V-Jun Targets Showing an Expression Pattern That Correlates with the Transformed Cellular Phenotype

Oncogene (2004) 23, 5703–5706 & 2004 Nature Publishing Group All rights reserved 0950-9232/04 $30.00 www.nature.com/onc v-Jun targets showing an expression pattern that correlates with the transformed cellular phenotype

Jason S Iacovoni1, Steven B Cohen2, Thorsten Berg3 and Peter K Vogt*,1

1Department of Molecular and Experimental Medicine, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA; 2Department of Protein Sciences, The Genomics Institute of the Novartis Research Foundation, 10675 John J Hopkins Drive, La Jolla, CA 92121, USA; 3Department of Molecular Biology, Max-Planck-Institute of Biochemistry, Am Klopferspitz 18A, D-82152 Martinsreid, Germany

Targets of the oncogenic transcription factor v-Jun in the by functional tests. Since the number of differentially murine cell line C3H 10T1/2 cells have been identified expressed genes in v-Jun-transformed cells is large, it is using DNA microarrays. Two targets, Akap12 and necessary to develop methods that filter relevant targets Marcks, are downregulated in transformed cells and are from innocuous bystanders. DNA microarray analyses known tumor suppressor genes. Overexpression of either were performed to find genes with altered expression Akap12 or Marcks in v-Jun-transformed cells reverses the levels in v-Jun transformed 10T1/2 cells (Cohen et al., transformed phenotype and leads to the re-expression of 2001). Each EST and cDNA sequence on the array was the other tumor suppressor gene, suggesting that these two used to extract gene identity information from the genes cooperate in the establishment of the nontrans- UniGene database using GeneHuggers (Cohen et al., formed state. Reverted cells continue to express v-Jun at 2001; Wheeler et al., 2002; Iacovoni, 2003). Genes high levels and also re-express c-Jun, which is normally induced or repressed by at least 2.5-fold are shown in repressed by v-Jun. A panel of six cell lines has been Tables 1 and 2, respectively. generated to evaluate the expression levels of other v-Jun We were particularly interested in downregulated targets in 10T1/2 cells. With these cells, we find that the genes, because their relevance for oncogenic transforma- upregulated target Sprr1a has an expression pattern that tion can be readily tested by re-expressing them in correlates with the transformed phenotype. transformed cells. If downregulation of the gene in Oncogene (2004) 23, 5703–5706. doi:10.1038/sj.onc.1207737 question is essential for oncogenic transformation, such Published online 10 May 2004 re-expression should cause at least a partial reversion of the transformed phenotype. In previous work, we have Keywords: Jun; AP-1; target genes; coordinate expres- shown that the protein Akap12 (A-kinase anchoring sion; reversion; oncogenic transformation; DNA protein), also known as SSeCKs (Src suppressed protein microarray kinase C substrate) or Gravin, is downregulated at the transcriptional level in Jun-transformed 10T1/2 cells and that re-expression of this target induces reversion of transformation (Cohen et al., 2001). Akap12 is repressed Introduction in NIH3T3 cells transformed by the Src, Ras, Fos, or Myc oncoproteins and is capable of reverting the Src- The retroviral oncogene v-jun of avian sarcoma virus 17 induced phenotype (Frankfort and Gelman, 1995; Lin is derived from the cellular jun (c-jun) gene which codes et al., 1995, 1996; Gelman, 2002). In the present for a bZip protein component of the AP-1 transcription communication, we extend our observations on Akap12 factor complex (Bohmann et al., 1987; Maki et al., 1987; and report on a second protein downregulated at the Angel et al., 1988; Lamph et al., 1988; Nishimura and transcriptional level in Jun-transformed 10T1/2 cells, Vogt, 1988). Expression of v-jun in chicken embryo Marcks (myristoylated alanine-rich C kinase substrate). fibroblasts (CEF) and in the murine fibroblast line C3H Like Akap12, Marcks is capable of reverting the Jun- 10T1/2 cells (10T1/2 cells) leads to oncogenic transfor- transformed phenotype, as determined by agar colony mation (Cavalieri et al., 1985; Maki et al., 1987; Cohen and growth rate assays (Cohen, 2001; data not shown). et al., 2001). Identification of target genes with altered Marcks is also downregulated in Src- and Ras-trans- expression levels in cells transformed by the v-Jun formed murine fibroblasts and is relocated in trans- protein is a prerequisite to understanding the mechan- formed cells from the cell periphery to the cytosol (Reed ism of transformation. Differentially expressed genes et al., 1991; Joseph et al., 1992). We have used the v-Jun need to be validated as transformation-relevant targets expressing 10T1/2 cells reverted by Akap12 or Marcks as indicators for the behavior of other target genes, *Correspondence: PK Vogt; E-mail: [email protected] arguing that correlation of expression levels with cellular Received 12 December 2003; revised 16 March 2004; accepted 16 March phenotype suggests a role in the determination of that 2004; published online 10 May 2004 phenotype. Genes that are upregulated in transformed v-Jun targets relevant for transformation JS Iacovoni et al 5704 Table 1 Genes upregulated in v-Jun transformed 10T1/2 cellsa Table 2 Genes downregulated in v-Jun-transformed 10T1/2 cells Fold Gene Description Fold Gene Description

