INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Signaling between stromal and epithelial cells directs specific growth and differentiation patterns in many tissues, including the lung (1, 2), gastrointestinal tract (3), liver (4), skin (5), the mammary gland (6), and the prostate gland (7-10). Growth factors comprise a key component of stromal-epithelial interactions. Growth factors implicated in prostate gland biology include: the fibroblast growth factor family (11, 12), the transforming growth factor- family (13), transforming growth factor- (14), the insulin-like growth factor family (15), epidermal growth factor (16, 17), heparin-binding epidermal growth factor-like growth factor (18), tumor necrosis factor- (19), and heparin growth factor-scatter factor (20). Each of these growth factors are bound to the extracellular matrix, cell membrane, or other binding proteins in an inactive or sequestered form and their activity regulated by secreted proteases and protease inhibitors. The actions of proteases and protease inhibitors are fundamental to tissue homeostasis through the regulation of both growth factor bioavailability and cell interaction with extracellular matrix. Alterations in protease and protease inhibitor actions affect tissue development patterns, angiogenesis, cell motility, and tumor invasion (21-24).

In previous studies to identify stromally derived paracrine-acting factors which regulate prostatic epithelial cell proliferation and differentiation, we identified a urogenital sinus mesenchymal secreted growth-inhibitory activity (UGIF¹ activity), which altered epithelial cell proliferation, morphology, and protein secretion in several epithelial cell lines (25, 26). A novel mesenchymally secreted protein, termed ps20 (20-kDa prostate stromal protein), responsible for UGIF activity was purified to homogeneity and its amino acid sequence determined (27). Purified ps20 inhibited PC-3 prostatic epithelial carcinoma cell proliferation in a dose-dependent and saturable manner, altered phenotypic morphology, and stimulated protein synthesis and secretion in vitro (27). Specific mechanisms of ps20 action have not yet been determined.

We report here the cloning of full-length cDNA encoding ps20, expression of biologically active recombinant ps20, development of a ps20 antibody, and analysis of in vivo expression patterns. The ps20 protein is a novel member of the whey acidic protein (WAP)-type "four-disulfide core" domain protein family (28, 29). This family is composed of small, secreted serine protease inhibitors, which exhibit a variety of growth- and differentiation-regulatory functions and affect extracellular matrix remodeling and carcinoma progression. We demonstrate here that recombinant ps20 inhibits the proliferation of prostate carcinoma PC-3 cells and COS-7 cells. In addition, we show that ps20 is expressed specifically in the smooth cell type of the prostate gland and other tissues including vascular smooth muscle. These studies suggest that ps20 may function as a mediator of local growth and differentiation mechanisms.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Preparation of the PS-1 (Adult Rat Prostate Smooth Muscle Cell Line) ZAP Express (PS-1_BfsZAP) cDNA Library-- An oligo(dT) reverse transcribed cDNA library of 10⁹ to 10¹² plaque-forming units/ml was prepared using the ZAP Express cDNA synthesis kit (Stratagene, La Jolla, CA). PS-1 cells (adult rat prostate smooth muscle-like cell line) (30), were grown to post-confluence in Bfs medium (90% Dulbecco's modified Eagle's medium (DMEM), 5% fetal calf serum, 5% NuSerum (Becton Dickinson, Bedford, MA), 5 µg/ml insulin, 100 units/ml penicillin, and 100 µg/ml streptomycin) supplemented with androgens (0.5 µg/ml testosterone) (26). RNA was prepared using the RNAgents total RNA isolation system and poly(A) RNA selected with the PolyATtract^® system (both from Promega, Madison, WI). RNA was size-fractionated on a Sephacryl column prior to ligation with vector arms.

ps20 Cloning-- The amino-terminal amino acid sequence of ps20 was determined and previously reported (27): N-Thr-Trp-Glu-Ala-Met-Leu-Pro-Val-Arg-Leu-Ala-Glu-Lys-Ser-X-X-X-X-Val-Ala-Ala-(Thr)-Gly-X-Arg-Gln-Pro-(His)-C. Degenerate ps20 primers were designed from this sequence: primer P34 (256-fold degenerate, amino acids 1-7), primer P36 (16-fold degenerate, amino acids 1-7), primer P37 (64-fold degenerate, amino acids 7-13), and primer P38 (16, 384-fold degenerate, amino acids 27-19). Inosines were incorporated at sites of complete degeneracy (31) in primers P34 and P38, whereas optimal mammalian codon usage was assumed in the design of primers P36 and P37. The ps20-specific primers P40 and P42 were designed from clone 3438-1pCRII sequence and ps20-specific primers P45 and P46 from clone 42T7pCRII sequence. Primer sequences (IUPAC nomenclature) are identified in terms of rat ps20 full-length cDNA sequence (Fig. 2): P34 (nt 130-150), 5'-ACI TGG GAR GCI ATG YTI CC-3'; P36 (nt 130-150), 5'-ACM TGG GAR GCY ATG CTS CC-3'; P37 (nt 148-168), 5'-CCY GTS MGR CTG GCY GAR AA-3'; P38 (nt 209-183), 5'-GGC TGY CKI IIS CCK GTK GCR GCN AC-3'; P40 (nt 169-151), 5'-TCA GCT TGG GAC TTC TCA G-3'; P42 (nt 153-171), 5'-GAG AAG TCC CAA GCT GAA G-3'; P45 (nt 425-444), 5'-GGA GGA AGT GTT ACA AGC AG-3'; and P46 (nt 1011-992), 5'-GGT GGG CAG ATT TAT TCG GG-3. Primers P34-P42 were synthesized and purified using high performance liquid chromatography by Keystone Labs, Inc. (Menlo Park, CA). Primers P45 and P46 were synthesized by Genosys (The Woodlands, TX).

