Ewing's sarcoma (ES) is an extremely aggressive, poorly differentiated neoplasm of uncertain histogenesis,1 genetically characterized by the presence of specific chromosomal translocations that produce EWS/ets gene rearrangements [in more than 95% of cases, the gene fusion is EWS/FLI-1, due to the t(11;22)(q24;q12), or EWS/ERG, due to the t(21;22)(q22;q12)].2 ES occurs primarily in late childhood and adolescence and, despite the use of multimodal treatments and very aggressive chemotherapeutic regimens,3, 4, 5 frequently shows a severe clinical history. The observation that one-third of patients with nonmetastatic disease and the great majority of patients with metastases at diagnosis will not survive, regardless of the therapy, is frustrating. Moreover, recent clinical studies have indicated that the survival rate of ES patients has reached a plateau phase and, very likely, the highest levels achievable by conventional therapy.3, 4 The identification of valuable new therapeutic targets for the design of innovative, more effective strategies is therefore urgently needed for this tumor.
In recent years, a better understanding of the control of ES cell proliferation, differentiation and death has been achieved.6 Although several growth factor circuits7, 8, 9 appeared to be involved in deregulated ES tumor cell growth, insulin-like growth factor I (IGF-I) and its corresponding receptor (IGF-IR) were found to be of major importance.9, 10, 11 IGF-IR pathway is necessary for transformation of cells transfected with EWS/FLI-1 cDNA.12 The blockage of IGF-IR-mediated circuit was found to reduce effectively the tumorigenic and metastatic ability of ESFT cells in athymic mice and increase the effectiveness of conventional cytotoxic drugs,13, 14, 15, 16 therefore representing a valuable therapeutic approach against ES. However, these approaches showed limited possibilities for a prompt application in ES therapy. In fact, despite the promising findings obtained in preclinical conditions, the delivery of a murine antibody is of limited practical value in clinical settings, and antisense IGF-IR and dominant negative mutants may represent the future rather than the present.17 In an effort to identify an attractive and suitable IGF-IR-related target for pharmacologic intervention in ES, we here analyzed the contribution of 2 major pathways of the intracellular IGF-IR signaling cascade,18 the mitogen-activated protein kinase (MAPK) and phosphatidylinositol-3-kinase (PI3-K) signaling pathways, to the overall effects elicited by IGF-I in ES. In addition to these 2 primary pathways, IGF-IR also activates other signaling cascades such as protein kinase-C-delta19 and signal transducers and activators of transcription, particularly Stat3.20, 21, 22, 23 All these pathways may interact at different levels in mediating the different IGF-I responses, whether proliferative, transforming, antiapoptotic, differentiating, or chemotactic.24, 25, 26, 27, 28, 29, 30, 31, 32 In this study, we analyzed the biologic consequences, with respect to proliferation, apoptosis, migration, and chemosensitivity, of inhibiting the MEK/MAPKs and PI3-K pathways in the proper context of ES cells by using the selective inhibitors PD98059 or U0126 (specific inibitors of MEK)33, 34 and LY294002 (specific inhibitor of PI3-K).35
MATERIAL AND METHODS
A panel of 8 human ES cell lines were analyzed. The ES TC-71 cell line was a generous gift from T.J. Triche (Children's Hospital, Los Angeles, CA). The ES H825 cell line was kindly provided by Dr. A. Llombart-Bosch (Department of Pathology, University of Valencia, Spain). The ES SK-ES-1, RD-ES and SK-N-MC and the human breast cancer SK-BR-3 cell lines were obtained from American Type Culture Collection (Rockville, MD). The ES LAP-35, IOR/BER and IOR/NGR cell lines were established at the Laboratorio di Ricerca Oncologica, Istituti Ortopedici Rizzoli, Bologna, Italy, and previously characterized.36 Cells were routinely cultured in Iscove's modified Dulbecco's medium (IMDM) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin (Life Technologies, Invitrogen, Carlsbad, CA) and 10% inactivated fetal bovine serum (FBS; Biowhittaker Europe, Verviers, Belgium) and maintained in 37°C in a humidified 5% CO2 atmosphere.
