Large-scale analysis by SAGE reveals new mechanisms of v-erbA oncogene action
© Bresson et al; licensee BioMed Central Ltd. 2007
Received: 21 May 2007
Accepted: 26 October 2007
Published: 26 October 2007
The v-erbA oncogene, carried by the Avian Erythroblastosis Virus, derives from the c-erbAα proto-oncogene that encodes the nuclear receptor for triiodothyronine (T3R). v-ErbA transforms erythroid progenitors in vitro by blocking their differentiation, supposedly by interference with T3R and RAR (Retinoic Acid Receptor). However, v-ErbA target genes involved in its transforming activity still remain to be identified.
By using Serial Analysis of Gene Expression (SAGE), we identified 110 genes deregulated by v-ErbA and potentially implicated in the transformation process. Bioinformatic analysis of promoter sequence and transcriptional assays point out a potential role of c-Myb in the v-ErbA effect. Furthermore, grouping of newly identified target genes by function revealed both expected (chromatin/transcription) and unexpected (protein metabolism) functions potentially deregulated by v-ErbA. We then focused our study on 15 of the new v-ErbA target genes and demonstrated by real time PCR that in majority their expression was activated neither by T3, nor RA, nor during differentiation. This was unexpected based upon the previously known role of v-ErbA.
This paper suggests the involvement of a wealth of new unanticipated mechanisms of v-ErbA action.
The Avian Erythroblastosis Virus (AEV) induces erythroleukemia and sarcomas in chickens . This retrovirus also transforms erythroid progenitors and chicken embryo fibroblasts in vitro. The AEV carries two viral oncogenes in its genome, v-erbA and v-erbB. The expression of v-ErbA alone is sufficient to block differentiation of erythrocytic progenitors in vitro and in vivo [2, 3]. v-erbA is derived from the c-erbA proto-oncogene, which encodes the alpha form of the nuclear receptor for the thyroid hormone triiodothyronine (T3Rα) . This oncogene is expressed as a gag-v-ErbA fusion protein that binds to the DNA but does not bind the hormone T3 .
The most widely accepted hypothesis is that v-ErbA antagonizes ligand-dependent activation by T3R, to silence T3 responsive genes even in the presence of T3 [5–7]. The finding that overexpression of T3R and addition of ligand can overcome the blockade of differentiation induced by v-ErbA supports this assumption . Moreover, it was shown that v-ErbA also inhibits transcription mediated by ligand-activated Retinoic Acid Receptors (RAR) and that this inhibition is also of importance since a non-transforming point mutant for v-ErbA, S61G, is unable to block transcription mediated by RAR .
It has therefore been proposed that the v-ErbA oncoprotein represses genes that are normally upregulated by T3 and Retinoic Acid (RA) and that are upregulated during the differentiation process [5, 9]. Nevertheless, this hypothesis remains speculative as long as v-ErbA target genes harboring those properties are not found. So far, four v-ErbA target genes have been described: Carbonic anhydrase II (CAII), erythrocyte anion transporter (Band3), integrin α2 subunit (VLA2) and δ-aminolevulinate synthase (ALA-S) [10–12]. It has been demonstrated only in the case of the CAII gene that v-ErbA acts as a dominant negative of T3R by occluding the DNA response element and ablating the T3 mediated activation . In contrast, the expression of Band3, ALA-S and VLA2 genes is not controlled by T3 suggesting that v-ErbA may act independently of T3R [10, 11]. This is consistent with the observation that v-ErbA also interferes with pathways controlled by other transcription factors [9, 14]. Furthermore, functional studies have clearly demonstrated that the repression of Band3 and CAII by v-ErbA is not sufficient to account for the differentiation blocking ability of this oncoprotein .
Therefore, the precise mechanisms of v-ErbA action remain unclear and the identification of relevant v-ErbA target genes represents a mandatory step in its understanding. We thus decided to use a large-scale transcriptomic approach by SAGE (Serial Analysis of Gene Expression) to identify v-ErbA target genes involved in its transforming activity. One of the most important advantages of this technique is the possibility to obtain a comprehensive view of all the transcripts present in a cell population at a given time point, allowing the discovery of new genes . We have used a cellular model of normal immature erythroid progenitors called T2ECs (TGF-α TGF-β induced erythrocytic cells)  which are the natural target cells of v-ErbA. In order to identify v-ErbA target genes responsible for the transformation process induced by v-ErbA, we have compared the transcriptome of T2ECs expressing an oncogenic form of v-ErbA with the transcriptome of T2ECs expressing the S61G mutant of v-ErbA [9, 18]. This mutant, which carries a glycine in the DNA binding domain, is equivalent to the c-erbA sequence at this position, and is defective in its ability to inhibit differentiation and to induce erythroid transformation .
Thus, the comparison between the transcriptome of cells expressing either the transforming form of v-ErbA or the S61G mutant of v-ErbA allowed us to generate a list of 110 differentially expressed genes between these two conditions, 15 of which were observed to be repressed by v-ErbA by real-time PCR on different cell populations. A bioinformatic approach revealed the presence of c-Myb binding sites in the promoters of v-ErbA target genes. The identification of these sites coupled with a functional c-Myb transcriptional assay point out a putative role of c-Myb in the transformation process. Unexpectedly, the global analysis of gene functions reveals that two main cellular functions, protein metabolism and chromatin structure/transcriptional regulation, could be deregulated by v-ErbA. We further show here that the expression of most newly identified v-ErbA target genes is neither under the control of T3 nor RA, nor upregulated during differentiation. In conclusion, we have identified several new v-ErbA target genes unknown up to now and we present evidence that v-erbA oncogene action involves new unanticipated mechanisms of action.
