Decreased expression of annexin A1 is correlated with breast cancer development and progression as determined by a tissue microarray analysis

Human Pathology (2006) 37, 1583 – 1591

www.elsevier.com/locate/humpath

Original contribution

Decreased expression of annexin A1 is correlated with breast cancer development and progression as determined by a tissue microarray analysisB Dejun Shen MD, PhDa,b,1, Farzad Nooraie MDc,1, Yahya Elshimali MDc, Victor Lonsberry BSc, Jianbo He MDa,b, Shikha Bose MDc,e, David Chia MDc,d, David Seligson MDc,d, Helena R. Chang MD, PhDa,b,d,*,2, Lee Goodglick PhDc,d,*,2 a

Gonda/UCLA Breast Cancer Research Laboratory, Revlon/UCLA Breast Center, David Geffen School of Medicine, University of California, Los Angeles, CA 90095, USA b Department of Surgery, David Geffen School of Medicine, University of California, Los Angeles, CA 90095, USA c Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California, Los Angeles, CA 90095, USA d Jonsson Comprehensive Cancer Center and David Geffen School of Medicine at UCLA, University of California, Los Angeles, CA 90095, USA e Department of Pathology and Laboratory Medicine, Cedars–Sinai Medical Center, Los Angeles, CA, USA Received 1 March 2006; revised 23 May 2006; accepted 1 June 2006

Keywords: Annexin A1; Tissue microarray; Breast cancer; Myoepithelium; Tumor marker

Summary Annexin A1 (ANXA1) is a calcium- and phospholipid-binding protein and a known mediator of glucocorticoid-regulated inflammatory responses. Using a combined multiple high-throughput approach, we recently identified a reduced expression of ANXA1 in human breast cancer. The finding was confirmed at the gene level by quantitative reverse transcription-polymerase chain reaction and at the protein level by immunohistochemical staining of normal, benign, and malignant breast tissues. In this study, we constructed and used a high-density human breast cancer tissue microarray to characterize the expressional pattern of ANXA1 according to histopathologies. The tissue microarray contains 1158 informative breast tissue cores of different histologies including normal tissues, hyperplasia, in situ and invasive tumors, and lymph node metastases. Our results showed that there was a significant decrease in glandular expression of ANXA1 in ductal carcinoma in situ and invasive ductal carcinoma compared with either normal breast tissue or hyperplasia ( P b .0001). Moreover, in benign breast tissue, myoepithelial cells showed strong expression of ANXA1. There was a decrease of ANXA1 expression

B This work was supported in part by the Early Detection Research Network, Rockville, MD (D. C., L. G., and D. S.), NCI CA-86366 and Gonda Family Foundation, Beverly Hills, CA, USA (H. R. C.). * Corresponding authors. Lee Goodglick is to be contacted at Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California, Los Angeles, CA, 90095-1747, USA. Helena R. Chang, Revlon/UCLA Breast Center, Department of Surgery, UCLA David Geffen School of Medicine, 200 Medical Plaza, B265, Los Angeles, CA 90095, USA. E-mail addresses: [email protected] (L. Goodglick), [email protected] (H. R. Chang). 1 These authors contributed equally to this study. 2 These authors contributed equally to this study.

0046-8177/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2006.06.001

1584

D. Shen et al. in myoepithelial cells in ductal carcinoma in situ lesions compared with the same cell population in either normal or hyperplastic lesions. These results suggest that suppressed ANXA1 expression in breast tissue is correlated with breast cancer development and progression. D 2006 Elsevier Inc. All rights reserved.

