The muscular dysgenesis mutation in mice leads to arrest of the genetic program for muscle differentiation

DEVELOPMENTAL

BIOLOGY

133,456-467

(1989)

The Muscular Dysgenesis Mutation in Mice Leads to Arrest of the Genetic Program for Muscle Differentiation Department

NIRUPA

CHAUDHARI

AND

of Physiology,

Colorado

State

Accepted

February

KURT

University,

G. BEAM Fort

Collins,

Colorado

80523

13, 1989

Muscular dysgenesis (mdg) is a mutation in mice which causes the failure of excitation-contraction coupling in skeletal muscle. Although the sarcolemma, the sarcoplasmic reticulum, and the contractile apparatus all maintain nearly normal function, sarcolemmal depolarization fails to cause calcium release from the sarcoplasmic reticulum. Recently, the primary genetic defect in this mutation was shown to be located in the structural gene for the dihydropyridine receptor. We have examined the developmental expression from Fetal Day 15 onward, in normal and mutant muscle, of several unidentified genes as well as genes which are known markers of muscle differentiation. We find that the majority of mRNA sequences are found at similar concentrations in normal and dysgenic muscles at birth. Many differentiation-related genes also are expressed at normal levels early during myogenesis in mutant mice. However, as late fetal development progresses in dysgenic muscle, the mRNA concentrations for these genes fail to undergo the rapid rise which is characteristic of normal muscle. Several additional, unidentified genes, which normally would be down-regulated during development, remain expressed at a high level in dysgenic muscle. Thus, the primary absence of a functional dihydropyridine receptor appears to prevent the changes in gene expression which are necessary for maturation of skeletal muscle. 0 1989 Academic Press, Inc.

INTRODUCTION

Muscular dysgenesis (mdg) is a recessive, single gene mutation of mice which arose spontaneously (Pai, 1965a). In the homozygous state the mutation causes complete paralysis of skeletal muscle and hence leads to respiratory failure and perinatal death. Many aspects of mutant skeletal muscle physiology are similar to the wild type. The sarcolemma of dysgenic skeletal muscle is electrically excitable, generating action potentials both spontaneously and in response to cholinergic stimulation (Powell and Fambrough, 1973). The contractile apparatus of mutant muscle is also functional, since elevation of myoplasmic calcium (elicited experimentally with ionophore) produces contractures (Klaus et aL, 1983). The experimentally observed contractures of mutant muscle in response to caffeine (Bowden-Essein, 1972) imply a functional sarcoplasmic reticulum (SR). Thus, the paralysis seems to involve a specific defect in excitation-contraction (E-C) coupling, the series of events that link sarcolemmal depolarization to release of calcium from the SR (Powell and Fambrough, 1973; Klaus et ah, 1983). Although the mechanism of E-C coupling is not yet understood in its entirety, it probably involves a voltage sensor protein with subunits or domains in common with the dihydropyridine-sensitive calcium channel of skeletal muscle (Rios and Brum, 1987). In keeping with 0012-1606/89$3.00 Copyright All rights

0 1989 by Academic Press, Inc. of reproduction in any form reserved.

456

this, electrophysiological studies have demonstrated that dysgenic skeletal muscle is specifically deficient in a slow, dihydropyridine-sensitive calcium current, whereas a dihydropyridine-insensitive calcium current remains unaffected (Beam et al, 1986). The mutation does not affect calcium currents of heart or sensory neurons. The specific absence of a slow calcium current in skeletal muscle suggests that the primary gene altered in muscular dysgenesis might be essential for the function of this channel (Beam et al, 1986). However, additional phenotypic characteristics have been noted for dysgenic animals, such as the loss of skeletal muscle mass (Pai, 1965b), poor organization of myofilaments (Klaus et al., 1983), altered basal lamina (Pincon-Raymond et ah, 1987), multiple innervation of single fibers (Rieger and Pincon-Raymond, 1981), incomplete maturation of motor synapses (Rieger et ah, 1984), and a retardation of motor neuron cell death (Oppenheim et al., 1986). These characteristics are perhaps all secondary effects stemming from paralysis of the muscle. Alternative models for the primary gene defect also have been forwarded. These models have invoked a missing trophic signal that is essential for normal muscle development and is supplied by motoneurons or extramuscular tissue (Pincon-Raymond et ah, 1987; Rieger et al, 1987). The mutation has also been viewed as simultaneously affecting the development of several separate cell lin-

CHAUDHARI

AND

BEAM

Muscular

eages (Wieczorek, 1984). These alternative interpretations are called into question by the results described below. Few molecular analyses of this mutation were carried out until recently. Pincon-Raymond et al. (1985) demonstrated that dysgenic skeletal muscle contains a fivefold reduced number of receptors for the dihydropyridine class of calcium channel blockers. They also found a concurrent reduction in transverse tubules (T-tubules) and an absence of triads, (the region of association between T-tubules and SR at which E-C coupling takes place). Furthermore, Tanabe et al. (1988) have shown that the primary gene defect in muscular dysgenesis is probably located within the structural gene for the skeletal muscle dihydropyridine receptor, as demonstrated through restriction fragment variations for this gene between normal and mutant genomes. In addition, a normal phenocopy could be obtained by inserting a functional copy of the gene in mutant myotubes (Tanabe et al., 1988). The most straightforward interpretation of these results is that all the phenotypic changes associated with muscular dysgenesis are a consequence of the loss of the dihydropyridine receptor of skeletal muscle. How does the mdg mutation lead to numerous phenotypic alterations of skeletal muscle? To gain insight into this question, we have determined how widespread the changes in gene expression are in mutant muscle. Additionally we have compared the expression of muscle-differentiation genes between mutant and normal skeletal muscles. These analyses suggest that the loss of a functional dihydropyridine receptor prevents appropriate expression of a variety of genes essential for muscle maturation. Preliminary descriptions of some of these results have appeared in abstract form (Chaudhari and Beam, 1987, 1988). MATERIALS

AND

METHODS

Animals

The muscular dysgenesis mutation was bred into the 129/ReJ strain of mice (Jackson Laboratory) by Dr. Jeanne Powell at Smith College (Northampton, MA), and a colony derived from this stock was established at Colorado State University. The mutation is maintained in heterozygotes, which appear functionally normal. Fetal mice were obtained by killing timed pregnant females and were further staged according to Rugh (1968). Dysgenic fetuses (mdg/mdg) were distinguished from their phenotypically normal littermates (+/mdg?) by the absence of muscle movements in response to electrical stimulation with a bipolar electrode.

