Inhibition Induces Regression of Human Tumor ...

Lyn Is a Target Gene for Prostate Cancer : Sequence-Based Inhibition Induces Regression of Human Tumor Xenografts Mirela Goldenberg-Furmanov, Ilan Stein, Eli Pikarsky, et al. Cancer Res 2004;64:1058-1066.

Updated version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/64/3/1058

Cited Articles

This article cites by 35 articles, 16 of which you can access for free at: http://cancerres.aacrjournals.org/content/64/3/1058.full.html#ref-list-1

Citing articles

This article has been cited by 24 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/64/3/1058.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on June 13, 2013. © 2004 American Association for Cancer Research.

[CANCER RESEARCH 64, 1058 –1066, February 1, 2004]

Lyn Is a Target Gene for Prostate Cancer: Sequence-Based Inhibition Induces Regression of Human Tumor Xenografts Mirela Goldenberg-Furmanov,1,4 Ilan Stein,1,4 Eli Pikarsky,2 Hila Rubin,4 Shafika Kasem,2 Marc Wygoda,3 Irina Weinstein,1,4 Hadas Reuveni,4 and Shmuel A. Ben-Sasson1 Departments of 1Experimental Medicine and Cancer Research, 2Pathology, and 3Oncology, The Hebrew University-Hadassah Medical School, and 4Keryx Biopharmaceuticals, Jerusalem, Israel

ABSTRACT The Src-related protein kinase Lyn plays an important role in B-cell activation. However, several lines of evidence suggest that it is also involved in the control of cellular proliferation and the inhibition of apoptosis. We have discovered that Lyn is expressed in normal prostate epithelia, in 95% of primary human prostate cancer (PC) specimens examined, and in all of the PC cell lines that we assayed. Moreover, Lyn knockout mice display abnormal prostate gland morphogenesis, which suggests that Lyn plays an important role in prostate epithelium development and implies that Lyn is a candidate target for specific therapy for PC. Using a drug-design strategy to construct sequence-based peptide inhibitors, a Lyn-specific inhibitor, KRX-123, targeting a unique interaction site within Lyn, was synthesized. KRX-123 was found to inhibit cellular proliferation in three hormone-refractory PC cell lines, DU145, PC3, and TSU-Pr1 with IC50 values of 2– 4 ␮M. In vivo, tumor volume of DU145 explants in nude mice was significantly reduced after once-a-week injections of KRX-123, at a dose of 10 mg/kg, for a period of 5 weeks. Histological analyses of the treated tumors indicated extensive apoptosis. Thus, we suggest that Lyn inhibition may serve as a prime target for the treatment of hormone-refractory PC.

INTRODUCTION Prostate cancer (PC) is the most common malignancy detected in men. Although hormone therapy may be effective in the early stages of PC, virtually all metastatic PCs become refractory to hormonal treatment (1). The emergence of hormone resistance eventually results in the patient’s death because, currently, there is no effective treatment for hormone-refractory PC (HRPC). Protein kinases (PKs) control major events related to cell proliferation, and their deregulated activity can lead to cellular transformation and is associated with tumor maintenance and progression (2). Therefore, the identification of key regulatory PKs involved in PC could be an initial step in developing new therapeutic options in which the activity of the identified kinases is modulated by novel designer agents. Site-specific phosphorylation by PKs acts as a chemical relay that turns intracellular protein targets on or off. PK cascades facilitate a variety of signal-transduction pathways. Members of the Src-related PK family collectively control a variety of cellular activities in a broad spectrum of tissues. These nonreceptor protein tyrosine kinases (PTKs) mediate signaling between membrane-anchored growth-factor receptors and intracellular cascades or between integrins and cytoskeletal elements (3). Lyn, a member of this family, was originally discovered in the context of

hematopoietic cells, in which it plays a role in the regulation of B-cell immune responses (4). Lyn-deficient mice display disrupted B-cell function, leading to autoimmunity and defective mast cell degranulation (5, 6). On the other hand, several recent studies have suggested that Lyn is a negative regulator of apoptosis in various cell systems (7, 8). In leukemic cells, Lyn is constitutively activated, and the inhibition of Lyn expression, using Lyn antisense oligonucleotides, reversed proliferation (9). In addition, it was shown that Lyn is expressed in colon and PC cells, and that overexpression of a dominant active Lyn in colon cancer cell lines induced chemoresistance via a phosphoinositide 3-kinase/protein kinase B-mediated cell survival pathway (7). An early study showed that there is an inverse relationship between transforming growth factor ␤ levels and Lyn expression in the PC cell line PC3 (10). Recently, Sumitomo et al. (11) demonstrated a role for Lyn in mediating the neutral endopeptidase inhibition of focal adhesion kinase in PC cell lines. Because of the indication of an involvement of Lyn in cancer in general, and of PC in particular, we conducted a series of studies designed to characterize the role of Lyn in PC. We first demonstrated that Lyn is expressed in normal prostatic epithelia and that Lyn expression is extensively manifested in the vast majority of primary human PCs. Our examination of Lyn knockout mice revealed a compromised prostate gland development. To further facilitate these studies, we have developed a sequence-based peptide inhibitor of Lyn that targets a unique interaction site within Lyn. Our studies show that interception of Lyn signaling in HRPC cell lines by the Lyn-derived peptide inhibits cell proliferation. Moreover, Lyn appears to be an outstanding therapeutic target for HRPC because the administration of the Lyn-specific inhibitor to nude mice bearing HRPC tumor explants induces both cell apoptosis and tumor regression. MATERIALS AND METHODS Reagents

Synthetic Peptides. Peptides were synthesized by Fmoc [N-(9-fluorenyl)methoxycarbonyl] solid-phase peptide synthesis (Novotide Ltd. and Genemed Synthesis Inc.) and were purified by reverse-phase high-performance liquid chromatography. Identity and purity were confirmed by high-performance liquid chromatography–mass spectroscopy. For cell-based assays, peptides were formulated (with approved excipients only) as follows: 15 mg peptide were dissolved in 0.25 ml of 4% benzyl alcohol, 4% pluronic L44, and 2% benzyl benzoate in propylene glycol. To this, 0.125 ml of 2.2% glycine in double-distilled water and 0.125 ml of 50 mM sodium bicarbonate (pH 7.5) were added while vigorously stirring the tube. The preparation was heated to 100°C for 15 min and then was homogenized with a Polytron (Kinematica) for 2 min, during which 0.5 ml of 0.3 M lactose were gradually added. Heating and Received 8/5/03; revised 10/16/03; accepted 11/6/03. homogenizing were repeated, and the preparation was sterilized by heating to Grant support: This research was supported by a grant from Keryx Biopharmaceuticals, Israel. 100°C for 30 min. No change in the peptide high-performance liquid chromaThe costs of publication of this article were defrayed in part by the payment of page tography–mass spectroscopy profile was noted after this treatment. The resultcharges. This article must therefore be hereby marked advertisement in accordance with ing preparation, designated BBlac, is a very fine emulsion, suitable also for i.v. 18 U.S.C. Section 1734 solely to indicate this fact. injection. Note: M. Goldenberg-Furmanov, I. Stein, and E. Pikarsky contributed equally to this work. Antibodies. Anti-Lyn (clone 44), for immunoblotting, immunoprecipitaRequests for reprints: Shmuel A. Ben-Sasson, Department of Experimental Medicine tion, and immunohistochemistry, anti-Syk (clone N-19), anti-Fyn (clone and Cancer Research, The Hebrew University-Hadassah Medical School Ein-Karem, FYN3), anti-Lck (clone 2102), and anti-mitogen-activated protein kinase (antiP. O. Box 12272, Jerusalem, Israel. Phone: 972-2-675-8338; Fax: 972-2-641-4583; MAPK; C-14) were purchased from Santa Cruz Biotechnology; anti-CD19 E-mail: [email protected]. 1058


