SIRT3 blocks myofibroblast differentiation and pulmonary fibrosis by preventing mitochondrial DNA damage

Abstract

Myofibroblast differentiation is a key process in the pathogenesis of fibrotic diseases. Transforming growth factor-β₁ (TGF-β₁) is a powerful inducer of myofibroblast differentiation and is implicated in pathogenesis of tissue fibrosis. This study was undertaken to determine the role of mitochondrial deacetylase SIRT3 in TGF-β₁-induced myofibroblast differentiation in vitro and lung fibrosis in vivo. Treatment of human lung fibroblasts with TGF-β₁ resulted in increased expression of fibrosis markers, smooth muscle α-actin (α-SMA), collagen-1, and fibronectin. TGF-β₁ treatment also caused depletion of endogenous SIRT3, which paralleled with increased production of reactive oxygen species (ROS), DNA damage, and subsequent reduction in levels of 8-oxoguanine DNA glycosylase (OGG1), an enzyme that hydrolyzes oxidized guanine (8-oxo-dG) and thus protects DNA from oxidative damage. Overexpression of SIRT3 by adenovirus-mediated transduction reversed the effects of TGF-β₁ on ROS production and mitochondrial DNA damage and inhibited TGF-β₁-induced myofibroblast differentiation. To determine the antifibrotic role of SIRT3 in vivo, we used the bleomycin-induced mouse model of pulmonary fibrosis. Compared with wild-type controls, Sirt3-knockout mice showed exacerbated fibrosis after intratracheal instillation of bleomycin. Increased lung fibrosis was associated with decreased levels of OGG1 and concomitant accumulation of 8-oxo-dG and increased mitochondrial DNA damage. In contrast, the transgenic mice with whole body Sirt3 overexpression were protected from bleomycin-induced mtDNA damage and development of lung fibrosis. These data demonstrate a critical role of SIRT3 in the control of myofibroblast differentiation and lung fibrosis.

idiopathic pulmonary fibrosis (IPF) is a progressive fatal disease characterized by parenchymal fibrosis and structural distortion of the lungs. Age-adjusted mortality due to pulmonary fibrosis is increasing, and it poses a vexing clinical challenge given the lack of effective medical therapy. IPF is thought to be a disorder of abnormal wound healing (59), in which the initial trigger to the fibrotic response is injury to alveolar epithelial cells, followed by an exuberant, nonresolving wound-healing response (55). Injury of alveolar epithelial cells is thought to result in the elaboration of a fibrinous provisional matrix and activation of several proinflammatory, procoagulant, and profibrotic mediators, among which transforming growth factor-β₁ (TGF-β₁) is the most established (49). Fibroblasts respond to TGF-β₁ by altering their gene expression profile with de novo expression of cytoskeletal and contractile proteins normally found within smooth muscle cells, modified focal adhesion complexes, and components of the extracellular matrix (25, 35). These smooth muscle-like fibroblasts are called myofibroblasts and display a phenotype that is an intermediate state between fibroblasts and smooth muscle cells (14, 26). Several cytoskeletal and smooth muscle proteins are expressed in myofibroblasts, including smooth muscle α-actin (SMA), the most established marker for myofibroblast differentiation (26, 57). Additionally, induction of myofibroblast phenotype is also associated with secretion of extracellular matrix proteins (collagen isoforms, cellular fibronectin, etc.) and of profibrotic factors [connective tissue growth factor (CTGF), insulin-like growth factor (IGF-1) etc.], thus perpetuating the ongoing tissue remodeling and fibrosis. Myofibroblasts are invariably found in histological sections of human lung specimens from patients with pulmonary fibrosis, and are thought to be a critical pathogenic mechanism responsible for the progressive nature of IPF (35). Therefore, disrupting cellular mechanisms that induce and maintain the myofibroblast phenotype may be a potential strategy to attenuate the ongoing fibrotic response during pulmonary fibrosis.

