LEADING ARTICLE Possible dominant-negative mutation of the SHIP

0 downloads 0 Views 1MB Size Report
The SH2 domain-containing inositol 5!-phosphatase (SHIP) is crucial in hematopoietic development. To evaluate the possible tumor suppressor role of the SHIP ...
Leukemia (2003) 17, 1–8  2003 Nature Publishing Group All rights reserved 0887-6924/03 $25.00 www.nature.com/leu

LEADING ARTICLE Possible dominant-negative mutation of the SHIP gene in acute myeloid leukemia J-M Luo1, H Yoshida2, S Komura3, N Ohishi3, L Pan1, K Shigeno1, I Hanamura4, K Miura4, S Iida4, R Ueda4, T Naoe5, Y Akao2, R Ohno6 and K Ohnishi1 1 Department of Medicine III, Hamamatsu University School of Medicine, Hamamatsu, Japan; 2Department of Genetic Diagnosis, Gifu International Institute of Biotechnology, Mitake, Japan; 3Institute of Applied Biochemistry, Mitake, Japan; 4Second Department of Internal Medicine, Nagoya City University Medical School, Nagoya, Japan; 5Department of Infectious Diseases, School of Medicine, Nagoya University, Nagoya, Japan; and 6Aichi Cancer Center Hospital, Nagoya, Japan

The SH2 domain-containing inositol 5⬘-phosphatase (SHIP) is crucial in hematopoietic development. To evaluate the possible tumor suppressor role of the SHIP gene in myeloid leukemogenesis, we examined primary leukemia cells from 30 acute myeloid leukemia (AML) patients, together with eight myeloid leukemia cell lines. A somatic mutation at codon 684, replacing Val with Glu, was detected in one patient, lying within the signature motif 2, which is the phosphatase active site. The results of an in vitro inositol 5⬘-phosphatase assay revealed that the mutation reduced catalytic activity of SHIP. Leukemia cells with the mutation showed enhanced Akt phosphorylation following IL-3 stimulation. K562 cells transfected with the mutated SHIPV684E cDNA showed a growth advantage even at lower serum concentrations and resistance to apoptosis induced by serum deprivation and exposure to etoposide. These results suggest a possible role of the mutated SHIP gene in the development of acute leukemia and chemotherapy resistance through the deregulation of the phosphatidylinositol-3,4,5-triphosphate (PI(3,4,5)P3)/Akt signaling pathway. This is the first report of a mutation in the SHIP gene in any given human cancer, and indicates the need for more attention to be paid to this gene with respect to cancer pathogenesis. Leukemia (2003) 17, 1–8. doi:10.1038/sj.leu.2402725 Keywords: SHIP; somatic mutation; myeloid leukemia

Introduction The SH2 domain-containing inositol 5⬘-phosphatase (SHIP) belongs to the inositol 5⬘-phosphatase family, and it is expressed exclusively in hematopoietic lineage cells.1,2 The human SHIP gene is located on chromosome 2q36-q37.1,2 and its primary structure predicts an amino-terminal SH2 domain followed by a catalytic domain with highly conserved motifs among the inositol 5⬘-phosphatase family, two NPXY motifs, and three potential polyproline motifs within the carboxyl terminal region.1,2 SHIP specifically cleaves the 5⬘phosphate group from phosphatidylinositol-3,4,5-triphosphate (PI(3,4,5)P3), the product of phosphoinositide 3-kinase (PI3K), and at least in vitro, from inositol-1,3,4,5-tetrakiphosphate (IP4), both of which act as second messengers (Ref. 1, and reviewed in Ref. 3). Recent studies have largely focused on phosphoinositide-mediated signaling pathways with respect to transforming and anti-apoptotic activity.3 Indeed, both PI3K and Akt (also called protein kinase B), which are serine/threonine-directed kinases, have been isolated as celluH Yoshida, Department of Genetic Diagnosis, Gifu International Institute of Biotechnology, Yagi Memorial Park, Mitake, Gifu 505-0116, Japan; Fax: +81-574-67-6627 J-M Luo and H Yoshida contributed equally to this study Received 29 April 2002; accepted 2 July 2002

lar counterparts of transforming retroviral oncogenes.4 Akt, whose activity is regulated by phosphatidylinositol phosphates, protects cells from apoptosis induced by various factors such as oxidative stress, serum deprivation, and genotoxic agents (reviewed in Ref. 5). Although the mechanism by which Akt promotes cell survival largely depends on the cellular context, Akt directly modulates the activity of pro-apoptotic molecules such as Bad, JNK and the Forkhead transcription factor FKHR-1.6 Recently, JNK and p38, two stressactivated mitogen-activated kinase molecules, have been revealed to have vital roles in chemotherapeutic-induced apoptosis.7 Based on these results, phosphoinositide phosphatases (PIPase) could function as tumor suppressors (reviewed in Ref. 8). The first example of this class discovered was PTEN (phosphatase and tensin homologue deleted on chromosome 10), which antagonizes PI3K activity by catalyzing the removal of the 3⬘-phosphate group from 3⬘-phosphoinositides.9 Whereas a high incidence of bi-allelic deletion or mutations of PTEN was reported in various tumors, including gliomas, breast and small lung cell cancers, and familial cancer predisposition disorders can be mimicked in PTEN heterozygous mice,8 recent studies failed to reveal frequent genetic aberrations in hematopoietic cell lines or in primary leukemias and lymphomas.10,11 Recent experiments have revealed that SHIP is another class of PIPase crucial for the negative regulation of proliferation and survival in hematopoietic cells.12 Of great interest is the finding that homozygous-null (SHIP−/−) mice display myeloproliferative disease and that their life span is shortened due to myeloid cell infiltration into vital organs.12 The bone marrow from SHIP−/− mice shows a chronic progressive hyperplasia of granulocytes and macrophages, whereas erythroid cells are little affected. The granulocyte/macrophage progenitors from these mice are capable of forming small colonies at high efficiencies without the addition of cytokines and also are less sensitive to apoptosis and substantially hypersensitive to multiple cytokine stimulation with increased PI(3,4,5)P3 production and Akt activity. Furthermore, SHIP expression is reduced in primary myeloid leukemia cells from patients with chronic myelogenous leukemia (CML),13 which is characterized by the expression of the BCR-ABL chimeric protein causing hyperleukocytosis and chemotherapy resistance.14 These observations strongly suggest that SHIP could function as a putative hematopoietic tumor suppressor during myelopoiesis and thus prompted us to investigate whether the loss of function of SHIP could be involved in leukemogenesis. Here, we give the first report of a mutation of the SHIP gene in acute myeloid leukemia and its functional analysis related to leukemia pathogenesis.

