Proc. Nati. Acad. Sci. USAVol. 77, No. 11, pp. 6922-6925, November 1980Neurobiology

Dependence of in vitro myogenesis on a trophic protein present inchicken embryo extract

(trophic influence/muscle differentiation)

TAE HWAN OH AND GEORGE J. MARKELONISDepartment of Anatomy, University of Maryland, School of Medicine, Baltimore, Maryland 21201

Communicated by Richard L. Sidman, August 21, 1980

ABSTRACT A trophic protein (sciatin) purified from sciaticnerves has been shown to enhance the morphological development and to promote the maintenance of sletal muscle cellsin vitro [Markelonis, G. J. & Oh, T. H. (1979) Proc. NatL Acad.Sci. USA 76,2470-24741. We have elicited a specific antiserumagainst purified sciatin in rabbits. By using this antiserum, wehave examined whether sciatin is also required for the initialdifferentiation of avian myogenic cells in vitro. Sciatin wasfound to be a component of chicken embryo extract, a constit-uent of culture medium required for myogenesis in vitro, by animmunodiffusion assay and by NaDodSO4 gel electrophoresisof immunoprecipitates. The removal of sciatin from chickenembryo extracts by immunoprecipitation with antiserumagainst sciatin completely inhibited myogenesis. When my-ogenic cells were grown in culture medium from which sciatinhad been removed, the cells failed to differentiate beyond themyoblast stage. However, when sciatin (25 pg/ml) was addedto the sciatin-absorbed culture medium, normal myogenesisensued. Furthermore, myogenic cells underwent normal myo-genesis in the absence of embryo extract if sciatin (25 ,g/ml)was added to the culture medium. These results demonstratethat sciatin is the component of chicken embryo extract requiredfor myogenesis and that the protein influences the initial dif-ferentiation of myogenic cells in vitro.

Innervation exerts a "trophic" influence on the morphologic,physiologic, and metabolic properties of skeletal muscle (forreviews, see refs. 1 and 2). To explain the mechanism by whichthe motor nerve conveys trophic influence to muscle, it has beenpostulated that such trophic influence is mediated in part bya trophic substance(s) released by the nerve (1-3).The best evidence for the trophic influence of a neurally

derived substance has been obtained from in vitro experiments(4-11). By using embryonic muscle culture, Oh (12, 13) dem-onstrated that extracts of central and peripheral nervous tissueenhanced the morphological development of aneural musclecells, increased protein synthesis and acetylcholinesterase ac-tivity, and maintained matured, cross-striated muscle fibers forseveral months in the absence of innervation. In the absence ofthe extracts or innervation, cultured muscle fibers atrophiedand degenerated rapidly. The active agent in peripheral ner-vous tissue was characterized as a protein of rather high mo-lecular weight (13) and was partially purified (14-16).We have purified a biologically active protein from adult

chicken sciatic nerves that duplicates many of the trophic ef-fects of innervation on cultured muscle (17). In preliminarypublications*t we have referred to this trophic protein as"neurotrophic factor" or NTF. We have now formally adoptedthe name sciatin in order to avoid confusion with the neuro-notrophic factors described by workers in other laboratories.The term sciatin is intended to designate the trophic proteinthat was isolated and purified from the sciatic nerve. Sciatin is

acidic, migrates with a molecular weight of 84,000 on Na-DodSO4 gel electrophoresis, and has a native molecular weightof 86,400 as estimated by sedimentation equilibrium centrif-ugation.* We immunized rabbits and elicited a specific anti-serum against purified sciatin. By using antiserum againstsciatin, we investigated the role of sciatin in myogenesis in vitro.In the present report, we show that sciatin is present in chickenembryo extract (EE) and that sciatin is the component of EErequired for myogenesis because the removal of sciatin fromEE by immunoprecipitation with antiserum against sciatincompletely inhibits myogenesis.

