- Email: [email protected]
Prostaglandin-dependent phosphatidylinositol signaling during embryonic chick myogenesis
Cell Differentiation and Development, Elsevier Scientific Publishers Ireland.
CELDIF
109
32 (1990) 109-116 Ltd.
00700
Prostaglandin-dependent phosphatidylinositol signaling during embryonic chick myogenesis H. Elgendy and R.E. Hausman Biological Science Center, Boston University, Boston, MA, U.S.A. (Accepted
16 August
1990)
Previous investigations suggested that binding of prostaglandin to a myoblast membrane receptor initiates a second messenger cascade which is essential for subsequent myogenesis. Initial evidence of the sensitivity of myogenesis to lithium suggested the involvement of inositol phosphate metabolism. That possibility is investigated here. The accumulation of inositol monophosphate in response to prostaghmdin binding was studied in aggregate cultures of chick embryo myoblasts in vitro. At 22 or 28 h in culture mononucleated myoblasts were labeled with [‘H]inositol, which was then incorporated into phosphoinositides. After experimental manipulations of prostaglandin metabolism and the addition of Li + prior to prostaglandin binding at 33 h, [ 3H]inositol monophosphate accumulation was measured by anion-exchange chromatography between 33 and 37 h. Inositol monophosphate was found to accumulate rapidly following 33 h. However, after 36 h of myogenesis, no inositol monophosphate accumulation was observed. The accumulation was dependent on prostaghmdin as indomethacin, which also blocks subsequent membrane events in myogenesis, blocked inositol phosphate accumulation. Like subsequent myogenesis, inositol phosphate accumulation was restored by the addition of exogenous prostaglandin. Finally, the accumulation of inositol phosphate began only after the binding of prostaghmdin. The results demonstrate that an inositol phosphate signal transduction mechanism connects prostaghmdin binding to membrane events in embryonic chick myogenesis. Cell adhesion; Myoblast fusion; Signal transduction; Receptor-mediated
Introduction Cell differentiation and expression of musclespecific proteins during embryonic myogenesis in
Abbreviations: taglandin.
PIP,,
polyphosphatidylinositol;
Correspondence address: ology, Boston University, 02215, U.S.A.
0922-3371/90/$03.50
R.E. Hausman, 5 Cummington
PG,
pros-
Department of BiStreet, Boston, MA
0 1990 Elsevier Scientific
Publishers
Ireland,
signaling
culture can be blocked or manipulated by a variety of environmental modifications (Hausman et al., 1986). These observations strongly suggest the existence of cell-cell signaling mechanisms to coordinate expression of the myogenic phenotype between the individual mononucleated cells. The cell membrane events termed fusion involve, at least, two steps: a characteristic increase in myoblast to myoblast adhesion and the actual fusion of the bilayers which may occur up to several hours later in culture (Knudsen and Horwitz, 1977, 1978). Ltd.
110
The increased adhesion occurs just after the expression of prostaglandin (PG) binding activity by the myoblasts and is sensitive to inhibitors of PG synthesis in the cultures (Hausman and Velleman, 1981; Hausman et al., 1986). Accompanying this binding of PG to the myoblasts, is a change in membrane organization (Santini et al., 1987) and the activity of a pertussis toxin-sensitive G protein (Hausman et al., 1990). Interference with either PG synthesis or G protein activity in the myoblasts at this critical time blocks the increase in cell-to-cell adhesion, the fusion of the bilayers and the characteristic decrease in myoblast membrane conductivity (Bonincontro et al., 1987; Santini et al., 1988; Hausman et al., 1989, 1990). These results suggested that PG, synthesized by the myoblasts, might be acting as an autocrine signal to coordinate changes in the cell membrane necessary for myogenesis (Hausman et al., 1986). Such a differentiation-staging signal would be able to affect cell surface events through a G protein without the intervention of a soluble message (Brown et al., 1988, 1989). However, if this same signal were to stage events at the level of the genome, a soluble second messenger pathway would seem to be necessary. Wakelam (1986) showed the existence of a polyphosphatidylinosito1 (PIP,) signal system in myoblasts, and we have shown indirectly by lithium sensitivity that turnover of PIP, occurs at the time of PG binding (Hausman et al., 1989). Here we demonstrate that binding of PG to embryonic chick myoblasts induces a rapid increase in inositol phosphate metabolism. This inositol phosphate flux occurs only under conditions which allow the increase in myoblast adhesion and the subsequent membrane events in myogenesis. Furthermore, this flux in inositol phosphate begins within 5 min after the addition of PG to the myoblasts. This suggests that the signal transduction process proceeds directly from the activated PG receptor through a G protein and a presumed phospholipase to cause the inositol phosphate flux. However, our present results cannot rule out a very limited number of rapidly occurring intervening steps.
Materials and Methods Materials
Fertile chicken eggs were obtained from Hardy’s (Essex, MA). Trypsin was obtained from Difco or Sigma, Dulbecco’s modified Eagle’s medium and penicillin/ streptomycin from Gibco. All other chemicals were obtained from Sigma, as previously (Hausman et al., 1986). [3H]Inositol, and [ ‘Hlinositol l-phosphate (inositol monophosphate) were obtained from American Radiolabeled Chemicals, Inc. (St. Louis, MO, U.S.A.), and [3H]PGE, was from New England Nuclear. Myoblast cell cultures
Tissue was prepared from pectoral and thigh muscles of ll-day embryonic chicks, and primary aggregate cultures enriched for myoblasts were established according to standard procedures (Hausman et al., 1986). Phosphoinositide labeling and hydrolysis
Myoblast aggregates at 22, 24 or 28 h of culture were exposed to [3H]inositol (5 pCi/ml) for 6-8 h. Longer periods of exposure did not affect the inositol phosphate accumulation measurements. At the end of the incubation period, the cultures were washed five times over a period of 60 min with fresh DME medium. The cultures then resuspended in fresh culture medium with or without 30 mM Li+, and the incubation continued for the time periods indicated in the figure legends. The incubations were arrested at the appropriate times by rapid cooling to 0°C removing the medium and adding 1 ml of cold 10% trichloroacetic acid. Acid extraction and ion-exchange chromatography of the inositol monophosphates were performed as described by Adamo et al. (1989). The water-soluble components of each sample were poured onto columns containing 1 ml of Dowex 1 (X8-200 resin; formate form) (Sigma) and eluted with 0.1 M Formic acid/O.2 M ammonium formate as described in Adamo et al. (1989). Details of individual experiments are given in the figure legends.
