Inward relocation of exogenous phosphatidylserine triggered by IGF-1 in non-apoptotic C2C12 cells is concentration dependent

The plasma membrane is composed of two leaﬂets that are asymmetric with regard to their phospholipid composition with phosphatidylserine (PS) predominantly located within the inner leaﬂet whereas other phospholipids such as phosphatidyl-choline (PC) are preferentially located in the outer leaﬂet. An intimate relationship between cellular physiology and the composition of the plasma membrane has been demonstrated, with for example apoptosis requiring PS exposure for macrophage recognition. In skeletal muscle development, differentiation also requires PS exposure in myoblasts to create cell– cell contact areas allowing the formation of multinucleate myotubes. Although it is clearly established that membrane composition/asymmetry plays an important role in cellular physiology, the role of cytokines in regulating this asymmetry is still unclear. When incubated with myoblasts, insulin-like growth factor I (IGF-1) has been shown to promote proliferation versus differentiation in a concentration dependent manner and therefore, may be a potential candidate regulating cell membrane asymmetry. We show, in non-apoptotic C2C12 cells, that relocation of an exogenous PS analogue, from the outer into the inner leaﬂet, is accelerated by IGF-1 in a concentration-dependent manner and that maintenance of membrane asymmetry triggered by IGF-1 is however independent of the PI3K inhibitor wortmannin. Copyright # 2005 John Wiley & Sons, Ltd.


INTRODUCTION
The two leaflets of the plasma membrane lipid bilayer, of normal non-apoptotic cells have been shown to be asymmetric with regard to phospholipid composition. For example, phosphatidylserine (PS) and phosphatidylethanolamine (PE) are preferentially accumulated in the inner leaflet whereas phosphatidylcholine (PC) and sphingomyelin (SM) are mainly confined to the outer leaflet. This non-random distri-bution of phospholipids was initially described in erythrocytes but has subsequently been observed in a number of cell types including myoblasts, fibroblasts, and Shawn cells. 1 Furthermore, studies have shown that addition of exogenous PS or PE to cells leads to the rapid (within a few minutes) relocation of these phospholipids from the outer to the inner leaflet of the plasma membrane thus promoting the membrane asymmetry, 2-4 following activation of an ATP-driven lipid transporter or aminophospholipid-translocase (flippase). [5][6][7] Nonetheless to date, it is unclear whether such ATP-driven transportation can be modulated by external cytokines.
Interestingly, when cells go through the apoptotic process a marker used for quantification is annexinV which interacts with PS in the outer leaflet of apoptotic cells. 8,9 Moreover, it has been shown that incubation with IGF-1 at relatively high concentrations promotes cell survival via the PI3K/Akt pathway and consequently relocation of PS from the outer to the inner leaflet of the plasma membrane. [9][10][11][12][13][14] Additionally, when cells are not apoptotic and incubated with IGF-1 at a high concentration, it has been shown that IGF-1 promotes cell proliferation. 14,15 During skeletal muscle cell differentiation, impairment of the membrane asymmetry has recently been suggested to play a key role in myotube formation. In fact, it has been proposed that when incubated in differentiation medium, myoblasts externalize their PS generating cell-cell contact areas required for myotube formation. [16][17][18] Support for this has been derived by adding annexinV to myoblasts in differentiation medium, which interacts with externalized PS and severely impairs the formation of multinucleate syncytial myotubes . These observations would therefore suggest that impairment of PS externalization in myoblasts should delay myotube formation, e.g. normal differentiation of myoblasts. 18 It is known that IGF-1 can promote either proliferation or differentiation in myoblasts in vitro dependent upon the concentration of this growth factor in the surrounding medium. It is also known that PS externalization is necessary for myotube formation and the progression of the differentiation pathway. However, although it is known that IGF-1 can promote cell survival, which is associated with outward relocation of PS in apoptotic cells, its role in the regulation of membrane asymmetry in non-apoptotic myoblasts has not been investigated. If at high concentrations IGF-1 were to inhibit outward relocation of PS in myoblasts and thus to impair cell-cell contact and myoblast fusion, then this would explain why this cytokine at high levels promotes proliferation as opposed to differentiation.
In this study we have therefore examined whether IGF-1 can accelerate the relocation of exogenous PS incorporated into the outer monolayer of the plasma membrane of non-apoptotic C2C12 myoblast cells and also, whether such a PS relocation is affected by IGF-1 concentration and the PI3K inhibitor Wortmannin.

