Ras is involved in nerve-activity-dependent regulation of muscle genes

articles

Ras is involved in nerve-activity-dependent regulation of muscle genes

Marta Murgia*†, Antonio L. Serrano*†, Elisa Calabria*, Giorgia Pallafacchina*, Terje Lømo‡ andStefano Schiaffino*§

*Department of Biomedical Sciences and CNR Center of Muscle Biology and Physiopathology, University of Padova, Viale G. Colombo 3, 35121 Padova, Italy‡Department of Physiology, University of Oslo, Oslo 0317, Norway

§e-mail: [email protected]†These authors contributed equally to this work

Gene expression in skeletal muscle is regulated by the firing pattern of motor neurons, but the signalling systems involved in excitation–transcription coupling are unknown. Here, using in vivo transfection in regenerating muscle, we show that constitutively active Ras and a Ras mutant that selectively activates the MAPK(ERK) pathway are able to mimic the effects of slow motor neurons on expression of myosin genes. Conversely, the effect of slow motor neurons is inhibited by a dominant-negative Ras mutant. MAPK(ERK) activity is increased by innervation and by low-frequency electrical stimulation. These results indicate that Ras–MAPK signalling is involved in promoting nerve-activity-dependent differentiation of slow muscle fibres in vivo.

he signalling pathways that convert motor neuron activityinto specific transcriptional programmes in the nuclei of mus-cle cells have not yet been identified. Previous studies have

mostly focused on the regulation of genes coding for the nicotinicacetylcholine receptor (AChR) subunits1. Nerve activity inducesdownregulation of extrajunctional AChR genes, and it has beensuggested that this process is controlled through the phosphoryla-tion of myogenin — a muscle-specific transcription factor that isalso regulated by electrical activity2 — by as-yet-unidentified pro-tein kinases3–5. Myogenin is more abundant in slow as compared tofast muscle fibres, and might play a part in the differentiation of theslow-fibre phenotype6,7. However, overexpression of myogenin infast muscles of transgenic mice has been shown to induce a switchto an oxidative metabolism but no increase in levels of slowmyosin8, the type of myosin protein characteristic of slow fibres.

Calcineurin has been implicated in the differentiation of theslow-fibre phenotype. Activated calcineurin stimulates the pro-moter/enhancer of the troponin-I-slow gene in cultured musclecells, and treatment with cyclosporin A, which blocks the cal-cineurin pathway, reduces the proportion of slow fibres in the ratsoleus muscle9. In addition, the proportion of slow fibres isincreased in the gastrocnemius muscle of transgenic mice express-ing a constitutively active form of calcineurin under the control ofthe promoter/enhancer of the muscle creatine kinase gene10. Thetranscription factor NF-AT, which is a major target of calcineurin,has been implicated as a downstream effector of calcineurin in skel-etal muscles9. However, it is controversial whether the slow-fibre-specific expression of the troponin-I-slow promoter/enhancer intransgenic mice is changed by deletion of the NF-AT-recognitionelement in the slow troponin enhancer11 (R.S. Williams, personalcommunication). On the other hand, calcineurin has been shownto cause a switch to a glycolytic metabolism, which is typical of fastand not slow muscle fibres, in cultured muscle cells12.

Ras may be involved in nerve-dependent regulation of musclegenes, as it has been shown that high-intensity physical exercise13–15

and short-term electrostimulation16,17 activate Ras-dependentmitogen-activated protein kinase (MAPK) pathways in rat andhuman skeletal muscle. However, it is not clear whether thesechanges are a prerequisite for transcriptional activation of musclegenes. To explore the role of Ras in activity-dependent muscle generegulation, we have used an in vivo model of muscle regeneration inwhich rapid changes in the relative expression of fast- and slow-muscle-type isoforms of myosin and other contractile proteins areinduced by innervation18. A specific advantage of the muscle-regen-

eration model is that efficient gene transfer can be obtained bydirect intramuscular DNA injection19. We therefore transfected dif-ferent Ras mutants into regenerating rat soleus muscle to establishwhether Ras is involved in promoting nerve-activity-dependent fibre-type differentiation.

T

Figure 1 MyHC-slow expression is induced by constitutively active RasV12 in denervated regenerating soleus muscle. a–c, Immunofluorescence analysis of MyHC-slow protein expression in cross-sections of a, denervated and c, innervated soleus transfected with empty vector, as compared with b, denervated soleus transfected with RasV12. Scale bar represents 100 µm. d, Number of MyHC-slow-positive fibres in denervated soleus muscles transfected with RasV12 or empty vector.

