Vol. 9, 155-163, February 1998 Cell Growth & Differentiation 155

Vasopressin and Insulin-like Growth Factors SynergisticallyInduce Myogenesis in Serum-free Medium’

Simona Minotti, Bianca Maria Scicchitano,Clara Nervi, Sigfrido Scarpa, Marco Lucarelli,Mario Molinaro, and Sergio Adamo2Dipartimento di Istologia ed Embnologia Medica [S. M., B. M. S., C. N.,M. M., S. A.] and Dipartimento di Biotecnologie Cellulari ed Ematologiae I Istituto di Clinics Chirurgica [S. S., M. L], Universit#{224}“La Sapienza,”00161 Rome, Italy

AbstractTerminal differentiation of myogenic cells has longbeen known to be positively regulated by insulin-likegrowth factors (IGFs). Arg8-vasopressin (AVP) has been

recently reported to potently Induce myogenicdifferentiation. In the present study, the effects and themechanisms of action of AVP and IGFs on myogeniccells have been investigated under conditions allowinggrowth and differentiation of myogenic cells in asimple serum-free medium. Under these conditions, L6and L5 myogenic cells slowly proliferate and do notundergo differentiation (less than I % fusion up to 7days). AVP rapidly (2-3 days) and dose-dependentlyinduces the formation of multinucleated myotubes.Creatine kinase activity and myosin accumulation arestrongly up-regulated by AVP. Insulin or IGF-I or IGF-II,at concentrations that cause extensive differentiationin serum-containing medium, induces a modest degreeof differentiation in serum-free medium. Thesimultaneous presence of AVP and of one of the IGFsin the synthetic medium induces maximaldifferentiation of L6, L5, and satellite cells. Theexpression of both myogenin and Myf-5 is dramaticallystimulated by AVP. Our results indicate that AVPinduces a significant level of myogenic differentiationin the absence of other factors. Furthermore, theysuggest that to express their full myogenic potential,IGFs require the presence of other factors normallypresent in serum and fully mimicked by AVP. Thesestudies support the conclusion that terminal myogenicdifferentiation may depend on the presence of

Received 3/4/97; revised 10/27/97; accepted 12/8/97.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to mdi-cate this fact.1 Supported by Grant 418 from Telethon Italy. The financial support ofConsiglio Nazionale delle Richerche, Progetto Finalizzato ApplicazioniCliniche della Ricerca Oncologica (Grant 96.00651 .39 to S. A.) and As-sociazione Italiana Ricerca sul Cancro (grant to M. M.) is also acknowl-edged.2 To whom requests for reprints should be addressed, at Dipartimento diIstologia ed Embriologia Medica, via Scarpa 14, Universit#{224}“La Sapienza,”00161 Rome, Italy. Phone: 39-6-49766756; Fax: 39-6-4462854; E-mail:adamo@axrma.uniromal it.

differentiation factors rather than the absence ofgrowth factors.

IntroductionTerminal differentiation of myogenic cells is regulated, atleast in vitro, by hormones and growth factors (henceforthcollectively referred to as hormones). Fibroblast growth fac-tor and transforming growth factor 13are potent inhibitors ofmyogenic differentiation (1-3), both of which act on relativelyearly steps of the differentiation process. Other factors suchas adrenocorticotropic hormone, platelet-derived growthfactor BB, -y-melanocyte-stimulating hormone, leukemia-inhibitory factor, and other cytokines have been reported toprimarily affect myoblast proliferation and have shown van-able effects on myogenic differentiation (4-6). IGF-I,3 IGF-Il,and insulin (IGFs) display unique properties, because theystimulate, rather than inhibit, myogenic differentiation in sev-eral myogenic cell lines and primary cultures of myoblastsand satellite cells (1 , 7, 8). One exception is represented bythe widely used C2 cell line, which does not seem to respondto exogenous lGFs, basically because C2 cells secrete largeamounts of these hormones (9). It is remarkable that the

differentiative action of the IGFs is accompanied by a mito-genic action (1). It has been proposed that the effects of IGFson myogenic cell differentiation and proliferation are medi-ated by the IGF-I receptor, which exhibits the binding spec-ificity IGF-l > IGF-Il > insulin (1, 10, ii).

Recently, AVP and related peptides have been shown toconstitute a new family of positive regulators of the terminaldifferentiation of myogenic cells (L6, mouse satellite cells;Refs. 12 and 13). By interacting with V1-type receptors, AVPcauses dose-dependent stimulation of the fusion of myo-blasts into myotubes and of the expression and accumula-tion of myogenin and Myf-5 products at both the RNA andprotein level. Other markers of myogenic differentiation, suchas myosin, nicotinic acetylcholine receptor function, andsubunit expression, were also found to be stimulated by AVPand related peptides (1 2, 13).

The mechanism by which IGFs and AVP stimulate myo-genesis is likely mediated by the enhancement of the tran-scniptional activity of MRF genes (12, 14). MRF proteins(myogenin, MyoD, Myf-5, and MRF4) share homology within

a basic helix-loop-helix motif that mediates dimenization andbinding to a consensus CANNTG sequence (E-box) presentin the promoter of many muscle-specific genes. Each of theMRF genes can activate myogenesis when expressed innonmyogenic cells (15, 16).

3 The abbreviations used are: IGF, insulin-like growth factor; AVP, arg8-vasopressin; CK, creatine kinase; FBS, fetal bovine serum; MRF, rnyo-genesis-regulatory factors; mAb, monoclonal antibody.

