Chromobox protein homologue 7 (CBX7) is a chromobox family protein encoding a novel polycomb protein, the expression of which shows a progressive reduction, well related with the malignant grade of the thyroid neoplasias. Indeed, CBX7 protein levels decreased in an increasing percentage of cases going from benign adenomas to papillary, follicular, and anaplastic thyroid carcinomas. To elucidate the function of CBX7 in carcinogenesis, we searched for CBX7 interacting proteins by a proteomic analysis. By this approach, we identified several proteins. Among these proteins, we selected histone deacetylase 2 (HDAC2), which is well known to play a key role in neoplastic cell transformation and down-regulation of E-cadherin expression, the loss of which is a critical event in the epithelial-to-mesenchymal transition. We confirmed by coimmunoprecipitation that CBX7 physically interacts with the HDAC2 protein and is able to inhibit its activity. Then, we showed that both these proteins bind the E-cadherin promoter and that CBX7 up-regulates E-cadherin expression. Consistent with these data, we found a positive statistical correlation between CBX7 and E-cadherin expression in human thyroid carcinomas. Finally, we showed that the expression of CBX7 increases the acetylation status of the histones H3 and H4 on the E-cadherin promoter. Therefore, the ability of CBX7 to positively regulate E-cadherin expression by interacting with HDAC2 and inhibiting its activity on the E-cadherin promoter would account for the correlation between the loss of CBX7 expression and a highly malignant phenotype. [Cancer Res 2009;69(17):7079–87]
Chromobox protein homologue 7 (CBX7) gene encodes a novel polycomb protein of 28.4 kDa and 251 amino acids, which contains a “chromodomain” between amino acids 10 and 46 ( 1, 2). CBX7 is a chromobox family protein and a member of the polycomb repressive complex 1, which, together with the polycomb repressive complex 2, maintains developmental regulatory genes in a silenced state ( 3– 5). Mouse Cbx7 associates with facultative heterochromatin and with the inactive X chromosome, suggesting a role of the Cbx7 protein in the repression of gene transcription ( 6, 7). We have found previously that the CBX7 gene was drastically down-regulated in six thyroid carcinoma cell lines versus normal thyroid cells. Subsequently, the analysis of CBX7 expression in a large number of thyroid carcinoma samples revealed a progressive reduction of CBX7 levels that was well related with the malignant grade of the thyroid neoplasias ( 8). Indeed, it decreased in an increasing percentage of cases going from benign adenomas to papillary, follicular, and anaplastic thyroid carcinomas ( 9, 10). The analysis of rat and mouse models of thyroid carcinogenesis gave rise to very similar results ( 8). More recent results confirmed a correlation between low CBX7 expression and a reduced survival in colon carcinoma patients. 5 Moreover, the association between lack of CBX7 expression and a more aggressive histotype has also been shown by our group in breast, ovary, and prostate carcinomas. 6 Restoration of CBX7 expression in thyroid cancer cells reduced their growth rate, indicating that CBX7 plays a critical role in the regulation of transformed thyroid cell proliferation ( 8).
The aim of the present work has been to elucidate the mechanisms by which the loss of CBX7 is involved in carcinogenesis, attempting to identify the CBX7 protein partners by performing a functional proteomic experiment. Here, we show that CBX7 interacts with histone deacetylase 2 (HDAC2) and inhibits its activity. Moreover, both HDAC2 and CBX7 bind the E-cadherin (CDH1) promoter, and CBX7 contrasts the inhibiting effect of HDAC2 on E-cadherin expression.
Therefore, the ability of CBX7 to positively regulate E-cadherin expression might account for the correlation of the loss of CBX7 expression with a highly malignant phenotype in cancer patients.
Materials and Methods
Cell culture and transfections. TPC1 and NPA (derived from thyroid papillary carcinomas) and PC Cl3, HEK 293, and HeLa cells were grown as described previously ( 8, 11). PC Cl3 cells are differentiated thyroid cells of 18-month-old rat Fischer origin depending on thyrotropine for the growth ( 12). For the inhibition of CBX7 expression, rat Cbx7 small interfering RNA (SI01495795 and SI01495802; Qiagen) and nonsilencing control small interfering RNA (1022076; Qiagen) were transfected using Oligofectamine (Invitrogen) according to the manufacturer's recommendations. CBX7-inducible NPA cells were generated by transfecting NPA cells with the pcDNA6/TR regulatory vector (Invitrogen). Trichostatin A (Sigma) treatments were done for 24 h after transfection.
