Mechanisms of mitochondrial DNA (mtDNA) maintenance have recently gained wide interest owing to their role in inherited diseases as well as in aging. Twinkle is a new mitochondrial 5'-3' DNA helicase, defects of which we have previously shown to underlie a mitochondrial disease, progressive external ophthalmoplegia with multiple mtDNA deletions. Mouse Twinkle is highly similar to the human counterpart, suggesting conserved function. Here, we have characterized the mouse Twinkle gene and expression profile and report that the expression patterns are not conserved between human and mouse, but are synchronized with the adjacent gene MrpL43, suggesting a shared promoter. To elucidate the in vivo role of Twinkle in mtDNA maintenance, we generated two transgenic mouse lines overexpressing wild-type Twinkle. We could demonstrate for the first time that increased expression of Twinkle in muscle and heart increases mtDNA copy number up to 3-fold higher than controls, more than any other factor reported to date. Additionally, we utilized cultured human cells and observed that reduced expression of Twinkle by RNA interference mediated a rapid drop in mtDNA copy number, further supporting the in vivo results. These data demonstrate that Twinkle helicase is essential for mtDNA maintenance, and that it may be a key regulator of mtDNA copy number in mammals.

Variants in mitochondrial DNA (mtDNA) could be associated with type 2 diabetes because ATP plays a critical role in the production and release of insulin. Diabetes can be precipitated both by mtDNA mutations and by exposure to mitochondrial poisons. The risk of inheriting diabetes from an affected mother is greater than that from an affected father, but this is not explained by maternally inherited diabetes and/or deafness (MIDD) caused by the 3243G : C mtDNA point mutation, which accounts for less than 0.5% of cases of diabetes. A common mtDNA variant (the 16189 variant) is positively correlated with blood fasting insulin, but there are no definitive studies demonstrating that it is associated with diabetes. We demonstrated a significant association between the 16189 variant and type 2 diabetes in a population-based case–control study in Cambridgeshire, UK (n=932, odds ratio=1.61 (1.0–2.7, P=0.048), which was greatly magnified in individuals with a family history of diabetes from the father's side (odds ratio=∞; P<0.001).