Growth Factors That Repress Myoblast Differentiation Sustain ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY (e) 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 3, Issue of January 25, pp. 1580-1585,1993 Printed in U.S.A.

Growth Factors That Repress Myoblast Differentiation Sustain Phosphorylation of a Specific Site on HistoneH l * (Received for publication, July 24, 1992)

Francesca Cole, Thomas M. Fasy, Sunkara S. Rao, Mary Anne dePeralta, and D. Stave KohtzS From the Department of Pathology, Mount SinaiSchool of Medicine, New York, New York 10029

A monoclonalantibody (12D11) is presented that binds an epitope on histone H1 only when it is phosphorylated. Skeletal myoblasts, which are cultured in high mitogen medium to induce proliferation and inhibit differentiation, contain histone H1 reactive with monoclonalantibody 12D11, whereas differentiated myocytes and adultskeletal muscle donot. Phosphorylation of H1 at the 1 2 D l l epitope, as assessed by antibody binding, is also induced and maintained in myoblasts cultured in low mitogen medium supplemented with transforming growth factor , B,which blocks differentiation but allows the cells to withdraw from the cell cycle (Olson, E., Sternberg, E., Hu, J., Spizz, G., and Wilcox, C. (1986) J. Cell Biol. 103, 1799-1805; Massague, J., Cheipetz, S., Endo, T., and Nadal-Ginard, B. (1986) Proc. Natl. Acad. Sci. U. S. A . 83, 8206-8210). These observations suggest that phosphorylation of histone H1 at the 1 2 D l l epitope is associated with the negative regulation ofmyoblast differentiation by growth factors.

of growth factors is reduced, are being actively investigated. The discovery of myogenic determination genes, which when transfected and constitutively expressed in cells of other phenotypes induce trans-determination into myoblasts, has provided a focus for understanding the regulation of differentiation. Four distinct myogenic determination genes have been described MyoDl (Daviset al., 1987), myogenin(Wright et al., 1988),myf-5 (Braun et al., 1989) and MRF4/myf-6/ herculin (Rhodes and Koznieczny, 1989; Braun et al., 1990; Miner and Wold, 1990). All four share a 68 residue DNAbinding motif referred to as a basic-helix-loop-helix (Murre et al., 1989a), also commonto an extended family of transcription factors relevant todevelopment. The myogenic determination proteins binda consensus sequence found oneor more times in the enhancer elements of muscle-specific genes (Buskin and Hauschka, 1989; Lassar et al., 1989; Murre et al., 198913; Lassar et al., 1991), and thereby trans-activate transcription of these genes in differentiated myocytes. Since the expression of myogenic determination genes can occur either natively or through forced expression inundifferentiated myoblasts (which do not express most muscle-specific genes) as well as in differentiatedmyofibers, their ability to activate Differentiation of proliferating myoblasts into muscle is transcription of muscle-specific genes must be regulated so inhibited in culture by components of serum (Konigsberg, that these genes are expressed only in differentiated cells. In 1963;Konigsberg et al., 1978) or embryoextracts(Slater, vivo footprint studies of the endogenous muscle creatine ki1976). Some inhibitorshave been identified as growth factors, nase gene enhancer haveshown that