The Suppression of SH3BGRL Is Important for V-Rel-Mediated Transformation

Oncogene (2006) 25, 756–768 & 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc ORIGINAL ARTICLE The suppression of SH3BGRL is important for v-Rel-mediated transformation

SM Majid1, AS Liss1, M You2 and HR Bose Jr1

1Section of Molecular Genetics and Microbiology, Institute of Cellular and Molecular Biology, University of Texas at Austin, Austin, TX, USA and 2Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA

The v-rel oncogene is the most efficient transforming 2004). The vertebrate Rel/NF-kB family consists of member of the Rel/NF-jB family of transcription factors. structurally related proteins that include NF-kB1 (p50/ v-Rel induces avian and mammalian lymphoid cell tumors p105), NF-kB2 (p52/p100), RelA (p65), c-Rel, and RelB and transforms chicken embryo fibroblasts in culture by (Gilmore, 1995). All Rel/NF-kB family members share a the aberrant regulation of genes under the control of Rel/ highly conserved N-terminal Rel homology domain NF-jB proteins. Here we report that the expression of (RHD), which is responsible for DNA-binding and SH3BGRL, a member of the SH3BGR (SH3 domain- dimerization (Bose, 1992; Gilmore, 1995). Rel/NF-kB binding glutamicacid-rich)family of proteins, is down- proteins are maintained as homodimers or heterodimers regulated in v-Rel-expressing fibroblasts, lymphoid in the cytoplasm of cells by the IkB family of inhibitory cells, and splenic tumor cells. Chromatin immunoprecipi- proteins (Cheng et al., 1994; Chiao et al., 1994). Upon tation experiments demonstrated that v-Rel binds to the proper external stimuli, the N-terminal regulatory sh3bgrl promoter in transformed cells. Coexpression of domains of IkBs are phosphorylated by the IkB kinase SH3BGRL with v-Rel in primary spleniclymphocytes complex (Cheng et al., 1994; Chiao et al., 1994). reduced the number of colonies formed by 76%. Muta- Phosphorylation targets the IkBs for ubiquitin-mediated tions in the predicted SH3-binding domain of SH3BGRL proteasomal degradation, resulting in the release and abolished the suppressive effect on v-Rel transformation nuclear translocation of NF-kB complexes (Kerr et al., and resulted in colony numbers comparable to those 1991). Once in the nucleus, the active Rel/NF-kB dimers formed by v-Rel alone. However, mutations in the can bind to 10 base pair DNA sequences (kB sites) and predicted EVH1-binding domain of SH3BGRL only had promote transcription of downstream genes such as ikba a modest effect on suppression of v-Rel transformation. (Cheng et al., 1994). The expression of IkBa is thus This study provides the first example of a gene that is autoregulated to maintain a basal level of NF-kB downregulated in v-Rel-expressing cells that also plays a activity in normal cells (Chiao et al., 1994). role in v-Rel transformation. Disruptions in the regulation of Rel/NF-kB proteins Oncogene (2006) 25, 756–768. doi:10.1038/sj.onc.1209107; have been implicated in a variety of cancers (Karin published online 26 September 2005 et al., 2002; Madrid and Baldwin, 2003). The first evidence of the involvement of the Rel/NF-kB family in Keywords: v-Rel; NF-kB; SH3BGRL; SH3BGR; trans- oncogenic cell transformation came from studies on formation the v-rel oncogene (Stephens et al., 1983). The v-rel oncogene, originally isolated from the avian reticuloen- dotheliosis virus REV-T, is the most efficient transforming member of the Rel/NF-kB family. v-Rel is able to induce avian and mammalian lymphoid cell tumors and Introduction transforms chicken embryo fibroblasts (CEFs) in culture (Franklin et al., 1977; Moore and Bose, 1988; Bose, The Rel/NF-kB family of transcription factors plays an 1992; Carrasco et al., 1996). v-Rel is evolutionarily important role in mediating immune responses, develop- derived from the avian proto-oncoprotein c-Rel and mental growth, inflammatory responses, and apoptosis differs by two N-terminal amino-acid deletions, several in response to a wide variety of stress-inducing stimuli point mutations, and a C-terminal truncation (Stephens (Denk et al., 2000; Chen et al., 2001; Li and Verma, et al., 1983; Wilhelmsen et al., 1984; Bull et al., 1990; 2002; Bonizzi and Karin, 2004; Hayden and Ghosh, Moore et al., 1993; Pahl, 1999). v-Rel also has 11 N-terminal and 18 C-terminal env-derived sequences Correspondence:Dr HR Bose Jr, Section of Molecular Genetics and (Hrdlickova´ et al., 1995a). Unlike normal Rel/NF-kB Microbiology, Institute of Cellular and Molecular Biology, University proteins, v-Rel has evolved to evade the negative of Texas at Austin, 1 University Station A5000, Austin, TX 78712- regulation of IkBa. The C-terminal truncation and 1095, USA. E-mail:[email protected] alterations in the RHD have made v-Rel less susceptible Received 8 March 2005; revised 8 August 2005; accepted 9 August 2005; to inhibition by IkBa and less efficient at activating published online 26 September 2005 transcription of ikba than c-Rel (Sarkar and Gilmore, Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 757 1993; Hrdlickova´ et al., 1995b; Kralova et al., 1996; extracts, and chromatin immunoprecipitation (ChIP) Sachdev and Hannink, 1998; Gilmore, 1999). Escape experiments revealed that v-Rel is bound to this from IkB regulation allows the constitutive nuclear promoter in a v-Rel-transformed cell line. Overexpres- localization of v-Rel homodimers and heterodimers with sion of SH3BGRL in primary splenic lymphocytes other members of the Rel/NF-kB family, where they diminished the transforming properties of v-Rel. More- bind to kB sites in the promoter of target genes normally over, mutations in the predicted SH3-binding domain in regulated by Rel/NF-kB proteins (Kabrun et al., 1991; SH3BGRL abolished this effect and led to equal Baldwin, 1996). numbers of colonies formed as v-Rel alone. These Although it is now well established that v-Rel experiments provide the first example of a down- transforms cells by affecting the transcription of regulated gene in v-Rel-expressing cells that also plays Rel/NF-kB targets, relatively few examples of genes a role in v-Rel transformation. have been identified with altered expression in v-Rel- transformed cells. Some examples of v-Rel upregulated genes include ikba, the transcription factors c-rel, Results nf-kb1, nf-kb2,c-jun,c-fos, IRF-4, and IRF-10, cyto- kines such as MIP-1b and IL-6, the cell surface Identification and characterization of a v-Rel molecules DM-GRASP and p75, and the antiapoptotic downregulated gene and related family members proteins ch-IAP1 and Nr13 (Zhang et al., 1995; Zhang Differential display was employed to identify genes with and Humphries, 1996; Petrenko et al., 1997; Kralova altered expression in v-Rel-transformed CEFs. Total et al., 1998, 2002; Lee et al., 1999; Hrdlickova´ et al., RNA was harvested from CEFs expressing v-Rel after 2001; Nehyba et al., 2002). Of these, only AP-1 and the cells became morphologically transformed. The IRF-4 have been directly implicated in v-Rel-mediated profile of expressed genes from normal CEFs, CEFs transformation (Kralova et al., 1998; Hrdlickova´ et al., transfected with an empty RCAS retroviral vector, or 2001). A few genes exhibit decreased expression patterns CEFs transfected with RCAS expressing v-rel was in v-Rel-transformed cells. These include nap-1, eif-2a, compared by differential display (Figure 1a). Among c-myb, and the AP-1 transcription factor fra-2 (Petrenko the cDNAs identified was an 89 bp DNA fragment that et al., 1997; Kralova et al., 1998). It is expected that the was downregulated in v-Rel-transformed cells. The band normal expression of these genes restricts or otherwise was excised, amplified by PCR, cloned into a plasmid diminishes the transforming effect of v-Rel, but this vector, and used as a probe in Northern blot analysis remains to be demonstrated. of total RNA from CEFs transfected with the empty Here we report that the expression of SH3BGRL is RCAS vector or with RCAS expressing v-rel downregulated in v-Rel-expressing cells. This protein (Figure 1b). The radiolabeled cDNA probe hybridized belongs to the SH3 domain-binding glutamic acid-rich with a 3.2 kb transcript that was expressed at lower protein (SH3BGR) family of proteins, which also levels in the v-Rel-transformed cells relative to control includes SH3BGR, SH3BGRL2, and SH3BGRL3, the cells. first of which was mapped to the region of chromosome In order to obtain the full-length cDNA sequence, the 21 responsible for congenital heart disease in down 89 bp DNA fragment was used as a probe to screen a syndrome patients (Scartenzzini et al., 1997; Egeo et al., CEF cDNA library. The longest positive clone included 1998; Mazzocco et al., 2001, 2002). The protein sequences of all the members except SH3BGRL3 contain two overlapping predicted proline-rich protein- binding domains, a Src homology 3- (SH3-) and an Ena/ Vasp Homology 1- (EVH1-) binding domain. Protein interactions involving SH3 domains have been implicated in signal transduction, cytoskeletal rearrange- ments, membrane trafficking, and other key cellular processes (Cesareni et al., 2002). The EVH1-binding domain has not been characterized extensively; however, the corresponding EVH1 domain-containing proteins have been shown to be involved in neuronal excitatory processes and in regulating the actin cytoskeleton (Ball et al., 2002). Since the normal biological functions of the SH3BGR family members have yet to be elucidated, the Figure 1 Identification of a downregulated gene in v-Rel- roles played by the putative SH3- and EVH1-binding transformed CEFs. (a) Differential display reactions were performed using total RNA isolated from normal CEFs (lane 1), domains in these proteins remain unclear. CEFs transfected with the empty RCAS vector (lane 2), and CEFs In this study, we report the downregulation of transformed by RCAS-v-Rel on the 8th and 14th passages, SH3BGRL as an important step for v-Rel-mediated respectively (lanes 3 and 4). The location of the downregulated transformation. SH3BGRL transcripts were downregu- cDNA band is indicated by the arrow. (b) RNA (20 mg) from CEFs transfected with empty RCAS vector (control) or RCAS vector lated in response to v-Rel in fibroblasts, lymphoid cells, expressing v-rel (v-Rel) was analysed by Northern blot using the and splenic tumors. The promoter of sh3bgrl contains a downregulated cDNA fragment from panel a as a probe. kB site that can be bound by v-Rel complexes in nuclear Methylene blue-stained 18S rRNA is shown as the loading control.