Sequencing-- Multiple pass DNA sequencing was performed on an Applied Biosystems model 377 Sequencer version 2.1.1 (University of Texas Houston Molecular Genetics Core Facility) using Ampli-Taq polymerase, 3 pmol of primer (pBKT3-1-, pBKT7-1-, M13-20-, and ps20-specific primers), and 1 µg of double-stranded DNA per reaction, or by manual sequencing with Sequenase version 2.0 DNA sequencing kit (U. S. Biochemica Corp./Amersham Life Science) (33). Sequences were assembled using MacVector version 4.1 and AssemblyLign version 1.0 (Kodak).

Sequence Analysis-- Nucleotide sequence searches were performed on all available data bases using the BLASTN and TBLASTN (blast enhanced alignment utility) algorithms (34). Secondary structure predictions of deduced amino acid sequence were made by SOPMA (Self Optimized Prediction Method from Alignment) (35) and SBASE (36) and confirmed by BLASTX and TBLASTX (34) searches of the available data bases (37). Secretory peptide prediction was confirmed by PSORT version 6.3.

Northern Analysis-- Total RNA was isolated from adult male Sprague-Dawley rat tissues (intact, castrate 2 weeks, and castrate 6 weeks), adult female rat tissues (kindly provided by J. S. Richards, Baylor College of Medicine) and from U4F cells grown to post-confluence in Bfs medium. Total RNA was prepared using RNA Stat-60 total RNA isolation reagent (Tel-Test "B," Inc., Friendswood, TX). RNA was electrophoresed through 1% agarose gels (containing 1 × MOPS and 5.1% of a 37% formaldehyde solution), which were transferred to neutral charged nylon membranes (Schleicher & Schuell) by downward capillary blotting. Inserts from T340pCRII and 42T7pCRII were excised, gel-purified, labeled by random priming with [-³²P]dCTP, purified, and denatured as described above. Blots were UV-cross-linked and exposed to probe at 1-5 × 10⁶ counts/ml, overnight, at 42 °C; washed with multiple 20 min washes at 60 °C in 0.1% SDS containing 2 to 0.1 × SSC; and exposed to Kodak XAR film for 24 h to 2 weeks, as necessary.

Preparation of ps20 Antisera and Affinity Purification-- The amino-terminal sequence of native ps20 was determined as described previously (27). A unique 14-amino acid synthetic peptide, ps20 peptide-(1-14), corresponding to the amino terminus of purified ps20 (see Fig. 2), was synthesized on an Applied Biosystems model 430A peptide synthesizer: N-Thr-Trp-Glu-Ala-Met-Leu-Pro-Val-Arg-Leu-Ala-Glu-Lys-Ser-C (Baylor College of Medicine Protein Core Facility). For initial immunization, ps20 peptide was solubilized in sterile, tissue culture grade H₂O (400 µg/ml), mixed with Freund's complete adjuvant (1:1 ratio), and injected in 500-µl (100-µg) aliquots at multiple subcutaneous and intramuscular sites of three female New Zealand rabbits. At 3 weeks following immunization, sera samples were analyzed by solid phase enzyme-linked immunoabsorbent assay (ELISA). Each rabbit received a booster of 100 µg of peptide in 500 µl of Freund's incomplete adjuvant injected subcutaneously and intramuscularly at 2 weeks and 5 weeks. Serum samples were tested for ps20-specific antibody every 2 weeks and immunoglobulin subtype determined by ELISA analysis. High titer antiserum (reactive at 1:1 × 10⁶ dilution) was pooled, and the IgG class was precipitated by ammonium sulfate (50% saturation), resolubilized in 1 × PBS, and dialyzed overnight against 1 × PBS at 4 °C. This antibody preparation is referred to as ps20ab-1.

Cell Culture-- The U4F urogenital sinus mesenchymal cell line (26) was derived from a urogenital sinus organ explant (25) and the PS-1 cell line (30) derived from an adult rat prostate organ explant. U4F cells and PS-1 cells were maintained in medium Bfs as described previously (26, 27, 30). PC-3 cells (ATCC CRL-1435, human prostate adenocarcinoma) were maintained as recommended by the ATCC and as described previously (27). The COS-7 cell line (ATCC CRL-1651, SV40 transformed African green monkey kidney fibroblast-like) was maintained in 90% DMEM, 10% fetal calf serum (DMEM-10) (BioWhittaker, Walkersville, MD), 100 units/ml penicillin, and 100 µg/ml streptomycin. Cultures were tested for mycoplasma every 6 months.

