JOURNAL OF CELLULAR PHYSIOLOGY 143524428 (1990)

Effect of Transforming Growth Factor Beta on Proliferation of 16 and Embryonic Porcine Myogenic Cells MARY S. PAMPUSCH, J O A N R. HEMBREE, MARCIA R. HATHAWAY, AND WILLIAM R. DAYTON* Department of Animal Science, University of Minnesota, St. Paul, Minnesota 55 108

We have examined the effect of Transforming Growth Factor (TGF) beta on proliferation of L6 and embryonic porcine myogenic cells. Proliferation of L6 cells was suppressed by both TCF beta-1 and TGF beta-2 in a dose-dependent manner. Half-maximal suppression of proliferation occurred at .036 ng TGF beta-1/ml and .06 ng TGF beta-Uml. Maximal inhibition (60% suppression of proliferation for TGF beta-1 and 52% for TGF beta-2) occurred between .1 and .3 ng/ml for each growth factor. Suppression of proliferation was completely abolished in the presence of an anti-TGF beta antibody that inhibited the biological activity of TGF beta-1 and TCF beta-2. When we evaluated the effect of TGF beta-1 on proliferation of embryonic porcine myogenic cells we obtained results which were very similar to those obtained for L6 cells. Insulin-like growth factor (11%-I stimulated proliferation of L6 cells in a dose-dependent manner in serum-free,defined medium. However as little as .02 ng TGF beta-l/ml detectably suppressed this stimulation and .3 ng TGF beta-l/ml caused a 60% reduction in cell number in cultures treated with 30 ng IGF-Vml. Thus TGF beta-1 significantly suppressed IGF-I-stimulated proliferation of L6 cells.

Members of the Transforming Growth Factor beta TGF beta-1 on proliferation of L6 cells as compared to (TGF beta) family significantly inhibit proliferation its effect on primary neonatal muscle or satellite cells. and differentiation of numerous types of cultured cells To test this hypothesis, we have compared the effects of including r a t satellite cells and neonatal rat myogenic porcine TGF beta-1 and beta-2 on proliferation of L6 cells (Allen and Boxhorn, 1987, 1989). In contrast, al- myogenic cells and embryonic porcine myogenic cells though TGF beta-l inhibits differentiation of the L6 in parallel assays. In contrast to earlier reports (Masmyogenic cell line, it has been reported to have little o r sague et al., 1986; Florini and Ewton, 1988; Florini and no effect on proliferation of these cells (Massague et al., Magri, 19891, our data show that TGF beta-1 and beta1986; Florini and Ewton, 1988; Florini and Magri, 2 significantly inhibit L6 cell proliferation a t concen1989). Based on the reported inability of TGF beta-1 to trations comparable to those required to inhibit embryinhibit proliferation of L6 cells, it has been suggested onic myogenic cells. Additionally, the magnitude of that the mechanisms regulating their proliferation inhibition of L6 cell proliferation is comparable to that may be fundamentally different than those regulating observed for primary myogenic cultures. proliferation of satellite cells and neonatal myogenic MATERIALS AND METHODS cells (Allen and Boxhorn, 1987). Even though L6 cells have been used extensively to study the effects of speMaterials cific growth factors and growth factor receptors on proliferation and differentiation of muscle (Beguinot et al., TGF beta-1 and TGF beta-2, derived from porcine 1985; Ewton et al., 1987; White et al., 1988; Kayalar platelets, and the TGF beta neutralizing antibody were and Wong, 1989; McCusker et al., 1989; Rahm et al., obtained from R&D Systems Inc. (Minneapolis, MN). 1989; Walker et al., 1989), this proposal raises poten- Recombinant IGF-I was obtained from AmGen (Thoutially important questions concerning their validity for sand Oaks, CA). Dulbecco's Modified Eagle Medium use in studying the effects of TGF beta on myogenic cell (DMEM), dexamethasone, Deutsch fetuin, antibioticantimycotic, and trypsin were obtained from Gibco proliferation. The protocols used to assay the effect of TGF beta-1 (Grand Island, NY). Ham's F12 medium was from on proliferation of L6 cells were significantly different Sigma Chemical Company (St. Louis, MO). Basement than those used to assay its effect on proliferation of satellite cells and neonatal myogenic cells. It appeared to us that this difference in assay procedure, not a fundamental difference in response to TGF beta-1, could Received October 23, 1989; accepted January 30, 1990. have been responsible for the relatively small effect of *To whom reprint requestdcorrespondence should be addressed. 0 1990 WILEY-LISS, INC.