156 Hormonal Control of Myogenic Differentiation

The recent discovery of AVP and related peptides as myo-

genic differentiation factors makes it interesting to compare

their effects with those of the lGFs and to analyze their

respective mechanisms of action. Such a study is compli-cated by the presence of serum in the culture medium, whichcontains a number of more or less defined factors, including

mitogens, which, in general, repress the expression of the

differentiated phenotype. This is particularly true for myo-

genic cells in culture (1 , 8), whose differentiation is classicallyinduced by decreasing the concentration of serum in theculture medium. These reasons have prompted us to use a

chemically defined medium in which the proliferation anddifferentiation of myoblasts can be modulated by the addi-

tion of specific hormones, including AVP. In this study, we

report that under carefully controlled conditions, AVP is able

to induce a significant level of myogenic differentiation in the

absence of other factors, and that AVP and IGFs synergis-tically stimulate the expression of the differentiated pheno-

type in cultured myogenic cells.

ResultsMorphological DifferentiationEffects of AVP and Insulin. L6 cells were plated at 5000cells/cm2 in DMEM + 10% FBS and shifted 24 h after plating

to DMEM + 1 % BSA (serum-free medium). AVP (0.1 �u�i) orinsulin (0.1 �M) or both were added, and the cells were

cultured for several days. Under serum-free conditions, cells

slowly proliferated, regardless of the treatment, and by thethird day of culture in serum-free medium, the cultures that

had received AVP alone displayed the presence of multinu-cleated myotubes (Fig. 1B). No fusion was detectable incontrol cultures (Fig. 1A). Insulin-treated cells (Fig. iC) pre-sented a very modest percentage of fusion, whereas exten-

sive fusion (approximately 75%) was obtained in the pres-ence of both AVP and insulin (Fig. iD). Insulin-treatedcultures consisted mostly of mononucleated myoblasts, withrare, thin myotubes containing less than 6 nucleVfiber. Con-versely, in cultures treated with 0.1 �.tM AVP, frequent, ratherlarge myotubes containing an average of 12 nuclei/fiber werepresent. Larger myotubes (average, 23 nucleVfiber) formed incultures treated with both AVP and insulin. Parallel expeni-ments conducted with L5 cells yielded a superimposable

morphological pattern of differentiation. Calculation of the

percentage of fusion for both L6 and L5 cultures (Fig. iE)confirmed that a significant level of myogenic differentiation

of both cell lines was induced by AVP and potentiated by thesimultaneous presence of insulin in serum-free medium.

Evaluation of the total number of nuclei (unfused + fused) per

microscopic field (Fig. iF) did not suggest dramatic effects of

either hormone or of their combination on either L6 or L5 cellproliferation under these culture conditions.

Although cell lines such as L6 and L5 represent valid

models for the study of myogenesis, it was of interest to

verify the hormonal response of primary cultures of satellite

cells under the conditions of this study. The effect of AVP or

insulin or their combination on satellite cell differentiation in

serum-free medium was assessed by evaluating the percent-

age of myosin heavy chain-expressing cells after 72 h ofhormone treatment (see “Materials and Methods”). The re-

suIts (Fig. 2) indicate that AVP (0.1 �LM) stimulated myosinheavy chain accumulation to a higher extent than 0.1 �M

insulin, whereas maximal stimulation of myosin heavy chain

expression was induced by simultaneous treatment withboth hormones.

The time course of differentiation of L6 cell cultures treated

with 0.1 �.1.MAVP or 0.1 �M insulin or both in serum-freemedium is shown in Fig. 3A. The data indicate that AVP-treated cells underwent a rapid burst of fusion during thethird day of treatment and, by the end of the 4th day oftreatment, reached approximately 40% fusion. Conversely,no significant amount of fusion was detected in control cellsduring the same period, and a modest percentage of fusion

was measurable in insulin-treated cultures. The simultane-ous presence of AVP and insulin induced extensive fusion,

starting after the second day of hormone treatment and

reaching approximately 80% fusion by the end of the fourth

day. No significant change in the percentage of fusion ofcontrol or hormone-treated cultures occurred after the fourthday, up to the seventh day (data not shown). Analysis of totalnucleVmicroscopic field during the period examined mdi-cated no striking difference among the various culture con-ditions, confirming that none of the treatments significantlyaffected L6 cell proliferation in the absence of serum (Fig.3C). Comparison of the effects of AVP and insulin in theabsence and in the presence of FBS shows that in the lattercondition, AVP induced a higher degree of fusion than inserum-free medium (Fig. 3, B versus A); insulin promoted

extensive differentiation (at variance with its effect in serum-free medium), and control cells differentiated to a significant

degree. Furthermore, it may be noted that in the presence of5% FBS, hormone-induced L6 cell differentiation was aslower process than in the absence of serum. Serum-con-taming medium induced L6 cells to proliferate to a much

higher extent than in serum-free conditions (Fig. 3, D versusC). However, in the presence of 5% FBS, neither AVP norinsulin induced significant effects on cell number at the con-

centration used, in agreement with our previous data (1 2, 13).The dose dependency of the effect of AVP alone or in

combination with insulin in serum-free medium is reported in

Fig. 4A. A significant effect of AVP on cell fusion was evidentat a concentration of 0.1 nM, reaching a plateau at the highest

AVP concentrations used. A similar dose dependency, al-though on higher values of fusion, was measurable when a

fixed concentration of insulin (0.1 �M) was present along withthe various AVP concentrations (Fig. 4A). Both in the ab-

sence and in the presence of insulin, AVP stimulated L6-C5myoblast fusion with an ECso � 0.3 nM. A similar experimentwas conducted to investigate the insulin concentration de-pendency of L6 cell fusion in serum-free medium, both in theabsence and in the presence of a fixed concentration (0.1 p.r�i)

of AVP (Fig. 4B). In the presence of AVP, insulin significantlystimulated L6 cell fusion at concentrations �30 n�i, and

maximal stimulation of fusion occurred at insulin concentra-tions in the 0.1-3 �tM range. A modest but reproduciblestimulation of L6 cell differentiation occurred at the sameinsulin concentrations, also in the absence of AVP (Fig. 4B).Microscopic evaluation of total nuclei in the experimentalconditions reported for Fig. 4, A and B, did not indicate

c . � .. �. .