In vitro growth
To study the effects of specific inhibitors of intracellular signaling pathways in standard or low-serum conditions, cells were plated into 24-well plates (cells/well: 50,000 for TC-71 and SK-N-MC and 100,000 for LAP-35) in IMDM plus 10% FBS. After 24 hr, medium was replaced by IMDM plus 10% FBS or 1% FBS with or without (control) various concentrations of the PI3-K inhibitor LY294002 (1, 10 μM; Calbiochem, Alexandria, Australia) and/or MEK/MAPK inhibitors PD98059 or U0126 (10, 50 μM; Calbiochem). Cell growth was evaluated on harvested cells by Trypan blue vital cell count. Two independent researchers performed the count in a blinded manner. To evaluate the ability of ES cell recovery from the cytotoxic effects of the selective inhibitors, cells treated with 10–50 μM of the inhibitors for 48 hr were exposed to low-serum medium for up to 6 days and counted at different times. To evaluate the effects of the PI3-K and/or MAPK inhibitors in IGF-I-mediated conditions, experiments were performed in low-serum conditions. A total of 20,000 cells/cm2 of TC-71 cells were seeded in IMDM plus 10% FBS. After 24 hr, medium was replaced by IMDM 1% FBS with or without (control) LY294002 (10 μM), PD98059 (50 μM) or a combination of both the inhibitors in the presence or not of human recombinant IGF-I (100 ng/ml; United Biomedical, Lake Placid, NY). Cell growth was evaluated 48 hr following the treatments by Trypan blue vital cell count.
Cell cycle analysis
A total of 20,000 cells/cm2 of TC-71 were seeded in IMDM plus 10% FBS. The day after, medium was changed in IMDM plus 10% FBS or IMDM plus 1% FBS without (control) or with LY294002 (1, 10 μM), PD98059 (10, 50 μM) or with a combination of both the inhibitors (LY294002 10 μM plus PD98059 50 μM) in the presence or not of IGF-I (100 ng/ml; United Biomedical). After 24–72 hr of treatments, cell cultures were incubated with 10 μM bromodeoxyuridine (BrdUrd; Sigma, St. Louis, MO) for 1 hr in CO2 atmosphere at 37°C. Harvested cells were fixed in 70% ethanol for 30 min. After DNA denaturation with 2 N HCl, 1 × 106 cells were processed for indirect immunofluorescence staining using α-BrdUrd monoclonal antibody (MAb) diluted 1:4 as a primary antibody (Becton Dickinson, San Jose, CA) and analyzed by flow cytometry (FACScan; Becton Dickinson). For the analysis of DNA content, cells were fixed with cold 70% ethanol, treated with 0.5 mg/ml RNase and stained with 20 μg/ml propidium iodide.
Analysis of apoptosis
For morphologic assessment of apoptotic nuclei, ES cell lines were seeded in 60 mm dishes in IMDM plus 10% FBS. The following day, medium was changed in IMDM plus 10% FBS or IMDM plus 1% FBS without (control) or with LY294002 (1, 10 μM), PD98059 (10, 50 μM), or with a combination of inhibitors, in the presence or not of IGF-I (100 ng/ml; United Biomedical). The p38 MAPK inhibitor SB203580 (50 μM; Calbiochem) was also used. Twenty-four to 72 hr from treatment, cells were fixed in methanol/acetic acid (3:1) for 15 min and stained with 50 ng/ml Hoechts 33258 (Sigma). Cells with 3 or more chromatin fragments were considered apoptotic. The percentage of apoptotic nuclei was evaluated out of 1,000–2,000 nuclei. Detection and quantification of apoptotic cells was also obtained by the flow cytometric analysis of annexin V-FITC-labeled cells. This test was performed according to the manufacturer's instructions. Annexin V is a Ca2+-dependent phospholipid protein with high affinity for phosphatidylserine (PS). This protein can hence be used as a sensitive probe for PS exposure on the outer layer of the cell membrane and is therefore suited to detect apoptotic cells. Since necrotic cells also expose PS according to the loss of membrane integrity, the simultaneous application of propidium iodide (PI) as a DNA stain is required to discriminate necrotic from apoptotic cells. The effects of the selective inhibitors were evaluated in the presence or not of IGF-I after 24–48 hr.
