Cardiac overexpression of 8-oxoguanine DNA glycosylase 1 protects mitochondrial DNA and reduces cardiac fibrosis following transaortic constriction

Construction of the OGG1 transgenic mouse model with cardiac-specific overexpression of OGG1.

First-strand cDNA was synthesized from human fibroblast total RNA by superscript II and oligo dT primers (Invitrogen). The cDNA was amplified by PCR using primers designed for the human mitochondrial isotype 2a of OGG1 (5, 26). The sense primer extends from nucleotide 78 of the OGG1–2a mRNA in the 5′-untranslated region and included an additional Sal I restriction site (gacttagtcgaccgggagaagataagtcgcaa). The antisense primer extended to position 1,782 in the OGG1–2a mRNA 3′-untranslated region and contained an additional Hind III restriction site (actactaagcttccagagggccacataatgtt). This PCR reaction produced a fragment containing all of the coding sequence of the human OGG1–2a isoform. The OGG1–2a isoform is predominantly localized in mitochondria (26) due to the presence of a putative mitochondrial targeting sequence and absence of a nuclear localization signal from exon 7 of the OGG1 gene. The purified OGG1 cDNA fragment was subcloned behind a 5.5-kb DNA fragment from the α-myosin heavy chain (MHC) promoter (12) that produces transcription exclusively in cardiac myocytes and in front of a 500 bp of poly(A) DNA sequence derived from the rat insulin II gene. The sequence of the entire subcloned OGG1 cDNA fragment was confirmed. The transgene was released from the plasmid by Not I digestion and microinjected into single cell fertilized FVB mouse embryos by standard embryo microinjection procedures. Male mice between the ages of 90 and 120 days were used for subsequent experiments. This age was chosen so that animals were large enough for surgery and to reduce variation between mice. All animal procedures conformed to the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Pub. No. 85-23, Revised 1996) and were approved by the United State Department of Agriculture-Certified Institutional Animal Care Committee.

TAC surgery.

TAC surgery was performed by a modification of a previously published technique (2). Mice were anesthetized with avertin (0.4 g/kg), maintained on a 37°C pad, and ventilated with 100% oxygen. An incision at the left second intercostal space was made to open the chest. A chest retractor was applied to facilitate the view. The thymus was pulled away, and transverse aorta was dissected from surrounding tissues. A 6–0 silk suture was passed around the transverse aorta and tightened against a 26-G needle. The needle was immediately removed to provide a lumen with a stenotic aorta. Lungs were inflated, and the chest cavity, muscles, and skin were closed layer by layer with 6–0 silk sutures. The whole procedure lasted 20–30 min. After surgery, mice were warmed with a heating lamp and received an intraperitoneal injection of 0.5 ml of saline at 37°C. For 48 h, mice were given subcutaneous injections of buprenorphine, 0.1 mg/kg every 12 h. Sham mice were done same as TAC animals except that banding of the transverse aorta was omitted.

Echocardiographic assessment of cardiac function.

Transthoracic echocardiography of the left ventricle (LV) was performed using a 15-MHz linear array transducer (15L8) interfaced with a Sequoia C512 system (Siemens, Malvern, PA) as previously described (42). Mice were anesthetized with 2% isoflurane, maintained under anesthesia with 1.25% isoflurane, and examined. Ventricular parameters were measured in M-mode with a sweep speed of 200 mm/s. The echocardiograms were captured from short-axis views of the LV at the midpapillary level. LV percent fractional shortening (LV%FS) was calculated according to the following equation: LV%FS = [(LVEDD − LVESD)/LVEDD] × 100, where LVEDD and LVESD are LV end diastolic and systolic diameter, respectively. All data were calculated from 10 independent cardiac cycles per experiment.

Mitochondrial and nuclear protein and DNA.

