospec), and the supernatants of skin homogenate were collected for cytokine evaluation. Cytokine production was measured by ELISA kits of IL-4 (Multisciences (Lianke) Biotech Co., Ltd.), IL-13 (eBioscience), TSLP (Multisciences (Lianke) Biotech Co., Ltd.), IL-23 (R&D), IL-17A (R&D), CCL20 (R&D), CXCL1 (R&D), IL-17F(BD Pharmingen) and TNF (BD Pharmingen) according to the manufacturer’s instructions. The samples with a low yield of protein were predetermined and excluded.
Histology and immunofluorescence staining
Formalin-fixed, paraffin-embedded tissue sections (~5 μm in thickness) mounted on glass slides were used for various methods of staining. The H&E staining was performed as previously described17. The epidermal hyperplasia (acanthosis) was evaluated in 12 independent regions of each section. For immunofluorescence, the sections were deparaffinized and pretreated with antigen retrieval solution (10 mM sodium citrate buffer, pH 6.0) for 20 min. The sections were then blocked by 3% bovine serum albumin in PBS for 1 h at room temperature and stained with DDX5 antibody (Abcam) at 4 °C overnight. Next day, the sections were reprobed with rabbit IgG FITC-conjugated antibody (Invitrogen) and then mounted in ProLong Gold antifade reagent with DAPI (Invitrogen) and visualized by confocal microscope (Leica).
Immunoblotting and immunoprecipitation
The 2-mm pieces of skin taken from patients with AD or psoriasis and MC903-treated, OVA-treated or IMQ-treated mice, or the cells with different treatments, were lysed with pre-cooled radioimmunoprecipitation (RIPA) buffer, pH 7.4, containing protease inhibitor cocktail (Roche). Then, 30 μg of total protein was subjected to sodium dodecylsulfate (SDS)–polyacrylamide gel electrophoresis (PAGE) and blotted using the indicated antibodies. For immunoprecipitation, HEK293T or HeLa cells were cultured in DMEM medium and transfected with plasmids containing epitope-tagged DDX5 or SF2. After 24 h, cells were lysed in lysis buffer (250 mM NaCl, 50 mM Hepes, pH 7.4, 1 mM EDTA, 1% Nonidet P-40 (NP-40)) containing protease inhibitor cocktail (Roche). Total protein, 40 μg, was used for immunoprecipitation with anti-Flag beads (Bimake) or anti-hemagglutinin (HA) beads (Bimake) and the precipitated protein complex was used for immunoblotting with antibodies against Flag (MBL) or HA (MBL).
NHEKs were lysed. Of the cell lysate, 10% was used for input control and 90% was incubated with DDX5 antibody (Abcam) at 4 °C. Next day, protein A/G agarose beads (Beyotime) were added and the incubation was continued for 2–3 h at 4 °C. DDX5-pulldown or rabbit IgG-pulldown proteins were loaded for SDS–PAGE. After Coomassie Brilliant Blue staining, bands with strong intense signals were cut and digested. The resulting peptides were analyzed on the high-pressure liquid chromatography (HPLC) liquid system Dionex Ultimate 3000 (Thermo Fisher Scientific) coupled to a Dionex Trap column (100 μm × 2 cm × 5 μm) with an in-house packed C18 column (75 μm × 15 cm × 3 μm), and the mass of peptides was analyzed by the maXis HD-UHR-TOF mass spectrometer (Bruker). The spectra from MS were automatically used for searching against the nonredundant International Protein Index human protein database (v.3.72) with the Bioworks browser (rev.3.1).
Skin cell preparation and flow cytometry
Skin cells were prepared according to previous studies with minor modifications58. In general, the epidermis and dermis were separated using dispase II (5 mg ml−1 in Hanks’ Balanced Salt Solution; 37 °C for 90 min). The dermal cells were separated by collagenase (Roche) and hyaluronidase (Sigma-Aldrich) digestion (10 mM Hepes, collagenase D (2.5 mg ml−1), hyaluronidase (100 U ml−1) and DNase (50 μg ml−1) in DMEM; 37 °C for 45 min). Isolated cells were stained with different cell surface markers (CD45 for leukocytes; CD45+ and IgE+ for basophils; CD45+ and SiglecF+ for eosinophils; CD45+, CD3+ and CD4+ for CD4+ T cells; CD45+, CD11b+ and Ly6G+ for neutrophils; CD45+, CD11c+ and MHCII+ for APCs; and CD45+, CD3+ and γδT cell receptor (TCR+) for γδT cells). The cells were then fixed and the relevant isotype control monoclonal antibodies were used. Samples were analyzed using LSR Fortessa (BD Biosciences) and FlowJo v.10 software (TreeStar).
