Gerard Boink has spent two decades studying how a single gene could reprogram the heart—and now he's dismantling everything the field believed about it. The cardiologist and principal investigator at Amsterdam UMC has just published findings that overturn more than a decade of assumptions about TBX18, a transcription factor once hailed as a potential biological pacemaker therapy. The results, published in the Journal of Clinical Investigation, show that TBX18 doesn't work—and that a different ion channel, Hcn2, actually does.
For years, high-impact papers had reported that TBX18 could reprogram ordinary heart muscle cells into pacemaker cells. But Boink's team decided to rigorously re-examine that claim using adeno-associated virus (AAV) vectors and detailed electrophysiology. What they discovered challenges a foundational assumption in regenerative cardiology.
When researchers used conventional high-level TBX18 expression, it was brutally toxic. The protein caused severe myocardial fibrosis and scarring in mouse hearts, while control groups showed no such damage. "Supraphysiological TBX18 expression is highly toxic for cardiomyocytes," Boink explains. But toxicity alone doesn't prove whether the gene actually works—it might have been the poison, not the ineffectiveness, causing problems.
To isolate TBX18's true function, the team engineered a new AAV cassette that reduced TBX18 protein levels to roughly 1% of conventional expression. This clever approach eliminated the fibrosis entirely while preserving the gene's ability to suppress known targets like Connexin43. Now they could ask: does TBX18 actually create pacemaker cells?
The answer was no. Even at non-toxic levels, cardiomyocytes never acquired genuine pacemaker properties. TBX18 suppressed multiple working-myocyte genes and created abnormal action potentials, but it failed to activate key pacemaker genes. It didn't induce Hcn4 protein or the critical pacemaker current If. "Even at realistic, non-toxic expression levels, you do not get pacemaker activity," Boink says. "Instead of providing a therapy, you risk arrhythmia by ion channel dysregulation and electrical instability."
The team uncovered another problem: earlier studies using adenoviral vectors, not AAV, had likely been fooled by the vector itself. In a rat model of complete atrioventricular block, both adenoviral TBX18 and adenoviral control produced similar ectopic pacing and extensive local fibrosis—suggesting the vector's inflammation, not the gene's function, had created the illusion of pacemaker activity.
Here's where the story pivots toward hope. When researchers used AAV to deliver Hcn2, an ion channel that helps generate pacemaker rhythm, it worked beautifully. In the same rat complete AV-block model, Hcn2 produced robust, autonomically responsive ventricular pacing. Adding TBX18 alongside Hcn2 provided no benefit, confirming that TBX18 cannot enhance gene-therapy-based pacemakers either.
The implications are profound. "An efficient biological pacemaker can be created with Hcn2 alone," Boink says. This finding has direct relevance for developing gene-therapy pacemakers for patients with congenital complete heart block and other rhythm disorders—work his team is now actively pursuing. Boink reflects on how high-impact papers, even well-intentioned ones, can harden assumptions into dogma. Ironically, his own 20-year-old studies had originally uncovered TBX18's role in sinus node development, unknowingly planting the seed for a field-wide misconception. By rigorously challenging the dogma his own work helped create, he's cleared the path toward therapies that actually work.