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 758 a poly(A) tail. This clone was extended by 50 Rapid Database searches using the protein sequence revealed Ampliﬁcation of cDNA Ends (RACE)-PCR to generate this to be the chicken ortholog of SH3BGRL (SH3- a 3123-bp-long cDNA sequence that corresponded in binding-glutamic acid-rich protein like). Chicken size to the RNA transcript observed by Northern blot SH3BGRL shares 86 and 85% identity with the human analysis (Figure 2a). Sequence analysis of the gene and mouse homologs, respectively (Figure 2b). revealed a 345-bp-long open reading frame (ORF) that SH3BGRL is a member of the SH3BGR family of predicts a 114 amino acid-long protein product. proteins, which includes three additional members:

a 1 tccgcgtccttcgctccccgccagcATGGTCATCAAGGTGTACATCGCCTCGTCCTCCGGATCCACGGCGATTAAAAAGCAACAGCAAGATGTTCTAGGC M V I K V Y I A S S S G S T A I K K Q Q Q D V L G 25 101 TTCCTGGAAGCCAACAAAATAGAATTTGAGGAAAAAGACATCGCTGCCAATGAGGAGAATCGGAAGTGGATGCGGGAGAACGTCCCTGAAGACAGGCGGC F L E A N K I E F E E K D I A A N E E N R K W M R E N V P E D R R P 58 201 CAGCCAGCGGGAACCCGCTGCCACCCCGGCTCTTCAACGACAGCCGCTACCTCGGGGATTATGAAGCTTTCTTTGAAGCTCGAGAGAACAATGCAGTATA A S G N P L P P R L F N D S R Y L G D Y E A F F E A R E N N A V Y 91 301 TGCATTTTTAGGCTTGACTGCGCCACCTGGTTCAAAGGAAGCTGAAGCACTGGCAAAGCAACAAGCATGAatttcaactgccttaatgttctgtaaaggg A F L G L T A P P G S K E A E A L A K Q Q A * 114 401 aatttccacggctttttataaaatagtaaaactgtctctgcttcaagaaatatagatgtgatttctaacatgtgattggtgcttgcagcattcaccatat 501 ataacacaaatttgaacacaagatgaaaatgtgaacataaaggtagaaagcatggggtgtactattacaaagttacaagcataggaaactgaagcaattg 601 gacatgaatggttttcacatagtttaaattgtcttggcaatttgtccattttgtgtaaacttcaagaagctattttaatacagatgtggttatacctagg 701 aatccacctatttaaagtctaggaaaaaaaagcagcagtatctaaaaccttgtcagatactagctttagtgcctgagttgcctcaaaaaatgttacagaa 801 atacaccatttttcagtgtctttaaattctggtttcagatgaaattgtaggtggtggtgtgtattataaccagtgtagtcactgttgaaagttatgtttg 901 ttcttttgtaggatggatatttaagatttgtattgtaacacactgtacagatgatgaaaaaaaaatgtcatgtatttctcatttaagtgatttgtgtggg 1001 ctcttttgacaaacaaagagagatgtgggcactggacagccttttcgttggcaagttcatcacacagctgcctaagcatgacttggttacacgtttatgt 1101 gcaagattcagtgcacctcaagaaatagatgctgattcccaggcaatgttgcaattaagccaaaatggccagaagtggcatttccttacctgtgtaaggg 1201 tgaacgcctttaatatatttctctacaaaaataccagttcacactgcttatagtattaggcatcttaatgcatgttcgagtagtcattagattcatatag 1301 tattttttcaaatgagtaaagttaaatgtttgaatagcaagaactattgtaattatccaagatttttttaaggtttaaaagcttatttccattgctgttg 1401 ccatatcatcaagctattacaggaattgtgaacccattaaggtgtcttgtattggtgaatttaccacgtgcaatattaacctgtctgttctcccagtgct 1501 catgagatagcctgaatttcctggccatgttctgcaaaaatgcaggtaattacaagtgccctgaaggataagaaacaggtttaaaaaatgtgtgagtaat 1601 ttaaaatgcattatgcaccacataatgtacttatcccctcaaagcatcaaaatacagaaaggtaattagtccttgctgtagttagaggatcattctaatt 1701 gctctgcatacacatattttgattttctttgctttatgaagaaacaggtggtgcggttcagcaaaaggctgcagagtatcatcgcagtgcggttatttcc 1801 tgctgggaaatgtatgcagtgcttttcagagactttacaaaaccgtttacagttcatagttttaaactctccaattaggaacaaggatagaagtaaaaac 1901 tcaggcttacaaagacggtttaaagtgcagtaattaaatacttccctcactgacgtttctttggggcttctgttcattcgctgaaaccgctgattttgtc 2001 aacatctgttaaagcagaaatgcagctccttgctgtggttggagagcagctctgcttctggagggttcacgttgaaactttcacagatgtcagagcagga 2101 tcgagcacagagctgagttaggtacatttaattcacgtttcatcttttctgttgcggcagccagcaagctggcttgaagcttgtgcttaaaaatgcaagc 2201 acgctgtcaaattacacacggatagattaatgttgccttaccttcaacatttacacacacatctgtgctttttttagtaacgagctttatcagcaattac 2301 tgaatttctgtgaaatgtatttttaagatcttttgagtgtgaaagttaagtattgctttaagtttagtgtgttatttttgtttactgtaatttcatcttt 2401 caaaacggaggtgatttttctttcttcatatttgtgtaatctttttttctgtgtctaaactgaaggccaagtttttgagaaagtatttaagaatactgag 2501 tgaagtcttgtgaagaaagtggaatttgacactcaagtcagtatcctcattccaatttcctgtgtaaggaaagtaagttctgtgaattacagtcattatt 2601 ttgaagaagctgtaattgaaagtcactgccagattctgtcagttcagcgacccagagcattctgaaatgaggtcagtctgtttctgatggtcaagagaat 2701 taccaggttaagaaagcgaattcgctttcctcttctggtcccccatctttgccatatggtagcattcagactgtatcagcattcctaagaaagctctgtt 2801 gagacagactgaattacctctttgtccagagaacaaggcaccaacaccaactgaaaaatacctgaaatatctttgttgtgggactgcgtgtgacagtttt 2901 gatgcagacatcttgtttgaatttgggagtgctgggaacgtgttgcctctatgctttttttctcgtccacagacctgagctaacgacgggcacaagaact 3001 gaatttgtttgtttctcattttctgtttcaccatgtagatggcaggcaatgaggactctacctgtattactaaccattcctgtatcattccaataaaacc 3101 gttttagaatgctttctcacgct

b EVH1 SH3 SH3BGRL (C) 1 MVIKVYIASSSGSTAIKKQQQDVLGFLEANKIEFEEKDIAA-----NEENRKWMRENVPED-RRPASGNPLPPRLFNDSRYLGDYEAFFEARENNAVYAFLGLT SH3BGRL (H) 1 MVIRVYIASSSGSTAIKKKQQDVLGFLEANKIGFEEKDIAA-----NEENRKWMRENVPEN-SRPATGYPLPPQIFNESQYRGDYDAFFEARENNAVYAFLGLT SH3BGRL (M) 1 MVIRVYIASSSGSTAIKKKQQDVLCFLEANKIGFEEKDIAA-----NEENRKWMRENVPED-SRPSTGYPLPPQIFNECQYRGDYDAFFEARENNAVYAFLGLT SH3BGR (C) 1 MVIKVFVATSSGSTAIKKKQQEVVGFLEANKIDFQQMDIAG-----DEDNRKWMRENVPGE-KKPQNGIPLPPQIFNEERYCGDFESFFSAKEENIIYSFLGLA SH3BGR (H) 64 MVIKVFVATSSGSIAIRKKQQEVVGFLEANKIDFKELDIAG-----DEDNRRWMRENVPGE-KKPQNGIPLPPQIFNEEQYCGDFDSFFSAKEENIIYSFLGLA SH3BGR (M) 1 MVIKVFVATSSGSIAIRKKQQEVVGFLEANKIDFKELDIAG-----DEDNRKWMRENVPGE-KKPQNGIPLPPQIFNEEQYCGDFDSFFSAKEENIIYSFLGLA SH3BGR (D) 1 MVLKVYVSGMSGNKEVKKRQQRVLMILDSKNIKYDTVDITEPGKESEKELMQNKSTSNGGTVSDPEPRHPLPPQLFNDDEYCGDYDAFDMANEIDTLEVFLKLA SH3BGRL2 (C) 1 MVIRVFVASSSGSVAIKKRQQDVVRFLEANRIEFEEVDITM-----SEEKRQWMYKNIPED-RQPAQGNPLPPQIFSDDRYCGDCFGRSVLTDIADARPVTCQ- SH3BGRL2 (H) 1 MVIRVFIASSSGFVAIKKKQQDVVRFLEANKIEFEEVDITM-----SEEQRQWMYKNVPPE-KKPTQGNPLPPQIFNGDRYCGDYDSFFESKESNTVFSFLGLK SH3BGRL2 (M) 1 MVVRVFVASCSGFVAIKKKQQDVVRFLEANKIEFEEVDITM-----SEEQRQWMYKNIPPE-KKPAQGNPLPPQIFNGDRYCGDYDSFFESKESNTVFSFLGLK SH3BGRL3 (C) 1 MSTLKVYSTSVTGSREIKSQQSEVTRILDGKNIKYELVDISQ-----DNALREEMRAKAG----NPK---AIPPQIVNGDHYCGDYELFVEAVEQNTLQEFLKLA SH3BGRL3 (H) 1 MSGLRVYSTSVTGSREIKSQQSEVTRILDGKRIQYQLVDISQ-----DNALRDEMRALAG----NPK---ATPPQIVNGDQYCGDYELFVEAVEQNTLQEFLKLA SH3BGRL3 (M) 1 MSGLRVYSTSVTGSREIKSQQSEVTRILDGKRIQYQLVDISQ-----DNALRDEMRTLAG----NPK---ATPPQIVNGNHYCGDYELFVEAVEQDTLQEFLKLA

