pRb is required for MEF2-dependent gene expression as well as cell-cycle arrest during skeletal muscle differentiation

Research Paper

449

pRb is required for MEF2-dependent gene expression as well as cell-cycle arrest during skeletal muscle differentiation Bennett G. Novitch*†, Douglas B. Spicer*‡, Paul S. Kim, Wang L. Cheung and Andrew B. Lassar Background: The onset of differentiation-specific gene expression in skeletal muscle is coupled to permanent withdrawal from the cell cycle. The retinoblastoma tumor-suppressor protein (pRb) is a critical regulator of this process, required for both cell-cycle arrest in G0 phase and high-level expression of late muscle-differentiation markers. Although the cell-cycle defects that are seen in pRb-deficient myocytes can be explained by the well-described function of pRb as a negative regulator of the transition from G1 to S phase, it remains unclear how pRb positively affects late muscle-gene expression. Results: Here, we show that the myogenic defect in Rb–/– cells corresponds to a deficiency in the activity of the transcription factor MEF2. Without pRb, MyoD induces the accumulation of nuclear-localized MEF2 that is competent to bind DNA yet transcriptionally inert. When pRb is present, MyoD stimulates the function of the MEF2C transcriptional activation domain and the activity of endogenous MEF2-type factors. Co-transfection of MyoD together with an activated form of MEF2C containing the Herpesvirus VP16 transcriptional activation domain partially bypasses the requirement for pRb and induces late muscle-gene expression in replicating cells. This ectopic myogenesis is nevertheless significantly augmented by co-expression of an E2F1–pRb chimeric protein that blocks the cell cycle. Conclusion: These findings indicate that pRb promotes the expression of latestage muscle-differentiation markers by both inhibiting cell-cycle progression and cooperating with MyoD to promote the transcriptional activation activity of MEF2.

Background Skeletal muscle differentiation (myogenesis) involves a cascade of muscle-specific gene expression that is coordinated with permanent withdrawal from the cell cycle. The commitment of cells to the myogenic lineage requires either of two members of the myogenic basic helix–loop–helix (bHLH) transcription-factor family, MyoD or Myf-5, which are expressed in proliferating myoblasts prior to the onset of muscle differentiation (reviewed in [1–3]). Further steps in myogenesis require another myogenic bHLH factor, myogenin, as well as the MEF2 transcription-factor family which cooperate with the myogenic bHLH proteins in the activation of many muscle structural genes (reviewed in [1–3]). Studies of the differentiation of cultured myoblasts have revealed that muscle-differentiation-specific gene expression occurs in a stereotypic pattern. Within 24 hours of serum removal, proliferating myoblasts initiate the expression of myogenin. Subsequently, these cells express the cyclin-dependent kinase (Cdk) inhibitor p21 and permanently exit the cell cycle. Once the cells have become post-mitotic, expression of myofibrillar proteins such as myosin heavy chain (MHC) and enzymatic genes such as muscle creatine kinase (MCK) begins, approximately

Address: Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, 240 Longwood Avenue, Boston, Massachusetts 02115, USA. Present address: †Howard Hughes Medical Institute, Department of Biochemistry and Molecular Biophysics, Columbia University College of Physicians & Surgeons, New York, New York 10032, USA. ‡Center for Molecular Medicine, Maine Medical Center Research Institute, 125 John Roberts Road, South Portland, Maine 04106, USA. *B.G.N. and D.B.S. contributed equally to this work. Correspondence: Andrew B. Lassar E-mail: [email protected] Received: 30 December 1998 Revised: 10 March 1999 Accepted: 18 March 1999 Published: 20 April 1999 Current Biology 1999, 9:449–459 http://biomednet.com/elecref/0960982200900449 © Elsevier Science Ltd ISSN 0960-9822

36–48 hours after the onset of differentiation, followed by fusion of cells into multinucleated myotubes [4–6]. With the exception of the earliest induction of myogenin, and perhaps p21, muscle-differentiation-specific transcription occurs exclusively in non-dividing cells, suggesting that the expression of most differentiation-specific genes requires an arrest in the G0 phase of the cell cycle. This idea is supported by findings that muscle-gene expression can be inhibited by peptide growth factors, activated oncogenes, or other proteins that promote cell-cycle progression (reviewed in [6–9]). Most of these studies indicate that perturbation of the cell cycle antagonizes the function of the myogenic bHLH proteins, resulting in a general absence of differentiation-specific gene expression in dividing cells. However, these studies have not clarified why early markers of the myogenic program, such as myogenin, can be expressed in proliferating cells whereas late markers of the differentiation program are induced only after cell-cycle withdrawal. The retinoblastoma tumor-suppressor protein (pRb) plays a critical role in establishing the G0 arrest observed in

450

Current Biology, Vol 9 No 9

and pRb in the activation of MEF2 transactivation-domain (TAD) function. The need for pRb and cell-cycle exit in the later stages of myogenesis can be partially fulfilled by artificially activating MEF2. Full restoration of the muscledifferentiation program in Rb–/– cells, however, requires both MEF2 activation and cell-cycle arrest. Together, these findings indicate that pRb regulates the later stages of myogenesis through its ability both to arrest cellular growth and, in conjunction with MyoD, to affect MEF2 function positively. Because MyoD and pRb are both required to functionally activate MEF2, our findings further suggest that modulation of pRb function may act to restrict the expression of MEF2-dependent late muscledifferentiation markers to post-mitotic cells.

differentiated myocytes; muscle cells lacking pRb fail to exit the cell cycle and are susceptible to apoptotic cell death [10–13]. In addition, pRb is required for certain aspects of the muscle differentiation program. Whereas pRb-deficient muscle cell lines express normal levels of both myogenin and p21, these cells express reduced levels of MHC and are unable to activate the expression of a transfected MCK reporter construct [10]. These in vitro findings parallel the muscle defects observed in mice that express reduced levels of pRb during embryogenesis. Skeletal muscle in these animals contains wild-type levels of the early muscledifferentiation markers myogenin and cardiac α-actin, but diminished levels of the late muscle-differentiation markers MCK and MRF4 [12]. Furthermore, fusion of pRb-deficient myocytes into multinucleated myotubes is markedly impaired [12]. Thus, pRb is specifically required for execution of the later steps in skeletal myogenesis, and its absence molecularly uncouples the early and late phases of this differentiation program.