the MyoD1 binding and include basic fibroblast growth factor (Gospodarowicz et sequence is unoccupied in myoblasts while it is occupied in al., 1976; Linkhart et al., 1980; Clegg et al., 1987) and trans- differentiated muscle fibers (Mueller and Wold, 1989). It is forming growth factor(TGF-P)’ (Olson et al., 1986; Ma- unlikely that this difference is due to post-translationalmodssague et ul., 1986). Homogeneous preparations of these fac- ifications of the MyoDl protein, since MyoDl isolated from tors, either aloneor in combination, inhibit differentiationof myoblasts or from myofibers displays equivalent binding acmyoblasts in culture, and their subsequentremoval results in tivity in gel shift assays (Lassaret al., 1989, 1991). Regulation spontaneous morphological and biochemical differentiation of MyoDl activitymay be explainedby the modulating action of the myoblasts into skeletal muscle fibers. Myoblasts cul- of c-jun (or junB),which inhibits myogenesis (when overextured in low mitogen medium at confluence in the presence pressed) and formsa complex with MyoDl in vivo (Bengal et of transforming growth factor P are mitogenically arrested al., 1992; Li et al., 1992). An alternate or additional regulatory but remain undifferentiated until the factor is removed (Olson mechanism, suggested by the present results, isbased on the et al., 1986), indicating that inhibition of myogenic differen- hypothesis that theaccessibility of muscle-specific enhancers tiation by TGF-0 is not inextricably linked to its activity as differs between myoblasts and differentiatedmyofibers. a mitogen. The molecular mechanisms which determine that Transcriptional repression by histone H1 is thought to be proliferating myoblasts will eventually differentiate into mus- a central mechanismof regulating eukaryotic gene expression. cle, but defer differentiation until the ambient concentration HistoneH1 represses transcription invitro (discussedin Weintraub (1985) and Felsenfeld (1992)); the ability of certain * This work was supported by National Institutesof Health Grants trans-activators toovercome this effect is referred to as antiHL40659 and HL43583. The costs of publication of this article were repression (Croston et al., 1991). Although many studieshave defrayed in part by the payment of page charges. This article must mediates the condensation of chrotherefore be hereby marked “aduertisement” in accordance with 18 indicated that histone H1 matin (Bradbury et al., 1974; Allan et al., 1980; Staynov et al., U.S.C. Section 1734 solely to indicate this fact. $ To whom correspondence should be addressed: Dept. of Pathol1988), the mechanism by which histone H1 represses tranogy, 1 Gustave Levy Place, Mount Sinai School of Medicine, New scription is not clear. The additionof histone H1 toXenopus York, NY 10029. Tel.: 212-241-9169; Fax: 212-534-7491. chromatin in vitro can selectively repress oocyte 5 S RNA The abbreviations used are: TGF-p, transforming growth factor genes whereas somatic 5 S RNA genes remain transcription/$; HUSK, primaryhumanskeletal myoblasts; sMHC, sarcomeric ally active (Schlissel and Brown, 1984; Wolffe, 1989). In this myosin heavy chain; mAb, monoclonal antibody: PAGE, polyacrylRNA polymerase I11 system, hyperphosphorylation of H1 by amide gel electrophoresis; MES, 4-morpholineethanesulfonic acid.

