Med Oncol (2011) 28:481–487 DOI 10.1007/s12032-010-9482-7

ORIGINAL PAPER

Reduced selenium-binding protein 1 is associated with poor survival rate in gastric carcinoma Jin Zhang • Wei-guo Dong • Jun Lin

Received: 24 January 2010 / Accepted: 4 March 2010 / Published online: 31 March 2010 Ó Springer Science+Business Media, LLC 2010

Abstract Human selenium-binding protein 1 (SBP1) is known to play a key role in the development and progression of many cancers. The role of SBP1 expression in gastric carcinoma (GC) is far from being fully established. The aim of the present study was to evaluate the expression of SBP1 in GC and correlate the findings with several clinicopathological features and prognosis. Tissue samples from 65 patients treated by gastric resection for GC with clinical stage II and III were used. Each sample was matched with the corresponding nonneoplastic epithelia tissues removed during the same surgery. Reverse transcriptasepolymerase chain reaction, immunostaining and Western blot analyses were used to detect the expression of SBP1 at the mRNA and protein levels, respectively. The associations between SBP1 expressions and clinicopathological features were analyzed. Expressions of SBP1 at both mRNA and protein levels were significantly lower in GC than those in the corresponding nonneoplastic epithelia tissues (P = 0.000). SBP1-negative expression had a significant relationship with high clinical stage (P = 0.038). Prognosis of SBP1-negative patients was significantly poorer than that of SBP1-positive patients (P = 0.001), and multivariate analysis further confirmed that SBP1 was an independent prognostic factor (P = 0.004). Thus, downregulation of SBP1 may play a key role in the tumorigenic

J. Zhang  W. Dong Department of Gastroenterology, Renmin Hospital of Wuhan University, 9 # Ziyang Road, 430060 Wuhan, Hubei Province, China J. Lin (&) Department of Gastroenterology, Zhongnan Hospital of Wuhan University, 169 # East Lake Road, 430071 Wuhan, Hubei Province, China e-mail: [email protected]

process of human GC. The correlation of SBP1 reduction in GC with clinical stage and survival proposes a prognostic role in GC. Keywords Gastric carcinoma  Selenium-binding protein 1  Prognosis

Introduction Nutrients and food have been related to malignant tumors of the gastrointestinal tract in many ways [1–3]. Epidemiologic studies have discovered that many cancers are all related to dietary selenium (75Se) deficiency [4, 5]. Many investigations have demonstrated that selenium can inhibit tumor cell growth and tumorigenesis [6, 7]. In mammals, 75Se is an essential nutrient. It is incorporated in the selenoaminoacid, Se–Cys which is required for the translation of several proteins involved in cell defense or hormone regulation [8, 9] and can as well be bound by binding proteins such as SBP1 [10] in a mode that is distinct from conventional selenocysteine-containing selenoproteins [11]. Although a physiological function of SBP1 was not clear at present, the abnormal suppression of SBP1 has been reported in several tumor types and was proposed to be a prognosis marker of many cancers [12–15]. SBP1 was one of the diverse proteomic alterations in gastric carcinoma (GC) analyzed by two-dimensional gel electrophoresis [16–18]. Nevertheless, there have been few reports about the relationship between gastric carcinoma and SBP1. The specific role of SBP1 in prognosis is unknown. In the present study, we systemically investigated the expression of SBP1 in GC and corresponding nonneoplastic epithelia tissues at both protein and mRNA levels. We further analyzed the relation of SBP1 with clinicopathological data of

123

482

patients with GC and attempted to determine whether SBP1 is prognostic of survival in patients with curatively resected GC.