Dysgenesis

457

Mutation

RNA PuriJication

RNA was obtained from tissues dissected from fetal and postnatal mice within 10 min of death of the mother or young. The only exception was for newborn dysgenic mice, since these die immediately following birth. For these animals, tissue was obtained as soon after birth as possible and in no instance longer than 2 hr after death. Limbs as well as trunk were used as sources of skeletal muscle for fetal and postnatal mice. Cytoplasmic RNA used for constructing the cDNA library was purified from a postmitochondrial supernatant (Bantle and Hahn, 1976) using guanidine isothiocyanate. RNAs used for generating complex cDNA probes and for blot hybridization analyses were total tissue RNA, purified by homogenizing frozen tissues directly into 4 M guanidine isothiocyanate and pelleting RNA through a 5.7 M CsCl cushion (Ullrich et ah, 1977). Poly(A)-containing RNAs were purified on oligo(dT)-cellulose (type 3, Collaborative Research, Lexington, MA) as described by Aviv and Leder (1972). Library Construction

Cytoplasmic poly(A) RNA was used as a template for the synthesis of double-strand cDNA essentially by the method of Gubler and Hoffman (1983). The cDNA was treated with EcoRI methylase and made blunt-ended with DNA Polymerase I. EcoRI linkers were ligated to the ends with T4 DNA polymerase and cut back with EcoRI, all following standard procedures. The cDNA was then ligated into the pKS.M13-plasmid (Stratagene Inc., La Jolla, CA), which had been cut with EcoRI and treated with alkaline phosphatase. A small library of approximately lo* transformants was obtained in E. coli HBlOl. Colony Hybridization

Screening

Colony hybridizations were performed essentially following the procedure of Hanahan and Meselson (1980). Approximately 5000 transformants on a total of four loo-mm petri plates were grown (i.e., at a moderate colony density) and lifted onto duplicate nitrocellulose filters. The filters were processed to lyse the bacteria and immobilize plasmid DNA. The duplicate sets of filters were hybridized with 32P-labeled complex cDNA probes generated from either normal or dysgenic newborn mouse skeletal muscle poly(A) RNA. These cDNA probes, representing all the mRNAs expressed in normal or mutant skeletal muscle, were synthesized using poly(A)-containing whole cell RNA as template, oligo(dT)12.18 as primer, and [a-32P]dCTP (ICN Radiochemicals, Irvine, CA) for internal labeling. Hybridizations were carried out at moderate stringency: 42°C in 50% formamide, 4X SSPE, 2~ Denhardt’s, 100 pg/ml

458

DEVELOPMENTALBIOLOGY VOLUME133.1989

(pMACT-o() of the mouse skeletal Lu-actin gene which single-strand herring DNA, and 1 pg/ml poly(A). (1X SSPE is 150 mM NaCl, 10 mM phosphate buffer, pH 6.9, recognizes all actin mRNAs was obtained from Dr. N. 1 mM EDTA; 1X Denhardt’s is 0.02% bovine serum Davidson (Hu et al., 1986); rat embryonic myosin heavy albumin, 0.02% Ficoll, 0.02% polyvinylpyrrolidone.) chain cDNA (pMHC25) from Dr. B. Nadal-Ginard Probe, lo7 dpm (=500 ng), in 20 ml hybridization buffer (Medford et al, 1980); P-tropomyosin cDNA (pRSK-22) was used for each set of four nitrocellulose filters. After from Dr. D. Helfman (Helfman et al., 1986); mouse hybridization, filters were washed in a high-salt buffer muscle creatine kinase cDNA (pH 0.4) from Dr. S. (2X SSC, 0.1% SDS) at room temperature and then in a Hauschka (Chamberlain et ah, 1985); rabbit skeletal low-salt buffer (0.2X SSC, 0.1% SDS) at 42°C. SSC, 1X, muscle cDNA for Ca2+ + Mg2+-dependent ATPase (F9) is 150 mM NaCl, 15 mM Na-citrate, pH 7.0. Autoradiofrom Dr. D. MacLennan (MacLennan et al., 1985); graphic images of duplicate filters were aligned and cDNAs for mouse muscle acetylcholine receptor subcolonies which exhibited different hybridization signal units LY,& y, and 6 (pBMA47, pBMB49, pBMG419, and intensities for normal and mutant cDNAs were picked. pMBD451) from Dr. J. Boulter (Patrick et ah, 1987); These colonies were clonally purified through two furmyoD (pVZCI1) from Drs. H. Weintraub and A. Lassar ther rounds of duplicate hybridizations with the same (Davis et al., 1987), and a mouse genomic adenine phosprobes. phoribosyl transferase (APRT) gene (pSAM 3.1) from Dr. P. Stambrook (Dush et al., 1985). RNA Blot Hybridizations RESULTS RNA Northern blots were carried out using the method of Fourney et al. (1988). Poly(A) RNAs, isolated Fraction of Gene Expression Altered from the requisite tissues, were denatured and electroMessenger RNAs present at dissimilar concentraphoresed on 1.2% agarose gels containing 20 mMMOPS tions in dysgenic and normal newborn skeletal muscle (3-(N-morpholino)propanesulfonic acid) buffer and 2% can be identified by hybridizing a normal muscle reformaldehyde. The electrophoresed RNAs were then combinant library with cDNA probes representing the transferred, by capillary blotting in 10X SSC, to BiomRNA populations of normal and mutant muscles. A trans membrane (uncharged nylon from ICN BiochemicDNA library was constructed from mRNA extracted cals, Irvine, CA). Membranes were prehybridized for from muscle of normal (+/mdg?) newborn sibs of dys2-6 hr in 50% formamide, 4X SSPE, 4X Denhardt’s, 250 genie (m&/m&) pups. (Because homozygous normal pg/ml single-strand herring DNA, 10 pg/ml poly(A) at (+/+) and heterozygous (+/mdg) newborn mice cannot 42°C. Hybridizations were carried out in the same solube distinguished from one another, they are designated tion, with 32P-labeled nick-translated plasmids at a as +/mdg?). This normal library, composed of about final concentration of 10 rig/ml. Moderate stringency 10,000 recombinants, should represent the majority of hybridization conditions (42-44’C) were used for probes abundant to moderate, and a fraction of infrequent, derived from mouse genes. Lower stringency hybridizamRNA sequences present in normal muscle at birth. To tion, at 38”C, was employed for heterologous probes. determine the fraction of genes that muscular dysgenPosthybridization washes were carried out in 2X SSC at esis affects at the level of mRNA accumulation, we room temperature followed by 0.1X SSC, 0.1% SDS at subjected this library to differential colony hybridizaeither 38°C (low stringency) or 44°C (moderate strintion. gency). Some blots were probed sequentially with sevIn developing normal muscle, highly and moderately eral independent DNA sequences. A high temperature prevalent mRNAs together comprise approximately wash (as recommended by the manufacturer) was used 55% of total mRNAs by mass (Ordahl and Caplan, to strip off the previously hybridized probe and was 1978). This is the subset which would be expected to followed by another series of hybridization steps as yield a signal when the recombinant library from norabove. Autoradiographic films were scanned using a mal muscle is probed with cDNA from normal muscle. GS-300 linear densitometer from Hoefer Scientific. Of the 5000 colonies analyzed, approximately 35% gave Peak areas, representing band intensities, were intea discernible signal with the normal muscle probe grated under the assumption of a Gaussian distribution (Table 1). Specifically, about 4% of detectably hybridusing the GS-360 Data System software also from ized colonies gave very intense signals and 96% gave Hoefer. moderate to light signals. Many of the colonies that gave an intense signal for normal muscle gave a less Recombinant Plasmids intense signal for mutant muscle. These clones were not Many investigators generously provided us with the analyzed further since they are likely to represent genes cloned DNAs used in this study. A genomic fragment for contractile proteins. Instead, we analyzed differen-

