- Email: [email protected]
Global histone modification of histone H3 in colorectal cancer and its precursor lesions
Human Pathology (2012) 43, 834–842
www.elsevier.com/locate/humpath
Original contribution
Global histone modification of histone H3 in colorectal cancer and its precursor lesions☆,☆☆ Tadao Nakazawa MD, PhD a,⁎, Tetsuo Kondo MD, PhD a , Defu Ma MD, PhD a , Dongfeng Niu MD a , Kunio Mochizuki MD, PhD a , Tomonori Kawasaki MD, PhD a , Tetsu Yamane MD, PhD a , Hiroshi Iino MD, PhD b , Hideki Fujii MD, PhD b , Ryohei Katoh MD, PhD a a
Department of Pathology, Interdisciplinary Graduate School of Medicine and Engineering, University of Yamanashi, Yamanashi 409-3898, Japan b Department of Surgery, Interdisciplinary Graduate School of Medicine and Engineering, University of Yamanashi, Yamanashi 409-3898, Japan Received 6 April 2011; revised 11 July 2011; accepted 21 July 2011
Keywords: Global histone modification; Histone H3; Immunohistochemistry; Western blotting; Colorectal tumor; Tubular adenoma; Adenocarcinoma
Summary Chromatin remodeling through histone modification is an important mechanism of epigenetic gene dysregulation in human cancers. However, little is known about global alteration of histone status during tumorigenesis and cancer progression. Histone H3 status was examined in benign and malignant colorectal tumors by immunohistochemistry and Western blotting. For immunohistochemical evaluation, 4 anti-histone H3 antibodies, specific to dimethylation at lysine 4 (H3K4me2), acetylation at lysine 9 (H3K9ac), dimethylation at lysine 9 (H3K9me2), and trimethylation at lysine 27 (H3K27me3), were used. On immunohistochemistry, H3K4me2, H3K9ac, and H3K27me3 showed no significant changes between normal and colorectal tumors. On the other hand, the global level of H3K9me2 was distinctly higher in neoplastic cells (adenoma and adenocarcinoma) than in normal glandular cells. In addition, it was significantly higher in adenocarcinoma than in adenoma. Correspondingly, Western blotting confirmed that H3K9me2 expression was significantly higher in adenocarcinomas than in normal colorectal mucosa. No alteration of H3K9me2 was observed with tumor differentiation and with the histological subtypes of colorectal cancers. These results suggest that aberration of the global H3K9me2 level is an important epigenetic event in colorectal tumorigenesis and carcinogenesis involved with gene regulation in neoplastic cells through chromatin remodeling. © 2012 Elsevier Inc. All rights reserved.
1. Introduction ☆
Conflict of interest statement: We declare that we have no conflict of interest. ☆☆ This work was supported by the Ministry of Education, Culture, Sports, Science and Technology, Japan: Grant in-Aid for Young Scientists (21790348) for T.N. ⁎ Corresponding author. E-mail address: [email protected] (T. Nakazawa). 0046-8177/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2011.07.009
Colorectal cancer is one of the most common malignancies in the world and is the major cause of cancer-related death [1]. In addition to environmental and alimentary factors, colorectal cancers are thought to occur via a multistep process, resulting in the accumulation of numerous genetic
Global histone modification of histone H3 in colorectal cancer Table 1
835
Immunohistochemical results of H3K9 in normal and neoplastic colorectal tissues
Histology
n
IHC score (H3K9me2)
IHC score (H3K9ac)
Negative
Low
Moderate
High
Negative
Low
Moderate
High
Normal colon mocusa Tubular adenoma Adenocarcinoma
85 25 60
0 0 0
57 (67.1%) 9 (36%) 0
21 (24.7%) 10 (40%) 11 (18.3%)
7 (8.2%) ⁎⁎ , ⁎⁎⁎ 6 (24% ) ⁎ 49 (81.7%)
0 0 0
0 0 0
23 (27.1%) 9 (36%) 17 (28.3%)
62 (72.9%) 16 (64%) 43 (71.7%)
⁎ P b .001 (adenoma vs adenocarcinoma). ⁎⁎ P = .055 (normal vs adenoma). ⁎⁎⁎ P b .001 (normal vs adenocarcinoma).
and epigenetic alterations, including activation of oncogenes and the inactivation of tumor suppressor genes [2-4]. Epigenetic events include DNA methylation and histone modifications. Histone protein, around which DNA is wrapped, can be chemically modified in residues by the addition of acetyl, methyl, phospholyl, or other groups. Histone modification is mainly accounted for by acetylation and methylation of histone core tails, which mostly occurs at lysine or arginine residues of NH2 termini. Histone modification plays an important role in reversible regulation of chromatin dynamics [5]. In addition, it is known that histone modification acts as an activator or suppressor of gene transcription through chromatin remodeling [6]. Generally, condensed chromatin (heterochromatin) is associated with gene inactivation, whereas sparse chromatin (euchromatin) promotes gene transcription. This depends on the site of modified residues and the type of modification (ie, acetylation, methylation, phosphorylation, and sumoylation), that acts on gene transcription [7]. So far, interest in epigenetic involvement in human epithelial cancers has focused on DNA hypermethylation of specific genes, resulting in carcinogenesis by silencing tumor suppressor genes [8]. It has also been shown that histone modification and DNA methylation are closely related to each other [9]. The correlation between the methylation status of CpG islands in the specific genes and histone modification has been investigated mainly using the chromatin immunoprecipitation (ChIP) assay [10,11]. Recently, global histone modification levels have been intensively studied in cancers of various organs using immunohistochemistry [12-21]. In these reports, global alteration of histone modification in certain cancers has been closely associated with patient prognosis and tumor aggressiveness. With regard to colorectal cancers, there has been only one report investigating global histone status and its implications in tumorigenesis and/or carcinogenesis of the neoplasms [21]. In the current study, to clarify the epigenetic environments in colorectal cancers and its precursor lesions, the global histone H3 status was studied in individual cell nuclei of 60 adenocarcinomas, 25 adenomas, 4 adenocarcinomas with adenoma component, and the adjacent nonneoplastic mucosa by immunohistochemistry with 4 specific histone markers (H3K4me2, H3K9ac, H3K9me2, and H3K27me3), along with Western blotting with one histone marker (H3K9me2).
2. Material and methods 2.1. Human colonic tissues A total of 56 surgically and 33 endoscopically resected specimens of colorectal tumors from 89 Japanese patients was studied. Specimens were collected from the pathological files of Yamanashi University Hospital. Informed consent had been obtained before resection in all patients. The materials consisted of 25 tubular adenomas and 60 adenocarcinomas, including 14 well-differentiated adenocarcinomas, 18 moderately differentiated adenocarcinomas, 14 poorly differentiated adenocarcinomas, 14 mucinous adenocarcinomas, and 4 adenocarcinomas accompanied by an adjacent tubular adenoma component. Adenocarcinomas limited to the mucosa were excluded, and the adenocarcinomas showing invasion to layers deeper than the lamina propria were selected. For these specimens, pathological diagnoses were
Fig. 1 The results of Western blotting for H3K9me2. Pairs of lanes (1, 5), (2, 6), (3, 7), and (4,8) are the samples obtained from the same patients with moderately differentiated adenocarcinoma, respectively. H3K9me2 expression is up-regulated in moderately differentiated adenocarcinomas (lanes 1-4) compared with individual-matched normal tissues (lanes 5-8), as shown by the densitometric assessments below (P = .0005). Columns, mean of densitometric values of normal colon mucosa (n = 4) or colon carcinomas (n = 4); bars, SE.