A somatic mutation of the SHIP gene in myeloid leukemia J-M Luo et al

2

Materials and methods

Cell lines and patients’ samples Eight myeloid leukemia cell lines (KG-1, NKM-1, HL-60, NB4, NOMO-1, U937, K562, and Meg01) were cultured in RPMI 1640 medium (Gibco, Grand Island, NY, USA), and COS-7 cells, in DMEM medium (Sigma, St Louis, MO, USA), with both media supplemented with 10% FCS, 1 mM L-glutamine, and appropriate antibiotics. All cell lines were obtained from the Japanese Cancer Research Bank. Primary leukemia cells were obtained from 30 patients with acute myeloid leukemia (AML) under informed consent according to our institutional review board-approved guidelines. The diagnosis was made according to the French–American–British classification (FAB) criteria. Bone marrow mononuclear cells (BMMNCs) or peripheral blood mononuclear cells (PBMNCs) of patients with active disease or in clinical remission were separated by density gradient centrifugation and cryopreserved until analysis could be performed. PBMNCs from 50 unrelated healthy volunteers were also examined for polymorphism analysis.

Mutation analysis Reverse transcription polymerase chain reaction/single-strand conformation polymorphism (RT-PCR/SSCP) analysis was performed as previously described.15 The following nine sets of oligonucleotide primers were used for PCR amplification: S1 (sense primer for fragment 1; 5⬘-CAGCCGGAGCCCGACAT3⬘) and AS1 (antisense for fragment 1; 5⬘-TCACACTGG TGATTTCTTGC-3⬘); S2 (5⬘-TGAGTGAGAAGGAGTGGCT-3⬘) and AS2 (5⬘-GTTGACGAACCCTAAGGAG-3⬘); S3 (5⬘CACACTGGGGAACAAGGG-3⬘) and AS3 (5⬘-GGTAGGCA GATCCACACG-3⬘); S4 (5⬘-CCTTTAACATCACTCACCG-3⬘) and AS4 (5⬘-GGTAGGTTGGGGCAAACG-3⬘); S5 (5⬘-CAC GACCAGCTGCTCAC-3⬘) and AS5 (5⬘-TACTGCCATAA GACTGACAC-3⬘); S6 (5⬘-AACTTCCTTCCTGGTGTG-3⬘) and AS6 (5⬘-GGACTTGGTCTTCAATGTG-3⬘); S7 (5⬘-GGACTG TTGACAGCCAAG-3⬘) and AS7 (5⬘-TCTAGCAGGTACTC AGGGT-3⬘); S8 (5⬘-GGTGGTGAAGTTTGGTGAG-3⬘) and AS8 (5⬘-CGTCTTGCCCTGAGAGGT-3⬘); and S9 (5⬘-CCATGG GGAGTTGACAGG-3⬘) and AS9 (5⬘-TGGGGTTGATGATTT CAGTG-3⬘). Sequencing was carried out on newly generated amplicons of fragments that had revealed aberrant conformers. The respective forward and reverse primers were used in independent sequencing reactions. The sample with a SHIP mutation was examined three times independently.

Western blot analysis Total cell lysates from leukemia cells were subjected to analysis. Anti-phospho-Akt (BD Biosciences Pharmingen, San Diego, CA, USA), anti-Akt (sc-8312) or anti-SHIP antibody (sc6244) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used as the first antibodies. Immunoblots were visualized with alkaline phosphatase-mediated (Bio-Rad, Hercules, CA, USA) or ECL (Amersham, Buckinghamshire, UK) detection according to the manufacturer’s instructions.

Fluorescence in situ hybridization (FISH) Leukemia cells were cytospin-prepared for FISH analysis. BAC clone (RP11-400N9, Gene Bank accession No. AC013726), Leukemia

which was derived from human chromosome region 2q36 containing the SHIP coding sequence, was kindly provided by Children’s Hospital, Oakland Research Institute (Oakland, CA, USA), directly fluorescence-labeled by nick translation and used as a FISH probe. Cells were counterstained with 4, 6diamino-2-phenylindole dihydrochloride. The hybridization and detection procedures were conducted according to the manufacturer’s instructions (Vysis, Downers Grove, IL, USA).