Lyn IS A TARGET GENE FOR PC

antibodies (clones MB19-1 and 1D3) were purchased from PharMingen; anti-phospho-tyrosine monoclonal antibody (mAb; clone 4G10), and anti-Lyn (catalogue no. 06-207), used for immunohistochemistry and immunoblotting, were purchased from Upstate Biotechnology; and anti-phospho-MAPK (clone MAPK-YT) was purchased from Sigma. Expression Constructs. To generate expression constructs for Syk and Lyn, we cloned reverse transcription-PCR fragments containing the entire open reading frames (GeneBank accessions L28824 and M16038, respectively) in pGEX-2T (Amersham), and pET-21b (Novagen), respectively. To introduce a lysine-to-leucine substitution at position 275 in the kinase domain of Lyn (kinase-dead Lyn, designated Lyn-K275L), in vitro mutagenesis was performed using GeneEditor in vitro site-directed mutagenesis system (Promega). All of the plasmid constructs were verified by sequencing. Syk was expressed in bacteria as a glutathione S-transferase fusion protein, which was isolated from lysates by chromatography on glutathione-agarose. Lyn-K275L was refolded from bacteria insoluble fraction by dialysis against decreasing pH gradient. Protein concentrations were calculated from the 280-nm UV absorbance. Lynⴚ/ⴚ Mice, Prostate Whole-Mount, and Histological Examination The Lyn⫺/⫺ mice were those generated by Chan et al. (6), using a targeting vector in which exons 3–7 of Lyn were replaced with a pol2s-neo cassette. Age-matched C57BL/6 wild-type mice were used as controls. Mice were kept in a specific pathogen-free barrier facility, and all procedures were approved by the Animal Care and Use Committee of the Hebrew University. For analysis of prostate morphology, mice were 10 weeks of age, whereas for proliferation assays, mice aged 3 and 7 weeks were examined. Mice were given injections of bromodeoxyuridine (Amersham; 100 ␮l/10g body weight) 1 and 24 h before sacrifice. Microdissection of prostate glands was performed essentially as described previously (12). Histological analysis of prostate glands was performed on en bloc specimens of male mouse urogenital systems marked with India ink to correctly identify the ventral, ventrolateral and dorsal prostate lobes. Prostates, fixed in 4% formaldehyde and embedded in paraffin, were sectioned at 4 ␮m and were either stained with H&E or immunohistochemically stained with anti-bromodeoxyuridine (NeoMarkers; clone BRD.3). Immunohistochemistry and Tissue Microarrays Initial sections were stained for H&E to verify histology. Sections of formalin-fixed paraffin-embedded specimens were stained with anti-Lyn antibody using a standard biotin-avidin complex immunohistochemistry (Zymed). Control staining for specificity was carried out in the presence of a Lynspecific blocking peptide (sc-15P; Santa Cruz Biotechnology). In all of the experiments, negative control, excluding primary antibody, was included. Tissue microarrays (including specimens of normal human prostate tissue) were purchased from Clinomics Inc. (PC tissue arrays PR200). Specimens of human PC metastasis to lymph node or bone were obtained from the archives of the Department of Pathology, The Hebrew University-Hadassah Medical Center. The use of human tissues was approved by the Institutional Review Board. Northern Blot Hybridization Total RNA was prepared with Tri-Reagent (Sigma). Human spleen total RNA and human prostate total RNA were purchased from Ambion. RNA samples were denatured with glyoxal and electrophoresed through a 1% agarose gel. RNAs were transferred onto a nylon-based membrane and were hybridized with a human Lyn cDNA probe, labeled with 32P by randomly primed DNA synthesis. For standardization, ribosomal RNAs were visualized by methylene blue staining before hybridization. Cell-Free Kinase Assays

tyrosine mAbs, followed by reprobing with anti-Syk or anti-Lyn antibodies as appropriate. Lyn Autophosphorylation. Each reaction contained KRX-123.302 dissolved in 1% DMSO, 720 ng/sample Lyn-K275L, and 13.5 ng/sample active Lyn in reaction buffer. The mixtures were treated as above but were maintained for 15 min after the addition of 100 ␮M ATP and were probed with anti-phospho-tyrosine mAbs, followed by reprobing with anti-Lyn antibodies. Lck and Fyn Autophosphorylation. Each reaction contained KRX-123.302 dissolved in 1% DMSO, 30 ng Lck or 40 ng Fyn (both from Upstate Biotechnology), in reaction buffer. The mixtures were agitated at 400 rpm for 20 min at 30°C, reactions were initiated by the addition of 100 ␮M ATP and were maintained agitated for 7 min (Lck) or 10 min (Fyn). Reactions were treated as described above, and probed with anti-phospho-tyrosine mAb, followed by reprobing with anti-Lck or anti-Fyn antibodies. B-Cell Activation, Immunoprecipitations, and Immunoblotting Splenic B cells were purified from 8 –12-week-old C57BL/6 mice, by removing CD90⫹ (Thy1.2⫹) cells with antibody-coated magnetic beads (Milteny Biotec). B-cell suspensions were ⬎95% B220⫹ as determined by flow cytometry analysis. Either splenic B cells or WEHI-231 cells were suspended (1 ⫻ 107/ml) in serum-free RPMI 1640, and were incubated with KRX-123.302 or with the appropriate control for 2 h. Cells were stimulated for 5 min with 40 ␮g/ml F(ab⬘)2 fragments of goat-antimouse IgM antibody (Jackson ImmunoResearch) and were subsequently lysed in lysis buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% NP40, 0.5% sodium deoxycholate, 2 mM Na3VO4, and protease inhibitors] at 4°C for 30 min, followed by centrifugation at 14,000 ⫻ g for 15 min. To follow MAPK phosphorylation, cells were treated as above but were stimulated with F(ab⬘)2 fragments of goat-antimouse IgM antibody for 20 and 40 s. Proteins of interest were immunoprecipitated from lysates by the addition of antibodies at optimized concentrations, followed by the addition of protein-G Sepharose beads. Each immunoprecipitation step was performed for 1 h at 4°C with rotation. Immune complex were washed three times in lysis buffer before SDS-PAGE analysis. Cell lysates, precipitated proteins, or cell-free kinase assays were subjected to SDS-PAGE with electrophoretic transfer to nitrocellulose membranes. Membranes were blocked in 3% BSA in Tris-buffered saline/Tween 20 [25 mM Tris (pH 7.4), 140 mM NaCl, and 0.2% Tween 20), and were incubated with primary antibody against the protein of interest with subsequent incubation with horseradish peroxidase-conjugated antirabbit or antimouse IgG antibodies (Jackson), or horseradish peroxidase-coupled antimouse IgA (Southern Biotechnology). When appropriate, blots were stripped, and reprobed with control antibodies. Immunoreactive proteins were visualized with enhanced chemiluminescence detection system (Pierce). NIH image software (version 1.26t) was used for quantification of band intensities. Cell Culture and Proliferation Assay DU145, PC3, TSU-Pr1, HT29, and WEHI-231 cells were maintained in RPMI 1640 containing 10% FCS, 2 mM glutamine, and 100 units/ml penicillin and streptomycin. Despite TSU-Pr1 cells, which were kindly provided by Dr. Shoshana Ravid, The Hebrew University, Jerusalem, Israel; all of the other cell lines were obtained from American Type Culture Collection. For cell proliferation assays, 4000 cells/well were plated in 96-well plates. Four h after plating, serial dilutions of peptides formulated in BBlac (see “Reagents” section), were added in hexaplicates, for an additional 72 h. Samples were fixed in 4% buffered formaldehyde, and the relative cell numbers were determined as described previously (13). Analysis of Intracellular Penetration