Plasmids. V5-tagged CBX7 expression plasmid was generated by the insertion of the PCR product CBX7 cDNA into pcDNA-DEST-40 Gateway Vector (Invitrogen). HA-tagged CBX7 expression plasmid was obtained by PCR amplification and subcloned into the pCefl-HA expression vector. CBX7-deletion mutants were cloned in pCefl-HA as well.
Protein extraction, Western blotting, and immunoprecipitation assays. Protein extraction, Western blotting, and coimmunoprecipitation procedures were carried out as reported elsewhere ( 13). The antibodies used for immunoprecipitation and Western blotting were anti-CBX7 ( 8), anti-HA (Roche), anti-V5 (Sigma), anti-pan-cadherin (Sigma), and anti-α-tubulin (Sigma).
GST pull-down experiments. GST fusion proteins were constructed by cloning the human cDNA sequence in a pGEX4T-1 vector (Promega). GST pull-down experiments were carried out as reported elsewhere ( 14).
HDAC activity assay. Cells were transfected with increasing amount of CBX7 expression vector and were used to assay the HDAC activity according to the manufacturer's instructions of Histone Deacetylase Assay Kit (Upstate).
Fresh human thyroid tissue samples. Neoplastic human thyroid tissues and normal adjacent tissue or the normal contralateral thyroid lobe were obtained as described previously ( 8).
RNA extraction, reverse transcription, and PCR analysis. Total RNA isolation and reverse transcription-PCR (RT-PCR) from human tissues were done as described previously ( 8, 15). Each reaction was carried out in duplicate. We used the 2−ΔΔCT method to calculate relative expression levels ( 16, 17). Detailed primer sequences are available as Supplementary Materials and Methods.
Immunohistochemistry. Immunohistochemical analysis were done as described previously ( 8).
Electrophoretic mobility shift assay. For gel shift analysis, nuclear extracts were prepared as described elsewhere ( 18) and electrophoretic mobility shift assay was done as described previously ( 19). The double-strand oligonucleotides covered a region spanning from nucleotide −70 to +54 of the human E-cadherin promoter with respect to the transcription start site (TSS).
Chromatin immunoprecipitation and re-chromatin immunoprecipitation assays. After transfection, chromatin samples were processed for chromatin immunoprecipitation and re-chromatin immunoprecipitation experiments as reported elsewhere ( 19). Samples were subjected to immunoprecipitation with the following specific antibodies: anti-HA (Roche) and anti-HDAC2, anti-H3K4m2, anti-H3K4m3, anti-H3K9me, anti-H3K9m3, and anti-H4K20 (Upstate). The sequences of the used primers are available as Supplementary Materials and Methods.
Transactivation assay. Cells were transiently transfected with the reporter construct in which the luciferase gene was driven by described fragments of E-cadherin promoter ( 20, 21) and normalized with the use of a cotransfected β-galactosidase construct. Luciferase activity was analyzed by Dual-Light System (Applied Biosystems).
DNA extraction and methylation analysis. DNA was prepared using QIAamp DNA Mini Kit (Qiagen) following the instruction manual. PCR primers to analyze E-cadherin promoter were designed by using Methprimer ( 22). The MassCLEAVE biochemistry was done as described previously ( 23). Mass spectra were acquired by using a MassARRAY Compact matrix-assisted laser desorption/ionization—time-of-flight (Sequenom) and spectra methylation ratios were generated by the Epityper software version 1.0 (Sequenom).
Statistical analysis. For the comparison between two groups of experiments, Student's t test was used. The statistical significant difference was considered when P < 0.05. A Pearson correlation coefficient (R2) close to 1 was considered indicative of a significant direct correlation. All experiments were done in triplicate and the data are mean ± SD of three independent experiments.
CBX7 physically interacts with HDAC2 protein. To investigate the mechanisms by which the loss of CBX7 expression correlates with the highly malignant phenotype, we searched for CBX7 interacting proteins by performing a functional proteomic analysis. Therefore, we transiently transfected the HEK 293 cells with a V5-tagged CBX7 expression vector, immunoprecipitated the nuclear protein lysates with anti-V5 antibodies (Supplementary Fig. S1A), and fractionated the immunoprecipitated material on a 12% one-dimensional gel and stained with colloidal Coomassie (Supplementary Fig. S1B). After SDS-PAGE, single components of the immunoprecipitated complexes were analyzed by mass spectrometry ( 24). In Supplementary Table S1, we report some representative CBX7 interacting proteins. Among them, we focused our attention on HDAC2 because of its relevance in tumor biology ( 25). HDACs catalyze the removal of acetyl groups from core histones and, because of their ability to induce local condensation of chromatin, are generally considered repressors of transcription.