Potential v-ErbA target genes identified by SAGE analysis
We constructed two SAGE libraries of T2ECs infected either with the XJ12 retrovirus which expresses the oncogenic form of v-erbA (VA) or with the S61G retrovirus which expresses a non-transforming form of v-erbA (NTVA) in order to identify v-ErbA target genes responsible for transformation. 59391 tags were obtained from the VA library and 37347 tags from the NTVA library. We obtained 5764 different tags appearing more than once in the VA library and 4153 different tags in the NTVA library.
Presence of c-Myb binding sites in the promoters of v-ErbA repressed target genes
In order to investigate the molecular mechanisms by which v-ErbA regulates gene expression, we next searched for common motifs in the promoters of the v-ErbA target genes identified by SAGE. For that purpose, we first identified SAGE tags by using Identitag  and then aligned the corresponding transcripts on the chicken genome. It has been demonstrated for CAII that v-ErbA binds to several domains in the proximal promoter sequence, up to 600 bp of the transcription start site (TSS) . Moreover, it has been demonstrated that the position of transcription factor binding sites often lie within a few kb 5' of the basal promoter . We therefore decided to consider as promoter regions 3 kbp 5' of the TSS and 1 kbp 3' of the TSS.
We obtained 31 promoter sequences from genes that appeared as being repressed by the transforming form of v-ErbA by SAGE and 21 sequences from genes that appeared as being activated by the same form of v-ErbA compared to the non transforming form of v-ErbA. We first searched for known v-ErbA response elements [22–25]. All the nuclear receptors of the c-ErbA family recognize derivatives of the same hexameric DNA core motif 5'-AGGTCA-3' . We found response elements based upon this motif only in 3 promoters. The response element PAL-0 (AGGTCATGACCT) is present in RPL11 and Succinate promoters while the response element DR-4 (AGGTCAnnnnAGGTCA) is present in RPS3 and Succinate promoters (data not shown). It has been demonstrated that T3R binds preferentially and rather specifically to T3 response element (TRE) of the DR-4 structure [25, 26]. The palindromic element without a spacer (PAL-0) is not specific to T3R since it works efficiently as a retinoic acid responsive element (RARE) [25, 26]. However, despite the presence of the T3 response element in RPS3 promoter, the expression of this gene did not vary in response to T3 (see below). To conclude, out of 52 promoters analyzed, we have only identified 3 promoters with known response element for T3R, RAR and v-ErbA. It is conceivable that the 4 kb region considered is not sufficient since binding sites that would be more distal to the TSS will be missed by this approach. Furthermore, it is also obvious that the simple presence of a transcription factor binding site (TFBS) does not imply an activity of that transcription factor. Finally, even if the protein binds the TFBS, we cannot be informed about the direction of the regulation (i.e. activation or inhibition). Nevertheless, we have found few v-ErbA/T3R/RAR response elements in this region that is in accordance with the experimental assessment of the expression pattern of newly identified v-ErbA target genes under the influence of T3R and RAR (see below).
Finally, we used a gene reporter assay to test the ability of v-ErbA to transactivate c-Myb. We transfected T2ECs with a reporter plasmid in which five strong c-Myb binding sites were placed upstream of a simple TATA box and a luciferase cDNA . This reporter plasmid (EW5) was clearly active in T2ECs (Figure 3C), which express c-Myb (see above). Moreover, a 1.7-fold increase of luciferase activity was seen with the transforming form of v-ErbA (VA) compared to the S61G mutant for v-ErbA (NTVA). A control reporter gene containing five mutant c-Myb binding sites (EM5) was much less active in these cells, and its activity was unaffected by the different forms of v-ErbA. It is thus conceivable that the c-Myb binding to a real promoter, presenting a higher complexity with several transcription factor binding sites, silencers and enhancers, could result in an inhibition of gene expression controlled by such a promoter. Therefore, although the direction of the functional interaction was unexpected (activation versus repression), these data nevertheless demonstrate that v-ErbA can indeed functionally interact directly or indirectly with the transcriptional activity of endogenous c-Myb in T2ECs.
v-ErbA mostly affects chromatin/transcription and protein metabolism genes
The use of functional annotations (mainly Gene Ontology (GO) annotations) allowed us to find that many v-ErbA repressed target genes are involved in two main cellular functions: chromatin structure/transcriptional regulation and protein metabolism.
The remodelling of chromatin and transcription regulation constituted a cellular function that was expected to be deregulated by v-ErbA. Indeed, v-ErbA represses transcription via association with corepressors such as NCoR and SMRT . The corepressors, in turn, recruit histone deacetylase (HDAC) containing complexes that stabilize repressive chromatin structure to assure efficient silencing of downstream genes [31, 32].
Protein metabolism is the most represented and the most unanticipated category. It suggests the importance of the translation and the protein degradation processes during transformation. Among the genes that are implicated in protein synthesis, we have observed an important number of ribosomal proteins. However, the expression of all ribosomal proteins is not affected by v-ErbA since a vast number of genes encoding ribosomal proteins (RPL3, RPL7, RPL12, RPL24, RPL29, RPS4, RPS6, RPS10 and RPS13) was shown not to vary by real time PCR (data not shown). This suggests that only a specific subset of some ribosomal proteins is affected by v-ErbA.
The other genes down-regulated by v-ErbA, which are not implicated in these two cellular functions, are either implicated in different functions or their functions are unknown. Among these genes, stem cell antigen 2 (sca2) and transforming growth factor-β1 (TGF-β1) encode proteins that are involved in the regulation of intracellular signalling pathways. Indeed, we have observed that v-ErbA is able to modulate the TGF-β signalling pathway in T2ECs (Gonin-Giraud et al., submitted). These data suggest a role for v-ErbA in intracellular signalling pathways in agreement with its cytoplasmic localisation .
In contrast, the candidate genes whose expression is up-regulated by v-ErbA are fewer and are involved in a greater variety of functions, such as sodium-calcium exchange, chaperone activity, nuclear transport and others.
