Inhibition of Myc-induced cell transformation by brain acid-soluble ...

Download PDF
30 downloads 0 Views 2MB Size Report
Comment
Apr 7, 2009 - Cell transformation by the Myc oncoprotein involves transcrip- tional activation or suppression of specific target genes with intrinsic oncogenic ...
Inhibition of Myc-induced cell transformation by brain acid-soluble protein 1 (BASP1) Markus Hartl1, Andrea Nist, M. Imran Khan, Taras Valovka, and Klaus Bister1 Institute of Biochemistry and Center for Molecular Biosciences, University of Innsbruck, A-6020 Innsbruck, Austria Edited by Peter K. Vogt, The Scripps Research Institute, La Jolla, CA, and approved February 20, 2009 (received for review November 27, 2008)

Cell transformation by the Myc oncoprotein involves transcriptional activation or suppression of specific target genes with intrinsic oncogenic or tumor-suppressive potential, respectively. We have identified the BASP1 (CAP-23, NAP-22) gene as a novel target suppressed by Myc. The acidic 25-kDa BASP1 protein was originally isolated as a cortical cytoskeleton-associated protein from rat and chicken brain, but has also been found in other tissues and subcellular locations. BASP1 mRNA and protein expression is specifically suppressed in fibroblasts transformed by the v-myc oncogene, but not in cells transformed by other oncogenic agents. The BASP1 gene encompasses 2 exons separated by a 58-kbp intron and a Myc-responsive regulatory region at the 5ⴕ boundary of untranslated exon 1. Bicistronic expression of BASP1 and v-myc from a retroviral vector blocks v-myc-induced cell transformation. Furthermore, ectopic expression of BASP1 renders fibroblasts resistant to subsequent cell transformation by v-myc, and exogenous delivery of the BASP1 gene into v-myc-transformed cells leads to significant attenuation of the transformed phenotype. The inhibition of v-myc-induced cell transformation by BASP1 also prevents the transcriptional activation or repression of known Myc target genes. Mutational analysis showed that the basic N-terminal domain containing a myristoylation site, a calmodulin binding domain, and a putative nuclear localization signal is essential for the inhibitory function of BASP1. Our results suggest that downregulation of the BASP1 gene is a necessary event in myc-induced oncogenesis and define the BASP1 protein as a potential tumor suppressor.

and metabolism, including protein synthesis, ribosomal biogenesis, glycolysis, mitochondrial function, and cell cycle progression (3, 11). Most of the genes consistently found to be repressed by Myc are involved in cell cycle arrest, cell adhesion, and cell-to-cell communication (3, 10, 11). In this article, we describe the identification of a Myc target that is specifically repressed in Myc-transformed cells and, conversely, has a strong potential to specifically inhibit cell transformation by Myc. The gene, termed BASP1 (or CAP-23, NAP-22), encodes a 25-kDa acidic protein that was originally isolated as a membrane and cytoskeleton-associated protein from rat and chicken brain (12, 13), but is also expressed in other tissues. The BASP1 protein is implicated in neurite outgrowth, maturation of the actin cytoskeleton, and organization of the plasma membrane (14), but its precise biochemical function is unknown. The protein binds to calmodulin, is a substrate of protein kinase C and N-myristoyltransferase, and shares several distinctive biochemical and biophysical properties with the cytosolic growth-associated proteins GAP-43 and MARCKS (13–15). Intriguingly, BASP1 was also found as a nuclear factor regulating the transcriptional activity of the Wilms’ tumor suppressor protein WT1 (16, 17). Our results show that (i) the BASP1 gene is consistently down-regulated in myc-transformed cells and (ii) the BASP1 protein is a strong inhibitor of myc-induced oncogenesis, defining this protein as a potential tumor suppressor.

cancer 兩 gene expression 兩 myc oncogene 兩 transcriptional control 兩 tumor suppressor

transformed by doxycycline-controlled v-myc alleles (18), displaying differential gene expression of putative Myc target genes like the strongly-activated WS5 gene that was fully characterized recently (19). One of the original partial cDNA clones (VCB2) isolated in these screens corresponded to a gene that was strongly suppressed in Myc-transformed quail embryo fibroblasts (QEFs) (18). We have now completely characterized this gene representing the avian homolog of the murine and human BASP1 genes encoding the neuronal brain acid-soluble protein 1 (13). Northern analysis showed that expression of the 2-kb BASP1 mRNA is abundant in normal chicken embryo fibroblasts (CEFs) or CEFs transformed by the NK24, ASV17, or Rous sarcoma virus (RSV) retroviruses carrying the v-fos, v-jun, or v-src oncogenes, but is specifically suppressed in CEFs transformed by the MC29 or MH2 retroviruses carrying the v-myc or v-myc/v-mil oncogenes (Fig. 1A). Specific suppression was also observed in the v-myc- and v-myc/v-miltransformed quail cell lines Q8 and MH2A10, but not in normal QEFs or QEFs transformed by v-jun (VJ) (Fig. 1 A). The specific