To verify the CBX7/HDAC2 interaction in vivo, HEK 293 cells were transiently transfected with the V5-tagged CBX7 expression vector. Protein lysates were immunoprecipitated with either anti-V5 or anti-HDAC2 antibodies and immunoblotted with both anti-V5 and anti-HDAC2 antibodies ( Fig. 1A ; Supplementary Fig. S1C). As shown in Fig. 1A, we detected the association between CBX7 and the endogenous HDAC2 protein, confirming that CBX7 and HDAC2 form complexes in vivo.
To further examine the specificity of this interaction, and to map the regions of CBX7 protein required for the binding to HDAC2, pull-down assays were done incubating total cell extracts deriving from HEK 293 cells, with the CBX7 recombinant protein fused to GST (GST-CBX7) and with two deletion mutants of CBX7: GST-CBX7-CHROMO (1-100 amino acids) and GST-CBX7-NOCHROMO (55-251 amino acids). As shown in Fig. 1B, HDAC2 interacts with both GST-CBX7 and GST-CBX7-NOCHROMO mutants but not with the GST alone or the GST-CBX7-CHROMO mutant. These results clearly show that the CBX7 chromodomain is not required for the interaction between CBX7 and HDCA2.
CBX7 inhibits HDAC activity. To evaluate the effects of the CBX7/HDAC2 interaction on the HDAC2 activity, nuclear extracts from HeLa and NPA cells were prepared and the ability of increasing amount of transfected CBX7 to modulate HDAC activity was tested using a HDAC activity assay. CBX7 expression significantly inhibited HDAC activity in a dose-dependent manner (the percentage of inhibition was 65.15% and 42.55% in HeLa and NPA cells, respectively, after the transfection of 10 μg CBX7 expression vector; Fig. 2A and B ). As positive control for HDAC inhibition, we used 250 mmol/L sodium butyrate, a strong HDAC inhibitory compound ( 26, 27). Nuclear extracts from NPA transfected with CBX7 were immunoprecipitated with anti-HDAC2 antibodies. Then, immunoprecipitated material was tested using a HDAC activity assay. As shown in Fig. 2C, CBX7 specifically inhibited HDAC2 activity.
CBX7 binds to the E-cadherin gene promoter. It has been recently shown that HDAC2 is involved in the repressive complex that silences the E-cadherin gene expression during tumor progression ( 28). E-cadherin is emerging as one of the caretakers of the epithelial phenotype because the loss of its expression has been shown to be a critical event of epithelial-to-mesenchymal transition ( 29, 30). Therefore, to evaluate whether CBX7 was able tobind the E-cadherin promoter in vitro, we performed an electrophoretic mobility shift assay. Nuclear extracts from HEK 293 cells transiently transfected with either V5-tagged CBX7 or empty vectors were incubated with a radiolabeled oligonucleotide corresponding to the E-cadherin promoter. As shown in Fig. 3A , the E-cadherin oligonucleotide forms a specific complex (indicated in the figure as A) with nuclear proteins of cells transfected with V5-tagged CBX7, which was not present in mock-transfected cells (compare lanes 1 and 2). Binding specificity was shown by incubating the nuclear extract with a 100-fold molar excess of unlabeled E-cadherin oligonucleotide (lanes 3 and 4).
Then, HEK 293 and HeLa cells were transiently transfected with HA-tagged CBX7 expression vector, tested by Western blotting for protein expression ( Fig. 3B), crosslinked, and immunoprecipitated with anti-HA or IgG antibodies. Immunoprecipitation of chromatin was subsequently analyzed by semiquantitative PCR using primers spanning the region of the E-cadherin promoter (−300 bp upstream to +40 bp downstream to the TSS). Anti-HA antibodies precipitated this E-cadherin promoter region from HEK 293 and HeLa cells transfected with HA-tagged CBX7 protein. No immunoprecipitation was observed with IgG precipitates, and when primers for the control promoter GAPDH were used ( Fig. 3B), indicating that the binding is specific for the E-cadherin promoter. Similar results were obtained when NPA cells were used (data not shown). These results indicate that CBX7 binds the E-cadherin promoter region in vivo.