SH3BGRL (C) 99 APPGS------KEAEALAKQQA------SH3BGRL (H) 99 APPGS------KEAEVQAKQQA------SH3BGRL (M) 99 APPGS------KEAEAQANQQA------SH3BGR (C) 99 PPPGT------KETEKSDTAGETEAHTFSLEQVFRKHKLEQPCQRKSL------PLKQTVTTFHACVHVNI------SH3BGR (H) 163 PPPDS------KGSEKAEEGGETEAQKEGSEDVGNLPEAQEKNEEEGE------TATEETEEIAMEGAEGEAEEEEET SH3BGR (M) 99 PPPGSKVTKSEEASSLPNGDVAGEAEGAAEGTEKAEKSGENEAQKEDSEDTGELSESQEKKEEEGEDGEEGEEGEEREEGGEGETTGETEEAPEEGAGGEAEEE SH3BGR (D) 104 PADTTAVS------TAQIELKQENGDAKKEEAETEAEDKKTEAGDGDVDVKEEAAEKAEV------SH3BGRL2 (C) ------SH3BGRL2 (H) 99 PRLASKAEP------SH3BGRL2 (M) 99 PRPA------SH3BGRL3 (C) 65 ------SH3BGRL3 (H) 65 ------SH3BGRL3 (M) 65 ------

SH3BGR (H) 228 AEGEEPGEDEDS------SH3BGR (M) 203 EPEEEAGEGEDS------

Figure 2 Sequence analysis of sh3bgrl.(a) The nucleotide sequence of the chicken sh3bgrl cDNA is shown with the 50 and 30 untranslated sequences in lower case letters and the ORF and corresponding protein sequences in upper case letters. The italicized sequence at the 30 end indicates the 89 bp fragment originally identiﬁed by differential display. The location of the polyadenylation signal is indicated by letters in bold. The primers used for 30 and 50 RACE are underlined. (b) Protein sequences for each SH3BGR family member from human (H), mouse (M), chicken (C), and Drosophila (D) were aligned using CLUSTALW. Conserved residues are shaded in black and amino acids with similar chemical properties are shaded in gray.

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 759 SH3BGR, SH3BGRL2, and SH3BGRL3. The BBSRC SH3BGR family members have different expression chickEST database was searched by BLAST analysis patterns in transformed cells using the human and mouse protein sequences in order To determine whether family members other than to identify the chicken orthologs of these family SH3BGRL exhibit expression patterns suggestive of a members (Figure 2b). The chicken SH3BGR protein is role in v-Rel-mediated transformation, their expression 157 amino acids long and shares 82% sequence identity patterns were analysed in cells following infection by with the mouse and human homologs. The published viruses expressing v-Rel and c-Rel. Since v-Rel is more human SH3BGR sequence used in the homology studies transforming than c-Rel, genes that are involved in the is 64 amino acids longer at the N-terminus than the transformation process are likely to exhibit unique or corresponding proteins from mouse and chicken. The more dramatic changes in expression in response to chicken SH3BGRL2 protein is 97 amino acids long and v-Rel, relative to c-Rel. Thus, CEFs and DT95 cells is about 83% identical to its human and mouse (a chicken B-cell line) were infected with retroviruses homologs. The SH3BGRL3 protein is 93 amino acids expressing c-Rel (REV-C) or v-Rel (REV-TW) or with long in all the three organisms examined. The chicken the helper virus chicken syntitial virus (CSV), and RNA SH3BGRL3 protein shares 90 and 88% identity with and protein were extracted 10–14 days post-infection the human and mouse proteins. As evident from the when v-Rel-expressing cells were morphologically trans- alignment, the SH3BGRL3 sequences are more diver- formed. Since the adaptation of primary cell cultures gent from the other members. All three SH3BGRL3 and transformed cell lines to conditions in vitro may homologs have a different N-terminus and lack the influence the expression patterns observed, the expres- conserved polyproline sequences that define potential sion of sh3bgr family members was also investigated in SH3- or EVH1-binding domains in the other SH3BGR tumors induced by v-Rel. Splenic tumors were induced proteins. in 1-day-old chicks by intraperitoneal injection with the To investigate whether there is a difference in the v-Rel-expressing retrovirus REV-TW. Control birds expression patterns of the chicken sh3bgr family were injected with the helper virus CSV. When the members, Northern blot analysis was performed using REV-TW-infected birds appeared moribund (7–10 days RNA isolated from tissues from 3-week-old chickens post-infection), protein and RNA were extracted from (Figure 3). The sizes of the sh3bgrl, sh3bgr, sh3bgrl2, and purified splenic lymphocytes from both groups of sh3bgrl3 transcripts are about 3.4, 1.4, 6 kb, and between chickens. The spleens of v-Rel-infected chickens were 1.3 and 0.8 kb, respectively. Similar to expression in approximately 50% larger than those of control birds human tissues, the endogenous expressions of sh3bgrl, and had visible tumors. The expression levels of the sh3bgrl2, and sh3bgrl3 were detected in all chicken sh3bgr family members in v-Rel-expressing CEFs, tissues examined, while the expression of sh3bgr was DT95 cells, and splenic tumor cells were analysed by restricted to muscle tissues (Scartenzzini et al., 1997; Northern blot analysis (Figure 4a). Overexpression of Egeo et al., 1998; Mazzocco et al., 2001, 2002). Both c-Rel in REV-C-infected cells, expression of v-Rel in sh3bgrl and sh3bgrl3 were highly expressed in lymphoid REV-TW-infected cells, and endogenous expression of tissues (spleen, bursa, and thymus), while exhibiting low SH3BGRL were confirmed by Western blot analysis levels of expression in the liver, skeletal muscle, and (Figure 4b). kidney. sh3bgrl2 was expressed highest in the brain and A good correlation was established between the lung, and relatively low levels of expression were expression of sh3bgrl RNA and protein levels in all the detected in the other tissues examined. cells examined. In both CEFs and DT95 cells, v-Rel was more efficient at downregulating the expression of sh3bgrl than was c-Rel. The levels of sh3bgrl RNA and protein were upregulated about two-fold in c-Rel- expressing CEFs, whereas they were downregulated three-fold in v-Rel-transformed CEFs, relative to control CSV-infected CEFs. In DT95 cells, sh3bgrl expression was downregulated to a greater extent by v-Rel (2.8-fold RNA and six-fold protein) than by c-Rel (2.2- fold RNA and about five-fold protein) relative to the CSV control. The expression of sh3bgrl was reduced about 2.5-fold in the v-Rel-induced splenic tumor cells at both the protein and RNA levels, compared to CSV- infected splenic cells. In contrast to sh3bgrl, the changes in expression of sh3bgrl2 and sh3bgrl3 were not consistent in response to v-Rel across the three cell types. sh3bgrl2 expression, although low in all three cell types, was unchanged in Figure 3 Expression of SH3BGR family members in chicken CEFs, elevated in DT95 cells, and reduced in the v-Rel- tissues. Northern blot analysis was performed using total RNA (10 mg) isolated from the indicated tissues from 3-week-old chickens induced splenic tumors. Expression of sh3bgrl3 was to determine expression of sh3bgrl, sh3bgr, sh3bgrl2, and sh3bgrl3. elevated in CEFs expressing v-Rel, reduced in DT95 Methylene blue-stained 18S rRNA is shown as the loading control. cells expressing v-Rel, and not significantly altered in the

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 760 were analysed in a panel of cell lines composed of an immortalized CEF cell line and lymphoid cell lines transformed by v-Rel or other oncogenes (Figure 4c). All of these cell lines expressed SH3BGRL, with MSB1 and BM2 expressing the highest amounts at both the RNA and protein levels. Relative to the other avian transformed cell lines, the expression of SH3BGRL was consistently lower in each of the v-Rel-transformed cell lines.