Results pRb is required for MEF2 activation during skeletal muscle differentiation

Previous observations that early muscle differentiation markers are normally expressed in the absence of pRb have suggested that the ability of MyoD to activate transcription per se might be pRb-independent. The specific loss of late muscle-specific gene expression in pRb-deficient myocytes might then result from the reduced function of another pRb-dependent component of the myogenic program. To test this possibility, we first determined whether MyoD required pRb to activate a simplified reporter construct containing four repeated MyoD-binding sites (E-box reporter). In Rb–/– fibroblasts, MyoD significantly activated this reporter, and co-expression of pRb only slightly enhanced this activity (Figure 1a, compare columns 2 and 3). A similar result was obtained with a p21 promoter construct that also contains multiple E-box elements (data

It is currently believed that pRb controls cell-cycle progression through its ability to silence E2F-dependent genes (reviewed in [14]), but it is unclear whether this or a different property of pRb stimulates late muscle-gene expression, as pRb mutants that fail to bind stably to E2F can still enhance muscle differentiation [15]. To better understand the mechanism(s) through which pRb positively affects myogenesis and to ascertain why the later steps of myogenesis are specifically altered in the absence of pRb, we have further characterized the differentiation defects that are seen in pRb-deficient muscle cells. We demonstrate that the failure to activate late muscle-gene expression in pRbdeficient myocytes is due to a requirement for both MyoD Figure 1

(a)

60

(b)

E-box reporter

56

MEF2 reporter

52

50 Relative activity

Relative activity

48 40 30 20

20 16 12 8

10

4

0

0 1

MyoD: MyoD–VP16: pRb: Current Biology

2 +

3 +

4 +

+

5 MyoD: MyoD–VP16: + + MEF2C: MEF2C–VP16: pRb:

1

2 +

3 +

+

4

5

+

+ +

6

7

+

+ +

8

9

+

+ +

Activation of endogenous MEF2 function by MyoD requires pRb. (a) Rb–/– fibroblasts were transiently transfected with a CAT reporter construct containing four repeated MyoDbinding sites (E-box reporter) along with expression plasmids for MyoD, the bHLH domain of MyoD fused to the herpesvirus VP16 transcriptional-activation domain (MyoD–VP16), and pRb, as indicated (+). (b) Rb–/– fibroblasts were transfected with a CAT reporter construct containing four repeated MEF2-binding sites (MEF2 reporter) along with expression plasmids for MyoD, MyoD–VP16, full-length MEF2C, the MADS and MEF2 domain of MEF2C fused to VP16 (MEF2C–VP16), and pRb, as indicated. Error bars represent the results from duplicate plates from a representative experiment.

Research Paper Requirement for pRb in skeletal muscle differentiation Novitch et al.

MEF2 expression, DNA binding, and nuclear localization are not modulated by pRb

To discern the step in the regulation of MEF2 activity at which Rb–/– myocytes were blocked, we investigated whether MEF2 expression levels, DNA-binding properties, or nuclear localization were modulated in a MyoDdependent and/or pRb-dependent manner. As reported previously [10], pRb-deficient fibroblasts infected with a MyoD retrovirus express wild-type levels of myogenin upon differentiation, but they express markedly reduced levels of MHC and MCK (Figure 2a, compare lanes 2 and 4). Interestingly, whereas MCK was only slightly induced by MyoD in the absence of pRb, the brain isoform of creatine kinase (BCK) was highly upregulated (Figure 2a, lane 2). BCK is normally expressed at low levels in proliferating

D yo

– M

Probe

Rb+/+

M

Rb–/–

yo – D

(b)

D yo

Immunoblot MyoD

Rb+/+

M

Rb–/–

(a)

yo – D

As muscle differentiation proceeds, myogenic bHLH proteins induce the expression and activity of members of the MEF2 transcription factor family [16,17], which are thought to be important for the expression of a number of skeletal muscle-specific genes (reviewed in [1–3,18]). Furthermore, in Drosophila, D-MEF2-deficient embryos contain myoblasts that fail to undergo the later phases of muscle differentiation [19–21], similar to the muscle phenotype of Rb–/– mice. We therefore thought that pRb could be required for MyoD either to induce the expression of MEF2 or perhaps to affect the DNA-binding or transcriptional-activation properties of the MEF2 proteins themselves. To test these possibilities, we transfected Rb–/– fibroblasts with a MyoD expression plasmid and assayed whether endogenous MEF2 activity was increased by monitoring the activity of a MEF2 reporter construct containing four repeated MEF2-binding sites (MEF2 reporter), in either the absence or the presence of pRb. As previously observed with the activation of the MCK promoter ([10], see also below), MyoD required pRb to activate the MEF2 reporter (Figure 1b, columns 2 and 3). Strikingly, activation of the MEF2 reporter by a fusion protein consisting of MyoD with the herpesvirus VP16 transactivation domain (MyoD–VP16) remained predominantly pRb-dependent (Figure 1b, columns 4 and 5), even though MyoD–VP16 alone activated the E-box reporter considerably more than did wild-type MyoD and pRb together (Figure 1a, columns 3 and 4). Moreover, MyoD–VP16 did not activate the MEF2 reporter to significantly greater levels in either the absence or the presence of pRb than did wild-type MyoD (Figure 1b, compare columns 2 and 3 with columns 4 and 5). Together, these findings indicate that pRb is specifically required for MyoD-mediated activation of MEF2 and that this process is not a direct function of the transactivation abilities of MyoD.

Figure 2

– M

not shown). These experiments indicate that, although cotransfected pRb can increase E-box reporter activity, MyoD substantially activates transcription in its absence.

451

MyoD– E protein Myogenin– E protein

E box

Myogenin MEF2 MHC

MEF2 Creatine B kinase M 1

1

2 3 4

2

3

(c)

MEF2 (d)

DAPI (e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

4

β-gal

MyoD

pRb

MyoD and pRb Current Biology

MyoD induces nuclear-localized MEF2 protein that can bind DNA in the absence of pRb. (a,b) Protein extracts were prepared from Rb–/– or wild-type fibroblasts infected with a retrovirus encoding MyoD, as indicated, and then allowed to differentiate for 3 days. (a) Western analysis of muscle-specific protein expression. The more slowly migrating brain isoform (B) and the faster-migrating muscle isoform (M) of creatine kinase are indicated. When normalized to MyoD content, the levels of muscle-protein expression in Rb–/– cells relative to wildtype determined by scanning densitometry were: myogenin, 75%; MEF2, 110%; MHC, 9%; BCK, 2000%; MCK, 1%. (b) MyoD induces MEF2 DNA-binding complexes in either the absence or presence of pRb. Electrophoretic mobility shift assay (EMSA) analysis of E-box and MEF2 DNA-binding activities. Complexes of MyoD and its E-protein binding partner (MyoD–E protein), myogenin–E protein, and MEF2, with DNA are indicated. (c–n) MyoD induces nuclear accumulation of MEF2 in either the absence or presence of pRb. Rb–/– fibroblasts were transiently transfected with expression vectors encoding either (c–e) β-galactosidase, (f–h) Myc-tagged MyoD, (i–k) pRb, or (l–n) the combination of MyoD and pRb. Cells were immunostained for (c,f,i) the transfected protein, (l) transfected MyoD, and (d,g,j,m) endogenous MEF2. (e,h,k,n) Nuclei were visualized by DAPI staining.

