Neuromuscular Development Following Tetrodotoxin-Induced Inactivity in Mouse Embryos Lucien J. Houenou,*t Martine Pinqon-Raymond, Luis Garcia, A. John Harris,' and Franqois Rieger Unite de Biologie et Pathologie Neuromusculaires, INSERM U.153, Paris, France; and 'Department of Physiology, University of Otago, Dunedin, New Zealand

SUMMARY Developmental aspects of the neuromuscular system in mouse embryos chronically paralyzed in utera with tetrodotoxin (TI'X) between embryonic days 14 and 18 were studied using biochemical and histological methods. T h e number of lumbar spinal motoneurons (MNs) was higher in inactive embryos than in controls suggesting a decreased motoneuron cell death. In association with the increase in MN number, choline acetyltransferase activity was significantly increased in both spinal cord and peripheral synaptic sites. Paralyzed muscles exhibited a decreased number of mature myofibers and the nuclei were centrally located. Creatine kinase activity was greatly decreased and total acetylcholine receptor and receptor cluster numbers per myofiber were significantly increased in paralyzed muscles. A similar pattern of changes occurs in the neuromuscu-

INTRODUCTION

The extent to which neuromuscular activity regulates spinal motoneuron (MN) and skeletal muscle differentiation and nerve-musclc trophic interactions during embryonic development is not well understood. In many vertebrate species, the normal development of spinal motoneuron connections with skeletal muscles is accompanied by the death of thousands of MNs. Periods of naturally Received July 18> 1990: accepted July 23, 1990 Journal of Neurobiology, Vol. 2 I , No. 8, pp. 1249- 126 I (1990) C 1990 John Wiley & Sons, Inc. ccc 0022-3034/90/0801249- 13$04.00 * To whom correspondence should be addressed. t Present address: Department of Neurobiology and Anatomy, Wake Forest University. Bowman Gray School of Medicine, Winston-Salem, NC 27103, U S A .

lar system of the mutant mouse muscular dysgenesis (mdg). However, in contrast to the mdg mutant, tetrodo-

toxin-treated muscles were similar to controls in their innervation pattern, in the ultrastructural aspects of the excitation-contraction coupling system (i.e., dyads and triads) and in the extent of dihydropyridine binding. Thus, neuromuscular inactivity is not sufficient to impair the pattern of muscle innervation or the appearance of either the triadic junctions or dihydropyridine receptors. These results indicate that alterations of dihydropyridine binding sites and triads in muscular dysgenesis cannot be accounted for by inactivity hut rather must reflect a more primary defect involving the structural gene(s) regulating the development of one or more aspects of muscle differentiation.

occurring MN cell death have been described i n thc frog (Prestige, 1967; Lamb. 1981), chick (Hamburger, 1975; Chu-Wang and Oppenheim, 197th; Oppenheim and Majors-Willard, 1978), rat (Harris and McCaig, 1984; Oppenheim, 1986), mouse (Lance-Jones, 1982; Oppenheim et al., 1986), and human (Forger and Breedlove, 1987) embryos. It is known that chronic paralysis of chick (Pittman and Oppenheim, 1978; 1979; Laing and Prestige, 1978; Oppcnheim and ChuWang, 1983;Oppenheim, 1984) or rat (Harris and McCaig, 1984) embryos with neuromuscular blocking agents results in a substantial reduction of the naturally occurring MN cell death as well as the alteration of other aspects of neuromuscular development. Another model for studying the developmental role of neuromuscular activity is the mouse mu1249

1250

Hoiienou t% ul.

tant, muscular dysgenesis, which is a lethal autosoma1 recessive mutation. '4ffected embryos (rndg/mdg)are characterized by a lack of maturation of the skeletal musculature and a total absence of spontaneous or induced muscle contractions during embryonic and fetal development (Glueckson-Waelsch, 1963; Pai, 1965a,b: Banker, 1977; Rieger and Pincon-Raymond, 1981). Dysgenic embryos exhibit several modifications of neuromuscular development including, reduced M N death, increased choline acetyltransferase (ChAT) activity in both spinal cord (Oppenheim et al., 1986) and skeletal muscle (Rieger and PinConRaymond, 198l), an increase of total acctylcholine receptors (AChR) and AChR clusters per myofiber (Powell, Rieger, Blondet, Dreyfus, and PinGon-Raymond, 1984), decreased creatine kinase (CK) activity, and a delay in muscle differentiation (Rieger and Pincon-Raymond. 1981; Oppenheim et al., 1986). The innervation density of dysgenic muscle is higher and the morphological maturation of neuromuscular junctions less advanced than in control embryos (Pincon-Raymond and Rieger, 1982). Dysgenic emhryos also display a drastic deficit in voltage-sensitive (slow) Ca2+ channels (Beam. Knudson, an d Powell. 1986; Romey, Rieger, Renaud. PinCon-Raymond, and Lazdunski, 1986), dihydropyridine receptors (DHP-R), and a lack of triadic junction organization (Pinqon-Raymond, Rieger, Fossct, and Lazdunski, 1985). It has been suggested that the abnormal neuromuscular development in rndg is due to the mutation of the Ca2+channel protein (Beam et al., 1986). Recently, Tanabe, Beam, Powell, and Numa (1988) have shown that contractile activity and slow Ca2+current can be restored by microinjections of DHP-R complementary DNA into dysgenic muscle cells. Their conclusion that the mutation mdg alters the structural gene for the muscle DHP-R is at odds with a previous report (Rieger et al., 1987a) which claimed that dysgenic muscle contraction and voltage-dependent Ca2+ channel function can be restored by coculturing mdg muscle cells with normal spinal cord cells. Rieger et al. ( 1987a) suggested that normal spinal cord neurons synthesize and secrete a factor essential for muscle differentiation and that this factor is not produced by mdg/mdg spinal neurons. In order to determine whether the cellular and molecular modifications of the neuromuscular system described in mdg/mdg could be accounted for by inactivity alonc, we undertook studies on spinal MN and skeletal muscle devclopment in normal mouse embryos chronically paralyzed in