MATERIALS AND METHODSPurification of Sciatin. Sciatin was purified from adult

chicken sciatic nerves by ion-exchange chromatography onDEAE-cellulose followed by gel filtration on Sephadex G-100superfine (17). The purity of sciatin was greater than 97%, asestimated by densitometric analysis of NaDodSO4 gels. *

Preparation of Antiserum Against Purified Sciatin. Puri-fied sciatin was lyophilized, dissolved in a small volume ofphosphate-buffered saline, and emulsified with an equal vol-ume of complete Freund's adjuvant (Difco). The emulsion,containing 1 mg of purified sciatin, was injected intradermallyin a total of eight sites on the axillary-inguinal lymphatic lineof a male rabbit. Similar injections were repeated at 2-weekintervals, except that the emulsion used incomplete Freund'sadjuvant. An initial anti-sciatin serum was obtained 2 weeksafter immunization and antiserum was then collected biweeklyfor 4 months. Preimmune rabbit serum served as a control. Bothpreimmune and anti-sciatin rabbit sera were heat-inactivatedat 560C for 30 min before use.

Addition of anti-sciatin serum (diluted 1:99 with standardculture medium) to myoblast cultures prevented fusion ofmyoblasts. When the antiserum was removed f~om these cul-tures, however, the myoblasts then began fusing to form mul-tinucleated myotubes within 24 hr. Addition of anti-sciatinserum (diluted 1:99) to cross-striated myotube cultures resultedin atrophy and degeneration of the myotubes. Within 48 hr,cross-striations disappeared and the myotubes became thinnerand attenuated. On the other hand, preimmune serum had nosuch deleterious effects upon the fusion of myoblasts or uponthe viability of myotubes (unpublished data).

Preparation of EE. The extract was prepared from 10-day-old chicken embryos (Spafas, Lancaster, PA) as describedby Paul (18). Briefly, whole embryos were rinsed three timeswith Hanks' balanced salt solution and pressed through a 50-mlsyringe opening, and an equal volume of Hanks' solution was

The publication costs of this article were defrayed in part by pagecharge payment. This article must therefore be hereby marked "ad-vertisement" in accordance with 18 U. S. C. §1734 solely to indicatethis fact.

6922

Abbreviation: EE, chicken embryo extract.* Markelonis, G. J., Kemerer, V. F. & Oh, T. H. (1980) Soc. Neurosci.,10th Meeting (abstr.) 6,376.

t Oh, T. H. & Markelonis, G. J. (1980) Int. Soc. Dev. Neurosci., IstAnnual Meeting (abstr.), p. 145.

Dow

nloa

ded

by g

uest

on

Janu

ary

15, 2

021

Proc. Nati. Acad. Scd. USA 77 (1980) 6923

added to this extract. The mixture was kept at room tempera-ture for 30 min and centrifuged at 2000 X g for 20 min. Thesupernatant was collected and stored at -200C. Before use, theextract was thawed and debris was removed by centrifugationat 25,000 X g for 20 min.Removal of Sciatin from EE by Immunoprecipitation.

Sciatin was removed from EE by immunoprecipitation withanti-sciatin serum as described (19). An aliquot of EE and anequal volume of anti-sciatin serum (heat-inactivated) weremixed well and the mixture was kept at room temperature for2 hr and then at 40C for 48 hr. The mixture was centrifuged at1500 X g for 20 min to remove the precipitated antibody-antigen complexes. The supernatant was collected and used inexperimental media as described below. Anti-sciatin antibodiesremaining in the supernatant were assessed by double-immu-nodiffusion assay, and no immunological reactivity was foundin the supernatant when tested against purified sciatin. Theprotein constituents of the precipitated antibody-antigencomplexes were analyzed by NaDodSO4/gel electrophoresis(see Results). EE was also absorbed with heat-inactivatedcontrol, preimmune rabbit serum as described above.

Muscle Culture. Trypsin-dissociated skeletal muscle cellswere prepared from breast muscles of 12-day-old chickenembryos (Spafas) as described (13, 14). Selective enrichmentof myogenic cells was obtained by the method of Richler andYaffe (20). The myogenic cells (5 X 105 cells) were plated ona Linbro plastic dish (35 X 15 mm) that had been coated withcollagen (Calbiochem; acid-soluble). The cultures were grownat 370C in a humidified atmosphere of 95% air/5% C02.