Increased reactive oxygen species (ROS) production from mitochondria and subsequent mitochondrial DNA (mtDNA) damage are important events for pathogenesis of tissue fibrosis in animal models (6, 17, 63). Excessive ROS production can oxidize guanine residues to 7,8-dihydro-8-oxoguanine (8-oxo-dG), and addition of these adducts to DNA can lead to DNA damage. The enzyme that hydrolyzes 8-oxo-dG from DNA is 8-oxoguanine-DNA glycosylase-1 (OGG1), a DNA repair enzyme (50). The mitochondrial DNA isolated from OGG1-null mice were shown to have 20-fold more 8-oxo-dG than the mtDNA from wild-type animals (12). Furthermore, tissue-specific overexpression of OGG1 was shown to protect mitochondrial DNA and reduce organ fibrosis following oxidative stress. Downregulation of mitochondrial OGG1 levels has been also shown to promote the aging process (56). In line with this, OGG1-null mice were shown to have increased mitochondrial DNA damage and exacerbated pulmonary fibrosis when exposed to asbestos (10). TGF-β₁ is also a well-known inducer of ROS, which is implicated in mitochondrial DNA damage (36, 54). Together, these studies support the rationale for hypothesizing an inverse relationship between OGG1 and TGF-β₁ in regulating fibrosis. One of the enzymes which was found to maintain cellular OGG1 and TGF-β₁ levels is the mitochondrial deacetylase SIRT3 (9, 51).

SIRT3 is a member of the sirtuin family of class III histone deacetylases, which is primarily localized in mitochondria. SIRT3 has received considerable attention in recent years for its ability to regulate wide range of cellular processes including cell growth, metabolism, apoptosis, and mitochondrial detoxification (8, 16, 52). Correspondingly, the substrates of SIRT3 are also diverse, including enzymes critical for maintaining health of mitochondria. SIRT3-mediated deacetylation increases activity of MnSOD, which is vital for controlling the cellular ROS levels (40, 53). SIRT3 also deacetylates NDUFA9 in complex I, SDHA in complex II, and the ATP synthase β in complex V of the electron transport chain, leading to increased bioenergetic capacity of mitochondria (2, 11, 41). SIRT3 also deacetylates and activates acetyl-CoA synthetase 2 that helps to provide substrates for TCA cycle (22, 47). It also deacetylates the TCA cycle enzyme isocitrate dehydrogenase-2 (IDH2) and urea cycle enzyme ornithine transcarbamylase (23, 64). In addition, an enzyme involved in long-chain fatty acid oxidation, long-chain acyl-coenzyme A dehydrogenase (LCAD), is also found to be a substrate of SIRT3 (27). SIRT3 also protects mitochondria from DNA damage by maintaining OGG1 levels and activating DNA repair mechanisms (5, 9, 61). Based on these and other studies SIRT3 is considered a mitochondrial fidelity protein.

In this study we show that during development of lung fibrosis SIRT3 levels were downregulated. Loss of SIRT3 caused decreased expression of OGG1 and accumulation of 8-oxo-dG adducts, which led to increased mitochondrial DNA damage and concomitant development of tissue fibrosis. Overexpression of SIRT3 maintained OGG1 levels, leading to reduced mitochondrial DNA damage, and thereby prevented the development of bleomycin-induced lung fibrosis.

MATERIALS AND METHODS

Cell culture.

Early-passage human lung fibroblasts (IMR-90; Institute for Medical Research, Camden, NJ) were used for performing in vitro experiments. The cells were maintained in medium consisting of Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and 100 units/ml penicillin-streptomycin. For experimental treatments, cells were plated on 35- or 100-mm dishes at a density of 10⁶ cells/dish and incubated in 5% CO₂-95% air. Once cells were 80% confluent they were infected with adenovirus encoding SIRT3 (Ad.SIRT3) or Ad.empty virus (Ad.mk) in growth medium for overnight. Next day cells were treated with TGF-β₁ (10 ng/ml) in DMEM containing 1% FBS and 1% antibiotics and harvested 48 h later. Adenovirus for recombinant human SIRT3 was purchased from Vector Biolabs. In studies with EUK134, cells were pretreated with EUK134 (10 μM) for 4 h before TGF-β₁ treatment. For immunofluorescence studies, cells were plated on glass coverslips placed in 12-well plates. All reagents were purchased from Sigma.

Animal experiments.

All animal work described in this study was reviewed and approved by the IACUC of the University of Chicago. Sirt3 knockout (129Sv) mice were originally provided by Dr. F. W. Alt. These mice are propagated and maintained in our colony. Sirt3 transgenic mice (whole body SIRT3 over expression) were generated and characterized by Eric Verdin's group and have been used and described by us previously (34). Bleomycin-induced lung fibrosis was developed as described before (30). Bleomycin (Bleocip, Cipla) at a dose of 1 unit/kg was instilled intratracheally to 10- to 12-wk-old mice. Mice were euthanized on day 21 and lungs were surgically harvested. Tissues were fixed in 10% formalin for 48 h and embedded into paraffin blocks to cut sections for histochemical analysis. A portion of lung tissue was also snap-frozen for biochemical analysis. CRi Pannoramic whole slide scanner was used to obtain images of the stained sections. The severity of lung fibrosis was graded on a scale of 0 to 7 by examining five randomly selected fields of the same lung at ×100 magnification.