a b

c d

Num

ber

of s

low

fibr

es

200

150

100

50

0Ras Vector

© 2000 Macmillan Magazines Ltd142 NATURE CELL BIOLOGY | VOL 2 | MARCH 2000 | cellbio.nature.com

articles

ResultsActivated Ras induces the expression of slow MyHC. Intramuscu-lar injection of the myotoxic drug bupivacaine into the slow-twitchrat soleus muscle causes necrosis followed by rapid regeneration.Expression of the myosin heavy chain (MyHC) genes in regenerat-ing soleus muscle fibres is controlled by innervation20. In theabsence of the nerve, regenerating fibres express by default the fastMyHC-2X and -2B transcripts, whereas slow nerve activity inducesa MyHC switch, with upregulation of MyHC-slow and downregu-lation of the fast transcripts21. Newly formed muscle fibres areinnervated at day 3 after bupivacaine injection, and MyHC-slow isdetectable by immunofluorescence by day 5–6. The effect of inner-vation on MyHC gene expression can be reproduced by direct elec-trostimulation of denervated regenerating soleus with a low-frequency impulse pattern resembling the firing pattern of slowmotor neurons (S.S. et al., unpublished observations).

We investigated whether constitutively active Ras could mimicthe effect of slow nerve activity. We transfected denervated regener-ating soleus muscles either with empty plasmids or with plasmidsexpressing a constitutively active form of Ras, RasV12, 3 days aftermuscle injury, and assessed the presence of MyHC-slow protein 7days later. We did not detect MyHC-slow in denervated musclesafter transfection with empty vectors (Fig. 1a). In contrast, RasV12was able to induce the expression of MyHC-slow in numerousfibres in denervated muscles (Fig. 1b, d). Fibres expressing MyHC-slow were first detected by immunofluorescence on day 5 aftertransfection with RasV12 and were still abundant 1 month later(data not shown). The relatively long lag period before the induc-tion of MyHC-slow by Ras is in part due to the transfection proce-dure per se. In fact, injection of empty vectors into innervatedregenerating soleus also caused a delay in the timing of appearanceof MyHC-slow protein, which was detectable at day 7–8 after bupi-vacaine injury compared with day 5–6 in uninjected innervatedcontrols. This effect presumably results from delayed functional re-innervation because of transient damage of the newly formed neu-

romuscular junctions by the injection itself. However, by day 10most fibres in muscles injected with empty vectors were reactive forMyHC-slow, like uninjected controls (Fig. 1c).Dominant-negative Ras blocks the MyHC switch induced by slowmotor neurons. To assess the physiological relevance of Ras signal-ling in mediating the effects of innervation, we investigated whetherthe inhibition of endogenous Ras by a dominant-negative Rasmutant would prevent the activation of the MyHC-slow gene byslow motor neurons. Innervated muscle fibres overexpressing thedominant-negative RasN17 were unreactive for MyHC-slow (Fig.2a) and expressed the fast MyHC-2X transcript (Fig. 2b). This effectwas seen both at 1 week and at 1 month after transfection. Ras isthus required for MyHC gene switching from the default fast pro-gramme to the slow gene programme induced by slow motor neu-rons. To determine the role of Ras signalling in the transcriptionalregulation of the MyHC-slow gene, we investigated whetherRasN17 inhibits the expression of a construct consisting of the pro-moter of the MyHC-slow gene fused to a luciferase gene. Transfec-tion of RasN17 resulted in a threefold reduction of luciferaseactivity as compared with transfection with empty vector (Fig. 2c).The MyHC switch is induced by RasS35 but not RasG37 andRasC40. To define the downstream pathways mediating the effectsof Ras, we used three Ras double mutants, each bearing the V12mutation and an extra mutation in the effector-site domain. RasS35selectively stimulates the MAPK(ERK) pathway; RasC40 stimulatesphosphatidylinositol-3-OH kinase; and RasG37 stimulates ralGDS,the guanine-dissociation stimulator for the GTPase Ral22–24.Although all three haemagglutinin (HA)-tagged Ras mutants wereexpressed in numerous fibres in denervated regenerating muscle,only RasS35 was able to induce MyHC-slow in a large number offibres (Fig. 3a, b). The number of transfected fibres reactive towardsHA-tagged RasS35 was comparable to that of slow fibres (Fig. 3b),and for most fibres there was a close correlation between the expres-sion of this Ras mutant and that of MyHC-slow (Fig. 3c). RasS35mimicked the reprogramming of myosin gene expression induced

Figure 2 The fast-to-slow myosin switching induced by slow motor neurons is inhibited by dominant-negative RasN17. a, Serial cross-sections of innervated regenerating soleus muscle transfected with RasN17 and examined 7 days after transfection. Sections were stained with anti-MyHC-slow (left) and anti-Ras (right) antibodies and studied at low (top) or high (bottom) magnification. Fibres overexpressing RasN17 are unreactive for MyHC-slow protein (arrows). Scale bars represent 100 µm (top) and 25 µm (bottom). b, Serial cross-sections of innervated

regenerating soleus muscle transfected with RasN17 were analysed by immunofluorescence with anti-Ras antibodies (right) and by in situ hybridization with a probe specific for MyHC-2X transcripts (left). Fibres overexpressing RasN17 contain MyHC-2X transcripts (arrows). Scale bar represents 25 µm. c, Activity of the MyHC-slow promoter in innervated soleus transfected with RasN17 or with empty vector. Luciferase activity is expressed as arbitrary units.