T

he myc oncogene was originally identified as a transduced allele (v-myc) in the genome of avian acute leukemia virus MC29 (1). The cellular c-myc protooncogene encodes the Myc protein, an intensively studied transcription factor with oncogenic potential (2, 3). Myc is the central part of a transcriptional regulator network controlling the expression of up to 15% of all human genes (2, 3), and several of the hitherto identified Myc target genes are critically involved in cellular growth, proliferation, apoptosis, metabolism, and differentiation (2, 3). Deregulation of the c-myc gene is a frequent event in animal and human tumorigenesis, and ⬇30% of all human cancers express Myc at elevated levels (4, 5). Accordingly, strategies have been proposed to use inhibition of Myc function as a potential cancer therapy (6, 7). Myc is a basic helix–loop–helix (bHLH)-Zip protein, forms heterodimers with the bHLH-Zip protein Max, binds to specific DNA sequence elements (E-boxes), and is part of a transcription factor network including additional proteins like Mad or Mnt (2). Myc–Max heterodimers are usually implicated in transcriptional activation of distinct target genes, but Myc has also been associated with transcriptional repression (2, 3). Known mechanisms of repression involve specific inhibition of transcriptional activators like C/EBP or Miz-1 (3, 8, 9). Although a large number of genes activated or repressed by Myc have been identified (3, 10, 11), the distinction between target genes that mediate the physiological functions of Myc and those that are directly relevant for tumorigenesis remains to be assessed. Many of the genes activated by Myc are related to processes of cell growth 5604 –5609 兩 PNAS 兩 April 7, 2009 兩 vol. 106 兩 no. 14

Results BASP1 Gene Expression Is Specifically Suppressed in Myc-Transformed Cells. We have previously established avian cell lines conditionally

Author contributions: M.H. and K.B. designed research; M.H., A.N., M.I.K., and T.V. performed research; M.H., T.V., and K.B. analyzed data; and M.H. and K.B. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. EU888907 and AF285876). 1To

whom correspondence may be addressed. E-mail: [email protected] or [email protected].

This article contains supporting information online at www.pnas.org/cgi/content/full/ 0812101106/DCSupplemental.

www.pnas.org兾cgi兾doi兾10.1073兾pnas.0812101106

down-regulation of BASP1 in Myc-transformed cells was also observed at the protein level. The BASP1 protein with an apparent Mr of 58,000 was detectable in normal QEF and VJ cells, but not in Q8 or MH2A10 cells (Fig. 1B). The specificity of BASP1 suppression in Myc-transformed cells is corroborated by comparison with the expression pattern of another gene (BRAK) isolated in our screens for putative Myc targets. BRAK represents the chicken homolog of the human CXCL14 chemokine gene (20), and its expression is suppressed at the mRNA and protein level in all transformed fibroblasts, irrespective of the transforming agent (Fig. 1 A and B). We have also analyzed the dynamics of transcriptional down-regulation of BASP1 by using the conditional cell transformation system described above. Conditional activation or deactivation of the v-myc oncogene correlates directly with deactivation or activation, respectively, of the BASP1 gene in a fully-reversible pattern, indicating a close molecular link between the mutuallyexclusive expression of these genes (Fig. 1C). Structure of the Chicken BASP1 Gene and Protein. Alignment of the

982-bp chicken BASP1 cDNA sequence and the 190-bp VCB2 sequence with the chicken genome database confirmed that VCB2 represents the 3⬘ end of the chicken BASP1 gene that encodes a Hartl et al.

Fig. 2. Topography of the chicken BASP1 gene and protein. (A) Structure of chicken BASP1. Exons (gray), the transcription start site (arrow), the coding region (black), and the regions corresponding to exon-specific cDNA probes or the original VCB2 clone (gray bars) are indicated. The gene segment used for the luciferase reporter construct pLUC-BASP1 is shown below. (B) Northern analysis of poly(A)⫹-selected RNA (2.5 ␮g) from CEFs, using 32P-labeled cDNA probes specifying exons 1 or 2 from chicken BASP1. (C) Mapping of the chicken BASP1 transcriptional start site by primer extension analysis. Five micrograms of poly(A)⫹-selected RNA from CEFs and an exon 1-specific 32P-labeled oligodeoxynucleotide were used as template and primer, respectively, for the reverse transcription. The same unlabeled oligodeoxynucleotide was used as primer in a sequencing reaction using a subcloned 787-bp SmaI fragment of BASP1 genomic DNA as a template. (D) Specific transcriptional repression of the BASP1 promoter. Aliquots (4.0 ␮g) of the pLUC-BASP1 reporter construct and the pSV-␤-galactosidase plasmid (2.0 ␮g) were transfected into QEFs and into QEFs transformed by v-myc (QEF/MC29) or v-jun (VJ). Luciferase activities expressed in relative light units (RLU)/␮g protein were determined for aliquots (5 ␮L) of cell extracts prepared 1 day after transfection. Cell extracts were also analyzed by immunoblot analysis using antibodies directed against v-Myc, v-Jun, or BASP1. (E) Amino acid sequence alignment of the chicken (c), human (h), and mouse (m) BASP1 proteins (GenBank accession nos. NP㛭989447, AAH00518, AND NP㛭081671). Residues in the mammalian BASP1 sequences identical to the corresponding residues in the chicken sequence are shown as dots, dashes indicate gaps. The calmodulin binding domain (asterisks), the myristoylation site (yellow), the potential nuclear localization signal (pink), and the acidic domain (blue) are indicated.