.#{176}�, .‘. - . ,. .

‘l� � �

�%

:.. � �

D AVP+Ins

.‘ .- ‘‘.� ., . .‘ . . -

% � � _ � . - ‘ -#{176}% ...

,‘‘ �a � � ‘�_ #{149}� - -,.. �: --.#{149}

..‘ , .% ..� ., . . % - ,- #{149}..-:.- ,. ‘?. � - . .. . �. S #{149}

� ‘1:_::�: � �‘:��: � : � � :��� “ ‘ � ‘. ... #{149}�* . . #{149}:

‘C � #{149}. , . �‘‘��: . � ‘-#{149}� � ; :%

� #{149}.l S .‘ %#{149} .. a

: #{149}; � , � ‘:‘�.#{149} �: � #{149}:�:1��0#{149}

:-�;::��3’:::::’:; � � � :.#{149} �:� #{149},� . - ‘ .-.. � ‘S _ _ �

%_�. � . ‘: . - #{149}� ‘ � : #{149} ‘

C Insulin 0.1 pM

‘ .� ,‘%� . - ..-. , , , .‘

I.. � #{149} __,��_ ..: � . . #{149}, � ...#{176} . �;b_ �- . � . .... ,�t. �.:‘�‘. ;�-. #{149}�.“..

‘ � . . . 0 -, � . ‘

‘ . �‘.. . .#{176}#{176}

F’ � #{149} �: �� . � #{149}�‘ . � ..... . :,#{149}

.‘ .. . - : S.... � �, � � � . . .., � .. #{149}

- .� #{149} ‘..�..; . . ,.. , a

,.‘� . . . - .�1 � ‘ � ,� - .. .

I 00

.� 80

1’ ::

a 20

0

200

160 �

0)

:140

0c�’ CO

cp� CO� �\� 0

�#{149}

A CtrI B AVPO.1 �tM

Cell Growth & Differentiation 157

I � �L51

Fig. 1 . Morphological analysis of L6 and L5 cells cultured in serum-free medium. A-D, photomicrographs of L6 cultures at the third day in serum-freemedium (fourth day of culture) in the absence (A) or in the presence of0.1 �M AVP (B), 0.1 �tM insulin (C), and AVP and insulin (D). Similarly prepared culturesof L6 and L5 cells were microscopically counted, and the results were expressed as the percentage of fusion (E) and the total (unfused + fused)nucleVmicroscopic field (F).

50

40

30

20

10

it?4)Ua)>

..�

(I)00.

C.)

-.--- Ctrl -A- Ins, 0.1 pM-U - AVP,0.1 pM -v-- Ins+AVPSatellite cells

I-

�

Ctrl AVP Ins AVP+ Ins

Fig. 2. Effect of AVP and insulin on the differentiation of mouse satellitecells in serum-free medium. Primary cultures of satellite cells were shiftedto serum-free medium and treated with 0.1 �M AVP and/or 0.1 �M insulinas indicated. After 72 h, the cells were fixed and subjected to immuno-cytochemical analysis using the anti-myosin heavy chain mAb MF2O.

I 00

!; 80

‘� 60

a)1::0

400

! 300‘�5 200C.)

C

I 00

0

day of culture

Fig. 3. Time course of the differentiation (percentage of fusion, A and B)and proliferation (nuclei/field, C and D) of L6 cells cultured in serum-freemedium (left panels) or in the presence of 5% FBS (right panels). Allcultures were seeded on day 0 at the same initial density (5000 ceIIs/cm�)in DMEM + 10% FBS and shifted 24 h later to either serum-free mediumc4 and C) or 5% FBS medium (B and D). #{149},control (no hormone addition);., 0.1 �M AVP; A, 0.1 j.tM insulin; V, AVP + insulin. The arrows indicate thetime of hormone addition.

0123456012345678

158 Hormonal Control of Myogenic Differentiation

significant effects of either hormone, alone or in combination,

on L6 cell proliferation at all concentrations tested (data not

shown), except for a modest increase (1 0-1 5%) in total

nuclei induced by insulin at 10 and 30 f.LM.

IGF-l and IGF-Il. In the experiments described, insulin

was used as an inexpensive and readily available alternative

to the physiological ligands IGF-I and IGF-Il (1). The dose-

dependent effect of these factors was investigated as shown

in Fig. 4, C and D, and compared with that of insulin. In the

presence of 0.1 .LM AVP, significant stimulation of cell fusion

was induced by IGF-I (Fig. 4C) and IGF-II (Fig. 4D) at con-

centrations of 1-1 0 n�. This effect reversed at higher IGF

concentrations, in agreement with previous data (1 7). It is to

be noted that the biphasic shape of the curves, more evident

for IGF-l and IGF-II than for insulin (Fig. 4B), was likely related

to the mitogenic effect of these factors at the highest con-

centrations (up to 2.5-fold stimulation of total nuclei at IGF-I

or IGF-II concentrations above 10 nM; data not shown).

The modest effect of the IGFs in serum-free medium (com-

pared to their effects in the presence of serum) could be

attributed to the lower densities attained by L6 cells in se-rum-free medium, such that even fusion-competent cells

were prevented from fusing. To verify this possibility, L6

cultures were set up at various initial densities up to four

times the density used in the experiments already described,

thus reaching cell densities similar or higher than those dis-

played by cultures maintained in serum-containing medium,

whose differentiation was strongly enhanced by IGFs. The

results obtained (data not shown) indicate that under fusion-

permissive conditions (e.g., in the presence of AVP ± IGFs),

a progressive acceleration of myoblast fusion occurred as

the initial seeding density of L6 cells was increased stepwise

from 2,500 to 20,000 cells/cm2; however, even at the highest

initial density, lGFs were unable to induce extensive fusion of

L6 cells in the absence of AVP.