Transfection of IL36R splicing reporter
IL36R-reporter minigene consisted of the genomic region of IL36R encompassing 139 bp of intron2, exon3, 201 bp of intron3 and exon4 was cloned in the pcDNA3.1 plasmid. PCR primers for IL36R reporter minigene constructs are listed in Supplementary Table 3. Two versions of the IL36R-reporter minigene containing the wild-type exon4 or mutation of the ESE in the context of exon4 were constructed and transfected into HeLa cells using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s recommendations. Total RNA was harvested 24 h after transfection using TRIzol reagent.
Characterization of IL36R splicing
For IL36R splicing analysis, total RNA was extracted and reverse transcription was performed using RevertAid First Strand cDNA Synthesis Kit (Roche) according to the manufacturer’s instructions. PCR reactions (20 μl) were prepared as follows: 2X Taq Master (Novoprotein), 10 μl; IL36R Exon3 Primer (forward), 0.5 μl; IL36R Exon4 Primer (reverse), 0.5 μl; cDNA, 2 μl; and ddH2O, 7 μl. PCR primers for endogenous IL36R were complementary to exon2 (forward) and exon5 (reverse) (Supplementary Table 3). PCR products were electrophoresed on 12% nondenaturing polyacrylamide/TBE gels.
RNA immunoprecipitation (RIP) was performed as previously described59 with minor modification. WT, DDX5–/– and SF2–/– HeLa cells were transfected with an IL36R-reporter minigene and plasmids containing HA-tagged DDX5 or Flag-tagged SF2, respectively. After 24 h, cells were collected and suspended in an equal volume of polysome lysis buffer (100 mM KCl, 5 mM MgCl2, 10 mM Hepes, pH 7.0, 0.5% NP-40, 1 mM dithiothreitol (DTT), 100 units ml−1 of RNase and 400 μM RVC (New England Biolabs)) supplemented with protease inhibitors. Cell lysates were centrifuged at 15,000g for 15 min to remove large particles. Before incubation with cell lysates, HA or Flag beads were washed by 1 ml of ice-cold NT2 buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM MgCl2 and 0.05% NP-40) 4× and then resuspended in 850 μl of the immunoprecipitation reaction solution (200 units of an RNase inhibitor, 400 μM RVC, 10 μl of 100 mM DTT, 30 μl of 0.5 mM EDTA and 800 μl of cold NT2 buffer). Cleared cell lysates (20 mg ml−1), 150 μl, were incubated with 850 μl of HA/Flag beads solution at 4 °C for 4 h and then centrifuged at 366g for 2 min to pellet beads. Beads were then washed with 1 ml of ice-cold NT2 buffer 4–5× and TRIzol was added to isolate total RNA from messenger ribonucleoprotein components binding on beads. Total RNA was reverse transcribed into cDNA by using PrimeScript RT Reagent Kit with gDNA Eraser (Takara) according to the manufacturer’s instructions and specific primers for exon3 and exon4 were used to amplify SF2- or DDX5-binding fragments.
Competition binding assay
HeLa cells were seeded into a 6-well plate and cultured in DMEM. The cells were grown to 70% confluence and then transfected with the indicated doses of plasmids containing Flag-tagged IL36R gene or Flag-tagged sIL36R gene, or transfected with plasmids containing green fluorescent protein (GFP)–IL36R gene or sIL36R gene. After 36 h, cells transfected with plasmids containing GFP–IL-36R or sIL-36R were stimulated by 100 ng ml−1 of His-tagged IL-36γ. After 30 min, cells were collected and stained with phycoerythrin–anti-His antibody for 30 min, followed by flow cytometric analysis. For competition binding analyzed by immunoprecipitation, cells transfected with plasmids containing Flag-tagged IL36R gene or Flag-tagged sIL36R gene were lysed with radioimmunoprecipitation assay buffer, and 1 μg of His-tagged IL-36γ was added. After 4 h, IL-36γ-binding IL-36R or sIL-36R was pulled down by anti-His beads (Bimake) and loaded on to SDS–PAGE for immunoblotting with anti-Flag antibody.
