Researchers from Amsterdam UMC have overturned a key assumption in the biological
pacemaker field. In a new preclinical study they show that the transcription factor TBX18 does
not generate true biological pacemaker activity, while the ion channel Hcn2 does produce
robust pacemaker function in the heart.
TBX18 revisited
For more than a decade, TBX18 has been reported to reprogram working ventricular cardiomyocytes
into sinoatrial‑node–like pacemaker cells, based on widely cited high‑impact papers. Using lowimmunogenic adeno-associated virus (AAV) vectors and detailed electrophysiology, the team from
Amsterdam UMC systematically re-examined this claim. They first showed that conventional
high‑level TBX18 overexpression is profoundly toxic, causing severe myocardial fibrosis and scarring
in mouse hearts, while the control did not. "We found that supraphysiological TBX18 expression is
highly toxic for cardiomyocytes," says senior author Gerard Boink, cardiologist and principal
investigator at Amsterdam UMC. "That toxicity alone already questions the feasibility of TBX18 as a
clinically relevant biological pacemaker strategy."
Safe TBX18 levels, no pacemaker
To separate toxicity from biological functionality, the researchers engineered an optimized AAV
cassette and reduced the TBX18 protein levels to about 1% of conventional CMV-driven expression.
This completely prevented fibrosis, yet preserved transcriptional activity of TBX18 with effective
repression of well-known targets such as Gja1 (Connexin43). Despite this controlled, non-toxic
expression, cardiomyocytes did not acquire a genuine pacemaker phenotype. TBX18 suppressed
multiple working-myocyte genes and lead to abnormal action potentials. However, it did not induce
key pacemaker gene programs, nor did it induced Hcn4 protein, or the pacemaker current If. "These
results give good reason to stop TBX18-based gene therapy efforts in the heart," says Boink. "Our
data show that even at realistic, non-toxic expression levels, you do not get pacemaker activity.
Instead of providing a therapy, you risk arrhythmia by ion channel dysregulation and electrical
instability."
Vector artefacts exposed
In a rat model of complete atrioventricular (AV) block, both Adenoviral (AdV)-TBX18 and the AdVcontrol produced similar ectopic pacing and extensive local fibrosis, pointing to AdV vector-related
inflammation and scarring as the true driver of the earlier reported "TBX18 pacing" signals. The AAV
system used in the present study avoided these confounders. "In some parts of the field, there has
been a tendency to over‑value single high‑impact papers," Boink notes. "That can unintentionally
promote artefactual concepts, like TBX18‑mediated reprogramming, to the status of a dogma.
"Ironically, this dogma is founded on our own 20-year-old studies, which originally uncovered the
role of TBX18 in sinus node development."
Hcn2 proves its value
In contrast to TBX18, AAV‑mediated expression of the pacemaker channel Hcn2 produced robust,
autonomically responsive ventricular pacing in the same rat complete AV‑block model. Co‑expression
of TBX18 did not improve Hcn2‑based pacing, underscoring that TBX18 can also not be employed to
further promote HCN-based pacing. "With this study we also show that an efficient biological
pacemaker can be created with Hcn2 alone," says Boink. "That has direct implications for
developing gene‑therapy–based pacemakers for patients with congenital complete heart block, and
other pacemaker indications, where we are now actively moving forward."
Valorization at Amsterdam UMC
Next to his work as cardiologist and group leader, Boink is also Chief Valorization Officer
for Amsterdam Cardiovascular Sciences and Chief Scientific Officer at PacingCure. "I feel
privileged to work in an environment where valorization and real‑world impact are just as
important as publishing in high impact journals, which obviously also has our priority," says
Boink.
