Entering a New Era of Rational Biomarker Discovery for Early Detection of Melanoma Metastases: Secretome Analysis of Associated Stroma Cells Verena Paulitschke,† Rainer Kunstfeld,† Thomas Mohr,‡ Astrid Slany,‡ Michael Micksche,‡ Johannes Drach,‡ Christoph Zielinski,‡ Hubert Pehamberger,† and Christopher Gerner*,‡ Department of Dermatology and Department of Medicine I, Medical University of Vienna, Vienna, A-1090, Austria Received December 17, 2008
Metastasis in melanoma is associated with poor prognosis. Early detection may thus substantially improve patient survival. Here we present a novel biomarker discovery strategy based on proteome profiling and secretome analysis of primary cells. Tumor associated stroma cells secrete proteins that may act as powerful tumor promoters. This cell cooperativity is reversible and may thus be directly accessible to therapeutic intervention. The onset of these characteristic events seems to precede tumor progression. Thus, proteins specifically secreted by these cells may serve as early disease biomarkers. Due to the leaky nature of newly formed blood vessels and the increased hydrostatic pressure within tumors, secreted proteins are most plausibly shed into the blood. Our analysis strategy is based on three different model systems, including established cultured cell lines, animal model systems, and clinical human samples. The feasibility is demonstrated with secretome and proteome profiles generated from normal human skin fibroblasts in comparison to melanoma-associated fibroblasts isolated from mouse xenografts and fibroblasts from bone marrow of multiple myeloma patients. Further mutual comparisons were enabled including proteome profiles of melanocytes and M24met melanoma cells. All shotgun proteomics data are accessible via the PRIDE database. Among others, the candidate biomarkers GPX5, secreted by melanoma cells, in addition to periostin and stanniocalcin-1, which are expressed by melanoma-associated fibroblasts were identified. In conclusion, this is a novel strategy to identify diagnostic marker proteins aiding early detection of metastatic melanoma and to improve our understanding of pathomechanisms involving the microenvironment to enable the design of novel therapeutic strategies. Keywords: AUTHOR • PLEASE • PROVIDE • KEYWORDS
Introduction Melanoma. Melanoma is a malignant tumor arising from melanocytes. During the last 10 years, the melanoma incidence has increased more rapidly than that of any other cancer. Although melanoma contributes accounts for only 4% of all dermatologic cancers, it is responsible for 80% of deaths from skin cancer. The strongest risk factors for melanoma are a family history of melanoma, a previous melanoma or multiple benign or atypical nevi.1 Additional risk factors are immunosuppression, sun sensitivity, and exposure to ultraviolet radiation. Genetic analysis in familial melanoma patients has identified germline mutations in CDKN2A, which encodes INK4A and ARF, and CDK4.1 With the discovery of highfrequency BRAF mutations in melanocytic neoplasms, activation of the RAS-RAF-ERK signaling pathway seems to be a very * To whom correspondence should be addressed. Christopher Gerner: phone, +43 1 4277 65 230; fax, +43 1 4277 9651, e-mail, Christopher.Gerner@ meduniwien.ac.at. † Department of Dermatology. ‡ Department of Medicine I. 10.1021/pr8010827 CCC: $40.75
2009 American Chemical Society
common event in melanoma development.2 The cytologic atypia in dysplastic nevi reflect lesions within the cyclindependent kinase inhibitor 2A (CDKN2A) and phosphatase and tensin homologue (PTEN) pathways. Further progression of melanoma is associated with decreased differentiation and the decreased expression of melanoma markers regulated by microphthalmia-associated transcription factor (MITF). Progression from the radial-growth phase to the vertical-growth phase of melanoma is marked by the loss of E-cadherin, the expression of N-cadherin3 and an enhanced expression of RVβ3 integrin.4 These alterations often lead to an autocrine and paracrine secretion of cytokines and growth factors which results in remodeling of the surrounded tissue and enhanced tumor progression. Local invasion and metastatic spread are actually mainly responsible for the morbidity and mortality in melanoma. Up to one-fifth of patients develop metastatic disease which is associated with poor prognosis and a median survival time of 7.5 months. More than 14% of patients with metastatic melanoma survive for five years. The effect of chemotherapy Journal of Proteome Research 2009, 8, 2501–2510 2501 Published on Web 02/17/2009
research articles of advanced melanoma is still not satisfying. Despite over 20 years of trials with various therapeutic combinations, dacarbazine (DTIC) and IL-2 remain the standard treatment for advanced melanoma.5 Early detection of processes supporting metastasis and appropriate interference may thus substantially improve patient survival. Metastasis is facilitated by angiogenesis and lymphangiogenesis. The induction of angiogenesis is generally considered as essential to ensure the supply of oxygen and nutritients for malignant tumor growth, invasion and metastasis, and an increase in vessel density correlates with poor prognosis in a number of tumor types.6 The ability to promote lymphangiogenesis is likely to enhance the metastatic spread of melanoma and recent studies revealed that tumor associated lymphangiogenesis is significantly correlated with poor disease-free and overall survival of patients with cutaneous melanoma.7 Metastatic cancer can thus be considered as product of the tissue microenvironment. In addition to vessels, the microenvironment consists of fibroblasts, macrophages and other immune cells. Cancer associated fibroblasts (CAFs) are known to stimulate cancer progression also by providing survival factors elevating the threshold for apoptosis.8 Tumor-associated macrophages contribute to tissue remodeling and repair.9 The tumor microenvironment, through the process of aberrant cell growth, cellular invasion and altered immune system function, represents a unique constellation of proteins secreted, of enzymatic (for example, kinases and phosphatases) and proteinase activity (for example, matrix metalloproteinases).10,11 This generates an unbalanced or altered stoichiometry of molecules within the tumor profile compared with the “normal” milieu and can represent characteristic fingerprints applicable as specific and sensitive biomarkers. Melanoma presents a substantial clinical challenge since there is a lack of adequate approaches to properly define disease subgroups for rational treatment design and selection. Current diagnostic methods are limited in their ability to diagnose early disease and accurately predict individual risk of disease progression and outcome. Various marker proteins have already been established for melanoma, mainly for immunohistochemical analysis. HMB45 to gp100 is the most commonly used melanocyte differentiation marker. Recently it was complemented by the antibodies to Melan-A/MART-1 and tyrosinase. Other reagents, whose reactivity is not strictly confined to melanocyte differentiation antigens, are also commonly used, the most prominent is S100.12 In contrast to the low specificity, S-100 remains the most sensitive marker for melanocytic lesions. HMB-45, MART-1/Melan-A, tyrosinase, and MITF demonstrate relatively good specificity but not as good sensitivity as S-100. None of these markers is known to have a high specificity and sensitivity or to exhibit prognostic value for melanocytic neoplasms.13 This may be attributed to the fact that there is a high heterogeneity in melanoma with a lot of parameters such as tumor size, location, histology, Clark level, Breslow depth, stage, grade, ulceration, age, sex etc. The emerging pattern of molecular complexity in melanoma tumors mirrors the clinical diversity of the disease. This highlights that melanoma, like other cancers, is not a single disease but a heterogeneous group of disorders that arise from complex molecular changes.14 Knowledge of molecular aberrations involving important cellular processes, such as cellular signaling networks, cell cycle regulation, and cell death, will be essential for better diagnosis, accurate assessment of prognosis, and rational design of effective therapeutics. The dermatological 2502
