Advertisement

S100P-Binding Protein, S100PBP, Mediates Adhesion through Regulation of Cathepsin Z in Pancreatic Cancer Cells

Published:February 13, 2012DOI:https://doi.org/10.1016/j.ajpath.2011.12.031
      Several S100 proteins are up-regulated in pancreatic ductal adenocarcinoma (PDAC), the most significant being S100P. We previously reported on S100PBP, a binding partner of S100P, that shows no homology to any described protein and whose functions are completely unknown. To determine S100PBP expression across human tissues and organs, immunohistochemistry was performed using both multiorgan- and in-house–constructed pancreatic tissue microarrays. To establish S100PBP functions, cell lines with either stably overexpressed or silenced S100PBP were generated and investigated using Affymetrix gene expression arrays and complementary functional assays. We show that S100PBP is differentially expressed in various healthy and tumor specimens, which is both cancer- and tissue-type dependent. In healthy pancreas, S100PBP is expressed in the nuclear/perinuclear region of both exocrine and endocrine compartments. In early precancerous lesions, S100PBP is translocated to the cytoplasm, whereas in PDAC and metastatic lesions, its expression is significantly diminished. The most pronounced phenotypic change after manipulation of S100PBP expression was seen in adhesion; this was significantly reduced after S100PBP up-regulation and increased after S100PBP silencing. Up-regulation or silencing of S100PBP also led to a concomitant change in the levels of the protease cathepsin Z, the silencing of which significantly reduced PDAC cell adhesion. We further demonstrate that the interaction of cathepsin Z with arginine-glycine-aspartic acid–binding integrins, specifically αvβ5, mediates the changes seen in adhesion of PDAC cells.
      Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of cancer-related death in the industrialized world, each year causing 227,000 deaths worldwide.
      • Olsen A.H.
      • Parkin D.M.
      • Sasieni P.
      Cancer mortality in the United Kingdom: projections to the year 2025.
      PDAC has an extremely poor 5-year survival rate of <5%, which has not improved during the past 30 years. This indicates that an increased understanding of the pathobiological characteristics of this malignancy, the development of an early diagnostic test, and the design of novel therapies are still critically needed.
      In 1998, a progression model from early intraductal hyperplasia to infiltrating carcinoma of the pancreas was proposed. PDAC develops by acquiring several genetic alterations and progressing through three stages of precursor lesions, called pancreatic intraepithelial neoplasia (PanIN 1 to 3), that are classified based on the degree of atypia.
      • Brat D.J.
      • Lillemoe K.D.
      • Yeo C.J.
      • Warfield P.B.
      • Hruban R.H.
      Progression of pancreatic intraductal neoplasias to infiltrating adenocarcinoma of the pancreas.
      The high prevalence of several members of the S100 family of calcium-binding proteins has previously been described in pancreatic cancer,
      • Crnogorac-Jurcevic T.
      • Missiaglia E.
      • Blaveri E.
      • Gangeswaran R.
      • Jones M.
      • Terris B.
      • Costello E.
      • Neoptolemos J.P.
      • Lemoine N.R.
      Molecular alterations in pancreatic carcinoma: expression profiling shows that dysregulated expression of S100 genes is highly prevalent.
      • Logsdon C.D.
      • Simeone D.M.
      • Binkley C.
      • Arumugam T.
      • Greenson J.K.
      • Giordano T.J.
      • Misek D.E.
      • Kuick R.
      • Hanash S.
      Molecular profiling of pancreatic adenocarcinoma and chronic pancreatitis identifies multiple genes differentially regulated in pancreatic cancer.
      • Iacobuzio-Donahue C.A.
      • Maitra A.
      • Olsen M.
      • Lowe A.W.
      • van Heek N.T.
      • Rosty C.
      • Walter K.
      • Sato N.
      • Parker A.
      • Ashfaq R.
      • Jaffee E.
      • Ryu B.
      • Jones J.
      • Eshleman J.R.
      • Yeo C.J.
      • Cameron J.L.
      • Kern S.E.
      • Hruban R.H.
      • Brown P.O.
      • Goggins M.
      Exploration of global gene expression patterns in pancreatic adenocarcinoma using cDNA microarrays.
      including S100P, which can also be detected in these early PanIN lesions.
      • Dowen S.E.
      • Crnogorac-Jurcevic T.
      • Gangeswaran R.
      • Hansen M.
      • Eloranta J.J.
      • Bhakta V.
      • Brentnall T.A.
      • Luttges J.
      • Kloppel G.
      • Lemoine N.R.
      Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
      S100P is a 10.4-kDa secreted protein expressed in most PDAC cases whose up-regulation correlates with increasing PanIN grade.
      • Dowen S.E.
      • Crnogorac-Jurcevic T.
      • Gangeswaran R.
      • Hansen M.
      • Eloranta J.J.
      • Bhakta V.
      • Brentnall T.A.
      • Luttges J.
      • Kloppel G.
      • Lemoine N.R.
      Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
      It promotes cancer cell growth and survival and is involved in increased invasiveness,
      • Arumugam T.
      • Simeone D.M.
      • Van Golen K.
      • Logsdon C.D.
      S100P promotes pancreatic cancer growth, survival, and invasion.
      • Whiteman H.J.
      • Weeks M.E.
      • Dowen S.E.
      • Barry S.
      • Timms J.F.
      • Lemoine N.R.
      • Crnogorac-Jurcevic T.
      The role of S100P in the invasion of pancreatic cancer cells is mediated through cytoskeletal changes and regulation of cathepsin D.
      suggesting that it is a promising diagnostic
      • Deng H.
      • Shi J.
      • Wilkerson M.
      • Meschter S.
      • Dupree W.
      • Lin F.
      Usefulness of S100P in diagnosis of adenocarcinoma of pancreas on fine-needle aspiration biopsy specimens.
      and therapeutic target.
      • Arumugam T.
      • Logsdon C.D.
      S100P: a novel therapeutic target for cancer.
      By using far-Western screening of a placental expression phage library, we isolated a novel binding partner for S100P, S100P-binding protein S100PBP (previously called S100PBPR).
      • Dowen S.E.
      • Crnogorac-Jurcevic T.
      • Gangeswaran R.
      • Hansen M.
      • Eloranta J.J.
      • Bhakta V.
      • Brentnall T.A.
      • Luttges J.
      • Kloppel G.
      • Lemoine N.R.
      Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
      The gene for S100PBP is located at 1p34.3, and in silico analysis using Vega Gene Tree (http://www.ensembl.org, last accessed April 2011) showed that, akin to S100 proteins, S100PBP is expressed only in vertebrates, primarily in mammals. S100PBP shows approximately 55% conservation and almost 100% homology at the C-terminus across all species. The predicted 45-kDa protein shows no homology to any protein present in available public databases. Pfam (http://pfam.sanger.ac.uk, last accessed April 2011) searches for potential functional domains showed a BSD-putative domain [named after the basic transcription factor (BTF)-2–like transcription factors, synapse-associated proteins, and DOS2-like protein] and pentatricopeptide repeats, although the probability for either domain did not reach statistical significance. Hence, the functional roles of S100PBP are completely unknown. Therefore, the aims of this study were to refine cellular localization, determine levels of expression, and define the functional roles of S100PBP in pancreatic cancer.