Statistics and reproducibility
In vitro experiments: experiments were done in triplicate and independently repeated three times with similar results, with few exceptions in which experiments were repeated twice. For each experiment every sample was processed identically and internal controls and normalization methods were included to avoid technical bias. In vivo experiments: experiments were independently repeated twice with similar results. An exact n for each experimental group/condition is shown in the figures by symbols and in Supplementary Table 4. Each symbol represents an individual mouse. All data are presented as mean ± s.e.m. We used unpaired, two-tailed Student’s t-test to determine significance between two groups. We did analyses of multiple groups by one-way or two-way analysis of variance (ANOVA) with Bonferroni’s posttest of GraphPad Prism v.9. For all statistical tests, we considered P < 0.05 to be statistically significant. P values of RT–qPCR, ELISA and flow cytometric analyses are reported in Supplementary Table 4. No statistical methods were used to predetermine sample sizes, but our sample sizes are similar to those reported in previous publications15,17,23. Data distribution was assumed to be normal but this was not formally tested.
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
The raw sequence data reported in the present paper have been deposited in the GEO database under accession nos. GSE208666, GSE208669 and GSE208671. The MS proteomics data have been deposited in the ProteomeXchange Consortium (http://www.proteomexchange.org)60 under accession no. PXD021379. All other data supporting the findings of the present study are available within the paper or from the corresponding author upon request. Source data are provided with this paper.
The raw sequence code reported in the present paper has been deposited in the GEO database under accession nos. GSE208666, GSE208669 and GSE208671. The MS proteomics code has been deposited in the ProteomeXchange Consortium (http://www.proteomexchange.org)60 under accession no. PXD021379. All other code supporting the findings of thhis present study are available within the paper or from the corresponding author upon request.
Frankiw, L., Baltimore, D. & Li, G. D. Alternative mRNA splicing in cancer immunotherapy. Nat. Rev. Immunol. 19, 675–687 (2019).
Montes, M., Sanford, B. L., Comiskey, D. F. & Chandler, D. S. RNA splicing and disease: animal models to therapies. Trends Genet. 35, 68–87 (2019).
Schaub, A. & Glasmacher, E. Splicing in immune cells-mechanistic insights and emerging topics. Int. Immunol. 29, 173–181 (2017).
Guttman-Yassky, E., Nograles, K. E. & Krueger, J. G. Contrasting pathogenesis of atopic dermatitis and psoriasis–part I: clinical and pathologic concepts. J. Allergy Clin. Immunol. 127, 1110–1118 (2011).
Guttman-Yassky, E. & Krueger, J. G. Atopic dermatitis and psoriasis: two different immune diseases or one spectrum? Curr. Opin. Immunol. 48, 68–73 (2017).
Tsoi, L. C. et al. Atopic dermatitis Is an IL-13-dominant disease with greater molecular heterogeneity compared to psoriasis. J. Invest. Dermatol. 139, 1480–1489 (2019).
Li, J. & Yu, P. Genome-wide transcriptome analysis identifies alternative splicing regulatory network and key splicing factors in mouse and human psoriasis. Sci. Rep. 8, 4124 (2018).
Schwingen, J., Kaplan, M. & Kurschus, F. C. Review—current concepts in inflammatory skin diseases evolved by transcriptome analysis: in depth analysis of atopic dermatitis and psoriasis. Int. J. Mol. Sci. 21, 699 (2020).
Xing, Z., Ma, W. K. & Tran, E. J. The DDX5/Dbp2 subfamily of DEAD-box RNA helicases. Wiley Interdiscip. Rev. RNA 10, e1519 (2019).
Hashemi, V. et al. The role of DEAD-box RNA helicase p68 (DDX5) in the development and treatment of breast cancer. J. Cell Physiol. 234, 5478–5487 (2019).
Legrand, J. M. D. et al. DDX5 plays essential transcriptional and post-transcriptional roles in the maintenance and function of spermatogonia. Nat. Commun. 10, 2278 (2019).
Fuller-Pace, F. V. DEAD box RNA helicase functions in cancer. RNA Biol. 10, 121–132 (2013).
Xu, J. et al. The RNA helicase DDX5 promotes viral infection via regulating N6-methyladenosine levels on the DHX58 and NFkappaB transcripts to dampen antiviral innate immunity. PLoS Pathog. 17, e1009530 (2021).