Journal of Proteome Research • Vol. 8, No. 5, 2009
Paulitschke et al. diagnostic and prognostic methods for early disease detection rely primarily on microscopic tissue morphology, the measurement of depth of involvement and the analysis of dermatoscopy criteria. For an accurate, individualized assessment of risk of disease progression it is obligatory to classify disease subgroups and rationally select treatments to substantially affect the outcome of advanced disease. Sekulic et al. discuss that the low overall response rates observed in clinical trials that rely on clinical disease features for patient selection might simply reflect a relatively low percentage of patients with the disease susceptible to a given therapeutic agent or combination. As a consequence, patient selection for clinical trials and selection of therapy on the basis of individual molecular attributes might be necessary to improve response rates to any therapy. These important concepts mark the beginning of a new era of rational therapy design and selection for treatment of malignant melanoma.14 Sekulic et al. offer a modern strategy to define patient’s unique tumor characteristics by genomic techniques with a complementary use of targeted analytical methods which may lead to personalized prediction of outcomes and selection of therapy. They propose that the reduction to a single patient will overcome the heterogeneity and lead to a new classification.14 However, it is a very elaborate procedure to screen every patient by genomic techniques and there is still an urgent need for further investigations. In addition the genomic techniques rather focus on hereditary predisposition. Proteome analysis, in contrast is able to detect when and to what extend the risks have become manifest. Model. The identification of potential marker proteins is not trivial.15 Comparative analysis of serum samples and tissue specimen is hindered by the natural complexity of protein expression. Disease like cancer means a variety of deregulated cell processes all of which causing characteristic aberrant protein expression. Different kinds of patho-physiological processes may be associated with tumor development, such as involvement of the immune system, alterations of the microenvironment and characteristic processes in the cancer cells themselves. This complexity is further enhanced by the individual heterogeneity in disease in addition to heterogeneities introduced by the involved experimental procedures. Low abundant proteins may be hard to identify as long as they are present in a complex protein mixture together with other proteins, several at million fold higher concentrations. Dependent on the protein mixture, positive identification of actually present, but low abundant proteins may fail. Statistical evaluation of comparative proteome analysis data may thus fail to identify the truly relevant proteins due to these methodological limitations. Here, we provide a new concept to overcome this inherent heterogeneity. Alternative to individual molecular profiling, since an individual person is still a complex system, it is based on the functional analysis of cell types in advance. It is predicated on the characterization of smallest independent units and tries to find a combination of independent units to match the molecular profile of an individual sample. In mathematics this strategy is called Fourier transformation which makes a complex function amenable. The smallest independent and potentially predictable protein synthesis machinery unit is a cell. Since every cell aberration is associated with protein aberrations, the cell is an optimal starting point for biomarker discovery. Like Fourier analysis in physics, the establishment of profiles of the smallest autonomous protein
Secretome Analysis of Associated Stroma Cells
research articles
Figure 1. Novel approach detecting biomarkers and defining potential therapeutic targets. The basic strategy for biomarker discovery is visualized. As model systems cultured cell lines, animal models for melanoma and squamous skin cancer and biopsy specimens of human skin cancer are presented. In all cases the secretome of the same isolated cell types (i.e., cancer cells, endothelial cells and fibroblasts) is analyzed. The present feasibility study includes data from primary and cultured cells in addition to cells isolated from mouse xenografts. Multiple comparisons of the proteome profiles as demonstrated in Table 1 support the identification of the most significant proteins.
production units in the body, i.e. cells, may greatly facilitate the interpretation of complex proteome profiles as derived from human serum or tissue samples. All proteomes, i.e. protein mixtures, should it be from tissues, blood, plasma or other body fluids can be expressed as a function of cellular proteomes. The assignment to cellular proteome reference maps will lead to a massive reduction of apparent complexity. Therefore possible candidates were extracted by defining the involved cell systems such as cancer cells and distinguished cell of the environment including fibroblasts and endothelial cells in a first step. With the aid of a specialized database, the CPL/ MUW-database, specificities and commonalities of protein expression profiles of such different cells can be quickly assessed. In further steps it is envisaged to analyze for specific cell-cell interactions mimicking characteristic tissue states for example by applying different cocultures starting from in vitro to in vivo models. In a last step these results shall then be evaluated in the human background in the tissue and blood profile (Figure 1). Our strategy is composed of seven independent steps: (1) Establishment of relevant model systems mimicking various functional cell states including characteristic in vitro cell activation experiments and (non-) contact cocultures (2) Standardization of protein isolation (3) Standardization of MS-procedures (4) Generation of proteome reference maps for human primary cells (5) Data organization via database
(6) Interpretation of data from diseased tissues by the use of multiple reference maps (7) Verification of biomarkers or possible therapeutic targets by i.e ELISA, immunhistochemistry, Western blot The underlying mechanisms can be analyzed focusing on secretome analysis. The secretome is defined as the set of secreted proteins.16,17 Secreted proteins may determine, control and coordinate many biological processes such as growth, cell division and differentiation or angiogenesis and lymphangiogensis. Protein secretion is fulfilling a biological function rather than maintenance of basic metabolism. Therefore, secreted proteins much better reflect biological functions compared to cytoplasmic proteins. Their key roles makes them good targets and sources for therapeutical and drug-based intervention as well as tools for diagnosis and prognosis. Thus, great interest is currently focused on the characterization of secreted proteins in order to find and identify novel biomarkers. An added benefit for the potential use of secreted proteins as cancer biomarkers is the leaky nature of newly formed blood vessels and the increased hydrostatic pressure within tumors.18 This pathological situation would tend to push molecules from the tumor interstitium into the circulation. Therefore it seems to be plausible that proteins produced by the microenvironment may be shed into the blood, making the ongoing processes of tumor development detectable.19 Combinations of markers that are indicative for the specific interactions of the tumor tissue microenvironment will achieve a higher specificity and a higher sensitivity than the application of single markers. Candidate biomarkers are expected to exist at very low concentrations Journal of Proteome Research • Vol. 8, No. 5, 2009 2503