Cells and materials
C2C12 cells were grown in DMEM-GlutamaxI containing penicillin (50 mU ml À1 ) and streptomycin (50 mg ml) supplemented with 10%FBS (Gibco BRL-Life Techonologies) and maintained at 37 C in 5% CO 2 . For all the experiments the cells were plated at a density of 10 5 cells per well, in 12-well plates precoated with collagen Type1 (Sigma). IGF-1 was purchased from Sigma.

Plasma membrane labelling with NBD-PS and NBD-PC
This method follows previous published and established protocols. 4,19,20 Short chain NBD-PC and NBD-PS (Avanti Polar lipids) were dried under nitrogen flow and incubated in serum-free medium at 37 C at a final concentration of 75 mM and sonicated for 1 min (Soniprep-150-MSE). One ml per well of the solution was subsequently added to cells for specific time periods. After three washes with PBS at 4 C, 500 ml of PBS-10% BSA(w/v) was incubated at room temperature for 20 s. The supernatant containing the fluorescent phospholipids was collected, and 500 ml of water was added. Phospholipid fluorescence intensity was determined using the Wallac Victor-1420 plate reader (excitation at 485 nm/emission at 535 nm). For the determination of total fluorescence, after the cold washes the cells were directly incubated in 500 ml water for 30 min to lyse the cells, which were then scraped off the wells. The solution was collected and 500 ml of PBS-10%BSA(w/v) was added and sonicated for 1 min to allow phospholipid interaction with BSA. Then phospholipid fluorescence intensity was determined using the same technique (plate reader). To measure whether NBD-PC and NBD-PS behave differently when incorporated within the outer leaflet of the plasma membrane, in a first experiment continuous incubation of either NBD-PS or NBD-PC phospholipid (75 mM) was performed at time points up to 30 min. At given time points (10,20 or 30 min), phospholipid in the outer leaflet of the plasma membrane was collected using the BSA exchange technique and the fluorescence proportional to the amount of phospholipid in the outer leaflet (N PS S or N PC S ) was measured. In a parallel experiment, using both the same protocol and the same time points, cells were scraped off and the overall cellular fluorescence proportional to the amount of phospholipid incorporated into the cellular membrane (N PS V or N PC V ) was measured for both phospholipids. Therefore, ratios N PS S / N PS V and N PC S /N PC V were plotted as a function of time, giving information on the location of each phospholipid analogue. To measure phospholipid internalization, NBD-PS or NBD-PC were dried under nitrogen flow and incubated in DMEM at a final 384 c. rauch and p. t. loughna concentration of 18.7 mM and sonicated for 1 min. One ml of the solution was subsequently added to individual wells for 3 min. After three washes with PBS at 4 C, cells were re-incubated in 1 ml of DMEM at 37 C for different times (0, 10, 20 or 35 min). At a given time point, extraction of phospholipids in the outer leaflet was carried out as described above. Each value of the intensity measured was plotted as follows: (I/I o À 1) Â 100, where I o is the fluorescence intensity measured for the control (0 min) and I the intensity measured at a given time (0, 10, 20 or 35 min), representing the internalization of phospholipids as a function of time. To inhibit the PI3K pathway, wortmannin (Sigma) was incubated at 1 mM for 30 min prior to NBD-phospholipid addition and also included in all the subsequent steps.

Data analysis
For histograms, data are represented by the mean of each experiment AE standard deviation. Student's t-test was performed to determine significance compared to control (*P < 0.05).

