Review

Periostin: A Novel Prognostic and Therapeutic Target For Genitourinary Cancer? Pier Vitale Nuzzo,1,2 Giulia Buzzatti,1,2 Francesco Ricci,1,3 Alessandra Rubagotti,1,2,3 Francesca Argellati,1,3 Linda Zinoli,1,2,3 Francesco Boccardo1,2,3 Abstract Many of the cellular abnormalities present in solid tumors are structural in nature and involve the proteins of the extracellular matrix (ECM). Periostin is a protein produced and secreted by the fibroblasts as a component of the ECM where it is involved in regulating intercellular adhesion. The expression of periostin has an important physiological role during embryogenesis and growth, namely at the level of bone, dental, and cardiac tissues. Many studies indicate that periostin plays an important role for tumor progression in various types of cancer, such as colon, lung, head and neck, breast, ovarian, and prostate. To the best of our knowledge, a limited number of studies have investigated periostin expression in urogenital cancer, such as prostate, bladder, penile, and renal cancer, and no studies were performed in testis cancer. In this review article, we summarize the most recent knowledge of periostin, its genetic and protein structure, and the role of the different isoforms identified and sequenced so far. In particular, we focus our attention on the role of this protein in genitourinary tumors, trying to emphasize the role not only as a possible prognostic marker, but also as a possible target for the development of future anticancer therapies. Clinical Genitourinary Cancer, Vol. -, No. -, --- ª 2014 Elsevier Inc. All rights reserved. Keywords: Biomarker, Extracellular matrix protein, Genitourinary tumors, Isoform, POSTN protein

Introduction Periostin, originally named as osteoblast-specific factor-2, was first identified in 1993 as a putative cell adhesion protein for preosteoblasts in a mouse osteoblastic MC3T3-E1 cell line.1 This protein was then renamed periostin because of its preferential location in the periosteum and periodontal ligament.2 Periostin is an extracellular matrix (ECM) protein involved in regulating intercellular adhesion via an interaction with other ECM proteins, such as fibronectin, tenascin-C, collagen V, and periostin itself.3,4 More recent studies have revealed that periostin is an important regulator of bone and tooth formation and is essential for heart development Pier Vitale Nuzzo and Giulia Buzzatti should be considered equally as first author. 1 Academic Unit of Medical Oncology (Medical Oncology B), University of Genoa, School of Medicine, Genoa, Italy 2 Department of Internal Medicine, University of Genoa, School of Medicine, Genoa, Italy 3 IRCCS San Martino University Hospital - IST National Cancer Research Institute, Genoa, Italy

Submitted: Dec 6, 2013; Revised: Jan 29, 2014; Accepted: Feb 12, 2014 Address for correspondence: Francesco Boccardo, MD, IRCCS San Martino University Hospital—IST National Cancer Research Institute, Academic Unit of Medical Oncology (Medical Oncology B), Largo Rosanna Benzi 10, 16132 Genoa, Italy Fax: þ39010352753; e-mail contact: [email protected]

1558-7673/$ - see frontmatter ª 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clgc.2014.02.005

and healing after acute myocardial infarction.4-6 Periostin is also reexpressed in adult tissues under stress conditions, for instance in the heart under pressure or volume overload,7 in the skeletal muscle after injury,8 in the bone after fracture,9 in pulmonary aortic smooth muscle cells in response to hypoxia,10 and in chronic sinusitis.11 Periostin protein expression was observed in a wide variety of normal adult and fetal tissues, such as embryonic periosteum, placenta, periodontal ligaments, cardiac valves, adrenal glands, lung, testis, thyroid, stomach, vagina, ovary, colon, prostate, and breast.1,12 Recently, it has been reported that periostin is frequently overexpressed in various types of human cancer cell lines in vitro and human cancer tissues in vivo (including breast cancer, colon cancer, nonesmall-cell lung carcinoma, head and neck cancer, ovarian cancer, pancreatic ductal adenocarcinoma, melanoma, gastric cancer, oral squamous cell carcinoma [SCC], thymoma and neuroblastoma), underlining the putative role of this protein in tumor development.13 Although the role of this protein in the process of carcinogenesis remains to be clarified, it is known that its overexpression in cancer stroma and/or the epithelium of the neoplasms is usually associated with the most malignant phenotypes and the poorest outcomes.13

Clinical Genitourinary Cancer Month 2014

-1

Periostin and Genitourinary Cancer Indeed, the matrix protein periostin was shown to be a marker and an inducer of epithelialemesenchymal transition (EMT), a process that is responsible for the dissemination of primary tumor epithelial cells to the sites of metastasis and the dedifferentiation program that leads to increased malignant behavior of tumors.14,15 To the best of our knowledge, a limited number of studies have investigated the expression of periostin in urogenital cancers, such as prostate, bladder, penile, and renal cancer, and nobody so far has reported about periostin in testicular cancer. In prostate, renal, and penile cancer, the upregulation of periostin was usually associated with a more aggressive tumor behavior and advanced stage; conversely, in bladder cancer, high-grade urothelial carcinomas appear to be associated with the lowest levels of periostin; moreover, in vitro transfection of urothelial tumor cells with the periostin gene was shown to neutralize their invasive potential.15 In cancer, namely in urogenital tumors, the following crucial issues related to the role of periostin and function have not been fully elucidated: (1) whether the production and secretion of periostin occurs in tumor or stromal cells or in both compartments; (2) where periostin is located at the cell level in epithelial and stromal cells; and (3) whether, and should this be the case, under which conditions, periostin functions as a tumor promoter or as a tumor suppressor. In this review, we will first summarize the current knowledge about the functional roles of periostin; we will then focus on findings related to the current understanding of the specific role of periostin in tumorigenesis, with a special focus on urogenital cancer; finally, current findings supporting the role of periostin as a prognostic marker and a putative target for anticancer therapy will be reviewed.

Literature Review A systematic literature search was performed using the Medical Subject Headings function on PubMed, from 1993 to October 2013, using the key words, “periostin,” “isoform,” “neoplasm,”

“prostate neoplasm,” “bladder neoplasm,” “renal neoplasm,” “testicular neoplasm,” and “penile neoplasm.” An independent search of the Cochrane electronic databases was also performed to ensure that no additional studies were overlooked. Our search strategy yielded 8 articles concerning prostate cancer (PCa), 4 articles concerning bladder neoplasms, 5 articles concerning renal cancer, and just 1 article relative to penile cancer. As previously mentioned, we were not able to find any article dealing with periostin and testicular cancer. We also identified a further 137 potentially relevant articles related to periostin and neoplasms in general, or relative to other tumor types, or with periostin isoforms. Only the most complete, recent, and updated reports were used if believed to fit in with the scopes of the present literature review.

