Human Bcl-2 activates ERK signaling pathway to regulate ... - Nature

Oncogene (2004) 23, 7310–7321

& 2004 Nature Publishing Group All rights reserved 0950-9232/04 $30.00 www.nature.com/onc

Human Bcl-2 activates ERK signaling pathway to regulate activating protein-1, lens epithelium-derived growth factor and downstream genes Hao Feng1, Hua Xiang2,4, Ying-Wei Mao2,5, Juan Wang2,6, Jin-Ping Liu3, Xiao-Qin Huang1, Yan Liu1, Shao-Jun Liu1, Chen Luo1, Xuan-Jie Zhang1, Yun Liu1 and David Wan-Cheng Li*,1,2,3 1 College of Life Sciences, Hunan Normal University, Changsha, Hunan, PR China; 2Department of Molecular Biology, University of Medicine and Dentistry of New Jersey, Stratford, NJ, USA; 3The Hormel Institute, University of Minnesota, Austin, MN, USA

The proto-oncogene, bcl-2, has various functions besides its role in protecting cells from apoptosis. One of the functions is to regulate expression of other genes. Previous studies have demonstrated that Bcl-2 regulates activities of several important transcription factors including NFjB and p53, and also their downstream genes. In our recent studies, we reported that Bcl-2 substantially downregulates expression of the endogenous aB-crystallin gene through modulating the transcriptional activity of lens epithelium-derived growth factor (LEDGF). In the present communication, we report that human Bcl-2 can positively regulate expression of the proto-oncogenes c-jun and c-fos. Moreover, it enhances the DNA binding activity and transactivity of the activating protein-1 (AP-1). Furthermore, we present evidence to show that Bcl-2 can also activate both ERK1 and ERK2 MAP kinases. Inhibition of the activities of these kinases or the upstream activating kinases by pharmacological inhibitors or dominant-negative mutants abolishes the Bcl-2-mediated regulation of AP-1, LEDGF and their downstream genes. Together, our results demonstrate that through activation of the ERK kinase signaling pathway, Bcl-2 regulates the transcriptional activities of multiple transcription factors, and hence modulates the expression of their downstream genes. Thus, our results provide a mechanism to explain how Bcl-2 may regulate expression of other genes. Oncogene (2004) 23, 7310–7321. doi:10.1038/sj.onc.1208041 Published online 23 August 2004 Keywords: Bcl-2; ERK1/2; c-jun; c-fos; AP-1; LEDGF; lens

*Correspondence: DW-C Li, The Hormel Institute, University of Minnesota, 801 16th Avenue NE, Austin, MN 55912, USA; E-mail: [email protected] 4 Current address: State Key Laboratory of Microbial Resources and Center for Molecular Microbiology, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100080, PR China 5 Current address: Department of Pharmacology, University of Michigan, Ann Arbor, MI 48109, USA 6 Current address: Division of Endocrinology, Diabetes, and Metabolism, Departments of Medicine and Genetics and The Penn Diabetes Center, University of Pennsylvania, Philadelphia, PA 19104, USA Received 19 May 2004; revised 9 July 2004; accepted 9 July 2004; published online 23 August 2004

Introduction The proto-oncogene, bcl-2, was identified by its translocation (t[14:18]) and elevated expression in the follicular B-cell lymphomas (Tsujimoto et al., 1985). Initially, Bcl2 was found to prevent interleukin-3-dependent cells from apoptotic death upon withdrawal of the cytokine (Vaux et al., 1988). Subsequently, work from many different laboratories has shown that Bcl-2 can protect a variety of cells against different stress-induced apoptosis (Korsmeyer, 1992; Reed, 1994; Gross et al., 1999). In addition to its well-established role in promoting cell survival, recent studies have revealed that Bcl-2 also has other important functions. Bcl-2 is involved in the regulation of cell cycle (O’Reilly et al., 1997). During T-cell development, proliferative expansion and apoptotic cell death play prominent roles. In this process, Bcl-2 inhibits apoptosis and slows down cell cycle progression in cortical thymocytes and mature T cells, particularly during the transition from the quiescent state to the cell cycle (O’Reilly et al., 1997). During the cell cycle of the fibroblasts, Bcl-2 retards transition from G0 to S phase, which occurs through upregulation of p27 and p130 proteins and downregulation of p107 protein (Vairo et al., 2000). Bcl-2 is found to modulate cell differentiation (Lu et al., 1995; Hilton et al., 1997). In human luminal epithelial cells, Bcl-2 affects the phenotype of the original epithelial cells, and promotes epithelial–mesenchymal conversion, accompanied by loss of the cell adhesion molecules E-cadherin and a2b1-integrin (Lu et al., 1995). Bcl-2 is also observed to regulate axonal growth rates in embryonic neurons (Hilton et al., 1997). Sensory neurons from the trigeminal ganglia of bcl-2-deficient mouse embryos, removed from the embryo on embryonic day 11 or 12, extend axons more slowly in vitro than do neurons from wild-type embryos of the same age. Serial measurements of axonal length in the same neurons revealed that there were marked differences in axonal growth rate between bcl-2-deficient and wildtype neurons, irrespective of whether the neurons were grown with nerve growth factor, brain-derived neurotrophic factor or neurotrophin-3 (Hilton et al., 1997). Bcl-2 is also able to regulate gene expression (Miyashita et al., 1995; Feng et al., 1999; Vairo et al.,

Bcl-2 activates ERK1/2 H Feng et al

7311

2000; Mao et al., 2001; Schwarz et al., 2002). Bcl-2 has been found to modulate the transactivity of several transcription factors including NF-kB (de Moissac et al., 1998; Ricca et al., 2000), an important transcription factor mediating multiple signaling pathways (Li and Karin, 1999), and p53 (Froesch et al., 1999; Zhan et al., 1999), a transcription factor involved in the regulation of many targets genes (El-Deiry, 2003). In our recent study (Mao et al., 2001), we have demonstrated that Bcl2 negatively modulates the lens epithelium-derived growth factor (LEDGF), a lens transcription factor implicated in the regulation of aB-crystallin and other genes (Shinohara et al., 2002). However, how Bcl-2 could regulate the activities of different transcription factors remains unknown. The present study aims at further exploring the mechanism by which Bcl-2 regulates gene expression. First, we demonstrate that expression of the human Bcl-2 gene in the rabbit lens epithelial cells (RLECs) can positively regulate expression of c-jun and c-fos, and enhance the DNA binding activity and transactivity of the activating protein-1 (AP-1) besides its negative modulation on LEDGF (Mao et al., 2001). Bcl-2mediated regulation of AP-1 further affects the downstream genes such as the bA3/A1-crystallin gene, which contains an AP-1 binding site. More importantly, we show here that Bcl-2 activates ERK1/2 kinases, which is necessary for Bcl-2 to modulate AP-1 and LEDGF. Inhibition of the ERK1/2 signaling pathway with pharmacological inhibitor or dominant-negative mutants abolishes Bcl-2 modulation of AP-1 and LEDGF. Furthermore, repression of ERK1/2 activation also prevents Bcl-2 regulation of the crystallin genes containing AP-1 or LEDGF binding sites. These results show that Bcl-2 regulates gene expression through activation of the ERK signaling pathway, thus providing a novel mechanism to explain how Bcl-2 may regulate expression of different genes. Results Expressions of c-jun and c-fos are upregulated in pSFFV-Bcl-2-N/N1003A cells than in pSFFV-N/N1003A cells Our previous studies have revealed that Bcl-2 downregulates expression of aB-crystallin gene through regulating the transactivity of LEDGF (Mao et al., 2001). To further explore the Bcl-2 regulation of gene expression, we examined the effects of Bcl-2 on expression of the proto-oncogenes c-jun and c-fos with Northern blot analysis. As shown in Figure 1, the mRNA levels for both c-jun and c-fos were substantially upregulated in Bcl-2-transfected cells compared with that in parental and vector-transfected cells. To further confirm that the observed upregulation of these protooncogenes was derived from the effect of Bcl-2, we analysed their expression patterns in cells where antisense-bcl-2 RNA was expressed to block Bcl-2 expression. Expression of Bcl-2 in these cells was substantially downregulated (Mao et al., 2001). Similarly, expression

