Decreased expression of ING2 gene and its clinicopathological significance in hepatocellular carcinoma

Decreased expression of ING2 gene and its clinicopathological significance in hepatocellular carcinoma

Available online at www.sciencedirect.com Cancer Letters 261 (2008) 183–192 www.elsevier.com/locate/canlet Decreased expression of ING2 gene and its...

744KB Sizes 0 Downloads 60 Views

Available online at www.sciencedirect.com

Cancer Letters 261 (2008) 183–192 www.elsevier.com/locate/canlet

Decreased expression of ING2 gene and its clinicopathological significance in hepatocellular carcinoma Hua-kun Zhang a, Ke Pan a, Hui Wang a, De-sheng Weng a, Hai-feng Song a, Jun Zhou a, Wei Huang a, Jian-jun Li a, Min-shan Chen b, Jian-chuan Xia a,* a

State Key Laboratory of Oncology in Southern China and Department of Experimental Research, Sun Yat-sen University Cancer Center, 651 Dongfeng Road East, Yuexiu District, Guangzhou, Guangdong Province 510060, PR China b Department of Hepatobilary Oncology, Sun Yat-sen University Cancer Center, Guangzhou 510060, PR China Received 6 September 2007; received in revised form 17 October 2007; accepted 10 November 2007

Abstract The inhibitor of growth (ING) family member 2 (ING2) is a newly discovered member of ING family that can regulate a wide range of cellular processes including cell growth arrest, apoptosis, and DNA repair. Researches have shown that ING2 can activate p53 and p53-mediated apoptotic pathway involved in the hepatocarcinogenesis. To investigate the role of ING2 in hepatocellular carcinoma (HCC) pathogenesis, we analyzed the correlations between the ING2 expression level and clinicopathologic factors and studied its prognostic role in primary HCC. Using reverse transcription-polymerase chain reaction (RT-PCR) and Western blot, ING2 transcription and post-transcription level was found to be downregulated in the majority of tumors compared with matched non-tumors liver tissues (p = 0.004 and p = 0.014, respectively). The immunohistochemistry data indicated significant reduction of ING2 expression level in 44 of 84 (52.4%) HCC cases. In addition, the expression level of ING2 correlated with tumor size, histopathologic classification, serum AFP (p < 0.05). Kaplan–Meier curves demonstrated that patients with reduced ING2 expression were at significantly increased risk for shortened survival time (p = 0.009). Using multivariate analysis, ING2 expression was found to be an independent prognostic factor. Our data suggest that ING2 is involved in the progression of HCC, therefore it is considered to be a candidate tumor suppressor gene and its significantly decreased expression in HCC may lead to an unfavorable prognosis.  2007 Elsevier Ireland Ltd. All rights reserved. Keywords: ING2; Hepatocellular carcinoma; Down-regulation; AFP; Prognosis

1. Introduction The inhibitor of growth (ING) family member 2 (ING2), as a candidate tumor suppressor isolated in 1998 through a homology search of p33ING1b * Corresponding author. Tel.: +86 20 87343173; fax: +86 20 87343392. E-mail address: [email protected] (J. Xia).

cDNA sequence with the Otsuka cDNA database, is an important member of the ING family which encode a series of proteins as important cofactors of p53, associated with cell cycle progression, apoptosis, and, DNA repair [1,2]. ING2, initially named INGL by one group, encodes a protein which shares 58.9% homology with p33ING1b [1].The C-terminus of ING2 contains a highly conserved plant homeodomain (PHD) zinc-finger that binds to

0304-3835/$ - see front matter  2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2007.11.019

184

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

histone-3 trimethylated at lysine 4 and acts as nuclear phosphoinositide receptors to active gene repression and regulate DNA damage [3–5]; the N-terminus contains a leucine zipper-like (LZL) motif critical for the proper functions of ING2 in DNA repair, apoptosis, and chromatin remodeling after UV irradiation [6]. ING2 proteins were mainly localized in the nucleus, also with a diffuse distribution in the cytoplasm [5]. It has been demonstrated that the expression of ING2 not only negatively regulates cell growth through induction of apoptosis and G1-phase cell-cycle arrest in a p53-dependent manner but also is involved in p53-mediated replicative senescence [7,8]. The arguments that ING2 can negatively regulates cell proliferation, induce apoptosis, and modulate cell cycle progression strongly support the notion that ING2 acts as a tumor suppressor gene [8,9]. Hepatocellular carcinoma (HCC) is a highly lethal cancer that typically has poor prognosis. It ranks the third cause of cancer death especially in East Asia and sub-Saharan Africa [10,11] and is the second leading cause of cancer death among males in China [12]. Now it is also increasing in United States and Europe [13,14]. Even with well developed surgical resection, the patient’s 5-year survival rate is still poor (about 35%) [15]. To improve patient’s outcome, it is of great clinical importance to find reliable prognostic markers, which could predict relapse and dissemination of HCC. The molecular pathogenesis of HCC seems to involve multiple genetic aberrations in the molecular control of hepatocyte proliferation, differentiation, and the maintenance of genomic integrity. This process is influenced by the cumulative activation and inactivation of oncogenes, tumor suppressor genes, and other genes. Clarification of the genes involved in the development of HCC will contribute to our understanding of mechanisms of hepatocarcinogenesis. ING2 gene mapped onto 4q35.1 [1], a frequent loss of heterozygosity (LOH) region in HCC [16,17], has been implicated as a potential tumor suppressor gene in a few human malignancies [18,19]. Furthermore, ING2 shares the structural similarities with p33ING1b which may play an important role as a tumor suppressor gene in HCC [20,21], however, the involvement of ING2 in HCC remains largely unclear. In the present research, we intended to investigate the expression of ING2 in primary HCC using reverse transcriptase-polymerase chain reaction (RT-PCR), Western blot and immunohistochemistry. Furthermore, we identified its relation-