Soft agar assay
Anchorage-independent growth was determined in 0.33% agarose (SeaPlaque; FMC BioProducts, Rockland, ME) with a 0.5% agarose underlay. ES cell suspensions (cells/60 mm Oslash; dish: 3,300–10,000 for TC-71 and SK-N-MC; 33,000–100,000 for LAP35) were plated in a semisolid medium (IMDM plus 10% or 1% FBS containing 0.33% agarose) with or without selective inhibitors (10 μM of LY294002, 50 μM of PD98059, 50 μM of U0126 or 50 μM of SB203580) in the presence or not of IGF-I (100 ng/ml; United Biomedical). Dishes were incubated at 37°C in a humidified atmosphere containing 5% CO2; colonies were counted after 7–10 days.
Expression of IGF-IR was determined by flow cytometry (FACScan; Becton Dickinson) after indirect immunofluorescence with the primary antibody αIR3 (Oncogene Research Product, San Diego, CA) diluted 1:10 and a fluorescein isothiocyanate-conjugated antimouse immunoglobulin antiserum (Kierkegaard and Perry Laboratories, Gaithersburg, MD) diluted 1:20 as secondary antibody.
A total of 20,000 cells/cm2 were seeded in IMDM 10% FBS. After 72 hr, supernatants were collected and the production of IGF-I in the conditioned medium of ES cells was measured using the Quantikine human IGF-I immunoassay according to the manufacturer's instructions (R and D System, Minneapolis, MN). Standard culture medium (IMDM 10% FBS) was analyzed as a control. IGF-I production by ES cells was also determined by using serum-free medium. Supernatants were collected after 72 hr from cell seeding.
Motility assay was performed using Transwell chambers (Costar, Cambridge, MA) with 8 μm pore size, polyvinylpyrrolidone-free polycarbonate filters (Nucleopore, Pleasanton, CA). IMDM plus 10% FBS alone or IMDM plus 10% FBS with IGF-I (100 ng/ml; United Biomedical) were placed in the lower compartment of the chamber; 105 ES cells were resuspended in IMDM plus 1% FBS with or without the MEK/MAPK or PI3-K inhibitors (50 μM PD98059, 10 μM LY294002, or a combination of both the compounds) and then seeded in the upper compartment. Chambers were incubated at 37°C in a humidified atmosphere containing 5% CO2 for 12 hr. Cells migrated toward the filter to reach the lower chamber base were counted after Giemsa staining. All the experiments were made in triplicate. Cell viability for the inhibitors at the used concentrations was analyzed and found to be ≥ 90%.
Combined treatments with doxorubicin
A total of 200,000 TC-71 cells were seeded in 6-well plates in IMDM plus 10% FBS. The following day, cells were treated with varying concentrations of doxorubicin (Sigma; range, 1–30 ng/ml) alone or in association with LY294002 (10 μM) or PD98059 (10 μM). After 72 hr of treatment, cell growth was evaluated by Trypan blue vital cell count. To analyze the effects of combined treatments in anchorage-independent conditions, 3,300 TC-71 cells were plated in 60 mm Oslash; dishes containing semisolid medium (IMDM 0.33% agarose plus 10% FCS) with or without varying concentrations of doxorubicin (range, 3–30 ng/ml) alone or in association with LY294002 (10 μM) or PD98059 (10 μM). Dishes were incubated at 37°C in a humidified atmosphere containing 5% CO2, and colonies were counted after 7 days.
