Insulin-like Growth Factors Require Phosphatidylinositol 3-Kinase to Signal Myogenesis: Dominant Negative p85 Expression Blocks Differentiation of L6E9 Muscle Cells

Abstract

Phosphatidylinositol 3 (PI 3)-kinases are potently inhibited by two structurally unrelated membrane-permeant reagents: wortmannin and LY294002. By using these two inhibitors we first suggested the involvement of a PI 3-kinase activity in muscle cell differentiation. However, several reports have described that these compounds are not as selective for PI 3-kinase activity as assumed. Here we show that LY294002 blocks the myogenic pathway elicited by insulin-like growth factors (IGFs), and we confirm the specific involvement of PI 3-kinase in IGF-induced myogenesis by overexpressing in L6E9 myoblasts a dominant negative p85 PI 3-kinase-regulatory subunit (L6E9-Δp85). IGF-I, des(1–3)IGF-I, or IGF-II induced L6E9 skeletal muscle cell differentiation as measured by myotube formation, myogenin gene expression, and GLUT4 glucose carrier induction. The addition of LY294002 to the differentiation medium totally inhibited these IGF-induced myogenic events without altering the expression of a non-muscle-specific protein, β1-integrin. Independent clones of L6E9 myoblasts expressing a dominant negative mutant of the p85-regulatory subunit (Δp85) showed markedly impaired glucose transport activity and formation of p85/p110 complexes in response to insulin, consistent with the inhibition of PI 3-kinase activity. IGF-induced myogenic parameters in L6E9-Δp85 cells, i.e. cell fusion and myogenin gene and GLUT4 expression, were severely impaired compared with parental cells or L6E9 cells expressing wild-type p85. In all, data presented here indicate that PI 3-kinase is essential for IGF-induced muscle differentiation and that the specific PI 3-kinase subclass involved in myogenesis is the heterodimeric p85-p110 enzyme.

INTRODUCTION

Growth factors are generally considered to inhibit myogenesis. However, it is well documented that insulin-like growth factors (IGFs) are crucial to this process (1). IGF-I and IGF-II are potent stimulators of muscle differentiation, and they are potential candidates for regulation of satellite cell function during regeneration, a characteristic response of adult muscle to exercise or injury (2, 3). It has been shown that IGF expression is increased during myoblast differentiation in response to serum withdrawal (4–8). Furthermore, the level of IGF-II secreted from muscle cells correlates with the rate of spontaneous differentiation, and antisense oligonucleotides complementary to IGF-II mRNA inhibited differentiation in the absence but not in the presence of exogenous IGF-II (6). The biological significance of the IGFs has also been analyzed by recombinant ablation studies. A common observation in mouse lines lacking IGF-I or its receptor is that embryos are viable, but embryonic development is impaired and neonates die immediately after birth because they cannot breathe (9, 10). Furthermore, expression of IGF-I in skeletal muscle results in myofiber hypertrophy (11), and overexpression of IGF-I in the heart leads to cardiomegaly mediated by an increased number of cells in the heart (12).

Much information has recently been gained on the role of IGFs in myogenesis (reviewed in Ref.13). However, the intracellular myogenic signaling process dependent on IGFs is poorly understood. We have recently reported that the phosphatidylinositol 3 (PI 3)-kinase inhibitors, wortmannin and LY294002, block differentiation of skeletal muscle cells, suggesting that phosphatidylinositol 3-kinase is essential for the terminal differentiation of muscle cells (14). In this context, it has recently been reported that LY294002 inhibits L6A1 muscle cell differentiation induced by IGF-I (15). Indeed, during the last few years, much insight has been gained on the cellular functions of PI 3-kinase by the use of wortmannin (for review see Ref.16) and LY294002 (17), both of which inhibit all PI 3-kinase subclasses so far described in the nanomolar or low micromolar range. However, several reports have described that these compounds are not as selective for PI 3-kinase activity as assumed. Indeed, wortmannin and its structural analog demethoxyviridin inhibit stimulated phospholipase A2 activity with an IC₅₀ of 2 nm (18). Moreover, wortmannin has also been reported to inhibit phosphatidylinositol 4-kinase (19), phospholipase C and D (20), and myosin light chain and pleckstrin phosphorylation (21) albeit at concentrations greater than those required to inhibit PI 3-kinase. On the other hand, the specificity of LY294002 for other lipid-metabolizing enzymes has not been examined. Accordingly, the assignation of a role for PI 3-kinase in a particular cellular pathway on the sole basis of its chemical inhibition may lead to incorrect conclusions.

In an attempt to identify a signaling intermediate for the myogenic actions of IGFs, here we analyze the effects of 1) LY294002 and 2) the expression of a dominant negative mutant of p85 PI 3-kinase-regulatory subunit in IGF-induced L6E9 muscle cell differentiation. We show that PI 3-kinase is an essential element for IGF-induced muscle differentiation and that the specific PI 3-kinase subclass involved in myogenesis is the heterodimeric p85-p110 enzyme.

