Transforming growth factor (TGF)-β has been associated with podocyte injury; we have examined its effect on podocyte bioenergetics. We studied transformed mouse podocytes, exposed to TGF-β1, using a label-free assay system, Seahorse XF24, which measures oxygen consumption rates (OCR) and extracellular acidification rates (ECAR). Both basal OCR and ATP generation-coupled OCR were significantly higher in podocytes exposed to 0.3–10 ng/ml of TGF-β1 for 24, 48, and 72 h. TGF-β1 (3 ng/ml) increased oxidative capacity 75%, and 96% relative to control after 48 and 72 h, respectively. ATP content was increased 19% and 30% relative to control after a 48- and 72-h exposure, respectively. Under conditions of maximal mitochondrial function, TGF-β1 increased palmitate-driven OCR by 49%. Thus, TGF-β1 increases mitochondrial oxygen consumption and ATP generation in the presence of diverse energy substrates. TGF-β1 did not increase cell number or mitochondrial DNA copy number but did increase mitochondrial membrane potential (MMP), which could explain the OCR increase. Reactive oxygen species (ROS) increased by 32% after TGF-β1 exposure for 48 h. TGF-β activated the mammalian target of rapamycin (mTOR) pathway, and rapamycin reduced the TGF-β1-stimulated increases in OCR, ECAR, ATP generation, cellular metabolic activity, and protein generation. Our data suggest that TGF-β1, acting, in part, via mTOR, increases mitochondrial MMP and OCR, resulting in increased ROS generation and that this may contribute to podocyte injury.
- oxygen consumption rate
- extracellular acidification rate
increased glomerular expression of transforming growth factor (TGF)-β has been linked to glomerulosclerosis in several kidney diseases, including focal segmental glomerulosclerosis (FSGS) and diabetic nephropathy. A common feature of these diseases is podocyte depletion, which likely arises through diverse molecular pathways. TGF-β induces apoptosis in cultured podocytes and FSGS in transgenic mice (15, 25, 27). Increasing evidence points to a role for TGF-β pathways involved in energy balance, metabolism, reactive oxygen species (ROS), and mitochondria. The available data suggest that increasing TGF-β activity is associated with mitochondrial dysfunction and increasing mitochondrial ROS synthesis (4). However, it is not known whether TGF-β modulates the podocyte bioenergetics profile. We have characterized podocyte bioenergetics and mitochondrial function (1) and set out to determine whether TGF-β alters mitochondrial function in ways that might contribute to podocyte loss.
We used transformed mouse podocytes and primary mouse podocytes, exposed to TGF-β using a label-free assay system, Seahorse XF24, which measures oxygen consumption rates (OCR) and extracellular acidification rates (ECAR). Our principal findings are that the TGF-β1 activates the mammalian target of rapamycin (mTOR) pathway, leading to increase mitochondrial membrane potential (MMP), OCR, and intracellular reactive oxygen species (ROS) levels.
MATERIALS AND METHODS
FVB/N mice were killed at the age of 4–6 wk to establish primary podocyte cultures, as described below. Mice were cared for under a protocol approved in advance by the National Institute of Diabetes and Digestive and Kidney Diseases Animal Care and Use Committee, and all animal care conformed to the National Institutes of Health's “Guide for the Care and Use of Laboratory Animals.”
The mouse podocyte cell line AI was established from glomeruli of transgenic mice bearing the podocin/rtTA gene and the SV40 temperature-sensitive T antigen (13). For these studies, clone 1–1P4G5 was cultured in growth medium consisting of RPMI 1640, 10% FBS, and 100 units/ml of penicillin and 100 μg/ml streptomycin, obtained from Gibco (Rockville, MD), under permissive conditions (33°C). Cells were studied after 5–8 days of culture at the nonpermissive temperature of 37°C.
Primary glomerular cells from isolated glomeruli were obtained from 4- to 6-wk-old FVB/N mice using Dynabeads (14, 31). Briefly, purified glomeruli were placed in culture in RPMI 1640 medium supplemented with 10% FBS. When colonies were established, the cells were harvested with trypsin and used at passage 3 or 4. For immunostaining, cells were fixed with 2% paraformaldehye and 4% sucrose, exposed to primary antibody (mouse monoclonal anti-WT1 antibody; Upstate, Lake Placid, NY; and mouse anti-nestin antibody; Millipore, Billerica, MA), exposed to fluorescent secondary antibody, and examined under fluorescence microscopy. Most cells in these primary glomerular cells cultures exhibited strong nuclear staining for WT-1 and for nestin within cytoplasmic filaments (data not shown); for simplicity, we will refer to these cultures as primary podocytes, although these cells are not a fully homogenous population.
All cell culture was at 37°C and with 5% CO2, except for culture under permissive conditions (33°C, 5% CO2) to allow cell replication and during the period of assay using XF24 (37°C, room air).
