View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Immunity, Vol. 6, 773–782, June, 1997, Copyright 1997 by Cell Press Regulation of the Oncogenic Activity of BCR-ABL by a Tightly Bound Substrate RIN1

Daniel E. H. Afar,* Limin Han,† Jami McLaughlin,* BCR-ABL is a chimeric generated by the ‡ † Stephane Wong, Ajay Dhaka, Kalindi Parmar,k reciprocal chromosomal translocation t(9;22) (q34;q11) ‡§ Naomi Rosenberg,k Owen N. Witte,* (reviewed by Kurzrock et al., 1988). BCR sequences are and John Colicelli†‡ fused upstream of c-, resulting in the activation of *Department of Microbiology and Molecular Abl tyrosine kinase activity (Konopka et al., 1984). Two Genetics BCR-ABL isoforms, P185 and P210, are associated with † Department of Biological Chemistry the pathology of human acute lymphocytic and chronic ‡ Molecular Biology Institute myelogenous leukemias (Kurzrock et al., 1988). § Howard Hughes Medical Institute BCR-ABL is a complex molecule with multiple func- University of California, Los Angeles tional domains, including a BCR oligomerization and Los Angeles, California 90095 SRC homology 2 (SH2)–interacting region, Abl-derived SH3 and SH2 domains, a tyrosine kinase domain, and k Departments of Molecular Biology and Microbiology and Pathology an actin-binding region at the carboxyl terminus (re- Tufts University School of Medicine viewed by Laneuville, 1995). The SH3 domain, which Boston, Massachusetts 02115 binds proline-rich sequences, and the SH2 domain, which binds phosphotyrosine-containing peptide se- quences, provide potential docking sites for BCR-ABL targets. The BCR sequences activate Abl kinase activity, Summary which is the most critical determinant of transforming activity (Lugo et al., 1990; McWhirter and Wang, 1991; RIN1 was originally identified by its ability to physically Muller et al., 1991). Activation of Ras is an essential bind to and interfere with activated Ras in yeast. Para- event for transformation by BCR-ABL (Sawyers et al., doxically, RIN1 potentiates the oncogenic activity of 1995). This is achieved by the binding and activation of the BCR-ABL tyrosine kinase in hematopoietic cells Shc, Crkl, Grb-2, and Grap, which are adapter and dramatically accelerates BCR-ABL–induced leu- involved in redundant pathways for Ras activation kemias in mice. RIN1 rescues BCR-ABL mutants for (Pendergast et al., 1993b; Puil et al., 1994; Matsuguchi transformation in a manner distinguishable from the et al., 1994; Oda et al., 1994; Tauchi et al., 1994; ten cell cycle regulators c-Myc and cyclin D1 and the Ras Hoeve et al., 1994; Goga et al., 1995; Feng et al., 1996). connector Shc. These biological effects require tyro- Mutations in the Grb-2–binding site (Y177F) (Pender- sine phosphorylation of RIN1 and binding of RIN1 to gast et al., 1993b; Puil et al., 1994), in the phosphotyro- the Abl-SH2 and SH3 domains. RIN1 is tyrosine phos- sine-binding site of the SH2 domain (R552L) (Afar et al., phorylated and is associated with BCR-ABL in human 1994), and in the major kinase autophosphorylation site and murine leukemic cells. RIN1 exemplifies a new (Y793F) (Pendergast et al., 1993a) have been shown to class of effector molecules dependent on the con- reduce BCR-ABL–mediated transformation. The trans- certed action of the SH3, SH2, and catalytic domains forming activity of the SH2 mutant is specifically re- of a cytoplasmic tyrosine kinase. stored by increasing the dosage of c-myc (Afar et al., 1994) or cyclin D1 (Afar et al., 1995), while transforma- Introduction tion by the Grb-2–binding site mutant is rescued by increasing Shc dosage (Goga et al., 1995). These differ- ential rescue experiments indicate that multiple signals The biological activity of tyrosine kinases is dependent on their ability to activate downstream signaling path- must be activated for cellular transformation, which in- clude signals to cell cycle regulators and Ras. ways. This function is often mediated by the physical Ras is a member of a family of small GTP-binding binding of tyrosine kinases with their targets via con- proteins that are central components of many tyrosine served protein domains (Pawson, 1995). Multiple tyro- kinase signaling pathways. They function by activating sine kinase substrates and binding proteins have been downstream effector molecules, which include serine/ identified (reviewed by Kazlauskas, 1994; Brown and threonine kinases (Van Aelst et al., 1993; Vojtek et al., Cooper, 1996), but not all prove to be biologically rele- 1993; Manser et al., 1994; Chou and Blenis, 1996). vant. However, generating mutations in the tyrosine ki- Recently, a novel Ras-binding protein was identified nases or their targets that prevent physical association from a human cDNA library in a Ras interference screen is an effective means to determine the importance of in yeast (Han and Colicelli, 1995). The RIN1 protein their interaction (Valius and Kazlauskas, 1993). We and blocked RAS2V19-induced cell death by interacting di- others have developed a genetic strategy to analyze rectly with the effector domain of GTP-bound RAS. Se- signaling by activated tyrosine kinases (Roussel et al., quencing of RIN1 revealed the presence of an amino- 1991; Afar et al., 1994, 1995; Goga et al., 1995; Roussel et terminal SH2 domain and a proline-rich region (Han et al., 1995). Partial loss-of-function mutants of a tyrosine al., 1997). The proline-rich region in RIN1 is homologous kinase are tested for rescue of the wild-type phenotype to the Abl-SH3–binding consensus (Ren et al., 1993; by increasing the gene dosage of a regulator or effector Rickles et al., 1994; Alexandropoulos et al., 1995). This molecule. Increasing the of effectors raised the possibility that RIN1 can interact with different compensates for the loss of a specific component or signaling molecules, including Ras and tyrosine kinases reduced strength of the tyrosine kinase signal. such as Abl. Immunity 774

Expression of BCR-ABL in Rat-1 cell fibroblasts per- mits anchorage-independent growth in soft agar (Lugo and Witte, 1989) that is dependent on a critical level of Ras activity (Mandanas et al., 1993; Pendergast et al., 1993b, Sawyers et al., 1995). The Ras interference activ- ity of RIN1 predicted that it would function as an indirect inhibitor of BCR-ABL function. Paradoxically, we found that BCR-ABL transformation activity is augmented by increased gene dosage of RIN1. Overexpression of RIN1 increased BCR-ABL –mediated transformation of hema- topoietic cells and increased the leukemogenic activity of BCR-ABL in mice. RIN1 is found complexed with BCR-ABL in human leukemia–derived cell lines. Deletion of the Ras-binding region of RIN1 increased its syner- gistic activity with BCR-ABL. This transformation syn- ergy required tyrosine phosphorylation of RIN1 and its physical binding to BCR-ABL, supporting a role as a novel binding and effector molecule for BCR-ABL.