13.7 Thbs2 Thrombospondin 2 À13.7 Serpine2 Serine/cysteine proteinase inhibitor, clade 10.5Sprr1a Small proline-rich protein 1A E, 2 4.9 AL024221 Expressed sequence AL024221 À5.6 Ptx3 Pentaxin related gene 4.7 Lgals3 Lectin, galactose binding, soluble 3 À5.1 Bgn Biglycan 3.9 Slc20a1 Solute carrier family 20, member 1 À5Meg3 Maternally expressed gene 3 3.7 4933403C17Rik RIKEN cDNA 4933403C17 gene À4.7 Ptn Pleiotrophin 3.6 Ndr4 N-myc downstream regulated 4 À4.1 Marcks Myristoylated alanine-rich PKC substrate 3.5 AA589632 Expressed sequence AA589632 À4 Spp1 Secreted phosphoprotein 1 3.3 Evi2 Ecotropic viral integration site 2 À3.9 Akap12 A kinase (PRKA) anchor protein (gravin) 3.1 Aprt Adenine phosphoribosyl transferase 12 3.1 Slc30a4 Solute carrier family 30, member 4 À3.6 Slc29a1 Solute carrier family 29, member 1 3.1 Pcolce2 Procollagen C-endopeptidase enhancer 2 À3.6 4833423D12Rik RIKEN cDNA 4833423D12 gene 2.8 AL024221 Expressed sequence AL024221 À3.4 Osf2-pending Osteoblast-speciﬁc factor 2 (fasciclin I- 2.7 Klf5Kruppel-like factor 5 like) 2.7 2410018A17Rik RIKEN cDNA 2410018A17 gene À3.3 2610209L21Rik RIKEN cDNA 2610209L21 gene 2.6 Cd9 CD9 antigen À3.2 6330406I15Rik RIKEN cDNA 6330406I15 gene 2.6 AW489850 Expressed sequence AW489850 À3.1 4921531G14Rik RIKEN cDNA 4921531G14 gene 2.6 Pdk2 Pyruvate dehydrogenase kinase, À3.1 Sepp1 Selenoprotein P, plasma, 1 isoenzyme 2 À3.1 C1s Complement component 1, s 2.6 1110005A05Rik RIKEN cDNA 1110005A05 gene subcomponent 2.5Suv39h1 Suppressor of variegation 3-9 homolog 1 À3 Cxcl12 Chemokine (C-X-C motif) ligand 12 2.5AW260363 Expressed sequence AW260363 À2.9 4921531G14Rik RIKEN cDNA 4921531G14 gene 2.5Psma7 Proteasome subunit, alpha type 7 À2.9 Lgals9 Lectin, galactose binding, soluble 9 2.5Glipr2 GLI pathogenesis-related 2 À2.9 4833423D12Rik RIKEN cDNA 4833423D12 gene 2.52410022L05Rik RIKEN cDNA 2410022L05gene À2.9 1110028E10Rik RIKEN cDNA 1110028E10 gene 2.5Basp1 Brain abundant, membrane attached À2.9 Stat3 Signal transducer and activator of signal protein 1 transcription 3 2.52410003B16Rik RIKEN cDNA 2410003B16 gene À2.7 Anxa6 Annexin A6 2.5 AI845668 Expressed sequence AI845668 À2.7 Rian RNA imprinted and accumulated in 2.52610200G18Rik RIKEN cDNA 2610200G18 gene nucleus 2.5Pi4k2b-pending Phosphatidylinositol 4-kinase type 2 beta À2.7 Tox Thymocyte selection-associated HMG box gene amRNA from v-Jun-transformed 10T1/2 murine cells and from their À2.6 Cbx5Chromobox homolog 5(Drosophila normal 10T1/2 progenitors was used to identify differentially regulated HP1a) genes in a commercial microarray (Incyte Pharmaceuticals). Known À2.6 Bicc1 Bicaudal C homolog 1 (Drosophila) internal controls were run with the experimental samples and were À2.6 9130005N14Rik RIKEN cDNA 9130005N14 gene used to normalize the results. Changes of 1.7-fold and above were À2.5Col6a1 Procollagen, type VI, alpha 1 found to be statistically signiﬁcant. Tables 1 and 2 use a 2.5-fold À2.5A230020K05Rik RIKEN cDNA A230020K05gene change as the cut off