15 v-ErbA target genes validated by real time-PCR on multiple experiments
We observed that genes were not repressed at the same fold variation in the five different experiments. This can be due to a biological variability between different cultures, as we showed in our previous analysis of the transcriptome of T2ECs in a self-renewal or in a differentiation state . By contrast, the expression of the 25 candidates, which have not been confirmed by real-time PCR, could be explained by different hypothesis (multiple or erroneous identification, variability in gene expression, specific gene variations during the culture that was used for preparing SAGE mRNAs, etc...) as developed in Damiola et al., 2004.
To conclude, out of the 40 candidates tested, we have identified 15 v-ErbA repressed target genes. None of these genes were previously known to be a v-ErbA target gene.
The expression of the majority of newly identified v-ErbA target genes is not activated nor repressed during differentiation
Expression of the v-ErbA target genes in T2ECs during the differentiation process or following a T3, RA or T3 and RA treatment.
T3 and RA do not activate the majority of newly identified v-ErbA target genes
The v-ErbA oncoprotein was initially described to be a transcriptional repressor for genes normally activated by T3 Receptor (T3R) and RA Receptor (RAR) [5, 9]. Therefore, we have analyzed the expression of v-ErbA target genes by real-time PCR following a treatment by T3 and RA (Table 1 and see Additional File 2). We observed that, among these genes, some are activated by RAR because their expression is induced by RA treatment in normal cells, such as sca2, LOC129881 and eIF2B2. This confirms that activation by RAR is indeed inhibited by v-ErbA. However, the expression of a vast majority of newly identified v-ErbA target genes is either repressed by T3 or does not vary in response to T3 and RA. These results suggest that v-ErbA must also act by T3R and RAR independent mechanisms in the transformation process.
Moreover, among the 11 differentiation target genes identified, the expression of only 3 genes is activated during the differentiation process and repressed by v-ErbA (Table 1 Figure 6B). To conclude, most of the newly identified v-ErbA target genes are not activated by T3, RA and differentiation.
The newly identified v-ErbA target genes are cell type specific
In this paper, we identified new v-ErbA target genes. All of these genes were found to be repressed by v-ErbA and involved in protein metabolism and in chromatin structure/transcriptional regulation. A role for c-Myb in the transformation process induced by v-ErbA could be one of the indirect mechanisms of v-ErbA action. Furthermore, this study clearly illustrates new mechanisms of v-ErbA action, independent from T3R, RAR and erythroid differentiation.
To identify the target genes of the v-ErbA oncoprotein, previous studies have analyzed either the expression of erythroid specific genes , of genes which are regulated by T3  or the expression of different membrane antigens  in cells transformed by v-ErbA. Thus, v-ErbA has been reported to repress the expression of only four genes: ALA-S, Band3 , CAII  and VLA2 . Repression of Band3 and CAII by v-ErbA cannot explain altogether the blockade of the differentiation induced by the oncogene, since reexpression of these two proteins had little (CAII) or no effect (Band3) on the differentiation of v-ErbA transformed cells . We searched for these four well-established v-ErbA target genes in the two SAGE libraries of T2ECs expressing either the oncogenic form of v-ErbA (VA) or the non-transforming form of v-ErbA (NTVA). We found only CAII in these libraries. Different reasons can explain this result, including polymorphism in the tag sequence, and a weak expression of the other genes under these two conditions. This suggests that more v-ErbA target genes could be found by a deeper sequencing of SAGE libraries. However, we quantified the expression of these previously identified targets of v-ErbA by real-time PCR in several independent experiments. Only VLA2 and Band3 are repressed by the transforming form of v-ErbA compared to the non-transforming one (data not shown). These results suggest that these genes could be indeed involved in the transformation process induced by v-ErbA as previously suggested for Band3 and VLA2 [11, 15]. Thus, the identity of cellular genes whose deregulation by the v-erbA oncogene is responsible for the transformation remained an open question. By using a global approach by SAGE, we compared the transcriptome of T2ECs expressing the oncogenic form of v-ErbA with the transcriptome of T2ECs expressing the non-transforming form of v-ErbA (S61G). This allowed us to identify several new v-ErbA target genes potentially involved in the transformation induced by the oncogene. The vast majority of expressed genes were not differentially expressed between the two conditions, suggesting that the switch from a normal state toward a pathological state involves a very limited number of genes and small variations.
The newly identified v-ErbA target genes can be either primary or secondary consequences of v-ErbA action. Indeed, v-ErbA target genes involved in the transformation process cannot be differentiated from genes whose expression is modulated consequently to the transformation process. The study of the biological function of each v-ErbA target gene could answer in part this question. Moreover, Genome wide Chromatin Immunoprecipitation (ChIP) assay could be performed to determine the amount of target genes that are susceptible to be under the direct influence of v-ErbA. Among the four previously established target genes of v-ErbA, only CAII was demonstrated to be a primary target of v-ErbA [13, 20]. The molecular mechanisms through which ALA-S, Band3 and VLA2 are repressed by v-ErbA remain elusive.
The bioinformatic analysis of the v-ErbA target gene promoter sequences revealed the presence of c-Myb binding sites as a motif that appears more frequently in the promoters of v-ErbA repressed target genes. We observed that v-ErbA upregulates c-Myb expression and that the level of c-myb expression was equivalent in cells expressing either wild-type or S61G forms of v-ErbA, as previously demonstrated by Bauer et al. . Since the non-transforming form of v-ErbA also induces c-Myb upregulation, this does not seem to be sufficient per se to induce the transformation process. In contrast, reporter gene experiments with a plasmid containing c-Myb binding sites upstream the luciferase gene suggest that v-ErbA, in contrast to S61G, could activate c-Myb by either direct or indirect interactions which will lead to the repression of a subset of key genes. These data propose that a functional interaction with c-Myb could play a role in the transforming process induced by v-ErbA. The c-myb proto-oncogene encodes a transcription factor that functions as an activator as well as a repressor of transcription and appears to play a critical role in the regulation of hematopoietic cell differentiation . Its transcription is typically down-regulated during the erythroid cell maturation process  and constitutive expression of c-myb blocks erythroid differentiation .
