Physical and functional interactions between Lyn and p34cdc2 kinases in irradiated human B-cell precursors.

Exposure of human B-cell precursors (BCP) to ionizing radiation results in cell cycle arrest at the G2-M checkpoint as a result of inhibitory tyrosine phosphorylation of p34cdc2 . Here, we show that ionizing radiation promotes physical interactions between p34cdc2 and the Src family protein-tyrosine kinase Lyn in the cytoplasm of human BCP leading to tyrosine phosphorylation of p34cdc2. Lyn kinase immunoprecipitated from lysates of irradiated BCP as well as a full-length glutathione S-transferase (GST)-Lyn fusion protein-phosphorylated recombinant human p34cdc2 on tyrosine 15. Furthermore, Lyn kinase physically associated with and tyrosine-phosphorylated p34cdc2 kinase in vivo when co-expressed in COS-7 cells. Binding experiments with truncated GST-Lyn fusion proteins suggested a functional role for the SH3 rather than the SH2 domain of Lyn in Lyn-p34cdc2 interactions in BCP. The first 27 residues of the unique amino-terminal domain of Lyn were also essential for the ability of GST-Lyn fusion proteins to bind to p34cdc2 from BCP lysates. Ionizing radiation failed to cause tyrosine phosphorylation of p34cdc2 or G2 arrest in Lyn kinase-deficient BCP, supporting an important role of Lyn kinase in radiation-induced G2 phase-specific cell cycle arrest. Our findings implicate Lyn as an important cytoplasmic suppressor of p34cdc2 function.

Exposure of human B-cell precursors (BCP) to ionizing radiation results in cell cycle arrest at the G 2 -M checkpoint as a result of inhibitory tyrosine phosphorylation of p34 cdc2 . Here, we show that ionizing radiation promotes physical interactions between p34 cdc2 and the Src family protein-tyrosine kinase Lyn in the cytoplasm of human BCP leading to tyrosine phosphorylation of p34 cdc2 . Lyn kinase immunoprecipitated from lysates of irradiated BCP as well as a full-length glutathione S-transferase (GST)-Lyn fusion protein-phosphorylated recombinant human p34 cdc2 on tyrosine 15. Furthermore, Lyn kinase physically associated with and tyrosine-phosphorylated p34 cdc2 kinase in vivo when coexpressed in COS-7 cells. Binding experiments with truncated GST-Lyn fusion proteins suggested a functional role for the SH3 rather than the SH2 domain of Lyn in Lyn-p34 cdc2 interactions in BCP. The first 27 residues of the unique amino-terminal domain of Lyn were also essential for the ability of GST-Lyn fusion proteins to bind to p34 cdc2 from BCP lysates. Ionizing radiation failed to cause tyrosine phosphorylation of p34 cdc2 or G 2 arrest in Lyn kinase-deficient BCP, supporting an important role of Lyn kinase in radiation-induced G 2 phase-specific cell cycle arrest. Our findings implicate Lyn as an important cytoplasmic suppressor of p34 cdc2 function.
B-cell precursor (BCP) 1 leukemia is the most common childhood malignancy and represents one of the most radiationresistant forms of human cancer (1)(2)(3)(4)(5)(6)(7)(8). Recent studies demonstrated that Ͼ75% of clonogenic BCP leukemia cells from more than one-third of the newly diagnosed patients and virtually all of the relapsed patients are able to repair potentially lethal or sublethal DNA damage induced by radiation doses that correspond to the clinical total body irradiation dose fractions (i.e. 2-3 Gy) (6). Consequently, the vast majority of BCP leukemia patients undergoing total body irradiation in the context of bone marrow transplantation relapse within the first 12 months and only 15-20% survive disease-free beyond the first 2 years (9,10).
Ionizing radiation and various DNA damaging agents cause an accumulation of cells in G 2 phase of the cell cycle (11)(12)(13)(14). Several lines of evidence indicate that this transient G 2 arrest allows the cells to repair potentially lethal or sublethal DNA lesions induced by radiation or other DNA damaging agents. Cells that are unable to show this response are more sensitive to DNA damaging agents, and drugs that abolish this response sensitize cells to DNA damaging agents (11,(15)(16)(17)(18)(19)(20)(21)(22). A human lymphoma cell line that displayed markedly enhanced sensitivity to DNA damage by nitrogen mustard was found to be defective in the G 2 phase checkpoint control (14). The elucidation of the mechanism by which ionizing radiation induces G 2 arrest in BCP leukemia cells could lead to a rational design of radiation sensitizers that impair the repair of radiation-induced DNA damage by leukemia cells and improve the outcome after total body irradiation and bone marrow transplantation.
The molecular mechanism by which ionizing radiation induces G 2 arrest in the human cell cycle and prevents entry into mitosis has not yet been deciphered, but preliminary evidence suggested that it may involve the inactivation of p34 cdc2 kinase by inhibitory tyrosine phosphorylation on tyrosine 15 (23)(24)(25). p34 cdc2 kinase is the catalytic subunit of mitosis promoting factor (MPF), and its activation is a prerequisite for induction of M phase (26 -28). Recent studies demonstrated that exposure of BCP leukemia cells to ␥-rays results in enhanced tyrosine phosphorylation of multiple substrates including p34 cdc2 kinase (25,29). Furthermore, the protein-tyrosine kinase (PTK) inhibitor herbimycin A was able to prevent radiationinduced tyrosine phosphorylation and inactivation of p34 cdc2linked histone H1 kinase activity as well as mitotic arrest (25), supporting the notion that radiation-induced cell cycle arrest of BCP leukemia cells at G 2 -M transition is likely triggered by inhibitory tyrosine phosphorylation of p34 cdc2 kinase.
Several mitotic control genes encoding for protein-tyrosine kinases or protein-tyrosine phosphatases have been shown to coordinately regulate MPF function by altering tyrosine phosphorylation of p34 cdc2 kinase (30 -33). Genetic experiments in fission yeast have shown that the WEE1 kinase negatively regulates mitosis by phosphorylating p34 cdc2 on Tyr 15 , thereby inactivating p34 cdc2 -cyclin B complex (32,33). Preliminary genetic studies in fission yeast initially suggested an important role for WEE1 kinase in radiation-induced mitotic arrest at G 2 -M transition (34). However, a more recent study using Schizosaccharomyces pombe cells lacking functional wee1 gene product provided convincing evidence that fission yeast WEE1 kinase is not required for radiation-induced mitotic arrest (35). Furthermore, we detected no increase of human WEE1 kinase activity after radiation of BCP leukemia cells, as measured by autophosphorylation, tyrosine phosphorylation of (a) recombinant human p34 cdc2 -cyclin B complex isolated from lysates of insect cells coinfected with recombinant viruses encoding GSTcyclin B and [Arg 33 ]p34 cdc2 , an inactive mutant of p34 cdc2 , (b) p34 cdc2 -cyclin B complex biochemically purified from starfish oocytes, or (c) a synthetic peptide derived from the p34 cdc2 amino-terminal region, [Lys 19 ]Cdc2(6 -20)NH 2 (25). Human WEE1 kinase isolated from unirradiated or irradiated BCP leukemia cells had minimal PTK activity toward the aforementioned substrates (25). Thus, the identity of radiation-responsive kinases which inactivate MPF in human BCP leukemia cells remains unknown.
