Gerald Crabtree is a developmental biologist, the kind of scientist who has spent decades thinking carefully about how cells live and die — and specifically about apoptosis, the biological process by which the body systematically eliminates cells it no longer needs. Every day, roughly 60 billion cells in the human body execute a kind of controlled self-demolition, a process so orderly and so essential that without it, tissue balance would collapse. Cancer, in its most fundamental character, is what happens when certain cells learn to ignore that instruction.
They stop dying. They keep dividing. And the longer they do it, the harder they become to stop. Crabtree’s question — one he apparently spent considerable time sitting with — was whether the very machinery cancer uses to evade death could be reversed and made into the mechanism of its destruction. The answer, at least in a laboratory at Stanford, appears to be yes.
| Category | Details |
|---|---|
| Stanford Breakthrough | Molecular glue technology developed at Stanford links two proteins — BCL6 and CDK9 — forcing cancer cells to reactivate their own self-destruct (apoptosis) genes; conceived by developmental biologist Gerald Crabtree |
| How the Molecule Works | BCL6 normally deactivates apoptosis genes, granting cancer cells a form of immortality; when artificially linked to CDK9, this same protein reverses course and triggers cell death — turning cancer’s survival mechanism against itself |
| Laboratory Results (Stanford) | Tested against 859 different cancer cell lines; demonstrated exceptional specificity — only lymphoma cells were killed; healthy cells were spared; no notable side effects observed in healthy mice during trials |
| Primary Target (Stanford) | Diffuse large B-cell lymphoma (DLBCL) — a blood cancer; researchers plan to develop similar molecules targeting other oncogenes in future studies |
| KAIST Breakthrough | Korea Advanced Institute of Science and Technology (KAIST) developed lipid nanoparticles injected directly into tumors; absorbed by macrophages already inside the tumor, which then reprogram themselves into cancer-killing CAR-macrophages — without being removed from the body |
| Why Solid Tumors Are Different | Gastric, lung, and liver solid tumors form dense masses that block immune cell entry; most immunotherapies fail here; macrophages — which can physically engulf cancer cells — are uniquely suited to penetrate these environments |
| KAIST Animal Study Results | Significant reduction in tumor growth in melanoma models; immune response extended beyond the injected tumor, suggesting potential for body-wide immune activation — a finding that raises cautious optimism for metastatic cancer treatment |
| Key Advantage Over Chemotherapy | Both approaches target cancer cells specifically without broadly poisoning healthy tissue — the fundamental limitation of conventional chemotherapy and many traditional targeted therapies |
| What Is Apoptosis | A natural biological process eliminating ~60 billion cells daily in the human body; cancer cells survive by disabling it; reactivating apoptosis selectively in cancer cells is the core principle behind both Stanford and KAIST approaches |
| Research Publication | KAIST study published in ACS Nano (Vol. 19, 2025); led by Prof. Ji-Ho Park and first author Jun-Hee Han, Department of Bio and Brain Engineering, KAIST |
The approach involves a “molecular glue” — a designed molecule that physically links two proteins, BCL6 and CDK9, which normally operate independently inside cells. BCL6 is a protein found commonly in lymphomas, and its normal function includes deactivating the genes responsible for triggering apoptosis — essentially, it helps cancer cells stay permanently switched to “survive.” CDK9 does something different. When researchers forced the two proteins together using this molecular connector, something unexpected but not entirely surprising happened: BCL6, now locked into a new configuration, began doing the opposite of what it usually does. Instead of suppressing the cell-death signal, it activated it. The cancer cell’s own proteins became the instrument of its destruction.
The laboratory results were tested against 859 different cancer cell lines. The molecule worked with what researchers described as exceptional specificity — killing lymphoma cells while leaving healthy cells largely untouched. In trials involving healthy mice, no significant side effects emerged, though scientists did note that the treatment eliminated a specific subset of healthy B cells in the animals, a type of immune cell, without apparent major consequence.
That distinction matters because it’s the kind of collateral effect that will require careful scrutiny as the research moves toward human trials. But it is an entirely different order of damage than chemotherapy, which saturates the body with toxins that attack dividing cells indiscriminately, making patients violently ill while fighting the disease. There’s a feeling, reading through this research, that oncology is slowly arriving somewhere it has been pointing toward for a long time — at treatments that can read the difference between a cancer cell and the tissue around it.
Meanwhile, roughly six thousand miles away at the Korea Advanced Institute of Science and Technology, a research team led by Professor Ji-Ho Park was working on a different version of the same underlying instinct: that cancer’s defeat might come from within the tumor itself, not from chemicals introduced from outside. Tumors are, it turns out, full of immune cells. Macrophages — large, aggressive immune cells capable of physically engulfing pathogens and damaged cells — gather naturally inside and around tumors in significant numbers. They are, in theory, exactly what you would want fighting cancer. The problem is that the tumor environment suppresses them. It chemically signals the macrophages to stand down, converting what should be a natural line of defense into a passive bystander population that the cancer learns to coexist with and even exploit.
Park’s team designed lipid nanoparticles — tiny molecular packages that macrophages readily absorb — loaded with mRNA encoding cancer-recognition proteins and an immune-boosting compound. When injected directly into a tumor, the nanoparticles were taken up by the macrophages already present. Those macrophages then began producing CAR proteins — the cancer-recognizing machinery — on their own, transforming into what the researchers called “enhanced CAR-macrophages” without ever being removed from the body, modified in a lab, and reintroduced. The significance of that distinction is considerable. Current CAR-macrophage therapies require extracting immune cells from a patient, genetically reprogramming them outside the body, and infusing them back — a process that is slow, expensive, and logistically complex enough that it remains largely inaccessible. Park’s approach skips all of that. The factory is already inside the tumor. The method just gives it better instructions.
In animal models of melanoma, one of the more aggressive and difficult skin cancers, tumor growth was significantly reduced. More intriguingly, the immune response appeared to extend beyond the site of injection — suggesting that the reprogrammed macrophages may be activating broader immune pathways that could, in principle, respond to cancer elsewhere in the body. It’s still too early to know whether that effect translates reliably to human patients or to different cancer types, and the gap between animal studies and clinical results in oncology is long and historically full of disappointments. But the consistency of the finding across multiple approaches — the molecular glue at Stanford, the in-situ reprogramming at KAIST, the Trojan horse strategies being explored at Mount Sinai — suggests something is shifting in how researchers are thinking about what cancer actually is and where its vulnerabilities lie.
For most of the history of cancer treatment, the dominant logic was subtraction: cut it out, burn it with radiation, poison it with chemotherapy. Those approaches saved and extended lives and continue to do so. But they were blunt in ways that caused enormous suffering, and they worked only as long as cancer couldn’t adapt — which it often could. The emerging logic is something different. It starts from the observation that the body already contains the machinery to destroy cancer, and that the disease’s primary achievement is not strength but deception. It convinces immune cells to stand down. It hijacks proteins that were supposed to trigger its death. It builds a local environment that protects it from attack. If researchers can consistently reverse those deceptions — at the molecular level, within the tumor itself — the treatment becomes a matter of restoring what the body was already trying to do, rather than introducing an outside force powerful enough to overwhelm both the cancer and the patient simultaneously.
It’s possible that neither the Stanford molecule nor the KAIST injection will survive the translation to clinical use intact. The history of oncology is lined with laboratory findings that generated genuine excitement before encountering complications in human trials that nobody anticipated. That caveat needs to sit alongside the results, not buried beneath them. But there’s a convergence happening across multiple research centers — different mechanisms, different targets, the same underlying philosophy — and that convergence is harder to dismiss than any single study taken alone.
