Loss-of-function mitochondrial DNA polymerase gamma variants cause vascular smooth muscle cells to secrete a diffusible mitogenic factor.

Abstract

Introduction: Mitochondrial dysfunction promotes vascular aging and disease through diverse mechanisms beyond metabolic supply, including calcium and radical signaling and inflammation. Mitochondrial DNA (mtDNA) replication by the POLG-encoded mitochondrial DNA polymerase (POLG) is critical for mitochondrial health. Loss-of-function POLG variants are associated with a predisposition to hypertension. We hypothesized that impaired POLG, through reduced mtDNA copy number or other mechanisms, would promote smooth muscle hypertrophy or hyperplasia that drives vascular remodeling associated with hypertension.

Methods: We characterized the effect of over-expressing POLG variants that were previously observed in a cohort of hypertensive patients (p.Tyr955Cys, p.Arg964Cys, p.Asn1098Ile, and p.Arg1138Cys) in A7r5 cells.

Results: AlphaFold modeling of the POLG holoenzyme complexed with DNA predicted changes in the catalytic site in the p.Tyr955Cys and p.Asn1098Ile variants, while p.Arg964Cys and p.Arg1138Cys showed minimal effects. The POLG variants reduced mtDNA copy number, assessed by immunofluorescence and droplet digital PCR, by up to 27% in the order p.Tyr955Cys > p.Arg964Cys > p.Asn1098Ile > p.Arg1138Cys relative to wild-type-transfected cultures. Loss of mtDNA was reduced in cultures grown in low serum and glucose media, but the cell density was increased in the same rank order in both 10% serum and 1% serum. POLG constructs contained a Myc epitope, the counterstaining for which showed that the mtDNA copy number was reduced in both transfected cells and untransfected neighbors. Live-cell imaging of mitochondrial membrane potential with TMRM and radical oxygen species production with MitoSOX showed little effect of the POLG variants. POLG variants had little effect on oxygen consumption, assessed by Seahorse assay. Live-cell imaging growth analyses again showed increased growth in A7r5 cells transfected with p.Tyr955Cys but a decreased growth with p.Arg1138Cys, while p.Tyr955Cys increased growth of HeLa cells. Conditioned media from HeLa cells transfected with POLG variants reduced doubling times in naïve cultures. Pharmacologically, wedelolactone and MitoTEMPOL, but not indomethacin or PD98059, suppressed the mitogenic effects of p.Tyr955Cys and p.Arg964Cys in A7r5 cells.

Discussion: We conclude that POLG dysfunction induces secretion of a mitogenic signal from A7r5 and HeLa cells even when changes in mtDNA copy number are below the limit of detection. Such mitogenic stimulation could stimulate hypertrophic remodeling that could contribute to drug-resistant hypertension in patient populations with loss-of-function POLG variants.