![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 19, Issue 11, 4717-4729, November 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
FosB/2FosB Encoded by Alternatively Spliced Forms of fosB TranscriptsDivision of Neurofunctional Genomics, Department of Immunobiology and Neuroscience, Medical Institute of Bioregulation, Kyushu University, Fukuoka 812-8582, Japan
Submitted August 9, 2007;
Revised August 8, 2008;
Accepted August 20, 2008
Monitoring Editor: William P. Tansey
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
fosB mRNAs, by alternative splicing of an exonic intron. The former encodes full-length FosB. The latter encodes FosB and 2FosB by alternative translation initiation, and both of these lack the C-terminal transactivation domain of FosB. We established two mutant mouse embryonic stem (ES) cell lines carrying homozygous fosB-null alleles and fosBd alleles, the latter exclusively encoding FosB/2FosB. Comparison of their gene expression profiles with that of the wild type revealed that more than 200 genes were up-regulated, whereas 19 genes were down-regulated in a FosB/2FosB-dependent manner. We furthermore found that mRNAs for basement membrane proteins were significantly up-regulated in fosBd/d but not fosB-null mutant cells, whereas genes involved in the TGF-β1 signaling pathway were up-regulated in both mutants. Cell-matrix adhesion was remarkably augmented in fosBd/d ES cells and to some extent in fosB-null cells. By analyzing ES cell lines carrying homozygous fosBFN alleles, which exclusively encode FosB, we confirmed that FosB negatively regulates cell-matrix adhesion and the TGF-β1 signaling pathway. We thus concluded that FosB and FosB/2FosB use this pathway to antagonistically regulate cell matrix adhesion. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Shaulian and Karin, 2001; Miura et al., 2005). Among fos family genes, only fosB produces two forms of mature transcripts, namely fosB and fosB mRNAs, by alternative splicing of an exonic intron in exon 4 (see Figure 1A). The former encodes full-length FosB protein, whereas the latter encodes FosB protein (Nakabeppu and Nathans, 1991). The FosB protein is a C-terminal-truncated form of FosB and lacks the C-terminal transactivation domain and the TATA-binding protein (TBP)-binding domain. It has been proposed that FosB dramatically enhances Jun transcription regulation of AP-1–dependent promoters, whereas, based on reporter assays, FosB suppresses this Jun function in a dominant-negative manner (Nakabeppu and Nathans, 1991).
We have reported that FosB has the potential to trigger cell proliferation of quiescent embryonic cell lines or primary neuronal precursor cells and to induce delayed morphological changes or apoptosis dependent on the cell type (Nakabeppu et al., 1993; Nishioka et al., 2002; Tahara et al., 2003; Kurushima et al., 2005), whereas Nestler's group have found that FosB plays an essential role in long-term adaptive changes in the brain associated with diverse conditions in their studies of mice with an inducible-FosB transgene (McClung et al., 2004). Their mice provide a suitable model for the inducible expression of FosB, but the authentic role of this protein remains unclear because the effects of exogenous FosB expression are likely to be modified by endogenous FosB or FosB expression, or vice versa. Furthermore, it has been reported that fosB mRNA also produces 2FosB protein lacking a N-terminal Fos homology domain (FH) by alternative translation initiation (Sabatakos et al., 2000). To elucidate the authentic function of each protein, it is essential to establish mutants expressing only one of these from an endogenous fosB gene.
In the present study, we initially established two mutant embryonic stem (ES) cell lines carrying homozygous fosB-null alleles and fosBd alleles. The former cells express neither FosB nor FosB/2FosB, and the latter express only FosB and 2FosB from artificially manipulated fosBd alleles. Both mutant ES cells are devoid of FosB; therefore common phenotypes shared by the two mutants but not by the wild type depict the effects of FosB deficiency, whereas opposite phenotypes in the two are considered to be due to FosB/2FosB. By comparing the gene expression profiles and cellular functions of these two mutants with the wild-type ES cells, we found that FosB/2FosB positively regulates cell-matrix adhesion. Finally, by establishing ES cell lines carrying homozygous fosBFN alleles that exclusively encode FosB, we confirmed that FosB negatively regulates cell-matrix adhesion.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
FosB protein, the two codons (GTG, AGA) after the last codon (GAG) for FosB (Glu237) were changed. These also function as an alternative splicing donor site for the exonic intron in exon 4 of the fosB gene to produce FosB mRNA and were replaced by two tandem stop codons (TAG, TGA) by PCR-mediated site-directed mutagenesis using a Mega-primer protocol (see Figure 1B; Sarkar and Sommer, 1990). The initial PCR was performed with FB8 and FB9, a set of primers with designed mutations, and the BglII fragment (4432 base pairs) containing exon 3 and exon 4 of the fosB gene was used as a template using high-fidelity Pyrobest DNA polymerase (Takara Bio, Tokyo, Japan). The second PCR was performed with primer FB10 and with the resulting PCR product as a Mega-primer and using the same template as with the first PCR using zTaq DNA polymerase (Takara Bio).
Next, the AatII-SgrAI fragment, excised from the PCR product, was reciprocally exchanged with the corresponding region surrounding exon 4 of the targeting vector. In the pTV-fosBGN vector, a fosB gene fragment extending from the NcoI site of exon 2 to the PstI site of exon 3 was replaced with d2EGFP cDNA in-frame and followed the inverted neo cassette flanked by two loxP sites. The d2EGFP cDNA used was excised from the d2EGFP-N1 vector, which encodes a destabilized variant of EGFP (Clontech, Palo Alto, CA). In each construct, the modified mouse genomic fragment was flanked by herpes simplex virus (HSV)-1 and HSV-2 thymidine kinase (TK) cassettes (Rancourt et al., 1995). In the pTVfosBFN vector, we substituted three bases (GAGGTGAGA
GAAGTTCGA) in the alternative splicing donor site and two bases (AGTGAATCTGAA) in the splice acceptor site by PCR-mediated site-directed mutagenesis in order to avoid alternative splicing, which results in the production of FosB (see Figure 7A). The initial PCR was performed with a primer set consisting of FB12 and FB15, the latter carrying designed mutations in the splice acceptor site, and the BglII fragment containing exon 3 and exon 4 of the fosB gene was used as a template. The second PCR was performed with primer FB16 carrying designed mutations in the splice donor site and with the resulting PCR product as a Mega-primer, using the same template. The third PCR was performed with primer FB10 and the resulting PCR product as a Mega-primer, using the same template as with the first PCR. Finally, the AatII-SgrAI fragment, excised from the third PCR product, was reciprocally exchanged with the corresponding region surrounding exon 4 of the targeting vector. PCR primers used for construction of targeting vectors are listed in Table 1.
|
; Ide et al., 2003). To obtain quiescent ES cells, cells subcultured twice without feeder cells were maintained in ES medium supplemented with 0.5% FBS for 3 d without changing the medium. To obtain serum-stimulated ES cells, the quiescent ES cells were stimulated by exchanging the medium for fresh ES medium supplemented with 20% FBS.