Structure of Periostin The periostin gene has been cloned in humans and in several animal species (including mouse, rat, chicken, bovine, and xenopus). The gene is located at locus 13q13.3 in humans and 3C in mice.16 The periostin gene in humans and mice has 23 exons, with a genomic footprint covering approximately 36 and 30 kilobases, respectively. The terminal exons in humans and mice are proteincoding. The length of the mouse periostin cDNA is 3187 base pair (bp), with an 18-bp 50 untranslated region, a 733-bp 30 untranslated region, and a 2436-bp open reading frame that encodes a protein of 811 amino acids with a molecular weight (MW) of 90.2 kDa (Fig. 1).1 The template of the 3D protein structure of periostin was obtained from the SWISS-MODEL Repository database (Fig. 2).17,18 The first 2 isolated human periostin cDNAs were screened from placental and osteosarcoma cDNA libraries using mouse periostin cDNA as a probe. The human placental periostin open reading frame encodes for a protein of 779 amino acids, with a MW of 87.0 kDa; the human osteosarcoma periostin open reading frame encodes for a protein of 836 amino acids with a MW of 93.3 kDa.1 Periostin is highly conserved between human and mouse, with 89.2%

Figure 1 Structure of Periostin and Its Isoforms. Periostin Protein Sequence Showing the Signal Sequence, the EMI Domain, the 4 FAS-1 Domains, and the Variable Domain With 9 Different Exons, Whose Combination Gives Rise to 4 Different Isoforms Sequenced and Annoted

2

-

Abbreviations: EMI ¼ EMILIN-like; FAS ¼ fasciclin.

Clinical Genitourinary Cancer Month 2014

Pier Vitale Nuzzo et al Figure 2 The Template of the 3D Protein Structure of Periostin Obtained From the SWISS-MODEL Repository Database

mediate big-h3 adhesion to a3b1.21 In addition, the FAS-1 domains contain an N-terminal recognition site for g-glutamylcarboxylase, which mediates the posttranslational modification of glutamate to g-carboxyglutamate.22 The C-terminal region (exons 15-23), containing proteolytic cleavage sites, regulates the organization of the cell matrix. Periostin interacts with collagen I,23 fibronectin,24 and Notch1,25 through its EMI domain and with tenascin-C24 and bone morphogenetic protein-126 through its FAS-1 domains. It also possesses 4 putative N-glycosylation sites and a heparin-binding domain and binds glycoproteins, glycosaminoglycans, and proteoglycans by its C-region.2,27 The interactions with ECM proteins and alternative splicing of its domains generate different isoforms of human periostin.21

Periostin Isoforms

homology for the entire protein and 90.1% homology for the mature form. However, compared with other regions within the mature periostin protein, the C-terminal region shows slightly less conservation, with 85.5% identity.1 The protein structure of periostin is composed of an N-terminal secretory signal peptide followed by an EMILIN-like (EMI) domain rich in cysteine, 4 internal repeated and conserved fasciclin (FAS)-1 domains, and a C-terminal variable hydrophilic domain.1,2 The N-terminal region (exon 1) is highly conserved; it contains a signal peptide to promote periostin secretion and regulates cell functions by binding to integrins at the plasma membrane of the cell through its FAS domain.1,2 The EMI domain (exons 2 and 3) was first named after its presence in proteins of the EMILIN family and is associated with other domains, such as C1q, laminin-type epidermal growth factorlike, fibronectin type 3, whey acidic protein, zona pellucida, or FAS-1.19,20 Its cysteine-rich region of approximately 75 amino acids promotes the formation of multimers in nonreducing conditions.19 Based on the conservation of the typical FAS-1 domains (exons 3-14), each consisting of 150 amino acids, periostin has been assigned to the fasciclin family, which includes transforming growth factor beta-induced gene clone 3 (big-h3), stabling I and II, MPB-70, algal-cell adhesion molecule, and periostin-like factor (PLF).16 In Drosophila, the ancestral fasciclin domain functions as an adhesion molecule linked to axonal guidance, migration, and differentiation.1,21 The presence of integrin-binding motifs in the second and fourth FAS-1 domains suggests that periostin is implicated in cell adhesion, because these domains have been shown to

Currently, 8 different isoforms of periostin have been identified, but only 4 (including the full-length variant) have been sequenced and annotated.1,28-30 The isoforms of periostin are between 83 and 93 kDa in mass, vary between 751 and 836 amino acids, and differ in their C-terminal sequences, which are characterized by the individual presence or absence of cassette exons 17-21 (UniProtKB/SwissProt, March 2013); the N-terminus is conserved (Fig. 1). Alternative splicing of the C-terminal sequences gives rise to the 4 known periostin isoforms, isoform 1 or osteoblast-specific factor (OSF)-2OS (full-length variant with all exons), isoform 2 or OSF-2p1 (exons 17 and 18 are absent), isoform 3 or PLF (exons 17 and 21 are absent), and isoform 4 (exons 17, 18, and 21 are absent), with the following UniProtKB/Swiss-Prot accession numbers: D13666, D13665, AY140646, and AY918092, respectively. The 4 periostin isoforms were identified from different tissues: isoform 1 from human osteosarcoma, isoform 2 from human placenta, isoform 3 from epithelial ovarian carcinoma, and isoforms 2 and 4 from normal or cancerous human bladder tissue.1,28,29,31 The isolation of different periostin isoforms from different tissues suggests that their expression might be tissue-specific. Based on comparative studies of periostin sequences among vertebrates, Hoersch and Andrade-Navarro suggested that the set of exons that is subjected to alternative splicing is species-dependent.21 These investigators found that in humans, exons 17, 18, 19, and 21, but never exon 20, are alternatively spliced; conversely, in mouse, alternative splicing does not affect exon 20 or exons 17 and 21. The biological effects of periostin isoforms remain to be fully clarified, but specific roles for certain isoforms have been suggested. Multiple isoforms of periostin are expressed in vivo during embryogenesis and in vitro in MC3T3-E1 cells and primary osteoblast cultures. The expression of these isoforms appears to be developmentally regulated, and findings suggest that they are involved in regulating the process of differentiation.16 Hoersch and Andrade-Navarro also suggested that the C-terminal domain of periostin is responsible for binding to the ECM and that the isoforms might exert their influence on the ECM fibrillogenesis differently.21 This hypothesis is supported by the finding that the down-regulation and/or loss by alternative splicing of the nonspliced periostin form, together with the expression of isoform 2, increases invasiveness in vitro and in vivo, as shown in nude mice and in experiments using the human bladder cancer cell line SBT991 and

Clinical Genitourinary Cancer Month 2014

-3

Periostin and Genitourinary Cancer the mouse malignant melanoma cell line B16F10.29 Furthermore, blocking the activity of periostin isoforms in the mouse osteoblast cell line MC3T3-E1 led to a significant reduction in osteoblastspecific differentiation markers.16

Role of Periostin in Carcinogenesis Because periostin was found to be upregulated in various type of human cancer, it has become evident that periostin plays an important role in the acquisition of most hallmarks of cancer cells, such as survival,32 proliferation,33 evasion of apoptosis,13,34,35 angiogenesis, invasion, and metastasis (Table 1).12,14,15,28,33-42 Furthermore, recent studies have focused attention on the putative role played by periostin in the development of EMT in addition to its effect on the ECM fibrillogenesis.38 Although the role of periostin in tumor development is still unclear, some physiopathological mechanisms have been demonstrated in recent years. Integrins are heterodimeric transmembrane receptors that are involved in cellecell and celleECM interactions43; the expression of integrins is frequently altered in cancer cells.44,45 The FAS-1 domain in the N-terminal region of periostin acts as a ligand for integrins to promote metastasis via the activation of the protein kinase B (Akt/PKB) cell survival signaling pathway in cancer cells.33 The FAS-1 domain also induces angiogenesis via the focal adhesion kinase (FAK)-mediated signaling pathway in endothelial cells.36 The integrin receptors of periostin that are thought to be involved in tumorigenesis include avb3, avb5, and a6b4.46-48 In ovarian cancer cells, periostin has been shown to be a ligand for the avb3 and avb5 integrins, resulting in cell adhesion and migration. Of note, monoclonal antibodies against avb3 or avb5 integrins were shown to inhibit adhesion when used as single agents and to completely suppress cell adhesion when used in combination.28

