Reprogramming of myosin light chain 13 expression in muscle heterokaryons

DEVELOPMENTAL

BIOLOGY

145,28-39

(19%)

Reprogramming

of Myosin Light Chain l/3 Expression in Muscle Heterokaryons

LAURA PAJAK, MUTHUCHAMY Department

of Molecular

Genetics,

Biochemistry,

MARIAPPAN,

und Microbiology, Accepted

University January

AND DAVID F. WIECZOREK’ of Ci?Lci?inuti

College oj’ Medicine,

Cincinnati,

Ohio 45267-05.2~

17, 1991

Fast myosin light chain (MLC) l/3 is one of the few genes which regulates transcript production at both transcriptional and post-transcriptional levels, utilizing two functionally distinct promoters coupled with alternatively spliced exons. The transcriptional process controlling expression from this single gene locus is developmentally regulated, such that MLC 1 precedes MLC 3 during myogenesis. Results from our RNA analyses demonstrate that in differentiated rat L6E9 muscle, MLC 3 is the sole isoform expressed from the MLC l/3 locus. However, we also show that by generating rat L6ES:mouse C2 muscle heterokaryons, MLC 1 expression from the L6E9 MLC locus can be induced. In addition to novel rat MLC 1 expression in the C2:L6E9 heterokaryons, we show that the synthesis profile of rat MLC 3 mRNA is also altered relative to L6E9 muscle cultured alone. Additional experiments demonstrate that the reprogramming of rat MLC 1 and 3 expression in the muscle heterokaryons requires that C2 and L6E9 nuclei be contained within a common ” and is not under the control of cytoplasm. These results demonstrate that expression from the MLC l/3 gene is “plastic, a strict developmental program but, rather, can be modified by the environmental milieu. cc: 1991 Academic press, IIIC.

ment, located over 24 kb downstream of the MLC 1 promoter, may play a dramatic role in regulating the expression of this gene (Donoghue et al., 1988). A common regulatory element shared by many of the members of the MLC gene family has also recently been identified (Uetsuki et ah, 1990). The differential expression of the two MLC promoters has been addressed recently. Employing a minigene construct in which the MLC 1 promoter, the transcription initiation site for MLC 1, and 6.3 kb of intron sequence were deleted, Garfinkel and Davidson (1987) demonstrated that MLC 3 mRNA was produced in both myogenie and nonmyogenic cells. Preliminary results by Taubman et al. (1989) suggest that expression of MLC in BC3Hl cells is restricted to the MLC 3 isoform, although it is unclear why there is restricted expression from this locus in these differentiation-deficient cells. In this report, we unequivocally demonstrate that MLC 3 is the sole isoform expressed from the MLC l/3 locus in differentiated L6E9 muscle. The expression of MLC 3, in the absence of MLC 1, is unlike the in viva expression which occurs in the rat or the in vitro expression obtained from culturing primary rat muscle cells. However, through the use of muscle heterokaryons, we show that this absence of MLC 1 expression in L6E9 cells is reversible. The generation of rat MLC 1 in C2:L6E9 muscle heterokaryons demonstrates that the MLC 1 promoter in L6E9 cells is not defective, but rather suggests that differentiated L6E9 muscle is deficient in producing trans-acting regulatory factors necessary to activate the MLC 1 promoter. In addition to the altered rat MLC 1 expres-

INTRODUCTION

Myosin light chain (MLC), a major protein of the myosin head region, is a component of thick striated muscle filaments. Encoded within a multigene family, the MLC genes exhibit both tissue-specific and developmental stage-specific expression (Cauthier et ab, 1982; Obinata et ab, 1983). Vertebrate fast skeletal muscles contain several members of this family, including the alkali myosin light chain isoforms termed MLC 1 (M,. 21,000) and MLC 3 (M, 16,000). Nucleotide and amino acid analyses have demonstrated that MLC 1 and MLC 3 are protein isoforms produced from a single gene locus which utilize two functionally distinct promoters coupled with alternatively spliced exons (Nabeshima et al., 1984; Periasamy et al., 1984; Robert et ab, 1984; Strehler et al., 1985). The transcriptional process controlling expression from this locus is developmentally regulated such that MLC 1 precedes MLC 3 during myogenesis. Recent investigations on the promoter activities for the MLC l/3 gene have defined the regulatory elements necessary to confer functional activity in a muscle differentiation-dependent fashion (Billeter et al., 1988; Daubas et al., 1988; Cohen et al., 1988). Results from Daubas et al. (1988) show that the 5’ upstream proximal promoter regions of 1.2 kb and 438 bp, respectively, for MLC 1 and MLC 3 are sufficient for tissue-specific expression. In addition, a muscle-specific enhancer ele1 To whom dressed. 0012.1606/91

correspondence

and

$3.00

Copyright CG1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

reprint

requests

should

be ad-

28

PAJAK.

MARIAPPAN,

AND

WIECZOREK

sion in the C2:L6E9 heterokaryons, we show that the synthesis profile of rat MLC 3 transcripts is altered. This MLC 3 mRNA synthesis occurs much earlier in the heterokaryons than when L6E9 cells are cultured alone, indicating that the production of rat MLC 1 and 3 transcripts is following the pattern dictated by C2 cells. Furthermore, additional experiments strongly suggest that the reprogramming of rat MLC 1 and 3 expression in the muscle heterokaryons is not due to stable diffusible factors produced by C2 cells, but rather requires that C2 and L6E9 nuclei be contained within a common cytoplasm. MATERIALS

