Loss of thymidine kinase 2 alters neuronal bioenergetics and leads to neurodegeneration

Abstract

Mutations of thymidine kinase 2 (TK2), an essential component of the mitochondrial nucleotide salvage pathway, can give rise to mitochondrial DNA (mtDNA) depletion syndromes (MDS). These clinically heterogeneous disorders are characterized by severe reduction in mtDNA copy number in affected tissues and are associated with progressive myopathy, hepatopathy and/or encephalopathy, depending in part on the underlying nuclear genetic defect. Mutations of TK2 have previously been associated with an isolated myopathic form of MDS (OMIM 609560). However, more recently, neurological phenotypes have been demonstrated in patients carrying TK2 mutations, thus suggesting that loss of TK2 results in neuronal dysfunction. Here, we directly address the role of TK2 in neuronal homeostasis using a knockout mouse model. We demonstrate that in vivo loss of TK2 activity leads to a severe ataxic phenotype, accompanied by reduced mtDNA copy number and decreased steady-state levels of electron transport chain proteins in the brain. In TK2-deficient cerebellar neurons, these abnormalities are associated with impaired mitochondrial bioenergetic function, aberrant mitochondrial ultrastructure and degeneration of selected neuronal types. Overall, our findings demonstrate that TK2 deficiency leads to neuronal dysfunction in vivo, and have important implications for understanding the mechanisms of neurological impairment in MDS.

Figure 1.

Reduction of mtDNA levels in brain due to TK2 deficiency leads to decreased steady-state levels of ETC proteins. (A) Relative quantification of mtDNA copy number in different brain regions of P7 and (B) P12 TK2^+/+ and TK2^−/− mice (n = 3). (*P < 0.05; **P < 0.01; Two-way ANOVA Bonferroni post-tests; error bars are SEM). Western blot and relative quantification of steady-state ETC proteins in the cerebellum of P7 (C and D) P12 (E and F) TK2^+/+ and TK2^−/− mice (n = 4 for each group; *P < 0.05; ***P < 0.001; Two-way ANOVA Bonferroni post-tests; error bars are SEM). Values are normalized to the TK2^+/+ control from the same group. Arrows indicate mtDNA-encoded subunits.

Figure 2.

ETC dysfunction in TK2^−/− PCs. (A) Anti-complex IV subunit I (COX I; mtDNA-encoded), (B) anti-complex IV subunit IV (COX IV; nDNA-encoded) in PCs of P12 of TK2^+/+ and TK2^−/− mice. TK2^−/− PCs have reduced expression of COX I whereas expression of COX IV is unaffected. Scale bars = 20 µm. (C) Sequential enzyme histochemistry for COX and SDH (complex II) activities in cerebella from P12 mice. TK2^+/+ PCs have normal COX activity and appear brown (upper panel), whereas TK2^−/− PCs lack COX activity and appear blue (low panel). Scale bar = 100 µm (low magnification) and 50 µm (high magnification), respectively.

Figure 3.

TK2 deficiency leads to reduction of mtDNA-encoded ETC subunits in cerebellar granule neurons (CGNs) in vitro. Dissociated CGNs were cultured from TK2^+/+ and TK2^−/− mice. (A) qPCR was used to confirm the decrease in mtDNA levels in DIV7 CGNs from TK2^−/− mice (student's t-test; n = 3 **P < 0.01; error bars are SEM). (B) Western blot analysis of ETC subunits in TK2^−/− CGNs, demonstrating a reduction of mtDNA-encoded subunits of complex IV and complex I. (C) CGNs from wild-type mice infected with a Turbo-GFP lentivector containing either a sequence targeted to a unique site of the mouse TK2 gene (shRNA-TK2) or non-silencing sequence (Scramble). qPCR was used to confirm the decrease in TK2 mRNA expression and the decrease of the protein level was confirmed by Western blot (student's t-test; n = 3 ***P < 0.005; error bars are SEM). (D) Western blot analysis of subunits of respiratory chain complexes of CGNs shown a lack of COX I subunit in shRNA-TK2 CGNs.

Figure 4.

Decreased mitochondrial bioenergetic capacity and aberrant mitochondrial structure in TK2^−/− neurons. (A and E) OCR and (B and F) ECAR in DIV7 and DIV14 CGNs, cultured from TK2^+/+ and TK2^−/− mice (student's t-test n = 3; ***P < 0.005; error bars are SEM). Both rates are normalized against cell number and expressed as rate per minutes. (C and G) Real-time analysis of OCR in CGNs. Mitochondrial uncoupler FCCP and mitochondrial complex I inhibitor rotenone were injected sequentially at the indicated time points into each well after baseline rate measurement. (D and H) Relative ATP levels in DIV7 and DIV14 CGNs, cultured from TK2^+/+ and TK2^−/− mice. Measurements were made in triplicate (student's t-test n = 3; ***P < 0.005; error bars are SEM) with two animals per group. (I) EM images showing mitochondria in the cell body of DIV10 CGNs from TK2^+/+ and TK2^−/− mice. TK2^−/− mitochondria have dramatic abnormalities in morphology and cristae organization. Scale bars = 2 and 1 µm, respectively.

Figure 5.

PC degeneration in TK2^−/− cerebellar sections. (A) Anti-calbindin immunocytochemistry to highlight PCs. TK2^−/− PC show reduced dentritic arbors at P7 and degeneration at P12. Scale bars = 50 µm. (B) Higher magnification of P12 PCs. TK2^+/+ PCs dendrites spread into molecular layer, whereas mutant PCs shown reduced dendritic arbors and cell loss. Scale bars = 20 µm. (C) TUNEL staining (green for apoptotic cells). Increased TUNEL positive cells are present throughout the cerebellar layers in P12 TK2^−/− mice. DAPI counterstaining (grey or blue) was used to demonstrate the cellular architecture. Scale bars = 50 µm.