Inhibition of Wnt16 in Human Acute Lymphoblastoid Leukemia Cells Containing the T(1;19) Translocation Induces Apoptosis

Home , WNT16, WNT8A, Wnt signaling pathway

Inhibition of Wnt16 in human acute lymphoblastoid leukemia cells containing the t(1;19) translocation induces apoptosis

Julien Mazieres1,2, Liang You1, Biao He1, Zhidong Xu1, Amie Y Lee1, Iwao Mikami1, Frank McCormick1 and David M Jablons*,1

1UCSF Comprehensive Cancer Center, San Francisco, CA 94115, USA; 2Department Innovation The´rapeutique et Oncologie Mole´culaire, INSERM U563, Institut Claudius Regaud, 31052 Toulouse Cedex, France

The Wnt family of secreted glycoproteins is widely reports have demonstrated aberrant activation of the involved in cell proliferation,differentiation and oncogen- Wnt signaling pathway in disparate human cancers esis. Many Wnt signaling genes are upregulated and (Morin et al., 1997; Rhee et al., 2002; Weeraratna et al., activated in chronic lymphocytic leukemia. Less is known 2002). We reported the overexpression of dishevelled concerning acute leukemia. One subtype of acute (Dvl) proteins in thoracic malignancies (Uematsu et al., lymphoblastoid leukemia (ALL) is characterized by a 2003a, b). It is also known that Wnt proteins regulate t(1;19) chromosomal translocation resulting in a fusion early B-cell growth and survival (Reya et al., 2003) and protein E2A–Pbx1 that promotes transformation and the activation of the Wnt signaling pathway has been leukemogenesis. Wnt16 has been shown to be targeted by demonstrated in chronic lymphocytic leukemia (Lu E2A–Pbx1. We performed a differential gene expression et al., 2004). array in acute leukemia cell lines displaying or not Another area of interest regarding the Wnt proteins is displaying the t(1;19) translocation. We found that Wnt16 the acute leukemias that are often characterized by and many Wnt signaling-related genes were upregulated translocations that can determine prognosis and treat- in the translocation-containing cells. As two isoforms of ment options. Muller-Tidow et al. (2004) reported that Wnt16,Wnt16a and Wnt16b,have been recently identi- translocations in acute myeloid leukemia could activate fied,we demonstrated by using RT–PCR and Western the Wnt signaling pathway. Among these chromosomal blot that Wnt16b (and not Wnt16a) is overexpressed in abnormalities, the t(1;19) translocation results in the t(1;19)-containing cell lines. We then directly addressed production of chimeric E2A–Pbx1 proteins that display the role played by both isoforms in this type of leukemia. oncogenic properties (Kamps et al., 1991). The mole- Using specific short interfering RNA (siRNA) and an anti- cular properties of E2A–Pbx1 have been extensively Wnt16 antibody,we showed that targeted-Wnt16b inhibi- characterized and several candidate target genes have tion leads to apoptotic cell death. We also demonstrated been identified. Light has been shed on Wnt16 in recent that Wnt16b mediates its effect through the canonical studies. Representational differential expression analysis Wnt pathway involving dishevelled-2, b-catenin and performed in leukemia cell lines containing the fusion survivin. We thus propose that Wnt16 plays an important protein vs others lacking E2A–Pbx1 identified Wnt16 as role in leukemogenesis,raising its therapeutic interest. a putative target gene (McWhirter et al., 1999). More- Oncogene (2005) 24, 5396–5400. doi:10.1038/sj.onc.1208568; over, gene expression profiling of leukemic blasts published online 20 June 2005 showed a marked overexpression of Wnt16 in E2A– Pbx1-expressing leukemias (Ross et al., 2003). Keywords: Wnt; Wnt16; leukemia; apoptosis; human Nevertheless, the specific role played by Wnt16 and antibody; RNA interference the Wnt-related proteins in acute leukemia has never been addressed so far. Moreover, Fear et al. identified two isoforms of Wnt16, Wnt16a and Wnt16b, and the respective role of each