244-aa protein (Fig. 2A). The sequence of the first 153 bp of the chicken BASP1 cDNA was not present in the database, suggesting that this sequence belongs to a yet-unidentified exon of the BASP1 gene. This idea was corroborated by Northern analysis showing that the 153-bp segment recognized the same mRNA as a probe derived from the coding region (Fig. 2B). The 153-bp probe was also used to screen a genomic library leading to the isolation of a 793-bp SmaI fragment. Sequence analysis (GenBank accession no. EU888907), transcription start site mapping (Fig. 2C), and partial overlap with a contig from the chicken genome (GenBank accession no. PNAS 兩 April 7, 2009 兩 vol. 106 兩 no. 14 兩 5605

BIOCHEMISTRY

Fig. 1. Specific suppression of BASP1 gene expression in v-myc-transformed cells. (A) Northern analysis of poly(A)⫹-selected RNAs (2.5 ␮g) from normal CEFs or QEFs, CEFs transformed by the retroviruses ASV17 (v-jun), NK24 (v-fos), MC29 (v-myc), RSV (v-src), or MH2 (v-myc/v-mil), and the quail cell lines Q8, MH2A10, or VJ, transformed by v-myc, v-myc/v-mil, or v-jun. (B) Immunoblot analysis using aliquots (15 ␮g of total protein) of cell extracts derived from QEFs or the transformed quail cell lines shown in A. Proteins were detected by polyclonal antibodies directed against a carboxyl-terminal peptide of the chicken BASP1 protein or against a recombinant BRAK protein. (C) Kinetics of BASP1 suppression monitored by Northern analysis using total RNAs from the cell line Q/tMON conditionally transformed by a v-myc allele controlled by a doxycycline-dependent transactivator. Cells were first grown continuously in the presence (⫹) of doxycycline; at day ⫺6, the drug was removed, readded at day 0, and removed again at day 6. RNAs were isolated before removal or addition of the drug and at the intermediate time points indicated. Filters in A and C were hybridized with 32P-labeled cDNA probes specific for chicken BASP1 or BRAK, v-myc, or quail GAPDH. Sizes of the mRNAs were: BASP1, 2.0 kb; BRAK, 1.8 kb; MC29, 5.4 kb; MH2, 5.3/2.3 kb; c-myc, 2.3 kb; v-myc, 1.9 kb; GAPDH, 1.4 kb.

Fig. 3. Inhibition of Myc-induced cell transformation by concomitant expression of BASP1. (A) Plasmid DNA constructs pRCAS-MC29 and pRCAS-ASV17 specify replication-defective retroviruses encoding Gag-Myc and Gag-Jun hybrid proteins, respectively. The bicistronic derivatives pRCAS-MC29-IRES-BASP1, pRCAS-MC29IRES-v-Mil, pRCAS-MC29-IRES-BRAK, and pRCAS-ASV17-IRES-BASP1 were generated by inserting EcoRI/NotI (blunt-ended) DNA fragments containing an IRES and the coding regions of BASP1, v-mil, or BRAK, into the KpnI site (blunt-ended) of the original vectors. LTR, long terminal repeat; gag, group-specific antigen; env, envelope; bla, ␤-lactamase. (B) SDS/PAGE analysis of Myc, Mil, BRAK, Jun, and BASP1 proteins ectopically expressed after transfection of QEFs with the constructs shown in A together with the empty RCAS vector specifying replication-competent helper virus. For detection of Myc, Mil, BRAK, and Jun proteins, aliquots (2 ⫻ 107 cpm) of lysates from cells metabolically labeled with [35S]methionine were immunoprecipitated with the antisera indicated. Expression of BASP1 proteins was monitored by immunoblotting using a BASP1-specific peptide antiserum. (C) QEFs were transfected with 8-␮g aliquots of DNA from the retroviral constructs shown in A together with empty pRCAS DNA (0.8 ␮g), kept under agar overlay for 2 weeks, and then stained with eosin methylene blue. Numbers of foci scored on 100-mm dishes are indicated.

NW㛭001471639) revealed that the isolated fragment contains the entire untranslated exon 1 of chicken BASP1 that is separated from exon 2 by an intron of ⬇58 kbp. A 511-bp 5⬘ DNA segment extending into exon 1 was inserted into a luciferase reporter plasmid (Fig. 2 A). The construct was transfected into normal QEFs or the cell lines QEF/MC29 or VJ transformed by v-myc or v-jun, respectively (Fig. 2D). The analysis showed that the BASP1 promoter was specifically down-regulated in v-myc-transformed cells, in accordance with the BASP1 mRNA expression profile. Further dissection of the regulatory region revealed that a 135-bp segment from the 5⬘ end of exon 1 is sufficient to mediate transcriptional activation in normal cells and suppression in Myc-transformed cells. This segment contains 2 essential binding sites for transcription factor Sp1, and ChIP showed that Sp1 and Myc occupy this minimal regulatory region (Fig. S1). An alignment of the amino acid sequences of the chicken BASP1 protein and the human and mouse homologs (Fig. 2E) revealed extensive sequence similarities, particularly in the N-terminal domain encompassing a myristoylation site, a calmodulin binding region, and a putative nuclear localization signal (14, 15). Another conserved feature is a large acidic region, presumably causing the anomalous electrophoretic mobility of BASP1 proteins in SDS/ PAGE (12, 13). The 244-aa chicken BASP1 protein exhibits a theoretical isoelectric point of 4.66, a calculated Mr of 25,438, and an apparent Mr of 58,000 (Fig. 1B). Myc-Induced Cell Transformation Is Specifically Inhibited by BASP1.