a b

c

MyHC-2X RasN17

MyHC-slow RasN17 RasN17

Luci

fera

se a

ctiv

ity

Vector

140

120

100

80

60

40

20

0

© 2000 Macmillan Magazines LtdNATURE CELL BIOLOGY | VOL 2 | MARCH 2000 | cellbio.nature.com 143

articles

by slow motor neurons: the fibres expressing RasS35 and MyHC-slow protein were also reactive for MyHC-slow transcripts andunreactive for the fast MyHC-2X transcripts (Fig. 3c). In addition,transfected fibres also expressed the myosin light chain 1-slow, asdo innervated fibres25 (data not shown). The effect of RasS35appeared to result from transcriptional regulation of the MyHC-slow gene, as the MyHC-slow promoter was activated more thantenfold in denervated muscle, that is, to levels similar to those foundin innervated regenerating soleus (Fig. 3d). Transcriptional activa-tion of the MyHC-slow promoter by RasS35 was already detected 3days after transfection — much sooner than endogenous MyHC-slow protein could be detected in transfected fibres (data notshown). A role of the Ras–MAPK(ERK) pathway in the establish-ment of the slow myosin gene programme is further supported bythe finding that a constitutively active MAPK kinase (MAPKK)mutant induced numerous MyHC-slow-positive fibres in dener-vated muscle (Fig. 4a, b). MAPKK was also able to transactivate theMyHC-slow promoter in regenerating denervated soleus (Fig. 4c).

We investigated whether RasS35 acts mainly on muscle fibres oron satellite cells. In agreement with previous studies of gene transferin skeletal muscle19,26, we observed that RasS35 driven by thecytomegalovirus (CMV) promoter is expressed only in musclefibres and not in mononucleated cells from the earliest stages exam-ined, that is, at 24 h post-transfection (data not shown). At thisstage, we never observed labelled satellite cells associated with unla-belled muscle fibres, indicating that satellite cells are not transfectedunder these conditions. To induce overexpression of activated Rasselectively in differentiated muscle fibres, we transfected regenerat-ing muscles with RasS35 driven by the muscle-specific myoglobinpromoter, which is activated in cultured myotubes but not inmyoblasts27. The finding that numerous slow fibres (70 ± 23) were

Figure 4 MyHC-slow expression is induced by constitutively active MAPKK in denervated regenerating muscle. a, Immunofluorescence staining with anti-MyHC-slow antibody of a section of denervated soleus transfected with constitutively active MAPKK. Scale bar represents 100 µm. b, Number of MyHC-slow-positive fibres in denervated soleus transfected with constitutively active MAPKK or empty vector. c, MyHC-slow promoter activity in innervated soleus muscle (Inn) and in denervated soleus transfected with empty vector (Den) or MAPKK (Den + MAPKK).

a

b c

Num

ber

of s

low

fibr

es

150

50

0

100

150

50

0

100

MAPKK Vector

Luci

fera

se a

ctiv

ity

Inn Den Den+MAPKK

Figure 3 Myosin switching is induced by RasS35 but not by RasG37 or RasC40. a, Serial cross-sections of denervated regenerating soleus transfected with HA-tagged RasS35 or RasG37 were stained with anti-HA (left) or anti-MyHC-slow (right) antibodies. Note that numerous slow fibres are seen only after transfection with RasS35. Scale bar represents 100 µm. b, Number of fibres expressing Ras and MyHC-slow after transfection with the three Ras double mutants. c, Serial cross-sections of denervated soleus transfected with HA-tagged RasS35 and analysed by

immunofluorescence with anti-HA (top left) and anti-MyHC-slow (top right) antibodies, and by in situ hybridization with probes specific for MyHC-slow (bottom left) and MyHC-2X (bottom right) messenger RNA transcripts. The dashed line marks the boundary of a muscle fascicle containing transfected fibres. d, Activity of the MyHC-slow promoter in innervated (Inn) and denervated (Den) soleus transfected with empty vector, and in denervated muscles transfected with RasS35.

a

b

c

d

Ras MyHC-slow

Ras

Ras

MyHC-slow

MyHC-slow

MyHC-slow mRNA MyHC-2X mRNA

Num

ber

of fi

bres

400

300

200

100

0RasS35 RasG37 RasC40

Luci

fera

se a

ctiv

ity

160

120

80

40

0Inn Den Den+RasS35

RasS35

RasG37

© 2000 Macmillan Magazines Ltd144 NATURE CELL BIOLOGY | VOL 2 | MARCH 2000 | cellbio.nature.com

articles

induced by this RasS35 construct points to a primary effect of Rason differentiated muscle fibres.

