Triple-AAV FLNC Replacement Restores Cardiac Structure in a Mouse Model
Filamin C sits at the center of cardiomyocyte mechanics. When the protein is truncated or absent, the sarcomere loses a key structural scaffold, and the heart pays for it with dilation, fibrosis, and electrical vulnerability. In this preclinical study, investigators tackled the problem at its source: they rebuilt the missing full-length protein rather than trying to work around the downstream damage.
The challenge was formidable from the outset. The FLNC coding sequence is about 8.2 kb, too large for standard adeno-associated virus cargo limits, which has long blocked straightforward gene replacement. To get around that ceiling, the team engineered a triple-AAV strategy designed to reconstitute full-length FLNC in cardiac tissue through split-intein–mediated protein assembly. In practical terms, the approach turns a packaging problem into a reconstruction problem, allowing separate viral vectors to deliver pieces that can be stitched back into a complete therapeutic protein inside the heart.
Orthogonal Intein Pairing Enabled Full-Length Protein Reconstitution
Before moving into mice, the investigators screened candidate intein components in HEK293 cells. Two elements, GP41-1 and IMPDH, emerged as a compatible orthogonal pair for efficient full-length FLNC reconstitution. That selection step mattered because split-protein systems fail if the components do not assemble cleanly and efficiently; the therapeutic payload has to be rebuilt with enough fidelity to recapitulate native protein localization and function.
Once the delivery system was optimized, the study moved into a tamoxifen-inducible, cardiomyocyte-specific Flnc knockout model. This design created a direct test of whether restoring FLNC after depletion could rescue the phenotype, rather than merely preventing a developmental defect. Systemic delivery of optimized MyoAAV2A vectors restored full-length FLNC expression and brought the protein back to its sarcomeric location after cardiomyocyte FLNC loss. The reappearance of sarcomeric localization is not a cosmetic detail; FLNC belongs in the contractile architecture, and misplaced protein would likely fail to support the cytoskeletal tension that cardiomyocytes experience beat after beat.
One dose stood out. A total vector dose of 1 × 10¹³ genome copies/kg restored FLNC to approximately endogenous abundance.
Cardiac Function Held Up as Remodeling Was Blunted
The biological consequences were broad and internally consistent. Restored FLNC preserved systolic function and ventricular dimensions, limited cardiac enlargement, attenuated fibrotic remodeling, and prevented premature death in the knockout mice. Those outcomes track logically with the protein’s role in mechanotransduction: when the sarcomere can better sense and distribute mechanical stress, the ventricle is less likely to remodel under strain, fibrose, and progress toward failure.
The benefit extended beyond gross anatomy. FLNC replacement also attenuated pathological transcriptional remodeling and dampened compensatory changes in cytoskeletal and junction-associated proteins. In other words, the intervention did not simply patch one missing structural element while leaving the rest of the cell to spiral into maladaptive compensation. It appears to have stabilized a wider network of mechanical and junctional pathways that had been perturbed by FLNC deficiency.
The survival finding is especially consequential. Premature death disappeared in the treated cohort, indicating that structural rescue translated into a hard biological endpoint, not just an improved imaging readout. For a genetic cardiomyopathy model, that is the kind of divergence that gives the strategy translational weight.
A Mutation-Agnostic Strategy Could Fit Multiple FLNC Variants
The logic of this approach is attractive because it is mutation-independent. Many inherited cardiomyopathies are difficult to treat at the variant level, especially when the causative gene is large or the pathogenic spectrum is broad. Here, the therapeutic concept does not depend on correcting one specific nucleotide change. Instead, it attempts to restore the missing functional protein regardless of which truncating variant caused the deficiency in the first place.
That matters for FLNC-associated disease because truncating FLNC variants cause haploinsufficiency linked to both dilated and arrhythmogenic cardiomyopathies. A replacement strategy that rebuilds the full-length protein could, in principle, apply across this spectrum so long as the core mechanism is deficiency rather than dominant toxic gain of function. The preclinical data support that premise by showing that reconstitution of the intact protein is sufficient to preserve cardiac structure and function in vivo.
Still, the work remains preclinical and the delivery system is complex. Triple-vector assembly raises immediate questions about manufacturing consistency, tissue targeting, dosing, and translational scalability in humans. Cardiac gene delivery also has to contend with immune considerations, long-term expression, and the challenge of reaching enough cardiomyocytes to reproduce the level of rescue seen in mice. Those are not abstract hurdles; they are the difference between a compelling biologic proof and a feasible therapy.
Why This Study Matters for FLNC-Associated Cardiomyopathy
The strongest message from this work is straightforward: full-length FLNC replacement was sufficient to prevent cardiomyopathy in a mouse model of cardiomyocyte FLNC depletion. The intervention corrected protein localization, preserved chamber geometry, protected systolic performance, reduced fibrosis, and improved survival. It also normalized broader transcriptional and cytoskeletal disturbances that typically accompany structural protein loss.
For inherited cardiomyopathy, that combination is unusually coherent. The heart did not just look better under one assay; its architecture, molecular response, and survival all moved in the same direction.
The unanswered issue is whether the same split-intein, triple-AAV framework can be adapted safely and efficiently to patients with FLNC-associated cardiomyopathy. Human translation will require proof that vector assembly remains efficient in the larger human heart, that expression levels can be controlled, and that the approach can sustain benefit beyond the short-term preclinical window. Those questions define the next stage of development, along with the practical limits of systemic cardiac gene delivery in a disease that may already be advanced by the time it is diagnosed.