Rotenone, carbonylcyanide-p-trifluoromethoxy phenylhydrazone (FCCP), 2-deoxyglucose (2-DG), sodium palmitate, l-carnitine, antimycin, etomoxir, and apocynin were obtained from Sigma-Aldrich (St. Louis, MO), and oligomycin was obtained from Calbiochem (San Diego, CA). Concentrated stocks of rotenone (1 mM), oligomycin (1 mM), and apocynin (10 mM) were prepared in DMSO. Concentrated stocks of FCCP (10 mM) and antimycin (10 mM) were prepared in ethyl alcohol. Concentrated stocks of 2-DG (1 M) and oxamate (500 mM) were prepared in assay medium. Concentrated stocks of l-carnitine (50 mM) and etomoxir (100 mM) were prepared in deionized water. Concentrated stocks of sodium palmitate (2 mM) were conjugated with 0.34 mM (2.267 g/dl) ultra fatty acid-free BSA. Ultra fatty acid-free BSA was purchased from Roche Diagnostics (Indianapolis, IN). Recombinant human TGF-β1 was obtained from R&D Systems (Minneapolis, MN). Rapamycin diluted in DMSO was obtained from Calbiochem (Darmstadt, Germany).
Measurements of oxygen consumption and extracellular acidification rates.
A Seahorse Bioscience XF24–3 extracellular flux analyzer was used to measure the rate changes of dissolved O2 and pH in medium immediately surrounding adherent cells cultured in an XF24 V7 cell culture microplates (Seahorse Bioscience, Billerica, MA) coated with collagen I (Becton Dickinson, Bedford, MA).
Transformed mouse podocytes were seeded in XF24-well microplates at 2.0 × 104 cells per well (area 0.32 cm2) in 100 μl of growth media and incubated overnight. The following day, an additional 100 μl of growth media was added and 2 days later, medium was replaced. On day 4, cells were exposed to rapamycin for 24 h when it was used for inhibition of the mTOR pathway. On day 5, cells were exposed to TGF-β1 for 48 h. After incubation for a total of 7 days, growth medium was removed and replaced with 675 μl of assay medium prewarmed to 37°C, comprising RPMI 1640 without bicarbonate supplemented with 10 mM KCl, 10 mM NaCl, 1 mM sodium pyruvate, and 3 mM lactate, and cultured at 37°C in room air (without supplemental CO2).
Primary mouse podocytes were seeded in XF24-well microplates at 2.0 × 104 cells per well in 200 μl of growth media and incubated overnight. The following day, growth medium was replaced with 675 μl of assay medium as above. To assess β-oxidation in podocytes, RPMI 1640 containing 11 mM glucose and 0.5 mM carnitine was employed as an assay media, and sodium palmitate was administered at the final concentration of 200 μM. All of these media included 2 mM l-glutamine.
Measurements of OCR and ECAR were performed after equilibration in assay medium (lacking supplemental CO2) for 0.5 to 1 h. Briefly, the Seahorse analyzer uses a cartridge with 24 optical fluorescent O2 and pH sensors that are embedded in a sterile disposable cartridge, one for each well. Prior to each rate measurement, the plungers mix assay media in each well for 8 min to allow the oxygen partial pressure to reach equilibrium. For measurements of rates, the plungers gently descend into the wells forming chambers that entrap the cells in ∼7-μl volume. Measurements of O2 concentration and pH are periodically made over 4 min, and the rates of oxygen consumption and extracellular acidification are obtained from the slopes of concentration change in these parameters vs. time. After the rate measurements, the plungers ascend, and the wells are gently mixed to equilibrate the medium. OCR are reported in the units of picomoles per minute, and ECAR are reported in milli-pH (mpH) units per minute. Baseline rates are measured four times. One or more testing chemicals are preloaded in the reagent delivery chambers of the sensor cartridge and then pneumatically injected into the wells to reach the desired final working concentration. After 2 min of mixing, postexposure OCR and ECAR measurements were made four to six times. The averages of four baseline rates and two to six test rates were used for data analyses.
Total RNA was isolated using miRNeasy mini kit and RNase-free DNase set, according to the manufacturer's protocol (Qiagen, Valencia, CA). RNA concentration was measured by an ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE). The first-strand cDNA synthesis was performed using 800 ng of RNA and first-strand cDNA synthesis kit for RT-PCR (avian myeloblastosis virus), according to the instructions of the manufacturer (Roche Diagnostics). To test for contamination by genomic DNA, additional reactions were done without adding reverse transcriptase. Real-time quantitative RT-PCR was performed using Power SYBR Green PCR master mix (Applied Biosystems, Carlsbad, CA) and 7900HT sequence detection system (Applied Biosystems). The primers used in this study were shown in Table 1. Each condition was studied with triplicate cell cultures. After sequential incubation at 95°C for 10 min, the amplification protocol consisted of 40 cycles of denaturing at 95°C for 15 s, annealing, and extension at 60°C for 60 s. Expression levels of each mRNA were calculated after normalizing with β-actin. The PCR product size was confirmed on a 1.5% agarose gel.
Transformed podocytes were seeded in collagen I-coated 96-well microplates at 2.0 × 104 cells per well (0.32 cm2) in 100 μl of growth media. Cells were incubated overnight, with the addition of another 100 μl of growth media the next day and replacement of media 2 days later. After incubation for 5 days, cells were exposed to 0.1–10 ng/ml of TGF-β1 or vehicle with fresh growth media for 24 to 72 h. Assays were initiated by removing the growth medium from each well and replacing it with 100 μl of assay medium. Cells were exposed to vehicle or compound for 45 min at 37°C before starting the ATP assay. The quantity of ATP present in the test cells in each well was measured by CellTiter-Glo luminescent cell viability assay (Promega, Madison, WI). In some experiments, transformed podocytes were exposed to 0.5 nM of rapamycin for 24 h on day 4. Luminescence intensity from each well was measured using a FLUOstar Optima plate reader (BMG Labtech, Offenberg, Germany).
Calcein AM cell viability assay.