We then asked whether RasS35 would be able to induce MyHC-slow gene expression in a regenerating fast muscle. However, nosignificant induction of slow fibres was observed after transfectionof RasS35 in the regenerating fast extensor digitorum longus (EDL)muscle (data not shown).Fibre growth is induced by RasC40 but not by RasS35. The differ-ential response of soleus and EDL muscle to overexpression ofRasS35 indicates that regenerating soleus muscle fibres may be pre-programmed to become slow fibres. It could be argued that the onlyrole of activated Ras or RasS35 would be to stimulate musclegrowth, and that the activation of the slow gene programme is a sec-ondary consequence of fibre growth. However, neither Ras norRasS35 stimulated growth of muscle fibres in regenerating dener-vated muscle. Interestingly, fibre growth was differentially affectedby the three Ras double mutants, and the effect on fibre size wasclearly dissociated from the effect on myosin gene expression (Fig.5). Transfected and untransfected fibres in denervated soleus mus-cles injected with RasS35 were similar in size (Fig. 5a). In contrast,fibres overexpressing RasG37 or RasC40, most of which do not con-tain MyHC-slow protein, were larger in size compared to untrans-fected fibres of the same muscles (Fig. 5b, c). The average cross-sectional area of the fibres expressing RasC40 was comparable tothat of the fibres present in innervated muscles, that is, about three

times larger than that of denervated fibres (Fig. 5d).ERK activity is increased by innervation and by electrostimula-tion. RasS35, which mimics the effect of slow motor neurons, selec-tively stimulates the MAPK(ERK) pathway. We therefore askedwhether ERK activity is increased by innervation in regeneratingsoleus muscle. To measure ERK activity specifically in muscle fibresand to avoid the contribution of proliferating satellite cells, wetransfected regenerating muscle with a plasmid expressing HA-tagged ERK2 (HA–ERK2); this plasmid, like that encoding RasS35as well as other expression vectors, is expressed in muscle fibres butnot in mononucleated cells (data not shown). ERK activity was notsignificantly increased at days 5 and 6 in innervated as compared todenervated muscle, but was significantly increased (by about 2.3-fold) at day 7 after injury (Fig. 6a), that is, at a time when a markedupregulation of the MyHC-slow promoter was observed in regen-erating innervated soleus (Fig. 6b).

We next determined whether the effect of muscle activity onERK is also detectable in adult, non-regenerating muscles and isrelated to specific patterns of activity. Previous studies showed thatERK is activated by electrostimulation in rat skeletal muscle16,17.However, only very short bouts of activity (up to 1 h) were studied,and activity patterns corresponding to the firing pattern of fast andslow motor neurons were not compared. We analysed endogenous

Figure 6 ERK activity is induced by nerve activity in regenerating soleus muscle and by electrostimulation in adult soleus muscle. a, Regenerating innervated or denervated muscles were transfected at day 3 after injury with pcDNA3 HA–ERK2, and kinase reactions were done at the indicated time periods using anti-HA immunoprecipitates from the corresponding muscle homogenates. Activity was normalized according to HA–ERK expression, as determined by western blot analysis of immunoprecipitates. Data are expressed as fold increase in ERK activity in innervated as compared with denervated muscles. b, Regenerating innervated or denervated muscles were transfected at day 3 after injury with a plasmid containing luciferase under the control of the MyHC-slow promoter. Luciferase activity was assayed at the indicated time periods. Data are expressed as fold increase in luciferase activity in innervated as compared with denervated muscles. c, ERK activity in adult denervated soleus muscles, stimulated for 24 h with low-frequency/high-amount (20 Hz), high-frequency/high-amount (100 Hz), or high-frequency/low-amount (150 Hz) impulse patterns. Total ERK activity was determined in muscle homogenates and is expressed as fold increase in ERK activity in stimulated as compared with unstimulated muscles.

a b

c

ER

K a

ctiv

ity (

fold

incr

ease

)

Luci

fera

se a

ctiv

ity (

fold

incr

ease

)

4

4 5 6 7

3

2

1

0

ER

K a

ctiv

ity (

fold

incr

ease

)

4

321

8

765

0

4

5

3

2

1

0

Days of regeneration4 5 6 7

Days of regeneration

Stimulation protocol150 Hz 100 Hz 20 Hz

Figure 5 Effect of Ras double mutants on muscle-fibre growth in regenerating muscles. a–c, Sections of denervated soleus muscles transfected with HA-tagged RasS35 (a), RasG37 (b) or RasC40 (c) were stained with anti-HA antibody. The striking hypertrophy of the fibres expressing RasC40 is apparent. Scale bar represents 50 µm. d, The size of the muscle fibres in innervated and denervated regenerating muscles is compared with the size of the transfected fibres (dark grey columns) and untransfected fibres (white columns) present in denervated muscles injected with one of the three Ras double mutants.