The promoter of SH3BGRL is bound by v-Rel in transformed cells Since the levels of sh3bgrl were consistently reduced in v-Rel-expressing cells, we examined whether this decreased expression may be a direct effect of v-Rel. v-Rel alters the expression of c-jun, ch-IAP1, and ikba by binding to kB sites in the promoters of these genes that reside in close proximity to their transcription start sites (Schatzle et al., 1995; Fujii et al., 1996; Kralova et al., 2002). To determine whether v-Rel may also directly alter the expression of sh3bgrl, the genomic sequences upstream of sh3bgrl were compared with the TRANS- FAC database using the transcription factor element search system (TESS) (Petsko, 2002). Two potential kB sites were identified 154 and 194 base pairs upstream of the transcription start site (Figure 5a). To evaluate whether these encode genuine kB sites that can be bound by v-Rel, DNA probes containing these sequences were used in electrophoretic mobility shift assays (EMSAs) with nuclear extracts from the v-Rel-transformed cell Figure 4 Expression of SH3BGR family members in transformed line 160/2 (Figure 5b). Supershift analyses using antisera cells. (a) Total RNA (10 mg) was isolated from CEFs (left panels) or specific for v-Rel demonstrated that three v-Rel-contain- DT95 cells (middle panels) infected with the helper virus CSV (H), ing complexes were able to bind to the kB site found or REV-based REV-C (C) or v-Rel (V). RNA was isolated from at 194. In contrast, no v-Rel-containing complexes primary splenic cells from chickens (right panels) infected with the À helper virus CSV (H) or with REV-TW (V). Northern blot analysis bound to the predicted kB site found at À154. was performed using probes specific for sh3bgrl, sh3bgrl2, and To ascertain whether v-Rel is bound to the promoter sh3bgrl3. Methylene blue-stained 18S rRNA (CEFs and DT95 of sh3bgrl in transformed cells, ChIP experiments were cells) or the hybridization of a probe specific for the 18S rRNA performed with antisera specific for v-Rel, normal rabbit (splenic cells) is shown as the loading control. (b) Protein from whole cell lysates (20 mg) from CEFs (left), DT95 cells (middle), or IgG, or no antibody (Figure 5c). DNA from these ChIPs splenic cells (right) in panel a were analysed by Western blot with was analysed by PCR with primers specific to the antiserum against SH3BGRL (top panels) or with antibody against sh3bgrl promoter (À156 to À467). Consistent with the Rel (bottom panels). (c) Protein (20 mg) from whole-cell lysates of ability of v-Rel to bind to the kB site at À194, the various chicken cell lines was analysed for the expression of sh3bgrl promoter was specifically amplified in ChIP SH3BGRL by Western blot (top panel). Total RNA (10 mg) from these cell lines was analysed for the expression of sh3bgrl by assays performed with v-Rel-specific antisera, but not Northern blot (middle panel). Methylene blue-stained 26S rRNA is with normal rabbit IgG or those performed in the shown as the loading control (bottom panel). The cell lines absence of antibody. Primers flanking a region of the included an immortalized CEF cell line (DF-1), two B-cell lines alpha-2 type I collagen (COLA) gene that is devoid of transformed by avian leukosis virus (DT40 and DT95), two T-cell lines transformed by Marek’s disease virus (MSB1 and RP1), a kB sites served as a negative control for these experi- macrophage-like cell line transformed by avian myeloblastosis ments. Although these sequences were amplified by PCR virus (BM2), a chicken erythroid cell line transformed by avian using control input DNA, significant amplification of erythroblastosis virus (AEV) and three v-Rel-transformed cell lines. these sequences was not observed in the various ChIP The v-Rel-transformed cell lines included a B-cell line (123/12), a samples. T-cell line (160/2), and a non-B/non-T-cell line (123/6T).

Overexpression of SH3BGRL inhibits v-Rel transformation splenic tumors. sh3bgr expression was undetectable in The levels of sh3bgrl RNA and protein were consistently both CEFs and DT95 cells (data not shown). reduced in cells expressing v-Rel and in v-Rel-induced To determine whether the decreased expression of splenic tumors, relative to control cells (Figure 4). To SH3BGRL was speciﬁc to the v-Rel transformation determine if the reduced expression of sh3bgrl con- pathway, the RNA and protein levels of SH3BGRL tributes to v-Rel-mediated transformation, SH3BGRL

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 761 was overexpressed with v-Rel in spleen cell transforma- fewer colonies were formed when SH3BGRL was tion assays. Viral constructs were made using a coexpressed with v-Rel than when v-Rel was expressed bicistronic spleen necrosis virus-based retroviral vector alone (Table 1). This result indicates that the over- pBIS, which allowed for expression of up to two genes expression of SH3BGRL significantly (Pp0.03) inhibits from a single virus (Figure 6a). Viruses were constructed the initiation of v-Rel transformation. to individually express v-Rel from the splice acceptor Since lower expression levels of SH3BGRL were site (BIS-V) or SH3BGRL from the internal ribosomal observed in v-Rel-transformed cell lines compared to entry site (BIS-BGRL), or to coexpress both v-Rel and cell lines transformed by other oncogenes (Figure 5c), SH3BGRL (BIS-V þ BGRL). The overexpression of the effect of SH3BGRL overexpression on the main- SH3BGRL and the expression of v-Rel by the appro- tenance of v-Rel transformation was analysed. The priate viruses were confirmed by Western blot analysis sh3bgrl gene was cloned into the pTZDS-XB retroviral (Figure 6b). Primary splenic lymphocytes from 3-week- vector to produce viruses (DS-BGRL) capable of old chickens were infected with these viruses and plated infecting the v-Rel-transformed cell lines. The DS in soft agar. Plates were scored for colony forma- viruses were used to infect (multiplicity of infection tion after 10 days. The results from three individual (MOI) ¼ 7) three different v-Rel-transformed lymphoid experiments demonstrated that approximately 75% cell lines:160/2 (a T-cell line), 123/12 (a B-cell line), and 123/6T (a non-B/non-T-cell line). Whole-cell lysates were prepared 5 days after infection and expression of v-Rel and overexpression of SH3BGRL was confirmed by Western blot analysis (Figure 7a). Cells were plated in media containing soft agar 7 days after infection and scored for colony formation after a further 10–14 days (Figure 7b). Colony counts revealed a 36% reduction (Pp0.03) in 160/2 cells, a 28% reduction (Pp0.01) in 123/12 cells, and a 31% reduction (Pp0.01) in 123/6T

Figure 5 Analysis of v-Rel binding to the promoter of sh3bgrl. (a) A Schematic representation of the first exon and 700 base pairs upstream of the transcription start site (arrow) of sh3bgrl are shown. The relative positions of two potential kB sites are indicated. (b) EMSAs were performed using nuclear extracts from Figure 6 Expression of v-Rel and SH3BGRL from bicistronic the v-Rel-transformed cell line 160/2 and 32P-labeled probes retroviruses. (a) The genomic structures of the retroviruses used to containing the potential kB sites identified above. Supershift coexpress v-Rel and SH3BGRL are shown. The parental bicis- analyses were performed by the addition of normal rabbit serum tronic SNV-based virus (BIS) contains 50 and 30 long-terminal or serum specific for v-Rel. The three DNA-binding complexes that repeats (LTRs), a Neomycin resistance cassette (NeoR), a splice were shifted by the addition of v-Rel-specific antiserum are donor (SD), and SA site used to generate subgenomic viral mRNA, numbered. (c) ChIP assays were performed with crosslinked and an internal ribosome entry site (IRES) that is necessary for DNA from 160/2 cells. DNA–protein complexes were immuno- translation of the GFP. In addition, the virus contains a MCS precipitated with normal rabbit IgG or serum specific for v-Rel. downstream of the SA. Additional viral constructs were created by PCR was performed using control input DNA, no template DNA, the insertion of v-rel into the MCS (BIS-V) or by replacing GFP or DNA recovered from ChIPs performed in the absence of with sh3bgrl (BIS-BGRL). A virus expressing both v-Rel and antibody (No IgG), normal rabbit IgG, or antisera specific for SH3BGRL (BIS-V þ BGRL) was also constructed. (b) Protein v-Rel. The products of PCR performed with primers specific for the extracts (20 mg) from CEFs expressing the viruses described in sh3bgrl promoter (top panel) or a region of the alpha-2 type I panel a were analysed for SH3BGRL expression (top panel) and collagen gene (COLA) (bottom panel) are shown. v-Rel expression (bottom panel) by Western blot.

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 762 Table 1 Effect of SH3BGRL overexpression on v-Rel-mediated are strongly conserved (Figure 2b). The proline-rich transformation of splenic lymphocytes domains include a predicted class I SH3-binding domain Retrovirusa Number of coloniesb that has a conserved core PXXP motif (amino acids 64–67; X is any amino acid) (Mayer, 2001). A proline is Exp. 1 Exp. 2 Exp. 3 Avg. (7s.d.) also found at amino acid 59, which has been observed BIS 0 0 0 0 in certain SH3-binding domains (Tong et al., 2001). BIS-BGRL 0 0 0 0 A predicted class II EVH1 recognition domain has BIS-V 531 549 728 603 (7109) also been found that contains the conserved PPXXF BIS-V+BGRL 189 99 136 141 (745) sequence (amino acids 66–70; X is any amino acid) (Tu et al., 1998; Barzik et al., 2001). An aspartic acid at aBicistronic retroviruses expressing v-Rel and SH3BGRL alone or in combination were used to infect splenic lymphocytes. bInfected splenic amino acid 72 is conserved in a subset of EVH1-binding lymphocytes were plated in soft agar and colonies scored micro- domains (Barzik et al., 2001; Ball et al., 2002). To scopically after 10 days. determine whether the ability of SH3BGRL to inhibit v-Rel transformation requires intact SH3- and EVH1- binding domains, mutations were introduced into these conserved regions (Figure 8a). Two mutants (SH3BGRLm1 and SH3BGRLm2) were specific for the SH3-binding domain (P59A and P64A), one mutant (SH3BGRLm3) contained two mutations in the EVH1- binding domain (F70A, D72A), and two mutants (SH3BGRLm4 and SH3BGRLm5) disrupted both SH3- and EVH1-binding domains (P67A and P64A, P67A). Similar mutations in other proteins have been demonstrated to disrupt interactions between SH3 and EVH1 domains and known binding partners (Grabs et al., 1997; Barzik et al., 2001). Viruses were generated using the pBIS bicistronic retroviral vector described above to enable expression of each SH3BGRL mutant protein alone or with v-Rel. The overexpression of SH3BGRL, SH3BGRL mutants, and the expression of v-Rel from these viruses were confirmed by Western blot analysis (Figure 8b). Each of the SH3BGRL mutants was expressed at levels com- Figure 7 Overexpression of SH3BGRL in v-Rel-transformed cell parable to that of ectopically expressed wild-type lines. (a) Whole-cell lysates were prepared from v-Rel-transformed SH3BGRL. Transformation assays were performed by cell lines that were uninfected (lanes 1, 4 and 7), infected with a infection of primary splenic lymphocytes from 3-week- control DS virus (lanes 2, 5 and 8) or infected with DS-BGRL virus old chickens. The cells were plated in soft agar after expressing SH3BGRL (lanes 3, 6 and 9). The v-Rel-transformed cell lines included the T-cell line 160/2 (lanes 1–3), the B-cell line 3 days and plates were scored for colonies after a further 123/12 (lanes 4–6), and the non-B/non-T-cell line 123/6T (lanes 14–16 days (Figure 8c). Coexpression of wild-type 7–9). Protein extracts (20 mg) were analysed by Western blot for SH3BGRL with v-Rel reduced colony formation by expression of SH3BGRL (top panel) and v-Rel (bottom panel). 62% (Pp0.00001), relative to expression of v-Rel alone. (b) The cells described above were plated in soft agar and scored for Expression of SH3BGRLm1 or m2, which have amino colony formation after 14 days. For each cell line, the number of colonies formed by uninfected cells (black bars) was standardized acids mutated in the predicted SH3-binding domain, to 100. The number of colonies formed by DS- (striped bars) or resulted in little or no change in colony numbers when DS-BGRL- (gray bars) infected cells were normalized based on this expressed with v-Rel. The expression of SH3BGRLm3, standard. The average and standard deviation of three separate which has mutations in the predicted EVH1-binding experiments is shown. domain, reduced colony formation by v-Rel by 48% (Pp0.03), a level comparable to that observed with colony numbers due to DS-BGRL infection when wild-type SH3BGRL. However, the ability of the compared to infection with the empty DS vector. This EVH1-binding domain mutant of SH3BGRL to inhibit difference in colony formation is not due to an effect on colony formation is significantly reduced (Pp0.001) the growth rate of the cells since cell counts during this relative to that of the wild-type protein. Expression of period did not reveal significant differences in growth SH3BGRLm4 or m5, which have mutations overlapping between infected and uninfected cells for each cell line both domains, resulted in colony numbers comparable (data not shown). to m1 or m2, but not m3.