Cell viability was determined following 24–72 h exposure to TGF-β1 or vehicle. Transformed podocytes were seeded in collagen I-coated 96-well black microplates (Becton Dickinson) at 2.0 × 104 cells per well (0.32 cm2) in 100 μl of growth media. Cells were incubated overnight, with the addition of another 100 μl of growth media the next day and replacement of media 2 days later. After incubation for 5 days, cells were exposed to 0.1–10 ng/ml of TGF-β1 or vehicle with fresh growth media for 24 to 72 h at 37°C. Calcein AM (Invitrogen, Carlsbad, CA) staining solution was prepared by diluting to 1 μM in Hank's balanced salt solution (HBSS) immediately prior to use. The assay was performed by first removing the medium containing 0.1–10 ng/ml of TGF-β1 or vehicle and then washing each well with 200 μl of HBSS. One-hundred microliters of 1 μM calcein AM staining solution was then added to each well. After incubation at 37°C for 30 min, the fluorescence intensity of each well was measured using a SpectraMax M2 microplate reader (Molecular Devices, Sunnyvale, CA). In some experiments, transformed podocytes were exposed to rapamycin for 24 h on day 4.
Apoptosis detection via caspase activity.
Transformed podocytes were seeded in collagen I-coated 96-well microplates (Becton Dickinson) at 1.0 × 104 cells per well (0.32 cm2) in 50 μl of growth media. Cells were incubated at 37°C overnight, with the addition of another 50 μl of growth media the next day and replacement of media 2 days later. After incubation for 5 days, cells were exposed to 0.1–10 ng/ml of TGF-β1 or vehicle with fresh growth media for 24 to 72 h. Apoptosis was detected by Caspase-Glo 3/7 assay (Promega). Luminescence intensity from each well was measured using a FLUOstar Optima plate reader.
3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay.
Cellular metabolic activity was assessed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), Cell-Titer 96 nonradioactive cell proliferation assay (Promega). Transformed podocytes were seeded in collagen I-coated 96-well microplates (Becton Dickinson) at 1.0 × 104 cells per well (0.32 cm2) in 50 μl of growth media. Cells were incubated overnight, with the addition of another 50 μl of growth media the next day and replacement of media 2 days later. After incubation for 5 days, cells were exposed to 0.1–10 ng/ml of TGF-β1 or vehicle with fresh growth media for 24 to 72 h and incubated with MTT for 1 h. In some experiments, transformed podocytes were exposed to 0.5 nM of rapamycin for 24 h on day 4. The absorbance at 570 nm and 650 nm was detected by a SpectraMax M2 microplate reader.
Quantification of total protein in podocytes.
Transformed podocytes were seeded in collagen I-coated 10-cm dishes (Becton Dickinson) at 100 × 104 cells per dish in 5-ml growth media. Cells were incubated overnight, with the addition of another 5 ml of growth media the next day and replacement of media 2 days later. After incubation for 5 days, cells were exposed to 3 ng/ml of TGF-β1 or vehicle with fresh growth media for 48 h. Whole-cell extracts were made using RIPA buffer. Total protein in podocytes was quantified by bicinchoninic acid protein assay kit (Pierce Biotechnology, Rockford, IL). In some experiments, transformed podocytes were exposed to 0.5 nM of rapamycin for 24 h on day 4. The absorbance at 562 nm was detected by Tunable Microplate Reader Model Versa Max (Molecular Devices).
Mitochondrial content assay.
Mitochondrial quantity was assayed by analysis of mitochondrial DNA (mtDNA) and mitochondrial mass (MitoTracker Green fluorescent probe). The amount of mtDNA was quantified by determining the ratio of mitochondrially encoded NADH dehydrogenase 1 (mt-Nd1), mitochondrially encoded cytochrome b (mt-Cytb), and mitochondrially encoded cytochrome c oxidase II (mt-Co2) to a nuclear DNA (hemoglobin intron) by quantitative PCR. Cells were incubated at 37°C for 30 min with 200 μl of serum-free medium, including 200 nM of MitoTracker Green FM (Invitrogen, Eugene, OR). Fluorescence intensity was detected using a SpectraMax M2 microplate reader with excitation and emission wavelength of 490 and 516 nm, respectively.
MMP was estimated by staining cells with 5,5′,6,6′-tetrachloro-1,1′,3,3,′-tetra-thylbenzimidazole carbocyanide iodide (JC-1) fluorescence dye (Biotium, Hayward, CA) or by using tetramethylrhodamine, methyl ester, perchlorate (TMRM; Invitrogen, Carlsbad, CA). Transformed mouse podocytes were incubated for 5 days at 37°C, exposed to TGF-β1 or vehicle with fresh growth media for 48 h. For the JC-1 experiments, cells were incubated with JC-1 at 37°C for 30 min. Red fluorescence (excitation 550 nm, emission 600 nm) and green fluorescence (excitation 485 nm, emission 535 nm) was detected using a SpectraMax M2 microplate reader. The ratio of red to green fluorescence was considered as a degree of MMP. In some experiments, transformed podocytes were exposed to rapamycin for 24 h on day 4. For the TMRM experiments, cells were incubated with TMRM at 37°C for 30 min. Fluorescent intensity was detected using a Spectra Fluor Plus fluorimeter (Tecan, Kawasaki, Japan) with excitation and emission wavelengths of 535 nm and 590 nm, respectively.
Determination of intracellular ROS level.