Generation of fosB Mutant ES Cell Lines
CCE cells were electroporated with the linearized targeting vector DNA, and colonies doubly resistant to G418 and gancyclovir were selected as described (Robertson, 1987; Ide et al., 2003). Correctly targeted clones were identified by Southern blotting and genomic PCR analyses. The introduced mutations were confirmed by direct sequencing of each genomic PCR product containing the entire exonic intron in fosB exon 4 amplified with a set of primers, FB7 and FB12, using FB11 as the sequencing primer. Twenty clones of fosB+/dN, 4 clones of fosB+/GN, and 3 clones of fosB+/FN heterozygous ES mutants were independently identified. Subsequently, each heterozygous ES mutant was cultured in the presence of increasing concentrations of G418 (1.2–1.5 mg/ml) for 9 d and then 1.0 mg/ml G418 for 5 d in order to isolate ES cells carrying the two mutant alleles with the neo cassette generated by homologous recombination of the wild-type and mutant alleles. Resistant colonies were picked up and homozygotes for each mutant allele were identified by Southern blotting and direct sequencing of their genomic PCR products. Five clones of fosBdN/dN, 13 clones of fosBGN/GN, and 3 clones of fosBFN/FN were independently established. To excise the neo cassette flanked by the loxP sites in the targeted allele, a Cre expression vector, pBS185 (Invitrogen, Carlsbad, CA), was introduced into fosBdN/dN or fosB+/dN ES cells by electroporation. Excision of the neo cassette in each colony formed in the absence of G418 was confirmed by Southern blotting and genomic PCR (see Figure 2, A and E). Thus, the generated allele was designated the fosBd allele, and 9 fosBd/d and 10 fosB+/d clones were independently isolated. No genomic PCR product was amplified from any of the established homozygous clones using two-primer sets for part of the Cre coding sequence and the CMV promoter region in pBS185.
Western Blot Analysis
Nuclei were isolated from ES cells maintained in the absence of feeder cells using NP-40 lysis buffer (10 mM Tris-HCl, pH 7.4, 10 mM KCl, 1.5 mM MgCl2, 0.5% NP-40, 0.1% DNase, 0.05% RNase, and protease/phosphatase inhibitors). Each sample was subjected to 12.5% SDS-PAGE and Western blot analysis as previously described (Nakabeppu et al., 1993; Tsuchimoto et al., 2001). Anti-FosB(N) was raised against amino acids 79–130 of the N-terminus of FosB or FosB, whereas anti-FosB(C) was raised against amino acids 245–315 of the C-terminus of FosB (Nakabeppu and Nathans, 1991). Anti-c-Fos (sc-52) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA).
cDNA Microarray Analysis
Total RNA was purified from ES cells grown in the absence of feeder cells using ISOGEN (Nippon Gene, Tokyo, Japan), according to the manufacturer's instructions. For DNA microarray experiments, total RNA (10 µg) was labeled using an Agilent Linear Amplification/Labeling kit (Agilent Technologies, Wilmington, DE) according to the manufacturer's instructions. A mixture was made of 1 µg of each Cy3-labeled wild-type and Cy5-labeled mutant cDNA or each Cy3-labeled mutant and Cy5-labeled wild-type cDNA, and these cDNAs were then hybridized to Agilent Mouse cDNA Microarrays (G4104A, design file number: 000522R000679, Agilent Technologies) with 8500 unique clones from the Incyte mouse UniGene 1 clone set, according to the manufacturer's hybridization protocol. After washing, the microarray slides were analyzed with an Agilent G2565AA microarray scanner system. These experiments were carried out in duplicate using exchanged dye-labeled cDNA probes (as in Cy3 and Cy5 dye-swapping experiments). Data analysis was performed using Agilent Feature Extraction software (Ver. A.6.1.1) and Excel 2002 (Microsoft, Redmond, WA). Genes whose expression levels in mutant ES cells were altered significantly in comparison to those in wild-type ES cells with a p value < 0.01 were retrieved.
Cell-Matrix Adhesion Assay
ES cells maintained in the absence of feeder cells were prepared for the cell-matrix adhesion assay as 1) exponentially growing, 2) quiescent, or 3) serum-stimulated for 4 h, as described above. Each preparation of ES cells was plated at various concentrations (0.5 x 105, 1.0 x 105, 2.0 x 105 and 4.0 x 105 cells per well [1.9 cm2]) on five types of coated and noncoated 24-well plates in ES medium supplemented with 20% FBS and incubated for 4 h. Surfaces of the 24-well plates were coated with five different matrices consisting of 0.1% gelatin, 10 µg/ml collagen type I, 10 µg/ml collagen type IV, 10 µg/ml fibronectin, and 10 µg/ml laminin (Sigma-Aldrich, St. Louis, MO). After fixation with 25% glutaraldehyde, attached cells were stained with 0.3% crystal violet, and the dye was extracted in 50% ethanol to measure absorbance at 595 nm, which provides a reading that is proportional to the number of cells attached. A one-way ANOVA with Dunnett's post hoc test was performed to assess the number of cells that attached at a concentration of 2.0 x 105 cells per well, using KaleidaGraph (Synergy Software, Reading, PA).
