THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists, Inc.

Vol. 261, NO. 1. Issue of January 5, pp. 376-380,1986 Printed In U.S.A.

Regulation of Nuclear and Mitochondrial Gene Expression by Contractile Activity in Skeletal Muscle*

(Received for publication, June 17, 1985)

R. Sanders Williams#& Stanley Salmonsll, Eric A. Newsholmell, Russel E. KaufmanS, and Jane Mellor 11 From the $Department of Medicine, Duke University Medical Center, Durham, North Carolina 27710, the TDepartment of Anatomy, University of Birmingham Medical School, Birmingham B15 2TJ, United Kingdom, and the 11 Department of Biochemistry, University of Oxford, Oxford OX1 3QU, United Kingdom

Increased contractile activity of skeletal muscle aug- ments the volume fraction and enzymatic capacity of mitochondria and suppresses the enzymatic capacity of several cytoplasmic enzymes of glycolysis. To examine the biochemical mechanisms underlying these effects, we measured the concentrations of cytochrome b mRNA and aldolase A mRNA in tibialis anterior mus- cles of adult rabbits that had been stimulated via the motor nerve to contract continuously at 10 Hz for 5 or 21 days; these were compared with the corresponding levels in the unstimulated limbs of the same animals. After 2 1 days of stimulation aldolase mRNA had fallen to one-fourth of control levels, while cytochrome b mRNA had increased by &fold. A reduction in aldolase mRNA was already evident after only 5 days of stim- ulation, whereas the level of cytochrome b mRNA was not elevated at this stage. Mitochondrial DNA was unchanged after 5 days but had increased by 4-fold after 21 days.

We conclude that contractile activity in skeletal mus- cle produces reciprocal changes in the expression of these two genes at a transcriptional level but via dif- ferent regulatory mechanisms. Enhancement of the expression of the mitochondrial cytochrome b gene appears to be proportional to the increase in its copy number and may not, therefore, depend upon changes in transcriptional or translational efficiency. The re- duction in aldolase A mRNA occurs at an earlier stage in the response to contractile activity and is probably mediated by a reduced transcriptional efficiency.

The biochemical, morphological, and physiological charac- teristics of mammalian skeletal muscle are not genetically fixed but are capable of profound modification in adult ani- mals or humans according to the pattern of contractile activity performed (1-3). Within 3-6 weeks of an increase in contrac- tile activity produced by as little as 30 min of daily exercise such as cycling or running, the capacity of the exercised muscle for producing energy for contractile work by oxidative metabolism is increased, in association with an increase in the mass and volume of mitochondria relative to other cellular constituents and a corresponding increase in the enzymatic

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

5 Supported by National Institutes of Health Grant 1FO6-TW- 00963 (Fogarty Senior International Fellowship) and by the PepsiCo Foundation. Established Investigator of the American Heart Associ- ation. To whom correspondence should be addressed.

capacities of mitochondrial enzymes of the tricarboxylic acid cycle, of fatty acid metabolism, and of the electron transport chain (1-3). This response is accompanied by a reduction in the enzymatic capacities of several cytoplasmic enzymes of the glycolytic pathway (1-3). In experimental animals chronic electrical stimulation of specific muscles via the motor nerve for 12-24 h daily has been used to produce a more radical increase in contractile activity than can be achieved by exer- cise. Such interventions produce effects on mitochondrial and cytoplasmic enzymatic capacities that are qualitatively similar to those produced by exercise, but these changes are produced more rapidly and are of a greater magnitude (2, 4-6).

The adaptive response of mitochondrial and cytoplasmic enzymes to contractile activity in skeletal muscle is of partic- ular interest from the standpoint of gene expression for two reasons: first, this response requires a coordinated regulation of the expression of a very large number of genes encoding the many specific proteins that participate in the process; second, it calls for a coordinated regulation of genes located both within nuclear chromatin and within mitochondrial DNA (7).