TGF BETA INHIBITS L6 MYOGENIC CELL PROLIFERATION

525

Membrane Matrigel and Dispase were from CollabHigh density L6 proliferation a s s a y orative Research (Bedford, MA). Bovine serum albuAssays done essentially a s described by Florini min (BSA) was from ICN Immunobiologicals (Lisle, and Ewtonwere (1988).L6 cells were plated in 8.6 cm2 mulIL). Fetal bovine serum (FBS) was obtained from Flow tiwell plates a t 12,500 cells/cm2 in DMEM containing Laboratories Inc. (McLean, VA). Swine serum ( S S )was 10% FBS. Cells were allowed to attach for 24 hours and collected in our laboratory as previously described then washed twice with DMEM prior to addition of test (Kotts et al., 1987b) and chick embryo extract (EE) medium. Test medium consisted of DMEM containing was prepared as described by Bischoff (1974). L6 cells 4% swine serum and TGF beta-1. In control medium were subcloned from the L6 cell line (White et al., TGF beta-1 was replaced with a n equivalent volume of 1988). the buffer in which the TGF beta-1 was dissolved (4 mM HC1, 1 mg/ml BSA). After a n additional 24 hour incubation cultures were trypsinized and released cells Low density L6 proliferation assays were counted by using a Coulter Counter. The culture methods and proliferation assay utilized for most of the experiments in this study have been Porcine p rima r y cultures described and validated by Kotts et al. (1987a,b). Myogenic cells were isolated from the hind limb musBriefly, L6 cells were plated in 25 cm2 flasks at 400 cells/cm2 in DMEM containing 10% FBS. Cultures cle of porcine fetuses (50 to 55 days gestation) by using were incubated for 24 hours in this medium to allow a procedure similar to th a t used to isolate adult rat attachment. Following this attachment period, cul- myogenic cells (Bischoff, 1974). All media contained tures were washed with DMEM and test medium was antibiotic/antimycotic unless otherwise indicated. applied. The composition of the test medium varied de- Hind limb muscles were aseptically excised, washed in pending upon the particular experiment. In experi- warm (37°C) Earle's Balanced Salt Solution (EBSS), ments designed to measure the inhibitory effect of TGF minced with scissors, and then digested with 10 volbeta on serum-stimulated proliferation, test media con- umes (volume/weight of minced muscle) of 0.2% (w/v) tained DMEM, 4% swine serum, and either TGF beta-1 trypsin in Ca-Mg free EBSS (no antibiotic/antimycotic or TGF beta-2. In control cultures TGF was replaced added) for 1 hour a t 37°C with frequent vortexing. Afwith a n equivalent volume of the buffer in which the ter incubation the digested tissue was pelleted by cenTGF was dissolved (4 mM HC1 containing 1 mgiml bo- trifugation a t 400g for 4 min a t 25°C. The pellet was vine serum albumin). In experiments designed to mea- washed by suspension in 10 volumes of EBSS and censure the effect of TGF on IGF-I-stimulated prolifera- trifugation was repeated. The pellet was suspended in 10 ml EBSS. The suspension was triturated 6 to 7 times tion, test medium contained Ham's F12 medium, lo-' M dexamethasone, .5 mg/ml Deutsch Fetuin, .5 mg/ml with a 10 ml pipet and then centrifuged at 1,400g for 4 BSA, and varying amounts of IGF-I and/or TGF beta-1 min. The resultant pellet was suspended in DMEM as indicated in figure legends. Control cultures con- containing 10% FBS to give .4 g original tissue weight tained appropriate amounts of the buffers in which per ml medium and was filtered sequentially through TGF andlor IGF-I was dissolved (4 mM HC1 containing 149 pm and 74 pm mesh Nitex cloth to remove connec1 mg/ml BSA or 100 mM acetic acid, respectively). Af- tive tissue and large cell clumps. The filtrate was diter addition of the appropriate test media, cultures luted with 2 volumes of DMEM containing 10% FBS were incubated for 72 hours and trypsinized, and re- and preplated on 75 cm2 t-flasks (10 ml per flask) for 1 leased cells were counted by using a Coulter Counter hour at 37"C, 5% CO,, 95% air in a water saturated (Model ZB). It should be noted th at these assays used environment. Unattached cells were removed from the MM-1 defined media (Florini and Roberts, 1979) con- flasks, pelleted by centrifugation a t 400g for 10 min, taining .5% (w/v) BSA. Cultures grown in DMEM alone and then suspended in DMEM containing 10% FBS or in Ham's F12 plus BSA did not proliferate during and 10% (viv) DMSO. To minimize clumping of cells, the 72 hour incubation period used to measure the ef- the cell suspension was triturated with a 10 ml pipe fect of IGF-I on proliferation. In MM-1 plus BSA, the and with a n 18 gauge needle fitted on a 35 cc syringe. L6 cells continued to proliferate indicating that they Aliquots were placed in Nunc vials and frozen a t -50°C remained viable (Fig. 4). When IGF-I was added to this in a Styrofoam box (1 inch sides) for 24 hours before medium, cells responded in a manner similar to that being transferred to a liquid nitrogen storage tank. reported by other researchers (Fig. 4) (Ewton et al., Effect of TGF beta-1 on proliferation of porcine 1987). BSA was added to the MM-1 defined medium in myogenic cells order to reduce potential adsorption of the growth factors to the tissue culture flasks and had no effect on Cultures were established by rapidly thawing frozen proliferation (data not shown). cells and diluting them in DMEM containing 10% FBS In the antibody neutralization experiments, test me- and 3% embryo extract (EE) to give .03 g original tisdium containing DMEM, 4% SS, anti-TGF antibody (4 sue weight per ml. The cell suspension (2 ml/dish) was pglml), and either TGF beta-1 or TGF beta-2 (.08 ng/ placed in 35 mm dishes precoated with Basement Memml) was incubated for 1hour at 37°C before addition to brane Matrigel (diluted 1 : l O in DMEM) and cells were cultures. Control medium containing TGF beta-1 or allowed to attach for 24 hours. All cultures were mainTGF beta-2 and the buffer in which the antibody was tained at 37"C, 5% COz, 95% air in a water saturated dissolved (137 mM NaC1,27 mM KC1,8 mM Na2HP0,, environment. After the attachment period, cultures pH 7.5) was also incubated and added to control cul- were refed with 2 ml DMEM containing 10%FBS, 3% tures. Inhibition of proliferation was assayed a s de- EE, and TGF beta-1. Cultures were incubated for 72 scribed above. hours and then individual cells were released from the