Mitochondria and nuclei were isolated by modification of a previously described protocol (34). A fresh heart was cut into small pieces and homogenized by a motor-driven Teflon pestle (Wheaton Scientific, Millville, NJ) in 5 ml of isolation buffer containing 10 mM HEPES, 200 mM mannitol, 70 mM sucrose, and 1 mM EGTA, pH 7.4. Nuclei were obtained by centrifugation at 700 g for 10 min at 4°C. The supernatant was moved to a new tube for centrifuging at 8,000 g for 15 min at 4°C. The pellet was collected and washed again with isolation buffer. The final mitochondrial pellet was used to prepare mitochondrial DNA and mitochondrial protein. The purity of nuclei and mitochondria was assessed by Western blots with antibodies to lamin A (Abcam, Cambridge, MA) and cyclooxygenase IV (Abcam). For purification of DNA, freshly isolated mitochondria or nuclei were lysed and extracted by QIAprep miniprep kit (Qiagen. Hilden, Germany). The concentration of DNA was determined with a spectrophotometer, and the quality of mtDNA was confirmed by Sac II digestion and agarose gel electrophoresis. Protein was isolated from freshly isolated mitochondria or nuclei by lysis in buffer containing 2% SDS, 10% glycerol, and 62.5 mmol/l Tris·HCl (pH 7.0), sonication, and centrifuging at 11,000 g at 4°C for 15 min. The supernatant containing protein was collected, and the concentration was determined by the Lowry method (Pierce kit, Rockford, IL). Western blot (41) protein was probed with rabbit anti-hOGG1 antibody at 1:1,000 dilution (Abcam) at 4°C overnight. The secondary antibody was incubated with the membrane for another hour at room temperature. Finally, the antigen-antibody complexes were visualized with use of an enhanced chemiluminescence kit (ECL; GE Healthcare, Niskayuna, NY).

Assay of OGG1 activity.

The fresh isolated mitochondrial or nuclear pellets were lysed in buffer containing 20 mM HEPES-KOH (pH 7.4), 1 mM EDTA, 1 mM DTT, 300 mM KCl, 5% glycerol, and 0.5% Triton X-100 on ice for 10 min. The lysate was centrifuged at 11,000 g for 30 min at 4°C (36). The supernatant was collected, and the concentration of protein was determined by the Lowry method. OGG1 activity was determined by excising 8-oxoG from a ³²P-labeled oligonucleotide substrate as described previously (36) with modification. Briefly, a synthetic oligonucleotide containing an 8-oxogunine adduct (5′-gaactagtgOatcccccgggctg-3′ where O is 8-oxoguanine; Trevigen, Gaithersburg, MD) was labeled with γ−³²P-ATP by T4 kinase (Invitrogen, Carlsbad, CA) and the labeled oligonucleotide was annealed with its complementary oligonucleotide (5′-gcagcccgggggatccactagttc-3′; Invitrogen). Unincorporated label was removed by passage over a NAP-5 column (GE Healthcare, Niskayuna, NY). Fifty micrograms of protein were incubated with 0.1 pmol of the ³²P-labeled duplex oligonucleotide in excision buffer [20 mM HEPES-KOH (pH 7.6), 5 mM EDTA, 1 mM DTT, 100 mM KCl, and5% glycerol] for 3 h at 37°C in a volume of 20 μl. After incubation, the reaction was terminated by addition of proteinase K (250 μg/ml) and SDS (0.5%) and heating at 50°C for 15 min. Olignucleotides were precipitated with 70% ethanol, 45 mM ammonium acetate, and 0.01 μg/μl glycogen and redissolved in loading buffer containing 70% formamide, 30 mM NaOH, and 0.05% bromophenol blue. The substrate and cleaved oligonucleotide product were separated on a gradient (4–15%) polyacrylamide gel containing 7 M urea (Invitrogen). Radioactivity in the separated DNA bands was quantitated with a PhosphoImager and Imagequant software (Molecular Dynamics, Sunnyvale, CA).

Quantitative RT-PCR.