AND

METHODS

Preptrrntim of ccl1 cultures. L6E9 rat muscle cells were maintained in Dulbecco’s modified Eagle’s medium (DME) (GIBCO Laboratories) supplemented with 20% fetal calf serum (Nadal-Ginard, 1978). For induction of myogenic differentiation, the medium of exponentially growing cultures was replaced with DME supplemented with 3% horse serum and 3% fetal calf serum. These cultures were maintained for 30 days with the initial exposure of the cells to differentiation medium designated as Day 0. The differentiation medium was replaced every 3-4 days. C2 mouse muscle cells were maintained in DME medium supplemented with 20% fetal calf serum and 0.5% chick embryo extract (Yaffe, 1968). For induction of myotube fusion, growing myoblasts were cultured in DME supplemented with 5%~horse serum. This differentiation medium was replaced every 3-4 days. To produce mixed cocultures of rat LGE9:mouse C2 heterokaryons, myoblasts of both species type were counted, mixed, and plated together in ratios of l:l, 3:1, and 1:3 at a cell density of 4.5 X lo6 cells per 150-mm dish, in DME containing 20% fetal bovine serum. After 2 days, the medium of exponentially growing cultures was replaced with DME supplemented with 3% horse serum and 3% fetal calf serum. This medium was also changed every 3-4 days. Rat primary muscle cultures dissociated from 19-day fetal limb muscle were established on collagen-coated dishes (4.5 x lo6 cells per 150-mm dish) according to established methods (Wieczorek, 1984). Fibroblast preplating and the addition of arabinofuranosylcytosine (10m5M) to the medium during myoblast fusion were used to decrease the concentration of fibroblast cells. Primary myogenic differentiation was obtained in DME containing 10% horse serum. Primary rat fibroblasts were established and passaged in DME supplemented with 10% fetal calf serum. Pr~pctm~tiov of’conditioned m,edium. Parallel cultures of C2 and L6E9 cells were grown under conditions to promote muscle differentiation (see above). After 5

Reproqrcr II,n/ irl.q MLC’ l/Y h?pressitm

29

days, the medium was removed from the muscle cultures. This enriched medium was supplemented with an equal volume of fresh DME containing 5% horse serum and placed on the muscle cultures of the reciprocal species type. After 2 days, cytoplasmic RNA was harvested from the C2 and L6E9 cell cultures and analyzed by Sl nuclease mapping. Nuclear staining. To stain the nuclei of the C2 and L6E9 muscle cultures, the cells were washed with phosphate-buffered saline (PBS) and fixed with 1% paraformaldehyde in PBS at 3’7°C for 20 min. These cells were then treated with 100% ethanol and incubated at -20°C for 20 min. After washing with deionized water, the cells were stained with Hoechst 33258 (0.12 yg/ml) for 5 min. The cultures were then rinsed with PBS and overlayed with 100% glycerol. Nuclear fluorescence was visualized at a wavelength of 370 nm. RNA isolation and analysis. Total cytoplasmic RNA was isolated from cells grown in culture according to established methods (Favaloro et al., 1980). Total rat skeletal muscle RNA was obtained from limb tissues using modifications of the hot phenol procedure (Soeiro et a,l., 1966). The procedure of Thomas (1980) was used for blotting and hybridization of RNAs electrophoresed in formaldehyde-agarosegels. Spectrophotometric measurements and formaldehyde-agarose gels were run to ensure that the quantitative amounts of the RNAs from the various samples were identical. Radioactive cDNA probes used in Northern blot hybridizations were prepared by nick-translation of purified cDNA insert sequences (Rigby et ah, 1977). RNA-DNA hybridization followed by Sl nuclease mapping analysis was performed according to the method of Berk and Sharp (1977), under the conditions used previously (Wieczorek et al, 1985). The probes used were generated by digestion with restriction endonucleases and labeled at the 5’ end with [Y-~‘P]ATP (Amersham Corp.) and T4 polynucleotide kinase (Bethesda Research Laboratories). Twenty five micrograms of total cellular RNA was hybridized to 2 X lo4 cpm of probe in 25 ~1 of 80% deionized formamide, 400 mMNaC1, 10 mMPipes, pH 6.4, 0.05% sodium dodecyl sulfate, 1 mM EDTA. The hybridization mixture was incubated at 65°C for 1 hr, the temperature was adjusted to 42”C, and the incubation continued for 16 hr. One hundred units of Sl nuclease (BRL) in 300 ~1 of 200 mM NaCl, 30 mM sodium acetate, pH 4.5, 3 mM ZnSO, was added to each sample and incubated at 25°C for 1 hr. The reaction was terminated with 10 mM EDTA and precipitated with ethanol. Dried pellets were dissolved in 80% formamide and electrophoresed on 6% polyacrylamide, 8 M urea sequencing gels. The gel was dried and exposed for autoradiography on Kodak X-Omat AR film. Slot-blot a,nd h ybridixntion of’ RNA. Fifteen micrograms of RNA was denatured with 6.15 M formalde-

30

DEVELOPMENTAL

BIOLOGY

hyde, 10x SSC, and slot-blotted to nitrocellulose. The nitrocellulose membranes were baked for 2 hr at 80°C under vacuum. Hybridization conditions for oligonucleotide probes to rat MLC 1 and rat MLC 3 transcripts were 6~ SSC, 0.1% Denhardts, 0.1 MNaH,PO,, 1% SDS, and 0.25 mg/ml salmon sperm DNA. Blots were hybridized at 42°C overnight, followed by washing with 2~ SSC, 0.2% SDS at 37°C. Synthetic oligonucleotides were 5’ end-labeled using [-y-32P]ATP. The sequence of the rat-specific MLC 1 oligonucleotide probe is: 5’-TGGGGCAGGAGCAGGAGCA3’; the sequence of the rat-specific MLC 3 oligonucleotide probe is: 5’-AACTGAAGACACCTCCAGTGGG-3’.

RESULTS

Organization

of the MLC l/3 Gene and Its Transcripts

Regulation of the fast MLC l/3 gene is controlled by both transcriptional and post-transcriptional processes. The rat MLC l/3 gene is organized into both common and isoform-specific exons, utilizing two differentially regulated promoters (Fig. 1) Periasamy et al., 1984; Strehler et al., 1985). MLC 1 mRNA transcripts are produced by alternatively splicing the first and fourth exons to a set of five common exons, including a single 3’ untranslated region. This message is initially expressed during fetal development and remains expressed in the adult. MLC 3 transcripts are produced from a distinct promoter (exon 2) and utilize alternative splicing to couple exons 2 and 3 to a common body of exons. Previous studies showed that in vivo expression of MLC 3 is first detected during the neonatal stages of development and remains expressed in the adult (Periasamy et ab, 1984). Thus, the regulation of MLC l/3 mRNA synthesis entails activation of distinct promoters coupled with differential exon incorporation.