Xing, X. et al. IL-17 responses a the dominant inflammatory signal linking inverse, erythrodermic, and chronic plaque psoriasis. J. Invest Dermatol. 136, 2498–2501 (2016).
Han, Y. et al. Genome-wide analysis highlights contribution of immune system pathways to the genetic architecture of asthma. Nat. Commun. 11, 1776 (2020).
Bantz, S.K., Zhu, Z. & Zheng, T. The atopic march: progression from atopic dermatitis to allergic rhinitis and asthma. J. Clin. Cell Immunol. 5, 202 (2014).
Lai, Y. et al. The antimicrobial protein REG3A regulates keratinocyte proliferation and differentiation after skin injury. Immunity 37, 74–84 (2012).
Xu, M. et al. An interleukin-25-mediated autoregulatory circuit in keratinocytes plays a pivotal role in psoriatic skin inflammation. Immunity 48, 787–798 e784 (2018).
McGeachy, M. J., Cua, D. J. & Gaffen, S. L. The IL-17 family of cytokines in health and disease. Immunity 50, 892–906 (2019).
Su, Y. et al. Interleukin-17 receptor D constitutes an alternative receptor for interleukin-17A important in psoriasis-like skin inflammation. Sci. Immunol. https://doi.org/10.1126/sciimmunol.aau9657 (2019).
Hvid, M. et al. IL-25 in atopic dermatitis: a possible link between inflammation and skin barrier dysfunction? J. Invest. Dermatol. 131, 150–157 (2011).
Chung, S. H., Ye, X. Q. & Iwakura, Y. Interleukin-17 family members in health and disease. Int. Immunol. 33, 723–729 (2021).
Huang, J. et al. Interleukin-17D regulates group 3 innate lymphoid cell function through its receptor CD93. Immunity 54, 673–686 e674 (2021).
Reynolds, G. et al. Developmental cell programs are co-opted in inflammatory skin disease. Science 371, eaba6500 (2021).
Langan, S. M., Irvine, A. D. & Weidinger, S. Atopic dermatitis. Lancet 396, 345–360 (2020).
Hawkes, J. E., Yan, B. Y., Chan, T. C. & Krueger, J. G. Discovery of the IL-23/IL-17 signaling pathway and the treatment of psoriasis. J. Immunol. 201, 1605–1613 (2018).
Noda, S. et al. The Asian atopic dermatitis phenotype combines features of atopic dermatitis and psoriasis with increased TH17 polarization. J. Allergy Clin. Immunol. 136, 1254–1264 (2015).
Shao, S. et al. IRAK2 has a critical role in promoting feed-forward amplification of epidermal inflammatory responses. J. Invest. Dermatol. 141, 2436–2448 (2021).
Ghoreschi, K. et al. Interleukin-4 therapy of psoriasis induces Th2 responses and improves human autoimmune disease. Nat. Med. 9, 40–46 (2003).
Gros, E., Bussmann, C., Bieber, T., Forster, I. & Novak, N. Expression of chemokines and chemokine receptors in lesional and nonlesional upper skin of patients with atopic dermatitis. J. Allergy Clin. Immunol. 124, 753–760.e751 (2009).
Lowes, M. A., Suarez-Farinas, M. & Krueger, J. G. Immunology of psoriasis. Annu Rev. Immunol. 32, 227–255 (2014).
Lee, Y. J., Wang, Q. & Rio, D. C. Coordinate regulation of alternative pre-mRNA splicing events by the human RNA chaperone proteins hnRNPA1 and DDX5. Genes Dev. 32, 1060–1074 (2018).
Han, J. et al. SR proteins induce alternative exon skipping through their activities on the flanking constitutive exons. Mol. Cell. Biol. 31, 793–802 (2011).
Krainer, A. R., Conway, G. C. & Kozak, D. The essential pre-mRNA splicing factor SF2 influences 5′ splice site selection by activating proximal sites. Cell 62, 35–42 (1990).
Graveley, B. R., Hertel, K. J. & Maniatis, T. A systematic analysis of the factors that determine the strength of pre-mRNA splicing enhancers. EMBO J. 17, 6747–6756 (1998).
O’Sullivan, T. et al. Interleukin-17D mediates tumor rejection through recruitment of natural killer cells. Cell Rep. 7, 989–998 (2014).