1580

Myogenesis Phosphorylation and Histone

1581

growth-associated H 1 kinase decreased the potency of H1 as full strength for immunofluorescence (medium for culture supernaa transcriptional inhibitor (Jerzmanowski and Cole, 1990). tants was supplemented with 20 mM HEPES, pH 7.4). Monoclonal antibody MF-20 (Bader et al., 1982) was grown as an Transcriptional repression by histone H 1 i n vitro does not ascites tumor in nude mice, and the ascites fluid was diluted 1:lOOO necessarily require assembled chromatin as a template. Puri- in TBS supplemented with 5%horse serum, 5% nonfat dry milk, and fied H1 is capable of repressing transcription from several 0.2% Tween 20) for Western blotting andimmunofluorescence applinaked DNA templates, including Drosophila jockey, alcohol cations. Monoclonal antibody MRA12, an IgG2a that binds generically to dehydrogenase (proximal), Drosophila Ultrabithorax, Drosophila Kruppel, adenovirusmajor late, andE4 promoters (see histone H1 (Monestier et al., 1989), waspurified,lyophilized, and provided by M. Monestier. It was used at a final concentration of 5 Croston et al. (1991) and references within). pg/ml (in the samebuffer as MF-20) for Western blotting. We have generated a monoclonal antibody (mAb 1 2 D l l ) Immunofluorescence and Western Blotting-Cells were cultured on that binds a phosphorylated subsetof histone H1. Binding of 35-mm culture plates for immunofluorescence analyses. Cells were mAb 12Dll is abolished by treating histone H1 with bacterialwashed twice at ambient temperature with phosphate-buffered saline alkaline phosphatase. This antibody aispowerful and unique ((PBS) 100 mM Na2HP04/NaH2PO*, pH 7.4, 154 mM NaCl), fixed rnin). The cells tool for in situ analysis of the phosphorylation state of a with freshly made 3.7% paraformaldehyde in PBS (15 specific site on histone H1,designated the 12Dll epitope. were then treated with0.2% Triton X-100 in PBS, washed once with PBS, then saturated for 30 min with 10% horse serum (EIA grade, Phosphorylation of histone H 1 at the 12Dll epitope is ob- GIBCO), all at ambient temperature. Hybridoma supernatant from served at high levels in skeletalmyoblasts, but is notobserved mAb 12Dll was applied directly to the cells, then incubated for 1 h i n differentiated myocytes. Mitogenically arrested myogenic at 37 "C. The cells were then washed twice, and incubated with as quiescent mononuclear fluorescein isothiocyanate-conjugated goat anti-mouse (Boehringer cells that are undifferentiated, such cells in differentiated cultures and myoblasts cultured in low Mannheim) diluted1:50 in 0.1% bovine serum albumin-supplemented mitogen medium supplemented with TGF-P, also maintain PBS, 45 min a t 37 "C. The cells were then washed three times with PBS and viewed on an inverted Nikon Diaphot fluorescence microhigh levels of histone H 1 phosphorylated at the 12Dll epi- scope. tope. In this reportwe present evidence that phosphorylation Equal cell numbers (or nuclei in fused cultures) were used for all of histone H1 ata specific site(s) is associated withregulation comparisons. Experiments were always performed sequentially with of myoblast differentiation. Phosphorylation at this site ap- identical starting cultures. Cell lysates for SDS-PAGE (sodium dopears to be a downstream event which follows exposure to decyl sulfate- polyacrylamide gel electrophoresis) were generated by growth factors that inhibit myogenic differentiation, and thus washing monolayers with PBS and adding sample buffer (100 mM Tris, pH, 6.8; 2% SDS (w/v), 0.25% phenol red (w/v), 10% glycerol, this inhibition may be linked to the regulation of transcrip- 5% (v/v) mercaptoethanol) directly to the tissue culture plates. Each tional repression. lysate was inspected for anomalies by SDS-PAGE and Coomassie stain beforebeingused for blotting experiments. Isolated protein samples were dissolved in an equalvolume of sample buffer. Samples Cell Culture-C2C12 myoblasts (Blau et al., 1983) were grown in were heated a t 100 "C priorto electrophoresis oneither9%(for myosin heavy chain) or 18% polyacrylamide gels (for histone H1) Dulbecco's modified Eagle's medium supplemented with 15% fetal using the discontinuous buffer system described previously (Laemmli, bovine serum (HyClone), 1% chick embryo extract (GIBCO), minimum Eagle's medium nonessential amino acids (GIBCO), minimum 1970). The gels were transferred tonitrocellulose for Western blotting (as described byBio-Rad, the manufacturer of the transfer apparatus) Eagle's medium