1. Introduction Breast cancer is the leading cancer among American women. More than 200 000 new breast cancer cases were reported annually, accounting for approximately one third of all female cancers in the country [1]. Clinically, breast cancer may evolve through different stages from normal ductal epithelium to hyperplasia, in situ carcinoma, invasive cancer, and metastatic carcinoma. Although an extensive research effort has been devoted to identifying the molecular abnormalities contributing to this process, the mechanisms underlying breast cancer development are not yet clear. Recently we have successfully developed a strategy of using combined multiple high-throughput technologies to search for the important molecular biomarkers characteristic of breast cancer. Through these studies, expression of annexin A1 (ANXA1) was found to be consistently reduced in breast cancer at both RNA and protein levels [2,3]. ANXA1, also known as lipocortin 1, is a 37-kDa calcium- and phospholipid-binding protein of the annexin superfamily found in a wide range of organisms, including vertebrates, invertebrates, and plants [4]. ANXA1 is an important mediator in glucocorticoid-regulated inflammatory response [5] and was found to be associated with dexamethasone-induced cell growth arrest [6]. However, the expression of ANXA1 has not been well studied in breast cancer. Whereas expression of ANXA1 was found reduced or lost in multiple cancers including esophageal [7-9], gastric [10], head and neck [11], prostate [7,12-14], and breast cancers [3,15] as well as B-cell non-Hodgkin lymphoma [16], others reported that ANXA1 is upregulated in a drug-resistant stomach cancer cell line [17], an androgen-independent prostate cancer cell line [18], pancreatic cancer [19], hepatocellular carcinoma [20], and mammary adenocarcinoma [21]. For each of these malignancies, the role of ANXA1 expression in tumor initiation or progression remains unclear. There are a number of theories on the possible function of ANXA1 in cancer development. For example, ANXA1 has been linked with reduced cell proliferation involving the regulation of the extracellular signal-regulated kinase (ERK)/mitogen-activated protein kinase (MAPK) signal transduction pathway. Forced expression of ANXA1 activated the ERK/MAPK pathway and reduced cell proliferation by disrupting actin skeleton and ablating cyclin D1 expression [22,23]. In addition, when treated with an anticancer agent, perillyl alcohol, ANXA1 expression in breast cancer was up-regulated, along with other apoptosisrelated genes, p21 (Cip1/WAF1), Bax, and Bad [24]. A

recent study showed that ANXA1 expression conferred drug resistance in breast cancer cells [25]. In our previous study, ANXA1 expression was found suppressed or lost, both at the messenger RNA level by realtime reverse transcription-polymerase chain reaction analysis and a breast cancer tissue profiling array, and at the protein level by immunohistochemical staining of a panel of tissue samples from patients with breast cancer [3]. In this report, we further characterized the expressional patterns of ANXA1 in breast cancer tissues of different types and stages by using a high-density breast cancer tissue microarray (TMA).

2. Materials and methods 2.1. Breast TMA The breast TMA was constructed similar to those arrays previously described [26-28]. The breast TMA was built with institutional review board approval from archived formalin-fixed, paraffin-embedded breast tissue samples (Department of Pathology and Laboratory Medicine at the UCLA Medical Center). The breakdown of histopathologies is summarized in Table 1. The breast cancer tissue encompasses materials from 242 surgical cases of 210 patients, whose procedures took place between 1995 and

Table 1 Histopathologic distribution of tissue spots in the breast TMA Histology

Number of spots

Normal Sclerosing adenosis Intraductal papilloma CAPS DH Lobular hyperplasia ADH ALH DCIS LCIS IDC Invasive lobular carcinoma Invasive tubular carcinoma Mucinous (colloid) carcinoma Medullary carcinoma Mixed ductal and lobular carcinoma Mixed ductal and tubular carcinoma Inflammatory carcinoma Total informative spots

314 2 3 10 30 1 5 11 209 11 474 28 9 14 6 19 10 2 1158

Annexin A1 expression is decreased in breast cancer 2000. The original archived hematoxylin-eosin–stained slides were reviewed by a pathologist (S. B.). At least 3 samples of each histology were taken from donor tissue blocks to fully represent each case. When available, matched benign tissue was also arrayed from each case. Of the 175 cases of patients with an invasive tumor, 70 were associated with metastases to lymph nodes.

2.2. Immunohistochemistry The breast TMA was evaluated for ANXA1 expression using immunohistochemical staining. Briefly, 4-lm-thick TMA sections were cut by using the Paraffin TapeTransfer System (Instrumedics, Hackensack, NJ). The sections were deparaffinized in xylenes and rehydrated through a series of graded alcohols. Endogenous peroxidase activity was quenched with 3% hydrogen peroxide in methanol for 10 minutes. For epitope retrieval, the TMA sections were immersed in a 968C solution of 0.01 mol/L sodium citrate buffer (pH 6.0) for 30 minutes. Nonspecific binding was blocked by incubating with 5% normal goat serum for 20 minutes at room temperature. Afterward, slides were incubated overnight at 48C with primary antiANXA1 monoclonal antibody (catalog no. 610067; BD Transduction Laboratories, Lexington, KY) diluted to 0.06 lg/mL. As a negative control, isotype-and concentrationmatched nonimmune IgG (Abcam Inc, Cambridge, MA) was also assayed. Staining was detected according to the manufacturer’s recommended protocol using a two-step, biotin-free, dextran-based EnVision+ immunostaining system with 3,3V-diaminobenzidine as chromogen (DAKO, Carpinteria, CA). After visualization, the TMA sections were counterstained with Harris’ hematoxylin, dehydrated, and coverslipped.