Culture Media. The standard culture medium consisted of87.5% (vol/vol) Dulbecco's modified Eagle's medium, 10%(vol/vol) horse serum (heat-inactivated), and 2.5% (vol/vol) EE.For the first 24 hr, all cultures were grown in standard culturemedium in order to permit all viable myogenic cells to attachfirmly to the substratum. After 24 hr, the standard culturemedium was removed and replaced with experimental media.The experimental media were: (i) 90% (vol/vol) Dulbecco'smodified Eagle's medium and 10% (vol/vol) horse serum(standard culture medium without EE); (ii) 85% Dulbecco'smodified Eagle's medium, 10% horse serum, and 5% EE ab-sorbed with preimmune serum or with anti-sciatin serum(standard culture medium without sciatin); (iti) standard culturemedium without sciatin supplemented with purified sciatin (25,og/ml); or (iv) standard culture medium without EE supple-mented with purified sciatin (25 ,ug/ml). The fresh experi-mental media were replaced every 3 days.

Biochemical Techniques. NaDodSO4/polyacrylamide gelelectrophoresis in thin-layer gels was performed according toa modification (17) of the method described by King and La-emmli (21). Protein was determined by the method of Lowryet al. (22), with bovine serum albumin as standard.

RESULTSThe specificity of anti-sciatin serum was demonstrated bydouble-immunodiffusion precipitin reactions. Anti-sciatinserum gave a single precipitin line upon double immunodif-fusion when tested against purified sciatin (Fig. 1A). As shownin Fig. IA, preimmune serum or anti-sciatin serum absorbedwith purified sciatin showed no immunological reactivity.The standard culture medium used in the present study

consisted of Dulbecco's modified Eagle's medium, horse serum,and EE. To determine whether sciatin was present in thestandard culture medium, each constituent of the medium wastested for its immunological reactivity against anti-sciatinserum. As shown in Fig. lB, double immunodiffusion demon-strated that only EE showed a single precipitin line, which di-

1

2 6

3 54

A

1 6

2 /5

3 4

B

FIG. 1. Double-immunodiffusion precipitin reaction demon-strating the specificity of anti-sciatin serum (A) and the presence ofsciatin in EE (B). (A) Center well, purified sciatin (125 Ag/ml); wells1 and 3, preimmune serum; wells 2,4, and 6, antiserum diluted 1:3 withphosphate-buffered saline; well 5, diluted antiserum (1:4) absorbed(vol/vol) with purified sciatin (86 ig/ml). (B) Center well, antiserumdiluted 1:3 with phosphate-buffered saline; well 1, undiluted Dul-becco's modified Eagle's medium; well 2, undiluted horse serum; well3, EE diluted 1:3 with phosphate-buffered saline; well 4, undilutedstandard culture medium; well 5, purified sciatin (125 jig/ml); well6, EE absorbed (vol/vol) with antiserum.

minished as the sciatin present in EE was selectively absorbedwith anti-sciatin serum.The presence of sciatin in EE was further demonstrated by

NaDodSO4 gel electrophoresis. When EE was separated intoprotein constituents by NaDodSO4 gel electrophoresis, the re-sulting gels showed the presence of a protein with a RF identicalto that of sciatin purified from adult chicken sciatic nerves (17)(Fig. 2). Furthermore, the protein precipitated from EE byanti-sciatin serum also migrated in a position identical to thatof sciatin (Fig. 2).

Fig. 3 compares the myogenesis of muscle cells grown in the

S I 2 3

aF __b a

HC

d

Le

FIG. 2. NaDodSO4 gel electrophoresis of EE, precipitated anti-gen-antibody complexes, and purified sciatin. Arrowhead indicatesthe position of sciatin. Lane 1, EE (25 jg). Lane 2, precipitated anti-gen-antibody complexes (23 ,ug) obtained by mixing EE and anti-sciatin serum (vol/vol): H, heavy chains of IgG (molecular weight%50,000); L, light chains of IgG (molecular weight %25,000). Lane 3,purified sciatin (5 jg). LaneS, molecular weight standards: a, phos-phorylase B (94,000); b, albumin (67,000); c, ovalbumin (43,000); d,carbonic anhydrase (30,000); e, trypsin inhibitor (20,100); f, a-lac-talbumin (14,400).