vitamins (GIBCO), 1 mM sodium pyruvate, 100 pg/ ml L-glutamine, and 100 pg/ml gentamicin. To induce differentiation, or fixed with 10% acetic acid, 25% methanol and stained with Coomassie Blue (Sigma). Transfer efficiency was monitored with presimilar medium was used in which fetal bovine serum and embryo stained marker lanes. extract were replaced by4% horseserum.Humanfetalskeletal After protein transfer, nitrocellulose sheets were treated with 10% myoblasts (HUSK cells) were cultured as described before (Feghali for 10 min, then washed with several e t al., 1992). Hybridoma cells were grown in RPMI medium (GIBCO) acetic acid, 25% isopropanol containing the supplements listed above with 10% fetal bovine serum changes of water. The sheets were then incubated with gentle rotary were then incubated 8shaking for 8-18 h (4 "C) in TBS. The sheets (HyClone) and no embryo extract. TGF-8 was purchased from R & D Systems andused at a concen- 18 h (4 "C) in TBS supplemented with 10% Carnation nonfat dry milk. The sheets were washed with TBS containing 0.2% Tween 20 tration of 5 ng/ml in 4% horse serum medium to inhibit myoblast and incubated overnight with primary antibody (buffers for each differentiation. antibody are described above). The sheets were then washed twice Antibodies-To generate mAb 12Dl1, a crude nuclear pellet was with TBS containing 0.2% Tween 20 and incubated overnight (4'C) isolated frombovine brain: whole bovine brain washomogenized (Waring blender) in an equal volume buffer A (100mM MES, pH6.0, with 5 FCi of '251-labeled rabbitanti-mouse IgG (DuPont-New bufferdescribedabove for mAb MF-20 0.5 mM MgCI,, 5 mM mercaptoethanol, 1 mM EGTA, 5 pg/ml leupep- EnglandNuclear)inthe tine, 0.2 mM phenylmethylsulfonyl fluoride, 0.5 mM benzamidine). supplemented with0.5 M NaCl. The blots were washed (with shaking) was discarded, twice (4 "C, 20 min each) with TBS supplemented with0.2% Tween After centrifugation (300 X g, 4 "C, 10 min), the pellet and the supernatant was centrifuged again (5000 X g, 4 "C, 30 min). 20 and 0.5 M NaC1, once (4 "C) with TBS supplemented with 0.2% T h e second pellet was resuspended in an equalvolume of 20 mM Tris Tween 20, and once (ambient temperature) with TBS, then dried and (pH 7.5), and heated to 100 "C for 10 min. The suspension was then exposed to Kodak XAR-5 film usinga Du Pont Lightening plus centrifuged (100,000 X g, 4 "C,60 min), and the supernatantused as intensifying screen. Phosphatase Treatment of Histone HZ-Histone H1 was purified immunogen. This supernatant was also used a t a protein concentration of 5 pg/ml for enzyme-linked immunosorbent assay screeningof as previously described (Fasy etal., 1979; Monestier, et al. 1989) from bovine thymus,humanplacenta,andhumanskeletal muscle (obthe individual hybridoma supernatants. Immunizations, generation of the hybridomas, and initial screening of the hybridoma superna- tained a t autopsy). Phosphatase reactions were performed with putants by enzyme-linked immunosorbent assay were performed using rified H1 from human placenta in 10 mM Tris, pH 8.3, 1 mM MgC12, 1 mM ZnC12. Samples were incubatedwith 0.7 unit of bacterial standard methods, asdescribed previously (Kohtz etal., 1987). Monoclonal antibodies testing positive in the enzyme-linked immunosor- alkaline phosphatase for 45 min (37 "C).An additional 1.4 units were bent assaywere then assessed by immunofluorescence for subcellular then added, and the mixturewas then incubated for another 45 min. Sham reactions were performed in parallel without added enzyme. localization. Monoclonal antibodies testing positive for nuclear imThe histone H 1 was then isolated by precipitatingthebacterial munofluorescence were then assessed by Western blotting for reactivity with known nuclear antigens (see text). Monoclonal antibody alkaline phosphatase with an equal volume of cold 10% perchloric 1 2 D l l was selected initially with this assay for its binding to histoneacid and pelleting the precipitate in a microcentrifuge a t 13,000 rpm H1. The hybridoma cells secreting mAb 12Dll were subcloned twice for 10 min a t 4 "C. The supernatant was collected and either neuby limiting dilution. For Western blotting, culture supernatants were tralized with 2 M Tris, pH 8, or 12 N HC1 was added at a dilution of diluted 1:2.5 in Tris-buffered saline ((TBS) 20 mM Tris, pH 7.5, 100 1:50 to the sample and the histone was precipitated with 10 volumes mM NaC1) containing 0.2% Tween. Culture supernatants were used of acetone. MATERIALS ANDMETHODS