The specific suppression of BASP1 gene expression in v-myctransformed cells prompted us to test whether constitutively expressed BASP1 interferes with Myc function. The plasmid constructs pRCAS-MC29 and pRCAS-ASV17 specify replicationdefective oncogenic retroviruses that encode Gag-Myc (p110gag-myc) or Gag-Jun (p83gag-jun) hybrid proteins, respectively (Fig. 3A). They 5606 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0812101106

were used to engineer bicistronic vectors by inserting an internal ribosome entry site (IRES) followed by the coding regions of BASP1, v-mil, or BRAK, allowing simultaneous expression of v-myc or v-jun and the second gene from the same retroviral genome. Protein expression in QEFs transfected with these constructs was analyzed by immunoprecipitation or immunoblotting, confirming synthesis of the exogenous proteins specified by the particular vector. The expression pattern of the endogenous p52c-myc, p42c-jun, p12BRAK, and p58BASP1 proteins is in agreement with the known negative autoregulation of c-Myc and c-Jun proteins by their viral counterparts (21) and with the suppression of BRAK and BASP1 in v-myc-transformed cells (compare Fig. 1), respectively. Specifically, endogenous p58BASP1 is expressed at high levels in normal QEFs or QEFs transformed by RCAS-ASV17, but suppressed in QEFs transformed by RCAS-MC29 or RCAS-MC29-IRES-vMil. QEFs transfected with pRCAS-MC29-IRES-BASP1 or pRCAS-MC29-IRES-BRAK contain exogenous BASP1 or BRAK proteins, equal in size to their endogenous counterparts (Fig. 3B). Transfection of QEFs with the pRCAS-MC29 construct led to efficient cell transformation monitored by focus formation (Fig. 3C). Coexpression of Myc and Mil oncoproteins from the pRCASMC29-IRES-v-Mil construct enhanced focus formation as reported (21), and coexpression of BRAK did not interfere with the transforming capacity of Myc. Intriguingly, coexpression of Myc and BASP1 from pRCAS-MC29-IRES-BASP1 did not induce focus formation (Fig. 3C). The reduced level of p110gag-myc in cells transfected with this construct (Fig. 3B) is presumably caused by their nontransformed state involving a lower rate of protein synthesis, because it has been observed for the expression of a nontransforming mutant Myc protein (8). The specificity of inhibition of Myc-induced cell transformation by BASP1 was strongly corroborated by the lack of inhibition of Jun by BASP1, as demonstrated by equal efficiency in cell transformation by pRCASASV17 and pRCAS-ASV17-IRES-BASP1 (Fig. 3C). Hartl et al.

transformation (Fig. 4B), although they retained their susceptibility to transformation by other oncogenes like src (Fig. S2). Overexpression of BASP1, BRAK, and v-Myc proteins from the RCAS or pRc vectors was confirmed by immunoblotting (Fig. 4B). In contrast to the experimental design in Figs. 1 and 3, v-Myc expression did not spread through the whole culture; hence endogenous BRAK and BASP1 proteins were detectable in the total cell extract. To test whether ectopic BASP1 expression can also interfere with established Myc-induced cell transformation, the fully transformed cell line QEF/Rc-Myc was transfected with the pRCAS, pRCASBASP1, or pRCAS-BRAK vectors, and the cultures were passaged several times. As expected, endogenous BASP1 and BRAK proteins were not detectable in these cells, whereas the ectopicallyexpressed proteins were efficiently overexpressed (Fig. 4C). The presence of exogenous BRAK did not interfere with the transformed phenotype of QEF/Rc-Myc cells, whereas ectopic BASP1 expression led to a more flattened morphology and a reduced capacity of these cells to grow in semisolid medium (Fig. 4C). Hence, BASP1 can also partially interfere with established transformation. To rule out that ectopic expression of BASP1 by itself effects morphology, proliferation, or viability of cells, BASP1 overexpressing cells were compared with normal or v-myc-transformed cells (Fig. S3 A–C). No significant differences between BASP1 overexpressing and normal cells were found. Both endogenous and ectopic BASP1 proteins were distributed in cytoplasmic and nuclear localizations (Fig. S3D), in agreement with previous reports (16, 17). Efficient knockdown of BASP1 expression by transient siRNA nucleofection in the chicken DF-1 cell line (Fig. S3E) did not lead to significant changes in cellular morphology or proliferation, nor to changes in c-myc expression.