Figure 1 Human Bcl-2 positively regulates expression of c-jun and c-fos. A 100 mg portion of total RNAs extracted from parental RLECs (lane 1), vector-transfected cells (lane 2), Bcl-2-transfected cells (lane 3) and both Bcl-2- and antisense-bcl-2-transfected cells (lane 4) was denatured and resolved on 1.2% formaldehyde– agarose gel. Then the resolved RNAs were transferred to supported nitrocellulose membranes and the RNA blot was sequentially hybridized to a-32P-dATP-labeled c-jun, c-fos and GAPDH cDNA probes, washed under high stringency conditions and then exposed to an X-ray film as previously described (Li et al., 1994; Li and Spector, 1997). The top panel shows ethidium bromide staining of the RNA gel to display the RNA quality

of c-jun and c-fos was also downregulated in these double-transfected cells (Figure 1). Thus, human Bcl-2 expressed in N/N1003A cells upregulates expression of c-jun and c-fos. DNA binding activity and transactivity of AP-1 are distinctly increased in pSFFV-Bcl-2-N/N1003A cells than in pSFFV-N/N1003A cells Since Bcl-2 upregulates expression of c-jun and c-fos, we next examined its effect on the activity of AP-1, an important general transcriptional factor derived from homodimers of Jun family proteins or heterodimers of Jun and Fos family proteins, and regulating expression of other genes in many types of cells (Bohmann et al., 1987; Curran and Franza, 1988; Johnson and McKnight, 1989; Angel and Karin, 1991; Kerppola Oncogene


7312

and Curran, 1991). First, we explored whether Bcl-2 is able to affect the DNA binding activity of AP-1. As shown in Figure 2b, when nuclear extracts prepared from vector- and Bcl-2-transfected cells were examined, gel mobility shifting assays revealed that the Bcl-2transfected cells contain a much higher level of DNA binding activity to the AP-1 binding site than the vectortransfected cells (lanes 1 and 2 of Figure 2b). This binding activity is contributed by AP-1 for three reasons. First, the unlabeled oligos containing AP-1 binding site (the top panel of Figure 2a) can compete off the labeled probe during the binding shifting assay (lanes 3 and 4 of Figure 2b). Second, the oligos containing a mutated AP-1 site (the bottom panel of Figure 2a) did not display competition effect (lanes 5 and 6 of Figure 1a). Finally, the precleared nuclear extracts with anti-c-Jun antibody greatly attenuated DNA binding activity (lanes 7 and 8 of Figure 2b). These results suggest that Bcl-2 enhances the DNA

Figure 2 Human bcl-2 increases the DNA binding activity and transactivity of AP-1 in RLECs. (a) Diagram of the two oligos containing a native AP-1 binding site (up panel) or mutant AP-1 binding site (bottom panel), which were used for gel mobility shifting analysis described in Figures 2b, 4a, 5a and 6a. (b) Gel mobility shifting assays. Nuclear extracts prepared from vector-, Bcl-2-, Bcl-2/antisense-bcl-2-transfected cells were incubated with g-32P-ATP-labeled oligonucleotides containing wild-type or mutated AP-1 binding sites (Figure 2a) under various conditions shown in the figure. The reaction mixtures were then separated with 5% native PAGE. The gel was dried and exposed to an X-ray film for 3 h. Shown here is a typical result of three independent experiments. Lanes 1 and 2: gel mobility shifting assays with labeled oligo containing native AP-1 site and nuclear extract from vector-transfected cells (S) or Bcl-2-transfected cells (B). Lanes 3 and 4: the same assays as described for lanes 1 and 2 except that 50fold of nonlabeled oligo containing the native AP-1 binding site was added to each reaction. Note that the AP-1 complex was competed off by the nonlabeled oligo. Lanes 5 and 6: the same assays as described for lanes 3 and 4 except that the nonlabeled competing oligo contained a mutated AP-1 binding site (bottom panel of Figure 2a), which could not compete off the AP-1 complex formed between AP-1 protein and the oligo containing native AP-1 binding site. Lanes 7 and 8: the same assays as described in lanes 1 and 2 except that the nuclear extracts were preincubated with antic-Jun monoclonal antibody to remove the c-Jun protein. Note that when c-Jun protein was largely precleared, the AP-1 complex was weakly formed. Lanes 9 and 10: gel mobility shifting assays with labeled oligo containing native AP-1 site and nuclear extract from double vector-transfected cells (pSFFV/pZeoSV-N/N1003A, V) or Bcl-2- and antisense-bcl-2-transfected cells (pSFFV-Bcl-2/pZeoSVAntisense-bcl-2-N/N1003A, A). Note that inhibition of Bcl-2 expression through antisense bcl-2 RNA (Mao et al., 2001) abolished the Bcl-2-mediated increase in the DNA binding activity of AP-1. (c) Diagram to show the luciferase reporter gene driven by a prolactin gene mini-promoter (37 to þ 36) with four copies of AP-1 binding sites inserted in front of the promoter. This reporter gene is used in experiments described in Figures 2d, 4b, 5b and 6b. (d) Assays for transient reporter gene expression. The stable lines of vector-transfected cells (pSFFV-N/N1003A), Bcl-2-transfected cells (pSFFV-Bcl-2-N/N1003A) and both Bcl-2- and antisensebcl-2-transfected cells (pSFFV-Bcl-2/pZeoSV-Antisense-bcl-2-N/ N1003A) were further transfected with a reporter gene (Figure 2c) together with a control plasmid containing the b-galactosidase gene driven by the pRSV promoter (Nieson et al., 1983) as described in Materials and methods. At 24 h after transfection, these cells were harvested for analysis of luciferase activity as described (Xiang et al., 2000) Oncogene

binding activity of AP-1. To further confirm that it is Bcl-2 that enhances AP-1 activity for DNA binding, we tested the AP-1 binding activity from either double vector-transfected cells (pSFFV/pZeoSV-N/N1003A, lane 9) and cells transfected with both Bcl-2 and antisense-bcl-2 expression constructs (pSFFV-Bcl-2/ pZeoSV-Antisense-bcl-2-N/N1003A, lane 10). Our results indicated that inhibition of Bcl-2 expression also abolished the observed increase in the DNA binding activity of AP-1 (lane 10). To determine whether Bcl-2 also regulates the transactivity of AP-1, we transfected pSFFV-N/N1003A, pSFFV-Bcl-2-N/N1003A and pSFFV-Bcl-2/pZeoSV-Antisense-bcl-2-N/N1003A cells with an expression construct containing a luciferase reporter gene driven by a mini-prolactin gene promoter (37 to þ 36) with four copies of the AP-1 binding site inserted in front of the promoter (Figure 2c) together with a control plasmid containing the b-galactosidase gene (Nieson et al., 1983). At 24 h after transfection,