The SH3-binding domain is important for SH3BGRL to Discussion function as an inhibitor of v-Rel transformation SH3BGRL proteins from human, chicken, and mouse v-Rel is a constitutively active member of the Rel/NF- share a region of overlapping proline-rich domains that kB family of transcription factors and transforms cells

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 763 by the inappropriate activation or suppression of genes normally under the regulation of Rel/NF-kB proteins. The majority of target genes of v-Rel identified to date are upregulated in response to direct or indirect regulation by v-Rel. Evidence has been provided for the involvement of only a few of these genes in the transformation of cells by v-Rel. However, none of the v-Rel downregulated genes identified so far have been biologically implicated in v-Rel transformation. In this study, we identify a novel downregulated gene in v-Rel- transformed cells, sh3bgrl, the suppression of which is important for v-Rel transformation. SH3BGRL was identified in differential display experiments to be downregulated in v-Rel-transformed CEFs. Subsequently, we have shown that RNA and protein levels of SH3BGRL are decreased in v-Rel- transformed CEFs, in DT95 cells expressing v-Rel, and in v-Rel-induced splenic tumor cells (Figures 1 and 4). Since v-Rel binds to a kB site in the promoter of sh3bgrl and is associated with this promoter in transformed cells (Figure 5), it is likely that v-Rel directly participates in its decreased expression. In contrast to SH3BGRL, most of the other SH3BGR family members did not exhibit patterns of expression that were consistent with a role in v-Rel-mediated transformation. The expression of sh3bgr was not detected in cell types transformed by v-Rel (Figures 3 and 4) and the expression of sh3bgrl2, while decreased in v-Rel-induced splenic tumors, was not significantly altered in CEFs or DT95 cells expressing v-Rel (Figure 4). Although the changes in sh3bgrl3 expression were not consistent, they were dramatic and in proportion to the oncogenic potential of the Rel proteins expressed in CEFs and DT95 cells. An increase in sh3bgrl3 expression was observed in CEFs expressing v-Rel, and to a lesser extent in CEFs overexpressing c-Rel. In contrast, the expression of v-Rel or c-Rel in DT95 cells resulted in a decrease in sh3bgrl3 levels. sh3bgrl3 is the only member of the sh3bgr family for which a potential biological role has been characterized (Berleth et al., 1999). The initial identification of SH3BGRL3 over- lapped with another study that characterized this protein as TNF inhibitory protein B1 (TIP-B1). Figure 8 Characterization of proline-rich sequences in the SH3BGRL3 (TIP-B1) was shown to be upregulated in function of SH3BGRL. (a) Mutations were introduced into response to TNFa and was able to block TNF-induced the SH3- and EVH1-binding domains of SH3BGRL as shown. The mutation in SH3BGRLm1 disrupts a proline five amino acids apoptosis when added exogenously to human fibroblast upstream of the core SH3-binding domain. The mutation in cell lines (Berleth et al., 1999, 2000; Henn et al., 2001). SH3BGRLm2 is specific for the SH3-binding domain. The Since these studies did not show that TIP-B1 was mutations in SH3BGRLm3 disrupt conserved amino acids in the secreted by cells, or that exogenously added TIP-B1 was EVH1-binding domain. Mutations in SH3BGRLm4 and m5 disrupt conserved amino acids in both the SH3- and EVH1- able to enter cells, further experiments are necessary to binding domains. (b) Protein extracts (20 mg) from CEFs infected determine whether the inhibition of TNF-induced with bicistronic retroviral vectors encoding each form of apoptosis is a normal biological function of SH3BGRL3. SH3BGRL alone or together with v-Rel were analysed by Western Interestingly, a separate study showed that induction of blot for expression of SH3BGRL (top panel) and v-Rel (bottom v-Rel in a tetracycline-regulated system blocked TNF- panel). (c) The viruses used in panel b were used to infect splenic lymphocytes from 3-week-old chickens. The cells were plated in induced apoptosis of HeLa cells (Zong et al., 1997). soft agar and scored for colony formation after 14–16 days. The Thus, the upregulation of SH3BGRL3 (TIP-B1) ob- colony-forming ability of the different viruses is shown relative to served in CEFs expressing v-Rel may play a role in this the colony-forming ability of cells infected with BIS-V (100%). The cell type. Future experiments will investigate the results presented here are average values from three different experiments (m1 and m2) or from six individual experiments (m3, significance of the cell type-specific expression of m4, and m5). Cells infected with viruses not expressing v-Rel did sh3bgrl3 in v-Rel-mediated transformation. In addition, not form colonies and are not represented on the graph. the strong homology between SH3BGRL3 and the other

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 764 members of the SH3BGR family argue for the which is conserved among all the SH3BGR members, possibility of their involvement in the TNF signaling and Pro64 disrupted the predicted SH3-binding domain. pathway. Mutations in Phe70 and Asp72 disrupted the EVH1- Human SH3BGRL2 and SH3BGRL3 proteins loca- binding domain. Splenic lymphocytes coexpressing lized to both the nucleus and cytoplasm in COS-7 cells v-Rel and SH3BGRL mutants with disrupted SH3- (Mazzocco et al., 2001, 2002). To analyse the subcellular binding domains formed colonies equal to or greater in distribution of chicken SH3BGRL, we expressed number than v-Rel alone. Thus, the SH3-binding SH3BGRL containing a C-terminal FLAG-tag in domain appears to be critical for SH3BGRL to function CEFs. The localization of this protein was analysed by as an inhibitor of v-Rel transformation. The mutant immunofluorescence and found to mimic that of the with the disrupted EVH1-binding domain, however, other family members (data not shown). This pattern maintained most of wild-type SH3BGRL function by was unaffected by v-Rel since SH3BGRL-FLAG was inhibiting v-Rel transformation by 48%. Since expres- also observed in both the nucleus and the cytoplasm in sion of this mutant form did not completely inhibit v-Rel-transformed CEFs (data not shown). These colony formation to the extent of wild-type SH3BGRL, results suggest that while v-Rel efficiently downregulates it is possible that interactions between SH3BGRL and the expression of SH3BGRL, it does not alter the an EVH1 domain-containing protein play an additional subcellular distribution of SH3BGRL. minor role in the ability of SH3BGRL to inhibit v-Rel Transformation assays performed in this study transformation. However, it is also possible that the demonstrate that the downregulated expression of F70A and D72A mutations caused weakened interac- SH3BGRL plays a significant function in v-Rel trans- tions between the neighboring SH3-binding domain and formation. Coexpression of SH3BGRL and v-Rel interacting SH3 domain-containing proteins. diminished the transforming ability of v-Rel by 76% The characterization of genes regulated by v-Rel in in splenic lymphocytes (Table 1). This reduction is not transformation is critical to understanding the mechan- the result of a change in target cell specificity since 70% isms by which v-Rel efficiently transforms cells. The of the cell lines established from colonies transformed by deregulated expression of some v-Rel target genes (AP- the coexpression of v-Rel with SH3BGRL were of T-cell 1, IRF-4, and IAP1) has been implicated in human origin (data not shown). These findings are consistent malignancies. Microarray analysis has revealed that with previous reports describing the target cell specificity SH3BGRL expression is downregulated in androgen- of v-Rel (Barth et al., 1990). Therefore, these combined independent human cell lines derived from prostate results indicate that the initiation of transformation by tumors (Shi et al., 2004). Interestingly, the activation of v-Rel was inhibited by SH3BGRL in 76% of the cells Rel/NF-kB temporally correlates with the progression infected. of prostate tumor cells to androgen independence, at In addition to its ability to initiate cell transforma- which point they become refractory to anticancer tion, the continued expression of v-Rel is also required therapies (Thalmann et al., 2000; Gasparian et al., to maintain the transformed phenotype (White et al., 2002; Suh et al., 2002; Suh and Rabson, 2004; Zhou 1995). The v-Rel-regulated genes required for the et al., 2004). Furthermore, recent studies have demon- initiation of transformation may or may not be required strated that in human breast tumor cell lines, the for the maintenance step. To assay its participation in expression of SH3BGRL is also suppressed relative to maintenance of v-Rel transformation, the effect of normal mammary epithelial cells (Mecham et al., 2004). SH3BGRL overexpression on the colony-forming The role of elevated c-Rel expression and activity has ability of v-Rel-transformed cell lines was analysed. been documented in the development of breast cancer Overexpression of SH3BGRL resulted in an average (Sovak et al., 1997; Cogswell et al., 2000; Cao and reduction in colony numbers of 32% for each cell Karin, 2003; Romieu-Mourez et al., 2003). Although the line tested (Figure 7). Thus, it appears likely that the suppression of SH3BGRL may be a common event in pathways impacted by SH3BGRL play a greater role in the development of cancers induced by aberrant Rel/ initiating v-Rel transformation rather than in maintain- NF-kB activity, the inappropriate activation of Rel/NF- ing the transformed phenotype. Since the exact biolo- kB proteins is not always sufficient to cause a reduction gical function(s) of SH3BGRL is unknown, it remains in the levels of sh3bgrl. While v-Rel suppressed sh3bgrl to be seen whether SH3BGRL is a positive regulator expression in all cell types examined, the overexpression of a pathway that inhibits v-Rel transformation or of the less oncogenic c-Rel had differential effects on a negative regulator of one that promotes v-Rel sh3bgrl levels in lymphocytes and fibroblasts (Figure 4). transformation. It is likely that the ability of a specific Rel/NF-kB Based on the homology between SH3BGRL and complex to suppress sh3bgrl expression enhances its other family members, we analysed the requirement of ability to transform a particular cell type. protein interaction domains of SH3BGRL for initiation of v-Rel transformation. A proline-rich region Materials and methods PASGNPLPPRLFND (amino acids 59–72) containing two potential protein interaction domains was identified General cell culture techniques in SH3BGRL (Egeo et al., 1998). Mutations were made CEFs were prepared from 10–11-day-old embryos (Charles in either the SH3- (PASGNPLPP) or the EVH1- River SPAFAS, Wilmington, MA, USA) and grown in (PPRLFND) binding domains. Mutations in Pro59, Dulbecco’s modified Eagle’s medium (DMEM) supplemented