Plasmid construction and transfection The original human SHIP expression vector was kindly provided by Dr Hirai (Tokyo University1). The SHIP-V684E mutant was constructed by oligonucleotide-primed sitedirected mutagenesis according to the manufacturer’s instruction (Stratagene, La Jolla, CA, USA), and the correctness of the DNA sequences was confirmed. Plasmid constructs encoding polyhistidine-tagged SHIP were generated by cloning the wild and mutant SHIP in-frame into the pcDNA3.1/His vector (Invitrogen, San Diego, CA, USA). Each expression vector was used to transfect cells with the aid of Superfect Transfection Reagent according to the manufacturer’s instructions (Qiagen, Valencia, CA, USA). Each SHIP-expressing clone was generated by three independent transfections.

In vitro inositol 5⬘-phosphatase assay COS-7 cells were transfected with the polyhistidine-tagged wild-type or mutant SHIP as described above. Forty-eight hours later, total cell lysates were subjected to purification under native conditions with a Ni-NTA affinity column according to the manufacturer’s instruction (Qiagen). Inositol 5⬘-phosphatase activity was assayed by incubating purified SHIP protein at 37°C for 15 min in 10 ␮l of 50 mM TrisHCl buffer (pH 7.4), containing 3 mM MgCl2 33 ␮M BODIPYlabeled PI(3,4,5)P3 (Molecular Probe, Eugene, OR, USA). The reaction was initiated by the addition of the substrate and terminated by adding 10 ␮l of 40% acetonitrile. PI(3,4)P2 and PI(3,4,5)P3 were separated by HPLC on an anion-exchange column (TSK-GEL DEAE-2SW) (Tosoh, Tokyo, Japan) with elution buffer of 0.55 M KH2PO4/acetonitrile (80:20, v/v). Results and discussion

A somatic mutation within the signature motif of 5⬘phosphatase domain was found in primary myeloid leukemia cells Consistent with the notion that the phosphatase activity of PTEN is required for the tumor suppressor activity, many cancer-associated mutations have been mapped within its conserved catalytic domain (reviewed in Ref. 8). In this context, the coding region of the putative 5⬘-phosphatase SHIP (codons 411–866) was examined by RT-PCR/SSCP (Figure 1a). Prior to mutation analysis, genetic polymorphisms were examined in PBMNCs from 50 healthy volunteers. One polymorphism was found at codon 583 (CAA to CAG), as reported previously,16 which was also detected in both hematopoietic cell lines and primary leukemia cells. In addition, a novel polymorphism was also found at codon 529, showing an amino acid replacement (Met to Val; ATG to GTG, Figure 1a). Then, eight acute myeloid leukemia cell lines and primary

A somatic mutation of the SHIP gene in myeloid leukemia J-M Luo et al

3

Figure 1 Leukemia cells from patient AML19 have a replacement mutation within the signature motif of the 5⬘-phosphatase domain of SHIP but no alleic loss. (a) Schematic representation of the 5⬘-phosphatase domain of SHIP. Amino acid numbers are given. Asterisks indicate the position of the signature motifs conserved among the 5⬘-phosphatase family. Bold bars show the regions of RT-PCR fragments: fragment 1, amino acid 397–474; 2, 455–537; 3, 516–595; 4, 571–643; 5, 615–691; 6, 666–742; 7, 721–798; 8, 776–859; 9, 839–912. Arrows show the positions of genetic polymorphisms. The closed triangle indicates the position of the missense V684E mutation. (b) DNA sequences of fragment 6 from patient AML19. Compared with the wild-type sequence from her PBMNCs in complete remission (left panel), the mutated allele was confirmed to have a single-nucleotide substitution in her leukemia cells (right panel). (c) Absence of loss of heterozygosity in patient AML19. Lymphocytes from a healthy volunteer were probed with RP11-400N9 BAC clone. One representative metaphase is given. Arrows show the position of chromosome region 2q36-37.1 (left panel). The BM cells from patient AML19 at initial diagnosis contained >95% of leukemia cells. Some representative cells in interphase are given (right panel).

leukemia cells from 30 AML patients were subjected to mutation analysis. RT-PCR/SSCP analysis of fragment 6 from patient AML19 detected an altered mobility band in samples from her leukemia cells but not in her PBMNCs in complete remission (data not shown). DNA sequence analysis revealed a T to A transversion, replacing Val with Glu at codon 684 (V684E, Figure 1b). DNA fragments eluted from gels were also subjected to cloning into plasmids. In addition to the mutated allele, approximately half of the distinct clones contained the wild-type allele (data not shown). Since the PBMNC sample from the same patient during her complete remission was found to have only the wild-type allele (wt, Figure 1b left panel), the mutation in the leukemia cells was confirmed to be somatic. Considering the putative tumor suppressor function of SHIP, analogous to that of PTEN, we concomitantly examined loss of heterozygosity by FISH, which revealed the absence of allelic loss at the SHIP locus in patient AML19 (Figure 1c). Within the phosphatase domain of SHIP, there are two signature motifs, motif 1 and motif 2 (Figure 1a and Ref. 17), both of which are highly conserved among the inositol 5⬘-phosphatase supergene family. Interestingly, codon 684 lies just within the signature motif 2, which is the phosphatase active site. Thus, the V684E mutation could reduce the phosphatase activity and be expected to cause deregulation of signal transduction pathways elicited by PI(3,4,5)P3.

