INTERNATIONAL JOURNAL OF IMMUNOPATHOLOGY AND PHARMACOLOGY
Vol. 27, DO. 4, 543-551 (2014)
CLINICAL SIGNIFICANCE OF HMGBI EXPRESSION IN HUMAN GASTRIC CANCER J. ZHANG, R. ZHANG, W.W. LV, J.S. ZHV, L.Q. XIA, Y.M. LV and N.W. CHEN
Department ofGastroenterology, Shanghai Sixth People s Hospital Affiliated to Shanghai Jiao Tong University, Shanghai, China Received August 8, 2014 - Accepted October 6, 2014 High mobility group box 1 (HMGB1) has been proved to be implicated in a variety of cell physiological and pathological behaviors including immune response, inflammation and cancer. Accumulating evidence suggests that HMGBl plays a critical role in the development and progression of multiple malignancies. However, the clinical significance and prognosis of HMGBl expression in some cancers remain controversial. The present study aimed to investigate whether overexpression of HMGBl is an independent prognostic factor in patients with gastric cancer. The correlation ofHMGBl expression with clinicopathologic characteristics and prognosis was assessed by immunohistochemical assay through tissue microarray procedure in 50 primary gastric cancer cases. Our results indicated that the positive expression of HMGBl was significantly increased in the nucleus of gastric cancer tissues compared with the adjacent non-cancerous tissues (ANCT) (64.0% vs 44.0%, P=0.025), but was not linked to the clinicopathologic features, including the TNM stage (P=0.533) and metastatic lymph node (P=0.771), in patients with gastric cancer. Kapalan-Meier and log-rank analysis demonstrated that overexpression of HMGBl did not exert significant impact on the overall survival of patients with gastric cancer (P=0.805). Furthermore, Cox regression analysis showed that high HMGBl protein expression did not represent an independent risk factor for patients with gastric cancer (P=0.677). Taken together, our findings suggest that high expression of HMGBl is not correlated with the clinicopathologic characteristics of gastric cancer, and can not serve as an independent prognostic biomarker for patients with gastric cancer. cDNAmicroarray analysis performed to compare gene expression profiles of the highly invasive and the low invasive subclones originating from ovarian cancer shows that high HMGB 1 expression is associated with poor clinicopathologic features (3). According to the cellular localization, HMGB 1, expressed in the nuclei andcytoplasm ofcancer cells aswell as the extracellular space, is correlated with lymph node metastases and prognosis in esophageal squamous cell carcinoma (4). HMGB 1 overexpression is significantly associated with tumor grade, stage, shorter disease-free survival and overall survival, representing an independent prognostic factor for patients with bladder cancer
Gastric cancer is one of the most common malignancies worldwide, with an estimated one million diagnosed each year (1). Complex genetic and physiological variations as well as environmental factors are central to the pathogenesis of human diseases and disorders and understanding this molecular foundation for tumorigenesis is crucial for searching for new diagnostic biomarkers, and identifying new therapeutic targets (2). High mobility group box I (HMGB1), a nuclear and extracellular protein, participates in a variety of physiologic and pathologic conditions, including immune response, inflammation and cancer. The
Key words: high mobility group box 1, gastric cancer, prognosis Mailing address: Prof. Jin-Shui Zhu, Department of Gastroenterology, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University, 600 Yishan Road, Shanghai 200233, China Tel.: +86-21-64369181, Fax: +86-21-64837019. e-mail: [email protected]
0394-6320 (2014)
543
Copyright © by BIOLlFE, s.a.s. This publication and/or article is for individual use only and may not be further reproduced without written permission from the copyright holder. Unauthorized reproduction may result in financial and other penalties
DISCLOSURE: ALL AUTHORS REPORT NO CONFLICTS OF INTEREST RELEVANT TO THIS ARTICLE.
J. ZHANG ET AL.
544
(5). Functionally, HMGB 1 is required for autophagy, mitochondrial quality control, the regulation of gene expression, and organ function in the adult organism (6). Knockdown of HMGBI inhibits proliferation, migration and invasion of cancer cells through the VEGF-C (4) and NF-KB pathway (7), whereas ectopic expression of HMGB 1 promotes cell growth by suppressing BTB-induced cell death (8) and cell migration by epigenetic silencing of semaphorin 3A (9), decreases Bax and p53 expression and enhances the expression of Bcl-xL, Bcl-2, CyclinDl and NFKB via activation of the FAK/PI3K/mTOR pathway (8). In addition, extracellular HMGB 1 subsequently enhances resistance to the P-gp-related drugs (adriamycin and vincristine) (10), while knockdown of HMGB 1 restores chemosensitivity and tumor cell death (11). However, few studies show that the expression of HMGBI is reduced in lung, lymph node and spleen tumor tissues (12), and negatively associates with the infiltration of tumor cells and prognosis in patients with advanced colon cancer (13). HMGBI functions as a tumor suppressor and radiosensitizer in breast cancer (14). To elucidate the clinical significance of HMGB 1 in cancer, in the present study, we examined the expression of HMGB 1 in primary gastric cancer tissues and determined whether overexpression of HMGB 1 was correlated with the clinicopathologic parameters and prognosis of patients with gastric cancer. MATERIALS AND METHODS Materials Human gastric cancer tissues and the corresponding ANCT were collected from the Department of gastrointestinal surgery of Shanghai Sixth People's Hospital. The tissue microarray of gastric cancer was made by Shanghai Outdo Biotech Co. Ltd (China). The antibody of HMGB 1 was purchased from Cell Signaling Technologies (USA). Clinical samples and data Tissue microarray was prepared for IHC test. Human gastric cancer tissues and the corresponding ANCT were obtained from biopsy in a total of 50 consecutive gastric cancer cases admitted to our hospital from January 2007 to December 2011.The baseline characteristics ofthe patients before neo-adjuvant chemotherapy are summarized in Table I. The study was approved by the Medical Ethics