myoblasts and, like myogenin and p21, is upregulated within the first 24 hours of myogenic differentiation. As differentiation proceeds, however, its expression normally declines as it is replaced by the muscle-specific isoform, MCK [22]. Together, the elevated expression of BCK and decreased expression of both MHC and MCK are evidence

452

Current Biology, Vol 9 No 9

of a specific blockade in the expression of late differentiation markers in MyoD-infected Rb–/– cells. Although these pRb-deficient myocytes failed to fully differentiate, MyoD induced MEF2 protein expression to wild-type levels (Figure 2a, compare lanes 2 and 4). Electrophoretic mobility shift assays (EMSA) revealed that MyoD also induced identical amounts of MEF2 DNAbinding activity in either the absence or the presence of pRb (Figure 2b), indicating that the observed defects in MEF2 function were not due to impaired DNA binding. Recently, it has been demonstrated that MEF2 proteins can be regulated by translocation to the cytoplasm [23]. We therefore investigated whether the localization of MEF2 was affected by pRb, and found that MyoD induced similar levels of nuclear MEF2 accumulation whether or not pRb was co-expressed (compare Figure 2g with 2m). These findings indicate that neither the protein levels, nor the DNA-binding properties, nor the nuclear localization of MEF2, is abnormal in Rb–/– fibroblasts that express MyoD, and suggest that the inability of MyoD to induce MEF2-reporter activity reflects a deficiency in some other aspect of MEF2 function. A MEF2–VP16 chimera provides MEF2 function in the absence of both MyoD and pRb

Because fibroblasts express very low levels of MEF2 protein in the absence of MyoD (Figure 2a), we investigated whether pRb affected the activity of exogenously supplied MEF2. When transfected into Rb–/– cells, MEF2C activated the MEF2 reporter poorly, even when pRb was co-expressed (Figure 1b, columns 6,7), suggesting that exogenously expressed MEF2 also requires additional factors or modifications to reveal its functional capacity. The effects seen here were not unique to MEF2C, as we obtained similar results with transfected MEF2A and MEF2D (data not shown). These findings indicate that both MyoD and pRb together are required to enhance MEF2 function. To see whether the induction of MEF2 activity was a property of the MEF2 TAD, we tested whether a chimeric protein containing the MEF2C DNA-binding domain (amino acids 1–117) linked to the VP16 TAD (MEF2C–VP16 [24]) required either MyoD or pRb for its function. We found that MEF2C–VP16 was a potent activator of the MEF2 reporter in the absence of MyoD, and that the activity of this construct was not significantly augmented by the co-expression of pRb (Figure 1b, compare columns 8 and 9). Similar results were obtained with a construct consisting of the VP16 TAD fused to full-length MEF2C (data not shown). Together, these findings demonstrate that MyoD-mediated induction of MEF2 function is specifically defective in pRb-deficient cells, and that the transcriptional activity of MEF2 can be restored by appending the VP16 TAD to MEF2.

MyoD stimulates the activity of the MEF2C transactivation domain

Given that fusing a VP16 TAD onto MEF2C activated MEF2C in the absence of both MyoD and pRb, we reasoned that the myogenic defect in pRb-deficient fibroblasts could result from a failure of MyoD to positively affect the function of the wild-type MEF2C TAD. To explore this possibility, we co-transfected pRb-positive 10T½ fibroblasts with an expression vector encoding a chimeric protein consisting of the yeast Gal4 DNA-binding domain fused to the TAD of human MEF2C (see Figure 3a [25]), in either the absence or the presence of expression vectors encoding MyoD and its E-protein binding partner, E12. Gal4–MEF2C alone weakly activated a co-transfected Gal4-reporter construct, but the inclusion of MyoD and E12 increased this activity at least 10-fold (Figure 3b, compare columns 3 and 4). In 10T½ cells, activation of MEF2 reporter constructs was maximally enhanced by the co-expression of MyoD with E12 (data not shown); we therefore included exogenous E12 in this and in subsequent experiments. Importantly, the expression of E12 by itself did not enhance Gal4–MEF2C activity (data not shown), indicating that MyoD was required to activate the MEF2C TAD. Furthermore, transfection of MyoD and E12 in the absence of Gal4–MEF2C failed to activate the Gal4 reporter (Figure 3b, column 2). Moreover, co-transfection with MyoD and E12 did not significantly augment the activity of a Gal4–VP16 TAD chimera (data not shown), indicating that MyoD and E12 expression specifically induces the function of the MEF2C TAD. It has been shown previously that MyoD and MEF2 are able to interact directly through their bHLH and MADS/MEF2 domains, respectively [24,26]. Since the Gal4–MEF2 construct that we employed lacks the MADS/MEF2 domain [25], these data further suggest that MyoD can modulate MEF2 function in the absence of a direct interaction. pRb promotes MyoD-mediated activation of the MEF2C transactivation domain

To test whether pRb is required for MyoD to activate the MEF2C TAD, we transfected Rb–/– fibroblasts with the Gal4–MEF2C fusion protein either alone or with MyoD and E12, in either the absence or the presence of exogenous pRb. When expressed alone, MyoD+E12 did not significantly stimulate Gal4–MEF2C function in these cells (Figure 3c, column 4). Similarly, the expression of pRb alone did not affect the activity of Gal4–MEF2C (Figure 3c, column 5); but, the combined expression of MyoD+E12 and pRb led to a 45-fold increase in the activity of Gal4–MEF2C (Figure 3c, column 6). These findings indicate that MyoDmediated activation of the MEF2C TAD requires pRb. MyoD-mediated activation of the MEF2C transactivation domain requires Ser387 of MEF2C

Recently, it was reported by Han et al. [25] that the TAD of MEF2C can be positively regulated through

Research Paper Requirement for pRb in skeletal muscle differentiation Novitch et al.

Expression of MyoD together with MEF2C–VP16 dissociates activation of the MCK promoter from G0 arrest

Given that MyoD-mediated activation of the MEF2 TAD required pRb, we evaluated whether artificial activation of MEF2 could bypass the requirement for pRb for late muscle-gene expression. We therefore tested whether MEF2C–VP16 could cooperate with MyoD in the activation of an MCK–luciferase reporter construct in Rb–/– fibroblasts. The combination of MyoD+E12 and MEF2–VP16 induced MCK–luciferase expression to levels significantly above that obtained with MyoD+E12 alone, to 30% of that observed with MyoD+E12 and pRb (Figure 4a, compare columns 2,3,5). In marked contrast, the combination of MyoD+E12 and wild-type MEF2C failed to induce MCK–luciferase expression in the absence of pRb (Figure 4a, column 4). Activation of the MCK–luciferase reporter required MyoD, as expression of MEF2C–VP16 alone did not activate it (data not shown). Interestingly, the addition of the VP16 TAD activated MCK–luciferase expression only when appended to MEF2; the combination of MyoD–VP16+E12 and wildtype MEF2C failed to drive expression of this reporter construct (data not shown). Because co-expression of MyoD with MEF2C–VP16 substantially activated the MCK promoter in the absence of pRb, we examined whether this induction of differentiation-specific gene expression coincided with cell-cycle exit. To evaluate this, we transfected pRb-deficient fibroblasts with MyoD and an MCK–CAT reporter construct along with either pRb or MEF2C–VP16. After

