Glucocorticoid Hormone Stimulates Mitochondrial Biogenesis ...

0013-7227/02/$03.00/0 Printed in U.S.A.

Endocrinology 143(1):177–184 Copyright © 2002 by The Endocrine Society

Glucocorticoid Hormone Stimulates Mitochondrial Biogenesis Specifically in Skeletal Muscle ¨ CK, ZSUZSANNA MIKES, JAN-HEINER KU ¨ PPER, KATHARINA WEBER, PATRICK BRU MARTIN KLINGENSPOR, AND RUDOLF J. WIESNER Department of Vegetative Physiology (K.W., R.J.W.), University of Ko¨ln, Ko¨ln 50931, Germany; Department of Physiology II (P.B., Z.M.), University of Heidelberg, Heidelberg 69120, Germany; Department of Molecular Pathology (J.-H.K.), University Hospital of Tu¨bingen, Tu¨bingen 72076, Germany; and Department of Animal Physiology (M.K.), University of Marburg, Marburg 35043, Germany High levels of circulating glucocorticoid hormone may be important mediators for elevating resting metabolic rate upon severe injury or stress. We therefore investigated the effect of dexamethasone on mitochondrial biogenesis in rats (6 mg/kg daily) as well as in cells in culture (1 ␮M) over a period of 3 d. A marked stimulation of mitochondrial DNA transcription and increased levels of cytochrome c oxidase activity were found in skeletal muscle of rats and differentiated mouse C2C12 muscle cells, but not in other tissues, myoblasts, or other cell lines. The effect was inhibited by RU486. Therefore, increased occupancy of glucocorticoid receptors is necessary, but not sufficient to increase mitochondrial biogenesis and

other, skeletal muscle specific factors are postulated. Expression of the mitochondrial transcription factor A was unchanged, suggesting a possible involvement of the recently described mitochondrial glucocorticoid receptor. Expression of uncoupling protein-3 was also unchanged. In conclusion, our results show that high levels of glucocorticoid hormone are sufficient to stimulate mitochondrial biogenesis; however, only in skeletal muscle. Increased mitochondrial mass in this tissue, without changes of the coupling state of the respiratory chain, might be the molecular basis for the elevated resting metabolic rate observed under high cortisol levels in humans. (Endocrinology 143: 177–184, 2002)

I

NCREASED LEVELS OF glucocorticoid hormones may be important for mediating the hypermetabolic state that is seen after traumatic injury or during stress. This state is characterized by accelerated protein turnover in skeletal muscle, stimulated gluconeogenesis in the liver, hyperglycemia and insulin resistance, and, finally, augmented energy expenditure. Indeed, cortisol infusion to healthy human subjects in a physiological to supraphysiological range showed, in a dose-dependent manner, an increase of resting energy expenditure (REE) ( 1, 2). This was accompanied by an elevation of circulating glucose as well as FFA, but rather a shift toward fatty acid oxidation. It is not known which energy demanding processes are responsible for the augmented REE and in which tissue these occur. Although a stimulation of protein turnover was observed, this energy demanding mechanism could only account for up to 10% of the observed augmentation of REE (1). Mitochondria are the main source of ATP in almost all cells and are responsible for approximately 90% of total oxygen consumption. However, 20 –30% of the resting metabolic rate (RMR) is not explained by mitochondrial ATP synthesis but rather due to futile proton cycling across the proton leak of the inner mitochondrial membrane (3). It was reported that dexamethasone, a glucocorticoid hormone analog, stimulates transcription of genes encoded on mitochondrial DNA (mtDNA) in some hepatoma cell lines (4, 5). A rise of mito-

chondrial transcripts was also found in colon epithelium of rats after dexamethasone injection. This was interpreted as a compensatory stimulation of mitochondrial biogenesis preventing limitation of ATP supply to the stimulated Na⫹ transport occurring under these conditions (6). In the same study, an increase of cytochrome c oxidase subunit I (CO I) mRNA was also reported for skeletal muscle. This led us to the hypothesis that high glucocorticoid hormone levels may stimulate mitochondrial proliferation mainly in skeletal muscle, thus contributing significantly to the rise of RMR occurring under such circumstances. Thus, the effect of dexamethasone on mitochondrial transcript levels was initially studied in skeletal muscle and other rat tissues as well as in various cell lines. Quadriceps muscle as well as the mouse skeletal muscle cell line C2C12 was chosen for further investigation. Materials and Methods Animal treatment Ten male Wistar rats (200 –250 g; Heidelberg University Animal House) were used for the in vivo study. Five animals received dexamethasone dissolved in saline (6 mg/kg body weight, 100 ␮l volume, sc) on 3 consecutive days at 1000 h; five control animals received saline alone. Three days after the first injection, rats were anesthetized with ketamine/xylazine (100 mg/kg and 4 mg/kg, respectively), and the body weight was determined. Liver, kidney, soleus, and quadriceps muscle were removed, freeze clamped with stainless steel prongs at the temperature of liquid nitrogen, and stored at ⫺80 C for later analysis. The distal colon was opened, cleaned with saline, and the epithelium was scraped off and transferred to the RNA extraction solution directly (see below). All experiments were conducted in accordance with institutional guidelines and the Guide for the Care and Use of Laboratory Animals put forth by the U.S. Department of Health and Human Ser-

Abbreviations: CO, Cytochrome c oxidase; GR, glucocortocoid hormone receptor; mtDNA, mitochondrial DNA; mtTFA, mitochondrial transcription factor A; OXPHOS, oxidative phosphorylation; REE, resting energy expenditure; RMR, resting metabolic rate; UCP, uncoupling protein.