zifero with tetrodotoxin (TTX) during early development of the neuromuscular system (i.e., embryonic day (E) 14 to E18). This period also corresponds to that of naturally occumng MN death in mouse embryos (Lance-Jones, 1982; Oppenheim et al., 1986). TTX causes a total block of nerve and muscle electrical activity by inhibiting the voltagedependent sodium channels (Nardhashi, Deguchi, Urakawa, a n d Ohkubo, 1960). In the present study, we have focused on the following aspects of neuromuscular development: MN survival and cholinergic (ChAT) differentiation of MNs, and biochemical and ultrastructural aspects of' muscle differentiation, with a special emphasis on DHP-R and the formation of triads (i.e., the apposition of transverse tubule and sarcoplasmic reticulum membranes). The pattern of muscle innervation in TTX-treated embryos was examined using double-staining of whole-mount diaphragm muscle by acetylcholinesterase (AChE) histochcmistry and silver-staining of nerve fibers. We also examined the level and distribution of AChR on myofibers from both TTX-treated and control embryos. Tetrodotoxin suppressed both muscle excitation and contraction. The resulting inactivity altered MN cell death and muscle development but did not impair the differentiation of either triadic junctions or DHP-R binding. The significance of these results is discussed in relation to the effects of inactivity in the mouse mutant niuscular dwgenesis.

MATERIALS AND METHODS

Paralysis of Embryos Pregnant mice (129/ReJ strain) were anesthetized with ether, and after exposure of onc uterus through an incision in the abdomen, embryos were paralyzed by inserting 5-mm-long glass capillaries filled with 6 m M TTX (Sigma, St. Louis. MO) inside the amniotic fluid as described previously by Harris and McCaig (1984). One end of each capillary had a pore of 12 pm diameter to allow a slow release of the drug by diffusion. Embryos wcre treated on E l 4 (the day a copulatory plug was present is designated EO) and remained paralyzed until they were collected on E 18. In another series of expcriments, embryonic MNs and nerve terminals were destroyed with a single injection of 0.2 pg of beta-Bungarotoxin (8-BTX) (Boehringer, France) into the amniotic fluid on E l 4 (Harris, 1981; McCaig, Ross, and Hanis, 1987). These experiments with [j-BTX were designed to determine the level of non-MN-related ChAT activity in spinal cord and muscles. Pregnant mice do not tolerate more than two treatments with TTX capillaries or /3-BTX injections. With one exception. all experiments

Telrodoloxin and

presented here were done with embryos that were treated on E I4 and collected on E 18; assays for creatine kinase activity measurement involved embryos paralyzed from E l 3 to E17. In all cases, treated embryos remained completely paralyzed for at least 4 days. Pregnant females were sacrificed by cervical dislocation and fetuses were collected by cesarean section. The criteria for paralysis were complete flaccidity and absence of reflex body flexions in response to strong electrical slimulation (50 V X 0.2 ms at 10 Hz) ofthe snout at the time of Sacrificc.

Biochemical Procedures Tissue Sample Preparation. Control and treated embryos were chilled on ice, and spinal cords and muscles were dissected. Muscle samples included the diaphragm, intrinsic muscles of the foot. and pooled limb muscles from the calf and thigh. Muscles were homogenized in a solution containing 1 M NaCI, 0.001 M EGTA, 1% Triton XIOO, and 0.01 MTris-HCI, pH 7.2. We used a weight/volume (mg/pl) ratio of 1:10. Homogenates were centrifuged at 20,000 X g for 15 min (Sorvall RC2), and thc supernatant was taken for assays. Spinal cords were transected between the first and second cervical vertebrae to obtain the entire cords (i.e., cervical to sacral), homogenized in 0.3 mL of solution, and processed in the same manner as the muscle samples. Choline Acetyltransferase (Acetyl-CoA: Choline 0-Acetyltransferase, ChAT; EC: 2.3.1.6) Activity. This was determined for intrinsic muscles of the foot, the diaphragm muscle, and for spinal cord homogenates following the method described by Fonnum (1975) using [14C]acetylcoenzymeA (coenzyme A, 50 mCi/mmol; Amersham, France) as a cofactor. The results were expressed as counts per minute (cpm) of synthetized [14C]-acetylcholine. Because 4 4 I naphthylvinyllpyridine (NVP) is a specific inhibitor of ChAT (White and Cavallito, 1970), the specificity of the enzyme activity was tested by running some samples in the presence of 4 m M NVP.

Acetylcholine Receptor Quantification. This was determined by incubating limb muscles for I h in 0.2 pg/ml [‘251]alpha-bungarotoxin(a-BTX) (New England Nuclear, Boston, MA), as described by Powell et al. ( I9 84), after which the tissue samples were prepared as described above. The homogenatcs were then loaded onto continuous sucrose gradients (5-20% in the same solution as the samples) and centrifuged for 15 h a1 246,310 X g and 4°C (Beckman ultracentrifuge rotor SW 41). The toxin-receptor complex sediments as a 9s peak (Merlie, Sobel, Changeux, and Gros, 1975) and the total binding sites were determined by evaluating the area of the radioactive peak after measuring the radioactivity in a gamma-counter (Kontron). Results were ex-

NeuromusczilarDevelopment

1251

pressed as cpm of bound toxin per embryo (four limb muscles).

Creatine Kinase (ATP Creatine Phosphotransferuse, CK; EC: 2.7.3.2) Activity. This was measured in El7 limb muscle homogenates by a modification of the method of Rosalki (1 967) using “CPK single vial N o 45- 1” from Sigma. For this assay, 1 mL of reconstituted test reagent and 33 p1 of samples were mixed and incubated at 25°C. After a preincubation of 4 min to allow reaction kinetics to become linear (zero order), the absorbance of the mixture was read at 340 nm at 5-min intervals using water as a reference. The rate of change of absorbance is related to enzyme activity which was expressed as international units (IU) per embryo (four limb muscles).