Myogenesis Phosphorylation and Histone

1582

displayed variable expression in different cell lines and tissues. For example, preparations of bovine thymus (Fig. l), Monoclonal Antibody 12011 Binds Exclusively to a Phos- bovine brain (not shown), and human placenta(Fig. 3, lanes phorylated Subset of Mammalian Histone HI Not Present in 2 and 4 ) all produced positive signals in Western blots with Skeletal Muscle-The hybridoma secreting mAb 1 2 D l l was mAb 12Dll. In contrast, preparations of histone H1 from generated by the fusion of splenocytes from a mouse immuadulthumanskeletal muscle did notcontaindetectable nized with a crude nuclear extract with Sp2/0 myeloma cells 12Dll+ histone H1 (Fig. 3, lanes 1 and 3 ) . (see "Material and Methods"). Monoclonal antibody 1 2 D l l Phosphorylation of Histone HI at the12011 Epitope Is wasanalyzed by Western blot for binding to a variety of Eliminated during Myoblast Differentiation-Expression of knownnuclear antigens(includingHMG1, 2, 14, and 17; 12Dll+ histone H1 in differentiatingC2C12 (murine; Blau et histones H1, H2A, H2B, H3, H4, and H5) and found to react al. (1983)) and human fetal primarymyoblasts (HUSK cells; exclusively with histone H1 (not shown).As shown in Fig. 1, Feghali et al. (1992)) was examined by Western blot (Fig. 4). mAb 12Dll reacts with a form of histone H1 that migrates Whole cell homogenates of myoblasts cultured in high mitoslower than thebulk of H1 during SDS-PAGE. To determinegen medium (0 point) and of myoblasts induced to differenif phosphorylation antigenically distinguishes this subset of tiate for increasing lengths of time in low mitogen medium histone H1 from other histone H1, human placental H 1 was were analyzed by Westernblot with mAbs MF-20 (antianalyzed by Western blot before and after treatment with sarcomeric myosin heavy chain; Bader etal. (1982)) andmAb bacterial alkaline phosphatase. As shown in Fig. 2 (lane 5), 12Dll. Homogenates from equal numbers of cells (or cell bacterial alkaline phosphatase treatmentof H1 eliminatedall nuclei in fused cultures) were compared. Sarcomeric myosin reactivity with mAb 12Dl1, without affecting the reactivity heavy chain (sMHC) proteinexpression was usedas a marker of mAb MRA 12 (which genericallyreacts with most formsof for myogenic differentiation, as the onset of its expression histone H1 (Monestier etal., 1989) (Fig. 2, lanes 7-9). These coincides with the formation of syncytia 24-96 h after myoresults show that mAb 12Dll reacts specifically with a phosblastsare switched to low mitogenmedium. In atypical phorylated subset of histone H1. experiment, HUSK cells produced high levels of sMHC after The phosphorylated subset of histone H1 recognized by 96 h in low mitogen medium (Fig. 4B), and C2C12 myoblasts mAb 12Dll (hereafter referred to as 12Dll+ histone H1) produced high levels after 48 h (Fig. 4B). The same set of samples was immunoblotted with mAb 12Dl1, revealing that A B 12Dll+ histone H1 expression started to diminish within 8 (C2C12) to 24 (HUSK) h after the cells were switched to low mitogen medium (Fig. 4.4). Corresponding to the induction of syncytia formation and high level sMHC expression, 12Dll+ histone H1 expression dropped to its lowest levels 48-96 h after switching the cells to low mitogen medium (Fig.4, C and 'I) TD). These experimentshave been repeated several times, and ,thetime periodfrom switchingthe cells to low mitogen medium to the onset of differentiation varied from 18 to 96 FIG. 1. Monoclonal antibody 1 2 D l l binds a subset of hish. Several factors produce this expected variability, including tone HI that migrates slower than the bulk of histone H1. Bovine thymus histone H1 (10 pg/lane)was resolved by SDS-PAGE cell density and variationsbetween different preparations of on a 15'70 polyacrylamide gel (migration from top to bottom). Lane A medium. In all cases, the onset of differentiation occurred as was stained for total protein with Coomassie Brilliant Blue; lane R 12Dll' histone H1 reached a minimum. was analyzed by Western blot with mAb 12Dll. Bar indicates the Immunofluorescence microscopy was performed with mAb position of antibody-boundhistoneH1 (12Dll+ histone Hl), and 1 2 D l l on HUSK cells. Only human, canine, and bovine cell d o t t d outline represents the positions of the bulk histone H1, as lines produced strong immunofluorescent signals with mAb determined after Amido Black staining of the blot. 12Dl1, whereas rodent cell lines, despite displaying strong binding in Western blot applications, produced weak immuA B C M A A B B C C nofluorescence signals.As shown inFig. 5R, HUSKmyoblasts . 58 48.5 in high mitogen medium displayed positive nuclear staining with mAb 12Dll. Skeletal myoblasts induced to differentiate .36.5 in culture by low mitogen medium produce two populations - .-r= . 26.5 of cells: terminallydifferentiated cells, which are mostly RESULTS

I-

'