Constitutive activation of MAPK or PI3-K pathways was evaluated on ES cell lines grown for 72 hr in complete or serum-free medium. To analyze the effects of the neutralizing IGF-IR antibody αIR3 as well as of the signaling pathway inhibitors in standard or low-serum conditions, semiconfluent TC-71 cells were treated for 2 hr with 1 μg/ml of the αIR3 antibody or with LY294002 (10 μM) and/or PD98059 (50 μM), respectively. For evaluation of IGF-I effects, starved cells (18 hr of serum starvation) were treated with IGF-I (100 ng/ml; 5-–60 min) in the presence or not of selective inhibitors. To determine the phosphorylation status of Erk and Akt, the downstream mediators of MEK/MAPK and PI3-K pathways, cells were washed twice with ice-cold PBS and cell lysates were prepared with a buffer containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% sodium dodecyl sulfate (SDS), 1% TritonX-100, 5 mM EDTA, 1% deoxycholate and protease inhibitors (1 mM phenylmethylsulfonyl flluoride, 1 mM sodium orthovanadate). Protein concentration was determined by Bio-Rad protein assay (Bio-Rad, Hercules, CA) and equivalent amounts of total cell lysate (50 μg) were separated by 12% SDS-polyacrylamide gel electrophoresis under denaturating conditions and transferred onto nitrocellulose membrane. Membranes were blocked for 1 hr at 25°C with Tris-buffered saline-Tween 20 (TBST) containing 5% nonfat dry milk and incubated overnight with primary antibody antiphospho-Akt (Ser473), antiphospho-p44-42 MAPK (Thr202/Tyr204; dilution 1:1,000), or antiphospho-SAPK/JNK (thr183/tyr185; dilution 1:1,000; New England Biolabs, Cell Signaling Technology, Beverly, MA). Analysis of Akt and Erk was also performed to verify the total proteins as control (primary antibodies for Akt or Erk diluted 1:1,000; New England Biolabs, Cell Signaling Technology). After washes with TBST, membranes were incubated with secondary antirabbit antibody conjugated to horseradish peroxidase (dilution 1:1,500; Dako, Glostrup, Denmark) and revealed by ECL Western blotting detection reagents (Amersham, Amersham Place, Buckinghamshire, U.K.).
Differences between means were analyzed using a 2-sided Student's t-test. The analysis of drug combination effects was performed by using the fractional product method.
Involvement of PI3-K or MEK/MAPK pathways in ES
Recently, the constitutive activation of the MAP kinases Erk 1 and Erk 2 and the survival factor Akt, a downstream mediator of PI3-K pathway, has been demonstrated in a variety of human cancers, including ES.37, 38, 39 Accordingly, we found activation of both pathways in a panel of ES cell lines maintained in standard medium (IMDM 10% FBS; Fig. 1a). Constitutive activation of Erk1/2 was also clearly evident in serum-free conditions, whereas Akt phosphorylation seemed to be more serum-dependent. Constitutive activation of intracellular pathways, at least the MAPK pathway, is not surprising since all cell lines express IGF-IR on cell surface (Fig. 1b) and produce IGF-I in the supernatants (Fig. 1c). The amount of IGF-I in ES-conditioned IMDM 10% FBS is much higher than that present in medium alone, indicating that ES cells are constantly exposed to IGF-I stimulation independently from the presence of serum. In serum-free medium, IGF-I production by ES cells was lower but still detectable in all the ES cell lines except SK-ES-1 and IOR/BER (data not shown). Treatments with neutralizing IGF-IR antibody αIR3 abolished constitutive and IGF-I-induced activation of Erk1/2 and Akt mediators, further supporting the idea that their status is related to the constant presence of functionally active IGF-IR on ES cells (Fig. 2a). Despite the presence of the autocrine loop, ES cells maintained the ability to respond to exogenous IGF-I by moderately increasing their proliferation.13 Exposure of starved TC-71 cells to exogenously added IGF-I (100 ng/ml) induced a marked phosphorylation of Erk1/2 (maximum stimulation was observed after 5 min) and Akt (continuous increase in Akt phsphorylation was observed up to 60 min, the last time point evaluated; Fig. 2b). Figure 2(c) shows that both pathways are involved in IGF-IR-mediated intracellular signaling. Treatments with 10 μM of LY294002 completely abolished constitutive and/or IGF-I-induced phosphorylation of Akt. Similarly, exposure to 50 μM of PD98059 completely and selectively blocked phosphorylation of Erk1/2. Inhibition of JUNK pathway by PD98059, but not by LY294002, was also observed in either basal or IGF-I-mediated condition (Fig. 2c).