Quantification of Dihydropyridine Binding Sites. Limb muscles were dissected from El8 control and E 14- 18 TTX-treated embryos, and homogenized in 10-20 vol. of ice-cold 20 m M Tris-HC1 buffer. pH 7.4, containing 0.3 M sucrose, using a Polytron apparatus at setting 4, for 1 min. Homogenates were then filtered through four layers of moist cheesecloth before assays were run. The binding experiments were done as described by Borsotto, Barhanin, Noman, and Lazdunski (1984) and Pinqon-Raymond et al. (1 985). Briefly, aliquots of homogenates were incubated at 25°C in 1 mL of a solution containing 50 mM Tris-HC1 buffer, pH 7.5, and the appropriate concentration of the dihydropyridine calcium-channel inhibitor (+)[methyl-’HI PN 200- 1 10 (82 Ci/mmol; Amersham, France). The final protein conccntration in the samples was 0.2-0.9 mg/mL. In all experiments, (+)[3H] PN 200-1 10 was added last and the mixtures were incubated for 40-45 min. The reaction was stopped by rapid filtration of 400 p1 of the mixtures through Whatman GF/B glass-fiber filters under reduced pressure. The filters were immediately washed and processed for radioactivity counts. To measure non-specific binding, 1 pA4 of unlabelled nitrendipine (Bayer Pharma, France) was added to the reaction mixture. Specific binding was calculated as the difference between ( o l d and non-spec[fificbinding. The maximum binding capacities (Bma) were expressed as femtomoles of radioactive ligand bound per mg of protein.

Protein Concentration. This was determined by the method of Lowry, Rosebrough, Fur, and Randall (1951).

Histology and Histochemistry The lumbar spinal cord was dissected out from control and TTX-treated embryos and fixed overnight in Carnoy’s solution. Tissues were dehydrated in graded EtOH, embedded in paraffin, and serially sectioned at 12 pm. and then processcd for thionin staining. Motor neurons, identified on the basis of their location, size,

Houenou ef al

I252 A 60

B

_I

pmw1

h B

I

I

I I

Figure 1 Motoneuron number (mean t- S.E.M.) in the lumbar spinal cord (A) and total ChAT activity in whole spinal cord (B) from control (CON3 ) and TPX-treated (TTX) embryos on E18. The numbers in the graph are sample sizes. p. Student's t test. (see graph for p value).

and staining characteristics (Lance-Jones, 1982; Oppenheim et al.. 1986) were counted in every tenth section throughout the lumbar spinal cord. Results wcrc cxpressed as the mean MN numbers (*SEMI per section. To assess the gross development of muscles, hindlimbs from control and paralyzed embryos wcrc scctioned transversally on a cryostat at 16 pm and stained with hematoxylin-eosin. To study the distribution ofAChR clusters on muscle fibers, whole diaphragms with attached adjacent ribs were removed from embryos, carefully dissected, and fixcd for 1 h in 2% formaldehyde. Single fibers were teased from the preparation, stained with rhodamineconjugated alpha-Rungarotoxin ((u-BTX) for 2 h, washed. and mounted in Moviol for light microscopic observation. Hemi-diaphragms from normal and paralyzed embryos were stained for both focal accumulations of AChE and for nerve fibers using the method described by Hopkins, Brown, and Keynes (1985). Briefly, tissues were fixcd overnight in 0.25 M paraformaldehyde in 0. I M phosphate buffer pH 7.25. washed several times, and incubated 30 min in icc-cold cholinesterase solution (30 mA4 maleate buffer, 10 mMtri-sodium citrate, 0.5 mM potassium femcyanide. 4.7 m M copper sulfate). After 3 washes of 5 min each, tissues were incubated for 10 min in 7.5 mhf potassium ferricyanide, washed, dehydrated in graded alcohols, and rehydrated. Samples were then stained in silvcr solution ( I W Ag in 20 mM boric acid and 37.7 m M sodium borate) for 20-35 min. Tissues were finally washed and incubated in ice-cold dcvcloper (0.39 M sodium sulfite, 90 m M hydroquinone, 20 m M sodium borate) for less than 2 min before immersion in 100%ethanol and in xylene. and mounted in Permount.

intercostal grid, and placed in the same fixative for an additional 2 h at 4°C. After a postfixation in 2% osmium for I h, samples wcrc embedded in Epon resin. At this time. the diaphragm was removed from the intercostal grid and laid flat on a glass slide for the final embedding and sectioning.

RESULTS

The TTX-treated embryos exhibited several abnormal morphological features: exaggerated body curvature, fixed o r immobile joints, cleft palate, underdevelopment of the inferior jaw. and the absence of spontaneous or evoked contractile activity. By contrast. cardiovascular activity appeared normal. A similar array of abnormalities have been previously described in the paralyzed chick embryo (Piltman and Oppenheim, 1979) and in the mouse mutant, muscular dysgenesi:, (Pai, 1965a; Rieger and Pincon-Raymond, 1981), and thus probably reflect common features resulting from chronic embryonic inactivity. Effects of Inactivity on MN Survival and Cholinergic Differentiation Motoneuron development and survival were examined in the lumbar segments of the spinal cord. Because of the small number of embryos used here (i.e.. two animals in each group), the results are expressed as MN numbers per section. Lumbar spinal cord sections from TTX-treated embryos have about 60c%more MNs than those from control embryos [Fig. l(A)]. The morphology of MNs in paralyzed embryos appears comparable to controls (not shown). In order to evaluate the cholin-

A

B

L

5

ON'

Electron Microscopy. Diaphragm muscles from control and TTX-treated embryos were fixed in situ for 30 min with a solution containing 2.5% glutaraldehyde, 0.5% tannic acid, and 0.1 A4 phosphate buffer, pH 7.4. The diaphragm was removed while still inserted in the

Figure 2 Motoneuron number (mean f S.E.M.) in the lumbar spinal cord (A) and total ChAT activity in whole spinal cord (B) from control (CONT) and P-BTXtreated ((3-BTX) embryos on E18. Sample sizcs are in the graph. p. Student's t test. p values are in the graph.