isoform in oncogenesis has also The Wnt family of secreted glycoproteins is a group of yet to be elucidated. signaling molecules that is widely involved in develop- In order to analyse the Wnt signaling pathway mental processes and oncogenesis. Currently, 19 human activation, we first performed gene expression profiling Wnt proteins have been identified. The proto-oncogenic using a custom array designed to profile the expression effects of Wnt were discovered more than 20 years ago of the Wnt signaling pathway-related genes (GEArray Q (Nusse and Varmus, 1982) and since then numerous Human Wnt Signaling Pathway Gene Array, Super- Array). Four acute lymphoblastic leukemia cell lines were analysed, two containing the t(1;19) translocation *Correspondence: DM Jablons, Department of Surgery, Cancer (697 and RCH-ACV) and two without the translocation Center, 1600 Divisadero St, C322C, Box 1674, San Francisco, CA 94115, USA; E-mail: [email protected] (CCL-119 and NALM-6). We first confirmed the Wnt16 Received 28 July 2004; revised 17 January 2005; accepted 24 January overexpression in both 697 and RCH-ACV cell lines, 2005; published online 20 June 2005 whereas no signal was seen in the control cell lines. Wnt16 in t(1;19)-leukemia J Mazieres et al 5397 Moreover, many other genes involved in the Wnt cell lines (Figure 2b), whereas Wnt16a was not detected canonical pathway, such as b-catenin, Dvl-2 and Tcf-4, (data not shown). As Fear et al. suggested that Wnt16a were overexpressed, underlining the activation of the was more likely to be targeted by E2A–Pbx1 than Wnt signaling pathway in those cells (Figure 1). We Wnt16b, we confirmed by sequencing analysis that the should note that other Wnt proteins (Wnt2 and Wnt6) amplified band corresponded strictly to Wnt16b are overexpressed in CCL-119, suggesting that various (Figure 2c). Wnt proteins may be upregulated in different leukemia As Wnt16b appeared to be a putative target to subtypes in agreement with the findings of Muller- mediate the effects of the fusion protein E2A–Pbx1, we Tidow et al. (2004). thus studied its ability to control apoptosis and cell To further confirm the previous data, we analysed survival. We designed specific siRNA for Wnt16a and Wnt16 expression by both Western blot and RT–PCR. Wnt16b. Four siRNA were used, one nonsilencing We first demonstrated, by using a commercially avail- siRNA, one siRNA targeting both Wnt-16a and able Wnt16 polyclonal antibody (Santacruz Biotechno- three siRNA targeting specifically Wnt-16b. SiRNAs logies), that Wnt16 is not expressed in both NALM-6 were purchased from Qiagen (Valencia, CA, USA) and and CCL-119 control cell lines and is highly over- sequences were the following: expressed in both 697 and RCH-ACV (Figure 2a). Wnt16a: r(AGAUGGAAAGGCACCCACC)d(TT) Moreover, as Fear et al. showed that Wnt16 consists of and r(GGUGGGUGCCUUUCCAUCU)d(TT) two isoforms, we analysed by RT–PCR the expression Wnt16b1: r(UGGCAUUGCAACCAGAGAG)d(TT) of both Wnt16a and Wnt16b in the same cell lines. and r(CUCUCUGGUUGCAAUGCCA)d(TT) RT–PCR (SuperScript One-Step RT–PCR with Plati- Wnt16b2: r(GGAAACUGGAUGUGGUUGG)d(TT) num Taq, Invitrogen) was performed with the same and r(CCAACCACAUCCAGUUUUCC)d(TT) primers as those used by Fear et al. (2000) (Wnt16aF: Wnt16b3: r(UGCAACCGUACAUCAGAGG)d(TT) CAGAAAGATGGAAAGGCACC, Wnt16bF: TGCT CGTGCTGTTCCCCTAC, Wnt16aR and Wnt16bR: and r(CCUCUGAUGUACGGUUGCA)d(TT) ATCATGCAGTTCCATCTCTC). We showed a We followed the protocol proposed by Tuschl et al. marked overexpression of Wnt16b in t(1;19)-containing (Elbashir et al., 2002). After siRNA transfection, plates

a CCL-119 697 b AES APC AX1 AX1 BMP4 BT Cyclin Cyclin N1 N2 RC D1 D1

Cyclin CDD CDX CLD COL1 CT CTB Botenin D1 44 1 N1 A1 BPL P1 CTN DKK DKK DKK DVL2 EFN EGR ENP NBIP 1 2 4 B1 1 P1

p300 HST Fre-1 FRA FRAT FRZ FST FZD T1 2 B 1 FZD FZD FZD FZD FZD6 FZD FZD FZD 10 2 3 4 6 7 8 FZD GAS GIA GSK GSK3 ICA ID2 ID3 9 1 3A M-1