177

178

Endocrinology, January 2002, 143(1):177–184

vices, NIH Publication No. 86 –23, and were approved by local authorities.

Cell culture Mouse C2C12 myoblasts were maintained in DMEM containing 10% FCS. Upon confluence, the medium was changed to DMEM containing 2% horse serum, which had been depleted of steroid hormones by treatment with activated charcoal (7). After 3 d, when myoblasts had fused to myotubes, medium containing dexamethasone or control medium was added and cells were harvested for analysis after further 3 d. In some experiments, RU 486 (10 ␮m) was added together with dexamethasone. Media were replaced daily, and lactate was analyzed in the removed samples. Reuber hepatoma cells (H-4-II-E) were cultivated in DMEM containing 10% FCS. SV40-tranformed Chinese hamster embryo cells (CO60), which had been engineered to constitutively overexpress the glucocorticoid hormone receptor were described in detail previously (COR cells) (8). For total protein and DNA analysis, cells were lysed in perchloric acid, centrifuged, the denatured protein pellet was dissolved in NaOH and assayed by the Bradford method (9). Total DNA was measured in the supernatant after complete nucleic acid hydrolysis by pentose analysis (10).

Isolation and blotting of nucleic acids RNA was isolated from tissue pulverized under liquid nitrogen according to Chomczynski and Sacchi (11). RNA was isolated from cultured cells using a commercially available RNA extraction kit (RNeasy, QIAGEN, Hilden, Germany). Mitochondrial transcript levels were assayed in rat tissues by slot blot analysis using a rat CO III probe. Steady-state levels of mitochondrial mRNAs were analyzed in detail by Northern blots (quadriceps muscle, C2C12 myotubes, hepatoma cells, COR cells) using formaldehyde containing agarose gels, nitrocellulose and the capillary transfer method, loading 5 ␮g of tissue RNA and 10 ␮g of cell RNA per lane. To measure mtTFA mRNA in rat quadriceps, 25 ␮g of total RNA was loaded onto the gel.

Hybridization of RNA blots Blots were prehybridized for 2 h and hybridized overnight at 42 C (prehybridization: 40% formamide; 5⫻ SSC (1⫻ SSC ⫽ 0.15 m NaCl, 0.015 m Na-citrate); 50 mmol/liter phosphate buffer, pH 7.4; 10⫻ Denhardt’s solution (1⫻ Denhardt’s ⫽ 0.2 g Ficoll/liter, 0.2 g polyvinylpyrolidone/liter, 0.2 g BSA/liter); 0.2% SDS; 500 ␮g/ml salmom sperm DNA; hybridization: 50% formamide; 3⫻ SSC; 10 mmol/liter phosphate buffer, pH 7.4; 2⫻ Denhardt’s solution; 0.2% SDS, 170 ␮g/ml salmom sperm DNA). cDNA probes isolated from appropriate plasmids or PCR products labeled to high specific radioactivity by the random priming method were used for hybridization (12). Probes for mitochondrial transcripts were described in detail previously (13). After hybridization, blots were washed at 42 C (2 ⫻ 15 min in 2⫻ SSC, 0.1% SDS followed by 2 ⫻ 15 min in 0.1⫻ SSC, 0.1% SDS). A mouse full-length uncoupling protein (UCP)-3 cDNA was cloned by RT-PCR. Poly A⫹ RNA isolated from skeletal muscle was reverse transcribed using SuperScript II with random hexamers (Life Technologies, Karlsruhe, Germany). A full-length UCP-3 fragment was amplified from cDNA by PCR using the primer 5⬘ CTA ATG GAG TGG AGC CTT AGG-3⬘ (forward) and primer 5⬘-GCC TGC TTG CCT TGT TCA-3⬘ (reverse). The PCR product at a size of 1093 bp was gel-extracted, ligated by T/A cloning in pGEM-T, and transformed into Escherichia coli DH5␣ for amplification. To measure mtTFA mRNA, a 1550-bp probe encoding mouse mTFA cloned into the EcoRI site of pBS-KS was used (14); in this case, hybridization temperature was 38 C, yeast total tRNA was used instead of salmon sperm DNA for blocking and only two washes were employed after hybridization (2 ⫻ 15 min, 0.1⫻ SSC, 0.1% SDS, 38 C). Between hybridizations, blots were stripped from the previous probe (4 ⫻ 5 min incubations in boiling 0.01⫻ SSC, 0.01% SDS) and were finally hybridized to cytosolic 28S tRNA for normalization. For this, hybridization temperature was 44 C and the last two washing steps were performed at 50 C. Blots were exposed to x-ray films, and the films were evaluated densitometrically using a video camera-based analysis system and AIDA software version 1.0 (Raytest, Straubenhardt, Germany). For quantitation, densitometric data for mitochondrial transcripts were normalized to the 28S rRNA signal, taking care that the signal was in the linear range of the film.