COS-7 Cell Transient Transfection-- COS-7 cells were seeded at 1.7 × 10⁶ cells/ml in 100-mm culture dishes in 90% DMEM, 10% fetal calf serum (DMEM-10) with no antibiotics 24 h prior to transfection. Cells were transfected at 80% confluence with 4.1 µg of ps20-2bpBKCMV-lac DNA or pBKCMV-lac control plasmid DNA (prepared by Qiagen tip-500) and 41.0 µl of LipofectAMINE reagent (Life Technologies, Inc.), following recommendations of the manufacturer.

COS-7 Cell Stable Transfectant Cell Lines-- To select for stably transfected COS-7 cell lines, 48 h following transfection with either ps20-2bpBKCMV-lac DNA or pBKCMV-lac control plasmid DNA cells were split 1:15 and replated in COS-7 cell medium containing 600 µg/ml G418 (Sigma). Cells were grown in G418-containing medium for 3 weeks when pools of G418-resistant colonies were collected. Clonal cell lines were obtained by serial dilution in 96-well plates and assayed for ps20 expression by Western analysis of the conditioned medium. Once established, stable cell lines were maintained in 400 µg/ml G418-containing medium (DMEM-10/G418).

Crystal Violet Staining-- Cells were grown to desired density and then incubated with a solution of crystal violet (0.5% (w/v) crystal violet, 3.2% formaldehyde, 0.17% (w/v) NaCl, 22% ethanol overnight, washed three times with H₂O, allowed to dry, and photographed.

[³H]Thymidine Incorporation Assays-- Cells were seeded in a 150-µl volume in 96-well plates at a density of 5.0 × 10⁴ cells/ml (PC-3) and 2.5 × 10⁴ cells/ml (COS-7-derived cell lines), and allowed to attach overnight. Sample was added in triplicate in a 50-µl volume (if applicable) and allowed to incubate an additional 24 h. Cultures were pulsed with 10 µl of 2 µCi/ml [³H]thymidine (Amersham) during the final 3 h of incubation, fixed, and processed as described previously (27). To assay for ps20 activity in the conditioned medium (CM) of transfectant lines, cultures were seeded in a 5-ml volume in 25-mm² flasks at 1 × 10⁵ cells/ml in standard medium, CM collected after 24 or 48 h, and 100 or 200 µl of CM added to pBKCMV-4 cells and cultures assayed as described above. All assays were performed in triplicate and each experiment repeated multiple times (n  3). Data are presented as mean ± S.E. (error bars) and significance determined by Student's t test.

Immunoprecipitation-- 75-mm² flasks of PS-1 and U4F cells, grown in Bfs medium, were allowed to reach post-confluence. Cells were then incubated in M medium for 6 h then for 24 h in 4 ml of medium M made with DMEM lacking Cys, Met, and glutamine (ICN, Costa Mesa, CA), 2 mM L-glutamine, and 0.6 mCi of ³⁵S (ICN, Boston, MA). 300 µl of ³⁵S-labeled medium was incubated overnight with an equal amount of RIPA buffer (50 mM Tris, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, pH 8.0) and 10 µl of ps20ab-1 or 50 µl preimmune sera followed by a 1-h incubation with 50 µl of Protein A-Sepharose (Pharmacia). Beads were washed four times in 500 µl of RIPA buffer, resuspended in SDS sample buffer with or without 100 mM dithiothreitol, and resolved by 12% SDS-PAGE. The gel was fixed for 30 min in 30% methanol with 10% acetic acid, incubated for 20 min in Enlightning (NEN Life Science Products), dried, and exposed to X-OMAT AR film (Kodak) for 12-48 h.

Expression and Purification of rps20-His-- PCR primers, P69 and P60, were used to synthesize a modified 585-nt fragment of the ps20 coding region for cloning into the bacterial expression vector, pET23c (Novagen, Madison, WI), a vector that allows incorporation of a His tag (six consecutive histidine residues) at the carboxyl terminus. At the amino terminus, the secretory peptide (amino acids 1-26) was replaced with a Met-Ala and the fifth codon was modified to optimize for bacterial expression. Primer P69 (nt 117-153; 5'-C TCT cat atg GCT ACT TGG GAA GCA ATG TTG CCG GTC-3') includes an NdeI site (lowercase) containing an internal ATG start site. Nucleotides modified from ps20 cDNA sequence are underlined. Primer P60 (nt 672-697; 5'-C TCC gaa ttc TGG AAA GTG CCT CTG T-3'), contains an EcoRI site (lowercase) immediately following the final amino acid for cloning into pET23c. Because the EcoRI site in the pET23c vector is out of frame with the His tag, the vector was modified by deleting the four-nucleotide SacI 3' overhang downstream of the EcoRI site; the entire construct was digested with SacI, digested with T4 polymerase in the presence of 100 µg/ml dNTPs at 37 for 10 min, re-ligated, and confirmed by sequencing. The ps20 cDNA clone ps20-2bpBKCMV (1 ng) was used as template using standard PCR procedures with primers P69 and P60 to amplify a PCR product that was cloned into the T-tailed vector, pCRII. The insert from one clone, rps20pCRII, confirmed by automated sequencing, was excised by a NdeI/EcoRI double digest and ligated with NcoI/EcoRI-digested pET23c. An insert-containing clone, rps20-1pET23c, was confirmed by sequencing and transformed into the BL21(DE3)pLysS bacterial expression cell line (Novagen). 500-ml cultures were grown to OD₆₀₀ = 0.6, protein expression induced by 0.4 mM isopropyl-1-thio--D-galactopyranoside for 3 h, and soluble cell extracts generated by sonication in SB buffer (50 mM phosphate, 300 mM NaCl, pH 8.0) on ice followed by centrifugation at 15,000 × g for 20 min. Soluble recombinant rat ps20 His-tagged protein (rps20-His) was purified on a nickel-charged resin (Ni⁺-NTA, Qiagen). The column was washed and eluted with an imidazole gradient, as recommended by the manufacturer. Fractions containing rps20-His with apparent >95% purity were pooled and dialyzed to remove imidazole.