Tetrodotouin and Nmronzusciilar Llevrloprnent

ergic differentiation of the surviving MNs, ChAT activity was measured in spinal cords from TTXtreated versus control embryos. Paralyzed embryos exhibit an increase of 40-50% in whole spinal-cord ChAT activity compared to control embryos [Fig. l(B)]. Paralyzed spinal cords also exhibit an increase in total protein content (TTX: 985.2 k 42.4 pg, n = 5 ; Control: 783.0 i 47.5 pg, n = 5, p = 0.01). Taken together, these data suggest that the increased ChAT activity cannot be accounted for by an increased enzyme activity in individual MNs, but rather very likely reflects an

increased number of cells. Because the spinal cord contains other cholinergic neurons besides MNs, ChAT activity from the whole spinal cord does not entirely reflect en7ymatic activity specific to MNs. To estimate the level of non-motoneuronal ChAT, embryos were treated at E l 4 with P-BTX to destroy selectively the MNS (see Methods). After 4 days of cxposurc to @-BTX,embryos did not appear grossly different from T I X-treated embryos (see above), although some were edematous. In these aneural embryos, virtually all (94%) of the MNs in the lumbar spinal cord were destroyed

Figure 3 Photomicrographs of cryostat sections through limb muscles from (A) control and (B) TTX-treated embryos on E 18. Note general loss of tissue in toxin-trcated muscle and increased number of sinall (mononucleated) cells (stars). Arrows indicate centrally located

nuclei in immature myotubes.

1253

1254

Houenoii et a1

0

CONT m< Figure 4 Creatine kinase activity (mean k S.E.M.) in pooled limb muscles from control and TTX-treated embryos on E 17 (TTX treatment was initiated on E 13). The numbers are sample sizes. p , Studcnt's f test. p value is in the graph.

[Fig. 2(A)]. These results are similar to those reported previously for rat embryos treated with @-BTX(Harris, 1981; McCaig et al., 1987). ChAT activity in the spinal cords from aneural embryos was 55% of control values [Fig. 2(B)], indicating that almost half of ChAT activity in the spinal cord is derived from cells other than MNs.

Effects of Paralysis on Muscle Development and Innervation Hematoxylin-eosin staining of sections from hindlimbs shows that El8 control muscles exhibit all the characteristics of well-differentiated normal mouse muscles (Wirsen and Larsson, 1964; Rieger and Pincon-Raymond. 198 I), including little intercellular space between the myofibers and myofibers with peripherally placed nuclei [Fig. 3(A)]. By contrast, E 14- 18 TTX-treated muscles lack cohesion, exhibit extracellular spaces, and contain many mononucleated (small) cells and immature myotubes with centrally located nuclei [Fig. 3(B)]. Creatine kinase activity, a quantitative measure of muscle differentiation (Goedde, Christ, Benkmann, Beckmann, and Lang, 1978; Oppenheim et al., 1986), is considerably lower (90%) in TTXtreated limb muscles compared to controls (Fig. 4). This biochemical index of maturation is in accord with the immature histological characteristics of TTX-treated muscle described above. At the ultrastructural level, myofibers from TTX-treated diaphragm exhibit a well-formed

basal laminae. However, the myofilament network was disorganized [Fig. 5(C)]. The triadic junctions, which are involved in excitation-contraction (E-C) coupling, appeared normal in TTX-treated myofibers (Fig. 5). The tritiated dihydropyridine (DHP) derivative (+)[3H] P N 200-1 10 was used to quantitate the putative DHP-sensitive calcium channels in crude muscle extracts from control and TTX-treated muscle tissue. The amounts of DHP binding were not significantly modified after activity blockade with TTX (B,,, = 164 fmol/mg protein as compared to 120 & 10 fmol/mg for controls).* Similarly, when normal muscle cells are cultured in v i m and treated with TTX added to the culture medium, the levels of DHP-R are also unaltered (Rieger et al., 1987b). The levels of AChR in normal active and paralyzed muscles was studied using ['251]a-BTXas a ligand. Figure 6 shows a typical determination of the total ['251]~-BTX binding in limb muscles homogenates. Extracts from TTX-treated muscles generally give a significantly larger peak (ca. 24270, n = 4, p < 0.02) than do age-matched control muscles. The distribution of AChR on diaphragm myofibers was examined using rhodamine-conjugated-a-BTX. The results show that control fibers have only one AChR cluster with high receptor density as described previously by Powell et al. (1984) [Fig. 7(A)]. By contrast, the majority of TTX-treated myofibers exhibited more than one receptor cluster [Figs. 7(B)]. No attempt was made to distinguish between junctional and extrajunctional AChR clusters. The qtaining of nerve fibers by silver nitrate impregnation combined with focal accumulation of AChE staining indicates that the pattern of muscle innervation (diaphragm) is not altered after TTX treatment (Fig. 8). In both the control and TTXtreated diaphragm, neuromuscular junctions are localized in the middle of the muscle. However, the extent of staining appeared to be greater in TTX-treated muscles suggesting that paralysis may induce an increase in axon and nerve terminal numbers (Pincon-Raymond and Rieger, 1982; Dahm and Landmesser, 1988). To estimate the innervation density in paralyzed versus normal active muscles, we measured peripheral ChAT ac* For B,,,, determinatlon, limb muscles from 7 to 10 embryos of each group were pooled, homogenized, and processed as descnbed under Matenals and Methods. Due to the small size of TTX-treated tissues, only one determinatlon has been made. For control. B,, represents the mean (+-SEM)of three determinations

Tetrodotoxin and Neurornwcular Developmcnl

tivity. The results show that there is 42% and 85% increase of enzymatic activity in TTX-treated diaphragm and intrinsic foot muscles, respectively [Fig. 9(A)]. To ascertain whether all the ChAT activity measured reflects the intramuscular motor nerves, we estimated the enzymatic activity present after treatment of embryos with P-BTX. Figurc 9(B) shows that ChAT activity in aneural muscles is very low (ca. 10%) as compared to controls.