d

1,200

800

400

0

Fib

re s

ize

(µm

)2

Inn Den Den+RasS35

Den+RasG37

Den+RasC40

c

a b

© 2000 Macmillan Magazines LtdNATURE CELL BIOLOGY | VOL 2 | MARCH 2000 | cellbio.nature.com 145

articles

ERK activity in adult soleus muscles that were electrostimulated for24 h with impulse patterns that resemble the firing patterns of threetypes of motor unit28 and which have been shown previously toaffect MyHC gene expression29. The following patterns were used:first, a low-frequency/high-amount impulse pattern (20 Hz), simi-lar to that of slow motor neurons; second, a high-frequency/high-amount pattern (100 Hz), similar to that of fast, fatigue-resistantmotor units composed of type-2A, oxidative fibres; and third, ahigh-frequency/low-amount pattern (150 Hz), similar to that offast, fatiguable motor units composed of type-2B, glycolytic fibres.Electrostimulation for 24 h with the 20-Hz impulse patternincreased ERK activity almost sixfold over unstimulated controls,whereas the effect of the 100-Hz pattern was significantly less andthe 150-Hz pattern had no effect at all (Fig. 6c).

DiscussionRas has been traditionally associated with the inhibition of myob-last fusion and muscle-cell differentiation in culture30, although ithas been shown that the MAPK pathway positively regulates laterstages of muscle differentiation31,32. Our results point to a new rolefor Ras in the control of the differentiation of muscle-fibre types bynerve activity. Activated Ras is sufficient to drive a slow-myosingene programme in denervated regenerating soleus muscle and,conversely, slow motor neurons cannot induce slow myosin whenRas signalling is inhibited by RasN17. Ras activation is thusrequired to mediate the switch in myosin gene expression inducedby slow motor neurons in the regenerating soleus muscle.

Experiments with Ras double mutants and activated MAPKKsuggest that MAPK(ERK) is a major downstream pathway involvedin this effect. Accordingly, ERK activity is higher in innervated thanin denervated regenerating muscle, and the time course of thisincrease in ERK activity parallels the activation of the MyHC-slowpromoter induced by slow motor neurons. The close temporal cor-relation between the upregulation of MAPK activity and the activa-tion of the MyHC-slow promoter may indicate that there is a directcausal relationship between the two events. Thus RasS35, whichrapidly activates MAPK, should also have a rapid effect on MyHC-slow transcription. We have observed a difference between thetransactivation of the MyHC-slow promoter, which is stronglystimulated 3 days after transfection with RasS35, and the expressionof endogenous MyHC-slow, which is detected 5 days after transfec-tion. A possible explanation for this discrepancy is that the expres-sion of the endogenous MyHC-slow gene is also controlled at thetranscriptional or post-transcriptional level by MAPK(ERK)-inde-pendent pathways.

MAPK(ERK) activity is markedly increased in adult muscle after24 hours of electrostimulation with an impulse pattern that resem-ble the firing pattern of slow motor neurons. In contrast, no changein ERK activity is seen when using a pattern similar to the firing pat-tern of motor neurons innervating fast, fatiguable muscle fibres,and an intermediate effect is seen when using a pattern similar tothat of motor neurons innervating fast, fatigue-resistant musclefibres. These results indicate that ERK activity is sensitive to theamount and/or pattern of activity imposed onto the muscles andmight thus act as one of the ‘activity sensors’ that mediate the regu-lation of the transcriptional machinery in response to the differentfiring patterns of motor neurons.

The response to Ras–MAPK activation is dictated by intrinsicmuscle properties, as RasS35 is able to induce expression of MyHC-slow in the regenerating slow soleus but not in the fast EDL muscle.Satellite cells from fast and slow muscles differ in their capacity togenerate myofibres expressing slow myosin in culture33–36. Further-more, fast and slow rat muscles differ in their response to electros-timulation: slow myosin is not induced in fast muscles after 2months of electrostimulation with a low-frequency impulse patternthat is able to maintain expression of slow myosin in denervatedsoleus29. Our findings are thus consistent with the idea that the

expression of the slow phenotype is dependent on both myoblastlineage and innervation37. On the other hand, the induction of theslow-myosin gene programme by Ras is not a secondary conse-quence of a general activation of gene expression and musclegrowth in fibres already committed to this differentiation pathway.In fact, RasS35 induces the myosin-gene switch but not musclehypertrophy, whereas RasC40 has the opposite effect. These find-ings indicate that the mechanisms regulating muscle-fibre growthare different from those involved in myosin-gene regulation.