Lyn kinase is the predominant PTK in human BCP leukemia cells (36,37). The enzymatic activity of Lyn in human BCP leukemia cells is rapidly stimulated by ionizing radiation (38). Similarly, exposure of myeloid leukemia cells to ionizing radiation has been reported to cause Lyn kinase activation (39). Lyn kinase was shown to physically associate with p34 cdc2 kinase in lysates of irradiated myeloid leukemia cells, however the significance of Lyn kinase activation or its association with p34 cdc2 kinase in myeloid cells has not been examined (39). These recent observations prompted the hypothesis that p34 cdc2 kinase may associate with and serve as a substrate for Lyn in BCP leukemia cells.
Here, we show that the Lyn kinase associates physically and functionally with p34 cdc2 in the cytoplasm of BCP. Immunoblotting of Lyn immune complexes with an anti-p34 cdc2 -Cter antibody (where Cter indicates COOH terminus) and immunoblotting of p34 cdc2 immune complexes with an anti-Lyn antibody provided evidence for an association between Lyn and p34 cdc2 kinases in lysates of BCP even before radiation exposure. Irradiation of BCP stimulated the Lyn kinase, and concomitant with Lyn kinase activation following radiation exposure, p34 cdc2 became detectable in the Lyn immune complexes as a tyrosine-phosphorylated protein substrate. The abundance of the Lyn protein, as estimated by anti-Lyn Western blot analysis, did not change during the course of the experiment, suggesting increased enzymatic activity of Lyn. However, the abundance of the p34 cdc2 protein in the same Lyn immune complexes, as determined by anti-Cdc2-Cter Western blot analysis, was significantly increased after radiation exposure, suggesting that enhanced tyrosine phosphorylation of p34 cdc2 which parallels the Lyn activation is at least in part due to radiation-induced promotion of the physical association between Lyn and p34 cdc2 in NALM-6 cells. Binding experiments with truncated GST-Lyn fusion proteins suggested a functional role for the SH3 rather than the SH2 domain of Lyn in Lyn-p34 cdc2 interactions in BCP. The first 27 residues of the unique amino-terminal domain of Lyn were also essential for the ability of GST-Lyn fusion proteins to bind to p34 cdc2 from BCP lysates. Lyn kinase immunoprecipitated from lysates of irradiated BCP as well as a full-length GST-Lyn fusion proteinphosphorylated recombinant human p34 cdc2 on tyrosine 15. The ability of the Lyn kinase to phosphorylate recombinant human p34 cdc2 on Tyr 15 was amplified following radiation exposure. Lyn kinase interacts with and tyrosine-phosphorylates p34 cdc2 in vivo when these kinases are coexpressed in COS-7 cells. Ionizing radiation failed to induce p34 cdc2 tyrosine phosphorylation or G 2 arrest in Lyn kinase-deficient BCP leukemia cells expressing Fyn, Blk, and Lck kinases. These convergent observations constitute a strong argument for an important role of a cytoplasmic signal transduction pathway intimately linked to the Lyn kinase in radiation-induced G 2 phase-specific cell cycle arrest of human BCP leukemia cells. Since the dura-tion of the G 2 arrest is a major determinant of radiation resistance in BCP leukemias, this knowledge may lead to the design of a leukemia-specific radiosensitization method.
Our findings implicate Lyn as an important cytoplasmic suppressor of p34 cdc2 function. Lyn kinase may serve as an integral component of a physiologically important surveillance and repair mechanism for DNA damage by delaying the G 2 -M transition in cells exposed to mutagenic oxygen free radicals, thereby allowing them to repair their DNA damage prior to mitosis. Lyn kinase may also protect the cell from the potentially catastrophic consequences of premature cytoplasmic p34 cdc2 activation by maintaining the p34 cdc2 -cyclin B complex in its inactive, tyrosine phoshorylated state.

EXPERIMENTAL PROCEDURES
Irradiation of Cells-NALM-6 pre-B leukemia cells and Lyn kinasedeficient leukemic BCP from acute lymphoblastic leukemia (ALL) patients were obtained from the Cell Bank of the Childrens Cancer Group ALL Biology Reference Laboratory in Minneapolis, MN. Where indicated, cells (5 ϫ 10 5 /ml in plastic tissue culture flasks) were irradiated ( 137 Cs irradiator; J. L. Shephard, Glendale, CA, model Mark I) with 1 Gy (ϭ 100 rads) or 2 Gy at a dose rate of 1 Gy/min under aerobic conditions, as described previously (5-7, 40, 41).
Immunoblot Analysis of Tyrosine Phosphorylation of p34 cdc2 and Its Interaction with Lyn Kinase in BCP Leukemia Cells-p34 cdc2 kinase or Lyn kinase were immunoprecipitated from Nonidet P-40 lysates of BCP leukemia cells using an anti-Cdc2-Cter antibody (Upstate Biotechnology, Inc., Lake Placid, NY) or an anti-Lyn antibody, according to previously published procedures (25,29,36). In brief, cells (5 ϫ 10 6 cells/ sample) were solubilized in 0.5 ml of 1% Nonidet P-40 lysis buffer (50 mM Tris-Cl, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, plus 1 mM EDTA) containing 0.1 mM sodium orthovanadate and 1 mM sodium molybdate as phosphatase inhibitors, and 10 g/ml leupeptin, 10 g/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride as protease inhibitors on ice for 30 min. Lysates were spun twice at 12,000 ϫ g for 15 min at 4°C prior to immunoprecipitation. Indicated amounts of the cell lysates were immunoprecipitated with anti-Cdc2-Cter (10 g of antibody/200 g of lysate) or anti-Lyn (2 l of antibody/200 g of lysate) overnight at 4°C.The immune complexes were collected with 50 l of a 1:1 (v/v) slurry of protein A-Sepharose (Repligen Corp., Cambridge, MA) in Nonidet P-40 buffer. The immunoprecipitates were washed four times with Nonidet P-40 buffer, resuspended in 2 ϫ SDS reducing sample buffer, and boiled. Samples were run on 10.5% SDS-polyacrylamide gel electrophoresis (PAGE) gels, transferred to PVDF membranes, and subsequently immunoblotted with either anti-phosphotyrosine (5 g/ ml) or anti-Cdc2-Cter (2 g/ml) antibodies. 125 I-Labeled protein A was used to detect tyrosine-phosphorylated proteins or p34 cdc2 kinase. In some experiments, we used immunoblotting with anti-Lyn antibody to detect Lyn kinase in p34 cdc2 immune complexes and immunoblotting with anti-Cdc2-Cter antibody to detect p34 cdc2 kinase in Lyn immune complexes. Blots were incubated with 1 Ci/ml 125 I-labeled protein A (specific activity ϭ 30 Ci/g; ICN Biomedicals) in blocking solution. After a 30-min incubation in 125 I-protein A, blots were washed, as indicated above, dried, and autoradiographed using a XAR-5 film (Eastman Kodak Co.). Molecular masses (in kDa) of the phosphotyrosyl protein substrates were calculated from prestained molecular size markers (Amersham Corp.) that were run as standards.
