The Magnitude of Akt/Phosphatidylinositol 3′-Kinase Proliferating Signaling Is Related to CD45 Expression in Human Myeloma Cells¹

Multiple myeloma (MM)³ is a rapidly fatal plasma cell malignancy that evolves mainly in the bone marrow. MM is a very heterogeneous disease in that the clinical features and subsequent outcome are quite variable because the survival of patients may range from a few days to >15 years (1). IL-6 is known to be an essential growth and survival factor in this malignancy (2, 3). Besides this well-characterized growth factor, it is now clear that insulin-like growth factor-1 (IGF-1) also plays an important role in the growth and survival of human myeloma cell lines (HMCLs) (4, 5, 6). In MM, activation of the IGF-1 receptor (IGF-1R) induces both insulin receptor substrate-1 (IRS-1) and Shc activation that ultimately results in activation of both the PI3K/Akt and the MAPK signaling cascades (5). Indeed, activation of IRS-1 triggers its association with and activation of PI3K. Activation of PI3K leads to generation of phosphatidylinositol and subsequent to activation of Akt by phosphorylation on Thr³⁰⁸ and Ser⁴⁷³. Although the Thr³⁰⁸ phosphorylation strictly governs the activation of Akt, the Ser⁴⁷³ phosphorylation is necessary to obtain a fully active multiphosphorylated Akt enzyme (7). A significant support for a role of PI3K/Akt pathway as a potential mediator of tumor expansion in MM was recently demonstrated. Indeed, selective inhibition of the Akt pathway resulted in inhibition of MM cell proliferation (8) or in sensitization to apoptosis (9). It was shown that Ser⁴⁷³ Akt phosphorylation was frequently activated in MM cells and the frequency of activation correlated with disease activity (10). Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a recently identified tumor suppressor gene that encodes a phosphatase that mainly dephosphorylates phosphatidylinositol 3,4,5-trisphosphate and is an important negative regulator controlling the Akt activation (11). It was recently shown that loss of PTEN expression is responsible for uncontrolled Ser⁴⁷³ Akt phosphorylation in two HMCLs (12). The critical role of PTEN in tumor growth was clearly demonstrated for OPM-2 HMCL in SCID mice. The lack of PTEN expression in OPM-2 cells facilitated in vivo tumoral growth in 100% of SCID mice. Furthermore, tumor growth could be reversed in a dose-dependent fashion by increasing amount of PTEN (13).

IGF-1R has a high degree of homology with the insulin receptor, and their cytoplasmic signaling after receptor autophosphorylation appears to be similar. In the U266 HMCL, it has been demonstrated that CD45, a transmembrane protein tyrosine phosphatase, can negatively regulate insulin receptor signal transduction (14) and that the expression of CD45 is induced by IL-6 stimulation (15). In other cell types, such as HepG2 or kidney fibroblast cells, CD45 functions as a negative modulator of IGF-1 signaling (16, 17). Based on these data, it is conceivable that CD45 could act directly on IGF-1 signaling in MM cells. Thus, it appears that PI3K/Akt activation could be very heterogeneous in MM and under the influence of several phosphatases like PTEN and CD45. Because the role of PTEN in Akt activation was clearly demonstrated, the purpose of this study was to elucidate the impact of CD45 expression in the control of the PI3K/Akt pathway. The expression of CD45 in myeloma cells is directly related to their proliferation rate and differentiation status (18, 19). The heterogeneity of CD45 expression is also found on HMCLs, thus CD45⁻ and CD45⁺ HMCLs have been identified, leading to an analysis of PI3K/Akt pathway for each group of cell lines.