Weber et al. • Glucocorticoids Stimulate Mitochondrial Biogenesis

Immunoblotting Small pieces of frozen tissue or cell pellets were homogenized in 62.5 mm Tris (pH 6.8), 2% SDS, 10% glycerol at 95 C using a small glass-Teflon homogenizer. Protein samples (20 ␮g) were run on 12.5% slab gels (8 ⫻ 7 x 0.15 cm) and a 3% stacking gel at 100 V, for 3 h. Proteins were transferred to nitrocellulose in an electroblot apparatus in 154 mm glycin, 20 mm Tris (pH 8.3) and 20% methanol at 30 V, 100 mA for 3 h. Blots were subsequently blocked in 20 mm Tris, pH 7.5, 150 mm NaCl, 0.5% Tween, 2% BSA, and 1% milk powder for 2 h and incubated overnight in the same buffer containing rabbit antiserum raised against mouse mtTFA (15) in a 1:1000 dilution or a polyclonal antibody against UCP-3 (Alexis Biochemicals, Grumberg, Germany) in a 1:1000 dilution. As a positive control, the mouse UCP-3 protein was overexpressed in HEK293 cells by transient transfection. For this, the UCP-3 insert was excised from pGEM-T-UCP-3 with NotI/ApaI and subcloned into a CMV-driven expression vector (pEGFP-N1, CLONTECH Laboratories, Inc.). After washing, protein bands were visualized by incubation with donkey-antirabbit IgG antiserum, horseradish-peroxidase conjugated and chemiluminescence detection (ECL, Amersham Pharmacia Biotech, Freiburg, Germany). Chemiluminescent blots were exposed to x-ray films and bands of interest were evaluated densitometrically using a video camera based analysis system and the AIDA, Version 1.0, software (Raytest).

Determination of CO activity For analysis of the mitochondrial marker enzyme, cytochrome c oxidase, a small piece of frozen muscle (20–30 mg) was homogenized with a glass homogenizer and pestle in 1 ml of ice-cold phosphate buffer (100 mmol/ liter, pH 7.0). C2C12 muscle cells were rinsed with PBS, harvested by scraping them off the culture plates and homogenized in 500 ␮l of the same buffer. Maximal enzyme activity was determined spectrophotometrically by measuring the rate of oxidation of reduced horse heart cytochrome c (Sigma, Taufkirchen, Germany), reflected by the change in absorbance at 550 nm (16). The protein concentration of the homogenates was measured using BSA as standard (9). Enzyme activity was then expressed as enzymatic units (␮mol cytochrome c min⫺1 mg protein⫺1), using the millimolar extinction coefficient of 29.5 for reduced horse heart cytochrome c.

Lactate production For quantitation of lactate production, the cell culture media were deproteinized with 1/10 volume of 6 m perchloric acid and centrifuged for 20 min at 20,000 ⫻ g. The supernatant was neutralized with 1/10 volume of 6 m KOH, KClO4 was pelleted and the resulting supernatant was used for assaying lactate concentration by a spectrophotometric test using lactate dehydrogenase coupled to NAD⫹-reduction.

Statistical evaluation of results Results are expressed as mean values ⫾ sd and groups were compared by t test; P ⬍ 0.05 was assumed to be statistically significant.

Results Effect of dexamethasone on mitochondrial biogenesis in rat tissues

The body weight of rats was 214 ⫾ 38 g in control vs. 234 ⫾ 29 g in dexamethasone-treated animals, indicating no gross changes in growth or total muscle mass within the observation period. The heart weight/body weight ratios were 2.40 ⫾ 0.37 mg/g in controls vs. 2.62 ⫾ 0.10 mg/g in dexamethasone treated animals (P ⫽ NS), indicating that no significant cardiac hypertrophy had occurred under the experimental conditions. Because levels of mtDNA encoded transcripts generally correlate well with oxidative phosphorylation (OXPHOS) capacity, in a first screen RNA of various rat tissues was blotted and hybridized to a cDNA-probe for CO III mRNA. Figure 1 shows that this mRNA was significantly increased in quadriceps and soleus muscle. These


Endocrinology, January 2002, 143(1):177–184 179

FIG. 1. Levels of cytochrome c oxidase subunit III mRNA in various rat tissues following treatment with dexamethasone for 3 d. Slot blots of tissue RNA were consecutively hybridized to radiolabeled cDNA probes for CO III followed by 28S rRNA. Data are densitometric values normalized to 28S rRNA (arbitrary units, mean ⫾ SD, n ⫽ 5). *, Significantly different from control tissue, P ⬍ 0.05.

were chosen because they represent skeletal muscle mainly composed of type II fast-glycolytic or type I slow-oxidative fibers, respectively. CO III mRNA was significantly decreased in liver but was rather unchanged in kidney and colon epithelium. The decrease of mitochondrial gene expression in liver was also confirmed by significantly lower levels of 12S rRNA (not shown). Quadriceps muscle was then further investigated in detail. A Northern blot (Fig. 2A) shows a marked stimulation of mtDNA transcription by dexamethasone. Quantitative data for mitochondrially encoded transcripts are given in Fig. 2B. As examples for respiratory chain subunits encoded by the nucleus, the mRNA for cytochrome c also slightly increased after dexamethasone treatment (0.61 ⫾ 0.11 vs. 0.81 ⫾ 0.05, control vs. dexamethasone, P ⬍ 0.01), whereas no significant increase was found for mRNA for CO subunit VIc (0.81 ⫾ 0.30 vs. 0.91 ⫾ 0.33). Levels of the mRNA for the nuclear encoded mitochondrial transcription factor A were also found to be similar after 3 d of dexamethasone treatment (0.45 ⫾ 0.13 vs. 0.55 ⫾ 0.13, control vs. dexamethasone, P ⫽ NS). Because this is a low abundance transcript, in contrast to mRNAs for respiratory chain subunits, mtTFA mRNA levels were also normalized to GAPDH mRNA to exclude nonlinearity of the 28S rRNA signal on these blots. However, also the mtTFA/ GAPDH ratio was similar in the two groups (0.76 ⫾ 0.15 vs. 0.80 ⫾ 0.10). In addition, levels of mtTFA protein were analyzed by Western blotting and were found to be similar in quadriceps muscle of control and dexamethasone treated rats (data not shown). Cytochrome c oxidase activity as a marker for total OXPHOS capacity was measured by a spectrophotometric method in quadriceps muscle extracts and was found to be elevated about 2-fold after 3 d of hormone treatment (59.9 ⫾ 10.1 vs. 105.2 ⫾ 14.2 ␮mol cytochrome c min⫺1 mg protein⫺1, control vs. dexamethasone: P ⬍ 0.01). Effects of dexamethasone on mitochondrial biogenesis in C2C12 cells