Immune Complex Kinase Assays-To evaluate the effects of ionizing radiation on the kinase activity of Lyn, exponentially growing cells (5 ϫ 10 6 /ml in ␣-minimal essential medium) were irradiated and lysed at the indicated time points in a Nonidet P-40 buffer. 200 g of cell lysates/ sample were immunoprecipitated with a rabbit anti-Lyn antibody, as described previously (25,29,36). Samples were assayed for kinase activity during a 10-or 20-min incubation in the presence of [␥-32 P]ATP (50 Ci/mol) in the presence and absence of synthetic Cdc2 peptides or human p34 cdc2 -cyclin B complex as exogenous substrates (25). In initial experiments, kinase reactions consisted of 10 l of Lyn immunoprecipitate, 5 l of assay buffer (0.25 M Tris-HCl pH 7.0, 0.125 M MgCl 2 , 0.025 M MnCl 2 , and 0.25 mM Na 3 VO 4 ) and 5 l of 1.5 mM substrate peptide [Lys 19 ]Cdc2(6 -20)-NH 2 (sequence: KVRKIGEGTYGVVKK) (Upstate Biotechnology, Inc.), a synthetic peptide derived from p34 cdc2 kinase. The reactions were initiated by the addition of 5 l of 0.5 mM [␥-32 P]ATP (specific activity ϭ 10 5 cpm/pmol) and incubation for 30 min at 30°C. The reaction was terminated by the addition of 10 l of glacial acetic acid, and then 25 l of the reaction mixture was spotted onto a P-81 phosphocellulose disc. The discs were washed four times with 0.75% phosphoric acid and once with acetone. [Val 12 ,Ser 14 ,Lys 19 ]Cdc2(6 -20)-NH 2 and [Phe 15 ,Lys 19 ]Cdc2(6 -20)-NH 2 (Upstate Biotechnology, Inc.) were included as control peptides. The PTK activity of Lyn toward the Cdc2 peptides was measured by incorporation of 32 P into the peptide substrates and expressed as the background-subtracted counts/min or pmol of PO 4 Ϫ incorporated/min. For subsequent experiments, human p34 cdc2 -cyclin B complex was isolated from lysates (100 g/sample) of insect cells coinfected with recombinant viruses encoding GST-cyclin B and [Arg 33 ]p34 cdc2 using glutathione-agarose beads (Sigma). The in vitro phosphorylation of p34 cdc2 by immunoprecipitated Lyn was assayed after a 20-min kinase reaction at 30°C in kinase buffer (50 mM Tris-Cl, pH 7.4, 5 mM MnCl 2 , 10 mM MgCl 2 , 1 mM DTT, and 50 M ATP). The kinase reaction was initiated by the addition of p34 cdc2 (50 l of the GST-cyclin B-[Arg 33 ]p34 cdc2 precipitate/sample) and [␥-32 P]ATP (20 Ci). Following the kinase reaction, these samples were fractionated on 9.5% polyacrylamide gels, detected by autoradiography and incorporation of 32 P was quantitated by a 4-min liquid scintillation counting of excised bands.
Phosphoamino Acid Analysis and Phosphotryptic Peptide Mapping-For phosphoamino acid analysis, protein bands were excised and hydrolyzed, as described previously (25,36,41). For two-dimensional phosphotryptic peptide mapping, 32 P-labeled protein bands were excised and subjected to enzymatic digestion with 100 g/ml trypsin (Sigma) overnight in 50 mM ammonium bicarbonate. Supernatants were dried by centrifugal evaporation, and dried samples were resuspended in 4 l of a buffer containing 7.8% glacial acetic acid and 2.5% formic acid. Labeled peptides were separated on thin layer phosphocellulose plates (Kodak) by electrophoresis at pH 1.9 for 30 min at 1,000 V, followed by ascending chromatography in a buffer containing 37.5% n-butanol, 7.5% glacial acetic acid, and 25% pyridine. Subsequently, air-dried plates were exposed to Kodak XAR-5 film. Prior to phosphoamino acid analysis and tryptic peptide mapping, protein bands of interest were excised and Cerenkov-counted for 32 P content.
Binding Assays with GST-Lyn Fusion Proteins-Truncated GST-Lyn fusion proteins corresponding to various domains of Lyn (44) were purchased from PharMingen, San Diego, CA. GST-Lyn fusion proteins were non-covalently bound to glutathione-agarose beads (Sigma) under conditions of saturating protein. In brief, 25 g of each fusion protein was incubated with 50 l of beads for 2 h at 4°C. The beads were washed three times with 1% Nonidet P-40 buffer. Nonidet P-40 lysates of NALM-6 cells were prepared as described above, and 250 g of the lysate was incubated with 50 l of fusion protein-coupled beads for 2 h on ice. The fusion protein adsorbates were washed with ice-cold 1% Nonidet P-40 buffer and resuspended in reducing SDS sample buffer. Samples were boiled for 5 min and then fractionated on SDS-PAGE, as described previously (36). SDS-PAGE gels were transferred to Immobilon-P (Millipore) membranes. Membranes were immunoblotted with anti-Cdc2-Cter (2 g/ml), as described (25,36). 125 I-Labeled protein A was used to detect p34 cdc2 kinase.
In Vitro Kinase Assays Using GST-WEE1 and GST-Lyn Fusion Proteins-A highly purified preparation of Lyn was prepared for these experiments by cloning a lyn cDNA (45) into the expression vector pBMS-1 (46), which directs the production of a recombinant baculovirus encoding a 83-kDa GST-Lyn fusion protein in insect cells. The GST-Lyn protein was purified to homogeneity using glutathione-Sepharose chromatography (46). The ability of GST-Lyn (1:100 dilution) and GST-p49 WEE1Hu (1:10 dilution, kindly provided by Dr. Laura Parker) to phosphorylate [Arg 33 ]p34 cdc2 was measured in a 20-min kinase reaction at 30°C in kinase buffer. The kinase reaction was initiated by the addition of p34 cdc2 (50 l/sample) and [␥-32 P]ATP (20 Ci). Following the kinase reactions, samples were boiled in 2 ϫ SDS reducing sample buffer, and proteins were fractionated on 15% polyacrylamide gels and visualized by autoradiography. Two-dimensional phosphoamino acid analysis and phosphotryptic peptide mapping of p34 cdc2 kinase were performed as described above.