Committee of Shanghai Jiao Tong University and written informed consent was obtained from the patients or their parents before sample collection. Two pathologists respectively reviewed all of the cases. Tissue microarray Advanced tissue arrayer (ATA-IOO, Chemicon International, Tamecula, CA, USA) was used to create holes in a recipient paraffin block and to acquire cylindrical core tissue biopsies with a diameter of 1 mm from the specific areas of the "donor" block. The tissue core biopsies were transferred to the recipient paraffin block at defined array positions. The tissue microarrays contained tissue samples from 50 formalin-fixed paraffinembedded cancer specimens with known diagnosis, and corresponding ANCT from these patients. The block was incubated in an oven at 45°C for 20 min to allow complete embedding of the grafted tissue cylinders in the paraffin ofthe recipient block, and then stored at 4°C until microtome sectioning. Immunohistochemical staining HMGBI antibody was used for IRC detection of protein expression in tissue microarrays. HMGBI antibody was used at 1:100 dilutions. Endogenous peroxidase was inhibited by incubation with freshly prepared 3% hydrogen peroxide with 0.1% sodium azide. Non-specific staining was blocked with 0.5% casein and 5% normal serum. Tissue microarrays were incubated with biotinylated antibodies and horseradish peroxidase. Staining was developed with diaminobenzidine substrate and sections were counterstained with hematoxylin. PBS replaced HMGB1 antibody in negative controls. Quantification ofprotein expression The expression of HMGB1 was semiquantitatively estimated as the total immunostaining scores, which were calculated as the product of a proportion score and an intensity score. The proportion and intensity of the staining was evaluated independently by two observers. The proportion score reflected the fraction of positive staining cells (0, none; 1,::;10%; 2,10% to 2:25%; 3, >25% to 50%; 4, >50%), and the intensity score represented the staining intensity (0, no staining; 1, weak; 2, intermediate; 3, strong). Finally, a total expression score was given ranging from 0 to 12. Based on the analysis, HMGBI expression was categorized in advance into two groups: low-level HMGBI expression (score 0-3) and highlevel HMGBI expression (score 4-12). The scoring was independently assessed by twopathologists. Statistical analysis Kruskal-Wallis H test and Chi-squared test were used
Int. J. Immunopathol. Pharmacol.
to analyze the expression rate in all groups . Survival curves were plotted by the Kaplan-Meier method and compared using the log-rank test. The Kaplan-Meier method was to analyze the association of HMGB I expression with the prognosis of patients with gastric cancer. Conventional clinicopathological factors and HMGBI protein level were assessed by Cox 's univariate hazard regression model. All statistical analyses were performed using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA) for Microsoft Windows. Probability values less than 0.05 were considered to be statistically significant.
RESULTS
Expression ofHMGBl in gastric cancer tissues The expression ofHMGBl protein was evaluated using IHC staining in gastric cancer tissues. As shown in Figs. 1-2, different levels of positive expression of HMGB 1 protein were detected in gastric cancer and ANCT tissues. Positive HMGB 1 immunostaining was mainly localized in the nucleus of gastric cancer tissue cells. According to the HMGB 1 immunoreactive intensity, the positive expression of
545
HMGB 1 in gastric cancer tissues was significantly increased compared with that of HMGB 1 in ANCT (P==O.025) (Table II).
Correlation of HMGBl expression with clinicopathologic parameters According to the HMGB I immunoreactive intensity, the patients were classified as low HMGB 1 group and high HMGB 1 group. We then analyzed the association between HMGB 1 expression and the clinicopathological data of the patients with gastric cancer. As summarized in Table III, we found that there was no significant correlation between HMGB I protein expression and the clinicopathologic factors including age, gender, tumor size , cancer TNM stage , depth of invasion, and metastatic lymph node (each P>O.05). Association ofHMGBl expression with the prognosis ofgastric cancer After surgery, all patients were followed-up for overall survival, and the mean duration of the
Fig. 1. The exp ression of HMGBJ protein in gastric cancer (magnification x ]OO and x400). Tissue sections were immunohistochemi cally stained with an anti-HMGBJ antibody and scored as (A) 0, (B) ] +, (C) 2+ and (D) 3+.
J. ZHANG ET AL.
546
Fig. 2. The expression of HMGB1 protein in adjacent non-cancerous tissues (magnification x 100 and x400) . Tissue section s were immunohistochemically stained with an anti-HMGB1 antibody and scored as (A) 0, (B) 1+, (C) 2+ and (D) 3+. 1.0
..r .sr:
,-.. ~ ~
group],
group2
0.8
!l
e
-a 0.6 ;>
.~
en
0.4
]o~ 0.2 1
o o
10
20
30
40
50
60
Months after operation (m)
Fig. 3. Kaplan-Meier analysis of overall survival rate ofgastric cancer patients with HMGB1 high expression (group 1) and HMGB1 low exp ression (group 2). The overall survival rate between these two groups showed no significant difference (P=0.805).
Int. J. Immunopathol. Pharmacal,
Table I. The baseline clinicopathologic characteristics.
Variables Number of patients Age :560 >60 Gender Male Female Tumor size (em) :53.5 > 3.5 Cancer TNM grade 1+11 III+IV Depth ofinvasion Mucosa/Submucosa Muscularis Serosa Serosal outer layer Metastatic lymph node Negative Positive Tstage Tl+T2 T3+T4 Nstage NO+Nl N2+N3
Number of cases (%) 50 (100%)
547
There was no need to perform the multivariate analysis. These results revealed that HMGB 1 protein expression was not an independent prognostic factor for overall survival of patients with gastric cancer (P=0.677, Table IV). DISCUSSION
16 (32.0%) 34 (68.0%) 35 (70.0%) 15 (30.0%) 9 (18.0%) 41 (82.0%) 14 (28.0%) 36 (72.0%) 4 (8.0%) 12 (24.0%) 27 (54.0%) 7 (14.0%) 32 (64.0%) 18 (36.0%) 7 (14.0%) 43 (86.0%) 19 (38.0%) 31 (62.0%)
follow-up period for 33 survivors was 48.12 months (range: 42-51 months). We investigated that there was no significant difference between the survival of patients with high HMGB 1 protein expression and that of patients with low HMGB 1 protein expression (P=0.805; Fig. 3). Univariate analysis indicated that all the clinicopathologic parameters including HMGB 1 expression were not associated with overall survival of patients with gastric cancer (P>0.05, Table IV).