Lyn-Syk. Each reaction contained KRX-123.302 (see Fig. 4B) or KRXTo facilitate intracellular permeabilization of the peptides, their NH2 ter107.110 (myristylated peptide containing the sequence VTREVPFA), which minus was acylated with a large hydrophobic tail of myristic acid. To enable were dissolved in 1% DMSO, 98 ng/sample glutathione S-transferase-Syk, 27 visualization, we used a COOH terminus-biotin-tagged peptide, Myristyl-Gng/sample active histidine-tagged recombinant human Lyn (Panvera) in reac- LVTYkKIK@-NH2, and a control peptide with an identical sequence except tion buffer [50 mM Tris (pH 7.5), 10 mM MgCl2, 1 mM DTT, and 0.1 mM for K& in its COOH terminus (where k ⫽ D-Lys, K@ ⫽ Lys-⑀ amino-biotin, sodium orthovanadate]. The mixtures were agitated at 900 rpm for 20 min at and K& ⫽ Lys-⑀ amino-benzoyl). DU145 cells were grown in chamber slides 30°C, reactions were initiated by the addition of 100 ␮M ATP and were (30,000 cells/slide) and were incubated for 2 h with 10 ␮M biotin-conjugated maintained agitated for another 10 min. Reactions were stopped by the addition peptide or the control peptide. Slides were fixed with 4% formaldehyde and of 2⫻ sample buffer, separated on SDS-PAGE, and probed with anti-phospho- permeabilized by 0.2% Triton X-100; avidin-FITC (Calbiochem), diluted in 1059



0.5% BSA/PBS, was added for 30 min, followed by washing with 0.5% BSA/PBS. Slides were viewed using fluorescent confocal microscopy. Treatment of Human HRPC Xenografts in Nude Mice CD1 nude mice were inoculated with DU145 cells (4 ⫻ 106/mouse) s.c. Peptide treatment was initiated when tumors reached ⬃100 mm3. The animals were treated once a week by i.v. injection of 10 mg/kg peptide, formulated in BBlac, for a period of 5 weeks. Tumor volume was calculated according to the equation: volume ⫽ 0.52 ⴱ (width)2 ⴱ length. Vehicle and KRX-123.719treated groups contained five mice each, whereas the KRX-123.302-treated and nontreated groups contained six mice each. Terminal deoxynucleotidyl transferase-mediated nick end labeling (TUNEL) analysis was performed on sections obtained at the end of treatment, using the DeadEnd fluorometric TUNEL system (Promega). To quantify the proportion of TUNEL-positive cells, at least three arbitrarily chosen high-power fields from four tumors of control and peptide-treated groups were photomicrographed, and the ratio of green:blue nuclei in each tumor was calculated.

RESULTS Lyn Is Expressed in Normal and Malignant Prostatic Epithelium. As a first step in determining the role of Lyn in PC, the pattern of Lyn expression in normal prostate glands was investigated. In an earlier study, Allard et al. (14) have demonstrated the presence of Lyn in whole extracts of normal and hyperplastic dog prostate glands. To explore the relevance of their results to the prostatic epithelium, immunohistochemical studies using specific anti-Lyn antibody were performed on either normal human or mouse prostate glands. Two commercially available anti-Lyn antibodies, directed against different epitopes, were initially used, both showed specific staining of prostate epithelium. Staining was abolished by a competing Lyn-specific blocking peptide (Fig. 1, A and B), thus proving the specificity of our staining procedure. To verify it further, we performed immunoblotting analysis on protein extracts from various human PC cell lines with the anti-Lyn antibody. Only a doublet of Mr 53,000 and 56,000 was detected in all of them, consistent with a specific staining of the two isoforms of Lyn (Fig. 5B). Using Lyn-specific antibody, we found that Lyn is expressed in the prostatic epithelia of both humans and mice (Fig. 1, A and C). In an overall survey, all of the normal human (11 of 11) and mouse (5 of 5) prostate epithelia examined were Lyn positive. Previous studies detected Lyn transcripts, in addition to lymphoid tissues, in a variety of adult mouse tissues, including liver, kidney, brain, heart, ovary, and testes (15, 16). Our immunohistochemical staining of human tissues confirms that Lyn expression can be detected in lymphoid tissues as well as in the kidney, brain, ovary, placenta, and liver (Kupffer cells). However, we did not detect positive cells in the myocardium. In addition, breast, stomach, and colon epithelia showed clear staining with anti-Lyn antibodies (data not shown). We then used immunohistochemistry to assess the expression of Lyn in human primary PC tissue arrays (Fig. 1, D–F), and in PC metastasis (Fig. 1, G and H). Lyn expression was evident in 137 (95%) of 145 samples of primary PC and in all of the analyzed PC metastasis to lymph node or bone samples (5 of 5 and 6 of 6, respectively). In all cases, Lyn staining in the neoplastic tissue was at least as intense as that of the adjacent normal prostate epithelium, when present (e.g., see Fig. 1D). Notably, in cancers showing two Gleason patterns in different parts of the same tumor, Lyn expression tended to be higher in the less differentiated regions. In addition, Lyn mRNA was detected using Northern blot analysis in human WISHPC2 tumor xenografts [derived from an androgen-independent neuroendocrine small cell prostate carcinoma (17)], cultured DU145 HRPC cells, and normal whole human prostate RNA (Fig. 1I).

Fig. 1. Lyn is expressed in prostate epithelia. Immunohistochemical staining of human (A, B, D–H) and mouse (C) prostate tissues using anti-Lyn antibody. Normal glands (A–C), a serial section of the gland shown in A exposed to a Lyn-specific blocking peptide, concomitantly with the anti-Lyn antibody (B), normal mouse prostate epithelium (C). D, human prostate carcinoma (T) with adjacent normal tissue (N). Human prostate cancer (PC) specimens (E–H). Two different primary PCs (E and F), PC metastasis to a lymph-node (G) or bone (H). The lower corner of G contains residual lymph node tissue. In this area, the B-cell-rich lymphoid follicles (white arrowhead) stain positively, whereas the T-cell-rich zone in the medulla (white asterisk) does not stain. A, B, ⫻100; C, D, H, ⫻400; E—G, ⫻200. I, Northern blot analysis of total RNA from human spleen (Lane 1), human prostate (Lane 2), human hormone-refractory PC neuroendocrine xenograft WISHPC2 (Lane 3), and DU145 cells (Lane 4). RNAs were hybridized with 32P-labeled Lyn probe.