So far as the mitochondrial genome is concerned, there are several potential mechanisms that, alone or in combination, could account for changes in the expression of mitochondrial genes. These fall into three categories that may be distin- guished experimentally: changes in the copy number of mi- tochondrial genes; changes in transcriptional efficiency or RNA processing; and changes in translational efficiency or protein stability. In the first case, the rate of mitochondrial DNA replication could vary independently of nuclear DNA replication, producing changes in the number of copies of mitochondrial genes per unit of mitochondrial mass or per unit of muscle mass. Should this happen, altered expression of mitochondrial genes could occur without changes in tran- scriptional or translational efficiency. On the other hand, changes in mitochondrial mass and in the concentrations of mitochondrial gene products could occur without alterations in the cellular concentration of mitochondrial DNA. In this event, altered expression of mitochondrial genes could be brought about by changes in transcriptional efficiency or RNA processing (when there would be changes in the concentra- tions of RNA transcripts of mitochondrial genes) or by changes in translational efficiency or protein stability (when cellular concentrations of mitochondrial RNA would be un- altered). One objective of this study was to determine which of these three general mechanisms is utilized to regulate expression of mitochondrial genes in mammalian muscle dur- ing activity-induced changes in mitochondrial mass.

A second purpose was to compare the major mechanisms governing expression of mitochondrial genes in response to

376

contractile activity to those governing expression of nuclear genes, using for this part of the investigation a protein whose expression is reduced under conditions of increased contrac- tile activity.

EXPERIMENTAL PROCEDURES

Electrical Stimulation of Skeletal Muscle-Adult New Zealand White rabbits weighing 2.9-3.1 kg were anesthetized by intramuscular injection of 0.3 ml/kg fentanyl/fluanisone (Hypnorm; Crown Chem- ical Co.), following an overnight fast and following premedication with 3 mg/kg atropine sulfate and 5 mg/kg. diazepam. Under sterile conditions, a miniature electrical stimulator was inserted into the peritoneal cavity via an incision on the trunk, and electrodes were tunneled subcutaneously to a second incision in one hind' limb and sutured close to the common peroneal nerve. The incisions were closed in layers and the animals returned to individual cages with access to food and water ad libitum. Each stimulator was programmed to deliver 0.2-ms rectangular wave pulses a t 10 Hz. Pulsating con- tractions of the anterior compartment muscles innervated by the common peroneal nerve were palpable, producing oscillations of the foot when the leg was suspended, but the animals undergoing stimu- lation exhibited no signs of distress, they gained weight normally, and showed normal eating, sleeping, and locomotor behavior.

In 3 animals, electrical stimulation was initiated 5 days after implantation and was maintained for 5 days. The animals were then anesthetized with 0.5 ml/kg of a 25% solution of urethane adminis- tered intravenously, fotiowed by sodium pentobarbital to effect, i.e. until withdrawal reflexes were abolished. The hind limbs were dis- sected, and the tibialis anterior, extensor digitorum longus, and soleus muscles of both hind limbs either used immediately for biochemical studies or frozen in liquid nitrogen and stored at -70°C. In 4 addi- tional animals, electrical stimulation was maintained for 21 days. In one further animal, the surgical insertion of a pulse generator and stimulating electrodes was performed as previously described, but the pulse generator was not activated, allowing the animal to serve as a sham-operated controf.

Enzyme Assays-Citrate synthase (EC 4.1.3.7) activity was quan- titated by the method of Srere (8) and aldolase (EC 4.1.2.13) by the method of Bergmeyer (9) in unfractionated muscle homogenates under conditions of substrate and cofactor excess. Muscle slices of approximately 100 mg were prepared from a cross-section of the muscle and were stored at -70 "C. Immediately prior to enzyme analysis they were thawed in 10 volumes of 25 mM Tris, pH 7.4, 1 mM EDTA, and homogenized on ice with three 10-s bursts of a mechanical homogenizer ~Brinkmann Polytron). These homogenates were diluted 10-fold in the same buffer, and 2-10 pl of dilute homog- enate containing 3-25 f ig of muscle protein were used in each 2-mI assay reaction.