It has been proposed that v-ErbA represses at least a subset of genes that are normally upregulated by T3 and/or RA and that are possibly linked to the differentiation program of the cell [5, 9]. However, our results suggest that v-ErbA could modulate the transcription of a third group of genes that are not under the control of RAR or T3R, or not affected by the differentiation process. The evidence that all TR target genes are not necessary v-ErbA target genes is also supported by DNA binding specificity studies strengthened by functional assays [23, 24]. The evidence that some RAR target genes are not v-ErbA target genes is also reinforced by experiments using different retinoic acid response elements and seems to depend upon the cell type tested [9, 39–41]. The v-erbA oncogene has accumulated mutations that could have created a v-ErbA specific subset of target genes with some TR target genes, some RAR target genes and possibly some target genes for other transcription factors. This could lead to the specific repression of a subset of relevant key genes.
The protein synthesis function is one of the most prominent category of v-ErbA deregulated genes, and one of the most unexpected. We observed a reduction in the expression of several genes encoding proteins involved in protein synthesis, suggesting an important role of variations in translation during transformation. However, the expression of all proteins involved in translation, in particularly ribosomal proteins, is not affected by v-ErbA. It is therefore unlikely that a global change in the amount of ribosomes might result from v-ErbA action. In support of this, several studies have suggested an extraribosomal function of at least some ribosomal proteins . Interestingly, ribosomal proteins, translation initiation and elongation factors have been found to play roles in regulating cell growth and transformation due to such extraribosomal activities . One should also note that variation in the amount of mRNA encoding ribosomal proteins has already been observed to be a common feature in large scale transcriptomic analysis ( and references therein).
Altogether, this paper suggests the involvement of a wealth of new unanticipated and possibly indirect mechanisms of v-ErbA action, independently of T3R, RAR and differentiation. The function of the deregulated genes is currently under study in our laboratory.
TGF-β1 and TGF-α were purchased from PeproTech as a lyophilized human recombinant protein and resuspended as previously described .
All-trans Retinoic acid was purchased from SIGMA, prepared as a 10-3 M solution in ethanol and subsequently diluted in α-MEM.
3,3',5-Triiodo-L-thyronine (Triiodothyronine) was purchased from SIGMA, prepared as a 10-3 M solution in 0,1N NaOH and subsequently diluted in α-MEM.
T2ECs were generated from SPAFAS white leghorn chickens (PA12 line from INRA, Tours, France) as previously described . These cells were expanded in LMI medium (α-MEM medium, 10% FBS (Fetal Bovine Serum), 10-3 M Hepes, 10-4 M β-mercaptoethanol, 10-6 M dexamethasone, 5 ng/ml TGF-α, 1 ng/ml TGF-β1, penicillin and streptomycin (100 U/ml)).
Differentiating cells were obtained by changing the medium of exponentially growing cells into DM17 medium (α-MEM medium, 10% FBS (Fetal Bovine Serum), 10-3 M Hepes, 10-4 M β-mercaptoethanol, penicillin and streptomycin (100 U/ml), 15% ACS (anemic chicken serum), 10 ng/ml of insulin) . ACS was prepared as previously described .
Chicken Embryo Fibroblasts (CEFs) were generated from SPAFAS white leghorn chickens (LD1 line from INRA, Tours, France) as previously described . These cells were expanded in growth medium (Ham's F10 medium, TPB, 5% FBS, 1% NCS (Normal Chicken Serum), 0,1% Sodium Bicarbonate, penicillin and streptomycin (100 U/ml) and 2,5 μg/ml Amphotericin B).
Viruses were produced by cotransfection of pXJ12  or pS61G  with pRAV-1 plasmid on CEFs. pXJ12 coexpresses the Neor gene and the v-erbA oncogene  and pS61G coexpresses the Neor gene and a mutant version of v-erbA oncogene . CEFs were infected with viral supernatant XJ12 and S61G previously obtained from CEFs transfected, selected in growth medium containing 0,4 mg/ml G418 (Gibco) for 4 days and then expanded in growth medium.
T2ECs were infected with viral supernatant obtained from CEFs infected and selected for 5 days in the presence of 3 mg/ml G418. The infected cells were purified by centrifugation through a density gradient on Lymphocytes Separation Medium (Eurobio) to get rid of dead cells and debris, and seeded in LMI medium.
SAGE libraries generation and data analysis
RNA extraction and reverse transcription assay were performed as previously described .
We prepared RNA samples from T2ECs infected with XJ12 (VA) or infected with S61G (NTVA). Both cultures were originated from the same cell preparation and divided just before the infection of cells. The infection was validated by PCR verification of v-erbA mRNA, and verification of the presence of the mutation S61G before proceeding with the construction of the libraries (data not shown).
SAGE library construction was performed with 10 μg of total RNA, using the I-SAGE™ kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. New PCR primers were redesigned for verification steps on chicken mRNAs. cDNA synthesis verification was performed using chicken Cyclin D1 primers (5'-ACGCTCAGGGACTATACAGG-3' and 5'-GTCTGATGGAGTTGTCGG-3'; AT = 62°C) and chicken Cyclin D2 primers (5'-GTGCTAGTAAGCAACCTTAGGC-3' and 5'-GAACAAGCCTGACTTTCTGTGC-3'; AT = 55°C). NlaIII digestion verification was performed using chicken Cyclin D1 primers and chicken c-rel primers (5'-CTGCTCGAATTCAAGCTTCT-3' and 5'-CCAGTTCTTTAATTACCAAACC-3'; AT = 55°C). Linker ligation and BsmFI digestion verification was performed using the DTP-1 and DTP-2 primers provided with the I-SAGE™ kit together with one Cyclin D2 primer. Concatemer clone sequencing was performed by Genome Express (Meylan, France) and by the Genoscope (Evry, France). Tags were extracted from concatemer sequences by using the R package Sagenhaft . The occurrence number of each tag was computed by taking into account the sequencing errors in concatemer sequences (estimated by the Phred quality score for each base), by using the modelling described in Beissbarth et al., 2004.