research articles
Paulitschke et al. a
Table 1. Proteins Identified in the Cell Supernatant Fraction of the Indicated Cells Accession
Protein name
P98160/Q05793
Basement membrane-specific heparan sulfate proteoglycan core protein precursor (HSPG) (Perlecan) (PLC) Collagen alpha-1(I) chain precursor Collagen alpha-1(III) chain precursor Collagen alpha-1(VI) chain precursor Collagen alpha-2(I) chain precursor Connective tissue growth factor precursor Dermatopontin precursor Epididymal secretory glutathione peroxidase GPX5 Fibronectin precursor (Cold-insoluble globulin) Follistatin-related protein 1 precursor Glia-derived nexin precursor (GDN) (Protease nexin I) (PN-1) (Protease inhibitor 7) Insulin-like growth factor-binding protein 2 precursor (IGFBP-2) (IBP-2) Insulin-like growth factor-binding protein 7 precursor (IGFBP-7) (IBP-7) Interleukin-20 (IL-20) Interstitial collagenase precursor (EC 3.4.24.7) (Matrix metalloproteinase-1) (MMP-1) Laminin alpha-2 chain precursor (Laminin M chain) Laminin subunit alpha-4 precursor Laminin subunit beta-1 precursor Lumican precursor (Keratan sulfate proteoglycan lumican) (KSPG lumican) Macrophage migration inhibitory factor (MIF) Matrix metalloproteinase-9 precursor (EC 3,4,24,35) (MMP-9) (92 kDa type IV collagenase) Melanocyte protein Pmel 17 Metalloproteinase inhibitor 1 precursor (TIMP-1) Metalloproteinase inhibitor 2 precursor (TIMP-2) Neuropilin-1 precursor (Vascular endothelial cell growth factor 165 receptor) (CD304 antigen) Pentraxin-related protein PTX3 precursor (Pentaxin-related protein PTX3) Periostin Pigment epithelium-derived factor precursor (PEDF) Plasminogen activator inhibitor 1 precursor (PAI-1) (Endothelial plasminogen activator inhibitor) (PAI) Proactivator polypeptide precursor [Contains: Saposin A (Protein A); Saposin B-Val; Saposin B Protein CYR61 precursor (Cysteine-rich, angiogenic inducer, 61) (Insulin-like growth factor-binding protein 10) (GIG1 protein) SPARC precursor (Secreted protein acidic and rich in cysteine) (Osteonectin) (ON) 40) Stanniocalcin-1 precursor (STC-1)
P02452/P11087 P02461/P08121 P12109/Q04857 P08123/Q01149 P29279 Q07507 O75715 P02751/P11276 Q12841 P07093/Q07235 P18065 Q16270/Q61581 Q9NYY1 P03956 P24043/Q60675 Q16363 P07942 P51884 P14174/P34884 P14780 P40967 P01033 P16035 O14786/P97333 P26022 Q15063/Q62009 P36955/P97298 P05121 P07602 O00622 P09486/P07214 P52823/O55183
F
MC
Mel
X X X X X
EC
DC
X
CAF M
CAF MM
X
X X
X X X X
X X X X
X
X X X
XX XX X X X
X
XX X
X
X
X X
X X
X
X
X
X
X X
X
X
X
X X X
X X
X X
X
X XX XX
X
X
X
X
X
X
X XX
XX X X
X
X X X
X X X
X
X X
X
X
X X X
X
X
X
X X X
X
X
X
XX X
XX X X
X
X
XX
XX
X
X
X
X
X
a F, normal human skin fibroblasts; MC, primary human melanocytes; Mel, cultured human M24met melanoma cells; EC, human umbilial vein endothelial cells; DC, immature human dendritic cells; CAF M, cancer-associated fibroblasts isolated from the mouse melanoma model; CAF MM, cancer-associated fibroblasts isolated from bone marrow of multiple myeloma patients. Accesion numbers are from the Uniprot database referring to the human and, if applicable, to the corresponding mouse orthologue. X, protein positively identified in the corresponding cell type. Note that several proteins were found secreted by all cells analyzed, whereas others show rather specific expression patterns (XX).
together with highly abundant blood proteins, such as albumin, which exist in million-fold excess. At early stages of disease, cancer-specific proteins will always constitute a minor subfraction of the proteome representing a true analytical challenge. Noteworthy, early stage disease lesions such as carcinoma in situ represent a tissue volume of much less than 0.10 mL. However, the affected microenvironment comprises more cells compared to the number of tumor cells. Thus proteins derived from tumor associated stroma cells will be produced by more cells and may accumulate to higher amounts. Consequently it can be expected that such proteins will be better accessible for diagnostic purposes than proteins derived from cancer cells themselves. Secretome analysis is applicable to 2504
Journal of Proteome Research • Vol. 8, No. 5, 2009
cultured cells as well as tissue specimens.20 The most comprehensive analysis results, however, are obtained in case of isolated and cultured cells. This is why cell culture model systems of established cell lines as well as short time cultures of isolated primary cells are employed. This is applied for both tumor cells as well as a variety of stroma cells. Since secreted proteins offer new candidates for blood biomarkers, specific proteins identified in the cytoplasm may represent biomarker candidates applicable for immunohistochemical analysis. Cytoplasmic proteins are a matter of consequence shedding light on the actual status of the tumor and tumor associated cells. Combining the information of both
Secretome Analysis of Associated Stroma Cells secreted and cytoplasm proteins give rise to a complex understanding of patho-physiological processes. In conclusion, the present concept is based on three hypotheses: 1. Tumor cells may recruit stromal cells for the secretion of growth factors, which serve as survival factors. The induction of apoptosis due to genetic aberrations, a natural mechanism to stop cancer cells, may thus be inhibited. Stroma cell secretion of bioactive molecules, which may serve as diagnostic biomarkers, are early events in carcinogenesis and may thus enable the early detection of cancer progression. 2. Transformation of cancer cells is an irreversible process which may be corrected only by apoptotic cell death. Tumor therapy usually targets cancer cells, modern therapy concepts include to target the stroma in an antiangiogenic and anti-inflammatory fashion. Cooperativity contributed by stromal cells is reversible and thus directly accessible to therapeutic intervention. Most importantly, stroma derived survival factors shall be decreased resulting in a higher chemosensitivity of the tumor cells. Detailed understanding of the responsible processes may thus enable the design of completely new therapeutic strategies. 3. The endothelium is an important part of the microenvironment which facilitates metastasis to distant lymph nodes and further into organs. Proteome profiling may identify molecular signatures of processes which promote metastasis. Our approach offers the opportunity to identify pathomechanisms and the contribution of the involved cell types. Here, we want to demonstrate the feasibility of this approach. Secretomes of primary melanocytes, cultured melanoma cells and representatives of the most prominent stroma cells including fibroblasts, endothelial cells and dendritic cells were analyzed by shotgun proteomics. Several of the identified proteins may indeed serve as biomarkers indicative for certain cells and characteristic cell activities.