ship to clinicopathological features and evaluated its prognostic value to post-resectional survival in HCC. 2. Materials and methods 2.1. Patients and tissue specimens All matched fresh HCC and adjacent non-cancerous tissue specimens (n = 87) were obtained from those patients underwent surgical resection in the Sun Yat-sen University Cancer Center between 2005 and 2006 of their diseases. The patients who had received any preoperative treatment such as chemotherapy, ethanol injection or transarterial chemoembolization were excluded. The 87 patients included 78 males and 9 females with a median age of 50 years (range, 21–75 years). The fresh tissues were immediately immersed in RNAlater (Ambion, Inc., USA) after surgical resection and stored at 4 C overnight to allow thorough penetration of the tissues, then frozen at 80 C until RNA extraction. Both tumor and corresponding non-tumors tissues not less than 2 cm away from the HCC were sampled, respectively, and proved by pathological examination. The other archival paraffin-embedded HCC samples (n = 84) were retrieved according to the 1999–2001 surgical pathology files in the Sun Yat-sen University Cancer Center of which included the patients were without pretreatment. The median age of the patients was 53 years (range, 22–76 years). All tissue blocks were cut into consecutive 4-lm thick sections. The histologic cell types were assigned according to the criteria of the WHO classification. Patients were followed up annually; the follow-up time ranged from 3 to 76 months, with a median follow-up time of 33 months. The study was approved by the Committee for the Conduct of Human Research of the Sun Yat-sen University Cancer Center and informed consent was obtained from each patient. 2.2. Extraction of total RNA and RT-PCR Total RNA was extracted using TRIzol solution (Invitrogen, USA) according to the manufacturer’s protocol. And RNAse-free DNase I was used to remove DNA contamination. Total RNA concentration and quantity were assessed by absorbency at 260 nm using a DNA/Protein Analyzer (DU 530, Beckman, USA). Reverse transcription (RT) was performed in a 20 ll reaction system with 2 lg total RNA treated by M-MLV Reverse transcriptase to synthesis first-strand cDNA (Promega, USA) according to the manufacturer’s recommendation, followed by cDNA amplification using the specific primer set for ING2 and GAPDH as an internal control. The sequences of the sense and antisense primers were as follows: 5 0 -AAACGTCTACAGCAGCTTCTC-3 0 (F) and 5 0 -TCATCACAGTCTTCAATCCCA-3 0 for ING2;

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

5 0 -CGGGAAGCTTGTCATCAATGG-3 0 (F) and 5 0 -G GCAGTGATGGCATGGACTG-3 0 (R) for GAPDH, and the corresponding PCR products are 287 bp and 354 bp, respectively. Each PCR was generally performed in 30 thermal cycles and then the PCR products were observed by electrophoresis on 1.5% agarose gel and visualized after staining with ethidium bromide. To quantify the densities of the bands, the gray values were measured using the Bio-Rad imaging system. After the values of ING2 were normalized by the corresponding values of GAPDH, the ratio of the tumor to the non-tumors liver tissues was calculated. 2.3. Western-blot analysis Tissues were homogenized in nitrogen liquid and then treated with lysis buffer containing 50 mmol/L Hepes (pH 7.0), 250 mmol/L NaCl, 0.1% Nonidet P-40, 5 mmol/L ethylenediaminetetraacetate (EDTA), 1 mmol/L phenylmethylsulfonyl fluoride (PMSF), 1 mmol/L dithiothreitol (DTT), and protease inhibitor cock tail (EMD Biosciences, Inc., USA). After centrifugation at 12,000g 4 C for 30 min, the supernatant was collected and the protein concentration measured using the bicinchoninic acid protein assay kit (Pierce, Rockford, IL, USA). Total protein (50 lg) was separated on 12% SDS–polyacrylamide gel electrophoresis gel, and transferred to PVDF membrane. After blocking non-specific binding sites for 60 min with 5% non-fat milk, the membranes were incubated overnight at 4 C with primary polyclonal antibody against ING2 (Abcam, UK, at 1:250 dilution), and b-actin (Abcam, UK, at 1:250 dilution, at 1:10,000 dilution). After washing, the membrane was then probed with the appropriate secondary antibody for 60 min at room temperature. The immunoblot was revealed by autoradiograph using a supersignal west pico chemiluminescent substrate kit (Pierce, Rockford, IL, USA) and the band intensity was measured by densitometry using the Quantity One software (Bio-Rad Laboratories, Inc. Hercules, CA, USA). The protein levels were normalized with respect to b-actin protein level.