2.3. Tissue array scoring Semiquantitative assessment of antibody staining was performed as previously described [26,27] by a pathologist (Y. E.). Three target cell types—normal glandular epithelial cells, myoepithelial cells, and malignant epithelial cells— were identified for scoring. The initial grading and histology assignments were evaluated on slides from the pathology archives and confirmed with hematoxylin-eosin–stained TMA slides. For each TMA tissue spot, the staining of the normal glandular cells, myoepithelial cells, and malignant cells were evaluated separately. The intensity of the staining for each target cell type was scored on a 0 to 3 scale (where 0 was negative for signal; 1, weak; 2, moderate; and 3, strong staining). In addition, the percentage of target cells stained at each intensity (range, 0%-100%) was recorded for each target cell type. Results are presented as the percentage of cells stained at each intensity and/or by an integrated value. Integrated staining represents a combined value of the frequency of staining at each intensity level and was determined by the following formula: 3(x%) + 2( y%) + 1(z%) [26,27].

1585

2.4. Statistical analysis For comparisons and statistical analyses, values of bpositivityQ or bintegrated stainingQ were used. Positivity refers to the percentage of cells that had any detectable ANXA1 expression (ie, intensities of 1, 2, and 3). For nonparametric 2-group and multigroup comparison, the Mann-Whitney and the Kruskal-Wallis tests were used, respectively, as previously described [26,27]. For comparison of paired primary tumors with corresponding lymph node metastases (both tissues removed from primary surgery), Wilcoxon signed rank test was used. All statistical analyses were performed with software package R (http:// www.r-project.org) or Statsview Version 5.0 (SAS Institute, Cary, NC).

3. Results Previously, we observed that the levels of ANXA1 protein were decreased in breast cancer cells compared with normal breast glandular cells [3]. In this study, we examined a much larger and more diverse set of samples using TMA technology. A breast cancer TMA was constructed from archived paraffin-embedded samples at the UCLA Medical Center. This TMA included 242 surgical cases from 210 patients seen at UCLA between 1995 and 2000. On average, for each case, we arrayed 3 to 4 representative areas of pathology and 3 to 4 areas of adjacent morphologically normal tissue. In addition, the breast TMA contained 31 spots of normal breast tissues obtained from breast reduction procedures. The characteristics of the breast TMA are summarized in Table 1.

3.1. Glandular and myoepithelial cells in nonmalignant breast tissues express ANXA1 The TMA was stained with a monoclonal antibody recognizing ANXA1 or with an isotype control. Representative images are shown in Fig. 1. In nonmalignant breast tissue, there was a modest expression of ANXA1 in ductal and glandular epithelial cells. More intense staining was observed in myoepithelial cells (Fig. 1A). The latter cells were classified as myoepithelial type based on their morphology, location, and positivity for p63 staining (data not shown). Other stromal cells such as fibroblasts and endothelial cells were also positive for ANXA1 expression (Fig. 1). Here, we will focus on the report of ANXA1 expression in glandular and myoepithelial cells.

3.2. ANXA1 expression in malignant glandular epithelium compared with nonmalignant counterpart We further assessed ANXA1 protein expression and cellular location in a large sample population by using TMA technology. As shown in Fig. 2, ANXA1 expression in

1586

D. Shen et al.

Fig. 1 ANXA1 expression in breast tissue samples. The breast TMA was prepared and stained as described in Materials and methods. A, Normal breast tissue. B, DH. C, DCIS. D, IDC. Arrows indicate examples of epithelial cells (e) and myoepithelial cells (me) (original magnification 200).