E x t r ~ ~ ~ o n of RNA from Muscle Specimeras-All manipulations of tissue and extracts were performed under sterile conditions taking care to avoid contamination with RNase. Acid-washed and baked glassware were used throughout. Muscle specimens from 1.1 to 2.6 g, wet weight, were used immediately after dissection or thawed after storage at -70 "C in 5 volumes of 5 M guanidinium isothiocyanate containing 50 mM Tris, pH 7.6, 10 mM EDTA, 2% sodium lauryl sarkosinate (Sarkosyl), and 1 % 4-mercaptoethanoi (10). The tissues were minced with scalpel blades and homogenized with three 10-s bursts of a mechanical homogenizer (Brinkmann Polytron). An equal volume of phenol preheated to 60 'C was added, and the mixture passed 10 times through a no. 19 gauge needle attached to a sterile syringe. After vigorous shaking at 60 "C for 5 min, 0.8 mI/g tissue of 50 mM 'his, pH 7.4,140 mM NaCl, 5 mM EDTA were added together with 24:l chloroform/isoamyl alcohol in a volume equal to the volume of phenol. The mixtures were shaken for a further 5 min at 60 "C, then chifled on ice, and centrifuged at 4 "C for 20 min at 13,000 x g. The aqueous phases were removed, re-extracted twice more with ~henoi/chloroform, and then twice with chloroform alone. The final extract was mixed with 1:50 volume of 5 M NaCl and 2.5 volumes of absolute ethanol and held at -70 "C for 2-16 h. The ethanol precip- itates were collected by centrifugation for 20 min at 9,000 X g and -10 "C, washed once with 70% ethanol, and dried under a vacuum. Pellets were resuspended in 0.4 ml of 50 mM Tris, pH 7.4, I40 mM NaCI, 5 mM EDTA, and sodium dodecyl sulfate and proteinase K iGibco-Bethesda Research Laboratories) were added to final concen- trations of 0.1% and 200 pg/ml, respectively. After incubation at 37 "C for 1 h, the samples were extracted with phenol/chloroform

and precipitated with ethanol as before. The ethanol precipitates were washed twice with 70% ethanol, dried under a vacuum, and resus- pended in water. An aliquot was stored at -70 "C, and the remaining material was purified further by digestion for 30 min at 37 "C with 20 pg/mk DNase I (Gibco-Bethesda Research Laboratories; RNase free) in a reaction mixture including 10 mM Tris, pH 8.0, 10 mM MgC12, 5 mM dithiothreitol, and 1,000 units/ml RNasin (P & S Biochemicals), The post-DNase samples were again digested with proteinase K, extracted 3 times with phenol/chloroform, and precip- itated twice with 2.5 volumes of ethanol at -70 'C for 2-16 h. The final pellets were collected by centrifugation, washed with 70% ethanol, vacuum dried, dissolved in water at a concentration of 4-8 gg/gl, and stored at -70 "C.

Polyadenylated RNA was prepared by affinity chromatography over 1-ml poly(U)Sepharose 4B columns (11). Columns were equili- brated with 200 mM Tris, pH 7.4, 500 mM NaCl, 10 mM EDTA, the RNA samples loaded, and the columns washed with 8 column volumes of the same buffer. Poly(A+) RNA was eluted with water, the fractions pooled, ethanol precipitated, resuspended in H20, and stored at -70 "C.

~ x t r a c ~ i o n of DNA from Muscte ~pccime~-Muscle specimens prepared as cross-sectional slices from each muscle and weighing 0.2- 0.4 g were thawed in 5 volumes of 10 mM Tris, pH 7.4,50 mM EDTA and homogenized on ice with three IO-s bursts of a motor-driven homogenizer ~Brinkmann Polytron). The homogenates were incu- bated in a shaking water bath for 16 h at 55 "C in the presence of 0.5% Sarkosyl and 100 ggJml proteinase K. The resulting digests were centrifuged at 1,800 X g for 10 min and the small pellets discarded. After three phenol/chloroform extractions the aqueous phases were mixed and precipitated twice with 150 volume of 5 M NaCl and 2 volumes of absolute ethanol for 2-16 h at -70 "C. The precipitates were collected by centrifugation at 9,000 x g for 20 min at -10 "C, vacuum dried, and finally resuspended in 10 mM Tris, pH 7.4, 10 mM NaC1, 1 mM EDTA at DNA concentrations ranging from 1-3 gg/gI. In early experiments this preparation was subjected to further purification by digestion with 100 pg/ml RNase (DNase free, Sigma) for 2 h at 37 "C followed by digestion with proteinase K, phenol/chloroform extraction, and ethanol precipitation, but because these additional steps had no influence on quantitative hybridization of the mitochondrial probe they were omitted subsequently.