185

4 C overnight. Following incubation with biotinylated secondary antisera (Haoyang, Tianjin, China) at 37 C 30 min, the streptavidin–biotin peroxidase was applied according to the manufacturer’s instruction. Finally, the visualization signal was developed with 3,3 0 -diaminobenzidine tetrahydrochloride (DAB) and then all of the slides were counterstained in hematoxylin. For negative controls, tissue sections were immunoreacted without anti-ING2 antibody under the same experimental conditions. The total ING2 immunostaining score was calculated as the sum of the percent positivity of stained tumor cells and the staining intensity. The percent positivity was scored as ‘0’ (<5%, negative), ‘1’ (5–25%, sporadic), ‘2’ (25–50%, focal), ‘3’ (>50%, diffuse). The staining intensity was score as ‘0’ (no staining), ‘1’ (weakly stained), ‘2’ (moderately stained), and ‘3’ (strongly stained). The final ING2 expression score was ranged from 0 to 9, we defined ING2 expression level as follows: ‘’ (score 0–1), ‘+’ (score 2–3), ‘++’ (score 4–6) and ‘+++’ (score >6). 2.5. Statistical analysis Quantitative values were expressed as means ± SD or median (range). Paired-samples t-test was used to compare mRNA and protein expression of ING2 in HCC with paired adjacent non-cancerous tissue samples. Categorical variables were enumeration data of counting the number of samples. The v2 test for proportion and Pearson’s correlation was used to analyze the relationship between ING2 expression and various clinicopathologic characteristics. For survival analysis, the main outcome was overall survival rates which were calculated from the date of surgery to the date of death. Survival curves were calculated using the Kaplan–Meier method and compared by the log-rank test. Cox proportional-hazard analysis was used for univariate and multivariate analysis to explore the effect of clinicopathological variables and the ING2 expression on survival. The SPSS 15.0 software (SPSS Inc., Chicago, IL, USA) was used for all statistical analyses and a p-value <0.05 was considered significant.

2.4. Immunohistochemistry analysis

3. Results

The sections were deparaffinized and rehydrated, through graded ethanol, then endogenous peroxidase was inhibited with 0.3% hydrogen peroxide. For antigen retrieval, slides were boiled in EDTA (1 mmol/L; pH 8.0) for three times of every 10 min in microwave oven. After rinsing with PBS, the sections were incubated with IHC Biotin Block kit (Maixin, Fujian,China) to detent endogenous biotins according to the manufacturer’s instruction at room temperature. Next, the sections were blocked with 5% normal goat serum before primary antibody (rabbit anti-ING2, Proteintech Group, Inc., USA) in blocking buffer (1:100) was applied to be incubated at

3.1. RT-PCR and Western blot analysis Semiquantitative RT-PCR was performed on 87 paired clinical samples to examine the mRNA expression of ING2. ING2 expression was found in hepatocellular cancer tissues as well as in matched non-cancerous tissues, and the density value ratio of tumor to non-tumors was evaluated. There was significant difference in the average level of expression of ING2 mRNA between HCC tissues and adjacent normal tissues using paired t-test (p = 0.004). Well-differentiated tumors tended to exhibit the relatively higher ING2 expression level, and its expres-

186

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

sion decreased in proportion to the tumor stage (Fig. 1). To further investigate if the expression of ING2 is reduced at protein level in HCC, Western blot was performed on 29 HCC specimens and corresponding adjacent non-cancerous livers from 87 paired samples, including 14 samples used RT-PCR detected without ING2 reduced expression. Western blot showed an ING2 band at the expected 30 kDa size and the amount of ING2 protein was measured by densitometry. Sixteen of 29 (55%) tumors showed decreased ING2 expression, as shown in Fig. 2A and the corresponding mRNA expression in Fig. 2B. The analysis of results displayed the density value (normalized to b-actin expression as a loading control) of tumor was significantly lower than that in corresponding non-cancerous tissue using paired t-test (p = 0.014, Fig. 2C). The 16 cases of reduced ING2 protein level of cancerous liver tissues compared with the normal matched tissues included 10 cases with reduced expression on mRNA level and 6 cases even without the transcriptional reduction.

(62%; Fig. 3C and D). ING2 immunostaining was mostly in cytoplasm; however, weak nuclear staining was also observed (Fig. 3E). In cases with adjacent hyperplastic tissue, we often observed a sharp contrast between infiltrative tumor areas of negative staining and the adjacent tissue of positive staining (Fig. 3F and G). Correlations between the expression of ING2 and various clinicopathologic parameters are listed in Table 1. Consistent with the former result of RT-PCR and Western blot, the expression of ING2 correlated with tumor stage (p = 0.025). In addition, the negative ING2 expression was positively associated with tumor size (P5 cm) and the increased AFP expression in serum of patients (p = 0.004 and p = 0.016, respectively). There was no statistically significant difference between ING2 expression and gender, age, liver cirrhosis, HBV infection or recurrence.