To determine the intracellular level of ROS, DCFH-DA was used, according to the manufacturer's instructions (Cell Biolabs, San Diego, CA). Transformed podocytes and primary podocytes were incubated for 5 days and 1 day at 37°C, respectively. After the incubation, cells were exposed to 3 ng/ml of TGF-β1 or vehicle with fresh growth media for 48 h at 37°C. Cells were washed with HBSS twice and further incubated in serum-free medium containing 50 μM DCFH-DA for 1 h at 37°C. Stained cells were washed with HBSS twice and incubated with 100 μl of serum-free medium. After incubation at 37°C for 30 min, the fluorescent intensity of each well was detected using a SpectraMax M2 microplate reader with excitation and emission wavelengths of 480 nm and 530 nm, respectively. In some experiments, transformed podocytes were exposed to rotenone or apocynin before staining with DCFH-DA for 30 min. The fluorescent intensity was detected using a Spectra Fluor Plus fluorimeter (Tecan) with excitation and emission wavelengths of 485 nm and 535 nm, respectively.
Western blot analysis.
Whole cell extracts (20–30 μg) were resolved eletrophoretically on 4–12% Bis-Tris NuPAGE gels (Invitrogen) and subsequently transferred to nitrocellulose membranes (Invitrogen). Western blot analysis was performed with following antibodies: β-actin (mouse monoclonal; Sigma), VDAC1 (goat polyclonal; Santa Cruz Biotechnology, Santa Cruz, CA), mitochondrial complex IV (mouse monoclonal; Invitrogen), p-mTOR (rabbit polyclonal; Cell Signaling Technology, Beverly, MA), p-tuberin (rabbit monoclonal; Cell Signaling Technology), mTOR (rabbit polyclonal; Cell Signaling Technology), tuberin/TSC2 (rabbit monoclonal; Cell Signaling Technology), and Nox4 (rabbit polyclonal, Dr. Tom Leto, National Institutes of Health; Novus).
Statistical analyses were performed using Prism (GraphPad, San Diego, CA). Data are presented as means ± SD, unless stated otherwise. Analyses included unpaired t-test and ANOVA. When we compared control and TGF-β1-exposed paired cultures, across multiple analyses (e.g., Figs. 1, 2, 3, 6, and 7), the paired groups were compared by Student's t-test, and adjustments were made for multiple analyses by the Bonferroni or Dunnett correction. A P value <0.05 was considered significant.
TGF-β1 increased oxidative phosphorylation and glycolysis.
We investigated cellular respiration in transformed mouse podocytes, assessed as OCR. To measure maximal oxidative capacity, oligomycin was injected into culture wells prior to injection of the uncoupling agent FCCP. TGF-β1 consistently increased basal OCR at 24, 48, and 72 h, OCR coupled to ATP generation, which is oligomycin-sensitive OCR, and total oxidative capacity, which is FCCP-sensitive OCR, with a dose-response curve that peaked at 3 ng/ml. Oxidative capacity was increased by 75%, and 96% relative to control after exposure of TGF-β1 (3 ng/ml) for 48 and 72 h, respectively (Fig. 1A). Peak effects were seen at 3 ng/ml after 48 h; the dose and time that we selected for subsequent experiments. As shown in Fig. 1, B and D, TGF-β1 increased basal OCR, the fraction of OCR coupled ATP generation, and oxidative capacity in transformed podocytes. Regardless of the presence or absence of TGF-β1, antimycin inhibited OCR to the same level. This antimycin-insensitive OCR, interpreted as nonmitochondrial respiration, was slightly but significantly higher with TGF-β1. Similar findings were observed in primary podocytes (Fig. 1, D and E).
ECAR increased with oligomycin and FCCP and decreased with antimycin; the extent of the increase in ECAR was greater in transformed podocytes exposed to TGF-β1 compared with control (Fig. 1, B and C). The antimycin-sensitive component of ECAR was likely acidification from carbon dioxide, the end product of mitochondrial respiration (1, 9, 15). On the other hand, antimycin-insensitive acidification is likely due to glycolytic acidification via extruded lactic acid. TGF-β1 increased the antimycin-insensitive ECAR in transformed podocytes (Fig. 1, B and C). These results imply that TGF-β1 enhances glycolytic lactic acid production. Similar results were observed in primary podocytes (Fig. 1, D and E).
To clarify the effect of TGF-β1 on glycolysis, we used 2-deoxyglucose (2-DG), a glucose analog to inhibit hexokinase, the first enzyme in the glycolytic pathway that converts glucose to glucose-6 phosphate. We first injected oligomycin (a complex V inhibitor) and FCCP (to reduce the mitochondrial proton gradient), and ECAR rose as a sign of increased glycolysis to compensate for loss of mitochondrial ATP generation (Fig. 2A). TGF-β1-pretreated cells exhibited much higher ECAR compared with control cells. The fraction of ECAR attributed to glycolysis was 69.1 ± 18.4% in TGF-β1-treated cells compared with 57.2 ± 15.9% in control cells (Fig. 2B). The difference in ECAR was abolished by the addition of 2-DG, supporting the hypothesis that TGF-β1 boosts glycolysis (Fig. 2A). The injection of 2-DG initially increased OCR, presumably due to increased demand upon the metabolic demand, but the 2-DG also blocks pyruvate production and, thus, the OCR increase was transient (Fig. 2A). These findings indicate that TGF-β1 increases glycolysis in podocytes, particularly when mitochondrial function is inhibited.