7313

these cells were harvested for analysis of luciferase activity. As shown in Figure 2d, Bcl-2 induced a 4.8-fold increase in luciferase activity. This increase was abolished by the antisense bcl-2 RNA. Thus, Bcl-2 is able to modulate the DNA binding activity and transactivity of AP-1 in lens epithelial cells, N/N1003A. Bcl-2 activates ERK1/2 but not p38 kinase and JNK1/2 in pSFFV-Bcl-2-N/N1003A cells Because MAP kinases are actively involved in the regulation of AP-1 (Karin, 1995; Kumar and Bernstein, 2001; Kyriakis and Avruch, 2001; Yuan et al., 2001; Kukushkin et al., 2002; Young et al., 2002; Casalino et al., 2003; Chen and Davis, 2003; Monje et al., 2003), we next examined whether Bcl-2 could activate these kinases to mediate its regulation of AP-1 and possibly other transcription factors. As shown in panels 1 and 2 of Figure 3, although the total protein levels of ERK1/2 in vector-, Bcl-2- and Bcl-2/antisense-bcl-2-transfected cells were similar, Bcl-2 induced a substantial increase in the activities of both ERK1 and ERK2. Quantitation of the bands with the automated digitizing system from the Silk Scientific Corporation revealed that the activity of ERK1 in Bcl-2 expression cells was increased threefold than that in both parental and vector-transfected cells. Bcl-2 also induced a six- and threefold increase in ERK2 activity in Bcl-2-transfected cells compared with that in the parental cells and vector-transfected cells, respectively. To further confirm that it was the expression of Bcl-2 that led to induction of ERK1/2, we again blocked Bcl-2 expression through antisense strategy and then examined the activities of ERK1/2. As shown in the right lane of panels 1 and 2 of Figure 3, inhibition of Bcl-2 expression abolished the observed activation of ERK1/2. Western blot analysis of p38 kinase (panels 3 and 4 of Figure 3) and JNK1/2 (panels 5 and 6 of Figure 3) in parental, vector-, Bcl-2- and Bcl-2/ antisense-bcl-2-transfected cells revealed the absence of activation of the two MAPKs by Bcl-2. Bcl-2 also had little effect on expression of a house-keeper protein, bactin (panel 7 of Figure 3). Thus, human Bcl-2 activates only ERK1/2 kinases in N/N1003A cells. Bcl-2-mediated increase in AP-1 activity requires activation of ERK1/2 Since Bcl-2 enhances AP-1 activity and also activates ERK1/2, we next investigated whether Bcl-2-mediated increase in AP-1 activity requires activation of ERK1/2. For this, both vector- and Bcl-2-transfected cells were grown to 100% confluence and then treated with 0.01% DMSO (mock treatment) or 25 mM PD98059 for inhibition of MEK1/2 kinases for 22 h. At the end of treatment, these different cells were harvested for gel mobility shifting assays. As shown in Figure 4a, inhibition of ERK1/2 activation by PD98059 almost abolished Bcl-2-mediated increase in the DNA binding activity of AP-1. To demonstrate that inhibition of ERK1/2 activity by PD98059 also abolishes the Bcl-2mediated enhancement of AP-1 transactivity, the

Figure 3 Human Bcl-2 activates ERK1/2 in RLECs, N/N1003A. The parental RLECs (lane 1), vector-transfected cells (lane 2), Bcl2-transfected cells (lane 3) and both Bcl-2- and antisense-bcl-2transfected cells (lane 4) were grown to 100% confluence and then harvested for preparation of total proteins. A 100 mg portion of total proteins from each type of cells was resolved in 10% SDS– PAGE, transferred to supported nitrocellulose membranes and then analysed with antibodies against phospho-ERK1/2 (p-ERK1/ 2, activated form, panel 1), total ERK1/2 (T-ERK1/2, panel 2), phospho-p38 (p-p38, panel 3), total p38 (T-p38, panel 4), phosphoJNK1/2 (p-JNK1/2, panel 5), total JNK1/2 (T-JNK1/2, panel 6) and b-actin (panel 7) as described in Materials and methods. Note that Bcl-2 activates ERK1/2 in Bcl-2-transfected cells (lane 3) and this activation was abolished when Bcl-2 expression was inhibited by antisense-bcl-2 RNA (lane 4)

luciferase reporter gene construct and the b-gal control construct were introduced into both vector- and Bcl-2transfected cells. At 2 h after the transfected cells were seeded, 25 mM PD98059 was added to the transfected cells. After an additional 22 h incubation, the transfected cells were harvested for measurement of luciferase and b-galactosidase activity. As shown in Figure 4b, inhibition of ERK1/2 activity indeed abolished Bcl-2mediated enhancement in AP-1 transactivity. Thus, Bcl-2-induced increase in AP-1 activity requires activation of ERK1/2. Oncogene


7314

Activation of MEK1/2 and RAF1 is necessary for Bcl-2 to regulate positively AP-1 activity Because ERK1/2 kinases are activated by the upstream activation kinases MEK1/2 and Raf-1, we next explored whether activation of MEK1/2 and Raf-1 was necessary for Bcl-2 to regulate positively AP-1 activity. For this purpose, the dominant-negative mutants Ras and Raf-1, which, respectively, interfere with the activation of Raf-1 and MEK1/2 (Rosen et al., 1994; Mischak et al., 1996), were introduced into both vector- and Bcl-2transfected cells. As shown in Figure 5a, both the dominant-negative mutants Ras and Raf-1 significantly attenuated Bcl-2-mediated enhancement in the DNA binding activity of AP-1. Moreover, the transactivity of AP-1 was also abolished by the two dominant-negative mutants (Figure 5b). These results indicate that Bcl-2

activates Raf-1 through Ras and the activated Raf-1 further activates MEK1/2, leading to activation of ERK1/2. Furthermore, activation of the RAS/RAF/ MEK/ERK pathway is crucial for Bcl-2 to regulate positively AP-1 activity. p38 and JNK1/2 are not involved in Bcl-2 regulation of AP-1 activity The results that Bcl-2 does not change the activity of both p38 and JNK1/2 (Figure 3) suggest that Bcl-2mediated positive regulation of AP-1 activity does not require participation of p38 and JNK1/2. To further confirm that this is indeed the case, we have utilized the specific inhibitors for p38 and JNK1/2 to pretreat both vector- and Bcl-2-transfected cells. As shown in Figure 6a, gel mobility shifting assays revealed that inhibition of JNK1/2 activation by SP600125 (lanes 3 and 4) and prevention of p38 activation by PD169316 (lanes 5 and 6) did not change the Bcl-2-mediated increase in AP-1 activity. Similarly, transient reporter gene expression assays demonstrated that the two inhibitors had virtually no effect on the Bcl-2-enhanced transactivity of AP-1 (Figure 6b). These results clearly show that activation of ERK1/2 but not p38 and JNK1/ 2 signaling pathways mediates positive regulation of AP-1 activity by Bcl-2. Bcl-2 regulates the bA3/A1-crystallin gene containing AP-1 binding site and such regulation requires activation of the RAF/MEK/ERK signaling pathway Since Bcl-2 regulates AP-1 activity in N/N1003A cells, we next asked whether Bcl-2 could also modulate expression of the crystallin gene containing AP-1 binding site. Previous studies have shown that chicken bA3/A1-crystallin gene promoter contains an AP-1 binding site located at 264 to 258 (Mcdermott et al., Figure 4 Bcl-2-mediated increase of AP-1 activity in RLECs requires activation of ERK1/2 kinases. (a) Gel mobility shifting assays. Nuclear extracts prepared from both vector-transfected cells (S) and Bcl-2-transfected cells (B) were incubated with g-32PATP-labeled oligonucleotides containing wild-type AP-1 binding site (shown in the up panel of Figure 2a). The reaction mixtures were then separated with 5% native PAGE. The gel was dried and exposed to an X-ray film for 3 h. Shown here is a typical result of three independent experiments. Lanes 1 and 2: gel mobility shifting assays with labeled oligo containing native AP-1 site and nuclear extract from vector-transfected cells (lane 1) or Bcl-2-transfected cells (lane 2) pretreated with 0.01% DMSO for 12 h. Lanes 3 and 4: gel mobility shifting assays with labeled oligo containing native AP-1 site and nuclear extract from vector-transfected cells (lane 3) or Bcl-2-transfected cells (lane 4) pretreated with 25 mM PD98059, MEK1/2 inhibitor, for 22 h. (b) Assays for transient reporter gene expression. The vector-transfected cells (vector) and Bcl-2-tranfected cells (Bcl-2) were further transfected with a reporter gene (Figure 2c) together with a control plasmid pRSV-b-Gal as described in Materials and methods. At 2 h after transfection, the cells were either pretreated with 0.01% DMSO (columns labeled ‘Vector’ or ‘Bcl-2’) or 25 mM PD98059 (columns labeled ‘Vector þ PD98059’ or ‘Bcl-2 þ PD98059’) for an additional 22 h and then harvested for analysis of luciferase activity as described (Xiang et al., 2000). Note that inhibition of ERK1/2 activation abolished the Bcl-2-mediated increase of AP-1 transactivity