Transfection Experiments-lyn and cdc2 cDNAs were expressed transiently in COS-7 cells by Lipofectamine lipid encapsulation (47,48). COS-7 cells were allowed to grow to Ͼ50% confluence by overnight incubation at 37°C in a humidified 5% CO 2 atmosphere and washed with serum-free and antibiotic-free DMEM. COS-7 cells were transfected with eukaryotic expression vectors for Lyn kinase (pSV7c-lynA) or cdc2 kinase (pT7f1A-cdc2; generously provided by Dr. Giulio Draetta, Mitotix Inc., Cambridge, MA). Specifically, 4 g of pSV7c/lynA, 3 g of pT7f1A/cdc2, or a combination thereof was diluted in 0.6 ml of serum/ antibiotic-free DMEM, mixed with 15 l of Lipofectamine reagent (Life Technologies, Inc.) (47,48), and the mixture was incubated for 30 min at room temperature to allow binding of DNA to cationic liposomes. Subsequently, the DNA-liposome complexes were diluted by addition of 1.4 ml of DMEM to the mixture, and 2 ml of DNA-liposome complex was added directly to COS-7 cells. Cells were incubated for 6 h at 37°C, followed by addition of 2 ml of DMEM supplemented with 20% fetal calf serum. After an 18-h culture at 37°C in a 5% humidified CO 2 atmosphere, the transfection mixture was removed and replaced with freshly prepared DMEM plus 10% fetal calf serum. COS-7 cells were harvested 72 h after the start of transfection and cell lysates were prepared using 1% Nonidet P-40 lysis buffer for immune complex kinase assays as well as immunoblotting with anti-Cdc2-Cter or anti-Lyn antibodies, as described (36).
Lyn Kinase-deficient BCP Leukemia Cells-Leukemic cells from all children with newly diagnosed BCP leukemia entered on the Childrens Cancer Group (CCG) treatment protocols are being examined in the CCG ALL Biology Reference Laboratory in Minneapolis for their Src family PTK profile (supported by National Cancer Institute Grant U01-CA-60437). These treatment protocols were approved by the National Cancer Institute as well as by the institutional review boards of the CCG-affiliated institutions. Informed consent was obtained from parents, patients, or both, as deemed appropriate for both treatment and laboratory studies according to the Department of Health and Human Services guidelines. Mononuclear cell fractions containing Ͼ90% leukemic cells were isolated from pretreatment bone marrow aspirate samples by centrifugation of the cell suspensions on Ficoll-Hypaque gradients. Leukemic BCP from 2 of 455 patients studied between 12/93 and 7/94 (designated as unique patient number (UPN) 1 lynϪ and UPN2 lynϪ ) were found to be Lyn kinase-deficient. These cells were used in the current study to examine the role of Lyn kinase in radiationinduced tyrosine phosphorylation of p34 cdc2 and cell cycle arrest.
Analysis of Radiation-induced Mitotic Arrest Using DNA Flow Cytometry-Cells were irradiated and then cultured at 5 ϫ 10 5 cells/ml in clonogenic medium (RPMI 1640 medium ϩ 1% penicillin/streptomycin ϩ 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, and 10 mM Hepes buffer) for up to 28 h at 37°C, 5% CO 2 . At the indicated time points, cells were washed two times in fresh clonogenic medium and stained with the UV-excited dye, Hoechst 33342, to quantify their DNA content as described previously (7,25). Quantitative DNA analysis was performed on a FACStar Plus flow cytometer equipped with a Consort 40 computer using the COTFIT program, which includes CELLCY, a cell cycle distribution function that fits DNA content histograms and calculates the percentages of cells in G 0/1 , S, and G 2 M phases of the cell cycle, as described (7,25).
Preparation of Cytoplasmic, Membrane, and Nuclear Protein Fractions of BCP Leukemia Cells-Enucleated cytoplasmic fractions and plasma membranes were prepared by nitrogen cavitation and differential centrifugation on Percoll (Pharmacia Biotech Inc.) gradients, as described previously (38,42). No nuclei or nucleated cells were seen on the cytospin preparations of the cytoplasmic or membrane fractions, and no DNA was detectable by PCR amplification of a 110-base pair fragment from the first exon of the human ␤-globin gene (37). Nuclear proteins were extracted according to previously reported procedures (41). In brief, cells were lysed by vortexing at 4°C for 10 min in 10 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM DTT, 0.1% Nonidet P-40. Nuclei were collected by centrifugation in a microcentrifuge at maximum speed for 5 min. Nuclei were then suspended in 20 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 0.5 mM DTT, 0.42 M NaCl, 0.2 mM EDTA, 25% (v/v) glycerol, 0.5 mM phenylmethylsulfonyl fluoride and incubated at 4°C for 15 min to allow the leakage of solubilized nuclear proteins. Higher salt concentrations were avoided to prevent release of DNA and histones. The mixture was briefly vortexed and centrifuged for 10 min at maximum speed in the microcentrifuge, and the supernatants were used for immunoprecipitations. NALM-6 cells express very high levels of the B-lineage-restricted CD19 antigen on their membrane and in their cytoplasm (36). Nuclear fractions of NALM-6 cells were free of CD19 antigen, as determined by Western blot analysis with a polyclonal anti-CD19 antibody raised against a GST-CD19 fusion protein corresponding to the cytoplasmic portion of CD19 (i.e. amino acids 410 -540).

RESULTS AND DISCUSSION
Lyn Kinase Associates Physically and Functionally with p34 cdc2 Kinase in the Cytoplasm of Human BCP Leukemia Cells-We investigated if Lyn kinase is capable of a physical association with p34 cdc2 kinase in human BCP leukemia cells by first examining the in vitro kinase reaction products of p34 cdc2 and Lyn immune complexes from the Nonidet P-40 lysates of unirradiated NALM-6 cells. Kinase reactions were performed in the presence of [␥-32 P]ATP to allow autophosphorylation of the 53-and 56-kDa Lyn isoforms (i.e. p53 lyn and p56 lyn ) that differ in sequences of their "unique" region. As shown in Fig. 1A, autophosphorylated Lyn kinase isoforms were detected not only in the Lyn immunoprecipitates that were used as a positive control but also in the p34 cdc2 immunoprecipitates. The degree of p34 cdc2 phosphorylation in unirradiated NALM-6 cells was very low in these kinase assays, which favor Lyn autophosphorylation. In order to better document the presence of p34 cdc2 in Lyn immune complexes, we subjected the Lyn immunoprecipitates to Western blot analysis with an anti-Cdc2-Cter antibody. The anti-Cdc2 antibody used in these experiments immunoprecipitates native enzyme better than it blots denatured enzyme. Therefore, in an attempt to increase the sensitivity of detection for p34 cdc2 , we used 5 times more cell lysate to prepare the Lyn immunoprecipitate that we did to obtain the p34 cdc2 immunoprecipitate that was used as a positive control. As shown in Fig. 1B, p34 cdc2 was detected not only in the p34 cdc2 immunoprecipitate but also in the Lyn immunoprecipitate. Similarly, Western blot analysis of the p34 cdc2 immunoprecipitate from the lysates of unirradiated NALM-6 cells with an anti-Lyn antibody raised against a GST-Lyn fusion protein corresponding to the 56-kDa isoform of Lyn confirmed the presence of Lyn kinase in p34 cdc2 immune complexes (Fig. 1C). Taken together, these results demonstrated that Lyn kinase is capable of association with p34 cdc2 in BCP leukemia cells and this association does not require exposure of cells to ionizing radiation. We next sought to determine the intracellular site of interactions between Lyn kinase and p34 cdc2 . To this end, we examined Lyn immune complexes from various fractions of Nonidet P-40 lysates of unirradiated NALM-6 cells in kinase assays for autophosphorylation of Lyn kinase (upper panel) and in Western blots for presence of Lyn protein (middle panel) as well as for presence of p34 cdc2 protein (lower panel). As expected, Lyn kinase activity was detected in whole cell lysates as well as the cytoplasmic and membrane fractions and the presence of Lyn protein in these immunoprecipitates was formally confirmed by anti-Lyn Western blot analysis (Fig. 1D). Lyn kinase was also detected in the nuclear fractions (Fig. 1D), a finding that was confirmed by immunofluorescent staining techniques. 2 Anti-Cdc2-Cter Western blot analysis of Lyn immune complexes revealed no significant association between Lyn and p34 cdc2 in the membrane or nuclear fractions. The detection of p34 cdc2 in the Lyn immune complexes from the cytoplasmic fraction as shown in Fig. 1D suggested the cytoplasm as the primary site of association between Lyn and p34 cdc2 .