CHAUDHARI

DIFFERENTIAL

AND BEAM

Muscular

intensity

Colonies Normal 4 41 40 8 4 4

+++ t+ + ++ + +

Dysgenic or ++ ++ + + or ~ -

Designation

+++

++

m&-equal I m&-under I m&-over

Note. A cDNA library constructed from normal newborn skeletal muscle was subjected to differential colony hybridization using cDNA probes for mRNAs of normal and dysgenic newborn muscles. Colonies yielding each type of signal are indicated as percentages of all clones detectably hybridized with the normal probe. Intensity of autoradiographic signal: dark (+++), medium (+t), light (t), and undetectable (-).

tial expression of genes for contractile proteins with defined cDNA probes (see below). The large majority of clones in the primary screening gave signals of similar intensity with both normal and mutant probes. About 12% of detectable colonies in the primary screening demonstrated a moderate to light intensity signal with the normal muscle probe and a significantly lower intensity signal with the dysgenic one. About 4% of detectably hybridized colonies examined in the primary screening exhibited a signal with the mutant probe darker than that with the normal one. These colonies were typically those that gave the lightest signal with normal probe. The initial screening by differential colony hybridization indicated that the muscular dysgenesis mutation affects mRNA level for only a subset of the genes expressed in newborn muscle. Based on this initial screening, 85 colonies were selected that appeared to represent mRNAs differentially expressed in newborn dysgenic muscle relative to normal (65 of these appeared to be under-expressed and 20 to be over-expressed). These colonies were subjected to second and then third rounds of the same screening procedure to confirm the differential signal for the two probes and also to achieve clonal purification of these recombinants. Following the third screening, six clones representing genes underexpressed and four clones representing genes overexpressed in mutant muscle were selected as representatives of each set and were used for subsequent analyses. A single clone was randomly selected as representative of the set of moderately abundant clones expressed equally in normal and dysgenic muscles.

459

Mutation

Tissue Speci$city

TABLE 1 EXPRESSION OF GENES REVEALED BY COLONY HYBRIDIZATION Signal

Dys,qenesis

of Expression

of Unknown

Clones

The relative concentrations of mRNAs corresponding to the unknown cDNA clones described above were compared between muscle and nonmuscle tissue using RNA blot hybridization (“Northern”) analysis. Liver and kidney were selected as representative tissues not expected to produce specialized proteins of muscle. Poly(A) RNA from each tissue was denatured, electrophoresed on formaldehyde-containing agarose gel, and transferred to nylon membrane. The membranes were hybridized with 32P-labeled plasmids selected above. Typical examples of such RNA blots are shown in Fig. 1. MM20 is a clone which showed a similar signal intensity, in colony hybridizations, for newborn normal and mutant muscle (“m&-equal”). MM29 is one of the six clones we examined that is underexpressed in mutant muscle (“mdg-under”) and MM81 is one of the four clones that were selected as overexpressed in mutant muscle (“mdg-over”). The similar signal intensities observed for normal and dysgenic RNAs hybridized with the mdg-equal clone demonstrate that muscular dysgenesis does not lead to large scale and nonspecific degradation of muscle mRNAs. It also demonstrates that significant degradation of RNA did not occur during the up to 2-hr postmortem period until tissue was collected from mutant neonates. The mdg-equal clone gave a hy-

FIG. 1. Comparative levels in skeletal muscle and nonmuscle tissues of representative, moderately abundant RNAs that are (A) comparably, (B) under-, or (C) overexpressed in dysgenic relative to normal skeletal muscle. These have been designated the “m&-equal,” “mdgunder,” and “m&-over” classes, respectively, in Table 1. Two microgram samples of poly(A) RNA, purified from each of the indicated tissues of newborn mice, were denatured, electrophoresed, and transferred to nylon membranes. The blots were probed with the indicated “P-labeled plasmids and subjected to autoradiography. The cDNA inserts of these plasmids correspond to unknown skeletal muscle mRNAs and were selected by colony hybridization (see Methods). The size of each mRNA was calculated based upon mobility relative to denatured DNA fragments (pBR322 digested with HinfI) run in parallel lanes. Averaging from two to four separate RNA blots for each probe, the mRNAs corresponding to MM20, MM29, and MM81 are 1700, 900, and 750 nucleotides long, respectively.

460

DEVELOPMENTALBIOLOGY

bridization signal not only in muscle but also in all other tissues tested including kidney and liver (Fig. 1A) and brain, lung, and heart (not shown). The mdg-under clone corresponds to a mRNA that is present in normal and dysgenic muscles but is not detectable in either liver or kidney (Fig. 1B). A similar pattern was observed for the other five mdg-under clones examined. Thus, the expression of all six genes in this class appears to be limited to muscle (although we cannot rule out the possible presence of these mRNAs in other tissues not tested). For the six mdg-under clones, mRNA concentration in normal newborn muscle exceeded that in dysgenic newborn muscle by a factor of 5-20. Each of the six mdg-under clones examined hybridized to a mRNA of a different size (approximately 800, 900, 1000, 1700, 2000, and 6000 nucleotides, respectively), implying that each clone represents a distinct gene. Steady-state concentrations of mRNA corresponding to the four mdg-over clones were three- to fivefold higher in dysgenic than in normal newborn mouse muscle (e.g., Fig 1C). In contrast to the mdg-under clones, none of the clones in this class showed a muscle-specific pattern of expression. Hybridizable bands of mRNA were detected in at least one other tissue for each of these four clones. For example, a faint band was visible in kidney for MM81 (Fig. 1C). Expression of UnidentiJied Clones during Normal Myogenesis Messenger RNAs underexpressed in neonatal dysgenie muscle may represent gene products essential for the mature muscle phenotype. To address this issue we examined the developmental pattern in normal muscle for expression of mRNAs corresponding to mdg-under clones. The pattern is illustrated for two mdg-under clones in Fig. 2A. For each, mRNA hybridizing to the cloned cDNA was very sparse or undetectable at the earliest stage tested (Fetal Day 14; term is 20 days) but accumulated progressively as normal muscle development proceeded. A similar progression was observed for the other four mdg-under clones, although variations were seen with respect to the stage at which maximum concentration was reached as well as the total concentration of the mRNA. The normal developmental expression of mRNA corresponding to two mdg-over clones is shown in Fig. 2B. For each, the concentration of its corresponding mRNA decreased during normal muscle development. This pattern was also exhibited by the other two mdg-over cDNAs which were analyzed. Variations were seen between the members of this class both in total concentration of mRNA and in how rapidly each species di-