836
T. Nakazawa et al.
Fig. 2 Representative photomicrograph and results of immunohistochemical staining for H3K9me2 in adenocarcinoma with adenoma component. A, The adenocarcinoma component shows submucosal invasion (left side) and the adenoma component is seen continuously (right side). B, In the adenoma component with low-grade dysplasia, weak immunoreactivity for H3K9me2 is observed in the tumor cell nuclei. C, The immunoreactivity for H3K9me2 increases in the adenoma component showing severe nuclear overlapping and swelling and loss of polarity (high-grade dysplasia). D, Diffuse and strong immunoreactivity for H3K9me2 is seen in the adenocarcinoma component.
made based on The World Health Organization Classification of Tumours, Pathology and Genetics of Tumours of the Digestive System [22]. All cases were selected for the immunohistochemical study on the basis of availability of routinely processed, paraffin-embedded, formalin-fixed tissue blocks. Four pairs of fresh tumor and normal colorectal tissues were obtained from the patients with moderately differentiated adenocarcinoma for Western blotting. The fresh tissue was immediately frozen with liquid nitrogen and stored at −80°C immediately after surgical resection.
2.2. Immunohistochemistry Four-micrometer-thick sections were cut from paraffin blocks. After deparaffinization, they were steamed for 20 minutes in sodium citrate buffer (diluted to 1× from 10× heatinduced epitope retrieval buffer). The sections were incubated with primary rabbit polyclonal antibodies (Cell Signaling Technology, Boston, MA) at room temperature, for dimethyl histone H3 Lys9 (H3K9me2) (catalogue number 9753, 1:200 dilution), acetyl histone H3 Lys9 (H3K9ac) (catalogue number
9671, 1:100 dilution), dimethyl histone H3 Lys4 (H3K4me2) (catalogue number 9725, 1:1500 dilution), and trimethyl histone H3 Lys27 (H3K27me3) (catalogue number 9733, 1:50 dilution) for 90 minutes. They were then incubated with a goat antirabbit biotinylated secondary antibody conjugated with avidin-biotin peroxidase complex for another 90 minutes. Finally, specific immunostaining was detected with 3′,3′diaminobenzidine tetrahydrochloride (Sigma, St Louis, MO) and counterstained with hematoxylin. Lymph nodes in which the absence of metastatic tumor was confirmed were stained as positive controls and concomitantly all sections without the primary antibodies as negative controls. The stained sections were examined by 2 pathologists (T.N. and T.K.), who agreed in all cases.
2.3. Evaluation of immunohistochemistry For each case, at least 500 cells were counted in both tumor and adjacent normal colonic mucosa. Only the nuclear immunopositivity of epithelial components was estimated. The percentage of immunopositive cells and the intensities of
Global histone modification of histone H3 in colorectal cancer H3K9me2, H3K9ac, H3K4me2, and H3K27me3 were analyzed. The degree of positive cells was semiquantitatively evaluated as follows: first, scoring according to intensity (negative, 0; weak, 1; intermediate, 2; strong, 3); then, scoring according to the percentage of positive cells (0%, 0; 1%-10%, 1; 11%-50%, 2; N50%, 3). Finally, each sample was divided into 4 groups depending on the total score (frequency of immunopositive cells plus staining intensity); total scores of 0,
837
1 to 3, 4, and 5 to 6 were classified as negative, low expression, moderate expression, and high expression, respectively.
2.4. Western blotting Fresh colonic tissue was homogenized using polytron homogenizer in radioimmunoprecipitation assay buffer with proteinase inhibitors (1× PBS, 1% Nonidet P-40, 0.5%
Fig. 3 Representative results of immunohistochemical staining for H3K9me2. A, In normal colorectal mucosa, a small number of positive cells is observed in the nuclei of the nonneoplastic epithelium. B, Positive cells are focally observed in tubular adenoma. Most of the cancer cells demonstrate strong positivity in well (C), moderately (D), and poorly (E) differentiated adenocarcinoma and mucinous adenocarcinoma (F).
838 Table 2
T. Nakazawa et al. Immunohistochemical results of H3K9me2 in colorectal adenocarcinomas
Differentiation and histological subtype
n
IHC score (H3K9me2) Negative
Low
Moderate
High
Well-differentiated adenocarcinoma Moderately differentiated adenocarcinoma Poorly differentiated adenocarcinoma Mucinous adenocarcinoma
14 18 14 14
0 0 0 0
0 0 0 0
2 2 3 4
12 16 11 10
sodium deoxycholate, 0.1% SDS, 1 mmol/L phenylmethylsulfonyl fluoride, 12 μg/mL aprotinin, and 1 mmol/L sodium ovarthovanadate). Equal amounts of protein (40 μg) solubilized in sample buffer were separated on 10% SDS polyacrylamide gels and transferred electrophoretically to a polyvinylidene difluoride membrane. The membrane was blocked in TBS containing 0.5% Tween 20 plus 5% nonfat dried milk for 1 hour at room temperature and probed with a primary antibody (primary antisera: dimethyl histone H3 Lys9, 1:1000; Cell Signaling Technology; anti-actin, 1:1000; Sigma) at 4°C overnight. The membrane was washed thrice for 10 minutes in TBS containing 0.5% Tween 20 and incubated with horseradish peroxidase–conjugated anti-rabbit secondary antibody (1:2000; Santa Cruz Biotechnology, Santa Cruz, CA) for 1 hour at room temperature. Targeted proteins were visualized using an enhanced detection system (Amersham, Buckinghamshire, UK).
2.5. Statistical analysis Statistical analysis was carried out using χ 2 tests, comparing the frequencies of cases that had high expression of H3K9me2 by immunohistochemistry (normal vs adenoma, normal vs adenocarcinoma, and adenoma vs adenocarcinoma) (Table 1). Student t test was used to compare the Western blots of H3K9me2 protein in paired colon carcinoma and normal colon mucosa (Fig. 1). Statistic significance was set at P b .05. Data analyses were carried out using SPSS version 11.0 for Windows (SPSS, Tokyo, Japan).
3. Results 3.1. H3K9me2 levels in normal, colorectal cancer, and its precursor lesions H3K9me2 expression in 4 moderately differentiated adenocarcinomas was compared with that in the normal colorectal tissue sampled from the same patients by Western blotting. After loading an equal amount of protein, parallel actin immunoblotting was performed, and signal quantification was performed by densitometric scanning. The results showed that H3K9me2 protein expression was significantly up-regulated in all of the analyzed adenocarcinomas compared with individual-matched normal colorectal tissues (P = .0005) (Fig. 1).