Then, to investigate whether the physical interaction between CBX7 and HDAC2 takes place on the human E-cadherin promoter, we performed re-chromatin immunoprecipitation analysis. HEK 293 cells transiently transfected with HA-tagged CBX7 were crosslinked and immunoprecipitated with anti-HDCA2 antibodies. The anti-HDAC2 complexes were released, re-immunoprecipitated with anti-HA antibodies, and then analyzed by PCR. The results shown in Fig. 3C reveal that the antibodies against HA precipitate the E-cadherin promoter after their release from anti-HDAC2, indicating that CBX7 occupies this promoter region together with HDAC2. The reciprocal experiment provided comparable results ( Fig. 3D). Taken together, these results indicate that CBX7 binds the human E-cadherin promoter in vivo and participates in the same DNA-bound complexes that contain HDAC2.
CBX7 positively regulates the E-cadherin promoter. To evaluate the effect of CBX7 expression on E-cadherin transcription, HEK 293 cells were transiently cotransfected with an expression vector encoding CBX7 and with a reporter vector carrying the luciferase gene under the control of the E-cadherin promoter. As shown in Fig. 4A (top), CBX7 increases the transcriptional activity of the E-cadherin promoter in a dose-dependent manner. The same results were obtained on the NPA and TPC1 cell lines (Supplementary Fig. S2A). The treatment of cells with trichostatin A, a potent inhibitor of HDAC activity ( 31), cooperates with CBX7 to induce E-cadherin gene transcription ( Fig. 4A, bottom). These results strongly suggest that CBX7 protein is involved in the E-cadherin gene transcription likely counteracting with the already known inhibitory effect of HDAC2 on this gene promoter.
To identify the region of CBX7 required for E-cadherin promoter activation, we constructed two CBX7 deletion mutants in the expression vector pCefl-HA: pCefl-HA-CBX7-CHROMO (1-100 amino acids) and pCefl-HA-CBX7-NOCHROMO (55-251 amino acids). Transfection of the mutant pCefl-HA-CBX7-CHROMO, containing only the chromodomain, did not induce transcriptional activation of the E-cadherin promoter. Conversely, the mutant lacking the chromodomain, pCefl-HA-CBX7-NOCHROMO, induced a moderate activation of the E-cadherin promoter. Thus, these data indicate that chromodomain is not essential for the CBX7 transcriptional activity on the E-cadherin promoter ( Fig. 4B).
Moreover, to show that the expression of E-cadherin is directly regulated by CBX7, we generated some clones of NPA cells (NPA 4-11 and NPA 5-11) in which CBX7 cDNA was under the control of a tetracycline-regulated promoter. Western blot analysis ( Fig. 4C, top) and quantitative RT-PCR experiment ( Fig. 4C, bottom) show that the expression of CBX7 increases the levels of E-cadherin only after treatment with tetracycline.
To further confirm the role of CBX7 in the modulation of the E-cadherin gene, we evaluated the expression of the E-cadherin gene in the normal rat thyroid cell line PC Cl3 in which the synthesis of Cbx7 was suppressed by RNA interference. The knockdown of the Cbx7 mRNA levels, observed at 48 h after treatment, resulted in the reduction of E-cadherin mRNA levels in comparison with the untreated cells or those treated with the nonsilencing control small interfering RNA ( Fig. 4D).
To identify the regulatory elements essential for the activity of CBX7 on the E-cadherin promoter, we also tested some deletion mutants of the E-cadherin promoter: E-cadherin 601 and E-cadherin 211 ( 20, 21). Activation of E-cadherin promoter by CBX7 protein was maintained on full-length and −601 mutants of E-cadherin promoter. Conversely, the −211 mutant, which lacks a portion of −70/+54 sequence (able to bind CBX7), was not activated by CBX7 protein (Supplementary Fig. S2B).
CBX7 expression results in increased histone acetylation of the E-cadherin promoter. During the last years, chromatin remodeling and histone modifications have emerged as the main mechanisms in the control of gene expression and the connection between DNA methylation and histone deacetylation in the silencing of genes has been established ( 32– 37).
Because we have shown previously that CBX7 (a) interacts with HDAC2 on the human E-cadherin promoter, (b) increases the transcriptional activity of the E-cadherin promoter in a dose-dependent manner, and (c) reduces the activity of HDACs, we hypothesized that the positive effect on E-cadherin activation by CBX7 may be due to its ability to reduce the HDAC activity on the E-cadherin promoter. Therefore, we have evaluated the lysine acetylation of histone tails at the E-cadherin promoter.