Oncogene


7315

Figure 5 Bcl-2-mediated increase of AP-1 activity requires activation of MEK1/2 and RAF-1. (a) Gel mobility shifting assays. Nuclear extracts prepared from both vector-transfected cells (S) and Bcl-2-transfected cells (B) were incubated with g-32P-ATPlabeled oligonucleotides containing wild-type AP-1 binding site (shown in the top panel of Figure 2a). Other procedures are the same as described in Figures 2b and 4a. Shown here is a typical result of three independent experiments. Lanes 1 and 2: gel mobility shifting assays with labeled oligo containing a native AP-1 site as shown in the top panel of Figure 2a and nuclear extracts from vector-transfected cells (lane 1) or Bcl-2-transfected cells (lane 2) with transfection of an additional control vector, pCMV. Lanes 3 and 4: gel mobility shifting assays with labeled oligo containing a native AP-1 site and nuclear extracts from vector-transfected cells (lane 3) or Bcl-2-transfected cells (lane 4) with expression of a dominant-negative Ras mutant driven by the pCMV promoter. Lanes 5 and 6: gel mobility shifting assays with labeled oligo containing a native AP-1 site and nuclear extracts from vectortransfected cells (lane 5) or Bcl-2-transfected cells (lane 6) with expression of a dominant-negative Raf mutant driven by the pCMV promoter. Note that the dominant-negative mutants Ras (which interferes with activation of Raf-1) and Raf-1 (which blocks activation of MEK1/2) significantly abolished Bcl-2-mediated increase in the DNA binding activity of AP-1. (b) Assays for transient reporter gene expression. The vector-transfected cells and Bcl-2-transfected cells were further transfected with a reporter gene (Figure 2c) and a control plasmid pRSV-b-Gal alone (columns labeled ‘Vector’ or ‘Bcl-2’), or together with a dominant-negative mutant Ras (columns labeled ‘Vector þ DNM Ras’ or ‘Bcl2 þ DNM Ras’) or a dominant-negative mutant Raf (columns labeled as ‘Vector þ DNM Raf’ or ‘Bcl-2 þ DNM Raf’), as described in Materials and methods. At 24 h after transfection, the cells were harvested for analysis of luciferase activity as described (Xiang et al., 2000). Note that the dominant-negative mutants Ras and Raf-1 significantly abolished Bcl-2-mediated increase in the transactivity of AP-1

Figure 6 Bcl-2-mediated increase in AP-1 activity does not require involvement of p38 and JNK1/2. (a) Gel mobility shifting assays. Nuclear extracts prepared from both vector-transfected cells (S) and Bcl-2-transfected cells (B) were incubated with g-32P-ATPlabeled oligonucleotides containing wild-type AP-1 binding site (shown in the top panel of Figure 2a). Other procedures are the same as described in Figures 2b and 4a. Shown here is a typical result of three independent experiments. Lanes 1 and 2: gel mobility shifting assays with labeled oligo containing native AP-1 site and nuclear extract from vector-transfected cells (lane 1) or Bcl-2transfected cells (lane 2) pretreated with 0.01% DMSO for 22 h. Lanes 3 and 4: gel mobility shifting assays with labeled oligo containing native AP-1 site and nuclear extract from vectortransfected cells (lane 3) or Bcl-2-transfected cells (lane 4) pretreated with 1 mM SP600125, JNK1/2 inhibitor, for 22 h. Lanes 5 and 6: gel mobility shifting assays with labeled oligo containing native AP-1 site and nuclear extract from vector-transfected cells (lane 5) or Bcl-2-transfected cells (lane 6) pretreated with 2 mM PD169316, p38 inhibitor, for 22 h. Note that inhibition of JNK1/2 activation by SP600125 and prevention of p38 activation by PD169316 had little effect on Bcl-2-mediated increase in the DNA binding activity of AP-1. (b) Assays for transient reporter gene expression. The vector- and Bcl-2-transfected cells were further transfected with a reporter gene (Figure 2c) and a control plasmid pRSV-b-Gal as described in Materials and methods. At 2 h after transfection, the cells were pretreated with 0.01% DMSO (columns labeled ‘Vector’ or ‘Bcl-2’), 1 mM SP600125 (columns labeled ‘Vector þ SP600125’ or ‘Bcl-2 þ SP600125’) or 2 mM PD169316 (columns labeled ‘Vector þ PD169316’ or ‘Bcl-2 þ PD169316’) for an additional 22 h, and then harvested for analysis of luciferase activity as described (Xiang et al., 2000). Note that inhibition of JNK1/2 activation by SP600125 and prevention of p38 activation by PD169316 had little effect on Bcl-2-mediated increase in the transactivity of AP-1

Oncogene


7316

1997). This promoter has been cloned and linked to a reporter gene, chloramphenicol acetyltransferase (CAT; Mcdermott et al., 1997), to generate chicken bA3/A1CAT (Figure 7a). To examine the possible modulation of Bcl-2 on this crystallin gene promoter, pSFFV-N/ N1003A, pSFFV-Bcl-2-N/N1003A and pSFFV-Bcl-2/ pZeoSV-Antisense-bcl-2-N/N1003A cells were transfected with chicken bA3/A1-CAT together with the control plasmid pRSV-b-Gal. At 48 h after transfection, these cells were harvested for analysis of CAT expression and activity. As shown in Figure 7b, Northern blot analysis demonstrated that the CAT mRNA was substantially upregulated in Bcl-2-transfected cells than that in parental or vector-transfected cells. Enzyme activity assay revealed that Bcl-2 enhanced expression of the chicken bA3/A1 promoter up to eightfold (column 2 of Figure 7c). Mutation of the AP-1 binding site almost abolished the enhancement (column 4 of Figure 7c). Thus, Bcl-2 not only affects AP-1 activity but also modulates the genes containing AP-1 binding site.

Since Bcl-2 modulates expression of the lens crystallin gene containing an AP-1 binding site, we next explored whether such modulation requires activation of the RAF/MEK/ERK signaling pathway. To do so, both the chicken bA3/A1-CAT and the control plasmid pRSV-bGal alone (columns 1 and 2 of Figure 7d), or together with a dominant-negative mutant ERK2 (columns 3 and 4 of Figure 7d), a dominant-negative mutant Raf-1 (columns 5 and 6 of Figure 7d) or a dominant-negative mutant Ras (columns 7 and 8 of Figure 7d), were introduced into pSFFV-N/N1003A and pSFFV-Bcl-2N/N1003A cells. At 48 h after transfection, these cells were harvested for analysis of CAT activity as described above. As shown in Figure 7d, the dominant-negative mutants ERK2, Raf-1 and Ras all blocked Bcl-2mediated enhancement in the CAT activity driven by the bA3/A1-crystallin gene promoter. These results demonstrate that Bcl-2-mediated modulation of the lens crystallin gene requires activation of the RAF/MEK/ ERK signaling pathway.

Figure 7 Bcl-2 regulates bA3/A1-crystallin gene containing an AP-1 binding site and such regulation requires activation of the RAF/MEK/ERK signaling pathway. (a) Diagram for the reporter gene construct of bA3/A1-CAT. The bA3/A1-crystallin gene promoter contains an AP-1 binding site located at 264 to 258. The promoter has been shown to be active in mouse lens epithelial cells (Mcdermott et al., 1997). (b) Northern blot analysis to show that Bcl-2 positively regulates the bA3/A1-crystallin gene promoter. A 100 mg portion of total RNAs extracted from parental RLECs (lane 1), vector-transfected cells (lane 2), Bcl-2-transfected cells (lane 3) and both Bcl-2- and antisense-bcl-2-transfected cells (lane 4) was denatured and separated on 1.2% formaldehyde– agarose gel. Then, the resolved RNAs were transferred to supported nitrocellulose membranes and the RNA blot was sequentially hybridized to a-32P-dATP-labeled CAT and GAPDH cDNA probes, washed under high stringency conditions and then exposed to an X-ray film as previously described (Li et al., 1994; Li and Spector, 1997). Note that Bcl-2 upregulated the CAT mRNA from bA3/A1-CAT (lane 3) and this upregulation is Bcl-2 dependent (lane 4). (c) Assays for transient reporter gene expression from chicken bA3/A1-CAT. The stable lines of vectortransfected cells (column 1), Bcl-2-transfected cells (columns 2 and 4) and both Bcl-2- and antisense-bcl-2-transfected cells (column 3) were further transfected with a CAT reporter gene (Figure 7a) driven by the bA3/A1-crystallin gene promoter (382 to þ 22) carrying a native AP-1 site (columns 1–3) or a mutated AP-1 site (column 4) located at 264 to 258 together with a control plasmid pRSV-b-Gal. At 48 h after transfection, the cells were harvested for analysis of CAT activity as described (Xiang et al., 2002). Note that Bcl-2 positively regulated the chicken bA3/A1crystallin gene promoter and this regulation is AP-1 dependent. (d) Bcl-2-mediated positive regulation of the chicken bA3/A1 crystallin gene requires activation of the RAF/MEK/ERK signaling pathway. The stable lines of vector-transfected cells (columns 1, 3, 5 and 7) and Bcl-2-transfected cells (columns 2, 4, 6 and 8) were further transfected with a CAT reporter gene (Figure 7a) and a control plasmid pRSV-b-Gal alone (columns 1 and 2), or with a dominantnegative mutant ERK2 (columns 3 and 4), a dominant-negative mutant Raf-1 (columns 5 and 6) or a dominant-negative mutant Ras (columns 7 and 8), as described in Materials and methods. At 48 h after transfection, these cells were harvested for analysis of CAT activity as described (Xiang et al., 2002). Note that Bcl-2 positively regulated the CAT activity driven by the bA3/A1crystallin gene promoter and the dominant-negative mutants ERK2, Raf-1 or Ras significantly abolished Bcl-2-mediated enhancement in the reporter gene activity from bA3/A1-CAT Oncogene