Cardiac RNA was extracted with Trizol reagent. The total RNA was transcribed to cDNA with Superscript II enzyme and random oligonucleotide primers (Invitrogen). The primers, probes, and reaction buffer for RT-PCR were purchased from AB (Applied Biosystems, Carlsbad, CA) including hOGG1 (Hs00213454_m1), α-MHC (Mm01313844_mH), β-MHC (Mm00600555_m1), atrial natriuretic peptide (ANP) (Mm01255748_g1), brain natriuretic peptide (BNP; Mm01255770_g1), procollagen 1α1(Mm01302043_g1), procollagen 3α1(Mm01254476_m1), 18S RNA (Hs99999901_s1), and 2× Master buffer. RT-PCR was carried out on AB 7300 thermocycler with 35 cycles, each cycle consisted of 95°C for 15 s, 55°C for 15 s, and 75°C for 30 s. 18S RNA was used as endogenous control. Relative abundance of transcripts was determined by the ΔΔCT method.

Histological experiments.

Cryostat sections (5 μm) were fixed in 10% formalin for 15 min and washed three times with PBS. The cryostat slides were incubated with a saturated solution of picric acid containing 0.1% Sirius red for staining collagen and 0.1% fast green for staining noncollagen proteins. Staining was performed in the dark for 2 h. The slides then were rinsed with distilled water, dehydrated with alcohol, and mounted with permount. The sections were visualized and photographed by a blinded observer. Interstitial fibrosis in the sections was scored by a blinded observer against reference images using a scale of 1 to 4 based on the severity of fibrosis with scores of 1 for low, 2 for mild, 3 for moderate, and 4 for severe.

Measurement of 8-oxo-dG content.

Content of 8-oxo-dG was measured using a commercial enzyme-linked immunosorbent assay kit (Trevigen, Gaitherburg MD) as described by Gao et al. (9) with modification. In brief, 0.5 μg of sample DNA or 8-oxo-dG standards from 0 to 60 ng/ml were mixed with anti-8-oxo-dG antibody overnight at 4°C, and the sample DNA and 8-oxo-dG standards were transferred to a 96-well microplate previously coated with albumin:8-oxo-dG adduct and kept in the dark at room temperature for 2 h. The secondary antibody coupled with peroxidase was added to each well and incubation followed for 1 h in the dark, and the plate was washed. After six rounds of washing, the substrate tetramethylbenzidene was added and the reaction continued for 15 min in the dark at room temperature before a stop solution was used to terminate the reaction. The wavelength of 450 nm was set to measure the absorbance of the final solution in each well. Results were calculated according to the standard curve.

Characterization of OGG1 transgenic mice.

Transgenic mice carrying a cardiac targeted transgene for overexpression of human OGG1 were produced as described in the materials and methods. A total of four positive founders were obtained from 22 pups. The founder lines were named OGG1–3, OGG1–6, OGG1–7, and OGG1–10. Lines OGG1–3 and OGG1–10 were fertile and passed on the transgene to ∼50% of offspring. These lines were tested for expression of human OGG1 mRNA and protein and OGG1 activity. The expression of human OGG1 mRNA in heart was significantly elevated (Fig. 1A; P < 0.01 vs. FVB control). Western blots showed that human OGG1 protein was markedly increased in mitochondria of both transgenic lines (Fig. 1B). A possible explanation for the two immunoreactive bands in mitochondria is that the larger band (∼45 kDa) derives from the full human OGG1 protein and the smaller band (∼42 kDa) comes from the OGG1 protein after removal of the mitochondrial targeting sequence. OGG1 activity measured in extracts of cardiac mitochondria was over twofold elevated over FVB in both transgenic lines (Fig. 1C; P < 0.01). Lines OGG1–3 and OGG1–10 had similar human OGG1 expression by all parameters, and line OGG1–10 was used for subsequent experiments. To confirm OGG1 activity in vivo, OGG1–10 mice were injected with 11 mg/kg doxorubicin or saline and 8-oxo-dG content of mtDNA was measured 48 h postinjection. As shown in Fig. 1D, doxorubicin increased 8-oxo-dG content in transgenic and control mtDNA. The OGG1 transgene produced significant reductions of 25% and 31 in 8-oxo-dG content in saline- and doxorubici- treated mice, respectively. These results established cardiac overexpression of active OGG1 in mitochondria of transgenic mice.

