AI and CRISPR Identify Psoriasis Treatment Targets, Paving the Way for Topical Therapies
Background
Psoriasis is a chronic inflammatory skin disease affecting over 125 million people worldwide. It is characterized by excessive proliferation of epidermal keratinocytes, resulting in red rashes and white scales. Current psoriasis treatments primarily involve biologics that block the signaling pathways of interleukin-17 (IL-17) and its receptor, interleukin-17 receptor A (IL-17RA). While biologics administered via injection have shown excellent efficacy, they also have limitations. These include the potential for systemic side effects and high treatment costs. Furthermore, they are not suitable for patients who are unwilling or unable to receive injections. The development of topical small-molecule therapies is being explored as an alternative. To achieve this, it is essential to thoroughly elucidate the mechanisms by which IL17RA is regulated within keratinocytes, the key cells involved in psoriasis lesions.
Key Findings
The researchers conducted a genome-wide CRISPR knockout screen using primary human epidermal keratinocytes. This was done to identify genes that regulate the level of IL17RA expressed on the cell surface. They also employed a multi-faceted analysis approach by integrating VirtualCRISPR, a large language model (LLM) framework trained on functional genomics data. By combining the predictive power of AI with actual gene editing experimental data, they were able to precisely select key regulatory genes from a large number of candidates. This process led to the identification of 5-lipoxygenase (ALOX5) and the oxytocin receptor (OXTR), two regulators that were previously not well-known to be associated with IL17RA regulation. These two regulatory genes operate through distinct mechanisms within the cell. To validate the therapeutic effect, the researchers induced dermatitis in mice by administering imiquimod, a psoriasis-inducing agent. Subsequently, they applied zileuton, an ALOX5 inhibitor, and cligosiban, an OXTR antagonist, directly to the inflamed areas of the mice. The results showed that these small-molecule compounds strongly suppressed skin inflammation. This level of inflammation reduction is comparable to that achieved with systemic anti-IL17RA antibody treatments.
Significance and Prospects
This study demonstrates that combining gene editing screening with AI technology can significantly reduce the time required for discovering new drug candidates. It presents a new paradigm for rapidly identifying targets from vast amounts of genetic data that can be used as actual drug targets. In particular, the discovery of topical psoriasis treatment candidates offers a new treatment option for patients who experience limitations with existing systemic injection therapies. However, it is important to note that success in animal models does not guarantee success in human clinical trials. Further research is needed to address and overcome the subtle physiological differences between mouse and human skin. Furthermore, the development of formulation technologies that can help the topically applied small-molecule compounds penetrate the thickened psoriatic skin barrier and reliably reach the keratinocytes is also required. Thorough validation through rigorous testing before clinical entry is essential to ensure both safety and efficacy.
Psoriasis affects over 125 million people globally. Biologics targeting the IL-17/IL-17RA axis are effective but require systemic administration, are costly, and are unsuitable for some patients. Developing topical small-molecule alternatives requires a better understanding of how IL-17 receptor A (IL17RA) is regulated in keratinocytes, the principal effector cells of psoriatic lesions. Here, we report a genome-wide CRISPR knockout screen for regulators of surface IL17RA in primary human epidermal keratinocytes. We prioritize hits using experimental enrichment together with VirtualCRISPR, a language-model framework trained on functional-genomics data, and validate two regulators with minimal prior connection to IL17RA: 5-lipoxygenase (ALOX5) and the oxytocin receptor (OXTR), which act through distinct cell-intrinsic mechanisms. Topical zileuton, an ALOX5 inhibitor, and cligosiban, an OXTR antagonist, suppress imiquimod-induced psoriasiform dermatitis in mice, mirroring systemic anti-IL17RA antibody efficacy. By linking AI- guided selection to genetic perturbation screening, this study provides an efficient route from candidate gene nomination to biological validation and therapeutic discovery.
Topical small-molecule compound therapies are expected to improve the quality of life for psoriasis patients. Patients will no longer need to regularly visit a hospital to receive expensive injections; instead, they can manage their symptoms by applying medication themselves at home. This is especially beneficial for patients with mild psoriasis who may be hesitant to use biologics due to concerns about systemic side effects, as it can provide a safe first-line treatment option. From a pharmaceutical industry perspective, small-molecule compounds have a lower production cost, which can significantly reduce the burden on healthcare budgets. Furthermore, the AI-based target discovery technique using VirtualCRISPR is not limited to psoriasis. This platform technology can be readily applied to the development of topical treatments for various other difficult-to-treat immune-mediated skin diseases, such as atopic dermatitis and alopecia areata, thereby accelerating the development of new drug pipelines.