MIX

1

MIX 3 FIG. 1. Diagram of the rat fast MLC l/3 gene organization (top) and its associated mRNA transcripts (bottom two) (Periasamy et al., 1984). Black boxes indicate constitutive exons; striped boxes mark 5 untranslated (UT) sequences; dotted boxes indicate isoform-specific coding exons.

VOLUME

145,1993

Expression of MLC l/3 in Skeletal Musculature Primary Muscle Cell Cultures

and

Results obtained with Northern blot analyses by previous investigators (Periasamy et al., 1984) show that MLC 1 transcripts are initially present during fetal development, with MLC 3 being expressed during neonatal stages. To more precisely determine the pattern of MLC l/3 expression in rat musculature and in cultured primary rat muscle cells, we conducted an analysis of MLC synthesis using the sensitive and stringent conditions of Sl nuclease mapping. Striated muscle RNA was prepared from embryonic, newborn, and adult limb muscles and hybridized to a MLC probe containing both common and MLC 3-specific sequences. The MLC 3 probe used is a double-stranded, 379-bp-long, AvaI-PvuI fragment that maps at the 5’ end of the cDNA clone (pMLC-35; Periasamy et al., 1984). It includes the codons for aa 9-51, which are common to both MLC 1 and MLC 3 messages, aa 1-8 which are specific for MLC 3, and 110 nucleotides which are specific for the MLC 3 5’ untranslated region. The probe also contains 144 bp of pBR322 sequence. Hybridization to the homologous mRNAs should generate a partial protection of 235 nt corresponding to MLC 3 message and a partial protection of 123 nt representing MLC 1 transcripts. Renaturation of the probe to itself would be seen in a 379-nt band. From the Sl nuclease protection pattern observed in Fig. 2, partially protected bands of 235 and 123 nt are visible in all developmental stages of rat musculature. However, from the intensity of the 123- and 235-nt bands, it is evident that there is a greater quantitative level of MLC 1 transcript than MLC 3 during development. This pattern of expression is in basic agreement with results obtained with Northern blot analyses by previous investigators (Periasamy et al., 1984), although MLC 3 transcripts were not previously detected during fetal development. To investigate the induction of MLC l/3 transcripts in cultured rat primary muscle cells, skeletal limb muscle was isolated and dissociated from Wday rat fetuses. Myoblast fusion began 3 days after the initial plating and was complete by 5 days in vitro. Spontaneous contractions began after 7 days and the muscle cultures were maintained for 21 days. Cytoplasmic RNA was isolated from myotubes both before (early mt) and after (mt) the onset of spontaneous contractions. Primary fibroblast cell cultures were established as a negative control. In an Sl nuclease analysis with the MLC probe described above, 123- and 235-nt protected fragments representing MLC 1 and 3 mRNA synthesis, respectively, are visible in the primary muscle culture lanes (Fig. 2, lanes 8-9). Both MLC 1 and 3 transcripts are visible in the early mt sample, thus demonstrating that myotube contractions are not requisite for MLC 3 ex-

31

4m- 379 3102781271234-2235

-+dds

118-

MLC3 CDNA

5’ UT RI

FIG. 2. Detection of MLC l/3 expression in rat skeletal muscle and primary hybridized to RNAs from limb skeletal muscle of l&day fetal, newborn (Nb), or Protection of 235 nt represents full protection of MLC 3 coding and 5’ untranslated transcripts. The band at 379 nt reflects reannealing of the probe (including pBR322 lanes 7710 for 5 days. RSM, rat skeletal muscle; Early mt, noncontracting mgotubes; 21 days

pression. This phenomenon is also illustrated by the fact that L6E9 myotubes, which do not spontaneously contract, express MLC 3 transcripts (see below). No MLC l/3 mRNA is visible in the fibroblast or tRNA negative control samples. Furthermore, in agreement with the results obtained from the muscle tissue RNA analysis, MLC 1 mRNA levels were quantitatively greater than the MLC 3 levels in the primary muscle cell cultures. MLC l/3 expression from the 21-day primary muscle cultures (late mt) shows a decreased amount of both of these mRNAs, perhaps due to an increased cellular necrosis associated with these long term primary muscle cultures (Fig. 2, lane 10). MLC 3 Is Solely Expressed

+-123

in L6E.9 Muscle Cells

Previous investigations have examined contractile protein gene expression in L6E9 muscle cells (Garfinkel ef al., 1982; Medford ef al., 1983). This rat cell line has

AI

muscle cultures. A 5’ end-labeled MLC 3 cDNA probe was adult rats, primary rat muscle and fibroblast cell cultures. sequences. Partial protection of 123 nt corresponds to MLC 1 vector sequences) to itself. Lanes l-6 were exposed for 1 day; mt, contracting myotubes; Late mt, myotubes in culture for

served as a model system for understanding the induction and coordinate regulation of muscle-specific genes. A previous investigation (Periasamy et al., 1984) had failed to detect expression from the MLC l/3 gene in L6E9 muscle even though screening of a cDNA library prepared from differentiated L6E9 cells lead to the isolation of a MLC l/3 clone (Garfinkel et al., 1982). To more thoroughly assess MLC l/3 expression in L6E9 cells, we cultured L6E9 cells in muscle differentiation medium (DME + 5% horse serum) for 30 days and assayed for MLC l/3 mRNA synthesis using more sensitive methods. Fusion of myoblasts into muscle syncytia was initiated 3 days after initial culturing in low-serum medium and was complete by 6 days in vitro. Cytoplasmic RNA was isolated from these cultures from Days 2-30. To ascertain the expression of MLC l/3 transcripts in differentiated L6E9 muscle, we used MLC 3 and MLC 1 probes in Sl nuclease protection analyses. L6E9 RNA