Lee, Y., Clinton, J., Yao, C. & Chang, S. H. Interleukin-17D promotes pathogenicity during infection by suppressing CD8 T cell activity. Front. Immunol. 10, 1172 (2019).
Starnes, T., Broxmeyer, H. E., Robertson, M. J. & Hromas, R. Cutting edge: IL-17D, a novel member of the IL-17 family, stimulates cytokine production and inhibits hemopoiesis. J. Immunol. 169, 642–646 (2002).
Lee, Y. & Rio, D. C. Mechanisms and regulation of alternative pre-mRNA splicing. Annu. Rev. Biochem. 84, 291–323 (2015).
Valcarcel, J. & Green, M. R. The SR protein family: pleiotropic functions in pre-mRNA splicing. Trends Biochem. Sci. 21, 296–301 (1996).
Spena, S., Tenchini, M. L. & Buratti, E. Cryptic splice site usage in exon 7 of the human fibrinogen Bbeta-chain gene is regulated by a naturally silent SF2/ASF binding site within this exon. RNA 12, 948–958 (2006).
Geuens, T., Bouhy, D. & Timmerman, V. The hnRNP family: insights into their role in health and disease. Hum. Genet. 135, 851–867 (2016).
Sachen, K. L., Arnold Greving, C. N. & Towne, J. E. Role of IL-36 cytokines in psoriasis and other inflammatory skin conditions. Cytokine 156, 155897 (2022).
Tsang, M. S., Sun, X. & Wong, C. K. The role of new IL-1 family members (IL-36 and IL-38) in atopic dermatitis, allergic asthma, and allergic rhinitis. Curr. Allergy Asthma Rep. 20, 40 (2020).
Ni, X. & Lai, Y. Keratinocyte: a trigger or an executor of psoriasis? J. Leukoc. Biol. 108, 485–491 (2020).
Patrick, G. J. et al. Epicutaneous Staphylococcus aureus induces IL-36 to enhance IgE production and ensuing allergic disease. J. Clin. Invest. 131, e143334 (2021).
Jiang, Z. et al. IL-36gamma Induced by the TLR3-SLUG-VDR axis promotes wound healing via REG3A. J. Invest. Dermatol. 137, 2620–2629 (2017).
Tortola, L. et al. Psoriasiform dermatitis is driven by IL-36-mediated DC-keratinocyte crosstalk. J. Clin. Invest. 122, 3965–3976 (2012).
Bassoy, E. Y., Towne, J. E. & Gabay, C. Regulation and function of interleukin-36 cytokines. Immunol. Rev. 281, 169–178 (2018).
Marrakchi, S. et al. Interleukin-36-receptor antagonist deficiency and generalized pustular psoriasis. N. Engl. J. Med. 365, 620–628 (2011).
Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120 (2014).
Kim, D., Langmead, B. & Salzberg, S. L. HISAT: a fast spliced aligner with low memory requirements. Nat. Methods 12, 357–360 (2015).
Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014).
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Wu, T. et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation 2, 100141 (2021).
Shen, S. et al. rMATS: robust and flexible detection of differential alternative splicing from replicate RNA-Seq data. Proc. Natl Acad. Sci. USA 111, E5593–E5601 (2014).
Cai, Y. et al. Pivotal role of dermal IL-17-producing gammadelta T cells in skin inflammation. Immunity 35, 596–610 (2011).
Keene, J. D., Komisarow, J. M. & Friedersdorf, M. B. RIP-Chip: the isolation and identification of mRNAs, microRNAs and protein components of ribonucleoprotein complexes from cell extracts. Nat. Protoc. 1, 302–307 (2006).
Ma, J. et al. iProX: an integrated proteome resource. Nucleic Acids Res. 47, D1211–D1217 (2019).
This work is supported by the National Natural Science Foundation of China (grant nos. 82071785 and 31670925 to Y.L.), the National Key Research and Development Program of China (grant no. 2016YFC0906200/2016YFC0906202 to Y.L.), the ECNU Multifunctional Platform for Innovation (011) and the Innovation Program of Shanghai Municipal Education Commission (grant no. 2019-01-07-00-07-E00046 to Y.S.). We thank C. Xu from Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences and Y. Zhang from University of California, San Diego for critical reading and helpful suggestions.