Mutant SHIP is catalytically defective in inositol 5⬘phosphatase activity To elucidate the functional alteration caused by the mutation, we first examined the enzymatic activity of the mutant SHIP as an inositol 5⬘-phosphatase. Both polyhistidine-tagged wildtype (SHIP-wt) and V684E mutant-type (SHIP-V684E) SHIP were transiently expressed in COS-7 cells and successfully purified (Figure 2b). A mock-transfected cell lysate was also subjected to the same purification procedure for use as a negative control in the inositol 5⬘-phosphatase assay. Purified proteins were incubated with fluorescence-labeled PI(3,4,5)P3 as the substrate. Varying the amount of SHIP protein or the time of incubation with substrates revealed that in our routine 15min assay, most of the substrate was efficiently dephosphorylated (data not shown). While the mock-transfected cell lysate showed the minimal capability to convert the PI(3,4,5)P3 to PI(3,4)P2, SHIP-wt efficiently hydrolyzed the PI(3,4,5)P3 more than 90%. As expected, SHIP-V684E was defective in inositol 5⬘-phosphatase activity compared with SHIP-wt (Figure 2a).

Mutation V684E caused enhanced Akt phosphorylation in response to IL-3 SHIP has been implicated in the regulation of Akt activation following IL-3 receptor engagement.12,18 To elucidate the Leukemia

A somatic mutation of the SHIP gene in myeloid leukemia J-M Luo et al

4

Figure 2 Inositol 5⬘-phosphatase activity of purified SHIP protein. (a) Results are presented as % conversion from PIP(3,4,5)P3 to PI(4,5)P2 per 15 min. (b) Western blot of purified SHIP protein with anti-SHIP antibody. Results are representative of at least two experiments.

effects of the SHIP mutation, we examined primary leukemia cells from patient AML19, together with those from a patient (UPN AML30; FAB M2) without any SHIP mutation, for the time-course profile of Akt phosphorylation following IL-3 stimulation. SHIP was expressed at approximately equivalent levels in both patients (data not shown). The Akt protein level was not altered throughout the IL-3 stimulation (Figure 3a, lower panel). The basal level of Akt phosphorylation was comparable between these samples regardless of the SHIP mutation. Whereas the maximal level was almost comparable, the elevated Akt phosphorylation in leukemia cells from patient AML19 was still sustained for at least 60 to 120 min after IL-3 stimulation compared with that of the leukemia cells from patient AML30, which showed a gradual decrease by 120 min (Figure 3a, upper panel). The temporary profile after withdrawal of IL-3 showed an even more striking difference. Although the level of phosphorylated Akt in leukemia cells from patient AML 30 was rapidly reduced to the minimally detectable one by 60 min, in the leukemia cells from patient AML19 a moderate level of phosphorylated form remained even after 120 min (Figure 3b, upper panel). To exclude the possible involvement of a PTEN mutation in the Akt hyperphosphorylation in this case, we conducted RT-SSCP analysis of PTEN as well, but failed to detect any somatic mutations (data not shown). It is intriguing to mention that the results of phenotypic analysis and response to multiple cytokines of hematopoietic progenitors derived from SHIP heterozygous mice were similar to those of the wild-type mice,19,20 although their SHIP expression was reduced by 50%.19 Thus, the one allelic V684E SHIP mutation was found to be linked to the deregulation of Akt phosphorylation and render leukemia cells hypersensitive to IL-3. The oncogenic role of the dysregulated Akt activity probably accounts for the ability to induce multiple simultaneous effects on both cell growth and survival.

SHIP-V684E caused a growth advantage even at a low concentration of serum The requirement of PI3K in cell cycle progression has been indicated by several lines of evidence, such as the microinjecLeukemia

Figure 3 Enhanced phosphorylation of Akt in leukemia cells from patient AML19 with the SHIP V684E mutation. (a) IL-3-induced enhanced and prolonged phosphorylation of Akt in leukemia cells. Leukemia cells were stimulated with 5 ng/ml of IL-3 for the time periods indicated. Western blots using phospho-specific Akt antibody revealed the temporary profile of Akt phosphorylation (upper panel). Other aliquots of the same lysates were loaded in a different gel, and equal loading was confirmed by anti-Akt antibody (lower panel). The precise procedure is described elsewhere (Ref. 12). Leukemia cells from another patient (UPN AML30) with wild-type SHIP were also analyzed. (b) Prolonged elevated Akt phosphorylation level following IL-3 withdrawal. Leukemia cells were stimulated with 5 ng/ml of IL3 for 10 min. Time-points after IL-3 withdrawal are indicated. Akt phosphorylation (upper panel) and the protein level (lower panel) are also shown.

tion of stimulating antibodies against PI3K or dominant-negative forms of p85␣ modulate DNA synthesis induced by growth factor stimulation.21,22 In parallel, LY294002, a selective inhibitor of PI3K, potently inhibited the phosphorylation of Akt and also led to a concomitant accumulation of a G1 phase population.23 Consistent with these observations, PTEN potently suppressed the growth and tumorigenicity of cancer cell lines (reviewed in Ref. 8). Such effects are closely correlated with its ability to modulate G1 cell cycle progression.23,24 We sought to examine the biological consequences of the SHIP mutation to identify its underlying role in leukemogenesis by reintroducing the SHIP cDNA into hematopoietic cell lines. K562 cells were employed to further address this issue. By Western blot analysis we found that K562 cells did express a low level of SHIP (Figure 4a). RTPCR/SSCP detected no mutations of SHIP or PTEN in this cell line (data not shown). Furthermore, PI3K is constitutively activated by BCR-ABL, through both direct phosphorylation of the p85 subunit of PI3K and indirect activation via the Ras pathway.25–27 Because SHIP operates its regulatory activity just downstream of PI3K, K562 was considered to be a desirable cell for functional analysis of SHIP. The V684E muation did not appear to change protein stability, because both SHIP-wt and SHIP-V684E were efficiently expressed at approximately equivalent levels in K562 derivatives (Figure 4a). To analyze their effects on cell growth, we cultured SHIP transfectants in media with different concentrations of serum. K562 cells transfected with SHIP-V684E (SHIP-V684E K562), SHIP-wt (SHIP-wt K562) or empty vector (Mock K562) as a control were seeded at 5 × 104 cells/ml, and duplicate samples were