Partial Mapping of the Sites of Interaction between Lyn and Cdc2 Kinases in BCP Leukemia Cells-Src family PTK are composed of an unique amino-terminal domain, a regulatory carboxyl-terminal domain, an SH3 domain, and an SH2 domain (49). SH3 domains, which bind to proline-rich sequences, as well as SH2 domains, which bind to phosphotyrosine, have been shown to facilitate protein-protein interactions and formation of intracellular signaling complexes (44,49,50). The amino-terminal 27 residues of the unique domain of Lyn have been shown to mediate the association of Lyn with phospholipase C␥2, mitogen-activated protein kinase, and GTPase-activating protein (44). We performed binding experiments with truncated GST-Lyn fusion proteins corresponding to various domains of Lyn kinase to generate preliminary information regarding the structural requirements for Lyn association with p34 cdc2 . Schematic diagrams and the inclusive amino acid se-quences of these truncated GST-Lyn fusion proteins are depicted in Fig. 2A. Purified GST-Lyn fusion proteins, which were non-covalently immobilized on glutathione-agarose beads, were incubated with Nonidet P-40 lysates of unirradiated NALM-6 cells, and the adsorbates were analyzed for the presence of p34 cdc2 kinase by immunoblotting with an anti-Cdc2-Cter antibody. As shown in Fig. 2B, p34 cdc2 in NALM-6 lysates was able to bind to GST-Lyn fusion protein Lyn 1-119 containing the unique amino-terminal domain plus the SH3 domain, but it did not bind to GST-Lyn fusion proteins corresponding to the SH2 domain (i.e. Lyn 131-243), amino-terminal 27 residues (i.e. Lyn 1-27), amino-terminal 61 residues (i.e. Lyn 1-61), or amino-terminal domain plus proximal portion of the SH3 domain (i.e, Lyn 1-92). The results of these experiments suggest a functional role for the SH3 rather than the SH2 domain of Lyn in Lyn-p34 cdc2 interactions in leukemic BCP. Notably, GST-Lyn fusion protein Lyn 27-131 did not exhibit any binding activity to p34 cdc2 . Thus, the first 27 residues of the unique amino-terminal domain of Lyn, while not sufficient for the Lyn-p34 cdc2 interaction, appear to be essential for the ability of Lyn to bind to p34 cdc2 from NALM-6 cell lysates. Further studies will be required to elucidate the exact structural basis for the Lyn-p34 cdc2 interactions.
Ionizing Radiation Promotes the Physical and Functional Interactions between Lyn Kinase and p34 cdc2 in BCP Leukemia  FIG. 1. Lyn kinase associates with p34 cdc2 in unirradiated BCP leukemia cells. A, unirradiated NALM-6 cells were lysed in Nonidet P-40 lysis buffer and 200-g samples of the cell lysate were immunoprecipitated with a polyclonal rabbit anti-Lyn antibody or anti-Cdc2-Cter antibody. Immune complexes were assayed for kinase activity during a 10-min incubation in the presence of 0.1 mM [␥-32 P]ATP to allow autophosphorylation of the 53-and 56-kDa Lyn isoforms. Samples were boiled in 2 ϫ SDS sample buffer and fractionated on 12.5% polyacrylamide gels. B, unirradiated NALM-6 cells were lysed as in A and 100 g of the lysate was immunoprecipitated with anti-Cdc2-Cter antibody, whereas 500 g of the lysate was immunoprecipitated with anti-Lyn antibody, as described in A. The immune complexes were collected, boiled in 2 ϫ SDS sample buffer, fractionated on 15% polyacrylamide gels, transferred to an Immobilon-PVDF membrane, and immunoblotted for 90 min with anti-Cdc2-Cter antibody. 125 I-Labeled protein A was used to detect p34 cdc2 . C, unirradiated NALM-6 cells were lysed in Nonidet P-40 lysis buffer and 200-g samples of the cell lysate were immunoprecipitated with anti-Cdc2-Cter antibody; immune complexes were collected, washed, boiled in 2 ϫ SDS sample buffer, fractionated on 15% polyacrylamide gels, transferred to an Immobilon-PVDF membrane, and immunoblotted with anti-Cdc2-Cter antibody (lanes 1 and 2) or with an anti-Lyn antibody raised against a GST-Lyn fusion protein corresponding to the 56-kDa isoform of Lyn (lanes 3 and  4). 125  Ϫ incorporated/min, and it was amplified by 30% within 30 s following exposure to 2 Gy ␥-rays (data not shown). The specificity of this reaction was confirmed using a mutated Cdc2 peptide [Phe 15 ,Lys 19 ]Cdc2(6 -20)NH 2 ; Tyr 15 3 Phe) as a negative control. The kinase activity of Lyn immunoprecipitate toward this single amino acid substitution analog of the Cdc2 peptide, which does not contain a target Tyr 15 residue, was only 0.005 pmol of PO 4 Ϫ incorporated/min. We next examined the effects of ionizing radiation on the ability of Lyn kinase to phosphorylate a recombinant human p34 cdc2 -cyclin B complex preparation in the presence of [␥-32 P]ATP.This complex was isolated from lysates of insect cells coinfected with recombinant viruses encoding GST-cyclin B and [Arg 33 ]p34 cdc2 , an inactive mutant form of p34 cdc2 mutated at lysine 33 (25,43). Lyn kinase was immunoprecipitated from unirradiated as well as irradiated NALM-6 cells and examined in kinase assays for autophosphorylation as well as its ability to phosphorylate recombinant human p34 cdc2 on tyrosine. As shown in Fig. 3A, ionizing radiation resulted in a Ͼ4-fold increase in Lyn kinase activity, as measured by autophosphorylation. The increased autophosphorylation was accompanied by Ͼ1.8-fold increased phosphorylation of [Arg 33 ]p34 cdc2 . Two-dimensional phosphoamino acid analysis of the excised Cdc2 bands confirmed that the increased label on p34 cdc2 reacted with Lyn from irradiated cells was caused by enhanced tyrosine phosphorylation (data not shown). Thus, exposure of NALM-6 cells to ␥-rays prior to the immunoprecipitation augmented the ability of Lyn kinase to utilize recom-binant human p34 cdc2 as an exogenous substrate during the in vitro kinase reactions.