HEK 293 cells were transiently transfected with V5-tagged CBX7 expression vector or empty vector. Then, the cells were crosslinked and DNA-chromatin was immunoprecipitated with anti-H3 or anti-H4 acetylated or IgG antibodies. The immunoprecipitated chromatin was subsequently analyzed by quantitative PCR using primers spanning the region of the E-cadherin promoter (−70 bp upstream to +54 bp downstream to the TSS; ref. 38). As shown in Fig. 5A , higher amounts of H3 and H4 acetylated tails were detected in the E-cadherin promoter in the cells transfected with CBX7, with respect to those detected in mock-transfected cells, indicating an increased histone acetylation in CBX7-transfected cells, likely due to the ability of CBX7 to reduce HDAC activity. We also treated the cells, transfected or not with CBX7, with trichostatin A to verify the HDAC activity on H3 and H4 tails. As shown in the same figure, there were higher amounts of H3 and H4 acetylated tails in CBX7-transfected samples treated with trichostatin A than in untreated cells. These results indicate that CBX7 protein regulates the E-cadherin expression by modifying histone acetylation at its promoter, likely reducing HDAC activity.
Increased methylation of H3K4 and decreased methylation of H3K9 and H4K20 in CBX7-transfected cells. Lysine methylation can have different effects depending on which residue is modified: methylation of H3K4 and H3K36 is generally associated with transcribed chromatin; in contrast, methylation of H3K9, H3K27, and H4K20 generally correlates with gene repression ( 38). Therefore, we have evaluated the lysine methylation status of histone tails of E-cadherin promoter in presence or absence of CBX7.
HEK 293 cells were transiently transfected with V5-tagged CBX7 or empty vectors and crosslinked, and DNA-chromatin was immunoprecipitated with anti-H3K4m2, anti-H3K4m3, anti-H3K9m2, anti-H3K9m3, and anti-H4K20m3 or IgG antibodies. The immunoprecipitated chromatin was subjected to PCR with specific primers for the E-cadherin promoter region (−70/+54).
Higher amounts of chromatin immunoprecipitated for H3K4m2 and H3K4m3 were observed in the CBX7-transfected cells compared with the control ones, indicating that this lysine is methylated in a higher proportion in CBX7-transfected cells with respect to that observed in control cells. Conversely, in the case of chromatin immunoprecipitated for the H3K9m2, H3K9m3, and H4K20m3, higher amounts of chromatin were detected in control cells, indicating that these sites were methylated at a higher level in the control cells versus CBX7-transfected cells ( Fig. 5B). These data indicate that CBX7 is able to alter the methylation status of specific lysines of E-cadherin promoter, promoting the transcriptional activity of E-cadherin promoter.
CBX7 and E-cadherin expression levels are correlated in human thyroid carcinomas. E-cadherin down-regulation, due to epigenetic mechanisms, including transcriptional repression, also mediated by HDAC activity, and promoter hypermethylation is a frequent event during human cancer progression ( 39, 40). Because previous experiments showed that CBX7 expression was lost in most advanced thyroid cancers, we hypothesized the down-regulation of E-cadherin as a possible mechanism by which loss of CBX7 is involved in advanced stages of thyroid carcinogenesis. This hypothesis was also supported by recent results showing a role of polycomb repressive complex 1/2 in the regulation of E-cadherin expression ( 41).
Therefore, we analyzed CBX7 and E-cadherin mRNA levels in human thyroid carcinomas of different histotypes ( Fig. 6A ). CBX7 and E-cadherin mRNA levels were drastically reduced in anaplastic thyroid carcinoma, whereas just a weak decrease was observed for both genes in papillary thyroid carcinoma. The epithelial-to-mesenchymal transition in our tumor samples was also confirmed by the increased N-cadherin expression ( Fig. 6A).
We found a positive statistical correlation between CBX7 and E-cadherin expression in human thyroid carcinomas as shown in Fig. 6B. We also compared CBX7 and E-cadherin at protein level, by immunohistochemical analysis, confirming that E-cadherin protein expression parallels that of CBX7 ( Fig. 6C).
Because hypermethylation of the E-cadherin promoter has been postulated to play a critical role in the loss of E-cadherin expression and has been reported previously in other cancers, we decided to investigate the DNA methylation status of E-cadherin promoter ( 42). We analyzed DNA methylation status of 26 CpG sites located in a 368-bp region spanning the E-cadherin gene TSS ( Fig. 6D, inset) on 16 papillary thyroid carcinomas, 15 follicular variants of papillary thyroid carcinomas, 4 anaplastic thyroid carcinomas, and 4 normal thyroid tissue samples. A very low level of methylation was present in both normal and tumor samples, with the exception of only one anaplastic thyroid carcinoma sample in which a high degree of methylation was detected ( Fig. 6D).