Cloning, Expression, and Purification of GST-rps20-- The rps20pCRII clone was modified by cloning the oligonucleotide 5'-CATAcGGATCCCTCGAGCATATG-3' into the NdeI site providing an artificial ATG prior to nucleotide 129 (amino acid 26) of the rps20 cDNA sequence. The lowercase nucleotide represents a mutation in the 5' NdeI site, and the underlined nucleotides represent BamHI and XhoI sites, respectively. This clone, rps20BXpCRII, was digested with BamHI and EcoRI and the 647-base pair fragment purified on a 0.8% agarose gel by the following procedure. The 647-base pair fragment was run onto DE-81 paper (Whatman) and incubated overnight at 25 °C in 500 µl of 0.5 × TBE, 1.25 M NaCl. The paper was then filtered through a 0.45-µm filter, and the DNA was ethanol-precipitated and resuspended in dH₂O. This rps20 fragment was then cloned into the BamHI and EcoRI sites in the pGEX2TK vector (Pharmacia) (39), containing a thrombin cleavage and cAMP-dependent protein kinase phosphorylation site, creating the vector rps20pGEX2TK.

Western Analysis-- CM from transfected COS-7 cells and from U4F cells were prepared identically for Western analysis. U4F cells were seeded in Bfs medium and COS-7 stable transfectants seeded in DMEM-10/G418. Both were switched to serum-free, fully defined medium M (MCDB-110, ITS, 0.5 µg/ml testosterone, 0.1 µg/ml epidermal growth factor, 100 units/ml penicillin, and 100 µg/ml streptomycin) (40). CM-M was collected every 48-72 h, clarified by centrifugation, and stored at 20 °C until use. Proteins from CM were precipitated with ammonium sulfate (0-45% saturation) for 3 h at 4 °C, centrifuged at 9.5 × g for 20 min at 4 °C, and pellets resuspended in 0.5 ml of 1 M acetic acid. Samples were dialyzed 12-16 h against 1 M acetic acid in Spectra/Por M_r 3500 dialysis tubing (Spectrum Medical Industries, Inc., Houston, TX), dried by Speed Vac (Savant Instruments), and resolubilized in 1 × SDS Laemmli sample buffer containing (700 mM) -mercaptoethanol at 37 °C. Bacterial cell extracts and purified protein were mixed with 3 × SDS-sample buffer with or without 700 mM -mercaptoethanol or 100 mM dithiothreitol. Samples were boiled at 95 °C for 10 min, and resolved by 0.75-mm or 1.5-mm 15% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) following the procedures of Laemmli (41), as described previously (27).