Materials and Methods Spontaneously Metastatic Human Melanoma Mouse Model. Pathogen-free, 4- to 6-week-old female CB17 scid/scid (SCID) mice (Charles River, Sulzfeld, Germany) were housed and used as described;21 2 × 106 M24met cells were injected s.c. in the flank of SCID mice. Tumors were measured with callipers and tumor volumes were calculated using the formula (a2 x b)/2, with a being the width and b the length of the tumor. When tumors reached a size of approximately 900 mm3, commonly 14-18 days post injection, animal were anesthetized, the primary tumors removed and skin defects sutured. All procedures were carried out in line with the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care and the Guide for the Care and Use of Laboratory Animals (Department of Health and Human Services, NIH). All experiments were approved by the Animal Welfare Committee of the Medical University of Vienna and the Austrian Government Committee on Animal Experimentation. Isolation of Melanoma Associated Fibroblasts. The removed tumor was swabbed with ethanol, washed in PBS, cut in 4 mm slices and transferred into RPMI 1640 tissue culture medium. For isolation of melanoma associated fibroblasts the sliced tumor were plated on plastic dishes and incubated for 30 min at 37 °C. During this time sprouting of the fibroblasts started.
research articles The fibroblast were then cultivated in a Fibroblast Basal Medium (Clonetics, CC-3131) supplemented with 1 mg/mL of hFGF (human recombinant Fibroblast Growth Factor) (CC4065), 5 mg/mL Insulin (CC-4021), 50 mg/ml Gentamicin, 50 mg/mL Amphotericin-B (CC-4081) and 10 mL FBS (Fetal Bovine Serum) (CC-4101). In the third passage fibroblast were examined for the cytoplasmic and secreted proteins. Prior examination for the secreted proteins the fibroblasts were cultured for six hours in a Fibroblast Basal Medium without supplements. Isolation and Cultivation of Skin Fibroblasts. Fibroblasts were cultured as described by Frazier et al.,22 except that cells were isolated from normal skin punches and were cultured in RPMI supplemented with 10% FCS at standard cell culture conditions. Establishing Short-Term Cultures of Melanoma Cells ex vivo. On part of the melanoma ex vivo (mouse) was used for tumor cell isolation. The tumor was swabbed with ethanol, washed in PBS, cut in 4 mm slices and transferred into RPMI 1640 tissue culture medium. Medium was changed 24 h later. Monocyte-Derived DC Preparation and Maturation. Peripheral blood mononuclear cells were isolated from heparinized whole blood of four different healthy donors by standard density gradient centrifugation with Ficoll-Paque (GE Healthcare Bio-Sciences AB, Uppsala, Sweden). Subsequently, monocytes were separated by magnetic sorting using the MACS technique (Miltenyi Biotec, Bergisch Gladbach, Germany) as previously described.23 Monocytes were enriched by using the biotinylated CD14 mAb VIM13 (purity 95%) as previously described.23 Dendritic cells (DC) were generated by culturing purified blood monocytes for 7 days with a combination of granulocyte-macrophage colony-stimulating factor (GM-CSF) (50 ng/mL) and interleukin-4 (IL-4) (100 U/mL). Isolation and Culture of HUVECs. Human umbilical vein endothelial cells (HUVECs) were isolated as described by Jaffe et al.24 Obtained cells where seeded into tissue culture flasks (Corning, Corning, NY) coated with 0.5 µg/cm2 human fibronectin (HFN, Chemicon, Temecula, CA, USA) and cultivated in EBM-2(MV) (Cambrex) supplemented with growth factors according to the instructions of the manufacturer and additionally with 10% FCS and 10 µg/ml ECGS (EBM-2MV complete). Proteome and Secretome Analysis. Isolation of cytoplasm and cell supernatant fractions: Cells were washed after treatment and further incubated in serum-free specialized media formulations supplemented with L-glutamine for 6 to 24 h at 37 °C. For isolation of the secreted protein fraction, the cell supernatant was collected, sterile filtrated to remove cellular debris and precipitated by the addition of ethanol. For the isolation of cytoplasmic proteins, all buffers were supplemented with protease inhibitors. Cells were lysed in hypotonic lysis buffer and pressed through a 26 g syringe in order to open the cells by rupture. The cytoplasmic fraction was separated from the nuclei by centrifugation and precipitated by the addition of ethanol. All protein samples were dissolved in sample buffer (7.5 M urea, 1.5 M thiourea, 4% CHAPS, 0.5% SDS, 100 mM DDT). For shotgun analysis, peptides were separated by nanoflow LC (1100 Series LC system, Agilent, Palo Alto, CA) using the HPLC-Chip technology (Agilent) equipped with a 40 nl Zorbax 300SB-C18 trapping column and a 75 µm × 150 mm Zorbax 300SB-C18 separation column at a flow rate of 400 nl/min, using a gradient from 0.2% formic acid and 3% ACN to 0.2% Journal of Proteome Research • Vol. 8, No. 5, 2009 2505
research articles formic acid and 50% ACN over 60 min. Peptide identification was accomplished by MS/MS fragmentation analysis with an ion trap mass spectrometer (XCT-Ultra, Agilent) equipped with an orthogonal nanospray ion source. The MS/MS data were interpreted by the Spectrum Mill MS Proteomics Workbench software (Version A.03.03, Agilent) and searched against the SwissProt Database (Version 14.3 containing 20 328 protein entries) allowing for precursor mass deviation of 1.5 Da, a product mass tolerance of 0.7 Da and a minimum matched peak intensity (%SPI) of 70%. Due to previous chemical modification, carbamidomethylation of cysteines was set as fixed modification. The listed peptides were identified with the indicated scores. The scores were essentially calculated from sequence tag lengths, but also mass deviations were considered. To assess the reliability of the peptide scores, searches were performed against the corresponding reverse database. 5.9% positive hits were found with peptides scoring >9.0, while 0.22% positive hits were found with peptides scoring >13.0. Consequently, the threshold for protein identification was set to at least one peptide scoring higher than 13.0. Further details are accessible via www.meduniwien.ac.at/ proteomics. The complete data generated by these experiments (normal human skin fibroblasts cytoplasm and secretome; normal human melanocytes cytoplasm and secretome; melanoma-associated fibroblasts isolated out of xenografts cytoplasm and secretome, human M24met melanoma cell line cytoplasm, human M24met melanoma cell line isolated out of xenografts cytoplasm and secretome, human multiple myeloma-associated bone marrow fibroblasts secretome) is accessible via PRIDE database (http://www.ebi.ac.uk/pride/),25,26 PRIDE accessions 8931-8948.