Results

RIN1 Accelerates Lymphoid Cell Transformation by BCR-ABL RIN1 interferes with activated RAS alleles in yeast. Since BCR-ABL is dependent on Ras for transforming activity, we expected that expression of RIN1 would lead to a block in BCR-ABL–mediated transformation. To test the ability of RIN1 to block BCR-ABL, we constructed ret- roviral vectors (Afar et al., 1995; Sawyers et al., 1995) that allow the simultaneous expression of RIN1 and P185 BCR-ABL. A RIN1 mutant, RIN1N (amino acids 1–295), with a deletion of the carboxyl-terminal Ras– interacting region, was used as a control. Retrovirally Figure 1. Stimulation of P185 Bone Marrow Transformation by RIN1 mediated expression of BCR-ABL into primary mouse and RIN1N bone marrow results in transformation of pre–B cells in Retroviruses containing wild-type P185 linked in cis to either neo, stromal cell dependent cultures (McLaughlin et al., RIN1, or RIN1N were generated. Expression of neo, RIN1, and RIN1N 1987). The growth rate of pre–B cells from infected mar- was driven by the retroviral LTR, while expression of P185 was controlled by the tk promoter. Primary mouse bone marrow was row is directly dependent on the strength of the tyrosine infected with the retroviruses and plated in liquid culture. Four repli- kinase activity (McLaughlin et al., 1989). To determine cates of each sample were monitored for pre–B cell growth. Cell whether RIN1 affects hematopoietic cell transformation counts were performed at different time intervals after infection. by P185 BCR-ABL, mouse bone marrow was infected Cultures with more than 2 ϫ 107 cells are confluent. with retroviruses encoding different gene combinations. Liquid cultures were plated in quadruplicate and were either neo, RIN1, or RIN1N. Two days after infection, the monitored for pre–B cell growth. Instead of interfering cells were removed from growth factor containing media with P185-mediated transformation, expression of full- and were plated onto 96-well plates. In the absence of length RIN1 resulted in an unexpected acceleration of growth factor or P210, DAGM cells die. In the presence P185-mediated transformation (Figure 1). Deletion of the of P210, 6 of 96 wells of DAGM cells become GM-CSF Ras-interacting region of RIN1 increased the potency of independent (Table 1 and Goga et al., 1995). Coexpres- this effect. Cells coexpressing RIN1N with P185 reached sion of P210 with RIN1 increased the efficiency of growth saturation densities (1–2 ϫ 107 cells/culture) 2 days prior factor independence 4- to 5-fold, permitting growth in to cells coexpressing P185 with RIN1. 30 of 96 wells. RIN1N increased P210-mediated growth factor independence more than 10-fold, resulting in The Efficiency of Growth Factor Independence growth in about 90 of 96 wells (Table 1). of Myeloid Cells Induced by BCR-ABL Is Increased by RIN1 RIN1 Accelerates BCR-ABL–Induced Leukemias P210, the larger BCR-ABL isoform, is predominantly ex- in Mice pressed in leukemias of the myeloid lineage (Kurzrock Several murine models have been developed to study et al., 1988). To test the effect of coexpression of RIN1 BCR-ABL–induced leukemias. These include generation and RIN1N with P210, we used the granulocyte/macro- of transgenic mice expressing BCR-ABL (Heisterkamp phage colony-stimulating factor (GM-CSF)–dependent et al., 1990) or reconstitution of lethally irradiated mice myeloid cell line DAGM (Rennick et al., 1987). DAGM with BCR-ABL–expressing bone marrow (Daley et al., cells were infected with retroviruses encoding P210 with 1990; Elefanty et al., 1990; Kelliher et al., 1990, 1991; Regulation of BCR-ABL by RIN1 775

Table 1. Induction of Growth Factor Independence of DAGM specific point mutations in functional domains of BCR- Cells by BCR-ABL and RIN1 ABL. Point mutants in the Grb-2–binding site of the BCR Retrovirus Experiment 1 Experiment 2 region (Y177F), the phosphotyrosine-binding site of the SH2 domain (R552L), and the major kinase autophos- Mock 0/96 0/96 RIN1 tk neo 0/96 0/96 phorylation site (Y793F) are defective in transforming neo tk P210 6/96 7/96 Rat-1 cells despite retaining tyrosine kinase activity RIN1 tk P210 31/96 27/96 (Pendergast et al., 1993a, 1993b; Afar et al., 1994). The RIN1N tk P210 86/96 93/96 SH2 mutant can be rescued for transforming activity by The ability of DAGM cells to be rendered growth factor independent increasing the gene dosage of c-myc (Afar et al., 1994) following expression of P210 BCR-ABL with neo, RIN1, or RIN1N or cyclin D1 (Afar et al., 1995). Increasing the expression was determined in two independent experiments. Wells that demon- of the adaptor protein Shc can rescue the Grb-2–binding strated the outgrowth of factor–independent cells after 14 days in mutant for transformation (Goga et al., 1995). culture were scored positive. To determine whether RIN1 exhibits rescue activity similar to or distinct from that of c-Myc or cyclin D1, different RIN1 forms were coexpressed with the alterna- Scott et al., 1991; Gishizky et al., 1993). To investigate tive P185 point mutants in Rat-1 fibroblasts. Infected whether RIN1 or RIN1N accelerates BCR-ABL–induced Rat-1 cells were harvested for protein analysis and plat- leukemias, sublethally irradiated CB.17scid/scid (severe ing in soft agar. Western analysis showed comparable combined immunodeficient) mice were reconstituted expression levels of P185 in each sample and similar with BALB/c-derived whole bone marrow infected with expression levels of RIN1 forms. Elevated expression titer-matched viruses of wild-type P185 plus neo, RIN1, of full-length RIN1 increased Rat-1 cell transformation or RIN1N. A fraction of infected marrow was plated in by P185 Y177F from an average of 30 colonies to an liquid culture to confirm the transfer and expression of average of 80 colonies (Table 2). Expression of P185 BCR-ABL by monitoring pre–B cell transformation (data Y793F in Rat-1 cells causes the growth of fewer than not shown). In all cases, pre–B cell transformation was 10 colonies in agar. In the presence of RIN1, the number observed with kinetics, as described in Figure 1. of colonies increased by greater than a log,to an average In four independent experiments, 25% of mice receiv- of 100 colonies (Table 2). No increase in soft agar colo- ing P185 with neo developed leukemia within the 90- nies was detected when RIN1 was coexpressed with day period of observation (Figure 2). In contrast, more the SH2 mutant of P185. Since expression of wild-type than 60% of mice receiving bone marrow coexpressing P185 in Rat-1 cells results in the growth of 500 colonies RIN1 with P185 and more than 80% of mice expressing in agar, overexpression of RIN1 did not further increase RIN1N with P185 succumbed to leukemias within 30–80 colony growth (Table 2). days after reconstitution (Figure 2). All leukemic mice Expression of RIN1N caused a striking rescue of the displayed splenomegaly, and in some cases lymph kinase autophosphorylation mutant (Table 2). Colonies nodes were enlarged (to 4–5 times normal size). Fluores- growing in agar as a result of RIN1N coexpression with cence activated cell sorting analysis using antibodies P185 Y793F were as large (0.5 –3 mm in diameter) and directed to the B cell marker B220 showed that the as numerous as colonies derived from cells expressing mice exhibited lymphocytic leukemias. Western blotting an equal level of wild-type P185 (50-fold more colonies analysis of cell lysates from spleen, bone marrow, and than cells expressing P185 Y793F alone). This demon- lymph nodes showed expression of P185, RIN1, and strates that RIN1N expression restored wild-type–like RIN1N (data not shown). These data demonstrate that transformation activity to the autophosphorylation mu- RIN1 expression effectively potentiates the develop- tant. RIN1N was also effective in synergizing with P185 ment of BCR-ABL –mediated leukemia. Y177F, increasing its activity 10-fold (Table 2). The colo- nies were large (0.5–2 mm in diameter) and the media Differential Rescue of P185 Point Mutants with RIN1 was significantly acidified after 3 weeks in culture. Cells BCR-ABL is known to activate multiple signals for trans- coexpressing P185 R552L and RIN1N exhibited a small formation. Some of these signals can be abrogated by (3–5 fold) increase in colony formation but showed no