The Prostate Gland in Lyn-Deficient Mice Is Underdeveloped. To find out whether Lyn expression is important for normal prostate gland development, we analyzed prostate morphology in Lyndeficient (Lyn⫺/⫺) mice. To that end, whole-mount prostate gland

1060



Fig. 2. Lyn-deficient mice exhibit compromised prostate-gland development. Top panels, representative whole-mount of the ventral lobe of the prostate gland in 10 weeks old Lyn⫹/⫹ versus Lyn⫺/⫺ mice. Middle panels, H&E-stained sections of prostates from 10-week-old Lyn⫹/⫹ and Lyn⫺/⫺ mice. The dorsal (D) and ventral (V) lobes are marked for orientation. Bottom panels, anti-bromodeoxyuridine (anti-BrdUrd) staining of the anterior prostate lobes from 7-week-old Lyn⫹/⫹ and Lyn⫺/⫺ mice. Middle and bottom panels, ⫻25.

specimens from six 10-week-old Lyn⫺/⫺ mice, and six age-matched wild-type (Lyn⫹/⫹) controls were prepared according to Sugimura et al. (12). It clearly showed that the prostate glands of Lyn⫺/⫺ mice were uniformly smaller and had a much lower degree of branching of the ductal networks, as compared with those of Lyn⫹/⫹ mice (Fig. 2, top panels). To study more carefully the morphology of the epithelial cells in Lyn⫺/⫺mice, we prepared en bloc preparations from four 10-week-old Lyn⫺/⫺ mice, and four age-matched controls. Histological analysis of the prostate gland in Lyn⫺/⫺ mice revealed an overall reduction in the epithelium complexity with marked attenuation of the thickness of the epithelial layer. This difference is accentuated in the dorsal lobes, which normally have a more complex architecture than their ventral counterparts (Fig. 2, middle panels). Prostate morphology of heterozygous Lyn⫹/⫺ mice was indistinguishable from that of Lyn⫹/⫹ mice (data not shown). To quantify the difference, serial sections of the entire gland were prepared from Lyn⫹/⫹ and Lyn⫺/⫺ mice. The calculated area of the largest cross-section in the Lyn⫹/⫹ mice was 6.74 mm2 ⫾ 0.34 mm2 (mean ⫾ SE; n ⫽ 10) as compared with 2.83 mm2 ⫾ 0.44 mm2 (n ⫽ 7) in Lyn⫺/⫺ mice (P ⬍ 0.001 by Student’s t test). To further analyze the role of Lyn in normal prostate development, we examined whether Lyn deficiency affected epithelial proliferation at early stages of prostate development. For this, Lyn⫹/⫹ and Lyn⫺/⫺ mice, aged 3 and 7 weeks, were given injections of bromodeoxyuridine, and their prostates were dissected. Staining for bromodeoxyuridine revealed that the proliferation rate is markedly higher in Lyn⫹/⫹ prostates compared with Lyn⫺/⫺ prostates, at both

ages (Fig. 2, bottom panels, and data not shown), suggesting that Lyn is involved in the signaling cascade responsible for the regulation of epithelial growth. Collectively, our findings suggest that Lyn plays an important role in the development of prostate epithelium and its physiology. This led us to explore the possibility that Lyn inhibition may prove beneficial for the treatment of prostate cancer. A Lyn-Derived Designer Peptide Specifically Inhibits LynDependent Phosphorylation. Currently, most available inhibitors of PKs target the ATP binding site of the enzyme, which is common to all PKs, a feature that entails an inherent lack of specificity by this type of compounds (18). On the other hand, it was shown that short peptides derived from the substrate interaction domain of MAPKkinase 1 (MEK1) can inhibit the phosphorylation of its substrate MAPK2, presumably by abrogating the association of the kinase with its substrate (19). With respect to Lyn, a detailed structural analysis, based on nuclear magnetic resonance data and high resolution modeling, pointed to the loop between ␣-helix F and ␣-helix G (“the HJ-loop”) as being a substrate binding site of Lyn (20). This work is supported by other observations that implicate the HJ-loop region of the kinase domain in kinase-substrate interaction and the region that determines substrate specificity (21–24). This loop is fully exposed and available for external interactions (see the three-dimensional illustration in Fig. 4A). Comparative analysis among different PKs revealed that PK subfamilies are typified by an exclusive combination of amino acids (aa) within the HJ-loop that are unique to each subfamily. We, therefore, hypothesized that peptide copies of that

1061



region might serve as specific inhibitors, acting via disruption of the interaction between the kinase and its substrates. To that end, we synthesized a series of short peptides derived from the HJ-loop of Lyn and used an in vitro kinase assay to test the efficacy and specificity of these peptides against Lyn and other members of the Src family. The activity of one of these peptides, designated KRX-123.302, is shown in Fig. 3. As a first step, we determined whether KRX-123.302 affects Lyn transphosphorylation in vitro. Incubation of active Lyn with kinase-dead Lyn (Lyn-K275L) and ATP resulted in robust phosphorylation of Lyn. Addition of KRX-123.302 to the reaction mixture resulted in a significant inhibition of Lyn transphosphorylation in a dose-dependent manner, showing 50% inhibition at around 1 ␮M (Fig. 3A). We further assessed the activity of KRX-123.302 toward two related PTKs, Lck and Fyn. Whereas Lyn (and KRX-123.302 accordingly) contains Tyr at position P4 of the HJ-loop (see Fig. 4B for details), in Lck and Fyn, the corresponding residues are His and Lys, respectively. In addition, Lyn has Lys, whereas Lck and Fyn have Arg at position P6. As shown in Fig. 3A, even 20 ␮M KRX-123.302 did not inhibit the transphosphorylation of either Lck or Fyn, thus suggesting that KRX-123.302 is a selective Lyn inhibitor. The next set of experiments was designed to test whether KRX-123.302 can inhibit the phosphorylation of Syk, a physiological substrate of Lyn (6). We expressed Syk in bacteria as a glutathione S-transferase fusion protein and incubated purified glutathione S-transferase-Syk with active Lyn and ATP, in the presence of KRX-123.302. KRX-123.302 markedly inhibited the phosphorylation of Syk in a dose-dependent manner showing 50% inhibition at 0.3 ␮M. The addition of a control nonrelated myristylated peptide designated KRX-107.110 (see “Materials and Methods”), did not inhibit Lyndependent phosphorylation at up to 5 ␮M (Fig. 3B). The phosphorylation of an additional Lyn target, Vav (25), was also inhibited by KRX-123.302 in a dose-dependent manner (data not shown). Thus, we have found that a specific sequence of 8 aa derived from a unique interaction site of Lyn, specifically inhibits target phosphorylation by Lyn, but not by closely related Src family members Lck and Fyn in vitro. We hypothesize that the synthetic peptide behaves like a pseudokinase and inhibits Lyn via binding to Lyn substrates. The inhibition of Lyn activity by KRX-123.302 in the cell-free assays prompted us to examine its effect on Lyn-dependent phosphorylation and signaling in intact B lymphocytes. We chose B-cell antigen receptor (BCR) activation as a model system because the signaling pathways mediated by Lyn have been thoroughly studied in this system (6, 25). In particular, not only is it known that Lyn directly phosphorylates CD19 in these cells, but it is the sole kinase responsible for CD19 phosphorylation, because CD19 phosphorylation is undetectable in Lyn-deficient B cells even after BCR ligation (25). CD19 phosphorylation was tested after stimulation of splenic B cells with F(ab⬘)2 fragments of goat-antimouse IgM antibody in the presence of either 10 ␮M KRX-123.302 or vehicle. Addition of KRX123.302 significantly inhibited CD19 phosphorylation and Lyn transphosphorylation of stimulated splenic B cells (Fig. 3C). The latter finding is consistent with the proposed model for Lyn participation in BCR signaling, which involves CD19 successive phosphorylation and the augmentation of Lyn transphosphorylation (25). It is also in agreement with the inhibition of transphosphorylation of Lyn observed in the cell-free system (Fig. 3A). Furthermore, KRX-123.302 suppressed phosphorylation of Syk, which is phosphorylated by Lyn on BCR ligation (6), in the B-lymphoma cell line WEHI-231, whereas the vehicle had no effect on Lyn-dependent tyrosine phosphorylation (Fig. 3D). In the cell-free kinase assay, the inhibition of Syk or Lyn phosphorylation was obtained at similar concentrations of KRX-123.302 (Fig. 3, A and B), whereas in the cell-based assay, the inhibition of Syk phosphorylation required a