Anti-actin was purchased from Chemicon International (Temecula, CA). Anti-phospho-Akt (Thr³⁰⁸), anti-PTEN (A2B1), and anti-IGF-1Rβ are from Santa Cruz Biotechnology (Tebu-Bio, Le Perray en Yvelines, France.) Anti-phospho Akt (Ser⁴⁷³) and anti-phospho-p44/42 MAPK from Cell Signaling Technology (Ozyme, Saint Quentin Yvelines, France). Anti-STAT3 and anti-CD45 are from BD Biosciences (Le Pont de Claix, France). Wortmannin and LY294002 were from Alexis (Tebu-Bio). Human recombinant IGF-1 and vanadate were purchased from Sigma-Aldrich (St. Louis, MI).

The XG-1, XG-6, BCN, MDN HMCLs have been previously established in our laboratory and are cultured in the presence of 3 ng/ml r-IL-6 (Novartis Pharmaceuticals, Basel, Switzerland). LP-1, L363, and NCI-H929 HMCLs were purchased from Deutsche Sammlung von Mikroorganismen and Zellkulturen (Braunschweig, Germany) and U266 from the American Type Culture Collection (Manassas, VA). Cell lines were maintained in RPMI 1640 medium supplemented with 10% FCS, 2 mM glutamine, antibiotics, and 5 × 10⁻⁵ M 2-ME.

Cells (0.5 × 10⁶) were sampled in each condition and washed with PBS before incubation with anti-CD45-FITC (Immunotech, Marseilles, France) or IGF-1R-PE (BD Biosciences) for 20 min. After two washes, cells were fixed in 1% formaldehyde. Flow cytometry analysis was performed on a FACSCalibur using the CellQuest program (BD Biosciences, San Jose, CA). The ratio of fluorescence was determined by dividing the mean fluorescence intensity by the mean fluorescence intensity of the respective control.

Cell viability was determined by vital dye (0.4% eosin) exclusion and assessed by visual inspection in an hemocytometer.

The vanadate solution was prepared in incubation buffer (30 mM HEPES (pH 7.5), 150 mM NaCl, 4 mM KCl, 0.8 mM MgSO₄, 1.8 mM CaCl₂, 10 mM glucose). Pervanadate was obtained by mixing vanadate with 1 mM H₂O₂ for 15 min at 22°C. This was followed by the addition of catalase (200 μM) to remove residual H₂O₂.

A total of 4 × 10⁶ cells were resuspended in lysis buffer (10 mM Tris-HCl (pH 7.6), 150 mM NaCl, 5 mM EDTA, 1 mM PMSF, 2 mM Na₃VO₄, 1 mM NaF, 2 μg/ml aprotinin, leupeptin 1 μg/ml, and 0.5% Nonidet P-40). After 40 min on ice, lysates were cleared by centrifugation at 12,000 × g for 30 min at 4°C. Protein concentration was measured using bicinchoninic acid (BCA protein assay; Pierce, Rockford, IL). A total of 100 μg of proteins was loaded for each lane. The proteins were separated by 10% SDS-PAGE and then electrotransferred to polyvinylidene difluoride membranes. Western blot analysis was performed by standard techniques with ECL detection (Pierce).

Cells (20 × 10⁶) were lysed in 1% CHAPS-containing lysis buffer. Whole cell lysates were obtained, precleared with protein A-Sepharose, and then incubated overnight with 2 μg of the specific Ab. The immunocomplexes were captured with protein A-agarose. Beads were pelleted, washed three times, and boiled in SDS sample buffer. The presence of immune complexes was determined by Western blotting.

For cell cycle analysis, cells were fixed in 70% cold ethanol for 30 min at 4°C, washed twice in PBS, and stained with propidium iodide (20 μg/ml) for 10 min at room temperature. Flow cytometry analysis was performed on a FACSCalibur using CellQuest Pro software (BD Biosciences). Data were gated on the FL2-Area vs FL2-Width cytogram to excluded doublets and aggregates, and a minimum of 2 × 10⁴ gated cells were collected per sample. Analysis of the cell cycle was performed using the Modfit LT (BD Biosciences).

For statistical analysis, we used the nonparametric Wilcoxon rank sum test or the Student t test.