Immunohistochemistry-- Two-month-old and 6-month-old male Sprague-Dawley rats were sacrificed (intact and 2-week and 8-week castrates), and tissues removed and fixed in 10% neutral buffered formalin overnight. Fixed tissues were dehydrated in a graded series of ethanol washes, embedded in paraffin, cut into 5-µm-thick sections, applied to poly-L-lysine-coated slides, and baked at 37 °C prior to staining. Tissue sections were deparaffinized by immersion in Hemo D (Fisher) twice for 10 min each; rehydrated by 5-min graded washes in 100%, 95%, and 70% ethanol; permeabilized by immersion in 1 × PBS, 0.1% Triton X-100 for 5 min; and treated 5 min with 3% hydrogen peroxide (H₂O₂) to neutralize endogenous peroxidase activity. Conditions for primary antibody incubations were determined empirically. Affinity-purified ps20ab-1 was incubated with tissue sections at a 1:4 dilution overnight at room temperature in a humid chamber. Negative controls included affinity-purified ps20ab-1 preabsorbed with peptide, preimmune sera (1:18.9), and secondary antibody alone. Mouse monoclonal antibody against the smooth muscle-specific isoform of -actin (clone asm-1, Boehringer Mannheim) was incubated at a 1:4 dilution for 1 h at room temperature. Immunoreactivity was visualized by a 45-min incubation with either biotinylated goat anti-rabbit or goat anti-mouse secondary antibodies (Sigma), diluted 1:15; followed by a 30-min incubation with ExtrAvidin-conjugated peroxidase (Sigma), diluted 1:15; and concluding with a 7-min incubation with diaminobenzadine and mounting on a glass slide with Gel Mount.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Cloning of ps20 cDNA-- The amino-terminal amino acid sequence of ps20 was previously determined and reported (27) (see "Experimental Procedures"). Degenerate ps20 primers, P34 (corresponding to amino acids 1-9) and P38 (corresponding to amino acids 19-27) (primer sequences given under "Experimental Procedures"), were designed from this amino acid sequence and used in a nested PCR strategy in combination with vector primers to amplify a fragment of ps20 cDNA from the PS-1_BfsZAP cDNA library, a cDNA library developed from the PS-1 adult rat prostate smooth muscle cell line (30). Degenerate ps20 primer P38 was used in a primary PCR amplification with vector primer T7 to amplify a product from PS-1 cDNA library phage lysate. This primary amplification reaction was used as template for a secondary PCR amplification with primer P38 and degenerate ps20 primer P34 to yield a single 81-bp reaction product (Fig. 1A). To confirm that the 81-bp fragment was derived from ps20 sequence, two additional ps20 degenerate primers, P36 (amino acids 1-7) and P37 (amino acids 7-13), were generated and used independently in combination with P38 to amplify expected 81-bp and 63-bp fragments, respectively, in analogous nested PCR reactions. The 81-bp fragment produced by P34 and P38 was cloned into the pCRII T-tailed vector (clone 3438pCRII), and sequenced (Fig. 1A). Clone 3438pCRII sequence (primer sequence omitted) was: 5'-GGTCAGGCTGGCTGAGAAGTCCCAAGCTGAAGAG-3'. The deduced amino acid sequence of clone 3438pCRII matched directly with native ps20 amino acid sequence (27). The clone 3438pCRII was therefore assumed to be a fragment of the DNA sequence encoding the amino terminus of mature ps20 protein.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   ps20 nested PCR cloning strategy. A nested PCR strategy was used to isolate ps20 PCR clones from the PS-1_BfsZAP cDNA library lysate. In this diagram, PCR primers (see "Experimental Procedures" for sequences) are shown as arrows, vector sequence as bold lines, and partial ps20 PCR clones as single lines. Two ps20 degenerate primers, P34 and P38, were generated based on ps20 amino acid sequence. A, for nested PCR, primers T3 and P38 were used in a primary PCR amplification, which was sequentially used as template for a secondary amplification with primers P34 and P38. The 81-bp fragment was cloned into vector pCRII (3438pCRII) and sequenced. The deduced amino acid sequence corresponded completely with purified ps20 amino acid sequence. From 3438pCRII sequence, ps20 primers P40 and P42 were designed. B, primers T3 and P40 were used to clone the ps20 5' end (184-bp clone T340pCRII) in a secondary nested reaction following initial amplification with primers T3 and P38. C, primers P42 and T7 were used to clone the ps20 3' end (868-bp clone 42T7pCRII) following a primary PCR reaction with P34 and T7. D, primers P45 and P46 were used to screen plates for the presence of ps20 clones prior to the isolation of full-length ps20 cDNA by plaque hybridization with 32P-labeled 42T7pCRII insert as probe.

View larger version (53K): [in this window] [in a new window]   Fig. 2.   Nucleotide sequence and deduced amino acid sequence of rat ps20 full-length cDNA. The 1029-bp ps20 cDNA contains a start codon at nt 52-54, a stop codon at nt 688-690 (asterisk), and a polyadenylation signal at nt 1000-1002 (thick underline). This open reading frame predicts expression of a 212-amino acid protein with a predicted signal peptide cleavage site immediately following Gly26 (arrow). The deduced protein sequence is numbered beginning with the first amino acid of the signal peptide. Thr27-His54 (underlined) directly correspond in sequence with Thr1-His28 of native ps20. The sequence of peptide-(1-14) used to generate ps20ab-1 is underlined (dotted line). The WAP-type four-disulfide core domain is shaded, and the eight cysteines composing the domain are underlined. Potential protein modifications are indicated: casein kinase II phosphorylation sites at Ser40, Ser77, Ser137, Thr138, and Ser150 (dot), and an O-linked glycosylation site at Thr48 (dagger).

View larger version (68K): [in this window] [in a new window]   Fig. 3.   Northern analysis of ps20 mRNA expression in U4F cells and in adult rat tissues. Total RNA (15.0 µg/lane) isolated from the U4F cell line and from adult rat tissues was analyzed by Northern analysis with 32P-labeled clone 42T7pCRII insert (868 bp) as probe. A single mRNA species of 1.1 kb was recognized in both U4F cells and in several rat prostate tissues. The highest expression was observed in the heart with detectable message observed in dorsal-lateral prostate, vas deferens, testis, uterus, and ovary. The size of RNA markers is indicated at left. The gel was stained with ethidium bromide to visualize ribosomal RNA as a loading control prior to transfer (lower panel).

View larger version (31K): [in this window] [in a new window]   Fig. 4.   Amino acid sequence comparison of the WAP-type four-disulfide core domains of ps20 and other family members. The ps20 WAP-type four-disulfide core domain amino acid sequence was aligned with the WAP domains of other family members. Conserved residues are in reversed contrast, and conservative substitutions are shaded. Within the entire WAP-type four-disulfide core domain, ps20 is 39.4% identical to chelonianin (IBP), 35.4% identical to antileukoproteinase 1 (ALK1), 35.3% identical to whey acidic protein (WAP), 33.3% identical to WDNM1, 33.3% identical to HE4, 31.2% identical to Kallmann syndrome protein (KALM), 29.2% identical to elafin (ELAF), and 27.1% identical to caltrin-like protein II (CALU). The multiple sequence alignment was assembled using the PILEUP program (Genetics Computing Group, Madison, WI), spaces introduced to optimize alignment, and presented using BOXSHADE (ISREC). Alignment scores were calculated by pairwise alignments using ALIGN (EERIE) with default settings.