Among the tissues investigated in this study, the stimulatory effect of glucocorticoid hormone on mitochondrial

FIG. 2. Levels of mitochondrial transcripts in rat quadriceps muscle following treatment with dexamethasone for 3 d. A, Northern blot showing hybridization results for CO III mRNA, its unprocessed precursor, and 12S rRNA encoded by mitochondrial DNA as well as 28S rRNA used for normalization. Each lane contains RNA from one individual sample. B, Quantitative data obtained by densitometry from the Northern blot shown in A. Data have been normalized to 28S rRNA (arbitrary units, mean ⫾ SD, n ⫽ 5).*, Significantly different from control tissue, P ⬍ 0.05.

biogenesis seemed to be rather specific for skeletal muscle. To further confirm this cell specificity, mouse C2C12 myotubes were treated with dexamethasone for 3 d. To maximize the effect of the hormone, cells were cultivated in medium containing serum which had been stripped of endogenous steroid hormones. No obvious differences in myotube morphology or fusion index was observed between the two groups. Also, the protein to DNA ratio (mg/mg) as a marker for myotube differentiation was similar, 116 ⫾ 19 in control vs. 100 ⫾ 22 (P ⫽ NS) in dexamethasone treated cells. Mitochondrial transcripts markedly increased in the presence of 1 ␮m dexamethasone (Fig. 3A), but not at the lower concentration of 0.1 ␮m. MtTFA protein was measured by Western blotting and, like in quadriceps muscle, was found to be rather unchanged under all conditions (Fig. 3B). In an independent experiment, the increase of mitochondrial transcripts was shown to be ablated by the presence of the glucocorticoid hormone receptor antagonist RU 486 (10 ␮m) (Fig. 3C). The normalized CO III mRNA rose from 1.35 ⫾ 0.11 in controls to 2.30 ⫾ 0.14 after dexamethasone (1 ␮m) treatment (P ⬍ 0.001) but remained unchanged in the presence of RU 486 during dexamethasone treatment (1.35 ⫾ 0.11 vs. 1.50 ⫾ 0.57, P ⬎ 0.67 control vs. dexamethasone ⫹ RU 486). Similar results were obtained for CO II mRNA, together providing strong evidence that hormone action was due to binding to steroid hormone receptors.

180



TABLE 1. Levels of transcripts for mitochondrial proteins in C2C12 myotubes after dexamethasone treatment for up to 3 d 24 h

48 h

72 h

CO II

Control Dexa

0.21 ⫾ 0.07 0.49 ⫾ 0.09a

0.12 ⫾ 0.09 0.43 ⫾ 0.04a

0.30 ⫾ 0.12 0.58 ⫾ 0.05a

CO III

Control Dexa

0.22 ⫾ 0.08 0.47 ⫾ 0.10a

0.11 ⫾ 0.01 0.31 ⫾ 0.16

0.12 ⫾ 0.08 0.30 ⫾ 0.03a

12S rRNA

Control Dexa

0.41 ⫾ 0.13 0.60 ⫾ 0.25

0.35 ⫾ 0.02 1.08 ⫾ 0.21a

0.61 ⫾ 0.15 1.00 ⫾ 0.15a

Cytochrome c Control Dexa

0.11 ⫾ 0.03 0.25 ⫾ 0.08a

0.05 ⫾ 0.01 0.38 ⫾ 0.15

0.15 ⫾ 0.05 0.28 ⫾ 0.03a

mtTFA

Control Dexa

0.47 ⫾ 0.13 0.65 ⫾ 0.03

0.64 ⫾ 0.06 0.83 ⫾ 0.21

0.66 ⫾ 0.04 0.69 ⫾ 0.12

UCP-3

Control Dexa

0.40 ⫾ 0.01 0.07 ⫾ 0.02a

0.23 ⫾ 0.00 0.24 ⫾ 0.11

0.27 ⫾ 0.13 0.15 ⫾ 0.02

Data are densitometric values normalized to 28S rRNA (arbitrary units, mean ⫾ SD, n ⫽ 3; a, significantly different from control; P ⬍ 0.05).

FIG. 3. Mitochondrial transcripts and mtTFA protein levels in C2C12 myotubes following treatment with dexamethasone. A, Northern blot showing hybridization results for CO II and CO III mRNA as well as 28S rRNA used for normalization. B, Western blot showing levels of mtTFA protein in the same samples. C, Influence of the glucocorticoid hormone receptor antagonist RU486 on expression of mitochondrial mRNAs encoding CO II and CO III. Dexamethasone (Dexa) or dexamethasone ⫹ RU 486 (RU 486) were present. Each lane contains RNA or protein from one individual sample after 3 d of treatment.