Moreover, the avb3 integrin complex acts as a receptor for periostin and is involved in the activation of the Akt/PKB survival pathway in breast cancer,36 colon cancer, and endothelial cells.33 In pancreatic ductal adenocarcinoma, the a6b4 integrin was demonstrated to act as a cell surface receptor for periostin. The interaction between periostin and this integrin was shown to promote the phosphorylation of FAK and Akt via the activation of the phosphoinoside 3-kinase (PI3-K) pathway.34 The activation of FAK promotes invasiveness, and Akt/PKB, a serine/threonine protein kinase, is now recognized as one of the most central regulators of cell survival and proliferation.32 Akt functions as a cardinal mediator for transducing extracellular and intracellular signals and is positively regulated by PI3-kinase and negatively regulated by phosphatase and tensin homolog (PTEN).32,49 Interestingly, in colon cancer, the Akt/PKB pathway activated by periostin can dramatically enhance tumor growth by augmenting cancer and endothelial cell survival and by protecting the cells from stress-induced apoptosis. This effect has been demonstrated after treatment with an Akt/PKB inhibitor or PTEN overexpression mediated by retrovirus infection, both of which were sufficient to abrogate the periostin effect in promoting cancer cell resistance to stress conditions.33 This effect was also shown after the exposure of colorectal cancer cells to anti-periostin antibodies, which were able to activate apoptosis and to potentiate the effects of 5-fluorouracil chemotherapy.12 Furthermore, periostin has been shown to promote resistance to metastatic stress conditions, such as serum starvation, hypoxia, and loss of adhesion in pancreatic,34,35 breast, and ovarian cancer.13 Recent data showed that in head and neck tumors, the overexpression of periostin promoted invasion and anchorageindependent growth by allowing the cancer cells to survive via the

Table 1 Role of Periostin in Carcinogenesis: Overview of Periostin Function and Correlated Signaling Pathway Tumor Type

Binding Protein Signaling Pathway

Ovarian Cancer

avb3 and avb5 integrins

Pancreatic Cancer

avb3 and avb5 integrins a6b4 integrin via PI3-K/Akt

Colon Cancer

a6b4 integrin via PI3-K/FAK avb3 integrin via Akt/PKB Via VEGF e

Breast Cancer

avb3 via Akt/PKB avb3 via FAK/VEGF receptor Flk-1/KDR

Head and Neck Cancer Including Oral Cancer

avb3 and avb5 integrins

e

NoneSmall-Cell Lung Cancer

e Via VEGF-C expression e e

Periostin Function Supports the adhesion of ovarian epithelial cells Enhances cancer cell motility and migration Promotes cell survival and proliferation and resistance to stress conditions Promotes invasiveness by stimulating motility Promotes cell survival, proliferation, and resistance to stress conditions Induces tumor angiogenesis Promotes hepatic metastases Promotes cell survival Induces tumor angiogenesis Promotes bone metastases Promotes migration, invasion, and anchorage-independent growth Induces tumor angiogenesis Induces tumor lymphangiogenesis Promotes lymph node and lung metastases Promotes lymph node and lung metastases

Reference Gillan et al28

Erkan et al35 Baril et al34 Tai et al12, Bao et al33

Yan and Shao14, Shao et al36

Shao et al36 Kudo et al40,42 Siriwardena et al41 Takanami et al37

Abbreviations: FAK ¼ focal adhesion kinase; Flk-1/KDR ¼ vascular endothelial growth factor receptor 2; PKB ¼ protein kinase B; PI3-K ¼ phosphoinositide 3-kinase; VEGF ¼ vascular endothelial growth factor.

4

-

Clinical Genitourinary Cancer Month 2014

Pier Vitale Nuzzo et al inhibition of the anoikis-related apoptotic pathway, even though periostin alone was not able to promote cell proliferation.40 Intriguingly, the proliferation rate of periostin-overexpressing cells in vitro was slower than the rate of control cells; in vivo, the same periostin-overexpressing tumor cell lines showed a phenotype of accelerated growth and angiogenesis.36 Periostin also has a relevant metastatic potential as a major contributor to EMT by increasing cancer cell motility, migration, invasion, and adhesion. During EMT, cancer cells acquire mesenchymal features, such as the upregulation of proteins including periostin, vimentin, fibronectin, matrix metalloproteinases (MMPs), and N-cadherin. Periostin can dramatically increase vimentin, fibronectin, and MMP9 expression, and the expression of cytokeratin, E- and N-cadherin, MMP2, and MMP3 are unaltered. Recently, Yan and Shao14 found that the ectopic expression of periostin in tumorigenic but not metastatic 293T cells was a strong inducer of EMT via the interaction between periostin and integrin avb5, which recruits and activates the epidermal growth factor receptor (EGFR). Therefore, the crosstalk between integrin and EGFR activates intracellular signaling pathways, like Akt/PKB, to regulate the expression of multiple genes involved in the acquisition of the mesenchymal phenotype. The biological plausibility of this physiopathological model is supported by the fact that the periostin-induced enhancement of cell adhesion, migration, and invasion can be blocked by incubation with an antiintegrin avb5 antibody or with an EGFR kinase inhibitor.14,38 The high expression level of periostin during EMT in cancer cells has also been documented in nonesmall-cell lung cancer.39 Interestingly, periostin was found to regulate EMT and cell invasiveness in an opposite way in PCa and bladder cancer. In PCa cells, periostin promotes invasiveness by downregulating E-cadherin via snail and augmenting the phosphorylation of Akt.15 Conversely, in bladder cancer cells, periostin suppresses cell invasiveness by downregulating Akt phosphorylation and upregulating E-cadherin via twist inhibition.15 Although EMT has achieved popularity on the basis of results from cultured cells and animal models, there is not convincing evidence of its importance in metastasis in humans; in fact pathologists have examined millions of tissue sections from tumors without seeing cells in transition.50 In pancreatic tumor cells, periostin was shown to induce invasiveness by increasing the motility of cells without stimulating the expression of proteases.34 Similarly, in oral cancer cells, periostin appears to enhance cell migration and invasion because it was identified as the most expressed gene in the most highly invasive tumor cell clones.40 Moreover, periostin can enhance angiogenesis through the upregulation of vascular endothelial growth factor (VEGF) receptor 2 (Flk-1/KDR) expression in endothelial cells via an integrin avb3eFAK-mediated signaling pathway.14,33 In breast cancer, the higher density of blood vessels and the higher abundance of the VEGF receptor are associated with periostin-producing tumor cells. The treatment of the periostin-overexpressing tumor cells with an anti-avb3 antibody was shown to block the induction of the Flk-1/ KDR receptor and its effect on inducing capillary formation.14 Periostin is also commonly overexpressed and promotes angiogenesis and lymphangiogenesis in oral cancers and nonesmall-cell