Subsequently, we evaluated the effects of ionizing radiation on the intracellular physical and functional interactions between Lyn and p34 cdc2 in NALM-6 cells. To this end, NALM-6 cells were irradiated, lysed with Nonidet P-40 lysis buffer, and Lyn kinase was immunoprecipitated from the lysates of unirradiated as well as irradiated cells. In vitro kinase assays were performed to examine Lyn autophosphorylation as well as phosphorylation of any co-immunoprecipitated p34 cdc2 kinase. As shown in Fig. 3B (b1), irradiation of NALM-6 cells stimulated the the Lyn kinase, as measured by autophosphorylation. Concomitant with Lyn kinase activation at 10 or 20 min following radiation exposure, p34 cdc2 became detectable in the Lyn immune complexes as a tyrosine-phosphorylated protein substrate (Fig. 3B, b1). The abundance of the Lyn protein, as estimated by anti-Lyn Western blot analysis, did not change GST-Lyn fusion proteins non-covalently bound to glutathioneagarose beads were used in binding assays to examine their ability to interact with p34 cdc2 in NALM-6 cells, as described under "Experimental Procedures." Samples (250 g) of the Nonidet P-40 lysates from unirradiated NALM-6 cells were incubated with the GST-Lyn fusion protein-coupled beads. The fusion protein adsorbates were washed, resuspended in SDS sample buffer, boiled, fractionated on 12.5% SDS-PAGE gels, transferred to Immobilon-P membranes, and membranes were immunoblotted with anti-Cdc2-Cter, followed by visualization using 125 I-labeled protein A and autoradiography.

FIG. 3. Ionizing radiation promotes the interaction between Lyn and Cdc2 kinases in BCP leukemia cells.
A, ␥-rays stimulate the PTK activity of Lyn toward recombinant human [Arg 33 ]p34 cdc2 . Lyn kinase was immunoprecipitated from Nonidet P-40 lysates of unirradiated (lane 1) and irradiated (lane 2, 5 min after 1 Gy ␥-rays; lane 3, 5 min after 2 Gy ␥-rays) NALM-6 cells. In vitro kinase assays were performed to examine the immunoprecipitated Lyn kinase for autophosphorylation as well as its ability to phosphorylate recombinant human p34 cdc2 -cyclin B complex, which was used as an exogenous substrate, on tyrosine. B, ␥-rays promote the physical and functional interactions between Lyn and p34 cdc2 in BCP leukemia cells. b1, Lyn kinase was immunoprecipitated from the Nonidet P-40 lysates (600 g/sample) of unirradiated (lane 2) as well as irradiated (lane 3, 10 min after 2 Gy ␥-rays; lane 4, 20 min after 2 Gy ␥-rays) NALM-6 cells and in vitro kinase assays were performed using one-third of the samples, as described in the legend of Fig. 1A, to examine Lyn autophosphorylation as well as phosphorylation of co-immunoprecipitated p34 cdc2 kinase. Arrows indicate the positions of the Lyn and p34 cdc2 kinases. b2, another third of the samples from the Lyn immunoprecipitations shown in b1 were boiled in 2 ϫ SDS sample buffer, fractionated on 12.5% polyacrylamide gels, transferred to an Immobilon-PVDF membrane, and immunoblotted with an anti-Lyn antibody raised against a GST-Lyn fusion protein corresponding to the 56-kDa isoform of Lyn. 125 I-Labeled protein A was used to detect the 56 kDa isoform of Lyn. b3, the remaining one-third of the samples from the Lyn immunoprecipitations shown in b1 were boiled in 2 ϫ SDS sample buffer, fractionated on 12.5% polyacrylamide gels, transferred to an Immobilon-PVDF membrane, and immunoblotted with anti-Cdc2-Cter antibody. 125 I-Labeled protein A was used to detect the p34 cdc2 kinase in the Lyn immune complexes. The purpose of the b2 portion of the experiment was to confirm that lanes 2, 3, and 4 contained equal amounts of Lyn and the lane-lane differences in Lyn autophosphorylation or amount of Cdc2 kinase detected by immunoblotting were not caused by loading unequal amounts of Lyn immune complexes in each lane. In b1-b3, no primary immunoprecipitating antibody was added to the control samples shown in lanes 1. during the course of the experiment, suggesting increased enzymatic activity of Lyn (Fig. 3B, b2). However, the abundance of the p34 cdc2 protein in the same Lyn immune complexes, as determined by anti-Cdc2-Cter Western blot analysis, was significantly increased after radiation exposure (Fig. 3, B, b3), suggesting that enhanced tyrosine phosphorylation of p34 cdc2 , which parallels the Lyn activation, is at least in part due to radiation-induced promotion of the physical association between Lyn and p34 cdc2 in NALM-6 cells.
Recombinant GST-Lyn Fusion Protein and Lyn Kinase Immunoprecipitated from Irradiated BCP Leukemia Cells Phosphorylate Recombinant Human p34 cdc2 on Tyrosine 15-For further analysis of the interactions between Lyn and p34 cdc2 kinases, we prepared a highly purified 83-kDa GST-Lyn fusion protein, as described under "Experimental Procedures." This GST-Lyn fusion protein was enzymatically active, as confirmed by its autophosphorylation and its ability to phosphorylate denatured rabbit enolase, which was used as an exogenous substrate, during a 10-min in vitro kinase reaction (Fig. 4A).