111
PGE, binding
Total and non-specific binding of [3H]PGE1 to myoblast aggregates was determined as previously (Hausman and Velleman, 1981; Hausman et al., 1986). Detection of inositol I -phosphate (inositol monophosphate)
We detected inositol monophosphate by anion-exchange chromatography (Berridge et al., 1982). [3H]Inositol monophosphate (5 x lo3 cpm) was added to an otherwise non-labeled myoblast culture and prepared as described above. Eight ml of this solution was applied to a Dowex-1 anionexchange column and eluted with 0.1 M formic acid/O.2 M ammonium formate. All the radioactivity eluted at fraction 10 or 11, precisely the inositol monophosphate fraction reported by Berridge et al. (1982), and recovery of radioactivity from the column was consistently greater than 85% (data not shown).
Results
The accepted demonstration that an inositol monophosphate second messenger system is activated as the result of receptor stimulation is the accumulation of [ 3H]inositol monophosphate in the presence of Li+ in cells which had been prelabeled with [3H]inositol (Berridge et al., 1982). On the basis of evidence in many systems, they suggested that this occurs because Li+ specifically blocks activity of the enzyme inositol-phosphate phosphatase. This causes the accumulation of inositol l-phosphate and the inability to regenerate the PIP, necessary for continued signaling. While we have previously shown that several of the membrane steps in myogenesis were inhibited by addition of Li+ (Bonincontro et al., 1987), this was not a direct demonstration of inositol phosphate accumulation in response to PG binding. Inositol monophosphate accumulation is blocked by Li + When myoblasts were labeled with [ 3H]inositol
from 22-28 h of culture and Li+ was added at 30 h, inositol monophosphate began to accumulate
_
E -
E
aI ;;
L % ,” a E .Z ;
2000 r 1800 1600 t 1400 1200 1000 800 600 -
.G
400
’
200
I
VI
0 L32
33 Time
34
35
in culture
36
37
(hours)
Fig. 1. Inositol monophosphate accumulation in presence or absence of 20 mM Lif. At 22 h of culture myoblast aggregates were exposed to 5 pCi/ml of [3H]inositol for 9 h. The myoblasts were washed and resuspended at 31, 32, 33, 34, 35 and 36 h of culture in fresh culture medium with (hatched bars) or without (open bars) 20 mM Li+. Inositol monophosphate accumulation was tested between 31-32, 32-33, 33-34, 34-35, 35-36 and 36-37 h of culture. Incubations were arrested at the specified times and analyzed by anion-exchange chromatography for [3H]inositol monophosphate as described in Materials and Methods. This figure represents one of two identical experiments.
only after 32 h in culture (Fig. 1). This accumulation of inositol monophosphate was dependent on the presence of Li+ (Fig. 1). This finding was consistent with our earlier results showing that this 32-36-h interval is when myogenesis, as measured by the change in adhesion or bilayer fusion, is also sensitive to Li+ (Hausman et al., 1989). Rescue of the indomethacin block to accumulation by added PGE,
Inositol phosphate accumulated between 32 and 36 h of culture as demonstrated above. This is the same time during which PG binds to the embryonic myoblast aggregates (Hausman and Berggrun, 1987). This raised the possibility that the inositol phosphate metabolism might be dependent on the availability of PG. To test this, we added 30 PM indomethacin (an inhibitor of cyclooxygenase and thus, PG synthesis (Needleman et al., 1986)) to myoblast aggregates at 30 h (in the presence of 20 mM Li+ as above) and measured inositol monophosphate accumulation.
112
7
g
-
al ;; r
3300 3000 2700 2400
R
2100
1~
1800
0
z c
M
3600
1500 1200
:: .g
900
I
300
600 0 ii+
Lif&
indo
Lif, 32 h
indo
& PGE, 37 h
Fig. 2. Time of rescue of indomethacin blockage of inositol monophosphate accumulation by added PGE,. At 24 h of culture the myoblast aggregates were washed and resuspended as four different sets. Sets l-3 were exposed to 5 nCi/ml [3H]inositol for 6 h, washed and resuspended at 30 h in fresh culture medium with 20 mM Li+. At this time, sets 2-4 had 30 uM indomethacin added. Set 4 was also exposed to 5 &i/ml [3H]inositol at this time for 6 h. In set 3, 10 pM PGE, was added at 32 h, 2 h after the addition of indomethacin. In set 4, 20 mM Li+ was added at 35 h and PGE, was added only at 37 h of culture. Incubations were arrested and analyzed by anion exchange chromatography as described in Materials and Methods. This figure represents one of two identical experiments.
This concentration of indomethacin allows the transient binding of exogenously added PG to the myoblasts but blocks subsequent myogenesis (Hausman and Berggrun, 1987). No inositol monophosphate accumulation was seen in the presence of indomethacin. However, if exogenous PGE, (1 PM) was added at 32 h, inositol monophosphate accumulation was restored (Fig. 2). In contrast, when myoblasts were labeled with [3H]inositol from 30-36 h and PGEl was added only at 37 h, no accumulation of inositol was seen (Fig. 2). These results demonstrated that phosphatidylinositol second messenger activity was dependent on PG metabolism during this stage of myogenesis in culture. Because PG binds to a myoblast PG receptor during this stage of myogenesis in culture (Velleman and Hausman, 1981; Hausman et al., 1986; Hausman and Berggrun, 1987), we have suggested that the requirement might actually be for PG binding (Santini et al., 1988).
Inositol monophosphate accumulation after PGE, binding at 32 h Here we asked two questions about the temporal relationship between PG binding and inositol phosphate metabolism. First, whether the accumulation of inositol monophosphate occurred before or after the binding of PG to PG receptors. Second, if there was a delay long enough to suggest that multiple steps or new macromolecular syntheses were involved. Therefore myoblasts were prelabeled with [3H]inositol from 22-28 h of culture, and 30 PM indomethacin was added at 30 h to block PG synthesis. PGE, and 20 mM Li+ were added at 32 h, and inositol monophosphate accumulation was tested at 2, 5, 10, 20 and 30 min after PG binding to its receptor at 32 h. The time of PG binding was assessed in parallel cultures
m-
OCTiTime in culture
c*l (hours)
Fig. 3. Determination of inositol monophosphate accumulation and PGE, binding. Open circles represent inositol monophosphate accumulation with the scale on the left. Open squares represent specific PGE, binding with the scale on the right. Myoblast aggregates were divided into two sets at 22 h of culture. One set was exposed to 5 pCi/ml of [3H]inositol for 6 h. These myoblasts were washed and resuspended at 28 h of culture in fresh culture medium with 30 pM indomethacin in five different flasks. At 32 h of culture 20 mM Lif and 10 aM PGE, were added to each of the five flasks. Incubations were arrested after 5, 10, 20, 40 and 60 min after PGE, addition (32 h). Samples were analyzed by anion-exchange chromatography as described in Materials and Methods. The other set of myoblast aggregates was left to differentiate in control medium until 28 h when they were resuspended in medium with 30 PM indomethacin in twelve different flasks. At 30 h all 12 flasks had unlabeled PGE, added to 10 pM, six flasks had [‘H]PGE, added to 10 nM. One flask from each of the two was analyzed at 30, 32, 33 and 36 h and specific binding determined by subtracting nonspecific binding from total binding (Hausman and Velleman, 1981). This figure represents one of two identical experiments.
