Single‐cell transcriptomic analysis of GPP patients treated with IL‐12/23 or IL‐17A blockade

The interleukin (IL)-36 cytokine axis is critical in generalized pustular psoriasis (GPP).1 IL-36α, IL-36β and IL-36γ are involved in the activation of pro-inflammatory signalling, while IL-36 receptor antagonist (encoded by IL36RN) inhibits downstream IL-36R signalling, and recessive mutations in IL36RN are the most commonly associated genetic mutation in GPP.2-4 Two sibling GPP patients carrying mutations in IL36RN (c.28C>T; p.Arg10X and C.115+6T>C; p.Arg10ArgfsX1), GPP1, an 11-year-old boy, and GPP2, a 4-year-old girl, were enrolled. Full thickness skin (4-mm punch biopsies) was obtained from the abdomen or hip and enzymatically dissociated. The molecular changes in the tissue were dissected using single-cell RNA sequencing (scRNAseq) before and after off-label treatment with IL-12/23 blockade (ustekinumab) or IL-17A blockade (secukinumab), respectively, after which both patients exhibited clear skin (Figure 1a). The two patients exhibited varied baseline expression patterns of cytokines and receptors associated with GPP, although they were siblings with similar genetic backgrounds and had the same genetic mutation in IL36RN (Figure 1b). This may be attributed to differences in age, sex or the fluctuating nature of GPP. IL36G, expressed in granular and spinous keratinocytes, was downregulated in both patients after treatment (Figure 1c). IL17C also tended to decrease in granular keratinocytes after treatment (data not shown). Of note, the proportion of granular and spinous keratinocytes decreased post-treatment (Figure 1d), which appears to be correlated with clinical response. These results indicate that although IL-36 was not directly targeted, both treatments result in reduced IL36G, suggesting that there is a pathogenic loop connecting IL-36/IL-17/IL-12/23 in GPP. We identified differentially expressed genes (DEGs) that were altered in both patients after treatment, compared with before treatment, and conducted gene set enrichment analysis (GSEA). The neutrophil degranulation pathway was downregulated in granular and spinous keratinocytes after treatment (Figure 1e). Of the DEGs, CXCL1, CXCL8 and CCL20, chemokines related to neutrophil recruitment and the Th17 pathway in GPP, were decreased after treatment (Figure 1f). In addition, CCL27, a crucial regulator of immune homeostasis in the skin, was increased post-treatment (Figure 1f). Interestingly, CCR10, the receptor of CCL27, increased in post-treatment regulatory T cells (Tregs) (Figure 1g), suggesting that increased CCR10-dependent Treg recruitment post-treatment may partly explain the clinical response. To further identify the effects of systemic administration of IL-12/23 or IL-17A blockade on keratinocytes and examine keratinocytes for changes in particular inflammatory signatures, we defined cytokine scores using DEGs reported to be elevated in keratinocytes when stimulated with IL-1β, TNF, IFNA, IL-13, IL-17A or IL-36G in vitro.5 As expected, the IL17A and IL36G scores were successfully dampened post-treatment (Figure 1h). Other module scores, including IL1B, TNF and IL13 scores, which were slightly higher in GPP2 than GPP1 before treatment, were similarly downregulated post-treatment. In addition, the IFNA score was elevated only in GPP1 before treatment, suggesting that the two patients exhibited some varied baseline transcriptomic profiles. While each treatment targeted a specific cytokine, both treatments affected multiple cytokine pathways, suggesting an interconnected pathogenic circuit involving these cytokines. The frequency of Tregs slightly increased, while Th17 cells decreased post-treatment (Figure 2a). T helper type responses were differentially affected depending on the treatment strategy. The Th17 score was decreased in both patients, while the Th22 score was decreased in GPP1, and the Th1 score was decreased in GPP2 after treatment (Figure 2b). In T cells, several cytokines were decreased, while in myeloid cells, inflammation- and costimulation-related genes were decreased after treatment (Figure 2c). Receptor-ligand interaction analysis revealed a lower number and strength of interactions post-treatment (Figure 2d). Of the pathways, ANGPTL signalling was downregulated in keratinocyte subsets after treatment, and the expression of ANGPTL4, which is known to promote keratinocyte proliferation and inflammatory responses in psoriasis vulgaris,6 was downregulated after treatment. The ANNEXIN pathway and ANXA1 expression were upregulated (Figure 2f). ANXA1 may have anti-inflammatory effects, restricting Th17 cells.7 In summary, after treatment with IL-12/23 blockade or IL-17A blockade, we discovered that there is attenuation of IL36G, CCL20, IL17A, IL17C, CXCL1, CXCL8 and CSF2 leading to a reduction in inflammatory signals, including Th17 signatures and downregulation of neutrophil responses. Therefore, by profiling paediatric GPP lesions before and after treatment using scRNAseq, we provide insight into the underlying molecular mechanism of clinical improvement after IL-12/23 or IL-17A blockade, which feeds into the circular node of IL-36/IL-17/IL-12/23 to restore normality. This work was supported by the New Faculty Startup Fund (800-20240280) from Seoul National University. This research was supported and funded by SNUH Lee Kun-hee Child Cancer & Rare Disease Project, Republic of Korea (grant number: 22B-001-0100). None declared. This study was approved by the Institutional Review Board of Seoul National University Hospital (IRB No.: 2210-035-1368). The patients in this manuscript have given written informed consent to the publication of their case details. The data that support the findings of this study are available from the corresponding author upon reasonable request.