HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 281, Issue 49, 38022-38037, December 8, 2006

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 281/49/38022    most recent M600353200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Notari, L. Articles by Becerra, S. P. Search for Related Content PubMed PubMed Citation Articles by Notari, L. Articles by Becerra, S. P. Social Bookmarking                      What's this?

Identification of a Lipase-linked Cell Membrane Receptor for Pigment Epithelium-derived Factor^*

Luigi Notari, Victoriano Baladron^¶, J. Daniel Aroca-Aguilar^¶1, Natalia Balko, Raul Heredia, Christina Meyer, Patricia M. Notario^||, Senthil Saravanamuthu, Maria-Luisa Nueda^¶1, Francisco Sanchez-Sanchez^¶, Julio Escribano^¶2, Jorge Laborda^¶2, and S. Patricia Becerra²³

From the National Eye Institute, National Institutes of Health, Bethesda, Maryland 20892, the Food and Drug Administration, Bethesda, Maryland 20892, ^¶School of Medicine/Centro Regional de Investigaciones Biomédicas, University of Castilla-La Mancha, Albacete 02071, Spain, and ^||Georgetown University, Washington, D. C. 20057

Received for publication, January 13, 2006 , and in revised form, October 5, 2006.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Pigment epithelium-derived factor (PEDF) is an extracellular multifunctional protein belonging to the serpin superfamily with demonstrable neurotrophic, gliastatic, neuronotrophic, antiangiogenic, and antitumorigenic properties. We have previously provided biochemical evidence for high affinity PEDF-binding sites and proteins in plasma membranes of retina, retinoblastoma, and CNS cells. This study was designed to reveal a receptor involved in the biological activities of PEDF. Using a yeast two-hybrid screening, we identified a novel gene from pigment epithelium of the human retina that codes for a PEDF-binding partner, which we term PEDF-R. The derived polypeptide has putative transmembrane, intracellular and extracellular regions, and a phospholipase domain. Recently, PEDF-R (TTS-2.2/independent phospholipase A₂ (PLA₂) and mouse desnutrin/ATGL) has been described in adipose cells as a member of the new calcium-independent PLA₂/nutrin/patatin-like phospholipase domain-containing 2 (PNPLA2) family that possesses triglyceride lipase and acylglycerol transacylase activities. Here we describe the PEDF-R gene expression in the retina and its heterologous expression by bacterial and eukaryotic systems, and we demonstrate that its protein product has specific and high binding affinity for PEDF, has a potent phospholipase A₂ activity that liberates fatty acids, and is associated with eukaryotic cell membranes. Most importantly, PEDF binding stimulates the enzymatic phospholipase A₂ activity of PEDF-R. In conclusion, we have identified a novel PEDF-R gene in the retina for a phospholipase-linked membrane protein with high affinity for PEDF, suggesting a molecular pathway by which ligand/receptor interaction on the cell surface could generate a cellular signal.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Pigment epithelium-derived factor (PEDF),⁴ a noninhibitory member of the serine protease inhibitor superfamily (SERPINS), is a multifunctional protein involved in neuronal survival and differentiation and in preventing angiogenesis and the growth of tumor cell metastases (1-6). It was discovered as a 50-kDa protein released by cultured pigment epithelial cells from fetal human retina (7). In the retina, PEDF is highly expressed in the retinal pigment epithelium (RPE) (8, 9), is enriched in the interphotoreceptor matrix (10-12), and has been implicated in photoreceptor morphogenesis (13) and in vascular and neural retinal disorders (1). It acts on photoreceptor survival (14) and in preventing the pathological invasion of neovessels (15). Interestingly, its levels are decreased in choroidal neovascularization, because of age-related macular degeneration (16), and in diabetic retinopathy (17). Its importance in the development, maintenance, and function of the retina is evident in animal models for inherited and light-induced retinal degeneration (14, 18), retinal ischemia (19), and ocular neovascularization (20, 21).

However, PEDF is not exclusive for the retina and is found in vitreous, aqueous humor, serum, cerebrospinal fluid, and in most tissues, supporting its functions in extra-retinal systems (8, 22-24). In this regard, PEDF induces neurite-outgrowth in developing motor neurons (25), inhibits apoptosis of cerebellar granule cell neurons (26), and protects hippocampal (27), cerebellar granule cell, and motor neurons against glutamate cytotoxicity (28, 29). In contrast, it has antigliastatic activity on microglia from newborn rat brain (30) and demonstrable proapoptotic effects on endothelial and tumor-derived cells, as well as in animal models for tumor growth (31-35). PEDF null mice show the importance of PEDF in vasculature (31, 35).

Although the structural determinants by which PEDF lost its ability to inhibit serine proteases are unknown, studies have demonstrated the following. 1) The PEDF protein has the three-dimensional structure of a serpin member with a homologous serpin-reactive center loop (5, 36). 2) Cleavage at this loop does not induce the typical serpin stress-to-relax conformational change of inhibitory serpins (37) and does not abolish its neurotrophic activities or binding affinities for glycosaminoglycans and collagens (26, 37-39). 3) Peptide regions toward its N terminus and distal to the homologous serpin-reactive center loop have demonstrable neurotrophic and antiangiogenic activities (25, 26, 40-44). Additional studies demonstrated that overexpression and local or systemic administrations of PEDF promote survival of photoreceptors, inner retina, and retinal ganglion cell neurons in experimental animal models for retina degeneration and ischemia and inhibit neovascularization in numerous models for ocular angiogenesis (14, 18-21, 45). It has been shown that the PEDF action is cell surface binding-mediated because its effects are blocked by antibodies that are cell surface-binding antagonists (12, 28, 40). We and others have also provided biochemical evidence for high affinity PEDF-binding sites (K_D = 2-8 nM) and proteins (80-90 kDa) in plasma membranes of retina, retinoblastoma, CNS, pericytes, and endothelial cells, consistent with a cell-surface PEDF receptor protein (12, 40, 41, 43, 44, 46). However, little is known about the identity of the receptor and molecular mechanism(s) by which PEDF functions to regulate neuronal and endothelial cell behavior.

In this study, we investigated potential binding partners for PEDF using yeast two-hybrid screening and found a novel RPE gene, which we termed PEDF-R. We examined its gene expression and protein distribution in the retina. We expressed the PEDF-R protein, demonstrated its binding affinity for PEDF, characterized its phospholipase A₂ activity, and mapped its transmembrane topology. These studies present evidence for the identification of a phospholipase on the surface of retina cells and imply a potential regulatory role for phospholipase-linked membrane protein in the biological actions of PEDF.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Yeast Two-hybrid Screening—A two-hybrid system of Saccharomyces cerevisiae CG-1945 strain (Clontech) was used as described previously by Baladron et al. (47) with the following modifications. The target was a human liver cDNA library (Human Liver MATCHMAKER cDNA Library; Clontech) with about 3 x 10⁶ independent clones of 2 kb (mean cDNA size) having their 5' ends proximal to the activation domain (AD) of the GAL4 transcription factor in a pACT2 vector. The baits were PEDF cDNAs cloned in pAS2-1 plasmid (Clontech) with the binding domain at their 5' end. The PEDF inserts were PCR-amplified fragments encoding amino acids 35-418, 35-266, 35-229, and 35-119 of the human PEDF (GenBank^TM accession number NM_002615 [GenBank] ; for primers see Fig. 1). The selection of the cotransformants was performed based on resistance to Leu^- and Trp^-. The selection of the interaction was performed based on His^- medium. The intensity of the interaction was analyzed by 3-aminotriazole (3-AT) assays in which yeast cells were cultured in minimal medium minus histidine and increasing concentrations of 3-AT (0-100 mM), a competitive inhibitor of the yeast His3p protein used to eliminate background growth. All of the two-hybrid assays were repeated at least twice. The selected clones were used to remove the bait plasmids (PEDF) and isolate the AD-containing plasmids followed by propagation in Escherichia coli DH5- cells (47). The AD-containing plasmids were candidate PEDF interactants, and their nucleotide sequences were determined using an ABI PRISM 310 sequencer (Applied Biosystems). The sequences were analyzed and compared with GenBank^TM data bases using a BLAST search engine (www.ncbi.nlm.nih.gov/BLAST/) (48). Several clones were false-positives and were discarded as positive potential PEDF interactants because of the lack of correct frame alignment or nonrelevant subcellular location (see supplemental Table S1). After verifying their open reading frames, the potential positive AD plasmids were isolated and cotransformed with shorter PEDF constructs, as described above.

RNA Isolation and cDNA Synthesis—Total RNA was extracted using RNeasy^TM (Qiagen) following the manufacturer's instructions. The RNA was treated with TURBO DNase^TM (Ambion) to remove genomic DNA contamination. An oli-go(dT) probe was used to reverse-transcribe mRNA from cell samples in a final volume of 20 µl using SuperScript first-strand synthesis system (Invitrogen) following the manufacturer's instructions.

PCR—Specific primers for screening the expression of human PEDF-R within the 12c sequence were 12-forward 5'-AAC CCC TTG CTG GCG TTG C 3' and 12-reverse 5' CCC GTC TGC TCC TTC ATC C-3'. Alternatively, specific human PEDF-R primers were In2F-forward 5'-GCA GTT TCC TGC TGA AGG TC-3' and In2R-reverse 5'-GCT CGT CCT TGG AGT TGA AG-3'. Specific rat PEDF-R primers were rIn2F-forward 5'-TGT GGC CTC ATT CCT CCT AC-3' and rIn2R-reverse 5'-TGA GAA TGG GGA CAC TGT GA-3' (GenBank^TM accession number XM_341960 [GenBank] ). Primers for 18 S were 18 S rRNA-F 5'-ATG CTC TTA GCT GAG TGG CCC G-3' and 18 S rRNA-R 5'-ATT CCT AGC TGC GGT ATC CAG G-3'. Templates were PEDF-R cDNA and cDNAs prepared from tissues and cells in PCR SuperMix reactions following instructions by manufacturer (Invitrogen). The reverse transcription-PCR was performed in a final volume of 50 µl containing 2 µl of cDNA and 10 µM primers specific for the PEDF-R gene with 18 S or GAPDH primers as a positive control using a thermocycler (GeneAmp® PCR System 9700; Applied Biosystems) programmed in the following manner: initial denaturation at 94 °C, 35 cycles of 15 s at 94 °C, 30 s at 61 °C, and 45 s at 72 °C, and 5 min at 72 °C. PCR products were resolved by agarose gel electrophoresis and visualized with ethidium bromide.