113
with [ 3H]PGE,. Inositol monophosphate accumulation was first seen 5 min after addition of PGE,, and the accumulation continued to increase during the period of PG receptor activation (32-36 h) (Fig. 3).
Discussion
Previous work, using indomethacin to block PG synthesis (Hausman et al., 1990), had shown that binding of PG to the myoblasts is required for the characteristic increase in cell-cell adhesion. However, we did not know whether there was any further requirement for PG after the disappearance of the transient PG receptor at about 35 h. In this study we tested this by adding indomethacin to myoblasts at 30 h but delaying addition of PGE, until 37 h. We showed that myogenesis was not rescued under these conditions. Thus, we conclude that PG is required for cell signaling purposes only between 30 and 36 h in culture, an interval when myogenesis is also sensitive to Li+ (Bonincontro et al., 1987) and to pertussis toxin (Hausman et al., 1990). Perhaps this is not surprising, since the immediate consequence of this signaling event, the increase in myoblast adhesion, has already begun by 36 h of culture. Our present finding that, when added after 37 h of differentiation in culture, exogenous PG no longer results in an inositol phosphate signal is consistent with that hypothesis and further suggests that the signaling cascade at this time during myoblast differentiation in culture is critical. We cannot say whether one or more of the signal pathway components, such as the PG receptor (Hausman and Berggrun, 1987), disappears or becomes inactive. Previous evidence did not allow us to rule out alternative signaling sequences as being causal at this time during embryonic chick myogenesis. Ligands other than PG have been suggested for myogenic signaling (Wakelam 1986; Entwistle et al., 1988a; 1988b; Sorokin et al., 1988; Kelvin et al., 1989). The sensitivity of the signal transduction at 33-35 h of myogenesis to indomethacin demonstrates a need for PG metabolism at this time but not that PG initiates the signal by bind-
ing to a receptor. In some cells PG metabolism is necessary for signal transduction, but it is involved after the initial receptor-mediated event (Jeremy et al., 1988; Wang et al., 1988; Chaudhry et al., 1989; Kurachi et al., 1989; Murphy and Welk, 1989). The necessity for PG metabolism to precede the inositol phosphate flux is directly demonstrated here. Inositol monophosphate accumulation takes place when PG binds to its receptor and is absolutely dependent on PG being available. However, it does not matter whether PG is synthesized by the myoblasts (Hausman et al., 1986) or provided exogenously. Thus, the inositol phosphate flux is clearly downstream from the need for PG in the signal cascade. This conclusion is supported by the evidence that the inositol phosphate flux is detectable only after PG binds to its receptor and is apparent within 10 min. Given the asynchrony of the system (Hausman et al., 1987; Hausman and Berggrun, 1987) and the likelihood that the small initial amounts of inositol phosphate go undetected, this is evidence that it is the binding of PG to its receptor which directly stimulates the G protein (Hausman et al., 1990) and causes the breakdown of PIP,. PG receptor-stimulated signal transduction is well established in other cell types (Watanabe et al., 1988; Hatanaka et al., 1989; Laychock, 1989; Negishi et al., 1989; Okumura et al., 1989). Our present results clearly rule out the need for multiple steps and new macromolecular synthesis in the cascade of events which must occur between PG binding and G protein activity. Therefore, we suggest that the sequence of events which signals the time to increase myoblast cell adhesion amongst the population of cells in culture, and perhaps in the developing muscle, is that described above. The evidence obtained with indomethacin (Hausman et al., 1986) and pertussis toxin (Hausman et al., 1990) shows that PG metabolism and G protein activity are absolutely necessary for the increase in myoblast adhesion, the fusion of the bilayers and likely the later decrease in membrane conductivity (Bonincontro et al., 1987). The absolute need for PIP, breakdown for the increase in myoblast adhesion is less clear. Li+ (which blocks
114
inositol-phosphate phosphatase and hence the necessary regeneration of inositol) does not always block the change in adhesion and the decrease in conductivity but often only delays them (Hausman et al., 1989). As we have noted (Hausman et al., 1989), there are two obvious explanations for this. The increase in myoblast adhesion, bilayer fusion and the decrease in conductivity are all membrane-localized events, and some G proteins interact directly with membrane effecters (Brown et al., 1988; 1989). However, since Li+ delays these myoblast membrane events, the metabolism of inositol phosphate must play some role in their regulation. By similar reasoning, the membrane events may be also staged by the additional signaling pathway initiated by PIP, breakdown, that going through diacylglycerol. The diacylglycerol and inositol phosphate pathways are, for many cellular responses, known to be synergistic. A connection between the inositol phosphate signal initiated by PG binding and genomic events during embryonic chick myogenesis is, at present, speculative. However, there is evidence that a similar receptor-stimulated system acts during Dictyostelium development to coordinate the differentiation of the separate cells (Darcy and Fisher, 1989; Ginsburg and Kimmel, 1989).
Acknowledgement
This work was supported in part by funds from the Graduate School of Boston University.