Fig. 3. Lyn-dependent phosphorylation is selectively inhibited by the sequence-based inhibitor KRX-123.302. The inhibitory effect of KRX-123.302 is tested in cell-free kinase assays (A, B), in primary B cells (C), and in WEHI-231 B-lymphoma cells (D, E). A, transphosphorylation of Lyn was performed by incubating active recombinant Lyn, kinase-dead Lyn (Lyn-K275L), and KRX-123.302 at the concentrations indicated, followed by addition of ATP. Transphosphorylation assays of Lck and Fyn were performed by incubating the active recombinant kinase with KRX-123.302, followed by addition of ATP. Samples were separated by SDS-PAGE, and the phosphorylated form of each kinase was visualized by immunoprobing with anti-phospho-tyrosine monoclonal antibody (pTyr mAb), followed by reprobing with the appropriate antikinase antibody to verify equivalent amounts of proteins within all of the samples. B, KRX-123.302 or a control peptide KRX-107.110 were incubated with active recombinant Lyn and glutathione S-transferaseSyk followed by addition of ATP. Samples were separated by SDS-PAGE, and the phosphorylated Syk (pSyk) was visualized by immunoprobing with anti-phospho-tyrosine (anti-pTyr) mAb, followed by reprobing with anti-Syk antibody (Syk). For both A and B, the level of phosphorylation was normalized to the level of the relevant protein. C, resting splenic B cells were incubated for 2 h with either vehicle (veh) or 10 ␮M KRX-123.302, followed by 5 min stimulation with F(ab⬘)2 fragments of goat-antimouse IgM antibody (anti-IgM) antibody. Lyn or CD19 were immunoprecipitated, separated by SDS-PAGE, and probed with anti-pTyr mAb, followed by reprobing with the precipitating antibody. The level of the phosphoprotein (pLyn, pCD19) was normalized to the level of the total protein (Lyn, CD19). D, WEHI-231 cells were incubated for 2 h with KRX-123.302 or the vehicle (veh), at the indicated concentrations, followed by 5-min stimulation with antiIgM antibody. Cell lysates, separated by SDS-PAGE, were probed with anti-pTyr mAb, followed by reprobing with anti-Syk antibody. E, WEHI-231 cells were incubated with 20 ␮M KRX-123.302 for 2 h. Cells were lysed at the indicated time points (sec) after anti-IgM stimulation, and samples, separated by SDS-PAGE, were probed with anti-phosphomitogen-activated protein kinase (pMAPK), followed by anti-MAPK (MAPK) antibodies. Bands were quantified by using NIH image software, and the graphs represent relative band intensities (means ⫾ SD) from three independent experiments.

higher concentration of KRX-123.302, as compared with the inhibition of Lyn transphosphorylation (Fig. 3, C versus D). This may be a result of the different cell types used (primary splenic B cells versus B-lymphoma WEHI-231 cell line). A phenomenon observed in Lyn⫺/⫺ B cells is the enhancement of BCR-induced activation of MAPKs (6). This finding is compatible with the role of Lyn as an

1062



Fig. 4. Structure-activity relationship of Lyn-derived peptides. A, location of the HJ-loop (the loop between ␣-helix F and ␣-helix G) within the canonical three-dimensional structure of the kinase domain: space-filling model of the crystal structure of Hck (35), the protein tyrosine kinase closest to Lyn, based on sequence similarity. Highlighted in blue, the seven-amino-acid (aa) residues corresponding to the HJ-loop of Lyn. B, detailed structure of the tested peptides alongside with their corresponding inhibitory activity of DU145 and PC3 cell proliferation. Highlighted in bold, changes from the native sequence. Lower-case letters, D configuration of the corresponding aa; Y#, di-iodo Tyr; K&, Lys-⑀ amino-benzoyl; K$, Lys-⑀ amino-dansyl. All peptides were C-amidated (-NH2). % inhibition, cell number relative to that of cells grown in the presence of vehicle alone. C, illustration of intracellular permeabilization. DU145 cells incubated with either the nonbiotinylated control peptide (left panel) or with the biotin-tagged peptide (right panel); ⫻2000.

attenuator of B-cell response. We, therefore, tested whether KRX123.302 has a similar effect in BCR-activated WEHI-231 cells. In full agreement with its proposed mode of action as a Lyn-specific inhibitor, KRX-123.302 enhances F(ab⬘)2 fragments of goat-antimouse IgM antibody-induced MAPKs activation (Fig. 3E). Thus, using both an in vitro kinase assay and stimulated intact B cells, we demonstrated the ability of our designer short mimic of Lyn to specifically inhibit Lyn kinase activity and to abrogate Lyn-mediated signaling. The Lyn-Derived Designer Peptides Inhibit HRPC Cell Proliferation in Vitro. To study the possibility that Lyn inhibition compromises prostate cancer cells, we tested the activity of the different synthetic peptides derived from the HJ-loop of Lyn, vis a` vis the proliferation of the HRPC cell lines, DU145 and PC3, as detailed below (Fig. 4B). Our initial study showed that a synthetic peptide, composed of a seven-aa sequence of the Lyn parental HJ-loop, IV-

TYGKI (designated KRX-123.101; see Fig. 4B), significantly inhibits HRPC cell proliferation. We further modified this peptide structure to optimize its biological activity, and we refer to this group of peptides collectively as KRX-123 (Fig. 4B). The covalent attachment of myristic acid, through additional Gly at the NH2 terminus, enables intracellular permeabilization of the peptide, as demonstrated by fluorescent labeling. The internalized peptide is evenly distributed in the cytoplasm and is excluded from the nucleus (Fig. 4C). As can be seen in Fig. 4B, the peptide activity is sequence dependent and a nonrelated control peptide, KRX-683.112 (a myristylated derivative of the HJ-loop of G-protein-coupled receptor kinase-2), had no effect on the proliferation of either DU145 or PC3. Likewise, some single aa modifications abolish the inhibitory activity (e.g., KRX123.156 and KRX-123.124). Other modifications of the original KRX-123.101 were associated with increased inhibitory efficacy; in particular, the di-iodo substitution of Tyr (KRX-123.135), and the replacement of the native Gly by D-Lys (KRX-123.134) resulted in enhancement of the inhibitory activity. In addition, the peptide activity can be enhanced by aromatic adducts at the COOH terminus (Fig. 4B, KRX-123.302 and KRX-123.719). It is noteworthy that the inverted peptide sequence gave a response similar to the parental one (KRX-123.202 versus KRX-123.101). This phenomenon, of a significant activity of the retro sequence of a short peptide, has been described in other systems (26), indicating the greater importance of the side-chain residues location relative to each other, rather than to the backbone. To determine whether the Lyn inhibitor was effective in other cell types, we tested human cancer cell lines from other epithelial tissues for the inhibition of their proliferation by KRX-123.302. We also tested Lyn and Src expression levels in these cell lines using Western blot analysis. All three of the PC lines tested (DU145, PC3, and TSU-Pr1), express considerable levels of Lyn and Src and were strongly inhibited by the Lyn-derived peptide (Fig. 5). On the other hand, a human colon cancer cell line (HT29) that barely expresses Lyn, yet expresses high levels of Src, was refractory to the treatment (Fig. 5). In Vivo Administration of Lyn-Derived Peptides Induces Regression of HRPC Xenografts. After structural optimization of the Lyn-derived peptide in HRPC cell lines in vitro, we chose the two most active derivatives, KRX-123.302 and KRX-123.719, for testing their efficacy in vivo. To that end, nude mice were inoculated s.c. with DU145 cells. When the tumors reached a volume of ⬃100 mm3, we initiated i.v. treatment with an inhibitory peptide or vehicle only (a single injection once a week, for a period of 5 weeks). An i.v. injection of KRX-123.302 or KRX-123.719, at a dose of 10 mg/kg resulted in significant tumor regression (Fig. 6A). On the other hand, in all of the control animals (vehicle-treated plus nontreated), the tumors continued to grow, with more than doubling in tumor volume within 1 month (Fig. 6A). Statistical analysis of tumor volume among the various groups showed a significant difference between KRX123.719-treated tumors and vehicle (P ⬍ 0.01, Mann Whitney test) or no treatment (P ⬍ 0.01). Tumor volume of KRX-123.302-treated mice was also significantly smaller than that of nontreated animals (P ⬍ 0.01); however, the difference from the vehicle-treated group did not reach statistical significance (P ⫽ 0.16). In an additional set of experiments with various modifications of the peptides, 56 animals were treated with inhibitory peptides, whereas 26 animals received vehicle only or no treatment. Cumulatively, complete regression of the tumors was noted in 25% (17 of 67) of the treated mice, after once-a-week injections of the inhibitory peptide for a period of 5– 8 weeks, whereas no regression was observed in the controls (0/37; P ⬍ 0.005, ␹2 test with one degree of freedom). To study the morphological changes induced in the tumors as a