Effect ofAVP and IGFs on Myosin Level and CKActivityTo further substantiate our findings, biochemical markers of

terminal myogenic differentiation were also analyzed. The

effect of AVP and IGFs on the accumulation of myosin in L6

cells grown in serum-free medium was investigated by im-

munoblot analysis using the anti-myosin heavy chain mAb

MF2O (Fig. 5A). After 3 days of culture in serum-free medium,

control cells (DMEM + 1 % BSA) did not display detectable

evidence of myosin heavy chain, whereas a strong signal

was obtained from 0.1 p.M AVP-treated cells. A weak positive

signal was induced by IGF-l (3 nM) and insulin (0.1 MM). Again,

the combination of AVP and any one of the IGFs induced

significant accumulation of myosin heavy chain (Fig. 5A).

The specific activity of CK, which was extremely low in

serum-free medium without hormone additions, was strongly

enhanced in the presence of 0.1 �LM AVP (Fig. 5B). In insulin-

treated cells, the specific activity of CK was also significantly

enhanced compared to that of controls (over 20-fold stimu-

lation), although we constantly found CK specific activityvalues lower than those in AVP-treated L6 cells. As already

noted for fusion, simultaneous addition of AVP and insulin

synergistically induced very high levels of CK activity (Fig.5B).

80

60

40

20

0

Bt

0 0.1 1 10 100

[AVP], nM

s> /* I�$�\c�’ �

� -

C D

A AA.&..A A A�- �0

0 0.3 3 30 0 0.3 3 30

(IGF-l], nM (IGF-Il], nM

Fig. 4. Dependency of L6 differentiation (percentage of fusion) on theconcentration of AVP, lGFs, and their combinations in serum-free me-dium. L6 cells were seeded at 5000 cells/cm2 in DMEM + 10% FBS, andat 24 h, the cultures were shifted to serum-free medium, and the hor-mones were added. The cultures were fixed and stained on the fifth dayof culture. A, increasing concentrations of AVP without (0) or with (#{149})0.1,.�M insulin. B-D, increasing concentrations of insulin and IGF-l or IGF-llwithout (Lx) or with (A) 0.1 �M AVP, respectively.

B

160

�120

�80

0 40

0�

Fig. 5. Effect of AVP and/or IGFs on the accumulation of myosin and theCK activity in serum-free medium. A, immunoblot of extracts of L6 cellscultured for 3 days in the absence (Ctrl’) or in the presence of 0.1 �M AVP,0.1 �LM insulin, 3 nM IGF-l, 3 n�.i IGF-ll, or their combinations. MHC Std, amyosin standard run along with the experimental samples. B, CK-specificactivity of extracts of L6 cells cultured in serum-free medium in theabsence (Ctri’) or in the presence of 0.1 �M AVP, 0.1 f.LM insulin, or both.

Cell Growth & Differentiation 159

C0(I)

0

0)CUC

a)0.

0 0.1 1 10 100 1K 10K

(Insulin], nM

�c)

80C0U)�2 600

0)(UC

�5 20�0.

Expression of the MRF Genes Myogenin and Myf-5Differentiation of L6 cells is accompanied by expression of

the MRF gene myogenin. The effects of the various treat-

ments on the steady-state level of the myogenin transcript

indicate a strong effect of AVP and a relatively modest effect

of IGF-I, insulin, and IGF-II (in order of potency), whereas the

maximal level of expression is evident in AVP + insulin-

treated cells (Fig. 6A). The time course of the accumulation ofthe myogenin transcript in control or 0.1 j.tM AVP-stimulated

cells is shown in Fig. 6B. The myogenin transcript was un-

detectable in cells grown for 24 h in DMEM + 10% FBS,

whereas a low level of myogenin mRNA was present in

control L6 cells at 1 , 2, and 3 days after shifting the culturesto serum-free medium. At the same times, a strong increase

in the level of myogenin mRNA was evident in AVP-treated

cultures. Comparison with the data obtained in the presence

of 5% FBS (1 2) confirmed that induction of the expression of

myogenin by AVP occurred much earlier in serum-free me-

dium (24 h) than in 5% FBS (48 h).

The presence and the subcellular localization of myogeninprotein was also investigated by immunofluorescence stain-

ing of L6 cells, using the myogenin-specific mAb F5D (Ref.

1 8; Fig. 7). After 3 days of culture in serum-free medium,

control cells displayed a very weak and diffuse signal. At the

same time, none of the IGFs induced a strong effect,

whereas a strong positive signal correctly localized in thenucleus was elicited in cells treated with AVP or with AVP

plus one of the lGFs (Fig. 7).

Myf-5 and MyoD are known to be expressed early in the

developing somite and to act upstream of myogenin (19).

Whereas MyoD is not expressed in L6 cells (12, 20), it is

interesting to investigate the hormonal regulation of Myf-5expression in L6 cells under the conditions of this study. As

shown in Fig. 6C, no expression of the Myf-5 gene product

was observed in L6 cells growing in proliferative medium

(10% FBS; Lane A), in agreement with our previous obser-

vations (1 2), whereas a very slight expression level was de-

tectable in L6 cells 1 8 h after shifting to serum-free medium

(Lane B). At the same culture time, AVP alone (Lane C), or in

combination with insulin (Lane E), IGF-I (Lane G), or IGF-II

(Lane I) strongly induced Myf-5 mRNA expression. Con-

versely, none of the IGFs alone (insulin, Lane 0; IGF-l, Lane

F; and IGF-II, Lane H) significantly stimulated the expression

of Myf-5 in the absence of serum.

