Expression and activation of SHC/MAP kinase pathway in primary ...

ã

The Hematology Journal (2001) 2, 70 ± 80 2001 The European Haematology Association All rights reserved 1466 ± 4680/01 $15.00 www.nature.com/thj

Expression and activation of SHC/MAP kinase pathway in primary acute myeloid leukemia blasts Paolo Lunghi1, Antonio Tabilio2, Silvana Pinelli1, Giuseppe Valmadre1, Erminia Ridolo1, Roberto Albertini1, Carmelo Carlo-Stella3, Pier Paolo Dall'Aglio1, Pier Giuseppe Pelicci1,4 and Antonio Bonati*,1 1

The Institute of Medical Pathology, University of Parma Medical School, Parma, Italy; 2The Chair of Hematology, University of Perugia Medical School, Perugia, Italy; 3The Tumors National Institute, Milan, Italy; 4The European Institute of Oncology, Milan, Italy

Introduction: We report the results of a study investigating signaling proteins in 26 cases of primary acute myelogenous leukemia. We studied the Shc adaptor proteins p52/p46Shc, which can activate the RAS/Mitogen Activated Protein kinase pathway, p66Shc which is uncoupled from RAS/MAP kinases and the MAP kinase family members Extracellular signal Regulated Kinase (ERK) and c-Jun NH2-terminal protein Kinase (JNK) or Stress Activated Protein Kinase (SAPK). Material and methods: CD34+ and CD347 fractions of four human normal bone marrow and unfractionated bone marrow samples were investigated. Immunoblottings, immunoenzymatic and in vitro assays were performed. Results: Shc protein isoforms were constitutively expressed in all the AML cases examined. Tyrosine-phosphorylation of p53/p46Shc isoforms were found in CD34+ but not in the majority of CD347 cases. p66Shc isoform was not tyrosine-phosphorylated in CD347, and was tyrosine-phosphorylated only in some CD34+ cases. Expression and activation of ERK was constitutively present in the majority of AML patients analysed. JNK/SAPK was expressed but not activated in the AMLs examined. Activation occurred after treatment of the leukemic cells by anisomycin, etoposide, and cytarabine. ERK and JNK/SAPK activation were not detectable in the hematopoietic precursors of human normal bone-marrow. Conclusion: These data bear implications for the role of Shc-MAP kinase pathway in normal hemopoiesis and AML leukemogenesis. The Hematology Journal (2001) 2, 70 ± 80 Keywords: acute myeloid leukemia; signal transduction; Shc adaptor proteins; ERK kinases; JNK/SAPK kinases

Introduction Acute myelogenous leukemia (AML) is a neoplastic disease of malignantly transformed hematopoietic progenitor cells. The disease is characterized by the clonal expansion of hematopoietic progenitor cells that are blocked in terminal dierentiation resulting in a predominance of the malignant clone outgrowing the normal hematopoietic cells.1,2 The transformation to leukemogenesis may require multiple steps including the acquisition by the leukemic cells of the capacity to produce self-supporting growth factors by autocrine mechanisms.3,4 Signi®cantly, it has been reported that the capacity for leukemic blasts for autonomous in

*Correspondence: A Bonati, Istituto di Patologia Medica dell'UniversitaÁ di Parma, Via Gramsci 14, 43100 Parma, Italy; Tel: +39 0521 991702; fax: +39 0521 941290; E-mail: [email protected] Received 13 September 2000; accepted 29 December 2000

vitro proliferation is associated with highly clinical aggressiveness of AML.5 Self growth mechanisms may act by the direct secretion of cytokines and growth factors by the neoplastic cell or by the constitutive expression of their substrates that are activated in the cytoplasm of the cell.6 These substrates are adaptor or catalytic proteins encoded by c-oncogenes participating in intracellular signal transduction pathways.7 The redundancy of cytokines and growth factors network suggests that a better understanding of self-growth mechanisms could be obtained by a direct analysis of intracellular proteins regulating cells signal transduction. There is evidence that many signaling proteins may represent common steps towards which a multiplicity of external stimuli are directed to create a pattern of response that is not dispersed.8,9 Moreover, it is mandatory to understand whether or not dierent signaling proteins are more activated in AML blasts

Shc/MAP kinases in AML P Lunghi et al

71

than in normal hematopoietic precursors so as to exploit their modulation to inhibit the outgrowth of the leukemic clone with little if any toxicity on normal hematopoietic precursors. A central role in the induction of intracellular activation is played by the RAS/MAPK kinase pathway. After stimulation by protein tyrosine kinase (PTK) or non PTK receptors, the activation of the pathway starts by the recruitment of p21 RAS protein at the inner face of the cell membrane.10 The phosphorylation of Shc adaptor proteins is a crucial step in this process since it creates a high anity binding site for the Grb2/SoS complex whose function is to activate RAS.11,12 Then, RAS activation stimulates a wave of protein kinase phosphorylations with determinant involvement of mitogen activated protein (MAP) kinases. By using human primary cells, our study was aimed at: (i) understanding whether or not a constitutive expression and tyrosine-phosphorylation of p66Shc isoform, which is uncoupled from RAS/MAP kinase pathway,13 and of p52/p46Shc isoforms, which conversely can activate the RAS/MAP kinase pathway14 is present in primary cases of AML; (ii) determining the expression and activation status of the MAP kinase pathway in AML; (iii) understanding whether or not a dierent behavior of the MAP kinases is detectable between AML blasts and the hematopoietic precursors of human normal bone marrow (BM). MAP kinase pathway was investigated by analysing extracellular signal regulated kinases (ERK) and c-Jun aminoterminal kinases (JNK) or stress-activated protein kinases (SAPK).

