Differential expression of EphA7 receptor tyrosine kinase in gastric carcinoma

Human Pathology (2007) 38, 1649 – 1656

www.elsevier.com/locate/humpath

Original contribution

Differential expression of EphA7 receptor tyrosine kinase in gastric carcinomaB Jiandong Wang PhDa, Guoli Li MD, PhDb, Henghui Ma BSa, Yang Bao MD, PhDb, Xulin Wang MD, PhDb, Hangbo Zhou BSa, Zhen Sheng MD, PhDa, Haruhiko Sugimura MD, PhDc, Jie Jin BSa, Xiaojun Zhou MD, PhDa,* a

Department of Pathology, Nanjing University School of Medicine/Nanjing Jinling Hospital, Nanjing 210002, P.R. China Department of Surgery, Nanjing University School of Medicine/Nanjing Jinling Hospital, Nanjing 210002, P.R. China c Department of Pathology of Hamamatsu University School of Medicine, Hamamatsu 431-3192, Japan b

Received 30 October 2006; revised 16 January 2007; accepted 18 January 2007

Keywords: EphA7; Gastric carcinoma; Hypermethylation; Real-time RT-PCR

Summary Overexpression of the Eph receptor tyrosine kinases in cancers is related to malignant transformation, metastasis, tumor differentiation, and prognosis. Down-regulation of EphA7 secondary to hypermethylation has been reported in colorectal cancer. In the present study, expression of EphA7 in gastric cancer cell lines and gastric carcinoma specimens was determined by quantitative real-time reverse transcriptase–polymerase chain reaction. The expression of EphA7 was reduced in all tested gastric cancer cell lines; however, there is marked variability in expression among gastric carcinoma specimens. Overexpression was observed more often in younger patients (aged V65 years) ( P = .03) and in patients with advanced cancer ( P = .031). Hypermethylation of the EphA7 promoter–associated CpG island was found using methylation-specific polymerase chain reaction. Down-regulation of EphA7 was significantly related to hypermethylation ( P = .001), and the level of hypermethylation was greater in moderately differentiated tumors than in poorly differentiated ones ( P = .03). We also performed immunohistochemical staining for EphA7 in 52 gastric carcinoma specimens and found that expression of the protein was consistent with its transcript expression, with the protein being significantly overexpressed in younger patients ( P = .016) and in patients with advanced tumors ( P = .033). Our data indicate that EphA7 may have roles in the pathogenesis and development of gastric carcinomas. D 2007 Published by Elsevier Inc.

1. Introduction Receptor tyrosine kinases of the Eph family have important roles in vascular development, tissue-border B This work was supported in part by the China Postdoctoral Science Foundation (no. 2005038578; Beijing), China Jiangsu Province Science Foundation for Six Kinds of Scientific Research (no. 2005A2), and the China Nanjing Medicine Science and Technology Research Project (no. 06Z37). * Corresponding author. E-mail address: [email protected] (X. Zhou).

0046-8177/$ – see front matter D 2007 Published by Elsevier Inc. doi:10.1016/j.humpath.2007.01.030

formation, cell migration, axon guidance, and angiogenesis. The Eph receptors are the largest subfamily of the receptor tyrosine kinases, which can be subdivided into EphA and EphB groups according to the sequence homology of their extracellular domains and their affinity for the corresponding ligands, EphrinA and EphrinB. The first family member identified, which was originally named eph but later renamed EphA1, was isolated from a human hepatocellular carcinoma cell line [1]. At present, the Eph family includes at least 14 Eph receptors and 8 ephrin ligands. EphA receptors