VOLUME 133,1989

minished postnatally. In the case of MM55 the fetal mRNA appears to be replaced by a slightly larger mRNA at later postnatal ages. Expression of Unident$ed Clones during Dysgenic Myogenesis The colony hybridization analysis had identified mRNAs which were differentially expressed in dysgenie and normal muscles at birth. Such differential mRNA concentrations might result either from a developmental arrest or from a developmental lag of dysgenie muscle. Thus, we examined the relative concentrations of these unidentified mRNAs in dysgenic and normal muscles during fetal development using RNA blot hybridization. As shown in Fig. 3, the concentration of mRNA specific for the “mdg-equal” clone, MM20 was similar in normal and dysgenic muscle at each stage. The concentration of mRNA for the “mdg-under” clone, MM29, was equivalent at the earliest stage tested but failed to rise very much in dysgenic relative to normal muscle. mRNA concentrations for the “mdg-over” clone declined in normal muscle from Fetal (F) Day 15 to Postnatal Day 0 (Figs. 2 and 3). The level is dysgenic muscle was comparable to normal at F15 and F17 but then experienced a sharp increase immediately before birth. A similar pattern was obtained with one additional “mdg-over” clone probe on an independent blot. These observations suggest that the muscular dysgenesis mutation leads to alteration of the mRNA concentrations for a subset of genes expressed in muscle. Genes whose products accumulate later in development and primarily in muscle (presumably the muscle differentiation genes) are underrepresented. Other genes normally expressed early during muscle development and in other tissues (presumably undifferentiated state markers) remain expressed at a high level in mutant muscle. The majority of mRNAs (the mdg-equal class) are represented at equal concentrations in dysgenic and normal muscles throughout fetal development. Expression of Known Muscle Diflerentiation

Genes

In addition to examining the developmental expression of representative unknown genes, we also examined the expression of defined markers of muscle differentiation. For this analysis, mRNA from normal and dysgenic muscles was subjected to blot hybridization with defined cDNA probes obtained from a number of investigators. Genesfor contractile proteins. Muscle a-actin, myosin heavy chain (MHC), and P-tropomyosin (P-TM) were selected as three proteins that are directly involved in the contraction process. Cloned cDNAs for these three proteins were used sequentially to probe a single RNA

CHAUDHARI

AND

BEAM

Muscular

Dysgenesis

461

Mutation

MM55

TLFcTT%

ml

DAYS

DAYS

MM74

DAYS

MM 81

rr” DAYS

FIG. 2. (Top) Expression, during development of normal muscle, of mRNAs corresponding to representative cDNA clones (A) underexpressed or (B) overexpressed in dysgenic relative to normal newborn skeletal muscle. Poly(A) RNA was purified from skeletal muscle of mice at Fetal (F) Days 14, 16, and 19 and Postnatal (P) Days 0 (i.e., newborn), 5, 10, 15, and 23. Five-microgram samples of each RNA were used for blot hybridization analysis with the indicated “P-labeled plasmids (see Materials and Methods). The size of each mRNA was calculated based upon mobility relative to denatured DNA fragments (pBR322 digested with HinfI) run in parallel lanes. The mRNAs corresponding to MM41, MM74, MM55, and MM81 are 1650,2000,800, and 740 nucleotides long, respectively. The apparently lower intensity at P15 is probably due to loading inaccuracy in this lane. Additional blots probed with these plasmids and others from the “mdgs-over” class did not substantiate a consistently lower concentration of specific mRNA at P15. (Bottom) Autoradiographs for each RNA blot were scanned densitometrically. The signal intensity for each developmental age was compared to the intensity obtained at the earliest age at which mRNA corresponding to the probe could be detected (F14 for MM41, MM55, and MM81; F16 for MM74).

blot containing muscle poly(A) RNA isolated from mice of varying gestational age. To allow direct comparison of normal and mutant muscle, RNA was prepared from affected and normal sibs from single litters at each age. The levels of a-a&in mRNA, MHC mRNA, and P-TM mRNA are similar in normal and dysgenic muscles at Fetal Day 15 (Fig. 4A). During the development of normal muscle, the concentrations of all three mRNAs increase substantially. In contrast, the concentration of

each mRNA in dysgenic muscle remains nearly constant from Fetal Day 15 through birth. The actin clone used in this study hybridizes not only with the musclespecific ol-actin mRNA, but also with two comigrating mRNAs (/3 and 7) that encode the ubiquitous cytoskeletal actins. This band thus serves as a control to demonstrate equal loading of mRNA in the normal and mutant pairs of lanes (Fig. 4A). Genes jbr muscle-speci&c enzymes. Creatine kinase

462

DEVELOPMENTAL

F15 fm+r

BIOLOGY

A

B

MM20

MM 29

F17

PO m

+

m

/!I;( F15 F17 DAYS

PO

VOLUME

F17

--- F15 +m

PO

fm

133, 1989

F15

F17

+m+r

+m

PO In

+

m

/!!;;L F15

F17

PO

DAYS

F15

F17 DAYS

PO

FIG. 3. (Top) Expression, during fetal development of normal and dysgenic skeletal muscle, of mRNAs corresponding to representative cDNA clones from the (A) “n&g-equal,” (B) “mdg-under,” and (C) “mdg-over” classes (as defined in Table 1). Poly(A) RNA was purified from skeletal muscle of normal (+) and dysgenic (m) siblings at each of Fetal (F) Days 15 and 17 and Postnatal (P) Day 0 (i.e., newborn). Eight micrograms of RNA was used for each lane. The blot was hybridized sequentially with each of the indicated 32P-labeled probes. The sizes of mRNAs corresponding to MM20, MM29, and MM81 are 1700, 900, and 750 nucleotides, respectively. (Bottom) The autoradiographs shown above were scanned densitometrically. The absence of loading variations was confirmed for each pair of lanes (normal and dysgenic) using APRT probe as in Fig. 5. Signal intensity at each stage is depicted relative to the intensity at Fetal Day 15 of normal skeletal muscle development. (A) Normal muscle RNA; (0) dysgenic muscle RNA. Similar plots of relative intensity were obtained with two independent blots hybridized with each of these probes.