M

1

2

3

4 5 6

7 8 9

FIG. 2. Binding of mAb 12D11 is eliminated byphosphatase treatment of histone H1. Purified human placental histoneH1 (2.5 pg/lane) was treated with bacterial alkaline phosphatase, repurified, and analyzed by Western blot with either mAb 12Dll or MRA 12 (thelatterreacts withall histoneH1 isoforms; Monestier et al. (1989)). A lanes, untreated histone H1; H lanes, bacterial alkaline phosphatase-digested histone H1; C lanes, sham-treated histone H1; M lane, markers (molecular mass is indicated in kDa). Untreated lanes ( A ) produced narrower hands because of the higher glycerol content of their samples. Samples I, 2, and 3 were resolved by SDSPAGE and stained with Coomassie Brilliant Blue. Samples 4 , 5,and 6' were analyzed by Western blot with mAb 12Dl1, while samples 7, 8, and 9 were analyzedwith mAb MRA 12. Single bar indicates position of 12Dll' histone H1; double bar indicates position of bulk histone doublet.

84 58 48.5 36.5 26.5

A

B

A B -.

-

-

~

"

1 2

3 4

FIG.3. Monoclonal antibody 1 2 D l l does not bind histone H1 purified from human skeletal muscle. Histone H1 was purified from humanadult skeletal muscle ( A lanes) andplacenta ( B lanes), and resolved hy 18% SDS-PAGE(2.5 wg/lane). M lane, markers (molecular mass represents kDa). Lanes I and 2 were stained with Coomassie Brilliant Blue, while lanes 3 and 4 were analyzed by Western blot with mAb 12Dll. Bar indicates the position of 12Dll' histone H1, which is not detected in the skeletalmuscle preparation. Skeletal muscle histone H1 contains a large amount of histone H1° which migrates asa single hand just aheadof the usual H1 doublet.

Histone Phosphorylation, and

Myogenesis

1583

Maintained by TGF-/3, an Inhibitor of MJogenic Differentiation-When cultured at confluence in low mitogen medium supplemented with TGF-P, skeletal myoblasts withdraw from the cell cycle (Olson et al., 1986; Massague et al., 1986). but neitherformsyncytianorexpress biochemical markers of differentiationsuchassMHC.Myoblastscultured in the presence of TGF-/3 maintain the ability to differentiate after HUSK c2c12 the factor is removed (Vaidya et al., 1989). To examine the effects of TGF-8 in culture on phosphorylationof histone H1 0 4 81%$!@abMO4 8"3'\22b@9b hours at the 12Dll epitope,HUSKand C2C12 myoblasts were . 180 " -T grown in high mitogen medium, then switched tolow mitogen 116 s4 medium or low mitogen medium supplemented with 5 ng/ml 58 4n.5TGF-(3 for 72 h. The cultures were then analyzed for expression of differentiation markers and 12Dll' histone H1. In36.5hibition of HUSK and C2C12 myoblast differentiation by 26.5TGF-8 was evident visually from a lack of syncytia in the HUSK c2c12 B cultures, a lack of troponin I mRNA accumulation (data not HUSK Cells shown), and in Western blots from the absence of sMHC expression after 72 h (Fig. 6, lanes 6C and 6 F ) . Parallel flasks of cells cultured in low mitogen medium not supplemented with TGF-(3 differentiated normally, and expressed sMHC (Fig. 6, lanes 5 B and 5 E ) . In the experiments presentedhere, low mitogen medium contained 4% horse serum, and when this medium was supplemented with TGF-@, differentiation was inhibitedwithoutareduction in the level of MyoDl mRNA (data not shown). Western blots with mAb 12Dll revealed that 12Dll+ histone H1 levels in confluent quiescent CZC12 Calls myoblastsculturedwithTGF-fi were comparabletothose observed in proliferating myoblasts (Fig.6, compare lanes 1A and 3C; compare lanes I D and 3 F ) . The amount of 12Dll' histone H1 in differentiatedculturesgenerated in parallel lanes 213 and 2 E ) . approachedundetectable levels(Fig.6, Replacement of TGF-/3 supplemented medium in differentiation-inhibitedcultureswith low mitogenmedium lacking TGF-(3 resulted in the induction of sMHC expression (in 48 loss of 12Dll'histoneH1(not h)andacorresponding shown). The datashow that TGF-8 inhibits differentiation of FIG.4. Differentiation of human fetal skeletal myoblasts myoblasts while sustaining histone H1 phosphorylation at the (HUSK cells) and murine skeletal myoblasts (C2C12 cells) is 1 2 D l l epitope.