lung cancer.37,41 In a recent study on head and neck cancer, it was reported that lymphangiogenesis could be promoted by periostin itself and by the periostin-induced upregulation of VEGF-C.42 Finally, periostin might represent a critical limiting factor during metastatic colonization. In a recent article, Malanchi et al51 reported that periostin plays an important role in regulating the maintenance and expansion of cancer stem cells (CSCs) during metastatic colonization. CSCs are a subpopulation of tumor cells that have a potential to initiate metastatic growth; however, many factors limit metastatic colonization. Stromal periostin is crucial for metastatic colonization by regulating the interaction between CSCs and their metastatic niche. Periostin increases Wnt signaling by interacting with the Wnt ligands, Wnt1 and Wnt3A, thus contributing to a CSC-supportive niche that promotes metastatic colonization. Interestingly, previous studies also showed a possible correlation between periostin and the metastatic process. Bao et al33 showed that periostin overexpression in human colon cancer cells can enhance the number and size of metastases in the liver. Tilman et al52 showed that periostin expression is not increased in primary melanoma samples, but periostin overexpression was detected in approximately 60% of melanoma metastases in the liver or lymph nodes in their study. Periostin overexpression is frequently observed in head and neck SCC and is believed to promote the metastasis of oral SCC.40 In breast cancer patients, serum periostin levels were shown to be increased in patients with bone metastases. However, this finding was not reported in patients affected by lung cancer.53 Furthermore, Sasaki et al observed that serum periostin levels were not significantly different between most thymoma patients and control subjects; however, the serum periostin level of stage IV thymoma patients was significantly greater than the level of control subjects, suggesting that serum periostin levels might be an indicator of tumor invasion and metastasis.54

Periostin and PCa Periostin expression in normal prostate tissue was detected for the first time using Western blot analysis12 and its upregulation in PCa has been reported.55-62 Laser-capture microdissections of tissue specimens obtained from patients affected by PCa and analyzed using genome-wide expression microarray showed that periostin was upregulated in tumorassociated stroma compared with benign epithelial cell-associated stroma. This observation was confirmed using immunohistochemistry (IHC) in prostatectomy specimens from patients with PCa.55,58-61,63 Periostin expression in the stroma might be correlated with the pathological grade and the progression of disease.55,57,59-61 In fact, periostin was found to be strongly expressed in tumor stroma in most of the primary and metastatic PCa specimens. Immunohistochemical analysis revealed that overexpression in tumoral stroma was particularly evident in poorly differentiated tumors (Gleason score of 8-10) or in bone metastasis.56,59-61 The staining exhibited an overexpression in tumoral stroma of Gleason 3 and strong overexpression in the tumoral stroma of Gleason 4.56 Despite that expression of periostin in the tumor epithelial compartment was less than that observed in the stromal compartment, only 2 studies reported the significant overexpression of periostin in epithelial tumor compared with epithelial benign tissue.60,61

Clinical Genitourinary Cancer Month 2014

-5

Periostin and Genitourinary Cancer

6

-

Tischler et al analyzed periostin expression using IHC in a training cohort of 93 PCa and in a test cohort of 325 PCa, showing significant periostin overexpression not only in tumor stroma but also in tumor epithelia compared with normal stroma and epithelia of prostate tissue.60 Indeed, in the test cohort, high stromal and epithelial periostin expression were both associated with high Gleason score, and in the training cohort a higher pT stage was significantly associated with high epithelial periostin expression.60 These data were partially confirmed by Nuzzo et al using IHC analysis in the tumor and peri-tumor tissue specimens obtained from 90 patients who received radical prostatectomy.61 In this cohort, stromal and epithelial periostin expression was significantly increased in tumor tissues compared with normal adjacent tissues, but high epithelial periostin expression was not associated with high Gleason score, but only with an extraprostatic extension.61 Nevertheless, periostin expression in the stromal component of normal tissues was approximately twice as high as the expression observed in the epithelial component of PCa tissues, indicating that stromal cells mostly contribute to periostin secretion in normal and neoplastic conditions.61 Noteworthy, the prognostic role of periostin seems to be confirmed by the correlation with clinical outcome, namely prostatespecific antigen (PSA) relapse-free survival and overall survival.60,61 Tischler et al revealed that higher stromal periostin expression was significantly associated with shorter PSA-free survival in the training cohort and in the test cohort.60 However, the difference between low and high periostin subgroups was statistically significant only in the training set. No relationship between PSA-free survival and epithelial periostin expression was reported in this study.60 Nuzzo et al61 observed a direct relationship between stromal periostin expression and PSA-free survival, even if the difference was not statistically significant, exactly as reported by Tischler et al60 in their validation. Furthermore, low epithelial periostin expression was associated with shorter PSA-free survival and the epithelial expression was not predictive of patient survival. In contrast, stromal expression was highly predictive of the risk of death, and it was only a weak predictor of PSA progression. It is certainly intriguing that the phenotype characterized by low periostin expression in the epithelium and high protein expression in the stroma showed a bleak prognosis, in terms of PSA-free and overall survival.61 These findings suggest that periostin might play a different biological role in tumor progression, depending on its compartmentalization. In fact, although periostin downregulation in PCa epithelium appears to be correlated with extraprostatic extension and biochemical failure, both of which represent early events in the natural history of the disease, periostin overexpression in the stroma appears to be highly predictive of the risk of death, a late event that usually follows distant spreading and the loss of hormone dependency. Studies in vitro and in vivo provided clues for a potential target of periostin for therapeutic intervention in the future.57,64 Analyzing the periostin expression in 2 hormone-refractory (DU145 and PC3) and 2 hormone-dependent (22RV1 and LNCaP) PCa cell lines, Sun et al showed that only LNCaP expressed periostin mRNA and protein.57 Stable expression of shortRNA (shRNA)-periostin LNCaP cells, obtained by transfecting shRNA-periostin lentiviral particles, significantly reduced the level of periostin mRNA by nearly 80%. Furthermore, silencing periostin expression by RNA

Clinical Genitourinary Cancer Month 2014

interference inhibited the proliferation and migration of LNCaP cells in vivo.57 In contrast, Argellati et al64 showed that LNCaP cell lines do not constitutively express the periostin gene and protein. These findings align with those reported by Kim et al, who also found that DU145, PC3, and LNCaP cell lines do not express periostin mRNA or protein.29 The discrepancy between these findings regarding the LNCaP cell line is most likely related to the different strains available to the groups. For this reason, a cell model expressing the periostin gene was generated by stably transfecting cells with the periostin gene. Although transfection with the periostin gene did not interfere with the stimulatory effect of treatment with dihydrotestosterone (DHT), treatment with bicalutamide (BIC), a nonsteroidal antiandrogen commonly used in the treatment of PCa, had an inhibitory effect on cell growth, in the presence and the absence of periostin. The fact that cell proliferation induced by periostin transfection could be almost completely reversed by the concurrent administration of BIC suggests that at least one of the putative periostin interactors in these experimental conditions might be under androgenic control. Furthermore, periostin mRNA expression did not appear to be influenced by DHT or BIC, suggesting that these drugs do not interfere directly with the transcription of the periostin gene.64 In conclusion, a significant upregulation of periostin in PCa and a correlation with malignant behavior and poor prognosis underline a potential usefulness of periostin as a prognostic marker, and the in vivo and in vitro studies offer clues of periostin as a valid option in anticancer therapy.

