New therapeutic strategy for atopic dermatitis by targeting CHI3L1/ITGA5 axis

Atopic dermatitis (AD) a chronic inflammatory skin disease that is difficult to treat despite the dis-covery of various disease targets and development of ther-apeutics. Previous clinical studies have shown that chitinase 3-like protein 1 (CHI3L1), also known as YKL-40, associated the onset and severity of AD. 1–3 a autoim-mune Our CHI3L1 Previ-ously, we first inhibition of CHI3L1 by K284-6111, a novel CHI3L1 inhibitor, amelio-rates AD-like skin inflammation. 5 Here, we further investigated the role of CHI3L1 in AD pathogenesis and its underlying mechanism using CHI3L1 knockout (KO) mice and a CHI3L1-blocking antibody (CHI3L1-Ab), and based on out-comes, propose that a CHI3L1-targeted therapeutic strategy can be used to treat AD.

mice ( Figure S2). Moreover, expression of genes encoding inflammatory cytokines and chemokines induced by treatment with a mixture of tumour necrosis factor (TNF)α and interferon (IFN)-γ, an inflammatory stimulant that induces AD-like features, was inhibited by CHI3L1 knockdown, whereas it was up-regulated by CHI3L1 overexpression in HaCaT cells ( Figure S3). Taken together, these findings indicate that CHI3L1 greatly contributes to the AD pathogenesis.
To explore the underlying mechanism of CHI3L1 in AD pathogenesis, we performed gene-network analysis using Humanbase, an interactive platform of data-driven predictions of gene regulation and interaction, which showed that CHI3L1 is linked to numerous genes of interest, including integrin beta2 (ITGB2) (Figure 2A). Integrins are transmembrane adhesion receptors with important roles in biological and pathological processes. 6 Previous studies have shown that integrin alpha (ITGA) is overexpressed in atopic skin, 7 and CHI3L1-mediated biological and pathological effects are through interactions with integrins. 8 To determine the association between CHI3L1 and integrins including ITGB2, ITGA2, ITGA5 and ITGA6, we compared the mRNA expression levels of these integrins in the skin of the PA-induced AD model. Expression of all tested integrins increased in WT mice but decreased in CHI3L1-KO mice; particularly, ITGA5 expressions were significantly down-regulated by CHI3L1-KO ( Figure 2B,C). Additionally, gene silencing and overexpression experiments in HaCaT cells demonstrated that ITGA5 expression was more affected by CHI3L1 compared to ITGA6 ( Figure 2D,E). Differences in protein expression of ITGA5 were further confirmed in TNF-α/IFN-γ-treated HaCaT cells ( Figure 2F) and skin tissues ( Figure 2G). We further analysed the gene expression levels of pro-inflammatory mediators after knocking down ITGA5 expression in TNFα/IFN-γ-treated HaCaT cells and found that the mRNA expressions of IL-1β, IL-6, TSLP and CCL22 were significantly decreased ( Figure 2H).  Notably, transcription factor prediction analysis showed that ITGA5 contains 13 putative nuclear factor (NF)-κB binding sites ( Figure S4A). Interestingly, ITGA5 expression was inhibited by not only NF-κB p65 siRNA in TNF-α/IFN-γ-treated HaCaT cells ( Figure S4B), but also in CHI3L1-overexpressing cells treated with an NF-κB inhibitor ( Figure S4C), indicating that NF-κB is a key transcription factor regulating ITGA5 expression. As NF-κB is a critical regulator in AD, 9 and CHI3L1 is known to bind to the RAGE (receptor for advanced glycation end-products) receptor to contribute to NF-κB activation, we investigated whether CHI3L1 modulates NF-κB signaling. 10 CHI3L1 overexpression promoted NF-κB signaling activation in resting and TNF-α/IFN-γ-treated HaCaT cells (Figure S5A,B). In contrast, CHI3L1 knockdown reduced both TNF-α/IFN-γ-induced phosphorylation of IκBα and nuclear translocation of p50/p65 ( Figure S5C,D). In addition, CHI3L1 knockdown together with the NF-κB inhibitor synergistically inhibited IL-1β, IL-6, TSLP and CCL22 production ( Figure S5E). Notably, PA-induced NF-κB activation in the skin tissues decreased in CHI3L1-KO mice compared to in WT mice ( Figure S5F,G). These results suggest that CHI3L1 induces inflammatory responses by activating NF-κB signaling, which may lead to AD-related skin inflammation. Collectively, our findings suggest that CHI3L1-related F I G U R E 4 Clinical relationship between Chi3L1 and ITGA5 in patients with atopic dermatitis (AD). (A and B) Serum levels and receiver operating characteristic curve of chitinase 3-like protein 1 (CHI3L1) (A) and ITGA5 (B) in patients with AD and healthy controls. n = 20. (C) Correlation between CHI3L1 and ITGA5. R value indicates Spearman's correlation co-efficient. Data are expressed as the mean ± SD. *p < .05, and ***p < .001 AD pathogenesis is associated with the regulation of the NF-κB/ITGA5 axis.
After confirming that CHI3L1 regulates AD-related skin inflammation, we examined the therapeutic effect of CHI3L1-antibody (Ab). Administration of a CHI3L1-Ab alleviated PA-induced AD development and the inflammatory response and suppressed the production of CHI3L1 and ITGA5 and activation of NF-κB ( Figure 3A-D). Furthermore, the analysis of the reconstructed human skin (RHS) tissue AD model confirmed that CHI3L1-Ab treatment reduced skin inflammation by inhibiting the CHI3L1/NF-κB/ITGA5 axis ( Figure 3E-I). Additionally, we demonstrated that ITGA5-Ab treatment reduced epidermal thickness in the same RHS AD model ( Figure S6).
Finally, we found that the serum levels of CHI3L1, ITGA5 and other AD biomarkers were significantly increased in sera from patients with AD ( Figure 4A,B [left] and Figure S7A). Receiver operating characteristic curve analysis indicated that CHI3L1 and ITGA5 are more suitable diagnostic markers for AD compared to other biomarkers ( Figure 4A,B [right] and Figure S7B). Additionally, Spearman's correlation analysis showed that serum level of CHI3L1 significantly correlated with the serum level of ITGA5 ( Figure 4C). These results suggest that CHI3L1 and ITGA5 are potent diagnostic and prognostic biomarkers for AD.
In conclusion, CHI3L1 mediates AD-like skin inflammation by regulating NF-κB-dependent ITGA5 expression. Furthermore, the inhibition of CHI3L1 by CHI3L1-Ab suppresses AD-like symptoms and inflammatory responses by regulating CHI3L1, the NF-κB/ITGA5 axis. Our findings provide experimental rationale for developing therapeutic strategies for AD by targeting CHI3L1.

A C K N O W L E D G E M E N T S
This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korean government (MSIP) (grant numbers: MRC, 2017R1A5A2015541 and NRF-2020R1F1A1073308).