Early cross-striation formation in twitching Xenopus myocytes in culture

Transcription

1 Proc. Nati. Acad. Sci. USA Vol. 85, pp , March 1988 Neurobiology Early cross-striation formation in twitching Xenopus myocytes in culture (spontaneous transmitter release/acetylcholine/trophic interaction) YOSHIAKI KIDOKORO AND MITSUYOSHI SAITO Jerry Lewis Neuromuscular Research Center, Department of Physiology, University of California, School of Medicine, Los Angeles, CA 924 Communicated by Susumu Hagiwara, November 9, 1987 ABSTRACT Spontaneous release of neurotransmitter has been demonstrated in various types of synapses. Its physiological significance, however, is still unknown. In nerve-muscle cultures of embryonic Xenopus laevis, we observed that acetylcholine, which is released spontaneously at the synaptic terminal, caused frequent twitches of muscle cells. These muscle cells developed cross-striations earlier than neighboring nontwitching cells. This effect of innervation was unaffected by tetrodotoxin but was blocked by a-bungarotoxin. Repeated iontophoretic application of acetylcholine or KCI to muscle cells caused twitches and also accelerated the formation of crossstriations. Thus twitching apparently promotes lateral alignment of myofibrils. It is also known that myosin synthesis is higher in twitching muscle cells. Therefore, successfully innervated twitching muscle cells may have an advantage for faster differentiation over neighboring non-twitching muscle cells. We suggest that spontaneously released transmitter may serve as a mediator for trophic interaction at forming synapses. Since the disvery of miniature endplate potentials at the neuromuscular junction (1), similar spontaneous release of synaptic transmitter has been observed in many other chemical synapses (2). Although the amplitude of miniature endplate potentials is small in adult muscle cells, at newly formed neuromuscular junctions it is nsiderably larger due to the higher input impedance of young myotubes (3). The amplitude of synaptic potentials increases as acetylcholine (AcCho) receptors accumulate at the postsynaptic membrane (4). Spontaneous synaptic potentials are detected soon after a growth ne ntacts a muscle cell (5). In fact, the growth ne releases AcCho even before ntacting a muscle cell (6, 7). The physiological significance of spontaneous release of transmitter is not yet known. AcCho thus released from growing neurites may serve as a mediator for trophic interaction. MATERIALS AND METHODS The dorsomedial portion of Xenopus embryos at stage 14 (8) was dissected, and the vering skin was removed by using a pair of needles. Neurons and myocytes were dissociated in Ca2+- and Mg2+-free saline as described (9) and plated on the llagen-ated ver glass in culture medium [.6x Dulbec's modified Eagle's medium (buffered by Hepes to ph 7.4, GIBCO) supplemented with 5% (vol/vol) horse serum]. These cultures were maintained at room temperature (21-23 C) in the glass chamber. Xenopus myocytes are mononucleated and nstitute tail musculature of the tadpole. Cultured myocytes are flat, and cross-striation is clearly seen under a phase-ntrast microspe (4 x water-immersion objective, Zeiss). About 17 hr The publication sts of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in acrdance with 18 U.S.C solely to indicate this fact after plating, cultures were rinsed three times with culture medium to remove unattached cells. The great majority of attached cells were myocytes, and some ofthem possessed an adherent neuron. Typically a few thousand myocytes per dish attached to the substrate, and 2-4 neurons adhered to the myocytes. AcCho and K+ were applied iontophoretically with a current pulse 1.3 msec long and 1-6 na in amplitude at.1-1 Hz. Electrodes were filled with 1 M AcChoCl dissolved in normal saline (12 mm NaCl; 3 mm KCl; 1 mm CaCl2;.5 mm MgCl2; 8 mm Hepes'NaOH; 1 mm gluse; ph 7.4) or with 3 M KCl dissolved in distilled water. For AcCho application, a ntinuous braking current of 3-8 na was applied to avoid desensitization of receptors. All experiments