Hybridization Procedures for RNA-RNA that had been extracted from the muscle samples was bound to O.l-pm nitrocellulose paper (Schleicher and Schuell) either by transfer from agarose-formal~e- hyde gels in the presence of 0.3 M sodium citrate, pH 7.0, 3 M NaCI, or by direct blotting in a vacuum filtration manifold (Schleicher and Schuell dot blot). Filters were baked in a vacuum oven at 80 "C for 4 h, boiled for 5 min in 10 m M EDTA, pH 8.0, and hybr id i~d for 24- 36 h at 42 "C in 40% fonnamide, 30 mM sodium citrate, pH 7.0, 0.3 M NaCI, 0.1% sodium dodecyl sulfate, 50 mM sodium phosphate, 0.2% Ficoll, 0.2% polyvinylpyrrolidone, 0.2% bovine serum albumin (Frac- tion V), and 200 pg/ml sonicated salmon sperm DNA in the presence of the appropriate DNA probe, nick translated (Bethesda Research Laboratories nick translation kit) with P2P]dTTP (Amersham Carp.) to a specific activity of 1 X 10' dpmlpg. Transcription products of mitochondrial DNA were detected by the binding of pMM26, an 8.2- kb' BamHI fragment of mouse mitochondrial DNA that includes the coding region of the cytochrome b gene, as well as the regions encoding the 12 and 16 S ribosomal subunits (12) (gift of Dr. Ian Craig, Oxford, United Kingdom). Aldolase A mRNA was detected by the binding of pRM223, which contains aldolase A cDNA prepared from rabbit skeletal muscle (13) (gift of Dr. Dean Tolan, Berkeley, CAI.

Binding of these probes to specific bands identified in Northern blots was quantified by integration of optical densitometric scanning (Vitatron) of autoradiographs. Binding to RNA dot blots was quan- tified either by optical densitometric scanning of autoradiographs or by liquid scintillation counting of each dot in a toluene-based fluor. In either case, estimates of RNA concentration were linear over a 10- fold range in the amount of RNA bound to the filters.

There was no systematic difference between the yield of total RNA extracted from stimulated muscle (mean 170 gg/g) than from unstim- ulated control muscle (mean 196 gg/g) as estimated from spectropho- tometric absorbance. The reproducibility of the RNA extraction method was further assessed by dividing both a stimulated and an unstimulated muscle into equal halves and processing each half as a separate preparation. Quantitative binding of both pRM223 and

- ' The abbreviation used is: kb, kilobase pair(s).

378 Regulation of Gene Expression in Skeletal Muscle

pMM26 differed by no more than 6% between the two halves of the same muscle, whether stimulated or unstimulated.

There remained the possibility that some feature of stimulated muscle other than a true variation in the cellular concentrations of a specific RNA species could affect the efficiency of RNA recovery, introducing errors. In one experiment, therefore, we added a known quantity of foreign RNA (Saccharomyces cereuisiae) at the time of the initial homogenization of samples from a stimulated muscle and its unstimulated contralateral control. When RNA dot blots derived from these preparations were hybridized to a 1.25-kb PuuII/BglII fragment from the ORFl region of the yeast Tyl-15 element (14) with no detectable homology to mammalian RNA, recovery from stimu- lated muscle was within 4% of that from unstimulated control muscle.