(14.3%) (11.1%) (21.4%) (28.6%)
(85.7%) (88.9%) (78.6%) (71.4%)
In the tubular adenomas, the immunopositivity for H3K9me2 gradually increased by the degree of nuclear overlapping and swelling and by loss of polarity. A similar pattern was also observed in all 4 cases having adenocarcinoma with adenoma component (Fig. 2). The H3K9me2 immunoreactivity showed a tendency to be higher in the adenomas with high-grade dysplasia (Fig. 2C) than in those with low-grade dysplasia (Fig. 2B). As seen in the tumors consisting of adenocarcinoma alone, diffuse immunopositivity for H3K9me2 was evident in the adenocarcinoma components (Fig. 2D). Fig. 3 shows the immunohistochemical results for H3K9me2 in the representative sections of the normal colonic mucosa and the colorectal tumors. The cell nuclei completely lacked immunoreactivity for H3K9me2 in the negative controls. The immunohistochemical scores for H3K9me2 are shown in Table 1. In the positive control sections, most of the cell nuclei showed low to moderate expression. In 85 normal colorectal tissue samples, there were no negative cases. More than half of the cases (57/85, 67.1%) had low expression, 21 (24.7%) of 85 cases had moderate expression, and 7 (8.2%) of 85 cases had high expression. In 25 tubular adenomas, moderate expression was the most frequently observed (10/25 cases, 40%), whereas 9 (36%) and 6 (24%) of the 25 tubular adenomas had low and high expression, respectively. In the adenocarcinomas, most cases (49/60, 81.7%) had high expression, and 11 (18.3%) of 60 cases had moderate expression. The frequency of cases with high expression was significantly higher in the adenocarcinomas than in normal mucosa (P b .001). The adenocarcinomas showed a significantly higher proportion of high expression than the adenomas (P b .001). H3K9me2 expression tended to be higher in the adenomas than in the normal mucosa, but the difference was not significant (P = .055). The immunohistochemical results for H3K9me2 in well, moderately, and poorly differentiated adenocarcinomas and mucinous adenocarcinoma are shown in Fig. 1C to F. The immunohistochemical scores of H3K9me2 in the adenocarcinomas by differentiation and histological subtypes are shown in Table 2. In the adenocarcinomas, 12 (85.7%), 16 (88.9%), and 11 (78.6%) cases had high expression in well, moderately differentiated, and poorly differentiated types, respectively. Less frequently, moderate expression was demonstrated in 2 well (14.3%), 2 moderately (11.1%), and 3 poorly (21.4%)
Global histone modification of histone H3 in colorectal cancer differentiated adenocarcinomas. Of 14 mucinous adenocarcinomas, 4 had moderate expression (28.6%), and 10 showed high expression (71.4%). In the adenocarcinomas, H3K9me2 was highly expressed regardless of differentiation and histological subtype.
839
3.2. H3K9ac, H3K4me2, and H3K27me3 levels in normal, colorectal cancer, and its precursor lesions Fig. 4 shows the immunohistochemical results for H3K9ac, H3K4me2, and H3K27me3 in normal colorectal
Fig. 4 Representative immunohistochemical results for H3K9ac, H3K4me2, and H3K27me3 in the normal and neoplastic nuclei. A,B, Representative immunohistochemical results for H3K9ac. Diffuse immunopositivity is observed in normal mucosa (A) and moderately differentiated adenocarcinoma (B). C,D, Representative immunohistochemical results for H3K4me2. Positive cells were focally observed with mild to moderate intensity in normal colorectal mucosa (C) and moderately differentiated adenocarcinoma (D). E,F, Representative immunohistochemical results for H3K27me3. Immunopositivity is diffusely observed in normal colorectal mucosa (E) and moderately differentiated adenocarcinoma (F).
840 Table 3
T. Nakazawa et al. Immunohistochemical results of H3K4me2 and H3K27me3 in normal and neoplastic colorectal tissues
Histology
n
IHC score (H3K4me2)
IHC score (H3K27me3)
Negative
Low
Moderate
High
Negative
Low
Moderate
High
Normal colon mocusa Tubular adenoma Adenocarcinoma
85 25 60
0 0 0
54 (63.5%) 15 (60%) 37 (61.7%)
28 (32.9%) 9 (36%) 20 (33.3%)
3 (3.5%) 1 (4%) 3 (5%)
0 0 0
0 0 0
30 (35.3%) 6 (24%) 23 (38.3%)
55 (64.7%) 19 (76%) 37 (61.7%)
mucosa, cancer, and its precursor lesions. The cell nuclei showed no immunoreactivity for H3K9ac, H3K4me2, and H3K27me3 in the negative controls. The immunohistochemical scores for H3K9ac in normal colorectal mucosa are shown in Table 1; moderate expression was observed in 23 (27.1%) of 85, and high expression was seen in 62 (72.9%) of 85. In the positive controls, most of the cell nuclei showed moderate expression. Negativity and low expression were found in none of the normal glandular cell nuclei. Similar patterns were observed both in tubular adenomas and adenocarcinomas. There was no significant difference among these 3 groups. Similarly, no significant differences in the scores of H3K4me2 and H3K27me3 were seen between the cell nuclei of the normal and the neoplastic colorectal tissue (Table 3). The expression levels were equivalent to those of the positive controls. There were no significant differences in H3K9ac, H3K4me2, and H3K27me3 scores among well, moderately, and poorly differentiated adenocarcinomas, mucinous adenocarcinoma, and adenocarcinoma with adenoma component (data not shown).
4. Discussion Histone modification and DNA methylation are regarded as important epigenetic events that regulate gene transcription [23]. Recently, there have been several reports investigating global alteration of histone modification in human epithelial cancers by immunohistochemistry using specific histone markers, but these immunohistochemical analyses do not provide a correlation to specific gene activities. Nevertheless, they allow us to visualize the global status of histone modification at a cellular level. Previous reports studying the global pattern of histone modification focused on the clinical prognosis and suggested that global aberration of histone modification could closely relate to patient prognosis and serve as an accurate indicator [12-21]. However, little has been said about how the global state of histone modification can be changed at a cellular level in tumorigenic and/or carcinogenic processes in human cancers. Consequently, to clarify epigenetic events in colorectal cancer and its precursor lesions, histone H3 status was examined using 4
specific histone markers by comparing neoplastic cells to normal cells. To the best of our knowledge, there has been only one report studying global histone modification in colorectal neoplasms. Ashktorab et al [21] reported that global levels of H4K12ac and H3K18ac increased in adenocarcinomas in comparison with those in normal tissue and adenomas using immunohistochemistry. They also demonstrated that HDAC2 and H4K12ac expressions in adenocarcinoma were higher than in adenoma, implying that these epigenetic changes have a role in the progression from adenoma to adenocarcinoma. In the present study, the global level of H3K9me2 expression was higher in neoplastic cells than in normal glandular cells on immunohistochemistry. Interestingly, H3K9me2 expression was up-regulated in the adenomas, which are recognized as precancerous lesions, with highgrade dysplasia compared with those with low-grade dysplasia. In addition, H3K9me2 expression was also significantly increased in the nuclei of adenocarcinoma cells compared with that in the nuclei of adenoma cells. These results may suggest that the increased H3K9me2 expression is associated with progression (adenoma to adenocarcinoma). It seems preferable to separate adenoma with high-grade dysplasia from that with low-grade dysplasia and adenoma with high-grade adenoma from well-differentiated adenocarcinoma. However, it is actually difficult to discriminate between these tumors strictly. In the present study, it was confirmed that the amount of H3K9me2 protein was greater in the adenocarcinoma tissue than in the normal colorectal tissues by immunohistochemistry and Western blot. Seligson et al [17] reported that a low level of H3K9me2 is significantly associated with aggressive phenotypes of prostatic and renal cancers and that this can independently predict poor survival. However, in the current study, the cellular global H3K9me2 levels were not altered regardless of the tumor differentiation and the histological subtypes of colorectal adenocarcinomas. It has been generally recognized that poorly differentiated adenocarcinoma and mucinous adenocarcinoma are positively correlated with poor patient clinical outcomes. Therefore, the global H3K9me2 state seems to be less relevant in more aggressive colorectal cancers leading to poor clinical outcomes. Although it is difficult to interpret the effects, these findings suggest that global alteration of H3K9me2