Southern Blot Analysis
Southern blot analysis was carried out as described (Ide et al., 2003; Xu et al., 2003). Isolated genomic DNA was digested with XhoI and SpeI, or with NotI and SacI, separated in 0.8% agarose gel, and transferred to nylon membranes (GE Healthcare Bio-Science, Piscataway, NJ) by capillary blotting. The membrane was hybridized with a random-primed 32P-labeled probe using standard methods.
RT-PCR Analysis
Total RNA was prepared from cultured cells using ISOGEN. First-strand cDNA was synthesized using a first-strand cDNA synthesis kit (GE Healthcare Bio-Sciences) according to the manufacturer's instructions. Subsequently, PCR was performed with the primers listed in Table 1. For quantitative real-time RT-PCR for Tgfb1i1 mRNA, reactions in triplicate were performed using an Applied Biosystems 7500 Real-Time PCR System with SYBR Green Master Mix (Applied Biosystems, Foster City, CA) and the primers listed in Table 1. The expression level was determined by normalization with Gapdh mRNA, and a relative level of Tgfb1i1 mRNA in each cell line to that in the wild-type ES cells was calculated.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
FosB and 2FosB, but not the FosB protein, we introduced two tandem stop codons into the fosB gene after the last codon for Glu237 of FosB (Figure 1B). These base substitutions also disrupt the splicing donor sequence, and it was therefore expected that the fosBd allele would produce only the longer form of the fosB transcript, designated fosBd mRNA. In pTVfosBdN, a targeting vector shown in Figure 1B, the altered fosB allele (fosBdN) contains a neo cassette for positive selection, which is flanked by loxP sites. The neo cassette can be excised by Cre recombinase, thereby generating the fosBd allele. We also constructed pTVfosBGN, a targeting vector in which d2EGFP cDNA encoding a destabilized variant of enhanced green fluorescent protein (EGFP) was placed in exon 2 in frame, and the remainder of exon 2 and the entire exon 3 were replaced by a neo cassette flanked by loxP sites.
|
FosB (Figure 2D, dN/dN and d/d), whereas the fosBGN allele produced the fosB:d2EGFP fusion transcript (Figure 2D, +/GN, GN/GN), which may encode a nonfunctional FosB:d2EGFP fusion protein. In fosB+/dN and fosBd/d mutant ES cells, we confirmed the base substitutions introduced at the alternative splicing site by direct sequencing of their genomic PCR fragments (Figure 2E).
|
fosB mRNAs, respectively, which encode FosB and FosB/2FosB, were transiently induced in wild-type cells (Figure 3A, +/+), whereas only the longer form of the transcripts, namely fosBd mRNA, which is expected to encode FosB and 2FosB, was transiently induced in fosBd/d cells (Figure 3A, d/d). In serum-stimulated fosBGN/GN ES cells, two types of fosB transcripts were detected, but the level of the shorter form was much lower than that of the longer form (Figure 3A, GN/GN), suggesting that replacement of exons 2–3 with an EGFP-neo cassette may alter the efficiency of the alternative splicing. After serum stimulation, levels of c-fos, fra-1, and fra-2 mRNAs were also transiently increased in fosBd/d and fosBGN/GN ES cells as they were in the wild type (Figure 3A). Among the three cell lines, the highest levels of fra-2 mRNA were observed in fosBd/d cells, with or without serum stimulation. Furthermore, it was noteworthy that the levels of c-fos and fosB transcripts in the quiescent fosBd/d and to a lesser extent fosBGN/GN cells or in these cells 4 h after serum stimulation, were apparently higher than in the wild type, probably because of their lack of the functional FosB protein, which can suppress transcription of c-fos and fosB genes (Nakabeppu and Nathans, 1991; Lazo et al., 1992).
|
FosB/2FosB proteins in each type of mutant ES cell, Western blotting analyses were performed as shown in Figure 3, B–D. Expression of fosB gene products was barely detectable in quiescent ES cells, irrespective of the fosB genotype. In the serum-stimulated wild-type, 43- and 32/36-kDa polypeptides were detected by Western blotting with anti-FosB(N), which reacts with both FosB and FosB, and the 43-kDa polypeptide was also detected by anti-FosB(C), which reacts only with FosB (Nakabeppu and Nathans, 1991). In the fosBd/d cells, however, serum stimulation induced a significantly increased expression of the 32/36-kDa polypeptide, namely FosB, whose peak levels were 10 times as high as those of the wild type. There were no detectable polypeptides in the serum-stimulated fosBGN/GN cells using either type of anti-FosB, and these cells were therefore designated fosB-null ES cells. It was apparent that serum-stimulated fosBd/d cells express a 24-kDa polypeptide that was recognized only by anti-FosB(N), and its level was higher than that in the wild type, indicating that this polypeptide is 2FosB produced by alternative translation initiation of fosB mRNA (Sabatakos et al., 2000). We further found that the expression level of c-Fos protein was apparently down-regulated in serum-stimulated fosBd/d ES cells only (Figure 3E) and that similar levels of Fra-1 and Fra-2 proteins were induced in all three ES cells after serum stimulation (data not shown).
Altered Gene Expression Profiles in fosB Mutant ES Cells
Despite the altered expression of fosB gene products, fosBd/d, and fosB-null and wild-type ES cells exhibited essentially the same growth rate under normal growth conditions and efficiently differentiated into a neuronal lineage in the presence of retinoic acids. These data are presented as Supplemental Figure S1. To expand our understanding of the biological significance of FosB and FosB/2FosB, we next compared the gene expression profile of each mutant with that of wild-type ES cells using total RNA prepared from cells 4 h after serum stimulation (Figure 4A) and obtained reliable data on 5290 of 8500 unique probes. As shown in Figure 4, B and C, expression of 267 genes (5.05%) was significantly altered in fosBd/d cells, whereas that of only 44 genes (0.83%) was affected in fosB-null cells, compared with the wild type. Expression of 203 genes was significantly up-regulated in the fosBd/d cells, with no remarkable change noted in the fosB-null cells (Figure 4D).