TGF-β1 increased intracellular ATP level and protein amount, while cell numbers remain constant.
We next investigated ATP content after the exposure to TGF-β1. In spite of the increase in OCR after 24 h, there were no significant differences in ATP content, but with exposure to TGF-β1, ATP content was increased 19% and 30%, relative to control after 48 and 72 h of exposure, respectively (Fig. 3A).
We considered whether these increases in OCR and ATP content might reflect an effect of TGF-β1 on cell number or mitochondrial biogenesis. The MTT assay is commonly used as a measure of cell number, although the assay also depends on cellular metabolic activity. We found that TGF-β1 increased MTT absorbance at 24, 48, and 72 h at the concentration shown (Fig. 3B). It has previously been reported that TGF-β does not stimulate cell proliferation in cultured podocytes (19, 37), while it induces apoptosis in podocytes in vivo and vitro (27). In our system, caspase 3/7 activities (reflecting apoptosis) were increased by TGF-β1 exposure at 24, 48, and 72 h at the concentration shown (Fig. 3C). Although cells were initially seeded in a confluent state, to directly assess cell numbers, we counted cell number using trypan blue exclusion after exposure to TGF-β1, 3 ng/ml, or control. Cell counts were unaffected by TGF-β1: after 24 h, 100.0 ± 17.8% (control) vs. 114.2 ± 18.5% (TGF-β1); after 48 h, 100.0 ± 17.2% vs. 95.2 ± 10.9%; and after 72 h; 100 ± 16.9% vs. 111.4 ± 17.0%. Furthermore, the calcein AM stain (reflecting cell number) did not show significant differences after exposure to TGF-β1 from 0.1 to 10 ng/ml at 24, 48, and 72 h compared with control (Fig. 3D). To address whether TGF-β1 stimulates glycolysis or oxidative phosphorylation to generate ATP in our system, we used 2-DG and rotenone. As shown in Fig. 2C, rotenone had a certain effect on blocking ATP generation in TGF-β1-treated podocytes, but there was a tendency to shift glycolytic metabolism to generate ATP. In TGF-β1-exposed cells, 2-DG reduced ATP to levels nearly similar to those of control cells, while rotenone did not prevent the TGF-β1-driven increase in cellular ATP. These findings suggest either that the actions of TGF-β1 to stimulate ATP production are primarily via the stimulation of glycolysis (rather than mitochondria), or alternatively that these effects are due to the existence in these cells of a larger glycolytic reserve capacity (when mitochondrial respiration is inhibited with rotenone) than a mitochondrial reserve capacity (when glycolysis is inhibited with 2-DG). These findings are consistent with the observation that TGF-β1 increased both OCR and ECAR. TGF-β1 increased total amount of protein in podocytes after 48 h (Fig. 3E). Thus, TGF-β1 increased OCR, ATP production, and protein production without any change in cell number, while proapoptotic activity was increased and WT1 mRNA as a podocyte marker was downregulated. These findings suggest that TGF-β1-induced podocyte injury, which is consistent with the previous report (24) (Fig. 5G).
TGF-β1 increased fatty acid oxidation.
The increased in OCR shown above could be due to increased TCA cycle activity or increased β-oxidation of fatty acids. To investigate whether these increases in OCR might be associated with increased fatty acid oxidation, palmitate was employed as a metabolic substrate in an assay medium that also contained glucose and glutamine (to maintain cell viability), and lacked pyruvate (a substrate for the TCA cycle) and lactate (a potential precursor of pyruvate). TGF-β1 increased OCR at the baseline, in assay medium only (Fig. 4A). After injection of palmitate, there was a small increase in OCR in TGF-β1-treated and control cells. In the presence of FCCP, TGF-β1 pretreatment increased both OCR and ECAR compared with control cells, and OCR, as well as ECAR, was higher in cells exposed to palmitate compared with cells lacking palmitate. TGF-β1 increased palmitate-driven OCR by 49% (P < 0.001, by one-way ANOVA). Antimycin and rotenone reduced ECAR, likely due to reduced carbonic acid generation. These data suggest that podocytes can use palmitate to generate ATP. Furthermore, TGF-β1 increases the use of palmitate (fatty acid oxidation) and also activates glycolysis when the electron transport chain is inhibited (oligomycin, antimycin, and rotenone) or uncoupled from ATP generation (FCCP).
Next, to further investigate the contribution of fatty acid oxidation to OCR, etomoxir (an irreversible O-carnitine palmitoyltransferase-1 inhibitor), oligomycin, and rotenone (complex I inhibitor) were employed. TGF-β1 exposure for 48 h increased the following: basal OCR, mitochondrial respiration (the rotenone-sensitive fraction of OCR), OCR coupled to ATP turnover (the oligomycin-sensitive fraction of OCR), and β oxidation (the etomoxir-sensitive fraction of OCR) (Fig. 4B). The extent of rotenone, oligomycin, and etomoxir OCR inhibition reflects mitochondrial respiration, OCR-coupled ATP turnover, and lipid oxidation-derived oxygen consumption, respectively. While basal OCRs were higher with TGF-β1 treatment as noted above, the percentage reduction with rotenone, oligomycin, and etomoxir were similar between control and TGF-β1-exposed cells: mitochondrial respiration was 80.8 ± 7.9% vs. 82.5 ± 3.9%, OCR coupled to ATP generation was 60.8 ± 7.5% vs. 62.7 ± 6.4%, and lipid oxidation-derived oxygen consumption was 38.9 ± 10.3% vs. 44.2 ± 8.1%.