were carried out at room temperature (21-23C). RESULTS About two-thirds of myocytes ntacted by neurons twitched spontaneously after 24 hr of culture. These spontaneous twitches were blocked by.1 um a-bungarotoxin, indicating that they were the result of spontaneous release of AcCho from neurons. Properties of spontaneous synaptic potentials in Xenopus nerve-muscle cultures were studied (4, 1). Their amplitudes were extremely variable, and some spontaneous synaptic potentials were large enough to cause twitches without muscle action potentials. About 3 myocytes not ntacted by a neuron and >1 myocytes ntacted by a neuron in each culture dish were examined closely under a phase-ntrast microspe for cross-striation formation, and the percentage of striated cells was determined. A set of six cultures was thus examined at various intervals after plating (Fig. 1A). Cross-striations in myocytes not ntacted by a neuron first developed =z23 hr after plating. At first the number of striated bands was low, and striation was nfined to a small portion of the myocyte. The percentage of myocytes with cross-striations as well as the extent of striation in each myocyte increased with time. Forty-four hours after plating virtually all myocytes were cross-striated. The time for 5%o of myocytes not ntacted by a neuron to acquire cross-striation was measured (Fig. 1A) and found to be hr (mean + SD) after culturing at room temperature (for all four sets of cultures). Myocytes, ntacted by neurons and twitching, developed cross-striations much earlier (Figs. 1A and 2B) than myocytes not ntacted by neurons (Fig. 2A). At 23 hr a majority of twitching myocytes showed clear cross-striations. Some myocytes ntacted by neurons did not twitch during the observation period of -5 min. Probably the neurons ntacting myocytes were not cholinergic or were cholinergic, but released AcCho very infrequently. These myocytes were not striated 23 hr after culturing (Fig. 1A and Fig. 2C), Abbreviation: AcCho, acetylcholine.

2 Neurobiology: Kidokoro and Saito A s Proc. Natl. Acad. Sci. USA 85 (1988) I.- 5- I-- LL HOURS AFTER PLATING B,_ z i CO I - L. HOURS FIG. 1. (A) Percentage of striated myocytes is plotted against hours after the myocytes were cultured. Filled circles, myocytes not ntacted by a neuron; open circles, myocytes ntacted by a neuron and twitching; and open squares, myocytes ntacted by a neuron but not twitching. A smooth curve was drawn "by eye" to estimate the time when 5% of the myocytes not ntacted by a neuron had acquired striation. In this case it was 29 hr after plating (marked by a short vertical bar). (B) Frequency of spontaneous twitching (twitches, no. per min) is plotted against time. The twitching was observed on the video screen, and twitches were registered with a photodiode placed on the monitor screen, and the output was rerded on a chart rerder. Twitches were variable in amplitude, and obviously some small twitches were not rerded. This myocyte had partial striation at the beginning of the rerding, and the cross-striation developed further during the observation period. indicating that neuron ntact by itself does not promote the formation of cross-striation. Furthermore, when culture medium ntained.1,um a-bungarotoxin, none of the myocytes ntacted by neurons twitched or developed cross-striations ahead of myocytes not ntacted by neurons, indicating that activation of AcCho receptors is required for accelerated formation of cross-striation. Another way to reduce spontaneous twitching is to increase Mg2+ ncentration in culture medium, which reduces the frequency of spontaneous transmitter release. In medium with added 1 mm Mg2 +, muscle cells ntacted by neurons rarely twitched. Only one myocyte ntacted by a neuron twitched infrequently out of 26 examined, whereas in ntrol cultures 15 myocytes twitched out of 28 cells examined that were ntacted by neurons. No accelerated formation of cross-striation was observed in high Mg2+ medium. In ntrast, 3,uM tetrodotoxin in culture medium neither affected twitching nor blocked the facilitatory effect