526

PAMPUSCH ET AL.

fin I

1

20

I

1

J 0

0.1

1

5

0.10

0.20

TGF BETA-1 (ng I ml) Fig. 1. Effect of TGF beta-1 on proliferation of L6 cells in the high density assay. Cells were plated at 12,500/cm2 and allowed to attach for 24 hours before addition of test medium containing DMEM, 4% swine serum, and the indicated amount of TGF beta-1. After an additional 24 hour incubation, cells were removed from the plate by trypsinization and counted by using a Coulter counter. Pooled data from two assays are shown.

plate by treating with Dispase (1 ml/plate) for 15 min at 37°C. Released cells were counted by using a Coulter counter.

0.30

0.40

0.50

TGF BETA (ng I ml) Fig. 2. Effect of TGF beta-1 (W and TGF beta-2 ( 0 )on proliferation of L6 cells in the low density proliferation assay. Cells were plated a t 400/cm2 and allowed to attach for 24 hours before addition of test medium containing DMEM, 4% swine serum, and the indicated amount of either TGF beta-I or TGF beta-2. After an additional 72 hour incubation, cells were removed from the flask by trypsinization and counted by using a Coulter counter. * indicates the number of cells/cm2 in cultures treated with 0.08 ngiml of either TGF beta-1 or TGF beta-2 and 4 pgiml of the anti-TGF beta antibody. Pooled data from three assays are shown. Standard error bars that are not shown are contained within the symbol for the data point.