Taken together, these data indicate that TGF-β1 enhances fatty acid oxidation, but that increased TCA activity also contributes to the increase in OCR. This latter observation led us to investigate whether TGF-β1 might increase mitochondrial number or total mitochondrial mass.
TGF-β1 effects on mitochondria were not mediated by Sirt1, Pgc1α, or Ucp2.
We next investigated expression of molecules that regulate mitochondrial number and mass, including sirtuin 1 (Sirt1) and peroxisome proliferator-activated receptor-gamma coactivator 1 (Pgc1a). We also examined expression of uncoupling protein 2 (Ucp2), which reduces MMP and reactive oxygen species (ROS) generation. As shown in Fig. 5A, TGF-β1 decreased Sirt1 and Pgc1a mRNA expression (at 24 and 48 h) and increased Ucp2 mRNA expression (at 24 and 48 h). Despite these mRNA changes, there were no changes in steady-state protein level for these three proteins, as judged by Western blots (data not shown). Although there were no significant changes in protein expression, protein turnover rates might affect these results.
TGF-β1 did not alter mitochondrial number, mass, or gene expression.
To investigate whether TGF-β1 affects mitochondria numbers, mitochondrial DNA content and mass were quantified by two methods: quantitative PCR and the MitoTracker Green probe, respectively. Three mitochondrial DNA genes (mt-Nd1, mt-Cytb, and mt-Co2) were examined. Expression was unchanged after exposure to TGF-β1 for 48 h (Fig. 5B). Regardless of the MMP, the MitoTracker Green probe accumulates in mitochondria and provides an assessment of mitochondrial mass (7). In the presence of this probe, fluorescent intensity was not changed after 48 h exposure to TGF-β1 (Fig. 5C).
We next examined gene expression for mitochondrial proteins. We examined the complex IV constituent, Cox5a, and found no significant differences in Cox5a mRNA expression after 24 and 48 h of treatment with TGF-β1 (Fig. 5C) and complex IV subunit I protein expression after 4 (Fig. 7B), 24 (data not shown), and 48 h (Fig. 5E) of exposure to TGF-β1. The protein expression of VDAC1, which is a component of mitochondrial outer membrane, did not change after 4 (Fig. 7B) and 48 h (Fig. 5E) exposure to TGF-β1, as assessed by Western blot analysis.
These results suggest that TGF-β1-stimulated increases in OCR were not due to the change in mitochondrial number, mass, or gene expression.
TGF-β1 increased mitochondrial membrane potential.
TGF-β1 decreases MMP in Mv1Lu epithelial cells and hepatocytes (11, 23, 39). TGF-β1 increased MMP dose dependently, as assessed by the molecular probe JC-1 and TMRM (Fig. 6, E and F). These findings suggest that TGF-β1 may increase OCR by increasing MMP.
TGF-β1 increased ROS levels.
Following TGF-β1 exposure for 48 h, intracellular ROS levels, assessed by the fluorescent molecule DCFH-DA, increased by 32.3% at 3 ng/ml and 36.1% at 10 ng/ml of TGF-β1 in transformed podocytes (Fig. 6A) and by 44.6% in primary podocytes (Fig. 6B). To address the origin of ROS, 1 μM rotenone, complex I inhibitor, was used. It significantly reduced the ROS generation in both control and TGF-β1-treated cells (Fig. 6C). However, 1–1,000 μM of apocynin, NADPH inhibitor, had no significant effect on TGF-β1 increased ROS generation (Fig. 6D). These findings support that the major origin of ROS was mitochondria.
TGF-β1 increased Nox4 mRNA expression, but protein levels were unaffected.
Apart from the mitochondrial respiratory chain, cellular oxidases, including NADPH oxidases, such as Nox4, are also important sources of cellular ROS. After 24 and 48 h of exposure, Nox4 mRNA expression was upregulated by TGF-β1 by 213 and 415%, respectively. Despite these mRNA changes, there were no changes in steady-state protein level, as judged by Western blots using two different antibodies (Fig. 5F).
TGF-β1 activates the mTOR pathway, and rapamycin partially reverses the TGF-β1 stimulated increase in OCR and ECAR.
The mTOR activation increases mitochondrial oxygen consumption in Jurkat T cells and C2C12 myoblasts (6, 26). TGF-β activates the mammalian target of rapamycin complex-1 (mTORC1) (18, 22, 33). To investigate whether TGF-β1 activates the mTOR pathway, p-tuberin/TSC2 and p-mTOR were analyzed by Western blot. As shown in Fig. 7A and B, p-tuberin/TSC2 and p-mTOR were phosphorylated following TGF-β1 exposure. Peak effects were seen at 3 h, consistent with activation.
