Metabolic Reprogramming Is Required for Myofibroblast Contractility and Differentiation^*

Reagents

Porcine platelet-derived TGF-β1 was purchased from R&D Systems (Minneapolis, MN), protease inhibitor mixture set III was from EMD Chemicals (San Diego, CA), p38 MAPK inhibitor SB202190 was from Tocris Bioscience (Minneapolis, MN), and MitoTracker® was from Life Technologies. We purchased antibodies to β-actin (clone AC-15), fibronectin (clone IST-4), and α-tubulin (clone B-5-1-2) from Sigma; α-SMA (clone ASM-1) from American Research Products (Belmont, MA); TFAM, total PGC-1α, hexokinase II (HKII), total MLC20, phospho-MLC20, GAPDH, VDAC, p38 MAPK from Cell Signaling Technology (Boston, MA); TOM20 and lamin A/C from Santa Cruz Biotechnology (Dallas, TX); and phospho-PGC-1α from R&D Systems. All other reagents were purchased from Sigma unless otherwise specified.

Cell Culture

Human fetal lung fibroblasts (IMR-90 cells) at low population doubling (PDL 7) were purchased from Coriell Cell Repositories (Camden, NJ). All cells were cultured in DMEM (Life Technologies) supplemented with 10% fetal calf serum (HyClone Laboratories, Logan, UT), 100 units/ml penicillin, 100 μg/ml streptomycin, and 1.25 μg/ml amphotericin B at 37 °C in 5% CO₂, 95% air. Cells treated with TGF-β1 were cultured in serum-free medium for 24 h prior to and during treatment.

Western Immunoblotting

We prepared cell lysates in radioimmune precipitation assay buffer (150 mm NaCl, 1.0% IGEPAL® CA-630, 0.5% sodium deoxycholate, 0.1% SDS, 50 mm Tris, pH 8.0; Sigma-Aldrich), and cytosolic and nuclear lysates were prepared using the Pierce NE-PER kit according to the manufacturer's recommendations. Mitochondria were isolated using the Mitochondria Isolation kit for mammalian cells (Thermo Scientific). The total protein concentration of lysates was quantitated using a Micro BCA Protein Assay kit (Pierce) or the DC Protein Assay kit (Bio-Rad). Lysates were then subjected to SDS-PAGE under reducing conditions, and Western immunoblotting was performed as described previously (3). Immunoblots were imaged using an Amersham Biosciences 600 Imager (GE Healthcare). Signals were quantitated using ImageQuant TL software.

Quantification of Mitochondrial DNA Copies

The ratio of mitochondrial to nuclear DNA was assessed using the Human Mitochondrial DNA Monitoring Primer Set Ratio kit (Takara Bio, Mountain View, CA).

Real Time PCR

Total RNA was isolated from cells using the RNeasy® Mini kit (Qiagen, Valencia, CA) and reverse transcribed using iScript Reverse Transcription SuperMix for RT-quantitative PCR (Bio-Rad). Real time PCRs were performed using SYBR® Green PCR Master Mix (Life Technologies) and gene-specific primer pairs for collagen 1A1, collagen 1A2, fibronectin, α-SMA, TFAM, and 18S rRNA (for primer sequences, see Table 1). Reactions were carried out for 40 cycles (95 °C for 15 s and 60 °C for 1 min) in a StepOnePlus Real Time PCR System (Life Technologies). Real time PCR data are expressed for each target gene as 2⁻ΔΔ^Ct versus endogenous 18S rRNA with error bars representing the standard error of the mean for three experiments. Two-tailed Student's t tests were used for pairwise comparisons.

Plasmid Transfection

A plasmid encoding a dominant negative form of p38 MAPK (pCMV-p38DN) was transfected into IMR-90 cells using Lipofectamine 2000 (Life Technologies) according to the manufacturer's instructions (1:2 ratio of DNA (in μg) to Lipofectamine 2000 (in μl)). IMR-90 cells were incubated overnight in DNA-lipid complexes, and then cells were allowed to recover for 24 h in DMEM containing 10% FBS prior to serum starvation and TGF-β1 treatment.

RNA Interference

siRNA targeting TFAM as well as non-targeting control (Life Technologies; for siRNA sequences, see Table 2) was transfected into lung fibroblasts using Lipofectamine RNAiMAX (Life Technologies) at a final concentration of 100 nm.