RESULTS

Effect of TGF beta-1 and beta-2 on proliferation of L6 cells When L6 cells were plated at a high density (12,500 cells/cm2) and treated with TGF beta-1 for 24 hours, proliferation was only slightly depressed (Fig. 1). At .1 ng TGF beta-llml medium cell number was only 7.5% lower than in control cultures and even at very high concentrations of TGF ( 5 ng/ml) there was only a 22% reduction in cell number (Fig. 1). These results are consistent with those reported by others (Florini and Ewton, 1988) and appear to confirm reports that TGF beta-1 has a limited effect on proliferation of L6 cells. In contrast, as little as .08 ng TGF beta-l/ml medium depressed L6 muscle cell proliferation by 50% in the low density proliferation assay (Fig. 2). TGF beta-2 also inhibited proliferation of L6 cells although it appeared to be somewhat less effective than TGF beta-1 (Fig. 2). Inhibition of cell proliferation was half maximal at .036 ng TGF beta-l/ml medium and .06 ng TGF beta-2/ml medium while maximal inhibition was achieved at .1 to .3 ng of either TGF beta-1 or beta-2lml medium. Inhibition of proliferation in cultures treated with either TGF beta-1 or beta-2 was completely blocked by an anti-TGF beta antibody and, therefore, appeared to be a specific response to TGF beta treatment (Fig. 2). Effect of TGF beta-1 on the proliferation of primary porcine myogenic cells in culture Treatment of porcine myogenic cultures with porcine TGF beta-1 resulted in decreased proliferation (Fig. 3) with maximal inhibition being reached at .1 to .3 ng TGF beta-l/ml medium. Consequently, the TGF beta-1 dose-response curves for L6 and primary porcine myogenic cells appear to be very similar (Figs. 2, 3).

0 1 " 0

"

"

"

'

"

2

0.6 0.8 1.o 1.2 TGF BETA-1 (ng / rnl) Fig. 3. Effect of TGF beta-1 on proliferation of embryonic porcine myogenic cells. Cells were allowed to attach for 24 hours before addition of test medium containing DMEM, 10% FBS, 3% embryo extract, and the indicated amount of TGF beta-1. Cultures were incubated for an additional 72 hours and removed from the dish with Dispase. The points shown represent the mean i_ SE of three cultures per treatment. Similar results have been obtained in four separate experiments. 0.2

0.4

Effect of TGF beta-1 o n IGF-I-stimulated proliferation of L6 cells Having established that TGF beta-1 and beta-2 inhibited serum-stimulated proliferation of L6 cells, we next examined the effect of TGF beta-1 on IGF-I-stimulated proliferation in serum-free, defined medium. Figure 4 shows the effect of specific concentrations of IGF-I (0-30 nglml medium) on proliferation of L6 cells in the presence of various concentrations of TGF beta1. In the absence of TGF beta-1, L6 cells showed a concentration-dependent increase in proliferation in response to IGF-I. Addition of increasing concentrations of TGF gradually decreased the IGF-I response