Figure 2. Kaplan-Meyer Plot of Leukemic Mice Expressing BCR-ABL with RIN1 Survival rates of lethally irradiated SCID mice reconstituted with bone marrow infected with retroviruses encoding P185 with neo, RIN1, or RIN1N. Leukemic diagnosis was based on gross and microscopic assessment. Cells wereclassified as lymphoid based on expres- sion of the B cell marker B220. Mice reconsti- tuted with bone marrow expressing RIN1 or RIN1N alone did not develop any signs of leukemia. Immunity 776

Table 2. Rat-1 Transformation by BCR-ABL Mutants with RIN1 To test whether autophosphorylation activity is signifi- Numberof Soft Agar Colonies/ cantly altered, BCR-ABL was immunoprecipitated with 104 Cells anti-Abl antibodies from lysates of cells coexpressing Rescuing Gene the different P185 forms with RIN1 or RIN1N. Autophos- P185 Form neo RIN1 RIN1N phorylation activity of P185 proteins was assayed in vitro on the immunoprecipitates and analyzed by SDS– neo Ͻ10 Ͻ10 Ͻ10 polyacrylamide gel electrophoresis (SDS-PAGE) with P185 500 500 Ͼ1000 (wild type) subsequent autoradiography. The results showed that P185 autophosphorylation was not significantly different P185 Y177F 30 80 300 in the presence or absence of the RIN1 forms (Figure (Grb-2 site mutant) 3). However, full-length RIN1 and RIN1N coimmunopre- P185 R552L Ͻ10 10 50 cipitated with P185 and were tyrosine phosphorylated (SH2 mutant) during the in vitro kinase reaction (Figure 3). This activity P185 Y793F Ͻ10 100 600 depended on the SH2 domain of BCR-ABL, since muta- (Auto-phos. mutant) tion of the SH2 domain significantly decreased the asso- Rat-1 cells were infected with retroviruses encoding different BCR- ciation and phosphorylation of RIN1 molecules. The ABL forms with either neo, RIN1, or RIN1N. Cells were plated in soft kinase autophosphorylation and Grb-2–binding site mu- 4 agar at 5 ϫ 10 cells/plate in media containing 10% FBS and were tants coimmunoprecipitated and phosphorylated the incubated for 3 weeks. The number of colonies represent the aver- RIN1 forms with efficiency equal to that of wild-type age colony counts of two to five separate experiments rounded off to two significant figures. P185 (data not shown). To determine whether RIN1 and RIN1N are tyrosine phosphorylated in vivo, different P185 forms were coex- media acidification (Table 2). The colonies were small, pressed with neo, RIN1, or RIN1N in Rat-1 fibroblasts. with average diameters of less than 0.5 mm. Western blotting analysis of cell lysates using anti-phos- The P185 point mutants retain their ability to transform photyrosine antibodies showed that both RIN1 and primary mouse bone marrow cells (Goga et al., 1995). RIN1N were the major tyrosine-phosphorylated species However, the latency of pre–B cell outgrowth and at- in BCR-ABL–expressing cells (Figure 4). Since RIN1 mol- tainment of saturation growth density (1–2 ϫ 107 cells/ ecules failed to coimmunoprecipitate with the P185 SH2 culture) are delayed in cultures expressing the P185 mutant, we expected to detect no tyrosine phosphoryla- mutants. To test whether their hematopoietic transfor- tion of RIN1 molecules in cells expressing the SH2 mu- mation activity can be augmented with RIN1 or RIN1N, tant of P185. However, RIN1 molecules were also signifi- the P185 mutants were coexpressed with the different cantly tyrosine phosphorylated in these cells (Figure 4). RIN1 in primary mouse bone marrow. Previously, This result suggests that the SH2 domain of BCR-ABL we have shown that cyclin D1 stimulates the ability of is not essential for recognition and phosphorylation of P185 R552L to transform bone marrow (Afar et al., 1995). RIN1, but it is important for formation of stable com- Coexpression of RIN1 or RIN1N accelerated pre–B cell plexes with RIN1. outgrowth with either the kinase autophosphorylation mutant or the Grb-2 –binding site mutant (data not shown). Much weaker synergy was observed when RIN1 was coexpressed with the P185 SH2 mutant. These re- sults suggest that RIN1 functions to replace or compen- sate for signals dependent on the autophosphorylation site and Grb-2 binding.

The Mechanism of RIN1 Action Involves Tyrosine Phosphorylation of RIN1 and Physical Interaction with BCR-ABL The major kinase autophosphorylation site in BCR-ABL (Y793) is homologous to Y416 in c-Src, which serves to activate the intrinsic kinase activity (Piwnica-Worms et al., 1987; Kmiecik and Shalloway, 1987). In BCR-ABL, mutation of the autophosphorylation site results in a 30%–50% decrease in in vitro kinase activity (Pender- gast et al., 1993a; Afar et al., 1994). Although the phos- Figure 3. Association of RIN1 and RIN1N with BCR-ABL photyrosine content of cells expressing the auto- Immune complex kinase assays were performed as described in phosphorylation mutant is not significantly different Experimental Procedures. 293T cell lysates were immunoprecipi- from that of cells expressing wild-type BCR-ABL, their tated with anti-Abl antibodies and were assayed for in vitro phos- transforming activity is profoundly decreased (Pender- phorylation by the addition of 32P-␥ ATP. Molecular size markers are gast et al., 1993a; Afar et al., 1994). The rescue of P185 indicated in kilodaltons. The positions of P185, RIN1, and RIN1N Y793F biological activity with RIN1 could be due to a are indicated. Samples are wild-type P185 with neo (lane 1), with RIN1 (lane 2), and with RIN1N (lane 3); P185 R552L (SH2 mutant) recovery of this signal by either increasing BCR-ABL with neo (lane 4), with RIN1 (lane 5), and with RIN1N (lane 6); and kinase activity or by substituting for the loss of a poten- P185 Y793F (kinase autophosphorylation mutant) with neo (lane 7), tial effector that interacts with phosphorylated Y793. with RIN1 (lane 8), and with RIN1N lane (9). Regulation of BCR-ABL by RIN1 777