We next performed in vitro kinase assays using GST-Lyn in order to determine whether [Arg 33 ]p34 cdc2 can serve as a direct substrate for Lyn in the absence of other proteins or kinases which are associated with Lyn kinase (27)(28)(29). GST-Lyn effectively phoshorylated [Arg 33 ]p34 cdc2 (Fig. 4B), and two-dimensional phosphoamino acid analysis confirmed that the increased phosphorylation of GST-Lyn-treated p34 cdc2 kinase was on tyrosine (Fig. 4C). GST-p49 WEE1Hu , a positive control fusion protein of human WEE1 kinase with GST, which was reported to phosphorylate [Arg 33 ]p34 cdc2 on Tyr 15 (25,43), increased the p34 cdc2 -associated label Ͼ10-fold (Fig. 4D), and phosphoamino acid analysis confirmed that the increased phosphorylation was on tyrosine (data not shown). In some experiments, [Arg 33 ]p34 cdc2 was phosphorylated even in the absence of GST-Lyn (Fig. 4D). The threonine phosphorylation of untreated p34 cdc2 seen in two-dimensional phosphoamino acid analyses, which is depicted in Fig. 4E, is caused by a kinase that sometimes coprecipitates from the insect cell and phosphorylates p34 cdc2 on Thr-161 (25,30). Similarly, phosphorylation of cyclin B in this substrate preparation is due to an endogenous insect cell kinase that binds to cyclin B and copurifies with the cyclin B-p34 cdc2 complex, as kinase activity is associated with cyclin B when it is expressed in insect cells in the absence of p34 cdc2 as well (25,30). When observed, this base-line threonine phosphorylation of p34 cdc2 in kinase reactions partially masked the magnitude of GST-Lyn-induced phosphorylation of p34 cdc2 (Fig. 4D). However, two-dimensional phosphoamino acid analysis of GST-Lyn-phosphorylated [Arg 33 ]p34 cdc2 confirmed that increased phosphorylation of GST-Lyn-treated p34 cdc2 was on tyrosine (Fig. 4E), thereby unmasking and validating the potent PTK activity of GST-Lyn fusion protein toward recombinant human [Arg 33 ]p34 cdc2 .
To further evaluate the effects of GST-Lyn as well as Lyn kinase immunoprecipitated from unirradiated and irradiated NALM-6 pre-B leukemia cells on the phosphorylation state of  Fig. 4D were subjected to two-dimensional tryptic phosphopeptide mapping, as described under "Experimental Procedures." The position of Tyr 15 -containing peptide was identified as the site of GST-WEE1-induced phosphorylation of [Arg 33 ]p34 cdc2 (30,43). Bottom panel, [Arg 33 ]p34 cdc2 was also used as a substrate for Lyn kinase, which was immunoprecipitated from Nonidet P-40 lysates of unirradiated (N6, 0 Gy) and irradiated (N6, 1 Gy ϭ 5 min after 1 Gy ␥-rays; N6, 2 Gy ϭ 5 min after 2 Gy ␥-rays) NALM-6 cells. Two-dimensional tryptic phosphopeptide mapping was performed as for the samples shown in the top panel. In both the top and bottom panels, the position of this Tyr 15 -containing peptide in each phosphotryptic map shown is indicated with an arrowhead.
[Arg 33 ]p34 cdc2 , we subjected p34 cdc2 excised from the gels of kinase reactions to two-dimensional tryptic phosphopeptide mapping. As shown in Fig. 5, a single threonine-containing phosphopeptide was detected upon phosphotryptic mapping of untreated p34 cdc2 . Consistent with a previous report, which identified Tyr 15 as the site of GST-WEE1-induced phosphorylation of [Arg 33 ]p34 cdc2 (43), one major tyrosine-containing phosphopeptide was detected after treatment of [Arg 33 ]p34 cdc2 with GST-WEE1. The position of this Tyr 15 -containing peptide in each phosphotryptic map shown in Fig. 5 is indicated with an arrowhead. Notably, treatment of [Arg 33 ]p34 cdc2 with GST-Lyn or Lyn immunoprecipitated from irradiated NALM-6 cells resulted in increased phosphorylation of the same Tyr 15 -containing peptide.
Taken together, these experiments provided direct evidence that Lyn kinase can directly phosphorylate p34 cdc2 on Tyr 15 . The radiation-enhanced ability of Lyn kinase from NALM-6 cells to phosphorylate recombinant p34 cdc2 on Tyr 15 strongly supports the hypothesis that Lyn may be one of the PTK responsible for radiation-induced inhibitory tyrosine phosphorylation and inactivation of p34 cdc2 kinase in human BCP leukemia cells.
In Vivo Tyrosine Phosphorylation of p34 cdc2 by Co-expression with Lyn Kinase in COS-7 Cells-To further study the inter-action of Lyn and p34 cdc2 in vivo, cDNAs encoding these kinases were transiently co-expressed in COS-7 cells using the Lipofectamine reagent (47,48). Compared to COS-7 cells transfected with cdc2 cDNA and mock-transfected with the empty expression vector, COS-7 cells co-transfected with cDNAs for both lyn and cdc2 showed markedly amplified expression of Lyn protein, as determined by anti-Lyn Western blot analysis of Nonidet P-40 lysates (Fig. 6, upper panel). Amplified lyn expression did not affect cdc2 expression in co-transfected COS-7 cells, as determined by anti-p34 cdc2 -Cter Western blot analysis of Nonidet P-40 lysates (Fig. 6, upper panel). Anti-Cdc2-Cter Western blot analysis of Lyn immune complexes from COS-7 cells co-transfected with cdc2 cDNA demonstrated the presence of p34 cdc2 kinase (Fig. 6, lower left panel). Kinase assays of Lyn immune complexes from lysates of COS-7 cells co-transfected with cDNAs for both lyn and cdc2 demonstrated the presence of phosphorylated p34 cdc2 (Fig. 6, lower right were used for immunoprecipitation and immune complex kinase assays of the indicated Src family PTK. C, p34 cdc2 was immunoprecipitated from Nonidet P-40 lysates of unirradiated as well as irradiated (2 Gy delivered 5 min prior to lysis) BCP leukemia cells of UPN1 lynϪ and UPN2 lynϪ using a rabbit anti-Cdc2-Cter antibody. Samples were run on 10.5% SDS-PAGE gels and subsequently immunoblotted with either anti-phosphotyrosine or anti-Cdc2-Cter. 125 I-Labeled protein A was used to detect tyrosine-phosphorylated proteins or p34 cdc2 kinase. The position of p34 cdc2 is indicated with arrowheads. D, for comparison, using the procedures outlined in C, radiation-induced tyrosine phosphorylation of p34 cdc2 was also examined in Lyn kinase-positive NALM-6 pre-B leukemia cells. panel), and phosphoamino acid analyses confirmed that the p34 cdc2 -associated label was on tyrosine (data not shown). Thus, Lyn kinase associates physically with p34 cdc2 kinase when these kinases are co-expressed in COS-7 cells and elevated Lyn kinase activity is sufficient for induction of p34 cdc2 tyrosine phosphorylation in vivo. These results confirmed the ability of Lyn kinase to interact with p34 cdc2 in vivo.
Failure of Ionizing Radiation to Induce Tyrosine Phosphorylation and Inactivation of p34 cdc2 in Lyn Kinase-deficient Human BCP Leukemia Cells-Lyn kinase has been consistently identified as the predominant member of the Src family PTK family in leukemic cells from BCP leukemia patients (36,37). In a survey of 455 BCP leukemia cases, we were able to identify only two patients, UPN1 lynϪ (Fig. 7A) and UPN2 lynϪ , whose leukemic cells did not contain any Lyn enzyme detectable by immune complex kinase assays or by Western blot analysis of Lyn protein expression. We used immune complex kinase assays to examine the relative abundance of other members of the Src PTK family in UPN1 lynϪ cells. As shown in Fig. 7B, Fyn and Blk were the predominant Src family PTK in these Lyn kinase-deficient cells.