Immunohistochemistry of Rat Retina—Albino rat eyes (kind gift of Ignacio Rodriguez) were fixed in paraformaldehyde and embedded in paraffin, and 7-µm thick sections were made using a microtome. The sections were deparaffinized with xylene for 30 min and hydrated with alcohol in decreasing grade (100, 90, 70, 50, 30, and PBS) 5 min for each step. Then the nonspecific binding sites were blocked with 5% BSA in PBS for 30 min. The sections were incubated1hin anti-R^C diluted 1:1000 with PBS containing 1% BSA, 0.02% Tween 20, except for the no primary control sections, which were incubated only in diluent without antibodies. After several washes with PBS, the samples were incubated in secondary Alexa 568 anti-chicken antibody diluted 1:300 with PBS containing 1% BSA, 0.02% Tween 20 for 1 h. The slides were washed in PBS and stained with a DAPI solution (diluted 1:1,000 in 1% BSA in PBS) for 2 min. After wash with PBS, the samples were mounted with Gel/Mount mounting media (Biomeda) and observed under a laser confocal microscope.

Construction of Expression Vectors—Expression vectors for PEDF-R were constructed using ATCC IMAGE clone: 4875483 (ATCC number, 6903398; GenBank^TM accession number BC017280 [GenBank] ) as template for PCRs and recombining the amplified DNA products into expression plasmids using the Gateway^TM technology (Invitrogen). Briefly, PCR-amplified fragments encoding amino acids 1-504 and 250-383 of PEDF-R (PEDF-R, 212-1726 nucleotides, and 12c, 957-1362 nucleotides) were used to create entry clones with pENRT/D-TOPO or pENTR/SD/D-TOPO by PCR-directional TOPO cloning. The entry clones were used to generate expression clones by recombination events. Primer sequences and expression constructs are listed in Table 1. Final entry vectors had gene inserts flanked by phage sitespecific recombination attL sites and destination vectors had phage site-specific recombination attR sites. The entry clones were used in a simple LR recombination reaction to transfer the PEDF-R or 12c cDNAs in pENTR/D to the destination vectors (pEXP1-DEST or pcDNA 6.2/N-Lumio DEST) to produce pEXP1-PEDF-R, pEXP1-12c, or pLmN-PEDF-R plasmids, with N-terminal epitope-tagged genes of interest. Similarly, the PEDF-R and 12c cDNAs in pENTR/S.D./D were transferred to the destination vectors (pEXP2-DEST or pcDNA 6.2/C-Lumio-DEST) to produce pEXP2-PEDF-R, pEXP2-12c, or pLmC-PEDF-R with C-terminal epitope-tagged genes of interest. The cDNAs of interest in destination vectors were downstream from the T7 transcription promoter and had ribosomal binding sites upstream from their start codons. The nucleotide sequences of the PEDF-R cDNA inserts in the entry and expression plasmids were confirmed (Lofstrand Labs Ltd.).

Epitope-tagged PEDF-R Protein Purification—Purification of His₆-tagged PEDF-R polypeptides from cell-free protein synthesis reactions was accomplished by cation-exchange column chromatography and/or His tag affinity column chromatography using S-Sepharose Fast Flow (Amersham Biosciences) resin and the ProBond purification system with Ni-NTA resin (Invitrogen) following the manufacturer's instructions. Automated purification of His-tagged polypeptides was accomplished with a POROS HS and/or a POROS MC column attached to a BioCad 700E computerized system (See Supplemental Material for details). Purification from in vitro translation reactions yielded 10 µg of His₆-tagged PEDF-R and 0.5 mg of His₆-tagged 12c per ml reaction with 50 and 90% purity, respectively. The final samples were used immediately or stored at 4 °C. The masses of tryptic peptides of purified N-end His₆-tagged PEDF-R polypeptide were in agreement with the coding sequence in pEXP1-PEDF-R, as confirmed by Nano LC-MS/Ms (by ProTech).

Proteins—Human recombinant PEDF protein and fluorescein-conjugated PEDF were prepared as described before (41). Human recombinant maspin was a gift of Sally Twining (Medical College of Wisconsin). Ovalbumin was purchased from Worthington. Synthetic peptides designed from the coding sequence of PEDF-R, R^A (amino acids Cys¹⁹⁴-Asn²⁰⁹) and R^C (amino acids Glu⁷¹-Arg⁷⁹), were generated and used to develop rabbit anti-R^A antibodies (AnaSpec) and chicken anti-R^C (Aves Labs Inc.).

Protein Analyses—Protein concentration determination was performed using protein assay (Bio-Rad). SDS-PAGE was performed using gradient 10-20% polyacrylamide gels in Tricine/SDS (Invitrogen). Western blots were as described before (12). The apparent molecular weight was estimated after SDS-PAGE and relative to prestained marker molecular weights (molecular weights given by manufacturers varied from lot to lot). The relative mobilities (R_f values) for the markers were determined from several gels and plotted on semi-log graph paper. The apparent molecular weight of PEDF-R proteins were determined from the curve in linear range, and each was the average of more than three determinations.

His Tag Pulldown Assays—Binding of PEDF to His-tagged PEDF-R polypeptides was assayed by His tag pulldown of bound complexes with Ni-NTA resin. For PEDF-R/PEDF interactions, soluble aliquots of in vitro translation RTS 100 reactions containing His₆-tagged PEDF-R protein (0.7 µg) were mixed with PEDF protein (4 µg) in binding buffer (50 mM sodium phosphate, pH 7.5, 500 mM NaCl, 1% Nonidet P-40, 104 µl final volume) and incubated at 4 °C with gentle rotation for 2 h. The Ni-NTA resin beads (50 µl) pre-equilibrated in binding buffer containing only 0.1% Nonidet P-40 were added, and the suspension was incubated for 1 h at 4°C with gentle rotation. For 12c/PEDF binding, Ni-NTA resin (50 µl) was mixed with purified N-end His₆-tagged 12c (0.8 µg, fraction 33 of BioCad purification) in 200 µl of binding buffer (50 mM sodium phosphate, pH 8.0, 0.5 M NaCl). Mixtures were incubated and rotated for 30 min at 4 °C and washed three times with binding buffer. PEDF (3 µg) in binding buffer was added to the resin suspensions and rotated for 3 h at 4°C. The resin beads were sedimented by centrifugation and washed three times with binding buffer, and the proteins were extracted with SDS-PAGE sample buffer (50 µl) and analyzed by Western blotting against anti-PEDF antibodies.

Complex Formation Assay—PEDF·PEDF-R complex formation was assayed by size-exclusion ultrafiltration using Centricon-100 devices (Millipore), as described (38). For PEDF-R·PEDF binding, soluble aliquots of epitope-tagged PEDF-R in RTS 100 reactions and PEDF were mixed as above for His tag pulldown assays. After incubation, mixtures were diluted in binding buffer containing 0.02% Nonidet P-40 (0.05% Nonidet P-40 final concentration), transferred to Centricon-100 devices, and centrifuged at 1000 x g at 4 °C for 40 min or until most solution went through the membrane. The concentrated material was washed with the same buffer two more times and analyzed by Western blotting against anti-PEDF antibodies.

Solid-phase Assay—This assay was modified from that described for ELISA Femto^TM (Pierce). PEDF was immobilized on 96-well plates by adding to each well 0.5 µg of PEDF diluted in 0.6 M sodium citrate, 0.1 M sodium carbonate, pH 9.0 (BupH Citrate-Carbonate Buffer Pack; Pierce), and incubated for 1 h at room temperature. After removal of excess solution, wells were washed with PBS. Blocking solution (1% BSA in PBS, 0.05% Tween 20 (PBST)) was added to block nonspecific sites for 1 h at room temperature. Soluble purified His-tagged 12c was added to the wells and incubated at 4 °C for 16 h. After washing with blocking solution, anti-HisG-HRP antibody (Invitrogen) diluted at 1:1000 or 1:10,000 was added and incubated for 1 h at room temperature. The wells were washed three times with PBST, and ELISA Femto working solution (100 µl) was added to each well and incubated for 1 min. Then the luminescence was recorded by a luminometer (Wallac, model 1450 Microbeta TRILUX). Negative controls included BSA immobilized on the plates instead of PEDF, no His-tagged 12c, or no antibody added to the reactions. Quantification and plotting were performed using Microsoft Excel software.

Surface Plasmon Resonance (SPR) Assays—The interactions between PEDF and PEDF-R polypeptides were analyzed by SPR using a BIAcore 3000 instrument (BIAcore) with immobilized PEDF, ovalbumin, or maspin ligands, as described previously (39). Each ligand (4 ng) was immobilized on a CM5 sensor chip, by N-hydroxysuccinimide/N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide activation, followed by covalent amine coupling of the protein to the surface. A reference surface without protein was prepared by the same procedure. All surfaces were then washed with 0.5 M NaCl and then re-equilibrated with binding buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, HBS-N (BIAcore) plus 0.1% Nonidet P-40). Nine different dilutions of PEDF-R solutions with concentrations ranging between 0 and 50 nM were injected on the surfaces. Each injection was followed by a 50 mM NaOH regeneration step. The results were analyzed using BIAevaluation software. The data were then fitted to several binding models for a kinetic analysis. The best fittings were obtained with a simple 1:1 Langmuir model for the PEDF surface binding assay.