3.2. Immunohistochemistry analysis

To investigate the prognostic value of the ING2 gene, the association between the negative ING2 staining and overall survival was initially evaluated using Kaplan– Meier survival curves with the log-rank test and then confirmed with univariate and multivariate Cox regression models. The overall survival rates were 43% and 17%, respectively, in patients with positive and negative ING2 expression. Patients showing negative ING2 expression had a significantly shorter overall survival than those with positive expression (p = 0.009, log-rank test; Fig. 4A). Univariate Cox regression analysis also identified that clinical variables including tumor size, tumor stage and serum AFP were significantly associated with overall survival (Table 2). Furthermore, to evaluate the potential of negative ING2 expression (negative vs. positive) as an independent predictor for overall survival of HCC, multivariate Cox regression analyses were performed. While

To further gain insight into the effect and the prognostic value of ING2 paraffin-embedded tissue blocks (n = 84) with histopathologically confirmed HCC were examined using immunohistochemistry. Various levels of immunoreactivity for ING2 were found in hepatoma and adjacent non-tumor regions. The expression level of ING2 was evaluated as before described in Section 2. Overall, 44 of 84 (52.4%) cases had negative expression (ING2) in tumor cells, 40 of 84 cases had positive expression (ING2+ OR++). There was positive ING2 staining in 19 cases of well-differentiated (66%; Fig. 3A), 21 cases of moderately and poorly differentiated HCC specimens (38%; Fig. 3B). However, ING2 staining was decreased profoundly even undetectable in 34 cases of moderately and poorly differentiated HCC specimens

01 N C

20

21

32

36

3.3. Negative ING2 staining correlates with disease-specific patient survival

39

48

70

73

81

N C N C N C N C N C N C N C N C N C

ING2

287bp

GAPDH

354 bp

well

moderate

poor

Fig. 1. Expression pattern of ING2 in HCC specimens by RT-PCR. Representative results of semi-quantitative RT-PCR of ING2 from 10 pairs of HCC and corresponding normal livers are shown, where GAPDH was employed as an internal control. The expression of ING2 gene was reduced in tumor tissues when compared with corresponding non-tumorous tissues. Well, well differentiated; moderate, moderately differentiated; Poor, poorly differentiated. C, HCC; N, corresponding non-cancerous tissues.

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

03

A N

B

65 C

N

187

16 C

N

C

ING2

30KDa

β -actin

42KDa

ING2

287bp

GADPH

354 bp

C Protein expression (ING2/actin)

1.2 1 Protein expression

0.8

p=0.014

0.6 0.4 0.2 0 N

T

Fig. 2. Expression pattern of ING2 in HCC and corresponding non-tumors specimens. (A) Representative results of three matched pairs using Western are shown. b-Actin was used as a loading control. (B) Corresponding ING2 mRNA expression using RT-PCR. Sample * 16: the ratio of gray values (T/N; confer Section 2) was 0.92 for mRNA and 0.44 for protein. (C) Densitometry Analyses of ING2 protein level quantified by compared with b-actin in HCC and corresponding normal liver samples. The expression of ING2 gene was reduced in tumor tissues when compared with corresponding non-tumorous tissues. C, HCC; N, corresponding non-cancerous tissues.

the others failed to demonstrate independence, only serum AFP level and negative ING2 expression may play a role to predict of the overall survival in HCC (p = 0.021 and p = 0.041, respectively; Table 2).

4. Discussion HCC was generally recognized as a life-threatening malignancy especially in Southeast Asia because of its dismal prognosis. Compared with research on activation of cellular oncogenes, more studies concentrate on inactivation of tumor suppressor genes including P53, PTEN, nm23, etc. [22–25]. Molecular genetic studies have shown that LOH at specific chromosomal loci of 4q occurs very frequently in HCC [16,26,27], and the tumor suppressor genes always locate on such loci and their inactivations are involved in the development

of HCC. The human ING2 gene located on ‘‘hot spot’’ deleted region 4q35.1; however, at present few tumor suppressor genes located in this regions have been identified. Considering the structural homology, p33ING2 may share functional similarities with p33ING1b, and the reduced expression of ING1 gene have been detected in various cancers including HCC [20,28,29]. There have been a few other studies demonstrating the involvement of ING2 as a tumor suppressor in certain malignancies, however, most of these studies were conducted in vitro models or in animals [2]. Thus, we presumed that ING2 may play a role in tumor progression of HCC. To our knowledge, there are no reports about studies of ING2 in human HCC samples. In our study, to evaluate its potential clinical relevance we investigated the expression pattern of ING2 in HCC.

188

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

Fig. 3. Immunohistochemical detection of the ING2 protein expression in HCC and surrounding non-cancerous tissues. (A) Welldifferentiated hepatocellular carcinoma (200·), scored as ING2 (++) according to the criteria defined in Section 2; (B) moderately differentiated hepatocellular carcinoma (200·), scored as ING2 (+); (C and D) poorly differentiated hepatocellular carcinoma at two different magnification (200·) and (400·), respectively, scored as ING2 (); (E) Immunostaining of HCC (400·) showing nuclear and cytoplasmic staining: the immunostaining is weakly present in the nucleus and the cytoplasm, but it is not present in all the cells. (F and G) Immunostaining of HCC and adjacent surrounding non-cancerous tissues showing a sharp contrast between infiltrative tumor areas of negative staining and the adjacent tissue of positive staining (100·) and (400·), respectively.

