Oncogene (1998) 16, 2029 ± 2032  1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00 http://www.stockton-press.co.uk/onc SHORT REPORT Bcr/Abl associated leukemogenesis in bcr null mutant mice

Jan Willem Voncken1, Vesa Kaartinen, John Gro€en and Nora Heisterkamp

Department of Pathology Ms#103, Section of Molecular and Childrens Hospital of Los Angeles Research Institute, 4650 Sunset Boulevard, Los Angeles, California 90027, USA

The BCR contributes to Philadelphia-positive translocation with 9 in Philadelphia leukemogenesis via a number of discrete mechanisms, (Ph)-positive leukemia (de Klein et al., 1982). This one of which may be through interaction of its normal [t(9;22)(q34;q11)] translocation fuses BCR gene se- gene product with the Bcr/Abl oncoprotein. In the quences to the ABL proto-oncogene. Depending on current study this hypothesis was tested in vivo by where in the BCR the breakpoint occurs, a P190 introducing a Bcr/Abl P190 transgene into mice lacking or a P210 kDa chimeric Bcr/Abl oncoprotein is endogenous bcr . Our ®nding, that the P190 produced. At the molecular level, the BCR gene BCR/ABL oncogene is still capable of producing contributes to leukemogenesis through a number of leukemia in these mice with indistinguishable latency discrete mechanisms. Firstly, the BCR promoter and clinical pattern as in genetically matched counter- controls expression of both the remaining non- parts, rules out any signi®cant or major contribution of rearranged BCR allele and the BCR/ABL oncogene the bcr protein as a whole to leukemia development in in human leukemia. The observation that the these mice. physiological e€ects of bcr gene ablation in mice are found in those cell types also involved in Ph-positive Keywords: P190 Bcr/Abl; bcr null mutant; pre-B leukemia in man (Voncken et al., 1995a) most likely leukemia/lymphoma; Ph translocation pertains to di€erential transcriptional regulation of the BCR gene in cells that have respiratory burst activity (granulocytes, macrophages and B-cells). Secondly, N-terminal Bcr sequences are directly The BCR gene product has been implicated in responsible for the activation of the Abl tyrosine regulation of the Rho family of small GTP-binding kinase in the chimeric Bcr/Abl gene products (Muller Rho, Rac and Cdc42 through its exon 3 ± 10 et al., 1991; McWhirter et al., 1993; McWhirter and encoded GDP/GTP exchange factor (GEF) domain Wang, 1993, 1997). Thus, through its fusion to Abl, (Eva and Aaronson, 1985; Hart et al., 1991; Chuang et Bcr also contributes at a biochemical level to neoplastic al., 1995) and its C-terminal GTPase activating protein transformation of white blood cells. Finally, Bcr is able (GAP) activity (Diekmann et al., 1991; Hart et al., to form homotetramers through an N-terminal 1992). BcrGAP activity toward Rac and CDC42 is oligomerization domain constituted by amino acids currently the only well-understood biochemical prop- 28 ± 68 (McWhirter et al., 1993). Because both P210 erty of p160Bcr. By means of gene knock-out and P190 Bcr/Abl proteins retain this moiety, Bcr-Bcr/ technology, p160Bcr was found indispensable in vivo Abl heteromerization will occur in cells expressing both for regulation of respiratory burst activity in neutro- proteins. Indeed, Bcr consistently co-precipitates with phils through control of the cellular Rac-GTP/Rac- Bcr/Abl from Ph-positive cell lines and is a target for GDP pool during priming and activation (Voncken et the Bcr/Abl tyrosine kinase (Liu et al., 1993; Lu et al., al., 1995a). The Bcr protein also harbors a novel 1993; Pendergast et al., 1993; Puil et al., 1994). serine/threonine kinase activity (Stam et al., 1987; Li et Phosphorylation of tyrosine residues in the normal al., 1989; Maru and Witte 1991). BCR exon 1 Bcr protein by Bcr/Abl inhibits the Bcr serine/ sequences that encode this catalytic activity are unique threonine kinase encoded by BCR exon 1 and in the mammalian genome. Finally, Bcr carries several conversely, a serine-phosphorylated Bcr peptide from domains, including a peptide sequence bearing some exon 1 inhibits the Bcr/Abl tyrosine kinase activity (Liu homology to pleckstrin which potentially direct et al., 1996a,b). subcellular localization and interaction with other We have previously generated mice transgenic for cellular proteins. The biological signi®cance of these BCR/ABL P190 and P210 which develop leukemia/ domains and of the serine/threonine kinase domain in lymphoma (Heisterkamp et al., 1990; Voncken et al., vivo is as of yet unclear. 1992, 1995b). In a separate study, we have eliminated BCR was ®rst identi®ed as a gene encompassing the the cellular bcr function in mice through gene targeting breakpoint cluster region, a region on human (Voncken et al., 1995a). The availability of both mouse which undergoes a reciprocal models for Philadelphia-positive leukemia and a bcr mutant lacking bcr protein allowed us to test the hypothesis that the Bcr protein as a whole contributes to leukemia through its interaction with Bcr/Abl. Correspondence: N Heisterkamp Leukemogenesis in the BCR/ABL P190 transgenic 1 Present address: The Netherlands Institute, Division of mice is usually accompanied by the development of Molecular Carcinogenesis, H-2, Plesmanlaan 121, 1066 CX lymphomas which largely consist of malignant Amsterdam, The Netherlands Received 10 September 1997; revised 19 November 1997; accepted 19 lymphoblasts. The bcr protein is expressed in most November 1997 normal tissues but the highest expression levels are Leukemia in bcr null mutants JW Voncken et al 2030 found in brain (Figure 1, lane 4). In the lymphomas, row involvement and asymmetric lymph node swelling, the endogenous p160bcr is expressed at a very high level, paralysis) developed. Soft tissue invasion was compar- which is comparable to that in brain (Figure 1 compare able in BCR/ABL-bcr(7/7) and in (+/+) animals. lanes 1 ± 3 with lane 4) and exceeds that found in bone We also introduced a BCR/ABL P210 transgene (1994- marrow of disease-free animals (not shown). The bcr line; Voncken et al., 1995) into the bcr(7/7) back- protein detected in these lymphoblasts migrates as a ground and obtained leukemic animals, indicating that double band (Figure 1). Fibroblasts transfected with the P210 protein does not require the presence of Bcr/Abl but not non-transfected controls also show a endogenous bcr protein for leukemogenesis (not double band of endogenous bcr protein (not shown). shown). This indicates that, similar to the CML cell lines, the Our ®nding, that the P190 BCR/ABL oncogene is endogenous bcr interacts with the Bcr/Abl protein and still capable of producing leukemia in transgenic mice becomes tyrosine phosphorylated in our transgenic when bred into the bcr(7/7) background, with mouse model (Liu et al., 1996a). indistinguishable latency and clinical pattern as in In earlier investigations, the development of genetically matched (+/+) P190 BCR/ABL counter- leukemia was studied in the P190 BCR/ABL trans- parts, rules out any signi®cant or major contribution of