32

DEVELOPMENTALBIOLOGY

isolated from differentiated muscle cultures of varying time periods was hybridized to the MLC 3 probe previously described. As shown in Fig. 3A, following Sl nuclease digestion, a partially digested band of 235 nt is obtained in the Day lo-30 L6E9 cells and in the positive control adult rat skeletal muscle (RSM) RNA. This band represents the expression of MLC 3. Surprisingly, a partial protection of 123 nt representing MLC 1 mRNA is not observed in these same L6E9 cells, but is present in the RSM. The results from this analysis demonstrate that the MLC 3 transcript can be synthesized in long term cultures of L6E9 muscle in the absence of MLC 1 mRNA expression. To corroborate the expression of MLC 3 in the absence of MLC 1 in L6E9 cells, we conducted a reciprocal analysis using a MLC 1 cDNA probe. A cDNA clone, pMLC-91, has been characterized and contains sequences corresponding to MLC 1 mRNA (Periasamy et al, 1984). The

V0~~~~145,1991

double-stranded probe used is a 545-bp-long, AVaI-BgZI fragment that maps at the 5’ end of the cDNA clone. The sequence in this clone includes 123 nt common to both MLC 1 and MLC 3, 278 nt unique to MLC 1, and 144 nt specific for pBR322. Hybridization to the homologous mRNAs should generate a partial protection of 401 nt corresponding to MLC 1 message and a partial protection of 123 nt representing MLC 3 transcripts. Renaturation of the probe to itself would show a 545-nt band. From the Sl nuclease protection pattern observed in Fig. 3B, only a 123-nt partially protected band is seen in the L6E9 muscle samples. This protection represents MLC 3 expression, MLC 1 expression, indicated by a 401~ nt band, is only seen in the positive control RSM sample, and not in any L6E9 muscle samples. The results of Figs. 3A and 3B unequivocally demonstrate that MLC 3 expression can occur in the absence of MLC 1, thereby demonstrating that the developmental regulation nor-

(A) -379 310-

-310

-278/271

278/271-

-234

234+235

123

-

-118

-123

MLCB CDNA P”“,

RI

AI

FIG. 3. Sl nuclease analysis of MLC l/3 expression in L6E9 muscle cells. (A) Detection of Sl nuclease-protected MLC 1 and 3 mRNAs hybridized with a 5’end-labeled cDNA probe containing common and MLC 3-specific sequences. Following Sl nuclease digestion, a protection of 235 nt represents MLC 3-specific sequences; a partial protection of 123 nt corresponds to MLC 1 transcripts. Reannealing of the probe to itself is shown in the 379.nt band. Lanes l-3 were exposed for 1 day; lanes 4-9 exposed for 3 days. (B) Analysis of MLC l/3 mRNA by using a MLC 1 cDNA containing common and MLC l-specific sequences. Protection of 401 nt designates full protection of MLC 1 coding and 5’ untranslated sequences. Partial protection of 123 nt represents MLC 3 transcripts. The band at 545 nt corresponds to self-reannealing of the probe.

PAJAK,

mally associated uncoupled. Induction

with

MARIAPPAN,

AND

WIECZOREK

the MLC l/3 expression

can be

of Rat MLC 1 in L6E9 Muscle

The inability of L6E9 muscle to synthesize MLC 1 transcripts could be caused by a wide variety of reasons, including defective sequences located within the MLC 1 promoter or the absence of a regulatory factor essential for MLC 1 promoter activity. To investigate the potential defect associated with MLC 1 expression in L6E9 cells, we decided to conduct cell fusion experiments between rat L6E9 muscle cells and C2 cells, a mouse muscle cell line which expresses both MLC 1 and 3 transcripts (Donoghue et al., 1988). Fusion of C2 and L6E9 cells was monitored by Hoechst 33258 staining which binds preferentially with DNA rich in adenine and thymine (Weisblum and Haenssler, 1974). Previous studies have demonstrated that C2 nuclei stained with Hoechst give a characteristic intense punctate staining pattern from the poly(A-T) rich regions in the mouse centromerit regions of the chromosomes (Blau et al, 1983). Our control experiments demonstrate similar results (Figs. 4A and 4B). C2 myoblasts formed well-differentiated, multinucleated myotubes when cultured under lowserum conditions (Fig. 4A). As is clearly visible in Fig. 4B, when stained with Hoechst dye, C2 myotube nuclei are well-defined and exhibit a punctate appearance. In contrast with the well-ordered myotubes seen with C2 cells, differentiated L6E9 muscle exhibits the morphology of a large syncytia (Fig. 4C). Elongated myotube structures are occasionally present in differentiated L6E9 muscle, but they are usually less well-defined and sometimes randomly oriented. Furthermore, in contrast to C2 cells, when L6E9 cells are stained with Hoechst, the nuclei exhibit low level fluorescence and there is no punctate appearance associated with them (Fig. 4D). At increased cell densities, a higher level of nuclear fluorescence is occasionally seen, but these nuclei are seldom punctate. Thus, through the use of Hoechst staining, we can readily distinguish between mouse C2 and rat L6E9 nuclei. To investigate the inability of L6E9 cells to express MLC 1 mRNA, we conducted cell fusion experiments between L6E9 and C2 muscle cells. L6E9 and C2 cells were plated together in different ratios (l:l, 1:3, and 3:l) under low-serum conditions to stimulate myogenic differentiation. When the two muscle cell types fuse, welldefined multinucleated myotubes are present in many areas, however, large muscle syncytia are also found (Figs. 4E and 4G). Nuclear staining with Hoechst dye shows nuclei exhibiting both a punctate and a diffuse pattern, sometimes clearly contained within a common cytoplasm (Figs. 4F and 4H: see arrows). The appearance of both mouse- and rat-derived nuclei residing