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 765 with 5% chicken serum (Invitrogen Life Technologies, 4.7% glycerol, 0.05% bromophenol blue, 53.5% deionized Carlsbad, CA, USA), 5% fetal bovine serum (Atlanta formamide and 7.9% formaldehyde), denatured at 651C for Biologicals, Atlanta, GA, USA), and 1% Antibiotic-Anti- 20 min, then loaded on a 1.0% agarose gel containing 2.1% mycotic (Invitrogen Life Technologies) in 8% CO2 at 371C. formaldehyde for electrophoretic separation at 120 V for 2 h. Transformed chicken cell lines used in this study were RNA was transferred overnight onto a Hybondt-N þ nylon maintained under the same conditions described above and membrane (Amersham Biosciences, Piscataway, NJ, USA). included an immortalized CEF cell line (DF-1), two B-cell lines The membrane was dried and RNA was crosslinked with UV transformed by avian leukosis virus (DT40 and DT95), two at 1.2 Â 105 mJ/cm2 using the FB-UVXL-1000 UV Crosslinker T-cell lines transformed by Marek’s disease virus (MSB1 and (Fisher Scientific, Hampton, NH, USA). Ribosomal RNA was RP1), a macrophage-like cell line transformed by avian visualized by staining the membrane with 0.02% Methylene myeloblastosis virus (BM2), and a chicken erythroid cell line blue in 0.3 M sodium acetate. transformed by avian erythroblastosis virus (AEV) (Akiyama Northern blot analysis was carried out by hybridizing the and Kato, 1974; Moscovici et al., 1982; Baba et al., 1985; Rao membrane with 32P-labeled probes specific for sh3bgr, sh3bgrl, et al., 1990). The v-Rel-transformed cell lines included a B-cell sh3bgrl2, sh3bgrl3, or 18S rRNA using ULTRAhyb (Ambion) line (123/12), a T-cell line (160/2), and a non-B/non-T-cell line hybridization buffer according to the manufacturer’s instruc- (123/6T) (Hrdlickova´ et al., 1994). tions. Chicken ESTs encoding the complete ORFs of sh3bgr For induction of splenic tumors in vivo, 1-day-old chickens (ChEST862I9), sh3bgrl2 (ChEST971A7), and sh3bgrl3 were injected intraperitoneally with the CSV or the reticulo- (ChEST894H8) were obtained from the BBSRC ChickEST endotheliosis virus (REV)-based retrovirus expressing v-Rel Database (Swindon, UK) for generating the DNA probes. The (REV-TW). Spleens were extracted from both groups of 345-bp-long DNA sequence encoding the sh3bgrl ORF was chickens 7–10 days after injection or at the first sign of ill- used as the probe to detect sh3bgrl expression. The cDNA ness in each REV-TW-infected bird. Cell suspensions were probes were labeled with [a-32P]dATP using the StripEZ DNA prepared by macerating the spleens through a nylon screen, kit (Ambion). A 20-mer oligonucleotide (50-CGGAACTAC and lymphocytes were isolated by differential gradient GACGGTATCTG-30) specific for the chicken 18S rRNA was centrifugation using Histopaque-1077 (Sigma-Aldrich, St end-labeled with [g-32P]ATP to be used as a loading control for Louis, MO, USA). Northern blot analysis utilizing RNA from splenic tumors (Deindl, 2001). Cloning of sh3bgrl and site-directed mutagenesis Differential display and cloning of a full-length sh3bgrl cDNA Western blot analysis were performed using techniques previously described (You Proteins from cell extracts were resolved by sodium dodecyl and Bose, 1998). Briefly, 10 mg of total RNA was purified using sulfate–polyacrylamide gel electrophoresis (SDS–PAGE). the RNeasy kit (QIAGEN, Valencia, CA, USA) from normal For the analysis of SH3BGRL, proteins were resolved on CEFs, CEFs transfected with empty Rous sarcoma virus- 14% tricine gels (Schagger and Jagow, 1987). The expression derived vector (RCAS), or CEFs transformed by RCAS- of c-Rel and v-Rel was detected by resolving proteins on expressing v-Rel. RNA differential display was carried out 10% glycine gels (Sambrook et al., 1989). Proteins were elec- according to the RNAimaget kit (GenHunter, Nashville, TN, trophoretically transferred to nitrocellulose membranes. USA). The 89 bp fragment identified by differential display SH3BGRL was detected by using a rabbit polyclonal was cloned and used to screen a chicken CEF cDNA library. antiserum raised against a GST fusion protein expressing the cDNAs that contained poly(A) tails were extended by RACE complete ORF of SH3BGRL. A peroxidase-conjugated using oligonucleotides designed identical to the sequence donkey anti-rabbit IgG (Jackson ImmunoResearch Labora- 105 bp downstream of the 50 end of the longest isolated cDNA tories, West Grove, PA, USA) was used as the secondary (50-TAACTACAGCAAGGACTAATTACC-30) or identical antibody. For detection of v-Rel and c-Rel, a 1:2 dilution of to the 50 end (50-GTGCCCTGAAGGATAAGAAACAGG-30) hybridoma supernatant fluids containing the mouse mono- to obtain a cDNA encoding the full-length ORF. clonal antibody HY87 was employed (Schwartz and Witte, Point mutations were made in the sh3bgrl ORF using 1988). The secondary antibody used was a peroxidase- 0 the QuikChange Site-Directed Mutagenesis Kit (Stratagene, conjugated F(Ab )2 fragment donkey anti-mouse IgG (Jackson La Jolla, CA, USA) using the following primers and their ImmunoResearch Laboratories). complementary sequences:P59Amut 5 0-GAACGTCCCTGA AGACAGGCGGgCAGCCAGCGGGAAC-30, P64Amut 50- CAGCCAGCGGGAACgCGCTGCCACCCCGGCTCTTC Electrophoretic mobility shift assays AAC-30, EVH1mut 50-CTGCCACCCCGGCTCgcCAACGc Nuclear extracts were prepared from 160/2 cells using a CAGCCGCTA-30, P67Amut 50-CCAGCGGGAACCCGCT modified NP-40 lysis protocol (Dyer and Herzog, 1995). GCCAgCCCGGCTCTT-30, and P64A/P67Amut 50-CGG EMSA reactions were performed using 3 or 6 mg of protein in CCAGCCAGCGGGAACgCGCTGCCAgCCCGGCTCTT-30. binding reactions consisting of 25 mM Tris pH 7.5, 75 mM Mutations (lower case letters) were confirmed by sequencing NaCl, 1.5 mM EDTA, 7.5% glycerol, 0.3% NP-40, 1.5 mM (DNA Core Facility, The University of Texas at Austin, DTT, 1 mg poly [dI Á dC], and 1.2 mg BSA. For supershift Austin, TX, USA). analyses, 5 mg of normal rabbit serum or antiserum specific for v-Rel (anti-R5) was added to the reactions and incubated on ice for 45 min (Kralova et al., 2002). Labeled oligonucleotide Northern blot analysis probe (50 000 cpm) was added to the reactions and incubated Cells grown in culture and tissues isolated from chickens were at room temperature for 20 min. EMSA reactions were harvested for RNA using RNAwiz (Ambion, Austin, TX, electrophoretically separated on a 5% acrylamide gel as USA) according to the manufacturer’s instructions. Each described previously (Liss and Bose, 2002). sample was normalized to a final concentration of 1 mg/mlin The top strands of the oligodeoxynucleotides used as probes RNase-free water. Samples were diluted 1:2 in 2 Â loading for EMSAs were as follows: kB-154 (50-GTTCGTCGA buffer (2 Â MOPS (4.62% w/v MOPS (Sigma-Aldrich), 1.1% GAATTCGCCTCACCCCTTTCTTTC-30) and kB-194 (50- w/v sodium acetate trihydrate, and 0.42% w/v EDTA, pH 7.0), GCAGCGGGCAGCCCCCTCCTCCGAGCACGTC-30). The

