Stanford study shows molecular glue converts cancer driver into built-in kill switch

A two-headed experimental drug has eradicated aggressive human lymphoma tumors in mice by turning one of cancer’s most important growth-driving proteins against the malignant cells it normally supports. The compound, called TCIP3, was developed by researchers at Stanford Medicine and works through a strategy that does not simply inhibit or destroy a cancer-associated protein. Instead, it redirects that protein toward a molecular system that activates genes responsible for programmed cell death. In laboratory experiments, the treatment eliminated tumors within 11 days when administered twice daily. The findings offer an unusually direct example of using a cancer’s own biological machinery as a weapon against it, while also suggesting that the same approach could eventually be adapted to autoimmune disorders and other cancers.

The study focused on diffuse large B-cell lymphoma, or DLBCL, the most common form of non-Hodgkin lymphoma. This aggressive blood cancer frequently depends on BCL6, a transcriptional repressor that controls the activity of genes involved in immune-cell growth and survival. Under normal conditions, BCL6 temporarily suppresses genes that would stop cell division or initiate apoptosis, allowing activated B cells to multiply during an immune response. Once the immune threat has passed, chemical modifications disable BCL6, allowing those protective genes to become active and eliminating unnecessary cells. In lymphoma, however, BCL6 can remain abnormally active. Its persistent gene-silencing activity prevents the malignant cells from receiving molecular instructions to stop growing or die.

BCL6 belongs to a class of proteins that have often proved difficult to target with conventional medicines. It does not function like an enzyme with a single obvious chemical pocket, nor does it act solely from within a fixed cellular compartment. Instead, it binds DNA and recruits other regulatory proteins to suppress selected genes. Previous BCL6-directed drugs have attempted either to block its interactions or to promote its degradation. Those strategies can release some of the protein’s repression, but they do not necessarily provide a strong signal that forces the cancer cell into apoptosis. Stanford researchers Gerald Crabtree, Nathanael Gray, Stephen Hinshaw and Michael Green, together with colleagues, pursued a more aggressive approach: they sought to convert BCL6 from a repressor into a trigger for the very death program it normally keeps silent.

TCIP3 was designed using a technique known as chemically induced proximity. The molecule contains two functional regions connected by a chemical linker. One region binds BCL6, while the other binds P300 or CBP, two closely related lysine acetyltransferases. These enzymes attach acetyl groups to nearby proteins, including transcriptional regulators and histones, the proteins around which DNA is packaged. By bringing P300 or CBP into close contact with BCL6, TCIP3 creates an artificial molecular neighborhood that would be unlikely to form under normal cellular conditions. The resulting complex changes the chemical state of both BCL6 and the surrounding chromatin, shifting the local environment from gene repression toward gene activation.

Acetylation is central to the compound’s activity. When P300 or CBP acetylates BCL6, the modification interferes with its ability to silence target genes. At the same time, acetylation of nearby histones weakens the interaction between those histones and DNA. The chromatin becomes more accessible, allowing transcription factors and other components of the gene-expression machinery to reach sequences that had been hidden. In lymphoma cells, this combination removes BCL6’s repression while actively opening the chromatin surrounding genes that promote apoptosis. The researchers describe the distinction as the difference between releasing a brake and pressing an accelerator: TCIP3 does not merely permit death genes to operate; it helps drive their expression to a level that overwhelms the cancer cell’s survival systems.

Structural studies revealed that TCIP3’s effectiveness depended on more than the planned interactions between its two ends and their respective protein targets. The team crystallized the molecular assembly and used X-ray diffraction to determine its atomic structure. The resulting map showed that once BCL6 and P300 or CBP were brought together, the proteins formed additional contacts with one another. These unanticipated interactions acted like molecular glue, stabilizing the entire complex. Rather than behaving as a flexible bridge that simply placed two proteins nearby, TCIP3 promoted the formation of a tightly connected three-part assembly. Biophysical measurements confirmed that these cooperative contacts substantially strengthened the interaction, helping explain why very low concentrations of the compound were sufficient to kill cultured lymphoma cells.

The investigators then tested TCIP3 in mice carrying tumors formed from human lymphoma cells. The animals received the compound twice a day, and the tumors responded rapidly. After 11 days, tumors in the treated group had disappeared, while those in untreated control animals continued to grow. The researchers reported no obvious signs of toxicity during the short treatment period. Blood tests also showed no marked increase in inflammatory signals, an important observation because the therapy affects germinal centers, specialized immune structures where B cells divide rapidly and undergo selection. These cells depend heavily on BCL6 and share biological features with the cells that become malignant in DLBCL. Although the study did not establish long-term safety or whether all tumor cells were permanently eliminated, the results demonstrate that the strategy can produce a powerful antitumor effect in a living organism.

The impact on germinal centers may also point toward applications beyond cancer. Abnormally persistent germinal-center reactions contribute to the production of self-reactive antibodies in autoimmune diseases such as rheumatoid arthritis and myasthenia gravis. By disrupting or eliminating the B cells that sustain these reactions, a carefully designed BCL6-directed molecular glue might eventually reduce the source of disease-causing antibodies. Such an application would require a delicate balance, since normal germinal-center activity is essential for generating effective immune memory after infection or vaccination. The current experiments provide an initial biological clue rather than evidence of clinical usefulness, and the consequences of repeated or prolonged treatment remain unknown.

TCIP3 is not yet a candidate for human treatment. The molecule will require additional chemical optimization to improve its drug-like properties, including stability, distribution through the body, absorption, and selectivity for diseased cells. It must also undergo testing in additional animal species and more extensive studies of toxicity, immune function and potential resistance. Cancer cells could potentially evade the therapy by altering BCL6, P300, CBP, chromatin regulators or downstream apoptotic pathways. Nevertheless, the work illustrates a broader concept in drug discovery: proteins that are difficult to inhibit may still be vulnerable to forced partnerships. By using bivalent compounds to redirect transcription factors and other regulatory proteins toward activating beneficial genes, researchers hope to develop therapies that do not merely suppress cancer’s drivers but reprogram their function. The Stanford team is now investigating whether similar molecular matchmaking can be applied to other proteins that maintain cancer or autoimmune disease.

News Publication Date: 20-Jul-2026

Web References: https://www.sciencedirect.com/science/article/abs/pii/S0092867426007579

References: Cell; Stanford Medicine researchers; Gerald Crabtree, Nathanael Gray, Stephen Hinshaw and Michael Green