Tag identification was performed using Identitag  and chicken transcript sequences from the TIGR Gene Index release 10 and from the Refseq databank.
In order to assess the significance of the differential expression of tags between the two libraries, we implemented the Z-test  using the R language. We adjusted the resulting p-values for multiple testing by using the method proposed by Benjamini and Hochberg  and implemented in the Bioconductor R package multtest.
Because the completion of annotation is higher for human genes than for chicken ones, the classification into functional groups is based on the annotation of human genes orthologous to chicken differentially expressed genes. The classification was made with the DAVID software .
Real-Time PCR was performed with a LightCycler system (Roche) or MX3000P Stratagene using the LightCycler FastStart DNA Master SYBR Green I (Roche) according to the manufacturer's instructions. Primer design was carried out using the Primer3 software system . In order to assess the quality and the efficiency of the RT reactions, we analysed the expression of different invariant genes, namely T-Complex1 (Sigenae accession number: BBSRC60295473F184.108.40.206), hnRNP (Sigenae accession number: AB038220.127.116.11), ATP Synthase subunit B1 (GenBank accession number: XM_417993.1), GAPDH (GenBank accession number: K01458) and eEF1α1 (GenBank accession number: NM_204157.2) (only in differentiation experiments) as internal standards. In order to quantify gene expression, standard curves were generated, for each pair of primers, by dilution series of a sample. The relative expression ratio (R) for each target gene was performed using a previously published mathematical method :
ratio = (Etarget) ΔCP target(control - sample)/(Eref) ΔCP ref(control - sample)
The reference gene is a house-keeping gene transcript. For the target gene and the reference gene, E is the real time PCR efficiency and ΔCP is the crossing point deviation of control versus sample. Real time PCR efficiencies were calculated according to E = 10(-1/slope). For information about the specific PCR conditions and primers sequences used, please contact the authors.
Western Blot Analysis
Cells were harvested and washed in cold PBS supplemented with 1% phosphatase inhibitor cocktail 2 (Sigma) and complete mini EDTA-free protease inhibitor cocktail (Roche). The protein concentration was determined by using the Dc protein Assay (BioRad). Cells were then lysed in Laemmli buffer (0,125 M Tris pH6,8, 2% SDS, 10% glycerol, 5% β-mercaptoethanol and 0,1% bromophenol blue) following 5 min at 100°C. 30 μg of proteins were analyzed in a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (12,5% acrylamide) and blotted on Protean BA85 nitrocellulose membrane. After blockage in 0,1% tween-20 tris-buffered saline (TBS) 5% nonfat milk during 1 hour at room temperature, the membrane was probed with primary antibodies (dilution 1/200 in TBS-0,1%Tween-5% nonfat milk for c-Myb antibody (M-19, Santa Cruz) with an incubation of 4 hours at room temperature; dilution 1/30000 in TBS-0,1%Tween-5% BSA for GAPDH antibody (14C10, Cell signaling) with an incubation of 1 hour at room temperature). The membranes were washed twice in TBS-0,1%Tween before incubation during 1 hour at room temperature with a horseradish-peroxidase-coupled secondary anti-rabbit IgG antibody (dilution 1/7000 in TBS-0,1%Tween-5% nonfat milk for c-Myb analysis or in TBS-0,1%Tween-5% BSA for GAPDH analysis). Proteins were detected using the ECL plus western blotting detection system (Amersham).
Promoter sequence analysis
The WASHUC1 chicken genome assembly was downloaded from the Ensembl ftp site  as of September 2006. Transcripts were aligned on the chicken genome using the BLAT algorithm . The most 5' position of this alignment was used as a putative transcription start site (TSS). 3 kbp 5' of the TSS and 1 kbp 3' of the TSS were extracted and stored as promoter regions. This gave us two sets of promoters: 31 sequences from genes that appeared as being repressed by v-ErbA by SAGE and 21 sequences from genes that appeared as being activated by v-ErbA. We then applied on those two sets of sequences an original algorithm called FAVST (Finite Automata-based VST construction) . Marguerite, an implementation of an extended FAVST algorithm will be described elsewhere (Mitasiunaité, I. et al., in preparation; see also ). In contrast with other motif-searching algorithm, Marguerite does not try to find statistically overrepresented motifs, as compared to a random distribution. Marguerite is an exact algorithm that searches exhaustively for regular expression-type patterns that are over-represented in a given set of sequences (positive set) and under-represented in a different but related set of sequences (negative set) without any a priori about these patterns excepted their minimal length. Three parameters are required for running Marguerite: the minimum length of the pattern, the maximum number of examples in the negative set, and the minimum number of examples in the positive set (see legend to figure 2). The identification of the putative transcription factor binding motifs found was done using the PATCH function of the commercial version of TRANSFAC .