Hybridization Procedures for DNA-DNA extracted from muscle samples was bound to 0.45-~m nitrocellulose paper either by transfer from agarose gels or by direct blotting as described above for RNA. The subsequent hybridization methods were also identical to those described above. No attempt was made physically to separate mito- chondria prior to preparation of mitochondrial DNA, since pilot experiments revealed that pMM26 hybridized specifically to a 17-kb band in extracts of total DNA and could, therefore, be used to quantitate mitochondrial DNA, avoiding the considerable difficulties associat.ed with purification of intact mitochondria from skeletal muscle.

RESULTS

Enzymatic Analyses-The morphological, histochemical, and biochemical responses to chronic electrical stimulation have been extensively characterized in previous studies (4-6, 15-17). After 5 days only minimal changes in mitochondria~ and cytoplasmic enzyme capacities have been observed, but, after 21 days there are 4-8-fold increases in mitochondrial volume and mass and 3-10-fold increases in V,, of mito- chondrial enzymes of the tricarboxylic acid cycle (e.g. citrate synthase, succinate dehydrogenase, and malate dehydrogen- ase), of fatty acid oxidation (e.g. carnitine acetyl transferase and ,!%hydroxyacyl-CoA dehydrogenase), and of the electron transport chain (e.g. cytochrome oxidase). Accompanying these elevations of mitochondrial enzymatic capacities, the activities of cytoplasmic enzymes of glycolysis (e.g. aldolase and phosphof~ctokinase) have been observed after 21 days of stimulation to be reduced to 20-50% of control levels.

Enzymatic analysis of citrate synthase as a mitochondrial marker enzyme and of aldolase as a cytoplasmic enzyme of glycolysis enabled us to confirm these effects in the muscles stimulated for these studies. As Fig. 1 shows, we could discern no consistent effect upon the V,,, of citrate synthase after 5 days of stimulation, but a striking &fold increase took place after 21 days. The V,,, of aldolase declined slightly after 5 days to 82% of control but was-markedly suppressed to 24% of control by 21 days. These effects occurred consistently in all of the animals studied. Enzyme activities in muscle from the sham-operated limb were indistinguishable from those present in unstimulated control muscles.

RNA Concentrations-Hybridization of the mitochondrial DNA probe pMM26 to RNA extracted from muscle stimu- lated for 21 days was markedly greater both in RNA dot blots (Fig. Z} and in Northern blots (Fig. 3); the increase in mito- chondrial RNA was about &fold in comparison to the concen- tration measured in the unstimulated contralateral limbs of the same animals, but no change was observed in the elec- trode-containing limb of the sham-operated animal.

This increase in mitochondrial RNA involved both mRNA and rRNA, since the relative proportions of 16 S rRNA, 12 s rRNA, and cytochrome b mRNA in stimulated muscle re- mained equivalent to those observed in control muscle (Fig. 3). Furthermore, the increase in polyadenylated RNA from stimulated muscle that hybridized to pMM26 was also 5-fold,

a - 5 days

7-

6- Q \

5-

4-

3-

0""" CS mtRNA mt UNA ALDO A L U O ~ R ~

21 days 8 - f 7-

6 -

5 -

I+-

3 -

2 -

1 -

FIG. 1. Effects of 5 or 21 days of electrical stimulation of contractile activity upon citrate synthase enzymatic capacity (CS), mitochondrial RNA (mtRNA), mitochondrial DNA (mtDNA), aldolase enzymatic capacity (ALDO), and aldolase A mRNA (ALDO mRNA) in rabbit skeletal muscles. Open symbols represent results from unstimulated control muscles, and closed symbots represent results from stimulated muscles. Circles represent data from tibialis anterior muscles, and squares represent the findings from extensor digitorum longus muscles. Lines connect data from muscles from the two limbs of an individual animal. Units on the ordinate represent multiples of the following constants: 0.04 FmoI/min/+g protein (citrate synthase); optical density integrated units per g of muscle, wet weight, multiples of lowest 21-day value from autoradiographs of RNA dot blots hybridized to pMM26 (mtRNA) or pRM223 (aldolase A mRNA) labeled with [32P]; cpm per g of muscle, wet weight, multiples of lowest 21-day value from liquid scintillation counting of DNA dot blots hybridized to pMM26 labeled with "P (mtDNA); and 0.3 +mol/min/Fg of protein (aldolase A).

approximating the increase in unfractionated mitochondrial RNA.