1063



DISCUSSION

treatment options for PC patients were achieved since the breakthrough finding, more than 50 years ago, that hormonal ablation induces regression of PC. In recent years, agents against a series of molecular targets, that positively control malignant growth, have been explored. For instance, targeting the epidermal growth factor receptor by mAbs as well as by inhibitors of the epidermal growth factor receptor-associated tyrosine kinase were shown to inhibit cellular proliferation and angiogenesis and to enhance apoptosis, either when applied alone or in combination with paclitaxel (27, 28). Downstream substrates of cell surface receptors that have been explored as targets for therapy include members of the Ras-Raf-MAPK pathway involved in cell proliferation, and the phosphoinositide 3-kinase/protein kinase B pathway that plays an important role in cell survival (29). Agents are also being developed against apoptosis-regulating pathways, such as Bcl-2 and nuclear factor ␬B using antisense oligonucleotides, or proteasome inhibitors. Nevertheless, the puzzle of human PC proteome is far from being resolved, and it is not yet clear which candidate will yield an effective therapy. Therefore, the identification of new target proteins and a validation of their role might be of great importance. In this study, we identified a novel molecular target for PC, the PTK Lyn. Moreover, we demonstrated our ability to design a specific inhibitor that acts efficiently against Lyn-expressing HRPC xenografts in vivo. Evidence presented in this study indicates that Lyn is involved in the control of the development and proliferation of the prostatic epithelium. First, Lyn has a role in prostate morphogenesis, as demonstrated by the aberrant development of the gland in Lyn⫺/⫺ mice (Fig. 2). Second, Lyn is extensively expressed in human PC specimens (Fig. 1). Third, the inhibition of Lyn activity using a specific sequence-based inhibitor decreases proliferation of HRPC cell lines (Figs. 4, 5), and significantly reduces tumor growth in a tumor explant model (Fig. 6A). Fourth, Lyn inhibition induced apoptosis in HRPC cells in vivo, as demonstrated by TUNEL analysis (Fig. 6B). In addition, RNA interference-mediated specific inhibition of Lyn expression significantly suppressed the proliferation of DU145 cells.5 Our finding that the known B-cell-specific PTK, Lyn, plays an important role in prostate epithelium physiology is an example for different assignments of allegedly cell-specific PTKs, in alternate tissues or cell types. Yet another such example is the PTK Syk, which, until recently, was considered an exclusive B-cell signaling molecule and was now shown to play a regulatory role in the breast epithelium (30). It is well known that Lyn activation plays a critical role in BCR signaling, in which it leads to attenuation of the immune response (5, 6). However, several lines of evidence suggest that Lyn regulates cell survival through the stimulation of proliferation and the inhibition of apoptosis in various hematopoietic cell lineages (31–33). In addition, overexpression of dominant-active Lyn in colon cancer cells conferred chemoresistance through activation of the phosphoinositide 3-kinase/protein kinase B pathway (7). Similarly, the expression of wild-type, but not kinase-inactive, Lyn was found to protect human embryonic kidney 293 and HeLa cells from drug or ionizing radiationinduced apoptosis (8). Importantly, Grishin et al. (8) have also shown that interaction between Lyn and growth arrest-DNA damage protein 34 negatively regulates genotoxic apoptosis. Several publications reported expression of Lyn in PC cell lines, and in dog’s normal and hyperplastic prostates (10, 11, 14). Lyn was shown to be ubiquitously expressed in a panel of human PC cell lines; however, its role in this system remained unclear (11). Another study

PC continues to be one of the leading causes of cancer deaths. Hence, the development of an innovative therapy against HRPC still constitutes an outstanding challenge. Only moderate improvements in

5 H. Rubin, M. Goldenberg-Furmanov, I. Stein, R. Porat, E. Pikarsky, I. Wexler, and H. Reuveni. Selective down-regulation of Lyn inhibits the ERK pathway and induces apoptosis of prostate cancer cells, manuscript in preparation.

Fig. 5. Inhibition of cell proliferation by KRX-123.302 correlates with levels of Lyn expression. A, dose-response curves of human prostate cancer (PC) cell lines (DU145, PC3, TSU-Pr1), and a human colon cancer cell line (HT29) treated with Lyn-derived peptide KRX-123.302. The results are expressed as % cell number in peptide-treated wells, relative to that of cells grown in the presence of vehicle only (blue rectangles, DU145; red circles, PC3; purple triangles, TSU-Pr1; green squares, HT29). B, Western blot of Lyn and Src levels in the indicated cell lines. Exponentially growing cells were collected and lysed. Total cellular extracts, normalized for protein concentration, were separated by SDS-PAGE, and were immunoblotted with anti-Lyn (top panel) or anti-Src antibodies (middle panel). Bottom panel, ␤-actin loading control.

result of the treatment, we removed tumors several weeks after treatment initiation, before complete regression was achieved, and submitted them for histological evaluation. H&E staining of the PC xenografts depicted more apoptotic bodies in the KRX-123-treated tumors. Accordingly, TUNEL assay demonstrated massive apoptosis in the nonnecrotic parts of the treated tumors, in sharp contrast to the occasional apoptotic cells in control tumors (Fig. 6B). Quantification of the proportion of TUNEL-positive nuclei revealed that only 1.5 ⫾ 1.4% (mean ⫾ SE) of the nuclei in vehicle-treated animals where apoptotic, compared with 8.4 ⫾ 2.2% of the nuclei in the peptide-treated group. Interestingly, the morphological features of tumor cells in treated mice are reminiscent of the changes induced in PC cells by hormone ablation therapy in humans, including pyknotic nuclei and vacuolization of the cytoplasm (data not shown).