|
1 (Lama1) and Collagen type IV 1 (Col4a1), which are major components of the basement membrane in embryonic tissue (Ekblom et al., 2003), were the most significantly up-regulated in fosBd/d cells. Three additional genes for Laminin B1 (Lamb1-1), 
1 (Lamc1) and Collagen type IV 2 (Col4a2), proteins comprising the basement membrane, were likely up-regulated in fosBd/d cells (Table 3). As shown in Figure 5A, RT-PCR analyses confirmed that expression of Lama1, Lamb1-1, Lamc1, Col4a, and Col4a2 genes was up-regulated in serum-stimulated fosBd/d ES cells, and to a much lesser extent in fosB-null cells, in comparison to the wild type.
|
|
|
FosB/2FosB in quiescent fosBd/d ES cells must also be higher than in the quiescent wild type, as expected from the data shown in Figure 3A. With a large quantity of nuclear extract (200 µg protein), small but significant amounts of FosB and 2FosB were detected not only in the quiescent but also in the exponentially growing fosBd/d cells as well (Figure 5B). In contrast, neither wild-type nor fosB-null cells exhibited a detectable expression of FosB or FosB/2FosB under these conditions. Therefore, up-regulation of mRNAs for the three types of Laminin-1 and two types of Collagen (IV), but not for GATA6 or Akt 1, had occurred in the exponentially growing fosBd/d ES cells (Figure 5C), and this was likely due to the increased level of FosB/2FosB. However, it is noteworthy that under conditions of exponential growth, expression levels of Lama1, Lamb1-1, Lamc1, Col4a, Col4a2, and Gata6 genes in wild-type cells were slightly higher than those in fosB-null cells, whereas the lowest expression of the Thbs-1 gene was observed in the wild-type cells (Figure 5C), suggesting that FosB or FosB/2FosB may be expressed in the exponentially growing wild-type cells, thus contributing to their gene expression regulation.
Increased Cell-Matrix Adhesion in fosBd/d and fosB-null ES Cells
As shown in Table 3, more than 10 genes encoding proteins involved in basement membrane regulation, such as Integrin-linked kinase (Ilk), Nidogen 1 (Nid1), Procollagen prolyl 4-hydroxylase 1 and 2 (P4ha1, P4ha2), Procollagen-lysine, 2-oxoglutarate 5-dioxygenase 1 and 2 (Plod1, Plod2), and Thsb1, were up-regulated in fosBd/d ES cells, as were those for major basement membrane components (Supplementary Figure S2). Thus, we examined whether fosBd/d ES cells undergo any alteration in cell-matrix adhesion (Figure 6). Exponentially growing fosBd/d cells exhibited significantly higher adhesion to uncoated plates or to plates coated with gelatin, collagen type I or IV, laminin, and fibronectin, compared with exponentially growing wild-type cells, whereas fosB-null cells exhibited essentially the same adhesive behavior as that of wild-type cells (Figure 6, A–F). Quiescent wild-type cells were found to have dramatically lost their adhesion to plates coated with or without collagen type I, IV or gelatin, and to have moderately lost adhesion to those coated with laminin or fibronectin. In contrast, quiescent fosBd/d cells sustained efficient adhesion under any condition, compared with the exponentially growing cells. Interestingly, quiescent fosB-null cells exhibited essentially the same high levels of adhesion to laminin and fibronectin as did fosBd/d cells, but only an intermediate adhesion to other coating materials or to uncoated plates (Figure 6, G–L). Four hours after serum stimulation of the quiescent cells showed increase in their adhesion to all matrices. Among these, fosBd/d cells exhibited the highest adhesion to uncoated plates or to plates coated with collagen type I and IV, whereas fosB-null cells exhibited the highest adhesion to plates coated with gelatin, laminin, and fibronectin (Figure 6, M–R).
|
Up-Regulation of TGF-β 1 Signaling in fosBd/d ES Cells
It has been reported that constitutively active Akt caused an up-regulation of mRNAs for Laminin 1 and Collagen type IV isotypes in ES cells (Li et al., 2001). Although the protein level of Akt was slightly up-regulated in fosBd/d ES cells with or without serum stimulation, phosphorylation of Akt (Ser473, Thr 308), which was barely detectable in the quiescent cells, was transiently induced after serum stimulation, as observed in fosB-null and wild-type cells. Furthermore, GSK3β, which is a target of Akt and is expressed constitutively, was also transiently phosphorylated after serum stimulation in all three lines (Supplementary Figure S3).
An increased expression of GATA6 in ES cells is also known to induce expression of laminin 1 (Li et al., 2004); however, among the three ES cell lines, the level of Gata6 mRNA was remarkably decreased in exponentially growing fosBd/d cells (Figure 5C). We suggest that the increased expression of mRNAs encoding the basement membrane proteins and related proteins in fosBd/d ES cells is dependent on neither the Akt signaling pathway nor transactivation by GATA6.
Thrombospondin 1/TGF-β1 signaling is known to up-regulate the expression of genes encoding basement membrane proteins, thereby increasing cell-matrix adhesion (Lawler, 2002). As shown in Figure 5, A and C, we found that expression levels of Thbs1 mRNA under quiescent conditions were lower, but that they differed significantly among the three ES cell lines, compared with those in exponentially growing or serum-stimulated cells, and their order (fosBd/d > fosB-null > wild type) was the same as that of their cell-matrix adhesion under the quiescent conditions (Figure 6, G–L). Because there was no probe for the TGF-β1 gene (Tgfb1) in the cDNA microarray used, we examined its expression level by RT-PCR analysis (Figure 5D) and confirmed an increased expression of Tgfb1 in serum-stimulated fosBd/d, but not in fosB-null ES cells, in comparison to the wild type. Furthermore, expression of TGF-β1–induced transcript 1 (Tgfb1i1) was increased in fosBd/d cells and to a lesser extent in fosB-null cells, compared with the wild type (Figure 5D). Again, we found that expression levels of Tgfb1 and Tgfb1i1 mRNA under quiescent conditions were low but differed significantly among the three ES cell lines and that fosBd/d ES cells exhibited the highest levels of their expression. Thrombospondin 1 is known to activate the latent form of TGF-β1 through protein–protein interaction (Ren et al., 2006). Therefore, TGF-β1 signaling is likely to be active in quiescent fosBd/d cells, and to a lesser extent in quiescent fosB-null ES cells as well, reflecting the expression levels of Tgfb1 and Thbs1 mRNAs.