hyperplastic lesions, including ductal hyperplasia, is slightly higher than that in normal epithelial cells, but it is not statistically significant ( P N .1). However, the expression of ANXA1 is significantly decreased in ductal carcinoma in situ (DCIS) and invasive breast cancers. In Fig. 2, data are expressed by an integrated value, which simultaneously quantified both staining intensity and the percentage of glandular cells stained at a given intensity (see Materials and methods). A breakdown of ANXA1 staining pattern of various histopathologies is summarized in Table 2. Several histologic subtypes also showed decreased ANXA1 expression compared with normal cells, such as columnar cell metaplasia with apical snouts (CAPS), mucinous carcinoma, and mixed invasive carcinoma. However, the number of samples as well as the number of patient cases tested were too small to be conclusive. Eighty-seven percent of the nonmalignant TMA spots were derived from tumor-adjacent bnormalQ tissues. The ANXA1 expression in these tissue spots appeared to be similar to normal tissues derived from breast reduction procedures. Normal tissues from both sources showed an indistinguishable pattern of ANXA1 intensity, frequency, and distribution ( P = .122, 31 spots examined).

Fig. 2 ANXA1 expression in breast glandular epithelial cells. The integrated maximum intensity (range of 0-3) is shown here for normal, DH, DCIS, IDC, and other breast cancers. Note that there is a significant difference between DCIS, IDC, and other breast cancers compared with either normal glandular epithelial cells or DH lesions ( P b .001 for all stated comparisons). 1Other breast cancers include invasive lobular carcinoma, mucinous (colloid) carcinoma, invasive tubular carcinoma, mixed invasive carcinoma (ductal tubular or ductal lobular), and medullary carcinoma.

Annexin A1 expression is decreased in breast cancer

1587

Table 2 The average integrated scoring of ANXA1 in glandular cells from various breast histopathologies Histologies

n

Average integrated staining (F SEM)

Normal DH ADH ALH CAPS LCIS DCIS IDC Invasive lobular carcinoma Mucinous (colloid) carcinoma Mixed invasive carcinoma (ductal, tubular) Invasive tubular carcinoma Mixed invasive carcinoma (ductal, lobular) Medullary carcinoma

314 30 5 11 10 11 209 474 28

0.9 1.0 1.0 0.9 0.2 0.6 0.4 0.2 0.1

14

0*

10

0

9 19

0y 0z

6

F F F F F F F F F

0.05 0.14 0.28 0.24 0.10* 0.38 0.05* 0.02* 0.03*

0.7 F 0.14

* P b .001 compared with normal tissues (nonmalignant) or DH. y P = .007 compared with normal tissues (nonmalignant). z P = .004.

We further examined the level of ANXA1 in primary breast cancer cells compared with metastases from these cases to regional lymph nodes. Interestingly, although the expression of ANXA1 in both primary tumor and lymph node metastases was lower than in normal breast epithelium ( P b .001 for both), the primary tumor showed significantly lower ANXA1 expression compared with the corresponding metastatic cells in the lymph node (Fig. 3; P = .0427). This difference of ANXA1 expression was not associated with tumor grade. After substratification by grade, ANXA1 expression in metastatic cancer cells is still higher than that in primary tumors. Fig. 4 shows a bivariate scattergram of ANXA1 expression for individual cases from which the expression data for both primary tumor and corresponding lymph node metastases of similar grade were available. In most cases, the trend of a higher expression of ANXA1 was observed in the metastatic cells (ANXA1 expression: metastases N primary tumor, 13 cases; primary tumor N metastases, 6 cases; no expression in both primary tumor and metastases in 10 cases).

Fig. 3 ANXA1 expression in primary tumor cells and lymph node metastases. The difference in integrated maximum intensity of ANXA1 is shown between primary breast invasive carcinomas and nodal metastasis with a statistical significance of P = .0427 by Wilcoxon analysis.

hyperplasia [DH], atypical ductal hyperplasia [ADH], and atypical lobular hyperplasia [ALH]), and early-stage malignant breast tissues (lobular carcinoma in situ [LCIS] and DCIS). As shown in Fig. 5, all benign histologies had comparably higher ANXA1 staining. However, in DCIS, the integrated maximum intensity staining of ANXA1 was significantly lower than that of the benign histologies ( P b .0001 for each pairwise comparison).

3.3. ANXA1 expression in myoepithelial cells in normal and malignant breast tissues As noted previously, ANXA1 is expressed in myoepithelial cells lining the mammary ducts (Fig. 1). Myoepithelial cell identity was confirmed by the expression of p63, a known myoepithelial cell marker [29]. We further examined whether ANXA1 expression in myoepithelial cells was different among normal, hyperplastic conditions (ductal

Fig. 4 Bivariate scattergram of ANXA1 expression in primary tumor cells and lymph node metastatic cells. The integrated maximum intensity of ANXA1 is compared between primary breast invasive carcinomas and their corresponding lymph node metastases. Twenty-nine cases are shown, 10 of which showed no ANXA1 expression in primary tissue or lymph node metastases.

