Both cells and primary tumour (Figure 2C). Moreover, in BT-20 cells, which overexpress P-cadherin, the binding in all sites was very strong compared with the one found in MCF-7/AZ breast cancer cells.C/EBPb binding sites are important for CDH3 gene IQ1 biological activity activity and are selectively activated by the different C/ EBPb isoformsIn order to evaluate the importance 25033180 of the aforementioned binding sites to the CDH3 gene activation, as well as the specificity of the different C/EBPb isoforms to the CDH3 promoter, point mutations were introduced in the specific C/EBPb binding sequences. Figure 3A illustrates the CDH3 point mutations and their position 16960-16-0 within the C/EBPb binding sites in relation to the wild-type CDH3 promoter. Interestingly, when MCF-7/AZ cells were transfected with the CDH3 promoter containing point mutations at the binding sites 1 and 4 (CDH3-BS1 and BS4), there was a statistically significant alteration in CDH3 promoter activity related to the wild-type promoter sequence (Figure 3B). In contrast, the activity of the CDH3 promoter was not affected by the mutation introduced at the BS3 site, and only slightly affected by the introduced mutation at the binding site 2 (BS2). These results were mostly confirmed in BT-20 cells, especially for the BS4 mutation, located at the transcription start site region of the CDH3 promoter, which also significantly induced its activity (Figure 3B). Although not significant, the reduction on CDH3 promoter activity observed with the BS1 mutant was also found in BT-20 cells, suggesting that this distal C/EBPb binding site is also important to CDH3 gene transcriptional activation. In addition, the BS2 mutant significantly reduced CDH3 promoter activity in BT-20 cells, showing that this is also a crucial site for the activation of P-cadherin transcription in this model. Finally, we could not find any effect of BS3 mutation in CDH3 promoter activity also in BT-20 
cells, proving that this site is not relevant for its regulation. Since the distinct C/EBPb isoforms have been documented has having different functions in cancer gene activation and in a cellspecific context, we co-transfected LAP1, LAP2 and LIP together with the different mutants of CDH3 promoter in both breast cancer cell lines. The results demonstrated that distal CDH3-BS1 and BS2 are significantly important for the induced promoter activity mediated by all C/EBPb isoforms. In contrast, BS3 did not play a significant role in C/EBPb-mediated CDH3 promoter activity, since mutations in this specific region were not important to impair the activation of CDH3 gene mediated by any of the distinct isoforms. Similar results were observed concerning BS4, which did not reveal to be important for CDH3 promoter activity mediated by LAP1, LAP2 or LIP isoforms. Finally, although not significant, the same trend was observed with BT-20 cells, proving that BS1 and BS2 are most 
likely the binding sites where all C/ EBPb isoforms bind to induce 16574785 P-cadherin transcription in breast cancer.C/EBPb physically interacts with endogenous CDH3 gene promoter in breast cancer cellsSince the three C/EBPb isoforms were able to transactivate the 1.8 Kb CDH3 promoter gene construct (Figure 1D), we decided to evaluate in detail the sequence of this putative regulatory region using distinct bioinformatic tools, which can predict for the binding of specific transcription factors. Four concordant C/ EBPb-putative binding sites were identified within the first 1400 nucleotides. Inter.Both cells and primary tumour (Figure 2C). Moreover, in BT-20 cells, which overexpress P-cadherin, the binding in all sites was very strong compared with the one found in MCF-7/AZ breast cancer cells.C/EBPb binding sites are important for CDH3 gene activity and are selectively activated by the different C/ EBPb isoformsIn order to evaluate the importance 25033180 of the aforementioned binding sites to the CDH3 gene activation, as well as the specificity of the different C/EBPb isoforms to the CDH3 promoter, point mutations were introduced in the specific C/EBPb binding sequences. Figure 3A illustrates the CDH3 point mutations and their position within the C/EBPb binding sites in relation to the wild-type CDH3 promoter. Interestingly, when MCF-7/AZ cells were transfected with the CDH3 promoter containing point mutations at the binding sites 1 and 4 (CDH3-BS1 and BS4), there was a statistically significant alteration in CDH3 promoter activity related to the wild-type promoter sequence (Figure 3B). In contrast, the activity of the CDH3 promoter was not affected by the mutation introduced at the BS3 site, and only slightly affected by the introduced mutation at the binding site 2 (BS2). These results were mostly confirmed in BT-20 cells, especially for the BS4 mutation, located at the transcription start site region of the CDH3 promoter, which also significantly induced its activity (Figure 3B). Although not significant, the reduction on CDH3 promoter activity observed with the BS1 mutant was also found in BT-20 cells, suggesting that this distal C/EBPb binding site is also important to CDH3 gene transcriptional activation. In addition, the BS2 mutant significantly reduced CDH3 promoter activity in BT-20 cells, showing that this is also a crucial site for the activation of P-cadherin transcription in this model. Finally, we could not find any effect of BS3 mutation in CDH3 promoter activity also in BT-20 cells, proving that this site is not relevant for its regulation. Since the distinct C/EBPb isoforms have been documented has having different functions in cancer gene activation and in a cellspecific context, we co-transfected LAP1, LAP2 and LIP together with the different mutants of CDH3 promoter in both breast cancer cell lines. The results demonstrated that distal CDH3-BS1 and BS2 are significantly important for the induced promoter activity mediated by all C/EBPb isoforms. In contrast, BS3 did not play a significant role in C/EBPb-mediated CDH3 promoter activity, since mutations in this specific region were not important to impair the activation of CDH3 gene mediated by any of the distinct isoforms. Similar results were observed concerning BS4, which did not reveal to be important for CDH3 promoter activity mediated by LAP1, LAP2 or LIP isoforms. Finally, although not significant, the same trend was observed with BT-20 cells, proving that BS1 and BS2 are most likely the binding sites where all C/ EBPb isoforms bind to induce 16574785 P-cadherin transcription in breast cancer.C/EBPb physically interacts with endogenous CDH3 gene promoter in breast cancer cellsSince the three C/EBPb isoforms were able to transactivate the 1.8 Kb CDH3 promoter gene construct (Figure 1D), we decided to evaluate in detail the sequence of this putative regulatory region using distinct bioinformatic tools, which can predict for the binding of specific transcription factors. Four concordant C/ EBPb-putative binding sites were identified within the first 1400 nucleotides. Inter.