1650 interact with several ligands of the ephrin-A family, a group of glycosyl-phosphatidylinositol–linked membrane proteins, whereas EphB receptors bind to ephrin-B ligands, a family of transmembrane proteins. The EphA4 receptor is an exception, as it can bind both ephrin-A and most of the ephrin-B ligands. After binding of their ligands, Eph receptors are phosphorylated at specific tyrosine residues in the cytoplasmic region. The phosphorylated tyrosine residue can then interact with many Src homology2 (SH2) domain– containing proteins, including cytoplasmic tyrosine kinases of the Src family [2,3] and adaptor proteins [4], to activate signal pathways such as mitogen-activated protein kinase/ extracellular signal-regulated protein kinase (MAPK/ERK) [3]. Many of the signaling proteins interact with Eph receptors and are involved in cell adhesion, cell migration, axon guidance, and cytoskeleton rearrangement. However, the mechanisms of these signaling pathways remain to be determined. Initial studies indicated that Eph receptors and ephrin ligands are central to many developmental processes, including embryo patterning, angiogenesis, and axon guidance [5-9]. It is now clear that the Eph molecules also have a role in adult tissues under physiologic conditions and in disease states such as cancer. Eph receptors and their ligands often are overexpressed in different types of cancer. EphA2 has been shown to be overexpressed in clinical specimens and cell models of breast cancer, and EphA2 overexpression has been shown to confer malignant transformation and tumorigenic potential on nontransformed mammary epithelial cells [10]. Overexpression of EphA2 has been detected in esophageal cancers, and there is a significant correlation between EphA2 expression and regional lymph node metastasis. Patients with esophageal cancer who have EphA2 overexpression have a worse prognosis than those without [11]. Overexpression of EphA2 also has been observed in patients with prostate [12], colorectal [13], ovarian [14], and gastric [15] cancers; and EphB4 has been found to be highly expressed in prostate [16,17] and bladder [18] cancers. Previously, overexpression of some Eph genes has been reported as typical for certain tumors. Recently, the genes for Eph receptors and ephrins have been found to be differentially expressed in various human tumors. Hypermethylation is the mechanism by which EphA3 is down-regulated in hematopoietic tumors [19]. We previously reported that EphA7 expression is reduced in colorectal cancer [20]. The EphA7 (MDK1) receptor was first identified in the murine nervous system, and several alternatively spliced variants were found [21]. Human EphA7 (HEK11) was isolated from a human fetal brain library and found to be distributed widely in human tissues [22]. Of the Eph family genes, less attention has been directed to EphA7 in human tumors, and its potential role in human oncology has not been addressed. In our previous study, we found that EphA7 is not expressed in the gastric cancer cell lines. So far, the possible role of EphA7 in human gastric cancers has not been investigated. In the present report, we describe the differential expression of

J. Wang et al. EphA7 in gastric carcinoma samples and the correlation between EphA7 expression and clinicopathologic features.

2. Materials and methods 2.1. Gastric cancer cell lines and gastric carcinoma specimens The gastric cancer cell lines TMK1, MKN28, MKN74, and KATOIII were maintained in Dulbecco modified Eagle medium (NISSUI Pharmaceutical Co, Tokyo, Japan) supplemented with 1 mmol/L l-glutamine, 10% fetal bovine serum (Life Technologies, Inc, Burlington, Canada), penicillin G 100 U/mL, and streptomycin 100 mg/mL. The cells were incubated at 378C in a humidified atmosphere of 95% air and 5% CO2. A total of 62 gastric carcinoma tissues and adjacent normal mucosa specimens were obtained from the surgical resections performed at the Nanjing Jinling Hospital (Nanjing, China) between 2005 and 2006 as part of a study approved by the research ethics board of the hospital. The tissue specimens were evaluated pathologically. The clinicopathologic features of the patients are shown in Table 1.

2.2. Quantitative real-time reverse transcriptase–polymerase chain reaction Messenger RNA (mRNA) was extracted from tumor tissues and normal mucosa by homogenization of the Table 1

Characteristics of 62 patients with gastric carcinoma

Variable

Number

Male-female Age (y) V65 N65 Tumor differentiation Moderate Poor Tumor size (cm) b5.5 z5.5 Depth of wall invasion Mucosa, submucosa Muscularis propria Subserosa, serosa Stage I + II III + IV Microscopic subtypes Intestinal Diffuse Atypical Lymph node metastasis Present Absent