JUN LEF1 LRP LRP MMP MM MSX MSX 5 6 26 P7 1 2 C- NKD NKD NOS PTTX uPA PPN PPP2 myc 1 2 2 R RSD c Upregulated genes in untreated CCL-119 and 697 cell lines Cox- C- SFR SFR SM0 S0 T Tcf-3 697 CCL-119 2 Ret P2 P4 H X17

APC CD44 Tcf-4 Tcf-7 VEG WIF WISP WIS WIS WNT cyclin D3 Ephrin D1 F 1 1 P1 P3 1 beta-catenin frizzled 5 WNT WNT WNT WNT WNT WNT WNT WNT dvl-2 gastrin 10A 11 14 15 16 2 2B 3 frizzled 3 SOX17 WNT WNT WNT WNT WNT WNT WNT WNT GSK-3 Wnt6 3A 4 5A 5 6 7A 7B 8A LRP6 Wnt2 p300 Wnt8a PUC PUC PUC Bkn Bknk Bkn GAP GAP Tcf-4 18 18 18 k k D D Wnt16 PP1A PP1A PP1A PP1A PPL1 RPL ACT ACT 3A 13A B B

Figure 1 Wnt signaling pathway gene expression profiling in leukemia cell lines. (a) RNA was first extracted from leukemia cell lines. Both of them, 697 and RCH-ACV, contain the t(1;19) translocation, whereas CCL-119 does not and serves as a control. Only the data obtained from CCL-119 and 697 are shown here. After extraction, RNA was subjected to a reverse transcriptase reaction and cDNA probes were labeled with biotin-16-dUTP, and hybridized with the Wnt-specific arrays. Detection was carried out using a chemiluminescent reaction and the membranes were exposed to X-ray film. Wnt16 is surrounded by a black square. The genes that are upregulated in the t(1;19)-containing cell line are surrounded by black circles on the arrays. (b) Data were then matched against the gene list of the GEArray Q series human Wnt signaling pathway array provided by the manufacturer. (c) Upregulated and downregulated genes in both cell lines are detailed