Figure 4. Phosphotyrosine Analysis of RIN1 and RIN1N Rat-1 cells were infected with retroviruses that coexpress neo, RIN1, or RIN1N with the different P185 forms. The cells were lysed directly in SDS-containing sample buffer (at 100ЊC) 72 hr postinfection. The lysates wereresolved on SDS-PAGE (10% polyacrylamide) andwere transferred to nitrocellulose. The blots were probed with antibodies directed to BCR (␣-BCR), RIN1 (␣-RIN1), and phosphotyrosine (␣-P-TYR). Molecular size markers are indicated in kilodaltons. The Figure 5. Deletion of the Abl-SH3 Domain Prevents BCR-ABL/RIN1 positions of P185, RIN1, and RIN1N are indicated. The samples Association and Tyrosine Phosphorylation of RIN1N include neo with wild-type P185 (lane 1), P185 Y177F (lane 2), P185 (A) Neo or RIN1 were coexpressed with P185 or P185⌬SH3 in 293T R552L (lane 3), and P185 Y793F (lane 4); RIN1 with wild-type P185 cells and were analyzed by immune complex kinaseassays. Molecu- (lane 5), P185 Y177F (lane 6), P185 R552L (lane 7), and P185 Y793F lar size markers are indicated in kilodaltons (KD). The positions of (lane 8); and RIN1N with wild-type P185 (lane 9), P185 Y177F (lane P185 and RIN1 are indicated. The samples include wild-type P185 10), P185 R552L (lane 11), and P185 Y793F (lane 12). with neo (lane 1) or RIN1 (lane 2) and P185⌬SH3 with neo (lane 3) or RIN1 (lane 4). Structural Requirements for Tyrosine (B) Cell lysates were prepared from Rat-1 cells coexpressing RIN1N with different P185 forms and were analyzed using anti-Abl (␣-ABL), Phosphorylation of RIN1 anti-RIN1 (␣-RIN1), and anti-phosphotyrosine (␣-P-TYR) antibodies. The amino terminus of RIN1 includes a proline-rich do- The samples include mock-infected cells (lane 1) and cells coex- main with high homology to the Abl-SH3 domain con- pressing RIN1N with wild-type P185 (lane 2), P185 Y177F (lane 3), sensus binding sequence (Ren et al., 1993; Rickles et P185 R552L (lane 4), P185 Y793F (lane 5), and P185⌬SH3 (lane 6). al., 1994; Alexandropoulos et al., 1995). Recently, we have shown that a peptide derived from the amino- terminal region of RIN1, which includes the sequence PPAVPPPPVP binds to GST-Abl-SH3 fusion proteins in and RIN1N to increase the transformation activity of vitro (Han et al., 1997). To determine whether the Abl- an alternative abl oncogene, v-abl. The v-Abl protein SH3/RIN1 amino-terminal interaction contributes to the exhibits an amino-terminal fusion to viral gag sequences ability of BCR-ABL to interact with RIN1, the Abl-SH3 and a deletion in the SH3 domain (reviewed by Rosen- domain was deleted in P185. RIN1 was coexpressed berg and Witte, 1988). Coexpression of RIN1 or RIN1N with P185⌬SH3 in 293T cells to test for coimmunopre- with either v-Abl or a transformation defective kinase cipitation and in vitro phosphorylation. The results show autophosphorylation mutant of v-Abl did not result in a that deletion of the SH3 domain blocked physical asso- significant increase in transformation activity in either ciation of RIN1 with P185, as determined by an immune NIH 3T3 cells or primary mouse bone marrow cells (data complex in vitro kinase assay (Figure 5A). In vivo tyrosine not shown). In addition, no complementation was ob- phosphorylation of RIN1N was greatly reduced when served with carboxyl terminal v-Abl truncation mutants, RIN1N was coexpressed in Rat-1 cells with P185⌬SH3 which are severely compromised in lymphoid cell trans- (Figure 5B). In addition, no significant increase in cellular formation. Combined with the P185⌬SH3 data, these transformation was observed when RIN1N was coex- results suggest that RIN1 phosphorylation and biologi- pressed with P185⌬SH3 (data not shown). cal activity require an interaction with the Abl-SH3 To confirm this result, we tested the ability of RIN1 domain. Immunity 778

motifs include the minimum sequence requirement for SH3-domain binding (reviewed by Feller et al., 1994) and may explain the retention of binding and biological activity of RIN1N⌬.

Tyrosine Phosphorylation of the RIN1 Amino Terminus Is Required for Biological Activity The sequence of the RIN1 amino terminus reveals the presence of three potential tyrosine phosphorylation sites: Y36, Y121, and Y148 (Han et al., 1997). The se- quence following Y36 (YDVP) is identical to the se- quence following multiple tyrosine phosphorylation sites in the Src kinase substrate p130CAS (Sakai et al., 1994) and its relative Sin (Alexandropoulos and Balti- more, 1996). To determine which of these potential tyro- sine phosphorylation sites are required for RIN1N activ- ity, all sites were mutated to phenylalanine either singly or in combination as a triple mutant. Coexpression of the different RIN1N phosphorylation mutants with P185 Y793F in 293T cells showed that single point mutants coimmunoprecipitated with P185 Y793F and served as in vitro substrates (Figure6A). Totest for biological activ- ity, each mutant was stably coexpressed with P185 Y793F in Rat-1 cells. Analysis of soft agar growth showed that the single point mutants rescued P185 Y793F transforming activity (Figure 6B). The triple phos- phorylation mutant failed to rescue transforming activity despite an expression level equal to that of the other mutants (Figure 6A). Therefore, the three potential tyro- sine phosphorylation sites in RIN1N are redundant for biological activity.

RIN1 Is Tyrosine Phosphorylated and Associated with BCR-ABL in Human Leukemic Cells BCR-ABL expression naturally occurs in hematopoietic Figure 6. Rescue of P185 Y793F with RIN1N Mutants cells and is associated with the pathogenesis of human RIN1N mutants were generated with mutations in the potential tyro- sine phosphorylation sites or the proline-rich region and were tested CML and ALL. To determine whether RIN1 is tyrosine for biochemical and biological interaction with the autophosphoryla- phosphorylated and a potential regulator of BCR-ABL tion mutant of P185. in human leukemia, we analyzed the association of RIN1 (A) Cell lysates from cells coexpressing P185 Y793F with different with BCR-ABL in several Philadelphia RIN1N mutants were analyzed by immune complex kinase assays (Ph1)–positive leukemic cell lines. RIN1 expression was (as described for Figure 4) and Western blotting using anti-RIN1 detected in the three Ph1-positive leukemic cell lines antibodies (␣RIN1). Molecular size markers are indicated in kilodal- tons. The samples include P185 K619R (kinase inactive BCR-ABL) tested (K562, ALL1, and KCL22). The highest levels of with RIN1N (lane 1); P185 Y793F with RIN1N (lane 2), RIN1N Y36F RIN1 expression were detected in KCL22 cells (data not (lane 3), RIN1N Y121F (lane 4), RIN1N Y148F (lane 5), RIN1N Y/F TM shown). (triple tyrosine mutant) (lane 6), and RIN1N⌬ (lane 7). Endogenous RIN1 and P210 BCR-ABL were immuno- (B) Rat-1 cells expressing the different RIN1N mutants were plated precipitated with anti-RIN1N or anti-Abl antibodies from 4 in soft agar at 5 ϫ 10 cells/plate in media containing 10% FBS. KCL22 cell lysates. The immune complexes were sub- The number of colonies represent the average colony counts of duplicate plates after 3 weeks in culture. jected to in vitro kinase reactions and analyzed by SDS- PAGE. The results show that P210 coimmunoprecipi- tates with and phosphorylates a 90 kDa protein (lane 1). Anti-RIN1N immunoprecipitates also contain two phos- When the presumed proline-rich SH3-binding region phoproteins with molecular weights of 210 and 90 kDa in RIN1N (residues 244–295) was deleted (RIN1N⌬), (Figure 7A, lane 2). The identity of the 210 and 90 kDa coexpression with P185 Y793F still resulted in coimmu- proteins as P210 BCR-ABL and RIN1 were confirmed noprecipitation and phosphorylation of RIN1N⌬ in vitro, by reimmunoprecipitating the samples with either anti- although at a reduced efficiency ( lane 7), and tyrosine Abl or anti-RIN1N antibodies (Figure 7A, lanes 3–6). phosphorylation of RIN1N in vivo (data not shown). Bio- To determine whether RIN1 is tyrosine phosphory- logical rescue of P185 Y793F was also observed (Figure lated in KCL22 cells in vivo, RIN1 and P210 were immu- 6B). This result suggests that RIN1N⌬ may interact with noprecipitated from cell extracts and subjected to phos- the Abl-SH3 domain via an alternate region. The amino photyrosine analysis by Western blotting. Anti-RIN1N terminus of RIN1 contains four additional PXXP motifs and anti-Abl immunoprecipitates show a remarkably at residues 12–15, 49–52, 201–204, and 227–230. These similar phosphotyrosine pattern of at least three distinct Regulation of BCR-ABL by RIN1 779