Periostin and Bladder Cancer Despite that in human tumors, periostin expression has been shown to be upregulated, this is not true for bladder cancer.15,29,65,66 In bladder cancer, the studies revealed that the downregulation of periostin, which is constitutively expressed in urothelial tissue, could be a prognostic marker of the disease, and the expression of specific isoforms of periostin might have a key role in the metastatic process.15,29,65,66 Immunohistochemical analysis in normal bladder and cancer bladder tissues revealed that periostin protein was localized in the stroma of normal and cancer tissue and that the expression was greater and more concentrated in normal stroma than in cancer stroma.65 In fact, reverse transcriptase polymerase chain reaction analysis revealed that periostin mRNA was expressed in the normal bladder tissues, but was barely detectable in bladder cancer cell lines SBT31A, T24, and HT1197.65 Moreover, the downregulation of periostin mRNA was significantly related to higher grade bladder cancer, considering that periostin was detected in 81.8% of Grade 1, 40.0% of Grade 2, and 33.3% of Grade 3 bladder cancer tissues.65 The analysis of periostin isoforms revealed that isoform 2 appears to have a key role in the carcinogenesis process. Although the 4 isoforms of periostin were expressed in normal bladder tissues, only isoform 2 and isoform 3 were isolated from bladder cancer tissues, and isoform 2, which is predominantly expressed in bladder cancer tissues, did not suppress cell invasiveness in vitro or the metastatic spreading of cancer cells in vivo.29

Pier Vitale Nuzzo et al Moreover, no isoform 1 was detected in the bladder cancer tissues, suggesting the hypothesis that the loss of isoform 1, due to downregulation and/or alternative splicing, which produces isoform 2, might be closely correlated with the development of bladder cancer.29 Studies in vitro and in vivo provided clues for a potential target of periostin and/or its interactors for therapeutic intervention in the future. The ectopic expression of periostin mRNA by a retrovirus vector suppressed the in vitro cell invasiveness of bladder cancer cells without affecting cell proliferation and tumor growth in nude mice. Concerning molecular mechanism, the studies indicated that periostin is involved in the suppression of cell invasiveness via the transforming growth factor-activated kinase (TAK1) binding protein 1/TAK1 signaling pathway, that mediates various intracellular signaling pathways.67,68 The discrepancy in cell invasiveness between bladder cancer and other tumors could be due to an opposite way in regulating some molecular pathways.15,29 For instance, in PCa cells, periostin promotes invasiveness by downregulation of E-cadherin via snail and augmenting phosphorylation of Akt; differently in bladder cancer, periostin suppresses cell invasiveness through the downregulation of Akt phosphorylation and the upregulation of E-cadherin via twist inhibition.15 Future therapies based on the inhibition of Akt activity might complement conventional treatments by controlling tumor invasiveness and metastasis.15

8 was expressed more frequently in RCC than in matched nonneoplastic tissue, and isoforms 3 and 4 were detected similarly in RCC and normal tissue. Differently, isoform 1, 2, 5, 6, and 7 have been detected only in fetal kidney, suggesting a role of those proteins during embryogenesis of the kidney.70 The clarified role of isoform 8 and the molecular pattern involved in the tumorigenesis might allow the development of a new specific tumor target therapy.

Periostin and Penile Cancer Recently only Gunia et al have investigated the role of periostin protein in penile cancer.72 Immunohistochemical analysis on tissue microarray of 89 SCCs detected the expression of periostin in 44% of SCCs. Noteworthy, periostin was frequently found to be increased in the stromal compartment especially at the epithelialestromal interface.72 High periostin expression in either stroma or tumor epithelia showed significant positive correlations with tumor size, histologic grade, and pT stage.72 The evidence that in the multivariable Cox models including pT stage, pN status, grading, and the patients’ age at the time of surgery, periostin expression independently predicted cancer-specific survival, supported the potential role of periostin as a prognostic marker.72

Discussion Periostin and Renal Cancer Periostin seems to be a prognostic marker in renal cell carcinoma (RCC), in particular periostin was found to be an accessible biomarker in human kidney with tumor, not only from the tissue specimens but also from the bloodstream.69 A significant upregulation of mRNA periostin was found in RCC compared with normal renal tissue, with a prominent greater expression of periostin in the clear-cell than in the papillary subtype.70 Immunohistochemical analysis of tissue microarrays in more than 800 clear-cell RCCs, and in tissue from 1007 RCC patients provided evidence that periostin was expressed in mesenchymal cells of the tumor stroma and epithelial tumor cells.70,71 The role of periostin as a relevant prognostic marker is supported by the evidence of a close association between the upregulation of periostin and a more aggressive tumor behavior.70 It is of note that epithelial rather than stromal-specific expression of periostin was significantly correlated with high tumor dedifferentiation grade, high stage, the presence of tumor necrosis, lymph node metastases, and sarcomatoid differentiation.70,71 In particular, the presence and the amount of tumor areas with sarcomatoid differentiation in clear-cell RCC has been shown to be a relevant prognostic marker because this pathological feature is correlated with a poor prognosis.70 Moreover, Morra et al have shown that the stromal expression of periostin was correlated with a lower rate lymphocyte infiltration and lower level of periostin in tumor epithelia, whereas a high level of epithelial periostin was correlated with a high rate of lymphocyte infiltration and a decreased overall survival, suggesting that periostin might be related to a dysfunction of the immune system.70 Sequence analysis of periostin isoforms revealed that isoform 8 might have a key role during tumor development. In fact, isoform

In this review, we summarized the most recent knowledge about periostin, focusing on the role of this protein in human carcinogenesis with a special emphasis on genitourinary tumors and on its putative role as a prognostic marker and a novel target for the development of anticancer therapies (Fig. 3).15,29,56,57,59-61,64-66,69-72 Genitourinary tumors represent a clinically relevant health problem and an important cause of cancer-related death.73 In Western countries, the incidence of these tumors is in fact continuously increasing, because of the progressive aging of the population, the radical changes in lifestyle and nutrition, and the exposure to environmental pollution.73 Currently, many studies have been focused on individual genes or proteins as putative tumor drivers, with the aim of identifying novel tumor markers and novel therapeutic targets. The identification of novel targets could reveal to be particularly important to manage the most aggressive cancer phenotypes. Many abnormalities involve ECM proteins, namely periostin, a protein produced and secreted by fibroblasts in the ECM, for which there is more and more strong evidence relative to its role in promoting tumor progression, locally and remotely, through the generation of a microenvironment that favors and supports metastatic colonization.51

Role of Periostin in Tumor Progression-Driving Most of the data taken into analysis in the present review, regarding the expression pattern of periostin in genitourinary neoplasms, revealed that in PCa, penile, and renal cancer, stromal and epithelial expression of periostin is significantly increased in tumor tissues compared with normal adjacent tissues59-61,70-72; in contrast, in bladder cancer, periostin is steadily expressed in normal tissues and it is weakly expressed in cancer tissues.15

Clinical Genitourinary Cancer Month 2014

-7

Periostin and Genitourinary Cancer Figure 3 Overview of Periostin Role in Genitourinary Cancer

Abbreviations: PSA ¼ prostate-specific antigen; TAB1 ¼ transforming growth factor-activated kinase binding protein; TAK1 ¼ transforming growth factor-activated kinase.