The time course of changes of mitochondrial transcripts was analyzed in more detail over a period of 3 d (Table 1). For representation, the data of d 2 are shown in Fig. 4. The mRNAs for CO subunits II and III were significantly increased already 24 h after dexamethasone treatment and remained elevated about 2-fold, whereas 12S rRNA levels rose with a delay. Also, the nuclear encoded cytochrome c mRNA was elevated, whereas mtTFA levels remained unchanged again. The mRNA for CO subunit VIc could not be measured in the mouse cell line with the rat probe that was used. Also given are the data for UCP-3 mRNA, which will be discussed below. Also in this system, cytochrome c oxidase activity was measured spectrophotometrically and was found to be elevated 2.5-fold after 3 d of treatment with hormone (2.2 ⫾ 0.3 vs. 5.7 ⫾ 1.1, control vs. dexamethasone, P ⬍ 0.01). To show that the expanded OXPHOS capacity was actu-

FIG. 4. Levels of transcripts encoding mitochondrial proteins and 12S rRNA in C2C12 myotubes following treatment with dexamethasone. The Northern blot shows hybridization results after 2 d of treatment. Each lane contains RNA from one individual sample. Quantitative data for these transcripts over a period of 3 d are given in Table 1.

ally used for oxidative energy metabolism, the release of lactate into the cell culture medium was analyzed. Lactate production increased gradually in control myotubes (Fig. 5, upper panel), reflecting increasing cell mass. It was significantly lower already after 1 d of dexamethasone treatment and decreased continually over the observation period of 3 d, indicating a shift to ATP production by oxidative phosphorylation. To exclude that this was due to the slightly different cell masses of control vs. dexamethasone-treated myotubes (see above), in another experimental series total cellular protein per dish was determined and lactate production was found to be 13.8 ⫾ 2.0 vs. 6.1 ⫾ 1.1 mmol mg protein⫺1 day⫺1 (control vs. dexamethasone, P ⬍ 0.01) at d 3. Because lactate production rate is an easily accessible indicator of oxidative vs. anaerobic metabolism, it was also measured in undifferentiated myoblasts (Fig. 5, lower panel).


FIG. 5. Lactate production in C2C12 myotubes (A) and myoblasts (B) following treatment with dexamethasone for 3 d. Cell culture media were replaced after 24 h and lactate release was measured by a spectrophotometric method (mean ⫾ SD, n ⫽ 3). *, Significantly different from control, P ⬍ 0.05.

Lactate production steadily increased over time, probably due to increasing cell number in control cells. This was ablated by dexamethasone; however it is clear that the hormone effect was much less pronounced compared with fully differentiated myotubes (Fig. 5, upper panel). Effects of dexamethasone on mitochondrial biogenesis in other cell lines

We found that dexamethasone stimulated mitochondrial biogenesis rather specifically in skeletal muscle in vivo. Thus, we investigated whether the glucocorticoid hormone effect was also specific for differentiated skeletal muscle cells in vitro. Indeed, in our hands, no changes of mitochondrial transcripts were found in hepatoma H4-II-E cells (21a). To exclude the possibility that the absence of hormone effect is due to low levels of glucocorticoid receptors in the cell strains we used, dexamethasone was also applied to COR cells, SV40-tranformed Chinese hamster embryo cells that had been engineered to constitutively overexpress the glucocorticoid receptor (8). However, also in this cell line, no changes of mitochondrial transcript levels were observed upon glucocorticoid hormone addition (Fig. 6A, P ⫽ NS), and lactate production was similar in both groups (Fig. 6B, P ⫽ NS). Effect of dexamethasone on the expression of UCP-3

High levels of UCP-3 might cause increased RMR under the influence of high circulating glucocorticoid levels. Thus, UCP-3 expression was analyzed in rat quadriceps muscle as well as C2C12 myotubes upon dexamethasone treatment. To


FIG. 6. Levels of mitochondrial transcripts and lactate production in COR cells following treatment with dexamethasone for up to 2 d. A, Northern blot showing hybridization results for CO I mRNA and 12S rRNA encoded by mitochondrial DNA as well as 28S rRNA used for normalization. Each lane contains RNA from one individual sample. B, Lactate production: cell culture media were replaced after 24 h and lactate release was measured by a spectrophotometric method (mean ⫾ SD, n ⫽ 3).

be able to unequivocally identify the protein in the skeletal muscle samples, UCP-3 was overexpressed in HEK293 cells and protein lysates of these cells were also loaded on the Western blots. However, no difference was found between skeletal muscle of control and dexamethasone treated rats, neither on the level of mRNA (Fig. 7A) nor protein (Fig. 7B). In C2C12 myotubes, there was even a tendency toward lower levels of UCP-3 mRNA following glucocorticoid hormone exposure (Fig. 7C and Table 1). Unfortunately, the UCP-3 protein was below the limits of detection in myotube lysates and even in fractions of these cells enriched in mitochondria (not shown). Discussion

Our results show that high levels of circulating glucocorticoid hormone stimulate mitochondrial biogenesis, and that among the cell types chosen for analysis, this phenomenon is rather specific for skeletal muscle. The increased mitochondrial content in muscle may thus be the molecular basis for the elevated RMR observed under high cortisol levels in humans (1, 2). Liver would also be large enough to contribute significantly to whole body energy turnover; however, we found rather decreases of mitochondrial transcripts. This is an interesting observation probably related to the switch to gluconeogenesis and deserves further study. Glucocorticoid hormones have also been reported to be essential for the

182


FIG. 7. Expression of UCP-3 in rat quadriceps muscle and C2C12 myotubes following treatment with dexamethasone for 3 d. A, Northern blot and (B) Western blot analysis of quadriceps muscle, (C) Northern blot analysis of C2C12 myotubes. Quantitative data for UCP-3 over a period of 3 d are given in Table 1. Each lane contains RNA or protein from one individual sample.