Y.L., X.N., Y.X., W.W., Y.W., B.K. and X.G. have filed provisional patents (202110117195.8 and 201910333393.0) disclosure based on the results in the paper. The remaining authors declare no competing interests.
Peer review information
Nature Immunology thanks Daniel Kaplan and Ram Savan for their contribution to the peer review of this work. Ioana Visan was the primary editor on this article and managed its editorial process and peer review in collaboration with the rest of the editorial team. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
a, GO and KEGG pathway analyses of 13,934 shared DEGs in nonlesional skin and lesional skin from AD and psoriasis in RNA-seq datasets from GEO database (GSE121212). b, RT-qPCR of CDC5L, PRPF18, DHX8, DHX15, HNRNPA1 and SRSF1 in lesional skin from AD and psoriasis. c, Immunofluorescence analysis of DDX5+ cells in healthy skin or skin from patients with squamous cell carcinoma (SCC) or basal cell carcinoma (BCC). Scale bar, 50μm. d-e, RT-qPCR of Ddx5 in lesional skin from MC903-treated (Day 0 n = 4, Day 4,8,12,16, n = 3) (d) or IMQ-treated (n = 3) (e) wild-type mice at indicated days. Data represent two independent experiments. *P < 0.05, **P < 0.01 and ****P < 0.0001. P values were analyzed by two-way ANOVA (c) or one-way ANOVA (d-e). Data are presented as mean ± s.e.m.
a, Representative photos of ears from Ddx5fl/fl(n = 5) and Ddx5∆KC mice (n = 5)16 days post-MC903 administration. b, H&E analysis of skin sections from Ddx5fl/fl and Ddx5∆KC mice treated as in (a) and quantification of epidermal thickness of skin sections. Scale bar, 50μm. c, Representative photos of dorsal skin from Ddx5fl/fl and Ddx5∆KC mice 5 days post-IMQ administration. d, H&E analysis of skin sections from Ddx5fl/fl and Ddx5∆KC mice treated as in (c) and quantification of acanthosis of skin sections. Scale bar, 50μm. e, RT-qPCR of Il23, Il17a, Ccl20, Cxcl1, Cxcl2 and S100a7 in lesional skin treated as in (c). f, Flow cytometry of CD45+ cells, CD11b+Ly6G+ neutrophils, MHCII+ APCs and CD3+γδTCR+ cells in lesional dorsal skin from IMQ-treated Ddx5fl/fl (n = 5) and Ddx5∆KC mice (n = 5). Data represent two independent experiments. *P < 0.05, ** P < 0.01, *** P < 0.001 and **** P < 0.0001. P values were evaluated by unpaired, two-tailed Student’s t-test. Data are presented as mean ± s.e.m.
a, RT-qPCR of DDX5 in keratinocytes (n = 3) treated with different doses of TNF, IL-1β, IL-36γ, IL-4, IL-13, IL-17A, IL-17B, IL-17C, IL-17E and IL-17F. Data represent three independent experiments. b, Immunoblot of DDX5 in keratinocytes treated with different doses of TNF, IL-1β, IL-36γ, IL-4, IL-13, IL-17A, IL-17B, IL-17C, IL-17E and IL-17F. Data represent three independent experiments. *P < 0.05 and **P < 0.01. P values were analyzed by one-way ANOVA. Data are presented as mean ± s.e.m.
a, Immunoblot of Ddx5 in lesional skin from IMQ-treated wild-type mice (n = 2) at day 0 to day 5. b, Immunoblot of Ddx5 and IL-17D in lesional skin from IMQ-treated wild-type (n = 3) and Il17d–/– mice (n = 3). c, Immunofluorescence analysis of Ddx5+ cells in lesional skin from IMQ-treated wild-type and Il17d–/– mice (n = 5). Scale bars, 25μm. The dotted lines indicate the edge between epidermis and dermis. d, Representative photos of ears from MC903- or IMQ-treated wild-type (n = 5) and Il17d–/– mice (n = 5). e, Flow cytometry of CD45+ cells in lesional skin from IMQ-treated wild-type (n = 5) and Il17d–/– mice (n = 5). Data represent two independent experiments. **P < 0.01 and ***P < 0.001. P values were analyzed by unpaired, two-tailed Student’s t-test. Data are presented as mean ± s.e.m.