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 766 underlined sequences indicate the location of the predicted kB DS-based retroviruses were constructed by cloning the sites. The bottom strands of the probe were synthesized by sh3bgrl gene into the pTZDS-XB vector (containing the 30 annealing a 9-bp oligonucleotide complementary to the 30 end portion of env and one LTR) within the XhoI/BssHII sites of the top strand and extending with Klenow in the presence of (Hrdlickova´ et al., 2001). This construct was then linearized [32P] dCTP (Hrdlickova´ et al., 1995a). by digestion with SalI and ligated to the SalI-digested pSR- REP-A vector (containing one LTR, gag, pol, and the 50 portion of env). The ligation reaction was then directly used for Chromatin immunoprecipitations transfection of CEF cultures. ChIP experiments were performed as described by Upstate Biotechnology (Chrlottesville, VA, USA), with some modifica- tions. 160/2 cells (3 Â 107) were fixed in 1% formaldehyde for Preparation of retroviral stocks 10 min at 371C. Glycine was added to a final concentration of All retroviral stocks were prepared by transfecting CEFs with 0.125 mM and cells were incubated for 10 min at 371C to stop the appropriate viral DNA as described before (Chen and the crosslinking. Fixed cells were washed in PBS and nuclei Okayama, 1987). CEFs were plated in 60 mm tissue culture isolated (Dyer and Herzog, 1995). Nuclei were lysed in 300 ml dishes at 6 Â 105 cells/ml in serum-containing media and SDS lysis buffer (1% SDS, 10 mM EDTA, and 50 mM Tris, pH equilibrated at 361C and 3% CO2 at least 30 min prior to 8.0) and sonicated for 15 s three times. A volume of lysate transfection. Viral DNA (10 mg) together with the pCSV11S3 equivalent to 1 Â 106 cells was diluted to 200 ml and precleared plasmid (0.3 mg, encoding the CSV helper virus) was mixed for 1 h by incubation with 50 ml of protein-G agarose (50% with 200 ml of 0.25 M CaCl2, to which an equal volume of N,N- slurry) that had been blocked with salmon sperm DNA and bis[2-hydroxyethyl]-2-aminoethanesulfonic acid (BES)-buf- BSA. Cleared lysates were incubated for 1 h with 2 mgof fered saline solution (50 mM BES, 280 mM NaCl, and 1.5 mM normal rabbit IgG (Upstate Biotechnology) or with IgG- Na2HPO4) was added. After 2 min, the DNA mixture was purified antibody specific for amino acids 488–503 of v-Rel added to the cells, which were then incubated at 361C and 3% (Hrdlickova´ et al., 1995a). CO2 overnight. For BIS-based retrovirus production, CEFs DNA recovered from the ChIPs was used in PCR with were treated with 0.5 mg/ml G418 (Fisher Scientific) after the primers specific for sequences between À156 and À467 of second passage and virus was harvested 10 days later. DS- the sh3bgrl promoter (sense primer:5 0-TGCCTCACACA based viruses were harvested 6–7 days post-transfection. All CGCTGTATTTCAC-30 and antisense primer:5 0-ACGAAC viral stocks were aliquoted and stored at À801C. CACGCGAGACGTGC-30), or that amplify a 100 base pair The viral harvests were titered by dot-blot hybridization region of exon 6 of the alpha-2 type I collagen (COLA) gene analysis using Hybondt-N þ nylon membrane (Amersham (accession number:M25961) (sense primer:5 0-AAAGGGT Biosciences) (Nelson et al., 1998). Stocks of viruses with titers CATGTGTACTGAG-30 and antisense primer:5 0-AGAAAG determined by immunohistochemical assays were employed as CACTGATAACCCTC-30). PCR was performed using 20 ng controls. Viral stocks (180 ml) were added to the membrane of input DNA or 10% of the recovered DNA from each ChIP using a dot-blot transfer unit. Membranes were washed twice and products resolved by agarose (sh3bgrl) or polyacrylamide with 10 Â SSC, dried, and viral RNA crosslinked with UV at (COLA) gel electrophoresis. 1.2 Â 105 mJ/CM2 using the FB-UVXL-1000 UV Crosslinker (Fisher Scientific). The membranes were hybridized with the appropriate probe in ULTRAhyb (Ambion) hybridization Construction of retroviral vectors buffer according to manufacturer’s instructions. For titration The pBIS bicistronic vector was derived from the spleen of BIS-based retroviruses, a neomycin resistance gene-specific necrosis virus-based retroviral vector pJD216NeoHy (Dough- probe was generated and randomly labeled with [a-32P]dCTP erty and Temin, 1986). This vector contains two LTRs flanking (Prime-A-Gene Labeling System, Promega Corporation, a neomycin resistance gene, a REV-A SA site, and a Madison, WI, USA). For titration of the DS-based retro- Sal Nco hygromycin resistance gene. A 0.6 kB I– I fragment viruses, an 875-nucleotide long fragment was generated by containing the internal ribosome entry site (IRES) from the digesting the pTZDS-XB vector with the SmaI and XhoI encephalomyocarditis virus was isolated from pCITE-1 0 restriction enzymes. This fragment, which includes the ORF (Novagen, Madison, WI, USA) and cloned at the 3 end of encoding the envelope protein of the virus, was randomly the hygromycin resistance gene. A 0.7 kB NcoI–NotI fragment labeled with [a-32P]dCTP (Prime-A-Gene Labelling System, encoding EGFP was isolated from pEGFP-1 (Clontech, Palo 0 Promega). The hybridized probe was detected the next day and Alto, CA, USA) and cloned at the 3 end of the IRES to create levels were normalized to the known titers of the control pJZ466 (kindly provided by Jiayou Zhang). The pBIS vector viruses by using ImageQuant 1.0. was constructed by replacing the hygromycin-resistance cassette with a multiple cloning site (MCS) that included XhoI and BssHII recognition sequences. The ORF encoding v-Rel In vitro transformation assays was cloned into these sites to form the pBIS-V vector. The In vitro transformation assays were performed using splenic pIRES adapter vector was created by cloning a SalI–NotI lymphocytes isolated from 3-week-old chickens. Spleens were fragment that contained the IRES and green fluorescent macerated through a nylon screen, and lymphocytes were protein (GFP) from pBIS into pBluescript SK that had been isolated by differential gradient centrifugation with Histopa- digested with SalI and NotI. The sequences encoding GFP que-1077 (Sigma-Aldrich, St Louis, MO, USA). Purified were excised by digestion with NcoI and NotI and replaced lymphocytes were resuspended at a density of 2 Â 108 cells/ml with a polylinker containing restriction sites for NcoI, XbaI, and infected with viruses having a minimum titer of 1 Â 105 EcoRV, and NotI. The wild-type or mutant sh3bgrl genes infectious units/ml. All viruses were diluted to the same titer in were cloned into the NcoI and XbaI sites in the pIRES a final volume of 5 ml before the addition of 250 ml of the cell adapter vector. These genes were excised out together with suspension. Cells infected with BIS-based retroviruses were the upstream IRES sequence by digestions with SalI and resuspended in fresh media (15% fetal bovine serum and 5% NotI. Each excised fragment was then cloned into SalI/NotI- chicken serum) 24 h after infection. After an additional 48 h, digested pBIS and pBIS-V vectors for making BIS-based the cells were pelleted by centrifugation and resuspended in retroviruses. 12 ml of plating media prewarmed at 421C. The plating media

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 767 consisted of DMEM containing 15% fetal bovine serum, 5% After 7 days, cell viability was determined by trypan blue chicken serum, 1% Antibiotic-Antimycotic (Invitrogen Life exclusion. The 123/12 and 160/2 cells were plated in 0.35% soft Technologies), and 0.35% Noble agar (Becton Dickinson, agar at a concentration of 104 cells/plate, while the slower- Sparks, MD, USA). The cell suspension was divided equally growing 123/6T cells were plated at 2 Â 104 cells/plate. The soft into three 60 mm tissue culture plates, allowed to solidify, and agar plating media consisted of DMEM containing 5% fetal then incubated at 371C and 8% CO2. Plates from each group bovine serum, 5% chicken serum, 1% Antibiotic-Antimycotic of infections were scored microscopically for colony formation (Invitrogen Life Technologies), and 0.35% Noble agar (Becton after 10–16 days. Dickinson). The plates were incubated for 10–14 days before colonies were counted. Colony formation assays of v-Rel-transformed cell lines v-Rel-transformed cell lines were infected with empty DS Acknowledgements retroviruses or DS retroviruses expressing SH3BGRL at a We thank William Bargmann, Radmila Hrdlickova´ , Jarmila MOI of 6–10. Cells were maintained in complete medium and Kralova, Jiri Nehyba, and Emin Ulug for technical advice and routinely split every two days. Uninfected cells from each cell critical reading of the manuscript. This work was supported by line were grown in parallel. Ectopic expression of SH3BGRL Public Health Service grant CA33192 from the National was measured by Western blot analysis 5 days post-infection. Cancer Institute.