Results The present concept of systematic proteome profiling of primary cells shall enable multiple comparisons. To support this concept, standardized protein isolation procedures were used.27 According to previous results, we hardly achieved high reproducibility and reliability when using total cell extracts. This is why subcellular fractions, here cytoplasmic and, even more importantly, secreted proteins were isolated. Analysis results obtained from shotgun proteomics are subjected to rigorous quality control, considering both high statistical confidence of appropriate assignment of amino acid sequences to a MS2 spectrum in addition to biological criteria. The latter aspect includes the removal of potential contaminants such as keratins derived from dust and the consideration of biological plausibility. All complete data analyzed in the present study including secretomes and cytoplasmic fractions of normal human skin fibroblasts, normal human melanocytes, melanoma-associated fibroblasts isolated out of xenografts, human M24met melanoma cell line, human M24met melanoma cell line isolated out of xenografts and human multiple myeloma-associated bone marrow fibroblasts is publicly accessible via PRIDE database (http://www.ebi.ac.uk/pride/),25,26 PRIDE accessions 8931-8948. In addition to those, we refer to previously published data of endothelial cells and dendritic cells.28 The proteome profiles can as well be accessed and downloaded via the CPL/MUW-database at www.meduniwien.ac.at/proteomics/database. Furthermore, searches for single proteins quickly provide the specificity of protein expression by listing all cell types expressing the protein. In melanocytes, a total of 867 proteins was identified (Table S1, Supporting Information), 546 proteins were recognized as commonly expressed proteins as described recently.28 As 2506
Journal of Proteome Research • Vol. 8, No. 5, 2009
Paulitschke et al.
Figure 2. Schematic view of the intersection of shotgun and 2DPAGE protein identification results. Numbers of proteins identified in cytoplasmic fractions of primary human melanocytes, human skin fibroblasts, human melanoma M24met cells and human melanoma M24met cells transplanted from the mouse melanoma model. Intersections indicate the number of identical proteins shared between the respective cells.
expected, however, several proteins characteristic for melanocytes were identified (Table S1, Supporting Information), including melanocyte protein Pmel 17 (P40967), melanoma antigen gp75 (P17643), melan-A protein (Q16655) and interleukin-20 (Q9NYY1). Compared to these primary normal cells, the transformed melanoma cell line M24met expressed several proteins characteristic for cancer cells including 86 kDa subunit of Ku antigen (P13010), proliferating cell nuclear antigen (P12004), tissue transglutaminase (P21980) and many others (Table S2, Supporting Information). When transplanted into mice to obtain an in vivo model for human melanoma and isolated back from the established tumors, the cells expressed several proteins including dihydropyrimidinase-related protein 3 (Q14195), plectin (Q15149) and Prolyl 4-hydroxylase alpha-1 (P13674) which may indicate increased invasiveness and involvement of the microenvironment (Table S3, Supporting Information). With the aid of the CPL/MUW-database multiple comparisons between the expression profiles of these and related cells was greatly supported (Figure 2). Cytoplasmic proteins comprise a large number of proteins essential for the basic functions of a cell. In contrast, secreted proteins may reflect a more specialized subset of proteins which may contain more specifically expressed proteins. To assess the apparent specificity in detail, the secretome of a variety of relevant cells was analyzed, including endothelial cells, dendritic cells, melanocytes (Table S4, Supporting Information), melanoma cells (Table S5, Supporting Information), and normal skin fibroblasts (Table S6, Supporting Information) in addition to fibroblasts isolated from the mouse melanoma model (Table S7, Supporting Information). In order to assess potential characteristics of tumor-associated fibroblasts, we included analysis data of bone marrow-derived fibroblasts isolated from multiple myeloma patients (Slany et al., manuscript in preparation) (Table S8, Supporting Information). Secretome analysis was accomplished by shotgun proteomics based on mass spectrometric analysis of tryptic digests of cell supernatant protein fractions. Secretome analysis data of primary human endothelial cells and dendritic cells are pre-
Secretome Analysis of Associated Stroma Cells
research articles
Figure 3. Identification details of the biomarker candidates GPX5, stanniocalcin-1 and periostin by mass spectrometry. GPX5 was identified in the supernatant fraction of transplanted M24met cells (A). Stanniocalcin-1 (B, C) and periostin (D, E) were identified in the supernatant fractions of cancer-associated fibroblasts from the mouse melanoma model (C, E) and from human multiple myeloma bone marrow (B, D). After prefractionation of proteins by SDS-PAGE, proteins were digested with trypsin and the peptides separated by nanoflow liquid chromatography. Peptide fragmentation was accomplished in an ion-trap mass spectrometer (Agilent XCT Ultra). Interpretation of spectra and assignment of y- and b- ion series was performed with the Spectrum Mill software.
sented in more detail elsewhere (Mohr et al., Gundacker et al., manuscripts submitted). Table 1 provides a list of identified proteins in the different cell types. It comprises both proteins commonly expressed by all analyzed cells in addition to specifically expressed proteins. Skin fibroblasts secreted a variety of extracellular matrix (ECM) proteins including collagens, fibronectin, perlecan and laminin. Dermatopontin, presently identified only in these cells, is characteristic for skin fibroblasts.29 Cancer-associated fibroblasts (CAFs) isolated from the mouse melanoma model expressed many orthologues compared to the human skin
fibroblasts. In addition, they expressed the angiogenesis regulating protein neuropilin,30 which was not identified by us in any other fibroblast or epithelial cell type. Specifically, they secreted periostin and stanniocalcin-1 (Figure 3C, E). These proteins may be relevant, as they were as well identified in bone marrow fibroblasts isolated from human multiple myeloma patients (Figure 3B, D). This finding indicates that periostin and stanniocalcin-1 may signify a characteristic tumor-associated fibroblast phenotype. Although, dealing with low abundance proteins, highly confident data are provided (Figure 3). To demonstrate the high quality of raw data, mass spectra are Journal of Proteome Research • Vol. 8, No. 5, 2009 2507
research articles shown. Searches against the corresponding reversed database revealed less than 0.2% false positives for peptides scoring higher than 13.0, which can be translated into a confidence rate better than 99.5%. All present peptides actually scored higher than 13.0, translating into confidence rates above 99.5%. Compared to the skin fibroblasts, primary melanocytes secreted less ECM proteins (Table 1). The specific expression of lumican and the melanocyte protein Pmel 17 corresponds to known properties of melanocytes. The transplanted melanoma M24met cells secreted lumican, but Pmel17, specific for melanocytes, and PEDF, presently identified in all other listed cells, were undetectable. In addition, they secreted an unexpected protein, epididymal secretory glutathione peroxidase (GPX5, Figure 3A), which was not identified in any other cell presently analyzed and never yet identified in melanoma cells. Our data indicate that normal stroma cells may secrete potential tumor promoting factors. Endothelial cells secreted connective tissue growth factor (CCN2), which has been described to support tumor growth.31 This maybe partially due to the employed to the experimental conditions. It is only feasible to identify secreted proteins out of serum free medium, which inherently puts cells under stress. Possibly, these proteins would not be secreted under in vivo conditions, but may indicate the accessible tools of these cells to support a tumor. Insulin-like growth factor-binding proteins, identified only by cells of the surrounding tissue, and more specifically IGFBP2, as well regulate cell growth and survival32,33 and may thus contribute to shaping the tumor microenvironment. Immature dendritic cells, included here to represent tissue macrophages, secreted, among other proteins, MMP-9, which also contributes to tumor progression.34