RESULTS

IGFs Induce Biochemical and Morphological Differentiation of L6E9 Myoblasts: Blockade by the PI 3-Kinase Inhibitor LY294002

Confluent L6E9 myoblasts were incubated in a serum-free medium supplemented with increasing concentrations of IGF-II. After 2 days in these conditions, cells expressed myogenin mRNA with an ED₅₀ for IGF-II of ≈20 nm, which was consistent with the activation of the IGF-I receptor (Fig. 1a). IGF-II increased myogenin mRNA levels up to 6-fold compared with cells maintained in serum-free medium alone (basal conditions). Fig 1b shows that IGF-I and its more potent analog des(1, 3)IGF-I also induced myogenin mRNA and they were at least 10-fold more potent than IGF-II. The ability of all three IGFs to induce myogenin gene expression was completely abolished by the PI 3-kinase inhibitor LY294002 (20 μm), suggesting that PI 3-kinase is an essential downstream effector for IGF induction of L6E9 muscle cell differentiation (Fig. 1b).

Figure 1.

IGFs-Induced Biochemical and Morphological Differentiation in L6E9 Myoblasts Is Blocked by the PI 3-Kinase Inhibitor LY294002 Confluent L6E9 myoblasts were allowed to differentiate in serum-free medium for 2 days in the absence or presence of IGFs with or without LY294002 (20 μm). (a) Myogenin mRNA was analyzed by Northern blots and quantitated by densitometry. Myogenin mRNA abundance in the absence of IGF-II was considered as basal expression, and data are expressed as fold-stimulation over basal. (b) Myogenin mRNA expression was analyzed in myoblasts (Mb) and after 2 days in serum-free medium without (0) or with 3 nm IGF-I, 3 nm des(1, 3)IGF-I, or 40 nm IGF-II supplementation, in the absence or presence of 20 μm LY294002. Representative autoradiograms from three separate experiments are shown. The integrity and relative amounts of RNA in each sample were checked by ethidium bromide staining (rRNA). (c) GLUT4 glucose transporter and β1-integrin content was analyzed by Western blot of total cell lysates. GLUT4 level after 2 days in serum-free medium in the absence of IGF-II was considered as basal expression, and data are expressed as percentage over basal. (d) GLUT4 glucose transporter andβ 1-integrin content was analyzed in myoblasts (Mb) and after 2 days in serum-free medium without (0) or with 3 nm IGF-I, 3 nm des(1, 3)IGF-I, or 40 nm IGF-II supplementation, in the absence or presence of 20 μm LY294002. Representative autoradiograms from three independent experiments are shown. (e) Cells were grown to confluence in a 10% FBS-containing medium and then allowed to differentiate in a serum-free medium (0.5 mg/ml BSA containing DMEM) (left), supplemented with 40 nm IGF-II (center), or supplemented with 40 nm IGF-II and 20 μm LY294002 (right). After 2 days in each condition, cells were photographed. Images shown are representative of 10–20 microscopic fields taken at random from each one of at least 10 independent experiments. Scale bars, 30 μm (the scale is the same for all panels).

Among the functional markers of skeletal muscle terminal differentiation is the insulin-sensitive glucose transporter GLUT4 (22). After 2 days in serum-free medium supplemented with IGF-II, L6E9 cells expressed GLUT4 with an IGF-II dose dependency reflecting the activation of IGF-I receptor (ED₅₀ of ≈20 nm) (Fig. 1c). The effect of IGF-II was specific for the muscle protein as it did not modify the expression of the ubiquitous membrane proteinβ 1-integrin (Fig. 1c, inset). IGF-I and des(1, 3)IGF-I also induced the expression of GLUT4, but they were at least 10-fold more potent than IGF-II (Fig. 1d). As observed for myogenin expression, PI 3-kinase inhibitor LY294002 blocked the effect of IGF-I (3 nm), des(1, 3)IGF-I (3 nm), and IGF-II (40 nm) on GLUT4 expression (Fig. 1d). LY294002 did not alter the expression of β1-integrin, indicating that the inhibitor did not affect the protein levels of a structural component of the cells and that it specifically blocked the IGF-induced expression of GLUT4 (Fig. 1d).

At the morphological level, confluent L6E9 myoblasts incubated in a 2% serum containing medium initiate a differentiation program that consists, at the morphological level, in myoblast elongation and alignment during the first 24 h, followed by multinucleate myotube formation (14). The presence of low serum concentrations in the differentiation medium was found to be essential for terminal differentiation of L6E9 cells since after 2 days in a serum-free medium (DMEM containing 0.5 mg/ml BSA), cells aligned to each other and showed an elongated morphology, although they did not fuse or fused very poorly into myotubes (Fig. 1e, left). Supplementation of serum-free medium with IGFs led to a potent induction of cell fusion. Fig 1e (center) shows large multinucleated myotubes induced by 40 nm IGF-II. IGF-I (3 nm) or des(1, 3)IGF-I (3 nm) induced cell fusion comparable to that induced by 40 nm IGF-II (data not shown). As observed for myogenin and GLUT4, after 2 days in serum-free medium supplemented with 40 nm IGF-II and 20 μm LY294002, L6E9 cells remained largely unfused (Fig. 1e, right). PI 3-kinase inhibitor also blocked the cell fusion induced by IGF-I (3 nm) or des(1, 3)IGF-I (3 nm) (data not shown). All these results suggest that PI 3-kinase activity is essential for IGF-induced biochemical and morphological differentiation of L6E9 cells.