      Materials and Methods

      Cell Lines and Tissues

      A panel of eight PDAC cell lines (MiaPaCa2, Panc1, AsPc1, RwP1, FA6, PaTu 8988s, PaTu 8988t, and CFPac1) was obtained from Cancer Research United Kingdom and cultured in Dulbecco's modified Eagle's medium (Cancer Research United Kingdom Media Services, Middlesex, UK) supplemented with 10% heat-inactivated fetal calf serum (AutogenBioclear, Wiltshire, UK). An HPV16-E6E7 immortalized, nontumorigenic, pancreatic ductal cell line was obtained from Dr. Ming-Sound Tsao (Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada) and cultured as previously described.
      • Ouyang H.
      • Mou L.
      • Luk C.
      • Liu N.
      • Karaskova J.
      • Squire J.
      • Tsao M.S.
      Immortal human pancreatic duct epithelial cell lines with near normal genotype and phenotype.
      The identity of all of the cell lines used was verified by short tandem repeat profiling.
      Formalin-fixed, paraffin-embedded samples from 14 healthy and 37 PDAC cases were obtained from the Pathology Department, Royal London Hospital (London, UK), with full ethical approval and a tissue microarray (TMA) constructed with each case spotted in triplicate, as previously described.
      • Froeling F.E.
      • Mirza T.A.
      • Feakins R.M.
      • Seedhar A.
      • Elia G.
      • Hart I.R.
      • Kocher H.M.
      Organotypic culture model of pancreatic cancer demonstrates that stromal cells modulate E-cadherin, beta-catenin, and Ezrin expression in tumor cells.
      In addition, seven metastatic lymph node tissues were obtained from the Royal London Hospital, and 10 matched primary PDAC and lymph node metastases were obtained from the Department of Pathology, Osijek, Croatia. A multiorgan TMA containing 40 cores of paraffin-embedded matched tumor and healthy tissues from 25 different organs was purchased from Pantomics (San Francisco, CA).

      Plasmids

      The CMV-Tag2B and CMV-TagS100PBP constructs used in this study were previously described.
      • Dowen S.E.
      • Crnogorac-Jurcevic T.
      • Gangeswaran R.
      • Hansen M.
      • Eloranta J.J.
      • Bhakta V.
      • Brentnall T.A.
      • Luttges J.
      • Kloppel G.
      • Lemoine N.R.
      Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
      These constructs were transfected into FA6 and RwP1 cells in a 2 μg:3 μL ratio with FuGene6 transfection reagent (Roche Applied Science, West Sussex, UK) for the establishment of S100PBP-overexpressing stable clones after selection with 1.1 mg/mL G418 antibiotic (InvivoGen, San Diego, CA). S100P-overexpressing and control Panc1 cells were previously described.
      • Whiteman H.J.
      • Weeks M.E.
      • Dowen S.E.
      • Barry S.
      • Timms J.F.
      • Lemoine N.R.
      • Crnogorac-Jurcevic T.
      The role of S100P in the invasion of pancreatic cancer cells is mediated through cytoskeletal changes and regulation of cathepsin D.

      Gene Silencing

      Cells (2 × 105) were seeded per well of a six-well plate. After 24 hours, the medium was changed and each well was independently transfected for 48 hours either with 10 nmol/L nontargeting (NT) control or S100PBP or with 50 nmol/L of NT control or cathepsin Z (CTSZ) duplexes. For the double-knockdown experiment, cells were treated with 10 nmol/L S100PBP small-interfering RNA (siRNA) in combination with either 50 nmol/L NT or CTSZ siRNA (all siRNAs were from Dharmacon, Thermo Fisher Scientific Inc., Chicago, IL). The siRNA duplexes were incubated with 8 μL of INTERFERin transfection reagent (PolyPlus Transfection, New York, NY) for 10 minutes before addition to the plate.

      Quantitative Real-Time PCR

      RNA was isolated using the RNAqueous Total RNA isolation kit (Ambion, Foster City, CA), and cDNA was synthesized from 1 μg of total RNA using the AffinityScript Multiple Temperature cDNA Synthesis Kit (Stratagene, La Jolla, CA). TaqMan gene expression assays (25 μL) were prepared containing an S100PBP-specific probe primer mix (Hs00898758_m1; Applied Biosystems, Carlsbad, CA), TaqMan universal master mix (Applied Biosystems), and approximately 40 ng of cDNA, according to the manufacturer's instructions. The quantitative PCR (qPCR) was run and analyzed on an ABI7500 machine (Applied Biosystems). cDNA synthesized from 1 μg of human universal RNA (Applied Biosystems) was serially diluted from 1 in 5 to 1 in 3125 in RNase-free water to produce a standard curve for analysis. Reactions were performed in triplicate, and 18S mRNA was used as an internal standard. Validation of CTSZ deregulation was performed using SYBR Green reagent (Qiagen, West Sussex, UK). cDNA was synthesized as described above, and assays were set up in 20 μL containing 10 μL of SYBR Green reagent, 25 nmol of primers, and 2 μL of cDNA. Universal cDNA was used as a standard, S16 was used as a control, and the PCR was run and analyzed as described above. The CTSZ primers used were as follows: forward, 5′-GAATTCATGGGGTGAACCAT-3′; and reverse, 5′-TCCTTATAGGTGCTGGTCACG-3′.

      Affymetrix Gene Expression Profiling

      Human genome U133 plus 2.0 arrays (Affymetrix, Santa Clara, CA), which contain >47,000 transcripts, were used for the gene expression profiling of both FA6 cell lines that stably overexpress S100PBP and MiaPaCa2 cells, in which S100PBP was silenced; each experiment was performed in triplicate. Double-stranded cDNA was synthesized from 10 μg of total RNA using the one-cycle cDNA synthesis kit; subsequently, biotin-labeled anti-sense cRNA was synthesized using the in vitro transcription labeling kit (Affymetrix) and fragmented. A hybridization cocktail containing the fragmented cRNA, probe array controls (Affymetrix), bovine serum albumin (BSA), and herring sperm DNA (Life Technologies, Grand Island, NY) was hybridized to human genome U133 plus 2.0 arrays for 16 hours. Hybridization, washing, and staining of the arrays were performed on a Fluidics station (Affymetrix); all protocols were performed in accordance with the Affymetrix gene expression profiling technical manual. After hybridization and scanning, the data were analyzed using Bioconductor packages (http://www.bioconductor.org) within the open source R statistical environment (http://www.r-project.org). The quality control metrics recommended by Affymetrix, box plots, and intensity histograms were used for quality assessment. After background correction by robust multiarray analysis, a filter using the SD of gene expression values was applied to select the top most variable 10,000 genes. For differential expression analysis, Limma was used.
      • Smyth G.K.
      Linear models and empirical bayes methods for assessing differential expression in microarray experiments.
      A double cutoff of false-discovery rate <0.05 and a fold change of two or greater was applied. Microarray data have been deposited to the National Center for Biotechnology Information's Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo, accession number GSE35199).

      Western Blot Analysis

      Cells (5 × 105) were lysed in 200 μL of NP40 lysis buffer (50 mmol/L Tris, pH 7.4; 150 mmol/L NaCl; 1% NP40; and 2× complete protease inhibitor tablets). Western blot analysis was performed as previously described.
      • Whiteman H.J.
      • Weeks M.E.
      • Dowen S.E.
      • Barry S.
      • Timms J.F.
      • Lemoine N.R.
      • Crnogorac-Jurcevic T.
      The role of S100P in the invasion of pancreatic cancer cells is mediated through cytoskeletal changes and regulation of cathepsin D.
      Quantification by densitometry was performed using ImageJ software (NIH, Bethesda, MD) and is represented as a percentage compared with control. All primary antibodies used are listed in Table 1.
      Table 1Primary Antibodies and the Concentrations Used for Different Applications
      ProteinSourceSpeciesWestern BlotIHCFlow CytometryBlocking
      S100PBP (ab68733)AbCam (Cambridge, UK)Mouse1:100
      S100PBD Transduction LaboratoriesMouse1:25
      S100PR&D BiosystemsGoat1:500
      CTSZR&D BiosystemsGoat1:7501:100
      ActinSanta CruzGoat1:2000
      Integrin β1 (TS2/16)ATCC (Teddington, UK)Mouse10µg/mL10µg/mL
      Integrin αv (L230)ATCCMouse10µg/mL
      Integrin αvβ3 (23C6)CRUKMouse10µg/mL
      Integrin αvβ5 (P1F6)Gift from Dean Sheppard (UCSF, USA)Mouse10µg/mL10µg/mL
      Integrin αvβ6 (10D5)Chemicon International (Massachusetts, USA)Mouse10µg/mL
      Integrin αvβ8 (14E5)Gift from Steve Nishimura (UCSF, USA)Mouse10µg/mL
      Conditioned media were collected by incubating 3 × 106 cells in serum-free media for 24 hours. These media were then concentrated by centrifuging at 400 × g at 4°C for 1 hour in a 3K filtration column (Millipore, Billerica, MA).