HMGB 1 is an evolutionarily ancient and critical regulator of cell death and survival. As a chromatinassociated nuclear protein molecule, it has been reported in many types of cancers. HMGB1, expressed in 55.6% of the colorectal cancer (CRC) patients, is significantly correlated with lymph node metastases and pathological stage (15). HMGBI expression is elevated in thymic epithelial tumors involved in the pathology and physiology of thymus (16). High expression ofHMGB 1 activates Tregs and Th2 polarization, promoting the immune evasion and the recurrence of HPV infection in cervical cancer (17). These findings indicate that HMGB 1 may be a useful biomarker for the diagnosis and therapy of some cancers. Intriguingly, the expression of HMGB 1 is increased in cancer cells with the intestinal type and positively correlated with the degree of macrophage infiltration inside the tumor microenvironment (18). In the present study, we also found that the positive expression of HMGB 1 was significantly increased in the nucleus of gastric cancer cells compared with the adjacent non-cancerous tissues. Differently from these studies (15, 18), HMGBI was found not correlated with the clinicopathologic features, including the TNM stage and metastatic lymph node, in patients with gastric cancer, suggesting that HMGB 1might be not implicated in the tumorigenesis of gastric cancer. However, in terms of the important function of HMGB 1 in cancer, some studies have shown that knockdown of HMGB 1 expression inhibits the growth and metastasis of gastric cancer cells via the AKTandNF-KBpathways(19,20).HMGBI released from gastric cancer cells induces the expression of pro-inflammatory cytokines in peritoneal fibroblasts through the HMGBI-TLR2/4 pathway (21). HMGB l-RAGE/PI3K-Akt/ErkIl2 pathways contribute to the proliferation and apoptosis arrest, indicating a new target for intervention and treatment
J. ZHANG ET AL.
548
Table II. The expression ofHMGBl protein in gastric cancer. Variables
Group
Cases (n)
Expression levels (n)
Positive rate
P
(% )
HMGBI
+
++
+++
GC
50
18
15
11
6
64.0
ANCT
50
28
13
7
2
44.0
4.990
0.025
GC: gastric cancer; ANCT: adjacent non-cancerous tissue
Table III. Association between HMGBl expression and clinicopathologic characteristics.
Variables
Cases (n)
HMGB 1 expression (n) Low High
"l
p
0.226
0.635
Age ~60
>60
16 34
5 13
11 21
35 15
11
24
7
8
1.037
0.309
9 41
5 13
4 28
1.785
0181
14
6 12
8 24
0.389
0533
1
3 7 18 4
0.591
0.898
0.085
0.771
Gender
Male Female Tumor size (em) ~3.5
> 3.5 Cancer TNM grade 1+11 1I1+IV Depth ofinvasion
36
Mucosa/Submucosa Muscularis Serosa Serosal outer layer
4 12 27 7
5 9 3
32
12
Positive Tstage
18
6
20 12
n+T2 T3+T4 Nstage NO+Nl N2+N3
7
3 15
4 28
0.163
0.687
10 8
9 23
3.606
0.058
Metastatic lymph node
Negative
43 19 31
Int. J. Irnrnnnopathol. PharrnacoL
549
Table IV. Univariate Cox regression analysis ofoverall survival.
Parameter P
Univariate analysis HR 95%CI
Age
:::;60 vs> 60 Gender Male vs Female Tumor size (em)
0.689
0.160
0.963 - 1.059
0.317
1.002
0.652 - 3.744
:::;3.5 vs > 3.5 Pathological grade
0.850
0.036
0.439 - 2.718
1+11 vs III+IV Depth ofinvasion Mucosa/Submucosa vs Muscularis vs Serosa vs Serosal outer layer Metastatic lymph node Negative vs Positive Tstage T1+ T2 vs T3+ T4 Nstage NO+Nl vs N2+N3 HMGBI expression High vs Low
0.850
0.196
0.466 - 3.352
0.400
0.708
0.052
3.761
0.988 - 10.808
0.859
0.031
0.362 - 2.337
0.234
1.415
0.229 - 1.435
0.677
0.173
0.399 - 1.817
of some cancers (22). Therefore, in contrast with these studies, ours had several disadvantages, which were summarized as follows: the relatively small number of analyzed patients weakens the power to detect the statistical associations between HMGB 1 expression and clinicopathologic features of patients with gastric cancer. In addition, IHC is a means of semi-quantitative detection of protein expression levels in tumor cells, lacking the high quantitative accuracy in examining the HMGBI expression compared to the Real-time PCR and Western blot assays. New experimental assays need be applied to a large number of patients with gastric cancer to verify the present study. Furthermore, few studies have been focused on the prognostic value of HMGBI expression in some cancers. It is shown thatHMGB 1not only contributesto the malignant progression ofhepatocellular carcinoma (23), pancreatic cancer (24) and prostate cancer (25),
0.087 - 2.646
but also serves as an independent prognostic factor predicting the poor prognosis for these cancers (23-25). HMGBI expression is observed in gastric cancer, and correlated with metastasis status, TNM stage and poor prognosis, revealing an independent prognostic factor (26). Differently from these studies, individual studies illustrate that overexpression of HMGBI may be a marker of good prognosis of gastric adenocarcinoma patients given curative resection combined with adjuvant chemotherapy (27). In the present study, Kaplan-Meier plots and Cox proportional hazards regression model were used to analyze the prognostic value ofHMGB 1 expression in gastric cancer, indicating that there was no significant correlation between the overall survival of patients with gastric cancer and HMGB 1 protein expression, and that HMGB 1 expression was not an independent prognostic factor for overall survival of patients with gastric cancer, which was supported by some studies
J. ZHANG ET AL.
550
(18). Nevertheless, this study had several limitations. In addition to the relatively small number of analyzed patients, data on cancer recurrences were not available in these cohorts beyond cause of mortality, and all patients did not receive the same chemotherapy regimen. In conclusion, the data from the current study have demonstrated that HMGB I is highly expressed in gastric cancer, but is not correlated with the clinicopathologic characteristics of gastric cancer, and can not serve as an independent prognostic biomarker for patients with gastric cancer. ACKNOWLEDGEMENTS This work is supported by National Nature Science Foundation of China (No. 81302093 and No. 81272752). REFERENCES 1. 2.
3.
4.
5.
6.
7.
Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J C1in2013; 63:11-30. Kang R, Chen R, Zhang Q, et al. HMGB1 in health and disease. Mol Aspects Med 2014. doi: 10.1016/j. mam.2014.05.001. Chen J, Xi B, Zhao Y, Yu Y, Zhang J, Wang C. Highmobility group protein B1 (HMGB1) is a novel biomarker for human ovarian cancer. Gynecol Oncol 2012; 126:109-17. Chuangui C, Peng T, Zhentao Y. The expression of high mobility group box 1 is associated with lymph node metastasis and poor prognosis in esophageal squamous cell carcinoma. Pathol Oncol Res 2012; 18:1021-27. Yang GL, Zhang LH, Bo 11, Huo XJ, Chen HG, Cao M, Liu DM, Huang YR. Increased expression of HMGB 1 is associated with poor prognosis in human bladder cancer. J Surg Onco12012; 106:57-61. Huebener P, Gwak GY, Pradere JP, et al. Highmobility group box 1 is dispensable for autophagy, mitochondrial quality control, and organ function in vivo. Cell Metab 2014; 19:539-47. Zhang J, Kou YB, Zhu JS, Chen WX, Li S. Knockdown ofHMGB 1 inhibits growth and invasion of gastric cancer cells through the NF-lCB pathway in vitro and in vivo. Int J Onco12014; 44:1268-76.
8.
Ko YB, Kim BR, Nam SL, Yang JB, Park SY, Rho SB. High-mobility group box 1 (HMGB1) protein regulates tumor-associated cell migration through the interaction with BTB domain. Cell Signal 2014; 26:777-83. 9. Nehil M, Paquette J, Tokuyasu T, McCormick F. High mobility group box 1 promotes tumor cell migration through epigenetic silencing of semaphorin 3A. Oncogene 2013. doi: 10.1038/onc.2013,459. 10. Yin Y, Li W, Deng M, Zhang P, Shen Q, Wang G, Tao K. Extracellular high mobility group box chromosomal protein 1 promotes drug resistance by increasing the expression of P-g1ycoprotein expression in gastric adenocarcinoma cells. Mol Med Rep 2014; 9:1439-43. 11. Liu K, Huang J, Xie M, et al. MIR34A regulates autophagy and apoptosis by targeting HMGB 1 in the retinoblastoma cell. Autophagy 2014; 10:442-52. 12. Sterenczak KA, Kleinschmidt S, Wefstaedt P, Eberle N, Hewicker-Trautwein M, Bullerdiek J, Nolte I, Murua Escobar H. Quantitative PCR and immunohistochemical analyses of HMGB 1 and RAGE expression in canine disseminated histiocytic sarcoma (malignant histiocytosis). Anticancer Res 2011; 31:1541-8. 13. Peng RQ, Wu XJ, Ding Y, et al. Co-expression of nuclear and cytoplasmic HMGB1 is inversely associated with infiltration of CD45RO+ T cells and prognosis in patients with stage IIIB colon cancer. BMC Cancer 2010; 10:496. 14. Jiao Y, Wang HC, Fan S1. Growth suppression and radiosensitivity increase by HMGB1 in breast cancer. Acta Pharmacol Sin 2007; 28:1957-67. 15. Snren D, Yildmm M, Demirpence o, Kaya V, Alikanoglu AS, Bulbuller N, Yildiz M, Sezer C. The role of high mobility group box 1 (HMGB1) in colorectal cancer. Med Sci Monit 2014; 20:530-7. 16. Moser B, Janik S, Schiefer AI, et al. Expression of RAGE and HMGB1 in thymic epithelial tumors, thymic hyperplasia and regular thymic morphology. PLoS One 2014; 9:e94118. 17. Pang X, Zhang Y, Wei H, Zhang J, Luo Q, Huang C, Zhang S. Expression and effects of high-mobility group box 1 in cervical cancer. Int J Mol Sci 2014; 15:8699-712. 18.: Akaike H, Kono K, Sugai H, Takahashi A, Mimura
Int. J. ImmunopathoL PharmacoL
19.
20.
21.
22.
K, Kawaguchi Y,Fujii H. Expression ofhigh mobility group box chromosomal protein-1 (HMGB-1) in gastriccancer. Anticancer Res 2007; 27:449-57. Zhang J, Zhu JS, Zhou Z, Chen WX, Chen NW. Inhibitory effects of ethyl pyruvate administration on human gastric cancer growth via regulation of the HMGB1-RAGE and Akt pathways in vitro and in vivo. Oncol Rep 2012; 27:1511-9. Zhang J, Kou YB, Zhu JS, Chen WX, Li S. Knockdown ofHMGB1 inhibits growth and invasion of gastric cancer cells through the NF-KB pathway in vitro and in vivo. Int J Oncol20l4; 44:1268-76. Abe A, Kuwata T, Yamauchi C, Higuchi Y, Ochiai A. High Mobility Group Box1 (HMGB1) released from cancer cells induces the expression of proinflammatory cytokines in peritoneal fibroblasts. Pathol Int 2014; 64:267-75. Xu X, Zhu H, Wang T, et al. Exogenous highmobility group box 1 inhibits apoptosis and promotes the proliferation of lewis cells via RAGE/TLR4dependent signal pathways. Scand J Immuno120l4; 79:386-94.
551
23. Xiao J, Ding Y, Huang J, et al. The association of HMGBI gene with the prognosis ofHCC. PLoS One 2014; 9:e89097. 24. Zhao CB, Bao lM, Lu YJ, Zhao T, Zhou XH, Zheng DY, Zhao sc. Co-expression of RAGE and HMGB1 is associated with cancer progression and poor patient outcome of prostate cancer. Am J Cancer Res 2014; 4:369-77. 25. Takeda T, Izumi H, Kitada S, et al. The combination of a nuclear HMGB1- positive and HMGB2negative expression is potentially associated with a shortened survival in patients with pancreatic ductal adenocarcinoma. Tumour Biol2014. [Epub ahead of print] 26. He W, Tang B, Yang D, et al. Double-positive expression of high-mobility group box 1 and vascular endothelial growth factor C indicates a poorer prognosis in gastric cancer patients. World J Surg Onco12013; 11:161. 27. Bao G, Qiao Q, Zhao H, He X. Prognostic value of HMGB1 overexpression in resectable gastric adenocarcinomas. World J Surg Onco1201O; 8:52.