References Adamo, S., C. Caporale, C. Nervi, R. Ceci and M. Molinaro: Activity and regulation of calcium-, phospholipid-dependent protein kinase in differentiating chick myogenic cells. J. Cell Biol. 108, 153-158 (1989). Berridge, M.J., C.P. Downes and M.R. Hanley: Lithium amplifies agonist-dependent phosphatidylinositol responses in brain and salivary glands. Biochem. J. 212, 473-482 (1982). Berridge, M.J., C.P. Downes and M.R. Hanley: Neural and developmental actions of lithium: a unifying hypothesis. Cell 59, 411-419 (1989). Bonincontro, A., C. Cametti, R.E. Hausman, P.L. Indovina and M.T. Santini: Changes in myoblast membrane electri-
cal properties during cell-cell adhesion and fusion in vitro. Biochim. Biophys. Acta 903, 89-95 (1987). Brown, A.M., A. Yatani, Y. Imoto, G. Kirsch, H. Hamm, J. Codina, R. Mattera and L. Bimbaumer: Direct coupling of G proteins to ionic channels. Cold Spring Harbor Symp. Quant. Biol. 53, 365-374 (1988). Brown, A.M., A. Yatani, Y. Imoto, J. Codina, R. Mattera and L. Bimbaumer: Direct G-protein regulation of Ca2+ channels. Ann. NY Acad. Sci. 560, 373-386 (1989). Chaudhry, A., B.R. Conway, S.G. Laychock and R.P. Rubin: Analysis of the regulation of phosphatidylinositol-4,5-bisphosphate synthesis by arachidonic acid in exocrine pancreas. Arch. Biochem. Biophys. 272, 488-495 (1989). Darcy, P.K. and P.R. Fisher: A role for G-proteins and inositol phosphate signalling in Dictyostelium discoideum slug behaviour. J. Gen. Microbial. 135, 1909-1915 (1989). Entwistle, A., R.J. Zalin, S. Bevan and A.E. Warner: The control of chick myoblast fusion by ion channels operated by prostaglandins and acetylcholine. J. Cell Biol. 106, 1693-1702 (1988a). Entwistle, A., R.J. Zalin, A.E. Warner and S. Bevan: A role for acetylcholine receptors in the fusion of chick myoblasts. J. Cell Biol. 106, 1703-1712 (1988b). Ginsburg, G. and A.R. Kimmel: Inositol trisphosphate and diacylglycerol can differentially modulate gene expression in Dictyostelium. Proc. Natl. Acad. Sci. U.S.A. 86, 93329336 (1989). Hatanaka, M., N. Yumoto, N. Miwa, H. Morii, M. Tanemura, R. Ueno, Y. Watanabe and 0. Hayaishi: Late-phase accumulation of inositol phosphates stimulated by prostaglandins D, and Fza in neuroblastoma X glioma hybrid NGlOS-15 cells. J. Neurochem. 53, 1450-1455 (1989). Hausman, R.E. and S.G. Velleman: Prostaglandin E, receptors on chick embryo myoblasts. B&hem. Biophys. Res. Comm. 103, 213-218 (1981). Hausman, R.E. and D.A. Berggrun: Prostaglandin binding does not require direct cell-cell contact during chick myogenesis in vitro. Exp. Cell Res. 168, 457-462 (1987). Hausman, R.E., E.T. Dobi, E.J. Woodford, S. Petridis, M. Ernst and E.B. Nichols: Prostaglandin binding activity and myoblast fusion in aggregates of avian myoblasts. Dev. Biol. 113, 40-48 (1986). Hausman, R.E., C. Cametti, A. Bonincontro, P.L. Indovina, M.T. Santini and H. ElGendy: Rescue of the Li+-induced delay of embryonic myogenesis in vitro by added inositol. B&him. Biophys. Acta 1013, 92-96 (1989). Hausman, R.E., H. ElGendy and F. Craft: Requirement for G protein activity at a specific time during embryonic chick myogenesis. Cell Differ. Dev. 29, 13-20 (1990). Jeremy, J.Y., D.P. Mikhailidis and P. Dandona: Excitatory receptor-prostanoid synthesis coupling in smooth muscle: mediation by calcium, protein kinase C and G proteins. Prostaglandins Leukotrienes Essen. Fatty Acids 34,215-227 (1988). Kelvin, D.J., G. Simard and J.A. Connolly: FGF and EGF act synergistically to induce proliferation in BC3Hl myoblasts. J. Cell. Physiol. 138, 267-272 (1989).
115 Knudsen, K.A. and A.F. Horwitz: Tandem events in myoblast fusion. Dev. Biol. 58, 328-338 (1977). Knudsen, K.A. and A.F. Horwitz: Differential inhibition of myoblast fusion. Dev. Biol. 66, 294-307 (1978). Kurachi, Y., H. Ito, T. Sugimoto, T. Shimizu, I. Miki and M. Ui: Arachidonic acid metabolites as intracellular modulators of the G protein-gated cardiac K+ channel. Nature 337, 555-557 (1989). Laychock, S.G.: Prostaglandin E, inhibits phosphoinositide metabolism in isolated pancreatic islets. B&hem. J. 260, 291-294 (1989). Murphy, S. and G. Welk: Arachidonic acid evokes inositol phospholipid hydrolysis in astrocytes. FEBS Lett. 257, 6870 (1989). Needleman, P., J. Turk, B.A. Jakschik, A.R. Morrison and J.B. Lefkowith: Arachidonic acid metabolism. Annu. Rev. Biothem. 55, 69-102 (1986). Negishi, M., S. Ito and 0. Hayaishi: Prostaglandin E receptors in bovine adrenal medulla are coupled to adenylate cyclase via G, and to phosphoinositide metabolism in a pertussis toxin-insensitive manner. J. Biol. Chem. 264, 3916-3923 (1989). Okumura, T., T. Sago and K. Saito: Pertussis toxin blocks an inhibition of hormone-stimulated glycogenolysis by prostaglandin E, and its analogue in cultured hepatocytes. Prostaglandins 36, 463-476 (1988).
Santini, M.T., P.L. Indovina and R.E. Hausman: Changes in myoblast membrane order during differentiation as measured by EPR. Biochim. Biophys. Acta 896, 19-25 (1987). Santini, M.T., P.L. Indovina and R.E. Hausman: Prostaglandin dependence of membrane order changes during myogenesis in vitro. B&him. Biophys. Acta 938, 489-492 (1988). Sorokin, L.M., E.H. Morgan and G.C.T. Yeoh: Differences in transferrin receptor function between normal developing and transformed myogenic cells as revealed by differential effects of phorbol ester on receptor distribution and rates of iron uptake. J. Biol. Chem. 263, 14128814133 (1988). Wakelam, M.J.O.: Stimulated inositol phospholipid breakdown and myoblast fusion. B&hem. Sot. Trans. 14, 253256 (1986). Wang, J., M. Kester and M.J. Dunn: Involvement of a pertussis toxin-sensitive G-protein-coupled phospholipase A2 in lipopolysaccharide-stimulated prostaglandin E2 synthesis in cultured rat mesangial cells. Biochim. Biophys. Acta 963, 429-435 (1988). Watanabe, T., K. Umegaki and W.L. Smith: Association of a solubilized prostaglandin E, receptor from renal medulla with a pertussis toxin-reactive guanine nucleotide regulatory protein. J. Biol. Chem. 261, 13430-13439 (1986).