Fig. 4. Suppression of Myc-induced cell transformation by independent ectopic expression of BASP1. (A) Plasmid DNA constructs pRCAS-BASP1 and pRCAS-BRAK were generated by the insertion of the BASP1 or BRAK coding sequences into the ClaI site of the retroviral RCAS vector and specify replication-competent retroviruses expressing BASP1 or BRAK proteins. The eukaryotic expression vector pRc-Myc was used to generate the v-myc-transformed cell line QEF/Rc-Myc. (B) QEFs were transfected with the RCAS constructs shown in A, passaged 5 times, and then supertransfected with pRc-Myc containing the v-myc oncogene or with the empty pRc vector. Cells were kept under agar overlay for 2 weeks and then stained with eosin methylene blue. Foci were counted on 100-mm dishes. Proteins were analyzed by immunoblotting using equal amounts (15 ␮g of protein) of cell extracts prepared 2 days after supertransfection and specific antisera directed against BASP1, BRAK, or Myc. (C) QEF/Rc-Myc cells were transfected with RCAS-BASP1, RCASBRAK, or the empty RCAS vector. After multiple passages, equal numbers of cells (2.5 ⫻ 104) were seeded in soft agar on 35-mm dishes and incubated for 2 weeks. Phase-contrast micrographs of cultured cells, bright-field micrographs of agar colonies, and a quantification of colonies per 1,000 cells seeded are shown. Proteins were analyzed by immunoblotting using equal amounts (15 ␮g of protein) of cell extracts and specific antisera directed against BASP1, BRAK, or Myc.

To confirm these observations, QEF were transfected with replication-competent RCAS constructs expressing BASP1 or BRAK (Fig. 4A), or with the empty vector, and passaged several times. The cells were then transfected with the nonviral Myc expression vector pRc-Myc (Fig. 4A), or with the empty pRc vector, and transformation was monitored by focus formation (Fig. 4B). Efficient focus formation was observed for cells preinfected with the RCAS or RCAS-BRAK viruses, whereas cells preinfected with RCAS-BASP1 were almost completely resistant to v-myc-induced Hartl et al.

BASP1 interferes with the activation of direct Myc target genes like WS5 or Q83 (19, 22), QEFs were transfected with pRCAS-MC29, pRCAS-MC29-IRES-BASP1, or pRCAS-BASP1. Normal QEF or QEF/RCAS-BASP1 cells did not contain WS5 or Q83 mRNAs, whereas high levels were found in the Myc-transformed QEF/ RCAS-MC29 cells, in agreement with previous reports (19, 22). Cells transfected with the bicistronic construct pRCAS-MC29IRES-BASP1 contained no or low levels of WS5 or Q83 mRNAs, indicating an inhibitory effect of BASP1 on target gene activation by Myc (Fig. 5A). Accordingly, expression of BRAK was abolished only in the RCAS-MC29-transformed cells. The Myc-transformed cell line QEF/MC29 was transfected with a luciferase reporter construct containing the WS5 promoter (pWS5-LUC) and a pRc expression vector containing the BASP1 coding region (pRcBASP1) or engineered deletion mutants (Fig. 5B). Ectopic WT or mutant BASP1 protein expression was monitored by immunoblotting (Fig. 5C). The promoter analysis showed that ectopic WT BASP1 expression efficiently interferes with Myc-mediated transcriptional activation (Fig. 5D). Deletion of 10 amino acid residues from the N-terminal BASP1 domain (⌬MN) almost abrogated this inhibitory function, whereas even large deletions in other regions (⌬PS, ⌬RP, ⌬CT) had no significant effect. As a control, ectopic expression of the transcription factor C/EBP␣ led to strong promoter activation (Fig. 5D), in agreement with our previous identification of binding sites for Myc and C/EBP␣ on the WS5 promoter (19). In normal CEFs containing low levels of C/EBP␣ (19), ChIP analysis revealed association of c-Myc, BASP1, and Sp1 with the WS5 promoter (Fig. S1D). The mutant BASP1 proteins (Fig. 5B) were also expressed from RCAS vectors (compare Fig. 4B) and tested for their potential to interfere with Myc-induced focus formation, including 2 constructs in which amino acids 1–11 or 12–244 from BASP1 were fused to the GFP. The results indicate that the N-terminal region of BASP1 is essential for its inhibitory function (Fig. 5 E and F). Further mutational analysis revealed that complete integrity of the N-terminal region including the myrisPNAS 兩 April 7, 2009 兩 vol. 106 兩 no. 14 兩 5607

BIOCHEMISTRY

BASP1 Prevents Myc Target Gene Activation. To test whether ectopic

Fig. 5. Inhibition of transcriptional activation of Myc target genes and mutational analysis of BASP1. (A) Northern analysis of poly(A)⫹-selected RNAs (2.5 ␮g) from QEFs or QEFs transfected with pRCAS-MC29, pRCAS-MC29-IRESBASP1, or pRCAS-BASP1 (compare Figs. 3 and 4). The filter was hybridized with 32P-labeled cDNA probes specific for WS5, Q83, BRAK, or GAPDH. Sizes of the mRNAs were: WS5, 2.8 kb; Q83, 0.9 kb; BRAK, 1.8 kb; GAPDH, 1.4 kb. (B) Schematic diagram of WT and engineered mutant BASP1 proteins. The Nterminal region (MGGKLSKKKKG) is shown in black, and fused GFP is in white. (C) WT and mutant BASP1 proteins were expressed from pRc vector constructs in the v-myc-transformed cell line QEF/MC29 and detected by immunoblot analysis. (D) Transcriptional activation of the luciferase reporter construct pWS5-LUC (2 ␮g) transfected into QEF/MC29 cells together with the empty pRc vector (4 ␮g) or the pRc constructs encoding WT or mutant BASP1 proteins or the C/EBP␣ transcription factor. One day after nucleofection, luciferase activities were determined in triplicate from 5-␮l aliquots of cell extracts. (E) Immunoblot analysis of cells expressing BASP1 or BASP1 mutant proteins encoded by RCAS vectors, supertransfected with pRc-Myc as in Fig. 4B. Proteins were detected with antisera directed against BASP1, GFP, or v-Myc proteins. (F) Focus formation of the pRc-Myc supertransfected cells described in E.