Given the potential role of CD45 as a negative regulator of IGF-1 signaling in HepG2 (14), we searched for a correlation between the expression of CD45 and the level of Akt-P in MM cells. We selected eight HMCLs to analyze Akt activation in response to IGF-1, three expressing CD45 on 100% of cells (XG-1, XG-6, MDN), one expressing CD45⁺ on a majority of cells (>90%) (U266), and four lacking CD45 expression (LP-1, NCI-H929, BCN, L363). Of note, all these cell lines express PTEN proteins as shown by Western blotting analysis using an Ab recognizing the C terminus of PTEN (Fig. 1 A). Following overnight serum starvation, we investigated Thr³⁰⁸ and Ser⁴⁷³ Akt phosphorylation (Akt-P) in response to IGF-1. The level of Akt-P was variable, the highest levels for both Ser and Thr phosphorylation were observed for the four CD45⁻ cell lines. In contrast, CD45⁺ HMCLs had a weak to a moderate level of Akt-P, especially for the Thr³⁰⁸ phosphorylation in which the difference is significant (p = 0.05). In conclusion, the level of Thr³⁰⁸ Akt-P appears to discriminate CD45⁻ from CD45⁺ cell lines.

Then, the functional impact of the expression of the IGF-1R on Akt-P was evaluated by determining the level of IGF-1R by flow cytometry. The results in Fig. 2 indicate that all cell lines express the IGF-1R. Whereas LP-1 has the highest level of IGF-1R (mean fluorescence ratio (MFR) = 11), and BCN and MDN have the weaker level (MFR = 1.8), the ratio of IGF-1R expression is almost identical for other cell lines (3.2 < MFR < 4). Thus, there is no discernable association between the level of IGF-1R expression and the level of Akt-P.

Pervanadate, a well-known powerful inhibitor of tyrosine phosphatases, inhibits CD45 phosphatase activity at the concentration of 100 μM (20). XG-1 (CD45⁺) and LP-1 (CD45⁻) were treated by pervanadate before IGF-1 stimulation. In XG-1, both Ser⁴⁷³ and Thr³⁰⁸ Akt-P were dramatically increased by 3- and 2-fold, respectively, whereas in LP-1 only a very weak increase was observed (Fig. 3). Thus, these results indicate that Akt-P is controlled by a phosphatase in CD45⁺ cell lines only.

Because IL-6 has been shown to induce CD45 expression on U266 (15), LP-1 cells were cultured with 10 ng/ml IL-6. After two months of culture with IL-6, LP-1 expressed CD45 on 50% of cells, as shown in Fig. 4,A. In LP-1 IL-6-treated cells, both Ser⁴⁷³ and Thr³⁰⁸ Akt-P was reduced by 50 and 25%, respectively, and conversely, Erk phosphorylation (Erk-P) was also reduced by 32% (Fig. 4 B). These data clearly indicate that PI3K/Akt and ERK/MAPK signaling in response to IGF-1 are reduced in CD45⁺ cell lines.

To compare the kinetic of Akt-P in response to IGF-1 in both CD45⁺ and CD45⁻ HMCLs, we selected two CD45⁺ (XG-6, XG-1) and two CD45⁻ (LP-1, L363) cell lines. After 30 min of IGF-1 stimulation, cells were washed and incubated in RPMI 1640 0.5% BSA for different times. In response to IGF-1, the Thr³⁰⁸ Akt-P was short lasting in both CD45⁺ and CD45⁻ HMCL but the persistency of the Ser⁴⁷³ Akt-P was very different between CD45⁺ and CD45⁻ cell lines (Fig. 5). Indeed, after 60 min the level of Ser⁴⁷³ Akt-P was almost undetectable in the CD45⁺ XG-6 and XG-1 cells (<3% of the initial signal), whereas 30% and 36% of the initial signal was still observed in CD45⁻ LP-1 and L363 cells, respectively. Moreover, in the CD45⁻ HMCLs, Ser⁴⁷³ Akt-P still remained detectable 3 h after IGF-1 stimulation. These results demonstrate that Ser⁴⁷³ Akt-P remained activated for a long time (>3 h) after IGF-1 stimulation in CD45⁻ HMCLs in contrast to CD45⁺ HMCLs, where Ser⁴⁷³ Akt-P returned to baseline level in around 1 h.