1588

D. Shen et al. from the TMA spots of normal, hyperplastic, and DCIS that had any degree of ANXA1 positivity. As shown in Fig. 6, the percentage of myoepithelial cells with strong staining decreased, whereas the percentage of myoepithelial cells without any ANXA1 staining increased accordingly when comparing nonmalignant tissues with DCIS lesions. In contrast, the distribution of myoepithelial staining in normal versus hyperplastic ductal lesions was similar. These results suggest that the loss of ANXA1 expression occurs early during malignant transformation.

4. Discussion

Fig. 5 ANXA1 expression in breast myoepithelial cells of various histologic subsets. ANXA1 staining was presented as the integrated maximum intensity (as described in Materials and methods). NL indicates normal. The only statistically significant difference was observed in the comparison of DCIS with other histologies ( P b .0001).

One hallmark of breast cancer is the disappearance of myoepithelial cells in invasive cancer [30]. To determine whether the decreased integrated maximal intensity of ANXA1 staining observed in DCIS samples was simply due to a reduction in the number of these cells or to additional mechanisms such as a decreased level or frequency of cellular ANXA1 protein expression, we examined the expression of ANXA1 in myoepithelial cells

The expression of ANXA1 was evaluated in tissue samples derived from women with breast cancer by using a high-density TMA linked to pathology and clinical information. Consistent with previous findings [3], ANXA1 expression was primarily observed in normal glandular cells and myoepithelium. In invasive ductal carcinoma (IDC), there was a consistent decrease in both the level and the frequency of ANXA1 expression in malignant cells. A significant decrease of ANXA1 was first detected in DCIS lesions, suggesting that it might be an early event in breast cancer development. The ANXA1 expression in hyperplastic lesions, including DH (Fig. 2) and atypical ductal and lobular hyperplasia (not shown) is slightly higher than that in normal epithelial cells, albeit not statistically significant. Although the significance of such a change is not clear, it may represent an early response to the cellular derangement, requiring further study. Surprisingly, whereas the suppressed ANXA1 is involved in breast cancer development, a higher level of ANXA1 was observed in lymph

Fig. 6 Distribution of breast myoepithelial cells with different ANXA1 staining intensities among different breast histopathologies. The ANXA1 staining was categorized as follows: 0, not detectable; 1, weak; 2, moderate; 3, strong) as broken down by normal, DH, and DCIS.

Annexin A1 expression is decreased in breast cancer node metastases when compared with primary breast cancer (Fig. 3). A similar finding was observed by Pencil and Toth in rat metastatic mammary cancer [31] and by Ahn et al in human breast cancer [21]. ANXA1 expression was also found to be down-regulated by breast cancer metastasis suppressor 1 gene in MDA-MB-435 breast cancer cells metastatic to the lungs in an athymic mouse model [32]. A more recent study found an autocrine/paracrine role for membrane ANXA1 in stimulating the migration of SKCO15 colorectal cancer cells through activation of n-formyl peptide receptors, a leukocyte migratory factor [33]. Therefore, ANXA1 may play a multifaceted role in breast cancer development, progression, and metastasis. This study also reports an abundant expression of ANXA1 in glandular myoepithelial cells. A similar staining pattern was previously observed by Schwarz-Albiez et al [15]. Not surprisingly, ANXA1 levels drop in IDC in part because of the known scant or absent myoepithelial cells in malignant lesions, but also largely because of decreased levels in this cell population. A previous in vitro study found that human myoepithelial cells exerted an antiproliferative effect on breast cancer cells by p21WAF1/CIP1 induction, G2/M arrest, and apoptosis [34]. The mRNA profiling of myoepithelial cells from diverse sources of benign myoepithelial lesions suggested a tumor-suppressor phenotype [35]. We are currently examining the extent of ANXA1 reduction in myoepithelial cells of DCIS as well as the significance of such a decrease in tumor progression.