p85α Is the Predominant PI 3-Kinase Adapter Subunit Isoform Expressed in L6E9 Cells

Fully differentiated muscle expresses a number of splice variants of p85α adapter subunit of PI 3-kinase, all of which are regulated by insulin and could therefore potentially be involved in IGF-mediated processes (23). In an effort to determine whether any of these isoforms was important in the differentiation of L6E9 muscle cells, we first analyzed, by Western blot, lysates and total membranes of L6E9 myoblasts and myotubes using a previously described antibody that recognizes p85β and all the splice variant forms of p85α (23). In human muscle lysates, the antibody recognized four identified major bands of 87 kDa (p85β), 85 kDa (p85α), 53 kDa (p55α/AS53), and 48 kDa (p50) (Fig. 2) (23). However, in both L6E9 myoblasts and myotubes the full-lengh p85α was the predominant adapter subunit expressed.

Figure 2.

Characterization of PI 3-Kinase Regulatory Subunits Present in L6E9 Skeletal Muscle Cells L6E9 myoblasts (Mb) and myotubes (Mt) lysates or total membrane fractions (75 μg) and a control of human muscle lysates (300 μg) were analyzed by Western blot with a polyclonal antibody raised against glutathione S-transferase fusion protein, corresponding to the N-SH2 domain of human p85α, and visualyzed by [¹²⁵I]protein A.

Expression of a Dominant-Negative p85α in L6E9 Myoblasts

All PI 3-kinase isoforms so far described are potently inhibited in the nanomolar or low micromolar range by two structurally unrelated membrane- permeant reagents: wortmannin (16) and LY294002 (17). By using these two compounds we first suggested the involvement of a PI 3-kinase activity in muscle cell differentiation (14). Furthermore, the results presented above seem to indicate that PI 3-kinase is essential for IGF-induced myogenesis. However, one question remains: whether the inhibition of myogenesis by wortmannin and LY294002 is a specific reflection of PI 3-kinase involvement. In an effort to clarify this aspect and taking into account that p85α is the predominant PI 3-kinase adapter subunit form expressed in L6E9 cells, we stably overexpressed in L6E9 myoblasts a p85α lacking a binding site for the p110 catalytic subunit of PI 3-kinase (L6E9-Δp85) and a wild-type p85α as a control (L6E9-Wp85) (24).

Screening of positive clones overexpressing p85 (Wp85 or Δp85) was performed by immunofluorescence assays using polyclonal rabbit antibodies against rat p85 PI 3-kinase. We selected five independent clones for each Wp85- or Δp85-transfected cells in which the level of expression of Δp85 was comparable to the level of expressed Wp85. Transfected proteins were 2- to 3 times overexpressed compared with the level of endogenous p85 in untransfected cells (Fig. 3, B and C vs. A). As a control, we analyzed the level of expression of β1-integrin, which was essentially identical for untransfected and transfected cells (Fig. 3, D–F). Figure 3 also shows that the subcellular distribution of p85 under basal conditions was mostly intracellular in both transfected and untransfected cells (Fig. 3, A–C). In contrast, β1-integrin exhibited a typical distribution pattern of a plasma membrane marker (Fig. 3, D–E).

Figure 3.

Immunofluorescence Localization of p85 andβ ₁-Integrin in Untransfected L6E9, L6E9-Δp85, and L6E9-Wp85 Myoblasts Untransfected L6E9 (panels A and D), L6E9-Δp85 (panels B and E), and L6E9-Wp85 cells (panels C and F) were grown on glass coverslips, fixed with methanol, and incubated with anti-p85 antibody (panels A–C) or with anti-β₁-integrin antibody (panels D–F). Cells were incubated with a rodamine-conjugated secondary antibody, as described in Materials and Methods. Results are representative of five independent clones of both Wp85- andΔ p85-cells analyzed in two independent experiments. Scale bar, 25μ m.

We compared the growth rate of the selected clones of Wp85- andΔ p85-transfected cells to analyze the impact of p85 dominant negative expression on L6E9 cell proliferation. Consistent with our previous observations in L6E9 myoblasts grown in the presence of wortmannin (14), Δp85-transfected cells proliferated normally in response to serum, and no differences in cell growth were detected when compared to Wp85-transfected cells (Table 1) and untransfected cells (data not shown).

Table 1.

Dominant-Negative p85α Expression in L6E9 Cells Does Not Affect Cell Growth

Culture day	Cell No. (fold increase over day 1)
	Δp85-L6E9	Wp85-L6E9
1	1	1
2	2,5	2,5
3	4,8	5,4
4	12,4	14,7

To quantify cell proliferation, cells were plated in multiwell culture dishes, grown from 1–4 days in 10% FBS-containing medium, and counted after pancreatinization. Data are representative of at least 10 independent experiments performed with five different clones of bothΔ p85- and Wp85-transfected L6E9 cells.