of twitching on the formation of cross-striation, indicating that Na+ action potentials are not involved in this phenomenon. Therefore, we nclude that cholinergic neurons, by releasing AcCho, often caused twitches in myocytes and thereby accelerated the formation of cross-striation. It has been reported in a rat skeletal muscle cell line that high K+ medium (25 mm KCl added to culture medium) promotes formation of cross-striations (11). In Xenopus cultures, however, addition of KCl to culture medium to 1 or 2 mm did not increase the percentage of cross-striated myocytes 23 and 31 hr after culturing mpared with ntrol. Thus, maintained depolarization does not accelerate the formation of cross-striations in Xenopus myocytes. Myocytes ntacted by a neuron twitched at various rates between.1 and 18 twitches per min. In one culture 23 hr after culturing, the average twitch frequency was twitches per min (n = 11). The average twitch frequency in each myocyte, however, increased gradually during the period when cross-striations were forming (Fig. 1B). The extent of cross-striation was generally greater in frequently twitching myocytes than those twitching infrequently. This is probably because frequently twitching cells have been twitching longer and not necessarily because the formation of cross-striations depends on the twitch frequency. As

3 198 Neurobiology: Kidokoro and Saito A Proc. Natl. Acad. Sci. USA 85 (1988) B C FIG. 2. Phase-ntrast micrographs of Xenopus myocytes in culture 28 hr after plating. Myocytes at this early stage have many yolk granules, which are phase-bright and surround the nucleus. (A) Typical myocyte without neuron ntact and without cross-striation. (B) Myocyte with an adherent neuron (arrow). This myocyte twitched frequently. (Inset) Portion of the cross-striation formation in the myocyte ntacted by a neuron in B. The other myocyte in the upper part of the micrograph does not ntact a neuron or have cross-striations. (C) Myocyte that is ntacted by a neuron but not twitching. The cell is not cross-striated. Other cells in this micrograph also are not cross-striated. (Bar = 1,um.) described below, iontophoretic application of AcCho at various frequencies between.1 and 1 Hz did not noticeably change the extent of cross-striation. To stimulate myocytes directly we applied AcCho iontophoretically. Fig. 3A shows an unstimulated myocyte -24 hr after culturing. After 3 hr of stimulation at.5 Hz, the myocyte had beme partially cross-striated (Fig. 3B). During this period neighboring myocytes did not develop new striations. Similar experiments were repeated on eight FIG. 3. Myocyte not ntacted by a neuron was stimulated with iontophoretically applied AcCho. (A) Before stimulation. (B) After a 3-hr stimulation. Stimuli were delivered at.5 Hz causing partial twitches. Braking current, 8 na; pulse amplitude, 4-6 na; pulse duration, 1.3 msec. Enlarged micrographs are from the portions of the lower magnification micrographs delineated by white lines. Arrows in the enlarge micrograph of B are pointing to dark bands of crossstriation that developed during stimulation. Note that there are no such bands in the enlarged micrograph of A before stimulation. No cross-striation formation was observed in neighboring myocytes not ntacted by a neuron during stimulation. (Bars = 1 jtm.) myocytes under similar nditions. In all cases, formation of cross-striation was observed within 3 hr. The frequency of stimulation between.1 and 1 Hz did not affect the extent of cross-striation. Even when AcCho was applied locally to a fixed point on a myocyte, cross-striations did not necessarily start to form there, suggesting that the formation of crossstriation is not promoted directly by molecules entering through AcCho receptor channels. Furthermore, similar induction of partial cross-striation was observed in myocytes not ntacted by a neuron when 3 M KC1 was applied iontophoretically to induce twitches (in all 1 cases examined). Therefore, this effect is not specifically stimulated by AcCho. DISCUSSION In Xenopus embryos cross-striation was first observed under the light microspe at stage 22 (12), which is close to the time when spontaneous synaptic potentials were first re-