47:15 41 21 28 34 37 25 3 13 46 25 37 40 19 3 42 20

Differential expression of EphA7 receptor tyrosine kinase in gastric carcinoma tissues in TRIzol reagent (Invitrogen, Carlsbad, CA) for 30 seconds, followed by extraction according to the protocol provided by the manufacturer. The quality of the RNA was checked using a denatured gel in 3-[Nmorpholine]-propanesulfonic acid buffer, and the concentration of total RNA was determined with a spectrophotometer. A 1-lg sample of total RNA was reverse transcribed to complementary DNA (cDNA) with oligo(dT) primers and a reverse transcription system (Promega, Madison, WI). To detect the expression of EphA7 in gastric cancers, we subjected the gastric cancer cDNAs to quantitative real-time reverse transcriptase–polymerase chain reaction (RT-PCR). The reactions were performed in triplicate. The sense and antisense primers and TaqMan probe for EphA7 were designed according to the mRNA sequence (GenBank accession number NM_004440). We used amplified PCR fragments spanning different exons to prevent amplification of contaminated genomic DNA. The sense primer was 5V-AACAGAGTTGGAGTGGATTTC-3V and the antisense primer was 5V-CACCTGGTATGTTCGTATCG-3V. The PCR products were 94 base pairs (bp) long. The TaqMan probe was 5V-(FAM) CCAATGGGTGGGAAGAAATTAGTG (Eclipse)-3V. The housekeeping gene GAPDH was used as an internal control. The sense primer was 5V-CCAGGTGGTCTCCTCTGACTT-3V and the antisense primer was 5V-GTTGCTGTAGCCAAATTCGTTGT3V. The PCR products were 130 bp long. The probe was 5V(FAM) AACAGCGACACCCACTCCTCCACC (Eclipse)3V. Primer sets and probes were synthesized by TaKaRa Biotechnology, Inc (Dalian, China). The mRNA extracted from normal parts of a renal tumor was reverse transcribed into cDNA (the expression level of EphA7 in the kidney is high). This cDNA was serially diluted to obtain standard curves for the detection of EphA7 and GAPDH. The values of EphA7 mRNA expression were normalized using the GAPDH expression. Quantitative real-time RT-PCR was performed using an ABI PRISM 7000 sequence detection system (Applied Biosystems, Foster City, CA). The RT-PCR was carried out in a total volume of 30 lL. The reaction mixture included 1 buffer, 200 lmol/L of deoxy-ribonucleoside triphosphates (dNTPs) (Invitrogen), 0.3 lmol/L of sense and antisense primers, 1 U of Takara ExTaq Hotstart Taq (TaKaRa Biotechnology), 0.6 lL of 5-carboxy-x-rhodamine (ROX) reference dye, and 2 lL of cDNA. The PCR cycle involved 2 minutes at 958C followed by 40 amplification cycles of denaturation at 948C for 30 seconds, annealing at 588C (for detection of GAPDH) or 558C (for detection of EphA7) for 30 seconds, and elongation at 728C for 1 minute.

2.3. Methylation-specific PCR Genomic DNA was modified by sodium bisulfite, as described by Clark et al [23]. Primers were designed to discriminate between methylated and unmethylated alleles after sodium bisulfite treatment. Primer sequences were

1651

chosen for the regions containing frequent CpG and CpG pairs near the 3V end of the primers to provide maximum discrimination between methylated and unmethylated DNA. Two-microliter aliquots were amplified in a 30-lL reaction mixture consisting of 1 buffer (10 mmol/L Tris-HCl, 2.0 mmol/L MgCl2, 50 mmol/L KCl, pH 8.3), 1 U of Takara ExTaq Hotstart Taq, 260 lmol/L dNTPs, and 0.3 lmol/L of the primer sets. The PCR involved 2 minutes at 958C, then 35 cycles of 948C for 30 seconds, 588C (for detection of methylated DNA) or 538C (for detection of unmethylated DNA) for 30 seconds, and 728C for 1 minute, and finally 10 minutes at 728C. The methylation-specific primers were 5V-ATTTGATTTCGTTCGGTATC-3V (forward) and 5VCTCCGACTACAAACCGACCG-3V (reverse). The unmethylation-specific primers were 5V-ATTTGATTTTGTTT GGTATT-3V (forward) and 5V-CTCCAACTACAAACCAACCA-3V (reverse). Primer sets for the detection of methylated and unmethylated DNA were located at the same sites of genomic sequence (forward primer 432 to 413 from the translation start site; reverse primer 223 to 204). The PCR products were 229 bp long. They were separated on 8% nondenaturing polyacrylamide gel, followed by ethidium bromide staining.