As a functional assay to test the inhibition of PI 3-kinase activity in L6E9-Δp85 cells, we analyzed the glucose transport activity. We and others have previously shown that PI 3-kinase activity is crucial to the regulation of glucose transport in L6E9 myoblasts (25) and other mammalian cell types (26–30). Moreover, studies from Hara et al. (24) showed that glucose uptake is markedly impaired in Chinese hamster ovary (CHO) cells overexpressing Δp85. We analyzed three independent clones of both L6E9-Δp85 and L6E9-Wp85 cells for glucose transport activity (Fig. 4). In L6E9-Wp85 maximal glucose transport activity was observed in the absence of insulin. This seems to indicate that the overexpression of Wp85 saturated the endogenous cell machinery sensitive to insulin which did not cause any further enhancement of glucose uptake. In contrast, cells overexpressing Δp85 showed a marked decrease in both basal and insulin-stimulated 2-deoxyglucose uptake compared with either untransfected or L6E9-Wp85 cells (Fig. 4). The fact that glucose uptake by L6E9-Δp85 cells remained sensitive to insulin is consistent with our observation that, in L6E9 and Sol8 myoblasts, wortmannin produced a parallel decrease in basal and insulin-stimulated glucose uptake, but that insulin action is abolished only at very high wortmannin concentrations (1 μm) (25).

Figure 4.

2-Deoxyglucose Uptake Is Impaired in L6E9 Myoblasts Overexpressing a Dominant-Negative p85α After serum starvation, untransfected (nt), L6E9-Wp85 (Wp85), and L6E9-Δp85 (Δp85) cells were incubated for 30 min in the absence (open bars) or in the presence of 1 μm insulin (solid bars). After addition of 0.1 mm [³H]2-deoxyglucose, the cellular hexose uptake at t = 20 min was measured as described in Materials and Methods. Results are expressed as a percentage of untransfected cell glucose uptake determined in the absence of insulin (basal activity). Three independent clones of both Wp85- and L6E9-Δp85 cells were analyzed for glucose transport; results are the means of four to five independent experiments, in which each point was run in triplicate.

To further characterize the dominant negative effect of Δp85 transfection in L6E9 cells, we compared the ability of p85 to bind to the catalytic p110α PI 3-kinase subunit after insulin stimulation in untransfected, Δp85 and Wp85 L6E9 myoblasts (Fig. 5). Consistent with the glucose transport experiment, in which insulin showed no further effect on L6E9-Wp85 cells compared with untransfected cells, the level of p110α complexed with p85 in Wp85-cells was essentially the same as in untransfected cells. In contrast, Δp85-transfected cells showed a 2-fold decrease in the level of p110 coimmunoprecipitated with p85 (Fig. 5).

Figure 5.

Recruitment of p110α into p85 Complexes by Insulin After serum starvation, untransfected (nt), L6E9-Δp85 (Δp85), and L6E9-Wp85 (Wp85) myoblasts were incubated for 10 min in the presence of 1 μm insulin. Cells were solubilized, and proteins (2.5 mg) were incubated with 5 μl of antiserum against rat p85 and Protein G-Sepharose. Immune complexes were analyzed by Western blot with an antibody against rat p110α. A representative autoradiograph is shown.

Myotube Formation Is Impaired in L6E9-Δp85 Cells

Cell differentiation was analyzed in five independent clones of both Wp85- and Δp85-transfected L6E9 cells. L6E9-Wp85 cells were morphologically indistinguishable from L6E9 parental cells at all the conditions tested, i.e. proliferation and differentiation (data not shown). Images shown are representative of 10–20 microscopic fields taken at random from each one of four independent experiments in which L6E9 parental cells and Wp85- and Δp85-L6E9 clones were cultured in parallel under identical conditions.

L6E9-Δp85 myoblasts proliferated normally in a 10% FBS-containing medium and were morphologically similar to L6E9-Wp85 (Fig. 6, a and f). Confluent cells were allowed to differentiate in a serum-free medium with or without IGF-II (0–100 nm IGF-II). Figure 6 shows the morphological changes undergone by L6E9-Wp85 (Fig. 6, b–e) and L6E9-Δp85 cells (Fig. 6, g–j) during a 4-day differentiation period. After 2 days in serum-free medium without IGF-II, L6E9-Wp85 and L6E9-Δp85 cells were aligned to each other and elongated compared with myoblasts (Fig. 6, b vs. a and g vs. f, respectively), but little or no fusion was observed in these conditions. In the presence of IGF-II, myotube formation was observed in L6E9-Wp85 cells (Fig. 6, c and d, for 20 and 100 nm IGF-II, respectively), whereas under the same conditions, L6E9-Δp85 did not fuse or fused very poorly (Fig. 6, h and i, for 20 and 100 nm IGF-II, respectively). Large multinucleated myotubes were observed in L6E9-Wp85 cells after 4 days in the presence of 100 nm IGF-II (Fig. 6e) while L6E9-Δp85 remained aligned and elongated, but fusion was largely prevented (Fig. 6j).