a, Immunoblot of Ddx5 in lesional skin from IMQ-treated wild-type and Cd93–/– mice (n = 3). b, Representative photos of ears from MC903-treated (n = 7) or IMQ-treated (n = 4) wild-type and Cd93–/– mice. c, RT-qPCR of Cxcl1, Cxcl2, Ccl20, Il17a, Il23 and Il36γ in lesional skin from IMQ-treated wild-type (n = 5) and Cd93–/– mice (n = 4). d, RT-qPCR of DDX5 in NHEKs (n = 3) treated with 2.3 nM IL-17D after CD93 silencing. e, Immunoblot of p38, Akt, Smad2/3 and their phosphorylation in lesional skin from IMQ-treated wild-type (n = 5) and Cd93–/– mice (n = 5). Data represent two independent experiments. *P < 0.05, ** P < 0.01 and *** P < 0.001. P values were determined by unpaired, two-tailed Student’s t-test (e) or two-way ANOVA (d). Data are presented as mean ± s.e.m.
a-e, RT-qPCR of CCL20, CXCL1, CXCL2, CXCL6, CCL3, CCL17 and CCL22 in WT and DDX5–/– HaCaT cells (n = 3) treated with IL-17A (a), IL-36γ (b), TNF (c), IL-25 (d) or IL-4 (e) for 0-10 hours. Data represent three independent experiments. f-g, Heatmap of top 40 up-regulated genes related to leukocyte migration in MC903-treated Ddx5fl/fl (n = 3) and Ddx5∆KC mice (n = 4) (f) or IMQ-treated Ddx5fl/fl (n = 3) and Ddx5∆KC mice (n = 3) (g) by RNA-seq analysis. h-i, RT-qPCR of IL36RN and IL38 in WT and DDX5–/– HaCaT cells (n = 3) (h) or in NHEKs (n = 3) with DDX5 silencing (i). Data represent three independent experiments. j-l, Immunoblot of IL-36R in lesional skin from IMQ-treated Ddx5fl/fl and Ddx5∆KC mice (n = 3) (j) or IMQ-treated wild-type and Il17d–/–mice (n = 3) (k) or IMQ-treated wild-type and Cd93–/– mice (n = 3) (l). m, Immunoblot of IL-36R in healthy skin and lesional skin from patients with psoriasis (n = 3). Data represent two independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. n.s. no significance. P values were analyzed by two-way ANOVA (a-e) or unpaired, two-tailed Student’s t-test (f,g). Data are presented as mean ± s.e.m.
a, Analysis of different splicing events in IL-36γ-stimulated wild-type and DDX5–/– HaCaT cells (n = 3), MC903-treated Ddx5fl/fl(n = 3) and Ddx5∆KC (n = 4) mice or IMQ-treated Ddx5fl/fl and Ddx5∆KC mice (n = 3). SE, skipped exon; MXE, mutually exclusive exon; A5SS, alternative 5’ splice site; A3SS, alternative 3’ splice site; RI, retained intron. b, DNA-PAGE of splicing variants of IL4R, TSLPR, IL7R, IL17RA, IL17RC, CD6, FGFR2 and IL20RB in wild-type and DDX5–/– HaCaT cells. Data represent three independent experiments. c-d, Agarose gel analysis of IL36R transcripts amplified by PCR in NHEKs (c) or primary murine keratinocytes (d). e-f, Schematic diagrams of the structure of human (e) or murine (f) IL-36R and sIL-36R. g, Immunoblot of full-length IL-36R or sIL-36R by the antibody against Flag (left panel) or the specific antibody against murine sIL-36R (Middle panel) or the specific antibody against human sIL-36R (Right panel). h, Immunoblot of sIL-36R in cell cultures or cell lysates from HaCaT cells in which Flag-tagged human sIL-36R was overexpressed. Data represent three independent experiments. i, DNA-PAGE of IL36R and sIL36R transcripts in wild-type and DDX5–/– HaCaT cells. j, Silencing efficiency of indicated genes in NHEKs (n = 3). k, DNA-PAGE of Il36R and sIl36R transcripts in primary murine keratinocytes isolated from Ddx5fl/fl and Ddx5∆KC newborn mice. Data represent three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. n.s. no significance. P values were analyzed by two-way ANOVA (j). Data are presented as mean ± s.e.m.