Oncogene Wnt16 in t(1;19)-leukemia J Mazieres et al 5398 a Yang at the NHGRI/NIH. This Wnt16 cDNA is unable to bind to the Wnt16 siRNA we are using that is 0 NALM-6 CCL-119 697 RCH-ACV specially designed to bind the 3 -UTR region of Wnt16 Cell lines mRNA; therefore, can only knock down endogeneous Wnt16b Wnt16. We demonstrated that re-expression of Wnt16 can rescue the cells from the siRNA antibody-induced apoptosis (Figure 3b). In the canonical Wnt pathway, Wnt proteins bind to β-actin frizzled receptors, activate Dvl proteins, which inhibits the ability of glycogen synthase kinase-3b to phosphory- b late b-catenin that can in turn enter the cell nucleus, and activate target genes. We thus analysed the expression of Cell lines NALM-6 CCL-119 697 RCH-ACV the Wnt pathway’s downstream effectors. Western blot analysis was performed on the three cell lines using Wnt16 commercially available antibodies (b-actin, SantaCruz Biotechnologies; b-catenin, Pharmingen; Dvl-2, Santa- Cruz Biotechnologies; Survivin, SantaCruz Biotechnol- β ogies). We showed that Wnt16b inhibition led to a -actin decrease in b-catenin, Dvl-2 and survivin expression in t(1;19)-containing cell lines, whereas no effect was observed on the control cell lines (Figure 3c). To further c confirm our findings, we also performed a Wnt signaling specific microarray as previously described in CCL-119 30 90 100 110 120 and 697 cell lines treated with either control (nonsilencing) siRNA or Wnt16b siRNA. We first confirmed the specific inhibition of Wnt16 and showed that this inhibition led also to the downregulation of key components of the Wnt pathway, such as Dvl-2, Tcf-4 and p300 and c-myc (Figure 3d). Figure 2 Expression of Wnt16 in leukemia cell lines. (a) RT–PCR To further confirm and extend our above findings, we was performed for the four cell lines (697 and RCH-ACV contain used a commercially available anti-Wnt16 antibody (BD the t(1;19) translocation, whereas CCL-119 and NALM-6 do not) pharmingen). Apoptotic assay was performed 3–4 days using specific primers for Wnt16b. The fragment of human Wnt16b amplified is 236 bp. Actin primers were used to amplify b-actin as a afterwards. Anti-Wnt16 antibody induced apoptosis in control. (b) Western blot analysis was also performed. Proteins RCH-ACV cell line, whereas no effect was noticed in the from all three cell lines were extracted and an equal amount control cell line (Figure 4). We then analysed if re- (20–30 mg/lane) was subjected to immunoelectrophoresis and expression of Wnt16b can rescue the cells from the probed with a commercially available polyclonal antibody against Wnt16. Membranes were reprobed with an anti-b-actin antibody monoclonal antibody-induced apoptosis. We demon- as a control. (c) We confirmed by sequence analysis that the strated that re-expression of Wnt16 can rescue the amplified band corresponds strictly to Wnt16b. We show here a cells from the monoclonal antibody-induced apoptosis portion of the first exon of Wnt16b, which is known to be different (Figure 4). from the first exon of Wnt16a The t(1;19) chromosomal translocation results in the production of a fusion production E2A–Pbx1. The activated oncoprotein E2A–Pbx1 induces transforma- were incubated from 72 to 96 h before further analysis. tion in several cell types in vitro and induces lympho- We first demonstrated by both RT–PCR and Western blastic lymphomas in transgenic mice (Kamps et al., blot that Wnt16b-specific siRNA inhibited Wnt16b, 1991). Recently, expression profiles obtained using whereas Wnt16a siRNA and control siRNA had no microarrays identified t(1;19) (E2A–Pbx1) as one of effect (Figure 3a). We then assessed apoptosis by the six prognostic subtypes of pediatric ALL (Yeoh staining the cells with annexin-V and propidium iodide et al., 2002). E2A–Pbx1 is known to induce tumour (PI) (Apotarget, BioSource International). Flow cyto- formation in nude mice (Kamps et al., 1991) and to metry analysis (FACScan, Decton Dickinson) revealed promote myeloid leukemias (Kamps et al., 1996). The that Wnt16b-specific siRNA induced apoptosis and cell mechanisms by which E2A–Pbx1 produces pre-B-cell death compared to control. We used three different ALL still remain unclear (Aspland et al., 2001). We Wnt16b siRNA and obtained similar results in various propose that Wnt16b is of paramount importance in this experiments. All Wnt16b-specific siRNAs led to a subtype of leukemia. The specific function of Wnt16 minimal of 35% apoptosis induction (Figure 3b). was first addressed by McWhirter et al. (1999), who Wnt16a siRNA had no effect on cell survival in the cell proposed that Wnt16 might contribute to the develop- lines tested (data not shown). We then tested if ment of t(1;19) pre-B ALL. Using custom diagnostic coexpression of Wnt16b cDNA can rescue the cells microarrays, Ross et al. (2003) reported significant from the siRNA antibody-induced apoptosis. We overexpression of Wnt16 ranging from a 569- to 2547- obtained the Wnt16 pcDNA3 plasmid from Dr Yingzi fold change when comparing the mean signal value in

Oncogene Wnt16 in t(1;19)-leukemia J Mazieres et al 5399 the E2A–Pbx1-expressing leukemia to other leukemia. E2a–Pbx1 have been proposed. Fu and Kamps (1997) Wnt16 is also overexpressed in chronic lymphocytic first analysed genes induced by E2a–Pbx1 in fibroblasts leukemia (Lu et al., 2004). Other putative targets for and identified some tissue-specific and developmentally regulated genes. In human t(1;19)-containing pre-B-cell ALL, they reported that EB-1 overexpression could a Cell lines CCL-119 697 RCH-ACV interfere with proliferation (Fu et al., 1999). Recently, siRNA cont. 16a 16b cont. 16a 16b cont. 16a 16b Smith et al. (2003) demonstrated that E2a–Pbx1 Wnt16b induction enhances expression of Bmi-1, a lymphoid oncogene whose product functions as a transcriptional β-actin Cell lines CCL-119 697 RCH-ACV siRNA cont. 16a 16b cont. 16a 16b cont. 16a 16b 40 Wnt16 35 Control mAb Wnt16 mAb + empty vector β -actin 30 Wnt16 mAb + Wnt16 plasmid