4.1. Potential role of ANXA1 in malignancy Although ANXA1 has been studied in many types of cancer, there is currently no consensus on the mechanism of how this protein influences tumor initiation and/or progression. The annexin family consists of a group of structurally related calcium-binding proteins. These proteins share a conserved domain (annexin domain) that binds to phospholipids. Each annexin has a unique amino terminal domain that determines specificity and provides functional diversity [4]. Because annexin family proteins were found to bind to many different ligands, there are numerous mechanisms by which ANXA1 can potentially be linked to malignancy. For example, ANXA1 was reported to interact with profilin, suggesting its role in regulating the membrane-associated cytoskeleton [36]. There is also evidence that annexin family members can participate in cell signaling [37,38]. For example, ANXA1 is a known substrate of tyrosine kinase of the epidermal growth factor receptor [39]. Recent studies suggested that ANXA1 expression is associated with activation of the ERK pathway [22]. Forced expression of ANXA1 activates the ERK/MAPK pathway and reduces cell proliferation by disruption of the actin skeleton and ablation of cyclin D1 expression [23]. Multiple members of the annexin family, including annexins 1, 2, 5, 6, and 7, were reported to be involved in negative regulation of cell proliferation. ANXA1 was therefore postulated to have

1589 tumor suppressor properties [4,40]. Consistent with this hypothesis, expression of ANXA1 was found to be reduced or lost in many cancers including esophageal [7-9], gastric [10], head and neck [11], prostate [7,12-14], and breast [3,15] cancers and B-cell non-Hodgkin lymphoma [16]. An induced differentiation of human colon carcinoma cells accompanied by increase in the levels of ANXA1 and ANXA5 further supports a tumor suppressive role of ANXA1 [41]. There is also other evidence that ANXA1 may play a role in proapoptotic pathway. For example, expression of ANXA1 was found to be an endogenous ligand that mediates apoptotic cell engulfment [42] and was associated with caspase-dependent apoptosis in BZR bronchoalveolar epithelial cells [43]. When treated with an anticancer agent perillyl alcohol, ANXA1 expression in breast cancer was up-regulated along with other apoptosis-related genes, including p21 (Cip1/WAF1), Bax, and Bad [24]. Whereas many studies supported a tumor suppressive role for ANXA1, inconsistent observations were also reported, including up-regulated ANXA1 in a drug-resistant stomach cancer cell line [17], an androgen-independent LNCaP prostate cancer cell line [18], hairy cell leukemia [44], pituitary carcinoma [45], pancreatic carcinoma [19], heptocellular carcinoma [20], and breast cancer [21]. Although these discrepancies have not been explained, reasons could include differences in organ specificity, alterations from cell culture, dynamic changes of the expression during tumor progression, and/or variance in antibody reagents used. The monoclonal antibody first used to document lower expression of ANXA1 in breast cancer was obtained from Oncogene Science, Inc (Manhasset, NY) [15]. The monoclonal antibody used to describe higher expression of ANXA1 in breast cancer was from Zymed (South San Francisco, CA) [21]. However, both antibodies are no longer commercially available for further comparison.

4.2. Conclusion This study has characterized the pattern of ANXA1 expression in both glandular epithelial cells and myoepithelial cells in normal, hyperplastic, and malignant breast tissues. The expression of ANXA1 decreases in both DCIS and invasive carcinoma compared with nonmalignant glandular cells. The expression of ANXA1 in myoepithelial cells is also decreased in DCIS compared with nonmalignant tissues. Decreasing ANXA1 expression is correlated with breast cancer development and histologic progression. Our study suggests that ANXA1 may negatively regulate the proliferation of breast epithelial cells and may contribute to maintaining normal breast biology.

References [1] Jemal A, Murray T, Ward E, et al. Cancer statistics. CA Cancer J Clin 2005;55:10 - 30.