(CK) is an enzyme involved in energy production in muscle cells and Ca2+ + Mg2+ ATPase is the enzyme responsible for transporting calcium from the myoplasm into the SR. The level of mRNA for each enzyme increases 5- to lo-fold during late fetal development in normal muscle (Fig. 4B). In contrast, mRNA concentrations for these enzymes in dysgenic muscle increase only approximately twofold during the same period. Genes for sarcolemmal ion channels. Receptors for acetylcholine are expressed on the sarcolemmal surface very soon after the formation of myotubes (Patrick et al., 1972) and are vital for neuromuscular communication. Figure 5A illustrates mRNA levels for three of the subunits of the acetylcholine receptor. Because the CYsubunit mRNA is present at a higher concentration than the other subunit mRNAs, the alteration in its expression in mutant muscle is clearly visible. In normal muscle, the concentration of a-subunit mRNA remained constant between Fetal Days 15 and 17 and increased significantly during late fetal development. In dysgenic muscle, on the other hand, the concentration of this species decreased toward the end of fetal development. Similar patterns were also observed when cDNAs for the /3 (Fig. 5A) and y (not shown) subunits of the acetylcholine receptor were used to probe the same blot.

For all the muscle differentiation genes examined (Figs. 4 and 5A), mRNA concentrations increased from Fetal Day 15 onward in normal muscle; mRNAs for these genes are present in dysgenic muscle at Fetal Day 15 but subsequently either fail to increase (Fig. 4) or actually decrease (Fig. 5A). The y-subunit of the acetylcholine receptor is an embryonic form that is replaced by the E-subunit during early postnatal development (Mishina et al., 1986). Thus, the concentration of its mRNA decreases in normal muscle as shown in Fig. 5C (leftmost three lanes). Interestingly, the concentration of this mRNA follows the same developmental pattern in dysgenic muscle. Neither random RNA degradation nor loading inaccuracies account for the decreased concentration for the acetylcholine receptor mRNAs observed during late development of dysgenic muscle. This was demonstrated by stripping the blot shown in Figs. 5A and 5B and reprobing it with cDNA for adenine phosphoribosyl transferase (APRT). This housekeeping enzyme is expressed in roughly equal concentrations in all mouse cells tested (not shown). This probe reveals a substantial level of mRNA in neonatal dysgenic muscle (Fig. 5C, rightmost lane), the same lane which illustrates markedly decreased mRNA concentrations for the acetylcholine receptor subunits.

CHAUDHARI

AND BEAM

A F15 -m+ +

F17 m

Muscular

+

m

F17

PO

+m

+m

ATPase MHC

463

Mutation

Myogenic D$Terentiation

B F15 --P +m

PO

Dysgenesis

and Determination

Recently, Davis et al. (1987) reported on a cDNA, called myoD, derived from a gene which appears to be located near the beginning of the route for myogenic determination. Expression of this gene induces nonmuscle cells in culture to initiate the muscle differentiation program. We examined the expression of this gene during muscle development and found that a fairly steady concentration of myoD mRNA could be detected in normal muscle from Fetal Day 15 through birth (Fig. 5D). In dysgenic muscle, this mRNA initially accumulates at levels comparable to normal muscle but is depleted substantially at the end of fetal development. It is not possible to determine from this analysis if the down-regulation of myoD in mutant muscle plays a causal or an incidental role in the alteration of expression of the muscle differentiation genes that we have examined. DISCUSSION

F15

F17

PO

“-

DAYS

DAYS

F15 DAYS

F15

F17

The major finding reported here is that muscular dysgenesis alters mRNA concentrations for several categories of genes normally expressed in skeletal muscle. Roughly speaking, the effect of the mutation is to arrest the expression of muscle differentiation genes. Thus, at Fetal Day 15, mRNAs for all the known differentiation-related genes that we have examined are expressed at similar levels in normal and dysgenic muscles. As dysgenic fetal development proceeds, concentrations of these muscle-specific mRNAs fail to show the large increase that occurs in normal muscle. By the time affected fetuses are born, alterations in expression of a variety of genes become apparent. These alterations appear to affect primarily genes involved in muscle differentiation, as demonstrated both with defined gene

F17

PO

DAYS

PO

DAYS

FIG. 4. (Top) Expression of genes encoding proteins of the contractile apparatus (A) and two muscle enzymes (B) during development of normal and dysgenic skeletal muscle. Poly(A) RNA was purified from skeletal muscle of normal (+) and dysgenic (m) siblings at each of Fetal (F) Days 15 and 17 and Postnatal (P) Day 0 (i.e., newborn). A single RNA blot, containing 8 fig of each poly(A) RNA, was prepared and sequentially hybridized with each of the indicated a2P-labeled probes. Each hybridization was followed by autoradiography and

washes to strip the probe. By eliminating loading and transferring variabilities, this procedure allowed direct comparison of all the probes: actin, myosin heavy chain (MHC), P-tropomyosin (TM), creatine kinase (CK), and Ca*+ + Mga+ ATPase. The actin probe hybridizes with the nonmuscle 0 and y isoforms (which comigrate) as well as with the muscle 01 isoform. (Bottom) Autoradiographs from each hybridization were scanned densitometrically. The signal intensity for each developmental age was compared to the intensity for normal muscle RNA at Fetal Day 15. Each graph depicts the change in concentration of a specific mRNA during normal (A) and dysgenic (0) skeletal muscle development. For the actin probe hybridization, the dense signal obtained for the muscle-specific o-actin band made quantification of the p- and y-actin bands difficult. The intensity of each band was estimated by fitting a Gaussian curve for these overlapping peaks (see Materials and Methods). The same blot was also probed with a a2P-labeled gene fragment for APRT, a housekeeping enzyme. Signal intensity for this mRNA was very similar between normal and dysgenic muscle at each stage of development (_+20%).

464

DEVELOPMENTAL f

BIOLOGY In

F17

PO

F15

133,1989 +

C

A F15

VOLUME

F17

PO

m

F15

F17

PO

F15

F17

PO

F15

F17

PO

F15

F17

PO

8”.,”

AChR

Oc

APRT

f

B

IJ F15

AChR

Ill

F17

PO

F15

F17 .