We examined 26 AML cases taken at diagnosis characterized by dierent FAB classi®cation and dierent immunological and genetic phenotype. Four cases of normal human BM were also investigated.

Materials and methods Patients and samples Fresh BM samples were taken at diagnosis before any therapy. The diagnosis of AML and the identi®cation of immunological and genetic characteristics were performed according to standard criteria and methodologies.15 ± 17 Seventeen patients had more than 30% of the blasts CD34+ and were considered CD34+, nine patients had no more than 3% and were considered CD347 (Table 1). A reference number was given to each of the cases studied. Normal hematopoietic cells were obtained from healthy donors undergoing BM harvest (n=4) for allogeneic transplantation. After informed consent, marrow was obtained by aspiration from the posterior iliac crest. Marrow mononuclear cells were separated by centrifugation (400 g for 30 min at 48C) on a Ficoll-Hypaque gradient (density=1.077 g/ml). Interface cells were washed in phosphate buered saline (PBS). The samples were directly lysed for protein detection after the separation of CD34+ and CD347 fractions (see below). Polymorphonuclear granulocytes (PMN) were obtained after Ficoll by resuspending the pellet in dextran. Normal bone-marrow cells were enriched according to CD34 expression by means of a magnetic cell

Table 1 Clinical characteristics of AML patients Case no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Age (years)

Sex

FAB

CD 34+ (%)

32 59 40 52 59 69 41 77 67 49 35 63 34 45 32 55 38 47 39 55 45 64 54 48 54 42

M M F M F F M F M F F M M M M F F F M F M M F M F M

M2 M4 M2 M2 M4 M2 M3 M2 M5 M5a M1 M4 M0 M2 M1 M4 M2 M0 M2 M5 M4 M0 M1 M2 M4 M4

61.1 96 86 64 51 2.5 1.5 58 75.8 0 0 31.3 38.2 85 52.3 0 0 40 1.8 50 39 44.6 42 1.2 53.7 3

Cytogenetics abnormalities 46,XY,t(2;6)(q14.3;p21) 46,XY 46,XX,t(2;7;13)(q14;q21;q22) 46,XY 46,XX,inv(16)(p13;q22) 46,XX 46,XY,t(15;17)(q21-q11-22) 46,XX 46,XY 46,XX,t(9;11)(p21;q23) 46,XX,t(6;9)(p23;q34) ND 46,XY 45XY,del(3)(q21;q27) 46,XY 46,XX,inv(16)(p13;q22) 46,XX,t(8;21)(q22;q22) 46,XX 46,XY,t(8;21)(q22;q22) 46,XX 47,XY,+8,inv(16)(p13;q22) 46,XY 46,XX,t(3;3)(q21;q26) 46,XY,t(8;21)(q22;q22) 46,XX,inv(16)(p13;q22) 46,XY

FAB, French-American-British classi®cation; ND, not done. The Hematology Journal


72

sorting methodology (MACS, Miltenyi Biotec, Germany) as previously described.18

Cell lysis The cells were lysed on ice in 50 mM Tris-HCl pH 8, 1.5 mM MgCl2, 150 mM NaCl, 5 mM EGTA pH 7.5, 5% (v/v) glycerol, 1% (v/v) Triton X-100 containing freshly added protease inhibitors (2 mg/ml Aprotinin, 2 mg/ml Leupeptin, 1 mg/ml Pepstatin, 1 mM phenylmethyl sulphonyl ¯uoride, 1 mM sodium orthovanadate, 50 mM sodium ¯uoride). Insoluble materials were removed by centrifugation for 10 min at 12 000 g at 48C and protein concentration determined by Bio-Rad protein assay reagent (Bio-Rad Laboratories, Hercules, CA, USA).

Immunoprecipitation For immunoprecipitation experiments, cell lysates were incubated and gently rocked for 60 min at 48C with appropriate antibodies and immune complexes isolated with Protein A Sepharose CL-4B (Pharmacia LKB, Uppsala, Sweden). Immune complexes were washed four times with lysis buer at 48C, eluted and denaturated by heating for 3 min at 958C in reducing Laemmli buer. Samples were then prepared for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS ± PAGE) according to standard protocols.

Immunoblotting Either speci®c immunoprecipitates or total cell lysates were electrotransferred onto PVDF ®lters (Millipore Intertech, Bedford, MA, USA) after SDS ± PAGE. After blocking non speci®c reactivity, ®lters were probed for 1 h at room temperature with speci®c antibodies diluted in TBS-T (25 mM Tris HCl pH 8, 150 mM NaCl, 0.05% Tween 20) containing 5% non fat milk. For anti-phosphotyrosine (anti-pTyr) experiments, 0.02% Tween 20 and 1% bovine serum albumin (BSA, fraction V, Boehringer, Mannheim, Germany) concentrations were utilized. After extensive washing, immunocomplexes were detected with horseradishperoxidase conjugated specie-speci®c secondary antiserum followed by enhanced chemiluminscence reaction (Amersham International plc, Buckinghamshire, UK). When necessary to con®rm the speci®city of our detection, after enhanced chemiluminescence, some membranes were stripped by incubation for 5 min in 0.2 M NaOH. After washing, the ®lters were treated with blocking solutions and reprobed with the appropriate antibody. The intensity of phosphorylation of the proteins examined was observed by direct visualization of the pictures, and was quanti®ed by densitometry.