cells but expressed at control levels in revertants and, vice versa, genes downregulated in transformed cells and re-expressed in revertants are likely to be related to the transformation process. In contrast, genes differentially expressed in Jun-transformed cells but failing to adjust their expression in revertants are probably unrelated to the transformation process. We have generated six stably transfected sublines of 10T1/2 cells expressing v-Jun, v-Jun and Akap12, v-Jun and Marcks, empty vector, Akap12 or Marcks res- Figure 1 Morphology of the panel of cell lines expressing various pectively (Figure 1). v-Jun expressing cells show combinations of v-Jun, Akap12 and Marcks. 10T1/2 cells (Reznik- transformed morphology and are capable of ancho- off et al., 1973) were infected with either empty pBabe viruses or pBabe containing v-Jun, Akap12, or Marcks, as indicated. Cell rage-independent growth. When Akap12 or Marcks is culture conditions, vectors and cloning of v-Jun and Akap12 were re-expressed in these cells, they acquire a ﬂattened shape described previously (Cohen et al., 2001). Marcks was obtained and expanded adherence to the substrate, similar to from PJ Blackshear. The BglII/BamHI fragment of Marcks was control 10T1/2 cells transfected with empty vector alone. cloned into the BamHI site of the p-Babe-hygro vector These Akap12- or Marcks-induced revertants of v-Jun- transformed cells have also lost the ability to form colonies in nutrient agar and show reduced growth rates examined the Akap12- and Marcks-induced revertants as compared to their transformed progenitors (Cohen, for re-expression of c-Jun (Figure 2). Reverse transcrip- 2001; data not shown). A hallmark of v-Jun-trans- tase (RT)–PCR was performed with primers that bind to formed cells is the drastic downregulation of c-Jun (Gao 50 and 30 ﬂanking sequences of the Jun delta region and et al., 1996; Kilbey et al., 1996; Hussain et al., 1998). We amplify both mouse c-Jun and the chicken-derived v-Jun