Discussion Profiling methods such as proteomics are performed to gain a better understanding for disease processes and to identify potential marker molecules, which may be used for diagnosis, prognosis, patient stratification and surveillance.35 Unfortunately, many clinical proteome studies are of uncertain value, hard to interpret or not reproducible due to a variety reasons.15 While better designs of clinical studies and improved quality control may overcome several of the obvious obstacles,15 here we additionally suggest the application of an alternative strategy to improve our means to identify relevant proteins and interpret the data more appropriately. This strategy is based on the analysis of primary cells in addition to cultured cells and in vivo and in vitro models for relevant pathologic processes (Figure 1). Such systematic analyses generate huge amounts of data, which need to be organized with the help of a database. Here, it is demonstrated that this strategy, applied to human melanoma, may indeed work as it reveals improved understanding as well as identifies relevant proteins. The application of our standardized methods including isolation of subcellular fractions and analysis by shotgun proteomics enabled meaningful comparisons of the resulting data sets (Figure 2). Secretome analysis is not yet a standard technique, actually rather few secretome profiles have been established by direct protein identification via mass spectrometry. This is the first time that secretome profiles of melanocytes as well as of tumor-associated fibroblasts are presented. As outlined in the results section, the secretomes of the various cells analyzed clearly comprised known cell type specific proteins, proving the reliability of the applied methods Several proteins, however, were identified which were unexpected in 2508
Journal of Proteome Research • Vol. 8, No. 5, 2009
Paulitschke et al. the corresponding cells. GPX5, an epididymis-specific glutathione peroxidase-like protein,36 protects cells and enzymes from oxidative damage.37 Melanoma has been associated with increased oxidative stress.38 Tumors are able to produce proteins which are envisioned for other cells and other purposes. It appears plausible that melanoma cells may express GPX5 to manage constitutive oxidative stress. The presently observed GPX5 expression by melanoma cells needs to be further evaluated with respect to a potential application as biomarker. Applying a highly sensitive screening method such as mass spectrometry and comparative analysis of a set of different cell types and specificities new and possibly also unexpected markers were detected. The assessment of cell type-specific secretion characteristics is a prerequisite before potential relevant alterations of tumorassociated stroma cells can be recognized. In case a tumorassociated fibroblast would secrete a protein not secreted by normal fibroblasts, but secreted e.g. by normal endothelial cells, such a protein would hardly be useful as biomarker. This is why we systematically analyzed the most important representatives of tumor-associated stroma cells. This strategy helped us to identify proteins which are aberrantly expressed by tumor-associated fibroblasts but not in any normal counterparts isolated from healthy background (Table 1). Stanniocalcin-1 was identified by us in melanoma-associated fibroblasts as well as in fibroblasts of multiple myeloma (Figure 3, Table 1), but not in fibroblasts from healthy background nor in a great variety of human epithelial cells, endothelial cells and leukocytes analyzed by us (not shown). Stanniocalcin (STC) is a glycoprotein hormone that regulates calcium levels in fish. In fish, the product of this gene is secreted by the organ of Stannius in response to hypercalcemia, resulting in the inhibition of calcium absorption and increased resorption of phosphorus. The related human protein stanniocalcin-1 (STC-1) is expressed in multiple organs including the ovary, prostate, kidney, and thyroid. A unique feature of STC-1 is its lack of homology to any other proteins, except for stanniocalcin-2 (STC-2) with a homology of 34%. In contrast to its effects as an endocrine hormone in fish, STC-1 appears to function in an autocrine/ paracrine fashion to regulate calcium and phosphate homeostasis in humans. Recently, STC-1 was shown to be involved in angiogenesis and cancer progression. STC-1 mRNA, detected by quantitative RT-PCR, appears to be highly sensitive and specific to occult cancer cells and is not expressed in normal blood and bone marrow cells. In addition, the STC-1 mRNA expression in blood is closely related to tumor size in breast cancer, micrometastases of hepatocellular carcinoma, and minimal residual disease in leukemia.39 Very recently, stanniocalcins (STC-1 and STC-2) were shown to be expressed in breast cancers of lower grade with a slow initial progression of the tumor and to act as a survival factor contributing to the extended survival or dormancy of micrometastases with a cumulative risk for late relapses.40 Another biomarker candidate presently identified is periostin. We as well identified periostin in other kinds of cancerassociated fibroblasts (Table 1 and Figure 3). Periostin is a 90kD secreted protein that has been suggested to function as a cell adhesion molecule for preosteoblasts and to participate in osteoblast recruitment, attachment and spreading. POSTN/ Periostin (gene/protein) was mostly found to be overexpressed and correlated with increased tumor aggressiveness in various human cancers such as colon, pancreas, thyroid, oral squamous cell carcinoma or neuroblastoma.41 Periostin, when transfected
research articles
Secretome Analysis of Associated Stroma Cells to cancer cells, was shown to promote tumor angiogenesis, metastatic growth, cancer cell motility and adhesion.42 In normal skin, melanocytes do not express periostin but fibroblasts may express the gene at high level. Periostin overexpression was detected in about 60% of melanoma metastatic tumors in the liver or lymph nodes as evaluated by quantitative RTPCR.43 We identified periostin also in secretomes of vascular endothelial growth factor-stimulated primary endothelial cells (Mohr et al., manuscript submitted), pointing to a growthpromoting context. As a conclusion, our data are in line with those results, which were obtained by a totally different experimental approach. The present study shall demonstrate the feasibility of the present approach rather than provide final conclusions. It is our intention to complete the present data with experiments outlined in Figure 1. The analysis of the secreted proteins of LECs (lymphatic endothelial cells) and BECs (blood endothelial cells) cultured on matrigel also in coculture with melanoma cells may provide new insights into the pathomechanisms of tumor angiogenesis and lymphangiogenesis and metastasis. This will offer valuable clues for possible therapeutic combination regimens. Since proteins are secreted into the interstitial fluid and blood, it is quite promising to further evaluate such determined secreted proteins via ELISA studies to investigate their potential as biomarkers.