H. Zhang et al. / Cancer Letters 261 (2008) 183–192 Table 1 Correlations between ING2 expression and clinicopathologic variables of 84 cases of HCC Clinicopathologic variables

na

ING2 expression Negative

Positive

All cases Age <50 > =50

84

44

40

50 34

25 19

25 15

Gender Female Male

17 67

9 35

8 32

Tumor size (cm) <5 P5

30 54

8 36

29 42 13

10 28 6

19 14 7

Liver cirrhosis Yes No

43 41

25 19

18 22

Recurrence Yes No

21 63

13 31

8 32

HBV Negative Positive

19 65

8 36

11 29

AFP (lg/L) <25 > =25

27 57

9 35

22 22

a

v2

0.476

0.508

0.959

0.003

0.004b

8.213

0.025b

7.363

0.234

1.416

0.313

1.018

0.308

1.039

0.016b

5.788

22 18

Histological differentiation Well Moderate Poor

b

p-Value

Number of cases in each group. Statistically significant (p < 0.05).

Firstly, we examined the mRNA and protein expression of ING2 in paired liver tumorous and non-tumors samples. The decreased expression of ING2 was found in most of liver cancer samples compared with adjacent normal tissues. In addition, there was a correlation of the down-regulation of the gene with tumor grading. ING2 was likely to be involved in HCC differentiation. Poorly differentiated tumors showed relatively low expression of ING2. Compared the results of Western-blot to RT-PCR, the post-transcriptional level was mostly in agreement with the transcriptional level. However, even without reduced expression on mRNA level, the amount of ING2 protein was expressed also less in cancer tissue than in normal matched tissue (6/14). Discordant results may be due to various factors such as the detected ING2 mRNA is not translated into protein or the sensitivity limit of RT-PCR. Con-

189

sistent with our results, Okano et al. [19] reported that the expression of ING2 was down-regulated in 6 of 7 lung cancer cell lines using real-time RTPCR, suggesting ING2 may play a role in carcinogenesis and tumor progression in lung cancer. Secondly, using immunohistochemistry technology, we provided evidences for decreased ING2 protein level in >50% of HCC samples (44/84) and demonstrated the diagnostic value of negative ING2 expression as a prognostic factor. In our series, loss of ING2 expression in HCC was significantly related with tumor grading and with tumor size, consistent with the results of mRNA analyses delivered. Moderately–poorly differentiated tumors and large size tumors were identified related with negative expression of ING2. Furthermore, the negative expression of ING2 also presents to be associated with high rates of elevated AFP serum level (p < 0.01). Aberrant expression of AFP is characteristic of HCC cases and serves as a diagnostic tumorspecific marker [30]. It has been proved that p53 negatively regulates AFP gene expression through alteration of chromatin structure at the core promoter [31,32]. Since ING2 can act as a cofactor of p300 for p53 acetylation and negatively regulate cell proliferation through activation of p53 by enhancing its acetylation [7], it is possible that ING2 regulate AFP gene through modifying p53 function. In fact, Kataoka et al. [33] found p33ING1b, ING2 strongly repressed AFP promoter activity. It is concordant with our results which showed that negative ING2 was also correlated with high level of serum AFP. However, the mechanism is awaited to elucidate. Contrast to ING2 as a nuclear protein, we found ING2 immunostaining was mostly in cytoplasm, however, weak nuclear staining was also observed. In accordance with our findings, Vieyra et al. [34] also found p33ING1b as a homolog of ING2 demonstrated a nuclear protein, aberrantly localized to the cytoplasm in brain tumor samples, and to a lesser extent, to the nucleus of glioma cells. Recent studies support the idea that mislocalization of p33ING1b contribute to the generation of specific tumors. In melanoma, papillary thyroid carcinoma, ductal breast carcinoma, and acute lymphoblastic leukemia, increased levels of cytoplasmic p33ING1b were noted, concomitant with loss of nuclear localization [35,36]. It is possible that mislocalization of ING2 from the nucleus to the cytoplasm could trigger degradation and loss of its function. We have known that the PHD motif, as important domains of ING2, are nuclear receptors of phosphatidylino-

190

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

Survival function

Overall survival ration

negative ING2 expression positive ING2 expression

p=0.009

Months from diagnosis Fig. 4. Estimated overall survival according to the expression of ING2 in 84 cases of HCCs (the Kaplan–Meier method). Based on the results of immunohistochemical staining, the expression of ING2 was classified as the negative expression (; n = 44) and the positive (+ or ++; n = 40). Log-rank test shows that HCC patients with the negative ING2 expression showed significantly poorer prognosis than those with the positive expression. Table 2 The overall survival univariate and multivariate Cox regression analysis Variables

Relative risk (95% CI)

p-Value

Univariate Gender Age Tumor size Histological differentiation Liver cirrhosis HBVDNA Serum AFP Recurrence ING2

0.650(0.318–1.328) 0.663(0.384–1.146) 1.689(0.943–3.025) 1.678(1.152–2.445) 0.693(0.408–1.177) 1.520(0.785–2.946) 2.329(1.283–4.227) 1.225(0.677–2.218) 0.497(0.289–0.855)

0.237 0.141 0.078 0.007a 0.175 0.214 0.005a 0.502 0.011a

Multivariate Serum AFP ING2

2.113(1.121–3.982) 0.540(0.299–0.974)