Vol. 27, DO. 4, 543-551 (2014)
CLINICAL SIGNIFICANCE OF HMGBI EXPRESSION IN HUMAN GASTRIC CANCER J. ZHANG, R. ZHANG, W.W. LV, J.S. ZHV, L.Q. XIA, Y.M. LV and N.W. CHEN
Department ofGastroenterology, Shanghai Sixth People s Hospital Affiliated to Shanghai Jiao Tong University, Shanghai, China Received August 8, 2014 - Accepted October 6, 2014 High mobility group box 1 (HMGB1) has been proved to be implicated in a variety of cell physiological and pathological behaviors including immune response, inflammation and cancer. Accumulating evidence suggests that HMGBl plays a critical role in the development and progression of multiple malignancies. However, the clinical significance and prognosis of HMGBl expression in some cancers remain controversial. The present study aimed to investigate whether overexpression of HMGBl is an independent prognostic factor in patients with gastric cancer. The correlation ofHMGBl expression with clinicopathologic characteristics and prognosis was assessed by immunohistochemical assay through tissue microarray procedure in 50 primary gastric cancer cases. Our results indicated that the positive expression of HMGBl was significantly increased in the nucleus of gastric cancer tissues compared with the adjacent non-cancerous tissues (ANCT) (64.0% vs 44.0%, P=0.025), but was not linked to the clinicopathologic features, including the TNM stage (P=0.533) and metastatic lymph node (P=0.771), in patients with gastric cancer. Kapalan-Meier and log-rank analysis demonstrated that overexpression of HMGBl did not exert significant impact on the overall survival of patients with gastric cancer (P=0.805). Furthermore, Cox regression analysis showed that high HMGBl protein expression did not represent an independent risk factor for patients with gastric cancer (P=0.677). Taken together, our findings suggest that high expression of HMGBl is not correlated with the clinicopathologic characteristics of gastric cancer, and can not serve as an independent prognostic biomarker for patients with gastric cancer. cDNAmicroarray analysis performed to compare gene expression profiles of the highly invasive and the low invasive subclones originating from ovarian cancer shows that high HMGB 1 expression is associated with poor clinicopathologic features (3). According to the cellular localization, HMGB 1, expressed in the nuclei andcytoplasm ofcancer cells aswell as the extracellular space, is correlated with lymph node metastases and prognosis in esophageal squamous cell carcinoma (4). HMGB 1 overexpression is significantly associated with tumor grade, stage, shorter disease-free survival and overall survival, representing an independent prognostic factor for patients with bladder cancer
Gastric cancer is one of the most common malignancies worldwide, with an estimated one million diagnosed each year (1). Complex genetic and physiological variations as well as environmental factors are central to the pathogenesis of human diseases and disorders and understanding this molecular foundation for tumorigenesis is crucial for searching for new diagnostic biomarkers, and identifying new therapeutic targets (2). High mobility group box I (HMGB1), a nuclear and extracellular protein, participates in a variety of physiologic and pathologic conditions, including immune response, inflammation and cancer. The
Key words: high mobility group box 1, gastric cancer, prognosis Mailing address: Prof. Jin-Shui Zhu, Department of Gastroenterology, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University, 600 Yishan Road, Shanghai 200233, China Tel.: +86-21-64369181, Fax: +86-21-64837019. e-mail: [email protected]
0394-6320 (2014)
543
Copyright © by BIOLlFE, s.a.s. This publication and/or article is for individual use only and may not be further reproduced without written permission from the copyright holder. Unauthorized reproduction may result in financial and other penalties
DISCLOSURE: ALL AUTHORS REPORT NO CONFLICTS OF INTEREST RELEVANT TO THIS ARTICLE.
J. ZHANG ET AL.
544
(5). Functionally, HMGB 1 is required for autophagy, mitochondrial quality control, the regulation of gene expression, and organ function in the adult organism (6). Knockdown of HMGBI inhibits proliferation, migration and invasion of cancer cells through the VEGF-C (4) and NF-KB pathway (7), whereas ectopic expression of HMGB 1 promotes cell growth by suppressing BTB-induced cell death (8) and cell migration by epigenetic silencing of semaphorin 3A (9), decreases Bax and p53 expression and enhances the expression of Bcl-xL, Bcl-2, CyclinDl and NFKB via activation of the FAK/PI3K/mTOR pathway (8). In addition, extracellular HMGB 1 subsequently enhances resistance to the P-gp-related drugs (adriamycin and vincristine) (10), while knockdown of HMGB 1 restores chemosensitivity and tumor cell death (11). However, few studies show that the expression of HMGBI is reduced in lung, lymph node and spleen tumor tissues (12), and negatively associates with the infiltration of tumor cells and prognosis in patients with advanced colon cancer (13). HMGBI functions as a tumor suppressor and radiosensitizer in breast cancer (14). To elucidate the clinical significance of HMGB 1 in cancer, in the present study, we examined the expression of HMGB 1 in primary gastric cancer tissues and determined whether overexpression of HMGB 1 was correlated with the clinicopathologic parameters and prognosis of patients with gastric cancer. MATERIALS AND METHODS Materials Human gastric cancer tissues and the corresponding ANCT were collected from the Department of gastrointestinal surgery of Shanghai Sixth People's Hospital. The tissue microarray of gastric cancer was made by Shanghai Outdo Biotech Co. Ltd (China). The antibody of HMGB 1 was purchased from Cell Signaling Technologies (USA). Clinical samples and data Tissue microarray was prepared for IHC test. Human gastric cancer tissues and the corresponding ANCT were obtained from biopsy in a total of 50 consecutive gastric cancer cases admitted to our hospital from January 2007 to December 2011.The baseline characteristics ofthe patients before neo-adjuvant chemotherapy are summarized in Table I. The study was approved by the Medical Ethics
Committee of Shanghai Jiao Tong University and written informed consent was obtained from the patients or their parents before sample collection. Two pathologists respectively reviewed all of the cases. Tissue microarray Advanced tissue arrayer (ATA-IOO, Chemicon International, Tamecula, CA, USA) was used to create holes in a recipient paraffin block and to acquire cylindrical core tissue biopsies with a diameter of 1 mm from the specific areas of the "donor" block. The tissue core biopsies were transferred to the recipient paraffin block at defined array positions. The tissue microarrays contained tissue samples from 50 formalin-fixed paraffinembedded cancer specimens with known diagnosis, and corresponding ANCT from these patients. The block was incubated in an oven at 45°C for 20 min to allow complete embedding of the grafted tissue cylinders in the paraffin ofthe recipient block, and then stored at 4°C until microtome sectioning. Immunohistochemical staining HMGBI antibody was used for IRC detection of protein expression in tissue microarrays. HMGBI antibody was used at 1:100 dilutions. Endogenous peroxidase was inhibited by incubation with freshly prepared 3% hydrogen peroxide with 0.1% sodium azide. Non-specific staining was blocked with 0.5% casein and 5% normal serum. Tissue microarrays were incubated with biotinylated antibodies and horseradish peroxidase. Staining was developed with diaminobenzidine substrate and sections were counterstained with hematoxylin. PBS replaced HMGB1 antibody in negative controls. Quantification ofprotein expression The expression of HMGB1 was semiquantitatively estimated as the total immunostaining scores, which were calculated as the product of a proportion score and an intensity score. The proportion and intensity of the staining was evaluated independently by two observers. The proportion score reflected the fraction of positive staining cells (0, none; 1,::;10%; 2,10% to 2:25%; 3, >25% to 50%; 4, >50%), and the intensity score represented the staining intensity (0, no staining; 1, weak; 2, intermediate; 3, strong). Finally, a total expression score was given ranging from 0 to 12. Based on the analysis, HMGBI expression was categorized in advance into two groups: low-level HMGBI expression (score 0-3) and highlevel HMGBI expression (score 4-12). The scoring was independently assessed by twopathologists. Statistical analysis Kruskal-Wallis H test and Chi-squared test were used
Int. J. Immunopathol. Pharmacol.