Our data, while clearly showing a central role for Ras–MAPK(ERK) signalling in the differentiation of slow muscle fibres,do not rule out a role of further pathways, including other MAPKpathways and pathways involving calcineurin. The downstreameffectors and upstream signals involved in Ras-dependent regula-tion of muscle genes remain to be identified. Calcium is a potentialcandidate as an upstream signal, for several reasons. First, inner-vated and denervated muscles are likely to receive different patternsof calcium signals; second, Ras can be activated by calcium38; andthird, gene expression can be specifically controlled by the temporalpattern of calcium transients and oscillations39–41. It is tempting tospeculate that the distinct pattern of action potentials, and associ-ated intracellular calcium changes, evoked by nerve activity in slowmuscle fibres may lead to the activation of a specific set of transcrip-tion factors through Ras–MAPK pathways. h

MethodsMuscle regeneration and in vivo transfection.The rat soleus muscle of 200–250-g male Wistar rats was exposed under ketamine anaesthesia, and 0.5

ml 0.5% bupivacaine (1-butyl-N(2,6-dimethylphenyl)2-piperidinecarboxamide; Marcaine) in saline was

injected into the muscle. Denervation was produced by cutting the sciatic nerve high in the thigh.

Plasmid DNA (50 µg) was injected at day 3 after bupivacaine treatment in innervated or denervated

muscles as described19. Muscles were removed at various time periods after injury and frozen in

isopentane cooled in liquid nitrogen. For electrostimulation experiments, adult rat soleus muscles were

denervated to abolish nerve-evoked muscle activity and stimulated, through electrodes placed on the

muscles, at 20 Hz (200 pulses every 30 s), 100 Hz (60 pulses every 60 s) or 150 Hz (25 pulses every 15 min)

for 24 h, as described29. Unstimulated denervated soleus muscles were used as controls.

Plasmids and transactivation.The activated Ras mutant constructs, H-RasV12, H-RasV12S35, H-RasV12G37 and H-RasV12C40, were

inserted under the control of the CMV promoter in the same pDCR expression vector42. To express

RasS35 under the control of the myoglobin promoter, RasS35 was excised with SalI and BamHI from the

pDCR vector, blunt-ended and inserted into pGL3-myoglobin27 after excision of the luciferase reporter

gene. Additional mutants used in these experiments were the dominant-negative H-RasN17, activated

MAPKK and HA-tagged ERK2. For transactivation experiments, the rat MyHC-slow promoter linked to

the luciferase gene43 (50 µg) was co-injected with mutant Ras or empty vector (50 µg). Transfection

efficiency was monitored by co-transfection with RSV-CAT (5 µg). Luciferase activity was normalized for

extract protein content. Results of transfection and transactivation experiments represent the mean from

at least five different muscles.

In vitro kinase assay.To measure ERK activity in adult, electrostimulated soleus muscles, aliquots of 300 µg muscle extracts

were immunoprecipitated with anti-ERK antibody (Santa Cruz Biotechnology) and ERK activity was

determined in the immunoprecipitate as described44. To measure ERK2 activity in regenerating soleus

muscles transfected with pcDNA-3/HA–ERK2, we immunoprecipitated aliquots of muscle extracts with

anti-HA antibody (12CA5, Boehringer Mannheim) and measured ERK activity in the

immunoprecipitate. HA–ERK2 expression levels were assessed in parallel samples by western blotting.

Data represent the mean of three to four different muscles.

Immunocytochemistry and in situ hybridization.Cryosections of regenerating muscles were processed for immunofluorescence with the monoclonal

antibody BA-D5, which is specific for the MyHC-slow protein45. The monoclonal antibodies anti-HA

and anti-pan-Ras (Ab-1, Calbiochem) were used on cryosections after fixation with 4%

paraformaldehyde. In situ hybridization with probes specific for the 3′-untranslated regions of MyHC-

slow and MyHC-2X was done as described46.

Data analysis.All data are expressed as means ± s.e.m. (represented as error bars). Comparisons were made using the

Student’s t-test, with P < 0.05 being considered statistically significant.

RECEIVED 23 AUGUST 1999; REVISED 21 DECEMBER 1999; ACCEPTED 21 JANUARY 2000; PUBLISHED 4 FEBRUARY 2000.

1. Duclert, A. & Changeux, J. P. Acetylcholine receptor gene expression at the developing

neuromuscular junction. Physiol. Rev. 75, 339–368 (1995).

2. Eftimie, R., Brenner, H. R. & Buonanno, A. Myogenin and MyoD join a family of skeletal muscle

genes regulated by electrical activity. Proc. Natl Acad. Sci. USA 88, 1349–1353 (1991).

3. Klarsfeld, A. et al. Regulation of muscle AChR α subunit gene expression by electrical activity:

© 2000 Macmillan Magazines Ltd146 NATURE CELL BIOLOGY | VOL 2 | MARCH 2000 | cellbio.nature.com

articles

involvement of protein kinase C and Ca2+. Neuron 2, 1229–1236 (1989).

4. Huang, C. F., Tong, J. & Schmidt, J. Protein kinase C couples membrane excitation to acetylcholine

receptor gene inactivation in chick skeletal muscle. Neuron 9, 671–678 (1992).

5. Chahine, K. G., Baracchini, E. & Goldman, D. Coupling muscle electrical activity to gene expression

via a cAMP-dependent second messenger system. J. Biol. Chem. 268, 2893–2898 (1993).

