发现一种新型选择性脑渗透CDK4抑制剂用于靶向癌症治疗

In breast and other malignancies, CDK4 overactivation commonly occurs as a result of sustained endocrine and mitogenic inputs, dysregulated cyclin D1 expression, or the genetic loss or functional impairment of INK4 family inhibitors, or loss of endogenous inhibitors such as p16. Approved CDK4/6 inhibitors (palbociclib, ribociclib, and abemaciclib) are associated with hematologic toxicity, including neutropenia, linked to CDK6 rather than CDK4 inhibition. Since these drugs block CDK6, drug dosing is restricted by bone marrow suppression rather than anti-tumor efficacy. HR+ breast cancer cells are more dependent on CDK4 than CDK6 for proliferation, whereas normal hematopoiesis relies predominantly on CDK6. Loss of CDK4 alone does not impair blood cell development in mice, while CDK6 and its partner cyclin D3 play a key role in stem cell activation, erythroid and myeloid differentiation, and neutrophil maintenance. This divergence establishes a therapeutic window: Selective blockade of CDK4 could suppress tumor growth while sparing bone marrow function. Clinical benefits in patients with brain metastases have been documented. However, dose-limiting toxicity may prevent the gain in overall outcome. CDK4-selective inhibition represents a strategic evolution in targeting cell-cycle dysregulation. By sparing CDK6, agents aim to overcome the main toxicity barrier of dual CDK4/6 inhibitors, enabling more potent and sustained CDK4 blockade. Here, we describe AL-0433, a novel, selective brain penetrant CDK4 inhibitor with excellent in vitro and in vivo activity. Methods: AL-0433 was developed using resistance-aware, structure-based design, via AI-enabled virtual screening and medicinal chemistry at Expert Systems. Potency was evaluated against CDK1, 2, 4, 6, 9, and GSK3b in biochemical assays, and cytotoxicity was assessed in breast, ovarian, and glioblastoma cell lines. Pharmacokinetic (PK) properties were characterized in murines and dogs. Anti-tumor efficacy was tested in MCF-7 xenograft models as a single agent and in combination with other anti-tumor agents. Results: AL-0433 exhibited low-nanomolar CDK4 inhibition (2.7 nM) with 30-fold selectivity over CDK6, and 1000-fold selectivity over CDK1, CDK2, and CDK9. AL-0433 showed potent cytotoxicity in breast, ovarian, and glioblastoma cells (20-950 nM). Mechanism of action studies in MCF-7 cells demonstrated inhibition of pRB phosphorylation. AL-0433 demonstrated excellent oral bioavailability with high blood-brain barrier (BBB) permeability, making it particularly advantageous for treating both primary and metastatic tumors, as well as brain tumors such as glioblastoma. Conclusion: AL-0433 is a promising, AI-designed, selective CDK4 inhibitor with broad activity across breast, ovarian, and glioblastoma cell lines. Its favorable PK profile, preclinical efficacy, and rational design strategy support further clinical development