PLA Activity Assay—The PLA activity was spectrophotometrically determined as described previously (49) with minor modifications. This assay uses [1,2-dilinoleoyl]-phosphatidylcholine as phospholipase substrate and lipoxygenase as coupling enzyme. The phospholipase activity releases linoleic acid from the substrate, and the lipoxygenase oxidizes the released linoleic acid to form a derivative hydroperoxide. The PLA activity was followed spectrophotometrically by measuring the increase in absorbance at 234 nm as a result of the formation of the linoleic acid hydroperoxide. Freshly prepared PEDF-R was incubated in 50 mM Tris-Cl buffer, pH 7.5, containing 3 mM deoxycholate in the presence of 0.26 mM phospholipase substrate (Sigma) and 12,226 units/ml of lipoxygenase (EC 1.13.11.12 [EC] ; Sigma) in a final volume of 1 ml. Buffers for pH curve contained 3 mM sodium deoxycholate in 50 mM sodium phosphate, pH 6.4-7.4, 50 mM Tris-Cl, pH 7.0-8.2, and 50 mM sodium borate, pH 8.0-9.0. Spectrophotometric measurements were performed using a Beckman DU 640 spectrophotometer (Beckman Coulter) A spectrophotometric scan of the products formed every minute for 10 min was obtained, and the rate of absorbance change at 234 nm per min was determined. The PLA activity, expressed as the rate of product formed, A₂₃₄/min, was obtained using software from the spectrophotometer, and rates were plotted using Microsoft Excel software.

Cell Cultures—ARPE-19 cells were cultured in medium recommended by the ATCC (1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F-12 medium containing 1.2 g/liter sodium bicarbonate, 2.5 mM L-glutamine, 15 mM HEPES, and 0.5 mM sodium pyruvate, 90%; fetal bovine serum, 10%) with 100 units/ml penicillin and 100 µg/ml streptomycin. R28 cells were a gift of Dr. Gail Seigel (University of Buffalo). RGC-5 cells were a gift of Dr. Neeraj Agarwal (University of North Texas Health Science Center). COS-7, RGC-5, and R28 cells were cultured in Dulbecco's modified Eagle's medium, 10% fetal bovine serum. 3T3-L1 cells were cultured in Dulbecco's modified Eagle's medium with 4 mM L-glutamine adjusted to contain 1.5 g/liter sodium bicarbonate, 4.5 g/liter glucose, 10% bovine calf serum, and 1% penicillin/streptomycin.

Eukaryotic Expression of PEDF-R Polypeptides—Mammalian expression DNA vectors for PEDF-R were propagated in DH5 competent bacterial cells. DNA plasmids were purified from the bacterial cells using HiSpeed plasmid purification midi kit (Qiagen), following the manufacturer's instructions, and concentrated by ethanol precipitation. Purified DNA plasmids (24 µg) for transfection were mixed with 60 µl of Lipofectamine 2000 (Invitrogen) diluted in Opti-MEM reduced serum media (Invitrogen) and allowed to form complexes for 20 min at 25 °C. DNA-Lipofectamine complexes were mixed with culture medium without antibiotics and added to mammalian cells (COS-7, ARPE-19, RGC-5, R28, or 3T3-L1) at 90% confluency in 10-cm plates in media without antibiotics. Cells were harvested or labeled with Lumio reagent 24-48 h after transfection. For stable transfection, cells containing plasmids were selected with culturing medium containing 3 µg/ml blasticidin. Preliminary "kill curves" with each cell type indicated successful selection of stably transfected cells with 1-10 µg/ml blasticidin for at least 3 days (data not shown).

Cell Fractionation—Cells were harvested and subjected to differential centrifugation as we described previously (41). Separation of final supernatant (cytosolic fraction) and particulate material (membrane fraction) for transfected cells was with 150,000 x g centrifugation and for ARPE-19 cells was with 80,000 x g centrifugation (see details in Supplemental Material). For Lumio in-gel visualization, fractionated cell samples were prepared for SDS-PAGE using the Lumio Green detection kit (Invitrogen) following the manufacturer's instructions. Lumio Green reagent was added to each sample, followed by addition of Lumio Enhancer. Lumio-labeled proteins in the gels were visualized using a Typhoon 9410 laser-based scanner (Amersham Biosciences) with a Green 532 nm laser and a 555 nm bandpass filter. Proteins in the gel were subjected to Western blotting, staining with Ponceau Red, and immunoreactions with specific antibodies.

In-cell detection of Lumio-tagged polypeptides was accomplished with Lumio Green reagent (Invitrogen) solution (2.5 µM) prepared in Opti-MEM. Culture medium was removed from the cells; Lumio solution was added, and the cells were incubated for 30 min in dim light. The Lumio solution was then removed and replaced with a 20 µM solution of Disperse Blue 3 in Opti-MEM. The labeled cells were visualized using fluorescence or confocal microscopy.

Immunocytochemistry—Cells were plated in 12-well plates at 30% of confluence and incubated overnight at 37 °C with 5% CO₂. Cell permeabilization was accomplished with incubations with 100% ethanol for 5 min, whereas other wells of cells (nonpermeabilized) remained in PBS. Cells were fixed with 4% paraformaldehyde in PBS for 30 min and washed with PBS. Blocking of nonspecific primary antibody binding was done with 5% BSA/PBS incubations for 30 min at room temperature. Primary antibodies (rabbit anti-R^A diluted 1:25 and chicken anti-R^C diluted 1:1000) in 1% BSA/PBS for the not permeabilized cells and in 1% BSA/PBS containing 0.02% Tween 20 for the permeabilized cells were incubated at room temperature for 1 h. Secondary fluorescence-conjugated antibodies (Alexa Fluor® 568 goat anti-rabbit and Alexa Fluor® goat anti-chicken; Molecular Probes) diluted 1:300 in 1% BSA/PBS were incubated at room temperature for 1 h. Incubations with DAPI (Molecular Probes) diluted 1:1000 in PBS were for 2 min. Control cells did not receive primary antibodies. The cells were observed under a Nikon fluorescence microscope (Inverted Microscope Eclipse TE2000-U) using a TRITC filter with maximum emission wavelength of 630 nm, a maximum excitation wavelength of 540 nm, and a UV filter for DAPI staining. Images from the microscope were collected with Nikon ACT-1 version 2.63 software (time of exposure 1/20 s for DIA images and 1/4 s for fluorescence captures). Comparisons were made of microscope images obtained with identical magnification, gains, and exposures. Then microscope images were imported into Photoshop where they were scaled, adjusted, and composed. All Photoshop images to be compared were adjusted with identical variables.

Fl-PEDF Binding to Cells—ARPE-19 and 3T3L1 cells were plated in 6- or 24-wells plates, respectively, at 30% of confluence and left overnight in an incubator at 37 °C with 5% CO₂. The cells were washed with Opti-MEM® reduced serum medium (catalog number 31985-070; Invitrogen) and incubated in Opti-MEM® with increasing concentrations ranging from 5 to 50 nM of Fl-PEDF in 300 µl or 1 ml for 24- or 6-well plates, respectively, for 10 min in the dark and at room temperature. The medium was removed, and the cells were washed twice with Opti-MEM®. The cells were incubated in 600 µl or 2 ml, respectively, of Disperse Blue 3 in Opti-MEM® (final concentration 20 µM) and observed under a Nikon fluorescence microscope using a fluorescein isothiocyanate filter with maximum wavelengths of emission and excitation of 520 and 494 nm, respectively. For ligand competition, the media were removed from the cultures (12-well plates), and the cells were washed once with Opti-MEM® and preincubated for 10 min with 300 µl of medium containing 0.2, 0.4, or 1 µM of PEDF. The cells washed once with the reduced serum medium and incubated in 300 µl of 20 nM Fl-PEDF plus competitor in Opti-MEM® for 10 min. The cells were washed twice with Opti-MEM®, incubated in 600 µl of Disperse Blue 3/Opti-MEM® (final concentration of 20 µM), and observed under a fluorescence microscope. Images were collected with Nikon ACT-1 version 2.63 software (time of exposure 1/20 s for DIA images and 1/4 s for fluorescence captures) and imported into Photoshop for scaling and adjustment.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Selection of PEDF-R by Yeast Two-hybrid Screening—A yeast two-hybrid screen performed with human PEDF cDNA fragments as baits and a human liver library as target revealed about 50 clones expressing proteins potentially interacting with GAL4BD-PEDF fusion proteins. Several of these clones had prospective positive coding sequences that matched known genes in GenBank^TM data bases. A clone termed 12c was selected among them for two reasons. First, it had a sequence that matched genes of potential interest. Second, colonies coexpressing it and pPEDF-1 tolerated 50 mM 3-AT, demonstrating strong interactions between PEDF and the protein encoded by the cDNA of clone 12c. New yeast cotransformations of clone 12c and four individual truncated PEDF constructs (Fig. 1) were plated with increasing 3-AT concentrations in selective media to confirm the PEDF/12c interactions. Only the colonies coexpressing clone 12c and an N-terminal half-region of PEDF (35-229 amino acids positions, pPEDF-3) tolerated 3-AT above 10 mM, suggesting that the N-terminal portion of PEDF is implicated in the interactions with the polypeptide encoded by the 12c cDNA.

View larger version (40K): [in this window] [in a new window]   FIGURE 1. PEDF cDNA constructs used as baits in the yeast two-hybrid system. A, PCR-amplified fragments encoding human PEDF amino acids (as indicated on top of each line) were cloned into pAS2-1 vector using EcoRI and SalI restriction sites. Fragments were obtained with primers hPEDF219U33 + hPEDF1373L36 for pPEDF-1, hPEDF219U33 + hPEDF914L35 for pPEDF-2, hPEDF219U33 + hPEDF803L39 for pPEDF-3, and hPEDF219U33 + hPEDF474L39 for pPEDF-4 (see B for primer sequences). Final constructs had GAL4-binding domain-PEDF fusion cDNAs. B, primers used to prepare PCR-amplified fragments encoding human PEDF. The italicized portion of the primer sequence indicates the restriction site present in a PCR product prepared with that primer. The underlined portion of the sequence anneals to the PEDF-R gene. The TTA in boldface of the primer sequence corresponds to the stop codon present in a PCR product prepared with that sequence. C, 3-AT assay for the interaction of PEDF with the truncated PEDF-R (clone 12c). CG1945 yeast colonies coexpressing the truncated human PEDF-R and one of the four different PEDF truncated proteins, encoded by plasmids pPEDF-1, -2, -3, or -4, were plated in selective media plates lacking leucine, tryptophan, and histidine with increasing concentrations of 3-AT (0-100 mM). Labels correspond as follows: a, pPEDF-3; b, pPEDF-4; c, pPEDF-1; d, pPEDF-2. Labels for controls for protein-protein interactions were: e, pTD1-1 + pVA3-1 (positive control); f, pTD1-1 + pLAM (negative control); and g, pACT2 + pAS2-1 (empty vectors).