Antibodies used for the experiments For immunoprecipitation, the following antibodies were used: rabbit polyclonal anti-human Shc (Santa The Hematology Journal

Cruz Biotechnology Inc, CA, USA); mouse monoclonal anti-human Shc (Santa Cruz Biotechnology); rabbit polyclonal p44/p42 MAP kinase antibody (New England BioLabs, Beverly, MA, USA). For immunoblotting, the following antibodies were used: rabbit polyclonal anti-human Shc (Santa Cruz Biotechnology); mouse monoclonal anti-pTyr (Upstate Biotechnology Inc, Lake Placid, NY, USA); goat polyclonal anti-human actin (Santa Cruz Biotechnology); mouse monoclonal anti-ERK2 (p42) (Santa Cruz Biotechnology), rabbit polyclonal Phospho-JNK/SAPK (Thr183/ Tyr185), rabbit polyclonal JNK/SAPK, rabbit polyclonal phosphospeci®c ELK1 (Ser383), rabbit polyclonal phosphospeci®c c-Jun antibody (Ser63) all provided by New England BioLabs; goat anti-rabbit IgG (H+L)-HRP conjugated (Bio-Rad); goat antimouse IgG (H+L)-HRP conjugated (Bio-Rad); donkey anti-goat IgG (H+L)-HRP conjugated (Santa Cruz Biotechnology).

ERK and JNK immunoenzymatic assays The rabbit polyclonal p44/42 MAP kinase antibody was used to selectively immunoprecipitate MAP kinase from cell lysates. The resulting immunoprecipitates were then incubated with a ELK1 fusion protein in the presence of ATP and kinase buer to allow immunoprecipitated MAP kinase to phosphorylate ELK1. Phosphorylation of ELK1 at Ser383 was detected by Western blottings using a rabbit polyclonal phosphospeci®c ELK1 (Ser383) antibody. Ser383 is a major phosphorylation site by MAP kinase and is required for ELK1 dependent transcriptional activity.19 All the reagents, including puri®ed MAP kinase for quantitative immunoblottings and the reagents for chemiluminescent detection, were provided by New England BioLabs. It has been observed that this assay provides very good detection and quantification of the signal.19 The quanti®cations were performed by comparing the intensity of activation observed in the leukemic samples with the intensity of activation induced by a known amount of activated ERK.19 One unit of activity is de®ned as the amount of MAP kinase required to catalyse the transfer of 1 pmol of phosphate to myelin basic protein (100 mM) in 1 min at 308C in MAPK buer in a 30 ml reaction volume. To perform JNK immunoenzymatic assay an Nterminal c-Jun (1 ± 89) fusion protein was bound to glutathione sepharose beads to selectively pull down JNK/SAPK from cell lysates.19 ± 21 c-Jun contains a high anity binding site for JNK/SAPK, just N-terminal to the two phosphorylation sites, Ser 63 and Ser73.21,22 After selectively pulling down JNK using the c-Jun fusion proteins beads and washing, the kinase reaction was carried out in the presence of ATP. c-Jun phosphorylation was selectively measured using a rabbit polyclonal phosphospeci®c c-Jun antibody. This antibody speci®cally measures JNK induced phosphorylation of c-Jun at Ser63, a site important for c-Jun transcriptional activity,21,22 with near zero background.19 All the reagents including c-Jun fusion


73

protein substrate and chemiluminescent detection system were provided by New England BioLabs.

In vitro cell assays Leukemic cells were incubated at 378C, 5% CO2, RPMI+penicillin/streptomycin, glutamine, 1% human albumin, without serum or growth factors. The treatments were performed with the dierent compounds by doses and times of exposure speci®ed in the Results section.

Results High tyrosine-phosphorylation status of p52/p46Shc adaptor proteins in CD34+ AML blasts Immunoblotting of the immunoprecipitated samples by anti-pTyr antibody showed, in the CD34+ cases, Shc tyrosine-phosphorylation. All the CD34+ cases expressed signi®cant levels of p52/p46Shc tyrosine phosphorylation which are isoforms capable of activating the RAS/MAP kinase pathway, but only 40% of these cases showed an evident tyrosinephosphorylation of p66Shc which is an isoform demonstrated to be uncoupled from the RAS/MAP kinase pathway. Tyrosine-phosphorylation of the Shc isoforms was not detectable or very low in eight of the nine cases of CD347 AML examined. Only case no. 16

(not reported in Figure 1), classi®ed as M4, showed tyrosine-phosphorylation of the three Shc isoforms. Stripping of the ®lters and reprobing by anti-Shc antibodies showed that the three Shc isoforms were expressed in all the samples, both CD34+ and CD347. Figure 1 shows representative results obtained in 15 cases. Furthermore, by immunoblotting the whole cell lysates, Shc expression was clearly detectable in all the leukemic samples (data not shown).