      IHC Data

      All immunohistochemistry (IHC) was performed on the automated Ventana Discovery System (Ventana Medical Systems, Tucson, AZ). Sections (5 μm thick) were treated with primary antibodies against S100PBP, S100P, and CTSZ (Table 1) and the appropriate horseradish peroxidase–conjugated secondary antibodies (Epitomics, Inc., Burlingame, CA) and visualized using an Olympus CKX41 (GX Optical, Suffolk, UK) light microscope. Scoring was performed as previously described.
      • Crnogorac-Jurcevic T.
      • Missiaglia E.
      • Blaveri E.
      • Gangeswaran R.
      • Jones M.
      • Terris B.
      • Costello E.
      • Neoptolemos J.P.
      • Lemoine N.R.
      Molecular alterations in pancreatic carcinoma: expression profiling shows that dysregulated expression of S100 genes is highly prevalent.

      Functional Assays

      For MTT assays, 5 × 103 cells were seeded in a 96-well plate, with five replicates for each condition. Cells settled for 2 hours, and after 0, 24, 48, 72, and 96 hours, 10 μL of MTT reagent (Sigma-Aldrich, St. Louis, MO) was added. After 2 hours incubation, cells were lysed using 100 μL of acidic isopropanol and read immediately on a plate reader using the Revelation 4.04 program (both from Dynex, West Sussex, UK) at a wavelength of 560 nm. Wound healing, migration, and invasion assays were performed as previously described.
      • Whiteman H.J.
      • Weeks M.E.
      • Dowen S.E.
      • Barry S.
      • Timms J.F.
      • Lemoine N.R.
      • Crnogorac-Jurcevic T.
      The role of S100P in the invasion of pancreatic cancer cells is mediated through cytoskeletal changes and regulation of cathepsin D.
      Adhesion assays were performed in 96-well plates. Wells were coated with 100 μL of 1% BSA (Sigma-Aldrich) control or 20 μg/mL fibronectin, 20 μg/mL collagen I, 20 μg/mL vitronectin (all from BD Biosciences, Oxford, UK), or 3 μg/mL recombinant CTSZ (R&D Biosystems, Minneapolis, MN) and left overnight at 4°C. Cells, 2 × 104, were then seeded onto each surface and left to settle for 1 hour. After removing unbound cells with PBS, bound cells were fixed in 30% methanol and 10% acetic acid and stained for 5 minutes with 0.1% Coomassie Blue, 30% methanol, and 10% acetic acid. Then, they were washed in 30% methanol, 10% acetic acid, and PBS. The stain was eluted in 1% SDS for 30 minutes at 37°C and absorbance was read at 595 nm. The optical density values for the BSA-only wells were subtracted from each reading, and results were represented as a percentage of the control, unless stated otherwise. For antibody blocking experiments, cells were preincubated with the appropriate antibodies for 10 minutes on ice before seeding into wells. The antibodies used are outlined in Table 1.

      Flow Cytometry Analysis

      Cells (2 × 105) in 50 μL of media (Dulbecco's modified Eagle's medium, 0.1% BSA, and 0.1% sodium azide) were incubated with antibodies outlined in Table 1 for 45 minutes on ice, washed, and incubated with the appropriate species of Invitrogen molecular probe Alexa-488 (1:1000). An isotype control was used (Dako, Cambridgeshire, UK), and this reading was subtracted from each result. Flow cytometric acquisition was performed and analyzed on the FACSCalibur using the Cell Quest Pro program (BD Biosciences). Duplicate samples were prepared for all analyses.

      Statistical Analysis

      Each functional assay was repeated at least in triplicate, with a minimum of three repeats for each condition within each experiment. For the analysis of these data, an unpaired t-test was performed using GraphPad Prism 5 software (La Jolla, CA). Statistical analysis of IHC data was performed using SPSS software (IBM, Armonk, NY) and a Mann-Whitney test. Throughout the article, *P < 0.05, **P < 0.01, and ***P < 0.001 were considered to be statistically significant.

      Results

      S100PBP Protein Expression

      To determine the expression of S100PBP in various tissues, a human multiorgan tissue array comprising healthy and cancerous specimens from 25 different organs was used; representative images are shown in Figure 1. S100PBP was expressed in most of the tissues, predominantly in the perinuclear and nuclear region of epithelial cells. Overall, clear differences in S100PBP immunoreactivity were seen between healthy and tumor tissues. In some cases, S100PBP was present in healthy, but absent or much weaker in the corresponding cancerous, tissues, such as in healthy prostate and adenocarcinoma, healthy cell and squamous cell carcinoma of lung (Figure 1), and healthy breast compared with adenocarcinoma (data not shown). In contrast, in liver hepatocellular carcinoma, thyroid carcinoma (Figure 1), and ovarian adenocarcinoma (data not shown), S100PBP expression was higher compared with the healthy tissues. Several metastatic lesions (liver and ovarian metastases from colon cancer, lung metastasis from gastric cancer, and lymph node metastases from breast cancer) included on this TMA showed no S100PBP immunoreactivity.
      Figure thumbnail gr1
      Figure 1S100PBP expression in various healthy and cancer tissues. IHC was performed on a multiorgan TMA containing healthy and tumor specimens from 25 different organs. Representative images are shown for healthy prostate and adenocarcinoma, healthy lung and squamous cell carcinoma, healthy liver and hepatocellular carcinoma, and healthy thyroid and adenocarcinoma. Original magnification, ×200 (all images).
      The expression of S100PBP in pancreatic tissues was explored in more detail: 13 of 14 healthy pancreatic tissues, 9 of 9 PanIN lesions found within three individual PDAC cases, 46 of 47 PDACs, and 15 of 17 lymph node metastases were successfully analyzed. The representative images are shown in Figure 2. S100PBP expression was seen in both the exocrine and endocrine (data not shown) cells. Most healthy (11/13) tissue cores showed a high level of predominantly nuclear S100PBP expression, and in early PanIN lesions, S100PBP was either absent or seen in the cytoplasm. In more advanced PanINs and in the cancer cores, S100PBP staining was weaker or absent. The difference in S100PBP expression between healthy and PDAC tissues was statistically significant (P < 0.0001). In lymph node metastases, S100PBP expression was also weaker or absent when compared with the healthy pancreas (P < 0.0001). There was no significant difference in the expression of S100PBP in PDAC compared with lymph node metastasis (P = 0.639).
      Figure thumbnail gr2
      Figure 2S100PBP expression in pancreatic and lymph node metastatic lesions. Representative images are shown for healthy pancreas, PanIN, PDAC, and lymph node metastasis. Original magnification, ×100 (all images). A significant decrease in S100PBP expression was seen between healthy and PDAC tissues (P < 0.0001) and between healthy and metastatic lesions (P < 0.0001). In PanIN-1 lesions, cytoplasmic S100PBP was seen. a, acinar cells; d, ductal cells.
      The expression of S100PBP was also directly compared with that of S100P in the same healthy and PDAC cases, and an inverse correlation of expression was seen between the two proteins. Of the 45 PDAC cases successfully analyzed for both S100PBP and S100P expression, 51% (23/45) showed weak or no immunoreactivity for S100PBP and 9% (4/45) showed strong expression. In contrast, for S100P, only 7% (3/45) of the cases showed weak or no immunoreactivity and 82% (37/45) had strong expression. Of the 13 successfully analyzed healthy pancreas specimens, most (11/13) exhibited strong S100PBP expression, whereas none of the cases (0/13) showed S100P immunoreactivity (see Supplemental Figure S1 at http://ajp.amjpathol.org).