1255

Thus, virtually all of the peripheral ChAT activity measured in muscles is most likely derived from motor neuron axons and terminals.

DISCUSSION One of the major goals of this study was to determine whether altering the normal physiological

Figure 5 Elcctron micrographs showing membrane couplings (triads) in (,4) control and (B) TTX-treated myofibers on E 18. Junctional feet (arrows) bridge the T-tubule (opened circles) and sarcoplasmic membranes. General ultrastructure of a toxin-treated myofiber is shown in (C). Arrowheads indicate the basal lamina; stars show disorganked sarcomers.

1256

0

Hournoii et al.

5

10

15 20 25 FRACTION NUMBER

30

35

Figure 6 Sedimentation profiles of [1’25]~r-BTXhinding from control and TTX-treated limb muscles on E l 8. Sedimentation was carried out with I00 pl of homogenates loaded onto 5-2076 continuous sucrose gradients and centrifuged as described under Materials and Methods. Only the 9s peak represents the specific binding to AChRs. Ordinates are proportionals and the break in the ordinate represents the baseline of the upper

graph. function of nerve-muscle interactions in mouse embryos would have any effects on motoncuron and muscle development and on the pattern of myofiber innervation. TTX-paralyzed embryos exhibit gross morphological abnormalities and alterations of some aspects of neuromuscular development similar to those seen in dysgenic mutant embryos. The number of (lumbar) MNs is significantly increased as a consequence of reduced naturally occurring cell dcath. The surviving MNs appear to differentiate normally and make contacts with target muscles as indicated by the increased activity of ChAT in both spinal cord [Fig. l(B)] and target muscles [Fig. 9(A)]. As noted above (Results), these increases in C U T appear to reflect both increased MN numbers and peripheral branches although we cannot entirely exclude the possibility that increased levels of ChAT per MN also contribute to these results. Similarly, in the dysgenic mutant there is also a substantial increase in MN number and in both spinal cord ChAT (Oppenheim et al., 1986) and muscle ChAT (Rieger and Pincon-Raymond. 198 1 ) activity. Reductions in naturally occurring MN

death following neuromuscular activity blockade have also been reported in chick (Pittman and Oppenheim, 1978; Pittman and Oppenheirn, 1979; Laing and Prestige, 1978; Oppenheim, 1984), frog (Oleck, 1980). and rat (Hams and McCaig, 1984) cmbryos. However, the specific mechanisms by which this effect is accomplished are not entirely understood. In chronically paralyzed embryos, the size and differentiation of target muscle is reduced or retarded (Figs. 3 and 4; Harris. 1981; Oppenheim et al., 1986), yet M N survival is increased. It thus appears unlikely that MN survival is regulated solely by the size of the target muscles. It has also been suggested that neuromuscular activity and/or muscle contraction regulates the production or availability of a putative neurotrophic factor produced by the target (Hamburger. 1975: Oppenheim el al., 1986; Oppenheim, 1989; Hamburger and Oppenheim, 1982). Accordingly, muscle inactivity would result in increased production or availability of this factor leading to enhanced MN survival. Therc is a considerable amount of evidence supporting the idea that muscle-derived trophic agents promote motor neuron survival (Bennett. Lai, and Nurcombe, 1980; Calof and Reichardt, 1984; O’Brieii and Fischbach, 1986; Dohrmann, Edgar, Sendtner, and Thoenen. 1986; Oppenheim, Haverkamp, Prevette, McManaman, and Appel, 1988). However. it is not clear whether activity regulates the production or synthesis of these trophic agents by muscle cells. Although there is some zn utro evidence suggesting that inactivity up-regulates the production of neuron survival- and neurite-promoting agents (Hsu, Natyzak, and Trupin, 1982: Henderson, Eluchet, and Changeux, 1983; Nurcombe, Hill, Eagleson, and Bennett. 1984), other zn vitro (Tanaka, 1987) and in vivo (Houenou, Prevette, and Oppenheim, 1989) studies have failed to confirm these reports. Therefore, although activity appears to somehow regulate the awilaDiZitj1 of a trophic agent, it may do so by some means other than increased sjwzlhesis or production (Dahm and Landmesser, 1988; Oppenheim, 1989). In contrast to mdg/mdg in which the diaphragm muscle is characterized by profuse innervation with abnormally increased axonal sprouting and multiple nerve-muscle contacts (Rieger and PinCon-Raymond, 198 l), the innervation pattern in TTX-treated muscle appears “normal.” However, each neuromuscular junction in toxlntreated embryos may be hyperinnervated. Similarly, it has been reported previously that the diaphragm muscle from rat embryos chronically treated with TTX, possesses more axons and nerve

Tetrodotoxin and Neuromzrsculur Development

terminals than do controls, and yet MNs establish the same number of muscle contacts (motor unit size) as in normal active muscle leading to a hyperinnervation of the neuromuscular junctions (Harris and McCaig, 1984). It is possible that the time-course of TTX treatment used in our studies (ie., beginning treatment on E14) is not optimal for altering the innervation pattern of the diaphragm. Regrettably, attempts to paralyze embryos beginning on El 1-12 were not successful because of the fragility of the embryos and the amnion. Although further studies are needed, our present results suggest that inacti\-rty alone may not be sufficicnt to induce a multijunctional innervation pattern in mouse embryos. If correct, then the alteration of muscle innervation in dysgenic mice may be somehow related to actions of the mutation other than those influencing inactivity (see below). Alteration of AChRs and AChR clusters in inactive skeletal muscle has been well documented (e.g., Burden, 1977; Harris, 198 1; Ziskind-Conhaim and Bennett, 1982; Oppenheim, Bursztajn, and Prevette, 1989). The present study shows that TTX-induced inactivity in mouse results in the appearance of multiple receptor clusters along the myofibers. A similar result was reported previously in rat embryos chronically treated with TTX (Harris, 198 1; Ziskind-Conhaim and Bennett, 1982). Thus, it appears that toxin-induced increases in AChR clusters is not accompanied by multiple nerve-muscle contacts in mouse or rat.