A somatic mutation of the SHIP gene in myeloid leukemia J-M Luo et al

5

Figure 4 Effect of SHIP mutation on growth and survival of K562 transfectants. (a) SHIP expression in three K562 transfectants. SHIP expression in U937 cells is also shown. Ten micrograms of protein from total cell lysates were subjected to analysis. (b) Akt phosphorylation induced by IL-3 in SHIP transfectants. Each transfectant was cultured in serum-free medium for 24 h prior to IL-3 stimulation. Ten micrograms of protein from total cell lysates were subjected to analysis. Akt phosphorylation was examined before (−) and after (+) IL-3 stimulation. After stripping of the phospho-specific antibody, the same membrane was reprobed with anti-Akt antibody to confirm the equivalent amounts of Akt protein (bottom panel). (c) Each transfectant was seeded into culture at 5 × 104 cells/ml in RPMI1640 supplemented with the indicated serum concentration. The number of viable cells was counted after 72 h and is shown as % proliferation. The number of cells transfected with empty vector alone and cultured in medium containing 5% serum was taken as 100%. Means ± s.d. of three independent experiments are given. (d) Each K562 transfectant was washed twice with PBS and seeded at 5 × 105 cells/ml in serum-deprived medium. The % viability was examined by trypan blue-exclusion test every 24 h up to 6 days. Means ± s.d. of three independent experiments are given.

counted following trypan blue staining every 12 h for 4 days. Figure 4c shows the relative cell counts after 72 h of culture when the cells were growing rapidly before the cultures became confluent. SHIP-wt suppressed the growth, when compared with the empty vector alone. In contrast, the SHIPV684E transfectants grew significantly faster than other transfectants at every concentration of serum tested. It is interesting to note the consistency of these findings with those on PTENtransfected cells: when catalytically inactive PTEN was introduced into U87MG, a glioblastoma cell line, or into AN3CA or RL-95, endometrial carcinoma cell lines lacking PTEN expression, the cells grew to a higher saturation density with anchorage independency than cells transfected with the vector alone.23,24 Furthermore, more aggressive tumor growth was observed in vivo, when mutated PTEN-expressing cells were inoculated into mice.23 Thus, the catalytically inactive PIPases can promote tumorigenesis. Consistent with the difference in proliferation profiles, the level of Akt phosphorylation in SHIPV684E K562 cells was significantly increased; whereas it was reduced in SHIP-wt K562 cells (Figure 4b). Moreover, in parallel with its ability to induce Akt phosphorylation, IL-3 slightly, but significantly, stimulated the growth of SHIP-V684E K562 cells; whereas that of Mock or SHIP-wt K562 cells was barely

affected (data not shown). Taken together, these findings indicate that Akt, at least in part, would participate in the regulation of proliferation of K562 myeloid leukemia cells and their derivatives. Although several mechanisms of the regulation of cell proliferation by inositol phospholipid signaling have been proposed, the most relevant and critical target is the cyclin-dependent kinase inhibitor p27 KIP1.28 Akt represses FKHR-1 activity by phosphorylation to reduce the amount of p27 KIP1 at the transcriptional level.28 Thus, the expression of PTEN reduces both cyclin D-, E-, and A-associated kinase activities.23,24,29 By functional analogy to PTEN, SHIP-V684E, through the deregulation of Akt activity, could drive cell cycle progression by disturbing the activity of these components of the G1/S cell cycle machinery.

SHIP-V684E promotes cell survival under conditions of serum deprivation and causes resistance to VP16induced apoptosis A modestly elevated level of SHIP significantly increases the rate of apoptosis; and although the ability of SHIP to compete with Grb2 for Shc and thereby modulate the Ras signal could Leukemia