Because CD45 molecules modulate IGF-1 signaling, we wondered whether CD45 and IGF-1R could be physically associated. Immunoprecipitation assays were conducted using Abs directed to IGF-1R or STAT3 as a control in XG-6 cells. The results shown in Fig. 6 demonstrate that CD45 is coimmunoprecipitated with the IGF-1R, whereas CD45 is not found associated to STAT3. The physical interaction between IGF-1R and CD45 let us think that CD45 could directly dephosphorylate IGF-1R or linked molecules.

To evaluate the consequences of Akt activation, we next compared the efficiency of two PI3K inhibitors, i.e., wortmannin and LY294002 in NCI-H929 cells. Wortmannin induced a complete inhibition of Akt-P at the concentration of 1 μM, whereas the inhibition induced by LY294002 at 50 μM was not complete (Fig. 7). Consequently, wortmannin appeared more efficient than LY294002 to inhibit Akt-P confirming the results obtained by Davies et al. (21). Moreover, it was shown that LY294002 inhibits the Casein kinase 2 (21), a kinase involved in multiple pathways. Thus, we selected wortmannin to evaluate the importance of the PI3K pathway in myeloma cell proliferation and survival of three CD45⁻ cell lines (LP-1, NCI-H929, and BCN) and three CD45⁺ cell lines (U266, XG-1, and XG-6). Four-day wortmannin treatment does not induce significant apoptosis in any of the cell lines (data not shown). In contrast, wortmannin induces an important growth inhibition of CD45⁻ HMCLs of 70, 85, and 75% for LP-1, BCN, and NCI-H929, respectively, whereas this inhibition was weaker in CD45⁺ cell lines being 21, 25, and 25% for XG-6, XG-1, and U266, respectively (Fig. 8). The growth inhibition is significantly different between CD45⁺ and CD45⁻ cell lines (p < 0.001). This result indicates that the growth of CD45⁻ cell lines was mainly controlled by the PI3K pathway, whereas the growth of CD45⁺ cell lines was modestly controlled by the PI3K pathway and thus correlated well with the magnitude of Akt-P in response to IGF-1.

The effect of wortmannin on the cell cycle was investigated in XG-1 and LP-1. After 40 h of wortmannin treatment, the cell cycle was unchanged in XG-1 but impaired in LP-1 (Fig. 9). Indeed, in LP-1, S phase was reduced (38 vs 11%) (p < 0.001) whereas G₁ phase was significantly increased (61 vs 88%) (p < 0.001). Of note, no significant hypodiploid DNA peak was observed, indicating that wortmannin did not induce any apoptosis. Altogether, these results indicated that wortmannin induces a significant growth arrest in the CD45⁻ HMCLs only.