High activation status of ERK in AML blasts. Low activation status in normal hematopoietic precursors Phosphorylation at both Thr-202 and Tyr-204 is required for full enzymatic activation. Dual phosphorylation increases the weight of the kinase producing a delay in its gel migration so that when phosphorylation occurs a second band appears over the expression band.23 ERK2 was signi®cantly expressed in all the samples examined including PMN. The second phosphorylation band was present in the major part of leukemic samples, both CD34+ and CD347, but not in PMN. Case no. 26 (M4) did not show detectable levels of the second phosphorylation band, and case no. 3 (M2), no. 8 (M2), no. 17 (M2), no. 24 (M2) showed lower levels than the other cases. The results are shown in Figure 2. As we evaluated ERK in a hematopoietic malignancy, we considered it of interest to have parallel information on the behavior of these kinases in normal

Figure 1 Anti-pTyr immunoblot of Shc-immunoprecipitate (upper side). The same ®lter was stripped and reprobed by anti-Shc antibody (lower side). All the CD34+ AML showed evident p52/p46 Shc tyrosine-phosphorylation whereas p66 Shc was not tyrosine-phosphorylated in four of the 11 CD34+ cases represented (no. 4, 15, 21, 23). None of the Shc isoforms was tyrosinephosphorylated in the CD347 cases reported in this picture. Both CD34+ and CD347 cases, but not PMN, expressed the Shc proteins constitutively. +=CD34+ samples; 7=CD347 samples; K=K562 leukemic cell line; PMN=polymorphonuclear granulocytes; FAB classi®cation of the cases is reported. The number reported at level of each line represents the reference number given to each patient. The Hematology Journal


74

hematopoietic precursors of human BM. We investigated cellular lysates of CD34+ and CD347 fractions taken from human normal BM by immunoblotting and probing the samples with the anti-ERK2 monoclonal antibody since there was not sucient material available to immunoprecipitate CD34+ fraction. Sideto-side comparison with AML samples showed a signi®cant expression of ERK2 (p42) in both the CD34+ and CD347 fractions, but neither of these populations showed an evident dual phosphorylation of ERK. A second band of dual phosphorylation over the expression band was evident in AML cases (Figure 2) but not in CD34+ or CD347 fractions of normal BM. After selective immunoprecipitation by p44/42 MAP kinase antibody, we incubated the immunoprecipitates with kinase buer and ELK1 fusion protein as substrate and analysed ELK1 phosphorylation at Ser383 by immunoblotting with phosphospeci®c ELK1 antibody. A high activation status of ERK1/2 kinase was detected in the majority of AML samples examined but not in PMN. The range of the levels of ERK1/2 activity was between 93 and 170 units. The mean+s.d. of the activity of the dierent cases was 134+25 units. The immunoenzymaticatic assay of case no. 26 con®rmed that this case did not express detectable levels of ERK activation. This case was not considered for evaluation of s.d. due to the great discrepancy between its levels and those of the other samples. Representative pictures of 12 cases and the quantitative data are shown in Figure 3A,B. In contrast with AML cases, immunoprecipitation of unfractioned normal BM with p44/42 MAP kinase antibody and the evaluation by immunoenzymatic assay of the levels of ERK activation in these samples gave low or undetectable levels of ERK activation. Representative results of immunoenzymatic assays and the mean+s.d. of the activity detected in all the cases studied are shown in Figure 3.

Expression but not activation of JNK/SAPK in AML blasts By immunoblotting the cell lysates with a rabbit polyclonal anti-JNK antibody, we detected p46 and p54 JNK expression in all the cases. Immunoblotting the same samples with a phospho anti-JNK (Thr183/Tyr185) rabbit polyclonal antibody, which speci®cally detects the dually phosphorylated isoforms of JNK, we demonstrated that, in the leukemic cells, p54 and p46 did not express any levels or very low levels of dual phosphorylation, except for the AML case no. 15 classi®ed as M1. The embryonic cell line 293 showed phosphorylation of JNK isoforms when the cells were stressed by UV light (positive control). In normal BM, JNK/SAPK were expressed at lower levels than in leukemic samples and dual phosphorylation was not detectable (Figure 4). The Hematology Journal

Figure 2 Anti ERK2 (p42) immunoblots by utilizing an antibody that analysed p42 expression in cell lysates of the AML cases and in the CD34+ and CD347 fractions of normal bone-marrow. When ERK2 dual phosphorylation occurred, a second band with lower migration capability was detectable. ERK2 was expressed in all the samples examined, both leukemic and normal. Conversely, the second dual-phosphorylation band was clearly seen only in the leukemic samples whereas both CD34+ and CD347 fractions of normal bone marrow did not show it. The dual phosphorylation band was not present in PMN. The dual phosphorylation band was lower in case no. 3, 8, 17, and 24 than in the other AML cases. Case no. 26 did not show the dual-phosphorylation band. pp ERK2=dual-phosphorylation band; N.BM=normal bone marrow; K=K562 leukemic cell line; PMN=polymorphonuclear granulocytes; +=CD34+ AML samples; 7=CD347 AML samples; FAB classi®cation of the cases is reported. The number reported at the level of each line represents the reference number given to each case studied.