      S100PBP Functions

      To determine its functions in pancreatic adenocarcinoma-derived cell lines, S100PBP was overexpressed in FA6 and RwP1 and silenced in MiaPaCa2 and Panc1 cells. These cells express low and high levels of S100PBP, respectively (see Supplemental Figure S2 at http://ajp.amjpathol.org).
      • Dowen S.E.
      • Crnogorac-Jurcevic T.
      • Gangeswaran R.
      • Hansen M.
      • Eloranta J.J.
      • Bhakta V.
      • Brentnall T.A.
      • Luttges J.
      • Kloppel G.
      • Lemoine N.R.
      Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
      The effectiveness of overexpression and down-regulation of S100PBP was confirmed by qPCR (Figure 3A) because the S100PBP antibody did not produce a strong enough signal using Western blot analysis. This could be because S100PBP was unstructured and, hence, an easily degradable protein.
      Figure thumbnail gr3
      Figure 3S100PBP functions. A: S100PBP was overexpressed in FA6 cells using the plasmid CMV-TagS100PBP (FLAG-S100PBP), with the empty CMV-Tag2B as control (FLAG), and silenced in MiaPaCa2 cells using S100PBP targeting siRNA (S100PBP siRNA), with scrambled NT siRNA as control (NT siRNA). This deregulation was confirmed by real-time qPCR. B: Adhesion assays were performed by seeding cells for 1 hour on plates coated with different ECM proteins. Results are represented in arbitrary units, whereby absorbance data at 595 nm were normalized to control after the subtraction of BSA-only absorbance values. Overexpression of S100PBP caused a significant decrease in cell adhesion to fibronectin, collagen I, and vitronectin (P < 0.0001 for all); congruently, a significant increase was seen in cell adhesion to fibronectin (P = 0.0233) and vitronectin (P = 0.0011) when S100PBP was silenced. MiaPaCa2 cells did not bind to collagen I. C: S100PBP was transiently overexpressed in RwP1 cells using the plasmid CMV-TagS100PBP (FLAG-S100PBP), again with empty vector as a control (FLAG), and silenced in Panc1 cells using S100PBP targeting siRNA (S100PBP siRNA), with NT siRNA as a control (NT siRNA). This was confirmed by qPCR. D: Adhesion assays were performed as previously described for the cell lines. When S100PBP is overexpressed, a significant decrease in adhesion of RwP1 cells on fibronectin (P = 0.0095), collagen I (P = 0.0006), and vitronectin (P = 0.0397) is seen. Conversely, a significant increase in adhesion of Panc1 cells after S100PBP silencing is observed on vitronectin (P = 0.0021) and collagen I (P = 0.016); a small increase in adhesion to fibronectin was also seen, although this did not reach statistical significance (P = 0.093). For all panels, *P < 0.05, **P < 0.01, and ***P < 0.001.
      Neither overexpression in FA6 cells nor silencing of S100PBP in MiaPaCa2 cells showed any effect on proliferation or wound healing (data not shown). Although cell migration was not affected in three of four tested cell lines after modulation of S100PBP expression, significant changes in invasion (increase in MiaPaCa2 and Panc1 cells after S100PBP silencing and decrease in RwP1 cells after overexpression) were seen (see Supplemental Figure S3 at http://ajp.amjpathol.org). However, the most pronounced and consistent change after modulation of S100PBP expression was seen in adhesion, and this was, therefore, further studied. Compared with control cells (empty plasmid transfected), S100PBP-overexpressing FA6 cells exhibited a significant reduction in adhesion to fibronectin (57%), collagen (54%), and vitronectin (34%) (Figure 3B). Similarly, when S100PBP was silenced in MiaPaCa2 cells, there was a highly significant increase (30-fold) in adhesion to fibronectin and a moderate, but statistically significant, increase (23%) in adhesion to vitronectin-coated surfaces (Figure 3B) compared with the NT control; MiaPaCa2 cells did not bind to collagen (data not shown). To further corroborate these data, S100PBP was transiently overexpressed in RwP1 cells and silenced in Panc1 cells (confirmation by qPCR is shown in Figure 3C), and S100PBP-dependent changes in cell adhesion were again observed in these two additional pancreatic cancer cell lines (Figure 3D). Neither MiaPaCa2 nor Panc1 cells express S100P (see Supplemental Figure S4 at http://ajp.amjpathol.org), indicating that the role of S100PBP in adhesion is S100P independent.

      Transcriptomic Changes after Modulation of S100PBP Expression

      To elucidate the potential mechanisms underlying the functional changes previously outlined, we profiled FA6 cells that stably overexpress S100PBP and MiaPaCa2 cells after S100PBP silencing using Human Genome U133 plus 2.0 gene expression arrays. In both experiments, S100PBP levels were significantly altered, with a 64.4-fold increase in S100PBP in the overexpressing cells and a 3.5-fold reduction in the silenced cells, further indicating the validity of S100PBP overexpression and silencing.
      On overexpression of S100PBP in FA6 cells, 193 known genes were up-regulated and 192 genes were down-regulated (see Supplemental Table S1 at http://ajp.amjpathol.org). When MiaPaCa2 cells transfected with NT siRNA were compared with cells in which S100PBP was silenced, 117 genes were up-regulated and 75 genes were down-regulated (see Supplemental Table S2 at http://ajp.amjpathol.org). Affymetrix array analysis indicated that expression of the S100P transcript is down-regulated on S100PBP overexpression (see Supplemental Table S1 at http://ajp.amjpathol.org). Changes in the expression of several other S100 proteins were also observed; however, these findings will be reported in more detail elsewhere. In the present study, we describe one of the commonly deregulated genes across both Affymetrix experiments, CTSZ. This protease has recently been linked to cell adhesion in human umbilical vein endothelial cells
      • Lechner A.M.
      • Assfalg-Machleidt I.
      • Zahler S.
      • Stoeckelhuber M.
      • Machleidt W.
      • Jochum M.
      • Nagler D.K.
      RGD-dependent binding of procathepsin X to integrin alphavbeta3 mediates cell-adhesive properties.
      ; therefore, we hypothesized that CTSZ may exhibit a similar role in mediating the adhesion of pancreatic cancer cells.

      S100PBP and CTSZ

      In the Affymetrix gene expression profiling, CTSZ was commonly deregulated, with a 3.7-fold down-regulation on S100PBP overexpression in FA6 cells and an up-regulation (2.7-fold) on S100PBP silencing in MiaPaCa2 cells. This was confirmed at both the mRNA (data not shown) and protein levels (Figure 4A). Densitometry analysis of the blot is shown in Figure 4B. To establish whether the down-regulation of CTSZ by S100PBP could be responsible for the observed reduction in cell adhesion, CTSZ was silenced in FA6 cells; the validation of CTSZ silencing was confirmed by using Western blot analysis (Figure 5A). Cell adhesion assays were subsequently performed on these cells, and a significant decrease in adhesion to fibronectin (96%), collagen I (20%), and vitronectin (50%) was seen, compared with the NT control (Figure 5A). To further confirm the role of CTSZ in S100PBP-dependent adhesion, a rescue experiment was performed, whereby MiaPaCa2 cells were transfected with S100PBP siRNA in combination with either NT or CTSZ siRNA. The increased CTSZ expression after S100PBP silencing and the subsequent decrease in CTSZ expression after CTSZ silencing were confirmed by using Western blot analysis (Figure 5B). These cells were subjected to adhesion assays on fibronectin and vitronectin. When CTSZ and S100PBP were both silenced, there was a significant decrease in adhesion on both surfaces compared with S100PBP plus NT siRNA treated cells (Figure 5B; P = 0.0075 and P = 0.035, respectively). Furthermore, silencing CTSZ in S100PBP-knockdown cells reduced the levels of adhesion to those seen in parental MiaPaCa2 cells treated with only NT siRNA. This suggests that the S100PBP-dependent changes seen in the adhesion of PDAC cells are at least partially mediated through CTSZ.
      Figure thumbnail gr4
      Figure 4Confirmation of CTSZ deregulation. A: Western blot analysis confirmation of CTSZ deregulation, as indicated by Affymetrix gene expression profiling, in both S100PBP-overexpressing FA6 cells and MiPaCa2 cells after S100PBP silencing. A band representing active CTSZ is present at 34 kDa; in addition, in FA6 cells, the precursor form at 38 kDa is seen. In MiPaCa2 cells, the precursor form can only be seen at high exposures. Actin is included as a loading control. B: Densitometric quantification of the blot using ImageJ software; optical density is represented relative to control (P = 0.0359 for FA6; P = 0.0464 for MiaPaCa2). For all panels, *P < 0.05.
      Figure thumbnail gr5
      Figure 5Verification of CTSZ involvement in S100PBP- dependent cell adhesion. A: CTSZ was silenced in FA6 cells, with up to 50 nmol/L targeting siRNA (CTSZ siRNA), with scrambled NT siRNA as a control (NT siRNA), and confirmed by using Western blot analysis. CTSZ-silenced cells were subjected to adhesion assays on fibronectin, collagen I, and vitronectin; results are represented as arbitrary units, with data normalized to control after the subtraction of BSA-only absorbance values. On all surfaces, a significant reduction in cell adhesion was seen after CTSZ silencing (P < 0.0001, P = 0.0002, and P = 0.0002 for fibronectin, collagen I, and vitronectin, respectively). B: MiaPaCa2 cells were treated with NT, S100PBP, or a combination of S100PBP with NT or CTSZ siRNA. The increase in CTSZ after S100PBP silencing and the subsequent decrease in CTSZ expression after CTSZ silencing were confirmed by using Western blot analysis; actin is included as a loading control. siRNA-treated cells were subjected to adhesion assays on fibronectin and vitronectin, with results presented in arbitrary units (data normalized to control after the subtraction of BSA-only absorbance values). A significant decrease in adhesion on both surfaces was seen when S100PBP and CTSZ were silenced together, compared with S100PBP plus NT siRNA–treated cells (P = 0.0075 and P = 0.035 for fibronectin and vitronectin, respectively). For all panels, *P < 0.05, **P < 0.01, and ***P < 0.001.

