Trogenix / Steve Pollard: synthetic super-enhancers for targeted viral immunotherapy in glioblastoma
Nature has published a paper from Trogenix and Steven Pollard’s team showing a viral therapy for glioblastoma switched on primarily in cells with the right tumor program
On 8 April, Nature published a study on synthetic super-enhancers, synthetic DNA switches that read a cell’s internal state. The authors tuned them so the therapy turns on mainly in stem-like glioblastoma cells and works much more weakly in normal brain tissue. In the preclinical arm, that cleared tumors in most mice and protected against regrowth after repeat re-implantation.
The fight against glioblastoma runs into one core problem: after surgery, radiation, and chemotherapy, a small population of cells remains, and the tumor rebuilds itself from there. Pollard’s team set out to hit that population directly. Instead of the usual rule, “deliver the toxin to the right place,” they built a different one: switch the therapy on only in the cell with the right internal program. That principle matters more than any single construct: the treatment is no longer reading the cell’s address, but its state. That is especially compelling in diseases driven not by one mutation, but by a stable cellular role the tumor keeps recreating again and again.
The synthetic super-enhancers in this работе are artificially assembled stretches of DNA that act as a gene-level logic switch. The authors built them from enhancer fragments and tuned them to cells where SOX2 and SOX9 are active. These are factors that keep glioblastoma in a stem-like mode, its most persistent and relapse-prone state.
What follows is careful engineering. For delivery, they used AAV1, a viral vector, essentially the packaging for a genetic instruction. В конце марта на канале был человеческий anti-aging протокол с AAV как транспортом, and в другом посте мы разбирали AAV-платформу, где главная опасность сидела во внецелевом чтении генетического кода. Here the bottleneck is different: not reading the wrong codon, but switching on in the wrong cell. The authors put two payloads under control of the switch: HSV-TK, an enzyme that makes the cell vulnerable to ganciclovir, and IL-12, a strong signal for the immune system. The tumor cell first triggers a cascade that makes it vulnerable, and then also calls in the immune system to finish clearing what the vector did not reach.
The numbers are better taken without hype. In an aggressive mouse model with tumors growing directly in the brain, the authors gave a single intratumoral dose of AAV1-SSE-7-HSV-TK-IL-12, then administered ganciclovir for 20 days and achieved complete tumor clearance in 20 of 24 mice. Over 11 months of follow-up, they saw no meaningful toxicity. At 5 months, surviving animals were re-implanted with tumor cells, and in 10 of 10 the tumor did not regrow. This is not proof for humans, but it is already more than the usual mouse picture of “the tumor shrank for a while.”
The human layer is no less important. The authors tested selectivity in primary glioblastoma samples, fresh tumor slices, and normal cortex. About 90% of tumor cells that turned on the reporter constructs also carried SOX2. The switch was tied not to a random marker, but to a specific cellular program. In a separate analysis, roughly 20% activated cells were enough for the immune component to mop up the rest of the tumor mass.
The next cold test is already in the registry. On ClinicalTrials.gov, ADePT for TGX-007 is listed as not yet recruiting: about 68 patients, Edinburgh and Ohio State, delivery directly into the tumor before standard surgery, followed by 14-21 day valaciclovir. That is an important qualification: the clinic will not be testing a universal shot “for brain cancer,” but a local therapy tied to a specific surgical pathway. That is where we will see how much of the mouse result comes from a real mechanism and how much from laboratory staging. For now, this is not a “cure for glioblastoma,” but an attempt to make therapy read cellular identity and fire where the disease itself is sending the right signal. If that principle survives the clinic, the main result may not even be this current mouse win, but a new template: target not a single mutation, but the cellular role from which the disease keeps rebuilding itself.