We performed enzymatic assay in 20 cases by pulling down JNK kinase using c-Jun fusion protein beads and making kinase reaction using c-Jun fusion protein as substrate. c-Jun fusion protein phosphorylation was seen at Ser63 using phosphospeci®c antibody by Western blotting and chemiluminescent detection. The AML cases analysed in basal conditions showed extremely low levels of JNK activation. There was not a sucient available amount of cell lysate case no. 15 to perform immunoenzymatic assay. Unfractioned human normal BM did not show any levels of JNK activation (Figure 5).


75

Figure 3 Immunoenzymatic assays detecting ERK1/2 activity in AML patients and in normal bone marrow. The levels of phosphorylation of the substrate ELK1 produced by ERK1/2 activity were evaluated. The quanti®cations were performed by comparing the intensity of activation observed in the samples studied with the intensity of activation induced by a known amount of activated ERK and expressed as Units. (A) and (B) show representative results of the immunoenzymatic assays performed in 12 AML patients and in three normal bone marrow. The histograms of the activities are reported. High levels of ERK1/2 activity were detectable in all the AML cases analysed and were comparable with the activity of GM-CSF-stimulated HL60 leukemic cell line. Very low or undetectable levels of ERK1/2 activity were seen in N. BM and in PMN. (C) reports the mean+s.d. of the activity detected in all the cases studied. N.BM=normal bone marrow; PMN=polymorphonuclear granulocytes; one unit of activity is de®ned as the amount of MAP kinase required to catalyse the transfer of 1 pmol of phosphate to myelin basic protein (100 mM) in 1 min at 308C in MAPK buer in a 30 ml reaction volume; FAB classi®cation of the cases is reported. The number reported at level of each line represents the reference number given to each case stud ied. GM-CSF was utilized at 10 ng/ml concentration. The Hematology Journal


76

Activation of JNK/SAPK in AML blasts and in leukemic cell lines by chemical stress and by using anti-leukemic drugs Primary AML blasts (15 cases) were treated with anisomycin, a well known compound capable of inducing chemical stress by inhibiting protein synthesis. After treatment of the leukemic cells for 30 min by anisomycin at 5 mg/ml (18.8 mM) and 10 mg/ml concentrations, we observed a dramatic increase of the levels of JNK activity in the treated leukemic cells even at 5 mg/ml concentration with respect to the non treated controls (Figure 5). Given these values, we investigated whether or not the mechanisms of JNK activation can be induced in AML cells by treatment of the blasts with common anti-leukemic drugs. For this purpose, we treated primary leukemic blasts (15 cases) with etoposide and cytarabine. The exposure of the leukemic cells to 1 and 10 mM etoposide and cytarabine for 2 h significantly increased the levels of JNK with respect to non treated samples. Samples treated with anisomycin showed higher levels of JNK activation with respect to the levels detected by etoposide or cytarabine administration. Analysis of internucleosomal DNA fragmentation after drugs exposure showed the appearance of this phenomenon in the treated leukemic cells (data not shown). Interestingly, the treatment of AML blasts (see Figure 5) with drugs not active against AML such as a and g interferons, at concentration of 2000 U/ml, did not produce any eects on JNK activation.

Discussion This study analysed Shc together with ERK and JNK/ SAPK in the same blast cells of patients aected by primary AML. Furthermore, the behavior of normal and leukemic hematopoietic precursors was compared. It is very important to investigate primary cells for better understanding of the behavior of signaling proteins in a particular disease since their expression and activity have been frequently observed to be dependent on speci®c cell types.24,25 Both ERK and JNK/SAPK have dierent and opposite activity depending on the experimental conditions such as the type and duration of the stimuli.24,25 For instance, epidermal growth factor activation of the ERK pathway leads to proliferation, whereas nerve growth factor-stimulated ERK activation leads to differentiation in pheocromocytoma PC12 cells.24 Activation of ERK is essential for erythropoietin-dependent cell growth but not for erythropoietin-induced erythroid dierentiation.25 In ®broblasts, a sustained activation of ERK is associated with proliferation but not dierentiation.26,27 Among reports addressing the role of JNK/SAPK in induction of the apoptotic response (see28), two out of three found a role for JNK/SAPK in this process, the other providing evidence that JNK/ SAPK is anti-apoptotic. Thus, investigations of fresh The Hematology Journal