Data Base Searches and Sequence Analysis of PEDF-R—The availability of multiple genome, protein, and EST data bases facilitated the identification of several factors that may be crucial in elucidating the role of PEDF-R in humans. ENSEMBL (50) searches indicated that in humans PEDF-R maps to locus 11p15.5 on chromosome 11 and predicted it has 10 exons and 9 introns. Inspection of the PEDF-R cDNA sequence revealed mRNA transcripts of 2122 bases long with a coding capacity for a polypeptide of 504 amino acids (an expected molecular mass of 55,315.11 daltons) and four N-glycosylation consensus sites (Fig. 2A). Hydrophobicity plots of the derived amino acid sequence anticipated up to four transmembrane domains (supplemental Fig. S1D). TMpred and TOPpred programs predicted transmembrane positions with preferred orientations for the protein (Fig. 2, B and C). TMAP Prediction, TMHMM, and Split Server predicted more than one position, whereas PSORT II was the only program that predicted a cytosolic protein.

In addition to the normal retinal pigment epithelium, the PEDF-R or derived fragments were found in nucleic acid sequences isolated from various organs and specific areas of interest and adipose tissue, such as optic nerve, choroid, foveal retina, macular retina, and retinoblastomas as well as in cancerous tumors of different origins, indicating a ubiquitous character for PEDF-R. A BLASTn search of the GenBank^TM data base revealed that the PEDF-R sequence (accession number BC017280 [GenBank] .1) was most closely related to comparable sequences in other species, including Mus musculus (AK031609 [GenBank] ), Rattus norvegicus (XP_341961 [GenBank] .1), Bos taurus (XM_873300), Canis familiaris (XM_849071 [GenBank] ), and Gallus gallus (XM_428618 [GenBank] ). Further searches revealed that these sequences are putative orthologues of the human PEDF-R sequence and that a series of organisms, including Caenorhabditis elegans (accession number NM_171168 [GenBank] ) and Anopheles gambiae (accession number XM_320608 [GenBank] ) also contain homologous similar sequences. BLASTp searches revealed that the relatedness is based upon an identified patatin-like domain, found through Pfam^TM searches, to be from residues 10-179 on the human PEDF-R polypeptide. Most importantly, this domain contains structural homologies to lipases GXSXG and DX(G/A) (Fig. 2D) and was found conserved along a series of human sequences, particularly those within the newly described Patatin-like phospholipase domain containing 2 (PNPLA2)/calcium-independent PLA₂ (iPLA₂)/nutrin family that includes PEDF-R, adiponutrin, and GS2. Recently, it was reported that the human TTS-2.2 (also known as iPLA₂) and mouse desnutrin/ATGL described in adipose cells are homologue members of this new family of proteins (51-53). Human TTS-2.2/iPLA (53) is identical to PEDF-R, whereas desnutrin (52) and ATGL (51) were discovered in the mouse. Among human, rat, and mouse species, PEDF-R is highly conserved (87% identity for both human/mouse and for human/rat, and 96% for mouse/rat), but the identity decreases significantly in the segment ²⁶⁷ARPH... EDH²⁹⁵ within the 12c region (supplemental Fig. S1B). Comparisons with the other family members showed that in human PEDF-R and adiponutrin sequences had 45.62% homology, and in mouse 45.12% homology (supplemental Fig. S1C). The homology between these two sequences decreased toward the 12c region (²⁵⁰QRNG... RAKR³⁸³) in all species. The GS2 sequence, with only 253 amino acids, aligned with the N-terminal half of PEDF-R and lacked the 12c region.

View larger version (28K): [in this window] [in a new window]   FIGURE 2. Organization of the human PEDF-R cDNA. A, superimposition of the predicted TM domains and post-translational modifications on the polypeptide with the cDNA coding sequence for PEDF-R is shown. The open reading frame is indicated by an open box, the predicted transmembrane (TM) domains by gray boxes, and N-glycosylation (N-glyco) sites by ticks at the top. The hatched box illustrates the position of the 12c region. B, hydrophobicity plot of the derived amino acid sequence of PEDF-R. Certainty analyses were performed using TOPpred. The upper cut-off line is indicated (Certain line), and values equal to or above this point predict certainty for four membrane-spanning segments in PEDF-R, indicated at the top of the figure. The lower cut-off line (Putative dashed line) is also indicated. C, diagram showing a model for PEDF-R topology. A membrane is represented by horizontal parallel lines with orientation of extracellular and intracellular environments as indicated. Polypeptide is represented by a black line. TM positions with preferred orientations were predicted using TMpred program using the PEDF-R amino acid sequence and are illustrated by thicker black lines. The C termini, amino acid positions, and location peptides RA (amino acids Cys194-Asn209) and RC (amino acids Glu71-Arg79) on the PEDF-R sequence are indicated. D, homologous patatin phospholipase A (PLA) active site in PEDF-R. Alignment of partial sequences around conserved residues of PEDF-R, patatin B2, and cytoplasmic PLA2. Active sites (a Ser-Asp catalytic dyad) of the catalytic domain of human cPLA2 (Ser228 and Asp549) and of patatin B2 (Ser54 and Asp192) were obtained from previous crystallographic and mutational studies of these proteins (54, 70). The homologous patatin phospholipase A active residues of human PEDF-R correspond to Ser47 and Asp166.

Immunohistochemistry of rat retina with antibodies developed against synthetic peptide R^C designed from a PEDF-R region with 100% amino acid sequence identity among human, mouse, and rat species showed distribution of PEDF-R protein in the RPE and, at lower intensity, in the inner segments of photoreceptors and in the ganglion cell layer of the neural retina (Fig. 3C). These observations were consistent with the expression pattern of PEDF-R transcripts described above.

Binding of PEDF to Bacterial Recombinant PEDF-R Polypeptides—To determine the binding affinity for the PEDF-R/PEDF interactions, we first expressed ectopically N-terminal His/Xpress-tagged PEDF-R and 12c polypeptides in an in vitro cell-free E. coli expression system. Western analysis of reaction extracts driven by plasmids encoding 12c and PEDF-R revealed that the recombinant polypeptides had apparent molecular weights of 22,000 ± 3,000 and 69,000 ± 3,000, respectively, with anti-Xpress antibodies (Fig. 4A). Centrifugation of extracts demonstrated that the recombinant 12c and PEDF-R precipitated with storage time. However, treatments with detergents or chaotropic agents revealed that both polypeptides increased their solubility with 0.1% Nonidet P-40 or 4 M urea (Fig. 4B), consistent with the presence of putative hydrophobic TM regions in their sequences. Hence, 0.1% Nonidet P-40 was included in purification and storage buffers. Purification by Ni-NTA affinity column chromatography yielded 0.5 mg of each epitope-tagged PEDF-R and 12c polypeptides per ml of expression reaction with 25-50 and 70-90% purity, respectively. Expression from the PEDF-R plasmid produced a second immunoreactive band of smaller molecular weight (about 20,000) likely because of premature termination or degradation product and was not purified by affinity column chromatography (see Fig. 4A). Proteins of about M_r 28,000 (Fig. 4A, lane 7) present in the final purified sample were likely components of cell-free extracts with affinity for Ni-NTA.⁵

View larger version (54K): [in this window] [in a new window]   FIGURE 3. Expression of PEDF-R in ocular tissues and cells. A, reverse transcription-PCR on mRNA from RPE (human retinal pigmented epithelium, fetal and adult), ARPE-19, and hTERT cells (human RPE cell lines), human adipose, human retina, R28, and RGC-5 (rat retinal precursors) and human retinoblastoma Y-79 cells with specific PEDF-R primers (12c, RIn2, and In2 primer pairs as under "Materials and Methods"). PCR products were resolved by agarose gel electrophoresis and visualized with ethidium bromide. GAPDH, 18 S, and RB1 primers were used as controls. B, quantitative real time reverse transcription-PCR of human ARPE-19, adipose and retina. Values are relative to those of 18 S transcripts. C, immunohistochemistry of rat retina with antibodies against PEDF-R. Anti-Rc antibody (red) was used for detection of PEDF-R. The nuclei, stained with DAPI, are shown in blue and the autofluorescence in green. Panel B shows PEDF-R immunoreactivity in the RPE layer; panel D in the photoreceptors and other layers of the inner retina; panels E and F are magnifications of areas of inner segments of the photoreceptors and the inner layers of the retina, respectively (see insets in panel D). Panels A and C are controls without primary antibody of RPE and inner retina sections. The abbreviations used are as follows: Cho, choroid; IS, inner segment of photoreceptors; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer.

View larger version (49K): [in this window] [in a new window]   FIGURE 4. Bacterially derived recombinant PEDF-R polypeptides. Coupled transcription/translation reactions for in vitro protein synthesis were performed in E. coli extracts from expression vectors pEXP1-12c and pEXP1-PEDF-R with cDNA fragments under the control of the T7 transcriptional promoter. Reactions were with 0.25 µg of plasmid DNA in a final volume of 25 µl and incubations for 6 h at 30°C. Aliquots of 5 µl of each reaction were acetone-precipitated, and the resulting pellet resuspended in SDS-PAGE sample buffer. Purification of recombinant proteins was by His-tagged affinity column chromatography. A, automated purification of epitope-tagged 12c polypeptides by POROS MC column chromatography attached to BioCad 700. Epitope-tagged PEDF-R polypeptides were purified by Ni-NTA-agarose column chromatography. Lanes 1-4 correspond to 12c; and lanes 5-8 correspond to PEDF-R. Lanes 1 and 5, Western of in vitro synthesis reactions immunostained with anti-Xpress. Lanes 2 and 6, SDS-PAGE of samples loaded on columns and stained with Coomassie Blue. Lanes 3 and 7, SDS-PAGE of purified samples and stained with Coomassie Blue. Lanes 4 and 8, Western of purified samples immunostained with anti-HisG-HRP and anti-Xpress, respectively. B, Westerns against anti-Xpress of soluble (Sn, top) and insoluble (pp, bottom) fractions from in vitro protein synthesis reactions with pEXP1-12c and pEXP1-PEDF-R plasmids after treatment with reagents indicated to the top. Soluble and particulate material were separated by centrifugation.