A somatic mutation of the SHIP gene in myeloid leukemia J-M Luo et al

6

Leukemia

be involved in part, the regulation of Akt activity is the most important aspect in the SHIP-induced apoptosis in myeloid cell lines.30 Given the increased activity of Akt in leukemia cells expressing SHIP-V684E (Figure 3 and Figure 4b), cell survival was examined under several stress situations. K562 cells undergo typical apoptosis after serum deprivation, even though cells with BCR-ABL chimeric protein are rendered more resistant than BCR-ABL-negative cell lines.14 Figure 4d shows cell viability after serum withdrawal. The temporary profile of % cell viability showed the resistance to serum deprivation in SHIP-V684E cells. Compared to Mock K562 cells, SHIP-V684E K562 cells were significantly more resistant to serum deprivation, especially during the early period from day 3 to day 5, although there was little difference at day 6. In contrast, SHIP-wt K562 cells were the most sensitive among the three transfectants. These results paralleled the level of Akt phosphorylation after IL-3 withdrawal (Figure 3b). Thus, the mutation of SHIP could contribute to the reduced susceptibility to cell death via the elevated Akt activity. Next, we examined genotoxic stress-induced cell death. As shown previously,31 a high concentration of etoposide induced a certain population of K562 cells to undergo typical apoptotic morphological changes, such as cell shrinkage, chromatin condensation, nuclear fragmentation, and internucleosomal DNA fragmentation by 24 h and well over 90% of all the cells by 48 h, although this apoptosis was markedly delayed when compared with that in other sensitive hematopoietic cell lines such as HL-60 (data not shown, and Ref. 31). As shown in Figure 5a and b, the Mock K562 cells showed 45% cell death after 24 h of exposure to etoposide, the same apoptotic response as the parent K562 cells. Strikingly, whereas 65% of the SHIP-wt K562 cells were prone to undergo apoptosis, the SHIP-V684E K562 cells exhibited a considerable resistance to apoptosis, with only 10% of the cells showing apoptotic changes. Furthermore, the results of DNA fragmentation in transfectants shown in Figure 5c, negatively correlated with the level of Akt phosphorylation (Figure 4b). Recent studies have revealed that Akt inhibits the activity of apoptosis signalactivated kinase 1 (ASK1), a mitogen-activated protein kinase kinase kinase, and thereby downstream p38 or JNK, the pathway of which has been shown to play a crucial role in mediating apoptosis induced by a variety of stresses against the cell, especially inflammatory cytokines, serum withdrawal, and genotoxic chemotherapeutic reagents.32,33 Given the growing number of Akt substrates, the molecular basis of the anti-apoptotic effects caused by SHIP-V684E should be explored. To further elucidate the role of SHIP-V684E in leukemogenesis, we used it to transfect U937, a pro-monocytic leukemia cell line with endogenous expression of wild-type SHIP (Figure 4a), and thereby reproduced the same growth-promoting effect as observed in K562 cells (data not shown). Although it seems somehow curious that the SHIP-V684E affected the cells in a dominant-negative manner, there have been some interesting reports in which catalytically defective phosphatases can paradoxically cause hyperphosphorylation of substrates. The first example is MKP-1, a dual specificity phosphatase that dephosphorylates p42-MAPK (mitogen-activated protein kinase). Sun et al34 revealed that the catalytically inactive mutant of MKP-1 specifically binds to the phosphorylated form of the substrate, thus blocking its subsequent dephosphorylation by wild-type phosphatase and augumenting MAP kinase phosphorylation in a dominant-negative manner. Such a ‘substrate-trapping’ action was also reported by Myers et al35 in the case of catalytically defective PTEN, which

Figure 5 Evaluation of SHIP mutation by the induction of apoptosis in K562 transfectants. (a) Morphological aspects of SHIP transfectants stained with Hoechst 33342. SHIP-wt K562 cells before treatment (upper left), and Mock (upper right), SHIP-V684E (lower left), and SHIP-wt (lower right) K562 cells after exposure to 34 ␮M etoposide for 24 h are shown. All three transfectants depicted the same feature before etoposide treatment (data not shown). The characteristic findings of apoptosis are evident. (b) The percentage of apoptotic cells was determined from nuclear staining with Hoechst 33342. (c) DNA fragmentation analysis of K562 cell transfectants. DNA extracted from cells cultured in the absence (−) or presence (+) of 34 ␮M etoposide for 24 h was subjected to agarose gel electrophoresis. Each panel is representative of at least three experiments.

causes the accumulation of PIP(3,4,5)P3 and the phosphorylated form of Akt. This might also be the case of the catalytically defective SHIP-V684E, although it remains to be clearly proven. The effect of SHIP-V684E might also be associated with perturbation of cross-talk between Akt and other signaling pathways through titration of adaptor molecules, and should be further investigated. The presence of a mutation implies a worse prognosis or highly malignant phenotype; however, patient AML19 suffered from acute promyelocytic leukemia, she was successfully treated with all-trans retinoic acid, a potent differentiation-inducing agent. Therefore, clinical resistance to chemotherapy is not relevant in this patient. In view of differential catalytic specificity for PI(3,4,5)P3,36 SHIP may have less potent tumor suppressor activity than PTEN. The SHIP homologue SHIP2, the expression of which is ubiquitous, might partially compensate for the loss of SHIP activity. Indeed SHIP hemizygous-deficient mice do not show tumor susceptibility.19 Loss of SHIP might not be a primary transforming event, but instead promote tumor progression, for example, by preventing cells from undergoing apoptosis while allowing them to accumulate genetic alterations. As in the

A somatic mutation of the SHIP gene in myeloid leukemia J-M Luo et al

case of the co-existence of PTEN and p16 mutations,8 loss of normal SHIP function might confer an additional selective advantage over loss of other tumor suppressor pathways such as RB. The negative regulatory function of SHIP in myeloid cells in vivo, together with biochemical evidences of Akt hyperphosphorylation and subsequent resistance against apoptosis caused by the mutation of SHIP in vitro, suggests that SHIP is another promising target gene responsible for leukemogenesis. Furthermore, SHIP contains functional domains outside the 5⬘-phosphatase domain that act as an organizing scaffold for regulatory molecules involved in phosphoinositide signaling, and so analysis of the entire coding sequence of SHIP should be conducted. Obviously, the SHIP mutation in leukemogenesis in vivo will require further investigation. In conclusion, a catalytically defective somatic mutation within the phosphatase domain of the SHIP gene in primary AML was found to protect cells from apoptosis and to promote cell proliferation. This report provides the first evidence of a mutation of the inositol 5⬘-phosphatase gene in any given human malignant neoplasm and further supports the need for extensive analysis of the SHIP gene. The integrated investigation of critical signaling pathways affected by SHIP dysfunction could yield results that might serve as a basis for a novel and rational strategy for chemotherapy.