0.021a 0.041a

Abbreviation: 95% CI, 95% confidence interval. a Statistically significant (p < 0.05).

sitol phosphates correlated with the ING family proteins to induce apoptosis. We proposed that loss of nuclear ING2 expression abrogates its proapoptotic functions and ultimately make it lose the tumor suppressive function [9,34]. However, the precise mechanism of ING2 in our observations remains elusive and need further investigation. HCC is one of the most deadly human carcinoma even with improved diagnosis and compositive ther-

apy, however, the prognosis of HCC remains dismal [37]. Prognostic molecular biomarkers are invaluable for the clinician to evaluate patients and to aid in tumor control. In our series, we found that decreased ING2 protein expressions were significantly correlated with poor patient outcome. Further using multivariate Cox regression analysis, the final model of HCC included both serum AFP and ING2. In summary, the decreased ING2 expression, as an independent factor of survival, may constitute a prognostic factor for patients with HCC after surgery. Our data that the expression of ING2 significantly correlates with patient overall survival is consistent with the former findings that ING2 reduction correlates with tumor pathological grading. Coincident with our results, decreased expression of p33ING1b was associated with poor prognosis in advanced neuroblastomas [29] and reduced nuclear ING3 expression is an independent prognostic factor to predict patient outcome in primary melanomas as well [38]. In addition, serum tumor marker of AFP has also been found to be as a prognostic index in other reports in agreement with our findings. Peng et al. [39] recently analyzed AFP elevation and reported that high AFP appears to contribute to identify subsets of HCC patients with increased risk for early recurrence and poor prognosis after hepatectomy. Okuda et al. [40] also

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

reported that AFP-L3 is a valuable marker for evaluation of curability of surgical treatment and for improving the accuracy of prognosis. HCC with ING2 reduction may have a growth advantage with a higher proliferative capacity related with increased AFP and additional alterations that together lead to poor patient survival. Hepatocellular carcinogenesis is a multi-step and complex process involving several aetiological agents that exert differential effects on the molecular pathways involved. TP53 as a tumor suppressor gene located in 17p13.1 region of chromosomal allelic loss was found mutations in HCCs [41,42]. It has been reported that ING2 cooperates with p53 to regulate apoptosis via activation of both the mitochondrial/intrinsic and death-receptor/extrinsic apoptotic pathways [43], and to induce p53-mediated replicative senescence and DNA damage that negatively regulates cell proliferation through the activation of p53 by enhancing its acetylation [7,8]. Another report showed that ING2 was a stable component of a mSin3A HDAC complex containing the metastasis suppressor protein BRMS1 (breast cancer metastasis suppressor-1) [2,4]. BRMS1 suppresses metastasis of multiple human and murine cancer cells and has been found with RBP1 in mSin3 HDAC complexes [44]. Therefore, ING2 may play a significant role in these crucial tumor suppressor pathways. In our results, the correlation of the ING2 expression level in HCC with the clinicopathological features suggested that ING2 may play an important role as a tumor suppressor candidate in the progression of HCC. Consistent with our findings, Lu et al. [18] found that ING2 expression is significantly reduced in human melanomas and that reduced ING2 may be an important molecular event in the initiation of melanoma development. Sironi et al. [45] performed the analysis on basal cell carcinoma (BCC) and the findings suggest that ING2 in minimal deleted region of 4q32–35 could be involved in sporadic BCC carcinogenesis. In human fibrosarcoma, lung cancer, colorectal carcinoma, and hepatocellular carcinoma cell lines, the studies also indicated the role of ING2 as a tumor suppressor candidate through regulation of gene transcription, induction of cell cycle arrest and apoptosis [5,8,19,33]. In conclusion, these findings suggest that ING2 down-regulation frequently occurs in HCC and its significantly decreased expression in HCC may lead to an unfavorable prognosis. We propose that ING2 may be a candidate tumor suppressor gene