Conclusion Here we present a novel approach to understand the mechanisms of tumor progression and metastasis by involving the microenvironment. The approach is of tremendous importance since it will offer us a new understanding of the pathophysiology of tumor progression, helps us to detect novel biomarkers for early detection and prognosis and leads to new targets for therapeutic intervention. It will be possible to define new relevant targets and the optimal tools which enable the design and screening of complementary tumor therapies. The plenty of data can offer new opportunities to develop a biomarker set for ELISA analysis for the clinical routine and shed light on the mechanisms of metastasis, invasion of tumor cells into the vessels and tumor progression for new therapeutic targets. The combination of a set of relevant markers will yield an improvement of sensitivity and specificity of the screenings. By focusing on secreted proteins which are early shed by the microenvironment into the blood, we will gain very specific information about the actual status of the patient and define a fingerprint of the tumor status in the patient. This strategy may enable early diagnosis of metastatic processes and offers an opportunity for a rational therapy selection. Candidate biomarkers shall be evaluated in clinical studies by correlation with the progression free and overall survival. This approach will offer a better understanding of the heterogeneity in melanoma and will lead to a reevaluation of the basic concepts. Possibly, this concept is able to add information for a new classification, to define patient subgroups and to enhance the often low overall response rates observed in clinical trials.
Acknowledgment. This study was supported by the ¨ sterreichischen Nationalbank”, Pro“Jubila¨umsfond der O ject No. 12215 dedicated to VP, the “Interdisziplina¨rer Krebsforschungsfond” dedicated to VP and “Bu ¨rgermeisterfond der Stadt Wien”, dedicated to VP. We thank Editha Bayer and Anna Pirker for technical assistance, Helge Wimmer and Johannes Griss for bioinformatics support and Agilent for
lending the HPLC-Chip-Cube, which proved to be essential for shotgun analyses.
Supporting Information Available: Supplementary Tables S1-S8. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Miller, A. J.; Mihm, M. C., Jr. Melanoma. N. Engl. J. Med. 2006, 355 (1), 51–65. (2) Omholt, K.; Platz, A.; Kanter, L.; Ringborg, U.; Hansson, J. NRAS and BRAF mutations arise early during melanoma pathogenesis and are preserved throughout tumor progression. Clin. Cancer Res. 2003, 9 (17), 6483–8. (3) Hsu, M. Y.; Wheelock, M. J.; Johnson, K. R.; Herlyn, M. Shifts in cadherin profiles between human normal melanocytes and melanomas. J. Investig. Dermatol. Symp. Proc. 1996, 1 (2), 188–94. (4) Danen, E. H.; Ten Berge, P. J.; Van Muijen, G. N.; Van ’t Hof-Grootenboer, A. E.; Brocker, E. B.; Ruiter, D. J. Emergence of alpha 5 beta 1 fibronectin- and alpha v beta 3 vitronectin-receptor expression in melanocytic tumour progression. Histopathology 1994, 24 (3), 249–56. (5) Tsao, H.; Sober, A. J. Melanoma treatment update. Dermatol. Clin. 2005, 23 (2), 323–33. (6) Weidner, N.; Folkman, J. Tumoral vascularity as a prognostic factor in cancer. Important Adv. Oncol. 1996, 167–90. (7) Dadras, S. S.; Paul, T.; Bertoncini, J.; Brown, L. F.; Muzikansky, A.; Jackson, D. G.; Ellwanger, U.; Garbe, C.; Mihm, M. C.; Detmar, M. Tumor lymphangiogenesis: a novel prognostic indicator for cutaneous melanoma metastasis and survival. Am. J. Pathol. 2003, 162 (6), 1951–60. (8) Schamberger, C. J.; Gerner, C.; Cerni, C. bFGF rescues 423-cells from serum starvation-induced apoptosis downstream of activated caspase-3. FEBS Lett. 2004, 573 (1-3), 19–25. (9) Mantovani, A.; Allavena, P.; Sica, A. Tumour-associated macrophages as a prototypic type II polarised phagocyte population: role in tumour progression. Eur. J. Cancer 2004, 40 (11), 1660–7. (10) Culp, W. D.; Neal, R.; Massey, R.; Egevad, L.; Pisa, P.; Garland, D. Proteomic analysis of tumor establishment and growth in the B16F10 mouse melanoma model. J. Proteome Res. 2006, 5 (6), 1332– 43. (11) Liotta, L. A.; Kohn, E. C. The microenvironment of the tumourhost interface. Nature 2001, 411 (6835), 375–9. (12) Jungbluth, A. A. Serological reagents for the immunohistochemical analysis of melanoma metastases in sentinel lymph nodes. Semin. Diagn. Pathol. 2008, 25 (2), 120–5. (13) Ohsie, S. J.; Sarantopoulos, G. P.; Cochran, A. J.; Binder, S. W. Immunohistochemical characteristics of melanoma. J. Cutan. Pathol. 2008, 35 (5), 433–44. (14) Sekulic, A.; Haluska, P., Jr.; Miller, A. J.; Genebriera De Lamo, J.; Ejadi, S.; Pulido, J. S.; Salomao, D. R.; Thorland, E. C.; Vile, R. G.; Swanson, D. L.; Pockaj, B. A.; Laman, S. D.; Pittelkow, M. R.; Markovic, S. N. Malignant melanoma in the 21st century: the emerging molecular landscape. Mayo Clin. Proc. 2008, 83 (7), 825– 46. (15) Mischak, H.; Apweiler, R.; Banks, R. E.; Conaway, M.; Coon, J.; Dominiczak, A.; Ehrich, J. H. H.; Fliser, D.; Girolami, M.; Hermjakob, H.; Hochstrasser, D.; Jankowski, J.; Julian, B. A.; Kolch, W.; Massy, Z. A.; Neusuess, C.; Novak, J.; Peter, K.; Rossing, K.; Schanstra, J.; Semmes, O. J.; Theodorescu, D.; Thongboonkerd, V.; Weissinger, E. M.; Van Eyk, J. E.; T., Y. Clinical proteomics: A need to define the field and to begin to set adequate standards. Proteomics: Clin. Appl. 2007, 1 (2), 148–156. (16) Tjalsma, H.; Bolhuis, A.; Jongbloed, J. D.; Bron, S.; van Dijl, J. M. Signal peptide-dependent protein transport in Bacillus subtilis: a genome-based survey of the secretome. Microbiol. Mol. Biol. Rev. 2000, 64 (3), 515–47. (17) Greenbaum, D.; Luscombe, N. M.; Jansen, R.; Qian, J.; Gerstein, M. Interrelating different types of genomic data, from proteome to secretome: ‘oming in on function. Genome Res. 2001, 11 (9), 1463–8. (18) Hagendoorn, J.; Tong, R.; Fukumura, D.; Lin, Q.; Lobo, J.; Padera, T. P.; Xu, L.; Kucherlapati, R.; Jain, R. K. Onset of abnormal blood and lymphatic vessel function and interstitial hypertension in early stages of carcinogenesis. Cancer Res. 2006, 66 (7), 3360–4. (19) Petricoin, E. F.; Belluco, C.; Araujo, R. P.; Liotta, L. A. The blood peptidome: a higher dimension of information content for cancer biomarker discovery. Nat. Rev. Cancer 2006, 6 (12), 961–7. (20) Zwickl, H.; Traxler, E.; Staettner, S.; Parzefall, W.; Grasl-Kraupp, B.; Karner, J.; Schulte-Hermann, R.; Gerner, C. A novel technique