      CTSZ Expression in PDAC

      To confirm the relevance of CTSZ in pancreatic cancer, its expression was investigated in the same panel of healthy pancreas, PanIN, PDAC, and lymph node metastatic lesions used for S100PBP; representative images are shown in Figure 6. In pancreatic tissues, no CTSZ expression was seen in healthy acinar cells, heterogeneous immunoreactivity from negative to moderate was seen in 20% to 30% of healthy ducts, and strong immunoreactivity was seen in the islets and in PanIN lesions; in the islets, ducts, and PanIN lesions, CTSZ expression was cytoplasmic and granular. In PDAC cells, CTSZ immunoreactivity was moderate to strong; however, the expression was more diffuse throughout the cytoplasm, with occasional membranous accentuation. Strong immunoreactivity was also seen in stromal cells. Metastatic lesions in lymph nodes displayed the same expression pattern to that seen in the PDAC samples, with 100% of cases showing immunoreactivity.
      Figure thumbnail gr6
      Figure 6IHC analysis of CTSZ expression in healthy pancreas, PanIN, PDAC, and metastatic lesions in the lymph nodes. Representative images are shown for each of the different tissues. In healthy pancreas, there was strong granular staining in the islets (i) and no immunoreactivity in the acinar cells (a). Ductal cells (d; 20% to 30%) showed granular CTSZ expression; PanIN lesions also showed granular staining. In both PDAC and metastatic lesions in the lymph node, CTSZ expression was diffused throughout the cytoplasm and showed membranous accentuation. Original magnification, ×400 (healthy pancreas and PanIN lesions); ×200 (PDAC and lymph node metastases).

      CTSZ Interaction with Integrins

      Integrins are key molecules in the regulation of cell adhesion, and CTSZ contains an arginine-glycine-aspartic acid (RGD) integrin binding site in its propeptide; it was also recently reported that CTSZ interacts with αvβ3 in human umbilical vein endothelial cells.
      • Lechner A.M.
      • Assfalg-Machleidt I.
      • Zahler S.
      • Stoeckelhuber M.
      • Machleidt W.
      • Jochum M.
      • Nagler D.K.
      RGD-dependent binding of procathepsin X to integrin alphavbeta3 mediates cell-adhesive properties.
      We, therefore, investigated whether a similar interaction could modulate the adhesion of pancreatic cancer cells. By collecting conditioned cell media, we demonstrated that the proform of CTSZ is secreted from both FA6 and MiaPaCa2 cells (see Supplemental Figure S5 at http://ajp.amjpathol.org) and could, therefore, interact with the extracellular RGD binding site of the integrins on the plasma membrane. This is also supported by our IHC, which shows frequent membranous CTSZ immunoreactivity on PDAC cells. We subsequently performed integrin-CTSZ binding assays by coating plates with recombinant CTSZ and observed that both FA6 (to a higher extent) and MiaPaCa2 cells could bind to the coated surfaces. When cells were treated with an αv blocking antibody before seeding on the plates, cell binding was disrupted and a significant reduction in binding to CTSZ was seen (Figure 7A). By using flow cytometry, FA6 and MiaPaCa2 cells were tested for the presence of different αv integrins, and αvβ5 was the only αv integrin expressed by both cell lines (Figure 7B). Further investigation showed that blocking of αvβ5, using a specific antibody, inhibited the adhesion of cells to CTSZ (Figure 7C). Thus, αvβ5 mediates binding to CTSZ and could present the potential mechanism underlying the observed S100PBP-dependent changes in cell adhesion of PDAC cells.
      Figure thumbnail gr7
      Figure 7CTSZ interaction with RGD-binding integrins. A: CTSZ-binding assays were performed by coating adhesion plates with 3 μg/mL recombinant CTSZ and leaving cells to settle for 1 hour. Both FA6 and MiaPaCa2 cells could bind to CTSZ-coated plates; however, blocking αv on the surface of both FA6 and MiaPaCa2 cells using specific antibodies before seeding significantly reduced their ability to bind to the CTSZ-coated wells by up to 73%. P = 0.0216 and P = 0.0062 for FA6; P = 0.0168 and P = 0.0012 for MiaPaCa2 (compared to IgG and untreated controls, respectively). B: FA6 and MiaPaCa2 cells were screened for the presence of different αv integrins by flow cytometry analysis, with β1 as a positive control. The αvβ5 was the only integrin present in both cell lines. C: CTSZ-binding assays were subsequently repeated with blocking of integrin αvβ5; binding of FA6 and MiaPaCa2 cells to CTSZ was significantly decreased by up to 75%. P = 0.0069 and P = 0.0003 for FA6; P = 0.0222 and P = 0.0057 for MiaPaCa2 (compared to IgG and untreated controls, respectively). For all binding assays, data were represented in arbitrary units as an average of spectrophotometer readings at 595 nm after subtraction of BSA background values. For all panels, *P < 0.05, **P < 0.01, and ***P < 0.001.