Effects of PI3-K or MEK/MAPK inhibitors on ES in vitro growth
To determine the effects of LY294002 or PD98059 on ES cell proliferation, 3 ES cell lines were treated with different concentrations of the inhibitors. A remarkable dose-dependent growth inhibitory effect of LY294002 or PD98059 was observed in low-serum medium (data not shown) as well as in 10% FBS-containing medium (Fig. 3). The growth inhibitory activity of the 2 compounds was maintained for at least 72 hr after their removal (data not shown).
PI3-K or MEK/MAPK inhibition not only impaired proliferation of ES cells in monolayer conditions but also significantly impacted their ability to grow in anchorage-independent conditions, i.e., colony formation in soft agar, which is an accepted criterion for transformation40 and an experimental condition that better represents tumor cell growth in vivo39 (Table I). Simultaneous inhibition of the 2 pathways highly reduced but did not completely abolish the ability of ES cells to form colonies in soft agar, indicating that, besides these 2 pathways, others may play a role in anchorage-independent growth of ES cells. p38 MAPK is probably involved since the use of the selective inhibitor SB203580 (50 μM)41 induced a significant reduction of the number of colonies in all 3 ES cell lines (Table I). The role of p38 MAPK appeared to be more relevant in one cell line (SK-N-MC) compared to the others.
Table I. Growth in Soft AGAR of ES Cells After Selective Inhibition of PI3-K, MEK/MAPK or P38 MAPK Signaling Pathways
Cells were seeded at a concentration of 3,300 cells for TC-71, 10,000 cells for SK-N-MC and 100,000 cells for LAP-35. The number of colonies was determined after 7–10 days of growth in 10% FBS. Data are expressed as means of 4–6 plates ± SE.
Student's t-test, with respect to untreated cells.
1,694 ± 89
LY294002 10 μM
750 ± 39
PD98059 50 μM
754 ± 86
LY204002 + PD98059
113 ± 57
SB203580 50 μM
900 ± 98
359 ± 15
LY294002 10 μM
144 ± 55
PD98059 50 μM
252 ± 21
LY204002 + PD98059
129 ± 31
SB203580 50 μM
58 ± 5
2,158 ± 170
LY294002 10 μM
861 ± 110
PD98059 50 μM
1,339 ± 101
LY204002 + PD98059
644 ± 47
SB203580 50 μM
1,568 ± 76
Two major mechanisms may explain the inhibitory effects of LY294002 and PD98059 on ES cell proliferation: inhibition of cell cycle progression and/or induction of apoptosis. Figure 4 shows the data obtained for TC-71 cell line. Similar results were also observed for the other 2 ES cell lines (data not shown). Both LY294002 and PD98059 inhibited cell cycle progression in a dose-dependent manner, inducing specific G1 arrest (Fig. 4a). With regard to apoptosis (Fig. 4b), only LY294002 gave significant induction of the apoptotic rate, indicating a minor role of MEK1/MAPK pathway in sustaining survival of ES cells.