A

rRNA

CtrI

0123

AVP

123

Fig. 6. Northern blot analysis of the expression of myogenin and Myf-5in L6 cells. A, L6 cell cultures, seeded as usual in DMEM + 10% FBS, wereshifted at 24 h to serum-free medium and cultured for an additional 72 hwithout (Ctr!) or with 0.1 �M AVP, 0.1 �tM insulin, AVP + insulin, 4 n�i IGF-l,and 4 n� IGF-ll. Hybridization of the blot with an rRNA probe was used forthe purpose of normalization. B, time course of the effect of 0.1 �M AVPon the expression of myogenin. Control cells at day 0 (24 h of culture inDMEM + 10% FBS) and at 1 , 2, and 3 days of culture in serum-freemedium are shown; 0.1 MM AVP-treated cells at 1 , 2, and 3 days ofhormone treatment in serum-free medium are shown. Ethidium bromidestaining was used to verify equal loading of the samples. C, L6 cells werecultured for 24 h in DMEM + 10% FBS (Lane A) and then shifted toserum-free medium, treated with hormones as indicated, and incubatedfor an additional 18 h (Lanes B-I). Lane B, control; Lane C, 0.1 �M AVP;Lane D, 0.1 �tM insulin; Lane E, AVP + insulin; Lane F, 3 n�.i IGF-l; Lane G,AVP + IGF-l; Lane H, 3 n�.i IGF-ll; Lane I, AVP + IGF-ll.

1&’�j Hormonal Control of Myogenic Differentiation

Myogenin

B

Myogen in � � -

CABC DEFGHI

Myf-5 � ,-�,

rRNA

DiscussionUntil recently, lGFs were considered “unique among growthfactors and hormones in that they stimulate, rather than

inhibit, myogenic differentiation” (1). Our recent discovery of

the positive role played by AVP in the differentiation of myo-

genic cells (12) opened the possibility that other factors, inaddition to IGFs, may positively regulate myogenesis. Our

present results indicate that AVP, but not IGFs, in the ab-

sence of other factors, significantly stimulates myogenic

differentiation.

We found that reproducibility of the effects of hormones in

our system depends on several factors. More or less com-

plex serum-free media were used by several groups for the

analysis of myogenic cell proliferation and differentiation (re-

viewed in Ref. 8). Using L6 and L5 cells, we found that in the

absence of other factors, myogenic differentiation was af-

fected by the particular preparation of BSA used. Indeed,

whereas one of the commercial preparations of BSA that we

tested behaved as the one we have used in this study, one

was toxic for our cells, and in the presence of two additional

commercial BSA preparations, the percentage of fusion was

12 and 18% under control conditions, 45 and 55% in the

presence of 0.1 LM AVP, and 30 and 34% in the presence of

0.1 j.tM insulin, respectively. We selected a commercial BSA

preparation that allowed minimal differentiation unless spe-

cific hormones were present. Furthermore, accurate removal

of serum when cells were shifted from 10% FBS medium to

serum-free medium seemed to be critical, particularly for the

reproducibility of the effect of IGFs. In fact, addition of as little

as 0.5% FBS to serum-free medium resulted in a significantincrement of the percentage of fusion in the presence of 0.1

MM insulin (an approximately 4-fold increase compared toserum-free medium, insulin-treated cultures; data not

shown).

Under these culture conditions, AVP-dependent differen-

tiation of L6 cells was lower than in the presence of FBS inabsolute terms (40% fusion versus over 60%) but was

strongly stimulated when compared to the control (less than

1 % fusion in serum-free controls and approximately 20% in5% FBS controls; Fig. 3). Our results on the effect of insulin

or IGF-I/IGF-II on L6, L5, and satellite cell differentiation are

unexpected in that: (a) neither IGF-I/IGF-II nor insulin stimu-lated differentiation to any major extent; and (b) they syner-

gistically activated myogenic differentiation in the presence

of AVP. These results were confirmed by measuring other

differentiation markers such as myosin heavy chain or CKactivity.

The effects of AVP and IGFs on myogenic terminal differ-

entiation parameters are similar to those observed when the

expression of Myf-5 and myogenin was investigated. Taken

together, these results suggest that the mechanism of action

of AVP involves the induction of the expression of these

regulatory genes. It is worth recalling that Myf-5 is known toact upstream of myogenin, and that its expression in L6 cells

declines when that of myogenin increases (21 , 22). Thesedata and the observation that the onset of differentiation in

the presence of AVP and insulin is faster than in the presence

of AVP alone (Fig. 3A) may help explain the relatively modest

effect of the combined treatment with AVP and IGF-I (corn-

pared to that with AVP alone) on the expression of Myf-5.

Some discrepancy seems to exist between our data andthose of Ewton et a!. (23). Using a serum-free medium very

similar to ours, they reported that both IGF-I and IGF-Il

Ctrl IGF-I IGF-lI Insulin

AVP AVP+IGF-l

#{149}!#_ � S

*. .4’

, a� � � 9 S

C

� �‘ �

Cell Growth & Differentiation 161

, � 1.�

AVP + IGF-ll

I, �

#{149}�; “� a

%

,

I

‘‘,

AVP + Insulin

I , �.%

a

�, ‘

‘.

.,‘ \�1��

-

Fig. 7. lmmunofiuorescence analysis of the expression of myogenin in L6 cells. L6 cells, seeded and cultured as described, were treated for 3 days without(Ctr() or with the following hormones: 0.1 �M AVP; 4 n� IGF-l; 4 n� IGF-ll; and 0.1 �M insulin or their combinations.

stimulated myogenic differentiation in terms of both myotube

formation and OK activity. More recently, Engert et a!. (24)

have shown that IGF-I induces first proliferation and then

differentiation and/or hypertrophy of L6 cells cultured in se-

rum-free medium. In addition, it must be noted that they

obtained a significant level of fusion in control conditions

(serum-free medium). As already pointed out, one possible

factor that could explain these discrepancies is the origin of

the BSA used by these groups (see previous discussion).