Repoqt~~

rrr PI iryl ML(’

l/.1 E.rpressiott

33

within common myotubes demonstrates that myoblast cell fusion occurs between the two muscle cell lines of different species. To assess the ability of the C2:L6E9 fused muscle to express rat MLC 1 transcripts, we isolated cytoplasmic RNA from these cultures and analyzed it in Northern slot-blot analyses. RNA was prepared from these heterokaryon cultures after these cells had been grown under differentiation conditions for various time intervals (5, 8,12, and 14 days). To probe this RNA, we synthesized a 19-bp oligonucleotide specific for rat MLC 1 5’ untranslated sequences. As seen in Fig. 5A, rat MLC 1 is expressed in the C2:L6E9 muscle heterokaryons after they have morphologically differentiated in the cocultures and not when L6E9 or C2 muscle cells are cultured alone. The absence of expression in the L6E9 and C2 cultures demonstrates the specificity of the probe and confirms the results of the Sl nuclease analyses (Figs. 3A and 3B). Interestingly, this rat MLC 1 transcript accumulation is significant by 8 days in vitro. Furthermore, it appears that the level of this mRNA accumulation is partially determined by the quantitative ratio of C2:L6E9 cells which are initially seededin the cultures. As the number of L6E9 cells increases, the expression of rat MLC 1 mRNA increases. These results conclusively illustrate that the rat MLC 1 promoter in L6E9 cells is not defective and that the associated transcripts can be produced by the endogenous gene in L6E9 cells, but only when its normal cellular environment has been modified. In the current situation, factors produced by the C2 cells can apparently complement those provided by L6E9 cells or activate factors not normally expressed in L6E9 cells, to allow for the transcription, splicing, and stabilization of the rat MLC 1 mRNA. Since MLC 3 transcripts are encoded in the same gene which produces MLC 1 mRNA, the novel synthesis of rat MLC 1 in the L6E9:C2 heterokaryons may have influenced the expression of rat MLC 3 in these cells. To address this issue, RNA from the C2:L6E9 heterokaryons was analyzed in a Northern slot-blot analysis similar to the one conducted for MLC 1 expression. A synthetic oligonucleotide, specific for rat MLC 3 sequences, was synthesized and hybridized to the C2:L6E9 RNAs. As seen in Fig. 5B, rat MLC 3 is expressed in the muscle heterokaryons, and to a lesser degree in the L6E9 cells cultured alone. No expression is visible in the C2 myotubes. In accordance with the results of Fig. 5A, there is also an increase in the expressed levels of rat MLC 3 as the ratio of L6E9 cells increases in the heterokaryons. However, unlike the expression of rat MLC 3 in L6E9 cultured alone, in which the transcripts were detected only after extended time periods (i.e., 10 daysFigs. 3A and 3B), in the C2:L6E9 heterokaryons, rat MLC 3 mRNA is clearly visible at high levels by 8 days. This increased level of rat MLC 3 transcripts demon-

34

DEVELOPMENTAL BIOLOGY

VOLUME 145, 1991

FIG. 4. Phase-contrast and fluorescent micrographs of mouse C2, rat L6E9, and CZ:L6E9 heterokaryon cell cultures. C2 myotubes (A), L6E9 differentiated muscle syncytia (C), and C2:L6E9 differentiated muscle heterokaryons (E,G) are shown in phase-contrast; the same cells, respectively, are shown stained with Hoechst 33258 to show nuclei (B,D,F,H). Large arrows in E-H show punctate mouse C2 nuclei within myotubes containing rat L6E9 nuclei (small arrows). Magnification, 285~.

PAJAK,MARIAPPAN,

ANDWIECZOREK

strates that both MLC 1 and 3 expression are being dramatically affected by the cellular environment in which the L6E9 cells are being cultured. Since previous investigations studying actin gene expression in mouse-human heterokaryon cell cultures (mouse C2 muscle fused

with human fibroblasts) (Hardeman et al., 1986) found reprogramming of contractile protein gene expression to occur in the mouse C2 cells, we addressed whether the L6E9 cells in the heterokaryons had been reprogrammed to follow a C2 schedule of contractile protein

36

DEVELOPMENTALBIOLOGY

(6) RSM L6E9

(A)

c2

C2: L6E9

VOLUME 145,lWl

protection at 123 nt). These data suggest that a secreted factor produced by C2 cells is not sufficient to activate rat MLC 1 expression in the L6E9 cells. Also, under these conditions, there does not appear to be a stable factor produced by L6E9 muscle which prohibits or destabilizes MLC l/3 mRNA synthesis in C2 cells. Thus, these results imply that the C2 and L6E9 muscle nuclei must exist within a common cytoplasm to allow for activation of rat MLC 1 expression, although mechanisms mediated by cell-cell contact have not been ruled out at this time. DISCUSSION

The results presented here demonstrate unequivocally that MLC 3 can be expressed in the absence of MLC

FIG. 5. Rat MLC 1 expression of C2:L6E9 heterokaryons. Northern slot-blot autoradiograms of mouse C2 and rat L6E9 RNA cultured either alone or as heterokaryons in the designated ratios (l:l, 1:3,3:1) for the indicated days (D). (A) The RNA was hybridized to a rat-specific MLC 1 19-base oligomer with the following sequence: 5’-TGGGGCAGGAGCAGGAGCA-3’. (B) RNA hybridized to a rat-specific MLC 3 22-base oligomer with the following sequence: 5’-AACTGAAGACACCTCCAGTGGG-3’.