Figure 3

293 300 387 441

87

(a) MADS

MEF2C (WT)

TT

S

MEF2

Gal4–MEF2C (WT)

Gal4(1–147)

TT

S

Gal4–MEF2C (T293;300A)

Gal4(1–147)

AA

S

Gal4–MEF2C (S387A)

Gal4(1–147)

TT

A

Relative activity

(b)

350 Gal4 reporter 300 250 200 150 100 50 0

Gal4–MEF2C (WT): Gal4–MEF2C (T293;300A): Gal4–MEF2C (S387A): MyoD+E12: (c)

1

2

+

3 +

4 +

5

6

+

+

+

50 Gal4 reporter 45 40 35 30 25 20 15 10 5 0 1 2 3 Gal4–MEF2C (WT): + MyoD+E12: + pRb: +

7

8

+

+ +

+

Relative activity

phosphorylation by members of the p38 mitogen-activated protein (MAP) kinase family at three residues within MEF2C: Thr293, Thr300, and Ser387 (Figure 3a). In monocytic cells, mutation of these residues to alanine prevents MEF2C activation in response to either lipopolysaccharide stimulation or constitutive activation of the p38 kinase pathway [25]. Because these residues are required for enhanced MEF2C activity in monocytic cells, we tested whether MyoD activation of Gal4–MEF2C required the same amino acids. The combined mutation of both Thr293 and Thr300 to alanine (T293;300A) did not affect the basal activity of Gal4–MEF2C (Figure 3b, compare columns 3 and 5), but did enhance the activity of Gal4–MEF2C when MyoD+E12 were co-expressed (Figure 3b, compare columns 4 and 6). In contrast, mutation of Ser387 to alanine (S387A) decreased the basal activity of Gal4–MEF2C and completely eliminated its activation by co-transfected MyoD+E12 (Figure 3b, columns 7,8). Importantly, activation of the p38 kinase pathway did stimulate the function of the S387A Gal4–MEF2C mutant (data not shown; see also [25,27]), indicating that this mutation does not simply inactivate the protein. Therefore, MyoD activation of the MEF2C TAD requires Ser387, but neither Thr293 nor Thr300, of MEF2C.

453

4 + +

5 + +

6 + + +

Current Biology

MyoD requires pRb to activate the MEF2C transactivation domain. (a) The Gal4–MEF2C mutants used in the subsequent panels. Gal4–MEF2C (T293;300A) contains two threonine residues within the MEF2C transactivation domain that have been substituted with alanine; Gal4–MEF2C (S387A) has a single serine-to-alanine change. (b) Rb-positive 10T½ fibroblasts were transfected with a reporter construct containing five Gal4-binding sites upstream of luciferase either in the absence or the presence of MyoD and E12 plus the following Gal4–MEF2C fusion proteins, as indicated: Gal4–MEF2C (WT), Gal4–MEF2C (T293;300A); Gal4–MEF2C (S387A). Reporter activity relative to that observed with empty plasmid vectors (column 1) is displayed. Error bars represent duplicate plates from a representative experiment. (c) The Gal4–luciferase reporter was transfected into Rb–/– fibroblasts either in the absence or the presence of Gal4–MEF2C (WT) plus empty vectors, MyoD and E12, and pRb, as indicated.

allowing the cells to differentiate under low serum conditions, we added mitogen-rich medium containing the thymidine analog bromodeoxyuridine (BrdU) and evaluated the frequency of BrdU uptake in MCK–CAT-positive

454

Current Biology, Vol 9 No 9

Figure 4

(a)

180

MCK–CAT (b)

MCK promoter

Relative activity

160

BrdU (c)

DAPI

0.0

pRb

140 120

(e)

(f)

(g) 69.4

MEF2C–VP16

100

BrdU+/CAT+ (%)

(d)

80 (h)

60

(i)

(j) 3.7

E2F1–pRb(SP)

40 20 0 1 MyoD+E12: pRb: MEF2C: MEF2C–VP16: E2F1–pRb(SP):

2 +

3 + +

4 +

5 +

6 +

7 +

+

+ +

MEF2C–VP16 + E2F1–pRb(SP)

(k)

(l)

(m) 6.1

+ +

Current Biology

High-level expression of an MCK reporter gene requires both activation of MEF2 and cell-cycle arrest. (a) Rb–/– fibroblasts were transfected with an MCK–luciferase reporter construct and either empty plasmid vectors only (column 1), or expression vectors encoding MyoD and E12 plus either pRb, MEF2C (WT), MEF2C–VP16, E2F1–pRb(SP), or the combination of MEF2C–VP16 and E2F1–pRb(SP), as indicated. Cells were allowed to differentiate for 2 days and subsequently assayed for luciferase activity. (b–m) Rb–/– fibroblasts were transfected with an MCK–CAT reporter construct and MyoD plus either (b–d) pRb, (e–g) MEF2C–VP16, (h–j) E2F1–pRb(SP), or (k–m) the combination of MEF2C–VP16 and E2F1–pRb(SP). After

differentiation, cells were stimulated with serum-rich medium containing BrdU for 24 h. Cells were subsequently fixed and immunostained for (b,e,h,k) CAT expression (green) and (c,f,i,l) BrdU incorporation (red). (d,g,j,m) Nuclei were visualized by DAPI staining (blue). The proportion of MCK–CAT-positive cells containing BrdU staining is displayed as a percentage in the right-hand column and was derived from three independent experiments. The total number of MCK–CAT-positive cells counted for each condition were: pRb, 1006; MEF2–VP16, 396; E2F1–pRb(SP), 134; and MEF2–VP16 + E2F1–pRb(SP), 475. In (b–m), nuclei in MCK–CAT-expressing cells are indicated by arrowheads.

cells. MyoD expression alone did not arrest the growth of pRb-deficient cells or induce significant levels of CAT expression (data not shown), consistent with an inability of the MCK promoter to be activated in cycling cells. In contrast, co-expression of MyoD with pRb produced numerous cells with strong CAT staining and a complete lack of BrdU incorporation (Figure 4b–d). When the same analysis was applied to cells co-transfected with MyoD and MEF2C–VP16, we found many CAT-positive cells in the culture. The intensity of CAT staining in these cells was generally weaker than in those expressing pRb, however, and strikingly, the majority of the MCK–CAT-positive cells had incorporated BrdU (Figure 4e–g). By combining MyoD with an activated form of MEF2, therefore, we were able to dissociate activation of the MCK promoter from permanent cell-cycle exit in pRb-deficient cells. We did, nevertheless, find a considerable number of cells in the culture transfected with the combination of MyoD and MEF2C–VP16 that contained high amounts of CAT staining and no BrdU incorporation (data not shown). This finding suggested that, although cells could initiate the later stages of muscle differentiation when MyoD was coexpressed with MEF2C–VP16, a cell-cycle arrest might still be necessary to achieve the highest levels of musclegene expression.