2.4. Immunohistochemical staining Sections from the surgical specimens that were fixed in 10% formalin and embedded in paraffin were studied. Immunohistochemical staining was performed according to the standard method. Briefly, each 4-lm tissue section was deparaffinized and rehydrated using a gradient of ethanol concentrations. The sections were autoclaved in 10 mmol/L citrate buffer (pH 6.0) at 1208C for 2 minutes for antigen retrieval, then cooled to 308C and washed with phosphate-buffered saline (PBS, pH 7.3). After nonspecific sites had been blocked with 10% normal calf serum in PBS for 10 minutes, the sections were incubated with an anti-EphA7 polyclonal antibody (ABGENT, San Diego, CA) at a dilution of 1:50 in antibody diluent solution (ZYMED, Invitrogen) at 48C overnight, followed by washing with PBS. The sections were then incubated with secondary antibody (Dako, Glostrup, Denmark) for 30 minutes at room temperature. Color development was performed with 3,3V-diaminobenzidine. Nuclei were lightly counterstained with hematoxylin.

2.5. Evaluation of staining for EphA7 Two pathologists independently assessed the immunostained slides. Any difference in the scores was resolved by consensus. Immunohistochemical staining of both normal mucosa and cancer cells was assessed according to both the intensity and the proportion of cells that were stained. Staining intensity was recorded as 0 for no staining, 1 for weak staining, 2 for moderate staining, and 3 for strong staining. The percentage of positive cells was classified semiquantitatively as 0 for tissue specimens without

1652

J. Wang et al.

Fig. 1 Bisulfite sequencing of 5V CpG islands upstream of the EphA 7 translation start site. A, Genomic sequence from 578 to 204 of translation start site contains 30 5V CpG sites. The arrows show locations of primer sets of bisulfite sequencing and methylation-specific PCR. B, Bisulfite sequencing results in 6 gastric cancer cell lines (MKN1, NKN28, MKN74, KATOIII, TMK1, and AGS) and the renal cell line 293T. Methylated and unmethylated 5V CpG sites are shown as solid and open circles, respectively, and incompletely methylated alleles are shown as meshed circles.

staining, 1 for fewer than 10% of cells stained, 2 for 10% to 50% of cells stained, and 3 for more than 50% of the tissue stained. Scores for expression and percentage of positive cells were added. The EphA7 expression was assessed by comparing the scores for tumor cells and adjacent normal mucosa cells.

2.6. Statistical analysis The statistical significance of intergroup differences was determined using the v 2 test. All statistical analyses were performed using the SPSS software (SPSS, Chicago, IL). For all of the statistical tests, a 2-sided P value of less than .05 was considered statistically significant.

3. Results 3.1. Loss of EphA7 transcript expression in the gastric cancer cell lines The expression of EphA7 in the gastric cancer cell lines (KATOIII, MKN28, MKN74, and TMK1) was assessed using the semiquantitative RT-PCR method we previously

reported [20] (Supplementary Fig. 1). Expression was not detected in any of these lines. The methylation status of the 5V CpG island upstream of the translation start site was detected by bisulfite DNA sequencing, as we reported previously [20], and the results indicated aberrant methylation status (Fig. 1). The bisulfite sequencing result of the cell line KATOIII is shown in Supplementary Fig. 2.

3.2. Expression of the EphA7 transcript in gastric carcinoma specimens Expression of the EphA7 transcript was measured using real-time RT-PCR in 62 gastric carcinoma specimens that

Table 2 Correlation of EphA7 transcript expression with methylation status Methylation status

EphA7 transcript expression N/T N2

N/T 0.5-2

N/T b0.5

Methylation Unmethylation

15 12

6 7

1 21

Abbreviation: N/T, normal mucosa–tumor ratio.