1590 [2] Shen D, He J, Chang HR. In silico identification of breast cancer genes by combined multiple high-throughput analyses. Int J Mol Med 2005;15:205 - 12. [3] Shen D, Chang HR, Chen Z, et al. Loss of annexin A1 expression in human breast cancer detected by multiple high-throughput analyses. Biochem Biophys Res Commun 2005;326:218 - 27. [4] Gerke V, Moss SE. Annexins: from structure to function. Physiol Rev 2002;82:331 - 71. [5] Buckingham JC, Flower RJ. Lipocortin 1: a second messenger of glucocorticoid action in the hypothalamo-pituitary-adrenocortical axis. Mol Med Today 1997;3:296 - 302. [6] Croxtall JD, Flower RJ. Lipocortin 1 mediates dexamethasoneinduced growth arrest of the A549 lung adenocarcinoma cell line. Proc Natl Acad Sci U S A 1992;89:3571 - 5. [7] Paweletz CP, Ornstein DK, Roth MJ, et al. Loss of annexin 1 correlates with early onset of tumorigenesis in esophageal and prostate carcinoma. Cancer Res 2000;60:6293 - 7. [8] Xia SH, Hu LP, Hu H, et al. Three isoforms of annexin I are preferentially expressed in normal esophageal epithelia but downregulated in esophageal squamous cell carcinomas. Oncogene 2002; 21:6641 - 8. [9] Hu N, Flaig MJ, Su H, et al. Comprehensive characterization of annexin I alterations in esophageal squamous cell carcinoma. Clin Cancer Res 2004;10(18 Pt 1):6013 - 22. [10] Hippo Y, Yashiro M, Ishii M, et al. Differential gene expression profiles of scirrhous gastric cancer cells with high metastatic potential to peritoneum or lymph nodes. Cancer Res 2001;61: 889 - 95. [11] Garcia Pedrero JM, Fernandez MP, Morgan RO, et al. Annexin A1 down-regulation in head and neck cancer is associated with epithelial differentiation status. Am J Pathol 2004;164:73 - 9. [12] Xin W, Rhodes DR, Ingold C, Chinnaiyan AM, Rubin MA. Dysregulation of the annexin family protein family is associated with prostate cancer progression. Am J Pathol 2003;162:255 - 61. [13] Kang JS, Calvo BF, Maygarden SJ, Caskey LS, Mohler JL, Ornstein DK. Dysregulation of annexin I protein expression in high-grade prostatic intraepithelial neoplasia and prostate cancer. Clin Cancer Res 2002;8:117 - 23. [14] Smitherman AB, Mohler JL, Maygarden SJ, Ornstein DK. Expression of annexin I, II and VII proteins in androgen stimulated and recurrent prostate cancer. J Urol 2004;171(2 Pt 1):916 - 20. [15] Schwartz-Albiez R, Koretz K, Moller P, Wirl G. Differential expression of annexins I and II in normal and malignant human mammary epithelial cells. Differentiation 1993;52:229 - 37. [16] Vishwanatha JK, Salazar E, Gopalakrishnan VK. Absence of annexin I expression in B-cell non-Hodgkin’s lymphomas and cell lines. BMC Cancer 2004;4:8. [17] Sinha P, Hutter G, Kottgen E, Dietel M, Schadendorf D, Lage H. Increased expression of annexin I and thioredoxin detected by twodimensional gel electrophoresis of drug resistant human stomach cancer cells. J Biochem Biophys Methods 1998;37:105 - 16. [18] Vaarala MH, Porvari K, Kyllonen A, Vihko P. Differentially expressed genes in two LNCaP prostate cancer cell lines reflecting changes during prostate cancer progression. Lab Invest 2000;80: 1259 - 68. [19] Bai XF, Ni XG, Zhao P, et al. Overexpression of annexin 1 in pancreatic cancer and its clinical significance. World J Gastroenterol 2004;10:1466 - 70. [20] Masaki T, Tokuda M, Ohnishi M, et al. Enhanced expression of the protein kinase substrate annexin in human hepatocellular carcinoma. Hepatology 1996;24:72 - 81. [21] Ahn SH, Sawada H, Ro JY, Nicolson GL. Differential expression of annexin I in human mammary ductal epithelial cells in normal and benign and malignant breast tissues. Clin Exp Metastasis 1997; 15:151 - 6.