Another possibility could be the incomplete removal of serum

when the cells are shifted from the serum-containing me-

dium (20% FBS medium; Ref. 24) to the serum-free medium.

This possibility is particularly relevant in light of our obser-

vation that as little as 0.5% FBS significantly increases the

effect of insulin on fusion. We have also observed that initial

plating densities above those examined in this study or pro-

longed culturing in serum-containing medium before shifting

the cultures to serum-free medium results in hormone effects

different from those reported in this study. All these possi-

bilities could explain the differences between our results and

those presented by the other groups.

The origin and the mechanism of action of the lGFs have

been extensively studied (reviewed in Ref. 25), and their rolein skeletal muscle development is basically not disputed.

Conversely, the physiological relevance of the effect of AVP

on myogenic cell differentiation (1 2, 26, 27) is not univocally

established. On one hand, several reports indicate an indeed

modest effect of AVP on adult muscle (28), and no primary

disorders of skeletal muscle are evident in the clinical con-

ditions characterized by either reduced or increased levels of

AVP. On the other hand, an unsuspected developmental role

for a hormone whose effects in the adult were well known

has been recently reported for endothelin-i (29), and high

levels of immunoreactive AVP were measured in extracts of

human embryonic skeletal muscle; its concentration de-

dined as gestational age increased (30). Furthermore, Hanley

eta!. (31) reported the presence of a vasopressin-like peptide

in the mammalian sympathetic nervous system, a finding

that may cast light on the question of the origin of AVP

(or a vasopressin-like peptide) during skeletal muscle

development.

It is commonly thought that myogenic cells differentiate

spontaneously unless they are exposed to sufficientamounts of growth factors; the presence of such factors

induces myogenic cells to proliferate and represses the ex-

pression of the myogenic phenotype. By establishing real

control conditions, our study, on the contrary, supports the

view that terminal differentiation of at least some myogenic

cell culture models (such as those used in this study, which

do not secrete large amounts of differentiation factors) de-

pends on the presence of specific differentiation signals

rather than the absence of proliferative signals.

In conclusion, by using carefully controlled culture condi-

tions and measuring parameters such as fusion, accumula-

tion of myosin, CK activity, Myf-5, and myogenin expression,

we have consistently found that IGFs and AVP represent

families of differentiation factors whose synergistic actions

induce maximal myogenic differentiation. These two families

of myogenic factors have distinct receptors and distinct sig-

naling pathways. Research is in progress in our laboratory to

elucidate the level and the molecular mechanism of the syn-

ergism between AVP and lGFs in regulating the expression of

the myogenic phenotype.

Materials and MethodsSynthetic AVP and related peptides, bovine insulin, CK assay kit, Tn-Reagent, and other reagents were purchased from Sigma Chemical Co.

(St. Louis, MO). IGF-l and IGF-ll were purchased from Chemicon(Temecula, CA) and from Intergen (Purchase, NY). Quick-Hyb Northernblot hybridization solution was obtained from Stratagene (Heidelberg,

Germany). Fatty acid-free BSA (Boehringer Mannheim) was selected bycomparing BSA from different commercial sources, on the basis that nosigns of toxicity appeared even after long-term culture of L6 cells, and no

162 Hormonal Control of Myogenic Differentiation

fusion could be detected in the control conditions. The anti-myosin heavy

chain mAb MF2O (32) was a kind gift of Dr. D. Fishman (Cornell University,New York, NY). The anti-myogenin Ab F5D (18) was kindly provided by Dr.

W. E. Wright (University of Texas, Dallas, TX).Cell Cultures. Subcloning and characterization of L6 rat myogenic

cells (33) was reported previously (26). Cells of the subclone CS (L6-C5),a clone that had shown significant differentiation ability when culturedunder appropriate conditions (26), were used throughout this study. Thecells were routinely seeded at the density of 5000 cells/cm2 (unlessotherwise indicated) in DMEM supplemented with 100 units/mI penicillin,100 �g/ml streptomycin, and 10% heat-inactivated FBS. Twenty-four h

after plating, cuitures were extensively washed with DMEM and shifted toserum-free medium consisting of DMEM supplemented with BSA, with or

without other additions. Media were changed every 3 days, as appropri-

ate. Preliminary experiments performed with a medium composed ofDMEM supplemented with BSA allowed us to asses the minimal concen-tration (1 %) of BSA compatible with survival and good morphology of the

cells. Furthermore, it was found that BSA preparations obtained from

different commercial sources were widely different from one another whendifferentiation of L6-C5 cells was morphologically evaluated, both in the

absence and in the presence of hormones (see “Results”).

Rat L5 myogenic cells were routinely cultured in F14 medium supple-mented with 10% FBS (34). For serum-free medium experiments, 15 cellswere cultured as described above for L6 cells, except that gelatin-coateddishes were used.

Primary cultures of satellite cells were prepared from young adult CD1

mice as reported previously (12). After 48 h of cuiture, the proliferationmedium (DMEM supplemented with 20% horse serum and 5% chickembryo extract) was substituted with the serum-free medium described

above, and hormones were added as appropriate.Measurement of Myoblast Fusion and Growth. May Grunwald-

Giemsa-stained cultures were evaluated for cell fusion. Cells were con-sidered fused only if cytoplasmic continuity and at least three nuclei werepresent in each myotube. Greater than 10 randomly selected fields andgreater than 300 total nuclei were counted for each sample. The ratiobetween the number of nuclei in myotubes versus the total number ofnuclei per microscopic field was expressed as the percentage of fusion.Each experimental point represented in the graphs is the mean ± SD of

the counts obtained from four to nine independent samples obtained in atleast two separate experiments.