191

for HCC and ING2 may have diagnostic and therapeutic potential for patients with HCC. However, to elucidate the molecular mechanism role of ING2 in liver carcinogenesis and the interaction of ING2 and AFP expression, more studies still should be done. References [1] Y. Shimada, A. Saito, M. Suzuki, E. Takahashi, M. Horie, Cloning of a novel gene (ING1L) homologous to ING1, a candidate tumor suppressor, Cytogenet. Cell Genet. 83 (1998) 232–235. [2] Y. Doyon, C. Cayrou, M. Ullah, A.J. Landry, V. Cote, W. Selleck, et al., ING tumor suppressor proteins are critical regulators of chromatin acetylation required for genome expression and perpetuation, Mol. Cell 21 (2006) 51–64. [3] P.V. Pena, F. Davrazou, X. Shi, K.L. Walter, V.V. Verkhusha, O. Gozani, et al., Molecular mechanism of histone H3K4me3 recognition by plant homeodomain of ING2, Nature 442 (2006) 100–103. [4] X. Shi, T. Hong, K.L. Walter, M. Ewalt, E. Michishita, T. Hung, et al., ING2 PHD domain links histone H3 lysine 4 methylation to active gene repression, Nature 442 (2006) 96– 99. [5] O. Gozani, P. Karuman, D.R. Jones, D. Ivanov, J. Cha, A.A. Lugovskoy, et al., The PHD finger of the chromatinassociated protein ING2 functions as a nuclear phosphoinositide receptor, Cell 114 (2003) 99–111. [6] Y. Wang, J. Wang, G. Li, Leucine zipper-like domain is required for tumor suppressor ING2-mediated nucleotide excision repair and apoptosis, FEBS Lett. 580 (2006) 3787– 3793. [7] R. Pedeux, S. Sengupta, J.C. Shen, O.N. Demidov, S. Saito, H. Onogi, et al., ING2 regulates the onset of replicative senescence by induction of p300-dependent p53 acetylation, Mol. Cell. Biol. 25 (2005) 6639–6648. [8] M. Nagashima, M. Shiseki, K. Miura, K. Hagiwara, S.P. Linke, R. Pedeux, et al., DNA damage-inducible gene p33ING2 negatively regulates cell proliferation through acetylation of p53, Proc. Natl. Acad. Sci. USA 98 (2001) 9671–9676. [9] W. Gong, K. Suzuki, M. Russell, K. Riabowol, Function of the ING family of PHD proteins in cancer, Int. J. Biochem. Cell Biol. 37 (2005) 1054–1065. [10] J.M. Llovet, A. Burroughs, J. Bruix, Hepatocellular carcinoma, Lancet 362 (2003) 1907–1917. [11] T.M. Block, A.S. Mehta, C.J. Fimmel, R. Jordan, Molecular viral oncology of hepatocellular carcinoma, Oncogene 22 (2003) 5093–5107. [12] R. Tung-Ping Poon, S.T. Fan, J. Wong, Risk factors, prevention, and management of postoperative recurrence after resection of hepatocellular carcinoma, Ann. Surg. 232 (2000) 10–24. [13] R. Capocaccia, M. Sant, F. Berrino, A. Simonetti, V. Santi, F. Trevisani, Hepatocellular carcinoma: trends of incidence and survival in Europe and the United States at the end of the 20th century, Am. J. Gastroenterol. 102 (2007) 1661– 1670. [14] H.B. El-Serag, Hepatocellular carcinoma: recent trends in the United States, Gastroenterology 127 (2004) S27–S34.

192

H. Zhang et al. / Cancer Letters 261 (2008) 183–192

[15] M.B. Thomas, A.X. Zhu, Hepatocellular carcinoma: the need for progress, J. Clin. Oncol. 23 (2005) 2892–2899. [16] O. Bluteau, J.C. Beaudoin, P. Pasturaud, J. Belghiti, D. Franco, P. Bioulac-Sage, et al., Specific association between alcohol intake, high grade of differentiation and 4q34–q35 deletions in hepatocellular carcinomas identified by high resolution allelotyping, Oncogene 21 (2002) 1225– 1232. [17] K. Bando, H. Nagai, S. Matsumoto, M. Koyama, N. Kawamura, M. Onda, et al., Identification of a 1-cM region of common deletion on 4q35 associated with progression of hepatocellular carcinoma, Genes Chromosomes Cancer 25 (1999) 284–289. [18] F. Lu, D.L. Dai, M. Martinka, V. Ho, G. Li, Nuclear ING2 expression is reduced in human cutaneous melanomas, Br. J. Cancer 95 (2006) 80–86. [19] T. Okano, A. Gemma, Y. Hosoya, Y. Hosomi, M. Nara, Y. Kokubo, et al., Alterations in novel candidate tumor suppressor genes, ING1 and ING2 in human lung cancer, Oncol. Rep. 15 (2006) 545–549. [20] Z. Zhu, J. Lin, J.H. Qu, M.A. Feitelson, C.R. Ni, F.M. Li, et al., Inhibitory effect of tumor suppressor p33(ING1b) and its synergy with p53 gene in hepatocellular carcinoma, World J. Gastroenterol. 11 (2005) 1903–1909. [21] T. Ohgi, T. Masaki, S. Nakai, A. Morishita, S. Yukimasa, M. Nagai, et al., Expression of p33(ING1) in hepatocellular carcinoma: relationships to tumour differentiation and cyclin E kinase activity, Scand. J. Gastroenterol. 37 (2002) 1440– 1448. [22] C.D. Mann, C.P. Neal, G. Garcea, M.M. Manson, A.R. Dennison, D.P. Berry, Prognostic molecular markers in hepatocellular carcinoma: a systematic review, Eur. J. Cancer 43 (2007) 979–992. [23] D. Moradpour, H.E. Blum, Pathogenesis of hepatocellular carcinoma, Eur. J. Gastroenterol. Hepatol. 17 (2005) 477– 483. [24] T.H. Hu, C.C. Huang, P.R. Lin, H.W. Chang, L.P. Ger, Y.W. Lin, et al., Expression and prognostic role of tumor suppressor gene PTEN/MMAC1/TEP1 in hepatocellular carcinoma, Cancer 97 (2003) 1929–1940. [25] Y.J. Yao, X.L. Ping, H. Zhang, F.F. Chen, P.K. Lee, H. Ahsan, et al., PTEN/MMAC1 mutations in hepatocellular carcinomas, Oncogene 18 (1999) 3181–3185. [26] S. Li, H. Wang, C. Zhang, Genome-wide loss of heterozygosity analyses in primary hepatocellular carcinoma, Zhonghua Yi Xue Za Zhi 80 (2000) 577–581. [27] S.H. Yeh, M.W. Lin, S.F. Lu, D.C. Wu, S.F. Tsai, C.Y. Tsai, et al., Allelic loss of chromosome 4q21 approximately 23 associates with hepatitis B virus-related hepatocarcinogenesis and elevated alpha-fetoprotein, Hepatology 40 (2004) 847–854. [28] D.H. Shen, K.Y. Chan, U.S. Khoo, H.Y. Ngan, W.C. Xue, P.M. Chiu, et al., Epigenetic and genetic alterations of p33ING1b in ovarian cancer, Carcinogenesis 26 (2005) 855– 863. [29] M. Takahashi, T. Ozaki, S. Todo, A. Nakagawara, Decreased expression of the candidate tumor suppressor gene ING1 is associated with poor prognosis in advanced neuroblastomas, Oncol. Rep. 12 (2004) 811–816. [30] M.L. Caruso, A.M. Valentini, Overexpression of p53 in a large series of patients with hepatocellular carcinoma: a