Transcriptional activation assay
Transient transfections were performed by nucleofection method (Amaxa Technology). For each nucleofection assay, 10 × 106 cells were resuspended in 100 μl of nucleofector buffer (Cell line Nucleofector kit V, reference VCA-1003; Amaxa Biosciences) and nucleofected with the T-16 program. The β-galactosidase expressing plasmid (pCMV-β-Gal; 0,5 μg), luc reporter plasmids containing either five wild-type (polyA-EW5-LUC; 5 μg) or mutant (poly-A-EM5-LUC; 5 μg) Myb binding sites , basic pGL2 (5 μg; Promega), pXJ12 expressing the v-erbA oncogene (VA; 5 μg) [35, 45] and pS61G expressing a mutant version of v-erbA (NTVA; 5 μg) [9, 18] were transfected into T2ECs. Twenty-four hours after transfection, the cells were assayed for luciferase activity (Bright-Glo™ Luciferase Assay system; Promega) and β-Gal activity (Gal-screen® System; Applied Biosystems). All transfections were performed two to three times and the pCMV-β-Gal control was used to normalize the transfection experiments. Relative luciferase activities were obtained.
We thank the Centre National de Séquençage (Génoscope), and especially Patrick Wincker and Carolle Dossat for their help in sequencing of one SAGE library. Sequencing at Genoscope was supported by CNRG. We are indebted toward Ieva Mitasiunaité, Christophe Rigotti and Jean-François Boulicaut for their invaluable help in using Marguerite, and for stimulating discussions. We are grateful to Amel Bendjeladi for technical help with the real time PCR experiments. We would like to warmly thank Hong Wen and Joseph Lipsick for their generous gift of c-Myb-responsive reporter plasmids, François Morle's and Guy Mouchiroud's groups for their gift of respectively pGL2 and pCMV-β-Gal. We thank also Edmund Derrington for manuscript corrections. This work was supported by grants from the Ligue contre le cancer, the CNRS, the UCBL, the Region Rhône Alpes, the Fondation de France and the Association pour la recherche contre le cancer (ARC).
- Graf T, Beug H: Avian leukemia viruses: interaction with their target cells in vivo and in vitro. Biochim Biophys Acta. 1978, 516 (3): 269-299.PubMedGoogle Scholar
- Casini T, Graf T: Bicistronic retroviral vector reveals capacity of v-erbA to induce erythroleukemia and to co-operate with v-myb. Oncogene. 1995, 11 (6): 1019-1026.PubMedGoogle Scholar
- Gandrillon O, Jurdic P, Pain B, Desbois C, Madjar JJ, Moscovici MG, Moscovici C, Samarut J: Expression of the v-erbA product, an altered nuclear hormone receptor, is sufficient to transform erythrocytic cells in vitro. Cell. 1989, 58 (1): 115-121. 10.1016/0092-8674(89)90408-X.PubMedView ArticleGoogle Scholar
- Sap J, Munoz A, Damm K, Goldberg Y, Ghysdael J, Leutz A, Beug H, Vennstrom B: The c-erb-A protein is a high-affinity receptor for thyroid hormone. Nature. 1986, 324 (6098): 635-640. 10.1038/324635a0.PubMedView ArticleGoogle Scholar
- Damm K, Thompson CC, Evans RM: Protein encoded by v-erbA functions as a thyroid-hormone receptor antagonist. Nature. 1989, 339 (6226): 593-597. 10.1038/339593a0.PubMedView ArticleGoogle Scholar
- Sap J, Munoz A, Schmitt J, Stunnenberg H, Vennstrom B: Repression of transcription mediated at a thyroid hormone response element by the v-erb-A oncogene product. Nature. 1989, 340 (6230): 242-244. 10.1038/340242a0.PubMedView ArticleGoogle Scholar
- Thormeyer D, Baniahmad A: The v-erbA oncogene (review). Int J Mol Med. 1999, 4 (4): 351-358.PubMedGoogle Scholar
- Disela C, Glineur C, Bugge T, Sap J, Stengl G, Dodgson J, Stunnenberg H, Beug H, Zenke M: v-erbA overexpression is required to extinguish c-erbA function in erythroid cell differentiation and regulation of the erbA target gene CAII. Genes Dev. 1991, 5 (11): 2033-2047. 10.1101/gad.5.11.2033.PubMedView ArticleGoogle Scholar
- Sharif M, Privalsky ML: v-erbA oncogene function in neoplasia correlates with its ability to repress retinoic acid receptor action. Cell. 1991, 66 (5): 885-893. 10.1016/0092-8674(91)90435-2.PubMedView ArticleGoogle Scholar
- Pain B, Melet F, Jurdic P, Samarut J: The carbonic anhydrase II gene, a gene regulated by thyroid hormone and erythropoietin, is repressed by the v-erbA oncogene in erythrocytic cells. New Biol. 1990, 2 (3): 284-294.PubMedGoogle Scholar
- Mey A, Gandrillon O, McNagny KM, Clegg DO, Samarut J: The v-erbA oncogene blocks expression of alpha2/beta1 integrin a normal inhibitor of erythroid progenitor proliferation. Oncogene. 2002, 21 (18): 2864-2872. 10.1038/sj.onc.1205411.PubMedView ArticleGoogle Scholar
- Zenke M, Kahn P, Disela C, Vennstrom B, Leutz A, Keegan K, Hayman MJ, Choi HR, Yew N, Engel JD: v-erbA specifically suppresses transcription of the avian erythrocyte anion transporter (band 3) gene. Cell. 1988, 52 (1): 107-119. 10.1016/0092-8674(88)90535-1.PubMedView ArticleGoogle Scholar