7317

Activation of ERK1/2 is necessary for Bcl-2 to downregulate expression of aB-crystallin in pSFFV-Bcl-2-N/N1003A cells We have previously demonstrated that through negative regulation of LEDGF activity, Bcl-2 downregulates aBcrystallin gene (Mao et al., 2001). Since Bcl-2-mediated modulation of the chicken bA3/A1-crystallin gene promoter requires activation of ERK1/2, we next asked whether Bcl-2-mediated downregulation of aB-crystallin also requires activation of ERK1/2 kinases. To do so, both vector- and Bcl-2-transfected cells were first treated with 25 mM PD98059 for 22 h and then harvested for extraction of total proteins. As shown in panels a and b of Figure 8A, both ERK1 and ERK2 activities were completely abolished by the inhibitor. When ERK1/2 activation was suppressed, the expression of aB-crystallin was distinctly recovered in vector-transfected cells, especially in Bcl-2-transfected cells (panel c of Figure 8A). In contrast, expression of b-actin, a house-keeper protein, was not affected (panel d of Figure 8A). Since we have previously shown that downregulation of aBcrystallin is linked to decreased activity of LEDGF (Mao et al., 2001), we next examined whether ERK1/2 activation is responsible for the decreased activity of LEDGF. As shown in Figure 8B, inhibition of ERK1/2 activation by 25 mM PD98059 clearly enhanced the DNA binding activity of LEDGF. Thus, Bcl-2 modulation of LEDGF activity is also mediated through the RAF/ MEK/ERK signaling pathway.

Discussion In the present communication, we have demonstrated the following: (1) Bcl-2, when expressed in RLECs, upregulates expression of c-jun, and c-fos, enhances the DNA binding activity and transactivity of AP-1 and modulates expression of the lens crystallin gene containing an AP-1 binding site; (2) Bcl-2 also activates the ERK1/2 kinases; (3) Bcl-2-mediated enhancement in AP-1 activity, attenuation in LEDGF activity, and positive or negative modulation of expression of the lens crystallin genes require activation of the RAF/MEK/ ERK signaling pathway. Together, our results reveal a mechanism to explain Bcl-2 regulation of gene expression (Figure 9). Bcl-2 modulates various transcription factors to regulate gene expression Regulation of gene expression by Bcl-2 was initially revealed through its enhancement of the half-life of p21Bax (Miyashita et al., 1995). Later, Feng et al. (1999) showed that Bcl-2 promotes expression of a differentiation-specific gene encoding the proteoglycan aggrecan in rat chondrocyte cell line, IRC cells. This regulation occurs at the mRNA and protein levels. More recently, Schwarz et al. (2002) demonstrated that overexpression of Bcl-2 in PC12 cells through an inducible turn-on system positively regulates expression of the gene

Figure 8 Downregulation of aB expression by Bcl-2 requires activation of the RAF/MEK/ERK signaling pathway. (A) Western blot analysis. Total proteins were prepared from vector-transfected cells (lanes 1 and 3) and Bcl-2-transfected cells (lanes 2 and 4) pretreated with 0.01% DMSO (lanes 1 and 2) or 25 mM PD98059, MEK1/2 inhibitor (lanes 3 and 4), for 22 h. A 100 mg portion of total proteins from each type of cells was resolved in 10% SDS– PAGE, transferred to supported nitrocellulose membranes and then sequentially blotted with antibodies against phospho-ERK1/2 (p-ERK1/2, activated form, panel a), total ERK1/2 (T-ERK1/2, panel b), aB-crystallin (panel c) and b-actin (panel d) as described in Materials and methods. Note that when ERK activation is inhibited, expression of aB-crystallin is upregulated. (B) Gel mobility shifting assays. The oligo containing a native LEDGF binding site is shown in the left. Nuclear extracts were prepared from both vector-transfected cells (lanes 1 and 3) and Bcl-2transfected cells (lanes 2 and 4), which were pretreated with 0.01% DMSO (lanes 1 and 2) or 25 mM PD98059, MEK1/2 inhibitor (lanes 3 and 4), for 12 h. These extracts were incubated with g-32P-ATPlabeled oligonucleotide containing the wild-type LEDGF binding site. The reaction mixtures were then separated with 5% native PAGE. The gel was dried and exposed to an X-ray film for 3 h. Shown here is a typical result of three independent experiments. Note that inhibition of ERK activation restored the DNA binding activity of LEDGF

encoding Ha-Ras. In contrast to its positive regulation on the aggrecan and Ha-Ras genes, we recently reported that Bcl-2 substantially attenuates expression of the aBcrystallin gene in rabbit N/N1003A cells. Such negative regulation also occurs at both mRNA and protein levels. Bcl-2 also downregulates expression of an exogenous mouse aB-crystallin gene promoter (Mao et al., 2001). In the present study, we present evidence to show that Bcl-2 displays positive regulation on the chicken b-A3/ A1-crystallin gene. Such dual effects of Bcl-2 in regulating gene expression have also been observed in the mouse fibroblasts. Vairo et al. (2000) have demonstrated that Bcl-2 upregulates accumulation of p27 and p130 proteins but downregulates the level of p107 Oncogene


7318

positively regulated by Bcl-2 (de Moissac et al., 1998; Ricca et al., 2000). By changing the affinity of IkBa to NF-kB, Bcl-2 can increase the transactivity of NF-kB and promotes expression of NF-kB-responsive genes such as the one encoding the matrix metalloproteinase-9 (de Moissac et al., 1998; Ricca et al., 2000). The tumor suppressor, p53, a transcription factor regulating genes for cell cycle control and apoptosis, is also a target of Bcl2. Zhan et al. (1999) found that in the human Burkett’s lymphoma WMN cell line, Bcl-2 specifically suppresses the p53-mediated transactivation of p21CIP1/WAF1 and GADD45 after treatment with methylmethane sulfonate or UV irradiation. In human kidney 293 cells and MCF7 cells, Froesch et al. (1999) observed that overexpression of Bcl-2 downregulates p53 transactivity. Bcl-2 activates the RAF/MEK/ERK signaling pathway to regulate activities of various transcription factors