Down-Regulation of Cell-Matrix Adhesion and TGF-β1 Signaling in fosBFN/FN ES Cells
To verify the role of FosB in cell-matrix adhesion and TGF-β1 signaling, we established fosBFN/FN ES cells in which FosB protein is exclusively translated from fosBF mRNA (Figure 7A). Mutations introduced at alternative splicing donor and acceptor sites to avoid alternative splicing were confirmed by genomic sequencing (Supplementary Figure S4A). The exclusive expression of the longer fosBF mRNA in exponentially growing fosBFN/FN cells was confirmed by RT-PCR, and its level was equivalent to those of fosB or fosB mRNAs in wild-type cells or to that of fosBd mRNA in fosBdN/dN cells (Figure 7B). In Western blotting analysis, the level of FosB protein in exponentially growing fosBFN/FN cells was again below the limit of detection (data not shown); however, we detected anti-FosB(C) immunoreactivity in exponentially growing fosBFN/FN and wild-type cells but not in fosBd/d cells (Supplementary Figure S4B), confirming that in exponential growth conditions, fosBFN/FN and wild-type cells do indeed express FosB protein.
|
FosB/2FosB up-regulates its expression.
Because it was apparent that the fosBd allele expresses an increased level of the modified transcript, namely fosBd mRNA, compared with the wild-type (fosB+) or to the modified allele with the neo cassette (fosBdN; Figure 7B), we also analyzed the expression levels of genes encoding basement membrane proteins and those involved in TGF-β1 signaling in fosBdN/dN ES cells and found that expression levels of most of the genes examined were not significantly altered compared with levels in wild-type ES cells (data not shown), suggesting that the increased expression of these genes in fosBd/d ES cells is dependent on the increased level of FosB/2FosB, but not on the loss of FosB expression.
We finally examined whether fosBFN/FN ES cells undergo any alteration in cell-matrix adhesion in comparison to other cell lines under exponential growth conditions (Figure 7D). Exponentially growing fosBd/d cells exhibited the highest adhesion to plates coated or not coated with gelatin or laminin, compared with wild-type, fosB+/d, fosBdN/dN, fosBFN/FN, and fosB-null ES cells. The fosBFN/FN cells exhibited the lowest adhesion to all matrices, especially to the uncoated and gelatin-coated plates, whereas fosB-null or fosB+/d ES cells again exhibited a slight increase in adhesion compared with wild-type ES cells (Figure 7D), indicating that FosB itself negatively regulates cell-matrix adhesion. The fosBdN/dN ES cells expressing the lower level of fosBd mRNA equivalent to that of fosB mRNA in wild-type cells exhibited essentially the same level of cell-matrix adhesion as that of wild-type ES cells.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
FosB, whereas in the latter alternative splicing of an exonic intron in exon 4 was avoided by introducing five base substitutions at splicing donor and acceptor sites without amino acid substitution. We established ES cell lines that are homozygous for the fosBd, fosBdN and fosBFN alleles after this knockin mutagenesis. These fosBd/d and fosBdN/dN homozygotes produce a modified fosBd mRNA that encodes only FosB and 2FosB, whereas the fosBFN/FN homozygotes produce a modified fosBF mRNA encoding only FosB. Interestingly, the fosBd allele is likely to produce more fosBd mRNA than is the fosBdN allele, which produces an amount of transcript that is equivalent to that of the fosB+ allele, resulting in greater production of its translation products. In cells with the fosBdN allele, a neo cassette placed in the opposite direction may interfere with efficient transcription or processing of the fosBd transcript. We confirmed that levels of FosB and to a lesser extent 2FosB proteins expressed in the fosBd/d homozygotes were significantly higher than the total amounts of FosB, FosB, and 2FosB proteins expressed in wild-type ES cells in any culture conditions examined. In addition to the increased level of the fosBd mRNA in fosBd/d cells, the higher stability of FosB protein (Chen et al., 1997) may account for this phenomenon. It is noteworthy that in serum-stimulated fosBd/d ES cells, the expression level of c-Fos but not Fra-1 and -2 proteins was apparently lower than in serum-stimulated wild-type and fosB-null cells (Figure 3E). It is most likely that the significantly higher levels of FosB and 2FosB proteins must cause depletion of free Jun proteins, which are also c-Fos protein partners; thus any liberated c-Fos, which is known to be a very unstable protein (Acquaviva et al., 2001), may be rapidly degraded even in serum-stimulated cells.
To explore the impact of FosB or FosB/2FosB on gene expression, we performed cDNA microarray analyses using RNA prepared from serum-stimulated wild-type, fosBd/d, and fosB-null ES cells. We found that 203 genes were up-regulated and 19 were down-regulated in fosBd/d cells compared with fosB-null ES cells. However, compared with wild type, only 10 genes were up-regulated in both types of mutant ES cells, whereas three genes were down-regulated (Figure 4D). Because both mutant ES cells were devoid of FosB, genes whose expression was similarly altered in the two mutants are likely to be regulated by FosB. In contrast, genes whose expression was different between the two are considered to be regulated by FosB/2FosB. These results are unexpected in that they imply that expression of a substantial number of genes is directly or indirectly regulated by FosB and that expression of a much smaller number of genes is regulated by FosB, because FosB is a more potent transactivator than FosB or 2FosB (Nakabeppu and Nathans, 1991; Sabatakos et al., 2008).