toylation site (G) and putative nuclear localization (KKKK) signal is necessary for the full potential of BASP1 to inhibit target promoter activation and cell transformation (Fig. S4). To test whether the inhibitory functions of BASP1 were mediated by direct interaction with the Myc protein, coimmunoprecipitation experiments were carried out. Under conditions that allow detection of Myc/Max interaction in normal fibroblasts, direct interaction between BASP1 and Myc could not be detected (Fig. S5). 5608 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0812101106

Discussion The oncogenic capacity of the Myc protein is based on its biochemical function as a regulator of gene expression, exhibiting both gene-activating and gene-repressing potential. Because transcriptional control by Myc involves either activation or repression of genes, regulation of both growth-promoting and -inhibiting targets is implicated in establishing the Myc-induced oncogenic phenotype (3, 10, 11). The specific identification of such genes is severely complicated by the large number of direct or indirect putative Myc targets identified by extensive expression profiling or genomic DNA binding studies (3). Furthermore, the possible role of Myc as a coactivator or corepressor of other transcriptional regulators and the recent evidence that some functions of Myc may be independent of Max or E-box binding even amplify the range of possible Myc targets (3, 8, 9, 23, 24). We have been aiming at the isolation of genes that are specifically regulated by oncogenic transcription factors and display either oncogenic or tumor-suppressive potential. Following this approach, we have recently described the isolation of the direct Jun-activated targets TOJ3 (MSP58) and JAC, both displaying intrinsic oncogenic activity (25, 26), and the direct Myc-activated target WS5, related to human melanoma glycoprotein genes showing cell- transforming potential (19). The BASP1 gene now identified as a specific Myc target has several unique properties. Repression of BASP1 gene transcription is highly specific for Myc-transformed cells and not a general cellular response to oncogenic transformation. The kinetics of immediate BASP1 repression and reactivation in a conditional Myc transformation system also point to a specific molecular link between oncogenic Myc and BASP1 transcriptional repression. However, the 5⬘ control region of BASP1 that is sufficient to mediate promoter repression in reporter gene assays does not contain a typical Myc binding site. As discussed above, this does not rule out control by Myc, because transcriptional repression by Myc may occur independently of E-box binding via protein–protein interactions on the core promoter as it has been reported for several genes suppressed by Myc (3, 9). The most striking property of BASP1 is its potential to block the initiation of Myc-induced cell transformation, when the repression by Myc is circumvented by ectopic expression of BASP1 from retroviral vectors. Again, this is a highly specific effect, because ectopic expression of BASP1 does not inhibit cell transformation by the jun or src oncogenes. Furthermore, ectopic expression of the transformation-sensitive BRAK gene does not interfere with Mycinduced cell transformation. The inhibitory effect was independently demonstrated by simultaneous expression of Myc and BASP1 proteins from a bicistronic vector or preinfection of cultured cells with a BASP1 vector before they were challenged with oncogenic Myc. Both routes prevented cell transformation by the Myc protein almost completely. Subsequent introduction of ectopic BASP1 into cells after Myc-induced transformation has been fully established leads to a significant, but only partial, reversion of the transformed phenotype. The BASP1 protein was originally isolated as a membrane- and cytoskeleton-associated protein from rat and chicken brain. Although the protein is expressed in various tissues, it is particularly abundant in neurons during brain development and implicated in neurite outgrowth (14). BASP1 is a member of the family of growthand motility-associated proteins, including GAP-43 and MARCKS, implicated in the regulation of actin dynamics and membrane structure (27). These proteins share characteristic properties, including N-terminal myristoylation, facilitating association with lipid rafts of the plasma membrane, and a basic effector domain that binds acidic phospholipids including phosphatidylinositol-4,5biphosphate, Ca2⫹-calmodulin, and actin filaments. These interactions are presumably regulated by protein kinase C-mediated phosphorylation (14). Intriguingly, BASP1 was also reported to be located in the nucleus of mouse and human cells of various tissues Hartl et al.