CELDIF
109
32 (1990) 109-116 Ltd.
00700
Prostaglandin-dependent phosphatidylinositol signaling during embryonic chick myogenesis H. Elgendy and R.E. Hausman Biological Science Center, Boston University, Boston, MA, U.S.A. (Accepted
16 August
1990)
Previous investigations suggested that binding of prostaglandin to a myoblast membrane receptor initiates a second messenger cascade which is essential for subsequent myogenesis. Initial evidence of the sensitivity of myogenesis to lithium suggested the involvement of inositol phosphate metabolism. That possibility is investigated here. The accumulation of inositol monophosphate in response to prostaghmdin binding was studied in aggregate cultures of chick embryo myoblasts in vitro. At 22 or 28 h in culture mononucleated myoblasts were labeled with [‘H]inositol, which was then incorporated into phosphoinositides. After experimental manipulations of prostaglandin metabolism and the addition of Li + prior to prostaglandin binding at 33 h, [ 3H]inositol monophosphate accumulation was measured by anion-exchange chromatography between 33 and 37 h. Inositol monophosphate was found to accumulate rapidly following 33 h. However, after 36 h of myogenesis, no inositol monophosphate accumulation was observed. The accumulation was dependent on prostaghmdin as indomethacin, which also blocks subsequent membrane events in myogenesis, blocked inositol phosphate accumulation. Like subsequent myogenesis, inositol phosphate accumulation was restored by the addition of exogenous prostaglandin. Finally, the accumulation of inositol phosphate began only after the binding of prostaghmdin. The results demonstrate that an inositol phosphate signal transduction mechanism connects prostaghmdin binding to membrane events in embryonic chick myogenesis. Cell adhesion; Myoblast fusion; Signal transduction; Receptor-mediated
Introduction Cell differentiation and expression of musclespecific proteins during embryonic myogenesis in
Abbreviations: taglandin.
PIP,,
polyphosphatidylinositol;
Correspondence address: ology, Boston University, 02215, U.S.A.
0922-3371/90/$03.50
R.E. Hausman, 5 Cummington
PG,
pros-
Department of BiStreet, Boston, MA
0 1990 Elsevier Scientific
Publishers
Ireland,
signaling
culture can be blocked or manipulated by a variety of environmental modifications (Hausman et al., 1986). These observations strongly suggest the existence of cell-cell signaling mechanisms to coordinate expression of the myogenic phenotype between the individual mononucleated cells. The cell membrane events termed fusion involve, at least, two steps: a characteristic increase in myoblast to myoblast adhesion and the actual fusion of the bilayers which may occur up to several hours later in culture (Knudsen and Horwitz, 1977, 1978). Ltd.
110
The increased adhesion occurs just after the expression of prostaglandin (PG) binding activity by the myoblasts and is sensitive to inhibitors of PG synthesis in the cultures (Hausman and Velleman, 1981; Hausman et al., 1986). Accompanying this binding of PG to the myoblasts, is a change in membrane organization (Santini et al., 1987) and the activity of a pertussis toxin-sensitive G protein (Hausman et al., 1990). Interference with either PG synthesis or G protein activity in the myoblasts at this critical time blocks the increase in cell-to-cell adhesion, the fusion of the bilayers and the characteristic decrease in myoblast membrane conductivity (Bonincontro et al., 1987; Santini et al., 1988; Hausman et al., 1989, 1990). These results suggested that PG, synthesized by the myoblasts, might be acting as an autocrine signal to coordinate changes in the cell membrane necessary for myogenesis (Hausman et al., 1986). Such a differentiation-staging signal would be able to affect cell surface events through a G protein without the intervention of a soluble message (Brown et al., 1988, 1989). However, if this same signal were to stage events at the level of the genome, a soluble second messenger pathway would seem to be necessary. Wakelam (1986) showed the existence of a polyphosphatidylinosito1 (PIP,) signal system in myoblasts, and we have shown indirectly by lithium sensitivity that turnover of PIP, occurs at the time of PG binding (Hausman et al., 1989). Here we demonstrate that binding of PG to embryonic chick myoblasts induces a rapid increase in inositol phosphate metabolism. This inositol phosphate flux occurs only under conditions which allow the increase in myoblast adhesion and the subsequent membrane events in myogenesis. Furthermore, this flux in inositol phosphate begins within 5 min after the addition of PG to the myoblasts. This suggests that the signal transduction process proceeds directly from the activated PG receptor through a G protein and a presumed phospholipase to cause the inositol phosphate flux. However, our present results cannot rule out a very limited number of rapidly occurring intervening steps.
Materials and Methods Materials
Fertile chicken eggs were obtained from Hardy’s (Essex, MA). Trypsin was obtained from Difco or Sigma, Dulbecco’s modified Eagle’s medium and penicillin/ streptomycin from Gibco. All other chemicals were obtained from Sigma, as previously (Hausman et al., 1986). [3H]Inositol, and [ ‘Hlinositol l-phosphate (inositol monophosphate) were obtained from American Radiolabeled Chemicals, Inc. (St. Louis, MO, U.S.A.), and [3H]PGE, was from New England Nuclear. Myoblast cell cultures
Tissue was prepared from pectoral and thigh muscles of ll-day embryonic chicks, and primary aggregate cultures enriched for myoblasts were established according to standard procedures (Hausman et al., 1986). Phosphoinositide labeling and hydrolysis
Myoblast aggregates at 22, 24 or 28 h of culture were exposed to [3H]inositol (5 pCi/ml) for 6-8 h. Longer periods of exposure did not affect the inositol phosphate accumulation measurements. At the end of the incubation period, the cultures were washed five times over a period of 60 min with fresh DME medium. The cultures then resuspended in fresh culture medium with or without 30 mM Li+, and the incubation continued for the time periods indicated in the figure legends. The incubations were arrested at the appropriate times by rapid cooling to 0°C removing the medium and adding 1 ml of cold 10% trichloroacetic acid. Acid extraction and ion-exchange chromatography of the inositol monophosphates were performed as described by Adamo et al. (1989). The water-soluble components of each sample were poured onto columns containing 1 ml of Dowex 1 (X8-200 resin; formate form) (Sigma) and eluted with 0.1 M Formic acid/O.2 M ammonium formate as described in Adamo et al. (1989). Details of individual experiments are given in the figure legends.