Neurobiology: Oh and Markelonis

Dow

nloa

ded

by g

uest

on

Janu

ary

15, 2

021

6924 Neurobiology: Oh and Markelonis

/ooh; /,.

_<7 ~ :._ Ac

.-X

B

ah'N ' i

N.< .. K.

N

- *.1..

I-I

,a-

- I,- -/

C / / NtAth

* < /~~

U /

/ 1

F-F .. /JI -

FIG. 3. Muscle cultures grown instandard culture medium and experi-mental media. For the first 24 hr, allcultures were maintained in standardculture medium and were then grown inexperimental media for 6 days. Seven-day-old cultures were fixed in 10%(vol/vol) formalin and stained withhematoxylin/eosin. (A) Muscle culturegrown in standard culture medium. (B)Muscle culture grown in standard cul-ture medium in the absence of EE. (C)Muscle culture grown in standard cul-ture medium in the presence ofEE ab-sorbed with preimmune serum. (D)Muscle culture grown in standard cul-ture medium in the presence ofEE ab-sorbed with anti-sciatin serum. (E)Muscle culture grown in standard cul-ture medium in the presence of EE ab-sorbed with anti-sciatin serum supple-mented with purified sciatin (25 ,g/ml).(F) Muscle culture grown in standardculture medium in the absence of EEsupplemented with purified sciatin (25Ag/ml).

standard culture medium and in experimental media. Themorphological development of muscle cultures grown instandard culture mediu is similar to that described previously(23). One day after plating all viable myogenic cells had firmlyattached to the substratum and were apparent as typical spin-dle-shaped myoblasts. After 2-3 days in culture, the myoblastpopulation increased as a result of active cell division, and atthe same time myoblasts began fusing into multinucleatedmyotubes. By 7-10 days, muscle cultures showed maturedmyotubes (Fig. 3A).

Muscle cultures grown in the absence of EE showed no my-otube formation and consisted of mononucleated myoblasts(Fig. 3B). By contrast, cultures grown in medium in which EEhad been absorbed with preimmune serum exhibited maturedmyotubes (Fig. 3C). However, when muscle cultures weregrown in medium in which EE had been absorbed with anti-sciatin serum, no myogenesis occurred (Fig. 3D). Tbe inhibitionof myogenesis observed in cultures grown in the medium inwhich EE had been absorbed with anti-sciatin serum could bedue to the absence of sciatin in the medium. To test this possi-bility, muscle cultures were grown in medium in which EE hadbeen absorbed with anti-sciatin serum supplemented withpurified sciatin (25 Ag/ml). As shown in Fig. 3E, the culturesshowed matured myotubes with no morphological aberrations,indicating that sciatin is required for myogenesis. In fact, asshown in Fig. 3F, when purified sciatin (25 ,ig/ml) was sub-stituted for EE, normal myogenesis occurred.

DISCUSSIONMyogenesis of muscle cells in vitro requires the addition ofnutritional supplements to a chemically defined medium. Forexample, EE is a well-known supplemental requirement of cellculture medium. De la Haba and Amundsen (24) reported thepresence of two factors in EE, a factor that promotes the fusionof myoblasts and a factor that promotes the development ofmyotubes. Slater (25) also demonstrated the growth-promotingactivity of EE in muscle cultures. Jabaily and Singer (7) foundthat embryonic brain and liver extracts are most effective inpromoting proliferation of myogenic cells in vitro. They alsofound the presence of two proteins in these extracts, one of highmolecular weight and one of low molecular weight, both of