25 b 50

40 20

30 15

20 apoptotic cells (%) 10

10 5 Apoptotic cells (%) 0 CCL-119 697 RCH-ACV 0 NALM-6 RCH-ACV Control siRNA Wnt16 siRNA + empty vector Figure 4 Wnt16 antibody induces apoptosis through the canoni- Wnt16 siRNA + Wnt16 plasmid cal Wnt pathway. NALM-6 and RCH-ACV were treated with an CCL-19 697 RCH-ACV anti-Wnt16 antibody (BD Pharmingen). Cells were plated in six- jm122303.003 j303009 jm122303.024 well plates and treated the day after as following: control antibody (white), Wnt16 antibody alone (grey) and Wnt16 antibody plus Wnt16 cDNA (black). At 4 days after the treatment, cells were

control FL2-H FL2-H FL2-H siRNA collected, stained by propidium iodide and annexin V-FITCand subjected to ﬂow cytometry. The mean value of apoptotic cell rate FL1-H P.I. FL1-H FL1-H was calculated from three independent experiments. The bar graph shows the average of apoptotic rate and error bars are s.d. Wnt16b siRNA FL2-H FL2-H FL2-H

FL1-H FL1-H FL1-H Figure 3 Wnt16b inhibition by siRNA. (a) Cells were transfected Annexin V-FITC by Wnt16b-specific siRNA (cont: nonsilencing siRNA, 16a: Wnt16a-specific siRNA and 16b: Wnt16b-specific siRNA) and, 3 c Cell lines CCL-119 697 RCH-ACV days afterwards, RNA was extracted, quantified and an RT–PCR was performed using the same primers as previously described. siRNA cont. 16a 16b cont. 16a 16b cont. 16a 16b Actin serves as a control. The same cells were also transfected Dvl-2 according to the same procedure. Proteins were extracted 3 days after the transfection and Western blot was performed with an anti- Wnt16 polyclonal antibody. (b) Cells were transfected by control β-catenin siRNA (white), Wnt16b siRNA þ empty vector (grey) and Wnt16b siRNA þ Wnt16b cDNA (black), and 3 days later were subjected to an apoptotic assay by flow cytometry as described in Materials and Survivin methods. The mean value of apoptotic cell rate was calculated from three independent experiments. The bar graph shows the average of β-actin apoptotic rate and error bars are s.d. We report examples of apoptosis analysis by flow cytometry. X-axis (FL1-H) represents annexin V-FITCstaining and Y-axis (FL-3H) represents propi- d CCL-119 697 dium iodide (PI) staining. The upper row shows cell lines treated by Cont 16b Cont 16b control siRNA and the lower row shows the same cell lines treated with Wnt16b siRNA. (c) Cell lines were transfected by control Actin siRNA, Wnt16a siRNA and Wnt16b siRNA, and, 3 days after the Wnt16 transfection, proteins were extracted and a Western Blot analysis was performed with Dvl-2, b-catenin and survivin antibodies as Dvl-2 described in Materials and methods. Actin blotting was carried out p300 as a control. These experiments have been performed three times with similar results. (d) Wnt signaling specific arrays were cmyc performed as described previously. Some differentially expressed genes after Wnt16b siRNA transfection are shown in CCL-119 and Tcf4 697 cell lines. Nonsilencing siRNA was used as a control (cont.)

Oncogene Wnt16 in t(1;19)-leukemia J Mazieres et al 5400 repressor of the INK4A-ARF tumour suppressor locus. of Wnt signaling in oncogenesis and extend our previous Based on microarray data, Downing also proposed C- findings in solid tumours to hematopoietic malignancies. Mer as testable potential target due to its overexpression We provide new insights into the role played by Wnt16 and its ability to cause transformation (Carroll et al., in this subset of leukemia displaying the t(1;19) 2003). The respective role of these putative target genes translocation. Our findings raise the therapeutic interest and their likely interactions remain largely unknown of targeting Wnt16 in this disease. and warrant further investigations in order to define which one(s) could represent a relevant clinical target. The data here demonstrate, for the first time, an Acknowledgements important function of Wnt16 in leukemogenesis. Here, We thank Drs Mignon Loh and Kevin Shannon for providing we used two different techniques to inhibit Wnt16, RNA us RCH-ACV cell lines, Sarah Meles and the flow cytometry small interference and antibody. Both approaches led to laboratory at UCSF Cancer Center for flow cytometry analysis the same observations: inhibition of Wnt16 induces and the genome core analysis laboratory at UCSF Cancer apoptosis through the canonical Wnt pathway. We can Center for sequence analysis. Wnt16 cDNA plasmid was kindly provided by Dr Yingzi Yang at the NHGRI/NIH. This thus hypothesize that aberrant Wnt signaling mediated work was supported by the Larry Hall memorial trust, the by Wnt16 contributes to the failing in apoptosis that is Kazan, McClain, Edises, Abrams, Fernandez, Lyons & Farrise the signature of this highly aggressive malignancy. Our Foundation and the Fondation pour la Recherche Medicale results added to other reports (Reya et al., 2000; Muller- (JM) and Association Nationale pour le Traitement A Tidow et al., 2004; Lu et al., 2004), raise the importance Domicile des Insuffisants Respiratoires (JM).