to analyze the expression rate in all groups . Survival curves were plotted by the Kaplan-Meier method and compared using the log-rank test. The Kaplan-Meier method was to analyze the association of HMGB I expression with the prognosis of patients with gastric cancer. Conventional clinicopathological factors and HMGBI protein level were assessed by Cox 's univariate hazard regression model. All statistical analyses were performed using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA) for Microsoft Windows. Probability values less than 0.05 were considered to be statistically significant.
RESULTS
Expression ofHMGBl in gastric cancer tissues The expression ofHMGBl protein was evaluated using IHC staining in gastric cancer tissues. As shown in Figs. 1-2, different levels of positive expression of HMGB 1 protein were detected in gastric cancer and ANCT tissues. Positive HMGB 1 immunostaining was mainly localized in the nucleus of gastric cancer tissue cells. According to the HMGB 1 immunoreactive intensity, the positive expression of
545
HMGB 1 in gastric cancer tissues was significantly increased compared with that of HMGB 1 in ANCT (P==O.025) (Table II).
Correlation of HMGBl expression with clinicopathologic parameters According to the HMGB I immunoreactive intensity, the patients were classified as low HMGB 1 group and high HMGB 1 group. We then analyzed the association between HMGB 1 expression and the clinicopathological data of the patients with gastric cancer. As summarized in Table III, we found that there was no significant correlation between HMGB I protein expression and the clinicopathologic factors including age, gender, tumor size , cancer TNM stage , depth of invasion, and metastatic lymph node (each P>O.05). Association ofHMGBl expression with the prognosis ofgastric cancer After surgery, all patients were followed-up for overall survival, and the mean duration of the
Fig. 1. The exp ression of HMGBJ protein in gastric cancer (magnification x ]OO and x400). Tissue sections were immunohistochemi cally stained with an anti-HMGBJ antibody and scored as (A) 0, (B) ] +, (C) 2+ and (D) 3+.
J. ZHANG ET AL.
546
Fig. 2. The expression of HMGB1 protein in adjacent non-cancerous tissues (magnification x 100 and x400) . Tissue section s were immunohistochemically stained with an anti-HMGB1 antibody and scored as (A) 0, (B) 1+, (C) 2+ and (D) 3+. 1.0
..r .sr:
,-.. ~ ~
group],
group2
0.8
!l
e
-a 0.6 ;>
.~
en
0.4
]o~ 0.2 1
o o
10
20
30
40
50
60
Months after operation (m)
Fig. 3. Kaplan-Meier analysis of overall survival rate ofgastric cancer patients with HMGB1 high expression (group 1) and HMGB1 low exp ression (group 2). The overall survival rate between these two groups showed no significant difference (P=0.805).
Int. J. Immunopathol. Pharmacal,
Table I. The baseline clinicopathologic characteristics.
Variables Number of patients Age :560 >60 Gender Male Female Tumor size (em) :53.5 > 3.5 Cancer TNM grade 1+11 III+IV Depth ofinvasion Mucosa/Submucosa Muscularis Serosa Serosal outer layer Metastatic lymph node Negative Positive Tstage Tl+T2 T3+T4 Nstage NO+Nl N2+N3
Number of cases (%) 50 (100%)
547
There was no need to perform the multivariate analysis. These results revealed that HMGB 1 protein expression was not an independent prognostic factor for overall survival of patients with gastric cancer (P=0.677, Table IV). DISCUSSION
16 (32.0%) 34 (68.0%) 35 (70.0%) 15 (30.0%) 9 (18.0%) 41 (82.0%) 14 (28.0%) 36 (72.0%) 4 (8.0%) 12 (24.0%) 27 (54.0%) 7 (14.0%) 32 (64.0%) 18 (36.0%) 7 (14.0%) 43 (86.0%) 19 (38.0%) 31 (62.0%)
follow-up period for 33 survivors was 48.12 months (range: 42-51 months). We investigated that there was no significant difference between the survival of patients with high HMGB 1 protein expression and that of patients with low HMGB 1 protein expression (P=0.805; Fig. 3). Univariate analysis indicated that all the clinicopathologic parameters including HMGB 1 expression were not associated with overall survival of patients with gastric cancer (P>0.05, Table IV).
