IF a single AAV9 vector encoding a triple-engineered human ND1 transgene — comprising (1) a codon-harmonized (not simply codon-maximized) nuclear recoding of all 7 non-universal mitochondrial codons with strategic synonymous slow-codon insertions at each of ND1's hydrophobic transmembrane domain boundaries, (2) the COX8A mitochondrial targeting sequence (MTS) at the N-terminus, and (3) the 3′ UTR of the nuclear-encoded mitochondrial gene AKAP1 appended to the mRNA to anchor ND1 transcripts to the outer mitochondrial membrane (OMM) for pseudo-co-translational import — is delivered via AAV9 at MOI 10⁵ vg/cell to LHON patient-derived fibroblasts (GM3460A, m.3460G>A, cultured in galactose medium),
THEN at day 14 post-transduction, mitochondrial colocalization of allotopic ND1 protein (measured by immunofluorescence with anti-ND1 Abcam ab183721 / MitoTracker Deep Red) will exceed ≥70% — compared to the ≤30–40% colocalization typical of standard single-component COX8A-MTS-only constructs — and Complex I-dependent oxygen consumption rate (Seahorse XF, rotenone-sensitive OCR in galactose media) will increase ≥50% versus untreated m.3460G>A fibroblasts, achieving ≥70% of healthy control OCR, with less than 40% deterioration of these OCR gains between days 14 and 28,
BECAUSE the following mechanistic chain addresses the primary failure mode — cytosolic aggregation of hydrophobic ND1 — through two independent and synergistic mechanisms:
- Cytosolic aggregation is the dominant import bottleneck for allotopic ND1. Standard COX8A MTS-only constructs depend entirely on post-





























