Mitochondrial dysfunction impairs osteogenesis, increases osteoclast activity, and accelerates age related bone loss

Abstract

The pathogenesis of declining bone mineral density, a universal feature of ageing, is not fully understood. Somatic mitochondrial DNA (mtDNA) mutations accumulate with age in human tissues and mounting evidence suggests that they may be integral to the ageing process. To explore the potential effects of mtDNA mutations on bone biology, we compared bone microarchitecture and turnover in an ageing series of wild type mice with that of the PolgA^mut/mut mitochondrial DNA 'mutator' mouse. In vivo analyses showed an age-related loss of bone in both groups of mice; however, it was significantly accelerated in the PolgA^mut/mut mice. This accelerated rate of bone loss is associated with significantly reduced bone formation rate, reduced osteoblast population densities, increased osteoclast population densities, and mitochondrial respiratory chain deficiency in osteoblasts and osteoclasts in PolgA^mut/mut mice compared with wild-type mice. In vitro assays demonstrated severely impaired mineralised matrix formation and increased osteoclast resorption by PolgA^mut/mut cells. Finally, application of an exercise intervention to a subset of PolgA^mut/mut mice showed no effect on bone mass or mineralised matrix formation in vitro. Our data demonstrate that mitochondrial dysfunction, a universal feature of human ageing, impairs osteogenesis and is associated with accelerated bone loss.

Figure 1

Micro-CT scan of femurs and lumbar spines demonstrate accelerated bone loss in PolgA^mut/mut mice. An exercise intervention was not associated with a change in trabecular bone mass. A significant decline in femoral trabecular BV/TV can be seen by 7 months of age in female (A) PolgA^mut/mut mice (p = 0.003) and by 11 months of age in male (B) PolgA^mut/mut mice (p < 0.0001) when compared to age matched wild type controls (unpaired 2 tailed t-test). A similar pattern of bone loss was seen in the lumbar spine with significant reductions in trabecular BV/TV in female PolgA^mut/mut mice at 7 and 11 months (C), and in male PolgA^mut/mut mice at 4, 7 and 11 months (D), compared to age and sex matched wild type controls. Cortical thickness was also significantly reduced in female PolgA^mut/mut mice at 7 and 11 months (E), and in male PolgA^mut/mut mice at 11 months (F), compared to wild type controls. Exercise had no significant effects (ns) on bone mass observed in 11 month old PolgA^mut/mut males in terms of trabecular BV/TV (B, D). A gradual decline in BV/TV levels is observed in wild type mice as a feature of advancing age, with females showing a significantly reduced femoral trabecular BV/TV in comparison to age matched males at 4, 7 and 11 months of age (p = 0.0004, p = 0.0015, p < 0.0001 respectively (unpaired 2 tailed t-test)).

Figure 2

Hitomorphometric analysis shows reduced bone density, reduced bone formation rate, reduced osteoblast population density, and increased osteoclast population density in PolgA^mut/mut mice. BV/TV (A) and mineralisation (OS/BS%) (B) is significantly lower in PolgA^mut/mut mice compared to age matched wild type (WT) mice (p = 0.017 and 0.010 respectively). von Kossa staining of WT and PolgA^mut/mut mice (C). BFR were significantly lower (p = 0.0088) in PolgA^mut/mut mice compared to age matched WT (D). Immunofluorescent imaging of tissue sections from tibiae of WT and PolgA^mut/mut demonstrate the typical difference in BFR observed between the two genotypes (E). Significantly reduced osteoblast population densities (p = 0.005) in PolgA^mut/mut mice compared to age matched WT mice are observed (F). Typical appearances of WT and PolgA^mut/mut tibiae sections stained with toluidine blue, with red arrows demarcating bone-lining osteoblasts (G). There are significantly increased population densities of osteoclasts in PolgA^mut/mut mice compared to age matched WT, with higher osteoclast numbers (H) and osteoclast surface area relative to bone surface (I) (p = 0.017 and 0.023 respectively). Typical appearance of TRAcP stained tibiae sections from WT and PolgA^mut/mut mice (J).