Conversely, muscle stimulated for 21 days contained mark- edly lower concentrations of aldolase A mRNA (Fig. l), whether determined by examination of Northern blots (Fig. 4) or RNA dot blots (Fig. 2). Quantitatively, aldolase mRNA decreased by about 70% compared to that present in the unstimulated limb. Furthermore, a decrease in the level of aldolase mRNA was already present after 5 days of stimula- tion, when there was no increase in mitochondrial RNA.

An additional observation based on analysis of Northern blots was that the primary RNA species hybridizing to pRM223 in RNA extracted from muscle stimulated for 22 days was not only present at a lower concentration but mi- grated more slowly on formaldehyde-agarose gels, suggesting an increase in its molecular size of about 50 bases (Fig. 4). This apparent qualitative difference in aldolase mRNA from stimulated muscle was not evident after 5 days.

Mitochondrial DNA Concentrations-Hybridization of the mitochondrial probe pMM26 to DNA extracted from muscle stimulated for 21 days indicated a 4-fold increase in the concentration of mitochondrial DNA relative to control tissue (Figs. 1 and 5), a change similar to the increase in mitochon- drial RNA. This effect was not evident after 5 days of stimu-

Regulation of Gene Expression in Skeletal Muscle 379

pMM 26 pRM 223 ' 10 5 1 10P ' 10 5 1 IOP

FIG. 2. Dot blots of RNA extracted from muscles of the electrically stimulated (S) or the unstimulated contralateral limb ( C ) of adult rabbits after 21 days of stimulation. The extensor digitorum longus muscle (EDL) is made up predominantly of fast-twitch glycolytic fibers in the control state and is stimulated to contract in this experimental model. The soleus muscle (SOL) is made up predominantly of slow-twitch oxidative fibers and is not innervated by the motor nerve that is stimulated in this model. The numerals refer to the amount of RNA (pg) applied to the nitrocellu- lose paper and the letter P refers to samples that were digested with DNase I, as described in the text. RNA dot blots were hybridized to either pMM26 or pRM223 as indicated. The autoradiographs shown were exposed 24 h to Kodak XAR-5 film with two intensifying screens. Five-fold increases in binding of the mitochondrial probe (pMM26) and five-fold decreases in the binding of the aldolase probe (pRM223) are evident in blots of RNA from stimulated muscle.

1.8-

1.0-

C c s

-1.0

FIG. 3. Northern blots of muscle RNA hybridized to pMM26. The leftmost lane from a 1.2% formaldehyde-agarose gel was loaded with 20 pg of RNA extracted from an unstimulated control (C) tibialis anterior muscle, and the autoradiograph was exposed 48 h. The rightward two lanes from a different 1.0% formaldehyde- agarose gel were loaded with 10 pg of RNA extracted from the unstimulated control (C) and 21-day stimulated ( S ) tibialis anterior muscles of the same animal and were exposed for 24 h. The size markers (kb) indicate the position of yeast Ty element RNA from parallel lanes of each gel. The 0.9- and 1.6-kb bands represent the 12 and 16 S mitochondrial ribosomal RNA transcripts, respectively, and the 1.2-kb band represents cytochrome b mRNA.

lation (Fig. 1). The change in DNA concentration was re- stricted to mitochondrial DNA, since there were no changes in the binding of pRM223 (aldolase probe) to DNA extracted from stimulated muscle either after 5 or after 21 days of stimulation.

c c c s c s s c s c

FIG. 4. Northern blots of muscle RNA hybridized to pRM223. Lanes represented in the leftward panel were loaded with 10 pg of RNA (lane I ) or 20 pg of RNA (lanes 2-6) extracted either from unstimulated control muscles (C) or from muscles stimulated electrically for 21 days (S), and the autoradiograph was exposed for 24 h. In the rightward panel, the leftmost lane was loaded with 20 pg of RNA, and the remaining lanes with 10 pg of RNA. The autoradi- ograph was exposed for 7 days.

e... 0 20

5

1

FIG. 5. Dot blots of DNA extracted from unstimulated con- trol muscles ( C ) or from muscles stimulated electrically for 21 days (S) and hybridized to pMM26. The two samples labeled C, represent DNA extracted from the two limbs of a sham-operated control animal. The numbers on the ordinate refer to the amounts of DNA applied to the filter in each pair of rows. Four-fold increases in the binding of pMM26 to DNA from stimulated muscles are evident.