View larger version (21K): [in this window] [in a new window]   FIGURE 5. Direct binding of PEDF to epitope-tagged PEDF-R polypeptides. A, His tag pulldown and complex formation assays of PEDF binding to epitope-tagged PEDF-R. Soluble fractions of cell-free expression reactions containing His-tagged PEDF-R (700 ng) were mixed with PEDF protein (4 µg) in binding buffer. Half of the binding reaction mixtures were assayed by His tag pulldown (Pull-down)(lanes 1-3) and the other half by complex formation after size-exclusion ultrafiltration through Centricon 100 (C-100)(lanes 6-8). Westerns versus anti-PEDF are shown. Lanes 4, 5, and 9 correspond to PEDF protein standards that were not subjected to binding assays. B, pulldown assays of PEDF binding to purified His-tagged 12c. His-tagged 12c (800 ng) was bound to Ni-NTA resin beads, and then PEDF (3 µg) was added. Bound PEDF (Pull-down) was extracted with SDS-sample buffer, and half was analyzed by Western blotting versus anti-PEDF. C, solid-phase assay. Histagged 12c was added to wells coated with PEDF or BSA. Bound His-tagged 12c was detected with anti-HisG-HRP, and ELISA femto working solution followed by luminescence recordings. D and E, real time binding by SPR analyses of PEDF-R polypeptides and PEDF interactions. Sensograms were recorded with PEDF immobilized on a CM5 sensor chip and His-tagged PEDF-R (50, 12.5, 3.125, 0.78, 0.39, and 0 nM)(D) or His-tagged-12c (1000, 500, 100, 50, 20, 10, and 0 nM)(E) using a BIAcore 3000 biosensor and BIAevaluation software. The SPR response difference with respect to the blank surface and subtracted by the 0 nM sensogram concentration during the BIAevaluation (Resp. Diff., y axis) is shown in function of time (x axis). F, comparison of PEDF-R affinities to PEDF and other serpins. Saturation binding curves of PEDF-R to immobilized maspin, ovalbumin, and PEDF are shown. Serpins were immobilized and PEDF-R was the analyte in SPR assays as in D and E. The plots were obtained using the steady state kinetic analysis from Biacore BIAevaluation software, and a superimposition of the three plots is shown. The data for the kinetic and thermodynamic parameters for the interactions were derived using the BIAevaluation kinetic analysis as follows: PEDF/PEDF-R, ka = 7.87 x 103 M-1 s-1; kd = 2.98 x 10-5 s-1; KA (ka/kd) = 2.64 x 108 M-1; KD (kd/ka) = 3.78 x 10-9 M; maspin/PEDF-R, ka = 9.1 M-1s-1 ; kd = 1.02 x 10-5s-1, KA (ka/kd) = 8.93 x 105 M-1; KD (kd/ka) = 1.12 x 10-6 M; ovalbumin/:PEDF-R, ka = 7.68 M-1s-1, kd = 1.69 x 10-4 s-1; KA (ka/kd) = 4.55 x 104 M-1; KD (kd/ka) = 2.2 x 10-5 M.

To compare the PEDF-R interactions among serpins, SPR was performed with immobilized maspin, ovalbumin, and PEDF and epitope-tagged PEDF-R as analyte. PEDF-R had significantly lower binding to maspin than to PEDF, and it did not have detectable binding to ovalbumin (Fig. 5F).

Phospholipase A₂ Activity in PEDF-R—Sequence alignment revealed a patatin phospholipase-like region with a Ser/Asp dyad in PEDF-R resembling active sites in the catalytic domains of potato patatin B2 and human cytosolic PLA₂ (54) (see Fig. 2D) and predicting phospholipase A₂ activity for PEDF-R. We investigated the potential PLA enzymatic activity of purified epitope-tagged PEDF-R with an assay that uses [1,2-dilinoleoyl]-phosphatidylcholine as PLA substrate and lipoxygenase as coupling enzyme (49). In this assay, the PLA activity releases linoleic acid from the substrate that is oxidized by the coupling enzyme to form linoleoyl hydroperoxide, which is followed spectrophotometrically by measuring the increase in absorbance at 234 nm. The rate of absorbance (234 nm) change per min increased linearly with purified epitope-tagged PEDF-R (Fig. 6A). In one experiment we found that recombinant protein translated in the presence of Nonidet P-40 was more active than PEDF-R refolded after synthesis. However, because the purification yields for the refolded protein were significantly larger, this preparation was chosen for the remainder of the assays. The estimated specific activity of the refolded PEDF-R protein was 5.7 A_{234 nm}/min/nmol, whereas that of hog pancreas PLA₂ under identical buffer conditions was about four times lower (1.3 A_{234 nm}/min/nmol) (Fig. 6B). Given that the iPLA₂ (here identified as PEDF-R) contains a nucleotide sequence motif (¹⁴GCGFLG¹⁹) (53), we examined whether ATP would have an effect on the PLA activity of PEDF-R by preincubating PEDF-R with 1 mM ATP for 15 min at 25 °C before assaying the PLA activity. The underlines correspond to the conserved nucleotides for a nucleotide ((G/A)XGXXG) consensus sequence motif. No differences were observed between samples pretreated with or without ATP (data not shown). A pH curve revealed optimum enzymatic activity at pH 7.5 for PEDF-R protein (Fig. 6C). PEDF-R retained its activity when stored at 4 °C but was lost immediately when stored at -20 °C.⁵ These results demonstrate that bacterially derived recombinant PEDF-R had a potent phospholipase activity, which was higher than that for hog pancreas PLA₂ when assayed under identical conditions.

View larger version (28K): [in this window] [in a new window]   FIGURE 6. PLA activity of PEDF-R. PLA activity assays were performed with [1,2-dilinoleoyl]-phosphatidylcholine as substrate and purified His-tagged PEDF-R and lipoxygenase as coupling enzyme in PLA reaction buffer (3 mM deoxycholate, 50 mM Tris-Cl, pH 7.5, 0.01% Nonidet P-40). The appearance of linoleoyl hydroperoxide, the coupled reaction product, was measured spectrophotometrically by increases in absorbance at 234 nm per min. A, dose response of the enzymatic PLA activity of purified His-tagged PEDF-R. Plot of activity given in dA234 nm/min as a function of picomoles per assay is shown. B, dose response of the hog pancreas PLA2 assayed, as for PEDF-R in A. C, effect of pH on the PLA activity of purified His-tagged PEDF-R. Reaction assays were with 0.15 pmol of His-tagged PEDF-R in buffer with indicated pH values. D, effect of bromoenol lactone (BEL) on the PLA activity of PEDF-R. E, effect of PEDF and BSA protein additions on the PLA activity of purified epitope-tagged PEDF-R. Mixtures of 0.15 pmol of PEDF-R and increasing amounts of PEDF were preincubated in PLA reaction buffer at room temperature for 10 min and then assayed for PLA activity. Concentrations of PEDF in PLA reactions are indicated in the x axis (closed circles). Controls were PEDF-R with increasing amounts of BSA (open circles). F, effect of PEDF and BSA additions on the PLA activity of hog pancreas PLA2. Assays were as in E but with 42 pmol of hog pancreas PLA2. PEDF (closed circles); BSA (open circles).

To investigate the effect of PEDF on the PLA activity of PEDF-R, aliquots of purified soluble PEDF-R (0.15 pmol) were preincubated with increasing amounts of PEDF to allow for binding. We found that the A_{234 nm}/min of the preincubated mixtures increased with the molar ratio (Fig. 6E) but did not increase when BSA replaced PEDF, and PEDF alone did not contribute to this activity. In contrast, the activity of hog pancreas PLA₂ was not stimulated by PEDF in reaction conditions identical as for PEDF-R, i.e. with 0.01% Nonidet P-40, or when Nonidet P-40 was omitted (detergent included for PEDF-R solubility) (Fig. 6F). Although inclusion of 0.01% Nonidet P-40 increased the hog PLA₂ activity, PEDF did not have a positive effect on it. Note that SPR kinetic parameters for the PEDF/PEDF-R interactions did not change significantly when binding was performed in the PLA reaction buffer containing 50 mM Tris-HCl, pH 7.5, and 3 mM deoxycholate, an ionic detergent that binds to proteins in solution through hydrophobic interactions of its cholesterol-like hydrocarbon structure (k_a = 1.07 x 10⁵ M^-1 s^-1, k_d = 4.96 x 10^-4 s^-1; K_A (k_a/k_d) = 2.17 x 10⁸ M^-1; K_D (k_d/k_a) = 4.61 x 10^-9 M). These results demonstrated that PEDF specifically stimulated the PLA enzymatic activity of the recombinant PEDF-R.

Subcellular Localization of PEDF-R—To investigate the subcellular localization of PEDF-R in eukaryotic cells, we ectopically expressed Lumio-/V5-tagged PEDF-R in epithelial COS-7, ARPE-19, retinal ganglion RCG-5, retinal R28, and fibroblastic 3T3-L1 cells. The Lumio tag, a 6-amino acid peptide region containing four cysteines, has affinity for arsenical Lumio reagents, which upon binding unquench fluorescence to be detected by laser scanning or microscopy. Laser scanning of gels after SDS-PAGE of protein extracts from cells transfected with N-terminal Lumio/V5-PEDF-R plasmids showed a band for protein migrating with an apparent M_r of 81,000 (Fig. 7A). Western analyses with anti-V5 also showed one band with an identical migration pattern.

Cytosolic and membrane proteins of transfected cells were isolated following a previously optimized protocol for plasma membrane proteins (41). Recombinant N-terminal Lumio/V5-tagged PEDF-R partitioned to both soluble cytosolic and particulated membrane fractions and was not detected in the conditioned media, as shown by laser scanning and Western blotting with anti-V5 antibodies (Fig. 7A). Quantitation of Lumio and V5 immunoreactivity in gels and Western blots of equivalently loaded fractions from cells transfected with Lumio/V5-PEDF-R plasmids showed a cytosol:membrane PEDF-R ratio of 1:3 (Fig. 7A, compare lanes 3 and 4), indicating preferential partitioning of PEDF-R to membranes. Fluorescence microscopy of cells transfected with Lumio/V5-PEDF-R plasmids showed homogeneous fluorescence in the cytosol, in a punctated fashion excluding the nuclei and bordering the cells (Fig. 7B). These results indicate that transiently overexpressed epitope-tagged PEDF-R partitioned in both cytosolic and membrane fractions but preferentially as proteins from membranes, either inserted within the endoplasmic reticulum or at the cell surface.