Reflecting the fact that more than a quarter of the genes up-regulated in fosBd/d ES cells encode cell-adhesion–related proteins (Table 2, Figure 5), cell-matrix adhesion of fosBd/d and to a lesser extent fosB-null ES cells was more efficient than that of the wild-type cells under the three respective culture conditions of exponential growth, quiescence, and serum stimulation. Cell-matrix adhesion of wild-type cells was the lowest at quiescence among the three conditions; however, fosBd/d or fosB-null cells exhibited significantly increased cell-matrix adhesion even in quiescence, compared with wild-type cells (Figure 6). Under conditions of exponential growth in the presence of serum or with serum stimulation of quiescent cells, all three cell lines exhibited significantly increased cell adhesion compared with that seen under serum-deprived quiescent conditions. As shown in Supplementary Figure S3, phosphorylation of Akt was also significantly increased in all three cell lines after serum stimulation. It is known that growth factors in serum activate Akt and that activated Akt induces expression of various cell-adhesion–related proteins and thus enhances cell adhesion (Li et al., 2001; Lim et al., 2003). Therefore, the absence of activated Akt in the quiescent wild-type cells may account for their having the lowest level of cell adhesion, whereas the increased expression of cell-adhesion–related proteins in the quiescent fosBd/d cells (Figure 5B), would contribute to the increased cell adhesion in the absence of Akt signaling.
Among the possible downstream targets of FosB or FosB/2FosB, we focused on genes encoding proteins comprising the basement membrane and the TGF-β1 signaling pathway. The fosBd/d cells expressing only FosB/2FosB exhibited a strongly increased expression of genes involved in the TGF-β1 signaling pathway including Tgfb1i1 as well as genes encoding the basement membrane proteins, under the various culture conditions. In exponentially growing fosBFN/FN cells, which exclusively express FosB, we found that the expression level of the Tgfb1i1 gene, but not those of genes encoding the basement membrane proteins, was the lowest among the wild-type, fosB+/d, fosBdN/dN, fosBd/d, and fosB-null cell lines. Furthermore, we confirmed that fosBFN/FN cells exhibited the lowest cell-matrix adhesion among all cell lines under exponential growth conditions (Figure 7D). We thus conclude that FosB and/or 2FosB positively regulate cell-matrix adhesion as well as TGF-β1 signaling, whereas FosB negatively regulates them. Recently, it has been reported that 2FosB and to a lesser extent FosB modulate the expression and phosphorylation of Smads independent of intrinsic AP-1 activity (Sabatakos et al., 2008), confirming that both FosB and 2FosB play important roles in the up-regulation of TGF-β1 signaling, as we observed in the present study.
TGF-β1 is known to stimulate the DNA-binding activity of AP-1 complex containing FosB or FosB (Lai and Cheng, 2002). Furthermore, TGF-β1 stimulates gene expression through the cooperation of Smad3/Smad4 and AP-1 as a complex (Liberati et al., 1999; Liang et al., 2002). In fosBd/d ES cells, a gene for the Smad4-interacting transcription cofactor, Cited1 (Plisov et al., 2005), was also up-regulated (Table 3), suggesting that FosB is involved in various steps of the TGF-β1 signaling pathway. Moreover, FosB is likely to regulate gene expression through modulating mRNA stability as well as by modulating transcription itself (Nakabeppu and Nathans, 1991; Nakabeppu et al., 1993; Oda et al., 1995; Miura et al., 2005). Thbs1 encoding Thrombospondin 1 is known to activate latent TGF-β1 (Ren et al., 2006). It has been reported that transcription of the Thbs1 gene is suppressed by overexpression of c-Jun (Dejong et al., 1999) and that FosB facilitates c-Jun function, whereas FosB suppresses it by forming a heterodimer with c-Jun (Nakabeppu and Nathans, 1991). It is likely that FosB in wild-type ES cells down-regulates Thbs1 expression together with c-Jun, thereby negatively regulating cell-matrix adhesion. In contrast, FosB or 2FosB represses the c-Jun function and thus may abrogate negative regulation of Thbs1 expression by c-Jun resulting in the promotion of cell-matrix adhesion (Figure 8).
|
). As upstream regulators of TGF-β1 signaling, KLF6, which transactivates genes for TGF-β1 signaling components (Botella et al., 2002), and ELF3, which stimulates type II TGF-β receptor promoter (Kopp et al., 2004), were up-regulated in fosBd/d ES cells (Table 3). Several downstream targets of TGF-β1 were also up-regulated in fosBd/d cells (Table 3), such as genes for RUNX1 (Runx1), a transcription factor known to induce a gene for a tissue inhibitor of metalloproteinase 1 (Bertrand-Philippe et al., 2004), cardiac ankyrin repeat protein (Ankrd1) related to the initiation and regulation of arteriogenesis (Boengler et al., 2003), and E-cadherin (cdh1), whose induction by TGF-β1 is mediated through the activation of focal adhesion kinase due to extracellular matrix remodeling and increased cell-matrix interactions (Wang et al., 2004). It is suggested that both the expression and function of FosB can be up-regulated downstream of the TGF-β1 signaling pathway and that these two features are in turn up-regulated as a result of antagonistic regulation by FosB of Jun/Fos functioning as active AP-1, thereby generating a positive feedback circuit (Figure 8). Increased expression of FosB in fosBd/d or fosB+/d ES cells but not in fosBdN/dN ES cells was observed even in quiescence, and this was most likely the result of such positive feedback. In fosBFN/FN cells, lack of the antagonistic action of FosB on the active AP-1 complexes may further increase the negative regulation by FosB of the TGF-β1 signaling pathway. However, the complete lack of the fosB gene in fosB-null ES cells disallows these positive and negative feedbacks, resulting in the limited activation of Thrombospondin 1/TGF-β1 signaling in these cells (Figures 5D and 8).