DISCUSSION

Our findings support several principal conclusions. First, electrical stimulation of skeletal muscle of the rabbit increases the level of cytochrome b mRNA but decreases that of aldolase A. This finding suggests that the rate of transcription of the cytochrome b gene is increased, whereas that of the aldolase A gene is decreased by contractile activity. Furthermore, since changes in the concentrations of the respective mRNAs after 21 days of stimulation are almost identical to the changes in the maximum catalytic activities of citrate synthase and al- dolase (which we assume are proportional to the concentra- tions of cytochrome b and aldolase, respectively), the findings suggest that there is no change in the efficiency of translation of these proteins. Our findings do not exclude the possibility that contractile activity alters RNA processing or stability, translational efficiency, post-translational processing of apo- proteins, or the stability of the mature proteins encoded by these genes. Indeed, effects upon translational mechanisms and rates of protein degradation may be important in the acute responses of muscle to brief periods of contractile activ- ity (18). However, our data suggest that in the long term, transcriptional regulation alone is sufficient to account for the observed changes in expression of these two genes.

Second, the increase in levels of RNA transcripts of mito- chondrial genes is similar to that of mitochondrial DNA. This suggests that the increase in mitochondrial DNA, and hence the copy number of mitochondrial genes, is responsible for the increase in mitochondrial transcripts and that regulation

380 Regulation of Gene Expression in Skeletal Muscle

of transcriptional efficiency may not be required to enhance the expression of mitochondrial genes in response to contrac- tile activity.

Third, it has been observed previously that the relative concentrations of mRNA and rRNA within mitochondria can differ markedly among different cell lines (19-21). These previous findings suggest that mitochondrial gene expression could potentially be regulated by factors that alter the relative rates of transcription of individual mitochondrial genes. How- ever, the fact that cytochrome b mRNA and mitochondrial ribosomal RNA increased in parallel in response to contractile activity in this study suggests that this mechanism is not utilized by mammalian skeletal muscle during a potent stim- ulus to enhancement of mitochondrial gene expression.

Fourth, our results suggest that there are differences be- tween the mechanisms responsible for altered expression of nuclear and mitochondrial genes induced by changes in con- tractile activity in skeletal muscle. There was a distinct tem- poral difference between the onset of changes in aldolase mRNA concentrations, which were readily discernible after only 5 days of stimulation, and cytochrome b mRNA concen- trations, which were not increased at this early time point. In addition, the copy number of mitochondrial genes was in- creased, whereas a change in transcriptional efficiency not involving a major change in copy number was the more likely explanation for activity-induced suppression of aldolase gene expression.

Finally, we present preliminary evidence of a change in apparent molecular size of aldolase A mRNA in muscle that had been electrically stimulated for 21 days. This finding, if confirmed by additional studies that are in progress, suggests that electrical stimulation may have produced an effect other than altering the rate of transcription of the aldolase A gene. For example, the aldolase A mRNA may have been tran- scribed from a different (and possibly less efficient) promoter, leading to transcription of a previously untranscribed region within the gene. Alternatively, electrical stimulation may have altered the processing of primary transcripts such that a previously deleted region was retained in the mature message. As a third possibility, this result may have occurred because of transcription from a different but homologous gene within the aldolase gene family.