View larger version (46K): [in this window] [in a new window]   FIGURE 7. Subcellular localization of PEDF-R in mammalian cells. A, transiently expressed epitope-tagged PEDF-R in COS-7 cells. COS-7 cells were transfected with pLmN-PEDF-R, an expression vector of Lumio/V5-tagged PEDF-R. Cytosolic and membrane fractions were separated (150,000 x g) and subjected to SDS-PAGE. Lumio and V5 tags were detected by laser scanning and immunostaining versus anti-V5, respectively. Proteins were detected with Ponceau Red. Samples were as follows: lane 1, conditioned medium; lane 2, lysate; lane 3, cytosolic fraction; lane 4, membrane fraction. Lane 5 was bacterially derived epitope-tagged PEDF-R immunodetected with anti-Xpress. B, transfected COS-7 cells were treated with Lumio Green detection reagent and visualized under fluorescence microscopy. pEXP1-PEDF-R, negative control with no Lumio tag; pLumio-GW-p64, positive control expressing a nucleolar protein with the Lumio tag; pLmN-PEDF-R, sample. C, ARPE-19 cells were fractionated as above except with high speed centrifugation being 80,000 x g. Western blotting versus antibodies to PEDF-R (anti-PEDF-RA), to Na+/K+-ATPase (plasma membrane marker); calreticulin (endoplasmic reticulum marker), GAPDH (cytosolic marker), and TIP47 (lipid droplet marker) are shown. Total protein loaded in each lane were as follows: lane 1, 19.8 µg protein of lysate; lane 2, 12 µg protein of cytosolic fraction; and lane 3, 6.5 µg of membrane fraction. D, specific biotinylation of cell-surface proteins form ARPE-19cells was performed using a membrane-impermeable, thiol-cleavable, and amine-reactive sulfo-N-hydroxysuccinimide-SS-biotin reagent (Pierce) with avidin affinity. Biotinylated proteins were isolated with Neu-trAvidinTM gel and analyzed by Western blotting against antibodies to PEDF-R (anti-PEDF-RA), N-cadherin (plasma membrane marker), and histone 3b (nuclear marker). The bottom panel was Western of total lysate against antibody to tubulin (loading control). Treatments were in lane 1, ARPE-19 cells without biotinylation; lane 2, ARPE-19 cell-surface biotinylation followed by stripping with 50 mM dithiothreitol; and lane 3, ARPE-19 cell-surface biotinylation.

Topology of PEDF-R—To investigate the potential transmembrane nature of PEDF-R, peptides R^A and R^C (Fig. 2C) were synthesized, and antibodies were developed. R^A was designed from a region within the extracellular loop between TM3 and TM4 and next to the N terminus of 12c. R^C was designed from the intracellular loop between TM2 and TM3. Both antibodies Ab-R^A and Ab-R^C immunoreacted mainly with a single band in Western blots of extracts from cells expressing PEDF-R, and their respective peptide antigens blocked these immunoreactivities (Fig. 8A). The predicted extracellular and intracellular regions of PEDF-R were mapped using these antibodies and ARPE-19 cells. Immunocytochemistry of nonpermeabilized ARPE-19 cells with Ab-R^A showed immunolabeling of the cell surfaces, whereas the Ab-R^C label was extremely low if any. In permeabilized ARPE-19 cells, both antibodies immunolabeled the cytoplasm and membranes with comparable intensities and patterns (Fig. 8B). A radial structure arising from the nucleus area toward the plasma membrane was labeled with higher intensity. Controls without primary antibodies did not label cells. Similar observations were obtained with mouse 3T3-L1 cells with PEDF-R expression vectors (Fig. 3C) and rat retinal RGC-5 cells. Cells overexpressing PEDF-R had a significantly higher level of staining than in those without PEDF-R expression vectors, revealing similar topology and subcellular location for the heterologously and endogenously expressed PEDF-R.⁶ Together, these results clearly demonstrated that PEDF-R has extracellular and intracellular loops in agreement with the predicted topology.

PEDF Binding to Cell-surface PEDF-R Protein—To examine the binding of extracellular PEDF ligand to cell membrane PEDF-R, human recombinant PEDF conjugated with fluorescein, Fl-PEDF, was added to cultures of ARPE-19, 3T3L1, and 3T3L1[pLmN-PEDF-R] cells. Live cells exposed to Fl-PEDF concentrations of 5, 15, 20, and 50 nM exhibited a fluorescent stain at their cell surfaces in a dose-dependent fashion and with optimum detection levels at concentrations about 20 nM Fl-PEDF (Fig. 9). The labeling pattern resembled the pattern obtained with the antibody to the extracellular peptide region R^A in nonpermeabilized cells. The staining in ARPE-19 cells was visibly more intense around the nuclei where the density/thickness of the cell is increased, and similar to that reported for N-CAM, a surface marker for ARPE-19 (55). Cells without Fl-PEDF did not show fluorescent signal. A similar Fl-PEDF labeling pattern was observed in 3T3-L1 cells (Fig. 9B). The Fl-PEDF labeling pattern in 3T3-L1 cells stably expressing PEDF-R from expression vectors was identical to that without expression plasmids but at a much higher degree (Fig. 9C), consistent with indiscriminate Fl-PEDF binding to the endogenous mouse PEDF-R and to heterologous expressed human PEDF-R. Furthermore, ligand competition showed that Fl-PEDF fluorescent signal in the presence of a 10- and 50-fold molar excess of unlabeled PEDF over Fl-PEDF decreased on the cell surfaces of L1[pPEDF-R] cells (Fig. 9C). These results indicate that extracellular PEDF bound specifically to cells expressing PEDF-R in plasma membranes, as expected for ligand binding to cell-surface receptors.

View larger version (50K): [in this window] [in a new window]   FIGURE 8. Mapping of predicted extracellular and intracellular loops of PEDF-R with antibodies. A, Western blot of 3T3-L1[PEDF-R] and ARPE-19 cell extracts against antibodies Ab-RA and Ab-RC in the absence and presence of peptide antigens RA and RC, and as indicated at the top of each lane. Four left lanes correspond to extracts from 3T3-L1[PEDF-R] cells stably transfected with plasmid pLmN-PEDF-R; right two lanes correspond to extracts from human ARPE-19 cells. B, immunocytochemistry of ARPE-19 cells with and without primary Ab-RA and Ab-RC (indicated to the left) and secondary antibody. Permeabilized cells were incubated in ethanol prior to fixation and immunodetection; this step was omitted in the nonpermeabilized cells. Bright fields are shown to the left of each immunofluorescent image, with immunofluorescence in red, and permeabilization treatment is indicated at the top. DAPI was added for nuclei staining (blue). C, immunocytochemistry of nonpermeabilized and permeabilized 3T3-L1[PEDF-R] cells with or without primary Ab-RA and Ab-RC antibodies as in B.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In our investigations toward elucidating the mechanisms of action of PEDF in neuronal/retinal systems, we have previously provided biochemical evidence for high affinity PEDF-binding sites and proteins for PEDF in membranes of retina and CNS cells (40, 41, 43, 44). Here, by using yeast two-hybrid screening, we identified a novel gene with characteristics for a PEDF receptor which we termed PEDF-R. This study is the first to identify PEDF-R, although its cDNA has been previously cloned and named independently by different groups. PEDF-R (TTS-2.2/iPLA₂ and mouse desnutrin/ATGL) has been described recently as a member of the new calcium-independent PLA₂/nutrin/Patatin-like phospholipase domain-containing 2 (PNPLA2) family that possesses triglyceride lipase and acylglycerol transacylase activities in adipose systems and is likely to facilitate energy mobilization and storage in adipocytes (51-53). Here we describe the PEDF-R gene expression in retina cells and its heterologous expression by bacterial and eukaryotic systems, and we demonstrate that the PEDF-R protein binds to the PEDF ligand with high affinity and that cell membranes contain the gene product. Most importantly, we demonstrate that PEDF-R exhibits phospholipase A₂ activity that liberates fatty acids, an activity that PEDF can stimulate.

We have performed an exhausting number of assays for the binding of PEDF to PEDF-R, in addition to yeast two-hybrid screening. It is demonstrated that these two proteins interact when they are in solution and when either one is immobilized. PEDF can bind to PEDF-R that is synthesized in vitro by bacterial extracts, in plasma membranes of mammalian cells overexpressing it, or in eukaryotically derived cytosolic extracts (Fig. S4). PEDF-R can retain binding affinity and PLA₂ activity after denaturing and refolding. The activities of bacterially derived PEDF-R imply that eukaryotic post-translation modifications (such as N-glycosylation) are not structural determinants for PEDF binding and PLA₂ activity of the PEDF-R polypeptide. Interestingly, the observed affinity of the interactions between the bacterially derived PEDF-R and PEDF (K_D =3nM)isin agreement with the reported affinities of PEDF to receptors on the surface of cells that respond to PEDF (40, 41, 43, 44, 46). Yeast two-hybrid screening provided information of a potential PEDF-binding domain in PEDF-R, the 12c coding region (250-383 amino acids positions). However, there are reasons to believe that the 12c region contains incomplete structural determinants for PEDF binding as follows. 1) Lower drug tolerance of the yeast colonies coexpressing plasmids for PEDF and isolated clone 12c in the 3-AT assay demonstrate weaker interactions between fragment 12c and PEDF. 2) The amino acid sequence of the 12c region is not well conserved among species (human, mouse, rat, bovine, canine, Xenopus, and Danio), in particular the segment ²⁶⁷ARPH...EDH²⁹⁵ (see supplemental Fig. S1B). 3) More conclusively, the affinity for the PEDF/12c interactions is 3 orders of magnitude lower than that of PEDF/PEDF-R. Consistent with extracellular ligand binding, the N-terminal half of 12c maps to extracellular loop 4 that is available to interact with extracellular ligands. Interestingly, the 12c region contains homology to rat collagen I,⁷ suggesting a molecular basis for interacting with PEDF. The fact that the PEDF affinity for PEDF-R is higher than for rat tail collagen I (K_D of 134 nM) (39) implies that the ligand/receptor binding might be mediated by stronger interactions in addition to those between PEDF and collagen-like regions in 12c of PEDF-R. It is assumed that additional regions from extracellular loops might contribute to the structural determinants in PEDF-R for extracellular ligand binding. It is also assumed that the decrease in sequence identity in the 12c region between PEDF-R and human adiponutrin/iPLA₂ (30% identity) and the lack of this region in the truncated GS2/iPLA₂ suggest that PEDF-R might be the only likely family member with structural determinants for binding PEDF. Moreover, comparisons of PEDF-R binding affinities with other serpins, maspin and ovalbumin, indicate that PEDF might be a serpin with optimal structural determinants for interacting with PEDF-R.

