Abstract
Epigenetic modifications play a crucial role in regulating essential cellular processes in biological organisms. As an important epigenetic regulator, lysine-specific demethylase 1 (LSD1) is actively involved in the process of histone demethylation. Recent studies have highlighted the critical function of LSD1 in various malignancies, including colorectal cancer, acute myeloid leukaemia, and small cell lung cancer, among others. LSD1 promotes the progression of these diseases via its demethylase activity and additional non enzymatic roles, making it an attractive target for therapeutic strategies.This thesis describes attempts to develop LSD1-targeting treatments, including inhibitors and degraders. In the initial phase of the study, virtual screening strategies were integrated with biological assays for the purpose of identifying potential inhibitors. Initially, molecular docking simulation was combined with machine learning to identify potential reversible LSD1 inhibitors. Through the integration of structural information and biological activity data, a number of potential LSD1 inhibitors were identified. The inhibitory activities of these compounds were subsequently validated through enzyme inhibition assays. Besides, the results showed that the combination of ligand-based and structure-based virtual screening was superior in discovering highly active compounds. A drug candidate demonstrating notable inhibitory activity was found by improving the sequential virtual screening strategy in conjunction with molecular dynamics simulations. This compound showed promising stability in molecular dynamics study, but its cytotoxic effect was less-than-ideal. This discrepancy may result from factors such as poor cell membrane permeability, insufficient metabolic stability, and the impact of the intracellular environment.
In addition, given the diverse functionalities of LSD1, which extend beyond its role as a demethylase, research efforts have shifted towards the development of proteolysis-targeting chimera (PROTAC) aimed at inducing the degradation of LSD1. Utilizing the known LSD1 inhibitor SP-2577, a series of PROTACs were designed and systematically screened. Among the compounds studied, 2-30 exhibited the most pronounced anti-tumour effects, evidenced by its significantly lower IC50 value compared to SP-2577. Additionally, 2-30 effectively facilitated the degradation of LSD1 via the ubiquitin-proteasome system (UPS) pathway. Subsequent cellular assays and tumour xenograft studies conducted in murine models demonstrated that compound 2-30 exhibited markedly greater anti-tumour efficacy in comparison to SP-2577. The RNA-seq analysis revealed significant differences in biological effects between the compounds 2-30 and SP-2577. In particular, the variations in ferroptosis pathways, peroxisomal functions, and immune evasion mechanisms may explain the stronger anti-tumour potency of compound 2-30 relative to SP-2577. Overall, an active PROTAC molecule targeting LSD1 was identified, and its anti-tumour activities were systematically evaluated. The potential underlying mechanisms of anti-tumour activities were also explored, which provided new insights for LSD1-targeting therapy in the future.
This thesis presents a refined virtual screening methodology for the identification of LSD1 inhibitors, integrating approaches such as machine learning, molecular docking, molecular dynamics simulations, and biological assessments. Notably, this study introduces and provides evidence for the concept of LSD1-targeting PROTACs for the first time. The development of compound 2-30 as a capable degrader emphasizes an innovative treatment method for the targeted inhibition of LSD1 in cancer care.
| Date of Award | 15 Oct 2025 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Bencan Tang (Supervisor), Binjie Hu (Supervisor), Michael J. Stocks (Supervisor) & Hainam Do (Supervisor) |