8

-

It is not clear whether periostin is produced and secreted by epithelial rather than by stromal cells or by both; however, in most of the studies analyzed, periostin overexpression appears to occur mainly in cancer stroma, although findings are somewhat contradictory. For instance, in PCa, Tsunoda et al found that periostin expression was indeed greater in the epithelium than in the stroma59; these findings were partially confirmed by Tischler et al in their test cohort, but not in their training cohort60; differently in the study by Nuzzo et al, periostin appeared to be expressed mainly in the stromal compartment.61 It is not easy to explain the differences in periostin compartmentalization observed in the 3 studies. Besides the number of patients and disease stage, these studies are different in some methodological aspects, namely concerning patient grouping, study end points, and staining evaluation. For instance, Nuzzo et al and Tischler et al used an immune score obtained by multiplying the intensity of staining by the percentage of stained cells.60,61 Differently, in the study by Tsunoda et al, IHC analysis was performed considering only the quantitative expression of periostin (positive: at least > 5% of stained cells), without consideration of staining intensity.59 Nevertheless, in all of these studies, periostin overexpression in stroma appears to be associated with the most malignant phenotypes or the poorest outcomes, supporting

Clinical Genitourinary Cancer Month 2014

the hypothesis that this protein might play a role in driving tumor progression. Although all previous studies agreed that epithelial overexpression is far less frequent than stroma overexpression, findings concerning epithelial overexpression in PCa are more controversial.59,60 In fact, although epithelial overexpression is associated with a more aggressive phenotype (ie, higher Gleason score or advanced-stage tumors) and poorer outcomes in the studies from Tischler et al and Tsunoda et al, in the Italian study, epithelial overexpression is correlated with a better outcome.59-61 Comparably with PCa, epithelial and stromal periostin staining is observed in renal and in penile cancer.70-72 However, although in penile cancer periostin overexpression in either stroma or tumor epithelia showed a significant correlation with poor clinical pathologic features, in renal cancer only epithelial overexpression was associated with poor prognostic features, in high-grade and high-stage renal tumors.70 In bladder cancer, IHC studies allowed the localization of periostin expression almost exclusively in stromal cells.29 Noteworthy, the evidence that bladder cancer cell lines do not express periostin, taken together with the inverse relationship observed between periostin expression in cancer tissues and tumor grade, has contributed to generate the hypothesis that periostin downregulation is correlated with the development of bladder cancer.15

Pier Vitale Nuzzo et al These controversial findings might be explained probably because periostin can act either as a tumor promoter or as a tumor suppressor gene through the expression of different isoforms generated by alternative splicing. Isoforms probably localize differently in the 2 compartments and interact with different signaling pathways, thus resulting in a different biological behavior. For instance, preclinical findings, discussed in previous sections of the present review, showed that although in bladder cancer cell lines periostin upregulates E-cadherin via Akt/twist inhibition, the opposite effect was obtained in PCa cell lines.15 At present it is not known which isoform or isoforms are expressed in PCa and their compartmental and subcellular localization, because monoclonal antibodies specifically directed to each single isoform were not available for the IHC studies published so far. Conversely, it is well recognized that isoform 2 is predominantly found in bladder cancer and that isoform 8 is overexpressed in renal cancer.15,70

Role of Periostin As a Prognostic Marker Virtually all of the studies available demonstrated a more or less strict relationship between periostin overexpression (in PCa, renal, and penile cancer) or downregulation (in bladder cancer) and poor prognostic tumor features and poor outcome. As previously mentioned, a correlation between periostin overexpression in cancer tissue and the poorest phenotypes, including high Gleason score or advanced tumor stage, was found in the 4 studies addressing this issue, although results were partially controversial, especially on periostin localization.57,59-61 We have already discussed the methodological and selection bias which might explain the main controversial aspects. Only 2 studies have directly compared, although on a retrospective basis, protein overexpression with clinical outcome, namely with progression-free and/or overall survival.60,61 Both studies confirm that periostin overexpression is strictly associated with either a shorter progression-free or overall survival. In the study reported by Nuzzo et al, periostin overexpression at the stromal level and downregulation at the epithelial level were identified in a subgroup of patients with a very poor prognosis, showing the shortest progression-free and overall survival.61 These findings were confirmed in multivariate analysis, suggesting that periostin might add to the prognostic information provided by Gleason score. Periostin overexpression was also found to independently predict cancer-specific survival in multiparametric models in penile cancer.72 Moreover, a statistically significant decreased survival was reported for patients affected by renal cancer who showed periostin overexpression at the epithelial level. These findings derive from univariate analysis which do not allow assessment of the predictive value of this variable.70 However, these data have been achieved restrospectively in a very large cohort of patients. All previous findings taken together provide strong evidence in support of the potential role of periostin as a prognostic marker, although they should still be taken with caution. In fact, before periostin could enter clinical practice, the role as a prognostic factor should be tested prospectively and validated in much larger series and methodological issues should be properly addressed, starting from standardization of IHC assays and possibly engineering specific monoclonal antibodies able to identify the isoform that appears to be mainly involved in each tumor type.

Periostin As a Novel Therapeutic Target Preliminary data concerning the possibility to achieve periostin silencing in experimental models, in vitro and in vivo, by genetic and epigenetic manipulations highlight the putative role of periostin as a target for novel therapies.57 These new approaches might be particularly useful to manage aggressive diseases like hormone refractory PCa, penile cancer, and sarcomatoid or papillary renal cancer. Indeed, the evidence available in genitourinary models is very limited. In fact, as previously described in detail, we were able to find only 2 articles on PCa. Both research groups used LNCaP cells constitutively expressing the periostin gene, although different mechanisms have been used to silence gene transcription, either through the transfection with shRNA or by directly blocking the protein using neutralizing antibodies.57,74 In both cases, tumor cell proliferation was achieved. Moreover, these findings are strengthened by those achieved using periostin antibodies, either alone or combined with chemotherapy in other experimental tumor models, including colorectal cancer.12 Although promising, all data available so far do not allow the expectation that these novel therapeutic approaches could be used in clinical testing, although they might play a relevant role in the management of specific situations; for example, in PCa patients who are diagnosed with intrinsically hormone-resistant cancers or who develop hormone resistance early in the course of their disease (both situations imply poor prognosis and a limited sensitivity to endocrine manipulations and/or chemotherapy), or in patients with penile cancer, which frequently bears a poor prognosis and is relatively resistant to chemotherapy, or patients with specific renal cancer phenotypes, like sarcomatoid tumors, which are virtually resistant to tyrosine kinase inhibitors or mammalian target of rapamycin inhibitors (ie, the standard treatments).

Conclusion The present evidence revealed that periostin is involved in the development of genitourinary tumors and emphasized the potential role of this protein not only as a prognostic marker but also as a target for therapeutic intervention. Nevertheless, further morphological and functional investigations in preclinical models and in a larger prospective series will be required to better clarify the biological and clinical role of periostin and/or its isoforms, depending on its different expression and compartmentalization in prostate, bladder, renal, and penile cancer.

Disclosure The authors have stated that they have no conflicts of interest.