Journal of Proteome Research • Vol. 8, No. 5, 2009 2509
research articles (21)
(22)
(23)
(24)
(25)
(26)
(27)
(28)
(29) (30) (31)
2510
to specifically analyze the secretome of cells and tissues. Electrophoresis 2005, 26 (14), 2779–85. Kunstfeld, R.; Wickenhauser, G.; Michaelis, U.; Teifel, M.; Umek, W.; Naujoks, K.; Wolff, K.; Petzelbauer, P. Paclitaxel encapsulated in cationic liposomes diminishes tumor angiogenesis and melanoma growth in a “humanized” SCID mouse model. J. Invest. Dermatol. 2003, 120 (3), 476–82. Frazier, M. L.; Pathak, S.; Wang, Z. W.; Cleary, K.; Singletary, S. E.; Olive, M.; Mackay, B.; Steck, P. A.; Levin, B. Establishment of a new human pancreatic adenocarcinoma cell line, MDAPanc-3. Pancreas 1990, 5 (1), 8–16. Pickl, W. F.; Majdic, O.; Kohl, P.; Stockl, J.; Riedl, E.; Scheinecker, C.; Bello-Fernandez, C.; Knapp, W. Molecular and functional characteristics of dendritic cells generated from highly purified CD14+ peripheral blood monocytes. J. Immunol. 1996, 157 (9), 3850–9. Jaffe, E. A.; Nachman, R. L.; Becker, C. G.; Minick, C. R. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J. Clin. Invest. 1973, 52 (11), 2745–56. Jones, P.; Cote, R. G.; Cho, S. Y.; Klie, S.; Martens, L.; Quinn, A. F.; Thorneycroft, D.; Hermjakob, H. PRIDE: new developments and new datasets. Nucleic Acids Res. 2008, 36 (Database issue), D878– 83. Jones, P.; Cote, R. G.; Martens, L.; Quinn, A. F.; Taylor, C. F.; Derache, W.; Hermjakob, H.; Apweiler, R. PRIDE: a public repository of protein and peptide identifications for the proteomics community. Nucleic Acids Res. 2006, 34 (Database issue), D659– 63. Gundacker, N.; Bayer, E.; Traxler, E.; Zwickl, H.; Kubicek, M.; Stockl, J.; Gerner, C. Knowledge-based proteome profiling: considering identified proteins to evaluate separation efficiency by 2-D PAGE. Electrophoresis 2006, 27 (13), 2712–21. Slany, A.; Haudek, V. J.; Gundacker, N., C.; Griss, J.; Mohr, T.; Wimmer, H.; Eisenbauer, M.; Elbling, L.; Gerner, C. Introducing a New Parameter for Quality Control of Proteome Profiles: Consideration of Commonly Expressed Proteins. Electrophoresis 2008, (in press). Okamoto, O.; Fujiwara, S. Dermatopontin, a novel player in the biology of the extracellular matrix. Connect Tissue Res. 2006, 47 (4), 177–89. Klagsbrun, M.; Takashima, S.; Mamluk, R. The role of neuropilin in vascular and tumor biology. Adv. Exp. Med. Biol. 2002, 515, 33– 48. Shi-Wen, X.; Leask, A.; Abraham, D. Regulation and function of connective tissue growth factor/CCN2 in tissue repair, scarring and fibrosis. Cytokine Growth Factor Rev. 2008, 19 (2), 133–44.
Journal of Proteome Research • Vol. 8, No. 5, 2009
Paulitschke et al. (32) Fukushima, T.; Kataoka, H. Roles of insulin-like growth factor binding protein-2 (IGFBP-2) in glioblastoma. Anticancer Res. 2007, 27 (6A), 3685–92. (33) Rajah, R.; Katz, L.; Nunn, S.; Solberg, P.; Beers, T.; Cohen, P. Insulin-like growth factor binding protein (IGFBP) proteases: functional regulators of cell growth. Prog. Growth Factor Res. 1995, 6 (2-4), 273–84. (34) Himelstein, B. P.; Canete-Soler, R.; Bernhard, E. J.; Dilks, D. W.; Muschel, R. J. Metalloproteinases in tumor progression: the contribution of MMP-9. Invasion Metastasis 1994, 14 (1-6), 246– 58. (35) Ludwig, J. A.; Weinstein, J. N. Biomarkers in cancer staging, prognosis and treatment selection. Nat. Rev. Cancer 2005, 5 (11), 845–56. (36) Hall, L.; Williams, K.; Perry, A. C.; Frayne, J.; Jury, J. A. The majority of human glutathione peroxidase type 5 (GPX5) transcripts are incorrectly spliced: implications for the role of GPX5 in the male reproductive tract. Biochem. J. 1998, 333 (Pt 1), 5–9. (37) The universal protein resource (UniProt). Nucleic Acids Res. 2008, 36 (Database issue), D190-5. (38) Wittgen, H. G.; van Kempen, L. C. Reactive oxygen species in melanoma and its therapeutic implications. Melanoma Res. 2007, 17 (6), 400–9. (39) Tohmiya, Y.; Koide, Y.; Fujimaki, S.; Harigae, H.; Funato, T.; Kaku, M.; Ishii, T.; Munakata, Y.; Kameoka, J.; Sasaki, T. Stanniocalcin-1 as a novel marker to detect minimal residual disease of human leukemia. Tohoku J. Exp. Med. 2004, 204 (2), 125–33. (40) Joensuu, K.; Heikkila, P.; Andersson, L. C. Tumor dormancy: elevated expression of stanniocalcins in late relapsing breast cancer. Cancer Lett. 2008, 265 (1), 76–83. (41) Kikuchi, Y.; Kashima, T. G.; Nishiyama, T.; Shimazu, K.; Morishita, Y.; Shimazaki, M.; Kii, I.; Horie, H.; Nagai, H.; Kudo, A.; Fukayama, M. Periostin is expressed in pericryptal fibroblasts and cancerassociated fibroblasts in the colon. J. Histochem. Cytochem. 2008, 56 (8), 753–64. (42) Kanno, A.; Satoh, K.; Masamune, A.; Hirota, M.; Kimura, K.; Umino, J.; Hamada, S.; Satoh, A.; Egawa, S.; Motoi, F.; Unno, M.; Shimosegawa, T. Periostin, secreted from stromal cells, has biphasic effect on cell migration and correlates with the epithelial to mesenchymal transition of human pancreatic cancer cells. Int. J. Cancer 2008, 122 (12), 2707–18. (43) 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.
PR8010827
