Turning Lymphoma Against Itself: How a Molecular Glue Reactivates Cancer Cell Death

Edited and Approved by Stephen C. Rose, PhD, MS

Cancer drugs usually try to block, poison, or remove something the cancer cell needs. A new study in Cell explores a different strategy: use a small molecule to force two proteins together, then turn the cancer's own gene-control machinery against it. The experimental compound is described as a bivalent molecular glue because its two ends bind different proteins and stabilize a three-part complex. In laboratory models of diffuse large B-cell lymphoma, that forced partnership reactivated silenced cell-death programs and drove malignant cells toward apoptosis, the cell's organized form of self-destruction [1].

The disease matters particularly in an aging population. Diffuse large B-cell lymphoma, or DLBCL, is the most common form of non-Hodgkin lymphoma, and a population-based study reported a median age at diagnosis of 66 years [2]. Standard chemoimmunotherapy can cure many patients, but not everyone. In the phase 3 POLARIX trial, replacing vincristine with the antibody-drug conjugate polatuzumab vedotin improved two-year progression-free survival from 70.2% to 76.7%, although two-year overall survival was essentially identical between groups [3]. New mechanisms are therefore valuable, especially if they can eventually produce strong tumor selectivity without simply adding more systemic toxicity.

What molecular glue means

Molecular glues are small compounds that create or strengthen a protein-protein interaction. The best-known examples include thalidomide-related drugs, which alter the protein-recognition behavior of the cereblon ubiquitin ligase and mark selected proteins for destruction [4]. But the new Cell study is not a conventional molecular glue degrader. Its compound does not primarily send a target to the cellular recycling machinery. Instead, it brings together a lymphoma-associated transcriptional repressor and a transcriptional activator. The result is gain of function: an activity is redirected to a new genomic location.

The cancer-associated protein is BCL6, a transcription factor that helps normal germinal-center B cells temporarily suppress genes involved in DNA-damage responses, cell-cycle arrest, and apoptosis. That temporary suppression supports antibody development. In many DLBCL cells, however, BCL6 becomes part of the malignant survival program. It sits on regulatory regions of genes that could otherwise restrain or kill the cancer cell. Simply inhibiting BCL6 is one approach. The investigators instead asked whether BCL6 could be converted from a lock on cell death into a docking site for the machinery that opens those genes.

Building a two-ended switch

The researchers developed transcriptional/epigenetic chemical inducers of proximity, abbreviated TCIPs. One end of the molecule binds BCL6. The other recruits p300 and its closely related partner CBP, lysine acetyltransferases that add acetyl groups to histones and other proteins. Histone acetylation often loosens chromatin and supports gene transcription. The lead KAT-TCIP, called TCIP3, therefore acts less like a simple inhibitor and more like a molecular adapter: it places gene-activating chemistry at BCL6-controlled sites where death-promoting genes have been repressed [1].

In DLBCL cell systems, TCIP3 produced half-maximal growth inhibition at approximately 0.8 nanomolar, an extremely low concentration. Structural studies showed the compound wedged between BCL6 and a p300 binding domain, while newly formed protein-protein contacts stabilized the three-part complex. Genomic experiments indicated that p300/CBP was redistributed toward BCL6-bound regions rather than indiscriminately activated across the entire genome. Local acetylation increased, and previously repressed genes began to turn on [1]. Potency in a dish is encouraging, but it does not predict the dose, distribution, or safety of a medicine in people.

The downstream response involved more than one pathway. TCIP3 increased the pro-apoptotic protein PUMA, reduced the growth-promoting protein c-MYC, produced arrest in the G0/G1 phase of the cell cycle, and activated markers of apoptosis. Twenty-four hours of exposure was enough to initiate signals that continued toward cell death over the following one to two days. This is important mechanistically: the compound did not merely slow metabolism or produce nonspecific cellular stress. The evidence linked formation of the BCL6-TCIP3-p300/CBP complex to a coordinated transcriptional program that first stopped proliferation and then activated death [1].

What happened in the animal models

The study also moved beyond isolated cells. In DLBCL xenografts, in which human lymphoma cells were grown as tumors in mice, TCIP3 produced strong antitumor activity and reportedly eliminated established tumors within 11 days under the tested dosing schedule [1]. The compound was less toxic to tested healthy lymphocytes and fibroblasts than to susceptible lymphoma cells. Yet another result deserves equal attention: TCIP3 depleted germinal-center B cells in immunized mice. Because normal germinal centers use BCL6, this finding shows both target engagement and a plausible on-target immune consequence. Tumor selectivity was substantial in these models, not absolute.