View larger version (76K): [in this window] [in a new window]   FIGURE 9. Fluoresceinated PEDF (Fl-PEDF) binding to the surface of ARPE-19, 3T3L1, and 3T3-L1[PEDF-R] cells. Live cells were treated with or without FL-PEDF as indicated to the left of each subpanel. For ligand competition (C), cells were previously incubated with a molar excess of unlabeled PEDF ligand competitor as indicated. Treatments were as follows: A, ARPE-19 with 100 nM Fl-PEDF at 4 °C for 90 min; B and C, 3T3-L1 and 3T3-L1[PEDF-R] (stably overexpressing epitope-tagged PEDF-R), respectively, with 20 nM Fl-PEDF for 10 min at room temperature. Cells were observed under fluorescence microscopy. Bright fields and PEDF treatments are shown to the left of each fluorescence images, with fluorescence in green.

View larger version (15K): [in this window] [in a new window]   FIGURE 10. PEDF stimulation of PLA activity in ARPE-19 cell. Cell extracts containing native PEDF-R were prepared from ARPE-19 membrane fractions. PLA activity of detergent-soluble membrane fractions was assayed as in Fig. 6. A, dose response of PEDF-R from ARPE-19 cells. Plot of activity given in dA234 nm/min as a function of membrane fraction volume per assay is shown. B, effect of PEDF additions on the PLA activity of PEDF-R from ARPE-19 cells. Mixtures of detergent-soluble membrane fractions and increasing amounts of PEDF were preincubated in PLA reaction buffer at room temperature for 10 min. Concentrations of PEDF (, ) or BSA () in PLA reactions are indicated in the x axis. Two batches of ARPE-19 cells were used, one was the one used in A and at 8 µl of membrane fraction per assay (illustrated here by circles, and ), and the other one was used with 20 µl membrane fraction per assay ().

Our results provide detailed information on the PEDF-R distribution of retina tissues and in cells. PEDF-R is a retina protein distributed to the RPE, the inner segments of photoreceptors and cells in the neural retina. The PEDF-R immunostaining in rat retina resembles the fluorescence of Fl-PEDF ligandbinding sites in bovine retina reported previously (41). In those studies the RPE exhibited autofluorescence and precluded visualizing Fl-PEDF-binding sites in RPE cells. In this study we observe that immunolabeled PEDF-R in the RPE is more intense toward the apical membrane than the basal suggesting a preferred apicolateral receptor distribution. Apicolateral preference is also suggested from the distribution of immunolabel around nuclei in nonpermeabilized ARPE-19 cells with antibodies against an extracellular region, implying protein localization on top of the cell, i.e. apical side. The distribution in inner segments of photoreceptors and ganglion cells of both PEDF-R- and Fl-PEDF-binding sites is in agreement with the PEDF action and ligand/receptor interactions in the retina (41). Our observations are also in agreement with those of Jacob et al. (56). They detected significant amounts of PLA₂ activity against 2-arachidonyl phospholipids in two subretinal fractions adjacent to the rod outer segments, the RPE, and another one presumably containing neuronal cells, Muller cells, and rod inner segments but not in rod outer segments. Although the identity of the protein in the RPE and retina with PLA₂ activity was not reported, they demonstrated that the activity is light-, Ca²⁺-, and Mg²⁺-independent and optimal at pH 7.5-8.0, suggesting that the enzyme responsible for such activity might be PEDF-R. Furthermore, PEDF-R has recently been found modulated in the brain of animals depleted of serotonin or cathecholamine,⁸ implying a neuronal role for PEDF-R in the central nervous system.

In previous reports, observations regarding subcellular localization of desnutrin/ATGL proteins (here PEDF-R) were not comprehensive, as Western and microscopy imagery illustrated their association to membranes, but claims were made only for cytosolic protein (51, 52). Like adiponutrin (57), several programs predict up to four TM domains interrupted by two extracellular loops and three intracellular regions for the PEDF-R protein. Consistent with this prediction, membrane fractions of PEDF-R-overexpressing cells contain most of the protein (estimated 75% of epitope-tagged PEDF-R of total COS-7, RGC-5, and ARPE-19 cell extracts), which is not detected as a secreted protein. Note that we used the same fractionation protocol followed for the isolation of PEDF-binding proteins from retina plasma membranes (40, 41), again implying similarities between these proteins. Biochemical fractionation of RPE cells provides evidence for separation of PEDF-R from lipid droplet protein TIP47, in contrast to immunofluorescence of antibodies to ATGL in HeLa cells (71). Most conclusively, cell-surface labeling and mapping of predicted extracellular and intracellular loops with specific anti-peptides demonstrates that PEDF-R is a transmembrane protein on the surface of cells having an intracellular region between TM2 and TM3 and an extracellular region between TM3 and TM4. The latter region is likely to contain a PEDF-binding domain for PEDF-R.

The phospholipase A₂ domain is of great interest given that the enzyme is known to release bioactive fatty acids that function as second messengers or precursor of eicosanoids that mediate signal transduction (58). Phospholipase PLA₂ can also release bioactive lysophospholipids involved in cell signaling and known to interact with their G protein-coupled receptors to influence the development, functions, and diseases of every mammalian organ system (59). In previous reports on the members of the nutrin/calcium-iPLA family, the authors demonstrated assiduously their triglyceride lipase activities that liberate fatty acids in paradigms for adipose systems. This study demonstrates that the PEDF-R protein is also a functional phospholipase enzyme with an activity severalfold that of hog pancreas PLA₂ (EC 3.1.1.4 [EC] ) and capable of being stimulated by PEDF. The substrate recognition site, formed by the homologous PLA₂ active serine (Ser⁴⁷) and aspartic (Asp¹⁶⁶) amino acids, would face extracellularly in the predicted transmembrane topology of PEDF-R, suggesting release of products to the extracellular side.

The most abundant fatty acids in membrane phospholipids of the retina and central nervous system are arachidonic acid and DHA, known to be essential for proper neuronal function and to exert survival effects on retinal and neuronal cells (60, 61). Like PEDF, they have also been reported to have antitumorigenic and antiangiogenic properties, e.g. DHA is proapoptotic for tumor cells and growth (62, 63). In this regard, Jenkins et al. (53) have claimed PLA₂ activity of iPLA₂ (PEDF-R) with arachidonyl-sn-glycerol-3-phosphocholine supporting the release of arachidonic acid by PEDF-R. Although there is no empirical information of DHA release from phospholipids by PEDF-R, there is recent evidence for this fatty acid as a precursor for neuroprotectin D1, a retinal neuroprotectant, anti-inflammatory, and antioxidative agent for the RPE, retina, and central nervous system (64-66). Most interesting, PEDF at subnanomolar concentrations is a potent activator of the synthesis and release of this DHA derivative from the apical side of human ARPE-19 (67, 72). These observations suggest that PEDF may exert survival actions in the neural retina via a lipidsignaling pathway mediated by PEDF-R. It is interesting to speculate that plasma membrane-associated PEDF-R may produce extracellular bioactive lipids, which can diffuse back into the cell as signaling molecules. PEDF, by binding to membranelocalized PEDF-R, may stimulate release of fatty acids and lysophospholipids, providing neuronal cells with higher survival or differentiation capability and also rendering endothelial and tumor cells with irreversible cell damage. Although detailed examination of this working hypothesis must be performed to claim such mechanisms of action for PEDF, our results suggest a molecular pathway by which receptor/ligand interaction on the cell surface could generate a cellular signal to promote its biological activities.

Additional studies are also necessary on PEDF internalization and on potential PEDF stimulation of the triglyceride lipase and acylglycerol transacylase activities of the cytosolic PEDF-R protein. PEDF may also stimulate the cellular release of bioactive lipids, e.g. free fatty acids and lysophosphatidic acid, via the cytosolic phospholipase and triglyceride activities of PEDF-R to act as secondary messengers or for lipid remodeling in the retina. There is evidence for lipid metabolism in the RPE and the existence of retinosomes (retinyl ester storage particles), particles reminiscent of lipid bodies identified in RPE as distinct structures compartmentalizing a metabolic intermediate involved in regeneration of the visual chromophore (68). The RPE undergoes phagocytosis of outer segments of photoreceptors as light induces transduction signals. It is suggested that age-related loss of efficiency in phagocytosis and lipid accumulation in the RPE cells are because of a likely alteration in lipid metabolism. In early age-related macular degeneration, lipid-containing deposits (drusen) accumulate in Bruch membrane underlying the RPE. Bruch membrane lies at the critical juncture between the outer retina and its blood supply, the choriocapillaris. Lipid deposition causes reduced hydraulic conductivity and macromolecular permeability in Bruch membrane and is thought to impair retinal metabolism. There is histochemical evidence for an age-related increase in esterified and free fatty acids in human Bruch membrane that accumulate in extracellular droplets (69). Further investigations are necessary to reveal whether PEDF-R together with PEDF may interact with intracellular retinosomes and/or with extracellular drusen adjacent to RPE cells.

In conclusion, we have identified a novel PEDF-R receptor that is a lipase-linked membrane protein with high affinity for PEDF. Binding of PEDF to the extracellular loop between TM3 and TM4 of PEDF-R on the cell surface may mediate its neurotrophic effects and the promotion of neuronal survival and differentiation. Our data suggest that PEDF supports neural survival and differentiation by stimulation of neural surface fatty acid/lipid metabolism, which in turn may regulate neural physiology by established mechanisms. The PEDF/PEDF-R interactions may also mediate its antiangiogenic and antitumorigenic activities by stimulation of lipid mediators with proapoptotic properties. These results provide a mechanistic insight with respect to the biological action and signaling pathways of PEDF.