1257

The mechanisms regulating muscle innervation in these species may be different from those in avian muscle in which chronic activity blockade during normal cell death period results in the apparent innervation of virtually all AChR clusters (Pittmann and Oppenheim, 1979; Oppenheim, 1984; Ding, Jansen, Laing, and Tonnesen, 1983; Oppenheim et al., 1989). Recent findings suggest that the triadic junction plays an important role in E-C coupling in the neuromuscular system (Franzini-Armstrong and Nunzi, 1983). Muscle contraction is triggered by thc release of Ca2+ from the sarcoplasmic reticulum, and thc appearance of the voltage-sensitive “L type” (long-lasting) Ca2+ current can be blocked by dihydropyridine derivatives (Potreau a n d Raymond, 1980; Almers a n d McClesky, 1984). Dysgenic muscle apparently lacks “L type” Ca2+ currents (Beam et al., 1986: Rieger et al., 1987a) and triadic junctions and exhibits significantly less (about 20% of control) DHP binding (PinCon-Raymond et al., 1985). In TTX-treated muscle. both triadic junctions and the amount of DHP binding are comparable to controls. Taken together with thc previous reports on mdg (PinCon-Raymond et al., 1985; Rieger et al., 1987a). the present results suggest that inactzvrty alone does not affect the differentiation of elements (i.e., triads and DHP-R) involved in E-C coupling. Consequently, the dysgenic mouse mutation mdg is not a faithful developmental model of muscle inactivity. It seems likely that the primary defect in

Figure 7 Isolated diaphragm fibers from ( A ) control and (B) TTX-treated embryos stained for AChR with rhodamine-conjugated LY-BTX on E18. Scale bars = 10 bm.

1258

Hownou et a1

Figure 8 Microphotographs of hemi-diaphragm from control and TTX-treated embryos stained for focal AChE and neurofibrillary network on E 18. Notc the normal pattern of focal AChE (neuromuscularjunction) distribution (arrows) and the apparent increased diameter of nerve trunks (arrowheads) in toxin-treated tissue. Identical procedures were used for both samples, which were both stained at the same time using the same solutions.

dysgenic mice may involve genes that regulate one or more components of the E-C coupling system. It has been reported that a primary defect in dys-

A

Figure 9 Total ChAT activity (mean percent i S.E.M.) on E l 8 in (A) Control and TTX-treated diaphragm ( I ) and intrinsic foot muscles (2). (B) Control and D-BTXtreated limb (calf thigh) muscles. The numbers in the graph represent sample sizes. p , Student’s t test. p values are in the graph.

+

genic muscle involves the a1 subunit of the Ca2+ channel which contains the DHP-binding site (Tanabe et al., 1988). Because the receptor for DH P also exists in neuronal cells (Nowycky, Fox, and Tsien. 1985: McCobb, Best, and Beam, 1989), it will be of considerable interest to determine whether the abnormal axonal sprouting and muscle innervation pattern described previously in dysgenic mice (Rieger and PinGon-Raymond, 198 1) are due to the mutation of the Ca’+ channel protein in motor neurons. In conclusion, although chronic blockade of neuromuscular activity in normal mouse embryos does not completely mimic the alterations in neuromuscular development observed in the paralytic dysgenic mouse mutant, this model nonetheless provides a valuable in vwu procedure for examining the role of activity in early development of the mammalian nervous system. We wish to thank Dr. M. Fardeau for his support and Dr. R. W. Oppenheim and Dr. M. Verdikre-Sahuqut for

7 etrodotoxin and Ncuromu vcular Development their interest and helpful comments on the manuscript. We are grateful to A. P. Villageois for technical assistance and D. Styers for help in typing the manuscript. A fellowship from the “Association Francaise contre les Myopathies” was awarded to L.J.H.

REFERENCES and MCCLESKEY,E. W. (1984). Non-selective conductance in calcium channels of frog muscle: calcium sensitivity in a single file pore. J. Physiology 353:585-608. BANKER,B. Q. (1977). Muscular dysgenesis in the mouse (rndglmdg).I. Ultrastructural studies on skeletal and cardiac muscles. J.Neuropathol. Exp. Neurol. 36:100- 127. BEAM,K. G., KNunsoN, C. M.; and POWELL,J. A. (1986). A lethal mutation in mice eliminates the slow calcium current in skeletal muscle cells. Nature 320: 168- I 70. BENNETT, M. R., LAI, K., and NURCOMBE, V. (1980). Identification of embryonic motoneurons in vitro; Their survival is dependent on skeletal muscle. Brain Res. 190537-542. BORSOTTO,M., BARHANIN, J., NORMAN, R. I., and LAZDUNSKI, M. (1 984). Purification of the dihydropyridine receptor of the voltage dependent Gaff channel from skeletal muscle transverse tubules using (+)[3H] P N 200- 1 10. Biochem. Bioyhys. Rex Commun. 122:1357-1366. BURDEN, S. ( 1977). Development of the neuromuscular junction of the chick embryo: Thc number, distribution and stability of acetylcholine receptors. Do.Biol. 57:3 17-329. CALOF,.4.L., and REICHARD’T, L. F. (1984). Motoneurons purified by cell sorting respond to two distinct activities in myotube-conditioned medium. Dev. Biol. 106:194-210. CHU-WANG,I. W., and OPPENHEIM, R. W. (1978a). Cell death ofmotoneurons in the chick embryo spinal cord. I. A light and electron microscopic study of naturally occurring and induced cell loss during development. J. Corny. Neurol. 177:35-58. DAHM, L. M., and LANDMESSEK, L. (1988). The regulation of intramuscular nerve branching during normal development and following activity blockade. Dev. Bid. 130:621-644. DING,R., JANSEN, J. K. S., LAING,N. G.: and TONNESEN, H. (1983). The innervation of skeletal muscles in chickens curarized during early development. J . Neurocvtol. 12:887-9 19. DOHRMANN: U., EDGAR, D., SENDTYER, M., and THOENEN, H. ( 1 986). Muscle-derived factors that support survival and promote fiber outgrowth from embryonic chick spinal motor neurons in culture. Dev. R i d . 118:209-221. FONNUM, F. (1975). A rapid radiochemical method for ALMERS, W.,