References 1. Takeshita S, Kikuno R, Tezuka K, et al. Osteoblast-specific factor 2: cloning of a putative bone adhesion protein with homology with the insect protein fasciclin I. Biochem J 1993; 294:271-8. 2. Horiuchi K, Amizuka N, Takeshita S, et al. Identification and characterization of a novel protein, periostin, with restricted expression to periosteum and periodontal ligament and increased expression by transforming growth factor beta. J Bone Miner Res 1999; 14:1239-49. 3. Takayama G, Arima K, Kanaji T, et al. Periostin: a novel component of subepithelial fibrosis of bronchial asthma downstream of IL-4 and IL-13 signals. J Allergy Clin Immunol 2006; 118:98-104. 4. Norris RA, Borg TK, Butcher JT, et al. Neonatal and adult cardiovascular pathophysiological remodeling and repair: developmental role of periostin. Ann NY Acad Sci 2008; 1123:30-40. 5. Kuhn B, del Monte F, Hajjar RJ, et al. Periostin induces proliferation of differentiated cardiomyocytes and promotes cardiac repair. Nat Med 2007; 13:962-9.

Clinical Genitourinary Cancer Month 2014

-9

Periostin and Genitourinary Cancer 6. Lie-Venema H, Eralp I, Markwald RR, et al. Periostin expression by epicardiumderived cells is involved in the development of the atrioventricular valves and fibrous heart skeleton. Differentiation 2008; 76:809-19. 7. Wang D, Oparil S, Feng JA, et al. Effects of pressure overload on extracellular matrix expression in the heart of the atrial natriuretic peptide-null mouse. Hypertension 2003; 42:88-95. 8. Goetsch SC, Hawke TJ, Gallardo TD, et al. Transcriptional profiling and regulation of the extracellular matrix during muscle regeneration. Physiol Genomics 2003; 14:261-71. 9. Nakazawa T, Nakajima A, Seki N, et al. Gene expression of periostin in the early stage of fracture healing detected by cDNA microarray analysis. J Orthop Res 2004; 22:520-5. 10. Li P, Oparil S, Feng W, et al. Hypoxia-responsive growth factors upregulate periostin and osteopontin expression via distinct pathways in rat pulmonary arterial smooth muscle cells. J Appl Physiol (1985) 2004; 97:1550-8. 11. Sehra S, Yao W, Nguyen ET, et al. Periostin regulates goblet cell metaplasia in a model of allergic airway inflammation. J Immunol 2011; 186:4959-66. 12. Tai IT, Dai M, Chen LB. Periostin induction in tumor cell line explants and inhibition of in vitro cell growth by anti-periostin antibodies. Carcinogenesis 2005; 26:908-15. 13. Ruan K, Bao S, Ouyang G. The multifaceted role of periostin in tumorigenesis. Cell Mol Life Sci 2009; 66:2219-30. 14. Yan W, Shao R. Transduction of a mesenchyme-specific gene periostin into 293T cells induces cell invasive activity through epithelial-mesenchymal transformation. J Biol Chem 2006; 281:700-8. 15. Kim CJ, Sakamoto K, Tambe Y, et al. Opposite regulation of epithelial-tomesenchymal transition and cell invasiveness by periostin between prostate and bladder cancer cells. Int J Oncol 2011; 38:1759-66. 16. Litvin J, Selim A, Montgomery M, et al. Expression and function of periostinisoforms in bone. J Biol Chem 2004; 92:1044-61. 17. Kiefer F, Arnold K, Künzli M, Bordoli L, Schwede T. The SWISS-MODEL Repository and associated resources. Nucleic Acids Res 2009; 37:D387-92. 18. Kopp J, Schwede T. The SWISS-MODEL Repository of annotated threedimensional protein structure homology models. Nucleic Acids Res 2004; 32: D230-4. 19. Doliana R, Bot S, Bonaldo P, et al. EMI, a novel cysteine-rich domain of EMILINs and other extracellular proteins, interacts with the gC1q domains and participates in multimerization. FEBS Lett 2000; 484:164-8. 20. Callebaut I, Mignotte V, Souchet M, et al. EMI domains are widespread and reveal the probable orthologs of the Caenorhabditis elegans CED-1 protein. Biochem Biophys Res Commun 2003; 300:619-23. 21. Hoersch S, Andrade-Navarro MA. Periostin shows increased evolutionary plasticity in its alternatively spliced region. BMC Evol Biol 2010; 10:30. 22. Coutu DL, Wu JH, Monette A, et al. Periostin, a member of a novel family of vitamin K-dependent proteins, is expressed by mesenchymal stromal cells. J Biol Chem 2008; 283:17991-8001. 23. Norris RA, Moreno-Rodriguez RA, Sugi Y, et al. Periostin regulates atrioventricular valve maturation. Dev Biol 2008; 316:200-13. 24. Kii I, Nishiyama T, Li M, et al. Incorporation of tenascin-C into the extracellular matrix by periostin underlies an extracellular meshwork architecture. Biol Chem 2010; 285:2028-39. 25. Tanabe H, Takayama I, Nishiyama T, et al. Periostin associates with Notch1 precursor to maintain Notch1 expression under a stress condition in mouse cells. PLoS One 2010; 5:e12234. 26. Maruhashi T, Kii I, Saito M, Kudo A. Interaction between periostin and BMP-1 promotes proteolytic activation of lysyl oxidase. J Biol Chem 2010; 285: 13294-303. 27. Sugiura T, Takamatsu H, Kudo A, et al. Expression and characterization of murine osteoblast-specific factor 2 (OSF- 2) in a baculovirus expression system. Protein Expr Purif 1995; 6:305-11. 28. Gillan L, Matei D, Fishman DA, Gerbin CS, Karlan BY, Chang DD. Periostin secreted by epithelial ovarian carcinoma is a ligand for alpha(V)beta(3) and alpha(V)beta(5) integrins and promotes cell motility. Cancer Res 2002; 62: 5358-64. 29. Kim CJ, Isono T, Tambe Y, et al. Role of alternative splicing of periostin in human bladder carcinogenesis. Int J Oncol 2008; 32:161-9. 30. Bai Y, Nakamura M, Zhou G, et al. Novel isoforms of periostin expressed in the human thyroid. Jpn Clin Med 2010; 1:13-20. 31. Litvin J, Zhu S, Norris R, et al. Periostin family of proteins: therapeutic targets for heart disease. Anat Rec A Discov Mol Cell Evol Biol 2005; 287:1205-12. 32. Song G, Ouyang G, Bao S. The activation of Akt/PKB signaling pathway and cell survival. J Cell Mol Med 2005; 9:59-71. 33. Bao S, Ouyang G, Bai X, et al. Periostin potently promotes metastatic growth of colon cancer by augmenting cell survival via the Akt/PKB pathway. Cancer Cell 2004; 5:329-39. 34. Baril P, Gangeswaran R, Mahon PC, et al. Periostin promotes invasiveness and resistance of pancreatic cancer cells to hypoxia-induced cell death: role of the beta4 integrin and the PI3k pathway. Oncogene 2007; 26:2082-94. 35. Erkan M, Kleeff J, Gorbachevski A, et al. Periostin creates a tumor-supportive microenvironment in the pancreas by sustaining fibrogenic stellate cell activity. Gastroenterology 2007; 132:1447-64. 36. Shao R, Bao S, Bai X, et al. Acquired expression of periostin by human breast cancers promotes tumor angiogenesis through up-regulation of vascular endothelial growth factor receptor 2 expression. Mol Cell Biol 2004; 24:3992-4003.