Acknowledgements The human SHIP cDNA was kindly provided by Dr Hisamaru Hirai (Tokyo University). We thank Dr Haruhiko Sugimura for useful discussion and Dr Yoshinori Nozawa for critical review of the manuscript. We also thank Drs Kensuke Naito, Shinya Fujisawa, Yota Fujita, Takayuki Matui and Akihito Takeshita, for the management of patients’ samples, as well as Kumi Nishizawa and Terumi Taniguchi for technical assistance. This work was supported in part by a grant-in-aid from the Japanese Ministry of Education, Culture, Sport, and Science, and a grant from the Leukemia Study Group of the Ministry of Health, Labor, and Welfare.

References 1 Odai H, Sasaki K, Iwamatsu A, Nakamoto T, Ueno H, Yamagata T, Mitani K, Yazaki Y, Hirai H. Purification and molecular cloning of SH2- and SH3-containing inositol polyphosphate-5-phosphatase, which is involved in the signaling pathway of granulocyte– macrophage colony-stimulating factor, erythropoietin, and BcrAbl. Blood 1997; 89: 2745–2756. 2 Ware MD, Rosten P, Damen JE, Liu L, Humphries RK, Krystal G. Cloning and characterization of human SHIP, the 145-kD inositol 5-phosphatase that associates with SHC after cytokine stimulation. Blood 1996; 88: 2833–2840. 3 Rohrschneider LR, Fuller JF, Wolf I, Liu Y, Lucas DM. Structure, function, and biology of SHIP proteins. Genes Dev 2000; 14: 505–520. 4 Blume-Jensen P, Hunter T. Oncogenic kinase signaling. Nature 2001; 411: 355–365. 5 Coffer PJ, Jin J, Woodgett R. Protein kinase B (c-Akt): a multifunctional mediator of phosphatidylinositol 3-kinase activation. Biochem J 1998; 335: 1–13. 6 Brunet A, Datta SR, Greenberg ME. Transcription-dependent and independent control of neuronal survival by the PI3K-Akt signaling pathway. Curr Opin Neurobiol 2001; 11: 297–305. 7 Ichijo H, Nishida E, Irie K, Dijke P, Saitoh M, Moriguchi T, Takagi M, Matsumoto K, Miyazono K, Gotoh Y. Induction of apoptosis by ASK1, a mammalian MAPKKK that activates SAPK/JNK and p38 signaling pathways. Science 1997; 275: 90–94.