Global histone modification of histone H3 in colorectal cancer expression may correlate with genetic alteration and instability involving tumorigenesis and/or carcinogenesis in colorectal adenoma and adenocarcinoma, in conjunction with other epigenetic events. Previous reports have demonstrated that methylation of histone H3K9 is associated with gene repression through sparse chromatin (heterochromatin) [7,24]. From this aspect, it is conceivable that the increased H3K9me2 level may repress transcriptional activity of certain genes that function as tumor suppressors and/or carcinostasis promoters in colorectal tumors. Conversely, it can also raise the possibility that increased H3K9me2 inactivates genes that suppress tumor cell proliferation after tumorigenesis and/or carcinogenesis as negative feedback. Quantitative analyses for the correlation with specific genes using other molecular methods (ie, ChIP assay in combination with real-time PCR) will be necessary to determine which hypotheses provide a correct explanation. With respect to histone H3 modifications at other residues, the global states of H3K9ac, H3K4me2, and H3K27me3 have been investigated with specific histone markers in a variety of cancers other than colorectal cancers. In more detail, low H3K9ac has been reported to be related to a poor prognosis in lung and breast cancer patients [12,13]. On the other hand, a high level of H3K9ac has been associated with the aggressive type of prostatic cancers [14]. Low H3K4me2 was observed in prostatic cancers with a high risk of recurrence [17]. In kidney cancers, decreased expression of H3K4me1-3 was correlated with shorter survival [19]. Low expression of H3K27me3 was positively related to better patient prognosis in the early stage of esophageal squamous cell carcinomas [20]. In the present study, there were no significant differences in the expression levels of H3K9ac, H3K4me2, and H3K27me3 between normal and neoplastic cell nuclei in the colorectal tumors. In addition, these levels were stable, irrespective of the degree of differentiation and the histological subtypes of colorectal adenocarcinomas (data not shown). As previously reported, some patterns of global histone modification are interestingly common among cancers of different organ origins [17]. Nevertheless, the present results show that a similar pattern could not be observed in the colorectal cancers for the previously investigated patterns of histone modification (H3K9ac, H3K4me2, and H3K27me3) in the other cancers. In summary, global H3K9me2 expression was verified to be up-regulated, whereas there were no differences in H3K9ac, H3K4me2, and H3K27me3 levels in colorectal tumors. The H3K9me2 expression levels were significantly higher in adenocarcinoma cells than in adenoma cells. Among adenocarcinomas, H3K9me2 expression was unchanged by differentiation and histological subtype. Aberrant H3K9me2 expression may play a role in tumorigenesis and progression (adenoma-carcinoma sequence), rather than in transformation to more aggressive pheno-
841
types of colorectal cancers, although the correlation of this epigenetic event to specific genes still remains unclear. From a practical perspective, diffuse H3K9me2 immunopositivity can serve as a useful tool for pathological diagnosis in differentiating between tubular adenoma and adenocarcinoma. Furthermore, it raises the possibility that H3K9me2 is a potential pharmacological therapeutic target for colorectal cancers.
Acknowledgments The authors would like to thank Miyuki Ito, Mikiko Yoda, and especially Yoshihito Koshimizu for their excellent technical assistance and Kayoko Kono for executive assistance.
References [1] Imamura Y, Sobue T. Cancer statistics digest. Mortality trend of colon, rectal, liver, “gallbladder and biliary tract” and pancreas cancer in Japan by birth cohort. Jpn J Clin Oncol 2004;34:491-3. [2] Ashktorab H, Smoot DT, Carethers JM, et al. High incidence of microsatellite instability in colorectal cancer from African Americans. Clin Cancer Res 2003;9:1112-7. [3] Ashktorab H, Smoot DT, Farzanmehr H, et al. Clinicopathological features and microsatellite instability (MSI) in colorectal cancers from African Americans. Int J Cancer 2005;116:914-9. [4] Carethers JM. Racial and ethnic factors in the genetic pathogenesis of colorectal cancer. J Assoc Acad Minor Phys 1999;10:59-67. [5] Marusige K. Activation of chromatin by acetylation of histone side chains. Proc Natl Acad Sci USA 1976;73:3937-41. [6] Vidali G, Ferrari N, Pfeffer U. Histone acetylation: a step in gene activation. Adv Exp Med Biol 1988;231:583-96. [7] Jenuwein T, Allis CD. Translating the histone code. Science 2001;293: 1074-80. [8] Esteller M. Epigenetic gene silencing in cancer: the DNA hypermethylene. Hum Mol Genet 2007;16:R50-9. [9] Kurdistani SK. Histone modifications as markers of cancer prognosis: a cellular view. Br J Cancer 2007;97:1-5. [10] Kondo Y, Shen L, Issa JP. Critical role of histone methylation in tumor suppressor gene silencing in colorectal cancer. Mol Cell Biol 2003;23: 206-15. [11] Fujii S, Luo RZ, Yuan J, et al. Reactivation of the silenced and imprinted alleles of ARHI is associated with increased histone H3 acetylation and decreased histone H3 lysine 9 methylation. Hum Mol Genet 2003;12:1791-800. [12] Elsheikh SE, Green AR, Rakha EA, et al. Global histone modifications in breast cancer correlate with tumor phenotypes, prognostic factors, and patient outcome. Cancer Res 2009;69:3802-9. [13] Barlési F, Giaccone G, Gallegos-Ruiz MI, et al. Global histone modifications predict prognosis of resected non–small-cell-lung cancer. J Clin Oncol 2007;25:4358-64. [14] Sligson DB, Horvath S, Shi T, et al. Global histone modification patterns predict risk of prostate cancer recurrence. Nature 2005;435: 1262-6. [15] Ellinger J, Kahl P, von der Gathen J, et al. Global levels of histone modifications predict prostate cancer recurrence. Prostate 2010;70: 61-9.
842 [16] Park YS, Jin MY, Kim YJ, Yook JH, Kim BS, Jang SJ. The global histone modification pattern correlates with cancer recurrence and overall survival in gastric adenocarcinomas. Ann Surg Oncol 2008;15: 1968-76. [17] Seligson DB, Horvath S, McBrian MA, et al. Global levels of histone modifications predict prognosis in different cancers. Am J Pathol 2009;174:1619-28. [18] Minardi D, Lucarini G, Filosa A, et al. Prognostic role of global DNA-methylation and histone acetylation in pT1a clear cell renal carcinoma in partial nephrectomy specimens. J Cell Mol Med 2009; 13:2115-21. [19] Ellinger J, Kahl P, Mertens C, et al. Prognostic relevance of global histone H3 lysine 4 (H3K4) methylation in renal cell carcinoma. Int J Cancer 2010;127:2360-6.