P .001

Differential expression of EphA7 receptor tyrosine kinase in gastric carcinoma contain paired normal mucosa and tumor tissues. The carcinomas showed marked interspecimen variability in their degrees of expression. The expression of EphA7 in gastric carcinoma tissues was compared with that in paired normal mucosa and classified into A, B, or C according to the ratio of the two: A, normal mucosa-to-tumor tissue ratio of greater than 2 (N/T N2); B, normal mucosa-to-tumor ratio of less than 0.5 (N/T b0.5); and C, normal mucosa-to-tumor ratio of between 0.5 and 2 (N/T 0.5-2) (Table 2). Downregulation of EphA7 (class A) was observed in 27 gastric carcinoma specimens (43.5%), and overexpression (class B) was observed in 22 samples (35.4%).

3.3. Hypermethylation of EphA7 in gastric carcinoma specimens The methylation status of the 5V CpG island located upstream of the translation start site was assessed by methylation-specific PCR, as previously reported [20]. Evidence of hypermethylation of the 5V CpG island was found in the gastric carcinoma samples (Fig. 2). Methylated EphA7 DNA was detected in 22 gastric carcinoma samples (35.5%). Of these, expression was markedly down-regulated in 15 tumors, slightly down-regulated in 6, and upregulated in 1 tumor. The relation between the expression of EphA7 and hypermethylation of the 5V CpG island was significant ( P = .001) (Table 2).

3.4. Correlation between EphA7 transcript expression/hypermethylation and clinicopathologic parameters Table 3 shows the correlation between several clinical variables and the expression of EphA7 transcript and methylation status. The transcript level was significantly related to the age of the patients ( P = .03), and the expression of the EphA7 protein was significantly associated with the tumor stage ( P = .031). There was no significant association between EphA7 transcript expression and differentiation, tumor size, depth of wall invasion, microscopic subtype, or lymph node metastasis. Hypermethylation was significantly associated with tumor differentiation ( P = .03). There was no significant relation between hypermethylation and other clinicopathologic parameters.

Fig. 2 Examples of EphA7 methylation in primary gastric carcinoma specimens detected by methylation-specific PCR.

1653

3.5. Expression of EphA7 protein in gastric normal mucosa and gastric carcinoma cells A total of 52 gastric carcinoma and normal adjacent mucosa specimens were stained with a polyclonal antiEphA7 antibody. The expression of the protein differed between gastric cancer cells and adjacent normal mucosa, and heterogeneous staining was observed. Immunoreactivity for EphA7 was observed in the cytoplasm or both the cytoplasm and the golgiosome (Fig. 3). The blood vessels and nerve fibers also stained positively for EphA7. The transcription level of EphA7 was associated with the protein expression level (down-regulation P b .001 and overexpression P = .011) (Table 4). Expression of EphA7 protein was down-regulated in 24 tumor samples (46.2%) and up-regulated in 18 (34.6%).

3.6. Association of EphA7 protein expression with clinicopathologic parameters Table 3 also shows the correlation between EphA7 protein expression and clinicopathologic characteristics. The EphA7 protein was significantly overexpressed in younger patients ( P = .016) and in those with advanced tumors ( P = .033). The overexpression of EphA7 protein was not related to other clinicopathologic characteristics. No significant correlation was found between the down-regulation of EphA7 protein and any clinicopathologic parameter.