1259

the determination of choline acetyltransferase. J . Neurochem. 24:407-409. FORGER,N. G., and BREEDLOVE, S. M. (1987). Motoneuronal death during human fetal development. J. Comp. Neurol. 264:118-122. FRANZINI-ARMSTRONG, C., and NUNZI,G. ( 1 983). Junctional feet and particles in the triads of a fast twitch muscle fibre. J. Musc. Res. Cell Motil. 4:233252. GLUECKSOHN-WAELSCH, S. (1963). Lethal genes and analysis of differentiation. Science (Washington, DC) 142: 1269-1 276. GOEDDE.H. W.. CHRIST,I., BENKMANN, H. G., BECKMA“, R., and LANG,H. (1978). Creatine kinase isoenzyme patterns in Duchenne muscular dystrophy. Clin. Genet. 14:257-260. HAMBURGER, V. (1975). Cell death in the development of lateral motor column of the chick embryo. J. Comp. Neurai. 160:5 35-546. HAMBURGER, V., and OPPENHEIM, R. W. (1982). Naturally occurring neuronal death in vertebrates. h’eurusci. Comm. 1:39-55. HARRIS,A. J. (198 1). Embryonic growth and innervation of rat skeletal muscles. Philos. Trans. R. Sue. London. Scr. B 293:251-3 14. HARRIS. A. J., and MCCAIG,C. D. (1984). Motoneuron death and motor unit size during embryonic development of the rat. .J. Neurosci. 4: 13-24. HENDERSON, C. E., HIJCHET, M., and CHANGEUX, J. P. ( 1983). Denervation increases a neurite-promoting activity in extracts of skeletal muscle. Yature 302~609-61 1. HOPKINS,W. G., BROWN,M. C., and KEYNES,R. J. (1985). Post-natal growth of motor nerve terminals in muscles of the mouse. J. Neurocytol. 14525-540. HOUENOU, L.: PREVETTE, D., and OPPENHEIM, R. W. (1989). Motoneuron survival in vivo following treatment with extracts from active and inactive muscle. Soc. Neurosci. Abst. 15436. Hsu, L., NATYZAK, E., and TRUPIN,G. L. (1982). Neurotrophic effects of skeletal muscle fractions on spinal cord differentiation. J. Embryvl. Exp. Morphol. 71:83-95. LAING,N. G., and PRESTIGE, M. C. (1978). Prevention of spontaneous motoneuron death in chick embryos. J. Physiol. 282:33-34. LAMB,A. H. (198 1). Target dependency of developing motoneurons in Xenopus laevis. J. Coinp. Neurol. 203: 15 1- 17 I . LANCE-JONES, C. (1982). Motoneuron cell death in the developing lumbar spinal cord of the mouse. Dev. Brain Res. 4:473-479. LOWRY,0. H., ROSEBROUGH, N. J., FARR.A. L., and RANDALL, R. J. (1951). Protein measurement with the Folin-phenol method. J. Bid. Chem. 291:265275. MCCAIG,C. D.. Ross, J. J., and HARRIS, A. J. (1987). Embryonic somatic nerve destruction with beta-bungarotoxin. Cell Tissue Rex 2474 1-50.

1260

IIouenou et ul.

MCCOBB,D. P., BEST.P. M., and BEAM,K. G. (1989). Development alters the expression of calcium currents in chick limb motoneurons. Neuron 2:16331643. MERLIE, J. P., SOBEL,A,, CHANGEUX, J . P., and GROS, F. (1975). Synthesis of acetylcholine receptor during differentiation of cultured embryonic muscle cells. Proc. Nutl. Acad. Sci. USA 72:4028-4032. NARAIIASHI, T., DEGUCHJ, T.. URAKAWA,N., and OHKUBO, Y . (1960). Stabilisation and rectification of muscle fibcr membrane by tetrodotollxin. Am. J . Phyriol. 198:934-938. NOWYCKY, M. C., Fox, A. P., and TSIEN,R. W. (1985). Three types of calcium channel with different agonist sensitivity. Nutiire 316:440-44 3. NURCOMBE, V., HILL, M. A., EAGLESON,K. L., and BENNETT, M. R. (1984). Motor neuron survival and neurite extension from spinal cord explants induced by factors relcased from denervated muscle. Bruin Rex 291:19-28. O'BRIEN,R. J., and FISCHBACH, G. D. (1986). Isolation of embryonic chick motoneurons and their survival in vitro. .J. Naumci. 65265-3274. OPPENHEIM, R. W., and MAJORS-WILLARD: C. (1978). Neuronal cell death in the brachial spinal cord of the chick is unrelated to the loss of polyneuronal innervation of the wing muscle. Brain Res. 154:148-152. OPPENHEIM. R. W., and CHU-WANG.1. W. (1983). Aspects of naturally-occurring motoneuron death in the chick spinal cord during embryonic development. In: Somatic and ..liitonomic 1Veriw-Mu.rcle Jnteractions. G. Burnstock and G. Vrbova, Eds.. Elsevier, Amsterdam, pp. 57-107. OPPENHEIM, R. W. (1984). Cell death of motoneurons in the chick embryo spinal cord. VIII. Motoneuron prevented from dying in the embryo persist after hatching. Dev. Hiol. 101:35-39. OPPENHEIM, R. W. ( I 986). The absence of significant postnatal motoneuron death in the brachial and lumbar spinal cord of the rat. J. Camp. Xeurol. 246:28 I 286. OPPENHEIM, R. W., HOUENO~J, L., PINCON-RAYMOND, M., POWELL,J. A., RIEGER,R. F., and STANDISH, L. J. (1986). The development of motoneurons in the embryonic spinal cord of the mouse mutant, muscular dysgenesis (mdg/mdg):survival, morphology and biochemical differentiation. Dev. Biol. 114:426-436. R. W., HAVERKAMP. L. J., PREVETTE, D., OPPENHEIM, MCMANAMAN, J. L., and APPEL,S. H. (1988). Reduction of naturally occurring motoneuron death in vivo by a target-derived neurotrophic factor. Science 240~919-922. OPPENHEIM, R. W. (1989). The neurotrophic theory and naturally occurring motoneuron death. Trends Neurosci. 1 2 ~ 22-2 5 55. OPPENHEIM, R. W., BURSZTAJN, S., and PREVETTE, D. (1 989). Cell death in the chick embryo spinal cord: XI, Acetylcholine receptors and synaptogenesis in