Assessment of Bioenergetics

Cells were plated on Seahorse Extracellular Analyzer XF96 plates. Cells were washed in XF assay buffer (DMEM with 1 mm pyruvate, 5.5 mm d-glucose, 4 mm l-glutamine, pH 7.4) and brought to 180-μl final volume in this assay medium. Bioenergetic measurements were interpreted and performed as described (42). In short, the basal oxygen consumption rate (OCR) was measured for 24 min after which oligomycin (1 μg/ml), carbonyl cyanide p-trifluoromethoxyphenylhydrazone (0.6 μm), and antimycin A (10 μm) were injected sequentially. The extracellular acidification rate (ECAR) was measured simultaneously.

Mitochondrial Assay

This assay was performed on the Extracellular Analyzer as adapted from Salabei et al. (43). In short, cells were washed and brought to 180-μl final volume in mannitol and sucrose (MAS) buffer (70 mm sucrose, 220 mm mannitol, 10 mm KH₂PO₄, 5 mm MgCl₂, 2 mm HEPES, 1 mm EGTA, pH 7.2). After three basal measurements were acquired, saponin (30 μg/ml) together with pyruvate (5 mm), malate (2.5 mm), and ADP (1 mm) were injected. Sequentially rotenone (1 μm), succinate (10 mm), ADP (1 mm), and then antimycin A (10 μm) were injected.

Contractility Assay

After cell harvesting, lung fibroblasts were mixed at a density of 2·10⁵ cells/ml in a solution composed of 8 volumes of rat tail collagen type I (BD Biosciences), 1 volume of 10× Ham's F-12 (Life Technologies), and 1 volume of reconstitution buffer (2% NaHCO₃, 4.77% HEPES). The solution was poured into a 24-well plate (0.5 ml/well), and gelation was performed at 37 °C for 30 min. Gels were then overlaid with 1.5 ml of 1% serum-containing cell culture medium. The next day gels were treated with or without TGF-β1.

Statistical Analysis

Statistical analysis was performed using GraphPad software. The data are presented as mean ± S.E. Statistical comparisons were made by performing unpaired Student's t tests unless otherwise indicated.

TGF-β1 Enhances Mitochondrial Oxygen Consumption without Affecting Electron Transport Activity

To determine potential changes in cellular bioenergetics in TGF-β1-differentiated myofibroblasts, we analyzed OCRs and ECARs of normal human diploid lung fibroblasts (IMR-90) treated without or with TGF-β1 (2.5 ng/ml for 48 h) (3). TGF-β1 enhanced overall bioenergetics with a marked increase in basal OCR, ATP-linked respiration, maximal and reserve capacity, proton leak, and non-mitochondrial OCR (Fig. 1, A and C). Furthermore, TGF-β1 also increased ECAR from 16 ± 0.23 milli-pH units/min/10,000 cells in the unstimulated state to 30 ± 0.55 milli-pH units/min/10,000 cells in the TGF-β1-stimulated state (Fig. 1B). Interestingly, when mitochondrial ATP production was inhibited with oligomycin, the increase in ECAR was 63% higher in the TGF-β1-treated cells compared with the control, consistent with a greater capacity for glycolysis (Fig. 1B). Plotting OCR as a function of ECAR showed TGF-β1-induced metabolic reprogramming with induction of mitochondrial respiration and glycolysis resulting in a condition of aerobic glycolysis (Fig. 1D).

The increase in mitochondrial respiration could be due to an increase either in mitochondrial mass or in substrate supply to the electron transport chain. To delineate the mechanism of this increase in mitochondrial respiration, the functioning of complex I or II was measured using a mitochondrial assay where the plasma membrane of fibroblasts was permeabilized and exogenous respiratory substrates were delivered to the electron transport chain. In this assay, after basal OCR was measured, saponin, ADP, succinate (complex II-linked substrate), pyruvate, and malate (complex I-linked substrate) were injected to stimulate respiration (Fig. 1E). Next, rotenone was injected to quantify complex I-linked respiration followed by antimycin A to measure complex II-linked respiration. TGF-β1 induced a small, but significant, increase in complex I-linked respiration without a significant change in complex II-linked OCR at 48 h after treatment (Fig. 1E).