To establish the involvement of PI3-K and MEK/MAPK pathways in IGF-I-stimulated ES proliferation and to verify the ability of IGF-I to rescue ES cells from apoptosis and cell cycle blockage, we exposed ES cells to exogenous IGF-I in the presence or not of selective inhibitors of PI3-K and MAPK cascade. In low-serum conditions, IGF-I induced a 22% increase in the number of TC-71 cells after 48 hr of treatment. As previously shown, LY294002 and PD98059 both significantly reduced the growth of TC-71 in basal conditions (60% and 61.7% of growth inhibition, respectively). However, growth inhibition induced by LY294002 was reverted by exposure of cells to IGF-I, percentages shifting back toward control values, whereas the factor did not appear to rescue the effects of PD98059 (Fig. 5). Simultaneous exposure of TC-71 cells to 10 μM of LY294002 and 50 μM of PD98059 gave a 79% of growth inhibition. In these conditions, IGF-I did not exert any significant rescue of ES cell growth, indicating that no other major pathways are involved in the regulation of IGF-I-mediated ES growth and that MEK/MAPK pathway is the most relevant intracellular cascade for sustaining proliferation of ES cells. Similar results were also obtained in soft agar conditions (Table II). The number of colonies observed after exposure of cells to IGF-I and both inhibitors was higher than that obtained with the 2 inhibitors in basal conditions but the percentage of inhibition was still similar to that obtained with the use of PD98059 alone, further indicating that IGF-I was able to rescue the inhibitory effects of LY294002 but not those of PD98059. Similar results were also obtained when a different selective inhibitor of MEK-1, U0126, was used (data not shown).
Table II. Soft-AGAR Growth of TC-71 ES Cells After Selective Inhibition of PI3-K and MEK/MAPK Signaling Pathways in the Presence of IGF-I
Cells were seeded at a concentration of 3,300 cells for TC-71 cell line. The number of colonies was determined after 7 days of growth in 1% FBS. Data are expressed as means of 4–6 plates ± SE.
Student's t-test, with respect to untreated cells.
1,595 ± 120
IGF-I 50 ng/ml
2,003 ± 64
IGF-I + LY294002 10 μM
1,446 ± 105
IGF-I + PD98059 50 μM
619 ± 134
IGF-I + LY204002 + PD98059
440 ± 22
With respect to cell cycle, IGF-I exposure induced a significant abrogation of the LY294002-induced G1 arrest, whereas the blockage was maintained when cells were treated with PD98059 and IGF-I (Table III). In the presence of both inhibitors, IGF-I was unable to revert G1 blockage, confirming that no other major intracellular pathways were involved in IGF-I regulation of ES cell cycle progression.
Table III. Effects, on Proliferative Rate of TC-71 Cells After Selective Inhibition of PI3-K and MEK/MAPK Signaling Pathways in the Presence or not of IGF-I
With regard to apoptotic effects, IGF-I completely rescued the proapoptotic effect of LY294002 (Fig. 6a). Exposure to PD98059 alone or to IGF-I and PD98059 only slightly affected apoptosis of TC-71, further indicating the minor role of MEK/MAPK pathway in the regulation of apoptosis of ES cells both in basal and IGF-I-stimulated conditions. When cells were exposed to both inhibitors, the apoptotic rate was significantly higher than controls but only slightly higher than that observed with LY294002 alone. The presence of IGF-I completely abrogated the apoptotic induction observed when cells had both PI3-K and MEK/MAPK signaling pathways blocked, strongly suggesting the existence of another cascade of intracellular events activated by IGF-I. This pathway, which completely reintroduced the IGF-I protective action against apoptosis, did not seem to involve p38MAPK,32, 42 since TC-71 ES cells treated with SB203580 (50 μM) showed an apoptotic rate similar to control in basal and IGF-I-exposed conditions (Fig. 6a). Moreover, IGF-I was still able to rescue the proapoptotic effects of LY294002 when cells were treated with LY294002 and SB203580 or LY294002 and SB203580 and PD98059 (Fig. 6).