Materials and methods Tissue samples The fresh tissues were obtained from 65 stage II–III (M0) patients with GC (women 26, men 39; median age 59.5 years; range 41–72 years) who had undergone gastrectomy without any neoadjuvant treatment at Zhongnan hospital between January 2002 and December 2004. Owing to tissue availability and the possible confounding factors from stage IV such as performance status, we only included stage II, III patients in the analysis. As variables possibly affect prognosis, we collected clinicopathological features including gender, age, tumor site, tumor differentiation, infiltration depth, lymph node invasion, and clinical stage determined according to the 6th edition of AJCC cancer staging manual [19]. All patients were selected at their first diagnosis; and none had received radiotherapy, chemotherapy and/or immunotherapy before gastrectomy. Patients who had palliative or noncurative operations were excluded. A curative surgical resection (subtotal or total gastrectomy) with D2 or more extensive lymph node dissection (curability class A or B, according to the Japanese Classification of Gastric Carcinoma [20]) had been performed in all patients. The samples comprised both liquid nitrogen snap-frozen specimens obtained immediately after surgical resection and paraffin blocks. Each sample was matched with the corresponding nonneoplastic epithelia tissues removed during the same surgery, usually 5–10 cm away from the border of the main tumor lesion. The diagnosis of primary GC and the corresponding nonneoplastic epithelia tissues were confirmed by two independent pathologists who were blinded to the original diagnosis. During this process, the strict criteria were used to diagnose the nonneoplastic epithelia tissue as there was no cancer, metaplasia, dysplasia, and atypical hyperplasia in them. All of the survival status was regularly evaluated checkups from the date of primary laparotomy to December 31, 2009. The mean follow-up period was 55 ± 3 months (range 8–81 months). Only the records of patients who had died of GC were considered as uncensored, patients who were alive at the end of follow-up interval or patients who died of a cause not related to GC were recorded as censored. All research involving human participants were approved by the Ethics Committee of Wuhan University. Consent was received from all patients. All clinical investigations were conducted according to the principles expressed in the Declaration of Helsinki.

123

Med Oncol (2011) 28:481–487

Immunohistochemistry analysis For immunohistochemistry, 65 pairs of formalin-fixed, paraffin-embedded GC and the corresponding nonneoplastic epithelia tissues were cut into 4-lm sections. The sections were deparaffinized in xylene and rehydrated in a series of descending ethanol concentrations. For antigen unmasking, sections were immersed in antigen-unmasking solution and boiled in a microwave oven for 15 s. Tissue sections were incubated using anti-SBP1 (1:500 dilution, Clone 4D4, MBL, Nagoya, Japan) at room temperature for 60 min, followed by the standard procedure for the S–P immunohistochemical kit (Fujian Maixin Biological Technology Ltd., Fujian, China) and stained with streptavidin–biotin– peroxidase complex at room temperature. Immunohistochemical reactions were developed in freshly prepared 3, 30 -diamino-benzidine tetrahydrochloride (DAB kit; Fujian Maixin Biological Technology Ltd.) for immune complex visualization and then lightly counterstained with hematoxylin for 30 s before mounting. The immunoscoring of SBP1 was based on the percentage of immunopositive levels based on preliminary results. Tumors with less than 10% SBP1 expression were regarded as negative [14]. Also, because the nuclear and cytoplasmic stains paralleled, total cellular (instead of separate nuclear and cytoplasmic) scores were assessed. In cases showing separate foci of the same histologic pattern, an average score of staining for the same histologic pattern was used. Final scores were assigned after discussion by 2 independent observers. Reverse transcription-polymerase chain reaction We selected 16 pairs of frozen samples from the former 65 pairs of GC and their corresponding nonneoplastic epithelia tissues to further examine the SBP1 mRNA levels. These 16 pairs of samples were examined by microscopy to ensure that at least 90% of each cancer sample composed of tumor cells. Total RNA was isolated using an RNeasy Mini Kit (Qiagen, Valencia, CA), according to the manufacturer’s recommendations. After the purity and integrity of the obtained RNA was assessed, cDNA was synthesized using Superscript III platinum kit (Invitrogen) according to manufacturer’s instructions. The PCR consisted of 1 DNA denaturation step (94°C, 3 min), 35 cycles of amplification (94°C, 30 s, 50°C, 45 s, 72°C, 60 s). The housekeeping gene b-actin was set as internal control. The sequences of the primers used were as follows: SBP1-F: 50 -TCAAGGGC AATGGCAAAG-30 , SBP1-R: 50 -TCCAGCCTCCACATC AGC-30 (GenBank accession number NM_003944), b-actinF: 5’-CAACTGGGACGACATGGAGAAAAT-3’; b-actinR: 50 -CCAGAGGCGTACAGGGATAGCAC-30 (GenBank accession number NM_001101). PCR products were