TGF-β1 Stimulates Mitochondrial Biogenesis and Glycolysis in Lung Fibroblasts

Next, we determined whether TGF-β1-induced mitochondrial biogenesis is associated with alterations in mitochondrial distribution and abundance. Staining of the mitochondrial compartment (with the fluorescent dye MitoTracker) demonstrated that the mitochondrial population was relatively homogeneous and mainly localized to the perinuclear region in unstimulated cells; in contrast, there was an apparent increase in mitochondrial abundance with a more diffuse cytoplasmic staining in TGF-β1-differentiated myofibroblasts (Fig. 2A). To determine whether these apparent changes in mitochondrial abundance are linked to mitochondrial biogenesis, we analyzed the effect of TGF-β1 on the expression of mitochondrial biogenesis markers. TGF-β1 induced a time-dependent increase in phosphorylated PGC-1α, a transcriptional co-activator that regulates nuclearly encoded mitochondrial genes (16) (Fig. 2B). Phosphorylation of PGC-1α occurred as early as 15 min after exposure to TGF-β1. Phosphorylation of PGC-1α was increased by 9.71-fold (n = 4, p < 0.05) compared with baseline following 48-h treatment with TGF-β1. VDAC resides in the outer membrane of the mitochondria where it participates in metabolism and, under conditions of stress, in the formation of the mitochondrial permeability transition pore (44–47). We detected a time-dependent increase in VDAC expression following exposure to TGF-β1 (Fig. 2C), supporting increased mitochondrial mass in differentiated cells.

Mitochondria contain between two and 10 copies of mitochondrial DNA (48). Therefore, an increase in mitochondrial DNA copy number per cell correlates with an increase in mitochondrial mass. We tested the effect of TGF-β1 treatment on the mitochondrial DNA content in IMR-90 fibroblasts. TGF-β1 induced a significant increase of the mitochondrial DNA copy number:nuclear DNA ratio as determined by real time PCR (Fig. 2D).

HKII catalyzes the first step of glycolysis (49) and showed a time-dependent increase in expression starting at 12 h after TGF-β1 exposure with maximal expression at 48 h (Fig. 2E). Together, these data suggest that TGF-β1 stimulates both mitochondrial biogenesis and glycolysis in human lung fibroblasts.

TGF-β1 Activates Metabolic Reprogramming via a p38 Mitogen-activated Protein Kinase-dependent Pathway

TGF-β1 signals via SMAD-dependent and -independent pathways (7). p38 MAPK is one of the effectors of the SMAD-independent actions of TGF-β1, and it is also a known activator of PGC-1α (50, 51). We have previously demonstrated the early activation of p38 MAPK in response to TGF-β1 in IMR-90 fibroblasts (52). We analyzed the effect of blocking p38 MAPK activity using the p38 MAPK inhibitor SB202190, which binds the ATP binding pocket of the active kinase (53, 54). IMR-90 cells were pretreated with SB202190 (5 μm) or vehicle 30 min prior to the addition of TGF-β1. In response to TGF-β1, we detected an increase in phosphorylated PGC-1α in the nuclear fraction (Fig. 3A) without an effect on total PGC-1α (Fig. 3B). This activation of PGC-1α was inhibited by p38 MAPK inhibition (Fig. 3A). TGF-β1-induced PGC-1α phosphorylation was decreased by 54% after SB202190 treatment (n = 3, p = 0.01).

TFAM is a downstream target gene of PGC-1α and is involved in mitochondrial DNA transcription and replication. We analyzed TFAM expression in total and mitochondrial fractions isolated from cells treated with or without TGF-β1. In untreated IMR-90 fibroblasts, TFAM was predominantly detected at 29 kDa, corresponding to its precursor form; upon stimulation with TGF-β1, TFAM expression was increased in the mitochondrial fraction and was also processed to a 25-kDa form. This truncated form corresponds to the cleaved active product of TFAM following its import into the mitochondrial matrix. We detected increased expression of the 25-kDa active form of TFAM upon TGF-β1 stimulation; this effect and the associated increase in mitochondrial proteins VDAC and TOM20 (Fig. 3C) as well as the increase in mtDNA copy number (Fig. 3D) were inhibited by p38 MAPK blockade. In concert, the TGF-β1-induced expression of HKII was also suppressed by p38 MAPK inhibition (Fig. 3C). To confirm the role of p38 MAPK in mediating TGF-β1-induced mitochondrial biogenesis and glycolysis, we evaluated the effect of overexpressing a dominant negative mutant form of p38 MAPK (55). TGF-β1-induced HKII and VDAC expression was significantly decreased in cells expressing the p38 MAPK mutant (Fig. 4, A, B, C, and D). Together, these data indicate a key role for p38 MAPK in mediating mitochondrial biogenesis and glycolysis.