111
PGE, binding
Total and non-specific binding of [3H]PGE1 to myoblast aggregates was determined as previously (Hausman and Velleman, 1981; Hausman et al., 1986). Detection of inositol I -phosphate (inositol monophosphate)
We detected inositol monophosphate by anion-exchange chromatography (Berridge et al., 1982). [3H]Inositol monophosphate (5 x lo3 cpm) was added to an otherwise non-labeled myoblast culture and prepared as described above. Eight ml of this solution was applied to a Dowex-1 anionexchange column and eluted with 0.1 M formic acid/O.2 M ammonium formate. All the radioactivity eluted at fraction 10 or 11, precisely the inositol monophosphate fraction reported by Berridge et al. (1982), and recovery of radioactivity from the column was consistently greater than 85% (data not shown).
Results
The accepted demonstration that an inositol monophosphate second messenger system is activated as the result of receptor stimulation is the accumulation of [ 3H]inositol monophosphate in the presence of Li+ in cells which had been prelabeled with [3H]inositol (Berridge et al., 1982). On the basis of evidence in many systems, they suggested that this occurs because Li+ specifically blocks activity of the enzyme inositol-phosphate phosphatase. This causes the accumulation of inositol l-phosphate and the inability to regenerate the PIP, necessary for continued signaling. While we have previously shown that several of the membrane steps in myogenesis were inhibited by addition of Li+ (Bonincontro et al., 1987), this was not a direct demonstration of inositol phosphate accumulation in response to PG binding. Inositol monophosphate accumulation is blocked by Li + When myoblasts were labeled with [ 3H]inositol
from 22-28 h of culture and Li+ was added at 30 h, inositol monophosphate began to accumulate
_
E -
E
aI ;;
L % ,” a E .Z ;
2000 r 1800 1600 t 1400 1200 1000 800 600 -
.G
400
’
200
I
VI
0 L32
33 Time
34
35
in culture
36
37
(hours)
Fig. 1. Inositol monophosphate accumulation in presence or absence of 20 mM Lif. At 22 h of culture myoblast aggregates were exposed to 5 pCi/ml of [3H]inositol for 9 h. The myoblasts were washed and resuspended at 31, 32, 33, 34, 35 and 36 h of culture in fresh culture medium with (hatched bars) or without (open bars) 20 mM Li+. Inositol monophosphate accumulation was tested between 31-32, 32-33, 33-34, 34-35, 35-36 and 36-37 h of culture. Incubations were arrested at the specified times and analyzed by anion-exchange chromatography for [3H]inositol monophosphate as described in Materials and Methods. This figure represents one of two identical experiments.
only after 32 h in culture (Fig. 1). This accumulation of inositol monophosphate was dependent on the presence of Li+ (Fig. 1). This finding was consistent with our earlier results showing that this 32-36-h interval is when myogenesis, as measured by the change in adhesion or bilayer fusion, is also sensitive to Li+ (Hausman et al., 1989). Rescue of the indomethacin block to accumulation by added PGE,
Inositol phosphate accumulated between 32 and 36 h of culture as demonstrated above. This is the same time during which PG binds to the embryonic myoblast aggregates (Hausman and Berggrun, 1987). This raised the possibility that the inositol phosphate metabolism might be dependent on the availability of PG. To test this, we added 30 PM indomethacin (an inhibitor of cyclooxygenase and thus, PG synthesis (Needleman et al., 1986)) to myoblast aggregates at 30 h (in the presence of 20 mM Li+ as above) and measured inositol monophosphate accumulation.
112
7
g
-
al ;; r
3300 3000 2700 2400
R
2100
1~
1800
0
z c
M
3600
1500 1200
:: .g
900
I
300
600 0 ii+
Lif&
indo
Lif, 32 h
indo
& PGE, 37 h
Fig. 2. Time of rescue of indomethacin blockage of inositol monophosphate accumulation by added PGE,. At 24 h of culture the myoblast aggregates were washed and resuspended as four different sets. Sets l-3 were exposed to 5 nCi/ml [3H]inositol for 6 h, washed and resuspended at 30 h in fresh culture medium with 20 mM Li+. At this time, sets 2-4 had 30 uM indomethacin added. Set 4 was also exposed to 5 &i/ml [3H]inositol at this time for 6 h. In set 3, 10 pM PGE, was added at 32 h, 2 h after the addition of indomethacin. In set 4, 20 mM Li+ was added at 35 h and PGE, was added only at 37 h of culture. Incubations were arrested and analyzed by anion exchange chromatography as described in Materials and Methods. This figure represents one of two identical experiments.
This concentration of indomethacin allows the transient binding of exogenously added PG to the myoblasts but blocks subsequent myogenesis (Hausman and Berggrun, 1987). No inositol monophosphate accumulation was seen in the presence of indomethacin. However, if exogenous PGE, (1 PM) was added at 32 h, inositol monophosphate accumulation was restored (Fig. 2). In contrast, when myoblasts were labeled with [3H]inositol from 30-36 h and PGEl was added only at 37 h, no accumulation of inositol was seen (Fig. 2). These results demonstrated that phosphatidylinositol second messenger activity was dependent on PG metabolism during this stage of myogenesis in culture. Because PG binds to a myoblast PG receptor during this stage of myogenesis in culture (Velleman and Hausman, 1981; Hausman et al., 1986; Hausman and Berggrun, 1987), we have suggested that the requirement might actually be for PG binding (Santini et al., 1988).
Inositol monophosphate accumulation after PGE, binding at 32 h Here we asked two questions about the temporal relationship between PG binding and inositol phosphate metabolism. First, whether the accumulation of inositol monophosphate occurred before or after the binding of PG to PG receptors. Second, if there was a delay long enough to suggest that multiple steps or new macromolecular syntheses were involved. Therefore myoblasts were prelabeled with [3H]inositol from 22-28 h of culture, and 30 PM indomethacin was added at 30 h to block PG synthesis. PGE, and 20 mM Li+ were added at 32 h, and inositol monophosphate accumulation was tested at 2, 5, 10, 20 and 30 min after PG binding to its receptor at 32 h. The time of PG binding was assessed in parallel cultures
m-
OCTiTime in culture
c*l (hours)
Fig. 3. Determination of inositol monophosphate accumulation and PGE, binding. Open circles represent inositol monophosphate accumulation with the scale on the left. Open squares represent specific PGE, binding with the scale on the right. Myoblast aggregates were divided into two sets at 22 h of culture. One set was exposed to 5 pCi/ml of [3H]inositol for 6 h. These myoblasts were washed and resuspended at 28 h of culture in fresh culture medium with 30 pM indomethacin in five different flasks. At 32 h of culture 20 mM Lif and 10 aM PGE, were added to each of the five flasks. Incubations were arrested after 5, 10, 20, 40 and 60 min after PGE, addition (32 h). Samples were analyzed by anion-exchange chromatography as described in Materials and Methods. The other set of myoblast aggregates was left to differentiate in control medium until 28 h when they were resuspended in medium with 30 PM indomethacin in twelve different flasks. At 30 h all 12 flasks had unlabeled PGE, added to 10 pM, six flasks had [‘H]PGE, added to 10 nM. One flask from each of the two was analyzed at 30, 32, 33 and 36 h and specific binding determined by subtracting nonspecific binding from total binding (Hausman and Velleman, 1981). This figure represents one of two identical experiments.