Figure 9 Diagram to show a novel mechanism for Bcl-2 to regulate gene expression in lens system and possibly other cells. Bcl-2 either directly or indirectly activates RAS, which then activates the RAF-1 kinase, leading to activation of the downstream kinases, MEK1/2, which then activate ERK1/2. Activation of ERK1/2 modulates transcription factors such as AP-1 and LEDGF and other transcriptional factors through phosphorylation. The changes in the modulated transcriptional factors then exert regulation of expression of the downstream genes. Up-arrow represents for upregulation; down-arrow in the side represents downregulation

protein from G0 to S transition during the cell cycle of the fibroblasts. This differential regulation allows Bcl-2 to retard fibroblast cells entering into the cell cycle. As a mitochondria-associated protein, how could Bcl2 regulate gene expression? Our results suggest that Bcl2 can modulate transcription factors to regulate gene expression. In the present study, we present evidence to show that Bcl-2 positively modulates both the DNA binding activity and the transactivity of AP-1. Through modulation of AP-1, Bcl-2 also regulates chicken bA3/ A1-crystallin gene expression. This conclusion is consistent with numerous earlier studies. We have previously shown that the DNA binding activity of LEDGF, an important transcription factor in lens (Fatma et al., 2001; Singh et al., 2001; Kubo et al., 2002; Shinohara et al., 2002), is distinctly suppressed in Bcl-2 expression RLECs (Mao et al., 2001). The decreased LEDGF activity contributes substantially to the downregulation of aB-crystallin gene by Bcl-2 since expression of the exogenous LEDGF in Bcl-2 expression cells significantly upregulates the expression level of aBcrystallin (Mao et al., 2001). Previous studies from other laboratories also support this conclusion. For example, NF-kB, a prominent transcription factor mediating multiple signaling pathways (Li and Karin, 1999), is Oncogene

How could Bcl-2 modulate the activities of various transcriptional factors? It is well established that phosphorylation plays a key role in altering the activities of various transcription factors (Hunter and Karin, 1992). In the present study, we have shown that Bcl-2 is able to activate ERK1/2 kinases. Similar observations have been reported in PC12 cells (Schwarz et al., 2002). Our results further show that activated ERK kinase activity is essential for Bcl-2 to regulate gene expression because inhibition of ERK activation with a pharmacological drug or the dominant-negative mutant RAS or RAF-1, upstream activators of ERK1/2, all abolished Bcl-2-mediated regulation of AP-1 and LEDGF, and hence their downstream crystallin genes. Thus, by activating the RAF/MEK/ERK signaling pathway, Bcl2 modulates the activities of various transcription factors. How could activation of the RAF/MEK/ERK signaling pathway lead to modulation of various transcription factors? Activation of AP-1 is a good example for the demonstration of such modulation by ERKs. In the mammalian system, three types of mitogenactivated protein kinases (MAPKs including ERKs, JNKs and p38) have been identified and all of them have been shown to regulate AP-1 transcription factors (Karin, 1995; Kumar and Bernstein, 2001; Kyriakis and Avruch, 2001; Yuan et al., 2001; Kukushkin et al., 2002; Young et al., 2002; Casalino et al., 2003; Chen and Davis, 2003; Monje et al., 2003). Although all three types of MAPKs are expressed in the ocular lens (Li et al., 2003), our results show that Bcl-2 activates only ERK1/2 but not p38 kinase and JNKs (Figure 3). Such specificity is also observed in PC12 cells (Schwarz et al., 2002). ERK1/2 kinases are known to modulate c-Jun, JunD, c-Fos, Fra-1, Fra-2 and FosB (Gruda et al., 1994; Karin, 1995; Rosenberger et al., 1999; Yue and Mulder, 2000; Kumar and Bernstein, 2001; Kyriakis and Avruch, 2001; Yuan et al., 2001; Dhawan and Richmond, 2002; Kukushkin et al., 2002; Schwarz et al., 2002; Young et al., 2002; Casalino et al., 2003; Chen and Davis, 2003; Monje et al., 2003; Jiang et al., 2004;) in various systems through phosphorylation of either the intermediate transcription factors (indirect phosphorylation) or the


7319

above AP-1 family members (direct phosphorylation). Both phosphorylation mechanisms exist for c-Fos activation by ERK1/2. These kinases can activate Elk1 through phosphorylation, which together with the serum response factor forms a transcription complex, binding to the serum response element of the c-Fos promoter and thus enhancing c-fos expression at the transcription level (Karin, 1995). More recently, Monje et al. (2003) have demonstrated that ERK1/2 can directly phosphorylate multiple sites in the carboxylterminal transactivation domain of c-Fos to enhance its AP-1 transactivity. Both Fra-1 and Fra-2 were initially found to be phosphorylated by ERKs (Gruda et al., 1994). A recent report confirms that Fra-1 activation is ERK dependent in mouse JB6 cells (Young et al., 2002). In papilloma-producing 308 mouse keratinocytes, ERK1/2-induced phosphorylation of FosB and JunD is essential for okadaic acid-induced AP-1 activity (Rosenberger et al., 1999). During TGFb3 induction of TGFb1, AP-1 activation is dependent on activation of the RAS/RAF/MEK/ERK pathway and its modulation on JunD and Fra-2 (Yue and Mulder, 2000). In PC12 cells, ERK kinases promote phosphorylation of cJun (Schwarz et al., 2002). In RLECs, our results demonstrate that Bcl-2-activated ERK1/2 positively regulates expression of c-jun and c-fos, which contributes to enhanced AP-1 activity. Our results also show that Bcl-2-activated ERK1/2 is necessary for Bcl-2 to downregulate LEDGF activity. When ERK1/2 activity is inhibited by PD98059, the DNA binding activity of LEDGF is substantially upregulated in Bcl-2 expression cells (Figure 8B). It remains to be determined whether ERK1/2 downregulates LEDGF through direct phosphorylation or through phosphorylation of the intermediate transcription factors such as Elk-1. Besides their modification on AP-1 and LEDGF, ERKs are also able to regulate other transcription factors such as NF-kB. In human melanoma cells, activation of ERK1/2 is essential for the induction of NF-kB activity (Dhawan and Richmond, 2002). In cultured rat vascular smooth muscle cells, sustained activation of ERK is required for interleukin-1b to persistently activate NF-kB (Jiang et al., 2004). Although the direct phosphorylation of p53 by ERK1/ 2 has not been demonstrated, a correlation of ERK activation and the related changes in p53 stability and activity has been established in different cell lines (Persons et al., 2000; Sablina et al., 2001). Thus, our finding that Bcl-2 activates the RAS/RAF/MEK/ERK signaling pathway to modulate various transcription factors provides a mechanism to explain how Bcl-2 regulates expression of different genes (Figure 9).

Stratagene (La Jolla, CA, USA) and Promega Biotech (Madison, WI, USA). DNA oligos, DNA and protein size markers were purchased from Invitrogen Life Technologies (Gaithersburg, MD, USA). Mammalian expression vectors and constructs were purchased from Clontech (Palo Alto, CA, USA), Promega Biotech (Madison, WI, USA) and Stratagene (La Jolla, CA, USA). Various antibodies were obtained from Cell Signaling Technology (Boston, MA, USA), Santa Cruz Biotechnology (Santa Cruz, CA, USA), Roche Molecular Biochemicals (Indianapolis, IN, USA) and Transduction Laboratories (San Diego, CA, USA). The culture medium and most other chemicals and antibiotics were purchased from Sigma (St Louis, MO, USA) and Invitrogen Life Technologies (Gaithersburg, MD, USA). Cell culture RLECs, N/N1003A (a kind gift from Dr John Reddan) (Reddan et al., 1986), were grown in Eagle’s minimum essential medium containing 10% rabbit serum as described previously (Mao et al., 2001). The medium was prepared in ion-exchanged double-distilled water to give an osmolarity of 30075 mosmol supplemented with 26 mM NaHCO3 and 50 mg/ ml gentamycin sulfate. Media and sera were sterilized by filtration through 0.22 mm filters with pH adjusted to 7.2. All cells were kept at 371C and 5% CO2 gas phase. Establishment of stable expression cell lines The mammalian expression vector, pSFFV-neo, and the Bcl-2 expression construct, pSFFV-hBcl-2, have been described before (Fuhlbrigge et al., 1988; Hockenbery et al., 1990). The antisense bcl-2 expression plasmid was constructed using a different expression vector, pZeoSV vector (Invitrogen, CAT#V850-01), with the Bcl-2 cDNA inserted at the EcoRI site in the 30 to 50 direction (Mao et al., 2001). These constructs were amplified in DH-5a and purified by two rounds of CsCl ultracentrifugation as previously described (Li et al., 1995). Transfection of N/N1003A cells was performed using electroporation with a BTX Electro Cell Manipulator as previously described (Xiang et al., 2000). The transfected cells were then subject to specific drug selection for 4–6 weeks before individual clones were obtained. Both bcl-2-transfected cells (pSFFV-Bcl-2-N/N1003A) and vector-transfected cells (pSFFV-N/N1003A) were selected and maintained in MEM medium containing G418 (400 mg/ml). The antisense bcl-2transfected Bcl-2 expression cells (pSFFV-Bcl-2/pZeoSV-Antisense-bcl-2-N/N1003A) and the parallel vector-transfected Bcl-2 expression cells (pSFFV/pZeoSV-N/N1003A) were selected and maintained in media containing both G418 (400 mg/ml) and Zeocin (300 mg/ml). DNA probe preparation