Oncogene v-Jun targets relevant for transformation JS Iacovoni et al 5705 Akap12 is toxic to 10T1/2 cells. We found that some selected clones became growth-arrested and that by passaging Akap12 expressing cells, we may have selected lines that produce low levels of exogenous Akap12. The expression of Marcks and Akap12 is therefore coupled, and it is correlated with the cellular phenotype, down in transformed cells, and restored in revertants. The small proline-rich protein 1A (Sprr1a) represents an additional Jun target with an expression pattern that correlates with transformation (Figure 3). DNA micro- Figure 2 Re-expression of downregulated targets in cells reverted arrays show this gene to be induced 10.5-fold in Jun- by Akap12 and Marcks. (a) c-Jun and v-Jun expression were transformed cells. It encodes a small proline-rich monitored simultaneously by reverse transcriptase (RT)–PCR. protein, originally identified as having a cross-bridging Primers were designed to amplify fragments of mouse and chicken function in the formation of cornified cell envelopes of jun. Due to the deletion of the delta region in v-Jun, this PCR amplification generates different sized products for c-Jun and v- stratified squamous epithelia (Kartasova et al., 1996). Jun. In total, 20 cycles of RT–PCR were carried out with 1 mg total Recently, this gene has been shown to be TPA-inducible RNA as described by the manufacturer (GibcoBRL, SuperScript and to be expressed at high basal levels in advanced One-Step RT–PCR with Platinum Taq). Products were resolved on stages of skin cancer (Schlingemann et al., 2003). The a 10% polyacrylamide gel and stained with ethidium bromide. (b) Marcks and (c) Akap12 expression were determined by Western fact that v-Jun-induced transcriptional upregulation of blotting of 15 mg total protein from whole-cell lysates as previously Sprr1a is lost in reverted cells suggests that this gene described (Cohen et al., 2001). Marcks was detected with a may play a role in transformation, but additional commercial antibody sc-6455 (Santa Cruz) and Akap12 with gravin antiserum obtained from MH Ginsberg (Grove et al., 1994). Abbreviations used to designate cell lines are as follows: J cells are infected with pBabe(puro)-v-Jun and pBabe(hygro), JM cells with pBabe(puro)-v-Jun and pBabe(hygro)-Marcks, JA cells with pBabe(puro)-v-Jun and pBabe(hygro)-Akap12, M cells pBabe(- puro) and pBabe(hygro)-Marcks, A cells pBabe(puro) and pBabe(hygro)-Akap12, and C cells pBabe(puro) and pBabe(hygro) used for transformation (Figure 2). Akap12- or Marcks- induced reversion led to re-expression of c-Jun albeit at levels that were below those of control 10T1/2 cells, whereas in v-Jun-transformed 10T1/2 cells, c-Jun mRNA levels were undetectable. Reversion by Akap12 or Marcks did not qualitatively affect expression levels of v-Jun. The extinction of c-Jun transcription in v-Jun- transformed cells is mediated by binding of v-Jun/Fra heterodimers to the TRE sequence located at position À72 in the c-Jun promoter (Hussain et al., 1998). The re- expression of c-Jun in revertants that also express undiminished levels of v-Jun therefore raises interesting questions regarding c-Jun transcriptional regulation. In Akap12- or Marcks-expressing cells, transformation- related transcriptional activities of v-Jun are attenuated, possibly by effects on v-Jun dimerization with other b- Zip proteins, on target promoter binding, or on Jun Figure 3 Expression profiles of four selected v-Jun targets interactions with co-activators and co-repressors. The determined by Northern blotting. Total RNA preparation and data in Figure 2 also show that expression of c-Jun is Northern blotting were performed as previously described (Fu correlated with the cellular phenotype – virtually absent et al., 1999). In total, 25 mg of total RNA were prepared using RNA Stat (Tel-Test (Friendswood, TX, USA)) from each of the six cell in transformed cells, partially restored in revertants and lines: J, JM, JA, M, A, and C. RNA samples were separated on 1% high in nontransformed 10T1/2 cells. We next deter- denaturing agarose gels and transferred on to nitrocellulose mined the protein levels of Akap12 and Marcks in the membranes (Amersham). Equal loading of each sample was six cell lines. Vector-mediated re-expression of Akap12 confirmed by ethidium bromide staining of ribosomal RNA. Individual primer pairs were designed to amplify 800–1000 base in v-Jun-transformed 10T1/2 cells resulted in coordinate pair regions of publicly available mouse mRNA sequences using partial re-expression of the cellular Marcks. Vice versa, Primer 2.0 software. PCR was performed on a first strand cDNA vector-mediated expression of Marcks in v-Jun cells preparation made from a mixture of v-Jun-transformed and induced coordinate expression of Akap12. A similar control 10T1/2 RNA. PCR products were gel-purified and 32P level of Akap12 expression is seen in each cell line. We labeled with a random primed labeling kit (Roche). Hybridization, washing and detection were performed as described previously (Fu attribute the lack of Akap12 in cells infected with the et al., 1999). Images were obtained with ImageQuant software and exogenous construct to the fact that overexpression of the Storm Phosphorimager (Molecular Dynamics)

Oncogene v-Jun targets relevant for transformation JS Iacovoni et al 5706 functional studies will be required to confirm this the cellular phenotype. Examples are Osf2 and Serpine2. proposal. (3) Targets that follow neither cellular phenotype nor The galactose-binding, soluble lectin 3 (Lgals3) v-Jun in their expression levels. One such example is (Figure 3), was found by DNA microarrays to be Lgals3. Category (1) is likely to contain targets that are upregulated in Jun-transformed cells. The upregulation relevant for transformation, because reversion of the was confirmed by Northern blotting, yet this upregula- transformed phenotype also cancels the differential tion was not changed by Marcks- or Akap12-induced expression induced by the oncoprotein without changing reversion. Lgals3 expression was even enhanced in cells levels of the oncoprotein itself. However, the three singly expressing either tumor suppressor. categories of targets are correlative and can only suggest Two downregulated targets, serine protease inhibitor, relevance or irrelevance for transformation. Category clade E, member 2 (Serpine2), and osteoblast-specific (1) could contain irrelevant targets and category (2) factor 2 (Osf2) were also examined. Osf2 is also down- relevant ones, necessary but not sufficient for transfor- regulated by Jun in avian cells (Hartl et al., 2003). mation. The unabated expression of v-Jun in Akap12- Serpine2 was found in DNA microarrays to be down- and Marcks-reverted cells is of special significance. It regulated 13.7-fold. However, neither of these targets implies cancellation of the transforming activities of have an expression pattern that correlates with trans- v-Jun. In the case of category (1) targets, this cancella- formation; rather, they show a v-Jun-dependent down- tion occurs at the level of transcriptional regulation; in regulation. All three cell lines expressing v-Jun, the revertants v-Jun is prevented from regulating its regardless of revertant status, show a reduction in target promoters. These findings suggest that analyses of Osf2 and Serpine2 expression. In cells expressing all six cell lines with DNA microarrays should identify Akap12 singly, Osf2 levels are unaltered, but those of all genes that posses correlative transcriptional expres- Serpine 2 are reduced. Cells expressing Marcks alone sion patterns. In conclusion, our observations offer show reduction of Osf2 levels while those of Serpine 2 a new method for sifting through and identifying are unchanged. transformation-relevant v-Jun targets. The data also The downregulation of Akap12 and of Marcks is an raise interesting questions on the control of v-Jun in essential requirement for Jun-induced transformation. revertant cells. Akap12- or Marcks-induced revertants of v-Jun-transformed 10T1/2 cells, together with transformed and Acknowledgements control cells, can be used to classify Jun targets into This work was supported by National Institutes of Health three categories: (1) Targets with expression levels that grants CA78230, CA78045and CA79616. We thank Hershey are correlated with the cellular phenotype. These include Filoteo for competent technical assistance and Gail Yoder for Akap12, Marcks, c-Jun and Sprr1a. (2) Targets whose critical help with the manuscript. The manuscript number is expression is correlated with that of v-Jun but not with 15492-MEM of The Scripps Research Institute.