      Discussion

      In this study, we demonstrate that the expression of S100PBP in various organs and disease states shows tissue specificity and differences in expression between healthy and tumor specimens. We focused on S100PBP in pancreatic tissues and showed that, in healthy pancreas, S100PBP protein localizes predominantly in a nuclear/perinuclear region, which correlates with our previously published in situ hybridization data.
      • Dowen S.E.
      • Crnogorac-Jurcevic T.
      • Gangeswaran R.
      • Hansen M.
      • Eloranta J.J.
      • Bhakta V.
      • Brentnall T.A.
      • Luttges J.
      • Kloppel G.
      • Lemoine N.R.
      Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
      We further show, using IHC on more samples, that, in addition to strong expression in endocrine cells, S100PBP protein is also expressed in the exocrine compartment. Furthermore, S100PBP appears to translocate to the cytoplasm during PDAC development (although this has to be substantiated using more PanIN samples), and its expression decreases in invasive cancer and in metastatic lesions in the lymph node, suggesting a potential anti-metastatic role of S100PBP. Our data from both Affymetrix gene expression profiling and IHC analysis also indicate that expression of S100PBP shows an inverse correlation to that of the metastasis-associated protein S100P, whose levels increase during PDAC progression.
      • Dowen S.E.
      • Crnogorac-Jurcevic T.
      • Gangeswaran R.
      • Hansen M.
      • Eloranta J.J.
      • Bhakta V.
      • Brentnall T.A.
      • Luttges J.
      • Kloppel G.
      • Lemoine N.R.
      Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
      Our in vitro data performed on cell lines after overexpression or silencing of S100PBP have implicated this protein in several functional roles associated with tumorigenesis. Changes in migration after modulation of S100PBP expression were seen in one of the four cell lines tested, yet no changes were seen in this cell line in wound healing assays. Possible reasons might lie in the differences in experimental conditions between the two assays; because transwell migration evaluates single-cell migration toward a chemoattractant, wound healing assays are better suited for study of group cell migration in complete media. Therefore, further investigation of the role of S100PBP in PDAC cell migration is warranted. Our data also suggest that loss of S100PBP may result in increased invasion, as seen in three of the four cell lines tested, which might have important implications in pancreatic tumorigenesis; studies addressing the underlying mechanisms are ongoing. However, because the only consistent functional change on modulation of S100PBP expression was observed in cell adhesion, this was studied in further detail.
      Adhesion to fibronectin, collagen I, and vitronectin was significantly decreased when S100PBP was overexpressed in FA6 and RwP1 cells and concomitantly increased when S100PBP was silenced in MiaPaCa2 and Panc1 cells. When S100PBP was overexpressed, there was a more significant and consistent decrease in cell adhesion on all surfaces compared with the changes witnessed when S100PBP was silenced. This is most likely because of the fact that silencing of S100PBP was performed transiently and produced a relatively lower deregulation of S100PBP compared with the stable overexpression experiments. This was also demonstrated in the Affymetrix gene expression profiling, in which there was a much greater increase in S100PBP expression after overexpression, than a decrease after S100PBP silencing. The mechanisms of the regulation of S100PBP expression are, however, still unknown and are the subject of further investigation.
      Interestingly, neither MiaPaca2 nor Panc1 cells express endogenous S100P, and FA6 cells only express low levels; therefore, S100PBP appears to regulate PDAC cell adhesion in an S100P-independent manner. To investigate the underlying mechanism behind the changes seen in pancreatic cancer cell adhesion, we further studied CTSZ. The CTSZ gene was found consistently deregulated across our transcriptome profiling experiments; it was up-regulated when S100PBP was silenced and down-regulated when S100PBP was overexpressed. CTSZ (previously known as cathepsin X, Y, and P) is a cysteine cathepsin belonging to a family of lysosomal proteolytic enzymes.
      • Santamaria I.
      • Velasco G.
      • Pendas A.M.
      • Fueyo A.
      • Lopez-Otin C.
      Cathepsin Z, a novel human cysteine proteinase with a short propeptide domain and a unique chromosomal location.
      Cysteine cathepsins are involved in protein processing, antigen presentation, bone remodeling, and epidermal homeostasis in healthy tissues.
      • Turk B.
      • Turk D.
      • Turk V.
      Lysosomal cysteine proteases: more than scavengers.
      • Turk V.
      • Turk B.
      • Guncar G.
      • Turk D.
      • Kos J.
      Lysosomal cathepsins: structure, role in antigen processing and presentation, and cancer.
      Although other cysteine proteases, cathepsins B and L, and aspartyl protease D have specifically been associated with cancer invasion and metastasis through degradation of the basement membrane and extracellular matrix (ECM),
      • Turk V.
      • Turk B.
      • Guncar G.
      • Turk D.
      • Kos J.
      Lysosomal cathepsins: structure, role in antigen processing and presentation, and cancer.
      • Klemencic I.
      • Carmona A.K.
      • Cezari M.H.
      • Juliano M.A.
      • Juliano L.
      • Guncar G.
      • Turk D.
      • Krizaj I.
      • Turk V.
      • Turk B.
      Biochemical characterization of human cathepsin X revealed that the enzyme is an exopeptidase, acting as carboxymonopeptidase or carboxydipeptidase.
      this is unlikely to be the case for CTSZ, because it is the only known cathepsin that acts as either a carboxymonopeptidase or a carboxydipeptidase.
      • Klemencic I.
      • Carmona A.K.
      • Cezari M.H.
      • Juliano M.A.
      • Juliano L.
      • Guncar G.
      • Turk D.
      • Krizaj I.
      • Turk V.
      • Turk B.
      Biochemical characterization of human cathepsin X revealed that the enzyme is an exopeptidase, acting as carboxymonopeptidase or carboxydipeptidase.
      CTSZ also differs from the other lysosomal cathepsins because it exhibits a short propeptide that contains an RGD integrin binding site.
      • Santamaria I.
      • Velasco G.
      • Pendas A.M.
      • Fueyo A.
      • Lopez-Otin C.
      Cathepsin Z, a novel human cysteine proteinase with a short propeptide domain and a unique chromosomal location.
      • Sivaraman J.
      • Nagler D.K.
      • Zhang R.
      • Menard R.
      • Cygler M.
      Crystal structure of human procathepsin X: a cysteine protease with the proregion covalently linked to the active site cysteine.
      CTSZ is predominately expressed in immune cells, where it promotes cell attachment via interactions with the integrin β2,
      • Obermajer N.
      • Repnik U.
      • Jevnikar Z.
      • Turk B.
      • Kreft M.
      • Kos J.
      Cysteine protease cathepsin X modulates immune response via activation of beta2 integrins.
      although high levels of CTSZ expression have also been seen in gastric carcinomas and prostate cancer, including its precursor lesion, prostatic intraepithelial neoplasia.
      • Buhling F.
      • Peitz U.
      • Kruger S.
      • Kuster D.
      • Vieth M.
      • Gebert I.
      • Roessner A.
      • Weber E.
      • Malfertheiner P.
      • Wex T.
      Cathepsins K, L, B, X and W are differentially expressed in normal and chronically inflamed gastric mucosa.
      • Nagler D.K.
      • Kruger S.
      • Kellner A.
      • Ziomek E.
      • Menard R.
      • Buhtz P.
      • Krams M.
      • Roessner A.
      • Kellner U.
      Up-regulation of cathepsin X in prostate cancer and prostatic intraepithelial neoplasia.
      Also, pro-CTSZ can modulate the adhesive properties of endothelial cells through interaction with integrin αvβ3,
      • Lechner A.M.
      • Assfalg-Machleidt I.
      • Zahler S.
      • Stoeckelhuber M.
      • Machleidt W.
      • Jochum M.
      • Nagler D.K.
      RGD-dependent binding of procathepsin X to integrin alphavbeta3 mediates cell-adhesive properties.
      thus indicating that CTSZ may play an important role in adhesion.
      There is little information concerning the expression of CTSZ and its functions in PDAC. By using IHC, we show that, in healthy pancreas, CTSZ is predominantly expressed in the islets and in approximately 20% to 30% of ducts, whereas it is expressed in all PDAC cases tested. This is an inverse pattern to that witnessed with S100PBP expression in the same panel of cases, which showed expression in all cellular compartments in healthy specimens but a significant decrease in expression in PDAC and lymph node metastatic samples. Interestingly, the pattern of the CTSZ staining and the localization differed as the lesions progressed; however, the significance of this finding is unknown. In addition, our in vitro data suggest that CTSZ may play a pivotal role in the modulation of pancreatic cancer adhesion after the deregulation of S100PBP because overexpression of this molecule led to a significant decrease in CTSZ expression and decreased adhesion; in congruence, silencing of CTSZ in FA6 led to a significant decrease in adhesion. Furthermore, the increase in adhesion after S100PBP silencing in MiaPaCa2 cells could be rescued by silencing CTSZ. This suggests that S100PBP can modulate the adhesion of PDAC cells to the ECM through CTSZ.
      To determine potential downstream effectors of the S100PBP-CTSZ interaction, we examined the role of the integrins, which are the principal mediators of adhesion to the ECM. Integrins are heterodimers composed of two noncovalently associated α and β chains and are expressed on the cell surface of all nucleated cells.
      • Arnaout M.A.
      • Goodman S.L.
      • Xiong J.P.
      Structure and mechanics of integrin-based cell adhesion.
      A common RGD motif is recognized by eight different integrins (αvβ1vβ3, αvβ5, αvβ6, αvβ8, α5β1, α8β1, andαIIbβ3).
      • Barczyk M.
      • Carracedo S.
      • Gullberg D.
      Integrins.
      Because it was previously shown that CTSZ interacts with αvβ3,
      • Lechner A.M.
      • Assfalg-Machleidt I.
      • Zahler S.
      • Stoeckelhuber M.
      • Machleidt W.
      • Jochum M.
      • Nagler D.K.
      RGD-dependent binding of procathepsin X to integrin alphavbeta3 mediates cell-adhesive properties.
      we used specific antibodies to block all αv integrins; this inhibited, to a large extent, binding of pancreatic cancer cells to CTSZ. Of the eight RGD binding integrins, αvβ3 and αvβ5 are commonly seen in cancer cells and have previously been associated with tumor growth and metastasis
      • Arnaout M.A.
      • Goodman S.L.
      • Xiong J.P.
      Structure and mechanics of integrin-based cell adhesion.
      ; however, only αvβ5 was expressed at significant levels in our cell lines. Interestingly, blocking with an αvβ5-specific antibody was sufficient to significantly diminish the ability of the tested pancreatic cancer cells to bind CTSZ. αvβ5 is not known to interact with collagen I directly,
      • Barczyk M.
      • Carracedo S.
      • Gullberg D.
      Integrins.
      but it is plausible that this binding could be mediated by cross talk with other integrins, as recently reported for αvβ5 and α5β1 in the interaction of osteosarcoma cells with the fibronectin matrix.
      • Vial D.
      • McKeown-Longo P.J.
      PAI1 stimulates assembly of the fibronectin matrix in osteosarcoma cells through crosstalk between the alphavbeta5 and alpha5beta1 integrins.
      This, however, warrants further investigation. Overall, these data suggest that the role of CTSZ in adhesion is dependent on the cell type–specific display of integrins.
      Our data demonstrate a novel regulatory mechanism of pancreatic cancer cell adhesion, modulated by S100PBP and CTSZ. We show that CTSZ is negatively regulated by S100PBP at the transcriptional level (although the detailed S100PBP functions in the nucleus are not known) and that CTSZ can subsequently bind several RGD integrins, including αvβ5, on the plasma membrane of pancreatic cancer cells. It is widely acknowledged that, to metastasize, cancer cells have to adhere to the ECM of secondary organs.
      • Hanahan D.
      • Weinberg R.A.
      The hallmarks of cancer.
      • Gupta G.P.
      • Massague J.
      Cancer metastasis: building a framework.
      In some cancers, such as prostate, it has also been demonstrated that disruption of the adhesion of circulating tumor cells to microvascular endothelium will reduce their metastatic potential as they roll along the endothelium to distant sites.
      • Hsu J.W.
      • Yasmin-Karim S.
      • King M.R.
      • Wojciechowski J.C.
      • Mickelsen D.
      • Blair M.L.
      • Ting H.J.
      • Ma W.L.
      • Lee Y.F.
      Suppression of prostate cancer cell rolling and adhesion to endothelium by 1alpha,25-dihydroxyvitamin d(3).
      The novel mechanism of the regulation of cell adhesion through S100PBP may, thus, be important in the suppression of dissemination of PDAC cells; conversely, loss of S100PBP might facilitate metastatic spread. In support of this hypothesis, the metastatic lesions tested herein showed decreased levels of S100PBP. We have previously shown that overexpression of S100P in PDAC cells causes up-regulation of cathepsin D,
      • Whiteman H.J.
      • Weeks M.E.
      • Dowen S.E.
      • Barry S.
      • Timms J.F.
      • Lemoine N.R.
      • Crnogorac-Jurcevic T.
      The role of S100P in the invasion of pancreatic cancer cells is mediated through cytoskeletal changes and regulation of cathepsin D.
      which is an important mediator of increased invasion.
      • Briozzo P.
      • Morisset M.
      • Capony F.
      • Rougeot C.
      • Rochefort H.
      In vitro degradation of extracellular matrix with Mr 52,000 cathepsin D secreted by breast cancer cells.
      The current findings support the concept of complex interplay and cooperation of different cathepsins, S100 proteins, and S100PBP in the pathobiological characteristics of dissemination of pancreatic cancer cells. Furthermore, our data provide a framework and pave the direction for further functional studies of this novel and important protein.