Materials and methods

Rat c-fos and c-jun cDNA clones (kindly provided by Dr Tom Curran; Curran et al., 1987; Macgregor et al., 1990) and GAPDH cDNA clone (kindly provided by Dr Meng-sheng Qiu; Tso et al., 1985) were all amplified in DH-5a and purified by CsCl ultracentrifugation (Li et al., 1994). The cDNA inserts were recovered by restriction digestion and agarose gel purification as previously described (Li et al., 1995). The cDNA fragments were labeled with [a-32P]dATP (Amersham # PB 10204) as described before (Li and Spector, 1997).

Chemicals

RNA preparation and analyses

Various molecular biology reagents were purchased from Invitrogen Life Technologies (Gaithersburg, MD, USA),

Total RNAs were extracted from parental RLECs (N/N1003A), vector-transfected cells (pSFFV-N/N1003A), Oncogene


7320 Bcl-2-transfected cells (pSFFV-Bcl-2-N/N1003A), and both Bcl-2- and antisense-bcl-2-transfected cells (pSFFV-Bcl-2/ pZeoSV-Antisense-bcl-2-N/N1003A) as previously described (Li et al., 2001a, b). For Northern blots, 100 mg of total RNA was denatured, electrophoresed on formaldehyde–agarose gels (1.2%) and transferred to supported nitrocellulose membranes. The RNA blots were then UV crosslinked for 5 min. Prehybridization was conducted at 421C for 12 h in the following solution: 50% formamide, 6 SSPE (0.9 M NaCl, 72 mM sodium phosphate, pH 7.4, 7.2 mM EDTA), 5 Denhardt’s solution, 1% SDS and 200 mg/ml denatured and sheared herring sperm DNA. Hybridization was conducted for 36–42 h at the same temperature and in the same buffer containing 32P-labeled specific probes at a concentration of 5 106 cpm/ml. After hybridization, the filter was washed under high stringency conditions and then exposed to Kodak XAR-5 film for 12–48 h as previously described (Li et al., 1994; Li and Spector, 1997). For re-probing, bound radioactive probes were removed from the previously hybridized filter by washing twice with Tris-EDTA buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) at 801C. Protein preparation and Western blot The total proteins were prepared from N/N1003A or various transfected cell lines using 250 ml extraction buffer as previously described (Li et al., 1998). A 50 or 100 mg portion of total proteins in each sample was resolved by 10% SDS– polyacrylamide gel electrophoresis (SDS–PAGE) and transferred to supported nitrocellulose membranes. The protein blots were blocked with 5% milk in TBS (10 mM Tris-HCl, pH 8.0, 150 mM NaCl) overnight at 41C, and incubated with various antibodies against c-Fos, c-Jun, b-actin, total ERK1/2, phospho-ERK1/2, total p38, phospho-p38, total JNK1/2 and phospho-JNK1/2 at a dilution of 1 to 500 to 2000 (mg/ml) in 5% milk prepared in TBS. The secondary antibody is antimouse IgG or anti-goat IgG at a dilution of 1 to 1000 to 2000 (Amersham). Immunoreactivitiy was detected with an enhanced chemilluminescence detection kit according to the company’s instructions (ECL, Amersham Corp.) (Li et al., 2001a, b). Preparation of nuclear extracts Both pSFFV-N/N1003A and pSFFV-Bcl-2-N/N1003A cells were cultured in 175-cm2 flasks till confluence. The cells were either pretreated with 0.01% DMSO (mock) or MEK1/2 inhibitor, PD98059 (25 mM), JNK1/2 inhibitor, SP600125 (1 mM) or p38 inhibitor, PD169316 (2 mM), then washed twice with 5 ml of ice-sold PBS, and harvested into a 1.5 ml centrifuge tube with a plastic scraper. Pelleted cells were rapidly suspended in 400 ml of hypotonic buffer (10 mM HEPES (pH 7.9), 1.5 mM MgCl2, 10% glycerol, 10 mM KCl, 1 mM DTT, 0.5% Nonidet P-40, 0.5 mM PMSF, 100 mg/ml aprotinin) and incubated on ice for 15 min. After incubation, the samples were centrifuged at 2200 g for 2 min. The pelleted nuclei were resuspended in buffer D (20 mM HEPES (pH 7.9), 400 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 20% glycerol, 1 mM DTT, 0.5 mM PMSF, 100 mg aprotinin, 0.5% Nonidet

P-40) and incubated on ice for 20 min. The nuclei were further centrifuged at 8800 g for 5 min. The supernatant was collected and stored at 701C for gel mobility shifting assay. Gel mobility shifting assays Gel mobility shifting assays were conducted as previously described (Mao et al., 2001). The following oligos were used: 50 -CGCTTGATGAGTCAGCCGGAA-30 for AP-1 binding site; 50 -CGCTTGAatcggtAGCCGGAA-30 for mutated AP-1 binding site; 50 -AAATATTTGGGGTTTTTTTT-30 for LEDGF binding site. A 40 mg portion of nuclear extracts prepared from pSFFV-N/N1003A and pSFFV-Bcl-2-N/ N1003A cells without or with various pretreatments was incubated with 1 105 cpm of 32P-labeled double-stranded synthetic oligonucleotides for 30 min at 301C in a binding shifting buffer (1 mg/ml poly(dI-dC), 25 mM HEPES (pH 7.9), 40 mM KCl, 3 mM MgCl2, 0.1 mM EDTA, 1 mM DTT, 10% glycerol). For competition experiments, 50-fold of unlabeled double-stranded synthetic oligonucleotides were preincubated with the nuclear extracts for 10 min before the labeled probe was added to the reaction. In the precleared experiments, 2 mg antibody against c-JUN was preincubated with the nuclear extracts on ice for 10 min prior to addition of the 32P-labeled oligonucleotides. After the binding reaction, the mixtures were loaded onto 5% native PAGE and detected by autoradiography. Analysis of transient gene expression For reporter gene activity, the constructs of either luciferase or CAT reporter gene driven by the mini-prolactin gene promoter (for luciferase reporter gene; Xiang et al., 2000), chicken bA3/A1-crystallin promoter (Mcdermott et al., 1997), together with the control construct expressing b-galactosidase (Nieson et al., 1983) were introduced into both vector- and Bcl-2-transfected cells using electroporation as described above. At 2 h after transfection, the transfected cells were pretreated with 0.01% DMSO or 25 mM PD98059 (inhibitor for MEK1/2), 2 mM PD169316 (inhibitor for p38) and 1 mM SP600125 (inhibitor for JNK1/2) and then incubated for additional 22 or 48 h before being harvested for assays of luciferase, b-galactosidase and CAT activities as described previously (Xiang et al., 2000, 2002). Acknowledgements We thank Dr Stanley Korsmeyer for the human Bcl-2 expression construct, Dr Tom Curran for the rat c-jun and cfos cDNA clones, Dr Meng-Sheng Qiu for the GAPDH cDNA, Dr John Westwick for the luciferase reporter gene expression construct, Dr John Reddan for the RLECs (N/ N1003A) and Dr Joram Piatigorsky for the chicken bA3/A1CAT and pRSV-b-galactosidase expression constructs. This work was partially supported by NIH/NEI 11372, the Hormel Foundation, University of Minnesota Graduate School, and the Lotus Scholar Program Funds from Hunan Province Government and Hunan Normal University.