Materials and Methods Cells and Retroviruses. Cell culture, DNA transfection, and transformation of CEFs and QEFs were performed as described (26). The avian retroviruses ASV17, NK24, MC29, MH2, and RSV, and the established quail cell lines Q8, MH2A10, VJ, and Q/tMON were used (18, 19, 21, 30). The retroviral constructs pRCAS-MC29 and pRCAS-MC29-IRES-v-Mil have been described (21). pRCAS-ASV17 was constructed by replacing the 1,248-bp v-myc coding region of pRCAS-MC29 with the 861-bp v-jun coding region of ASV17. To generate pRCAS-MC29-IRES-BASP1, pRCASMC29-IRES-BRAK, and pRCAS-ASV17-IRES-BASP1, the coding regions of chicken BASP1 or BRAK were amplified by PCR from CEF total cDNA and inserted into the NcoI/BamHI sites of the adaptor plasmid pA-CLA12NCO as described (21), yielding pA-BASP1 and pA-BRAK. BASP1 or BRAK were excised from the adapter constructs as XbaI fragments and inserted into pIRES (BD Biosciences). EcoRI/NotI (blunt-ended) fragments containing either IRES-BASP1 or IRES-BRAK were then ligated into the unique KpnI sites (blunt-ended) of pRCAS-MC29 or pRCASASV17, respectively. Derivatives of the adapter pA-BASP1 plasmid were engineered by deleting 288-bp PvuII/StuI (⌬PS), 324-bp RsaI/PvuII (⌬RP), and 34-bp CviQI/NcoI (blunt-ended) (⌬MN) fragments followed by religation. To construct pA-⌬CT, the 90-bp StuI/BamHI fragment was replaced by an oligonucleotide containing a stop codon (5⬘-GGTAAGGCCATTACGGCCG-3⬘). For construction of pA-GFP-⌬MN, a 711-bp NcoI/BsrGI(blunt-ended) fragment from pIRES2-EGFP (Clontech) was ligated together with a 718-bp amplified fragment from pABASP1 into the NcoI/SalI sites of pA-CLA12NCO, resulting in fusion of codons 1–239 (GFP) with codons 12–244 (BASP1). To construct pA-NT-GFP, an oligonu-

1. Bister K, Jansen HW (1986) Oncogenes in retroviruses and cells: Biochemistry and molecular genetics. Adv Cancer Res 47:99 –188. 2. Eisenman RN (2001) Deconstructing Myc. Genes Dev 15:2023–2030. 3. Eilers M, Eisenman RN (2008) Myc’s broad reach. Genes Dev 22:2755–2766. 4. Nesbit CE, Tersak JM, Prochownik EV (1999) MYC oncogenes and human neoplastic disease. Oncogene 18:3004 –3016. 5. Dang CV, Kim JW, Gao P, Yustein J (2008) The interplay between MYC and HIF in cancer. Nat Rev Cancer 8:51–56. 6. Berg T, et al. (2002) Small-molecule antagonists of Myc/Max dimerization inhibit Myc-induced transformation of chicken embryo fibroblasts. Proc Natl Acad Sci USA 99:3830–3835. 7. Soucek L, et al. (2008) Modeling Myc inhibition as a cancer therapy. Nature 455:679 – 683. 8. Mink S, Mutschler B, Weiskirchen R, Bister K, Klempnauer KH (1996) A novel function for Myc: Inhibition of C/EBP-dependent gene activation. Proc Natl Acad Sci USA 93:6635–6640. 9. Wanzel M, Herold S, Eilers M (2003) Transcriptional repression by Myc. Trends Cell Biol 13:146 –150. 10. Dang CV (1999) c-Myc target genes involved in cell growth, apoptosis, and metabolism. Mol Cell Biol 19:1–11. 11. Dang CV, et al. (2006) The c-Myc target gene network. Semin Cancer Biol 16:253– 264. 12. Widmer F, Caroni P (1990) Identification, localization, and primary structure of CAP-23, a particle-bound cytosolic protein of early development. J Cell Biol 111:3035–3047. 13. Maekawa S, Maekawa M, Hattori S, Nakamura S (1993) Purification and molecular cloning of a novel acidic calmodulin binding protein from rat brain. J Biol Chem 268:13703–13709. 14. Korshunova I, et al. (2008) Characterization of BASP1-mediated neurite outgrowth. J Neurosci Res 86:2201–2213. 15. Matsubara M, Nakatsu T, Kato H, Taniguchi H (2004) Crystal structure of a myristoylated CAP-23/NAP-22 N-terminal domain complexed with Ca2⫹/calmodulin. EMBO J 23:712–718. 16. Carpenter B, et al. (2004) BASP1 is a transcriptional cosuppressor for the Wilms’ tumor suppressor protein WT1. Mol Cell Biol 24:537–549. 17. Green LM, Wagner KJ, Campbell HA, Addison K, Roberts SGE (2009) Dynamic interaction between WT1 and BASP1 in transcriptional regulation during differentiation. Nucleic Acids Res 37:431– 440.

Hartl et al.