- Braliou GG, Ciana P, Klaassen W, Gandrillon O, Stunnenberg HG: The v-ErbA oncoprotein quenches the activity of an erythroid-specific enhancer. Oncogene. 2001, 20 (7): 775-787. 10.1038/sj.onc.1204159.PubMedView ArticleGoogle Scholar
- Bauer A, Ulrich E, Andersson M, Beug H, von Lindern M: Mechanism of transformation by v-ErbA: substitution for steroid hormone receptor function in self renewal induction. Oncogene. 1997, 15 (6): 701-715. 10.1038/sj.onc.1201208.PubMedView ArticleGoogle Scholar
- Fuerstenberg S, Leitner I, Schroeder C, Schwarz H, Vennstrom B, Beug H: Transcriptional repression of band 3 and CAII in v-erbA transformed erythroblasts accounts for an important part of the leukaemic phenotype. Embo J. 1992, 11 (9): 3355-3365.PubMed CentralPubMedGoogle Scholar
- Velculescu VE, Zhang L, Vogelstein B, Kinzler KW: Serial analysis of gene expression. Science. 1995, 270 (5235): 484-487. 10.1126/science.270.5235.484.PubMedView ArticleGoogle Scholar
- Gandrillon O, Schmidt U, Beug H, Samarut J: TGF-beta cooperates with TGF-alpha to induce the self-renewal of normal erythrocytic progenitors: evidence for an autocrine mechanism. Embo J. 1999, 18 (10): 2764-2781. 10.1093/emboj/18.10.2764.PubMed CentralPubMedView ArticleGoogle Scholar
- Bonde BG, Sharif M, Privalsky ML: Ontogeny of the v-erbA oncoprotein from the thyroid hormone receptor: an alteration in the DNA binding domain plays a role crucial for v-erbA function. J Virol. 1991, 65 (4): 2037-2046.PubMed CentralPubMedGoogle Scholar
- Keime C, Damiola F, Mouchiroud D, Duret L, Gandrillon O: Identitag, a relational database for SAGE tag identification and interspecies comparison of SAGE libraries. BMC Bioinformatics. 2004, 5: 143-10.1186/1471-2105-5-143.PubMed CentralPubMedView ArticleGoogle Scholar
- Rascle A, Ghysdael J, Samarut J: c-ErbA, but not v-ErbA, competes with a putative erythroid repressor for binding to the carbonic anhydrase II promoter. Oncogene. 1994, 9 (10): 2853-2867.PubMedGoogle Scholar
- Wray GA, Hahn MW, Abouheif E, Balhoff JP, Pizer M, Rockman MV, Romano LA: The evolution of transcriptional regulation in eukaryotes. Mol Biol Evol. 2003, 20 (9): 1377-1419. 10.1093/molbev/msg140.PubMedView ArticleGoogle Scholar
- Bonde BG, Privalsky ML: Sequence-specific DNA binding by the v-erbA oncogene protein of avian erythroblastosis virus. J Virol. 1990, 64 (3): 1314-1320.PubMed CentralPubMedGoogle Scholar
- Judelson C, Privalsky ML: DNA recognition by normal and oncogenic thyroid hormone receptors. Unexpected diversity in half-site specificity controlled by non-zinc-finger determinants. J Biol Chem. 1996, 271 (18): 10800-10805. 10.1074/jbc.271.18.10800.PubMedView ArticleGoogle Scholar
- Subauste JS, Koenig RJ: Comparison of the DNA binding specificity and function of v-ErbA and thyroid hormone receptor alpha 1. J Biol Chem. 1995, 270 (14): 7957-7962. 10.1074/jbc.270.14.7957.PubMedView ArticleGoogle Scholar
- Wahlstrom GM, Sjoberg M, Andersson M, Nordstrom K, Vennstrom B: Binding characteristics of the thyroid hormone receptor homo- and heterodimers to consensus AGGTCA repeat motifs. Mol Endocrinol. 1992, 6 (7): 1013-1022. 10.1210/me.6.7.1013.PubMedGoogle Scholar
- Rascle A, Gandrillon O, Cabello G, Samarut J: The v-erbA oncogene. Oncogenes as transcriptional regulators. Edited by: Ghysdael J, Yaniv M. 1997, Basel , Birkhäuser Publishing Ltd, Vol. 1: Retroviral oncogenes: 119-165.Google Scholar
- Lee SD, De Raedt L: An efficient algorithm for mining string databases under constraints. Proceedings KDID’04. 2004, Springer-Verlag: 108-129.Google Scholar
- Howe KM, Watson RJ: Nucleotide preferences in sequence-specific recognition of DNA by c-myb protein. Nucleic Acids Res. 1991, 19 (14): 3913-3919. 10.1093/nar/19.14.3913.PubMed CentralPubMedView ArticleGoogle Scholar
- Smarda J, Sugarman J, Glass C, Lipsick J: Retinoic acid receptor alpha suppresses transformation by v-myb. Mol Cell Biol. 1995, 15 (5): 2474-2481.PubMed CentralPubMedView ArticleGoogle Scholar
- Chen JD, Evans RM: A transcriptional co-repressor that interacts with nuclear hormone receptors. Nature. 1995, 377 (6548): 454-457. 10.1038/377454a0.PubMedView ArticleGoogle Scholar
- Heinzel T, Lavinsky RM, Mullen TM, Soderstrom M, Laherty CD, Torchia J, Yang WM, Brard G, Ngo SD, Davie JR, Seto E, Eisenman RN, Rose DW, Glass CK, Rosenfeld MG: A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature. 1997, 387 (6628): 43-48. 10.1038/387043a0.PubMedView ArticleGoogle Scholar
- Wong J, Patterton D, Imhof A, Guschin D, Shi YB, Wolffe AP: Distinct requirements for chromatin assembly in transcriptional repression by thyroid hormone receptor and histone deacetylase. Embo J. 1998, 17 (2): 520-534. 10.1093/emboj/17.2.520.PubMed CentralPubMedView ArticleGoogle Scholar
- Bonamy GM, Guiochon-Mantel A, Allison LA: Cancer promoted by the oncoprotein v-ErbA may be due to subcellular mislocalization of nuclear receptors. Mol Endocrinol. 2005, 19 (5): 1213-1230. 10.1210/me.2004-0204.PubMedView ArticleGoogle Scholar