ROS and cell death detection.

Cellular ROS levels and cell death were detected using CM-H₂DCFDA reagent (Invitrogen) and 7- aminoactinomycin D (7AAD) (ThermoFisher Scientific) respectively as per the manufacturer’s instructions. In brief, for ROS, lung fibroblasts were infected with Ad.SIRT3 or Ad.mk empty virus vector. Twenty-four hours after infection cells were treated with TGF-β₁ or vehicle for additional 48 h. Cells were stained with CM-H₂DCFDA. For cell death assay only cells treated with or without TGF-β₁ were analyzed. Cells were acquired by FACSCalibur and analyzed with use of FlowJo software. The mean fluorescence intensity of cells positive for CM-H₂DCFDA and 7AAD staining were determined.

Antibodies.

Primary antibodies of fibronectin, tubulin, GAPDH, actin, and OGG1 and secondary antibodies of anti-mouse, anti-rabbit, and anti-goat were bought from Santa Cruz Biotechnology. Secondary Alexa Fluor antibodies were purchased from Invitrogen. Anti-SIRT3 antibody was from Cell Signaling. SMA antibody was bought from Sigma, while collagen type I antibody was from Calbiochem.

RNA isolation and Real-time PCR analysis.

Total RNA was isolated from cells as well as from mouse lungs by use of TRIzol reagent (Invitrogen). The residual genomic DNA was digested by incubating the RNA preparation with 0.5 units of RNase-free DNase-1 per microgram of RNA in a reaction buffer for 15 min at room temperature, followed by heat inactivation at 90°C for 5 min. Two micrograms of DNase-treated RNA was reverse transcribed by use of Fermentas, RevertAid First Strand cDNA Synthesis Kit. The resultant cDNA was diluted 10-fold before PCR amplification. A reverse transcriptase minus reaction served as a negative control. The mRNA levels were measured by SYBR Green real-time PCR. Primer sequences, for human, were as follows: fibronectin forward CAAGTATGAGAAGCCTGGGTCT, fibronectin reverse 5′-TGAAGATTGGGGTGTGGAAG-3′; collagen-I forward 5′-GCTTCACCTACAGCGTCAC-3′, collagen-I reverse 5′-TGGGATGGAGGGAGTTTACA-3′; β-actin forward 5′-AAGGCCAACCGCGAG-3′, β-actin reverse 5′-TAATGTCACGCACGATTCCCG-3′; SIRT3 forward 5′-ACCCAGTGGCATTCCAGAC3, SIRT3 reverse 5′-GGCTTGGGGTTGTGAAAGAAG-3′; OGG1 forward 5′-GTGCCCGTTACGTGAGTGCCAGTGC-3′, OGG1 reverse 5′-AGAGAAGTGGGGAATGGAGGGGAAGGTG-3′.

Mitochondrial DNA damage assay.

Genomic DNA was isolated using Qiagen Genomic-tip 20/G and Qiagen DNA Buffer Set (Qiagen, Gaithersburg, MD) per the manufacturer’s instruction. Eluted DNA was precipitated with isopropanol overnight at −80°C and centrifuged 12,000 g for 60 min. DNA was washed with 70% ethanol and dissolved in Tris-EDTA (TE) buffer. PCR was performed using Ex Taq (Clontech, Mountain View, CA). Primer sequences for long PCR are forward 5′-TGAGGCCAAATATCATTCTGAGGGGC-3′; reverse 5′-TTTCATCATGCGGAGATGTTGGATGG-3′. Short PCR was performed using forward primer sequence 5′-CATGCCCATCGTCCTAGAAT-3′, reverse primer 5′-ACGGGCCCTATTTCAAAGAT-3′ (43, 62). Resultant PCR products were quantified using PicoGreen (Life Technologies). Values obtained from the long fragments were normalized using values from short fragments. The lesion frequency per amplicon was then calculated as λ = −ln(AD/AO), where AD/AO is the ratio of the amplification of the treated samples (AD) to the amplification of the control samples (AO).

Immunofluorescent microscopy.

Human lung fibroblasts grown on 12-mm coverslips were infected with Ad.SIRT3 or Ad.empty virus vector. Twenty-four hours after infection cells were treated with 10 ng/ml TGF-β₁ for 48 h. Cells were washed with PBS, and fixed with 3.7% formaldehyde in PBS for 15 min, followed by permeabilization with 0.1% Triton X-100 for 5 min. Cells were then blocked with 10% BSA in PBS followed by incubation with primary antibody overnight at 4 °C. Thereafter, cells were incubated with a secondary antibody conjugated with either Alexa Fluor 594 or FITC for 1 h. Cells were washed and mounted in ProLong Gold antifade reagent with DAPI. Cells were visualized using a Leica SP2 laser scanning microscope. To quantify the myofibroblast transformation total fluorescence of each cell (100 cells in each group) were measured using the Image J software, and the results are presented as relative expression levels of α-SMA, collagen, or fibronectin.