Figure 3

PolgA^mut/mut osteogenic cells produce less mineralised matrix in vitro. Ratio of mineralised matrix surface area to surface area occupied by osteogenic ALP positive cells. A gradual decline in mineralised matrix formation is seen with increasing age in wild type cells extracted from both females (A) and males (B), with a significant reduction observed by 7 and 11 months in comparison to cells from mice aged 4 months (p < 0.001, one way ANOVA, Bonferroni). At all three ages of study, and in both sexes, mineralisation by PolgA^mut/mut cells is vastly reduced (p < 0.0001, unpaired, 2 tail t-testing). The functional capacity of PolgA^mut/mut cells from 11-month old exercised male mice to produce mineralised matrix was not significantly (ns) different to that observed in cells derived from non-exercised 11 month old male PolgA^mut/mut mice (p = 1). Typical appearances of wild type cell lines (C, D) and PolgA^mut/mut cell lines (E, F) taken from mice aged 11 months are shown.

Figure 4

Population density of ALP positive, osteogenic cells, derived from bulk unfractionated mesenchymal cell harvests. The ratio of ALP positive cells to total cell number was recorded as a measure of osteogenic cell population density, for female (A) and male (B) mice. Significant variation in the population density of ALP positive cells occurred in all groups. There was no significant difference (one-way ANOVA/Bonferroni) between wild type cells and PolgA^mut/mut cells (regardless of sex or age of mouse of origin), in their propensity to produce osteogenic cells, suggesting that reduced mineralised matrix formation in vitro by PolgA^mut/mut cell lines is not accounted for by reduced population density of osteogenic cells.

Figure 5

Osteoclast resorption assays using cells extracted from wild type and PolgA^mut/mut mice aged 11 months. Comparisons between wild type (WT) and PolgA^mut/mut cell lines. Average size of resorption pits formed by WT osteoclasts (A) was larger than those formed by PolgA^mut/mut osteoclasts (p = 0.0008). The ratio of resorption pits to osteoclasts was higher (B) in PolgA^mut/mut cells (p < 0.0001). The ratio of resorbed dentine to osteoclast number (C) was also higher in PolgA^mut/mut cells (p = 0.0134). Examples of appearances of osteoclasts on dentine (D) and visualisation of resorption pits (E).

Figure 6

NDUFB8:Porin and COX-I:Porin in mouse osteoclasts. Log transformed background corrected signal intensities of porin, COX-I and NDUFB8 within individual osteoclasts were recorded. A linear association of NDUFB8 and COX-I to porin is present which is strongest in wild type mice aged 4 months. With increasing age in wild type mice, increasing variation in COX-I and NDUFB8 signal occurs with some loss of the linear relationship. This change is more marked in PolgA^mut/mut mice aged 11 months in which many cells exhibited low COX-I and NDUFB8 protein expression (different colours within each age group signify different cell lines).

Figure 7

Representative graphs of Z-scored NDUFB8:Porin and COX-I:Porin for wild type and PolgA^mut/mut mice osteoclasts. Z-scores are derived for porin, NDUFB8 and COX-I from the mean and standard deviations of NDUFB8:porin and COX-I:porin relationships, in wild type controls aged 4 months. The Z scores for NDUFB8:porin and COX-I:porin are plotted against each other for each mouse, with each dot representing an individual osteoclast and it’s colour signifying the porin level (dark purple, very low; light purple, low; grey, normal; orange, high; red, very high). The typical appearances for a wild type mouse aged 4 months (A), wild type mouse aged 11 months (H), and PolgA^mut/mut mice aged 11 months (T and U) are shown (lettering corresponds to that in Supplemental Data S5 and S6). Most data points are no more than 3SD from the mean in the control group (wild type mice aged 4 months). However, in wild type mice aged 11 months reduced NDUFB8 and COX-I protein expression is observed and this deficit occurs to a greater degree in PolgA^mut/mut mice of the same age. The percentage of osteoclasts which are positive, intermediate positive, intermediate negative or negative for COX-I and NDUFB8 for each mouse are also shown, in the smaller panels to the right of each graph.