Discussion

RIN1 Accelerates BCR-ABL–Mediated Leukemias by Activating a Novel Signaling Pathway Here we describe the identification of RIN1 as a novel substrate effector molecule for BCR-ABL. The dramatic biological activity in in vitro hematopoietic transforma- tion assays and especially in the mouse leukemogenesis assay demonstrates the functional importance of RIN1. The significant expression levels in human leukemic cell lines suggest that RIN1 contributes to the pathogenesis of human BCR-ABL–induced leukemias. RIN1 expression rescued cellular transformation ac- tivity of both the autophosphorylation site and Grb-2 site mutants of BCR-ABL. The physical interaction of RIN1 with BCR-ABL may replace an interaction with a proximal effector that is lost by mutation of the auto- phosphorylation site. RIN1 may function as an adaptor protein that recruits molecules to form a signaling com- plex, a role proposed for the tyrosine kinase substrates Sin (Alexandropoulos and Baltimore, 1996), p130Cas (Sa- kai et al., 1994), and p62dok (Carpino et al., 1997; Yama- nashi and Baltimore, 1997). RIN1 itself contains a func- tional SH2 domain in the amino terminus and exhibits three potential tyrosine phosphorylation sites that could function as protein docking sites (Han et al., 1997). RIN1 may recover the Grb-2 signaling pathway by physically recruiting Grb-2 to the BCR-ABL complex. However, coexpression of RIN1 or RIN1N with either wild-type P185 or P185 Y177F did not significantly in- crease Grb-2 association as detected by coimmuno- precipitation (D. E. H. A. and O. N. W., unpublished data). This is in contrast to Shc, which rescues P185 Y177F by recruiting Grb-2 (Goga et al., 1995), and indicates Figure 7. Tyrosine Phosphorylation of RIN1 and Association with that RIN1 functions in a distinct pathway that can substi- P210 BCR-ABL in Human Leukemic Cells tute for Grb-2–mediated signaling. Lysates prepared from KCL22 cells were analyzed by (A) immune RIN1, through its carboxyl terminus, interacts specifi- complex kinase assays and (B) immune complex Western blot cally with GTP-bound Ras and may serve as a Ras ef- analysis. fector (Han and Colicelli, 1995). Thus, RIN1 may link (A) For immune complex kinase analysis, samples were immunopre- BCR-ABL directly to signals downstream of Ras. How- cipitated using antibodies directed toward Abl (lane 1) and RIN1N (lane 2). Anti-Abl immune complexes were reimmunoprecipitated ever, RIN1N does not interact with Ras and is more using antibodies directed toward Abl (lane 3) and RIN1N (lane 4). effective in rescuing BCR-ABL mutants. This suggests Anti-RIN1N immune complexes were reimmunoprecipitated using a function for RIN1 that may be unrelated to direct Ras antibodies directed toward Abl (lane 5) and RIN1N (lane 6). I.P., binding. RIN1 may have a dual purpose in acting as an immunoprecipitation. effector for both Ras and activated tyrosine kinases. In (B) Anti-Abl (lanes 1 and 3) and anti-RIN1N (lanes 2 and 4) immune vivo association studies have shown that RIN1 mole- complexes were analyzed by Western blotting using antibodies di- cules associated with BCR-ABL do not coimmunopre- rected toward phosphotyrosine (␣-P-TYR), Abl (␣-ABL), and RIN1 (␣-RIN1). The relative positions of P210 BCR-ABL and RIN1 are cipitate Ras, and, similarly, Ras associated RIN1 does indicated. not bind BCR-ABL (L. H. and J. C., unpublished data). These alternative functions possibly compete with each other, helping to explain why full-length RIN1 is less proteins with molecular weights of 210, 120, and 90 kDa effective in synergizing with BCR-ABL than RIN1N. (Figure 7B). The identity of the 120 kDa protein is not known and was not further analyzed. Anti-Abl Western RIN1 Functions by Physical Interaction analysis of the samples confirms that the 210 and 90 with BCR-ABL kDa proteins correspond to P210 BCR-ABL and RIN1, Our results show that RIN1 interacts with BCR-ABL by respectively (Figure 7B). at least three different mechanisms that involve the Abl- Tyrosine phosphorylation of RIN1 was also detected SH3, SH2, and kinase domains. One possible interpreta- in K562 and ALL1 cells (data not shown). Since 10 times tion of our results is that the interaction of RIN1 with more cell lysate was used to immunoprecipitate RIN1 BCR-ABL is initiated by the Abl-SH3 domain, followed and since the amount of P210 is approximately equiva- by RIN1 tyrosine phosphorylation by the Abl-kinase do- lent in the anti-Abl and the anti-RIN1 immune com- main and subsequent association of RIN1 with the Abl- plexes, we estimate that about 10% of BCR-ABL mole- SH2 domain. Abl-SH2 binding is critical for RIN1 func- cules are stably associated with RIN1. tion, since RIN1 phosphorylation alone is insufficient Immunity 780

for biological activity. Alternatively, all three interactions Generation of Virus Stocks may cooperate with each other to promote RIN1–BCR- Helper-free retroviruses were generated by transient cotransfection ABL association. of 293T cells (Pear et al., 1993) (provided by D. Baltimore, Massachu- setts Institute of Technology) with retroviral vectors and a PsiϪ eco- Deletion of the Abl-SH3 domain potentiates the trans- tropic packaging vector (Muller et al., 1991). Supernatant from forming activity of abl (Franz et al., 1989; transfected 293T cells was collected 36–50 hr posttransfection. Viral Jackson and Baltimore, 1989; Muller et al., 1991). This titers were determined indirectly by measuring protein expression demonstrates that the Abl-SH3 domain is not essential and/or retroviral integration into the genome 48–72 hr after infection for transformation and suggests that RIN1–Abl interac- of Rat-1 cells by Western blotting (Muller et al., 1991) or Southern tions may also not be critical for in vitro transformation analysis, respectively. Viral stocks were normalized to give equiva- lent protein expression for the various samples. of cells. However, coexpression of RIN1N with BCR- ABL is as effective as deletion of the SH3 domain of Soft Agar Transformation Assays and Growth Factor BCR-ABL in increasing transforming activity (D. E. H. Independence Assays A. and O. N. W., unpublished data). In addition, SH3 Soft agar transformation assays were performed as described pre- deletions in BCR-ABL are rarely detected in human dis- viously (Lugo and Witte, 1989; Muller et al., 1991). Retrovirally in- ease (Soekarman et al., 1990). This suggests that the fected cells were grown for 48–72 hr and were plated in soft agar at a density of 1 ϫ 104 cells/6 cm dish. The samples were plated in SH3 domain may provide a unique signal necessary for duplicate in medium containing 10% fetal calf serum. Colonies 0.5 leukemogenesis that may be mediated by molecules mm in diameter or larger were counted 2–3 weeks after plating. such as RIN1. Growth factor independence assays were performed as pre- RIN1 may normally interact with cellular forms of Abl viously described (Goga et al., 1995). DAGM cells were grown in and other tyrosine kinases that contain SH3 and SH2 media containing growth factor for 2 days postinfection. Cells were domains, such as the Src family kinases, which are in- then washed twice in media without growth factor and were plated in 96-well dishes at a density of 1 ϫ 104 cells/well. At 2 weeks volved in a range of cellular signaling pathways (re- postinfection, wells that demonstrated an outgrowth of factor-inde- viewed by Erpel and Courtneidge, 1995). c-Abl is ex- pendent cells were scored positive. pressed in the nucleus and cytoplasm (Van Etten et al., 1989; Wen et al., 1996), while Src-family kinases and Bone Marrow Transformation and Reconstitution RIN1 are localized to the cytoplasm and inner surface of Irradiated Mice of the plasma membrane (Erpel and Courtneidge, 1995; Bone marrow was isolated from tibias and femurs of 4–8 week old male BALB/c mice and infected with matched retroviral stocks. Cells Han and Colicelli, 1995). It may be that during certain were plated at a density of 5 ϫ 106 cells/6 cm dish and the number stimuli, when c-Abl is induced to relocate to the cyto- of nonadherent pre–B cells were counted 10–14 days postinfection plasm (Lewis et al., 1996), a normal interaction with RIN1 (McLaughlin et al., 1987). A culture was scored positive if it reached may occur. Oncogenic tyrosine kinases, such as BCR- a cell density of greater than or equal to 1 ϫ 106 cells/culture. ABL, may exploit molecules such as RIN1 to provide Saturation density was reached at greater than 1 ϫ 107 cells/culture. them with aberrant functions and to feed into multiple One day prior to reconstitution, severe combined immunodefi- signaling pathways. Tyrosine phosphorylation of RIN1 cient (SCID) mice were sublethally irradiated with 275 rad. Whole bone marrow was isolated and infected with retrovirus as described and binding to BCR-ABL are clearly detected in human above. Three hours postinfection, the bone marrow was injected 1 Ph -positive leukemic cells. These observations, to- intravenously into the tail veins of recipient SCID animals. Animals gether with the finding that RIN1 accelerates BCR-ABL– were monitored for signs of sickness over a 12-week period. Sick induced leukemias in mice, suggests that expression of mice were sacrificed and tissues were analyzed for BCR-ABL, RIN1, RIN1 in CML contributes to the pathology of the disease. and RIN1N expression using Western blotting. Blood and spleen samples were analyzed by fluorescence-activated cell sorting. Blood smears were analyzed by Wright/Giemsa staining. Experimental Procedures