      Acknowledgment

      We thank Prof. Helen C. Hurst for critical reading of the manuscript.

      Supplementary data

      • Supplemental Figure S1

        Link between S100PBP and S100P expression. A: Direct comparison of S100PBP and S100P protein expression in the same pancreatic tissues was performed. Representative images of two normal pancreas, two pancreatic intraepithelial neoplasia (PanIN) lesions, and two pancreatic ductal adenocarcinoma (PDAC) cases are shown. Brown staining indicates immunoreactivity in normal pancreas, early PanINs (p), and in PDAC. Normal pancreatic acinar cells (a), ducts (d) and islets (i) are indicated. Original magnification, ×100. B: Tables summarizing the overall IHC scores for both S100PBP and S100P expression. In total, 13 normal and 45 PDAC cases were successfully analyzed.

      • Supplemental Figure S2

        S100PBP RNA expression in pancreatic cell lines. One normal and eight PDAC immortalized ductal cell lines (HPDEs) were profiled for S100PBP expression using quantitative real-time PCR. The cell line expression correlates to that reported previously.

        • Dowen S.E.
        • Crnogorac-Jurcevic T.
        • Gangeswaran R.
        • Hansen M.
        • Eloranta J.J.
        • Bhakta V.
        • Brentnall T.A.
        • Luttges J.
        • Kloppel G.
        • Lemoine N.R.
        Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
        Results are represented as Ct values relative to an 18S internal standard. MiaPaCa2 cells were used for S100PBP silencing and FA6 cells for S100PBP over-expression experiments. Of note, the transfection efficiency of AsPc1 and PaTu8988s cells was too low.

      • Supplemental Figure S3

        S100PBP functional assays. After S100PBP over-expression (FLAG, control; FLAG-S100PBP, over-expressing cells) and silencing (NT siRNA, control; S100PBP siRNA, silenced) migration and invasion assays were performed. A significant increase was seen in migration (***P < 0.0001) of FA6 cells over-expressing S100PBP compared to control cells, but no significant change was seen in migration when S100PBP was silenced in MiaPaCa2 or Panc1 cells, nor when it was over-expressed in RwP1 cells. No significant changes were seen in invasion when S100PBP was over-expressed in FA6 cells, however, when over-expressed in RwP1, a significant decrease was seen (*P = 0.017). When S100PBP was silenced a significant increase in both MiaPaCa2 (**P = 0.0063) and Panc1 (*P = 0.015) cells was seen.

      • Supplemental Figure S5

        Western blot of both cell pellets and conditioned media (collected after 24 hours) indicated that pro-CTSZ, which is known to contain an arganine-glycine-aspartic acid (RGD) integrin binding site, is secreted from both FA6 and MiaPaCa2 cells.