References Angel P and Karin M. (1991). Biochim. Biophys. Acta, 1072, 129–157. Bohmann D, Bos TJ, Admon A, Nishimura T, Vogt PK and Tjian R. (1987). Science, 238, 1386–1392. Oncogene

Casalino L, De Cesare D and Verde P. (2003). Mol. Cell. Biol., 23, 4401–4415. Chen DB and Davis JS. (2003). Mol. Cell. Endocrinol., 200, 141–154.


7321 Curran T and Franza Jr BR. (1988). Cell, 55, 395–397. Curran T, Gordon MB, Rubino KL and Sambucetti LC. (1987). Oncogene, 2, 79–84. de Moissac D, Mustapha S, Greenberg AH and Kirshenbaum LA. (1998). J. Biol. Chem., 273, 23946–23951. Dhawan P and Richmond A. (2002). J. Biol. Chem., 277, 7920–7928. El-Deiry WS. (2003). Oncogene, 22, 7486–7495. Fatma N, Singh DP, Shinohara T and Chylack Jr LT. (2001). J. Biol. Chem., 276, 48899–48907. Feng L, Balakir R, Precht P and Horton Jr WE. (1999). J. Cell. Biochem., 74, 576–586. Froesch BA, Aime-Sempe C, LeberB, Andrews D and Reed JC. (1999). J. Biol. Chem., 274, 6469–6475. Fuhlbrigge RC, Fine SM, Unanue ER and Chaplin DD. (1988). Proc. Natl. Acad. Sci. USA, 85, 5649–5653. Gross A, McDonnell JM and Korsmeyer SJ. (1999). Genes Dev., 13, 1899–1911. Gruda MC, Kovary K, Metz R and Bravo R. (1994). Oncogene, 9, 2537–2547. Hilton M, Middleton G and Davies AM. (1997). Curr. Biol., 7, 798–800. Hockenbery D, Nunez G, Milliman C, Schreiber RD and Korsmeyer SJ. (1990). Nature, 348, 334–336. Hunter T and Karin M. (1992). Cell, 70, 375–387. Jiang B, Xu S, Hou X, Pimentel DR, Brecher P and Cohen RA. (2004). J. Biol. Chem., 279, 1323–1329. Johnson PF and McKnight SL. (1989). Annu. Rev. Biochem., 58, 799–839. Karin M. (1995). J. Biol. Chem., 270, 16483–16486. Kerppola TK and Curran T. (1991). Curr. Opin. Struct. Biol., 1, 71–79. Korsmeyer SJ. (1992). Blood, 80, 879–886. Kubo E, Fatma N, Sharma P, Shinohara T, Chylack Jr LT, Akagi Y and Singh DP. (2002). J. Mol. Biol., 320, 1053–1063. Kukushkin AN, Abramova MV, Svetlikova SB, Darieva ZA, Pospelova TV and Pospelov VA. (2002). Oncogene, 21, 719–730. Kumar NV and Bernstein LR. (2001). J. Biol. Chem., 276, 32362–32372. Kyriakis JM and Avruch J. (2001). Physiol. Rev., 81, 807–869. Li DW-C, Fass U, Huizar I and Spector A. (1998). Eur. J. Biochem., 257, 351–361. Li DW-C, Liu J-P, Wang J, Mao Y-W and Hou L-H. (2003). Invest. Ophthalmol. Vis. Sci., 44, 5277–5286. Li DW-C and Spector A. (1997). Mol. Cell. Biochem., 173, 59–69. Li DW-C, Xiang H, Fass U and Zhang X-Y. (2001a). Invest. Ophthalmol. Vis. Sci., 42, 2603–2609. Li DW-C, Xiang H, Mao Y-W, Wang J, Fass U, Zhang X-Y and Xu C. (2001b). Exp. Cell Res., 266, 279–291. Li N and Karin M. (1999). FASEB J., 13, 1137–1143. Li W-C, Kuszak JR, Dunn K, Wang R-R, Ma W-C, Wang G-M, Spector A, Leib M, Cotliar AM, Weiss M, Espy J, Howard G, Farris RL, Auran J, Donn A, Hofeldt A, Mackay C, Merriam J, Mittl R and Smith TR. (1995). J. Cell Biol., 130, 169–181. Li W-C, Wang G-M, Wang R-R and Spector A. (1994). Exp. Eye Res., 59, 179–190.

Lu PJ, Lu QL, Rughetti A and Taylor-Papadimitriou J. (1995). J. Cell Biol., 129, 1363–1378. Macgregor PF, Abate C and Curran T. (1990). Oncogene, 5, 451–458. Mao Y-W, Xiang H, Wang W, Korsmeyer SJ, Reddan J and Li DW-C. (2001). J. Biol. Chem., 278, 43435–43445. McDermott JB, Cvekl A and Piatigorsky J. (1997). Invest. Ophthalmol. Vis. Sci., 38, 951–959. Mischak H, Seitz T, Janosch P, Eulitz M, Steen H, Schellerer M, Philipp A and Kolch W. (1996). Mol. Cell. Biol., 16, 5409–5418. Miyashita T, Kitada S, Krajewski S, Horne WA, Delia D and Reed JC. (1995). J. Biol. Chem., 270, 26049–26052. Monje P, Marinissen MJ and Gutkind JS. (2003). Mol. Cell. Biol., 23, 7030–7043. Nieson DA, Chou J, MacKrell AJ, Casadaban MJ and Steiner DF. (1983). Proc. Natl. Acad. Sci. USA, 80, 5198–5202. O’Reilly LA, Harris AW and Strasser A. (1997). Int. Immunol., 9, 1291–1301. Persons DL, Yazlovitskaya EM and Pelling JC. (2000). J. Biol. Chem., 275, 35778–35785. Reddan JR, Cheplinsky AB, Dziedzic DC, Piatigorsky J and Goldenberg EM. (1986). Differentiation, 33, 168–174. Reed JC. (1994). J. Cell Biol., 124, 1–6. Ricca A, Biroccio A, Del Bufalo D, Mackay AR, Santoni A and Cippitelli M. (2000). Int. J. Cancer, 86, 188–196. Rosenberger SF, Finch JS, Gupta A and Bowden GT. (1999). J. Biol. Chem., 274, 1124–1130. Rosen LB, Ginty DD, Weber MJ and Greenberg ME. (1994). Neuron, 12, 1207–1221. Sablina AA, Chumakov PM, Levine AJ and Kopnin BP. (2001). Oncogene, 20, 899–909. Schwarz CS, Seyfried J, Evert BO, Klockgether T and Wullner U. (2002). NeuroReport, 13, 2439–2442. Shinohara T, Singh DP and Fatma N. (2002). Prog. Retin. Eye Res., 21, 341–358. Singh DP, Fatma N, Kimura A, Chylack Jr LT and Shinohara T. (2001). Biochem. Biophys. Res. Commun., 283, 943–955. Tso JY, Sun XH, Kao TH, Reece KS and Wu R. (1985). Nucleic Acids Res., 13, 2485–2502. Tsujimoto Y, Cossman J, Jaffe E and Croce CM. (1985). Science, 228, 1440. Vairo G, Soos TJ, Upton TM, Zalvide J, DeCaprio JA, Ewen ME, Koff A and Adams JM. (2000). Mol. Cell. Biol., 20, 4745–4753. Vaux D, Cory A and Adams J. (1988). Nature, 335, 440. Xiang H, Wang J, Mao Y-W and Li DW-C. (2000). Biochem. Biophys. Res. Commun., 278, 503–510. Xiang H, Wang J, Mao Y-W, Liu M, Reddy V and Li DW-C. (2002). Oncogene, 21, 3784–3791. Young MR, Nair R, Bucheimer N, Tulsian P, Brown N, Chapp C, Hsu TC and Colburn NH. (2002). Mol. Cell. Biol., 22, 587–598. Yuan PX, Huang LD, Jiang YM, Gutkind JS, Manji HK and Chen G. (2001). J. Biol. Chem., 276, 31674–31683. Yue J and Mulder KM. (2000). J. Biol. Chem., 275, 30765–30773. Zhan Q, Kontny U, Iglesias M, Alamo Jr I, Yu K, Hollander MC, Woodworth CD and Fornace Jr AJ. (1999). Oncogene, 18, 297–304.

Oncogene