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

clinicopathological correlation, Anticancer Res. 19 (1999) 3853–3856. K.C. Lee, A.J. Crowe, M.C. Barton, p53-Mediated repression of alpha-fetoprotein gene expression by specific DNA binding, Mol. Cell. Biol. 19 (1999) 1279–1288. S.K. Ogden, K.C. Lee, K. Wernke-Dollries, S.A. Stratton, B. Aronow, M.C. Barton, p53 targets chromatin structure alteration to repress alpha-fetoprotein gene expression, J. Biol. Chem. 276 (2001) 42057–42062. H. Kataoka, P. Bonnefin, D. Vieyra, X. Feng, Y. Hara, Y. Miura, et al., ING1 represses transcription by direct DNA binding and through effects on p53, Cancer Res. 63 (2003) 5785–5792. D. Vieyra, D.L. Senger, T. Toyama, H. Muzik, P.M. Brasher, R.N. Johnston, et al., Altered subcellular localization and low frequency of mutations of ING1 in human brain tumors, Clin. Cancer Res. 9 (2003) 5952–5961. G.S. Nouman, B. Angus, J. Lunec, S. Crosier, A. Lodge, J.J. Anderson, Comparative assessment expression of the inhibitor of growth 1 gene (ING1) in normal and neoplastic tissues, Hybrid. Hybridomics 21 (2002) 1–10. G.S. Nouman, J.J. Anderson, K.M. Wood, J. Lunec, A.G. Hall, M.M. Reid, et al., Loss of nuclear expression of the p33(ING1b) inhibitor of growth protein in childhood acute lymphoblastic leukaemia, J. Clin. Pathol. 55 (2002) 596–601. L.X. Qin, Z.Y. Tang, The prognostic molecular markers in hepatocellular carcinoma, World J. Gastroenterol. 8 (2002) 385–392. Y. Wang, D.L. Dai, M. Martinka, G. Li, Prognostic significance of nuclear ING3 expression in human cutaneous melanoma, Clin. Cancer Res. 13 (2007) 4111–4116. S.Y. Peng, W.J. Chen, P.L. Lai, Y.M. Jeng, J.C. Sheu, H.C. Hsu, High alpha-fetoprotein level correlates with high stage, early recurrence and poor prognosis of hepatocellular carcinoma: significance of hepatitis virus infection, age, p53 and beta-catenin mutations, Int. J. Cancer 112 (2004) 44–50. K. Okuda, M. Tanaka, N. Kanazawa, J. Nagashima, S. Satomura, H. Kinoshita, et al., Evaluation of curability and prediction of prognosis after surgical treatment for hepatocellular carcinoma by lens culinaris agglutinin-reactive alpha-fetoprotein, Int. J. Oncol. 14 (1999) 265–271. B. Bressac, M. Kew, J. Wands, M. Ozturk, Selective G to T mutations of p53 gene in hepatocellular carcinoma from southern Africa, Nature 350 (1991) 429–431. I.C. Hsu, R.A. Metcalf, T. Sun, J.A. Welsh, N.J. Wang, C.C. Harris, Mutational hotspot in the p53 gene in human hepatocellular carcinomas, Nature 350 (1991) 427–428. M.Y. Chin, K.C. Ng, G. Li, The novel tumor suppressor p33ING2 enhances UVB-induced apoptosis in human melanoma cells, Exp. Cell Res. 304 (2005) 531–543. W.J. Meehan, R.S. Samant, J.E. Hopper, M.J. Carrozza, L.A. Shevde, J.L. Workman, et al., Breast cancer metastasis suppressor 1 (BRMS1) forms complexes with retinoblastoma-binding protein 1 (RBP1) and the mSin3 histone deacetylase complex and represses transcription, J. Biol. Chem. 279 (2004) 1562–1569. E. Sironi, A. Cerri, D. Tomasini, S.M. Sirchia, G. Porta, F. Rossella, et al., Loss of heterozygosity on chromosome 4q32–35 in sporadic basal cell carcinomas: evidence for the involvement of p33ING2/ING1L and SAP30 genes, J. Cutan. Pathol. 31 (2004) 318–322.