      References

        • Olsen A.H.
        • Parkin D.M.
        • Sasieni P.
        Cancer mortality in the United Kingdom: projections to the year 2025.
        Br J Cancer. 2008; 99: 1549-1554
        • Brat D.J.
        • Lillemoe K.D.
        • Yeo C.J.
        • Warfield P.B.
        • Hruban R.H.
        Progression of pancreatic intraductal neoplasias to infiltrating adenocarcinoma of the pancreas.
        Am J Surg Pathol. 1998; 22: 163-169
        • Crnogorac-Jurcevic T.
        • Missiaglia E.
        • Blaveri E.
        • Gangeswaran R.
        • Jones M.
        • Terris B.
        • Costello E.
        • Neoptolemos J.P.
        • Lemoine N.R.
        Molecular alterations in pancreatic carcinoma: expression profiling shows that dysregulated expression of S100 genes is highly prevalent.
        J Pathol. 2003; 201: 63-74
        • Logsdon C.D.
        • Simeone D.M.
        • Binkley C.
        • Arumugam T.
        • Greenson J.K.
        • Giordano T.J.
        • Misek D.E.
        • Kuick R.
        • Hanash S.
        Molecular profiling of pancreatic adenocarcinoma and chronic pancreatitis identifies multiple genes differentially regulated in pancreatic cancer.
        Cancer Res. 2003; 63: 2649-2657
        • Iacobuzio-Donahue C.A.
        • Maitra A.
        • Olsen M.
        • Lowe A.W.
        • van Heek N.T.
        • Rosty C.
        • Walter K.
        • Sato N.
        • Parker A.
        • Ashfaq R.
        • Jaffee E.
        • Ryu B.
        • Jones J.
        • Eshleman J.R.
        • Yeo C.J.
        • Cameron J.L.
        • Kern S.E.
        • Hruban R.H.
        • Brown P.O.
        • Goggins M.
        Exploration of global gene expression patterns in pancreatic adenocarcinoma using cDNA microarrays.
        Am J Pathol. 2003; 162: 1151-1162
        • Dowen S.E.
        • Crnogorac-Jurcevic T.
        • Gangeswaran R.
        • Hansen M.
        • Eloranta J.J.
        • Bhakta V.
        • Brentnall T.A.
        • Luttges J.
        • Kloppel G.
        • Lemoine N.R.
        Expression of S100P and its novel binding partner S100PBPR in early pancreatic cancer.
        Am J Pathol. 2005; 166: 81-92
        • Arumugam T.
        • Simeone D.M.
        • Van Golen K.
        • Logsdon C.D.
        S100P promotes pancreatic cancer growth, survival, and invasion.
        Clin Cancer Res. 2005; 11: 5356-5364
        • Whiteman H.J.
        • Weeks M.E.
        • Dowen S.E.
        • Barry S.
        • Timms J.F.
        • Lemoine N.R.
        • Crnogorac-Jurcevic T.
        The role of S100P in the invasion of pancreatic cancer cells is mediated through cytoskeletal changes and regulation of cathepsin D.
        Cancer Res. 2007; 67: 8633-8642
        • Deng H.
        • Shi J.
        • Wilkerson M.
        • Meschter S.
        • Dupree W.
        • Lin F.
        Usefulness of S100P in diagnosis of adenocarcinoma of pancreas on fine-needle aspiration biopsy specimens.
        Am J Clin Pathol. 2008; 129: 81-88
        • Arumugam T.
        • Logsdon C.D.
        S100P: a novel therapeutic target for cancer.
        Amino Acids. 2011; 41: 893-899
        • Ouyang H.
        • Mou L.
        • Luk C.
        • Liu N.
        • Karaskova J.
        • Squire J.
        • Tsao M.S.
        Immortal human pancreatic duct epithelial cell lines with near normal genotype and phenotype.
        Am J Pathol. 2000; 157: 1623-1631
        • Froeling F.E.
        • Mirza T.A.
        • Feakins R.M.
        • Seedhar A.
        • Elia G.
        • Hart I.R.
        • Kocher H.M.
        Organotypic culture model of pancreatic cancer demonstrates that stromal cells modulate E-cadherin, beta-catenin, and Ezrin expression in tumor cells.
        Am J Pathol. 2009; 175: 636-648
        • Smyth G.K.
        Linear models and empirical bayes methods for assessing differential expression in microarray experiments.
        Stat Appl Genet Mol Biol. 2004; 3 (Article3)
        • Lechner A.M.
        • Assfalg-Machleidt I.
        • Zahler S.
        • Stoeckelhuber M.
        • Machleidt W.
        • Jochum M.
        • Nagler D.K.
        RGD-dependent binding of procathepsin X to integrin alphavbeta3 mediates cell-adhesive properties.
        J Biol Chem. 2006; 281: 39588-39597
        • Santamaria I.
        • Velasco G.
        • Pendas A.M.
        • Fueyo A.
        • Lopez-Otin C.
        Cathepsin Z, a novel human cysteine proteinase with a short propeptide domain and a unique chromosomal location.
        J Biol Chem. 1998; 273: 16816-16823
        • Turk B.
        • Turk D.
        • Turk V.
        Lysosomal cysteine proteases: more than scavengers.
        Biochim Biophys Acta. 2000; 1477: 98-111
        • Turk V.
        • Turk B.
        • Guncar G.
        • Turk D.
        • Kos J.
        Lysosomal cathepsins: structure, role in antigen processing and presentation, and cancer.
        Adv Enzyme Regul. 2002; 42: 285-303
        • Klemencic I.
        • Carmona A.K.
        • Cezari M.H.
        • Juliano M.A.
        • Juliano L.
        • Guncar G.
        • Turk D.
        • Krizaj I.
        • Turk V.
        • Turk B.
        Biochemical characterization of human cathepsin X revealed that the enzyme is an exopeptidase, acting as carboxymonopeptidase or carboxydipeptidase.
        Eur J Biochem. 2000; 267: 5404-5412
        • Sivaraman J.
        • Nagler D.K.
        • Zhang R.
        • Menard R.
        • Cygler M.
        Crystal structure of human procathepsin X: a cysteine protease with the proregion covalently linked to the active site cysteine.
        J Mol Biol. 2000; 295: 939-951
        • Obermajer N.
        • Repnik U.
        • Jevnikar Z.
        • Turk B.
        • Kreft M.
        • Kos J.
        Cysteine protease cathepsin X modulates immune response via activation of beta2 integrins.
        Immunology. 2008; 124: 76-88
        • Buhling F.
        • Peitz U.
        • Kruger S.
        • Kuster D.
        • Vieth M.
        • Gebert I.
        • Roessner A.
        • Weber E.
        • Malfertheiner P.
        • Wex T.
        Cathepsins K, L, B, X and W are differentially expressed in normal and chronically inflamed gastric mucosa.
        Biol Chem. 2004; 385: 439-445
        • Nagler D.K.
        • Kruger S.
        • Kellner A.
        • Ziomek E.
        • Menard R.
        • Buhtz P.
        • Krams M.
        • Roessner A.
        • Kellner U.
        Up-regulation of cathepsin X in prostate cancer and prostatic intraepithelial neoplasia.
        Prostate. 2004; 60: 109-119
        • Arnaout M.A.
        • Goodman S.L.
        • Xiong J.P.
        Structure and mechanics of integrin-based cell adhesion.
        Curr Opin Cell Biol. 2007; 19: 495-507
        • Barczyk M.
        • Carracedo S.
        • Gullberg D.
        Integrins.
        Cell Tissue Res. 2010; 339: 269-280
        • Vial D.
        • McKeown-Longo P.J.
        PAI1 stimulates assembly of the fibronectin matrix in osteosarcoma cells through crosstalk between the alphavbeta5 and alpha5beta1 integrins.
        J Cell Sci. 2008; 121: 1661-1670
        • Hanahan D.
        • Weinberg R.A.
        The hallmarks of cancer.
        Cell. 2000; 100: 57-70
        • Gupta G.P.
        • Massague J.
        Cancer metastasis: building a framework.
        Cell. 2006; 127: 679-695
        • Hsu J.W.
        • Yasmin-Karim S.
        • King M.R.
        • Wojciechowski J.C.
        • Mickelsen D.
        • Blair M.L.
        • Ting H.J.
        • Ma W.L.
        • Lee Y.F.
        Suppression of prostate cancer cell rolling and adhesion to endothelium by 1alpha,25-dihydroxyvitamin d(3).
        Am J Pathol. 2011; 178: 872-880
        • Briozzo P.
        • Morisset M.
        • Capony F.
        • Rougeot C.
        • Rochefort H.
        In vitro degradation of extracellular matrix with Mr 52,000 cathepsin D secreted by breast cancer cells.
        Cancer Res. 1988; 48: 3688-3692