OGG1 RNA interference.

Small interfering RNAs (siRNAs) for OGG1 and scrambled siRNA were obtained from Qiagen. Human lung fibroblasts were transfected with OGG1 siRNA (5 nmol/l) using Lipofectamine RNAiMAX transfection reagent (Invitrogen) for 48 h. Scrambled siRNA (20 nmol/l)-transfected cells were used as a negative control. OGG1 depletion was confirmed by Western blotting.

Measurement of 8-oxo-dG levels.

8-Oxo-dG levels were measured using 8-oxo-dG ELISA kit (Trevigen) as per manufacturer’s instructions (38).

Statistical analysis.

All values are expressed as means ± SE. At least three replicates were applied for each experiment. Statistical difference between groups were determined by Student's t-test.

RESULTS

SIRT3 overexpression inhibits TGF-β₁-induced myofibroblast differentiation.

As shown in Fig. 1, A–D, treatment of human lung fibroblasts (HLFs) with TGF-β₁ resulted in an increased expression of profibrotic markers including α-SMA, fibronectin, and collagen-1a, as determined by Western blotting. TGF-β₁-induced expression of these proteins was blocked by transduction of cells with Ad.SIRT3, but not with control adenovirus Ad.mk. Consistent with the results obtained from Western blotting, we also observed increased mRNA levels of α-SMA, fibronectin, and collagen-1a in TGF-β₁-treated cells, which was attenuated by transduction of Ad.SIRT3, but not of Ad.mk (Fig. 1, E–G). We also observed a significant decrease in Sirt3 (both protein and mRNA) expression in TGF-β₁-treated cells (Fig. 1, A and H). This implies a TGF-β₁-mediated transcriptional downregulation of SIRT3. To further validate the role of SIRT3 in TGF-β₁-induced myofibroblast differentiation, we analyzed expression of profibrotic markers by immunocytochemistry followed by confocal microscopy. HLFs transduced with control adenovirus and stimulated with TGF-β₁ showed marked increase in the expression of α-SMA, fibronectin, and collagen-1a. The expression of these markers was again blocked in cells transduced with Ad.SIRT3 (Fig. 2, A–C). We also noticed that, in the same field, cells negative for SIRT3 overexpression had increased α-SMA expression in response to TGF-β₁ compared with those positive for SIRT3 expression (Fig. 2D). These results thus indicate that SIRT3 has the ability to block lung fibroblast differentiation.

Fig. 1.

SIRT3 overexpression blocks TGF-β₁ induced expression of fibrotic markers in lung fibroblasts. A: human lung fibroblasts were transduced with Ad.SIRT3 or the control adenovirus (Ad.mk). Following 24 h of transduction, cells were treated with 10 ng/ml TGF-β₁ for 48 h. Cell lysates were subjected to Western blotting with antibodies against the indicated proteins. B–D: quantification of relative protein levels in different groups of cells. E–H: relative mRNA levels of α-SMA, fibronectin, collagen 1a, and SIRT3 as measured by RT-PCR. Values are means ± SE, n = 3. Col1a, collagen-1a; α-SMA, α-smooth muscle actin; Ad.T3, Ad.SIRT3.

Fig. 2.

SIRT3 overexpression blocks TGF-β₁-induced myofibroblast differentiation. A–C: human lung fibroblasts were transduced with Ad.SIRT3 or the control adenoviruses and then treated with TGF-β₁ (10 ng/ml) for 48 h. Cells were immunostained for α-SMA, collagen 1a, and fibronectin and costained with DAPI to mark the nuclei. D: after TGF-β₁ treatment cells were costained with α-SMA, DAPI, and SIRT3. Note that SIRT3-overexpressing cells (yellow arrows) show markedly reduced expression of α-SMA, compared with cells negative for SIRT3 expression (red arrow).

SIRT3 overexpression blocks myofibroblasts differentiation independent of ROS synthesis.