1064



Fig. 6. Sequence-based Lyn inhibitors induce a regression of human tumor xenografts. CD1 nude mice were inoculated s.c. with DU145 cells. When tumor volume reached ⬃100 mm3, the peptide was administered by i.v. injection once a week. A, relative change in tumor volume with time ⫾ SE in KRX-123.719-treated (red circles), KRX-123.302-treated (yellow triangles), vehicle-treated (green squares), and nontreated (blue rectangles) mice. Arrows, time of injection. B, the Lyn-derived peptides induce apoptosis in prostate tumor xenografts. Terminal deoxynucleotidyl transferase-mediated nick end labeling assay for apoptosis of prostate cancer xenografts treated with KRX-123.302, KRX-123.719, or vehicle; ⫻400.

has shown that not only is Lyn expressed in a PC cell line, but that it is down-regulated within 30 min after transforming growth factor ␤ treatment (10). This finding, in concert with data illustrating that transforming growth factor ␤ is a negative regulator of PC cell growth (34), suggests that transforming growth factor ␤ may exert its negative regulatory role in PC cells, at least partially, via the inhibition of Lyn expression. The above findings, in conjunction with our own data, collectively support the view that Lyn may play a role in the evolvement of PC and, therefore, can serve as a target for the development of anti-PC therapy. It could be argued that Lyn inhibition in PC patients will prove to be a double-edged sword because Lyn-deficient mice develop autoimmune disease and show defective mast cell function. However, our preliminary results based on treating dozens of immune-competent animals with high doses of KRX-123.302 detected only slight reductions in platelet numbers that did not fall below the physiological level, and minimal focal liver inflammation that did not result in altered liver function tests (data not shown). In addition, no signs of autoimmune kidney disease were noted after 5 weeks of KRX-123.302 treatment. Obviously, additional studies are needed to identify more subtle effects of reversible Lyn inhibition in immunecompetent animals before human studies are contemplated. Protein-protein interactions govern many important cellular functions and constitute a major challenge for drug discovery attributable to their complexity. We have devised a rational strategy for inhibiting the binding of PKs to their respective substrates. This process involves first identifying a key kinase-substrate interaction site, and then developing sequence-based peptide inhibitors to interfere with this interaction. Strong emphasis is placed on the structure–activity relationship with the peptide being modified in a manner that enhances efficacy. This strategy of sequence-based modulation is widely appli-

cable for multiple PKs because the protein–protein interaction site targeted is easily identifiable in most PKs, yet is distinct for each kinase subfamily, so that selective inhibitors can be specifically devised. Our structure-based approach, contrasts with the more common high-throughput screening of random combinatorial-chemistry libraries. In this study, we targeted the putative substrate-binding site of Lyn. This region within the kinase domain is composed of a short linear sequence, the HJ-loop, residing between ␣ helix F and ␣ helix G (Fig. 4A). Several studies point to the direct involvement of this region in substrate binding by PKs (21–24). For example, it is sufficient to exchange a short sequence that includes the HJ-loop, between Jun NH2-terminal kinase 1 and 2, to confer Jun NH2-terminal kinase 2 substrate specificity to an otherwise Jun NH2-terminal kinase 1 construct (22). Our comparative analysis across dozens of available PK sequences revealed that there is a consistent pattern of distinct composition of aa at the HJ-loop, unique to each PK subfamily and individual kinase within. Thus, each PK subset displays a unique combination of contact residues, embedded within a conserved framework. We thus postulated that peptides derived from this region can serve as kinase-specific inhibitors and would interfere with PK signaling in a predictive manner by serving as a kinase decoy. In agreement with this notion, we showed here that short peptides derived from this region of the Lyn kinase are able to (a) inhibit Lyn transphosphorylation and Lyn kinase activity in cell-free assays (Fig. 3, A and B); (b) inhibit Lyn transphosphorylation and interfere with Lyn-dependent signaling in intact B cells (Fig. 3, C–E); (c) inhibit the proliferation of Lyn-expressing HRPC cells (Figs. 4 and 5); (d) induce apoptosis in such cells (Fig. 6B); and (e) inhibit growth and even induce regression of tumors after injections to mice bearing

1065



HRPC xenografts (Fig. 6A). Although our results suggest that the Lyn-derived peptides inhibit PC explants in vivo through Lyn inhibition, at present we are unable to study the effect of the inhibitor on downstream Lyn targets in PC, because these targets are not yet identified. Several key issues still require further investigation. The identity and biochemical function of the immediate molecular partners upstream and downstream of Lyn in the normal prostate and PC cells are yet unknown. It will be interesting to characterize them and analyze their phosphorylation status under different physiological and pathological conditions. Another intriguing question is related to the spectrum of normal and neoplastic epithelial cells of other lineages expressing Lyn, or other members of the Src PTK family. Our approach offers a powerful research tool to study the involvement of PKs in various systems and opens the way for a systematic generation of designer inhibitors against a variety of PKs in general, and those involved in PC in particular. ACKNOWLEDGMENTS Lyn⫺/⫺ mice were a gift from Dr. Clifford A. Lowell, University of California at San Francisco, San Francisco, CA. The WISH-PC2 xenograft was a gift from Dr. Zelig Eshhar, the Weizmann Institute, Rehovot, Israel. We gratefully acknowledge the technical assistance of Alex Tartakovsky, Mali Sapir, and Tamara Golub.

Note Added in Proof A detailed account of structure-function relationships related to the protein kinases HJ-loop, the site targeted by us, was recently accepted for publication (M. Y. Niv et al., J. Biol. Chem., 279: 1242–1255, 2004). The full text of this paper is available at JBC online.

REFERENCES 1. Gopalkrishnan, R. V., Kang, D. C., and Fisher, P. B. Molecular markers and determinants of prostate cancer metastasis. J. Cell. Physiol., 189: 245–256, 2001. 2. Sachsenmaier, C. Targeting protein kinases for tumor therapy. Oncologie, 24: 346 – 355, 2001. 3. Thomas, S. M., and Brugge, J. S. Cellular functions regulated by Src family kinases. Ann. Rev. Cell Dev. Biol., 13: 513– 609, 1997. 4. Bolen, J. B., Rowley, R. B., Spana, C., and Tsygankov, A. Y. The Src family of tyrosine protein kinases in hemopoietic signal transduction. FASEB J., 6: 3403–3409, 1992. 5. Hibbs, M. L., Tarlinton, D. M., Armes, J., Grail, D., Hodgson, G., Maglitto, R., Stacker, S. A., and Dunn, A. R. Multiple defects in the immune system of Lyndeficient mice, culminating in autoimmune disease. Cell, 83: 301–311, 1995. 6. Chan, V. W., Meng, F., Soriano, P., DeFranco, A. L., and Lowell, C. A. Characterization of the B lymphocyte populations in Lyn-deficient mice and the role of Lyn in signal initiation and down-regulation. Immunity, 7: 69 – 81, 1997. 7. Bates, R. C., Edwards, N. S., Burns, G. F., and Fisher, D. E. A CD44 survival pathway triggers chemoresistance via Lyn kinase and phosphoinositide 3-kinase/Akt in colon carcinoma cells. Cancer Res., 61: 5275–5283, 2001. 8. Grishin, A. V., Azhipa, O., Semenov, I., and Corey, S. J. Interaction between growth arrest-DNA damage protein 34 and Src kinase Lyn negatively regulates genotoxic apoptosis. Proc. Natl. Acad. Sci. USA, 98: 10172–10177, 2001. 9. Roginskaya, V., Zuo, S., Caudell, E., Nambudiri, G., Kraker, A. J., and Corey, S. J. Therapeutic targeting of Src-kinase Lyn in myeloid leukemic cell growth. Leukemia (Baltimore), 13: 855– 861, 1999. 10. Atfi, A., Drobetsky, E., Boissonneault, M., Chapdelaine, A., and Chevalier, S. Transforming growth factor ␤ down-regulates Src family protein tyrosine kinase signaling pathways. J. Biol. Chem., 269: 30688 –30693, 1994. 11. Sumitomo, M., Shen, R., Walburg, M., Dai, J., Geng, Y., Navarro, D., Boileau, G., Papandreou, C. N., Giancotti, F. G., Knudsen, B., and Nanus, D. M. Neutral endopeptidase inhibits prostate cancer cell migration by blocking focal adhesion kinase signaling. J. Clin. Investig., 106: 1399 –1407, 2000.