6. Hughes, S. M. et al. Selective accumulation of MyoD and myogenin in fast and slow adult skeletal

muscle is controlled by innervation and hormones. Development 118, 1137–1147 (1993).

7. Voytik, S. L., Przyborski, M., Badylak, S. F. & Konieczny, S. Differential expression of muscle

regulatory factor genes in normal and denervated adult rat hindlimb muscles. Dev. Dyn. 98, 214–224

(1993).

8. Hughes, S., Chi, M., Lowry, O. & Gundersen, K. Myogenin induces a shift of enzyme activity from

glycolytic to oxidative metabolism in muscles of transgenic mice. J. Cell Biol. 145, 633–642 (1999).

9. Chin, E. et al. A calcineurin-dependent transcriptional pathway controls skeletal muscle fibre type.

Genes Dev. 12, 2499–2509 (1998).

10. Naya, R. J. et al. Stimulation of slow skeletal muscle fiber gene expression by calcineurin in vivo.

J. Biol. Chem. 275, 4545–4548 (2000).

11. Calvo, S., Venepally, P., Cheng, J. & Buonanno, A. Fibre-type-specific transcription of the troponin

I slow gene is regulated by multiple elements. Mol. Cell. Biol. 19, 515–525 (1999).

12. Semsarian, C. et al. Skeletal muscle hypertrophy is mediated by a Ca2+-dependent calcineurin

signalling pathway. Nature 400, 576–581 (1999).

13. Goodyear, L. J., Chung, P.-Y., Sherwood, D., Dufresne, S. D. & Moller, D. E. Effects of exercise and

insulin on mitogen-activated protein kinase signalling pathways in rat skeletal muscle. Am. J. Physiol.

271, E403–E408 (1996).

14. Aronson, D. et al. Exercise stimulates the mitogen-activated protein kinase pathway in human

skeletal muscle. J. Clin. Invest. 99, 1251–1257 (1997).

15. Widegren, U. et al. Divergent effects of exercise on metabolic and mitogenic signalling pathways in

human skeletal muscle. FASEB J. 12, 1379–1389 (1998).

16. Aronson, D., Dufresne, S. D. & Goodyear, L. J. Contractile activity stimulates the c-Jun NH2-terminal

kinase pathway in rat skeletal muscle. J. Biol. Chem. 272, 25636–25640 (1997).

17. Sherwood, D. et al. Differential regulation of MAP kinase, p70(S6K), and Akt by contraction and

insulin in rat skeletal muscle. Am. J. Physiol. 276, E870–E878 (1999).

18. Esser, K., Gunning, P. & Hardeman, E. Nerve-dependent and -independent patterns of mRNA

expression in regenerating skeletal muscle. Dev. Biol. 159, 173–183 (1993).

19. Vitadello, M., Schiaffino, M. V., Picard, A., Scarpa, M. & Schiaffino, S. Gene transfer in regenerating

muscle. Hum. Gene Ther. 5, 11–18 (1994).

20. Schiaffino, S. & Reggiani, C. Molecular diversity of myofibrillar proteins: gene regulation and

functional significance. Physiol. Rev. 76, 371–423 (1996).

21. Jerkovic, R., Argentini, C., Serrano-Sanchez, A., Cordonnier, C. & Schiaffino, S. Early myosin

switching induced by nerve activity in regenerating slow skeletal muscle. Cell Struct. Funct. 22,

147–153 (1997).

22. White, M. et al. Multiple Ras functions can contribute to mammalian cell transformation. Cell 80,

533–541 (1995).

23. Joneson, T., White, M., Wigler, M. & Bar-Sagi, D. Stimulation of membrane ruffling and MAP kinase

activation by distinct effectors of RAS. Science 271, 810–812 (1996).

24. Rodriguez-Viciana, P. et al. Role of phosphoinositide 3-OH kinase in cell transformation and control

of the actin cytoskeleton by Ras. Cell 89, 457–467 (1997).

25. Jerkovic, R., Vitadello, M., Kelly, R., Buckingham, M. & Schiaffino, S. Fibre type-specific and nerve-

dependent regulation of myosin light chain 1 slow promoter in regenerating muscle. J. Muscle Res.

Cell Motil. 18, 369–373 (1997).

26. Wolff, J. A. et al. Direct gene transfer into mouse muscle in vivo. Science 247, 1465–1468 (1990).

27. Devlin, B. H., Wefald, F. C., Kraus, W. E., Bernard, T. S. & Williams, R. S. Identification of a muscle-

specific enhancer within the 5'-flanking region of the human myoglobin gene. J. Biol. Chem. 264,

13896–13901 (1989).

28. Hennig, R. & Lømo, T. Firing patterns of motor units in normal rats. Nature 314, 164–166 (1985).

29. Ausoni, S., Gorza, L., Schiaffino, S., Gundersen, K. & Lømo, T. Expression of myosin heavy chain

isoforms in stimulated fast and slow rat muscles. J. Neurosci. 10, 153–160 (1990).