Effects of PI3-K or MEK/MAPK inhibitors on ES in vitro migration
Migratory ability of TC-71 and SK-N-MC was significantly stimulated by IGF-I (p < 0.01, Student's t-test; Fig. 7). Treatments with PD98059 significantly inhibited the migratory ability of ES cells both when complete medium alone was used as a chemoattractant or when IGF-I was added, indicating a role for MAPK signaling pathway in the basal and IGF-driven migration of ES cells. In contrast, PI3-K signaling pathway appeared to be involved only in IGF-I-mediated migratory ability of ES cells (Fig. 7). The simultaneous exposure to both inhibitors showed no advantage with respect to the use of PD98059 alone (data not shown).
Growth inhibitory effects of PI3-K or MEK/MAPK inhibitors in combination with doxorubicin
TC-71 cells were simultaneously exposed to increasing concentrations of doxorubicin and to a concentration of inhibitors that gave around 30% growth inhibition after 72 hr. The combined treatments with doxorubicin and LY294002 or PD98059 resulted in a significantly enhanced inhibition of TC-71 cell growth with respect to the therapeutic efficacy of doxorubicin alone (Fig. 8a). According to the fractional product method, an additive cytotoxic effect was observed with both the inhibitors,. The IC50 dose of doxorubicin, the concentration of drug that reduces growth of 50% of untreated control cells, was 3-fold reduced: 10.76 ng/ml when doxorubicin was used as a single agent, 3 ng/ml when doxorubicin was used in combination with LY294002, 3.7 ng/ml when doxorubicin was used in combination with PD98059. Experiments in anchorage-independent conditions showed similar results (Fig. 8b).
IGF-IR-mediated circuit is a major autocrine loop for ES cells and appears to be particularly important in the pathogenesis of this tumor.9, 10, 11, 12 Several reports9, 10, 11, 12, 13, 14, 15, 16 have clearly indicated that inhibition of IGF-IR may be implicated in the clinical treatment of ES patients, provided that effective strategies are developed to inactivate the receptor-mediated functions. In an effort to identify which pathway may be better exploited for therapeutic purposes, we analyzed the constitutive and IGF-I-activated contribution of MAPK and PI3-K signaling pathways to proliferation, apoptosis, migration and chemosensitivity of ES cells. Recently, a large number of small molecule inhibitors targeting these 2 pathways have been developed; of these, the ones that are currently under phase 1 or 2 clinical trails43 represent reasonable prospects for innovative treatments.
The actions of IGFs derive primarily from the activation of IGF-IR. Although other intracellular pathways have recently been described as activated by IGF-IR,19, 21, 22, 23 2 major different cascades have clearly emerged: one pathway involves Erk1/2 through Ras/Raf/MEK, the other proceeds through PI3-K.18, 25 ES cells produce IGF-I and show an autocrine IGF-IR-mediated loop. Consistently, cells showed a clearly detectable constitutive activation of Erk1/2 and Akt, 2 crucial mediators of the MAPK and PI3-K signaling pathways, respectively. Treatments with the neutralizing IGF-IR antibody αIR3 abolished the constitutive activation of the 2 mediators, supporting the idea that their status is related to the constant presence of functionally active IGF-IR on ES cells. Although IGF-IR was originally shown to drive cell proliferation through MAPK and survival through PI3-K, this division has recently emerged as being oversimplified.44, 45 In ES, our findings indicated that MAPK and PI3-K were both involved in sustaining proliferation of ES cells either in monolayer or in soft agar conditions. PD98059 and LY294002 inhibited ES cell proliferation and cell cycle progression by inducing G1 blockage. By contrast, only PI3-K but not MAPK or p38 MAPK was involved in the regulation of apoptosis in ES cells. Supporting the therapeutic potential is the fact that the growth inhibitory effects of the 2 signaling inhibitors were reversible but maintained for at least 72 hr after their removal.