10

-

Clinical Genitourinary Cancer Month 2014

37. Takanami I, Abiko T, Koizumi S. Expression of periostin in patients with nonesmall-cell lung cancer: correlation with angiogenesis and lymphangiogenesis. Int J Biol Markers 2008; 23:182-6. 38. Morra L, Moch H. Periostin expression and epithelial-mesenchymal transition in cancer: a review and an update. Virchows Arch 2011; 459:465-75. 39. Soltermann A, Tischler V, Arbogast S, et al. Prognostic significance of epithelialmesenchymal and mesenchymal-epithelial transition protein expression in none small-cell lung cancer. Clin Cancer Res 2008; 14:7430-7. 40. Kudo Y, Ogawa I, Kitajima S, et al. Periostin promotes invasion and anchorageindependent growth in the metastatic process of head and neck cancer. Cancer Res 2006; 66:6928-35. 41. Siriwardena BS, Kudo Y, Ogawa I, et al. Periostin is frequently overexpressed and enhances invasion and angiogenesis in oral cancer. Br J Cancer 2006; 95:1396-403. 42. Kudo Y, Iizuka S, Yoshida M, et al. Periostin directly and indirectly promotes tumor lymphangiogenesis of head and neck cancer. PLoS One 2012; 7:e44488. 43. Hynes RO. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 1992; 69:11-25. 44. Varner JA, Cheresh DA. Integrins and cancer. Curr Opin Cell Biol 1996; 8:724-30. 45. Mizejewski GJ. Role of integrins in cancer: survey of expression patterns. Proc Soc Exp Biol Med 1999; 222:124-38. 46. Brooks PC, Montgomery AM, Rosenfeld M, et al. Integrin alpha v beta 3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. Cell 1994; 79:1157-64. 47. Nicholson KM, Anderson NG. The protein kinase B/Akt signalling pathway in human malignancy. Cell Signal 2002; 14:381-95. 48. Stupack DG, Cheresh DA. Get a ligand, get a life: integrins, signaling and cell survival. J Cell Sci 2002; 115:3729-38. 49. Cheng GZ, Park S, Shu S, et al. Advances of AKT pathway in human oncogenesis and as a target for anti-cancer drug discovery. Curr Cancer Drug Targets 2008; 8: 2-6. 50. Ledford H. Cancer theory faces doubts. Nature 2011; 472:273. 51. Malanchi I, Santamaria-Martínez A, Susanto E, et al. Interactions between cancer stem cells and their niche govern metastatic colonization. Nature 2011; 481:85-9. 52. Tilman G, Mattiussi M, Brasseur F, van Baren N, Decottignies A. Human periostin gene expression in normal tissues, tumors and melanoma: evidences for periostin production by both stromal and melanoma cells. Mol Cancer 2007; 6:80. 53. Sasaki H, Yu CY, Dai M, et al. Elevated serum periostin levels in patients with bone metastases from breast but not lung cancer. Breast Cancer Res Treat 2003; 77: 245-52. 54. Sasaki H, Dai M, Auclair D, et al. Serum level of the periostin, a homologue of an insect cell adhesion molecule, as a prognostic marker in nonesmall-cell lung carcinomas. Cancer 2001; 92:843-8. 55. Furusato B, Tsunoda T, Shaheduzzaman S, et al. Osteoblast-specific factor 2 expression in prostate cancer-associated stroma: identification through microarray technology. Urology 2010; 75:768-72. 56. Tian Y, Bova GS, Zhang H. Quantitative glycoproteomic analysis of optimal cutting temperature-embedded frozen tissues identifying glycoproteins associated with aggressive prostate cancer. Anal Chem 2011; 83:7013-9. 57. Sun C, Zhao X, Xu K, et al. Periostin: a promising target of therapeutical intervention for prostate cancer. J Transl Med 2011; 9:99. 58. Chen J, Xi J, Tian Y, et al. Identification, prioritization, and evaluation of glycoproteins for aggressive prostate cancer using quantitative glycoproteomics and antibody-based assays on tissue specimens. Proteomics 2013; 15:2268-77. 59. Tsunoda T, Furusato B, Takashima Y, et al. The increased expression of periostin during early stages of prostate cancer and advanced stages of cancer stroma. Prostate 2009; 69:1398-403. 60. Tischler V, Fritzsche FR, Wild PJ, et al. Periostin is up-regulated in high grade and high stage prostate cancer. BMC Cancer 2010; 10:273. 61. Nuzzo PV, Rubagotti A, Zinoli L, et al. Prognostic value of stromal and epithelial periostin expression in human prostate cancer: correlation with clinical pathological features and the risk of biochemical relapse or death. BMC Cancer 2012; 12:625. 62. Planche A, Bacac M, Provero P, et al. Identification of prognostic molecular features in the reactive stroma of human breast and prostate cancer. PLoS One 2011; 6:e18640. 63. Sun C, Song C, Ma Z, et al. Periostin identified as a potential biomarker of prostate cancer by iTRAQ-proteomics analysis of prostate biopsy. Proteome Sci 2011; 9:22. 64. Argellati F, Nuzzo PV, Ricci F, et al. Dihydrotestosterone and bicalutamide do not affect periostin expression in androgen-dependent LNCaP prostate cancer cell lines. Anticancer Res 2013; 33:815-20. 65. Kim CJ, Yoshioka N, Tambe Y, et al. Periostin is down-regulated in high grade human bladder cancers and suppresses in vitro cell invasiveness and in vivo metastasis of cancer cells. Int J Cancer 2005; 117:51-8. 66. Isono T, Kim CJ, Ando Y, Sakurai H, Okada Y, Inoue H. Suppression of cell invasiveness by periostin via TAB1/TAK1. Int J Oncol 2009; 35:425-32. 67. Shibuya H, Yamaguchi K, Shirakabe K, et al. TAB1: an activator of the TAK1 MAPKKK in TGF-beta signal transduction. Science 1996; 272:1179-82. 68. Sakurai H, Miyoshi H, Toriumi W, et al. Functional interactions of transforming growth factor beta-activated kinase 1 with IkappaB kinase to stimulate NF-kappaB activation. J Biol Chem 1999; 274:10641-8. 69. Castronovo V, Waltregny D, Kischel P, et al. A chemical proteomics approach for the identification of accessible antigens expressed in human kidney cancer. Mol Cell Proteomics 2006; 5:2083-91. 70. Morra L, Rechsteiner M, Casagrande S, et al. A relevance of periostin splice variants in renal cell carcinoma. Am J Pathol 2011; 179:1513-21.

Pier Vitale Nuzzo et al 71. Dahinden C, Ingold B, Wild P, et al. Mining tissue microarray data to uncover combinations of biomarker expression patterns that improve intermediate staging and grading of clear cell renal cell cancer. Clin Cancer Res 2010; 16: 88-98. 72. Gunia S, Jain A, Koch S, et al. Periostin expression correlates with pT-stage, grading and tumour size, and independently predicts cancer-specific survival in

surgically treated penile squamous cell carcinomas. J Clin Pathol 2013; 66: 297-301. 73. Siegel R, Naishadham D, Jemal A. Cancer statistics. CA Cancer J Clin 2012; 62: 10-29. 74. Orecchia P, Conte R, Balza E, et al. Identification of a novel cell binding site of periostin involved in tumour growth. Eur J Cancer 2011; 47:2221-9.

Clinical Genitourinary Cancer Month 2014

- 11