T. Nakazawa et al. [20] Tzao C, Tung HJ, Jin JS, et al. Prognostic significance of global histone modifications in resected squamous cell carcinoma of the esophagus. Mod Pathol 2009;22:252-60. [21] Ashktorab H, Belgrave K, Hosseinkhah F, et al. Global histone H4 acetylation and HDAC2 expression in colon adenoma and carcinoma. Dig Dig Sci 2009;54:2109-17. [22] Bosman FT, Carneiro F, editors. Pathology and genetics of tumours of the digestive system. The World Health Organization Classification of Tumours. Lyon: IARC Press; 2010. p. 131-82. [23] Gronbaek K, Hother C, Jones PA. Epigenetic changes in cancer. APMIS 2007;115:1039-59. [24] Rea S, Eisenhaber F, O'Carroll D, et al. Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 2000; 406:593-9.
www.elsevier.com/locate/humpath
Original contribution
Global histone modification of histone H3 in colorectal cancer and its precursor lesions☆,☆☆ Tadao Nakazawa MD, PhD a,⁎, Tetsuo Kondo MD, PhD a , Defu Ma MD, PhD a , Dongfeng Niu MD a , Kunio Mochizuki MD, PhD a , Tomonori Kawasaki MD, PhD a , Tetsu Yamane MD, PhD a , Hiroshi Iino MD, PhD b , Hideki Fujii MD, PhD b , Ryohei Katoh MD, PhD a a
Department of Pathology, Interdisciplinary Graduate School of Medicine and Engineering, University of Yamanashi, Yamanashi 409-3898, Japan b Department of Surgery, Interdisciplinary Graduate School of Medicine and Engineering, University of Yamanashi, Yamanashi 409-3898, Japan Received 6 April 2011; revised 11 July 2011; accepted 21 July 2011
Keywords: Global histone modification; Histone H3; Immunohistochemistry; Western blotting; Colorectal tumor; Tubular adenoma; Adenocarcinoma
Summary Chromatin remodeling through histone modification is an important mechanism of epigenetic gene dysregulation in human cancers. However, little is known about global alteration of histone status during tumorigenesis and cancer progression. Histone H3 status was examined in benign and malignant colorectal tumors by immunohistochemistry and Western blotting. For immunohistochemical evaluation, 4 anti-histone H3 antibodies, specific to dimethylation at lysine 4 (H3K4me2), acetylation at lysine 9 (H3K9ac), dimethylation at lysine 9 (H3K9me2), and trimethylation at lysine 27 (H3K27me3), were used. On immunohistochemistry, H3K4me2, H3K9ac, and H3K27me3 showed no significant changes between normal and colorectal tumors. On the other hand, the global level of H3K9me2 was distinctly higher in neoplastic cells (adenoma and adenocarcinoma) than in normal glandular cells. In addition, it was significantly higher in adenocarcinoma than in adenoma. Correspondingly, Western blotting confirmed that H3K9me2 expression was significantly higher in adenocarcinomas than in normal colorectal mucosa. No alteration of H3K9me2 was observed with tumor differentiation and with the histological subtypes of colorectal cancers. These results suggest that aberration of the global H3K9me2 level is an important epigenetic event in colorectal tumorigenesis and carcinogenesis involved with gene regulation in neoplastic cells through chromatin remodeling. © 2012 Elsevier Inc. All rights reserved.
1. Introduction ☆
Conflict of interest statement: We declare that we have no conflict of interest. ☆☆ This work was supported by the Ministry of Education, Culture, Sports, Science and Technology, Japan: Grant in-Aid for Young Scientists (21790348) for T.N. ⁎ Corresponding author. E-mail address: [email protected] (T. Nakazawa). 0046-8177/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2011.07.009
Colorectal cancer is one of the most common malignancies in the world and is the major cause of cancer-related death [1]. In addition to environmental and alimentary factors, colorectal cancers are thought to occur via a multistep process, resulting in the accumulation of numerous genetic
Global histone modification of histone H3 in colorectal cancer Table 1
835
Immunohistochemical results of H3K9 in normal and neoplastic colorectal tissues
Histology
n
IHC score (H3K9me2)
IHC score (H3K9ac)
Negative
Low
Moderate
High
Negative
Low
Moderate
High
Normal colon mocusa Tubular adenoma Adenocarcinoma
85 25 60
0 0 0
57 (67.1%) 9 (36%) 0
21 (24.7%) 10 (40%) 11 (18.3%)
7 (8.2%) ⁎⁎ , ⁎⁎⁎ 6 (24% ) ⁎ 49 (81.7%)
0 0 0
0 0 0
23 (27.1%) 9 (36%) 17 (28.3%)
62 (72.9%) 16 (64%) 43 (71.7%)
⁎ P b .001 (adenoma vs adenocarcinoma). ⁎⁎ P = .055 (normal vs adenoma). ⁎⁎⁎ P b .001 (normal vs adenocarcinoma).
and epigenetic alterations, including activation of oncogenes and the inactivation of tumor suppressor genes [2-4]. Epigenetic events include DNA methylation and histone modifications. Histone protein, around which DNA is wrapped, can be chemically modified in residues by the addition of acetyl, methyl, phospholyl, or other groups. Histone modification is mainly accounted for by acetylation and methylation of histone core tails, which mostly occurs at lysine or arginine residues of NH2 termini. Histone modification plays an important role in reversible regulation of chromatin dynamics [5]. In addition, it is known that histone modification acts as an activator or suppressor of gene transcription through chromatin remodeling [6]. Generally, condensed chromatin (heterochromatin) is associated with gene inactivation, whereas sparse chromatin (euchromatin) promotes gene transcription. This depends on the site of modified residues and the type of modification (ie, acetylation, methylation, phosphorylation, and sumoylation), that acts on gene transcription [7]. So far, interest in epigenetic involvement in human epithelial cancers has focused on DNA hypermethylation of specific genes, resulting in carcinogenesis by silencing tumor suppressor genes [8]. It has also been shown that histone modification and DNA methylation are closely related to each other [9]. The correlation between the methylation status of CpG islands in the specific genes and histone modification has been investigated mainly using the chromatin immunoprecipitation (ChIP) assay [10,11]. Recently, global histone modification levels have been intensively studied in cancers of various organs using immunohistochemistry [12-21]. In these reports, global alteration of histone modification in certain cancers has been closely associated with patient prognosis and tumor aggressiveness. With regard to colorectal cancers, there has been only one report investigating global histone status and its implications in tumorigenesis and/or carcinogenesis of the neoplasms [21]. In the current study, to clarify the epigenetic environments in colorectal cancers and its precursor lesions, the global histone H3 status was studied in individual cell nuclei of 60 adenocarcinomas, 25 adenomas, 4 adenocarcinomas with adenoma component, and the adjacent nonneoplastic mucosa by immunohistochemistry with 4 specific histone markers (H3K4me2, H3K9ac, H3K9me2, and H3K27me3), along with Western blotting with one histone marker (H3K9me2).
2. Material and methods 2.1. Human colonic tissues A total of 56 surgically and 33 endoscopically resected specimens of colorectal tumors from 89 Japanese patients was studied. Specimens were collected from the pathological files of Yamanashi University Hospital. Informed consent had been obtained before resection in all patients. The materials consisted of 25 tubular adenomas and 60 adenocarcinomas, including 14 well-differentiated adenocarcinomas, 18 moderately differentiated adenocarcinomas, 14 poorly differentiated adenocarcinomas, 14 mucinous adenocarcinomas, and 4 adenocarcinomas accompanied by an adjacent tubular adenoma component. Adenocarcinomas limited to the mucosa were excluded, and the adenocarcinomas showing invasion to layers deeper than the lamina propria were selected. For these specimens, pathological diagnoses were
Fig. 1 The results of Western blotting for H3K9me2. Pairs of lanes (1, 5), (2, 6), (3, 7), and (4,8) are the samples obtained from the same patients with moderately differentiated adenocarcinoma, respectively. H3K9me2 expression is up-regulated in moderately differentiated adenocarcinomas (lanes 1-4) compared with individual-matched normal tissues (lanes 5-8), as shown by the densitometric assessments below (P = .0005). Columns, mean of densitometric values of normal colon mucosa (n = 4) or colon carcinomas (n = 4); bars, SE.