Figure 6.

Myotube Formation Is Impaired in L6E9-Δp85 Cells Cell differentiation was analyzed in five independent clones of both Wp85- and Δp85-transfected L6E9 cells. Results shown are representative of 10–20 microscopic fields taken at random from each of four independent experiments in which L6E9 parental cells and Wp85- and Δp85-L6E9 clones were cultured in parallel under identical conditions. L6E9-Wp85 cells were indistinguishable from L6E9 parental cells in all conditions assayed (data not shown), and in all the experiments, they were both considered as controls for proliferation and differentiation. Control (a–e) and L6E9-Δp85 (f–j) cells were grown to confluence in a 10% FBS-containing medium (a and f) and then allowed to differentiate in a serum-free medium (0.5 mg/ml BSA containing DMEM) without (b and g) or with IGF-II at concentrations of 20 nm (c and h) or 100 nm (d, e, i, and j). Cells were photographed after 2 days (b, c, d, g, h, and i) or after 4 days (e and j). Scale bars, 30 μm (the scale is the same for all panels).

Myogenin and Glucose Transporter GLUT4 Expression Is Decreased in L6E9-Δp85 Cells

L6E9-Δp85 and L6E9-Wp85 cells were grown to confluence. Cells were then incubated at increasing doses of IGF-II (0–100 nm). The IGF-II dose-dependence for myogenin expression in L6E9-Wp85 cells was similar to that observed for untransfected cells (Fig. 7vs.Fig. 1a). In L6E9-Δp85 cells, the maximal response to IGF-II for myogenin gene induction was reduced by 62 ± 13% (n = 3) compared with L6E9-Wp85 cells (Fig. 7). Figure 8 shows glucose transporter GLUT4 expression in L6E9-Δp85 and L6E9-Wp85 after 4 days in a serum-free medium with or without IGF-II. Little or no induction of GLUT4 was observed in L6E9-Δp85 or L6E9-Wp85 cells in the absence of IGF-II. As determined for untransfected cells (Fig. 1 c), maximal expression of GLUT4 was detected at 50 nm IGF-II concentrations. However, the maximal expression of the glucose transporter was decreased by 80 ± 7% (n = 3) in L6E9-Δp85 cells compared with control cells (Fig. 8). The expression of β1-integrin, a ubiquitous plasma membrane component, remained unaltered in all the conditions tested, indicating that IGF-II-induced GLUT4 protein expression was specifically associated with muscle differentiation and that PI 3-kinase was required for this process.

Figure 7.

IGF-II-induced Myogenin Gene Expression Is Decreased in L6E9-Δp85 Cells Confluent L6E9-Wp85 and L6E9-Δp85 cells were allowed to differentiate in serum-free medium for 2 days in the absence or presence of increasing concentrations of IGF-II (10, 50 and 100 nm). Total RNA was obtained from the different experimental groups, and 10μ g of RNA were laid on gels. After blotting, myogenin mRNA was detected by hybridization with a 1,100 bp EcoRI fragment as a cDNA probe. Representative autoradiograms after 30 min of exposure from three separate experiments are shown. The integrity and relative amounts of RNA in each sample were checked by ethidium bromide staining (rRNA).

Figure 8.

IGF-II-induced Glucose Transporter GLUT4 Expression Is Decreased in L6E9-Δp85 Cells Confluent L6E9-Wp85 and L6E9-Δp85 cells were allowed to differentiate in serum-free medium for 4 days in the absence or presence of IGF-II (50 and 100 nm). GLUT4 glucose transporter andβ 1-integrin content were analyzed by immunoblotting 30 μg of solubilized proteins from the different experimental groups. Representative autoradiograms from three independent experiments are shown.

DISCUSSION

In a previous study, the use of the cell-permeant inhibitors, wortmannin and LY294002, suggested that PI 3-kinase was essential for terminal differentiation of muscle cells (14). In this study, we show that PI 3-kinase is an essential second messenger for the myogenic actions of insulin-like growth factors (IGFs), and we identify the heterodimeric p85-p110 PI 3-kinase as the PI 3-kinase subclass involved in myogenesis.

We have previously shown that PI 3-kinase activity in L6E9 cells is stimulated by insulin at concentrations that correlate with the activation of the IGF-I receptor, this stimulation being inhibited in a dose-response manner by wortmannin (25). Indeed, most of insulin actions in these cells are mainly signaled through the IGF-I receptor since L6 cells express insulin and IGF-I receptors in a ratio about 1:400 in myoblasts and 1:50 in myotubes (31). Here we show that the effect of IGFs on cell fusion and myogenin and GLUT4 expression was totally blocked by the PI 3-kinase inhibitor LY294002. Wortmannin was not used in this study because of its short half-life in aqueous solution (32), which renders it unsuitable for experiments involving 2- to 4-day incubations. The dose-response studies presented here seem to indicate that the IGF-II receptor is not relevant for GLUT4 or myogenin expression in L6E9 cells. However, the contribution of IGF-II receptor to myogenesis cannot be ruled out since in mouse BC3H-1 muscle cells an IGF-II receptor-selective analog of IGF-II promoted cell differentiation (33).