The Tgfb1i1 gene, also known as Hic5/ARA55, encodes a LIM-only member of the paxillin superfamily that serves as a component of focal adhesions as well as a steroid receptor coactivator (Yang et al., 2000; Nishiya et al., 2001), indicating that the stronger expression of Tgfb1i1 may account in part for the higher level of cell-matrix adhesion. It has been shown that expression of the Tgfb1i1 gene is regulated by RAR as well as by TGF-β1, both of which can modulate AP-1 function (Liberati et al., 1999; Zhou et al., 1999; Suzukawa and Colburn, 2002; Zhuang et al., 2003), suggesting that expression of Tgfb1i1 is indirectly regulated by FosB or FosB/2FosB through their interaction with steroid receptors or Smad proteins, as reported recently (Sabatakos et al., 2008). Furthermore, it has been shown that Hic5/ARA55 itself negatively regulates Smad3 signaling (Wang et al., 2005). At the present time, we cannot explain why the expression level of the Tgfb1i1 gene in fosB-null cells is as high as in fosB+/d or fosBd/d cells that express increased levels of FosB, although it does appear that FosB positively regulates the expression of the Tgfb1i1 gene in a dose-dependent manner (Figure 7C). Availability of various cell lines that express different levels of either FosB or FosB/2FosB would help to shed light on the complex regulation of gene expression and cell-matrix adhesion. Furthermore, the establishment of cell lines that exclusively express FosB or 2FosB with or without FosB is also greatly important.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
Address correspondence to: Yusaku Nakabeppu (yusaku{at}bioreg.kyushu-u.ac.jp)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bertrand-Philippe, M., Ruddell, R. G., Arthur, M. J., Thomas, J., Mungalsingh, N., and Mann, D. A. (2004). Regulation of tissue inhibitor of metalloproteinase 1 gene transcription by RUNX1 and RUNX2. J. Biol. Chem 279, 24530–24539.
Boengler, K., Pipp, F., Fernandez, B., Ziegelhoeffer, T., Schaper, W., and Deindl, E. (2003). Arteriogenesis is associated with an induction of the cardiac ankyrin repeat protein (carp). Cardiovasc. Res 59, 573–581.
Botella, L. M. et al. (2002). Transcriptional activation of endoglin and transforming growth factor-β signaling components by cooperative interaction between Sp1 and KLF6, their potential role in the response to vascular injury. Blood 100, 4001–4010.
Chen, J., Kelz, M. B., Hope, B. T., Nakabeppu, Y., and Nestler, E. J. (1997). Chronic Fos-related antigens: stable variants of FosB induced in brain by chronic treatments. J. Neurosci 17, 4933–4941.
Deacu, E. et al. (2004). Activin type II receptor restoration in ACVR2-deficient colon cancer cells induces transforming growth factor-β response pathway genes. Cancer Res 64, 7690–7696.
Dejong, V., Degeorges, A., Filleur, S., Ait-Si-Ali, S., Mettouchi, A., Bornstein, P., Binetruy, B., and Cabon, F. (1999). The Wilms' tumor gene product represses the transcription of thrombospondin 1 in response to overexpression of c-Jun. Oncogene 18, 3143–3151.[CrossRef][Medline]
Ekblom, P., Lonai, P., and Talts, J. F. (2003). Expression and biological role of laminin-1. Matrix Biol 22, 35–47.[CrossRef][Medline]
Ide, Y., Tsuchimoto, D., Tominaga, Y., Iwamoto, Y., and Nakabeppu, Y. (2003). Characterization of the genomic structure and expression of the mouse Apex2 gene. Genomics 81, 47–57.[CrossRef][Medline]
Kopp, J. L., Wilder, P. J., Desler, M., Kim, J. H., Hou, J., Nowling, T., and Rizzino, A. (2004). Unique and selective effects of five Ets family members, Elf3, Ets1, Ets2, PEA3, and PU1, on the promoter of the type II transforming growth factor-β receptor gene. J. Biol. Chem 279, 19407–19420.
Kurushima, H. et al. (2005). Selective induction of FosB in the brain after transient forebrain ischemia accompanied by an increased expression of galectin-1, and the implication of FosB and galectin-1 in neuroprotection and neurogenesis. Cell Death Differ 12, 1078–1096.[CrossRef][Medline]
Lai, C. F., and Cheng, S. L. (2002). Signal transductions induced by bone morphogenetic protein-2 and transforming growth factor-beta in normal human osteoblastic cells. J. Biol. Chem 277, 15514–15522.
Lawler, J. (2002). Thrombospondin-1 as an endogenous inhibitor of angiogenesis and tumor growth. J. Cell. Mol. Med 6, 1–12.[Medline]
Lazo, P. S., Dorfman, K., Noguchi, T., Mattei, M. G., and Bravo, R. (1992). Structure and mapping of the fosB gene. FosB downregulates the activity of the fosB promoter. Nucleic Acids Res 20, 343–350.
Li, L., Arman, E., Ekblom, P., Edgar, D., Murray, P., and Lonai, P. (2004). Distinct GATA6- and laminin-dependent mechanisms regulate endodermal and ectodermal embryonic stem cell fates. Development 131, 5277–5286.
Li, X., Talts, U., Talts, J. F., Arman, E., Ekblom, P., and Lonai, P. (2001). Akt/PKB regulates laminin and collagen IV isotypes of the basement membrane. Proc. Natl. Acad. Sci. USA 98, 14416–14421.
Liang, C. L., Chen, J. L., Hsu, Y. P., Ou, J. T., and Chang, Y. S. (2002). Epstein-Barr virus BZLF1 gene is activated by transforming growth factor-β through cooperativity of Smads and c-Jun/c-Fos proteins. J. Biol. Chem 277, 23345–23357.
Liberati, N. T., Datto, M. B., Frederick, J. P., Shen, X., Wong, C., Rougier-Chapman, E. M., and Wang, X. F. (1999). Smads bind directly to the Jun family of AP-1 transcription factors. Proc. Natl. Acad. Sci. USA 96, 4844–4849.
Lim, I. J., Phan, T. T., Tan, E. K., Nguyen, T. T., Tran, E., Longaker, M. T., Song, C., Lee, S. T., and Huynh, H. T. (2003). Synchronous activation of ERK and phosphatidylinositol 3-kinase pathways is required for collagen and extracellular matrix production in keloids. J. Biol. Chem 278, 40851–40858.
McClung, C. A., Ulery, P. G., Perrotti, L. I., Zachariou, V., Berton, O., and Nestler, E. J. (2004). DFosB: a molecular switch for long-term adaptation in the brain. Brain Res. Mol. Brain Res 132, 146–154.[Medline]
Miura, T., Ohnishi, Y., Kurushima, H., Horie, H., Kadoya, T., and Nakabeppu, Y. (2005). Regulation of the neuronal fate by FosB and its downstream target, galectin-1. Curr. Drug Targets 6, 437–444.[CrossRef][Medline]
Nakabeppu, Y., and Nathans, D. (1991). A naturally occurring truncated form of FosB that inhibits Fos/Jun transcriptional activity. Cell 64, 751–759.[CrossRef][Medline]
Nakabeppu, Y., Oda, S., and Sekiguchi, M. (1993). Proliferative activation of quiescent Rat-1A cells by FosB. Mol. Cell. Biol 13, 4157–4166.