skeletal muscle following the reduction of motoneuron death by neuromuscular blockade. Development 107~33 1-341. PAI, A. I. ( 1 9654. Developmental genetics of a lethal mutation ( m d g )in the mouse. I. Genetic analysis and gross morphology. Dev. Biol. 11932-92. PAI. A. I . (1965b). Developmental genetics of a lethal mutation ( m d g ) in the mouse. 11. Developmental analysis. Dev. Bliol. 11:93- 109. PINCON-RAYMOND, M., and RIEGER,F. (1982). Extensive multiple innervation and abnormal synaptogenesis in muscular dysgenesis (mdg/mdg) in the mouse embryo. Reprod. Nulr. Dev. 22:2 17-226. PINCON-RAYMOND, M., RIEGER,F., FOSSET, M., and LAZDUNSKI, M. (1985). Abnormal transverse tubule system and abnormal amount of receptors for Catchannel inhibitors of the dihydropyridine family in skclctal muscle from mice with embryonic muscular dysgenesis. Dev.B i d . 112:458-466. PITTMAN, R. N., and OPPENHEIM, R. W. (1978). Neuromuscular blockade increases motoneuron survival during cell death in the chick embryo. Nuttire 271:364-366. P I T M A N , R. N., and OPPENHEIM, R. W. (1979). Cell death of motoneurons in the chick embryo spinal cord. IV. Evidence that a functional neuromuscular interaction is involved in the regulation of naturally occumng cell death and thc stabilization of synapses. J. Cornp. Neiirol. 187:425-446. D., and RAYMOND, G. (1980). Calcium-dePOTREAU, pendent electrical activity and contraction of voltageclampcd frog single muscle fibers. J. Physiol. 309:922. POWELL, J. A., RIECER,F., BLONDET,B., DREYFUS, P., and PINCON-RAYMOND, M. (1984). Distribution and quantification of ACh receptors and innervation in diaphragm muscle of normal and rndg mouse cmbryos. Dev. B i d . 101:168-180. PRESTIGE, M. C. (1967). The control of cell number in the lumbar ventral horns during the development of Xenopus laevis tadpoles. ,I. Embryol. Lxp. Morphol. 18r3.59-387. RIEGER.F., BOURNAUD, R., SHIMAHARA. T., GARCIA, L., PINCON-RAYMOND, M., ROMEY,G., and LAZDUNSKI, M. ( 1 987a). Restoration of dysgenic muscle contraction and calcium channel function by co-culture with normal spinal cord neurons. Nuture 330563-566. RIEGER,F., DREYFUS.P., GARCIA, L., PINCON-RAYMOND, M., TASSIN, A. M., HOUENOU,L., HARRIS, A. J., ROMEY,G., FOSSET,M., and LAZDUNSKt, M. (1987b). Genetic and pharmacological models of muscle inactivity. In: Model S~7zzcm.sin hTmrotoxiro1ogy: Alternulive Approaches to Animal Testing. A. Shahar and A. M. Goldberg, Eds.. Alan R. Liss. Inc., New York, 253: pp. 3 15-327. M. (1981). Muscle RIEGER,F.. and PINCON-RAYMOND, and nerve in muscular dysgenesis in the mouse at

birth: sprouting and multiple innervation. Dev. B i d TANAK~, H. (1 987). Chronic application of curare does 87:85- 101. not increase the level of motoneuron survival-proROMEY,G., RIEGER,F., R ~ . N A U D J. , F., PINCON-RAYmoting activity in limb muscle extracts during the MOND, M., and LAZDUNSKT, M. ( 1 986). The electronaturally occurring motoneuron cell death period. physiological expression of Ca++ channels and of a Dev. Bid. 124:347-357. pain sensitive Ca'+ activated K+ channels is abolished WHITE, I-I. L., and CAvA21,LIT0, c, J . 970). Choline in skeletal muscle cells from mouse with muscular acetyltransferase enzyme mechanism and mode of indysgenesis. Biochem. Hiophys. Res. Commun. hibition by a styrylpyridine analogue. Biochi. 136~935-940. Biophy.r Acta 206:343-358. ROSAI.KI,1. B. (1967). An improved procedure for K. s. ( 1964). Histochemical serum creatine phosphokinase determination. J , ~ , ~ b , WIRSEN. E.. and LARSSON, diferentiation of skeletal muscle in foetal and newClin. Med. 69596-705. born mice. J. Ernbryol. Ekp. Murphol. 12(4):759-767. TANABE, T., BEAM,K. G., POWELL, J. A., and N I I M A , S. ZISKIND-CONHAIM, I,., and BENNEI'T, J. I. (1982). The ( 1 988). Restoration of excitation-contraction couefects of electrical inactivity and denervation on the pling and slow calcium current in dysgenic muscle by distribution of acetylcholine receptors in developing dihydropyridine receptor complementary DNA. Nurat muscle. Dev. B i d . 90:185-197. ture 336:134-139.