A family of distinct PI 3-kinase enzymes has been cloned and characterized in mammals, and these can be distinguished on the basis of structure, function, and mechanisms of activation (reviewed in Ref.34). A well characterized class of PI 3-kinases are heterodimers composed of a regulatory p85 subunit (isoforms: α, β, p55PIK, and other p85 splice variants) (35–40) and a catalytic p110 subunit (isoforms α and β) (40, 41), which possesses a Ser/Thr protein kinase activity in addition to its lipid kinase activity (42–44). This group of enzymes is regulated by cell surface receptors via intrinsic or associated tyrosine-kinase activities. Here, we stably transfected L6E9 cells with a dominant negative p85α-subunit (Δp85) that lacks the binding site for the p110 catalytic subunit of PI 3-kinase. As expected, Δp85-cells showed impaired ability to form p85/p110 complexes in response to insulin, and they also showed reduced basal and insulin-stimulated glucose transport activity, which is known to be dependent on intact PI 3-kinase activity in L6E9 cells (25). However, probably due to the low level of overexpression of transfected proteins, Δp85-cells remained insulin-sensitive for both parameters, although to a much lesser extent than untransfected cells.

IGF-induced myogenic parameters in L6E9-Δp85, i.e. cell fusion, myogenin gene, and GLUT4 expression, were severely impaired compared with parental cells or L6E9-Wp85 cells. As for glucose transport activity, the effect of maximal doses of PI 3-kinase chemical inhibitors on cell differentiation blockade was more dramatic than the effect of a 2- to 3-fold overexpression of Δp85. However, the absence of large multinucleated myotubes and the reduction by 62% in myogenin mRNA and by 80% in GLUT4 protein expression in Δp85-transfected cells indicate that the heterodimeric PI 3-kinase is essential for IGF-induced L6E9 cell differentiation. In this context, several splice variants of p85 are present in fully differentiated human muscle, and each of these is stimulated by insulin to a different extent, indicating that they could have distinct roles in insulin and IGF-I signaling (23). However, in the current study we find that p85α is the predominant PI 3-kinase adapter subunit expressed in both L6E9 myoblasts and myotubes. This, together with the inhibition of differentiation by Δp85, indicates that IGF stimulation of full-length p85α is sufficient to activate the PI 3-kinase required for myogenesis. However, in human muscle, other adapter subunits that are also abundant (23) may play a role in cell differentiation. Indeed, observations from our laboratory show that PI 3-kinase is essential for myotube formation in human skeletal muscle cells (our unpublished observations).

There is scarce information regarding the downstream elements activated by PI 3-kinase or its PI 3-phosphate products. It has recently been shown that p70^S6k activity is increased substantially during skeletal muscle cell diffentiation in the absence, but not in the presence, of LY294002 and that rapamycin, an inhibitor of p70^S6k activity, abolishes IGF-I-induced differentiation (15). These results strongly suggest that p70^S6k is involved in the IGF/PI 3-kinase myogenic pathway. Other putative downstream elements of this pathway may include the Ser/Thr protein kinase PKB (also known as Akt/RAC) and some protein kinase C (PKC) isoforms. PKB is activated by insulin in L6 myotubes, and this activation is prevented by PI 3-kinase inhibitors (45). Furthermore, the relationship of PKB and PKC kinase families is particularly interesting in light of the ability of novel and atypical PKC isoforms (PKC ε, -δ, -ζ, and -η) to interact with PI 3-kinase products PI 3,4,5-triphosphate and PI 3,4-diphosphate (46, 47). Moreover, PKCδ specifically associates wih PI 3-kinase after cytokine stimulation (48). In the context of these findings combined with our results, it is tempting to hypothesize that PKB and/or PKC isoforms could be targets of PI 3-kinase in the myogenic signaling pathway. Moreover, it has recently been described that ERK6, a mitogen-activated protein kinase, is involved in C2C12 myoblast differentiation (49). ERK6 seems to be specifically expressed in skeletal muscle and to signal differentiation through phosphotyrosine-mediated pathways distinct from those activating other members of the mitogen-activated protein kinase family such as ERK1 and ERK2. It would be of interest to determine whether ERK6 and PI 3-kinase are convergent signals for myogenesis or whether ERK6 defines an alternative myogenic pathway.

Overall, the results presented here provide evidence that p85-p110 PI 3-kinase is an essential mediator for IGF-induced muscle cell differentiation through the IGF-I receptor. Additional work is required to identify the downstream elements involved in the myogenic signaling cascade.