8 Suzuki A, de la Pompa JL, Stambolic V, Elia AJ, Sasaki T, Barco BI, Ho A, Wakeham A, Itie A, Khoo W, Fukumoto M, Mak TW. High cancer susceptibility and embryonic lethality associated with mutation of the PTEN tumor suppressor gene in mice. Curr Biol 1998; 8: 1169–1178. 9 Maehama T, Dixon JE. The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol-3,4,5-trisphosphate. J Biol Chem 1998; 273: 13375–13378. 10 Sakai A, Thieblemont C, Wellmann A, Jaffe ES, Raffeld M. PTEN gene alterations in lymphoid neoplasms. Blood 1998; 92: 3410– 3415. 11 Aggerholm A, Gronbaek K, Guldberg P, Hokland P. Mutational analysis of the tumour suppressor gene MMAC1/PTEN in malignant myeloid disorders. Eur J Haematol 2000; 65: 109–113. 12 Liu Q, Sasaki T, Kozieradzki I, Wakeham A, Itie A, Dumont DJ, Penninger JM. SHIP is a negative regulator of growth factor receptor-mediated PKB/Akt activation and myeloid cell survival. Genes Dev 1999; 13: 786–791. 13 Sattler M, Verma S, Byrne CH, Shrikhande G, Winkler T, Algate PA, Rohrschneider LR, Griffin JD. BCR/ABL directly inhibits expression of SHIP, an SH2-containing polyinositol-5-phosphatase involved in the regulation of hematopoiesis. Mol Cell Biol 1999; 19: 7473–7480. 14 McGahon A, Bissonnette R, Schmitt M, Cotter KM, Green DR, Cotter TG. BCR-ABL maintains resistance of chronic myelogenous leukemia cells to apoptotic cell death. Blood 1994; 83: 1179– 1187. 15 Hongyo T, Buzard GS, Calvert RJ, Weghorst CM. ‘Cold SSCP’: a simple, rapid and non-radioactive method for optimized singlestrand conformation polymorphism analyses. Nucleic Acids Res 1993; 21: 3637–3642. 16 Damen JE, Liu L, Rosten P, Humphries RK, Jefferson AB, Majerus PW, Krystal G. The 145-kDa protein induced to associate with Shc by multiple cytokines is an inositol tetraphosphate and phosphatidylinositol-3,4,5-triphosphate 5⬘-phosphatase. Proc Natl Acad Sci USA 1996; 93: 1689–1693 17 Jefferson AB, Majerus PW. Properties of type II inositol polyphosphate 5-phosphatase. J Biol Chem 1995; 270: 9370–9377. 18 Liu L, Damen JE, Ware MD, Krystal G. Interleukin-3 induces the association of the inositol 5-phosphatase SHIP with SHP2. J Biol Chem 1997; 272: 10998–11001. 19 Helgason CD, Damen JE, Rosten P, Grewal R, Sorensen P, Chappel SM, Borowski A, Jirik F, Krystal G, Humphries RK. Targeted disruption of SHIP leads to hemopoietic perturbations, lung pathology, and a shortened life span. Genes Dev 1998; 12: 1610– 1620. 20 Huber M, Helgason CD, Damen JE, Liu L, Humphries RK, Krystal G. The src homology 2-containing inositol phosphatase (SHIP) is the gatekeeper of mast cell degranulation. Proc Natl Acad Sci USA 1998; 95: 11330–11335. 21 McIlroy J, Chen D, Wjasow C, Michaeli T, Backer JM. Specific activation of p85-p110 phosphatidylinositol 3⬘-kinase stimulates DNA synthesis by ras- and p70 S6 kinase-dependent pathways. Mol Cell Biol 1997; 17: 248–255. 22 Jhun BH, Rose DW, Seely BL, Rameh L, Cantley L, Saltiel AR, Olefsky JM. Microinjection of the SH2 domain of the 85-kilodalton subunit of phosphatidylinositol 3-kinase inhibits insulininduced DNA synthesis and c-fos expression. Mol Cell Biol 1994; 14: 7466–7475. 23 Li DM, Sun H. PTEN/MMAC1/TEP1 suppresses the tumorigenicity and induces G1 cell cycle arrest in human glioblastoma cells. Proc Natl Acad Sci USA 1998; 95: 15406–15411. 24 Zhu X, Kwon CH, Schlosshauer PW, Ellenson LH, Baker SJ. PTEN induces G(1) cell cycle arrest and decreases cyclin D3 levels in endometrial carcinoma cells Cancer Res 2001 ; 61: 4569–4575. 25 Sattler M, Salgia R, Okuda K, Uemura N, Durstin MA, Pisick E, Xu G, Li JL, Prasad KV, Griffin JD. The proto-oncogene product p120CBL and the adaptor proteins CRKL and c-CRK link c-ABL, p190BCR/ABL and p210BCR/ABL to the phosphatidylinositol-3⬘ kinase pathway. Oncogene 1996; 12: 839–846. 26 Goga A, McLaughlin J, Afar DE, Saffran DC, Witte ON. Alternative signals to RAS for hematopoietic transformation by the BCR-ABL oncogene. Cell 1995; 82: 981–988. 27 Melo JV. The molecular biology of chronic myeloid leukaemia. Leukemia 1996; 10: 751–756.

7

Leukemia

A somatic mutation of the SHIP gene in myeloid leukemia J-M Luo et al

8

Leukemia

28 Bruni P, Boccia A, Baldassarre G, Trapasso F, Santoro M, Chiappetta G, Fusco A, Viglietto G. PTEN expression is reduced in a subset of sporadic thyroid carcinomas: evidence that PTEN-growth suppressing activity in thyroid cancer cells mediated by p27kip1. Oncogene. 2000; 19: 3146–3155. 29 Nakamura N, Ramaswamy S, Vazquez F, Signoretti S, Loda M, Sellers WR. Forkhead transcription factors are critical effectors of cell death and cell cycle arrest downstream of PTEN. Mol Cell Biol 2000; 20: 8969–8982. 30 Liu L, Damen JE, Hughes MR, Babic I, Jirik FR, Krystal G. The Src homology 2 (SH2) domain of SH2-containing inositol phosphatase (SHIP) is essential for tyrosine phosphorylation of SHIP, its association with Shc, and its induction of apoptosis. J Biol Chem 1997; 272: 8983–8988. 31 Martins LM, Mesner PW, Kottke TJ, Basi GS, Sinha S, Tung JS, Svingen PA, Madden BJ, Takahashi A, McCormick DJ, Earnshaw WC, Kaufmann SH. Comparison of caspase activation and subcellular localization in HL-60 and K562 cells undergoing etoposideinduced apoptosis. Blood 1997; 90: 4283–4296.

32 Yamamoto K, Ichijo H, Korsmeyer SJ. BCL-2 is phosphorylated and inactivated by an ASK1/Jun N-terminal protein kinase pathway normally activated at G(2)/M. Mol Cell Biol 1999 ; 19: 8469–8478. 33 Kim AH, Khursigara G, Sun X, Franke TF, Chao MV. Akt phosphorylates and negatively regulates apoptosis signal-regulating kinase 1. Mol Cell Biol 2001; 21: 893–901. 34 Sun H, Charles CH, Lau LF, Tonks NK. MKP-1 (3CH134), an immediate early gene product, is a dual specificity phosphatase that dephosphorylates MAP kinase in vivo. Cell 1993; 75: 487– 493. 35 Myers MP, Pass I, Batty IH, Van der Kaay J, Stolarov JP, Hemmings BA, Wigler MH, Downes CP, Tonks NK. The lipid phosphatase activity of PTEN is critical for its tumor suppressor function. Proc Natl Acad Sci USA 1998; 95: 13513–13518. 36 Stokoe D, Stephens LR, Copeland T, Gaffney PR, Reese CB, Painter GF, Holmes AB, McCormick F, Hawkins PT. Dual role of phosphatidylinositol-3,4,5-trisphosphate in the activation of protein kinase B. Science 1997; 277: 567–570.