The myogenic functions of pRb can be mimicked by a combination of MEF2 activation and cell-cycle arrest

Given that pRb can both inhibit proliferation and support the expression of muscle-differentiation-specific genes, we reasoned that MEF2C–VP16 might be less effective than pRb in promoting muscle differentiation because MEF2–VP16 cannot stop the cell cycle. To investigate whether blocking the cell cycle could augment myogenesis, we co-transfected pRb-deficient cells with the combination of MyoD+E12 and E2F1–pRb(SP), a chimeric protein consisting of the DNA-binding domain from the transcription factor E2F1 (amino acids 1–368) fused to the ‘small pocket’ (SP) domain of pRb (amino acids 379–792) [28]. The pRb(SP) domain exhibits potent transcriptional silencing activity when fused to DNA-binding proteins, but does not stimulate myogenesis on its own ([28] and data not shown). This E2F1–pRb(SP) chimera has previously been shown to mimic the anti-proliferative functions of pRb by binding to E2F promoter sites and repressing the expression of E2F-responsive genes [28]; yet it fails to fully substitute for pRb in the activation of differentiation programs [15]. Here, the co-expression of MyoD+E12 with E2F1–pRb(SP) increased MCK-promoter activity to 30% of that achieved following co-expression of MyoD+E12 with pRb (Figure 4a, column 6; Figure 4h),

Research Paper Requirement for pRb in skeletal muscle differentiation Novitch et al.

455

Figure 5 High-level expression of endogenous muscle genes requires both activation of MEF2 and cell-cycle arrest. Rb–/– fibroblasts were transfected with MyoD plus either (a–c) empty expression vector, (d–f) pRb, (g–i) MEF2C–VP16, (j–l) E2F1–pRb(SP), or (m–o) the combination of MEF2C–VP16 plus E2F1–pRb(SP). After 2 days under differentiation conditions, cells were fixed and immunostained for expression of (a,d,g,j,m) myosin heavy chain (MHC; green) and (b,e,h,k,n) muscle creatine kinase (MCK; red). (c,f,i,l,o) Nuclei were visualized by DAPI staining (blue). At least 300 myosin-positive cells were evaluated for co-expression of MCK in each panel. Representative fields are shown.

MHC

MCK

DAPI

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

Vector

pRb

MEF2C–VP16

E2F1–pRb(SP)

MEF2C–VP16 + E2F1–pRb(SP)

Current Biology

even though this fusion protein efficiently prevented BrdU uptake in MCK–CAT-positive cells (Figure 4h–j). As E2F1–pRb(SP) retained the anti-proliferative property of pRb but did not fully activate myogenesis on its own, we tested whether it could cooperate with MEF2C–VP16 in our myogenesis assays. The combination of MyoD+E12, MEF2C–VP16, and E2F1–pRb(SP) activated MCK–luciferase expression to a level equal to or greater than that achieved following co-transfection of MyoD+E12 with pRb (Figure 4a, compare columns 3 and 7), and led to the appearance of numerous cells in the culture that displayed very high levels of CAT staining and no BrdU incorporation (Figure 4k–m), very similar to what was observed following transfection of MyoD with pRb (Figure 4b–d). Cotransfection of E2F1–pRb(SP) and MEF2C–VP16 in the absence of MyoD did not induce the expression of MCK–CAT (data not shown). Thus, we were able to fully substitute for the loss of pRb by cotransfecting MyoD with an activated form of MEF2 and the cell-cycle-arresting molecule E2F1–pRb(SP).

muscle genes. In pRb-deficient fibroblasts, MyoD transfection by itself induced low amounts of MHC expression and little or no detectable MCK (Figure 5a–c). In contrast, cells transfected with both MyoD and pRb produced high levels of MHC and MCK (Figure 5d–f). Cotransfection of MyoD with either MEF2C–VP16 or E2F1–pRb(SP) alone led to a relatively small increase in the intensity of MHC staining and low amounts of endogenous MCK expression (Figure 5g–l). However, the combination of MyoD, MEF2C–VP16, and E2F1–pRb(SP) produced cells that expressed very high levels of both MHC and MCK (Figure 5m–o), again mimicking the phenotype of cells transfected with MyoD and pRb. In summary, these results indicate that the ability of pRb to promote muscle differentiation can be fully substituted by the combination of E2F1–pRb(SP) and MEF2C–VP16, suggesting that the requirement for pRb to promote expression of late muscle-differentiation markers reflects the ability of this molecule both to arrest cells in G0 and to potentiate MyoD-mediated activation of the MEF2 TAD.

Because our results indicated that the combination of MEF2C–VP16 and E2F1–pRb(SP) could support the activity of MCK-reporter constructs as efficiently as could pRb, we further evaluated the ability of these two proteins to synergistically activate the expression of endogenous

Discussion Suppression of cell-cycle progression and activation of MEF2 synergistically activate myogenesis

In this study, we have determined that the block in terminal muscle differentiation observed in Rb–/– cells corresponds to

456

Current Biology, Vol 9 No 9

an inability of MyoD to activate MEF2 function. When pRb is absent, MyoD induces normal levels of nuclearlocalized MEF2 protein that is fully competent to bind DNA yet transcriptionally inert. Activation of MEF2 function requires both MyoD and pRb and correlates with the induction of the transactivation properties of the MEF2C TAD. The requirement for pRb to promote muscle-gene expression can be partially bypassed by co-expressing MyoD with an artificially activated form of MEF2C. Under these conditions, late muscle-gene expression can be activated in replicating cells. Nevertheless, the precocious activation of both the MCK reporter and endogenous late muscle-differentiation markers is significantly augmented by agents that block the cell cycle, such as pRb or E2F1–pRb(SP). Together, these findings indicate that high-level expression of late muscle-differentiation markers requires both activation of MEF2 TAD function as well as a G0 arrest, and argue that pRb plays an essential role during terminal muscle differentiation by regulating both of these processes (summarized in Figure 6). It has been observed previously that ectopic expression of either full-length E2F1 or the E2F1 DNA-binding domain alone can potently inhibit the differentiation of wild-type myoblasts [29–31], suggesting that suppression of E2F activity may be a critical step in myogenic differentiation. Here, we have found that an E2F1–pRb(SP) chimeric protein that efficiently suppresses E2F-responsive promoters and arrests cellular growth can only partially substitute for pRb in driving the expression of either a transfected MCK reporter or endogenous late muscledifferentiation markers in Rb–/– fibroblasts. Thus, whereas suppression of E2F activity and G0 arrest are normally necessary for myogenesis, the cell-cycle-regulatory activities conveyed by the E2F1–pRb(SP) chimeric protein are apparently not sufficient to promote complete myogenic

differentiation. This finding suggests that pRb may contain a differentiation-promoting activity that is distinct from its anti-proliferative functions. This idea is substantiated by the finding that pRb mutants that are unable to associate stably with E2F and are defective in cell-cycle regulation are nevertheless able to stimulate cellular differentiation programs [15]. Given our results, a second muscle-differentiation-promoting activity of pRb can be accounted for by the requirement for this factor to synergistically activate the function of the MEF2C TAD in combination with MyoD. MyoD modulates MEF2 function through two independent pathways