Analysis of Endogenous ps20 Secreted by U4F and PS-1 Cells-- Prior to expression of recombinant ps20 protein, CM from U4F cells (the cell line from which ps20 was identified) (25-27) was analyzed by Western analysis using ps20ab-1. CM from U4F cells was collected, partially purified (as described under "Experimental Procedures"), and analyzed by Western analysis under reducing conditions with ps20ab-1. Two specific bands of apparent molecular mass 21 and 27 kDa were specifically recognized by ps20ab-1 (Fig. 5A, lane 1), relative to preabsorbed control (Fig. 5A, lane 2).

View larger version (45K): [in this window] [in a new window]   Fig. 5.   Analysis of native and recombinant ps20 protein by SDS-PAGE. A, concentrated U4F CM was analyzed by Western analysis under reducing conditions and probed with ps20ab-1 (lane 1) and ps20ab-1 preabsorbed with ps20 peptide-(1-14) (lane 2). ps20 was detected as both 21- and 27-kDa bands. B, ps20 was immunoprecipitated from PS-1 CM with ps20ab-1 (lane 1) or preimmune sera (lane 2) or from U4F CM with ps20ab-1 (lane 3) or preimmune sera (lane 4) and resolved by 12% SDS-PAGE gels under reducing conditions. Identical samples were resolved under non-reducing conditions: lane 5, PS-1 CM immunoprecipitated with ps20ab-1; lane 6, PS-1 CM immunoprecipitated with preimmune sera; lane 7, U4F CM immunoprecipitated with ps20ab-1; lane 8, U4F immunopreciptated with preimmune sera. From both PS-1 and U4F cell CM, 27- and 21-kDa bands are specifically detected under reducing conditions when immunoprecipitated with ps20ab-1 (lanes 1 and 3) and a 21-kDa protein detected under non-reducing conditions (lanes 5 and 7). C, ps20 was expressed as a His-tagged fusion protein, rps20-His, in the BL21(DE3)pLysS bacterial cell line. Crude extracts of cells prior to induction (lane 1) and following induction (lane 2) were resolved by reducing SDS-PAGE and visualized by Coomassie Blue stain. Western analysis with ps20ab-1 (lane 3) demonstrated an intact amino terminus, and Western analysis with anti-His (COOH-terminal) antibody (lane 4) confirmed the protein was full-length. Soluble protein extracts were purified by nickel chromatography, resolved by SDS-PAGE analysis under reducing (lane 5) and non-reducing (lane 6) conditions, and visualized by silver staining. The 29-kDa rps20-His protein observed under reducing conditions (lane 5) migrated with an apparent mass of approximately 23 kDa under non-reducing conditions (lane 6). The apparent migration rate (molecular mass) of protein markers (in kDa) is indicated at left.

Expression of Recombinant ps20 Protein-- To further clarify the molecular forms of ps20, the ps20 open reading frame without signal peptide was subcloned into a bacterial expression vector and expressed as a carboxyl-terminal histidine (His)-tagged fusion protein (rps20-His) in bacteria. A protein of 29 kDa was observed by SDS-PAGE under reducing conditions in induced cells (Fig. 5C, lane 2). Western analysis with ps20ab-1 (ps20, amino-terminal) and anti-His antibody (His tag, carboxyl-terminal), demonstrated specific immunoreactivity of both antibodies with the 29-kDa protein (Fig. 5C, lanes 3 and 4, respectively), confirming intact amino and carboxyl termini. rps20-His protein was purified by nickel affinity chromatography, resolved by 15% SDS-PAGE under both reducing and non-reducing conditions, and visualized by silver staining. Consistent with immunoprecipitation of endogenous protein (Fig. 5A), rps20-His migrated at an apparent molecular mass of 29 kDa under reducing conditions (Fig. 5C, lane 5) and at 22-24 kDa (Fig. 5C, lane 6) under non-reducing conditions. These data confirm the apparent molecular mass of ps20 resolved by SDS-PAGE is sensitive to reducing agents. In addition, these data suggest that the bacterial and mammalian expressed forms of ps20 are equivalent in their structural properties as resolved by SDS-PAGE. Given the presence of the WAP four-disulfide core domain, it is not surprising that ps20 in its folded, non-reduced (21-23 kDa) form exhibits a more rapid migration in SDS-PAGE as compared with the reduced (27-29 kDa) form.