Previous studies have implicated ROS in cellular responses to TGF-β₁ (1, 24, 34, 60). TGF-β₁ treatment has been also shown to induce ROS production in HLF (35, 54). Because SIRT3 is known to activate MnSOD (SOD2) and reduce ROS synthesis from mitochondria, we tested whether antifibrotic activity of SIRT3 is dependent on its ability to reduce cellular ROS levels. HLFs were transduced with Ad.SIRT3 or control vector followed by treatment with TGF-β₁. ROS levels were measured by staining cells with the CM-H2DCFDA, a nonfluorescent dye that fluoresces upon oxidation by ROS. We also tested effects of an antioxidant EUK134, a MnSOD/catalase mimetic. As shown in Fig. 3, A and B, both SIRT3 overexpression and EUK134 treatment reduced the TGF-β₁-induced ROS synthesis to a similar degree. However, while augmenting SIRT3 by overexpression blocked the TGF-β₁-induced expression of α-SMA (Fig. 1A), EUK134 had no effect in myofibroblasts (Fig. 3, C and D). Since increased ROS synthesis is often implied in inducing apoptosis we also tested effect of TGF-β₁ treatment on HLF cell death. However, we observed no significant cell death after TGF-β₁ treatment (Fig. 3E). These results thus suggest that SIRT3 inhibits myofibroblast differentiation independently of its ability to prevent ROS synthesis.

Fig. 3.

SIRT3 controls myofibroblast differentiation independently of its ability to reduce ROS levels. A: human lung fibroblasts were transduced with Ad.SIRT3 or the control adenovirus and then stimulated with TGF-β₁ for 48 h. In another group, TGF-β₁-stimulated myofibroblasts were treated with the antioxidant EUK134 (10 µM). Cells were stained with CM-H2DCFDA and ROS levels were measured by fluorescence-activated cell sorter. B: quantification of mean fluorescence intensity in different groups of cells. Means ± SE, n = 4. C: cell lysates were analyzed by Western blotting with desired antibodies. D: quantification of relative protein levels in different groups of cells, means ± SE, n = 4. E: quantification of TGF-β₁-mediated cell death. Means ± SE, n = 4. ns, Not significant.

SIRT3 prevents mitochondrial DNA damage.

Given that reduced ROS synthesis is not the main mechanism through which SIRT3 blocks TGF-β₁-induced fibrosis, we considered alternative mechanisms. It has been shown previously that reduced SIRT3 levels lead to mitochondrial DNA damage owing to defects in DNA repair machinery (9, 38). In our experiments, we observed that TGF-β₁ treatment resulted in a significant decrease of SIRT3 expression in HLFs (Figs. 1A and 4A). Therefore, we examined mitochondrial DNA lesions by PCR. As shown in Fig. 4B, TGF-β₁ treatment induced a significant mtDNA damage, which was significantly reduced when SIRT3 was overexpressed, suggesting that SIRT3 activation protects mtDNA from TGF-β₁-induced damage. Previously, SIRT3 has been shown to deacetylate and stabilize OGG1, a DNA repair enzyme maintaining integrity of mtDNA (38). We therefore measured OGG1 levels in cells treated with TGF-β₁. Consistent with reduced SIRT3 levels, OGG1 levels were also decreased in cells treated with TGF-β₁, whereas SIRT3 overexpression reversed this effect of TGF-β₁ on OGG1 levels (Fig. 4, C and D). To determine whether OGG1 levels were regulated transcriptionally or posttranscriptionally, we analyzed OGG1 mRNA levels in TGF-β₁-treated cells. Real-time qPCR analysis showed that TGF-β₁ treatment did not significantly change the OGG1 mRNA levels (Fig. 4E). Finally, we assessed the role of ROS production in TGF-β₁-induced downregulation of OGG1, using antioxidant EUK134 that inhibited ROS production induced by TGF-β₁. As shown in Fig. 4F, pretreatment with EUK134 did not rescue downregulation of OGG1, whereas SIRT3 overexpression restored OGG1 expression in TGF-β₁-treated cells, suggesting that downregulation of SIRT3, and not ROS production, contributes to decrease in OGG1 expression. To further confirm that SIRT3 blocks TGF-β₁-induced fibroblast differentiation by maintaining OGG1 levels, we repeated the same experiments in cells subjected to OGG1 knockdown. As shown in Fig. 4G, SIRT3 overexpression failed to block TGF-β₁ induced α-SMA expression in OGG1 depleted cells, while it was capable of doing so in the wild-type cells (Fig. 1A). These data suggested that SIRT3 prevents mitochondrial DNA damage in TGF-β₁-treated cells by maintaining OGG1 levels, which could be a mechanism for the antifibrotic function of SIRT3.

Fig. 4.