HMGB 1 is an evolutionarily ancient and critical regulator of cell death and survival. As a chromatinassociated nuclear protein molecule, it has been reported in many types of cancers. HMGB1, expressed in 55.6% of the colorectal cancer (CRC) patients, is significantly correlated with lymph node metastases and pathological stage (15). HMGBI expression is elevated in thymic epithelial tumors involved in the pathology and physiology of thymus (16). High expression ofHMGB 1 activates Tregs and Th2 polarization, promoting the immune evasion and the recurrence of HPV infection in cervical cancer (17). These findings indicate that HMGB 1 may be a useful biomarker for the diagnosis and therapy of some cancers. Intriguingly, the expression of HMGB 1 is increased in cancer cells with the intestinal type and positively correlated with the degree of macrophage infiltration inside the tumor microenvironment (18). In the present study, we also found that the positive expression of HMGB 1 was significantly increased in the nucleus of gastric cancer cells compared with the adjacent non-cancerous tissues. Differently from these studies (15, 18), HMGBI was found not correlated with the clinicopathologic features, including the TNM stage and metastatic lymph node, in patients with gastric cancer, suggesting that HMGB 1might be not implicated in the tumorigenesis of gastric cancer. However, in terms of the important function of HMGB 1 in cancer, some studies have shown that knockdown of HMGB 1 expression inhibits the growth and metastasis of gastric cancer cells via the AKTandNF-KBpathways(19,20).HMGBI released from gastric cancer cells induces the expression of pro-inflammatory cytokines in peritoneal fibroblasts through the HMGBI-TLR2/4 pathway (21). HMGB l-RAGE/PI3K-Akt/ErkIl2 pathways contribute to the proliferation and apoptosis arrest, indicating a new target for intervention and treatment
J. ZHANG ET AL.
548
Table II. The expression ofHMGBl protein in gastric cancer. Variables
Group
Cases (n)
Expression levels (n)
Positive rate
P
(% )
HMGBI
+
++
+++
GC
50
18
15
11
6
64.0
ANCT
50
28
13
7
2
44.0
4.990
0.025
GC: gastric cancer; ANCT: adjacent non-cancerous tissue
Table III. Association between HMGBl expression and clinicopathologic characteristics.
Variables
Cases (n)
HMGB 1 expression (n) Low High
"l
p
0.226
0.635
Age ~60
>60
16 34
5 13
11 21
35 15
11
24
7
8
1.037
0.309
9 41
5 13
4 28
1.785
0181
14
6 12
8 24
0.389
0533
1
3 7 18 4
0.591
0.898
0.085
0.771
Gender
Male Female Tumor size (em) ~3.5
> 3.5 Cancer TNM grade 1+11 1I1+IV Depth ofinvasion
36
Mucosa/Submucosa Muscularis Serosa Serosal outer layer
4 12 27 7
5 9 3
32
12
Positive Tstage
18
6
20 12
n+T2 T3+T4 Nstage NO+Nl N2+N3
7
3 15
4 28
0.163
0.687
10 8
9 23
3.606
0.058
Metastatic lymph node
Negative
43 19 31
Int. J. Irnrnnnopathol. PharrnacoL
549
Table IV. Univariate Cox regression analysis ofoverall survival.
Parameter P
Univariate analysis HR 95%CI
Age
:::;60 vs> 60 Gender Male vs Female Tumor size (em)
0.689
0.160
0.963 - 1.059
0.317
1.002
0.652 - 3.744
:::;3.5 vs > 3.5 Pathological grade
0.850
0.036
0.439 - 2.718
1+11 vs III+IV Depth ofinvasion Mucosa/Submucosa vs Muscularis vs Serosa vs Serosal outer layer Metastatic lymph node Negative vs Positive Tstage T1+ T2 vs T3+ T4 Nstage NO+Nl vs N2+N3 HMGBI expression High vs Low
0.850
0.196
0.466 - 3.352
0.400
0.708
0.052
3.761
0.988 - 10.808
0.859
0.031
0.362 - 2.337
0.234
1.415
0.229 - 1.435
0.677
0.173
0.399 - 1.817
of some cancers (22). Therefore, in contrast with these studies, ours had several disadvantages, which were summarized as follows: the relatively small number of analyzed patients weakens the power to detect the statistical associations between HMGB 1 expression and clinicopathologic features of patients with gastric cancer. In addition, IHC is a means of semi-quantitative detection of protein expression levels in tumor cells, lacking the high quantitative accuracy in examining the HMGBI expression compared to the Real-time PCR and Western blot assays. New experimental assays need be applied to a large number of patients with gastric cancer to verify the present study. Furthermore, few studies have been focused on the prognostic value of HMGBI expression in some cancers. It is shown thatHMGB 1not only contributesto the malignant progression ofhepatocellular carcinoma (23), pancreatic cancer (24) and prostate cancer (25),
0.087 - 2.646
but also serves as an independent prognostic factor predicting the poor prognosis for these cancers (23-25). HMGBI expression is observed in gastric cancer, and correlated with metastasis status, TNM stage and poor prognosis, revealing an independent prognostic factor (26). Differently from these studies, individual studies illustrate that overexpression of HMGBI may be a marker of good prognosis of gastric adenocarcinoma patients given curative resection combined with adjuvant chemotherapy (27). In the present study, Kaplan-Meier plots and Cox proportional hazards regression model were used to analyze the prognostic value ofHMGB 1 expression in gastric cancer, indicating that there was no significant correlation between the overall survival of patients with gastric cancer and HMGB 1 protein expression, and that HMGB 1 expression was not an independent prognostic factor for overall survival of patients with gastric cancer, which was supported by some studies
J. ZHANG ET AL.
550
(18). Nevertheless, this study had several limitations. In addition to the relatively small number of analyzed patients, data on cancer recurrences were not available in these cohorts beyond cause of mortality, and all patients did not receive the same chemotherapy regimen. In conclusion, the data from the current study have demonstrated that HMGB I is highly expressed in gastric cancer, but is not correlated with the clinicopathologic characteristics of gastric cancer, and can not serve as an independent prognostic biomarker for patients with gastric cancer. ACKNOWLEDGEMENTS This work is supported by National Nature Science Foundation of China (No. 81302093 and No. 81272752). REFERENCES 1. 2.
3.
4.
5.
6.
7.
Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J C1in2013; 63:11-30. Kang R, Chen R, Zhang Q, et al. HMGB1 in health and disease. Mol Aspects Med 2014. doi: 10.1016/j. mam.2014.05.001. Chen J, Xi B, Zhao Y, Yu Y, Zhang J, Wang C. Highmobility group protein B1 (HMGB1) is a novel biomarker for human ovarian cancer. Gynecol Oncol 2012; 126:109-17. Chuangui C, Peng T, Zhentao Y. The expression of high mobility group box 1 is associated with lymph node metastasis and poor prognosis in esophageal squamous cell carcinoma. Pathol Oncol Res 2012; 18:1021-27. Yang GL, Zhang LH, Bo 11, Huo XJ, Chen HG, Cao M, Liu DM, Huang YR. Increased expression of HMGB 1 is associated with poor prognosis in human bladder cancer. J Surg Onco12012; 106:57-61. Huebener P, Gwak GY, Pradere JP, et al. Highmobility group box 1 is dispensable for autophagy, mitochondrial quality control, and organ function in vivo. Cell Metab 2014; 19:539-47. Zhang J, Kou YB, Zhu JS, Chen WX, Li S. Knockdown ofHMGB 1 inhibits growth and invasion of gastric cancer cells through the NF-lCB pathway in vitro and in vivo. Int J Onco12014; 44:1268-76.
8.
Ko YB, Kim BR, Nam SL, Yang JB, Park SY, Rho SB. High-mobility group box 1 (HMGB1) protein regulates tumor-associated cell migration through the interaction with BTB domain. Cell Signal 2014; 26:777-83. 9. Nehil M, Paquette J, Tokuyasu T, McCormick F. High mobility group box 1 promotes tumor cell migration through epigenetic silencing of semaphorin 3A. Oncogene 2013. doi: 10.1038/onc.2013,459. 10. Yin Y, Li W, Deng M, Zhang P, Shen Q, Wang G, Tao K. Extracellular high mobility group box chromosomal protein 1 promotes drug resistance by increasing the expression of P-g1ycoprotein expression in gastric adenocarcinoma cells. Mol Med Rep 2014; 9:1439-43. 11. Liu K, Huang J, Xie M, et al. MIR34A regulates autophagy and apoptosis by targeting HMGB 1 in the retinoblastoma cell. Autophagy 2014; 10:442-52. 12. Sterenczak KA, Kleinschmidt S, Wefstaedt P, Eberle N, Hewicker-Trautwein M, Bullerdiek J, Nolte I, Murua Escobar H. Quantitative PCR and immunohistochemical analyses of HMGB 1 and RAGE expression in canine disseminated histiocytic sarcoma (malignant histiocytosis). Anticancer Res 2011; 31:1541-8. 13. Peng RQ, Wu XJ, Ding Y, et al. Co-expression of nuclear and cytoplasmic HMGB1 is inversely associated with infiltration of CD45RO+ T cells and prognosis in patients with stage IIIB colon cancer. BMC Cancer 2010; 10:496. 14. Jiao Y, Wang HC, Fan S1. Growth suppression and radiosensitivity increase by HMGB1 in breast cancer. Acta Pharmacol Sin 2007; 28:1957-67. 15. Snren D, Yildmm M, Demirpence o, Kaya V, Alikanoglu AS, Bulbuller N, Yildiz M, Sezer C. The role of high mobility group box 1 (HMGB1) in colorectal cancer. Med Sci Monit 2014; 20:530-7. 16. Moser B, Janik S, Schiefer AI, et al. Expression of RAGE and HMGB1 in thymic epithelial tumors, thymic hyperplasia and regular thymic morphology. PLoS One 2014; 9:e94118. 17. Pang X, Zhang Y, Wei H, Zhang J, Luo Q, Huang C, Zhang S. Expression and effects of high-mobility group box 1 in cervical cancer. Int J Mol Sci 2014; 15:8699-712. 18.: Akaike H, Kono K, Sugai H, Takahashi A, Mimura
Int. J. ImmunopathoL PharmacoL
19.
20.
21.
22.
K, Kawaguchi Y,Fujii H. Expression ofhigh mobility group box chromosomal protein-1 (HMGB-1) in gastriccancer. Anticancer Res 2007; 27:449-57. Zhang J, Zhu JS, Zhou Z, Chen WX, Chen NW. Inhibitory effects of ethyl pyruvate administration on human gastric cancer growth via regulation of the HMGB1-RAGE and Akt pathways in vitro and in vivo. Oncol Rep 2012; 27:1511-9. Zhang J, Kou YB, Zhu JS, Chen WX, Li S. Knockdown ofHMGB1 inhibits growth and invasion of gastric cancer cells through the NF-KB pathway in vitro and in vivo. Int J Oncol20l4; 44:1268-76. Abe A, Kuwata T, Yamauchi C, Higuchi Y, Ochiai A. High Mobility Group Box1 (HMGB1) released from cancer cells induces the expression of proinflammatory cytokines in peritoneal fibroblasts. Pathol Int 2014; 64:267-75. Xu X, Zhu H, Wang T, et al. Exogenous highmobility group box 1 inhibits apoptosis and promotes the proliferation of lewis cells via RAGE/TLR4dependent signal pathways. Scand J Immuno120l4; 79:386-94.
551
23. Xiao J, Ding Y, Huang J, et al. The association of HMGBI gene with the prognosis ofHCC. PLoS One 2014; 9:e89097. 24. Zhao CB, Bao lM, Lu YJ, Zhao T, Zhou XH, Zheng DY, Zhao sc. Co-expression of RAGE and HMGB1 is associated with cancer progression and poor patient outcome of prostate cancer. Am J Cancer Res 2014; 4:369-77. 25. Takeda T, Izumi H, Kitada S, et al. The combination of a nuclear HMGB1- positive and HMGB2negative expression is potentially associated with a shortened survival in patients with pancreatic ductal adenocarcinoma. Tumour Biol2014. [Epub ahead of print] 26. He W, Tang B, Yang D, et al. Double-positive expression of high-mobility group box 1 and vascular endothelial growth factor C indicates a poorer prognosis in gastric cancer patients. World J Surg Onco12013; 11:161. 27. Bao G, Qiao Q, Zhao H, He X. Prognostic value of HMGB1 overexpression in resectable gastric adenocarcinomas. World J Surg Onco1201O; 8:52.
