For decades, oncogenic transcription factors like BCL6 have been considered largely undruggable — their surfaces lack the convenient pockets that conventional inhibitors require. A strategy that doesn't merely block BCL6 but weaponizes it against the cancer cells that depend on it would represent a meaningful conceptual shift in how targeted therapies can be designed.
Published in Cell, this work describes the development of KAT-TCIPs — a class of bivalent small molecules engineered to simultaneously bind BCL6, the dominant oncogenic driver in diffuse large B cell lymphoma (DLBCL), and the transcriptional coactivator p300/CBP, a lysine acetyltransferase. DLBCL is the most prevalent form of non-Hodgkin lymphoma, and BCL6 actively represses gene networks involved in apoptosis, enabling malignant B cells to survive. The KAT-TCIP lead compound induces artificial proximity between p300 and BCL6, effectively turning BCL6 into an activator of the very cell-death programs it normally silences. Crystallographic resolution of the chemically induced p300-BCL6 complex reveals precise structural features explaining potency and selectivity. Notably, analogous molecules recruiting BRD4 or CDK9 instead of p300 produced distinct transcriptional signatures, suggesting modular adaptability.
This belongs to the rapidly expanding chemically induced proximity (CIP) paradigm, which includes PROTACs and molecular glues, but this application is distinctive: rather than degrading a target protein, it subverts its function. The gain-of-function mechanism — co-opting oncogenic proteins to execute tumor-suppressive programs — is conceptually elegant and could theoretically generalize to other transcription-factor-driven cancers. Key limitations to contextualize: results appear to be cell-line based, and the translation gap from crystal structures and in vitro apoptosis to preclinical animal models and ultimately human trials remains substantial. Nevertheless, the structural clarity provided by the crystallographic data, combined with the mechanistic logic of the approach, positions this as more than incremental — it is a credible proof-of-concept that transcriptional oncogenes can be converted into therapeutic liabilities.