The finding that expression of MyoD induces the transcriptional-activation function of the MEF2C TAD reveals a novel mechanism by which MyoD may positively affect MEF2 activity and muscle-gene expression. Previous work has suggested that synergism between these two classes of transcription factors is mediated through direct protein–protein contact between the bHLH domain of MyoD and the MADS/MEF2 domain of MEF2 [2,24,26]. Importantly, the ability of MyoD to activate MEF2C in our experiments did not require the MADS/MEF2 domain, suggesting that MyoD can affect MEF2 function without a need for direct protein binding. Although our results cannot exclude the possibility that a physical interaction between MyoD and MEF2 facilitates the activation process, Olson and colleagues [32] have recently demonstrated that an interaction between MyoD and MEF2 can be uncoupled from MEF2 activation by employing MyoD basic-region mutants that are unable to functionally activate MEF2 yet can still physically interact with it [32]. It is notable that these MyoD basic-region mutants are also unable to establish the myogenic program [33–35], raising the possibility that the ability of MyoD to

Figure 6

MyoD or Myf-5

Myogenin MEF2 p21

Late muscle-specific gene expression

MEF2 TAD function

pRb

E2F

G0 growth arrest

Current Biology

MyoD and pRb synergistically activate the function of the MEF2 TAD. MyoD can induce expression of early muscle-differentiationspecific genes such as myogenin, p21, and MEF2 in the absence of pRb; pRb is then required both to provide a permissive state for MyoD to activate the function of the MEF2 TAD and to block E2F transcriptional activity.

Research Paper Requirement for pRb in skeletal muscle differentiation Novitch et al.

stimulate the MEF2 TAD may underlie the ability of MyoD to recruit cells to the myogenic fate. Substitution of the MEF2 TAD with that of VP16 produced an activated MEF2 protein that did not require pRb for its function. When combined with MyoD, MEF2C–VP16 was able to activate expression of latestage differentiation markers in cells lacking pRb. In contrast, a MyoD–VP16 chimeric protein, which could strongly activate expression of a multimerized E-box reporter construct in pRb-deficient cells, was unable to activate expression of a MEF2 reporter construct or to induce the expression of late-stage differentiation markers without pRb (data not shown). These findings emphasize the requirement for an activated TAD within MEF2 and suggest that direct association of wild-type MEF2 with MyoD–VP16 by a ‘piggy-back’ interaction [2,24] is not sufficient either to convey transcriptional activity to MEF2 or to induce late muscle-gene expression in the absence of pRb. Although it is currently unclear how MyoD activates the function of the MEF2 TAD, our findings indicate that this process requires both pRb and the Ser387 in the MEF2 TAD. Given that phosphorylation of this serine by MAP-kinase family members can increase the activity of the MEF2 TAD [25,27], one possibility that we are currently investigating is that MyoD indirectly induces the phosphorylation of this residue in a pRb-dependent manner. Interestingly, the ectopic expression of the broad-specificity MAP-kinase phosphatase MKP-1 has recently been shown to interfere with the fusion of myocytes into myotubes [36], a phenotype reminiscent of the fusion defect seen in pRb-deficient myocytes [12]. It is tempting to speculate that this effect may result from the ability of the phosphatase to inhibit kinases that normally stimulate MEF2 TAD activity. pRb controls entry into the later stages of the muscle differentiation program

Myogenesis, like many other differentiation programs, entails a coordination of cellular growth arrest and activation of differentiation-specific gene expression. Here, we have found that the coupling of these two processes is brought about because they are regulated by the same protein, pRb. In Rb–/– myocytes, genes that are normally induced in proliferating cells within the first 24 hours of muscle differentiation, such as myogenin and p21 [4], are expressed to levels that are indistinguishable from those of wild-type cells. In contrast, genes that are expressed later in the differentiation program and never normally observed in dividing cells are either reduced or completely absent from cells lacking pRb [10,12]. It is interesting to note that MyoD can activate the expression of both myogenin and p21 in the absence of new protein synthesis [37,38], suggesting that at least the initial expression of these pRb-independent genes may occur

457

without the need for de novo MEF2 synthesis. In contrast, analysis of the promoters of late muscle genes, such as MCK and MRF-4, has indicated that an intact MEF2binding site, and therefore presumably MEF2 function, is required for their expression (reviewed in [2,3,18]). Thus, the absence of MEF2-dependent late-stage muscle-gene expression in pRb-deficient skeletal muscle is completely consistent with our findings that MEF2 transactivation function is specifically defective in these cells, and may explain why the expression of most muscle markers is normally confined to post-mitotic cells. These findings further indicate that agents that modulate pRb function, such as the G1 cyclins and Cdks, may act to control entry into the later stages of myogenesis by modulating MEF2 activity during normal embryonic development.

Materials and methods Cell culture and transfection Wild-type and Rb–/– mouse embryo fibroblasts isolated from day-13 embryos were kindly supplied by Tyler Jacks and cultured in medium (DMEM) supplemented with 10% fetal bovine serum (FBS), as described previously [10]. 10T½ fibroblasts were obtained from the American Type Culture Collection and propagated in the same growth medium. Fibroblasts were seeded at 1.5–1.75 × 105 cells per 60 mm dish the day before transfection, and lipofected using 2–3 µg total plasmid DNA and 10 µl Lipofectamine reagent (Life Technologies), as described previously [10]. Myogenic differentiation was induced by incubating the cells for 2–3 days in DMEM supplemented with 2% horse serum and 10 µg/ml insulin. BrdU uptake experiments were performed as previously described [10]. In these experiments, more than 80% of the differentiated myocytes incorporated BrdU. Differentiated myocytes were identified by their weak expression of MHC.