primary cells of dierent cell types may add novel and mandatory information to our knowledge. The demonstration that ERK is constitutively activated in AML but barely activated or undetectable in normal, unstimulated, bone marrow hemopoietic precursors is an interesting observation. Recently, a highly potent and selective inhibitor of the upstream kinase MEK has been discovered.29 This compound, called PD 184352, speci®cally inhibits ERK activity as no substrates for MEK1 have been identi®ed other than ERK1 and ERK2.30 In vivo, tumor growth is inhibited as much as 80% in mice with colon carcinomas of both mouse and human origin after treatment with this inhibitor.29 Ecacy was achieved with a wide range of doses with no signs of hemopoietic toxicity. The undetectable levels of ERK activity found in normal hemopoietic precursors can explain this behavior. The data concerning Shc expression and tyrosine phosphorylation in AML are interesting, and for some aspects intriguing. In a previous report,31 high levels of Shc expression were found in the CD34+ cells taken from primary chronic myelogenous leukemias in blast crisis, in chronic phase (CML-CP), and in normal bone marrow. In contrast, CD347 fractions from CML-CP and normal BM expressed only low levels of p46Shc. p66Shc and p53/46Shc were strongly and constitutively tyrosine-phosphorylated in the CD34+ fraction from CML cells (both CP and BC), but not in the normal CD34+ cells. These data suggest that Shc expression correlates with the status of differentiation, and the dramatic changes of Shc expression during terminal dierentiation of hemopoietic cells adds a further level of regulation to the signal transduction function of Shc. Furthermore, the constitutive tyrosine-phosphorylation of all the three Shc isoforms might contribute to the expansion of the small CD34+ fraction of CML-CP during the BC transformation and to the more aggressive phenotype of CD34+ Ph+ cells in BC.31 Accordingly, p66Shc and p52/p46Shc were constitutively expressed in the AML cases studied. AML blasts are immature or poorly dierentiated cells: our data are in agreement with other observations that demonstrated that Shc is linked to speci®c subsets of integrins (a5 b1, av b3) at immature stages of cell dierentiation.32 Marked variations in Shc levels were also demonstrated at the transition from immature to dierentiated neural cells, with almost undetectable levels of Shc expression in postmitotic neurones.33 In contrast with CML behavior, AML showed high p52/p46Shc tyrosine-phosphorylation in the CD34+ but not in the majority of CD347 cases. p66Shc was not tyrosine-phosphorylated in some of the CD34+ cases, and we did not observe a tyrosine phosphorylated p66Shc in CD347 cases. These data are intriguing and need further investigation. Clinical studies may investigate whether or not AML patients who express high levels of Shc tyrosine-phosphorylation have a worse prognosis than the patients who do not. In fact, primary chronic myelogenous leukemia in blast crisis


77

Figure 4 Immunoblotting of cell lysates of AML patients and normal bone-marrow to analyse JNK/SAPK expression (A) or dualphosphorylation (B). The cases of AML examined had detectable levels of JNK/SAPK expression, whereas lower levels were seen in normal bone marrow. Dual phosphorylation was very low or undetectable in all the cases examined, both leukemic and normal, except for the AML case no. 15. Embryonic cell line 293 (positive control) was dual phosphorylated by UV treatment. For abbreviations see ®gure 2 legend.

that characteristically expresses high levels of tyrosinephosphorylation of all the three Shc isoforms,31 is a very aggressive and uncontrollable leukemic proliferation. Biological studies may investigate the relationships between CD34 receptor and Shc tyrosine-phosphorylation. We suggest that the recruitment by the stimulated CD34 receptor of speci®c phosphatases34 could be involved.

The fact that p52/p46Shc tyrosine-phosphorylation was associated with p66Shc tyrosine-phosphorylation, in the patients examined, is not surprising since the latter isoform derives from dierential translational initiation and has dierent biochemical activity being independent from RAS/MAP kinase pathway.13 While p66Shc serine-phosphorylation can stimulate oxidative processes, leading to cell apoptosis,35 we suggest that p66Shc tyrosine-phosphorylation could, conversely, The Hematology Journal


78

Figure 5 Immunoenzymatic assays detecting JNK/SAPK activity. JNK/SAPK activity was evaluated by analysing the levels of phosphorylation of a JNK/SAPK substrate, the oncogene c-Jun. (A) shows representative results obtained in ®ve cases of AML before or after treatment by anisomycin and in normal bone marrow samples. The results of a representative case before and after treatment by anisomycin, etoposide, cytarabine or a and g-interferons are also shown. (B) Histogram showing the mean+s.d. of the JNK/SAPK activity detected in the experiments performed. JNK/SAPK activity was very low or undetectable in leukemic samples before any treatment. JNK activity dramatically increased after administration of anisomycin, or of anti-leukemic drugs etoposide and cytarabine. Both a and g interferon had no eects on JNK/SAPK activity. a and g interferons were utilized at 2000 U/ml concentration. N.BM=normal bone marrow; NT=not treated; aniso-anisomycin; eto=etoposide; ARA=cytarabine; IFN=interferon; HR=hours.

contribute to a more aggressive phenotype, given the extremely high levels of tyrosine-phosphorylation found in CML-BC.31 Our data support the existence of alternative pathways of ERK activation that can be used instead of Shc-Grb2-RAS. The existence of these pathways has been suggested by others9,36 but the signaling proteins involved have not been clearly identi®ed. Interestingly, leukemic samples and normal hemopoietic precursors had a similar behavior with regard The Hematology Journal

to JNK/SAPK activation. The JNK/SAPK expression was higher in AML cases than in normal hemopoietic precursors but the same levels of JNK/SAPK activation were detected. Furthermore, JNK/SAPK retains the capability to be activated by toxic compounds or drugs in normal hemopoietic precursors (data not shown) and in leukemic cells. Our ®ndings suggest that JNK/SAPK could participate in the processes of tumor cell apoptosis after drugs exposure. Accordingly, JNK/SAPK may play a role