PO

7

Myo D

F15

F17

PO

DAYS

F15

F17

F15

F17 DAYS

PO

PO

DAYS

DAYS

DAYS

FIG. 5. (Top) Expression of genes encoding muscle acetylcholine receptor subunits (A, B), a housekeeping enzyme to be involved in the determination of cells toward a muscle lineage (D). Skeletal muscle of normal (+) and dysgenic prepare poly(A) RNA for Fetal Days 15 and 17 and Postnatal Day 0 (i.e., newborn). A single RNA blot, containing

(C), and a protein thought (m) fetal sibs was used to 8 fig of each poly(A) RNA,

CHAUDHARI

AND BEAM

Muscular

probes as well as with representative unknown cDNA probes. Although the overall differentiation program may be seen as arrested by muscular dysgenesis, diverse points of arrest are seen for specific genes. Thus, starting from Fetal Day 15, the concentrations of mRNA in mutant muscle for creatine kinase and Ca2+ + M$+ ATPase increase slightly, those for the contractile proteins remain essentially constant, and those for the acetylcholine receptor subunits and myoD experience a substantial drop. That muscular dysgenesis does not produce a uniform, simple arrest of development is emphasized by the effects of the mutation on mRNA for the y-subunit of the acetylcholine receptor. During normal muscle development, mRNA for the y-subunit decreases as this subunit is replaced by the c-subunit (Mishina et al., 1986). If muscular dysgenesis simply arrested muscle development, one would expect levels of the y-subunit mRNA to remain high in mutant muscle, whereas we found that these levels actually fall at the end of fetal development. Although the genes for muscle differentiation are often thought of as coordinately regulated during myogenesis, our results demonstrate that these genes are still subject to diverse control mechanisms. Such diverse mechanisms have also been demonstrated for liver-specific genes (Isom et al., 1987). On the basis of the expression of known housekeeping genes (p- and y-a&in and APRT) as well as an unidentified, ubiquitously expressed gene (MMZO), there is little or no random degradation of mRNA. This was also evidenced by the similar signal intensities with dysgenic and normal probes for the majority of cDNA clones examined by differential colony hybridization (see Table 1). The absence of random degradation of mRNA in dysgenic muscle is significant in light of early reports on the gross loss of muscle mass and disorganization of myofibrils (Pai, 196513). In this context, it is also important to note that although the yield of muscle per newborn dysgenic mouse was low, the yield of poly(A) RNA per mg of muscle was similar between normal and mutant animals. Newborn dysgenic muscle contains high concentrations of several mRNAs which appear to be part of the

Dysgenesis Mutation

465

immature muscle cell’s repertoire. Levels of these mRNAs decrease during maturation of normal muscle. Newborn dysgenic muscle continues to express such mRNAs at a high level, probably as an indication of its immature state. At a functional level, the most pronounced effect of muscular dysgenesis is a complete disruption of E-C coupling in skeletal muscle. In normal skeletal muscle, depolarization of the T-tubules causes calcium to be released from the sarcoplasmic reticulum and thus leads to contraction. In dysgenic muscle, depolarization fails to produce the elevation of myoplasmic calcium necessary for contraction. Current models of E-C coupling envision a voltage-sensing protein, in the T-tubular membrane, which controls the release of calcium from the sarcoplasmic reticulum. The slow calcium channel located in the T-tubular membrane or a protein having subunits or domains in common with the slow calcium channel is a likely candidate for this voltagesensing function (Rios and Brum, 1987). Indeed dysgenie muscle lacks not only E-C coupling but also the slow calcium current (Beam et al., 1986). It now appears (Tanabe et al., 1988) that both these deficits are a direct result of the primary gene defect in muscular dysgenesis, namely, an alteration of the structural gene for the skeletal muscle receptor for dihydropyridines (organic compounds that block the slow calcium current of skeletal muscle). The dihydropyridine receptor, whose primary sequence shows substantial similarity to the voltage-dependent sodium channel (Tanabe et ab, 198’7), is a major protein of the T-tubule membrane (Fosset et al., 1983). Although the abolition of E-C coupling is the most striking deficit produced by muscular dysgenesis, additional significant alterations have also been described. At the ultrastructural level, dysgenic muscle has a paucity of T-tubules and lacks triads (the region of specific association of T-tubules and the sarcoplasmic reticulum). One might imagine that these effects on membrane ultrastructure follow directly from the absence of the dihydropyridine receptor. For example, the structural association, during development, between primitive T-tubules and sarcoplasmic reticulum may be un-

was prepared and sequentially hybridized with all the =P-labeled probes in this figure in order to eliminate loading and transferring variabilities. Results similar to those shown here were observed with blots of RNA obtained from two additional independent samples of receptor subunits; APRT, adenine phosphoribosyl transferase. MyoD is a cDNA normal and dysgenic muscle. AChR u, & y, acetylcholine which, upon transfection into a variety of cell types, induces myogenic differentiation in culture, and is therefore thought to represent a gene involved in myogenic determination (Davis et al, 198’7). (Bottom) The autoradiographs shown above were scanned densitometrically. Quantitation of APRT mRNA band intensities indicated that similar amounts of mRNA had been loaded in normal and dysgenic muscle lanes at each stage of development (?20’%). The only exception was that the dysgenic F15 lane contained a twofold higher level of APRT mRNA. The values obtained for band intensities for all probes (including APRT) were corrected by twofold only for the dysgenic F15 lane in order to compensate for this loading inaccuracy. Signal intensity at each stage is depicted relative to the intensity at Fetal Day 15 of normal skeletal muscle development. (A) Normal muscle RNA; (0) dysgenic muscle RNA.