SIRT3 controls OGG1 levels and protects cells from TGF-β₁-induced mtDNA damage. Human lung fibroblasts were transduced with Ad.SIRT3 or the control adenoviruses or pretreated with the antioxidant EUK134 as indicated, and then stimulated with TGF-β₁ for 48 h. A: relative SIRT3 levels in control and TGF-β₁-treated cells as measured by Western blotting, means ± SE, n = 3. B: mtDNA damage was assessed by PCR amplification of DNA as described in materials and methods. Values are means ± SE, n = 4. C: cell lysate was subjected to Western blotting with antibodies against OGG1 and actin. D: quantification of relative OGG1 protein levels in different groups of cells, means ± SE, n = 3. E: quantification of OGG1 mRNA levels in control and TGF-β₁-treated cells, means ± SE, n = 3. F: cell lysates were analyzed by Western blotting for the indicated proteins. Note that, while SIRT3 overexpression restores OGG1 levels, EUK134 treatments did not. G: OGG1 was depleted in human lung fibroblasts by using OGG1 siRNA. These cells were later transduced with Ad.SIRT3 or the control adenoviruses for 24 h followed by TGF-β₁ treatment. SIRT3 augmentation failed to suppress α-SMA expression in absence of OGG1. ns, Not significant.

Protective role of SIRT3 in pathogenesis of pulmonary fibrosis.

To test the antifibrotic function of SIRT3 in vivo, we utilized the bleomycin model of pulmonary fibrosis. Bleomycin or saline was instilled intratracheally to Sirt3 knockout (Sirt3-KO) or transgenic (Sirt3-Tg) mice having whole body Sirt3 overexpression or wild-type littermates to induce fibrosis. Control mice received saline instillation. Consistent with previous results, adult (12-wk-old) Sirt3-KO mice, but not wild-type controls, showed the presence of fibrosis without bleomycin treatment, as estimated by Masson’s trichrome staining (51). Bleomycin instillation induced significant lung fibrosis in wild-type mice and further exacerbated the fibrosis in Sirt3-KO mice (Fig. 5, A and C). In contrast, Sirt3-Tg mice had significantly attenuated fibrosis in response to bleomycin (Fig. 5, B and D). To further support these findings, we analyzed lung extracts by Western blotting. There was increased expression of collagen-1a in Sirt3-KO lungs as compared with wild-type controls, and it was further enhanced after treatment of Sirt3-KO mice with bleomycin (Fig. 6, A and B). In contrast, we found significantly reduced levels of collagen-1a in Sirt3-Tg mice treated with bleomycin as compared with wild-type controls, thus suggesting an antifibrotic function of Sirt3 (Fig. 6, C and D).

Fig. 5.

Control of bleomycin-induced lung fibrosis by SIRT3. Sirt3-KO (A) or Sirt3-tg (B) mice along with corresponding wild-type (WT) control littermates were treated with vehicle or bleomycin (Bleo) and analyzed as described in materials and methods. Representative lung sections stained with Masson’s trichrome for determining fibrosis (blue). Scale bar 100 μm. C and D: quantification of lung fibrosis in different groups of mice. Means ± SE, n = 10.

Fig. 6.

Expression of collagen 1a and OGG1 levels in the lung tissues of the bleomycin model. Sirt3-KO (A, B, E) or Sirt3-tg (C, D, F) mice along with corresponding wild-type control littermates were treated with vehicle or bleomycin as described in materials and methods. A and C: lung extracts were analyzed by Western blotting with antibodies against collagen 1a, OGG1, and SIRT3, or actin. B, D, E, F: quantification of protein levels in different treatment groups of mice as indicated. Means ± SE, n = 5.

Given our in vitro studies showing significant reduction of OGG1 levels during myofibroblast differentiation, we also analyzed OGG1 levels in mice subjected to bleomycin instillation. Bleomycin treatment reduced the lung OGG1 levels in wild-type mice, which was additionally reduced in Sirt3-KO mice (Fig. 6, A and E). However, Sirt3-Tg mice retained OGG1 expression in the lung following bleomycin treatment (Fig. 6, C and F). Reduced OGG1 levels may indicate increased mtDNA damage, which would be manifested by elevated levels of 8-oxo-dG (38). We therefore measured 8-oxo-dG levels in the DNA isolated from the mouse lungs. Compared with wild-type controls, Sirt3-KO lungs exhibited significantly higher levels of 8-oxo-dG, which were additionally elevated after bleomycin treatment (Fig. 7A). In contrast, we found no significant accumulation of 8-oxo-dG adducts in lungs from Sirt3-Tg mice following bleomycin treatment (Fig. 7B). These results suggest that SIRT3 protects lungs from bleomycin-induced fibrosis by maintaining OGG1 levels and preserving integrity of the mtDNA.