Plasmids and reporter-gene assays

For transactivation experiments, 0.5 µg of the following cytomegalovirus (CMV) expression plasmids were used: pCSA–MyoD [10]; pCSA–MyoD–VP16, created by subcloning a MyoD–VP16 fusion [35] into the EcoRI site of pCSA; pCS2–MT6–MyoD (Myc-tagged MyoD [39]); pCDNA1–MEF2C and pCDNA1–MEF2C(1–117)–VP16 [24]; pCDNA3–Gal4–MEF2C(TAD) constructs [25]; and pCS2–nuclear-βGal, provided by R. Rupp and D. Turner. In some experiments, 0.1 µg pCSA–E12 [39] was also included. These transactivators were combined with 0.5 µg pCMV–pRb or 0.05–0.1 µg of either pCMV–E2F1(1–368)–pRb(SP) or pCDNA1–HA–E2F1(1–368)–pRb(SP) fusion protein constructs [15,28]. The amount of expression plasmids encoding pRb and E2F1–pRb(SP) transfected in these experiments was determined by titrating the amount of each plasmid and assaying for the greatest effect on myogenesis. This amount corresponded to approximately two to three-fold greater expression of E2F1–pRb(SP) than pRb as measured by protein immunoblotting (data not shown). All transfections contained 0.5 µg of one of the following reporter constructs: MCK–luciferase, constructed by inserting the 3300 bp upstream enhancer and promoter sequence from the mouse MCK gene [40] into the plasmid pGL3–basic (Promega). Side-by-side comparisons indicate that both MCK–CAT [40] and MCK–luciferase constructs have identical promoter activities in both Rb–/– and 10T½ fibroblasts; E2F1-promoter luciferase [15,28]; E-box reporter, 4R–TK–CAT [41]; MEF2-reporter, MEF2x4–CAT [42]; Gal4–Luciferase, provided by A. Bonni and M. Greenberg. CAT assays were performed as described previously [10], and luciferase assays were performed using commercially available kits (Analytical Luminescence Laboratories and Promega).

Viral infection, EMSA, and immunoblotting Subconfluent wild-type and Rb–/– fibroblasts were infected with media containing ecotropic MyoD virus from transfected BOSC 23 cells as

458

Current Biology, Vol 9 No 9

described previously [10]. After 3 days’ culture under differentiation conditions, whole-cell extracts were prepared and used for EMSA and immunoblotting. The following antibodies were used: MyoD, monoclonal antibody 5.8A [43]; myogenin, monoclonal antibody F5D [44]; MHC, monoclonal antibody MF20 [45]; MCK, rabbit antiserum to mouse MCK (kindly provided by Steve Hauschka) [22]; MEF2, affinitypurified rabbit antibodies to MEF2 (C-21; Santa Cruz Biotechnology). Note that the Santa Cruz MEF2C rabbit antibodies were generated against human MEF2A, but they cross-react with both human and mouse MEF2A, MEF2C, and MEF2D proteins (data not shown).

Immunostaining of myocytes Cells were fixed and stained as described previously [10] using the following primary antibodies: Myc-tagged MyoD, monoclonal antibody 9E10; pRb, monoclonal antibody XZ-91 (Pharmingen); MEF2, affinitypurified rabbit antibodies to MEF2 (C-21); CAT, affinity-purified rabbit antibodies to CAT (5 Prime–3 Prime, Inc.); MHC, monoclonal antibody MF20; MCK, rabbit antiserum to mouse MCK; BrdU, monoclonal antibody G3G4 (Developmental Studies Hybridoma Bank). Nuclei were visualized by staining cells with 4',6'-diamidino-2-phenylindole (DAPI). Cells were examined and photographed using a Zeiss Axioskop microscope equipped with a 40× water-immersion objective.

Acknowledgements We thank Jean Buskin, Azad Bonni, Don Fischman, Mike Greenberg, Jiahuai Han, Steve Hauschka, Marshall Horowitz, Peter Houghton, Tyler Jacks, Bill Kaelin, James Martin, Jeffrey Molkentin, Eric Olson, Michael Perry, Ralph Rupp, Bill Sellers, David Turner, and Woody Wright for the generous gifts of reagents. We also thank Bill Kaelin, Bill Sellers, and members of the Lassar laboratory for helpful discussions. This work was supported by grants to A.B.L. from the National Science Foundation, the NIH, and the Council for Tobacco Research. This work was done during the tenure of an established investigatorship from the American Heart Association to A.B.L; B.G.N. was supported by a fellowship from the Albert J. Ryan Family Foundation. D.B.S. was supported by a fellowship from The Medical Foundation, Inc., Boston, Massachusetts.

References 1. Yun K, Wold B: Skeletal muscle determination and differentiation: story of a core regulatory network and its context. Curr Opin Cell Biol 1996, 8:877-889. 2. Molkentin JD, Olson EN: Defining the regulatory networks for muscle development. Curr Opin Genet Dev 1996, 6:445-453. 3. Ludolph D, Konieczny S: Transcription factor families: muscling in on the myogenic program. FASEB J 1995, 9:1595-1604. 4. Andrés V, Walsh K: Myogenin expression, cell cycle withdrawal, and phenotypic differentiation are temporally separable events that precede cell fusion upon myogenesis. J Cell Biol 1995, 132:657-666. 5. Halevy O, Novitch BG, Spicer DB, Skapek SX, Rhee J, Hannon GJ, et al.: Correlation of terminal cell cycle arrest of skeletal muscle with induction of p21 by MyoD. Science 1995, 267:1018-1021. 6. Walsh K, Perlman H: Cell cycle exit upon myogenic differentiation. Curr Opin Genet Dev 1997, 7:597-602. 7. Lassar AB, Skapek SX, Novitch B: Regulatory mechanisms that coordinate skeletal muscle differentiation and cell cycle withdrawal. Curr Opin Cell Biol 1994, 6:788-794. 8. Olson EN: Interplay between proliferation and differentiation within the myogenic lineage. Dev Biol 1992, 154:261-272. 9. Maione R, Amati P: Interdependence between muscle differentiation and cell cycle control. Biochim Biophys Acta 1997, 1332:M19-M30. 10. Novitch BG, Mulligan GJ, Jacks T, Lassar AB: Skeletal muscle cells lacking the retinoblastoma protein display defects in muscle gene expression and accumulate in S and G2 phases of the cell cycle. J Cell Biol 1996, 135:441-456. 11. Schneider JW, Gu W, Zhu L, Mahdavi V, Nadal-Ginard B: Reversal of terminal differentiation mediated by p107 in Rb–/– muscle cells. Science 1994, 264:1467-1471. 12. Zacksenhaus E, Jiang Z, Chung D, Marth JD, Phillips RA, Gallie BL: pRb controls proliferation, differentiation, and death of skeletal muscle cells and other lineages during embryogenesis. Genes Dev 1996, 10:3051-3064.