TGF BETA INHLBITS L6 MYOGENIC CELL PROLIFERATION

527

and Allen for cultures of rat neonatal myogenic cells and satellite cells (Allen and Boxhorn, 1987, 1989). C m The assay procedures utilized by Boxhorn and Allen u) a were similar to those used in our study in that TGF 0 beta-1 was applied to relatively low density, exponen5 tially proliferating cultures for 3 days prior to deterN mining cell number. In contrast previous studies showE 0 ing little or no effect of TGF beta-1 on L6 cells have generally utilized assays employing either high denIA? sity cultures and a shorter treatment period (24 hours) W (Florini et al., 1986; Florini and Ewton, 1988)or fusion0 0 4 8 12 16 20 24 28 32 promoting medium (Massague et al., 1986). The exception to this statement is a report that TGF beta-1 only IGF- 1 (ng I ml ) slightly inhibits DNA accumulation in exponentially Fig. 4. Effect of TGF beta-1 on IGF-I-stimulated proliferation of L6 growing L6 cells (Massague et al., 1986). Unfortumyogenic cells. Cells were plated a t 400/cm2 and allowed to attach for nately, because neither specific data nor a detailed de24 hours before addition of test medium containing Ham's F12 meM dexarnethasone, .5 mgiml Deutsch fetuin, .5 mgiml scription of the assay in which the data were obtained dium, BSA, and varying amounts of IGF-I and/or TGF beta-1. TGF beta-1 is provided in this report, it is not possible to compare was added at 0 (w), .02 (A), .08 (*), or .3 ( 0 )ngiml medium. After an these data with our findings. additional 72 hour incubation, cells were released from the flask by It has been reported that TGF beta-1 has little effect trypsinization and released cells were counted by using a Coulter on stimulation of proliferation or inhibition of proteolcounter. indicates the number of cells/cm2 present at the time of test medium addition to the cultures. Pooled data from two assays are ysis by IGF-I in L6 cells and this finding has led to shown. Standard error bars that are not shown are contained within suggestions that TGF beta-1 may be a useful probe for the symbol for the data point. separately examining the effects of IGF-I on proliferation and differentiation in these cells (Florini and Ewton, 1988). Utilizing the assay system described in this study, however, we have found that even low concenuntil, a t .3 ng TGFlml medium, very little stimulation trations of TGF beta-1 very significantly inhibit IFGof proliferation was observed even in the presence of 30 I-stimulated proliferation of L6 myoblasts (Fig. 4). Beng IGF-I/ml medium. cause i t has been well established that TGF beta-1 does not interfere with binding of IGF-I to its receptor, we must assume, as suggested by Florini and Ewton DISCUSSION (1988), that TGF beta-1 interferes with IGF-I action at Previous studies have shown that TGF beta-1 only some point beyond the ligand-receptor interaction. slightly inhibits proliferation of L6 myogenic cells (Flo- However, while these workers suggest that this interrini et al., 1986; Massague et al., 1986; Florini and ference occurs a t a point which affects only differentiEwton, 1988; Florini and Magri, 1989). In contrast to ation-related effects of IGF-I, our data suggest that these reports, our current data show that both TGF stimulation of proliferation is also affected. Thus, it is beta-1 and TGF beta-2, a t concentrations below .1 ngt unlikely that TGF beta-1 can be used as a probe to ml, greatly suppressed proliferation of L6 myogenic separate effects of IGF-I on proliferation and differencells. Because this inhibition is completely abolished tiation of L6 cells. by a n antibody that inhibits the biological activity of Our results suggest that the previously reported inTGF beta-1 and TGF beta-2, i t appears to be a specific ability of TGF beta-1 to inhibit proliferation of L6 cells effect of TGF. Additionally, when we utilized similar as compared to neonatal muscle cells or satellite cells assay protocols to compare the inhibitory activity of reflects differences in the assay protocols used in varTGF beta-1 on L6 cells and on embryonic myogenic ious laboratories rather than a true difference in recells, we found that both the concentration of TGF sponse to TGF beta-1. Consequently, L6 cells appear to beta-1 needed to cause inhibition and the magnitude of be a valid system in which to examine the effects of inhibition were similar for both cell types. TGF beta-1 or beta-2 on proliferation of pure cultures It is possible that proliferation of the L6 subclone of myogenic cells. These findings also illustrate that utilized in our studies was more sensitive to TGF beta- valid comparisons of the effects of growth factors on 1 than was proliferation of subclones utilized in other cultured cells can only be made by using similar, if not laboratories. This is unlikely, however, because prolif- identical, assay protocols. Comparison of results oberation of our L6 subclone is only slightly inhibited by tained in different laboratories and utilizing different TGF beta-1 in the high density proliferation assay (Fig. assay protocols can lead to misleading conclusions and 1) and these results are similar to those reported by should be made with caution. others using this assay procedure with other L6 subclones (Florini and Ewton, 1988). Consequently, it apACKNOWLEDGMENTS pears that the assay procedure, not differences in the sensitivity of various L6 subclones to TGF beta-1, may Published as paper No. 18061 of the Scientific Jourbe responsible for discrepancies between our results nal Series of the Minnesota Agricultural Experiment and those obtained in other laboratories. This hypoth- Station on research conducted under Minnesota Experesis is supported by our finding that the effect of TGF iment Station Project No. 4816-83. This work was SUDbeta-1 on proliferation of both L6 cells and embryonic ported by USDA kompetitive Research Grant No. 87myogenic cells was similar to that reported by Boxhorn CRCR-1-2531. '0

-

.