CD45, initially described as leukocyte common Ag, is a transmembrane protein-tyrosine phosphatase that plays a crucial role for both T and B cell activation through the corresponding Ag receptors (22). CD45 appears to be a positive mediator of cellular tyrosine phosphorylation by activating members of the src family protein-tyrosine kinases (23). The expression of CD45 has been shown to decline during normal plasma cell differentiation (24). In MM cells, the expression of CD45 is variable and correlates with the proliferation rate and differentiation status (18, 19, 25). In the 5T2MM mouse model, CD45⁻ cell populations have the most important invasive, proliferative, and angiogenic capacities (26, 27). As in this model, in humans, the loss of expression of CD45 is associated with advanced disease and reflects the phenotype of more malignant cells (19). Finally, we recently demonstrated that this CD45⁻ phenotype is a strong predictor of poor clinical outcome in humans (28), as in the 5T2MM myeloma model (26). In our study, we show that HMCLs reflect this heterogeneity because some are exclusively CD45⁺, some express both CD45⁺ and CD45⁻ subpopulations, and finally, most lack CD45. In MM, Ishikawa et al. (29) have demonstrated that CD45 is required for activation of src kinase in IL-6-dependent myeloma cell proliferation. However, several studies indicate that CD45 may also function as a negative modulator of tyrosine kinase signaling (14, 16, 17). In MM, CD45 is associated with decreased insulin receptor signaling (14). IGF-1 signaling is also dependent on tyrosine phosphorylation and dephosphorylation of tyrosine residues by protein tyrosine phosphatase that counteracts the signaling of IGF-1R. Indeed, it has been demonstrated that CD45 inhibited both IGF-1-dependent phosphorylation of IRS-1 and its association with PI3K, leading to decreasing Akt-P in mammary tumor cells (17). Our study extends these findings to MM cells. Indeed, Akt multiphosphorylation in response to IGF-1 is more important in the CD45⁻ than CD45⁺ myeloma cells. Primarily, the level of Thr³⁰⁸ phosphorylation that strictly governs the Akt activation is very high in CD45⁻ HMCLs, whereas its level is always weak in CD45⁺ HMCLs. Moreover, by using a phosphatase inhibitor, we demonstrate that a phosphatase triggers an important Akt dephosphorylation in CD45⁺ HMCL only. Furthermore, we show that the expression of CD45 induced by IL-6 on LP-1 cells correlates with a significant decrease of both Akt-P (Ser and Thr) and Erk-P, indicating that the two downstream pathways of IGF-1 signaling are down-regulated. Of interest, we demonstrated that Akt activation is longer in CD45⁻ than in CD45⁺ myeloma cells, suggesting that CD45 is involved in the dephosphorylation of IGF-1R as previously shown (17). Finally, we provide the first evidence of a physical association between the IGF-1R and CD45, indicating that CD45 decreases IGF-1 signaling in CD45⁺ myeloma cells by acting directly on IGF-1R or associated proteins. In a recent study, the authors have documented Ser⁴⁷³ Akt-P in purified myeloma cells and concluded that Akt can be constitutively activated in MM (10). However, they exposed myeloma cells to high serum concentration, and then they starved cells for only 2 h. Considering the kinetic of Ser Akt activation in CD45⁻ cells, we think that the observed Ser⁴⁷³ Akt-P is the result of a long residual Akt activation induced by high serum concentration rather than a true basal activation. Inhibition of the PI3K pathway by wortmannin induced a dramatic growth inhibition with a G₁ growth arrest of CD45⁻ cell lines but affected only weakly the CD45⁺ cell growth. Thus, this indicates that myeloma cell proliferation due essentially to PI3K pathway was clearly associated with the CD45⁻ phenotype. Of note, the CD45⁺ phenotype of HMCLs seems to be associated either with an autocrine production of IL-6 (U266) or an IL-6 dependence for their growth. This may indicate that the IL-6 signaling pathway exclusively controls the CD45⁺ cell proliferation. Finally, in a recent study it was proposed that cyclin D1 is involved in the regulation of HMCLs proliferation by the PI3K pathway (30). Indeed, it was shown that inhibition of PI3K by LY29400 in the NCI-H929 induced a dramatic down-regulation of cyclin D1. Because three of the four CD45⁺ cell lines express very high level of cyclin D1 due to their translocation t(11;14), this abnormal level of cyclin D1 may also contribute to the weak effect of wortmannin in these cell lines. In conclusion, the delimitation of two groups of HMCLs by CD45 expression has a functional relevance demonstrated by the different implication of the PI3K pathway in their proliferation and suggests that CD45 negatively regulates IGF-1-dependent activation of PI3K. Recent study provides evidence that inhibition of IGF-1R using selective small molecules represents a novel potential anticancer treatment (31). Thus, this strategy of blocking IGF-1R signaling and consequently the Akt/PI3K pathway could be a priority in the treatment of patients with MM, especially those lacking CD45 expression that have a very poor outcome (28).