Med Oncol (2011) 28:481–487

separated by 2% agarose gel electrophoresis, then scanned by image analyzer and evaluated by the relative intensity ratio of SBP1/b-actin mRNA. Western blotting analysis The tissues selected in the RT-PCR were further subjected to Western blotting analysis. Briefly, 100 mg tissue from each sample was suspended in 1 mL cold cell disruption buffer containing 1 mM phenylmethyl sulfonylfluoride (PMSF) and centrifuged at 15,0009g for 5 min at 4°C. The protein concentration was measured with the BCA protein assay (Pierce Chemical, Rockford, USA). Protein (25 lg) from each sample was subjected to sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS–PAGE) and then transferred onto a polyvinylidene fluoride (PVDF) membrane. Target proteins were detected by using specific antibody anti-SBP1 antibody (1:500 in blocking solution; Clone 4D4, MBL, Nagoya, Japan) at room temperature for 2 h and the bound antibodies were visualized using diaminobenzidine (DAB) finally. The protein quantity was analyzed by UTHSCSA Image Tool Version 3.0. The SBP1 protein expression was evaluated by the relative intensity ratio of SBP1/b-actin protein.

483

analysis. Probability values below 5% were considered statistically significant.

Results Expression of SBP1 in GC and normal gastric mucosa

All analyses were performed using the SPSS software package, version 13.0 (SPSS Inc., Chicago, IL, USA). Correlation analyses of the expression of SBP1 with clinicopathological features were performed using the two-sided chi-squared test or Fisher’s exact test. The independentsample t-test was used to evaluate the significance of the difference in the expression level of SBP1/b-actin mRNA and protein. Cumulative survival was calculated by the Kaplan–Meier method, and differences in univariate survival were examined using the log rank test. The Cox proportional hazards model was used for multivariate regression

Immunohistochemical analysis demonstrated that SBP1 protein expression decreased significantly in GC compared to its abundance in the corresponding nonneoplastic epithelia tissues. Among the 65 stage II–III (M0) tested cancer samples, only 29.2% (19/65) stained positive. SBP1 expression was located primarily in the cytoplasm and nucleus of glandular cells, while became undetectable in germinal centers and connective tissue (Fig. 1). To verify whether the down-regulation of SBP1 is clinically correlated with GC progression, comparative analysis of SBP1 expression was further conducted on 16 pairs of primary GC tissues and corresponding nonneoplastic epithelia tissues. The analysis revealed that the levels of SBP1/b-actin mRNA expression in GC tissues were significantly lower than those in corresponding nonneoplastic epithelia tissues (0.075 ± 0.032 vs. 1.243 ± 0.115, P = 0.000) and the normal/tumor ratio of SBP1 mRNA expression could be as high as 22.5-fold in one case of 16 paired primary GC tissues (Fig. 2a, b). Similarly, the expression of SBP1 protein was also found to be downregulated in all 16 human primary GC tissues compared with their corresponding nonneoplastic epithelia tissues. The Western blotting analysis revealed that the levels of SBP1/b-actin protein were markedly down-regulated in GC tissues compared to the corresponding nonneoplastic epithelia tissues (0.118 ± 0.0746 vs. 1.228 ± 0.171, P = 0. 000). (Fig. 2c, d). It is of note that the levels of SBP1 protein in the GC tissues were correlated with the mRNA expression levels, indicating that the down-regulation of SBP1 in GC may be primarily caused by transcriptional down-regulation.

Fig. 1 Immunohistochemical staining of SBP1 showed the reduction of SBP1 in gastric carcinoma. (a) Negative staining of SBP1 in the gastric carcinoma while positive in corresponding nonneoplastic epithelia tissues. The arrow indicates the immunoactive glandular

cells (SP 9 200). (b) Positive staining of SBP1 in corresponding nonneoplastic epithelia tissues. The arrow indicates the location of SBP1 in the cytoplasm and nucleus of glandular cells (SP 9 400). (C) Negative staining of SBP1 in gastric carcinoma (SP 9 400)

Statistical analyses

123

484

Med Oncol (2011) 28:481–487

Fig. 2 Expression of SBP1 is reduced in gastric carcinoma tissues. Expression (a) and quantification (b) of SBP1 mRNA expression analyzed by RT-PCR in each of the primary gastric carcinoma (T) and corresponding nonneoplastic epithelia tissues (ANT) paired from the same patient. b-actin was used as a loading control. Expression

(c) and quantification (d) of SBP1 protein in each of the primary gastric carcinoma (T) and ANT paired from the same patient by Western blotting. Expression levels were normalized with b-actin. Error bars represent standard deviations calculated from three parallel experiments