which must be present to produce full mitogenic activity. Oh(12) suggested that the active component of EE that stimulatesmyogenesis might be neural in origin because a high molecularweight protein obtained from embryonic brain and spinal cordalso stimulated muscle development in vitro. In the presentstudy, we demonstrated that sciatin, a trophic protein purifiedfrom adult chicken sciatic nerves (17), is present in EE and isrequired for myogenesis in vitro. Furthermore, immunocyto-chemical studies at the light microscopic level by the peroxi-dase-antiperoxidase technique have revealed that sciatin is lo-calized in sciatic nerve axons and spinal cord neurons of adultchickens but not in such non-nervous tissues as liver and mus-cle.t Thus, our results would suggest that, for avian myogeniccells at least, myogenesis in vitro is under neural influence.Numerous studies of muscle development in vitro have

suggested that the early stages of myogenesis are independentof neural influence. In tissue culture, myogenic cells dissociatedfrom developing muscle can differentiate to form multinu-cleated myotubes in the absence of innervation (for reviews,see refs. 26-29). Such myogenesis in vitro involves essentiallythe same morphological, physiological, and biochemicalchanges known to occur in developing muscle in vivo. Also,experiments on developing muscle in situ suggested that theinitial differentiation of myogenic cells is independent of neuralinfluence (30-33). Although myogenic cells can develop intocross-striated myotubes in the absence of innervation, theysubsequently undergo atrophy and degeneration. In denervatedembryonic muscle, there are relatively few myotubes but manyunfused myoblasts. Further differentiation of myotubes andmaintenance of muscle fibers require innervation in vito as wellas in vitro (34).

In contrast to the observations described above, several studies(35-39) have indicated that nerve tissue exerts a "trophic" in-fluence on the initial differentiation of myogenic cells. Inchorioallantoic grafting experiments, Avery et al. (35) foundthat spinal cord tissue exerts a profound influence on the initialdifferentiation of myogenic cells. The presence of spinal cordalso accelerates the growth and differentiation of myogenic cellsin culture (36). In a study of muscle grafts, Eastlick (37) alsoobserved that innervation influences the proliferation of my-ogenic cells. Bonner (38) found that clonable myoblast popu-

Proc. Nati. Acad. Sci. USA 77 (1980)

N

Dow

nloa

ded

by g

uest

on

Janu

ary

15, 2

021

Proc. Natl. Acad. Sci. USA 77 (1980) 6925

lations obtained from denervated developing muscle are re-duced to about 60% of the level found in normally innervatedmuscle. Filogamo et al. (39) proposed the theory that the de-termination of myogenic cell lines during muscle developmentin vvo is nerve dependent. They observed contacts betweenthe myogenic cells of somites and primitive axons growing fromthe neural tube as early as 35 hr of incubation in chicken em-bryos. Our results agree with these observations that neuralinfluence is required both for the initial differentiation andsubsequent maintenance of muscle cells in vitro.

Note Added in Proof. While this paper was in press, the purificationand characterization of sciatin were published in full (40).

We thank Drs. L. Guth and J. D. Gearhart for their invaluable ad-vice. We also thank Mr. D. Derr and Ms. C. Sofia for their technicalassistance and Mrs. E. DeLong for preparation of the manuscript. Thiswork was supported by the National Institutes of Health (NS 15013 andNS 16076), the Muscular Dystrophy Association, and the ParalyzedVeterans of America.

1. Guth, L. (1968) Physiol. Rev. 48,645-687.2. Gutmann, E. (1976) Annu. Rev. Physiol. 38, 177-216.3. Singer, M. (1978) Am. Zool. 18,829-841.4. Lentz, T. L. (1971) Science 171, 187-189.5. Oh, T. H., Johnson, D. D. & Kim, S. U. (1972) Science 178,

1298-1300.6. Younkin, S. C., Brett, R. S., Davey, B. & Younkin, L. H. (1978)

Science 200, 1292-1295.7. Jabaily, J. & Singer, M. (1978) Dev. Biol. 64, 189-202.8. Podleski, T. R., Axelrod, D., Ravdin, P., Greenberg, I., Johnson,