Extended Data Fig. 8 SF2 binds to DDX5 and regulates IL-36R splicing and inflammatory responses in keratinocytes.
a, Mass spectrum analysis of DDX5-binding splicing factors in NHEKs. b, Co-immunoprecipitation analysis of the interaction of DDX5 and SF2. c, Immunoblot of HA-tagged or Flag-tagged proteins that bind to IL36R pre-mRNA by Flag-tagged antibody or HA-tagged antibody. d, RT-qPCR of CCL20, CXCL1, CXCL6, CCL17, CCL3, CCL11 and CCL27 in NHEKs in response to 100 ng/mL IL-36γ before and after SF2 was silenced. e, SF2-binding ESEs on exon 2, exon 3 and exon 4 of human IL1RL2 gene analyzed by the ESEfinder program (http://exon.cshl.edu/ESE). f, Schematic diagram represents IL36R pre-mRNA splicing regulated by SF2. Strong interaction of SF2 on the flanking exon 4 is responsible for skipping of the alternative exon 3 to generate sIL-36R. When the interaction of SF2 on exon 4 is weakened, either by ESE mutation in exon 4 or SF2 depletion, the alternative exon 3 is selected for IL-36R generation. g, Immunoblot of IL36R pre-mRNA-binding SF2 in nuclear extracts prepared from SF2–/– HeLa cells by Flag-tag antibody, in which IL36R reporter minigene or IL36R reporter minigene containing ESE mutation in Exon 4 was co-transfected with exogenous Flag-tagged SF2. Data represent at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. P values were determined by two-way ANOVA. Data are presented as mean ± s.e.m.
Extended Data Fig. 9 sIL-36R antagonizes IL-36R signaling to inhibit skin inflammation in keratinocyte, AD and psoriasis.
a, Flow cytometry analysis of the interaction between IL-36γ and IL-36R with or without sIL-36R overexpression in HeLa cells (n = 6). Data represent three independent experiments. b, RT-qPCR of CCL20, CXCL1, CXCL2, CXCL3, CCL3, CCL11 CCL17 and CCL27in NHEKs in response to IL-36γ after DDX5 was silenced and/or sIL-36R was overexpressed. Data represent three independent experiments. c, Representative photos of ears from MC903-treated (n = 5) or IMQ-treated (n = 5) wild-type mice injected with PBS in left ears or 1 μg per ear recombinant sIL-36R in right ears. d, RT-qPCR of Ccl20, Cxcl1, Il23 and Il17a in lesional skin from wild-type mice (n = 5) treated by IMQ as in (b). e, H&E staining of ear skin from wild-type mice treated by IMQ as in (b) and quantification of acanthosis of skin sections. f, Flow cytometry of CD45+ cells, CD11b+Ly6G+ neutrophils, CD11c+ APCs and CD3+γδTCR+ cells in ears from wild-type mice (n = 6) treated by IMQ as in (b). Data represent two independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. P values were analyzed by two-way ANOVA (b) or unpaired, two-tailed Student’s t-test (a,d,e,f).
Extended Data Fig. 10 sIL-36R restoration inhibits inflammatory responses amplified by DDX5 deficiency.
a-b, Representative photos of ears from MC903-treated (n = 4) (a) or IMQ-treated (n = 10) Ddx5∆KC mice (b) with intradermal injection with PBS in left ears or 1 μg per ear recombinant sIL-36R in right ears. c, Representative photos of ears from MC903-treated Ddx5fl/fl, Ddx5∆KC and Ddx5∆KCsIL36RTg/KC mice (n = 8). d, H&E staining of ear skin from MC903-treated Ddx5fl/fl, Ddx5∆KC and Ddx5∆KCsIL36RTg/KC mice (n = 3). e, Representative photos of ears from IMQ-treated Ddx5fl/fl, Ddx5∆KC and Ddx5∆KCsIL36RTg/KC mice (n = 8). f, H&E staining of ear skin from IMQ-treated Ddx5fl/fl, Ddx5∆KC and Ddx5∆KCsIL36RTg/KC mice (n = 3). Data represent at least two independent experiments.
About this article
Cite this article
Ni, X., Xu, Y., Wang, W. et al. IL-17D-induced inhibition of DDX5 expression in keratinocytes amplifies IL-36R-mediated skin inflammation. Nat Immunol (2022). https://doi.org/10.1038/s41590-022-01339-3
Received: 12 October 2021
Accepted: 14 September 2022
Published: 21 October 2022