MATERIALS AND METHODS

Materials

IGF-I and IGF-II were kindly provided by Eli Lilly (Indianapolis, IN). des(1–3)IGF-I was from Angelika F. Schutzdeller (Tubingen, Germany). LY294002 was from BIOMOL Research Laboratories (Plymouth, Meeting, PA). L6E9 rat skeletal muscle cell line was kindly provided by Dr. B. Nadal-Ginard (Harvard University, Boston, MA). The polyclonal antibody OSCRX was raised against the C terminus of GLUT4 (50). A rabbit polyclonal antibody against β1-integrin was kindly given by Dr. Carles Enrich (University of Barcelona, Barcelona, Spain) (51). Polyclonal antibodies against rat p85 and p110α subunits of PI 3-kinase were from Upstate Biotechnology, Inc. (Lake Placid, NY). A polyclonal antibody was raised to a glutathione S-transferase fusion protein corresponding to the N-SH2 (p85α-NSH2) domain of human p85α as described previously (23).

cDNA encoding for myogenin was kindly given by Dr. Eric Olson (University of Texas, Houston, TX).

Cell Culture

Rat skeletal muscle L6E9 myoblasts were grown in monolayer culture in DMEM containing 10% (vol/vol) FBS and 1% (vol/vol) antibiotics (10,000 U/ml penicillin G and 10 mg/ml streptomycin). Confluent myoblasts were differentiated by serum depletion in DMEM containing 0.5 mg/ml BSA and antibiotics. IGFs and/or LY294002 were added at the concentrations and times indicated for each experiment. Images shown are representative of 10–20 microscopic fields taken at random from each of at least four independent experiments.

Plasmids and Expression of Wild-Type and Mutant p85α in L6E9 Myoblasts

SRα-Wp85 and SRα-Δp85 were kindly provided by Dr. Masato Kasuga (Kobe University, Kobe, Japan). Wp85 was the entire coding sequence of bovine p85α. Δp85 encompasses a deletion mutant bovine p85α that lacks a binding site for the p110 catalytic subunit of PI 3-kinase. Both cDNAs were subcloned into SRα expression vector (24). The mutant p85α has a deletion of 35 amino acids (residues 479 to 513) and an insertion of two amino acids (Ser-Arg) replacing the deleted sequence. To obtain L6E9 myoblasts stably overexpressing Wp85 or Δp85, L6E9 cells were cotransfected with pcDNA3, a plasmid conferring geneticin resistance and either the SRα-Wp85 or the SRα-Δp85 plasmid.

For transfections and clone selection, subconfluent L6E9 cell monolayers (day 0) were pancreatinized and seeded 1:7 in two 25-cm² flasks. On day 1 (40–50% of confluence) cells were washed three times and then covered with 3 ml of serum-free medium. Cells were then transfected by adding dropwise 120 μl of DNA-Lipofectin mixture to each flask and swirling gently. DNA-Lipofectin mixture (1:1, vol/vol) was prepared with pcDNA3 together with Wp85 or Δp85 constructs in a 1:15 concentration ratio (45 μg total DNA/60 μl) and Lipofectin (30 μg/60 μl), following the supplier’s protocol (Life Technologies, Inc). Cells were incubated with DNA-Lipofectin-containing medium for 16 h under standard cell culture conditions. Medium was removed and replaced by complete medium (i.e. with 10% serum). Cells were grown to subconfluence, pancreatinized, and seeded in the presence of 0.4 mg/ml Geneticin (G418; Life Technologies, Inc, Gaithersburg, MD) to a very low density (1:200) so that single clones could be isolated by picking the clones with sterile pancreatin-embedded cotton swabs. G418-resistant clones were continuously grown in the presence of G418 (0.4 mg/ml). The culture time for transfected cells did not exceed the time for which the ability of L6E9 cells to differentiate is preserved. Screening of positive clones overexpressing p85 (Wp85 or Δp85) was performed by immunofluorescence assays using polyclonal rabbit antibodies against rat p85 PI 3-kinase (1:100) as primary antibody and rodamine-conjugated goat anti-rabbit Igs (1:100) as secondary antibody, as described below.

To quantify cell proliferation, cells were plated in multiwell culture dishes, grown from 1–4 days in 10% FBS-containing medium, and counted after pancreatinization.

Cell differentiation was analyzed in five independent immunofluorescence-positive clones of both Wp85- and Δp85-transfected L6E9 cells.

RNA Isolation and Northern Blot Analysis

Total RNA from cells was extracted using the phenol/chloroform method as described by Chomczynski and Sacchi (52). All samples had a 260:280 absorbance ratio above 1.7.

After quantification, total RNA (10 μg) was denatured at 65 C in the presence of formamide, formaldehyde, and ethidium bromide (53). RNA was separated on a 1% agarose/formaldehyde gel and blotted on Hybond N⁺ filters. The RNA in gels and filters was visualized with ethidium bromide and photographed by UV transillumination to ensure the integrity of RNA, to check the loading, and to confirm proper transfer. RNA was transferred in 10 × standard saline citrate (0.15 m NaCl and 0.015 m sodium citrate, pH 7.0).