Rabinowitz and colleagues (22) observed an increase in mitochondrial DNA in the hearts of rats subjected to aortic banding, indicating that functional overload may evoke a response within the mitochondrial genome of cardiac muscle similar to that we have observed in skeletal muscle. Clayton and colleagues (23) have subjected mouse lung macrophages to hypoxia in vitro, an intervention that produced a 50% decrease in the catalytic capacity of many of the same mito-

chondrial enzymes whose activities are augmented by con- tractile activity in skeletal muscle. It may seem contradictory that, in these cells, mitochondrial DNA concentrations were unchanged, for this would indicate that mitochondrial enzy- matic capacity can vary considerably without associated changes in mitochondrial DNA. However, the hypoxic stim- ulus to these cells was administered over only 4 days, and it should, therefore, be compared to our shorter term experi- ments in which stimulation of skeletal muscle for 5 days likewise produced no changes in mitochondrial DNA concen- trations. Although the experimental models are different, these findings both suggest that relatively long periods of time may be required to permit the mitochondrial genome to re- spond, even to potent regulatory stimuli, by changes in rates of DNA replication.

REFERENCES 1.

2.

3.

4. 5.

6. 7.

8. 9.

10.

11.

12.

13.

14.

15.

16. 17.

18. 19.

20. 21.

22.

23.

Holloszy, J . O., and Coyle, E. F. (1984) J. Appl. Physiol. 56,831- 838

Jolesz, F., and Sreter, F. A. (1981) Annu. Reu. Physiol. 43, 531- 552

Saltin, B., and Gollnick, P. D. (1983) in Handbook of Physiology: Skeletal Muscle (Peachey, L. D., ed) pp. 555-631, American Physiological Society, Bethesda, MD

Salmons, S., and Sreter, F. A. (1976) Nature 263, 30-34 Pette, D., Ramirez, B. U., Muller, W., Simon, R., Exner, G. U.,

and Hildebrand, R. (1975) Pflugers Arch. 361, 1-7 Salmons, S., and Henriksson, J. (1981) Muscle Nerve 4, 94-105 Bibb, M. J., Van Etten, R. A., Wright, C. T., Walberg, M. W.,

and Clayton, D. A. (1981) Cell 26,167-180 Srere, P. A. (1969) Methods Enzymol. 13,3-11 Bergmeyer, H. U. (1974) Methods of i h y m a t i c Analysis, pp. 430-

Feramisco, J. R., Smart, J . E., Burridge, K., Helfman, D. M., and

Holland, M. J., Hager, G . L., and Rutter, W. J. (1977) Biochem-

Kearsey, S. E., Flanagan, J. G., and Craig, I. W. (1980) Gene 12,

Tolan, D. R., Amsden, A. B., Putney, S. D., Urdea, M. S., and

Mellor, J., Fulton, S. M., Dobson, M. J., Wilson, W., Kingsman,

Eisenberg, B. R., and Salmons, S. (1981) Cell Tissue Res. 220,

Pette, D. (1984) Med. Sci. Sports 16, 517-528 Brown, W. E., Salmons, S., and Whalen, R. G. (1983) J . Biol.

Booth, F. W., and Watson, P. A. (1985) Fed. Proc. 44,2293-2300 Canatore, P., Gadaleta, M. N., and Saccone, C. (1984) Biochem.

Attardi, G. (1981) Trends Biochern. Sci. 6,86-89 Tabak, H. F., Grivell, L. A., and Borst, P. (1983) Crit. Reu.

Biochem. 14, 297-317 Rajamanickam, C., Merten, S., Kwiatkowska-Patzer, B., Chuang,

C., Zak, R., and Rabinowitz, M. (1979) Circ. Res. 45, 505-515 Murphy, B. J., Robin, E. D., Tapper, D. P., Wong, R. J., and

Clayton, D. A. (1984) Science 223,707-709

431, Academic Press, New York

Thomas, G . P. (1982) J. Biol. Chern. 257, 11024-11028

i s t ~ 16,8-16

249-255

Penhoet, E. E. (1984) J. Biol. Chem. 259, 1127-1131

S. M., and Kingsman, A. J. (1985) Nature 313, 243-246

449-471

Chem. 258, 14686-14692

Biophys. Res. Comrnun. 118,284-291