“se

c545

-401

a%

310

-401

II 2701271

310 * 278/27l 234 234

gene expression. Analysis of RNA from long term cultures of C2 myotubes demonstrates that both MLC 1 and 3 mRNAs are transcribed soon after muscle differentiation (low transcript levels are visible within 4 daysdata not shown). These results suggest that the program of rat MLC 1 and 3 expression in the heterokaryons resembles the pattern observed in C2 cultures. To address the importance of the intracellular milieu and diffusible factors in altering rat MLC l/3 expression, we conducted experiments using conditioned medium. Medium from differentiated C2 myotube cultures which are expressing MLC 1 and 3 transcripts was removed and mixed in equal amounts with freshly made medium; this conditioned medium was put into cultures of differentiated L6E9 muscle. A reciprocal experiment was also conducted by culturing C2 myotubes with L6E9 conditioned medium. After 2 days, the L6E9 and C2 muscle RNA was isolated and assayed for MLC expression. Using a 5’ end-labeled probe derived from pMLC91, Sl nuclease mapping was conducted on the RNA. As seen in Fig. 6, there was no activation of rat MLC 1 transcripts detected in L6E9 cultures when grown with medium from C2 cell cultures. Conversely, C2 cultures supplemented with medium derived from L6E9 cells continued to produce MLC 1 (evident by protection at 401 nt) as well as MLC 3 (as demonstrated by a partial

194 194

c

119

MLCI cDNA

@K322 El@,

123

51 HI

92 I AI

FIG. 6. Expression of MLC l/3 transcripts in L6E9 and C2 cultures incubated with conditioned medium. Autoradiograph of differentiated L6E9 and C2 muscle culture RNA probed in an Sl nuclease analysis for MLC l/3 expression. The muscle cultures were grown for 2 days with a 1:l mixture of conditioned:fresh medium. Total RNA from these cultures was probed with a 5’ end-labeled MLC 1 probe. Full protection of the probe is 401 nt, corresponding to MLC 1; a partial protection of 123 nt corresponds to MLC 3 expression. The artifactual band at 545 nt is due to self-reannealing of the probe. The partial protections at -301 and -215 nt in the C2 RNA samples correspond to unknown transcripts.

1. Using two structurally distinct promoters, the MLC gene produces two mRNA transcripts by a combined process of diflerential transcription and alternative splicing. In rat, MLC 1 transcripts are generated in fetal, newborn, and adult skeletal muscle, with MLC 3 expression principally expressed in newborn and adult musculature. The presence of MLC 3 mRNA in L6E9 cells demonstrates that the mechanism or endogenous factors which regulate the alternative splicing pattern to selectively produce MLC 1 or MLC 3 transcripts are operative in these cells. However, expression from the MLC l/3 gene is not under t.he control of a strict developmental program, but can be modified by regulatory factors. In fact, the generation of L6ES:C2 heterokaryons induces expression of rat MLC 1, the isoform normally expressed at the earlier developmental stage. Developmental and tissue-specific factors regulate the expression of various multigene families, including fi-globin (Hardison et ul., 1979; Wood e,f ul., 1985), immunoglobulin (Perlmutter et al., 1985), myosin heavy chain (MHC) (Mahdavi et al., 19X6), tropomyosin (Wieczorek ef crl., 1988; Lees-Miller et (xl., 1990), and MLC (Cauthier et trl., 1982; Obinata et c/l., 1983). This regulation often exhibits sequential activation of genes clustered on a single chromosome or multiple isoforms encoded within a single gene locus. Many of the contractile protein genes, however, can reversibly modulate their developmental isoform transitions by various influences, such as hormonal (Mahdavi et (xl., 1986; Whalen et crl., 1985) and neuronal (Jolesz and Sreter, 1981; Eccles, 1963) stimuli. For example, developmental stage-specific MHC genes can be transcriptionally reactivated in hypothyroid adult musculature (Izumo et ul., 1986; Russell ef al., 1988); cytoskeletal/myoblast cu-tropomyosin isoforms can be preferentially reinduced in differentiated muscle treated with nerve extracts (Wieczorek, 1988). The present studies illustrate that expression from the MLC l/3 locus can also be modified depending upon the cellular environment. Cell culture systems have proven invaluable in advancing our understanding of molecular regulatory events involved in the differentiation and development of eukaryotic cells and tissues. Analyses of gene products expressed in cell heterokaryons have illustrated that reprogramming of “determined/differentiated” cellular pathways can occur (Hardernan ef (xl., 1986). Furthermore, these changes in gene expression can lead to morphological changes (Miller et ul., 1988), with the distribution of gene products encoded by different nuclei becoming intermixed (Tassin et al., 1985) or localized (Pavlath d ul., 1989; Ralston and Hall, 1989) in nuclear domains. By fusing L6E9 and C2 muscle cells, we detect the novel synthesis of rat MLC 1. Previous studies on MLC expression have implied that the trot/,.+acting signals recognizing the distant 3’ MLC l/3 enhancer are

functionally identical in both rat and mouse species (Donoghue et (Al., 1988). Furthermore, recent investigations suggest that mouse C2 muscle cells express very high levels of myogenic regulatory factors (Wright and Lin, 1990). Thus, the activation of rat MLC 1 in the L6E9:C2 heterokaryons could result from the activation of the rat MLC 1 promoter by fruns-acting factors produced from the mouse C2 nuclei. The structural similarity of the mouse MLC 1 sequences to the rat MLC 1 region renders this scenario a distinct possibility, although we cannot rule out the possibility that a factor produced by C2 cells acts to stabilize the rat MLC 1 message. Experiments in which L6E9 cells were treated with 5-azacytidine (a cytidine analog which when incorporated into replicating DNA results in hypomethylation) (Jones and Taylor, 1980) resulted in MLC 1 mRNA synthesis in these cells (L. Pajak and D. Wieczorek, unpublished results). Transcriptionally active genes are often hypomethylated, indicating that methylation may influence gene expression. In these experiments, 5-azacytidine may directly demethylate the MLC locus making it accessible to the transcriptional machinery; conversely, 5-azacytidine may demethylate the locus of a trolls-acting regulatory factor which allows production of MLC 1 transcripts. Preliminary studies involving nuclear run-on experiments using protein extracts from C2 cells with L6E9 nuclei suggest that an induction of rat MLC 1 synthesis occurs in the rat nuclei, lending further support for the presence of a mouse frcxns-activating factor(s) acting on the rat MLC locus (M. Mariappan and D. Wieczorek, unpublished results). It is well-documented that there is coordinate induction of muscle-specific genes associated with the process of muscle differentiation. The L6E9 muscle cell line, a model system for investigating gene regulation, also exhibits this pattern of coordinate contractile protein gene expression. For example, when L6E9 muscle is cultured for extended time periods under differentiation conditions (i.e., low serum in the medium) and the RNA is analyzed at each time point, MHC transcripts are initially visible by Day 6 and reach peak levels soon thereafter (Day lo-data not shown). A similar pattern of expression occurs for cu-actin, troponin T, and the striated muscle isoform of cu-tropomyosin (data not shown). One interesting phenomenon is that by 14 days iz vitro, transcripts from each of these genes are at very low levels and are undetectable by 21 days. Surprisingly, MLC 3 mRNA synthesis is induced only after the deinduction of embryonic MHC (the only MHC isoform expressed in L6E9; Wieczorek ef ul., 1985). In ZGVO,MLC 3 synthesis occurs only after embryonic MHC gene transcription is downregulated, namely, during the perinatal stages of development. Thus, embryonic MHC expression and MLC 3 expression appear to be regulated in an antithetic fashion, suggestive of a common mechanism which