associated with a loss of 1 2 D l 1 + histone H1. A , HUSK and C2C12 myohlasts were cultured in high mitogen medium (0 h), then DISCUSSION switched to low mitogen medium to induce differentiation. A t several times postinduction (indicated in hours), the cells were homogenized Transcriptionalrepression by histoneH1 is thought to in sample huffer,resolved by SDS-PAGE (18% acrylamide for 12Dll' result from inaccessibility of the compact higher order strucH1; 9% for sMHC), and analyzed by Western blot with either mAb tures to transcriptional machinery (detailed by M'eintraub I2Dl1(12Dll' H I ) or MF-20(anti-sMHC).Equalcellnumbers (1985) and Felsenfeld (1992)). Compelling evidence has also (number of nuclei were consideredinfusedcells)wereusedper been presented that histone H1 acts as general a repressor of experimental determination. Molecular mass markers are indicated in kDa. R, microdensitometry scans of results in A for HUSK cells. transcriptioninthepresence(Shimamura et al., 1989) or C , microdensitometry scansof results in A for C2C12 cells. absence (Croston et al., 1991) of core histones. Activation of

transcription by some DNA-binding proteins hasbeen shown in vitro to be associated with their ability to prevent repression multinucleated and express sarcomeric proteins, and quiescent mononuclear cells which do not express sarcomeric pro- by histone H1; in this context, these transcription factors can teins. When HUSK cells were cultured in low mitogen me- be regarded as anti-repressors (Crostonet al., 1991). Consistdiumandinducedtodifferentiate,anobviouseffect was ent with this h-ypothesis, studies in vivo have shown that H1 regions of some observed: differentiated cells, as determined from the presenceis repositioned or depleted from the promoter et al., 1989). Finally, of multinucleated syncytia,were negative for staining by mAb activeorcompetentgenes(Nacheva 12Dl1, whereas as quiescent mononuclearcells were positive histone H1 moreefficiently inhibits transcription of Xenopus S RNA for nuclear staining (Fig.5, E and H ) . T h e loss of staining by laevis oocyte 5 S RNA genes than it does somatic .i mAb 12Dll was not associated merely with the formationof genes, and the selectivity of this effect depends on the pressyncytia but rather with the expression of sarcomeric proteins. ence of A + T-rich flanking sequences in the oocyte gene As shown in Fig. 5K, differentiated mononuclear myocytes (Jerzmanowski and Cole, 1990). This indicates that the reexpressing sarcomeric myosin heavy chain also are negative pressive effect of histoneH1 may be essential for proper loss developmental regulation of the expression of certain genes. for stainingby 12Dll. Together, these data indicate that In this report, dephosphorylation of a specific epitope on of 12Dll' histone H1 is associated with myogenic differentiation and not with mitogenic arrest of myoblasts in the absencehistone H1 is shown to be tightly associated with the inducof differentiation. tion of myogenic differentiation. To directly observe phosPhosphorylation of HistoneHIatthe12011Epitope Is phorylation at this epitope, anovel histone H1 mAb (12Dll)

1584

Myogenesis Historw Phosphorylation and M A

Bc

D E F

A B C

45.5 :>6.5 26.5 84

A

9s

P

1 2

3

4 5 6

D E F

-

A

-, -C

1 2 3

4

5

6

FIG. 6. Culturing HUSK and C2C12 myoblasts with TGF-L3 inhibits differentiation and maintains high levels of 1 2 D l l + histone H 1 . H1'SK (Inrws A-('1 and C'LCI'L tlnncs l)-E:) myohlasts were cultured as myohlasts in high mitogen medium ( A and 11 Innc.9). then switched at confluence to low mitogen medium IH and E; l n n ~ s ) . or t o low mitogen medium supplemented with 5 ng/ml TGF-d (Cand I.' Inncs). \!'hole cellhomogenates(representingequalnumhers of cells or cell nllclei in fused cells) were resolved hy SDS-PXGE and analyzed hy Western hlot aith either mAh 121111 (Inncs I - d ) or mAh MF-20 (lnncs d - 6 ) . flpcn circlcs indicatethepositionof 12Dllhistone H1. and o p m !rinnglcs indicate the position of sarcomeric MHC. M r markers ( M Inn