Plasmid Construction and Mutagenesis Western Blotting The different P185 BCR-ABL forms (Afar et al., 1994) were cloned Western blotting was performed as described previously (Muller et into the retroviral vector pSR␣MSVtk (Muller et al., 1991) using a al., 1991). Cells were lysed in boiling SDS-containing sample buffer. unique XbaI site 3Ј to the herpes simplex thymidine kinase (tk) pro- Solubilized protein was resolved by SDS-PAGE, transferred to nitro- moter. The different RIN1 cDNAs (Han and Colicelli, 1995; L. H. et cellulose filters, and probed with specific antibodies. Samples were al., unpublished data) or neo were inserted into a unique EcoRI site analyzed using monoclonal antibodies directed toward ABL (pex5), 3Ј to the retroviral LTR and 5Ј to the tk promoter and BCR-ABL. phosphotyrosine (4G10, UBI, Lake Placid, NY), and RIN1 (Transduc- This vector allows the simultaneous expression of both genes from tion Laboratories, Lexington, KY) or polyclonal antibodies directed a single retrovirus (Sawyers et al., 1995). toward the amino terminus of RIN1 (L. H. et al., unpublished data) The RIN1N⌬ mutant was created from pGEX-RIN1N (L. H. et al., or BCR (Timmons and Witte, 1989). unpublished data) by digestion with HindIII and KpnI, which releases a 150 bp fragment encoding the proline-rich region. The plasmid Immune Complex Kinase Assays was religated in the presence of oligonucleotide (AGCTTGTAC) and For in vitro kinase assays, 293T cells were transfected with retroviral dATP to generate pGEX-RIN1N⌬. This mutant was then cloned into vectors expressing the different cDNAs.Cells were lysed in immuno- the retroviral vector. precipitation buffer (10 mM Tris [pH 7.5], 130 mM NaCl, 20 mM NaF, The RIN1 point mutants were created using in vitro oligonucleo- 1 mM sodium vanadate, 1% Triton X-100, 10% glycerol) at 5 ϫ 106 tide-directed mutagenesis (Muta-Gene, BioRad) on single-stranded cells/ml and were clarified by centrifugation (100,000 ϫ g) for 30 template of RIN1N in pBluescriptSK (Stratagene). The oligonucleo- min. The lysates were incubated with polyclonal antibodies directed tides used to generate each mutation were 5Ј-AGGACCCACTGTTC toward Abl (pex5) for 1 hr with subsequent immunoprecipitation GACGTGCCCAAT (Y36 to F), 5Ј-TCTCCAGCCACTTCATCCTCGAG using protein A–Sepharose beads (Pharmacia). The immunoprecipi- AGCCCTGGC (Y121 to F), 5Ј-CCAGCTCATCTGCGCCTTCTGCCAC tates were washed three times with immunoprecipitation buffer and

ACCCG (Y148 to F). The triple tyrosine mutant was created using once with kinase buffer (10 mM Tris [pH 7.5], 10 mM MgCl2, 100 ␮M all three oligonucleotides in the same reaction. In each case a re- sodium vanadate). Kinase reactions were initiated by the addition striction endonuclease site was introduced to facilitate mutant of 10 ␮Ci 32P-␥ ATP at 30ЊC. After 10 min the reactions were termi- screening. nated with SDS-containing sample buffer. Regulation of BCR-ABL by RIN1 781