836
T. Nakazawa et al.
Fig. 2 Representative photomicrograph and results of immunohistochemical staining for H3K9me2 in adenocarcinoma with adenoma component. A, The adenocarcinoma component shows submucosal invasion (left side) and the adenoma component is seen continuously (right side). B, In the adenoma component with low-grade dysplasia, weak immunoreactivity for H3K9me2 is observed in the tumor cell nuclei. C, The immunoreactivity for H3K9me2 increases in the adenoma component showing severe nuclear overlapping and swelling and loss of polarity (high-grade dysplasia). D, Diffuse and strong immunoreactivity for H3K9me2 is seen in the adenocarcinoma component.
made based on The World Health Organization Classification of Tumours, Pathology and Genetics of Tumours of the Digestive System [22]. All cases were selected for the immunohistochemical study on the basis of availability of routinely processed, paraffin-embedded, formalin-fixed tissue blocks. Four pairs of fresh tumor and normal colorectal tissues were obtained from the patients with moderately differentiated adenocarcinoma for Western blotting. The fresh tissue was immediately frozen with liquid nitrogen and stored at −80°C immediately after surgical resection.
2.2. Immunohistochemistry Four-micrometer-thick sections were cut from paraffin blocks. After deparaffinization, they were steamed for 20 minutes in sodium citrate buffer (diluted to 1× from 10× heatinduced epitope retrieval buffer). The sections were incubated with primary rabbit polyclonal antibodies (Cell Signaling Technology, Boston, MA) at room temperature, for dimethyl histone H3 Lys9 (H3K9me2) (catalogue number 9753, 1:200 dilution), acetyl histone H3 Lys9 (H3K9ac) (catalogue number
9671, 1:100 dilution), dimethyl histone H3 Lys4 (H3K4me2) (catalogue number 9725, 1:1500 dilution), and trimethyl histone H3 Lys27 (H3K27me3) (catalogue number 9733, 1:50 dilution) for 90 minutes. They were then incubated with a goat antirabbit biotinylated secondary antibody conjugated with avidin-biotin peroxidase complex for another 90 minutes. Finally, specific immunostaining was detected with 3′,3′diaminobenzidine tetrahydrochloride (Sigma, St Louis, MO) and counterstained with hematoxylin. Lymph nodes in which the absence of metastatic tumor was confirmed were stained as positive controls and concomitantly all sections without the primary antibodies as negative controls. The stained sections were examined by 2 pathologists (T.N. and T.K.), who agreed in all cases.
2.3. Evaluation of immunohistochemistry For each case, at least 500 cells were counted in both tumor and adjacent normal colonic mucosa. Only the nuclear immunopositivity of epithelial components was estimated. The percentage of immunopositive cells and the intensities of
Global histone modification of histone H3 in colorectal cancer H3K9me2, H3K9ac, H3K4me2, and H3K27me3 were analyzed. The degree of positive cells was semiquantitatively evaluated as follows: first, scoring according to intensity (negative, 0; weak, 1; intermediate, 2; strong, 3); then, scoring according to the percentage of positive cells (0%, 0; 1%-10%, 1; 11%-50%, 2; N50%, 3). Finally, each sample was divided into 4 groups depending on the total score (frequency of immunopositive cells plus staining intensity); total scores of 0,
837
1 to 3, 4, and 5 to 6 were classified as negative, low expression, moderate expression, and high expression, respectively.
2.4. Western blotting Fresh colonic tissue was homogenized using polytron homogenizer in radioimmunoprecipitation assay buffer with proteinase inhibitors (1× PBS, 1% Nonidet P-40, 0.5%
Fig. 3 Representative results of immunohistochemical staining for H3K9me2. A, In normal colorectal mucosa, a small number of positive cells is observed in the nuclei of the nonneoplastic epithelium. B, Positive cells are focally observed in tubular adenoma. Most of the cancer cells demonstrate strong positivity in well (C), moderately (D), and poorly (E) differentiated adenocarcinoma and mucinous adenocarcinoma (F).
838 Table 2
T. Nakazawa et al. Immunohistochemical results of H3K9me2 in colorectal adenocarcinomas
Differentiation and histological subtype
n
IHC score (H3K9me2) Negative
Low
Moderate
High
Well-differentiated adenocarcinoma Moderately differentiated adenocarcinoma Poorly differentiated adenocarcinoma Mucinous adenocarcinoma
14 18 14 14
0 0 0 0
0 0 0 0
2 2 3 4
12 16 11 10
sodium deoxycholate, 0.1% SDS, 1 mmol/L phenylmethylsulfonyl fluoride, 12 μg/mL aprotinin, and 1 mmol/L sodium ovarthovanadate). Equal amounts of protein (40 μg) solubilized in sample buffer were separated on 10% SDS polyacrylamide gels and transferred electrophoretically to a polyvinylidene difluoride membrane. The membrane was blocked in TBS containing 0.5% Tween 20 plus 5% nonfat dried milk for 1 hour at room temperature and probed with a primary antibody (primary antisera: dimethyl histone H3 Lys9, 1:1000; Cell Signaling Technology; anti-actin, 1:1000; Sigma) at 4°C overnight. The membrane was washed thrice for 10 minutes in TBS containing 0.5% Tween 20 and incubated with horseradish peroxidase–conjugated anti-rabbit secondary antibody (1:2000; Santa Cruz Biotechnology, Santa Cruz, CA) for 1 hour at room temperature. Targeted proteins were visualized using an enhanced detection system (Amersham, Buckinghamshire, UK).
2.5. Statistical analysis Statistical analysis was carried out using χ 2 tests, comparing the frequencies of cases that had high expression of H3K9me2 by immunohistochemistry (normal vs adenoma, normal vs adenocarcinoma, and adenoma vs adenocarcinoma) (Table 1). Student t test was used to compare the Western blots of H3K9me2 protein in paired colon carcinoma and normal colon mucosa (Fig. 1). Statistic significance was set at P b .05. Data analyses were carried out using SPSS version 11.0 for Windows (SPSS, Tokyo, Japan).
3. Results 3.1. H3K9me2 levels in normal, colorectal cancer, and its precursor lesions H3K9me2 expression in 4 moderately differentiated adenocarcinomas was compared with that in the normal colorectal tissue sampled from the same patients by Western blotting. After loading an equal amount of protein, parallel actin immunoblotting was performed, and signal quantification was performed by densitometric scanning. The results showed that H3K9me2 protein expression was significantly up-regulated in all of the analyzed adenocarcinomas compared with individual-matched normal colorectal tissues (P = .0005) (Fig. 1).