Biological Activity of Recombinant ps20 Protein-- As reported previously for ps20/UGIF-treated cells (25-27), growth inhibition of the ps20 transfectant cell lines was measured using [³H]thymidine incorporation assays. Four ps20 stable transfectant COS-7 cell lines (rps20-1, rps20-3, rps20-4, and rps20-6pBKCMV COS-7), one mock-transfected cell line (pBKCMV-4 COS-7), and the parent cell line (COS-7) were seeded at identical densities, and the rate of [³H]thymidine incorporation by each cell line measured daily over a 4-day time course (Fig. 6A). The rate of [³H]thymidine incorporation by the ps20 transfectant lines was significantly reduced (p < 0.01) relative to the mock transfectant line and the COS-7 parent line at each time point assayed. Consistent with previous data indicating that ps20-induced inhibition of [³H]thymidine incorporation correlated directly with cell number (27), direct cell counting of these cultures indicated 14-45% fewer cells in cultures of the ps20 transfectant cell lines relative to the mock transfectant line when counted 24-96 h after seeding (data not shown). To demonstrate that the growth inhibition of the ps20 transfectant COS-7 cell lines was due to an activity secreted into the growth medium, mock transfectant cells were incubated with CM from ps20 transfectant or mock cell lines and assayed for [³H]thymidine incorporation. Incorporation of [³H]thymidine in cultures incubated with CM from ps20 transfectant cell lines was reduced (p < 0.01) relative to cultures incubated with CM from mock transfectant cells (Fig. 6B).

View larger version (73K): [in this window] [in a new window]   Fig. 6.   Inhibition of [3H]thymidine uptake by ps20 stable transfectant COS-7 cell lines. A, the stable transfectant COS-7 cell lines rps20-1, rps20-3, rps20-4, and rps20-6pBKCMV COS-7; the mock stable transfectant cell line pBKCMV-4; and COS-7 cells were seeded in 96-well plates at identical densities and assayed for [3H]thymidine incorporation. Incorporated [3H]thymidine (cpm) was plotted at 24-h intervals. B, pBKCMV-4 cells were seeded in 96 well plates, incubated with increasing amounts of CM from ps20 transfectant cell lines and the mock transfectant cell line, and assayed for [3H]thymidine incorporation. Each experiment was repeated three times. At all time points, samples were assayed in triplicate ± S.E. (error bars) and significance determined by Student's t test (p < 0.01 for all samples at all time points shown). C, COS-7, pBKCMV-4, and rps20-4 were exposed to crystal violet stain/fixative and photographed.

View larger version (25K): [in this window] [in a new window]   Fig. 7.   Inhibition of COS-7 and PC-3 cell [3H]thymidine uptake by bacterial produced recombinant ps20. A, GST (lane 1), GST-rps20 (lane 2), and rps20 (lane 3) were resolved by reducing SDS-PAGE and analyzed by Western analysis with ps20ab-1. GST-rps20 was expressed as a 50-kDa immunoreactive protein and rps20 as two migratory forms of 23 and 29 kDa. Molecular masses of standard protein markers are shown at left. B, COS-7 cells were seeded at 2.5 × 104 cells/ml in 96 well plates and incubated with: thrombin-treated GST (1, labeled GST), GST-rps20 (2), or thrombin-cleaved GST-rps20 (3, labeled rps20) for 24 h. PC-3 cells were seeded at 5 × 104 cells/ml and incubated with GST (4), GST-rps20 (5), or rps20 (6) for 24 h and assayed for [3H]thymidine incorporation (counts/min). GST-rps20 inhibited PC-3 and COS-7 cells by 9.6% and 15%, and rps20 inhibited by 47% and 61%, respectively. The 70-kDa mildly cross-reactive band is apparently a bacterial protein previously reported to co-purify with GST fusion proteins (56). Each experiment was repeated three times. At all time points, samples were assayed in triplicate ± S.E. (error bars).

Localization of ps20 Protein in Vivo-- Immunohistochemical analysis with affinity-purified ps20ab-1 and adult rat tissues was performed to assess ps20 protein expression patterns in specific cell types. In adult rat prostate, specific immunoreactivity was apparent only in the periacinar smooth muscle cells of the stroma (Fig. 8c), as determined by comparison of a serial section probed with a monoclonal antibody recognizing the smooth muscle-specific isoform of -actin (SM -actin), a smooth muscle cell marker (Fig. 8d). The smooth muscle staining pattern was eliminated by preabsorption of ps20ab-1 with ps20 peptide-(1-14) (Fig. 8b). No immunoreactivity was observed in the absence of primary antibody, and only nonspecific background reactivity was observed with preimmune sera (data not shown). No immunoreactivity was observed in the stromal fibroblast population. A weak immunoreactivity, noted in some epithelial cells, was eliminated by preabsorption with ps20 peptide. ps20 was detected in smooth muscle of both normal and castrate animals, with an apparent moderate increase of immunoreactivity noted in the smooth muscle of castrate animals (data not shown).

View larger version (117K): [in this window] [in a new window]   Fig. 8.   ps20 immunolocalization in rat prostate smooth muscle and vascular smooth muscle. Affinity-purified ps20ab-1 was used to immunolocalize ps20 protein expression in adult rat prostate tissue: a, hematoxylin and eosin counterstain; b, affinity-purified ps20ab-1 preabsorbed with ps20 peptide-(1-14) (1:4); c, affinity-purified ps20ab-1 (1:4); and d, SM -actin (1:4) (magnification, ×520). Immunoreactivity with ps20ab-1 (c) corresponded with SM -actin immunolocalization in smooth muscle (d) adjacent to acinar epithelial cells. Affinity-purified ps20ab-1 immunolocalized ps20 protein expression in vascular smooth muscle: e, hematoxylin and eosin counterstain of a typical artery (labeled A) and a typical vein (labeled V) embedded in parathyroid tissue; f, affinity-purified ps20ab-1 (1:4) in parathyroid tissue (magnification, ×235); g, hematoxylin and eosin counterstain of aorta; and h, affinity-purified ps20ab-1 (1:4) in aorta (magnification, ×470). ps20ab-1 shows specific immunoreactivity in the tunica media of arteries and veins and in the tunica media (labeled TM) of aorta and not in the other tissue layers, including the tunica adventitia (TA) and tunica intima (TI) composed of fibroblasts and endothelial cells, respectively).