CK Assay. Cells were homogenized in 30 m�a HEPES and 1 m�i EDTA(pH 7.2), and the 20,000 x g supernatant was used to measure CK activity

according to Szasz et a!. (35). Protein content of cell extracts was meas-ured as described previously (36).

Immunochemistry. Myogenin expression was analyzed in cuitured

cells after fixation in 4% paraformaldehyde in PBS and permeabilization in0.5% Triton X-100 in PBS. The monolayers were washed in 1% BSA inPBS and incubated overnight at room temperature with the undiluted

supernatant of F5D hybridoma cells (18). After extensive washing with 1%

BSA in PBS, the cells were incubated for 1 h at room temperature withfluorescein-conjugated goat antimouse immunoglobulmn (Cappel Labora-tories; diluted 1:50; Ref. 18).

Sarcomeric myosin expression was determined by Western blot anal-ysis using the MF2O mAb, and detection was performed by the enhancedchemiluminescence method (Amersham) according to the manufacturer’sinstructions. The same mAb was used for the immunocytochemical stain-ing of myosin heavy chain in primary satellite cell cultures. The reactionwas developed as reported previously (1 2), using a sheep antimouse

immunoglobulin, horseradish peroxidase-linked antibody obtained fromAmersham (Milan, Italy).

RNA Preparation and Northern Blot Analysis. Total RNA was iso-

lated from the cells by the Tri-Reagent procedure as indicated by themanufacturer. Equal amounts of total RNA (30 �g) were separated byelectrophoresis in 0.66 M formaldehyde-i .2% agarose slab gel, trans-ferred to Nytran membranes (Schleicher & Shuell, Hayward, CA) by cap-illary blotting. and cross-linked to the blots by UV irradiation (37). Otherprocedures were as described in Ref. 12, except that Quick-Hyb was usedas prehybridization and hybridization buffer. The rat myogenin probe(1 500 bp), kindly provided by Dr. W. E. Wright, and the human Myf-5

probe (1250 bp; Ref. 20) were the EcoRl restriction fragments of therespective cDNA clones. Equal loading of the samples was verified eitherby ethidium bromide staining of the gels (37) or by hybridization of the

blots with a ribosomal 18S RNA (rRNA) probe (American Type CultureCollection, Rockville, MD).

AcknowledgmentsWe thank G. Cossu and M. Fiszman for critical reading of the manuscript

and F. Naro and A. Di Noi for helpful advice and discussion. Mouse

satellite cells were kindly provided by M. Coletta and L Capece.

References1 . Florini, J. R., Ewton, D. 1, and Magri, K. A. Hormones, growth factors,and myogenic differentiation. Annu. Rev. Physiol., 53: 201-216, 1991.

2. Clegg, C. H., Linkhart, T. A., Olwin, B. B., and Hauschka, S. D. Growthfactor control of skeletal muscle differentiation: commitment to terminaldifferentiation occurs in G1 phase and is repressed by fibroblast growthfactor. J. Cell Biol., 105: 949-956, 1987.

3. Olson, N. E., Sternberg, E., Hu, J. S., Spizz, G., and Wilcox, C. Reg-ulation of myogenic differentiation by type (3 transforming growth factor.J. Cell Biol., 103: 1799-1806, 1986.

4. De Angelis, L, Cusella De Angelis, M. G., Monaco, L, Raschell#{224},G.,and Cossu, G. Pro-opiomelanocortin gene is expressed in post-implan-tation mouse embryos and enhances growth potential of myogenic cells.Dev. Dyn., 198: 265-272, 1993.

5. Yablonka-Reuveni, Z., Balestreri, T. M., and Bowen-Pope, D. F. Reg-ulatmon of proliferation and differentiation of myoblasts derived from aduftmouse skeletal muscle by specific isoforms of PDGF. J. Cell Biol., 111:1623-1629, 1990.

6. Austin, L, and Burgess, A. W. Stimulation of myoblast proliferation Incufture by leukemia inhibitory factor and other cytokines. J. Neurol. Scm.,

101: 193-197, 1991.

7. de Ia Haba, G., Cooper, G. W., and Elting, V. Hormonal requirementsfor myogenesis in vitro: insulin and somatotropin. Proc. NatI. Acad. Scm.USA, 56: 1719-1723, 1966.

8. Flormni, J. R. Hormonal control of muscle growth. Muscle & Nerve, 10:

577-598, 1987.

9. Tollefsen, S. E., Sadow, J. L, and Rotwein, P. Coordinate expressionof insulin like growth factor II and its receptor during muscle dlfferentla-tion. Proc. NatI. Aced. Sci. USA, 86: 1543-1547, 1989.

10. Ewton, D. Z., Falen, S., and Flonni, J. R. The type II IGF receptor haslow affinity for IGF-l analogs: pleiotypic actions of lGFs on myoblasts areapparently mediated by the type I receptor. Endocrinology, 120:115-124,1987.

11 . Cohick, W. S., and Clemmons, D. R. The insulin-like growth factors.Annu. Rev. Physiol., 55: 131-153, 1993.

12. Nervi, C., Benedeth, L, Minasi, A., Molinaro, M., and Adamo, S.Argmnmne-vasopressmn induces differentiation of skeletal myogenic cellsand up-regulates myogenin and myf-5. Cell Growth Differ., 6: 81-89,1995.

13. Nervi, C., Benedetti, L, Donchenko, V., Naro, F., Di Noi, A., Molinaro,M., and Adamo, S. Regulation of skeletal muscle differentiation byarginine-vasopressin. In: S. Waxman (ed.), Differentiation Therapy, pp.193-197, Rome: Ares Serono Symposia Publications, 1995.