528

PAMPUSCH ET AL

LITERATURE CITED Allen, R.E., and Boxhorn, L.K. (1987) Inhibition of skeletal muscle satellite cell differentiation by transforming growth factor-beta. J. Cell. Physiol., 1 3 3 5 6 7 4 7 2 , Allen, R.E., and Boxhorn, L.K. (1989) Regulation of skeletal muscle satellite cell proliferation and differentiation by transforming growth factor-beta, insulin-like growth factor I, and fibroblast growth factor. J. Cell. Physiol., 1381311-315. Beguinot, F., Kahn, C.R., Moses, A.C., andSMith, R.J. (1985)Distinct biologically active receptors for insulin, insulin-like growth factor I, and insulin-like growth factor I1 in cultured skeletal muscle cells. J. Biol. Chem., 260:15892-15898. Bischoff, R. (1974) Enzymatic liberation of myogenic cells from adult rat muscle. Anat. Rec., 180:645-662. Ewton, D.Z., Falen, S.L., and Florini, J.R. (1987) The type I1 insulinlike growth factor (IGF) receptor has low affinity for IGF-I analogs: Pleiotypic actions of IGFs on myoblasts are apparently mediated by the type I receptor. Endocrinology, 120:115-123. Florini, J.R., and Ewton, D.Z. (1988) Actions of transforming growth factor-p on muscle cells. J. Cell. Physiol., 135t301-308. Florini, J.R., and Magri, K.A. (1989) Effects of growth factors on myogenic differentiation. Am. J . Physiol., 256tC701-C711. Florini, J.R., Roberts, A.B., Ewton, D.Z., Falen, S.L., Flanders, K.C., and Sporn, M.B. (1986) Transforming growth factor-p: A very potent inhibitor of myoblast differentiation, identical to the differentiation inhibitor secreted by buffalo rat liver cells. J. Biol. Chem., 261:16509-16513. Florini, J.R., and Roberts, S.B. (1979) A serum-free medium for the growth of muscle cells in culture. In Vitro, 15r983-992.

Kayalar, C., and Wong, W.T. (1989) Metalloendoprotease inhibitors which block the differentiation of L, myoblasts inhibit insulin degradation by the endogenous insulin-degrading enzyme. J. Biol. Chem., 264:8928-8934. Kotts, C.E., White, M.E., Martin, F., Allen, C.E., and Dayton, W.R. (1987a) A statistically standardized muscle cell culture bioassay measuring the effect of swine serum on muscle cell proliferation. J. Anim. Sci., 64:615-622. Kotts, C.E., White, M.E., Allen, C.E., and Dayton, W.R. (198713)Stimulation of in vitro muscle cell proliferation by sera from swine injected with porcine growth hormone. J . Anim. Sci., 641623. Massague, J., Cheifetz, S., Endo, T., and Nadal-Ginard, B. (1986)Type p transforming growth factor is an inhibitor of myogenic differentiation. Proc. Natl. Acad. Sci. USA, 83:8206-8210. McCusker, R.H., Camacho-Hubner, C., and Clemmons, D.R. (1989) Identification of the types of insulin-like growth factor-binding proteins that are secreted by muscle cells in uitro. J. Biol. Chem., 264: 7795-7800. Rahm, M., Jin, P., Siimegi, J., and Sejersen, T. (1989) Elevated c-,fos expression inhibits differentiation of L6 rat myoblasts. J . Cell. Physiol., 139:237-244. Walker, P.S., Ramlal, T., Donovan, J.A., Doering, T.P., Sandra, A,, Klip, A., and Pessin, J.E. (1989) Insulin and glucose-dependent regulation of the glucose transport system in the rat L6 skeletal muscle cell line. J. Biol. Chem., 26416587-6595, White, M.E., Allen, C.E., and Dayton, W.R. (1988)Effect of sera from fed and fasted pigs on proliferation and protein turnover in cultured myogenic cells. J. Anim. Sci., 66.34-40,