The correlation between the decreased expression of SBP1 and clinicopathological features

survival rate than the SBP1-negative group (n = 46) (v2 = 11.217, P = 0.001). The cumulative survival curve for the patients with positive or negative SBP1 expression is shown in Fig. 3. To identify independent predictors for survival, univariate and multivariate Cox regression analyses were performed. Among the entire clinicopathological features possibly affecting prognosis, univariate analysis showed that negative SBP1 and high clinical stage were significant negative prognostic factors for cancer-specific survival. Age and sex of patients, tumor site, tumor differentiation, infiltration depth and lymph node invasion did not influence survival (Table 2). The 2 significant prognostic factors determined by univariate analysis were included in a subsequent stepwise multivariate analysis. In this analysis, SBP1 (P = 0.004; relative risk: 3.238; 95% CI: 1.458–7.191) as well as clinical stage (P = 0.001; relative risk: 3.495; 95% CI: 1.682–7.265) had a significant independent prognostic effect.

SBP1 down-regulation had no correlation with patients’ gender, age, tumor site, tumor differentiation, infiltration depth or lymph node invasion (P [ 0.05). However, the percentage of SBP1-negative expression rate increased as the clinical stage advanced; 52.0% was immunoreactive in clinical stage II; 82.5% in III, a significant difference was found between them (P = 0.038) (Table 1). The reduction of SBP1 was correlated to clinical stage. These results may indicate that the reduction of SBP1 was associated with the progression of the disease. Prognostic impact of SBP1 expression in GC To test the hypothesis that the changes in SBP1 expression are relevant for the outcome of patients with GC, we evaluated SBP1 as a prognostic factor in 65 patients with GC after surgical resection according to immunohistochemical SBP1 expression. The median survival time for patients with positive SBP1 expression was 67 (95% CI: 60.85–73.08) months compared to only 49 (95% CI: 42.53–54.76) months for patients with negative SBP1 expression. The 3-year and 5-year survival rates were 94.7% and 73.7% in the SBP1-positive expression group (n = 19) compared with 71.5%, and 27.2% in the negative expression group (n = 46), respectively. Survival in patients with positive SBP1 expression was significantly better than that in patients with negative SBP1 expression. The SBP1-positive group (n = 19) revealed a higher

123

Discussion Gastric carcinoma (GC) is fourth on the incidence list of cancers worldwide with a high disease-related mortality rate. To improve outcome and the quality of life of patients with GC, sensitive biomarkers for early diagnose and predicting prognosis were badly needed. The highly conserved sequences of selenium-binding proteins among diverse species and kingdoms suggest that selenium-binding proteins share a fundamental biological role. SBP1 was first characterized in mouse liver [10] and was cloned in 1997 [21].

Med Oncol (2011) 28:481–487

485

Table 1 Correlations between SBP1 expression and clinicopathological variables Clinicopathological variables

N

SBP1 expression (%) Positive

Overall frequency Age (years) Gender Tumor site

Tumor differentiation

Tumor infiltration

Lymph node invasion Clinical stage

v2

P value

43.144

0.000*

0.793

0.373

0.793

0.373

1.098

0.578

0.935

0.627

6.189

0.103

3.493

0.062

4.284

0.038*

Negative

Nonneoplastic

65

56 (44.6)

9 (55.4)

Gastric carcinoma

65

19 (29.2)

46 (70.8)

C65

26

6 (23.1)

20 (76.9)

\65

39

13 (33.3)

26 (66.7)

Male

39

13 (33.3)

26 (66.7)

Female

26

6 (23.1)

20 (76.9)

Distal 1/3 Middle 1/3

35 27

12 (34.3) 6 (22.2)

23 (65.7) 21(77.8)

Proximal 1/3

3

1 (33.3)

2 (66.7)

Well

7

3 (42.9)

4 (57.1)

Moderate

26

8 (30.8)

18 (69.2)

Poor

32

8 (25)

24 (25)

T1

13

1 (7.7)

12 (92.3)

T2

29

12 (41.4)

17 (58.6)

T3

15

5 (12.5)

10 (87.5)

T4

8

1 (29.2)

7 (70.8)

N0/N1

23

10 (43.5)

13 (56.5)

N2/N3

43

9 (21.4)

33 (78.6)

II

25

11 (44.0)

14 (56.0)

III

40

8 (20.0)

32 (80.0)

* P \ 0.05 Cancer staging was determined according to the 6th edition of AJCC cancer staging manual

Fig. 3 Overall survival curves of patients with gastric carcinoma after surgical resection according to immunohistochemical SBP1 expression. Among the 65 patients with gastric carcinoma, the SBP1positive group (n = 19) revealed a higher survival rate after surgical resection than the SBP1-negative group (n = 46) (v2 = 11.217, P = 0.001). Cumulative survival curve was analyzed by the Kaplan– Meier method

Researches have demonstrated the abundance of SBP1 in several tissues, including the normal human prostate, ovaries, small intestine, colon and lung.