genic mouse line 623, which was of a B6 CBA genetic the bcr protein as a whole to the oncogenic process. It background (Voncken et al., 1992). Transgenic 623 remains formally possible that some of the enzymatic o€spring died reproducibly of pre-B cell acute activity exhibited by the endogenous bcr protein is lymphoblastic leukemia (ALL) or lymphoblastic leukemia lymphoma; 50% mortality was approxi- mately at 10 weeks. The bcr(7/7) genotype was established in an 129Sv background. Introduction of a

129Sv into the B6 CBA P190 transgenic background resulted in a signi®cantly increased latency of BCR/ABL-mediated leukemia development both in female and male wild types (Figure 2, compare open diamonds with small dots). There was no signi®cant di€erence in tumor incidence, nor latency to development of leukemia between genetically matched BCR/ABL-bcr(+/+), BCR/ABL-bcr(7/7) and BCR/ABL-bcr(+/7) ani- mals of either sex. Fifty percent mortality for all `129Sv-background' transgenic female genotypes was about 22 weeks (Figure 2a) whereas for males this was at more than 1 year (Figure 2b). The di€erence between males and females may be caused by possible hormone-responsive elements in the transgenic P190 construct, e.g., in the MT promoter segment. Gross examination of terminally ill animals revealed b comparable pathology, independent of the genetic status of the bcr locus. The type of neoplasia that developed was uniform, regardless of genotype: typically, either ALL (high peripheral white blood cell count, splenomegaly, limited lymph node and bone marrow involvement, hind limb paralysis) or lympho- blastic leukemia/lymphoma (splenomegaly, bone mar-