4. Discussion In this study, we analyzed the expression of EphA7 in gastric cancer cell lines and gastric carcinoma specimens using semiquantitative RT-PCR and immunohistochemical staining. To our knowledge, this is the first report to document the relation between the expression of EphA7 and clinicopathologic characteristics in patients with gastric cancer. We detected the expression of the EphA7 transcript in gastric cancer cell lines and 62 gastric carcinoma specimens. Expression was reduced in all of the cell lines. The expression levels of EphA7 differed markedly among the gastric carcinoma specimens, which could be divided into 3 groups (A, B, and C) according to the extent of expression. Overexpression of EphA7 correlated significantly with age ( P = .03) and tumor stage ( P = .031). No significant relation was found between expression and lymph node metastasis, but our data indicate that patients with tumors overexpressing EphA7 have a higher likelihood of lymph node metastasis ( P = .063). Expression was down-regulated in 27 gastric carcinoma specimens (43.5%). Loss of expression of specific genes can be explained by genetic and epigenetic mechanisms. We screened the gastric cancer cell lines for EphA7 genomic mutations using a single-strain conformation polymorphism analysis, but no mutations were found in any of the lines. Dottori et al [19] have cloned and characterized the EphA3

1654

Table 3

Correlations of EphA7 transcript expression, methylation status, and protein expression with clinicopathologic parameters

Variable

EphA7 transcript expression N/T N2

Sex Male Female Age (y) V65 N65 Differentiation Moderate Poor Tumor size (cm) b5.5 z5.5 Depth of wall invasion Mucosa, submucosa Muscularis propria Subserosa, serosa Stage I + II III + IV Microscopic subtypes Intestinal Diffuse Atypical Lymph node metastasis Present Absent

N/T 0.5-2

P N/T b0.5

Methylation status M

U

P

Down-regulation Yes

No

P

Overexpression Yes

No

P

23 4

8 5

16 6

.24

17 5

30 10

.842

19 5

20 8

.521

14 4

25 9

.736

13 14

10 3

18 4

.03

13 9

28 12

.385

13 11

22 6

.061

16 2

19 15

.016

15 12

6 7

7 15

.251

14 8

14 26

.03

11 13

13 15

.966

8 10

16 18

.857

17 10

7 6

13 9

.857

14 8

23 17

.637

14 10

17 11

.862

11 7

20 14

.873

1 6 20

2 3 8

0 4 18

.319

2 6 14

1 7 32

.296

2 5 17

1 7 20

.739

1 3 14

2 9 23

.718

14 13

7 6

4 18

.031

11 11

14 26

.249

13 11

9 19

.109

4 14

18 16

.033

18 9 0

9 4 0

13 6 3

.217

17 5 0

23 14 3

.23

18 6 0

16 9 3

.18

10 7 1

24 8 2

.504

16 11

7 6

19 3

.063

14 8

28 12

.608

14 10

21 7

.202

14 4

21 13

.242

Abbreviations: M, methylation; U, unmethylation.

J. Wang et al.

Differential expression of EphA7 receptor tyrosine kinase in gastric carcinoma

1655

Fig. 3 Immunohistochemical analysis of expression of EphA7 in gastric carcinoma and normal tissue. A, Normal gastric tissue: expression of EphA7 is present in the cytoplasm (original magnification 400). B, Expression of EphA7 is not seen in normal gastric tissue (original magnification 400). C, Strong expression is present in gastric cancer cell cytoplasm as particles (original magnification 200). D, Expression in cytoplasm of gastric cancer cells (original magnification 400). E, No expression in gastric cancer cells (original magnification  400). F, Down-regulation of EphA7 in gastric cancer. EphA7 is expressed in normal cells (right) but not in gastric cancer cells (left) (original magnification 400).

promoter, and they identified a region of 86 bp, located 348 to 362 bp upstream from the transcription start site, as the basal promoter. Both EphA7 and EphA3 belong to the same group of EphA receptor tyrosine kinases, and EphA7 has 56% mRNA homology with EphA3. A 5V CpG island has been shown to be located upstream of the EphA7 transcription start site (Fig. 1). In our previous study of colorectal cancer, we reported that hypermethylation of the CpG island was related to the down-regulation of EphA7 expression. We therefore assumed that the down-regulation of EphA7 in gastric carcinoma cells is likewise the result of an epigenetic alteration. In the present study, hypermethylation of the 5V CpG island was confirmed in gastric cancer cell lines and clinical specimens. Of the 27 gastric carcinomas that showed down-regulation of EphA7 (normal mucosa–tumor ratio N2), hypermethylation was detected in 15 (55.5%) and not detected in 12 (44.5%). Other mechanisms, such as histone Table 4