Nakabeppu, Y., Ryder, K., and Nathans, D. (1988). DNA binding activities of three murine Jun proteins: stimulation by Fos. Cell 55, 907–915.[CrossRef][Medline]
Nishioka, T., Sakumi, K., Miura, T., Tahara, K., Horie, H., Kadoya, T., and Nakabeppu, Y. (2002). fosB gene products trigger cell proliferation and morphological alteration with an increased expression of a novel processed form of galectin-1 in the rat 3Y1 embryo cell line. J. Biochem 131, 653–661.
Nishiya, N., Tachibana, K., Shibanuma, M., Mashimo, J. I., and Nose, K. (2001). Hic-5-reduced cell spreading on fibronectin: competitive effects between paxillin and Hic-5 through interaction with focal adhesion kinase. Mol. Cell. Biol 21, 5332–5345.
Oda, S., Nishida, J., Nakabeppu, Y., and Sekiguchi, M. (1995). Stabilization of cyclin E and cdk2 mRNAs at G1/S transition in Rat-1A cells emerging from the G0 state. Oncogene 10, 1343–1351.[Medline]
Plisov, S., Tsang, M., Shi, G., Boyle, S., Yoshino, K., Dunwoodie, S. L., Dawid, I. B., Shioda, T., Perantoni, A. O., and de Caestecker, M. P. (2005). Cited1 is a bifunctional transcriptional cofactor that regulates early nephronic patterning. J. Am. Soc. Nephrol 16, 1632–1644.
Rancourt, D. E., Tsuzuki, T., and Capecchi, M. R. (1995). Genetic interaction between hoxb-5 and hoxb-6 is revealed by nonallelic noncomplementation. Genes Dev 9, 108–122.
Ren, B., Yee, K. O., Lawler, J., and Khosravi-Far, R. (2006). Regulation of tumor angiogenesis by thrombospondin-1. Biochim. Biophys. Acta 1765, 178–188.[Medline]
Robertson, E. J. (1987). Tetracarcinomas and Embryonic Stem Cells: A Practical Approach, New York: IRL Press.
Sabatakos, G. et al. (2000). Overexpression of FosB transcription factor(s) increases bone formation and inhibits adipogenesis. Nat. Med 6, 985–990.[CrossRef][Medline]
Sabatakos, G., Rowe, G. C., Kveiborg, M., Wu, M., Neff, L., Chiusaroli, R., Philbrick, W. M., and Baron, R. (2008). Doubly truncated FosB isoform (2FosB) induces osteosclerosis in transgenic mice and modulates expression and phosphorylation of Smads in osteoblasts independent of intrinsic AP-1 activity. J. Bone Miner. Res 23, 584–595.[CrossRef][Medline]
Sarkar, G., and Sommer, S. S. (1990). The "megaprimer" method of site-directed mutagenesis. Biotechniques 8, 404–407.[Medline]
Shaulian, E., and Karin, M. (2001). AP-1 in cell proliferation and survival. Oncogene 20, 2390–2400.[CrossRef][Medline]
Suzukawa, K., and Colburn, N. H. (2002). AP-1 transrepressing retinoic acid does not deplete coactivators or AP-1 monomers but may target specific Jun or Fos containing dimers. Oncogene 21, 2181–2190.[CrossRef][Medline]
Tahara, K., Tsuchimoto, D., Tominaga, Y., Asoh, S., Ohta, S., Kitagawa, M., Horie, H., Kadoya, T., and Nakabeppu, Y. (2003). FosB but not FosB induces delayed apoptosis independent of cell proliferation in the Rat1a embryo cell line. Cell Death Differ 10, 496–507.[CrossRef][Medline]
Tsuchimoto, D., Sakai, Y., Sakumi, K., Nishioka, K., Sasaki, M., Fujiwara, T., and Nakabeppu, Y. (2001). Human APE2 protein is mostly localized in the nuclei and to some extent in the mitochondria, while nuclear APE2 is partly associated with proliferating cell nuclear antigen. Nucleic Acids Res 29, 2349–2360.
Wang, H., Radjendirane, V., Wary, K. K., and Chakrabarty, S. (2004). Transforming growth factor β regulates cell-cell adhesion through extracellular matrix remodeling and activation of focal adhesion kinase in human colon carcinoma Moser cells. Oncogene 23, 5558–5561.[CrossRef][Medline]
Wang, H., Song, K., Sponseller, T. L., and Danielpour, D. (2005). Novel function of androgen receptor-associated protein 55/Hic-5 as a negative regulator of Smad3 signaling. J. Biol. Chem 280, 5154–5162.
Xu, P., Yoshioka, K., Yoshimura, D., Tominaga, Y., Nishioka, T., Ito, M., and Nakabeppu, Y. (2003). In vitro development of mouse embryonic stem cells lacking JSAP1 scaffold protein revealed its requirement during early embryonic neurogenesis. J. Biol. Chem 278, 48422–48433.
Yang, L., Guerrero, J., Hong, H., DeFranco, D. B., and Stallcup, M. R. (2000). Interaction of the tau2 transcriptional activation domain of glucocorticoid receptor with a novel steroid receptor coactivator, Hic-5, which localizes to both focal adhesions and the nuclear matrix. Mol. Biol. Cell 11, 2007–2018.
Zhou, X. F., Shen, X. Q., and Shemshedini, L. (1999). Ligand-activated retinoic acid receptor inhibits AP-1 transactivation by disrupting c-Jun/c-Fos dimerization. Mol. Endocrinol 13, 276–285.
Zhuang, Y., Faria, T. N., Chambon, P., and Gudas, L. J. (2003). Identification and characterization of retinoic acid receptor β2 target genes in F9 teratocarcinoma cells. Mol. Cancer Res 1, 619–630.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