1 2 3 4

Figure 2 Survival curves of BCR/ABL P190 transgenics. The BCR/ABL transgenes were bred into the bcr(7/7) background by mating double transgenic (TG) males of a B6CBA background to bcr(7/7) females of a hybrid B6CBA/129Sv background. Heterozygous bcr(+/7) transgenic animals were subsequently bred to each other to obtain, with exception of the bcr null , genetically matched wild type/TG and null mutant/TG Figure 1 Western blot analysis of p160bcr in lymphomas of BCR/ animals. (a) Females. TG P190 bcr null mutants [(7/7), n=15; ABL P190 transgenic mice. Lanes 1 ± 3 contain lysates from three open circles], TG bcr heterozygotes [(7/+), n=19, open squares] di€erent terminally ill animals, lane 4 a brain lysate from a and TG wild types [(+/+), n=42; open diamonds] are compared. disease-free wild type animal. The arrows point to the bcr TG wild type animals of a B6CBA background [(+/+), n=115; doublets in the lymphomas and to the single bcr band in brain. 20 small dots] are also shown. (b) Males. TG P190 bcr null mutants mg of total cellular protein was loaded per lane. Bcr protein was open circles, n=12; TG bcr heterozygotes, open squares, n=21; detected with monoclonal antibody AB-2 (Oncogene Science, Inc., TG wild types, open diamonds, n=33. TG animals of a B6CBA NY) which does not react with Bcr/Abl P190 background (small dots, n=56) are also shown Leukemia in bcr null mutants JW Voncken et al 2031 1) of the endogenous bcr protein appears to be phosphorylated by Bcr/Abl in the P190 bcr(+/+) mice and the remaining 50% would be still active. The biological signi®cance of the intrinsic serine/ threonine kinase activity of Bcr is unknown since the bcr null mutant has no phenotypic or other evident abnormalities which are easily associated with lack of this activity. However, the GTPase activating protein (GAP) activity of Bcr does have a clear biological function (Voncken et al., 1995) and its substrates, members of the Rho family of , are implicated in cell proliferation and tumorigenesis (Olson et al., 1995; Qui et al., 1995; Khosravi-Far et al., 1995). To investigate the possible e€ect of tyrosine phosphorylation of Bcr by Bcr/Abl on the bcrGAP activity, both proteins were individually expressed in Figure 3 Comparison of GAP activity of phosphorylated and insect cells and isolated. Bcr protein was subsequently non-phosphorylated Bcr. GTP-loaded GST-Rac1 was incubated tyrosine-phosphorylated by Bcr/Abl. When phosphory- in the presence of phosphorylated GST-Bcr (squares), non- lated and non-phosphorylated Bcr were compared for phosphorylated GST-Bcr (circles) and in the absence of GST- Bcr (triangles). GST-Rac1, GST-Bcr and GST-Bcr/Abl P210 their ability to stimulate the GTPase activity of Rac1, produced in Sf9 cells were puri®ed as described (Haataja et al., no di€erences were observed (Figure 3). Therefore 1997). GST-Bcr (500 ng) was tyrosine phosphorylated by 100 ng these results do not support the hypothesis that of GST-Bcr/Abl in 20 mM PIPES bu€er, pH 7.0, tyrosine-phosphorylation of Bcr e€ects one of its containing 20 mM MnCl and 1 mM ATP. Ecient transphos- 2 major biological functions. phorylation of Bcr by Bcr/Abl was veri®ed by kinase assay as described (Li et al., 1988). GAP activity was determined by Within the context of the Bcr/Abl protein, the size measuring GTP hydrolysis as described (Chuang et al., 1995). and primary sequence of the Bcr moiety was shown to Brie¯y, Rac1 was loaded with [g-32P]GTP by incubation of the be important for the generation of leukemia. The GST-Rac1 fusion protein in 25 mM Tris-HCl, pH 7.5, 4.7 mM results presented in this study limit the contribution of EDTA, 1 mM DTT, 0.1 mM MgCl2 0.1mg/ml bovine serum 32 Bcr in Ph-positive leukemia, since we found no albumin, 20 mM unlabeled GTP and 10 mCi [g- P]GTP (5000 Ci/ mmol, Amersham) at 308C for 4 min. Equally loaded aliquots (as evidence to support the hypothesis that the endogen- measured by c.p.m. bound) of Rac were used for GAP assays. ous normal cellular Bcr protein as a whole has an e€ect Hydrolysis was initiated by addition of 50 ng of either non- on the process of leukemogenesis caused by Bcr/Abl. phosphorylated or tyrosine-phosphorylated GST-Bcr and MgCl2 and GTP to ®nal concentrations of 17 mM and 170 mM, respectively. The [g-32P]GTP bound to Rac1 was determined by ®ltration analysis (Chuang et al., 1995) Acknowledgements Supported by PHS NIH grants CA47456 (JG), CA 50248 down-regulated in vivo through tyrosine-phosphoryla- (NH); ACS grant DB-70A (NH) and a Childrens Hospital tion by Bcr/Abl. However, only up to 50% (see Figure Los Angeles Research Institute Fellowship Award (JWV).