Fig. 7.

Relative 8-oxo-dg levels in the DNA of lungs isolated from different groups of mice. Sirt3-KO (A) and Sirt3-Tg (B) mice along with corresponding wild-type controls were treated with vehicle or bleomycin for 21 days and 8-oxo-dg levels measured as described in materials and methods. Means ± SE, n = 10; NS, not significant.

DISCUSSION

In this study we demonstrated a role of SIRT3 in protecting lungs from developing fibrosis. We found that SIRT3 expression is inversely related to TGF-β₁-induced lung myofibroblast differentiation in vitro and bleomycin-induced pulmonary fibrosis in vivo. Loss of SIRT3 correlated with downregulation of OGG1 and concomitant damage to mtDNA. Overexpression of SIRT3 was associated with stabilization of OGG1 and protection from mtDNA damage, both in TGF-β₁-treated fibroblasts and in the bleomycin model of pulmonary fibrosis. Together, our studies suggest the role of SIRT3 as a negative regulator of myofibroblast differentiation and lung fibrosis and suggest a potential mechanism for this role through stabilization of OGG1 and protection from mtDNA damage.

TGF-β₁ is one of the most established profibrotic cytokines (7, 31, 49). TGF-β₁ has been localized to areas of fibrosis in both the experimental pulmonary fibrosis and human disease (7, 37, 46). Lung-targeted transgenic overexpression of TGF-β₁ results in the development of airway fibrosis in animals (32, 49). Conversely, inhibition of TGF-β₁ via soluble TGF-β₁ receptors can inhibit in vivo fibrogenesis (19, 58). We have previously shown that SIRT3 deficiency is associated with increased TGF-β₁ production in the mouse heart (51). Our present studies suggest that TGF-β₁ itself induces loss of SIRT3 in lung fibroblasts, which may be a mechanism for bleomycin-induced loss of SIRT3 in the lungs. In this model, TGF-β₁ induces the fibrotic response by shutting down the antifibrotic processes. In this study we also found a reduction in SIRT3 mRNA levels after TGF-β₁ treatment, suggesting that downregulation of SIRT3 gene transcription could be one of the mechanisms through which TGF-β₁ induces myofibroblast transformation and subsequent pathogenesis. This observation is consistent with a recent study where downregulation of SIRT3 was shown in TGF-β₁-treated skin fibroblasts and during systemic sclerosis (4). However, the mechanism of TGF-β₁-mediated downregulation of SIRT3 gene transcription is not known and warrants further investigation.

Increased ROS levels can induce oxidation of proteins, lipids, and DNA (36). Since mitochondrial DNA is in close proximity to the site of ROS synthesis, it is highly susceptible to damage (48). In this study we observed that TGF-β₁-mediated transformation of lung myofibroblasts was associated with increased ROS levels, which is consistent with others and could be a potential mechanism for mtDNA damage (35, 54). However, our studies showed that treatment of cells with an antioxidant, EUK134, which is a biomimetic of SOD2 and catalase, was unable to prevent TGF-β₁-induced myofibroblast differentiation, even though it was capable of blocking cellular ROS levels. These results suggest that ROS-independent factors are likely to be contributing to increased mtDNA damage following TGF-β₁ treatment, and SIRT3 might employ additional mechanisms to prevent tissue fibrosis besides its ability to suppress cellular ROS levels. This notion is supported by several previous reports where ROS scavengers were found ineffective in reducing tissue fibrosis in patients (21, 28). Similarly, the role of oxidative damage as the cause of tissue fibrosis associated with the aging process has been challenged by recent studies (15, 42). These findings led us to hypothesize that SIRT3 deficiency might have initiated tissue fibrosis by compromising integrity of mitochondrial DNA due to defects in mtDNA repair machinery.