38

DEVELOPMENTAL BIOLOGY

may control the expression of both of these genes. Furthermore, the transcription of MLC 3 in the absence of any MHC gene transcript demonstrates that the synthesis of MLC 3 mRNA is not dependent on either the activation of the neonatal MHC gene or the presence of any detectable MHC gene transcript. Thus, these results, in addition to those of this report, suggest that the mechanisms and/or factors controlling MLC l/3 expression may differ from those coordinately regulating other contractile protein genes. The authors acknowledge Dr. B. Nadal-Ginard for assistance in the initial phases of this research. We thank Drs. E. Choi and T. Doetschman for their critical reviews of the manuscript. This work was partially supported by NIH Grant AR 39423 awarded to D.F.W.

REFERENCES BERK, A. and SHARP, P. (1977). Sizing and mapping of early adenovirus mRNAs by gel electrophoresis of Sl endonuclease-digested hybrids. Cell 12, 721-732. BILLETER, R., QUITSCHKE, W., and PATERSON, B. (1988). Approximately 1 kilobase of sequence 5’ to the two myosin light-chain lf/3f gene cap sites is sufficient for differentiation-dependent expression. &fo[. Cell. Biol. 8, 1361-1365. BLAU, H., CHIU, C., and WEBSTER, C. (1983). Cytoplasmic activation of human nuclear genes in stable heterocaryons. Cell 32,1171I1180. CAUTHIER, F., LOWEY, S., BENFIELD, P., and HOBBS, A. (1982). Distribution and properties of mgosin isozgmes in developing avian and mammalian skeletal muscle fibers, .I. &/I. Biol. 92, 471-484. COHEN, A., BARTON, P., ROBERT, B., GARNER, I., ALONSO, S., and BUCKINGHAM, M. (1988). Promoter analysis of myosin alkali light chain genes expressed in mouse striated muscle. Nucleic Acids Res. 16, 10,037-10,052. DAUBAS, P., KLARSFELD, A., FARNER, I., PINSET, C., Cox, R., and BUCKINGHAM, M. (1988). Functional activity of the two promoters of the myosin alkali light chain gene in primary muscle cell cultures: Comparison with other muscle gene promoters and other culture systems. Xuclric Acids Rm. 16, 1251-1271. DONOGHUE, M., ERNST, H., WENTWORTH, B., NADAL-GINARD, B., and ROSENTHAL, N. (1988). A muscle-specific enhancer is located at the 3’ end of the myosin light-chain l/3 gene locus. Genes Dw. 2,17791790.

ECCLES, J. (1963). Specificity of neural influence on speed of muscle contraction. 171 “The Effect of Use and Disuse on Neuromuscular Functions” (E. Gutmann and P. Hnik, Eds.), pp. 111-128, Elsevier, Amsterdam. FAVALORO, J., TREISMAN, R., and KAMEN, R. (1980). Transcription maps of polyoma virus-specific RNA: Analysis by two-dimensional nuclease Sl gel mapping. Al “Methods in Enzymology” (L. Grossman and K. Moldave, Eds.), Vol. 65, pp. 718-749. Academic Press, San Diego, CA. GARFINKEL, L., and DAVIDSON, N. (1987). Developmentally regulated expression of a truncated myosin light-chain lF/3F gene. Mol. Cell. Biol. 7,3826-3829. GARFINKEL, L., PERIASAMY, M., and NADAL-GINARD, B. (1982). Cloning and characterization of cDNA sequences corresponding to myosin light chains 1, 2, and 3, troponin-C, troponin-T,
VOLUME 145, 1991