A developing family of cancer-rewiring drugs

The KAT-TCIP did not appear from nowhere. Earlier work linked a BCL6-binding molecule to a ligand for BRD4, another transcriptional activator. That TCIP recruited BRD4 to BCL6-bound DNA, activated pro-apoptotic genes, and killed BCL6-dependent lymphoma cells, including a chemotherapy-resistant, TP53-mutant cell line [5]. A subsequent study redirected the transcriptional kinase CDK9 to the same general genomic neighborhood. Instead of globally inhibiting kinase activity, the compound relocated it to BCL6-bound genes and promoted transcription of cell-death programs [6].

These experiments suggest that the recruited partner is modular. BRD4 helps transcriptional elongation, CDK9 phosphorylates RNA polymerase II, and p300/CBP changes local acetylation and chromatin accessibility. Each route can awaken overlapping but distinct gene programs. Even the linker connecting the two binding ends matters. A 2026 medicinal-chemistry study tested 66 BCL6 TCIP analogues and found that linker rigidity and cyclic features could improve cellular selectivity, solubility, and plasma exposure in mice [7]. The active drug is therefore not merely two useful inhibitors tied together; the geometry of the induced three-protein complex can determine whether the system works.

Why this is promising - and why it is not yet a treatment

The conceptual advance is the conversion of an oncogenic dependency into a liability. A lymphoma cell with abundant, chromatin-bound BCL6 provides many docking sites for the molecular glue. The same protein that helps repress apoptosis becomes the address used to deliver an activating enzyme to those silenced genes. In principle, that can create context-dependent selectivity: cells organized around the BCL6 program receive a stronger effect than cells lacking that program. It may also allow drug designers to activate combinations of death and growth-arrest genes rather than relying on inhibition of a single survival pathway.

However, no patients received TCIP3 in this study. The evidence is preclinical and includes biochemical assays, cultured cells, genomic profiling, structural biology, and mouse experiments. Human pharmacology could differ in absorption, distribution, metabolism, immune effects, and toxicity. p300 and CBP regulate many genes in normal tissues, while BCL6 is important in normal immune biology. A therapeutic window observed in cell panels or mice must be demonstrated in formal toxicology studies and then in carefully staged clinical trials. The treatment would also need biomarkers identifying tumors that truly depend on the relevant BCL6 circuitry.

The commercial context should remain visible. The authors reported extensive biotechnology relationships, and the TCIP technology has been licensed from Stanford; several investigators are inventors or have relevant company ties [1]. Such connections are common in translational drug development and do not invalidate the data, but they strengthen the case for independent replication and transparent comparison with existing therapies. A dramatic mouse response is a reason to continue development, not a reason to skip the evidentiary steps between a laboratory discovery and a safe medicine.

For now, KAT-TCIPs offer a persuasive demonstration of a broader idea: some cancer drivers may be more useful when rewired than when merely blocked. By forcing BCL6 and p300/CBP into a productive partnership, TCIP3 turned epigenetic repression into activation of cell-cycle arrest and apoptosis. Whether that mechanism becomes a lymphoma therapy will depend on selectivity, dosing, resistance, manufacturability, and human safety. The molecular glue has assembled an intriguing scientific case. Clinical medicine must still determine whether the bond holds.

References

[1] Nix MN, Gourisankar S, Bowman KJ, et al. A bivalent molecular glue linking lysine acetyltransferases to oncogene-induced cell death. Cell. 2026 Jul 20:S0092-8674(26)00757-9. doi:10.1016/j.cell.2026.06.037.

[2] Maguire FB, Li Q, Morris CR, et al. Treatment Patterns and Survival in Older Adults with Diffuse Large B-cell Lymphoma: A Population-Based Study. J Registry Manag. 2020;47(3):135-145.

[3] Tilly H, Morschhauser F, Sehn LH, et al. Polatuzumab Vedotin in Previously Untreated Diffuse Large B-Cell Lymphoma. N Engl J Med. 2022;386(4):351-363. doi:10.1056/NEJMoa2115304.

[4] Oleinikovas V, Gainza P, Ryckmans T, Fasching B, Thoma NH. From Thalidomide to Rational Molecular Glue Design for Targeted Protein Degradation. Annu Rev Pharmacol Toxicol. 2024;64:291-312. doi:10.1146/annurev-pharmtox-022123-104147.

[5] Gourisankar S, Krokhotin A, Ji W, et al. Rewiring cancer drivers to activate apoptosis. Nature. 2023;620(7973):417-425. doi:10.1038/s41586-023-06348-2.

[6] Sarott RC, Gourisankar S, Karim B, et al. Relocalizing transcriptional kinases to activate apoptosis. Science. 2024;386(6717):eadl5361. doi:10.1126/science.adl5361.

[7] Chang FC, Shoba VM, Shyshlyk O, et al. Design of Potent and Selective BCL6 Transcriptional Chemical Inducers of Proximity through Linker Optimization. J Med Chem. 2026;69(13):15668-15684. doi:10.1021/acs.jmedchem.6c00756.