We studied the effect of rapamycin on cell viability, assessed by the calcein AM stain; at 0.1 nM and 0.5 nM rapamycin, there was no change in cell viability, whereas higher doses reduced cell viability (Fig. 7C). Using JC-1 to assess MMP, we found that rapamycin reduced MMP at these concentrations, as well as at higher concentrations (Fig. 7D). We next asked whether rapamycin decreases TGF-β1-driven ATP generation. In rapamycin (0.5 nM)-pretreated cells, the TGF-β1 effect, which significantly increases ATP generation compared with control (P < 0.0001), was abolished. TGF-β1 also increased glycolytic ATP generation when oxidative phosphorylation was blocked by rotenone. This TGF-β1-driven ATP generation was reduced in rapamycin (0.5 nM)-pretreated cells (Fig. 7E). These findings indicate that TGF-β1-driven ATP is partially generated by mTOR pathway and oxidative phosphorylation. To investigate whether rapamycin inhibits the effects of TGF-β1 on OCR and ECAR, we studied transformed mouse podocytes exposed to rapamycin 0.5 nM (a nontoxic concentration) for 24 h prior to TGF-β1 exposure. As shown in Fig. 7, F and G, rapamycin inhibited by ∼30–50% the effect of TGF-β1 on OCR (basal rate, coupled OCR to ATP generation, and oxidative capacity). With regard to ECAR, rapamycin completely inhibited the effect of TGF-β1 on basal ECAR but not the other fractions (oligomycin-insensitive, FCCP-enhanced, and antimycin-insensitive ECAR). Furthermore, rapamycin had an inhibitory effect on metabolic activity assessed by MTT assay and on TGF-β1-driven increase in the amount of protein in podocytes (Fig. 7, H and I). Taken together, these results indicate that the TGF-β1 effect on ATP generation, OCR, ECAR, metabolic activity, and protein generation is mediated, in part, by mTOR pathway.
In this study, we have shown that TGF-β1 increased OCR, ECAR, and cellular ATP levels in cultured mouse podocytes in a dose-dependent manner. TGF-β1 decreased RNA levels Sirt1 and Pgc1a, whose gene products enhance mitochondrial biogenesis. Expression of mitochondrial genes mt-Nd1, mt-Cytb, and mt-Co2 were unchanged, and mitochondrial mass assessed by a molecular probe was unchanged. Taken together, these data suggest that mitochondrial number and mass were likely unaffected by TGF-β1 and did not account for the increased mitochondrial activity. Analysis of MMP using the molecular probes JC-1 and TMRM indicated an increased MMP, which could explain the increase in OCR and ATP generation. The mechanisms for the increased glycolysis remain unclear. TGF-β1 increased intracellular ROS levels, which may reflect the increased oxidative phosphorylation. Rapamycin reduced the TGF-β1-stimulated increases in OCR, ECAR, ATP generation, cellular metabolic activity, and protein generation, suggesting that the mTOR pathway contributed to the effects on mitochondrial function.
Prior work lends support to the pathway we propose: TGF-β1 activates mTOR, which then stimulates mitochondrial oxidative phosphorylation. First, mTOR increases mitochondrial oxygen consumption in Jurkat T cells and C2C12 myoblasts (6, 26). Second, TGF-β activates the mTORC1 in fibroblasts (22) and murine mammary epithelial NMuMG cells (18), although not in Mv1Lu epithelial cells (22). Third, Wang et al. (33) reported that tuberin (TSC2) and mTOR are activated in unilateral obstructive nephropathy in rats in association with TGF-β1. Regarding TGF-β1-stimulated increases in OCR, Stieger et al. (30) showed podocytes cultured in prolonged high-glucose condition increased OCR that was even more pronounced when TGF-β was added (30). The novel contribution of the present work is to demonstrate that TGF-β1-stimulated OCR is mediated, at least in part, by the mTOR pathway.
Mitochondria are responsible for the production of cellular ATP and also are the chief source of ROS in most cell types (17). ROS are vital second messengers, as well as having antipathogenic effects in phagocytic cells, but excess levels are associated with cellular toxicity. Thus, an appropriate balance between ATP production and ROS is essential to maintain cellular homeostasis. ROS mediate glomerular injury in experimental glomerular diseases (28, 32, 35). TGF-β1 generates ROS in various types of cells. Yoon et al. (39) reported that TGF-β1 arrests mink lung epithelial MvLu cells at G1 cell cycle phase with acquisition of senescence, which is accompanied by prolonged generation of ROS and persistent disruption of MMP. Further, TGF-β1 decreased complex IV activity (but without affecting other components of the respiratory chain), and inhibition of complex IV activity also increased ROS, which together suggest that decreased complex IV activity was a likely mechanism for increased ROS levels. We also found that TGF-β1 increased ROS levels but through an opposite mechanism, increased OCR, reflecting increased mitochondrial activity in both primary and transformed podocytes. These cell type-specific differences may be due to differences in TGF-β1 effects on the mTOR pathway, as TGF-β1 activates mTOR in fibroblasts and podocytes but not Mv1Lu cells (22).
ROS are also produced outside of mitochondria. In particular, NADPH oxidase is an important source of ROS in cultured human podocytes (10). It is known that Nox4 is primary NADPH oxidase in the kidney (29). Bondi et al. (3) reported that TGF-β1 generates ROS via Nox4 activation in kidney myofibroblasts. NADPH oxidase contributes to nonmitochondrial oxygen consumption, which is chiefly composed of cell surface oxygen consumption, peroxisomal oxygen consumption, and substrate oxidation (2, 12). Although nonmitochondrial oxygen respiration was high after exposure to TGF-β1 and a portion of ROS might be generated by Nox4, Nox4 protein expression did not change in our present study. Furthermore, apocynin, an NADPH inhibitor, did not reduce TGF-β1-induced ROS generation. Therefore, it is unlikely that NADPH oxidase is highly activated by TGF-β1 in mouse podocytes. As Bishop and Brand (2) have suggested, high nonmitochondrial respiration could be a protective mechanism to remove oxygen when it is present at potentially harmful concentrations. Taken together, these results indicate that the increased mitochondrial oxidative phosphorylation likely results in generation of mitochondrial ROS and can be a potential mechanism contributing to podocyte injury.