The subcellular localization of human OGG1 was assessed on Western blots loaded with total, nuclear, and mitochondrial protein extracts from hearts of control and OGG1–10 transgenic mice (Fig. 1E). Mitochondria contained the largest amount of the small 42-kDa band. Unexpectedly, nuclei from OGG1 hearts contained a strong OGG1 band of ∼46 kDa that was due to the transgene. The reason this was unexpected is that the amino acid sequence of the OGG1–2a isoform (26) used to construct the OGG1 transgene does not contain a nuclear localization signal. The additional OGG1 protein present in transgenic nuclei resulted in a 28% increase in OGG1 enzyme activity (Fig. 1F) compared with FVB nuclei.

Evaluation of heart size and function.

TAC mimics human aortic stenosis with development of pressure-overload-induced left ventricular hypertrophy. TAC or sham surgery was performed on OGG1 and FVB mice and mice were killed for evaluation after 13 wk. As shown in Table 1, TAC surgery significantly increased heart to tibia length and heart to body wt ratio in FVB and OGG1 mice. When compared by heart to body weight ratio, there was less hypertrophy in OGG1 mice (P < 0.05). However, this was partially due to differences in body weight. When heart weight was compared with tibia length, the difference in hypertrophy of OGG1 and FVB mice did not quite reach significance (P = 0.07).

Echocardiography was used to assess the effect of TAC and the OGG1 transgene on cardiac function 13 wk after TAC. The data reported in Table 2 show that TAC mice had significantly impaired cardiac function compared with sham mice. This was evident by several parameters including increased left ventricular end diastolic and systolic diameter as well as reduced fractional shortening. However, there were no significant differences between FVB and OGG1 mice, indicating that OGG1 overexpression did not alter cardiac function nor protect it from the effect of TAC.

Hypertrophic mRNA markers β-MHC, α-MHC, ANP, and BNP were measured to determine whether OGG1 could reduce the fetal pattern of RNA expression induced by pressure overload and hypertrophy (Fig. 2). In sham-treated mice, there were no differences between OGG1 and FVB RNA expression. In both OGG1 and FVB mice, TAC produced the expected increases in expression for β-MHC, ANP, and BNP as well as the expected reduction in the expression of α-MHC. Compared with FVB-TAC mice, OGG1-TAC mice had 41% lower expression of β-MHC (P < 0.05) and OGG1-TAC mice tended toward lower BNP expression (35% lower than FVB-TAC; P = 0.12).

TAC induction of cardiac fibrosis.

Semiquantative analysis showed that 13 wk after TAC surgery, fibrosis was significantly increased in both OGG1 and FVB mice (Fig. 3, A and B; P < 0.05). This analysis also showed that interstitial fibrosis staining was lower in OGG1-TAC mice compared with FVB-TAC mice (P < 0.05). The staining analysis was supported by quantitative RT-PCR assays of cardiac collagen mRNA expression (Fig. 3, C and D). Compared with FVB-TAC mice, expression of collagen 1α1 mRNA and collagen 3α1 mRNA was 49 and 48% lower in OGG1-TAC mice, respectively (P < 0.05 for both mRNAs).

8-oxo-dG content in DNA after TAC.

Thirteen weeks after surgery, we tested whether TAC increased 8-oxo-dG content in nDNA and mtDNA and whether OGG1 overexpression could reduce 8-oxo-dG content. In mitochondria (Fig. 4A), TAC produced a 30% increase in 8-oxo-dG (P < 0.05) of FVB mice. This TAC-induced increase in mtDNA 8-oxo-dG content was completely prevented by overexpression of OGG1. In addition, OGG1-sham mice had 25% lower 8-oxo-dG content in mtDNA than FVB-sham mice (P < 0.05). In nuclear DNA 8-oxo-dG levels were much lower than in mtDNA (Fig. 4B). TAC produced a significant 10% increase in 8-oxo-dG content in both types of mice but there were no significant differences for nuclear DNA 8-oxo-dG content between OGG1 and FVB mice. Overall, these results show that OGG1 provides protection against some components of cardiac dysfunction in the TAC model including reductions in fibrosis, β-MHC expression, and 8-oxo-dG content in mtDNA.