(14.3%) (11.1%) (21.4%) (28.6%)
(85.7%) (88.9%) (78.6%) (71.4%)
In the tubular adenomas, the immunopositivity for H3K9me2 gradually increased by the degree of nuclear overlapping and swelling and by loss of polarity. A similar pattern was also observed in all 4 cases having adenocarcinoma with adenoma component (Fig. 2). The H3K9me2 immunoreactivity showed a tendency to be higher in the adenomas with high-grade dysplasia (Fig. 2C) than in those with low-grade dysplasia (Fig. 2B). As seen in the tumors consisting of adenocarcinoma alone, diffuse immunopositivity for H3K9me2 was evident in the adenocarcinoma components (Fig. 2D). Fig. 3 shows the immunohistochemical results for H3K9me2 in the representative sections of the normal colonic mucosa and the colorectal tumors. The cell nuclei completely lacked immunoreactivity for H3K9me2 in the negative controls. The immunohistochemical scores for H3K9me2 are shown in Table 1. In the positive control sections, most of the cell nuclei showed low to moderate expression. In 85 normal colorectal tissue samples, there were no negative cases. More than half of the cases (57/85, 67.1%) had low expression, 21 (24.7%) of 85 cases had moderate expression, and 7 (8.2%) of 85 cases had high expression. In 25 tubular adenomas, moderate expression was the most frequently observed (10/25 cases, 40%), whereas 9 (36%) and 6 (24%) of the 25 tubular adenomas had low and high expression, respectively. In the adenocarcinomas, most cases (49/60, 81.7%) had high expression, and 11 (18.3%) of 60 cases had moderate expression. The frequency of cases with high expression was significantly higher in the adenocarcinomas than in normal mucosa (P b .001). The adenocarcinomas showed a significantly higher proportion of high expression than the adenomas (P b .001). H3K9me2 expression tended to be higher in the adenomas than in the normal mucosa, but the difference was not significant (P = .055). The immunohistochemical results for H3K9me2 in well, moderately, and poorly differentiated adenocarcinomas and mucinous adenocarcinoma are shown in Fig. 1C to F. The immunohistochemical scores of H3K9me2 in the adenocarcinomas by differentiation and histological subtypes are shown in Table 2. In the adenocarcinomas, 12 (85.7%), 16 (88.9%), and 11 (78.6%) cases had high expression in well, moderately differentiated, and poorly differentiated types, respectively. Less frequently, moderate expression was demonstrated in 2 well (14.3%), 2 moderately (11.1%), and 3 poorly (21.4%)
Global histone modification of histone H3 in colorectal cancer differentiated adenocarcinomas. Of 14 mucinous adenocarcinomas, 4 had moderate expression (28.6%), and 10 showed high expression (71.4%). In the adenocarcinomas, H3K9me2 was highly expressed regardless of differentiation and histological subtype.
839
3.2. H3K9ac, H3K4me2, and H3K27me3 levels in normal, colorectal cancer, and its precursor lesions Fig. 4 shows the immunohistochemical results for H3K9ac, H3K4me2, and H3K27me3 in normal colorectal
Fig. 4 Representative immunohistochemical results for H3K9ac, H3K4me2, and H3K27me3 in the normal and neoplastic nuclei. A,B, Representative immunohistochemical results for H3K9ac. Diffuse immunopositivity is observed in normal mucosa (A) and moderately differentiated adenocarcinoma (B). C,D, Representative immunohistochemical results for H3K4me2. Positive cells were focally observed with mild to moderate intensity in normal colorectal mucosa (C) and moderately differentiated adenocarcinoma (D). E,F, Representative immunohistochemical results for H3K27me3. Immunopositivity is diffusely observed in normal colorectal mucosa (E) and moderately differentiated adenocarcinoma (F).
840 Table 3
T. Nakazawa et al. Immunohistochemical results of H3K4me2 and H3K27me3 in normal and neoplastic colorectal tissues
Histology
n
IHC score (H3K4me2)
IHC score (H3K27me3)
Negative
Low
Moderate
High
Negative
Low
Moderate
High
Normal colon mocusa Tubular adenoma Adenocarcinoma
85 25 60
0 0 0
54 (63.5%) 15 (60%) 37 (61.7%)
28 (32.9%) 9 (36%) 20 (33.3%)
3 (3.5%) 1 (4%) 3 (5%)
0 0 0
0 0 0
30 (35.3%) 6 (24%) 23 (38.3%)
55 (64.7%) 19 (76%) 37 (61.7%)
mucosa, cancer, and its precursor lesions. The cell nuclei showed no immunoreactivity for H3K9ac, H3K4me2, and H3K27me3 in the negative controls. The immunohistochemical scores for H3K9ac in normal colorectal mucosa are shown in Table 1; moderate expression was observed in 23 (27.1%) of 85, and high expression was seen in 62 (72.9%) of 85. In the positive controls, most of the cell nuclei showed moderate expression. Negativity and low expression were found in none of the normal glandular cell nuclei. Similar patterns were observed both in tubular adenomas and adenocarcinomas. There was no significant difference among these 3 groups. Similarly, no significant differences in the scores of H3K4me2 and H3K27me3 were seen between the cell nuclei of the normal and the neoplastic colorectal tissue (Table 3). The expression levels were equivalent to those of the positive controls. There were no significant differences in H3K9ac, H3K4me2, and H3K27me3 scores among well, moderately, and poorly differentiated adenocarcinomas, mucinous adenocarcinoma, and adenocarcinoma with adenoma component (data not shown).
4. Discussion Histone modification and DNA methylation are regarded as important epigenetic events that regulate gene transcription [23]. Recently, there have been several reports investigating global alteration of histone modification in human epithelial cancers by immunohistochemistry using specific histone markers, but these immunohistochemical analyses do not provide a correlation to specific gene activities. Nevertheless, they allow us to visualize the global status of histone modification at a cellular level. Previous reports studying the global pattern of histone modification focused on the clinical prognosis and suggested that global aberration of histone modification could closely relate to patient prognosis and serve as an accurate indicator [12-21]. However, little has been said about how the global state of histone modification can be changed at a cellular level in tumorigenic and/or carcinogenic processes in human cancers. Consequently, to clarify epigenetic events in colorectal cancer and its precursor lesions, histone H3 status was examined using 4
specific histone markers by comparing neoplastic cells to normal cells. To the best of our knowledge, there has been only one report studying global histone modification in colorectal neoplasms. Ashktorab et al [21] reported that global levels of H4K12ac and H3K18ac increased in adenocarcinomas in comparison with those in normal tissue and adenomas using immunohistochemistry. They also demonstrated that HDAC2 and H4K12ac expressions in adenocarcinoma were higher than in adenoma, implying that these epigenetic changes have a role in the progression from adenoma to adenocarcinoma. In the present study, the global level of H3K9me2 expression was higher in neoplastic cells than in normal glandular cells on immunohistochemistry. Interestingly, H3K9me2 expression was up-regulated in the adenomas, which are recognized as precancerous lesions, with highgrade dysplasia compared with those with low-grade dysplasia. In addition, H3K9me2 expression was also significantly increased in the nuclei of adenocarcinoma cells compared with that in the nuclei of adenoma cells. These results may suggest that the increased H3K9me2 expression is associated with progression (adenoma to adenocarcinoma). It seems preferable to separate adenoma with high-grade dysplasia from that with low-grade dysplasia and adenoma with high-grade adenoma from well-differentiated adenocarcinoma. However, it is actually difficult to discriminate between these tumors strictly. In the present study, it was confirmed that the amount of H3K9me2 protein was greater in the adenocarcinoma tissue than in the normal colorectal tissues by immunohistochemistry and Western blot. Seligson et al [17] reported that a low level of H3K9me2 is significantly associated with aggressive phenotypes of prostatic and renal cancers and that this can independently predict poor survival. However, in the current study, the cellular global H3K9me2 levels were not altered regardless of the tumor differentiation and the histological subtypes of colorectal adenocarcinomas. It has been generally recognized that poorly differentiated adenocarcinoma and mucinous adenocarcinoma are positively correlated with poor patient clinical outcomes. Therefore, the global H3K9me2 state seems to be less relevant in more aggressive colorectal cancers leading to poor clinical outcomes. Although it is difficult to interpret the effects, these findings suggest that global alteration of H3K9me2