deacetylation, therefore, may also be involved in the downregulation of EphA7 expression in gastric carcinomas. The relation between EphA7 down-regulation and hypermethylation of the 5V CpG island was significant ( P = .001). Our data indicate that hypermethylation of the 5V CpG island contributes to the down-regulation of EphA7 in gastric carcinomas. Of the 35 gastric carcinoma specimens in which EphA7 was not down-regulated (N/T 0.5-2 and N/T b0.5), hypermethylation was detected in 7 (20%). We presumed that the hypermethylation is a heterogeneous event in gastric cancer. If hypermethylation occurred in only part of a cancer cell, down-regulation of EphA7 may not be detected. Methylated DNA was also weakly detected in some samples of normal mucosa. This may be secondary to inflammation or infiltration of cancer cells. In the present study, a significant correlation was observed between hypermethylation of EphA7 and differ-

Correlation of EphA7 transcript expression with protein expression

EphA7 transcript expression level

Yes (n = 24)

EphA7 down-regulation No (n = 28)

N/T N2 N/T 0.5-2 N/T b0.5

18 3 3

6 7 15

P

b.001

EphA7 overexpression Yes (n = 18)

No (n = 34)

4 3 11

20 7 7

P .011

1656 entiation of gastric carcinoma ( P = .03). Thus, hypermethylation often was detected in moderately differentiated specimens, whereas nonmethylation often was detected in poorly differentiated specimens. The level of EphA7 protein expression was assessed using an anti-EphA7 polyclonal antibody in 52 gastric carcinoma specimens. The transcript level was consistent with the protein expression. The EphA7 expression levels differed markedly between gastric carcinoma and adjacent normal mucosa. The protein was distributed diffusely throughout the cytoplasm or stained as particles in the golgiosome. The protein was significantly overexpressed in younger patients ( P = .016) and in those with advanced tumors ( P = .033). No significant relations were found between down-regulation of EphA7 protein and any clinicopathologic parameter. The relation between the down-regulation of EphA7 protein and hypermethylation was not significant ( P = .062). This may be because the number of tested specimens was small (52). The Eph receptors are the largest subfamily of the receptor tyrosine kinases and include many oncogenes and protooncogenes that are involved in cell proliferation, differentiation, migration, and metastatic behavior. Many Eph receptor tyrosine kinases have been reported to be highly expressed in tumor tissue compared with paired normal tissue. In the present study, EphA7 was found to be differentially expressed in gastric carcinoma specimens, and overexpression correlated significantly with patient age, tumor stage, and extent of metastasis. Overexpression of EphA7 often is observed in younger patients at both the transcript and protein levels. We presume that EphA7 has a role in gastric carcinogenesis from an early age. Hypermethylation of the 5V CpG island contributes to down-regulation of EphA7 in parts of the gastric carcinoma specimens. Methylated DNA was detected more often in moderately differentiated gastric cancer specimens than in poorly differentiated cancers, and therefore EphA7 overexpression may also be related to poor differentiation. Briefly, our data indicate that EphA7 may have roles in the pathogenesis and development of gastric cancer.

Acknowledgments The authors thank Dr Xiaohong Liu, Minghong Pan, Qiu Rao, Zhiyi Zhou, and Yingchun Dong for their aid in collection of specimens.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.humpath. 2007.01.030.