M. M. & Salpeter, M. M. (1978) Proc. Natl. Acad. Sci. USA 75,2035-2039.

9. Jessell, T. M., Siegel, R. E. & Fischbach, G. D. (1979) Proc. Nati.Acad. Sci. USA 76,5397-5401.

10. Kumori, H., Gonoi, T. & Hasegawa, S. (1979) Brain Res. 175,109-118.

11. Rathbone, M. P., Vickers, J. D. & Logan, D. M. (1979) J. Exp.Zool. 210, 463-472.

12. Oh, T. H. (1975) Exp. Neurol. 46,432-438.13. Oh, T. H. (1976) Exp. Neurol. 50,376386.14. Markelonis, G. J. & Oh, T. H. (1978) Exp. Neurol. 58, 285-

295.15. Oh, T. H. & Markelonis, G. J. (1978) Science 200,337-339.

16. Oh, T. H. & Markelonis, G. J. (1979) in Muscle Regeneration,ed., Mauro, A. (Raven, New York), pp. 417-427.

17. Markelonis, G. J. & Oh, T. H. (1979) Proc. Natl. Acad. Scd. USA76,2470-2474.

18. Paul, J. (1961) Cell and Tissue Culture (Livingstone, Edin-burgh).

19. Garvey, J. S., Cremer, N. E. & Sussdorf, D. H. (1977) Methodsin Immunology (Benjamin, Reading, PA), pp. 313-321.

20. Richler, C. & Yaffe, D. (1970) Dev. Biol. 23, 1-22.21. King, J. & Laemmli, U. K. (1971) J. Mol. Biol. 62,465-473.22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.

(1951) J. Biol. Chem. 193,265-275.23. Oh, T. H. & Johnson, D. D. (1972) Exp. Neurol. 37,360-370.24. De la Haba, G. & Amundsen, R. (1972) Proc. Natl. Acad. Scd.

USA 69, 1131-1135.25. Slater, C. N. (1976) Dev. Biol. 50,264-284.26. Konigsberg, I. R. (1963) Science 140, 1273-1284.27. Hauschka, S. D. (1968) in The Stability of the Differentiated

State, ed. Ursprung, H. (Springer, Berlin), pp. 37-57.28. Holtzer, H. (1970) in Cell Differentiation, eds. Schjeide, 0. &

De Vellis, J. (Van Nostrand-Reinhold, New York), pp. 476-503.

29. Fischman, D. A. (1972) in The Structure and Function ofMuscle, ed. Bourne, G. H. (Academic, New York), pp. 75-149.

30. Zelena, J. (1962) in The Denervated Muscle, ed. Gutmann, E.(Pub. House Czech. Acad. Sci., Prague), pp. 103-126.

31. Drachman, D. B. (1974) Ann. N.Y. Acad. Sci. 228, 160-176.32. Giacobini-Robecchi, M. G., Giacobiiti, G., Filogamo, G. &

Changeux, J. P. (1975) Brain Res. 83, 107-121.33. Popiela, H. (1976) Exp. Neurol. 53,214-226.34. Crain, S. M. & Peterson, E. R. (1974) Ann. N.Y. Acad. Sci. 228,

6-34.35. Avery, G., Chow, M. & Holtzer, H. (1956) J. Exp. Zool. 132,

409-423.36. Murray, M. R. (1965) in Cells and Tissue in Culture, ed. Willmer,

E. N. (Academic, New York), Vol. 2, pp. 311-372.37. Eastlick, H. L. (1943) J. Exp. Zool. 93, 27-49.38. Bonner, P. H. (1978) Dev. Biol. 66,207-219.39. Filogamo, G., Peirone, S. & Sisto Daneo, L. (1978) in Maturation

of Neurotransmission, eds. Vernadakis, A., Giacobini, E. &Filogamo, G. (Karger, Basel), pp. 1-9.

40. Markelonis, G. J., Kemerer, V. F. & Oh, T. H. (1980) J. Biol.Chem. 255, 8967-8970.

Neurobiology: Oh and Markelonis

Dow

nloa

ded

by g

uest

on

Janu

ary

15, 2

021