postnatal rise in mitochondrial enzymes in the developing rat kidney (17, 18); however, we found no effect in the adult rats used here. The colon was investigated because it was postulated that mitochondrial biogenesis may be induced by the glucocorticoid action of high aldosterone levels in distal colon epithelial cells to cover the costs of increased salt and water retention (6); however, we could not find any significant changes of mitochondrial transcript levels in this organ. As it was shown before in several other models, also upon high glucocorticoid levels, mitochondrial biogenesis was stimulated primarily by up-regulation of transcription of the mitochondrial genome. Transcript levels of mtDNA, in particular mitochondrial mRNAs, generally correlate well with functional mitochondrial mass, either when different tissues are compared (19, 20) as well as under conditions of stimulated mitochondrial biogenesis (reviewed in Ref. 21). This was confirmed in the present study. Concomitant with 2-fold elevated levels of mitochondrial mRNAs, we found a 2-fold increase of cytochrome c oxidase activity in quadriceps muscle. Similarly, mitochondrial mRNAs rose 2- to 3-fold in dexamethasone treated myotubes together with a 2.5-fold induction of cytochrome c oxidase activity. To show that


elevated respiratory chain activity was actually used for enhanced aerobic metabolism, lactate production was measured in myotubes and found to be clearly decreased upon dexamethasone treatment. In contrast, no significant changes of mitochondrial parameters were found in undifferentiated myoblasts nor in H4-II-E hepatoma cells (21a), although this cell line had been used before by others to study dexamethasone effects on mtDNA transcript levels (4, 5). This may be due to strain differences, different serum lots, different cell confluence or other factors. To exclude the possibility that it is due to low levels of glucocorticoid receptors in the strain used, cells were also used that constitutively overexpress the glucocorticoid receptor (8). However, also in the COR cells, dexamethasone did not affect mitochondrial biogenesis. Thus, increased occupancy of glucocorticoid receptors alone is not sufficient to increase mitochondrial gene expression. One possibility is that some unknown, additional permissive factor(s) is/are present in differentiated skeletal muscle cells that, together with occupied GR, act(s) on nuclear genes controlling mitochondrial biogenesis. Alternatively, energy demanding processes may be up-regulated specifically in skeletal muscle by glucocorticoids, which then modulate mitochondrial biogenesis to cover the expanded ATP-turnover. Strong support for the involvement of additional factors comes from the fact that cyt c mRNA was increased in both rat muscle and mouse myotubes upon dexamethasone, but no glucocorticoid responsive element can be found in the rat and mouse cyt c promoters, using the MatInspector software (22). How does dexamethasone stimulate transcription of mtDNA, which is an important step in mitochondrial biogenesis? The mtTFA gene [tfam (23)], which was shown to be up-regulated in several models, does not seem to be a target for the glucocorticoid pathway because the mtTFA mRNA was unchanged in both models. Indeed, also the mtTFA promoter of the rat and mouse does not contain obvious glucocorticoid responsive elements. Recently, convincing evidence was provided that a glucocorticoid receptor is present within mitochondria in several cell types (24), and that mtDNA contains putative GR binding consensus sequences (25). It was thus proposed that glucocorticoids stimulate mtDNA transcription by a direct interaction with a mitochondrial GR. Such a mechanism was convincingly shown to operate for thyroid hormone via the mitochondrial T3 receptor p43 (26). In this case, mtTFA seems to regulate the overall activity of the mtDNA transcription machinery, whereas mitochondrial p43 upon T3 -binding regulates in a very subtle way the mRNA/rRNA ratio due to selection of the mtDNA transcription start site (27). The functionality of the mitochondrial GR, however, has not been demonstrated yet. Upon addition of RU486, the stimulatory effect of dexamethasone was ablated, so no evidence indicative for some novel way of glucocorticoid hormone action could be shown using this tool. However, because the mitochondrial GR is highly homologous to the cytosolic receptor, RU486 may also antagonize the mitochondrial GR. Thus, at the moment, our data cannot contribute to the question whether the mitochondrial GR is functional and is involved in up-regulation