Since all Ewing's sarcomas express IGF-IR9, 10, 11 and IGF-I is stored in bone matrix,46 which could be released by osteolysis induced by growing ES, it can be envisaged that ES cells are locally exposed to paracrine, in addition to autocrine, stimulation by IGF-I. Although it is obviously difficult to verify whether this really occurs in vivo, the analysis of the effectiveness of MAPK and PI3-K inhibitors in IGF-I-mediated conditions seems to be more representative of the in vivo condition. In fact, despite the existence of an autocrine loop, ES cells remained sensitive to exogenous IGF-I.
In the presence of IGF-I, in vitro growth blockage due to inhibition of PI3-K signaling pathway was significantly reverted. IGF-I induced G1-arrested cells to reenter the cell cycle and abolished apoptosis induced by LY294002. The protective action of IGF-I from apoptosis driven by LY294002 was clearly due to the existence of an alternative signaling pathway not involving MEK/MAPK or p38 MAPK. Although these pathways have been implicated in protection from apoptosis,24, 32, 42 IGF-I was still able to rescue ES cells when PI3-K, MEK/MAPK and p38 MAPK pathways were simultaneously blocked by their selective inhibitors. Whether or not this alternative pathway may involve translocation of Raf-1 to mitochondria20 will be the subject of further studies. In contrast with what was observed with LY294002, IGF-I was incapable of overcoming the inhibitory growth effects deriving from the blockage of MEK/MAPK by PD98059. Cell cycle blockage was maintained in all the conditions tested, indicating that MEK/MAPK is necessary, and probably sufficient, for IGF-I-mediated ES proliferation.
The third section of the study was devoted to analyzing MAPK and PI3-K signaling pathways with respect to migration. MEK/MAPK was found to be involved in the migration of ES cells both in basal and IGF-I-driven conditions, whereas PI3-K was involved only in the chemotactic response of ES cells to IGF-I. IGF-I-induced migration was significantly reduced by LY294002 or PD98059 but the inhibition was only partial, indicating, in agreement with previous suggestions,47 that at least another pathway is involved.
As a final step, we investigated the in vitro growth effects of the 2 selective inhibitors, PD98059 and LY294002, in combination with doxorubicin. It is now generally accepted that IGF-I attenuates the response of cancer cells to several chemotherapeutic agents.48 Thus, inhibition of IGF-I action could be a useful adjuvant to cytotoxic chemotherapy. With respect to ES, a recent report37 clearly indicated that ES cells, when treated with doxorubicin, used an IGF-IR-initiated signaling pathway through PI3-K and Akt for survival. Therefore, it was quite predictable that the inhibition of PI3-K by LY294020 rendered the ES cells more sensible to doxorubicin. Our findings showed that the blockage of MAPK signaling also significantly enhanced the chemosensitivity of ES cells toward doxorubicin, suggesting that any new drug targeting these 2 pathways could be effectively combined with this conventional anticancer agent and might be of significant therapeutic value.
In conclusion, we demonstrated that disruption of either the MEK/MAPK or the PI3-K modules has profound functional consequences in ES cells. However, in the presence of IGF-I, the inhibition of MEK/MAPK pathway may have an additional advantage. The critical role of MEK/MAPK pathway in ES, also supported by the fact that interference with the constitutive activation of members of the MAPK signaling pathway impairs EWS/FLI-1-dependent transformation,38 provides impetus for future studies testing the in vivo therapeutic value and the general toxicity of specific inhibitors to be considered for innovative treatments of ES patients.
The authors thank Dr. Alba Balladelli, Laboratorio di Ricerca Oncologica, Istituti Ortopedici Rizzoli, Bologna, Italy, for critical reading of the article. Supported by fellowships from the Italian Foundation for Cancer Research (to V.C. and R.S.).