Blots were probed with fluorescein-labeled probes prepared with the Gene Image (Amersham, Buckinghamshire, U.K.) random prime labeling module and were detected with the CDP-Star detection module (Amersham, Buckinghamshire, U.K.). The mouse cDNA probe for myogenin was a 1,100-bp EcoRI fragment.

Electrophoresis and Immunoblotting of Membranes

SDS-PAGE was performed as described by Laemmli (54). Proteins were transferred to Immobilon in buffer consisting of 20% methanol, 200 mm glycine, 25 mm Tris, pH 8.3. After transfer, the filters were blocked with 5% nonfat dry milk in PBS for 1 h at 37 C and then incubated overnight at 4 C with antibodies against GLUT4 (1:400) and β1-integrin (1:1000) in PBS containing 1% nonfat dry milk and 0.02% sodium azide. β1-integrin and PI 3-kinase-adapted subunits were detected using[ ¹²⁵I]protein A for 3 h at room temperature. GLUT4 and p110 were detected by ECL chemiluminiscence system (Amersham).

p85-p110α Complex Formation in Untransfected and Transfected L6E9 Cells

Cells were incubated in DMEM containing 0.2% BSA for 2 h before treatment with insulin to a final concentration of 1μ m (10 min at 37 C). After being washed twice in PBS solution, cells were scraped and solubilized for 30 min at 4 C in a buffer containing 50 mm HEPES, 150 mm NaCl, 10 mm EDTA, 10 mm Na₄P₂O₇, 100 mm NaF, 2 mm vanadate, 0.5 mm PMSF, 2 mm leupeptin, and 2 mm pepstatin, supplemented with 1% NP40 (buffer A). The solubilizates were centrifuged at 10,000 × g for 20 min at 4 C and 2.5 mg of the supernatants were immunoprecipitated with 5 μl of polyclonal antibodies against rat p85 or nonimmune serum as controls (not shown). Antibodies were preadsorbed on protein-G-Sepharose at 4 C for 1 h and washed twice in 30 mm HEPES, 30 mm NaCl, 0.1% Triton X-100, pH 7.4, before being incubated with the solubilized proteins for 90 min at 4 C. The immunopellets were washed three times in buffer A before being resuspended in SDS-PAGE sample buffer under reduction conditions and analyzed by Western blot using polyclonal antibodies against rat p110α as described above.

Immunofluorescence Analysis

For immunofluorescence labeling, cells were grown on glass coverslips. Coverslips were rinsed in PBS, fixed with methanol (−20 C) for 2 min, washed twice in PBS, and processed. Cells fixed on coverslips were incubated with 30 μl of primary antibody (1:100 anti-β₁-integrin, 1:100 polyclonal antibodies against rat p85 or 1:100 nonimmune serum controls in PBS containing 0.5% BSA) for 45 min at 37 C. Coverslips were washed three times in PBS, the last one for 15 min, before incubating with the secondary antibody (1:100 rodamine-conjugated goat anti-rabbit Igs in PBS containing 0.5% BSA) for 30 min at 37 C. Coverslips were then washed three times in PBS; the third wash was for 15 min in the presence of nuclear stain Hoechst 33342. Finally, coverslips were mounted with immunofluorescence medium. Confocal microscopy was performed at the confocal microscopy facility of the Serveis Cinentífico Tècnics of the Universitat de Barcelona.

Glucose Transport Measurements

Before transport experiments, cells were starved for 2 h in DMEM containing 0.5 mg/ml BSA and then treated or not with 1μ m insulin for 30 min. Cells were then washed and transport solution was added (20 mm HEPES, 150 mm NaCl, 5 mm KCl, 5 mm MgSO₄, 1.2 mm KH₂PO₄, 2.5 mm CaCl₂, 2 mm pyruvate, pH 7.4), together with 100 μm 2-deoxy-d-[³H]glucose (96 mCi/mmol). After 20 min, transport was stopped by addition of 2 vol of ice-cold 50 mm glucose in PBS. Cells were washed three times in the same solution and lysed with 0.1 n NaOH, 0.1% SDS. Radioactivity was determined by scintillation counting. Protein was determined by the Pierce method. Each condition was run in triplicate, and the nonspecific uptake (time zero) was determined by incubation of the 2-deoxy-d-[³H]glucose in stop solution (50 mm glucose in PBS) instead of transport solution. In all cases, time zero represented 4% of the basal transport activity at t = 20 min. Glucose transport under basal and stimulated conditions was linear over the time period assayed (data not shown).

We thank Mr. Robin Rycroft for his editorial support and Dr. Marta Camps, Dr. Ricardo Casaroli, and Susana Castel (Servei Científico-Tècnics, University of Barcelona) for expert advice in microscopy techniques. We are grateful to Mr. Quinzaños for art work.

This work was supported by research grants from the Dirección General de Investigación Científica y Técnica (PB92/0805; PB95/0971) from “Fondo de Investigación Sanitaria” (97/2101), Cost Action B5 and Generalitat de Catalunya (GRQ 94–1040), Spain. P.K. is supported by a postdoctoral fellowship from Comissió Interdepartamental i Innovació Tecnologica, Generalitat de Catalunya.