We next compared the ability of ionizing radiation to trigger tyrosine phosphorylation of p34 cdc2 in Lyn kinase expressing NALM-6 cells versus Lyn kinase-deficient UPN1 lynϪ or UPN2 lynϪ cells by anti-phosphotyrosine Western blot analysis of p34 cdc2 immunoprecipitates from the Nonidet P-40 cell lysates prepared 5 min after radiation exposure. Ionizing radiation induced tyrosine phosphorylation of p34 cdc2 in NALM-6 cells, but not in UPN1 lynϪ or in UPN2 lynϪ cells (Fig. 7, C and D).
We next compared the ability of ␥-rays to cause a G 2 arrest in Lyn kinase expressing NALM-6 cells versus Lyn kinase-deficient UPN1 lynϪ cells. Asynchronously dividing NALM-6 cells and uncultured UPN1 lynϪ cells were irradiated with 2 Gy ␥-rays and then cultured at 5 ϫ 10 5 cells/ml in a clonogenic medium, as described under "Experimental Procedures." At the indicated time points, cells were stained with Hoechst 33342 to quantify their DNA content on a FACStar Plus flow cytometer. Prior to radiation, 25% of NALM-6 cells and 29% of UPN1 lynϪ cells were in the G 2 phase of the cell cycle, which corresponds to the second peak of the DNA histogram (Fig. 8). In NALM-6 cells, a radiation-induced accumulation in G 2 phase was first detectable at 8 h after radiation, when the DNA flow cytometric analysis showed 38% of the cells to be in the G 2 phase. The cell cycle arrest at the G 2 -M transition checkpoint was further evident from the decreased percentage of G 0/1 cells. The percentage of cells accumulated in G 2 phase was further increased to 54% at 22 h. This cell cycle arrest at G 2 -M transition checkpoint was transient, as evidenced by the decreased percentage of G 2 cells and increased percentage of G 0/1 cells at 28 h after radiation. In contrast to NALM-6 cells, UPN1 lynϪ cells did not show any evidence of a cell cycle arrest at G 2 -M transition after exposure to 2 Gy ␥-rays (Fig. 8). These findings support the hypothesis that Lyn is the PTK responsible for radiationinduced inhibitory tyrosine phosphorylation and inactivation of p34 cdc2 kinase in human BCP leukemia cells.
Radiation-induced G 2 arrest allows the cells to repair potentially lethal or sublethal DNA lesions induced by radiation or other DNA damaging agents. Cells that are unable to show this response are more sensitive to DNA damaging agents, and drugs that abolish this response sensitize cells to DNA damaging agents (11,(15)(16)(17)(18)(19)(20)(21)(22). Abrogation of radiation-induced G 2 arrest by caffeine exposure induces premature mitosis before DNA repair is complete and results in enhanced cell death (8). Similarly, pentoxyfylline, a caffeine analog that shortens the duration of G 2 arrest, also displays radiosensitizing properties (40). A human lymphoma cell line that displayed markedly enhanced sensitivity to DNA damage by nitrogen mustard was found to be defective in the G 2 phase checkpoint control (9).
We provide experimental evidence for an important role of a cytoplasmic signal transduction pathway intimately linked to the Lyn kinase in radiation-induced G 2 phase-specific cell cycle arrest of human BCP leukemia cells. Because Lyn kinase maintains p34 cdc2 in an inactive state in irradiated BCP leukemia cells, thereby allowing them to repair sublethal radiation damage, we postulate that inhibition of Lyn kinase in BCP leukemia cells may result in radiosensitization. To accomplish this goal, a PTK inhibitor could be targeted to Lyn kinase in BCP leukemia cells with a monoclonal antibody, which binds to and remains complexed with CD19 receptor. CD19 receptor is physically associated with the Lyn kinase (36). Our recent results show that treatment of CD19 ϩ BCP leukemia cells with nanomolar concentrations of B43-Gen immunoconjugate causes sustained inhibition of CD19-associated Lyn kinase (37).
Role of Lyn Kinase in Surveillance and Repair of DNA Damage in Human B-lineage Lymphoid Cells-Our results implicate Lyn as an important cytoplasmic suppressor of p34 cdc2 function and extend previous observations that Lyn kinase may play an important role in anti-IgM or anti-CD19-induced G 1 arrest of B lymphoma cells (51). Recent studies indicate that p34 cdc2 is activated in the cytoplasm and premature activation of p34 cdc2 at an inappropriate time during the cell cycle leads to apoptotic cell death, underscoring the importance of regulatory events governing p34 cdc2 activation and deactivation (52). Lyn FIG. 8. Ionizing radiation does not cause G 2 arrest in Lyn kinase-deficient BCP leukemia cells. Lyn kinasepositive NALM-6 pre-B leukemia and Lyn kinase-deficient UPN1 lynϪ cells were irradiated with 2 Gy ␥-rays and then cultured in clonogenic medium for 8, 22, and 28 h at 37°C/5% CO 2 . Cells were washed two times in fresh clonogenic medium and stained with the UV-excited dye, Hoechst 33342, as described previously (5,25). Quantitative DNA analysis was performed on a FACStar Plus flow cytometer equipped with a Consort 40 computer using the COTFIT program, which includes CELLCY, a cell cycle distribution function that fits DNA content histograms and calculates the percentages of cells in G 0/1 , S, and G 2 M phases of the cell cycle.
kinase may protect the cell from the potentially catastrophic consequences of premature p34 cdc2 activation by maintaining the p34 cdc2 -cyclin B complex in its inactive, tyrosine phoshorylated state.
Several studies have documented the ability of B-lineage lymphoid cells to produce reactive oxygen intermediates in response to various activation signals (53)(54)(55)(56)(57)(58). Recent evidence suggests that production of reactive oxygen intermediates in response to various mitogenic stimuli may regulate the proliferative responses of peripheral blood mononuclear cells (59). It has been proposed that generation of reactive oxygen intermediates upon activation of B-lineage lymphoid cells may contribute to somatic mutations (53,(55)(56)(57)60). Lyn kinase may serve as an integral component of a physiologically important surveillance and repair mechanism for DNA damage by delaying the G 2 -M transition in cells exposed to mutagenic oxygen free radicals, thereby allowing them to repair their DNA damage prior to mitosis. Without this surveillance, the likelihood of malignant transformations leading to BCP leukemias as well as impaired survival and self-renewal capacity of BCP populations leading to immunodeficiency disorders may be increased. Therefore, it will be important to conduct appropriate epidemiologic studies designed to test the hypothesis that low activity levels of Lyn in BCP populations may be associated with increased risk of development of BCP leukemia or B-cell immunodeficiency during childhood.