113
with [ 3H]PGE,. Inositol monophosphate accumulation was first seen 5 min after addition of PGE,, and the accumulation continued to increase during the period of PG receptor activation (32-36 h) (Fig. 3).
Discussion
Previous work, using indomethacin to block PG synthesis (Hausman et al., 1990), had shown that binding of PG to the myoblasts is required for the characteristic increase in cell-cell adhesion. However, we did not know whether there was any further requirement for PG after the disappearance of the transient PG receptor at about 35 h. In this study we tested this by adding indomethacin to myoblasts at 30 h but delaying addition of PGE, until 37 h. We showed that myogenesis was not rescued under these conditions. Thus, we conclude that PG is required for cell signaling purposes only between 30 and 36 h in culture, an interval when myogenesis is also sensitive to Li+ (Bonincontro et al., 1987) and to pertussis toxin (Hausman et al., 1990). Perhaps this is not surprising, since the immediate consequence of this signaling event, the increase in myoblast adhesion, has already begun by 36 h of culture. Our present finding that, when added after 37 h of differentiation in culture, exogenous PG no longer results in an inositol phosphate signal is consistent with that hypothesis and further suggests that the signaling cascade at this time during myoblast differentiation in culture is critical. We cannot say whether one or more of the signal pathway components, such as the PG receptor (Hausman and Berggrun, 1987), disappears or becomes inactive. Previous evidence did not allow us to rule out alternative signaling sequences as being causal at this time during embryonic chick myogenesis. Ligands other than PG have been suggested for myogenic signaling (Wakelam 1986; Entwistle et al., 1988a; 1988b; Sorokin et al., 1988; Kelvin et al., 1989). The sensitivity of the signal transduction at 33-35 h of myogenesis to indomethacin demonstrates a need for PG metabolism at this time but not that PG initiates the signal by bind-
ing to a receptor. In some cells PG metabolism is necessary for signal transduction, but it is involved after the initial receptor-mediated event (Jeremy et al., 1988; Wang et al., 1988; Chaudhry et al., 1989; Kurachi et al., 1989; Murphy and Welk, 1989). The necessity for PG metabolism to precede the inositol phosphate flux is directly demonstrated here. Inositol monophosphate accumulation takes place when PG binds to its receptor and is absolutely dependent on PG being available. However, it does not matter whether PG is synthesized by the myoblasts (Hausman et al., 1986) or provided exogenously. Thus, the inositol phosphate flux is clearly downstream from the need for PG in the signal cascade. This conclusion is supported by the evidence that the inositol phosphate flux is detectable only after PG binds to its receptor and is apparent within 10 min. Given the asynchrony of the system (Hausman et al., 1987; Hausman and Berggrun, 1987) and the likelihood that the small initial amounts of inositol phosphate go undetected, this is evidence that it is the binding of PG to its receptor which directly stimulates the G protein (Hausman et al., 1990) and causes the breakdown of PIP,. PG receptor-stimulated signal transduction is well established in other cell types (Watanabe et al., 1988; Hatanaka et al., 1989; Laychock, 1989; Negishi et al., 1989; Okumura et al., 1989). Our present results clearly rule out the need for multiple steps and new macromolecular synthesis in the cascade of events which must occur between PG binding and G protein activity. Therefore, we suggest that the sequence of events which signals the time to increase myoblast cell adhesion amongst the population of cells in culture, and perhaps in the developing muscle, is that described above. The evidence obtained with indomethacin (Hausman et al., 1986) and pertussis toxin (Hausman et al., 1990) shows that PG metabolism and G protein activity are absolutely necessary for the increase in myoblast adhesion, the fusion of the bilayers and likely the later decrease in membrane conductivity (Bonincontro et al., 1987). The absolute need for PIP, breakdown for the increase in myoblast adhesion is less clear. Li+ (which blocks
114
inositol-phosphate phosphatase and hence the necessary regeneration of inositol) does not always block the change in adhesion and the decrease in conductivity but often only delays them (Hausman et al., 1989). As we have noted (Hausman et al., 1989), there are two obvious explanations for this. The increase in myoblast adhesion, bilayer fusion and the decrease in conductivity are all membrane-localized events, and some G proteins interact directly with membrane effecters (Brown et al., 1988; 1989). However, since Li+ delays these myoblast membrane events, the metabolism of inositol phosphate must play some role in their regulation. By similar reasoning, the membrane events may be also staged by the additional signaling pathway initiated by PIP, breakdown, that going through diacylglycerol. The diacylglycerol and inositol phosphate pathways are, for many cellular responses, known to be synergistic. A connection between the inositol phosphate signal initiated by PG binding and genomic events during embryonic chick myogenesis is, at present, speculative. However, there is evidence that a similar receptor-stimulated system acts during Dictyostelium development to coordinate the differentiation of the separate cells (Darcy and Fisher, 1989; Ginsburg and Kimmel, 1989).
Acknowledgement
This work was supported in part by funds from the Graduate School of Boston University.