The specific role of SBP1 was not clear now, but researches have suggested its important role in detoxification mechanisms [22]. Down-regulation of SBP1 expression has been identified in many cancers. Other research has demonstrated that SBP1 is correlated to lung cancer differentiation [12] and intestinal cell maturation [15]. SBP1 expression was also closely correlated to the prognosis of many malignant tumors. And recently, overexpression of SBP1 in human colon cancer cell line HCT116 was identified to induce H2O2-mediated apoptosis, inhibit cell migration in vitro and inhibit tumor growth in nude mice, which provide the evidence to show that SBP1 has anti-cancer functions [23]. Furthermore, SBP1 was characterized as a biomarker for schizophrenia [24]. However, the diagnosis and prognosis value of SBP1 in GC has not been elucidated. In our research, SBP1 was greatly reduced in GC while abundant in the corresponding nonneoplastic epithelia tissues. The result was concordant with previous studies. The reduction of SBP1 may play a significant role in gastric carcinogenesis. With the previous results, this may suggest that SBP1 was important to keep the normal function of gastric epithelial and may have tumor suppressive functions.

123

486 Table 2 Univariate analysis of predictive factors for survival (Cox proportional hazards model)

Med Oncol (2011) 28:481–487

Prognostic factors

Relative risk (95% confidence interval)

P values

Univariate Age (C60 years/\60 years)

1.416 (0.731–2.743)

0.302

Gender (male/female)

1.307 (0.659–2.592)

0.443

Tumor site

0.605 (0.336–1.090)

0.095

SBP1(?)/(–)

3.196 (1.349–7.572)

0.008* 0.286

Tumor differentiation

1.265 (0.821–1.950)

* P \ 0.05

Infiltration depth

1.273 (0.929–1.744)

0.133

Cancer staging was determined according to the 6th edition of AJCC cancer staging manual

Lymph node invasion

1.032 (0.469–2.268)

0.938

Clinical stage

3.707 (1.662–8.269)

0.001*

However, more cases should be examined to estimate whether the decrease of SBP1 is an early or a late event in gastric carcinogenesis. Although no statistically significant differences of SBP1 expression levels were found among 65 GC tissues at various stages of differentiation in our research, whether SBP1 could be proposed as a marker in GC cell differentiation still need further study. We did not find any association between SBP1 downregulation and clinicopathological features such as age, gender, tumor site, infiltration depth or lymph node invasion. However, our study showed that the reduction of SBP1 was significantly correlated with the clinical stage of GC. The percentage of SBP1-positive expression rate in III GC was much lower than that of II, which suggested that loss of SBP1 expression may be involved in more malignant characters. So, the reduction of SBP1 might play an important role in the progression of GC. Our study also showed a positive prognostic value of immunohistologically determined SBP1 protein expression in GC. Using univariate and multivariate Cox regression analysis, we demonstrated that negative SBP1 was an independent risk factor for worse prognosis in patients with GC. Further studies including more patients are still needed to accommodate these findings with the function of SBP1. Although we have not demonstrated clearly the mechanisms underlined the tumor suppressor functions of SBP1, it has been evidenced that SBP1 promoter was hypermethylated in human colon cancer cells and indicated that SBP1 expression in colon cancer was in part controlled through gene silencing [23]. Although decreased expression of SBP1 was demonstrated in GC, more evidence should be provided to identify the tumor suppressive activities of SBP1. Further investigations both in vitro and in vivo were needed. In conclusion, we demonstrated that aberrant SBP1 mRNA and protein are expressed in GC. Reduction of SBP1 may be involved in GC progression and is associated with relatively poor prognosis in patients with GC after surgical resection.

123

Acknowledgments This work was supported by Project Supported by Clinical Research Center of Hubei Province (No. XJ4D01).