For second-cycle immunoprecipitations, sample buffer–termi- Erpel, T., and Courtneidge, S.A. (1995). Src family protein tyrosine nated kinase reactions were diluted in immunoprecipitation buffer kinases and cellular signal transduction pathways. Curr. Opin. Cell to a final SDS concentration of 0.1%. The samples were incubated Biol. 7, 176–182. with polyclonal antibodies directed toward Abl (pex5) or RIN1N for 1 Feller, S.M., Ren, R., Hanafusa, H., and Baltimore, D. (1994). SH2 hr with subsequent immunoprecipitation using protein A–Sepharose and SH3 domains as modular adhesives: the interaction of Crk and beads. The samples were washed three times with immunoprecipi- Abl. Trends Biochem. Sci. 19, 453–458. tation buffer and were eluted using sample buffer. Feng, G.S., Ouyang, Y.B., Hu, D.P., Shi, Z.Q., Gentz, R., and Ni, J. Autokinase activity and RIN1 phosphorylation was analyzed by (1996). Grap is a novel SH3-SH2-SH3 adaptor protein that couples SDS-PAGE (10% polyacrylamide). The gels were incubated in fixing tyrosine kinases to the Ras pathway. J. Biol. Chem. 271, 12129– solution (10% acetic acid, 10% methanol) for 10 min and then were 12132. treated with 1 M potassium hydroxide for 2 hr at 55ЊC. After alkali treatment the gels were treated in fixing solution, dried, and sub- Franz, W.M., Berger, P., and Wang, J.Y.J. (1989). Deletion of an jected to autoradiography. N-terminal regulatory domain of the c-abl tyrosine kinase activates its oncogenic potential. EMBO J. 8, 137–147. Analysis of RIN1 in Leukemic Cells Gishizky, M.L., Johnson-White, J., and Witte, O.N. (1993). Efficient The human Ph1-positive leukemic cell lines KCL22 (CML in transplantation of BCR/ABL induced chronic myelogenous leuke- lymphoblast crisis), K562 (CML in erythroblast crisis), and ALL1 (ALL mia-like syndrome inmice. Proc. Natl. Acad. Sci. USA 90, 3755–3759. in lymphoblast crisis) were grown in RPMI supplemented with 10% Goga, A., McLaughlin, J., Afar, D.E.H., Saffran, D.C., and Witte, O.N. fetal bovin serum (FBS). Lysates were prepared from 5 ϫ 107 cells (1995). Alternative signals to RAS for hematopoietic transformation using 5 ml of immunoprecipitation buffer. RIN1 and P210 BCR- by the BCR-ABL oncogene. Cell 82, 981–988. ABL were immunoprecipitated with anti-RIN1N and anti-Abl (pex5) Han, L., and Colicelli, J. (1995). A human protein selected for interfer- polyclonal antibodies, respectively. The immunoprecipitates were ence with ras function interacts directly with ras and competes with subjected to immune complex kinase assays as described above raf1. Mol. Cell. Biol. 15, 1318–1323. or were analyzed by Western blotting using monoclonal antibodies Han, L., Wong, D., Dhaka, A., Afar, D., White, M., Xie, W., Herschman, directed to phosphotyrosine, RIN1, and Abl. H., Witte, O., and Colicelli, J. (1997). Protein binding and signaling properties of RIN1 suggest a unique effector function. Proc. Natl. Acknowledgments Acad. Sci. USA 94, 4954–4959. Heisterkamp, N., Jenster, G., ten Hoeve, J., Zovich, D., Pattengale, Correspondence should be addressed to O. N. W. (e-mail: owenw@ P.K., and Groffen, J. (1990). Acute leukaemia in /abl transgenic microbio.lifesci.ucla.edu). We are grateful to Larry Zipursky, Arnie mice. Nature 344, 251–253. Berk, Charles Sawyers, Zhigang Weng, and Stefan Mahlmann for Jackson, P., and Baltimore, D. (1989). N-terminal mutations activate helpful discussions and critical reading of the manuscript; to J. C. the leukemogenic potentialof the myristoylated form of c-abl. EMBO White and Luke Hsieh for photography; to Julia Shimaoka for prepa- J. 8, 449–456. ration of the manuscript; and to James Johnson for excellent techni- cal assistance. D. E. H. A. is a Special Fellow of the Leukemia Society Kazlauskas, A. (1994). Receptor tyrosine kinases and their targets. of America. O. N. W. is an Investigator of theHoward Hughes Medical Curr. Opin. Genetics Dev. 4, 5–14. Institute. This work was supported in part by National Institutes of Kelliher, M.A., McLaughlin, J., Witte, O.N., and Rosenberg, N. (1990). Health grants R35CA53867 to O. N. W., CA56301 to J. C., and Induction of a chronic myelogenous leukemia-like syndrome in mice CA24220 to N. R. with v-abl and BCR/ABL. Proc. Natl. Acad. Sci. USA 87, 6649–6653. Kelliher, M., Knott, A., McLaughlin, J., Witte, O.N., and Rosenberg, Received May 8, 1997. N. (1991). Differences in oncogenic potency but not target cell speci- ficity distinguish the two forms of the BCR/ABL oncogene. Mol. Cell. References Biol. 11, 4710–4716. Kmiecik, T.E., and Shalloway, D. (1987). Activation and suppression Afar, D.E.H., Goga, A., McLaughlin, J., Witte, O., and Sawyers, C.L. of pp60c-src transforming ability by mutation of its primary sites of (1994). Differential rescue of BCR-ABL point mutants with c-MYC. tyrosine phosphorylation. Cell 49, 65–73. Science 264, 424–426. Konopka, J.B., Watanabe, S.M., and Witte, O.N. (1984). An alteration Afar, D.E.H., McLaughlin, J., Sherr, C.J., Witte, O.N., and Roussel, of the human c-abl protein in K562 leukemia cells unmasks associ- M.F. (1995). Signaling by ABL oncogenes through cyclin D1. Proc. ated tyrosine kinase activity. Cell 37, 1035–1042. Natl. Acad. Sci. USA 92, 9540–9544. Kurzrock, R., Gutterman, J., and Talpaz, M. (1988). The molecular Alexandropoulos, K., and Baltimore, D. (1996). Coordinate activation genetics of Philadelphia chromosome-positive leukemias. N. Engl. of c-Src by SH3- and SH2-binding sites on a novel, p130Cas-related J. Med. 319, 990–998. protein, Sin. Genes Dev. 10, 1341–1355. Laneuville, P. (1995). Abl tyrosine protein kinase. Semin. Immmunol. Alexandropoulos, K., Cheng, G., and Baltimore, D. (1995). Proline- 7, 255–266. rich sequences that bind to Src homology 3 domains with individual Lewis, J.M., Baskaran, R., Taagepera, S., Schwartz, M.A., and Wang, specificities. Proc. Natl. Acad. Sci. USA 92, 3110–3114. J.Y.J. (1996). Integrin regulation of c-Abl tyrosine kinase activity Brown, M.T., and Cooper, J.A. (1996). Regulation, substrates and and cytoplasmic-nuclear transport. Proc. Natl. Acad. Sci. USA 93, functions of src. Biochim. Biophys. Acta 1287, 121–149. 15174–15179. Carpino, N., Wisniewski, D., Strife, A., Marshak, D., Kobayashi, R., Lugo, T., and Witte, O.N. (1989). The BCR/ABL oncogene transforms Stillman, B., and Clarckson, B. (1997). p62doc: a constitutively tyro- Rat-1 cells and cooperates with v-myc. Mol. Cell. Biol. 9, 1263–1270. sine-phosphorylated, GAP-associated protein in chronic myeloge- Lugo, T.G., Pendergast, A., Muller, A.J., and Witte, O.N. (1990). Tyro- nous leukemia progenitor cells. Cell 88, 197–204. sine kinase activity and transformation potency of bcr-abl oncogene Chou, M.M., and Blenis, J. (1996). The 70 kDa S6 kinase complexes products. Science 247, 1079–1082. with and is activated by the Rho family G proteins Cdc42 and Rac1. Mandanas, R.A., Leibowitz, D.S., Gharehbaghi, K., Tauchi, T., Bur- Cell 85, 573–583. gess, G.S., Miyazawa, K., Jayaram, H.N., and Boswell, H.S. (1993). Daley, G.Q., Van Etten, R.A., and Baltimore, D. (1990). Induction of Role of p21 RAS in p210 bcr-abl transformation of murine myeloid chronic myelogenous leukemia in mice by the P210bcr/abl gene of cells. Blood 82, 1838–1847. the Philadelphia chromosome. Science 247, 824–830. Manser, E., Leung, T., Salihuddin, H., Zhao, Z., and Lim, L. (1994). Elefanty, A.G., Hariharan, I.K., and Cory, S. (1990). bcr-abl, the hall- A brain serine/threonine protein kinase activated by Cdc42 and Rac mark of chronic myeloid leukaemia in man, induces multiple haemo- 1. Nature 367, 40–46. poietic neoplasms in mice. EMBO J. 9, 1069–1078. Matsuguchi, T., Salgia, R., Hallek, M., Eder, M., Druker, B., Ernst, Immunity 782