466

DEVELOPMENTAL BIOLOGY

stable in the absence of the dihydropyridine receptor and such instability might prevent the normal maturation of T-tubules. Many other changes which have been reported in muscular dysgenesis are probably indirect results of the absence of E-C coupling and the attendant changes in calcium metabolism. These might include the disorganization of myofilaments and loss of muscle mass (Pai, 1965b, Klaus et ah, 1983). Presumably, such changes would follow from the primitive state of expression of many muscle differentiation genes as we have shown here. In addition to examining mRNA for genes of known function, we also examined mRNA levels for myoD. Activation of this gene is postulated to commit stem cells to the myogenic pathway (Davis et al, 1987). No previous reports have appeared describing mRNA levels for this gene during in viva muscle development. We find that the mRNA level for myoD in skeletal muscle remains fairly constant between Fetal Day 15 and Postnatal Day 15, a period during which substantial myogenesis and muscle maturation are occurring (Platzer, 1978; Betz et ah, 1979). The continued expression of myoD could be involved in maintaining normal muscle development. In dysgenic muscle, myoD mRNA levels are equivalent to those in normal muscle through Fetal Day 17 and then abruptly drop off during the last few days of development in utero. This is the same period during which we observed pronounced changes in expression of muscle differentiation genes. Obviously, a comparison of the time course of mRNA levels for different genes cannot by itself establish causal relationships. The present report on mRNA concentrations is consistent with studies on proteins expressed in dysgenic muscle. For instance, Oppenheim et al. (1986) noted that creatine kinase is present at similar low levels in control and dysgenic muscle until Fetal Day 15, and then remains at low levels in the mutant while rising substantially in normal muscle. Recently, Knudson et al. (1988) have shown that the major dihydropyridine binding subunit of the dihydropyridine receptor is missing from muscle membranes of newborn dysgenic mice. On the other hand, several other important constituents of triad and SR membranes, including the Ca2+ + Mgz+ ATPase, are present, but at lower concentrations than in normal muscle. The reductions in concentrations of creatine kinase and Ca2’ + Mg2+ ATPase parallel our results with mRNA levels for these enzymes. Superficially, one might think that the effects of the muscular dysgenesis mutation would be mimicked by experimental paralysis of normal skeletal muscle. However, important differences should be noted between experimental paralysis and the mdg mutation. In the former case, mature myotubes are subjected to tempo-

VOLUME 133. 1989

rary paralysis through the use of blockers of cholinergic or sodium channel activity; most in vivo studies have been carried out with adult muscle following denervation or nerve block. Dysgenic muscle, on the other hand, is completely paralyzed (and also lacks a functional dihydropyridine receptor) throughout its development. A characteristic effect seen in experimental paralysis is the appearance of excess acetylcholine receptors in the sarcolemma (reviewed by Fambrough, 1979). This effect is brought about by a several-fold increase in the mRNA concentration for the receptor subunits (Merlie et al., 1984). Levels of actin mRNA in denervated muscle (Merlie et ah, 1984) and creatine kinase mRNA in paralyzed myotubes (Birnbaum et al., 1980) are essentially unaltered. In contrast, muscular dysgenesis alters the expression of a variety of muscle differentiation genes, including both actin and creatine kinase. Further, the fall in mRNA levels for acetylcholine receptor subunits seen in muscular dysgenesis contrasts with the rise induced by paralysis. Thus, the important question remains whether normal muscle (or myotubes) paralyzed throughout development more closely resembles dysgenie muscle. The present study demonstrates that genes for contractile proteins, for enzymes necessary for muscle function, and for membrane ion channels are all affected in their expression by the loss of a functional dihydropyridine receptor. Thus, the muscular dysgenesis mutation presents a novel system in which to examine the dependence of normal development on E-C coupling and to examine the molecular pathways through which important genes are regulated. This research was supported by grants from the Graduate School and the College of Veterinary Medicine and Biomedical Sciences to N.C. and by grants from MDA and NIH (NS-24444 and R.C.D.A. NS-01190) to K.G.B. REFERENCES AVIV, H., and LEDER, P. (1972). Purification of biologically active globin messenger RNA by chromatography on oligothymidylic acid-cellulose. Proc. Natl. Acad. Sci. USA 69, 1408-1412. BANTLE, J. A., and HAHN, W. E. (1976). Complexity and characterization of polyadenylated RNA in mouse brain. Cell 8,139-150. BEAM, K. G., KNUDSON, C. M., and POWELL, J. A. (1986). A lethal mutation in mice eliminates the slow calcium current in skeletal muscle cells. Nature (London) 320,168-1’70. BETZ, W. J., CALDWELL, J. H., and RIBCHESTER, R. R. (1979). The size of motor units during post-natal development of rat lumbrical muscle. J. Physiol 297, 463-478. BIRNBAUM, M., REIS, M. A., and SHAINBERG, A. (1980). Role of calcium in the regulation of acetylcholine receptor synthesis in cultured muscle cells. Pjugers Arch. 385, 37-43. BOWDEN-ESSIEN, F. (1972). An in vitro study of normal and mutant myogenesis in the mouse. Dev. Biol. 27,351-364. CHAMBERLAIN, J. S., JAYNES, J. B., and HAUSCHKA, S. D. (1985). Regulation of creatine kinase induction in differentiating mouse myoblasts. Mol. Cell. BioL 5,484-492.

CHAUDHARI AND BEAM CHAUDHARI, N., and BEAM, K. G. (1987). Genes aberrantly expressed in muscular dysgenesis in mouse. J. Cell BioL 105,196a. CHAUDHARI, N., and BEAM, K. G. (1988). Muscular dysgenesis in mice is a developmental mutation. J. Cell. Biochem. (Suppl.) 12C, 378. DAVIS, R. L., WEINTRAUB, H., and LASSAR, A. B. (1987). Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 51,987-1000. DUSH, M. K., SIKELA, J. M., KHAN, S. A., TISCHFIELD, J. A., and STAMBROOK, P. J. (1985). Nucleotide sequence and organization of the mouse adenine phosphoribosyltransferase gene: Presence of a coding region common to animal and bacterial phosphoribosyltransferases that has a variable intron/exon arrangement. Proc. Natl. Acad.

Sci. USA 82,2731-2735.

FAMBROUGH, D. (1979). Control of acetylcholine receptors in skeletal muscle. Physio2. Rev. 59,165-227. FOSSET, M., JAIMOVITCH, E., DELPONT, E., and LAZDUNSKI, M. (1983). [3H]Nitrendipine receptors in skeletal muscle. Properties and preferential localization in transverse tubules. J. BioL Chem. 258, 6086-6092.

FOURNEY, R. M., MIYAKO, J., DAY, R. S., and PATERSON, M. C. (1988). Northern blotting: Efficient RNA staining and transfer. Focus 10, 5-7.

GUBLER, U., and HOFFMAN, B. J. (1983). A simple and very efficient method for generating cDNA libraries. Gene 25,263-269. HANAHAN, D., and MESELSON, M. (1980). Plasmid screening at high colony density. Gene 10, 63-67. HELFMAN, D. M., CHELEY, S., KUISMANEN, E., FINN, L. A., and YAMAWAKI-KATAOKA, Y. (1986). Nonmuscle and muscle tropomyosin isoforms are expressed from a single gene by alternative RNA splicing and polyadenylation. Mol. Cell. Biol 6,3582-3595. Hu, M. C., SHARP, S. B., and DAVIDSON, N. (1986). The complete sequence of the mouse skeletal o-a&in gene reveals several conserved and inverted repeat sequences outside of the protein-coding region. MoL Cell. BioL 6, 15-25. ISOM, H., GEORGOFF, I., SALDITT-GEORGIEFF, M. and DARNELL, J. E. (1987). The persistence of liver-specific messenger RNA in cultured hepatocytes: different regulatory events for different genes. J. Cell BioL

105,2877-2885.