Previous studies have shown that OGG1 deficiency causes mitochondrial DNA damage. OGG1 is a DNA glycosylase involved in the hydrolysis of 8-oxo-dG, an oxidized version of guanine residue from the DNA (12), which is the most abundant DNA lesion. Downregulation of OGG1 is reported in different types of cancers, diabetes, neurological disorders, and many aging-associated diseases (45). OGG1 is an acetylated protein and this posttranslational modification makes the protein susceptible to degradation by proteases (9). In this study we found that TGF-β₁-induced mtDNA damage can be attenuated by overexpression of SIRT3, in parallel with stabilization of OGG1 in lung fibroblasts. TGF-β₁ treatment did not alter the mRNA levels of OGG1, suggesting that OGG1 was degraded posttranscriptionally. These data are in accordance with our previous studies where we demonstrated that SIRT3 physically binds to and deacetylates OGG1 and thereby prevent its degradation (38). We also found that degradation of OGG1 is independent of ROS, since EUK134 treatments did not help to increase OGG1 levels following TGF-β₁ treatment (Fig. 4F). Furthermore, in OGG1-depleted fibroblasts overexpression of SIRT3 was unable to attenuate fibroblasts to myofibroblasts transformation, suggesting that mtDNA damage indeed induces fibrosis in cells. Here we speculate that increased mitochondrial DNA damage may compromise the structural integrity of mitochondrial proteins involved in oxidative phosphorylation, cellular respiration, and ATP production. Reduced ATP levels and NAD⁺/NADH ratio create pseudohypoxia in cells, leading to further increase in activation of genes including HIF-1α that can induce fibrotic changes in cells (20). It is also possible that reduced OGG1 and increased ROS levels generated by loss of SIRT3, in response to fibrotic stimuli, contributed cumulatively to induce mtDNA damage and concomitantly fibrosis. These findings suggested that SIRT3 prevents mtDNA damage by blocking degradation of OGG1.

In our in vivo experiments intratracheal instillation of bleomycin induced severe lung fibrosis characterized by increased expression of collagen 1 and distortion of lung architecture. Consistent with our in vitro observations, SIRT3 deficiency exacerbated bleomycin induced fibrosis in mice, whereas Sirt3-Tg mice showed significantly reduced fibrosis in response to bleomycin treatment, suggesting that SIRT3 is an endogenous negative regulator of lung fibrosis. These results are in agreement with a recent study showing that, as compared with wild-type, SIRT3-deficient mice are more susceptible to asbestos-induced mtDNA damage and induction of pulmonary fibrosis (29). The mechanism by which SIRT3 may mediate its antifibrotic effects in vivo is also consistent with our in vitro data, as we observed that bleomycin, potentially through the induction of TGF-β, reduced SIRT3 and OGG1 levels and elevated 8-oxo-dG levels (indicative of mtDNA damage) in mouse lungs. Accumulation of 8-oxo-dG adducts was highest in Sirt3-KO mice, while SIRT3 overexpression prevented the accumulation of 8-oxo-dG adducts in bleomycin-treated mice. These results suggest that damage of mtDNA repair could be a significant factor responsible for the development of lung fibrosis.

SIRT3 is considered a mitochondrial fidelity protein regulating many different functions of mitochondria (18, 33). Based on our results, we propose a model showing cascade of events leading to bleomycin-induced lung fibrosis, and how SIRT3 activation blocks toxic effects of this drug to lungs (Fig. 8). SIRT3 has been shown to regulate mitochondrial ATP synthesis, fusion-fission dynamics of mitochondria as well as mitochondrial-dependent apoptosis (2, 39, 44). Although in our study we presented data showing that SIRT3 blocks bleomycin-induced lung fibrosis by preserving integrity of mtDNA, involvement of other SIRT3-dependent factors helping to maintain the health of mitochondrial population cannot be excluded for its antifibrotic effects.

Fig. 8.

A model for the mechanism of antifibrotic function of Sirt3 in lungs. Bleomycin treatment induces TGF-β₁ levels leading to reduced expression of SIRT3. Loss of SIRT3 subsequently leads to loss of DNA repair enzyme OGG1, resulting in mtDNA damage. Loss of SIRT3 also leads to increased ROS synthesis from mitochondria, which can again contribute to mtDNA damage. These changes promote myofibroblast transformation and induction of lung fibrosis. Our present results show that SIRT3 activation blocks the induction of lung fibrosis by stabilizing OGG1 levels, thereby mitigating mtDNA damage.

In summary, our results described here illustrate that mitochondrial DNA damage is a potential mediator of lung fibrosis. Overexpression of SIRT3 restored the levels and activity of OGG1 and prevented mtDNA damage to rescue the transformation of lung fibroblasts to myofibroblasts. Therefore, SIRT3 activation could be a potential therapeutic strategy to protect the lungs from developing fibrosis and/or to alleviate fibrosis where it already exists.

GRANTS

This study was supported by the National Heart, Lung, and Blood Institute RO1 grants HL-117041 and HL-111455 to M. P. Gupta and 1R56HL127395 to N. O. Dulin.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

S.B., V.B.P., A.K., and S.S. performed experiments; S.B., V.B.P., and M.P.G. analyzed data; S.B., V.B.P., and N.O.D. interpreted results of experiments; S.B. prepared figures; S.B. drafted manuscript; G.M., N.O.D., and M.P.G. edited and revised manuscript; E.V. and M.P.G. approved final version of manuscript.