Mitochondrial Biogenesis Is Required for Expression of α-SMA in Response to TGF-β1

Next, we determined whether mitochondrial biogenesis during the process of myofibroblast differentiation was necessary to maintain the differentiated phenotype. To answer this question, we analyzed the effect of inhibiting TGF-β1-induced mitochondrial biogenesis by silencing TFAM and determined the effect on the expression of α-SMA, a key marker of myofibroblast differentiation. TFAM silencing decreased the protein expression of α-SMA in response to TGF-β1 at 48 h (Fig. 5A); densitometric analysis of α-SMA expression in non-targeting (NT) or TFAM siRNA-transfected cells showed that TFAM silencing reduced TGF-β1-induced α-SMA expression (by 55%; n = 6, p < 0.05) (Fig. 5B). TFAM silencing did not alter TGF-β1-induced fibronectin (FN) expression (-fold change in FN expression following 48-h TGF-β1 was 2-fold in controls and in TFAM-silenced cells; n = 3, p = 0.18). TFAM silencing significantly reduced VDAC expression (Fig. 5C; 50% decrease by densitometric analysis of n = 3 independent experiments; p < 0.05) in response to TGF-β1, thus confirming inhibition of mitochondrial biogenesis when TFAM is silenced.

To determine whether the effect of TFAM silencing on α-SMA down-regulation occurred at the mRNA level, we analyzed α-SMA mRNA expression under the same conditions. TFAM silencing reduced TGF-β1-induced expression of α-SMA mRNA by 70% compared with control cells (n = 3, p < 0.005) (Fig. 5E). Silencing of TFAM did not significantly alter TGF-β1-induced expression of fibronectin or collagen 1A1 (Col1A1) (Fig. 5, F and G) and modestly decreased the expression of Col1A2 (Fig. 5H). Together, these data indicate that induction of TFAM and mitochondrial biogenesis is critical for maintaining the differentiated state of myofibroblasts.

Myofibroblast Contraction and Differentiation Are Dependent on Mitochondrial Bioenergetics and Glycolysis

Contraction of myofibroblasts has been shown to promote myofibroblast differentiation (9, 10, 56). We determined whether the increase in mitochondrial biogenesis/bioenergetics is necessary to mediate contractility of differentiated myofibroblasts. In a three-dimensional type I collagen gel model, we treated fibroblasts with or without TGF-β1 in the absence or presence of rotenone, which inhibits complex I (57). TGF-β1-induced contraction was abolished in the presence of rotenone (Fig. 6, A and B), suggesting that mitochondrial electron transport is required to mediate myofibroblast contractility.

Because phosphorylation of MLC20 is a prerequisite for contraction (58), we studied the effect of rotenone on the phosphorylation of MLC20. The presence of rotenone markedly inhibited the phosphorylation of MLC20 in response to TGF-β1 (Fig. 6C). This effect on reduced contractile activity correlated with a reduction in α-SMA protein expression, whereas fibronectin expression was not affected, similar to the observed effects with TFAM silencing (Fig. 5, A and B). Additionally, inhibition of mitochondrial biogenesis by silencing TFAM decreased MLC20 phosphorylation (Fig. 6, D and E). Together, these data support the requirement for mitochondrial biogenesis and oxidative phosphorylation for contractile signaling and myofibroblast differentiation.

Based on our findings of an increase in aerobic glycolysis following TGF-β1 treatment of lung fibroblasts (Figs. 1B and 2E), we investigated the contribution of TGF-β1-enhanced glycolysis in supporting myofibroblast differentiation. Treatment of lung fibroblasts with 2-deoxy-d-glucose (2-DG), a structural analogue of glucose that acts as a non-competitive inhibitor of hexokinase (59), markedly reduced the expression of TGF-β1-induced α-SMA expression without significantly affecting TGF-β1-induced fibronectin expression (Fig. 7, A, B, and C). Together, these data suggest that both mitochondrial biogenesis and aerobic glycolysis coordinately support myofibroblast differentiation and contractility induced by TGF-β1.