of mtDNA transcription under high circulating glucocorticoid levels. Finally, what causes elevated RMR upon high cortisol levels? One alternative mechanism to augmented ATP-turnover may be increased expression of UCP. UCP-3 is a potential candidate because this protein is expressed in skeletal muscle (28) and was shown to be induced by thyroid hormone, which also increases RMR (29). Mice overexpressing human UCP-3 in muscle are hyperphagic, but lean, providing strong evidence that high levels of UCP-3 indeed stimulate RMR in vivo (30). A 2-fold, although insignificant increase of UCP-3 mRNA has been reported in muscle of rats treated with dexamethasone (31) and also, UCP-3 expression is under the control of circulating fatty acids (32) and insulin (33). Thus, we reasoned that a direct effect of dexamethasone on UCP-3 expression alone, the hyperlipedemia and hyperinsulinemia following high circulating glucocortocoids, or a combination of these factors may up-regulate the UCP-3 gene. However, no changes of UCP-3 expression could be seen after dexamethasone treatment, neither in muscle nor in myotubes. Thus, glucocorticoids do not significantly change UCP-3 expression in skeletal muscle, neither directly, as shown in vitro nor indirectly via other systemic and metabolic effects, as shown in vivo. Uncoupling therefore does not seem to contribute to elevated RMR under high glucocorticoid levels. Alternatively, can the increase of mitochondrial mass alone, without a concomitant rise of ATP-turnover, explain the elevated RMR induced by high cortisol levels? In humans, at a constant infusion of 200 ␮g/kg䡠h of hydrocortisone, which led to a 6-fold elevation of circulating cortisol, a 13% increase of whole body REE was reported (1). These are cortisol levels that are reported during severe stress (34). It is reasonable to postulate that glucocorticoid receptors were saturated under this regime, but also in the models used here, so that maximal effects were probably observed in all systems. It is also reasonable to assume that the 2-fold elevation of cytochrome c oxidase in our models reflects a 2-fold increase of electron transport chain activity. Thus, we have to ask whether a 2-fold increase of mitochondrial content in skeletal muscle tissue might be able to cause a 13% elevation of total body RMR. Muscle contributes to total body RMR to about 30% (13– 42%) (35) in rats, and about 30% of it is due to the futile cycling of protons, the proton leak (36), which thus accounts for about 10% of RMR. Therefore, increasing mitochondrial mass in muscle by a factor of two, without changing coupling of electron flow to ATP-synthesis, could lead to a rise of total body RMR by about 20%. In conclusion, high levels of circulating glucocorticoid hormones alone, without the contribution of hyperlipedemia and hyperinsulinemia, are sufficient to increase mitochondrial mass; however, only in skeletal muscle cells, and this alone might explain the increased RMR observed in humans. The molecular mechanisms have to be studied in more detail because no glucocorticoid responsive elements are found in important nuclear genes encoding mitochondrial key proteins. Lack of stimulation of mtTFA expression, in the presence of stimulated transcription of mtDNA, may indicate the direct involvement of the mitochondrial glucocorticoid receptor. If stimulation of this pathway alone would expand


mitochondrial functional equivalents, this would imply that an increase of mtDNA transcription is sufficient to stimulate overall mitochondrial biogenesis, which has not been shown under physiological conditions so far. Acknowledgments The gift of antimouse-mtTFA antiserum and plasmid encoding mouse mtTFA by Dr. David Clayton, Stanford, is gratefully acknowledged. The expert help of Steffi Goffart with promoter analysis using MatInspector software is appreciated. Received July 10, 2001. Accepted October 1, 2001. Address all correspondence and requests for reprints to: Katharina Weber, Ph.D., Department of Vegetative Physiology, University of Ko¨ ln, Robert-Koch-Strasse 39, 50931 Ko¨ ln, Germany. E-mail: katharina. [email protected]. This work was supported by Deutsche Forschungsgemeinschaft (DFG Wi 889/3–2) and by Stiftung VERUM, Verhalten und Umwelt.

References 1. Brillon DJ, Zheng B, Campbell RG, Matthews DE 1995 Effect of cortisol on energy expenditure and amino acid metabolism in humans. Am J Physiol 268:E501–E513 2. Tataranni PA, Larson DE, Snitker S, Young JB, Flatt JP, Ravussin E 1996 Effects of glucocorticoids on energy metabolism and food intake in humans. Am J Physiol 271:E317–E325 3. Stuart JA, Brindle KM, Harper JA, Brand MD 1999 Mitochondrial proton leak and the uncoupling proteins. J Bioenerg Biomembr 31:517–525 4. Kadowaki T, Kitagawa Y 1988 Enhanced transcription of mitochondrial genes after growth stimulation and glucocorticoid treatment of Reuber hepatoma H-35. FEBS Lett 233:51–56 5. VanItallie CM 1992 Dexamethasone treatment increases mitochondrial RNA synthesis in a rat hepatoma cell line. Endocrinology 130:567–576 6. Rachamim N, Latter H, Malinin N, Asher C, Wald H, Garty H 1995 Dexamethasone enhances expression of mitochondrial oxidative phosphorylation genes in rat distal colon. Am J Physiol 269:C1305–C1310 7. Samuels HH, Stanley F, Shapiro LE 1979 Control of growth hormone synthesis in cultured GH1 cells by 3,5,3⬘-triiodo-l-thyronine and glucocorticoid agonists and antagonists: studies on the independent and synergistic regulation of the growth hormone response. Biochemistry 18:715–721 8. Kupper JH, Muller M, Jacobson MK, Tatsumi-Miyajima J, Coyle DL, Jacobson EL, Burkle A 1995 Trans-dominant inhibition of poly(ADP-ribosyl)ation sensitizes cells against gamma-irradiation and N-methyl-N⬘-nitro-Nnitrosoguanidine but does not limit DNA replication of a polyomavirus replicon. Mol Cell Biol 15:3154 –3163 9. Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248 –254 10. Schneider WC 1957 Determination of nucleic acids by pentose analysis. Methods Enzymol 3:680 – 684 11. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162: 156 –159 12. Sambrook J, Fritsch EF, Maniatis T 1989 Molecular cloning: A laboratory manual, ed 2. New York: Cold Spring Laboratory Press 13. Wiesner RJ, Kurowski TT, Zak R 1992 Regulation by thyroid hormone of nuclear and mitochondrial genes encoding subunits of cytochrome-c oxidase in rat liver and skeletal muscle. Mol Endocrinol 6:1458 –1467 14. Larsson NG, Oldfors A, Holme E, Clayton DA 1994 Low levels of mitochondrial transcription factor A in mitochondrial DNA depletion. Biochem Biophys Res Commun 200:1374 –1381 15. Larsson NG, Garman JD, Oldfors A, Barsh GS, Clayton DA 1996 A single mouse gene encodes the mitochondrial transcription factor A and a testisspecific nuclear HMG-box protein. Nat Genet 13:296 –302 16. Cooperstein SJ, Lazarow A 1951 A microspectrophotometric method for the determination of cytochrome c oxidase. J Biol Chem 189:665– 670 17. Djouadi F, Bastin J, Gilbert T, Rotig A, Rustin P, Merlet-Benichou C 1994 Mitochondrial biogenesis and development of respiratory chain enzymes in kidney cells: role of glucocorticoids. Am J Physiol 267:C245–C254 18. Djouadi F, Bastin J, Kelly DP, Merlet-Benichou C 1996 Transcriptional regulation by glucocorticoids of mitochondrial oxidative enzyme genes in the developing rat kidney. Biochem J 315(Pt 2):555–562 19. Hood DA 1990 Co-ordinate expression of cytochrome c oxidase subunit III and VIc mRNAs in rat tissues. Biochem J 269:503–506 20. Gagnon J, Kurowski TT, Wiesner RJ, Zak R 1991 Correlations between a nuclear and a mitochondrial mRNA of cytochrome c oxidase subunits, enzy-