Global histone modification of histone H3 in colorectal cancer expression may correlate with genetic alteration and instability involving tumorigenesis and/or carcinogenesis in colorectal adenoma and adenocarcinoma, in conjunction with other epigenetic events. Previous reports have demonstrated that methylation of histone H3K9 is associated with gene repression through sparse chromatin (heterochromatin) [7,24]. From this aspect, it is conceivable that the increased H3K9me2 level may repress transcriptional activity of certain genes that function as tumor suppressors and/or carcinostasis promoters in colorectal tumors. Conversely, it can also raise the possibility that increased H3K9me2 inactivates genes that suppress tumor cell proliferation after tumorigenesis and/or carcinogenesis as negative feedback. Quantitative analyses for the correlation with specific genes using other molecular methods (ie, ChIP assay in combination with real-time PCR) will be necessary to determine which hypotheses provide a correct explanation. With respect to histone H3 modifications at other residues, the global states of H3K9ac, H3K4me2, and H3K27me3 have been investigated with specific histone markers in a variety of cancers other than colorectal cancers. In more detail, low H3K9ac has been reported to be related to a poor prognosis in lung and breast cancer patients [12,13]. On the other hand, a high level of H3K9ac has been associated with the aggressive type of prostatic cancers [14]. Low H3K4me2 was observed in prostatic cancers with a high risk of recurrence [17]. In kidney cancers, decreased expression of H3K4me1-3 was correlated with shorter survival [19]. Low expression of H3K27me3 was positively related to better patient prognosis in the early stage of esophageal squamous cell carcinomas [20]. In the present study, there were no significant differences in the expression levels of H3K9ac, H3K4me2, and H3K27me3 between normal and neoplastic cell nuclei in the colorectal tumors. In addition, these levels were stable, irrespective of the degree of differentiation and the histological subtypes of colorectal adenocarcinomas (data not shown). As previously reported, some patterns of global histone modification are interestingly common among cancers of different organ origins [17]. Nevertheless, the present results show that a similar pattern could not be observed in the colorectal cancers for the previously investigated patterns of histone modification (H3K9ac, H3K4me2, and H3K27me3) in the other cancers. In summary, global H3K9me2 expression was verified to be up-regulated, whereas there were no differences in H3K9ac, H3K4me2, and H3K27me3 levels in colorectal tumors. The H3K9me2 expression levels were significantly higher in adenocarcinoma cells than in adenoma cells. Among adenocarcinomas, H3K9me2 expression was unchanged by differentiation and histological subtype. Aberrant H3K9me2 expression may play a role in tumorigenesis and progression (adenoma-carcinoma sequence), rather than in transformation to more aggressive pheno-
841
types of colorectal cancers, although the correlation of this epigenetic event to specific genes still remains unclear. From a practical perspective, diffuse H3K9me2 immunopositivity can serve as a useful tool for pathological diagnosis in differentiating between tubular adenoma and adenocarcinoma. Furthermore, it raises the possibility that H3K9me2 is a potential pharmacological therapeutic target for colorectal cancers.
Acknowledgments The authors would like to thank Miyuki Ito, Mikiko Yoda, and especially Yoshihito Koshimizu for their excellent technical assistance and Kayoko Kono for executive assistance.
References [1] Imamura Y, Sobue T. Cancer statistics digest. Mortality trend of colon, rectal, liver, “gallbladder and biliary tract” and pancreas cancer in Japan by birth cohort. Jpn J Clin Oncol 2004;34:491-3. [2] Ashktorab H, Smoot DT, Carethers JM, et al. High incidence of microsatellite instability in colorectal cancer from African Americans. Clin Cancer Res 2003;9:1112-7. [3] Ashktorab H, Smoot DT, Farzanmehr H, et al. Clinicopathological features and microsatellite instability (MSI) in colorectal cancers from African Americans. Int J Cancer 2005;116:914-9. [4] Carethers JM. Racial and ethnic factors in the genetic pathogenesis of colorectal cancer. J Assoc Acad Minor Phys 1999;10:59-67. [5] Marusige K. Activation of chromatin by acetylation of histone side chains. Proc Natl Acad Sci USA 1976;73:3937-41. [6] Vidali G, Ferrari N, Pfeffer U. Histone acetylation: a step in gene activation. Adv Exp Med Biol 1988;231:583-96. [7] Jenuwein T, Allis CD. Translating the histone code. Science 2001;293: 1074-80. [8] Esteller M. Epigenetic gene silencing in cancer: the DNA hypermethylene. Hum Mol Genet 2007;16:R50-9. [9] Kurdistani SK. Histone modifications as markers of cancer prognosis: a cellular view. Br J Cancer 2007;97:1-5. [10] Kondo Y, Shen L, Issa JP. Critical role of histone methylation in tumor suppressor gene silencing in colorectal cancer. Mol Cell Biol 2003;23: 206-15. [11] Fujii S, Luo RZ, Yuan J, et al. Reactivation of the silenced and imprinted alleles of ARHI is associated with increased histone H3 acetylation and decreased histone H3 lysine 9 methylation. Hum Mol Genet 2003;12:1791-800. [12] Elsheikh SE, Green AR, Rakha EA, et al. Global histone modifications in breast cancer correlate with tumor phenotypes, prognostic factors, and patient outcome. Cancer Res 2009;69:3802-9. [13] Barlési F, Giaccone G, Gallegos-Ruiz MI, et al. Global histone modifications predict prognosis of resected non–small-cell-lung cancer. J Clin Oncol 2007;25:4358-64. [14] Sligson DB, Horvath S, Shi T, et al. Global histone modification patterns predict risk of prostate cancer recurrence. Nature 2005;435: 1262-6. [15] Ellinger J, Kahl P, von der Gathen J, et al. Global levels of histone modifications predict prostate cancer recurrence. Prostate 2010;70: 61-9.
842 [16] Park YS, Jin MY, Kim YJ, Yook JH, Kim BS, Jang SJ. The global histone modification pattern correlates with cancer recurrence and overall survival in gastric adenocarcinomas. Ann Surg Oncol 2008;15: 1968-76. [17] Seligson DB, Horvath S, McBrian MA, et al. Global levels of histone modifications predict prognosis in different cancers. Am J Pathol 2009;174:1619-28. [18] Minardi D, Lucarini G, Filosa A, et al. Prognostic role of global DNA-methylation and histone acetylation in pT1a clear cell renal carcinoma in partial nephrectomy specimens. J Cell Mol Med 2009; 13:2115-21. [19] Ellinger J, Kahl P, Mertens C, et al. Prognostic relevance of global histone H3 lysine 4 (H3K4) methylation in renal cell carcinoma. Int J Cancer 2010;127:2360-6.
T. Nakazawa et al. [20] Tzao C, Tung HJ, Jin JS, et al. Prognostic significance of global histone modifications in resected squamous cell carcinoma of the esophagus. Mod Pathol 2009;22:252-60. [21] Ashktorab H, Belgrave K, Hosseinkhah F, et al. Global histone H4 acetylation and HDAC2 expression in colon adenoma and carcinoma. Dig Dig Sci 2009;54:2109-17. [22] Bosman FT, Carneiro F, editors. Pathology and genetics of tumours of the digestive system. The World Health Organization Classification of Tumours. Lyon: IARC Press; 2010. p. 131-82. [23] Gronbaek K, Hother C, Jones PA. Epigenetic changes in cancer. APMIS 2007;115:1039-59. [24] Rea S, Eisenhaber F, O'Carroll D, et al. Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 2000; 406:593-9.