Effects of Altering Mitochondrial Antioxidant Capacity on Molecular and Phenotypic Drivers of Fibrocalcific Aortic Valve Stenosis

Abstract

Background: While a small number of studies suggest that oxidative stress has an influential role in fibrocalcific aortic valve disease (FCAVD), the roles of specific antioxidant enzymes in progression of this disease remain poorly understood. Here, we focused on selectively altering mitochondrial-derived oxidative stress-which has been shown to alter progression of a myriad of age-associated diseases-on the progression of molecular and phenotypic drivers of FCAVD. Methods: We generated low-density lipoprotein receptor-deficient, Apolipoprotein B100-only mice (LA) that were either haploinsufficient for MnSOD (LA-MnSOD ^+/-) or genetically overexpressing MnSOD (LA-MnSOD ^Tg/0). After 6 months of Western diet feeding, mice underwent echocardiography to assess valvular and cardiac function and tissues were harvested. Quantitative-RT PCR, immunohistochemistry, and histopathology were used to measure changes in molecular pathways related to oxidative stress, calcification, and fibrosis. Results: While reductions in MnSOD increased oxidative stress, there was not an overt phenotypic effect of MnSOD deficiency on valvular and cardiac function in LA-MnSOD ^+/- mice. While markers of canonical bone morphogenetic protein signaling tended to increase in valve tissue from LA-MnSOD ^+/- (e.g., p-SMAD1/5/8 and osterix), we did not observe statistically significant increases in osteogenic signaling. We did, however, observe highly significant reductions in expression of osteopontin, which were associated with significant increases in calcium burden in LA-MnSOD ^+/- mice. Reciprocally, genetically increasing MnSOD did not preserve valve function in LA-MnSOD ^Tg/0, but we did observe slight reductions in p-SMAD1/5/8 levels compared to their non-transgenic littermates. Interestingly, overexpression of MnSOD dramatically increased expression of osteopontin in valve tissue from LA-MnSOD ^Tg/0 mice, but was not sufficient to attenuate calcium burden when compared to their LA-MnSOD ^0/0 littermates. Conclusions: Collectively, this study demonstrates that maintenance of mitochondrial antioxidant capacity is important in preventing accelerated disease progression in a mouse model of FCAVD, but that effectively altering mitochondrial antioxidant capacity as a monotherapeutic approach to slow key histopathological and molecular drivers of FCAVD remains biologically and therapeutically challenging.

Keywords: MnSOD; calcification; fibrocalcific aortic valve disease; mitochondria; oxidative stress.

Figure 1

Effects of MnSOD deficiency on valvular function and valvular oxidative stress in hyperlipidemic mice. (A,B) Reductions in mitochondrial antioxidant capacity have no effect on cusp separation distance or transvalvular peak velocity compared to wild-type littermates (n = 40, 22). (C) As expected, MnSOD expression levels were reduced by ~50% in the aortic valve of LA-MnSOD^+/− mice compared to their wild-type littermates (n = 11, 12). (D) Losses of mitochondrial antioxidant capacity tended in to increase 3-nitrotyrosine levels—a marker of oxidative stress—compared to wild-type littermates (n = 8, 8). For all panels, * denotes p < 0.05.

Figure 2

Effects of MnSOD deficiency on valvular osteogenic signaling and calcium burden in hyperlipidemic mice. (A) Bone morphogenetic protein-2 (BMP2) expression levels remained unchanged by reductions in mitochondrial antioxidant capacity (n = 11, 12). (B) Expression levels of ID1—a canonical BMP signaling target—did not change between LA-MnSOD^+/+ and LA-MnSOD^+/− mice (n = 11, 12). (C) Expression levels of RUNX2, a BMP-related transcription factor, was modestly reduced in LA-MnSOD^+/− compared to their LA-MnSOD^+/+ littermates (n = 11, 12). (D) Interestingly, osteopontin mRNA levels were significantly reduced in aortic valves from LA-MnSOD^+/− mice compared to LA-MnSOD^+/+ littermates (n = 11, 12). (E) Immunohistochemical assessment of p-SMAD1/5/8 suggested moderately increased canonical BMP signaling in LA-MnSOD^+/− mice compared to LA-MnSOD^+/+ mice (n = 8, 8). (F) Histological evaluations of calcium burden using Alizarin Red staining showed significant increases in valvular calcium burden in LA-MnSOD^+/− compared to their wild-type littermates (higher magnification images are available in the Supplementary Figure 4) (n = 8, 8). For all panels, * denotes p < 0.05.