Dev. BioL 99,152-165.

KNUDSON, C. M., CHAUDHARI, N., SHARP, A. H., POWELL, J. A., BEAM, K. G., and CAMPBELL, K. P. (1988). Specific absence of the o 1 subunit of the dihydropyridine receptor in mice with muscular dysgenesis. J. Biol. Chem., in press. MACLENNAN, D. H., BRANDL, C. J., KORCZAK, B., and GREEN, N. M. (1985). Amino acid sequence of a Ca*+ + Mga+ ATPase from rabbit muscle sarcoplasmic reticulum, deduced from its complementary DNA sequence. Nature (Lundun) 316,696-700. MEDFORD, R. M., WYDRO, R. M., NGUYEN, H. T., and NADAL-GINARD, B. (1980). Cytoplasmic processing of myosin heavy chain messenger RNA: Evidence provided by using a recombinant DNA plasmid. Natl.

Acad.

Dysgenesis

Mutation

467

PAI, A. C. (1965a). Developmental genetics of a lethal mutation, muscular dysgenesis (mdg) in the mouse. 1. Genetic analysis and gross morphology. Dew. Biol. 11, 82-92. PAI, A. C. (196513). Developmental genetics of a lethal mutation, muscular dysgenesis (mdg) in the mouse. 2. Developmental analysis. Dev. BioL

11,93-109.

PATRICK, J., BOULTER, J., GOLDMAN, D., GARDNER, P., and HEINEMANN, S. (1987). Molecular biology of nicotinic acetylcholine receptors. Ann. N. Y Acad. Sci. 505,194-207. PATRICK, J., HEINEMANN, S. F., LINDSTROM, J., SCHUBERT, D., and STEINRACK, J. H. (1972). Appearance of acetylcholine receptors during differentiation of a myogenic cell line. Proc. Natl. Acad. Sci. USA 69,2762-2766.

PINCON-RAYMOND, M., MURAWSKY, M., MEGE, R.-M., and RIEGER, R. (1987). Abnormal enwrapment of intramuscular axons by distal Schwann cells with defective basal lamina in the muscular dysgenic mouse embryo. Dev. BioL 124,259-268. PINCON-RAYMOND, M., RIEGER, G., FOSSET, M., and LAZDUNSKI, M. (1985). Abnormal transverse tubule system and abnormal amount of receptors for Ca2+ channel inhibitors of the dihydropyridine family in skeletal muscle from mice with embryonic muscular dysgenesis. Dev. Biol. 112, 458-466. PLATZER, A. C. (1978). The ultrastructure of normal myogenesis in the limb of the mouse. Anat. Rec. 190,639-658. POWELL, J. A., and FAMBROUGH, D. M. (1973). Electrical properties of normal and dysgenic mouse skeletal muscle in culture. J. Cell. Physiol. 82, 21-38. RIEGER, F., BOURNAUD, R., SHIMAHARA, T., GARCIA, L., PINCON-RAYMOND, M., ROMEY, G., and LAZDUNSKI, M. (1987). Restoration of dysgenic muscle contraction and calcium channel function by coculture with normal spinal cord neurons. Nature (London) 330, 563-566.

RIEGER, F., CROSS, D., PETERSON, A., PINCON-RAYMOND, M., and TRETJAKOFF, I. (1984). Disease expression in +/+ - mdg/mdg mouse chimeras: Evidence for an extramuscular component in the pathogenesis of both dysgenic abnormal diaphragm innervation and skeletal muscle 16s acetylcholinesterase deficiency. Dev. Biol. 106, 296-306.

KLAUS, M. M., SCORDILIS, S. P., RAPALUS, J. M., BRIGGS, R. T., and POWELL, J. A. (1983). Evidence for dysfunction in the regulation of cytosolic Ca2+ in excitation-contraction uncoupled dysgenir muscle.

Proc.

Muscular

Sci. USA 77,5749-5753.

MERLIE, J. P., ISENBERG, K. E., RUSSELL, S. D., and SANES, J. R. (1984). Denervation supersensitivity in skeletal muscle: analysis with a cloned cDNA probe. J. Cell Biol. 99, 332-335. MISHINA, M., TAKAI, T., IMOTO, K., NODA, M., TAKAHASHI, T., and NUMA, S., METHFESSEL, C., and SAKMANN, B. (1986). Molecular distinction between fetal and adult forms of muscle acetylcholine receptor. Nature (London) 321,406-411. OPPENHEIM, R. W., HOUENOU, L., PINCON-RAYMOND, M., POWELL, J. A., RIEGER, F., and STANDISH, L. J. (1986). Dew. Biol. 114,426-436. ORDAHL, C. P., and CAPLAN, A. I. (1978). High diversity in the polyadenylated RNA populations of embryonic myoblasts. J. BioL Chem. 253, ‘7683-7691.

RIEGER, F., and PINCON-RAYMOND, M. (1981). Muscle and nerve in muscular dysgenesis in the mouse at birth: Sprouting and multiple innervation. Dev. BioL 87,85-101. RIEGER, F., POWELL, J. A., and PINCON-RAYMOND, M. (1984). Extensive nerve overgrowth and paucity of the tailed asymmetric form (16s) of acetylcholinesterase in the developing skeletal neuromuscular system of the dysgenic (mdg/mdg) mouse. Dev. BioL 101, 181-191. RIOS, E., and BRUM, G. (1987). Involvement of dihydropyridine receptors in excitation-contraction coupling in skeletal muscle. Nature (London)

325,717-720.

RUGH, R. (1968). “The Mouse. Its Reproduction and Development.” Burgess, Minneapolis. TANABE, T., TAKESHIMA, H., MIKAMI, S., FLOCKERZI, V., TAKAHASHI, H., KANGAWA, K., KOJIMA, M., MATSUO, H., HIROSE, T., and NUMA, S. (1987). Primary structure of the receptor for calcium channel blockers from skeletal muscle. Nature (London) 328, 313-318. TANABE, T., BEAM, K. G., POWELL, J. A., and NUMA, S. (1988). Restoration of excitation-contraction coupling and slow calcium current in dysgenic muscle by dihydropyridine receptor cDNA. Nature (London) 336,134-139.

ULLRICII, A., SHINE, J., CHIRGWIN, J., PICTET, R., TISCHER, E., RUTTER, W. J., and GOODMAN, H. M. (1977). Rat insulin genes: Construction of plasmids containing the coding sequences. Science 196, 1313-1316. WIECZOREK, D. F. (1984). Early effects in vitro of the muscular dysgenesis mutation on nervous tissue in the mouse. Muscle Nerve 7, 170-193.