Oxidative stress during mitochondrial biogenesis compromises mtDNA integrity in growing hearts and induces a global DNA repair response

Abstract

Cardiomyocyte development in mammals is characterized by a transition from hyperplastic to hypertrophic growth soon after birth. The rise of cardiomyocyte cell mass in postnatal life goes along with a proportionally bigger increase in the mitochondrial mass in response to growing energy requirements. Relatively little is known about the molecular processes regulating mitochondrial biogenesis and mitochondrial DNA (mtDNA) maintenance during developmental cardiac hypertrophy. Genome-wide transcriptional profiling revealed the activation of transcriptional regulatory circuits controlling mitochondrial biogenesis in growing rat hearts. In particular, we detected a specific upregulation of factors involved in mtDNA expression and translation. More surprisingly, we found a specific upregulation of DNA repair proteins directly linked to increased oxidative damage during heart mitochondrial biogenesis, but only relatively minor changes in the mtDNA replication machinery. Our study paves the way for improved understanding of mitochondrial biogenesis, mtDNA maintenance and physiological adaptation processes in the heart and provides the first evidence for the recruitment of nucleotide excision repair proteins to mtDNA in cardiomyocytes upon DNA damage.

Figure 1.

Pronounced upregulation of genes involved in mitochondrial and oxidative metabolism in adult rat hearts. (A) Heat-map clustering of embryonic Day 18 (E18), neonatal (Neo), 10-day old (P10) SD, adult SD (Ad-SD) (>4 months), adult WS (Ad-WS) and old (>18 months) WS rats. All three independent samples in each cohort cluster together and are separable from samples in other cohorts. (B) 11 173 out of 31 042 genes included in the Affymetrix microarray possess an annotated GO term. Mitochondrial and oxidative metabolism genes show significant upregulation at P10 compared to neonatal heart (324 out of 1153 annotated genes) and in adults compared to P10 heart (376 genes out of 1153). A total of 63 NEM genes upregulated at P10 are further upregulated in adults (see also Supplementary Table S1). P10 rat hearts show significant enrichment of genes involved in DNA repair when compared to any other age group (see also Supplementary Table S2). See (33) for the details of the GO term analysis. P-values are based on the hypergeometric score analysis of target (upregulated) and background (all GO terms) data sets (34).

Figure 2.

Transcriptional networks controlling heart mitochondrial biogenesis. (A) PPARGC1B ([Ppargc1b] PGC-1β in humans) is the only member of the PGC-1 family of transcriptional coactivators highly upregulated in P10 rat hearts. PPARGC1A ([Ppargc1a] or PGC-1α) maintains relatively stable expression while PPRC1 ([Pprc1] PRC in humans) as well as HIF-1α (Hif1a) are downregulated. (B) Expression of many known targets of PGC-1 s, such as NRFs (Nrf1, Gabps) is unchanged in growing rat hearts with exception of PPARG (Pparg), an important regulator of β-oxidation genes. (C) Stat3 expression increases concomitantly with postnatal heart growth. (D) Western blot analysis of the expression of some transcriptional regulators during postnatal heart development to validate transcriptome data. (E) Crem and Crebl2 but no other members of the Creb gene family are significantly upregulated during postnatal heart growth (a: P = n.s., b: P < 0.05 and c: P < 0.01, one-way ANOVA with Tukey’s multiple comparison test).

Figure 3.

mtDNA maintenance and expression in growing rat heart. No significant differences in mtDNA copy number (A), topological organization (B) or 7S DNA quantity (C) were detected between neonatal and adult rat hearts. Aged rats had about 2-fold higher mtDNA levels than other age groups. Of the known factors involved in the expression of mtDNA, mitochondrial RNA polymerase POLRMT, TFB2M (D), MTERFD2 and MTERFD3 (E) show a significant positive response to increasing mitochondrial function in postnatal rat heart. The accessory subunit of POLG, POLG2 (F) and the topoisomerase TOP3A (G) were the only proteins involved in mtDNA maintenance that were induced during increased mitochondrial biogenesis. (H) TFAM and TFB2M were used to confirm the transcriptome results on protein level. Western blots were quantified and normalized against neonatal rats as loading control. Results are presented as mean ± SD. (**P < 0.01, one-way ANOVA with Tukey’s multiple comparison test).

Figure 4.

Induction of DNA repair genes corresponds to increased oxidative damage during postnatal heart development. (A) Factors involved in nDNA replication (Pcna, Mcm5), single- (Xrcc1) or double-strand break repair (Rad51, Rad51c and Xrcc5) are either downregulated or unchanged in growing heart. (B) Factors either involved in NER (Rad23a, Xpa) or BER (Tdg, Mpg) are transiently upregulated in P10 rat hearts (Supplementary Table S2). (C) Upregulation of NER and BER genes coincides with increased sensitivity of mtDNA to Fpg, an enzyme capable of cleaving a wide range of oxidized purine-bases. Neonatal and P10 rat heart mtDNAs were probed for the presence of 8-oxoG modifications by South-Western blot analysis. (D) Cellular antioxidant defenses are upregulated only in the adult animals although increased levels of mtDNA damage are evident only in P10 rats. (E) Changes in the expression of nDNA maintenance genes (PCNA), NER (RAD23A) and antioxidant defenses (SOD2) were validated on protein level. Western blots were quantified and normalized against neonatal rats as loading control (**P < 0.01, *P < 0.05, one-way ANOVA with Tukey’s multiple comparison test).

Figure 5.

RAD23A, a key NER protein is localized to cardiomyocyte mitochondria. (A) Western blot analysis of RAD23A presence in sucrose-gradient purified rat heart mitochondrial preparations subjected to trypsin digestion. TOM20 a mitochondrial outer membrane protein, partially accessible to trypsin, and the mitochondrial matrix protein TFAM were used as positive and negative controls, respectively. (B) Immunofluorescent analysis of the localization of RAD23A in the mitochondrial network (TOM20 and DAPI were used to label mitochondria and nucleus, respectively). (C) DXR for 16 h causes a marked increase in RAD23A protein levels in whole cell as well as in mitochondrial preparations. Sarcomeric actin demonstrates the relative enrichment of mitochondria in the preparations. (D) RAD23A co-localizes with mtDNA partially with nucleoids (cytoplasmic DAPI foci) under basal conditions without DNA damage induction. All nucleoids that had incorporated EdU stained positive for RAD23A (white arrows, N = 3 cells, 300 nucleoids). Open arrows indicate nucleoids that were positive for RAD23A but not obviously for EdU. Closed arrows show nucleoids negative for both EdU and RAD23A. (E) Exposure of cells to 1 µM DXR for 16 h results in the uptake of RAD23A into nuclei (N), together with more contrasted mitochondrial nucleoid staining. The nuclear signal in the EdU panels is due to auto-fluorescence of DXR when bound to nDNA, although some EdU incorporation due to nDNA repair cannot be excluded. The nucleus in the lower panels is mostly out of the confocal plane. Arrows as above. All scale bars 25 µm.