References

Aspland SE, Bendall HH and Murre C. (2001). Oncogene, 20, Ruthardt M, deVos S, Hiebert SW, Koefﬂer HP, Berdel WE 5708–5717. and Serve H. (2004). Mol. Cell. Biol., 24, 2890–2904. Carroll WL, Bhojwani D, Min DJ, Raetz E, Relling M, Nusse R and Varmus HE. (1982). Cell, 31, 99–109. Davies S, Downing JR, Willman CL and Reed JC. Reya T, Duncan AW, Ailles L, Domen J, Scherer DC, Willert (2003). Hematology (Am. Soc. Hematol. Educ. Program), K, Hintz L, Nusse R and Weissman IL. (2003). Nature, 423, 102–131. 409–414. Elbashir SM, Harborth J, Weber K and Tuschl T. (2002). Reya T, O’Riordan M, Okamura R, Devaney E, Willert K, Methods, 26, 199–213. Nusse R and Grosschedl R. (2000). Immunity, 13, 15–24. Fear MW, Kelsell DP, Spurr NK and Barnes MR. (2000). Rhee CS, Sen M, Lu D, Wu C, Leoni L, Rubin J, Corr M and Biochem. Biophys. Res. Commun., 278, 814–820. Carson DA. (2002). Oncogene, 21, 6598–6605. Fu X and Kamps MP. (1997). Mol. Cell. Biol., 17, 1503–1512. Ross ME, Zhou X, Song G, Shurtleff SA, Girtman K, Fu X, McGrath S, Pasillas M, Nakazawa S and Kamps MP. Williams WK, Liu HC, Mahfouz R, Raimondi SC, Lenny (1999). Oncogene, 18, 4920–4929. N, Patel A and Downing JR. (2003). Blood, 102, 2951–2959. Kamps MP, Look AT and Baltimore D. (1991). Genes Dev., 5, Smith KS, Chanda SK, Lingbeek M, Ross DT, Botstein 358–368. D, van Lohuizen M and Cleary ML. (2003). Mol. Cell, 12, Kamps MP, Wright DD and Lu Q. (1996). Oncogene, 12, 393–400. 19–30. Uematsu K, He B, You L, Xu Z, McCormick F and Jablons Lu D, Zhao Y, Tawatao R, Cottam HB, Sen M, Leoni LM, DM. (2003a). Oncogene, 22, 7218–7221. Kipps TJ, Corr M and Carson DA. (2004). Proc. Natl. Acad. Uematsu K, Kanazawa S, You L, He B, Xu Z, Li K, Peterlin Sci. USA, 101, 3118–3123. BM, McCormick F and Jablons DM. (2003b). Cancer Res., McWhirter JR, Neuteboom ST, Wancewicz EV, Monia BP, 63, 4547–4551. Downing JR and Murre C. (1999). Proc. Natl. Acad. Sci. Weeraratna AT, Jiang Y, Hostetter G, Rosenblatt K, Duray P, USA, 96, 11464–11469. Bittner M and Trent JM. (2002). Cancer Cell, 1, 279–288. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Yeoh EJ, Ross ME, Shurtleff SA, Williams WK, Patel D, Vogelstein B and Kinzler KW. (1997). Science, 275, Mahfouz R, Behm FG, Raimondi SC, Relling MV, Patel A, 1787–1790. Cheng C, Campana D, Wilkins D, Zhou X, Li J, Liu H, Pui Muller-Tidow C, Steffen B, Cauvet T, Tickenbrock L, Ji P, CH, Evans WE, Naeve C, Wong L and Downing JR. (2002). Diederichs S, Sargin B, Kohler G, Stelljes M, Puccetti E, Cancer Cell, 1, 133–143.

Oncogene