- Damiola F, Keime C, Gonin-Giraud S, Dazy S, Gandrillon O: Global transcription analysis of immature avian erythrocytic progenitors: from self-renewal to differentiation. Oncogene. 2004, 23 (46): 7628-7643. 10.1038/sj.onc.1208061.PubMedView ArticleGoogle Scholar
- Gandrillon O, Jurdic P, Benchaibi M, Xiao JH, Ghysdael J, Samarut J: Expression of the v-erbA oncogene in chicken embryo fibroblasts stimulates their proliferation in vitro and enhances tumor growth in vivo. Cell. 1987, 49 (5): 687-697. 10.1016/0092-8674(87)90545-9.PubMedView ArticleGoogle Scholar
- Graf T: Myb: a transcriptional activator linking proliferation and differentiation in hematopoietic cells. Curr Opin Genet Dev. 1992, 2 (2): 249-255. 10.1016/S0959-437X(05)80281-3.PubMedView ArticleGoogle Scholar
- Kuehl WM, Bender TP, Stafford J, McClinton D, Segal S, Dmitrovsky E: Expression and function of the c-myb oncogene during hematopoietic differentiation. Curr Top Microbiol Immunol. 1988, 141: 318-323.PubMedGoogle Scholar
- Clarke MF, Kukowska-Latallo JF, Westin E, Smith M, Prochownik EV: Constitutive expression of a c-myb cDNA blocks Friend murine erythroleukemia cell differentiation. Mol Cell Biol. 1988, 8 (2): 884-892.PubMed CentralPubMedView ArticleGoogle Scholar
- Hermann T, Hoffmann B, Piedrafita FJ, Zhang XK, Pfahl M: V-erbA requires auxiliary proteins for dominant negative activity. Oncogene. 1993, 8 (1): 55-65.PubMedGoogle Scholar
- Rascle A, Ferrand N, Gandrillon O, Samarut J: Myb-Ets fusion oncoprotein inhibits thyroid hormone receptor/c-ErbA and retinoic acid receptor functions: a novel mechanism of action for leukemogenic transformation by E26 avian retrovirus. Mol Cell Biol. 1996, 16 (11): 6338-6351.PubMed CentralPubMedView ArticleGoogle Scholar
- Gandrillon O, Samarut J: Role of the different RAR isoforms in controlling the erythrocytic differentiation sequence. Interference with the v-erbA and p135gag-myb-ets nuclear oncogenes. Oncogene. 1998, 16 (5): 563-574. 10.1038/sj.onc.1201550.PubMedView ArticleGoogle Scholar
- Naora H: Involvement of ribosomal proteins in regulating cell growth and apoptosis: translational modulation or recruitment for extraribosomal activity?. Immunol Cell Biol. 1999, 77 (3): 197-205. 10.1046/j.1440-1711.1999.00816.x.PubMedView ArticleGoogle Scholar
- Becquet C, Blachon S, Jeudy B, Boulicaut JF, Gandrillon O: Strong-association-rule mining for large-scale gene-expression data analysis: a case study on human SAGE data. Genome Biol. 2002, 3 (12): RESEARCH0067-10.1186/gb-2002-3-12-research0067.PubMed CentralPubMedView ArticleGoogle Scholar
- Hanafusa H: Rapid transformation of cells by Rous sarcoma virus. Proc Natl Acad Sci U S A. 1969, 63 (2): 318-325. 10.1073/pnas.63.2.318.PubMed CentralPubMedView ArticleGoogle Scholar
- Benchaibi M, Mallet F, Thoraval P, Savatier P, Xiao JH, Verdier G, Samarut J, Nigon V: Avian retroviral vectors derived from avian defective leukemia virus: role of the translational context of the inserted gene on efficiency of the vectors. Virology. 1989, 169 (1): 15-26. 10.1016/0042-6822(89)90036-6.PubMedView ArticleGoogle Scholar
- Beissbarth T, Hyde L, Smyth GK, Job C, Boon WM, Tan SS, Scott HS, Speed TP: Statistical modeling of sequencing errors in SAGE libraries. Bioinformatics. 2004, 20 Suppl 1: I31-I39. 10.1093/bioinformatics/bth924.PubMedView ArticleGoogle Scholar
- Kal AJ, van Zonneveld AJ, Benes V, van den Berg M, Koerkamp MG, Albermann K, Strack N, Ruijter JM, Richter A, Dujon B, Ansorge W, Tabak HF: Dynamics of gene expression revealed by comparison of serial analysis of gene expression transcript profiles from yeast grown on two different carbon sources. Mol Biol Cell. 1999, 10 (6): 1859-1872.PubMed CentralPubMedView ArticleGoogle Scholar
- Benjamini Y, Hochberg Y: Controlling the false discovery rate : a practical and powerful approach to multiple testing. J Roy Stat Soc. 1995, 57 (B): 289-300.Google Scholar
- Dennis G, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA: DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 2003, 4 (5): P3-10.1186/gb-2003-4-5-p3.PubMedView ArticleGoogle Scholar
- Primer3 Software system. [http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi]
- Pfaffl MW: A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001, 29 (9): e45-10.1093/nar/29.9.e45.PubMed CentralPubMedView ArticleGoogle Scholar
- Ensembl ftp site. [ftp://ftp.ensembl.org/pub/]
- Kent WJ: BLAT--the BLAST-like alignment tool. Genome Res. 2002, 12 (4): 656-664. 10.1101/gr.229202. Article published online before March 2002.PubMed CentralPubMedView ArticleGoogle Scholar
- Mitasiunaité I, Boulicaut JF: Looking for monotonicity properties of a similarity constraint on sequences. Proceedings of 2006 ACM SAC'2006, Data Mining track. 2006, ACM Press, 546-552.Google Scholar
- TRANSFAC. [http://www.biobase-international.com/pages/index.php]
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.