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

We report here the cloning of full-length cDNA encoding ps20, expression of biologically active recombinant ps20 protein, and characterization of ps20 mRNA and protein expression patterns in vivo. The cloning of ps20 from the PS-1 (prostate smooth muscle-like cell line) cDNA library and immunolocalization of ps20 expression in smooth muscle in vivo is consistent with the previous characterization of ps20 secretion by prostatic mesenchymal and smooth muscle cell lines in culture (27). We have demonstrated here that both bacterial and mammalian expressed recombinant ps20 protein exhibit growth-inhibitory effects on mammalian cells in culture, consistent with properties reported for UGIF/native ps20 protein (25-27). The ps20 amino acid sequence contains a whey acidic protein (WAP)-type four-disulfide core motif (29), which places ps20 in this family of small secreted proteins.

Biological activity of ps20 is consistent with the biological activity of other WAP family proteins, the majority of which have protease inhibitor activity. Protease inhibitors have been reported previously to have growth-inhibitory effects both in vitro and in vivo. Pefabloc, a synthetic serine protease inhibitor, inhibited prostatic PC-3 carcinoma cell proliferation through inhibition of an IGFBP-3 protease that regulated availability of active insulin-like growth factor-II (12). Other serine protease inhibitors, including maspin (50), plasminogen activator inhibitors (51), and tissue inhibitors of matrix metalloproteinases (52), have exhibited growth-inhibitory activities. The WAP family member elafin inhibited abnormal vascular smooth muscle proliferation in an in vivo rabbit arteriopathy model of coronary neoinitimization (12). In vitro, the mechanism for elafin-induced inhibition of pulmonary artery smooth muscle cell proliferation involves inhibition of elastase-mediated proteolytic release of mitogenic biologically active FGF-2 from extracellular matrix (12).

Members of the WAP signature motif family exhibit fundamental roles in growth control, differentiation, and tissue remodeling in development and maintenance of homeostasis in adult tissue. The Kallmann syndrome protein is a peripheral membrane protein, which is proteolytically released in soluble form and may be a neuronal axon chemoattractant (53). Mutations or deletions in the Kallmann syndrome protein lead to defective embryonic migration and differentiation of olfactory and gonadotrophin-releasing hormone-synthesizing neurons in development, resulting in the hypogonadotropic hypogonadism and anosmia of Kallmann syndrome (49). Expression of elafin is down-regulated in poorly differentiated squamous cell carcinoma (48) and in mammary carcinoma cells (54), and its loss of expression has been postulated to facilitate tumor progression in these tissues. WDNM1 was identified in a subtractive hybridization screen to identify potential tumor suppressor genes in mammary cancer (47, 55) as a transcript down-regulated in metastatic breast cancer cell lines. Although the significance of ps20 expression in vivo is not yet known, alterations in the expression of other WAP family members clearly influence growth control in vivo.

We initially identified ps20 protein as the major protein component of the mesenchymally produced UGIF activity, which has a growth-inhibitory effect on epithelial cell lines in vitro (25-27). We now report that ps20 localizes to the periacinar ring of smooth muscle in immediate contact with the secretory epithelial acini in the prostate gland. Whether or not the weak immunoreactivity observed in epithelial cells represents a low level of expression by these cells or stromally produced ps20 incorporated by these cells is not yet clear. The in vivo localization of ps20 in smooth muscle supports a putative role for ps20 in stromal-epithelial interactions in the prostate gland.

The significance of the apparent differential expression of ps20 by different types of smooth muscle cells is unknown, but is evident at both the mRNA and protein levels. Expression of ps20 message and protein was relatively low in many organs containing visceral smooth muscle, including the stomach and small and large intestines, which have thick walls of smooth muscle layers. Of interest, the staining intensity of ps20 in vascular smooth muscle, particularly in both larger elastic arteries and small muscular arteries, was consistently high and similar to or higher than in prostate smooth muscle. The elevated mRNA expression observed in heart tissue by Northern analysis is not likely due to expression by cardiac muscle cells, which were not immunoreactive, but is likely due to high expression by coronary artery and large elastic artery smooth muscle cells, which were highly immunoreactive with ps20ab-1. Vascular diseases are typified by medial hypertrophy and neointimal proliferation. The localization of ps20 in vascular smooth muscle suggests ps20 may play an autocrine role in regulation of smooth muscle proliferation and maintenance of homeostasis in the vasculature.

The specific mechanisms of ps20 biological action and functional role in vivo are not yet known. ps20 could potentially inhibit cell proliferation through alterations of growth factor bioavailability, extracellular matrix remodeling, or alterations of cell adhesion. Based on our data to date, no specific ps20 mechanism can be ruled in or ruled out. Definitive characterization of ps20 mechanisms of biological action awaits a more complete functional characterization of activity in vitro and in vivo.