References

Chuang T-H, Xu X, Kaartinen V, Heisterkamp N and Li W, Draezen O, Kloetzer W, Gale RP and Arlinghaus RB. Bokoch GM. (1995). Proc. Acad. Natl. Sci. USA, 92, (1989). Oncogene, 4, 127 ± 138. 10282 ± 10286. Liu J, Campbell M, Guo JQ, Lu D, Xian M and Arlinghaus de Klein A, van Kessel AG, Grosveld G, Bartram CR, RB. (1993). Oncogene, 8, 101 ± 109. Hagemeijer A, Bootsma JR, Spurr NK, Heisterkamp N, Liu J, Wu Y, Ma GZ, Lu D, Haataja L, Heisterkamp N, Gro€en J and Stephenson JR. (1982). Nature, 300, 765 ± Gro€en J and Arlinghaus R. (1996a). Mol. Cell. Biol., 16, 767. 998 ± 1005. Diekmann D, Brill S, Garrett MD, Totty N, Hsuan J, Liu J, Wu Y and Arlinghaus RB. (1996b). Cancer Res., 56, Monfries C, Hall C, Lim L and Hall A. (1991). Nature, 5120 ± 5124. 351, 400 ± 403. Lu D, Liu J, Campbell M, Guo JQ, Heisterkamp N, Gro€en Eva A and Aaronson SA. (1985). Nature, 316, 273 ± 275. J, Canaani E and Arlinghaus R. (1993). Blood, 82, 1257 ± Haataja L, Gro€en J and Heisterkamp N. (1997). J. Biol. 1263. Chem., 272, 20384 ± 20388. Maru Y and Witte ON. (1991). Cell, 67, 459 ± 468. Hart MJ, Maru Y, Leonard D, Witte ON, Evans T and McWhirter JR and Wang JYJ. (1993). The EMBO J., 12, Cerione RA. (1992). Science, 258, 812 ± 815. 1533 ± 1546. Hart MJ, Eva A, Evans T, Aaronson SA and Cerione RA. McWhirter JR and Wang JYJ. (1997). Oncogene, 15, 1625 ± (1991). Nature, 354, 311 ± 314. 1634. Heisterkamp N, Jenster G, ten Hoeve J, Zovich D, McWhirter JR, Galasso DL and Wang JYJ. (1993). Mol. Pattengale PK and Gro€en J. (1990). Nature, 344, 251 ± Cell. Biol., 13, 7587 ± 7595. 253. Muller AJ, Young JC, Pendergast AM, Pondel M, Landau Khosravi-Far R, Solski PA, Clark GJ, Kinch MS and Der NR, Littman DR and Witte ON. (1991). Mol. Cell. Biol., CJ. (1995). Mol. Cell. Biol., 15, 6443 ± 6453. 11, 1785 ± 1792. Li W, Kloetzer WS and Arlinghaus RB. (1988). Oncogene, 2, Olson MF, Ashworth A and Hall A. (1995). Science, 269, 559 ± 566. 1270 ± 1272. Leukemia in bcr null mutants JW Voncken et al 2032 Pendergast AM, Quilliam LA, Cripe LD, Bassing CH, Dai Z, Voncken JW, van Schaick H, Kaartinen V, Deemer K, Li N, Batzer A, Rabun KM, Der CJ, Schlessinger J and Coates T, Landing B, Pattengale P, Dorseuil O, Bokoch Gishizky ML. (1993). Cell, 75, 175 ± 184. GM, Gro€en J and Heisterkamp N. (1995a). Cell, 80, Puil L, Liu J, Gish G, Mbalamu G, Arlinghaus R, Pelicci PG 719 ± 728. and Pawson T. (1994). EMBO J., 13, 764 ± 773. Voncken JW, Kaartinen V, Pattengale PK, Germeraad Qiu RG, Chen J, Kirn D, McCormick F and Symons M. WTV, Gro€en J and Heisterkamp N. (1995b). Blood, 12, (1995). Nature, 374, 457 ± 459. 4603 ± 4611. Stam K, Heisterkamp N, Reynolds FH. Jr and Gro€en J. (1987). Mol. Cell. Biol., 7, 1955 ± 1960. Voncken JW, Morris C, Pattengale P, Dennert G, Kikly C, Gro€en J and Heisterkamp N. (1992). Blood, 79, 1029 ± 1036.