References

Angel P, Allegretto EA, Okino ST, Hattori K, Kartasova T, Darwiche N, Kohno Y, Koizumi H, Osada S, Boyle WJ, Hunter T and Karin M. (1988). Nature, 332, Huh N, Lichti U, Steinert PM and Kuroki T. (1996). 166–171. J. Invest. Dermatol., 106, 294–304. Bohmann D, Bos TJ, Admon A, Nishimura T, Vogt PK and Kilbey A, Black EJ, Unlu M and Gillespie DA. (1996). Tjian R. (1987). Science, 238, 1386–1392. Oncogene, 12, 2409–2418. Cavalieri F, Ruscio T, Tinoco R, Benedict S, Davis C and Lamph WW, Wamsley P, Sassone-Corsi P and Verma IM. Vogt PK. (1985). Virology, 143, 680–683. (1988). Nature, 334, 629–631. Cohen SB, Waha A, Gelman IH and Vogt PK. (2001). Lin X, Nelson PJ, Frankfort B, Tombler E, Johnson R and Oncogene, 20, 141–146. Gelman IH. (1995). Mol. Cell Biol., 15, 2754–2762. Frankfort BJ and Gelman IH. (1995). Biochem. Biophys. Res. Lin X, Tombler E, Nelson PJ, Ross M and Gelman IH. (1996). Commun., 206, 916–926. J. Biol. Chem., 271, 28430–28438. Fu S, Bottoli I, Goller M and Vogt PK. (1999). Proc. Natl. Maki Y, Bos TJ, Davis C, Starbuck M and Vogt PK. (1987). Acad. Sci. USA, 96, 5716–5721. Proc. Natl. Acad. Sci. USA, 84, 2848–2852. Gao M, Morgan I and Vogt PK. (1996). Cancer Res., 56, Nishimura T and Vogt PK. (1988). Oncogene, 3, 659–663. 4229–4235. Reed JC, Rapp U and Cuddy MP. (1991). Cell Signal., 3, Gelman IH. (2002). Front Biosci., 7, d1782–1797. 569–576. Grove BD, Bowditch R, Gordon T, del Zoppo G and Reznikoff CA, Brankow DW and Heidelberger C. (1973). Ginsberg MH. (1994). Anat. Rec., 239, 231–242. Cancer Res., 33, 3231–3238. Hartl M, Bader AG and Bister K. (2003). Curr. Cancer Drug Schlingemann J, Hess J, Wrobel G, Breitenbach U, Gebhardt Targets, 3, 41–55. C, Steinlein P, Kramer H, Furstenberger G, Hahn M, Angel Hussain S, Kilbey A and Gillespie DA. (1998). Cell Growth P and Lichter P. (2003). Int. J. Cancer, 104, 699–708. Differ., 9, 677–686. Wheeler DL, Church DM, Lash AE, Leipe DD, Madden TL, Iacovoni JS. (2003). Bioinformatics, 19, 2316–2318. Pontius JU, Schuler GD, Schriml LM, Tatusova TA, Joseph CK, Qureshi SA, Wallace DJ and Foster DA. (1992). Wagner L and Rapp BA. (2002). Nucleic Acids Res., 30, J. Biol. Chem., 267, 1327–1330. 13–16.

Oncogene