J. Wang et al.

References [1] Hirai H, Maru Y, Hagiwara K, Nishida J, Takaku F. A novel putative tyrosine kinase receptor encoded by the eph gene. Science 1987;238: 1717 - 20. [2] Zisch AH, Kalo MS, Chong LD, Pasquale EB. Complex formation between EphB2 and Src requires phosphorylation of tyrosine 611 in the EphB2 juxtamembrane region. Oncogene 1998;16:2657 - 70. [3] Vindis C, Cerretti DP, Daniel TO, Huynh-Do U. EphB1 recruits c-Src and p52Shc to activate MAPK/ERK and promote chemotaxis. J Cell Biol 2003;162:661 - 71. [4] Noren NK, Foos G, Hauser CA, Pasquale EB. The EphB4 receptor suppresses breast cancer cell tumorigenicity through an Abl-Crk pathway. Nat Cell Biol 2006;8:815 - 25. [5] Smith A, Robinson V, Patel K, Wilkinson DG. The EphA4 and EphB1 receptor tyrosine kinases and ephrin-B2 ligand regulate targeted migration of branchial neural crest cells. Curr Biol 1997;7:561 - 70. [6] Durbin L, Brennan C, Shiomi K, et al. Eph signaling is required for segmentation and differentiation of the somites. Genes Dev 1998;12: 3096 - 109. [7] Flanagan JG, Vanderhaeghen P. The ephrins and Eph receptors in neural development. Annu Rev Neurosci 1998;21:309 - 45. [8] Tanaka M, Wang DY, Kamo T, et al. Interaction of EphB2-tyrosine kinase receptor and its ligand conveys dorsalization signal in Xenopus laevis development. Oncogene 1998;17:1509 - 16. [9] Wang HU, Chen ZF, Anderson DJ. Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4. Cell 1998;93:741 - 53. [10] Zelinski DP, Zantek ND, Stewart JC, Irizarry AR, Kinch MS. EphA2 overexpression causes tumorigenesis of mammary epithelial cells. Cancer Res 2001;61:2301 - 6. [11] Miyazaki T, Kato H, Fukuchi M, Nakajima M, Kuwano H. EphA2 overexpression correlates with poor prognosis in esophageal squamous cell carcinoma. Int J Cancer 2003;103:657 - 63. [12] Zeng G, Hu Z, Kinch MS, et al. High-level expression of EphA2 receptor tyrosine kinase in prostatic intraepithelial neoplasia. Am J Pathol 2003;163:2271 - 6. [13] Kataoka H, Igarashi H, Kanamori M, et al. Correlation of EPHA2 overexpression with high microvessel count in human primary colorectal cancer. Cancer Sci 2004;95:136 - 41. [14] Thaker PH, Deavers M, Celestino J, et al. EphA2 expression is associated with aggressive features in ovarian carcinoma. Clin Cancer Res 2004;10:5145 - 50. [15] Nakamura R, Kataoka H, Sato N, et al. EPHA2/EFNA1 expression in human gastric cancer. Cancer Sci 2005;96:42 - 7. [16] Lee YC, Perren JR, Douglas EL, et al. Investigation of the expression of the EphB4 receptor tyrosine kinase in prostate carcinoma. BMC Cancer 2005;5:119. [17] Xia G, Kumar SR, Masood R, et al. EphB4 expression and biological significance in prostate cancer. Cancer Res 2005;65:4623 - 32. [18] Xia G, Kumar SR, Stein JP, et al. EphB4 receptor tyrosine kinase is expressed in bladder cancer and provides signals for cell survival. Oncogene 2006;25:769 - 80. [19] Dottori M, Down M, Huttmann A, Fitzpatrick DR, Boyd AW. Cloning and characterization of EphA3 (Hek) gene promoter: DNA methylation regulates expression in hematopoietic tumor cells. Blood 1999;94:2477-86. [20] Wang J, Kataoka H, Suzuki M, et al. Downregulation of EphA7 by hypermethylation in colorectal cancer. Oncogene 2005;24:5637 - 47. [21] Ciossek T, Millauer B, Ullrich A. Identification of alternatively spliced mRNAs encoding variants of MDK1, a novel receptor tyrosine kinase expressed in the murine nervous system. Oncogene 1995;10:97 - 108. [22] Fox GM, Holst PL, Chute HT, et al. cDNA cloning and tissue distribution of five human EPH-like receptor protein-tyrosine kinases. Oncogene 1995;10:897 - 905. [23] Clark SJ, Harrison J, Paul CL, Frommer M. High sensitivity mapping of methylated cytosines. Nucleic Acids Res 1994;22:2990 - 7.