References Adamo, S., C. Caporale, C. Nervi, R. Ceci and M. Molinaro: Activity and regulation of calcium-, phospholipid-dependent protein kinase in differentiating chick myogenic cells. J. Cell Biol. 108, 153-158 (1989). Berridge, M.J., C.P. Downes and M.R. Hanley: Lithium amplifies agonist-dependent phosphatidylinositol responses in brain and salivary glands. Biochem. J. 212, 473-482 (1982). Berridge, M.J., C.P. Downes and M.R. Hanley: Neural and developmental actions of lithium: a unifying hypothesis. Cell 59, 411-419 (1989). Bonincontro, A., C. Cametti, R.E. Hausman, P.L. Indovina and M.T. Santini: Changes in myoblast membrane electri-
cal properties during cell-cell adhesion and fusion in vitro. Biochim. Biophys. Acta 903, 89-95 (1987). Brown, A.M., A. Yatani, Y. Imoto, G. Kirsch, H. Hamm, J. Codina, R. Mattera and L. Bimbaumer: Direct coupling of G proteins to ionic channels. Cold Spring Harbor Symp. Quant. Biol. 53, 365-374 (1988). Brown, A.M., A. Yatani, Y. Imoto, J. Codina, R. Mattera and L. Bimbaumer: Direct G-protein regulation of Ca2+ channels. Ann. NY Acad. Sci. 560, 373-386 (1989). Chaudhry, A., B.R. Conway, S.G. Laychock and R.P. Rubin: Analysis of the regulation of phosphatidylinositol-4,5-bisphosphate synthesis by arachidonic acid in exocrine pancreas. Arch. Biochem. Biophys. 272, 488-495 (1989). Darcy, P.K. and P.R. Fisher: A role for G-proteins and inositol phosphate signalling in Dictyostelium discoideum slug behaviour. J. Gen. Microbial. 135, 1909-1915 (1989). Entwistle, A., R.J. Zalin, S. Bevan and A.E. Warner: The control of chick myoblast fusion by ion channels operated by prostaglandins and acetylcholine. J. Cell Biol. 106, 1693-1702 (1988a). Entwistle, A., R.J. Zalin, A.E. Warner and S. Bevan: A role for acetylcholine receptors in the fusion of chick myoblasts. J. Cell Biol. 106, 1703-1712 (1988b). Ginsburg, G. and A.R. Kimmel: Inositol trisphosphate and diacylglycerol can differentially modulate gene expression in Dictyostelium. Proc. Natl. Acad. Sci. U.S.A. 86, 93329336 (1989). Hatanaka, M., N. Yumoto, N. Miwa, H. Morii, M. Tanemura, R. Ueno, Y. Watanabe and 0. Hayaishi: Late-phase accumulation of inositol phosphates stimulated by prostaglandins D, and Fza in neuroblastoma X glioma hybrid NGlOS-15 cells. J. Neurochem. 53, 1450-1455 (1989). Hausman, R.E. and S.G. Velleman: Prostaglandin E, receptors on chick embryo myoblasts. B&hem. Biophys. Res. Comm. 103, 213-218 (1981). Hausman, R.E. and D.A. Berggrun: Prostaglandin binding does not require direct cell-cell contact during chick myogenesis in vitro. Exp. Cell Res. 168, 457-462 (1987). Hausman, R.E., E.T. Dobi, E.J. Woodford, S. Petridis, M. Ernst and E.B. Nichols: Prostaglandin binding activity and myoblast fusion in aggregates of avian myoblasts. Dev. Biol. 113, 40-48 (1986). Hausman, R.E., C. Cametti, A. Bonincontro, P.L. Indovina, M.T. Santini and H. ElGendy: Rescue of the Li+-induced delay of embryonic myogenesis in vitro by added inositol. B&him. Biophys. Acta 1013, 92-96 (1989). Hausman, R.E., H. ElGendy and F. Craft: Requirement for G protein activity at a specific time during embryonic chick myogenesis. Cell Differ. Dev. 29, 13-20 (1990). Jeremy, J.Y., D.P. Mikhailidis and P. Dandona: Excitatory receptor-prostanoid synthesis coupling in smooth muscle: mediation by calcium, protein kinase C and G proteins. Prostaglandins Leukotrienes Essen. Fatty Acids 34,215-227 (1988). Kelvin, D.J., G. Simard and J.A. Connolly: FGF and EGF act synergistically to induce proliferation in BC3Hl myoblasts. J. Cell. Physiol. 138, 267-272 (1989).
115 Knudsen, K.A. and A.F. Horwitz: Tandem events in myoblast fusion. Dev. Biol. 58, 328-338 (1977). Knudsen, K.A. and A.F. Horwitz: Differential inhibition of myoblast fusion. Dev. Biol. 66, 294-307 (1978). Kurachi, Y., H. Ito, T. Sugimoto, T. Shimizu, I. Miki and M. Ui: Arachidonic acid metabolites as intracellular modulators of the G protein-gated cardiac K+ channel. Nature 337, 555-557 (1989). Laychock, S.G.: Prostaglandin E, inhibits phosphoinositide metabolism in isolated pancreatic islets. B&hem. J. 260, 291-294 (1989). Murphy, S. and G. Welk: Arachidonic acid evokes inositol phospholipid hydrolysis in astrocytes. FEBS Lett. 257, 6870 (1989). Needleman, P., J. Turk, B.A. Jakschik, A.R. Morrison and J.B. Lefkowith: Arachidonic acid metabolism. Annu. Rev. Biothem. 55, 69-102 (1986). Negishi, M., S. Ito and 0. Hayaishi: Prostaglandin E receptors in bovine adrenal medulla are coupled to adenylate cyclase via G, and to phosphoinositide metabolism in a pertussis toxin-insensitive manner. J. Biol. Chem. 264, 3916-3923 (1989). Okumura, T., T. Sago and K. Saito: Pertussis toxin blocks an inhibition of hormone-stimulated glycogenolysis by prostaglandin E, and its analogue in cultured hepatocytes. Prostaglandins 36, 463-476 (1988).
Santini, M.T., P.L. Indovina and R.E. Hausman: Changes in myoblast membrane order during differentiation as measured by EPR. Biochim. Biophys. Acta 896, 19-25 (1987). Santini, M.T., P.L. Indovina and R.E. Hausman: Prostaglandin dependence of membrane order changes during myogenesis in vitro. B&him. Biophys. Acta 938, 489-492 (1988). Sorokin, L.M., E.H. Morgan and G.C.T. Yeoh: Differences in transferrin receptor function between normal developing and transformed myogenic cells as revealed by differential effects of phorbol ester on receptor distribution and rates of iron uptake. J. Biol. Chem. 263, 14128814133 (1988). Wakelam, M.J.O.: Stimulated inositol phospholipid breakdown and myoblast fusion. B&hem. Sot. Trans. 14, 253256 (1986). Wang, J., M. Kester and M.J. Dunn: Involvement of a pertussis toxin-sensitive G-protein-coupled phospholipase A2 in lipopolysaccharide-stimulated prostaglandin E2 synthesis in cultured rat mesangial cells. Biochim. Biophys. Acta 963, 429-435 (1988). Watanabe, T., K. Umegaki and W.L. Smith: Association of a solubilized prostaglandin E, receptor from renal medulla with a pertussis toxin-reactive guanine nucleotide regulatory protein. J. Biol. Chem. 261, 13430-13439 (1986).