T.J., and Griffin, J.D. (1994). Shc phosphorylation in myeloid cells forms stable complexes in vivo with v-Crk and v-Src in a tyrosine is regulated by granulocyte macrophage colony-stimulating factor, phosphorylation-dependent manner. EMBO J. 13, 3748–3756. interleukin-3, and steel factor and is constitutively increased by Sawyers, C.L., McLaughlin, J., and Witte, O.N. (1995). Genetic re- BCR/ABL p210 . J. Biol. Chem. 269, 5016–5021. quirement for ras in thetransformation of fibroblasts andhematopoi- McLaughlin, J., Chianese, E., and Witte, O.N. (1987). In vitro transfor- etic cells by the Bcr-Abl oncogene. J. Exp. Med. 181, 307–313. mation of immature hematopoietic cells by the P210 BCR/ABL onco- Scott, M.L., Van Etten, R.A., Daley, G.Q., and Baltimore, D. (1991). gene product of the Philadelphia chromosome. Proc. Natl. Acad. v-abl causes hematopoietic disease distinct from that caused by Sci. USA 84, 6558–6562. bcr-abl. Proc. Natl. Acad. Sci. USA 88, 6506–6510. McLaughlin, J., Chianese, E., and Witte, O.N. (1989). Alternative Soekarman, D., van Denderen, J., Hoefsloot, L., Moret, M., Meeuw- forms of the BCR-ABL oncogene have quantitatively different poten- sen, T., van Baal, J., Hagemeijer, A., and Grosveld, G. (1990). A novel cies for stimulation of immature lymphoid cells. Mol. Cell. Biol. 9, variant of the bcr-abl fusion product in Philadelphia chromosome- 1866–1874. positive acute lymphoblastic leukemia. Leukemia 4, 397–403. McWhirter, J.R., and Wang, J.Y.J. (1991). Activation of tyrosine ki- Tauchi, T., Boswell, H.S., Leibowitz, D., and Broxmeyer, H.E. (1994). nase and microfilament-binding functions of c-abl by bcr sequences Coupling between p210bcr-abl and Shc and Grb2 adaptor proteins in bcr/abl fusion proteins. Mol. Cell. Biol. 11, 1553–1565. in hematopoietic cells permits growth factor receptor-independent Muller, A.J., Young, J.C., Pendergast, A.-M., Pondel, M., Landau, link to Ras activation pathway. J. Exp. Med. 179, 167–175. N.R., Littman, D.R., andWitte, O.N. (1991). BCR first exon sequences ten Hoeve, J., Kaartinen, V., Fioretos, T., Haataja, L., Voncken, J.W., specifically activate Bcr/Abl tyrosine kinase oncogene of Philadel- Heisterkamp, N., and Groffen, J. (1994). Cellular interactions of CRKL phia chromosome positive human leukemias. Mol. Cell. Biol. 11, and SH2-SH3 adaptor protein. Cancer Res. 54, 2563–2567. 1785–1792. Timmons, M.S., and Witte, O.N. (1989). Structural characterization Oda, T., Heaney, C., Hagopian, J.R., Okuda, K., Griffin, J.D., and of the BCR gene product. Oncogene 4, 559–567. Druker, B.J. (1994). Crkl is the major tyrosine-phosphorylated protein in neutrophils from patients with chronic myelogenous leukemia. J. Valius, M., and Kazlauskas, A. (1993). Phospholipase C-gamma1 Biol. Chem. 269, 22925–22928. and phosphatidylinositol 3 kinase are the downstream mediators of the PDGF receptor’s mitogenic signal. Cell 73, 321–334. Pawson, T. (1995). Protein modules and signaling networks. Nature 373, 573–580. Van Aelst, L., Barr, M., Marcus, S., Polverino, A., and Wigler, M. (1993). Complex formation between RAS and RAF and other protein Pear, W.S., Nolan, G.P., Scott, M.L., and Baltimore, D. (1993). Pro- kinases. Proc. Natl. Acad. Sci. USA 90, 6213–6217. duction of high titer helper-free retroviruses by transient transfec- tion. Proc. Natl. Acad. Sci. USA 90, 8392–8396. Van Etten, R.A., Jackson, P., and Baltimore, D. (1989). The mouse type IV c-abl gene product is a nuclear protein, and activation of Pendergast, A.M., Muller, A.J., Havlik, M.H., Clark, R., McCormick, transforming ability is associated with cytoplasmic localization. Cell F., and Witte, O.N. (1991). Evidence for regulation of the ABL tyrosine 58, 669–678. kinase by a cellular inhibitor. Proc. Natl. Acad. Sci. USA 88, 5927– 5931. Vojtek, A.B., Hollenberg, S.M., and Cooper, J.A. (1993). Mammalian ras interacts directly with the serine/threonine raf. Cell 74, 205–214. Pendergast, A.M., Gishizky, M.L., Havlik, M.H., and Witte, O.N. (1993a). SH1 domain autophosphorylation of P210 BCR/ABL is re- Wen, S.T., Jackson, P.K., and Van Etten, R.A. (1996). The cytostatic quired for transformation but not growth factor-independence. Mol. function of c-Abl is controlled by multiple nuclear localization signals Cell. Biol. 13, 1728–1736. and requires the p53 and Rb tumor suppressor gene products. EMBO J. 15, 1583–1595. Pendergast, A.M., Quilliam, L.A., Cripe, L.D., Bassing, C.H., Dai, Z., Li, N., Batzer, A., Rabun, K.M., Der, C.J., Schlessinger, J., et al. Yamanashi, Y., and Baltimore, D. (1997). Identification of the Abl- (1993b). BCR-ABL-induced oncogenesis is mediated by direct inter- and rasGAP-associated 62 kDa protein as a docking protein, Dok. action with the SH2 domain of the GRB-2 adaptor protein. Cell 75, Cell 88, 205–211. 175–185. Piwnica-Worms, H., Saunders, K.B., Roberts, T.M., Smith, A.E., and Cheng, S.H. (1987). Tyrosine phosphorylation regulates the bio- chemical and biological properties of pp60c-src. Cell 49, 75–82. Puil, L., Liu, J., Gish, G., Mbamalu, G., Bowtell, D., Pelicci, P.G., Arlinghaus, R., and Pawson, T. (1994). Bcr-Abl oncoproteins bind directly to activators of the Ras signaling pathway. EMBO J. 13, 764–773. Ren, R., Mayer, B.J., Cicchetti, P., and Baltimore, D. (1993). Identifi- cation of a ten-amino acid proline-rich SH3 binding site. Science 259, 1157–1161. Rennick, D., Yang, G., Gemmell, L., and Lee, F. (1987). Control of hemopoiesis by a bone marrow stromal cell clone: lipopolysaccha- ride- and interleukin-1-inducible production of colony-stimulating factors. Blood 69, 682–691. Rickles, R.J., Botfield, M.C., Weng, Z., Taylor, J.A., Green, O.M., Brugge, J.S., and Zoller, M.J. (1994). Identification of Src,Fyn,Lyn, PI3K and Abl SH3 domain ligands using phage display libraries. EMBO J. 13, 5598–5604. Rosenberg, N., and Witte, O.N. (1988). The viral and cellular forms of the Abelson (abl) oncogene. Adv. Virus Res. 39, 39–81. Roussel, M.F., Cleveland, J.L., Shurtleff, S.A., and Sherr, C.J. (1991). Myc rescue of a mutant CSF-1 receptorimpaired in mitogenic signal- ing. Nature 353, 361–363. Roussel, M.F., Theodoras, A.M., and Sherr, C.J. (1995). Rescue of defective mitogenic signaling by D-type cyclins. Proc. Natl. Acad. Sci. USA 92, 6837–6841. Sakai, R., Iwamatsu, A., Hirano, N., Ogawa, S., Tanaka, T., Mano, H., Yazaki, Y., and Hirai, H. (1994). A novel signaling molecule, p130,