Figure 3

Effects of MnSOD deficiency on valvular fibrogenic signaling, tissue fibrosis, and collagen fiber thickness in hyperlipidemic mice. (A) TGFβ-1 expression levels tended to decrease in aortic valves from mice with MnSOD haploinsufficiency compared to their wild-type counterparts (n = 11, 12). (B) Immunofluorescence of phospho-SMAD3 protein was unchanged with reductions in MnSOD (n = 8, 8). (C–E) Picrosirius red staining was used to evaluate changes in total collagen burden. (C). Losses of MnSOD in LA-MnSOD^+/− mice tended to reduce total collagen fibers compared to their LA-MnSOD^+/+ littermates. (D,E) Genetic reductions in MnSOD did not alter collagen fiber composition in LA-MnSOD^+/− mice compared to their LA-MnSOD^+/+ littermates (n = 8, 8).

Figure 4

Effects of MnSOD overexpression on valvular function and valvular oxidative stress in hyperlipidemic mice. (A,B) Increasing mitochondrial antioxidant capacity had no impact on cusp separation distance and peak velocity compared to LA-MnSOD^0/0 mice (n = 17, 15). (C) mRNA levels for MnSOD were increased by three-fold in LA-MnSOD^Tg/0 mice compared to their wild-type littermates (n = 12, 10). (D) Interestingly, increasing mitochondrial antioxidant capacity did not reduce 3-nitrotyrosine levels in valves from LA-MnSOD^Tg/0 mice compared to their LA-MnSOD^0/0 littermates (n = 10, 10). For all panels, * denotes p < 0.05.

Figure 5

Effects of MnSOD overexpression on osteogenic signaling and calcium burden in valve tissue from hyperlipidemic mice. (A) Increasing mitochondrial antioxidant capacity levels did not alter BMP2 expression levels in the aortic valve compared to wild-type littermates (n = 12, 10). (B) Expression levels of ID1 were significantly increased in LA-MnSOD^Tg/0 compared to LA-MnSOD^0/0 littermates (n = 12, 10). (C) Expression levels of RUNX2 were significantly increased in LA-MnSOD^Tg/0 compared to the LA-MnSOD^0/0 littermates (n = 12, 10) (D) Osteopontin expression levels were significantly increased in LA-MnSOD^Tg/0 mice compared to LA-MnSOD^0/0 littermates (n = 12, 10). (E) Increasing MnSOD tended to reduce protein levels of phospho-SMAD1/5/8 in the aortic valve but failed to reach significance (n = 10, 10). (F) Interestingly, calcium deposition in the aortic valve was unchanged between genotypes despite dramatic increases in MnSOD expression in LA-MnSOD^Tg/0 mice (higher-magnification images are available in the Supplementary Figure 8) (n = 10, 10). For all panels, * denotes p < 0.05.

Figure 6

Effects of MnSOD overexpression on fibrogenic signaling and collagen deposition in aortic valves from hyperlipidemic mice. (A) Overexpression of MnSOD modestly increased TGFβ-1 expression levels in the aortic valve compared to wild-type mice (n = 12, 10). (B) Immunofluorescence of phospho-SMAD3 protein was unchanged between LA-MnSOD^0/0 and LA-MnSOD^Tg/0 mice (n = 10, 10). (C) Picrosirius red staining to evaluate changes in collagen burden resulted in modest increases in LA-MnSOD^Tg/0 mice compared to wild-type mice (n = 10, 10). (D,E) The increase in total collagen fibers in LA-MnSOD^Tg/0 mice was associated with a tendency toward greater collagen fiber thickness compared to their LA-MnSOD^Tg/0 littermates (n = 10, 10).