4 Neurobiology: Kidokoro and Saito rded (stage 21; ref. 13). It appears that formation of functional synaptic ntacts and cross-striation develop almost simultaneously in situ. This is almost the same time as nerve-induced receptor accumulation occurs at the junction (stages 22-26; ref. 14). In the Zebra fish, Myers et al. (15) observed twitching in one or two muscle fibers in the vicinity of the growth ne of the motor nerve emerging from the spinal rd. These muscle fibers were cross-striated. It is, therefore, probable that a similar phenomenon exists in the animal. However, Cohen et al. (16) reported that embryos reared in tricaine-ntaining solution during the initial period of nerve-muscle interaction (stages 18-41) developed normally and swam in a manner indistinguishable from untreated ntrol embryos when returned to tricaine-free water. Since the tricaine treatment decreased the amplitude and frequency of spontaneous synaptic potentials, this observation seemingly eliminated the physiological significance of twitching caused by spontaneous release of AcCho. However, tricaine treatment did not abolish spontaneous synaptic potentials. Remaining infrequent spontaneous synaptic potentials uld be adequate to stimulate cross-striation, or the effect of tricaine treatment on the animal behavior is too subtle to be observed casually. An obvious question is: How does twitching accelerate cross-striation formation? Twitching does not distort myocytes. The accelerated striation formation, therefore, is not likely to be due to the simple mpaction and nsequent visualization of existing striation. The assembly mechanism of myofibrils and formation of cross-striation are not well understood (17). In cultured Xenopus muscle cells, the process of myofibril assembly has been studied by electron microspy in thin-section and in whole-mount preparations (18). In newly cultured myocytes (dissociated between stages 2 and 22) nascent nyofibrils are found among abundant microfilament bundles in the area close to the surface membrane. New sarmeres are added at the nonstriated end of a nascent myofibril. Bundles of myofibrils assemble and align to form a wider cross-striated structure, visible with the light microspe. For lateral alignment of myofibrillar bundles, the involvement of a special protein, such as desmin, is postulated (19-21). In our culture nditions, myocytes were capable of twitching even before cross-striations became visible. Therefore, small bundles of assembled myofibrils must have been already formed, but lateral alignment had not been mpleted. Twitching, therefore, must accelerate the lateral alignment process. The molecular mechanism of this lateral alignment of myofibrillar bundles is yet to be elucidated. In cultured heart cells the role of ntraction in regulating myosin subcellular organization has been studied (22). Addition of culture medium ntaining 5 mm KCI resulted in a mplete arrest of otherwise actively twitching heart cells in culture. After >2 days in high K+ culture medium, the fluorescent staining pattern with the myosin light chain antibody was the same in nontwitching cells and ntrols. It was thus ncluded that twitching does not affect myosin assembly or lateral alignment of myofibrillar bundles. However, as we have demonstrated in this study the formation of cross-striations occurs in the absence of twitching; twitching merely accelerates it. Therefore, it is possible that in ntraction-arrested heart cells cross-striation formation was slightly delayed; but when the myosin distribution was examined, there was no difference between twitching cells and nontwitching cells. During neuromuscular junction formation, electrical activities of the muscle cell regulate developmental events. The AcCho receptor density in the extrajunctional region decreases sometime after establishment of the functional junction (3, 23). This is most likely due to neuron-induced activities in the