12. Sugimura, Y., Cunha, G. R., and Donjacour, A. A. Morphogenesis of ductal networks in the mouse prostate. Biol. Reprod., 34: 961–971, 1986. 13. Oliver, M. H., Harrison, N. K., Bishop, J. E., Cole, P. J., and Laurent, G. J. A rapid and convenient assay for counting cells cultured in microwell plates: application for assessment of growth factors. J. Cell. Sci., 92: 513–518, 1989. 14. Allard, P., Atfi, A., Landry, F., Chapdelaine, A., and Chevalier, S. Estradiol activates p60src, p53/56Lyn and renatured p50/55 protein tyrosine kinases in the dog prostate. Mol. Cell. Endocrinol., 126: 25–34, 1997. 15. Yi, T. L., Bolen, J. B., and Ihle, J. N. Hematopoietic cells express two forms of lyn kinase differing by 21 amino acids in the amino terminus. Mol. Cell. Biol., 11: 2391–2398, 1991. 16. Yamanashi, Y., Mori, S., Yoshida, M., Kishimoto, T., Inoue, K., Yamamoto, T., and Toyoshima, K. Selective expression of a protein-tyrosine kinase, p56lyn, in hematopoietic cells and association with production of human T-cell lymphotropic virus type. Proc. Natl. Acad. Sci. USA, 86: 6538 – 6542, 1989. 17. Pinthus, J. H., Waks, T., Schindler, D. G., Harmelin, A., Said, J. W., Belldegrun, A., Ramon, J., and Eshhar, Z. WISH-PC2: a unique xenograft model of human prostatic small cell carcinoma. Cancer Res., 60: 6563– 6567, 2000. 18. Morin, M. J. From oncogene to drug: development of small molecule tyrosine kinase inhibitors as anti-tumor and anti-angiogenic agents. Oncogene, 19: 6574 – 6583, 2000. 19. Xu, B. E., Stippec, S., Robinson, F. L., and Cobb, M. H. Hydrophobic as well as charged residues in both MEK1 and ERK2 are important for their proper docking. J. Biol. Chem., 276: 26509 –26515, 2001. 20. Gaul, B. S., Harrison, M. L., Geahlen, R. L., Burton, R. A., and Post, C. B. Substrate recognition by the Lyn protein-tyrosine kinase. NMR structure of the immunoreceptor tyrosine-based activation motif signaling region of the B cell antigen receptor. J. Biol. Chem., 275: 16174 –16182, 2000. 21. Taylor, S. S., Zheng, J., Radzio-Andzelm, E., Knighton, D. R., Ten Eyck, L. F., Sowadski, J. M., Herberg, F. W., and Yonemoto, W. M. cAMP-dependent protein kinase defines a family of enzymes. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 340: 315–324, 1993. 22. Smeal, T., Hibi, M., and Karin, M. Altering the specificity of signal transduction cascades: positive regulation of c-Jun transcriptional activity by protein kinase A. EMBO J., 13: 6006 – 6010, 1994. 23. Wang, Z., Harkins, P. C., Ulevitch, R. J., Han, J., Cobb, M. H., and Goldsmith, E. J. The structure of mitogen-activated protein kinase p38 at 2.1-A resolution. Proc. Natl. Acad. Sci. USA, 94: 2327–2332, 1997. 24. Xie, X., Gu, Y., Fox, T., Coll, J. T., Fleming, M. A., Markland, W., and Caron, P. R. Crystal structure of JNK3: a kinase implicated in neuronal apoptosis. Structure, 6: 983–991, 1998. 25. Fujimoto, M., Fujimoto, Y., Poe, J. C., Jansen, P. J., Lowell, C. A., DeFranco, A. L., and Tedder, T. F. CD19 regulates Src family protein tyrosine kinase activation in B lymphocytes through processive amplification. Immunity, 13: 47–57, 2000. 26. Yamada, K. M., and Kennedy, D. W. Amino acid sequence specificities of an adhesive recognition signal. J. Cell. Biochem., 28: 99 –104, 1985. 27. Karashima, T., Sweeney, P., Slaton, J. W., Kim, S. J., Kedar, D., Izawa, J. I., Fan, Z., Pettaway, C., Hicklin, D. J., Shuin, T., and Dinney, C. P. Inhibition of angiogenesis by the antiepidermal growth factor receptor antibody ImClone C225 in androgenindependent prostate cancer growing orthotopically in nude mice. Clin. Cancer Res., 8: 1253–1264, 2002. 28. Sirotnak, F. M., Zakowski, M. F., Miller, V. A., Scher, H. I., and Kris, M. G. Efficacy of cytotoxic agents against human tumor xenografts is markedly enhanced by coadministration of ZD1839 (Iressa), an inhibitor of EGFR tyrosine kinase. Clin. Cancer Res., 6: 4885– 4892, 2000. 29. Baselga, J., and Norton, L. Focus on breast cancer. Cancer Cell, 1: 319 –322, 2002. 30. Coopman, P. J., Do, M. T., Barth, M., Bowden, E. T., Hayes, A. J., Basyuk, E., Blancato, J. K., Vezza, P. R., McLeskey, S. W., Mangeat, P. H., and Mueller, S. C. The Syk tyrosine kinase suppresses malignant growth of human breast cancer cells. Nature (Lond.), 406: 742–747, 2000. 31. Wei, S., Liu, J. H., Epling-Burnette, P. K., Gamero, A. M., Ussery, D., Pearson, E. W., Elkabani, M. E., Diaz, J. I., and Djeu, J. Y. Critical role of Lyn kinase in inhibition of neutrophil apoptosis by granulocyte-macrophage colony-stimulating factor. J. Immunol., 157: 5155–5162, 1996. 32. Katagiri, K., Yokoyama, K. K., Yamamoto, T., Omura, S., Irie, S., and Katagiri, T. Lyn and Fgr protein-tyrosine kinases prevent apoptosis during retinoic acid-induced granulocytic differentiation of HL-60 cells. J. Biol. Chem., 271: 11557–11562, 1996. 33. Daigle, I., Yousefi, S., Colonna, M., Green, D. R., and Simon, H. U. Death receptors bind SHP-1 and block cytokine-induced anti-apoptotic signaling in neutrophils. Nat. Med., 8: 61– 67, 2002. 34. Cardillo, M. R., Petrangeli, E., Salvatori, L., Ravenna, L., and Di Silverio, F. Transforming growth factor ␤1 and androgen receptors in prostate neoplasia. Am. J. Epidemiol., 22: 403– 410, 2000. 35. Schindler, T., Sicheri, F., Pico, A., Gazit, A., Levitzki, A., and Kuriyan, J. Crystal structure of Hck in complex with a Src family-selective tyrosine kinase inhibitor. Mol. Cell, 3: 639 – 648, 1999.

1066