Therefore, epigenetic mechanisms, other than hypermethylation of the E-cadherin promoter, have a critical role in the down-regulation of E-cadherin expression.
Our group has reported previously that CBX7 gene is drastically down-regulated in thyroid carcinomas and its expression progressively decreases with malignant grade and neoplastic stage ( 8). These data suggest that the loss of CBX7 expression may be strictly correlated with the acquisition of invasiveness accompanied by the loss of the epithelial features and the gain of a mesenchymal phenotype, a process known as epithelial-to-mesenchymal transition. E-cadherin is a main component of the cell-cell adhesion junctions that plays a principal role in maintaining normal epithelial cell morphology, therefore emerging as one of the caretakers of the epithelial phenotype. In most cancers, E-cadherin down-regulation during neoplastic progression occurs by epigenetic mechanisms, including transcriptional repression, in some cases mediated by HDAC activity ( 40), and hypermethylation of the promoter ( 42). Only in a few cases mutations have been found in the E-cadherin gene leading to the absence or the expression of a nonfunctional protein ( 29). For this reason, we evaluated E-cadherin expression in human thyroid carcinomas of different histotypes: an evident correlation was found between CBX7 and E-cadherin expression levels in human thyroid carcinomas, both being drastically down-regulated in anaplastic thyroid carcinomas in comparison with the normal thyroid tissue. Interestingly, no hypermethylation of E-cadherin promoter was observed in thyroid carcinomas. Therefore, we have hypothesized a role for CBX7 as a transcriptional repressor of E-cadherin. We have shown that CBX7 binds to E-cadherin gene promoter in vitro by electrophoretic mobility shift assay and in vivo by chromatin immunoprecipitation and is able to positively regulate the activity of E-cadherin promoter. At the same time, a proteomic approach, aimed to unravel the mechanism by which the loss of CBX7 expression is involved in cancer progression, has identified HDAC2 among the CBX7 interacting proteins. HDACs regulate the expression and activity of numerous proteins involved in both cancer initiation and cancer progression, inducing a nonpermissive chromatin conformation that prevents the transcription of genes encoding proteins involved in tumorigenesis. HDACs are often overexpressed in many tumors ( 25).
Here, we show that CBX7 physically interacts with HDAC2 protein inhibiting its activity. Chromatin immunoprecipitation shows that both HDAC2 and CBX7 bind the E-cadherin promoter. We also show the ability of CBX7 to positively regulate E-cadherin expression by interacting with HDAC2 and inhibiting its activity on the E-cadherin promoter. Further, we show that, in the presence of the CBX7 protein, there is an increased histone acetylation of the E-cadherin promoter, validating our hypothesis that CBX7 recruits HDCA2 on the E-cadherin promoter. Moreover, we showed modifications in the histone methylation state on the E-cadherin promoter in the cells transfected with CBX7 confirming a relationship between acetylation and DNA methylation ( 36).
It is noteworthy that, among CBX7 interacting proteins, we have also identified an arginine methylation protein that can have a regulator effect on HDACs activity, because a cross-talk between histone acetylation and arginine methylation has also been observed ( 43). In fact, histone deacetylation is a prerequisite for PRMT5-mediated H3 and H4 arginine methylation, whereas histone acetylation enhances H3R17 methylation by PRMT4 ( 44– 46).
In conclusion, here we propose a novel pathway regulating the progression step of carcinogenesis in which the CBX7 protein, the loss of expression of which correlates with a highly malignant phenotype, is a key molecule. Indeed, our results indicate that the loss of CBX7 expression contributes to cancer progression by down-regulating E-cadherin expression because of the lack of its inhibitory effect on HDAC activity on the E-cadherin promoter.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Grant support: Associazione Italiana Ricerca sul Cancro.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Dr. E.R. Fearon for the E-cadherin-luciferase reporter construct and Konstantina Vergadou (Scientific Communication) for editing the text.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
↵5 Pallante et al., submitted for publication.
↵6 Pallante, Terracciano, and Troncone, manuscript in preparation.
- Received April 27, 2009.
- Revision received June 23, 2009.
- Accepted June 25, 2009.
- ©2009 American Association for Cancer Research.