79

in the transcriptional regulation of interferon regulatory factor-7 in response to chemotherapeutic DNAdamaging drugs.37 Experiments are in progress to inhibit JNK/SAPK activity by anti-sense oligonucleotides after drugs exposure of AML blasts to stringently demonstrate whether the role of JNK/SAPK is essential or not to obtain an anti-leukemic eect. In this study, we examined patients at diagnosis, with uniform clinical characteristics, and who were responsive to induction chemotherapy. It will be interesting to investigate whether blast cells of cases of AML, taken at relapse or resistant to chemotherapy, retain the capacity to activate JNK/SAPK. To perform ERK immunoenzymatic assays, we immunoprecipitated both ERK1 (p44) and ERK2

(p42). In this way, we detected a total ERK1/2 activity. However, it is highly likely that a signi®cant part of the ERK activity determined was due to p42 kinase since a much less abundant ERK1 than ERK2 expression has been reported in hematopoietic cells.25,36

Acknowledgements This work was supported by grants from `Associazione Italiana per la Ricerca sul Cancro' (AIRC) and `Ministero dell'UniversitaÁ e Ricerca Scienti®ca e Tecnologica' (MURST 40% and 60%). P Lunghi was supported by a grant from Roche Company, Basel, Switzerland.

References 1 Appelbaum FR, Gilliland DG, Tallman MS. The biology and treatment of acute myeloid leukemia, p 15 In: McArthur JR, Schechter GP, Schrier SL (eds) Hematology 1998. WB Saunders Company, Orlando, 1998. 2 Sutherland HJ, Blair A, Zapf RW. Characterization of a hierarchy in human acute myeloid leukemia progenitor cells. Blood 87: 4754, 1996. 3 Young DC, Wagner K, Grin DJ. Constitutive expression of the granulocyte-macrophage colonystimulating factor gene in acute myeloblastic leukemia. Journal of Clinical Investigation 79: 100, 1987. 4 Grin JD, Rambaldi A, Vallengo E, Young DC, Ostepovicz D, Cannistra SA. Secretion of interleukin-1 by acute myeloblastic leukemia cells in vitro induces endothelial cells to secrete colony stimulating factors. Blood 70: 1218, 1987. 5 Lowenberg B, Van Putten WL, Touw IP, Delvel R, Santini V. Autonomous proliferation of leukemic cells in vitro as a determinant of prognosis in adult acute myeloid leukemia. New England Journal of Medicine 328: 614, 1993. 6 Kanakura Y, Druker B, Cannistra SA, Furukawa Y, Torimoto Y, Grin JD. Signal transduction of the human granulocyte-macrophage colony stimulating factor and interleukin-3 receptors involves tyrosine phosphorylation of a common set of cytoplasmic proteins. Blood 76: 706, 1990. 7 Krontiris TG. Oncogenes. New England Journal of Medicine 333: 303, 1995. 8 Ihle JN. Cytokine receptor signalling. Nature 377: 591, 1995. 9 Corey SJ, Anderson SM. Src-related protein tyrosine kinases in hematopoiesis. Blood 93: 1, 1999. 10 Leever SJ, Paterson HF, Marshall CJ. Requirements for Ras in Raf activation is overcome by targeting Raf to the plasma membrane. Nature 369: 411, 1994. 11 Lanfrancone L, Pelicci G, Brizzi MF, Arouica MG, Casciari C, Giuli S, Pegoraro L, Pawson T, Pelicci PG. Overexpression of Shc proteins potentiates the proliferative response to the granulocyte-macrophage colonystimulating factor and recruitment of Grb2/SoS and Grb2/p140 complexes to the b receptor subunit. Oncogene 10: 907, 1995.

12 Plyte S, Majolini MB, Pacini S, Scarpini F, Bianchini C, Lanfrancone L, Pelicci PG, Baldari CT. Constitutive activation of the Ras/MAP kinase pathway and enhanced TCR signaling by targeting the Shc adaptor to membrane rafts. Oncogene 19: 1229, 2000. 13 Migliaccio E, Mele S, Salcini AE, Pelicci G, Venus Lai KM, Superti-Furga G, Pawson T, Di Fiore PP, Lanfrancone L, Pelicci PG. Opposite eects of the p52shc/p46shc and p66 splicing isoforms on the EGF receptor-MAP kinase-fos signalling pathway. EMBO Journal 16: 706, 1997. 14 Rozakis-Adcock M, McGlade J, Mbamalu G, Pelicci G, Daly R, Li W, Batzer A, Thomas S, Brugge J, Pelicci PG, Schlessinger J, Pawson T. Association of the Shc and Grb2/Sem5 SH2-containing proteins is implicated in activation of the Ras pathway by tyrosine kinases. Nature 360: 689, 1992. 15 Bonati A. Biochemical, immunological and molecular markers of hemopoietic precursor cells, p 1 In: Harris JR (ed.). Blood Cell Biochemistry. vol 3. Plenum Publishing Corporation, New York, 1991. 16 Strout MP, Caligiuri MA. Developments in cytogenetics and oncogenes in acute leukemia. Current Opinions in Oncology 9: 9, 1997. 17 Bonati A, Zanelli P, Ferrari S, Plebani A, Starcich B, Savi M, Neri TM. T-cell receptor b-chain gene rearrangement and expression during human thymic ontogenesis. Blood 79: 1472, 1992. 18 Carlo Stella C, Dotti G, Mangoni L, Regazzi E, Garau D, Bonati A, Almici C, Sammarelli G, Savoldo B, Rizzo MT, Rizzoli V. Selection of myeloid progenitors lacking BCR/ABL mRNA in chronic myelogenous leukemia patients after in vitro treatment with the tyrosine kinase inhibitor genistein. Blood 88: 3091, 1996. 19 Tan YI, Rouse J, Zhang A, Cariati S, Cohen P, Comb MJ. FGF and stress regulate CREB and ATF-1 via a pathway involving p38 MAP kinase and MAPKAP kinase-2. EMBO Journal 15: 4629, 1996. 20 Dai T, Rubie E, Franklin CC, Kraft A, Gillespie DA, Avruch J, Kyriachis JM, Woodget JR. Stress activated protein kinases bind directly to the delta domain of c-Jun in resting cells: implications for repression of c-Jun function. Oncogene 10: 849, 1995.