D. Shen et al. [22] Alldridge LC, Harris HJ, Plevin R, Hannon R, Bryant CE. The annexin protein lipocortin 1 regulates the MAPK/ERK pathway. J Biol Chem 1999;274:37620 - 8. [23] Alldridge LC, Bryant CE. Annexin 1 regulates cell proliferation by disruption of cell morphology and inhibition of cyclin D1 expression through sustained activation of the ERK1/2 MAPK signal. Exp Cell Res 2003;290:93 - 107. [24] Ariazi EA, Satomi Y, Ellis MJ, et al. Activation of the transforming growth factor beta signaling pathway and induction of cytostasis and apoptosis in mammary carcinomas treated with the anticancer agent perillyl alcohol. Cancer Res 1999;59:1917 - 28. [25] Wang Y, Serfass L, Roy MO, Wong J, Bonneau AM, Georges E. Annexin-I expression modulates drug resistance in tumor cells. Biochem Biophys Res Commun 2004;314:565 - 70. [26] Seligson DB, Horvath S, Shi T, et al. Global histone modification patterns predict risk of prostate cancer recurrence. Nature 2005;435: 1262 - 6. [27] Seligson D, Horvath S, Huerta-Yepez S, et al. Expression of transcription factor Yin Yang 1 in prostate cancer. Int J Oncol 2005; 27:131 - 41. [28] Kononen J, Bubendorf L, Kallioniemi A, et al. Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med 1998;4:844 - 7. [29] Barbareschi M, Pecciarini L, Cangi MG, et al. p63, a p53 homologue, is a selective nuclear marker of myoepithelial cells of the human breast. Am J Surg Pathol 2001;25:1054 - 60. [30] Man YG, Sang QX. The significance of focal myoepithelial cell layer disruptions in human breast tumor invasion: a paradigm shift from the bprotease-centeredQ hypothesis. Exp Cell Res 2004;301: 103 - 18. [31] Pencil S, Toth M. Elevated levels of annexin I protein in vitro and in vivo in rat and human mammary adenocarcinoma. Clin Exp Metastasis 1998;16:113 - 21. [32] Cicek M, Samant RS, Kinter M, Welch DR, Casey G. Identification of metastasis-associated proteins through protein analysis of metastatic MDA-MB-435 and metastasis-suppressed BRMS1 transfected-MDAMB-435 cells. Clin Exp Metastasis 2004;21:149 - 57. [33] Babbin BA, Lee WY, Parkos CA, et al. Annexin I regulates SKCO-15 cell invasion by signaling through formyl peptide receptors. J Biol Chem 2006;281:19588 - 99. [34] Shao ZM, Nguyen M, Alpaugh ML, O’Connell JT, Barsky SH. The human myoepithelial cell exerts antiproliferative effects on breast carcinoma cells characterized by p21WAF1/CIP1 induction, G2/M arrest, and apoptosis. Exp Cell Res 1998;241:394 - 403. [35] Barsky SH. Myoepithelial mRNA expression profiling reveals a common tumor-suppressor phenotype. Exp Mol Pathol 2003;74: 113 - 22. [36] Alvarez-Martinez MT, Porte F, Liautard JP, J. Sri Widada. Effects of profilin-annexin I association on some properties of both profilin and annexin I: modification of the inhibitory activity of profilin on actin polymerization and inhibition of the self-association of annexin I and its interactions with liposomes. Biochim Biophys Acta 1997;1339: 331 - 40. [37] Schmitz-Peiffer C, Browne CL, Walker JH, Biden TJ. Activated protein kinase C alpha associates with annexin VI from skeletal muscle. Biochem J 1998;330(Pt 2):675 - 81. [38] Wen Y, Edelman JL, Kang T, Sachs G. Lipocortin V may function as a signaling protein for vascular endothelial growth factor receptor-2/ Flk-1. Biochem Biophys Res Commun 1999;258:713 - 21. [39] Rothhut B. Participation of annexins in protein phosphorylation. Cell Mol Life Sci 1997;53:522 - 6. [40] Bastian BC. Annexins in cancer and autoimmune diseases. Cell Mol Life Sci 1997;53:554 - 6. [41] Guzman-Aranguez A, Olmo N, Turnay J, et al. Differentiation of human colon adenocarcinoma cells alters the expression and

Annexin A1 expression is decreased in breast cancer intracellular localization of annexins A1, A2, and A5. J Cell Biochem 2005;94:178 - 93. [42] Arur S, Uche UE, Rezaul K, et al. Annexin I is an endogenous ligand that mediates apoptotic cell engulfment. Dev Cell 2003;4:587 - 98. [43] Debret R, El Btaouri H, Duca L, et al. Annexin A1 processing is associated with caspase-dependent apoptosis in BZR cells. FEBS Lett 2003;546:195 - 202.

1591 [44] Falini B, Tiacci E, Liso A, et al. Simple diagnostic assay for hairy cell leukaemia by immunocytochemical detection of annexin A1 (ANXA1). Lancet 2004;363:1869 - 70. [45] Mulla A, Christian HC, Solito E, Mendoza N, Morris JF, Buckingham JC. Expression, subcellular localization and phosphorylation status of annexins 1 and 5 in human pituitary adenomas and a growth hormone-secreting carcinoma. Clin Endocrinol (Oxf) 2004; 60:107 - 19.