karyons suggests that human muscle regulatory factors are produced. Cell 47, 123-130. HARDISON, R., BUTLER, E., LACY, E., and MANIATIS, T. (1979). The structure and transcription of four linked rabbit B-like globin genes. Cell 18, 1285-1297. IZUMO, S., NADAL-GINARD, B., and MAHDAVI, V. (1986). All members of the mgosin heavy chain multi-gene family respond to thyroid hormone in a highly tissue specific manner. S’cic~rcce231,597-600. JOLESZ,F., and SRETER, F. (1981). Development, innervation, and activitg-induced changes in skeletal muscle. ilnt/?r. Rev. Physiol. 43,531552. JONES, P., and TAYLOR, S. (1980). Cellular differentiation, cytidinc analogs and DNA methplation. Cell 20, 85-93. LEES-MILLER, J., GOODWIN, L., and HELFMAN, D. (1990). Three novel brain tropomyosin isoforms are expressed from the rat a-tropomyosin gene through the use of alternative promoters and alternative RNA processing. ,blol. Cell. Biol. 10, 1729-1742. MAHDAVI, V., STREHLER, E., PERIASAMY, M., WIECZOREK, D., IZUMO, S., and NADAL-GINARD, B. (1986). Sarcomeric myosin heavy chain gene family: Organization and pattern of expression. Med. Sri. Spor~ts Med. 18, 299-308. MEDFORD, R., NGUYEN, H., and NADAL-GINARD, 8. (1983). Transcriptional and cell cycle-mediated regulation of myosin heavy chain gene expression during muscle cell differentiation. J. Biol. Chcnc. 258, 11,063~11,07:3. MILLER, S.. PAVLATH, G., BLAKELY, B., and BLAU, H. (1988). Muscle cell components dictate hepatozyte gene expression and the distribution of the Golgi apparatus in heterokarpons. Gerles De,*. 2, 330m 340. NABESHIMA, Y., FUJII-KURIYAMA, Y., MURAMATSU, M., and OGATA, K. (1984). Alternative transcription and two modes of splicingresult in two myosin light chains from one gene. No:ature 308,333-338. NADAL-GINARD, B. (1978). Commitment, fusion, and biochemical differentiation of a myogenic cell line in the absence of DNA synthesis. C’el/ 15, 855-864. OBINATA, T., MASAKI, T., TAKANO-OHMURO, H., TANAKA, T., and SHIMIZU, N. (1983). Coexistence of cardiac-type and fast skeletal-type myosin light chains in embryonic chicken cardiac muscle. J. Biorhen~. ITokyo) 94, 1025-1028. PAVLATH, G., RICH, K., WEBSTER, S., and BLAU, H. (1989). Localization of muscle gene products in nuclear domains. ,?rclture 337, 570-573. PERIASAMY, M., STREHLER, E., GARFINKEL, L., GUBITS, R., RUIZOPAZO, N., and NADAL-GINARD, B. (1984). Fast skeletal muscle myosin light chains 1 and 3 are produced from a single gene by a combined process of differential RNA transcription and splicing. J. Bid. Chrrn. 259, 13,595-13,604. PERLMUTTER, R., KEARNEY, J., CHANG, S., and HOOD, L. (1985). Developmentally controlled expression of immunoglobutlin VH genes. Sciercce 227, 1597-1601. RALSTON, E., and HALL, Z. (1989). Transfer of a protein encoded by a single nucleus to nearby nuclei in multinucleated myotubes. Science 244, 1066-1069. RIGBY, P., DIECKMANN, M., RHODES, C., and BERG, P. (1977). Labeling deoxyribonucleic acid to high specific activity in uitm by nick translation with DNA polymerase I. J. Mol. Biol. 113, 237-251. ROBERT, B., DAUBAS, P., AKIMENKO, M., COHEN, A., GARNER, I., GUENET, J., and BUCKINGHAM, M. (1984). A single locus in the mouse encodes both myosin light chains 1 and 3, a second locus corresponds to a related pseudogene. Cell 39,129-140. RUSSELL, S., CAMBON, N., NADAL-GINARD, B., and WHALEN, R. (1988). Thyroid hormone induces a nerve-independent precocious expression of fast myosin heavy chain mRNA in rat hindlimb skeletal muscle. J. Bid. Chmx. 263, 6370-6374. SOEIRO, R., BIRNBOIM, H., and DARNELL, J. (1966). Rapidly labeled

HeLa cell nuclear RNA. II. Base composition and cellular localization of a heterogenous RNA fraction. J. &i’ol. Biol. 19, 362-3’72. STREHLER, E., PERIASAMY, M., STREHLER-PAGE, M., and NADAL-GINARD, B. (1985). Myosin light-chain 1 and 3 gene has two structurally distinct and differentially regulated promoters evolving at different rates. Jfol. G/I. Biol 5, 3168-3182. TASSIN, A., MARO, B., and BORNENS, M. (1985). Fate of microtubule-organizing centers during myogenesis in t~itro. J Cell Rid. 100, 35-46. TAUBMAN, M., SMITH, C., IZUMO, S., GRANT, J., ENDO, T., ANDREADIS, A., and NADAL-GINARD, B. (1989). The expression of sarcomeric muscle-specific contractile protein genes in BC3Hl cells: BC3Hl cells resemble skeletal myohlasts that are defective for commitment to terminal differentiation. J. c’e// Riol. 108, 1799-1806. THOMAS, P. (1980). Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. E’roc. 1V~tl. ilctrrl. Sri. USA 78,2843-2847. I:ETSUKI, T., NABESHIMA, Y.. FUJISAWA-SEHARA, A., and NABESHIMA, Y. (1990). Regulation of the chicken embryonic myosin light-chain (L23) gene: Existence of a common regulatory element shared by myosin alkali-chain genes. ~Uol. C’fU. Biol. 10, 2562-2569. ~VEISBLUM, B., and HAENSSLER, E. (1974). Fluorometric properties of the bibenzimidazole derivative Hoechst 33258, a fluorescent probe specific for AT concentration in chromosomal DNA. C’hrowoso~~~tr 46, 25,i-260. WHALEN, R., TOUTANT. M., BUTLER-BROWNE, G., and WATKINS, S.

(1985). Hereditary pituitary dwarfism in mice affects skeletal and cardiac myosin isozyme transitions differently. ,I. Cd. Bid. 101, 603-609. WIECZOREK, D. (1984). Early effects its vitroof the muscular dvsgenesis mutation on nervous tissue in the mouse. dfwscle Xvrw 7, 179193. WIECZOREK, D. (1988). Regulation of alternatively spliced // Biol. 101, 618-689. WIECZOREK, D., SMITH, C., and NADAL-GINARD, B. (1988). The rat wtropomyosin gene genera& a minimum of six different mRNAs coding for striated, smooth, and nonmuscle isoforms hy alternative splicing. Jfol. CWI. Biol. 8, 679-694. WOOD, W., BUNCH, C., KELLY, S., GUNN, Y., and BRECKON, G. (1985). Control of haemoglohin switching by a developmental clock? ,Vcrtuw 313, 320-323. WRIGHT, W., and LIN, \‘. (1990). Uncoupling of muscle-specific protein expression in myocyte X myoblast htterokaryons. Cc~ll Lh$br. DPP. 29 7 113~122. c YAFFE, D. (1968). Rrtention of differentiation potentialities during prolonged cultivation of rn>-ogcnic cells. 1’roc. Avert/. ilccctl. Sci. lB.4 61, 477-483.