184


matic activity and total mRNA content, in rat tissues. Mol Cell Biochem 107:21–29 21. Wiesner RJ 1997 Adaptation of mitochondrial gene expression to changing cellular energy demands. News Physiol Sci 12:178 –184 21a.Garstka HL 1992 Diploma Thesis, University of Heidelberg 22. Quandt K, Frech K, Karas H, Wingender E, Werner T 1995 MatInd and MatInspector: new fast and versatile tools for detection of consensus matches in nucleotide sequence data. Nucleic Acids Res 23:4878 – 4884 23. Larsson NG, Barsh GS, Clayton DA 1997 Structure and chromosomal localization of the mouse mitochondrial transcription factor A gene (Tfam). Mamm Genome 8:139 –140 24. Scheller K, Sekeris CE, Krohne G, Hock R, Hansen IA, Scheer U 2000 Localization of glucocorticoid hormone receptors in mitochondria of human cells. Eur J Cell Biol 79:299 –307 25. Demonacos C, Djordjevic MR, Tsawdaroglou N, Sekeris CE 1995 The mitochondrion as a primary site of action of glucocortocoids: the interaction of the glucocorticoid receptor with mitochondrial DNA sequences showing partial similarity to the nuclear glucocorticoid respnsive elements. J Steroid Biochem Mol Biol 55:43–55 26. Casas F, Rochard P, Rodier A, Cassar-Malek I, Marchal-Victorion S, Wiesner RJ, Cabello G, Wrutniak C 1999 A variant form of the nuclear triiodothyronine receptor c-ErbA␣1 plays a direct role in regulation of mitochondrial RNA synthesis. Mol Cell Biol 19:7913–7924 27. Enriquez JA, Fernandez SP, Garrido PN, Montoya J 1999 Direct regulation of mitochondrial RNA synthesis by thyroid hormone. Mol Cell Biol 19:657– 670 28. Pecqueur C, Couplan E, Bouillaud F, Ricquier D 2001 Genetic and physiological analysis of the role of uncoupling proteins in human energy homeostasis. J Mol Med 79:48 –56 29. Barbe P, Larrouy D, Boulanger C, Chevillotte E, Viguerie N, Thalamas C,


30.

31.

32.

33.

34.

35.

36.

Trastoy MO, Roques M, Vidal H, Langin D 2000 Triiodothyronine-mediated upregulation of UCP2 and UCP3 mRNA expression in human skeletal muscle without coordinated induction of mitochondrial respiratory chain genes. FASEB J 15:13–15 Clapham JC, Arch JR, Chapman H, Haynes A, Lister C, Moore GB, Piercy V, Carter SA, Lehner I, Smith SA, Beeley LJ, Godden RJ, Herrity N, Skehel M, Changani KK, Hockings PD, Reid DG, Squires SM, Hatcher J, Trail B, Latcham J, Rastan S, Harper AJ, Cadenas S, Buckingham JA, Brand MD 2000 Mice overexpressing human uncoupling protein-3 in skeletal muscle are hyperphagic and lean. Nature 406:415– 418 Gong DW, He Y, Karas M, Reitman M 1997 Uncoupling protein-3 is a mediator of thermogenesis regulated by thyroid hormone, ␤3-adrenergic agonists, and leptin. J Biol Chem 272:24129 –24132 Cabrero A, Alegret M, Sanchez RM, Adzet T, Laguna JC, Vazquez M 2000 Down-regulation of uncoupling protein-3 and -2 by thiazolidinediones in C2C12 myotubes. FEBS Lett 484:37– 42 Pedersen SB, Kristensen K, Bruun JM, Flyvbjerg A, Vinter-Jensen L, Richelsen B 2000 Systemic administration of epidermal growth factor increases UCP3 mRNA levels in skeletal muscle and adipose tissue in rats. Biochem Biophys Res Commun 279:914 –919 Dallman MF, Akana SF, Bhatnagar S, Bell ME, Strack AM 2000 Bottomed out: metabolic significance of the circadian trough in glucocorticoid concentrations. Int J Obes Relat Metab Disord 24(Suppl 2):S40 –S46 Rolfe DF, Newman JM, Buckingham JA, Clark MG, Brand MD 1999 Contribution of mitochondrial proton leak to respiration rate in working skeletal muscle and liver and to SMR. Am J Physiol 276:C692–C699 Rolfe DF, Brown GC 1997 Cellular energy utilization and molecular origin of standard metabolic rate in mammals. Physiol Rev 77:731–758