cleotide containing the BASP1 codons 1–11 (5⬘- CATGGGAGGCAAACTGAGCAAGAAGAAGAAGGGG-3⬘) was inserted together with a PCR fragment containing the GFP codons 2–239 into NcoI/SalI-cut pA-CLA12NCO. To construct pRCASBASP1, pRCAS-BASP1 derivatives, and pRCAS-BRAK, ClaI fragments from pABASP1, pA-BASP1 derivatives, or pA-BRAK were inserted into the pRCAS-BP vector as described (26). Colony assays and focus assays were done as described (26, 30). To monitor focus formation, cells were fixed with ethanol and incubated for 30 min in Giemsas eosin methylene blue solution diluted 1:10. DNA Cloning and Nucleic Acid Analysis. Molecular cloning, DNA sequencing, Northern analysis, and primer extension have been described (21, 26). To isolate a 793-bp SmaI genomic fragment containing the 281-bp chicken BASP1 exon 1 including 379 bp of 5⬘-untranscribed region (GenBank accession no. EU888907), a genomic library was screened by using a cDNA probe corresponding to nucleotides 1–153 of the chicken BASP1 mRNA sequence (GenBank accession no. NM㛭204116). BRAK was isolated by subtractive hybridization as described (26) using cDNAs synthesized on poly(A)⫹ RNA from normal CEFs, followed by hybridization with total mRNA from ASV17-transformed CEFs. The subtracted probe was used to screen a CEF cDNA library. Transactivation Analysis. To construct pLUC-BASP1, a 511-bp SmaI/SacII (bluntended) segment from the genomic fragment was inserted into the pGL3-Basic vector (Promega). pLUC-WS5 was generated by transferring the 282-bp HindIII/ XbaI (blunt-ended) fragment of pCAT-WS5 (19) into the HindIII (blunt-ended) site of pGL3-Basic. The expression vector pRc-C/EBP␣ has been described (19). To construct pRc-BASP1 and mutant derivatives, the NcoI (blunt-ended)/XbaI inserts from adapter constructs were transferred into pRc/RSV cut with HindIII (bluntended)/XbaI. DNA transfer into cells by nucleofection was done as described (19) using liposome solution V and electroporation program T-20. Protein Analysis. A 15-mer peptide corresponding to the C-terminal residues 229 –243 of the chicken BASP1 protein (H2N-SEAPATNSDQTIAVQ-COOH) and full-length recombinant BASP1 or BRAK proteins were used to generate rabbit polyclonal antisera. Immunoblotting was done as described (18, 25). Immunoprecipitation of 35Smethionine-labeled proteins and SDS/PAGE were done as described with antisera directed against Myc, Mil, and Jun proteins (21). For additional information, see SI Text. ACKNOWLEDGMENTS. We thank Michael Witting for help with the luciferase assays and Doris Bratschun for technical assistance. This work was supported by Austrian Science Fund Grants P17041 and P18148.

18. Oberst C, Hartl M, Weiskirchen R, Bister K (1999) Conditional cell transformation by doxycycline-controlled expression of the MC29 v-myc allele. Virology 253:193–207. 19. Reiter F, Hartl M, Karagiannidis AI, Bister K (2007) WS5, a direct target of oncogenic transcription factor Myc, is related to human melanoma glycoprotein genes and has oncogenic potential. Oncogene 26:1769 –1779. 20. Hromas R, et al. (1999) Cloning of BRAK, a novel divergent CXC chemokine preferentially expressed in normal versus malignant cells. Biochem Biophys Res Commun 255:703–706. 21. Hartl M, Karagiannidis AI, Bister K (2006) Cooperative cell transformation by Myc/ Mil(Raf) involves induction of AP-1 and activation of genes implicated in cell motility and metastasis. Oncogene 25:4043– 4055. 22. Hartl M, et al. (2003) Cell transformation by the v-myc oncogene abrogates c-Myc/Maxmediated suppression of a C/EBP␤-dependent lipocalin gene. J Mol Biol 333:33– 46. 23. Gomez-Roman N, Grandori C, Eisenman RN, White RJ (2003) Direct activation of RNA polymerase III transcription by c-Myc. Nature 421:290 –294. 24. Orian A, et al. (2007) A Myc–Groucho complex integrates EGF and Notch signaling to regulate neural development. Proc Natl Acad Sci USA 104:15771–15776. 25. Bader AG, Schneider ML, Bister K, Hartl M (2001) TOJ3, a target of the v-Jun transcription factor, encodes a protein with transforming activity related to human microspherule protein 1 (MCRS1). Oncogene 20:7524 –7535. 26. Hartl M, Reiter F, Bader AG, Castellazzi M, Bister K (2001) JAC, a direct target of oncogenic transcription factor Jun, is involved in cell transformation and tumorigenesis. Proc Natl Acad Sci USA 98:13601–13606. 27. Wiederkehr A, Staple J, Caroni P (1997) The motility-associated proteins GAP-43, MARCKS, and CAP-23 share unique targeting and surface activity-inducing properties. Exp Cell Res 236:103–116. 28. Ohsawa S, Watanabe T, Katada T, Nishina H, Miura M (2008) Novel antibody to human BASP1 labels apoptotic cells postcaspase activation. Biochem Biophys Res Commun 371:639 – 643. 29. Wagner KJ, Roberts SGE (2004) Transcriptional regulation by the Wilms’ tumor suppressor protein WT1. Biochem Soc Trans 32:932–935. 30. Bister K, Hayman MJ, Vogt PK (1977) Defectiveness of avian myelocytomatosis virus MC29: Isolation of long-term nonproducer cultures and analysis of virus-specific polypeptide synthesis. Virology 82:431– 448.

PNAS 兩 April 7, 2009 兩 vol. 106 兩 no. 14 兩 5609

BIOCHEMISTRY

(16, 17, 28), and it was proposed that myristoylated and nonmyristoylated forms of the protein have different subcellular locations. It was reported that the nuclear form of BASP1 associates with the WT1 transcription factor and attenuates its transcriptional activity (16, 17, 29). WT1 is associated with tumor suppressive and oncogenic potential, and BASP1 was proposed to function as a cosuppressor for WT1 (29). The exact biochemical function of BASP1 is unknown, and it remains to be determined which molecular mechanisms lead to inhibition of Myc-induced oncogenesis by ectopic expression of BASP1 and by which precise mechanism endogenous BASP1 expression is shut off in Myc-transformed cells. Elucidation of the apparently mutually-exclusive functions of BASP1 and oncogenic Myc may facilitate the further delineation of the oncogenic pathways induced by this oncoprotein.