References

Akiyama Y, Kato S. (1974). Biken J 17:105–116. Gasparian AV, Yao YJ, Kowalczyk D, Lyakh LA, Karseladze Baba TW, Giroir BP, Humphries EH. (1985). Virology 144: A, Slaga TJ et al. (2002). J Cell Sci 115:141–151. 139–151. Gilmore TD. (1995). Regulation of Rel transcription com- Baldwin Jr AS. (1996). Annu Rev Immunol 14:649–683. plexes. In:Goodbourn S (ed). Frontiers in Molecular Ball LJ, Jarchau T, Oschkinat H, Walter U. (2002). FEBS 513: Biology: Eukaryotic Transcription. Oxford, England:Oxford 45–52. University Press, pp 104–133. Barth CF, Ewert DL, Olson WC, Humphries EH. (1990). J Gilmore TD. (1999). Oncogene 18:6925–6937. Virol 64:6054–6062. Grabs D, Slepnev VI, Songyang Z, David C, Lynch M, Barzik M, Carl UD, Schubert W-D, Frank R, Wehland J, Cantley LC et al. (1997). J Biol Chem 272:13419–13425. Heinz DW. (2001). J Mol Biol 309:155–169. Hayden MS, Ghosh S. (2004). Genes Dev 18:2195–2224. Berleth ES, Henn AD, Gurtoo HL, Wollman R, Alderfer JL, Henn AD, Berleth ES, Mihich E, Ehrke MJ. (2001). FASEB Mihich E et al. (2000). Int J Immunopharmacol 22: 15:1315–1317. 1137–1142. Hrdlickova´ R, Nehyba J, Bose Jr HR. (1995a). J Virol 69: Berleth ES, Nadadur S, Henn AD, Eppolito C, Shiojiri S, 3369–3380. Gurtoo HL et al. (1999). Cancer Res 59:5497–5506. Hrdlickova´ R, Nehyba J, Bose Jr HR. (2001). Mol Cell Biol 21: Bonizzi G, Karin M. (2004). Trends Immunol 25:280–288. 6369–6386. Bose Jr HR. (1992). Biochim Biophys Acta 1114:1–17. Hrdlickova´ R, Nehyba J, Humphries EH. (1994). J Virol 68: Bull P, Morley KL, Hoekstra MF, Hunter T, Verma IM. 308–319. (1990). Mol Cell Biol 10:5473–5485. Hrdlickova´ R, Nehyba J, Roy A, Humphries EH, Bose Jr HR. Cao Y, Karin M. (2003). J Mammary Gland Biol Neoplasia 8: (1995b). J Virol 69:403–413. 215–223. Kabrun N, Hodgson JW, Doemer M, Mak G, Franza Jr BR, Carrasco D, Rizzo CA, Dorfman K, Bravo R. (1996). EMBO Enrietto PJ. (1991). Proc Natl Acad Sci USA 88:1783–1787. J 15:3640–3650. Karin M, Cao Y, Greten FR, Li ZW. (2002). Nat Rev Cancer Cesareni G, Panni S, Nardelli G, Castagnoli L. (2002). FEBS 2:301–310. 513:38–44. Kerr LD, Inoue JI, Davis N, Link E, Baeuerle PA, Bose Jr HR Chen C, Okayama H. (1987). Mol Cell Biol 7:2745–2752. et al. (1991). Genes Dev 5:1464–1476. Chen F, Castranova V, Shi X. (2001). Am J Pathol 159: Kralova J, Liss AS, Bargmann W, Bose Jr HR. (1998). Mol 387–397. Cell Biol 18:2997–3009. Cheng Q, Cant CA, Moll T, Hofer-Warbinek R, Wagner E, Kralova J, Liss AS, Bargmann W, Pendleton C, Varadarajan Birnstiel ML et al. (1994). J Biol Chem 269:13551–13557. J, Ulug E et al. (2002). J Virol 76:11960–11970. Chiao PJ, Miyamoto S, Verma IM. (1994). Proc Natl Acad Sci Kralova J, Schatzle JD, Liss AS, Bargman W, Bose Jr HR. USA 91:28–32. (1996). Oncogene 12:2595–2604. Cogswell PC, Guttridge DC, Funkhouser WK, Baldwin ASJ. Lee RM, Gillet G, Burnside J, Thomas SJ, Neiman P. (1999). (2000). Oncogene 19:1123–1131. Genes Dev 13:718–728. Deindl E. (2001). Biotechniques 31:1250–1252. Li Q, Verma IM. (2002). Nat Rev Immunol 2:725–734. Denk A, Wirth T, Baumann B. (2000). Cytokine Growth Factor Liss AS, Bose Jr HR. (2002). J Virol 76:4928–4939. Rev 11:303–320. Madrid LV, Baldwin Jr AS. (2003). Methods Mol Biol 223: Dougherty JP, Temin HM. (1986). Mol Cell Biol 6:4387–4395. 523–532. Dyer RB, Herzog NK. (1995). Biotechniques 19:192–195. Mayer BJ. (2001). J Cell Sci 114:1253–1263. Egeo A, Mazzocco M, Arrigo P, Vidal-Taboada JM, Oliva R, Mazzocco M, Arrigo P, Egeo A, Maffei M, Vergano A, Di Lisi Pirola B et al. (1998). Biochem Biophys Res Commun 247: R et al. (2001). Biochem Biophys Res Commun 285:540–545. 302–306. Mazzocco M, Maffei M, Egeo A, Vergano A, Arrigo P, Di Lisi Franklin RB, Kang C-Y, Wan KM-M, Bose Jr HR. (1977). R et al. (2002). Gene 291:233–239. Virology 83:313–321. Mecham BH, Klus GT, Strovel J, Augustus M, Byrne D, Fujii M, Minamino T, Nomura M, Miyamoto KI, Tanaka J, Bozso P et al. (2004). Nucleic Acids Res 32:e74. Seiki M. (1996). Oncogene 12:2193–2202. Moore BE, Bose Jr HR. (1988). Virology 162:377–387.

Oncogene Suppression of SH3BGRL in v-Rel-mediated transformation SM Majid et al 768 Moore PA, Ruben SM, Rosen CA. (1993). Mol Cell Biol 13: Schatzle JD, Kralova J, Bose Jr HR. (1995). J Virol 69: 1666–1674. 5383–5390. Moscovici C, Zeller N, Moscovici MG. (1982). Continuous Schwartz RC, Witte ON. (1988). Virology 165:182–190. lines of AMV-transformed non-producer cells:growth and Shi X-B, Ma A-H, Tepper CG, Xia L, Gregg JP, Gandour- oncogenic potential in the chick embryo. In:Revoltella RF, Edwards R et al. (2004). Prostate 60:257–271. Pontieri GM, Basiliico C, Rovera G, Gallo RC, Subak- Sovak MA, Bellas RE, Kim DW, Zanieski GJ, Rogers AE, Sharpe JH (eds). Expression of Differentiated Functions in Traish AM et al. (1997). J Clin Invest 100:2952–2960. Cancer Cells. Raven Press:New York, pp. 435–449. Stephens RM, Rice NR, Hiebsch RR, Bose Jr HR, Gilden RV. Nehyba J, Hrdlicˇ kova´ R, Burnside J, Bose Jr HR. (2002). Mol (1983). Proc Natl Acad Sci USA 80:6229–6233. Cell Biol 22:3942–3957. Suh J, Payvandi F, Edelstein LC, Amenta PS, Zong WX, Nelson DM, Wahlfors JJ, Chen L, Onodera M, Morgan RA. Gelinas C et al. (2002). Prostate 52:183–200. (1998). Hum Gene Ther 9:2401–2405. Suh J, Rabson AB. (2004). J Cell Biochem 91:100–117. Pahl HL. (1999). Oncogene 18:6853–6866. Thalmann GN, Sikes RA, Wu TT, Degeorges A, Chang SM, Petrenko O, Ischenko I, Enrietto PJ. (1997). Oncogene 15: Ozen M et al. (2000). Prostate 44:91–103. 1671–1680. Tong AH, Evangelista M, Parsons AB, Xu H, Bader GD, Page Petsko G. (2002). Current Protocols in Bioinformatics. N et al. (2001). Science 294:2364–2368. Hoboken:John Wiley & Sons, Inc. Tu JC, Xiao B, Yuan JP, Lanahan AA, Leoffert K, Li M et al. Rao A, Kline K, Sanders BG. (1990). Cancer Res 50: (1998). Neuron 4:717–726. 4764–4770. White DW, Roy A, Gilmore TD. (1995). Oncogene 10: Romieu-Mourez R, Kim DW, Shin SM, Demicco EG, 857–868. Landesman-Bollag E, Seldin DC et al. (2003). Mol Cell Wilhelmsen KC, Eggleton K, Temin HM. (1984). J Virol 52: Biol 23:5738–5754. 172–182. Sachdev S, Hannink M. (1998). Mol Cell Biol 18:5445–5456. You M, Bose Jr HR. (1998). Methods Enzymol 16:373–385. Sambrook J, Fritsch EF, Maniatis T. (1989). Molecular Zhang G, Slaughter C, Humphries EH. (1995). Mol Cell Biol Cloning: A Laboratory Manual, 2nd edn. Cold Spring 15:1806–1816. Harbor:Cold Spring Harbor Laboratory Press. Zhang GQ, Humphries EH. (1996). Oncogene 12:1153–1157. Sarkar S, Gilmore TD. (1993). Oncogene 8:2245–2252. Zhou JR, Yu L, Zerbini LF, Libermann TA, Blackburn GL. Scartenzzini P, Egeo A, Colella S, Fumagalli P, Arrigo P, (2004). Int J Cancer 110:800–806. Nizetic D et al. (1997). Hum Genet 99:387–392. Zong W-X, Farrell M, Bash J, Ge´ linas C. (1997). Oncogene 15: Schagger H, Jagow GV. (1987). Anal Biochem 166:368–379. 971–980.

Oncogene