RESULTS AND DISCUSSION
In these studies we have used the well-characterized C2C12 myoblast cell line which does not become apoptotic in serum-free medium but over a period of several days, differentiates to form multinucleate myotubes. 21 Although it has been shown in a number of cell types that PS, but not PC, is actively translocated from the outer to the inner leaflet of the plasma membrane, such a process has not been described in myoblasts. Verification that exogenous PC and PS display different behaviours once inserted within the outer layer of the plasma membrane of C2C12 cells was first established using the BSA exchange technique which has been widely used to measure the rate of NBD-phospholipid translocation, using an NBD-PS analogue, from the outer to the inner leaflet of the plasma membrane in living cells. 4,19,22 This procedure has also been used to quantify plasma membrane internalization, via fluid phase endocytosis, using non-translocatable phospholipids such as NBD-glucosphingosine. 20 In this study, we used both a non-translocatable phospholipid NBD-PC analogue and a translocatable NBD-PS phospholipid analogue. To demonstrate that these phospholipids display different behaviours when incorporated into the plasma membrane of C2C12 myoblasts, cells were incubated with 75 mM of self-fluorescent phospholipid analogue. The amount of each phospholipid incorporated into the outer layer of the plasma membrane was measured at different time points and compared to that in the total cell using the BSA exchange technique. Measurement of the surface/volume ratio of the fluorescence intensity for each NBD-phospholipid ( Figure 1) demonstrated that NBD-PC accumulates mainly in the outer layer of the myoblasts whereas NBD-PS was equally distributed between the outer leaflet of the plasma membrane and the internal compartments of these cells. Such a discrepancy between the repartitioning of these two NBD-phospholipids suggests that there is an active clearance of NBD-PS from the outer leaflet probably involving a specific translocase (flippase). However, the enzymic conversion of NBD-phospholipids into lyso-derivates and free fatty acids over time cannot be excluded as a possible contributor to a loss of fluorescence. Nevertheless, as Figure 1 shows, over the short period of time examined in the present study, NBD-PC accumulated almost exclusively in the outer leaflet of the plasma membrane, where there is limited access to intracellular degradative enzymes. In contrast NBD-PS was translocated into the inner leaflet (cytosolic leaflet) of the plasma membrane where such enzymic degradation could take place. If such degradation were to occur, then the quantity of NBD-PS translocated when measuring the overall cell fluorescence would be underestimated and as such would also mean an underestimation of NBD-PS translocation, thus further emphasizing the conclusion that a potential translocase is involved in NBD-PS translocation. Prior to examination of the role of IGF-1 in NBD-PS relocation we compared the time course of internalization (outer layer clearance) for both NBD-PS and NBD-PC in the absence of this cytokine, with the PC phospholipid acting as a control to determine non-specific internalization, via endocytosis, as opposed to the active translocase activity of a flippase. To this end each phospholipid was incubated for 3 min and clearance measured as a function of time. The very low concentration of phospholipids used was chosen to minimize potential cellular responses due to phospholipid incorporation. Comparison of the relative amounts of NBD-PS or NBD-PC in the outer layer at different time points, to the initial quantity after an acute 3-min incubation, were plotted as a time course of internalization from the outer layer. The quantity of NBD-PS in the outer layer fell more rapidly when compared to NBD-PC over 35 min with approximately 80% of NBD-PS internalized as opposed to only 50% for NBD-PC (Figure 2), which is in agreement with measurements shown in Figure 1.
As anti-apoptotic pathways associated with IGF-1 have been shown to be activated in cells only at high concentrations, 11 we incubated the C2C12 myoblasts with 250 ng ml À1 of IGF-1 to determine the effect on internalization for both NBD-phospholipids ( Figure 3). Interestingly, within 10 min the NBD-PS internalization reached about 80%, whereas NBD-PC internalization was only 30%, the latter being not significantly different to the control levels (NBD-PC without IGF-1 Figure 2). Subsequently, NBD-PC internalization gradually increased to be not significantly different to the NBD-PS level at 35 min. These results suggest that IGF-1-induced NBD-PS clearance from the outer leaflet of the plasma membrane is independent of any plasma membrane internalization since after 10 min of acute stimulation using IGF-1, NBD-PC internalization was not significantly altered compared to the control whereas NBD-PS is almost entirely relocated into the inner leaflet of the plasma membrane. In conclusion, clearance of NBD-PC from the outer leaflet of the plasma membrane is more likely to be a product of fluid phase endocytosis (non-specific membrane internalization) in C2C12 myoblasts upon IGF-1 incubation. In contrast NBD-PS seems to be internalized both by endocytosis and the action of a translocase with the latter process accelerated by high concentrations of IGF-1.
As both anti-apoptotic and proliferative effects associated with IGF-1 incubation are known to occur only at high concentrations, we also examined the effects of a low IGF-1 concentration on NBD-PS relocation. Interestingly, incubation of IGF-1 at 25 ng ml À1 did not trigger the rapid relocation of NBD-PS (Figure 4). These results suggest that only at a high concentration of IGF-1 is there a rapid relocation of exogenous NBD-PS in non-apoptotic C2C12 myoblasts.