The Hematology Journal


80

21 Gupta S, Barret T, Whitmarsh AJ, Cavanagh J, Sluss HK, Derijard B, Davis RJ. Selective interaction of JNK protein kinase isoforms with transcription factors. EMBO Journal 15: 2760, 1996. 22 Canagarajah BJ, Khokhlatchev A, Cobb MH, Goldsmith EJ. Activation mechanisms of the MAP kinase ERK2 by dual phosphorylation. Cell 90: 859, 1997. 23 Yen A, Roberson MS, Varvajanis S, Lee AT. Retinoic acid induced Mitogen-activated Protein (MAP)/Extracellular Signal Regulated (ERK) kinase-dependent MAP kinase activation needed to elicit HL-60 cell differentiation and growth arrest. Cancer Research 58: 3163, 1998. 24 Marshall CJ. Speci®city of receptor tyrosine kinase signaling: transient versus sustained extracellular signalregulated kinase activation. Cell 80: 179, 1995. 25 Nagata Y, Takahashi N, Davis RJ, Todokoro K. Activation of p38 MAP kinase and JNK but not ERK is required for erythropoietin-induced erythroid differentiation. Blood 92: 1859, 1998. 26 Cowley S, Paterson H, Kemp P, Marshall CJ. Activation of MAP kinase kinase is necessary and sucient for PC12 dierentiation and for transformation of NHI 3T3 cells. Cell 77: 841, 1994. 27 Mansour S, Matten WT, Hermann AS, Candia JM, Rong S, Fukasawa K, Vande Woude GF, Ahn NG. Transformation of mammalian cells by constitutively active MAP kinase kinase. Science 265: 966, 1994. 28 Lieberman DA. Normal development, oncogenesis and programmed cell death. Oncogene 17: 1189, 1998. 29 Sebolt-Leopold JS, Dudley DT, Herrera R, Van Becelaere K, Wiland A, Gowan RC, Tecle H, Barret SD, Bridges A, Pryzybranowsky S, Leopold WR, Saltiel AR. Blockade of the MAP kinase pathway suppresses growth of colon tumors in vivo. Nature Medicine 5: 810, 1999. 30 Seger R, Ahn NG, Posada J, Munar ES, Jensen AM, Cooper JA, Cobb MH, Krebbs EG. Puri®cation and characterization of mitogen-activated protein kinase activator(s) from epidermal growth factor-stimulated AA31 cells. Journal of Biological Chemistry 267: 14373, 1992.

The Hematology Journal

31 Bonati A, Carlo-Stella C, Lunghi P, Albertini R, Pinelli S, Migliaccio E, Sammarelli G, Savoldo B, Tabilio A, Dall'Aglio PP, Pelicci PG. Selective expression and constitutive phosphorylation of Shc proteins in the CD34+ fraction of chronic myelogenous leukemias. Cancer Research 60: 728, 2000. 32 Wary KK, Mainiero F, Isako JI, Marcantanio EE, Giacotti FG. The adaptor protein Shc couples a class of integrins to the control of cell cycle progression. Cell 87: 733, 1996. 33 Conti L, De Fraia C, Gulisano M, Migliaccio E, Govoni S, Cattaneo E. Expression and activation of SH2/PTB containing ShcA adaptor protein re¯ects the pattern of neurogenesis in the mammalian brain. Proceedings of the National Academy of Sciences USA, 94: 8185, 1997. 34 Paulson RF, Vesely S, Siminovitch KA, Bernstein A. Signaling by the W/Kit receptor tyrosine kinase is negatively regulated in vivo by the protein tyrosine phosphatase Shp1. Nature Genetics 13: 309, 1996. 35 Migliaccio E, Giorgio M, Mele S, Pelicci G, Reboldi P, Pandol® PP, Lanfrancone L, Pelicci PG. The p66shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402: 309, 1999. 36 Rojnuckarin P, Drachman JG, Kaushansky K. Thrombopoietin-induced activation of the Mitogen-Activated Protein Kinase (MAPK) pathway in normal megakaryocytes: role in endomitosis. Blood 94: 1273, 1999. 37 Kim TK, Kim T, Kim TY, Lee WG, Yim J. Chemotherapeutic DNA-damaging drugs activate interferon regulatory factor-7 by the mitogen-activated protein kinase kinase-4-c-Jun NH2-terminal kinase pathway. Cancer Research 60: 1153, 2000.