muscle fiber (24). Another example is the Proc. Natl. Acad. Sci. USA 85 (1988) 1981 loss of electrical upling among myotomal muscle cells in developing Xenopus larvae (25, 26), which is believed to be due to electrical, rather than ntractile, activities (27). Electrical activity in muscle fibers may also influence the elimination of multiple innervation (28). However, the precise underlying molecular mechanism for any of these examples is not known. Trophic interactions between nerve and muscle cells occur in multiple forms at various developmental stages. At the initial stage of neuron-muscle interaction, the growth ne releases AcCho and generates synaptic potentials in the target muscle cell (5). At this stage the ntact between nerve and muscle is simple, involving no specialized structures detected by electron microspy (29). However, as the result of large transient depolarizations generated by spontaneous release of AcCho, the muscle cell twitches, which may in turn promote muscle differentiation. It is known that ntraction increases myosin synthesis (3, 31). Also it is plausible that other structural proteins and enzymes are synthesized at an accelerated rate in twitching cells. Thus the successful active innervation of a muscle cell may give this cell advantages in differentiation over the muscle cell that is not innervated or innervated but relatively inactive. In Aplysia neuronal cultures it has been reported that a neurotransmitter, serotonin, inhibits the growth ne motility of a certain type of neuron and prevents formation of electrical synaptic nnection (32). Thus, spontaneous release of neurotransmitters may play an important role in the formation of neuronal nnections. This study is supported by grants from the National Institutes of Health (NS23753) and from the Muscular Dystrophy Association to Y.K. We thank Mr. Jeff Rohbough and Drs. Steven Young and Ida Chow for their criticism of the manuscript. 1. Fatt, P. & Katz, B. (1952) J. Physiol. (London) 117, Kuno, M. (1971) Physiol. Rev. 51, Diamond, J. & Miledi, R. (1962) J. Physiol. (London) 162, Kidokoro, Y., Anderson, M. J. & Gruener, R. (198) Dev. Biol. 78, Kidokoro, Y. & Yeh, E. (1982) Proc. Natl. Acad. Sci. USA 79, Young, S. H. & Poo, M.-m. (1982) Nature (London) 35, Hume, R. L., Role, L. W. & Fischbach, G. D. (1982) Nature (London) 35, Nieuwkoop, P. D. & Faber, J. (1967) Normal Table of Xenopus laevis (Daudin) (Elsevier, Amsterdam). 9. Anderson, M. J., Cohen, M. W. & Zorychta, E. (1977) J. Physiol. (London) 268, Kidokoro, Y. (1984) Neurosci. Res. 1, Klier, G., Schubert, D. & Heinemann, S. (1977) Dev. Biol. 57, Muntz, L. (1975) J. Embryol. Exp. Morphol. 33, Kullberg, R. W., Lentz, T. L. & Cohen, M. W. (1977) Dev. Biol. 6, Chow, I. & Cohen, M. W. (1983) J. Physiol. (London) 339, Myers, P. Z., Eisen, J. S. & Westerfield, M. (1986) J. Neurosci. 6, Cohen, M. W., Greschner, M. & Tucci, M. (1984) J. Physiol. (London) 348, Fishman, D. A. (1982) in Muscle Development: Molecular and Cellular Control, eds. Pearson, M. L. & Epstein, H. F. (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY), pp Peng, H. B., Wolosewick, J. J. & Cheng, P.-C. (1981) Dev. Biol. 88, Lazarides, E. (198) Nature (London) 283, Tokuyasu, K. T., Dutton, A. H. & Singer, S. J. (1983) J. Cell Biol. 96,

5 1982 Neurobiology: Kidokoro and Saito 21. Dento, S. I. & Fishman, D. A. (1984) J. Cell Biol. 98, Klein, I., Daood, M. & Whiteside, T. (1985) J. Cell. Physiol. 124, Bevan, S. & Steinbach, J. H. (1977) J. Physiol. (London) 267, Burden, S. (1977) Dev. Biol. 61, Blackshaw, S. E. & Warner, A. E. (1976) J. Physiol. (London) 255, Armstrong, D. L. & Warner, A. E. (198) J. Physiol. (London) 3, 64p (abstr.). Proc. Nati. Acad. Sci. USA 85 (1988) 27. Armstrong, D. L., Turin, L. & Warner, A. E. (1983) J. Neurosci. 3, Dennis, M. J. (1981) Annu. Rev. Neurosci. 4, Nakajima, Y., Kidokoro, Y. & Klier, F. G. (198) Dev. Biol. 77, Bandman, E. & Strohman, R. C. (1982) J. Cell Biol. 93, McDermott, P., Daood, M. & Klein, I. (1985) Am. J. Physiol. 249, H Haydon, P. G., McCobb, D. P. & Kater, S. B. (1984) Science 226,