References 1. Van den Brandt PA, Goldbohm RA. Nutrition in the prevention of gastrointestinal cancer. Best Pract Res Clin Gastroenterol. 2006;20:589–603. 2. Holmes S. Nutrition and the prevention of cancer. J Fam Health Care. 2006;16:43–6. 3. Alaejos MS, Diaz Romero FJ, Diaz Romero C. Selenium and cancer: some nutritional aspects. Nutrition. 2000;16:376–83. 4. Reid ME, Duffield-Lillico AJ, Garland L, Turnbull BW, Clark LC, et al. Selenium supplementation and lung cancer incidence: an update of the nutritional prevention of cancer trial. Cancer Epidemiol Biomarkers Prev. 2002;11:1285–91. 5. Combs GF Jr. Current evidence and research needs to support a health claim for selenium and cancer prevention. J Nutr. 2005; 135:343–7. 6. Brinkman M, Buntinx F, Muls E, Zeegers MP. Use of selenium in chemoprevention of bladder cancer. Lancet Oncol. 2006;7: 766–74. 7. Meuillet E, Stratton S, Prasad Cherukuri D, Goulet AC, Kagey J, et al. Chemoprevention of prostate cancer with selenium: an update on current clinical trials and preclinical findings. J Cell Biochem. 2004;91:443–58. 8. Behne D, Kyriakopoulos A. Mammalian selenium-containing proteins. Annu Rev Nutr. 2001;21:453–73. 9. Papp LV, Lu J, Holmgren A, Khanna KK. From selenium to selenoproteins: synthesis, identity, and their role in human health. Antioxid Redox Signal. 2007;9:775–806. 10. Bansal MP, Mukhopadhyay T, Scott J, Cook RG, Mukhopadhyay R, et al. DNA sequencing of a mouse liver protein that binds selenium: implications for selenium’s mechanism of action in cancer prevention. Carcinogenesis. 1990;11:2071–3. 11. Jeong JY, Wang Y, Sytkowski AJ. Human selenium binding protein-1 (hsp56) interacts with vdu1 in a selenium-dependent manner. Biochem Biophys Res Commun. 2009;379:583–8. 12. Chen G, Wang H, Miller CT, Thomas DG, Gharib TG, et al. Reduced selenium-binding protein 1 expression is associated with poor outcome in lung adenocarcinomas. J Pathol. 2004;202:321–9. 13. Huang KC, Park DC, Ng SK, Lee JY, Ni X, et al. Selenium binding protein 1 in ovarian cancer. Int J Cancer. 2006;118: 2433–40. 14. Kim H, Kang HJ, You KT, Kim SH, Lee KY, et al. Suppression of human selenium-binding protein 1 is a late event in colorectal

Med Oncol (2011) 28:481–487

15.

16.

17.

18.

19.

carcinogenesis and is associated with poor survival. Proteomics. 2006;6:3466–76. Li T, Yang W, Li M, Byun DS, Tong C, et al. Expression of selenium-binding protein 1 characterizes intestinal cell maturation and predicts survival for patients with colorectal cancer. Mol Nutr Food Res. 2008;52:1289–99. He QY, Cheung YH, Leung SY, Yuen ST, Chu KM, et al. Diverse proteomic alterations in gastric adenocarcinoma. Proteomics. 2004;4:3276–87. Zeng X, Liao AJ, Tang HL, Yi L, Xie N, et al. Screening human gastric carcinoma-associated antigens by serologic proteome analysis. Ai Zheng. 2007;26:1080–4. Wu C, Luo Z, Chen X, Wu C, Yao D, et al. Two-dimensional differential in-gel electrophoresis for identification of GC-specific protein markers. Oncol Rep. 2009;21:1429–37. Greene FL, Page DL, Fleming ID, et al. AJCC cancer staging manual. 6th ed. New York: Springer; 2002.

487 20. Research Society for Gastric Cancer. Japanese classification of gastric carcinoma. Tokyo: Kanehara; 1998. (2nd English edn). 21. Chang PW, Tsui SK, Liew C, Lee CC, Waye MM, et al. Isolation, characterization, and chromosomal mapping of a novel cDNA clone encoding human selenium binding protein. J Cell Biochem. 1997;64:217–24. 22. Agalou A, Roussis A, Spaink HP. The Arabidopsis seleniumbinding protein confers tolerance to toxic levels of selenium. Functional Plant. 2005;32:881–90. 23. Pohl NM, Tong C, Fang W, Bi X, Li T, et al. Transcriptional regulation and biological functions of selenium-binding protein 1 in colorectal cancer in vitro and in nude mouse xenografts. PLoS One. 2009;4(11):e7774. 24. Glatt SJ, Everall IP, Kremen WS, Corbeil J, Sasik R, et al. Comparative gene expression analysis of blood and brain provides concurrent validation of SELENBP1 up-regulation in schizophrenia. Proc Natl Acad Sci USA. 2005;102:15533–8.

123