14. Flormni, J. R., Ewton, D. Z., and Rcof, S. L Insulin-like growth factor-istimulates terminal myogenic differentiation by induction of myogeningene expression. Mel. Endocrinol., 5: 718-724, 1991.

15. Olson, N. E. The MyoD family: a paradigm for development? Genes

Dev., 4: 1454-1461, 1990.

16. Wright, W. E., Binder, M., and Funk, W. Cyclic amplification and targetselection (casting) for the myogenin consensus binding site. Mol. Cell.Biol., 11: 4104-4111, 1991.

17. Florini, J. R., Ewton, D. Z., Falen, S., and Van Wyk, J. J. Biphasicconcentration dependency of the stimulation of myoblast differentiation

by somatomedins. Am. J. Physiol., 250: C771-C778, 1986.

18. CuselIa De Angelis, M. G., Lyons, G., Sonnino, C., De Angelis,L, Vivarelli, E., Farmer, K., Wright, W. E., Molinaro, M., Bouche, M.,Buckingham, M., and Cossu, G. MyoD, myogenmn-mndependent differen-

Cell Growth & Differentiation 163

tiation of primordial myoblasts in mouse somites. J. Cell Biol., 1 16: 1243-1255, 1992.

19. Buckingham, M. Making muscle in mammals. Trends Genet., 8: 144-149, 1992.

20. Braun, T., Bober, E., Buschhausen-Denker, G., Kotz, S., Grzeschik,K., and Arnold, H. H. Differential expression of myogenic determination

genes in muscle cells: possible autoactivation by the Myf gene products.EMBO J., 8: 3617-3625, 1989.

21 . Buckingham, M. Skeletal muscle development and the role of themyogenic regulatory factors. Biochem. Soc. Trans., 24: 506-509, 1996.

22. Mangiacapra, F. J., Roof, S. L, Ewton, D. Z., and Florini, J. R.Paradoxical decrease in myf-5 messenger RNA levels during induction ofmyogenic differentiation by mnsulmn-likegrowth factors. Mol. Endocrinol., 6:2038-2044, 1992.

23. Ewton, D. 1, Roof, S. L, Magri, K. A., McWade, F. J., and Florini, J. R.IGF-ll is more active than IGF-l in stimulating L6A1 myogenesis: greatermitogenic actions of IGF-I delay differentiation. J. Cell. Physiol., 16: 277-284, 1994.

24. Engert, J. C., Berglund, E. B., and Rosenthal, N. Proliferation pre-cedes differentiation in IGF-l-stimulated myogenesis. J. Cell Biol., 135:431-440, 1996.

25. Florini, J. R., Ewton, D. 1, and Coolican, S. A. Growth hormone andthe insulin-like growth factor system in myogenesis. Endocr. Rev., 17:481-517, 1996.

26. Teti, A., Naro, F., Molinaro, M., and Adamo, S. Transduction of the

arginine-vasopressin signal in skeletal myogenic cells. Am. J. Physiol.,265: C1i3-Ci21, 1993.

27. Naro, F., Donchenko, V., Minotti, S., Zolla, L, Molinaro, M., andAdarno, S. Role of phospholipase C and D signalling pathways In vaso-pressmn-dependent myogenic differentiation. J. Cell. Physiol., 171: 34-42,

1997.

28. Wakelam, M. J. 0., and Pette, D. The control of glucose 1,6-biphos-

phate by developmental state and hormonal stimulation in cultured mus-cle tissue. Biochem. J., 204: 765-769, 1982.

29. Kurihara, Y., Kurihara, H., Suzuki, H., Kodama, T., Maernura, K.,Nagai, R., Oda, H., Kuwaki, T., Cao, W. H., Kamada, N., Jishage, K.,Ouchi, Y., Azuma, S., Toyoda, Y., Ishikawa, T., Kumada, M., and Yazachi,

Y. Elevated blood pressure and craniofacial abnormalities in mice defi-

cient of endothelin-1. Nature (Lond.), 368: 703-710, 1994.

30. Smith, A., Stephen, A. J., Arkley, M. M., and Mcintosh, N. Immuno-reactive argmnmnevasopressmn in human fetal and neonatal skeletal muscle.Early Hum. Dev., 28: 215-222, 1992.

31 . Hanley, M. R., Benton, H. P., Lightman, S. L, Todd, K., Bone, E. A.,Fretten, P., Palmers, S., Kirk, C. J., and Michell, R. H. A vasopressin-likepeptide in the mammalian sympathetic nervous system. Nature (Lond.),309: 258-261, 1984.

32. Bader, D., Masaki, T., and Fishman, D. A. Immunochemical analysisof myosmn heavy chain during avian myogenesis in vivo and in vitro. J. CellBiol., 95: 763-770, 1982.

33. Yaffe, D. Retention of differentiation potentialities during prolonged

cuitivation of myogenic cells. Proc. NatI. Acad. Sci. USA, 61: 477-483,

1968.

34. Scarpa, S., Uhlendorf, B. W., and Cantoni, G. The differentiation ofL5/A10 myoblast cell line (a subclone of L5 line) is controlled by changesof cuiture conditions. Cell Differ., 17: 105-1 14, 1985.

35. Szasz, C., Gruber, W., and Bemt, E. Creatine kinase in serum: deter-mination ofoptimum reaction conditions. Clin. Chem., 22: 350-356, 1978.

36. Bensadoun, A., and Weinstein, D. The assay of proteins in the pres-ence of interfering materials. Anal. Biochem., 70: 241-250, 1976.

37. Sambrcok, J., Fritsch, E. F., and Maniatis, T. Molecular Cloning: ALaboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Labo-

ratory, 1989.