13. Wang J, Guo K, Wills KN, Walsh K: Rb functions to inhibit apoptosis during muscle differentiation. Cancer Res 1997, 57:351-354. 14. Sellers WR, Kaelin WG: RB as a modulator of transcription. Biochim Biophys Acta 1996, 1288:M1-M5. 15. Sellers WR, Novitch BG, Miyake S, Heith A, Otterson GA, Kaye FJ, et al.: Stable binding to E2F is not required for the retinoblastoma protein to activate transcription, promote differentiation, and suppress tumor cell growth. Genes Dev 1998, 12:95-106. 16. Lassar AB, Davis RD, Wright WE, Kadesh T, Murre C, Voronova A, et al.: Functional activity of myogenic HLH proteins requires hetero-oligomerization with E12/E47-like proteins in vivo. Cell 1991, 66:305-315. 17. Cserjesi P, Olson EN: Myogenin induces the myocyte-specific enhancer binding factor MEF-2 independently of other musclespecific gene products. Mol Cell Biol 1991, 11:4854-4862. 18. Olson EN, Perry M, Schulz RA: Regulation of muscle differentiation by the MEF2 family of MADS box transcription factors. Dev Biol 1995, 172:2-14. 19. Ranganayakulu G, Zhao B, Dokidis A, Molkentin JD, Olson EN, Schulz RA: A series of mutations in the D-MEF2 transcription factor reveal multiple functions in larval and adult myogenesis in Drosophila. Dev Biol 1995, 171:169-181. 20. Lilly B, Zhao B, Ranganayakulu G, Paterson BM, Schulz RA, Olson EN: Requirement of MADS domain transcription factor D-MEF2 for muscle formation in Drosophila. Science 1995, 267:688-693. 21. Bour BA, O'Brien MA, Lockwood WL, Goldstein ES, Bodmer R, Taghert PH, et al.: Drosophila MEF2, a transcription factor that is essential for myogenesis. Genes Dev 1995, 9:730-741. 22. Chamberlain JS, Jaynes JB, Hauschka SD: Regulation of creatine kinase induction in differentiating mouse myoblasts. Mol Cell Biol 1985, 5:484-492. 23. De Angelis L, Borghi S, Melchionna R, Berghella L, Baccaranicontri M, Parise F, et al.: Inhibition of myogenesis by transforming growth factor beta is density-dependent and related to the translocation of transcription factor Mef2 to the cytoplasm. Proc Natl Acad Sci USA 1998, 95:12358-12363. 24. Molkentin JD, Black BL, Martin JF, Olson EN: Cooperative activation of muscle gene expression by MEF2 and myogenic bHLH proteins. Cell 1995, 83:1125-1136. 25. Han J, Jiang Y, Li Z, Kravchenko VV, Ulevitch RJ: Activation of the transcription factor MEF2C by the MAP kinase p38 in inflammation. Nature 1997, 386:296-299. 26. Kaushal S, Schneider JW, Nadal-Ginard B, Mahdavi V: Activation of the myogenic lineage by MEF2A, a factor that induces and cooperates with MyoD. Science 1994, 266:1236-1240. 27. Kato Y, Kravchenko VV, Tapping RI, Han J, Ulevitch RJ, Lee J-D: BMK1/ERK5 regulates serum-induced early gene expression through transcription factor MEF2C. EMBO J 1997, 16:7054-7066. 28. Sellers WR, Rodgers JW Kaelin WG: A potent transrepression domain in the retinoblastoma protein induces a cell cycle arrest when bound to E2F sites. Proc Natl Acad Sci USA 1995, 92:11544-11548. 29. Guo K, Walsh K: Inhibition of myogenesis by multiple cyclin-cdk complexes. J Biol Chem 1997, 272:791-797. 30. Wang J, Huang Q, Tang W, Nadal-Ginard B: E2F1 inhibition of transcription activation by myogenic basic helix-loop-helix regulators. J Cell Biochem 1996, 62:405-410. 31. Wang J, Helin K, Jin P, Nadal-Ginard B: Inhibition of in vitro myogenic differentiation by the cellular transcription factor E2F-1. Cell Growth Diff 1995, 6:1299-1306. 32. Black BL, Molkentin JD, Olson EN: Multiple roles for the MyoD basic region in transmission of transcriptional activation signals and interaction with MEF2. Mol Cell Biol 1998, 18:69-77. 33. Davis RL, Weintraub H: Acquisition of myogenic specificity by replacement of three amino acids from MyoD into E12. Science 1992, 256:1027-1030. 34. Davis RL, Cheng P-F, Lassar AB, Weintraub H: The MyoD DNA binding domain contains a recognition code for muscle-specific gene activation. Cell 1990, 60:733-746. 35. Weintraub H, Dwarki V, Davis R, Hollenberg S, Snider L, Lassar A, et al.: Muscle-specific transcriptional activation by MyoD. Genes Dev 1991, 5:1377-1386. 36. Bennett AM, Tonks NK: Regulation of distinct stages of skeletal muscle differentiation by mitogen-activated protein kinases. Science 1997, 278:1288-1291.

Research Paper Requirement for pRb in skeletal muscle differentiation Novitch et al.

37. Otten AD, Firpo EJ, Gerber AN, Brody LL, Roberts JM, Tapscott SJ: Inactivation of MyoD-mediated expression of p21 in tumor cell lines. Cell Growth Diff 1997, 8:1151-1160. 38. Hollenberg SM, Cheng PF, Weintraub H: Use of a conditional MyoD transcription factor in studies of MyoD trans-activation and muscle differentiation. Proc Natl Acad Sci USA 1993, 90:8028-8032. 39. Spicer DB, Rhee J, Cheung WL, Lassar AB: Inhibition of myogenic bHLH and MEF2 transcription factors by the bHLH protein Twist. Science 1996, 272:1476-1480. 40. Jaynes J, Johnson J, Buskin J, Gartside C, Hauschka S: The muscle creatine kinase gene is regulated by multiple upstream elements, including a muscle-specific enhancer. Mol Cell Biol 1988, 8:62-70. 41. Weintraub H, Davis R, Lockshon D, Lassar A: MyoD binds cooperatively to two sites in a target enhancer sequence: occupancy of two sites is required for activation. Proc Nat Acad Sci USA 1990, 87:5623-5627. 42. Wong M-W, Pisegna M, Lu MF, Leibham D, Perry M: Activation of Xenopus MyoD transcription by members of the MEF2 protein family. Dev Biol 1994, 166:683-695. 43. Dias P, Parham D, Shapiro D, Webber B, Houghton P: Myogenic regulatory protein (MyoD1) expression in childhood solid tumors. Amer J Path 1990, 137:1283. 44. Cusella-De Angelis MG, Lyons G, Sonnino C, De Angelis L, Vivarelli E, Farmer K et al.: MyoD, myogenin independent differentiation of primordial myoblasts in mouse somites. J Cell Biol 1992, 116:1243-1255. 45. Bader D, Masaki T, Fischman DA: Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and in vitro. J Cell Biol 1982, 95:763-770.

Because Current Biology operates a ‘Continuous Publication System’ for Research Papers, this paper has been published on the internet before being printed. The paper can be accessed from http://biomednet.com/cbiology/cub — for further information, see the explanation on the contents page.

459