TGF-β1 reduces MMP in Mv1Lu cells and hepatocytes (11, 23, 39). By contrast, TGF-β1 increased MMP in mouse podocytes. These differences may reflect cell type-specific differences in cellular response. Because ROS induce mitochondrial membrane permeabilization both in vitro and in vivo (16), prolonged mitochondrial ROS generation could decrease MMP in mouse podocytes, when the exposure lasts longer than the period we studied. MMP is positively correlated with OCR, and it is dependent on supplied substrates, such as glucose and pyruvate (5). It is also known that rapamycin decreases MMP, as well as OCR (26). In the present study, TGF-β1 increased OCR and MMP, and rapamycin decreased OCR induced by TGF-β. Therefore, our data suggest a model in which TGF-β1 increases MMP, leading to increased OCR.
SIRT1 and PGC1α activation increases mitochondrial biogenesis, while UCP2 expression is negatively correlated with SIRT1 (34, 38, 40). Although UCPs lower MMP and ROS production (8, 38), UCPs are activated by superoxide, which is the proximal mitochondrial ROS (21). TGF-β1 decreased mRNA expression of Sirt1 and Pgc1α, but increased that of Ucp2, MMP, and ROS generation in our system. Recently, it was reported that SIRT1 negatively regulates mTOR signaling, potentially through the TSC1/2 complex (9). Thus, it is possible that the effects of TGF-β1 on mTOR observed here might be mediated via Sirt1 downregulation.
It is notable that TGF-β1 significantly increases basal OCR and that the increased OCR is coupled to ATP generation and increased ATP content. Clearly, some ATP-demand processes are upregulated, but it was shown that the cells are not proliferating in response to TGF-β1. In the absence of any increased workload, these results suggest that the cells have increased protein synthesis and/or growth, i.e., hypertrophy. Although TGF-β1-mediated fibrosis is typically associated with fibroblast hypertrophy and extracellular matrix remodeling, postmitotic cells, such as the podocyte, are not immune to the mTOR-mediated hypertrophic stimuli afforded by TGF-β1, as is seen with podocytes in this study. Indeed, TGF-β1 signaling along the Akt-mTOR axis mediates the pathology of diabetic nephropathy and much of the associated damage is ameliorated via inhibition of the mTORC1 complex (20, 36). What is unique to our study is the observation that hypertrophy arises in the absence of any mitochondrial biogenesis, although the mitochondria have increased capacity via increased capacity for fatty acid utilization. Ultimately, the hypertrophy in absence of mitochondrial biogenesis results in increased ROS and associated oxidative stress. This dysregulated growth in postmitotic cells can ultimately result in pathological states, such as those seen in hypertrophic cardiomyopathies, neurodegenerative diseases, and perhaps kidney failure.
Our study has certain limitations. First, much of our data was derived from a transformed mouse podocyte cell line, which has the advantage of providing a stable cell phenotype over time. It is possible that transformation alters cellular bioenergetics. To address this possibility, we confirmed key aspects of our work in primary mouse podocytes. Second, the time scale of our experimental manipulations was necessarily short, so the effect of particular interventions over a period of days or weeks could not be assessed. Third, these data may or may not reflect the bioenergetic profile of podocytes in vivo, where these cells receive cues from other cells and a specialized extracellular matrix and are bathed in a highly complex plasma milieu, with lower oxygen tension than is typical in cell culture. Fourth, rapamycin was used to verify the role of mTOR in our system. While rapamycin is widely used as an inhibitor of mTOR, it could have off-target effects and this remains a limitation.
In conclusion, we propose that TGF-β1 alters bioenergetic profile in mouse podocytes. Specifically, TGF-β1 increases ATP generation, OCR (oxidative phosphorylation), and fatty acid oxidation, and also activates glycolysis when the mitochondrial function is inhibited. These changes are associated with increased ROS levels, which is likely due to increased oxidative phosphorylation and may also be due to increased cellular oxidase activity. Activation of the mTOR pathway by TGF-β1 is at least partly responsible for the increased mitochondrial function. We have summarized a proposed pathway by which TGF-β1 stimulates ROS generation induced by TGF-β1 (Fig. 8). These findings may shed light on TGF-β1-induced podocyte injury in vivo in the setting of glomerular injury.
No conflicts of interest, financial or otherwise, are declared by the authors.
Author contributions: Y.A., T.S., and J.B.K. conception and design of research; Y.A. performed experiments; Y.A., T.S., C.B., and J.B.K. analyzed data; Y.A. and J.B.K. interpreted results of experiments; Y.A. prepared figures; Y.A. and J.B.K. drafted manuscript; Y.A. and J.B.K. edited and revised manuscript; Y.A., T.S., C.B., and J.B.K. approved final version of manuscript.
This study was supported by the National Institute of Diabetes and Digestive and Kidney Diseases Intramural Research Program, under project ZO1-DK043308. We are grateful to Hideko Takahashi and Huiyan Lu for technical assistance with animal care, Satoshi Watanabe for technical assistance with electron microscopy, Dr. Hisashi Hasumi and Dr. Masaya Baba for valuable advice, and Dr. Kevin Bittman and Dr. Min Wu, both of Seahorse Bioscience, for valuable advice in the design and interpretation of particular experiments. We wish to thank Dr. Tom Leto for a critical review of the manuscript.
- Copyright © 2013 the American Physiological Society