30. Olson, E. N., Spizz, G. & Tainski, M. A. The oncogenic forms of N-ras and H-ras prevent skeletal

myoblast differentiation. Mol. Cell. Biol. 7, 2104–2111 (1987).

31. Bennet, A. M. & Tonks, N. K. Regulation of distinct stages of skeletal muscle differentiation by

mitogen-activated protein kinases. Science 278, 1288–1291 (1997).

32. Gredinger, E., Gerber, A. N., Tamir, Y., Tapscott, S. J. & Bengal, E. Mitogen-activated protein kinase

pathway is involved in the differentiation of muscle cells. J. Biol. Chem. 273, 10436–10444 (1998).

33. Feldman, J. L. & Stockdale, F. E. Skeletal muscle satellite cell diversity: satellite cells form fibres of

different types in cell culture. Dev. Biol. 143, 320–334 (1991).

34. Dusterhoft, S. & Pette, D. Satellite cells from slow rat muscle express slow myosin under appropriate

culture conditions. Differentiation 53, 25–33 (1993).

35. Barjot, C., Cotten, M.-L., Goblet, C., Whalen, R. G. & Bacou, F. Expression of myosin heavy chain

and of myogenic regulatory factor genes in fast or slow rabbit muscle satellite cell cultures. J. Muscle

Res. Cell Motil. 16, 619–628 (1995).

36. Rosenblatt, J., Parry, D. & Partridge, T. Phenotype of adult mouse muscle myoblasts reflects their

fibre type of origin. Differentiation 60, 39–45 (1996).

37. DiMario, J. X., & Stockdale, F. E. Both myoblast lineage and innervation determine fiber type and are

required for expression of the slow myosin heavy chain 2 gene. Dev. Biol. 188, 167–180 (1997).

38. Finkbeiner, S. & Greenberg, M. E. Ca2+-dependent routes to Ras: mechanisms for neuronal survival,

differentiation, and plasticity? Neuron 16, 233–236 (1996).

39. Gu, X. & Spitzer, N. C. Distinct aspects of neuronal differentiation encoded by frequency of

spontaneous Ca2+ transients. Nature 375, 784–787 (1995).

40. Dolmetsch, R., Lewis, R. S., Goodnow, C. C. & Healy, J. I. Differential activation of transcription

factors induced by Ca2+ response amplitude and duration. Nature 386, 855–858 (1997).

41. Fields, R. D., Eshete, F., Stevens, B. & Itoh, K. Action potential-dependent regulation of gene

expression: temporal specificity in Ca2+, cAMP-responsive element binding proteins, and mitogen-

activated protein kinase signalling. J. Neurosci. 17, 7252–7266 (1997).

42. Ramocki, M. B. et al. Signalling through mitogen-activated protein kinase and Rac/Rho does not

duplicate the effects of activated Ras on skeletal myogenesis. Mol. Cell. Biol. 17, 3547–3555 (1997).

43. Hasegawa, K., Lee, S. J., Jobe, S. M., Markham, B. E. & Kitsis, R. N. Cis-acting sequences that mediate

induction of beta-myosin heavy chain gene expression during left ventricular hypertrophy due to

aortic constriction. Circulation 96, 3943–3953 (1997).

44. Crespo, P. et al. Signalling through transforming G protein-coupled receptors in NIH 3T3 cells

involves c-Raf activation. Evidence for a protein kinase C-independent pathway. J. Biol. Chem. 269,

21103–21109 (1994).

45. Schiaffino, S. et al. Three myosin heavy chain isoforms in type 2 skeletal muscle fibres. J. Muscle Res.

Cell Motil. 10, 197–205 (1989).

46. De Nardi, C. et al. Type 2X myosin heavy chain is coded by a muscle fibre type-specific and

developmentally regulated gene. J. Cell Biol. 123, 823–835 (1993).

ACKNOWLEDGEMENTS

This work was supported by grants from the European Commission (BIOTECH contract number

ERBBIO4CT960216), the Giovanni Armenise–Harvard Foundation for Advanced Scientific Research,

Telethon-Italy, the Italian Space Agency (ASI) and the Italian Ministry of University and Scientific and

Technological Research (MURST). A.L.S. is a TMR Marie Curie research training grant recipient. We

thank A. Hall, E. Taparowski and S. Alemà for Ras mutants; J.S. Gutkind for HA–ERK; R. Kitsis for

MyHC-slow promoter-luciferase; R.S. Williams for myoglobin–luciferase; T. Pozzan, P.P. di Fiore and

G. Scita for discussions; and A. Picard for help with some experiments.

Correspondence and requests for materials should be addressed to S.S.

© 2000 Macmillan Magazines LtdNATURE CELL BIOLOGY | VOL 2 | MARCH 2000 | cellbio.nature.com 147