To determine whether the clearance of NBD-PS and NBD-PC is mediated through the PI3K pathway, we pre-incubated C2C12 cells with the PI3K inhibitor wortmannin for 30 min, prior to acute incubation with NBD-phospholipids and the measurement of NBDphospholipid analogue incorporation. Wortmannin Following three cold washes cells were re-incubated with DMEM-free medium at 37 C in the presence of IGF-1 at 250 ng ml À1 for different times. At a given time point, NBD phospholipids incorporated into the outer leaflet were removed using the BSA exchange technique. Then the quantity of phospholipids removed was quantified using spectrofluorimetry. Fluorescence intensity was plotted according to the procedure described for Figure 2 386 c. rauch and p. t. loughna incubation significantly reduced the clearance of both NBD-PS and NBD-PC from the outer leaflet over the complete time course studied ( Figure 5), when compared to the absence of this drug as shown in Figure 2. The fact that wortmannin altered the internalization of NBD-PC is in accordance with the known ability of wortmannin to inhibit the internalization of horseradish peroxidase, a marker of fluid phase endocytosis. 4,22,23 However the rate of clear-ance of NBD-PS was significantly greater than for NBD-PC between 10 and 35 min suggesting that while wortmannin dramatically retarded endocytosis its effect upon NBD-PS was less pronounced. To determine whether the relocation of NBD-PS under IGF-1 stimulation is linked to PI3K activity, IGF-1 was added at a high concentration (250 ng ml À1 ) following pre-incubation with wortmannin (1 mM). Figure 6 shows that in the presence of wortmannin, the characteristic rapid relocation of NBD-PS in response to IGF-1 incubation is still observed when compared to wortmannin alone ( Figure 5). This rapid induction of NBD-PS clearance over the initial 10 min with a subsequent plateau is qualitatively similar to that observed in response to IGF-1 in the absence of wortmannin ( Figure 3). Although the overall NBD-PS clearance from the outer leaflet is lower in response to IGF-1 with wortmannin ( Figure 6) than in response to IGF-1 alone, the increase in clearance when compared with that of NBD-PC is remarkably similar due to the dramatic decrease in clearance of the latter in the presence of wortmannin. These data strongly suggest that whereas internalization of both phospholipid analogues through endocytosis is PI3K dependent, this pathway does not play a significant role in the IGF-1-inducible translocation of NBD-PS. Membrane asymmetry has been shown to be involved and to control different key physiological processes such as endocytosis, by generating a   Prior to incubation with phospholipids, wortmannin (Wort.) was incubated with cells at 1 mM for 30 min. Then, NBD-PS and NBD-PC were incubated at 18.5 nmol per 10 5 cells ml À1 , containing 1 mM of wortmannin for 3 min. Following three cold washes cells were reincubated with DMEM-free medium at 37 C in the presence of both wortmannin at 1 mM and IGF-1 at 250 ng ml À1 for different times. At a given time point, NBD phospholipids incorporated into the outer leaflet were removed using the BSA exchange technique. Then the quantity of phospholipids removed was quantified using spectrofluorimetry. Fluorescence intensity was plotted according to the procedure described for Figure 2 igf-1-dependent activation of ps relocation 387 difference of surface pressure between leaflets, 4,19,24 the regulation of cytokine-dependant genetic expression through destruction of the cytokine/receptor complex within internal compartments, 25 the clearance of apoptotic cells through macrophage activation 8 and myotube formation through cell-cell contact areas. [16][17][18] Therefore, membrane asymmetry is an integral part of cellular physiology and must be consistent with the subcellular processes involved.
In this present paper, we have shown that in response to acute IGF-1 exposure, a PS analogue is rapidly relocated from the outer to the inner leaflet of the plasma membrane, suggesting that this cytokine may directly or indirectly act on a phospholipid translocase, and for the first time have demonstrated that PS relocation is increased under IGF-1 stimulation in non-apoptotic cells. Moreover, the relocation is also dependent on IGF-1 concentration, as a low concentration does not produce any change in PS analogue relocation when compared to the control. Additionally, the results demonstrate that rapid PS analogue relocation induced by IGF-1 is not altered by wortmannin suggesting the PI3K pathway is not involved in the maintenance of membrane asymmetry under such stimulation. Alternative signalling pathways by which IGF-1 could mediate its effects upon membrane asymmetry are currently under investigation. Therefore, as IGF-1 promotes proliferation at high concentration in non-apoptotic cells, we propose that such an effect is in accordance with a lack of PS externalization impairing the creation of cell-cell contact areas needed for myoblast fusion, and therefore multinucleate myotube formation.