	FOOTNOTES

 
^* This work was supported in part by the Intramural Research Program of the NEI, National Institutes of Health, by research grants from the "Universidad de Castilla-La Mancha," and by Grant SAF2002-03086 from the "Ministerio de Ciencia y Tecnologia," Spain. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The on-line version of this article (available at http://www.jbc.org) contains supplemental Materials and Methods, Table S1, and Figs. S1-S4.

¹ Recipient of a fellowship from the "Consejeria de Sanidad de la Junta de Comunidades de Castilla-La Mancha."

² These authors should be considered as senior authors for this work.

³ To whom correspondence should be addressed: NEI, National Institutes of Health, Bldg. 7, Rm. 304, 7 Memorial Dr., MSC 0706, Bethesda, MD 20892-0706. Tel.: 301-496-6514; Fax: 301-451-5420; E-mail: becerrap{at}nei.nih.gov.

⁴ The abbreviations used are: PEDF, pigment epithelium-derived factor; PEDF-R, PEDF receptor; PLA, phospholipase A; iPLA₂, independent PLA₂; RPE, retinal pigment epithelium; 3-AT, 3-aminotriazol; TM, transmembrane; SPR, surface plasmon resonance; BSA, bovine serum albumin; PBS, phosphate-buffered saline; DAPI, 4,6-diamidino-2-phenylindole; ELISA, enzyme-linked immunosorbent assay; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; AD, activation domain; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; Ni-NTA, nickel-nitrilotriacetic acid; TRITC, tetramethylrhodamine; DHA, docosahexaenoic acid.

⁵ L. Notari, unpublished observations.

⁶ R. Heredia and S. P. Becerra, unpublished observations.

⁷ P. Notario, L. Notari, and S. P. Becerra, unpublished observations.

⁸ G. Chen, personal communication.

	ACKNOWLEDGMENTS

 
We thank Gail Seigel for providing R28 cells; Neeraj Agarwal for providing the RGC-5 cells; Fei Wang for providing human fetal RPE cDNA; Ignacio Rodriguez for providing human retina cDNA and albino rat eyes; Peggy Zelenka for insightful discussions on cell-surface proteins; Dean Londos and Amy Wagner for providing insightful information on the biochemistry of cellular lipid droplet and the antibody to TIP47; and Catherine Jackson for discussion on ATGL in HeLa cells.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Barnstable, C. J., and Tombran-Tink, J. (2004) Prog. Retin. Eye Res. 23, 561-577[CrossRef][Medline] [Order article via Infotrieve]
2. Bouck, N. (2002) Trends Mol. Med. 8, 330-334[CrossRef][Medline] [Order article via Infotrieve]
3. Crawford, S. E., Stellmach, V., Ranalli, M., Huang, X., Huang, L., Volpert, O., De Vries, G. H., Abramson, L. P., and Bouck, N. (2001) J. Cell Sci. 114, 4421-4428[Medline] [Order article via Infotrieve]
4. Garcia, M., Fernandez-Garcia, N. I., Rivas, V., Carretero, M., Escamez, M. J., Gonzalez-Martin, A., Medrano, E. E., Volpert, O., Jorcano, J. L., Jimenez, B., Larcher, F., and Del Rio, M. (2004) Cancer Res. 64, 5632-5642[Abstract/Free Full Text]
5. Gettins, P. G., Simonovic, M., and Volz, K. (2002) Biol. Chem. 383, 1677-1682[CrossRef][Medline] [Order article via Infotrieve]
6. Wang, L., Schmitz, V., Perez-Mediavilla, A., Izal, I., Prieto, J., and Qian, C. (2003) Mol. Ther. 8, 72-79[CrossRef][Medline] [Order article via Infotrieve]
7. Tombran-Tink, J., Chader, G. G., and Johnson, L. V. (1991) Exp. Eye Res. 53, 411-414[CrossRef][Medline] [Order article via Infotrieve]
8. Becerra, S. P., Fariss, R. N., Wu, Y. Q., Montuenga, L. M., Wong, P., and Pfeffer, B. A. (2004) Exp. Eye Res. 78, 223-234[CrossRef][Medline] [Order article via Infotrieve]
9. Perez-Mediavilla, L. A., Chew, C., Campochiaro, P. A., Nickells, R. W., Notario, V., Zack, D. J., and Becerra, S. P. (1998) Biochim. Biophys. Acta 1398, 203-214[Medline] [Order article via Infotrieve]
10. Karakousis, P. C., John, S. K., Behling, K. C., Surace, E. M., Smith, J. E., Hendrickson, A., Tang, W. X., Bennett, J., and Milam, A. H. (2001) Mol. Vis. 7, 154-163[Medline] [Order article via Infotrieve]
11. Tombran-Tink, J., Shivaram, S. M., Chader, G. J., Johnson, L. V., and Bok, D. (1995) J. Neurosci. 15, 4992-5003[Abstract]
12. Wu, Y. Q., Notario, V., Chader, G. J., and Becerra, S. P. (1995) Protein Expression Purif. 6, 447-456[CrossRef][Medline] [Order article via Infotrieve]
13. Jablonski, M. M., Tombran-Tink, J., Mrazek, D. A., and Iannaccone, A. (2000) J. Neurosci. 20, 7149-7157[Abstract/Free Full Text]
14. Cayouette, M., Smith, S. B., Becerra, S. P., and Gravel, C. (1999) Neurobiol. Dis. 6, 523-532[CrossRef][Medline] [Order article via Infotrieve]
15. Dawson, D. W., Volpert, O. V., Gillis, P., Crawford, S. E., Xu, H., Benedict, W., and Bouck, N. P. (1999) Science 285, 245-248[Abstract/Free Full Text]
16. Holekamp, N. M., Bouck, N., and Volpert, O. (2002) Am. J. Ophthalmol. 134, 220-227[CrossRef][Medline] [Order article via Infotrieve]
17. Duh, E. J., Yang, H. S., Haller, J. A., De Juan, E., Humayun, M. S., Gehlbach, P., Melia, M., Pieramici, D., Harlan, J. B., Campochiaro, P. A., and Zack, D. J. (2004) Am. J. Ophthalmol. 137, 668-674[Medline] [Order article via Infotrieve]
18. Cao, W., Tombran-Tink, J., Elias, R., Sezate, S., Mrazek, D., and McGinnis, J. F. (2001) Investig. Ophthalmol. Vis. Sci. 42, 1646-1652[Abstract/Free Full Text]
19. Duh, E. J., Yang, H. S., Suzuma, I., Miyagi, M., Youngman, E., Mori, K., Katai, M., Yan, L., Suzuma, K., West, K., Davarya, S., Tong, P., Gehlbach, P., Pearlman, J., Crabb, J. W., Aiello, L. P., Campochiaro, P. A., and Zack, D. J. (2002) Investig. Ophthalmol. Vis. Sci. 43, 821-829[Abstract/Free Full Text]
20. Mori, K., Gehlbach, P., Ando, A., McVey, D., Wei, L., and Campochiaro, P. A. (2002) Investig. Ophthalmol. Vis. Sci. 43, 2428-2434[Abstract/Free Full Text]
21. Mori, K., Gehlbach, P., Yamamoto, S., Duh, E., Zack, D. J., Li, Q., Berns, K. I., Raisler, B. J., Hauswirth, W. W., and Campochiaro, P. A. (2002) Investig. Ophthalmol. Vis. Sci. 43, 1994-2000[Abstract/Free Full Text]
22. Kuncl, R. W., Bilak, M. M., Bilak, S. R., Corse, A. M., Royal, W., and Becerra, S. P. (2002) J. Neurochem. 81, 178-184[CrossRef][Medline] [Order article via Infotrieve]
23. Singh, V. K., Chader, G. J., and Rodriguez, I. R. (1998) Mol. Vis. 4, 7[Medline] [Order article via Infotrieve]
24. Ortego, J., Escribano, J., Becerra, S. P., and Coca-Prados, M. (1996) Investig. Ophthalmol. Vis. Sci. 37, 2759-2767[Abstract/Free Full Text]
25. Houenou, L. J., D'Costa, A. P., Li, L., Turgeon, V. L., Enyadike, C., Alberdi, E., and Becerra, S. P. (1999) J. Comp. Neurol. 412, 506-514[CrossRef][Medline] [Order article via Infotrieve]
26. Araki, T., Taniwaki, T., Becerra, S. P., Chader, G. J., and Schwartz, J. P. (1998) J. Neurosci. Res. 53, 7-15[CrossRef][Medline] [Order article via Infotrieve]
27. DeCoster, M. A., Schabelman, E., Tombran-Tink, J., and Bazan, N. G. (1999) J. Neurosci. Res. 56, 604-610[CrossRef][Medline] [Order article via Infotrieve]
28. Taniwaki, T., Hirashima, N., Becerra, S. P., Chader, G. J., Etcheberrigaray, R., and Schwartz, J. P. (1997) J. Neurochem. 68, 26-32[Medline] [Order article via Infotrieve]
29. Bilak, M. M., Corse, A. M., Bilak, S. R., Lehar, M., Tombran-Tink, J., and Kuncl, R. W. (1999) J. Neuropathol. Exp. Neurol. 58, 719-728[Medline] [Order article via Infotrieve]
30. Sugita, Y., Becerra, S. P., Chader, G. J., and Schwartz, J. P. (1997) J. Neurosci. Res. 49, 710-718[CrossRef][Medline] [Order article via Infotrieve]
31. Doll, J. A., Stellmach, V. M., Bouck, N. P., Bergh, A. R., Lee, C., Abramson, L. P., Cornwell, M. L., Pins, M. R., Borensztajn, J., and Crawford, S. E. (2003) Nat. Med. 9, 774-780[CrossRef][Medline] [Order article via Infotrieve]
32. Guan, M., Pang, C. P., Yam, H. F., Cheung, K. F., Liu, W. W., and Lu, Y. (2004) Cancer Gene Ther. 11, 325-332[CrossRef][Medline] [Order article via Infotrieve]
33. Streck, C. J., Zhang, Y., Zhou, J., Ng, C., Nathwani, A. C., and Davidoff, A. M. (2005) J. Pediatr. Surg. 40, 236-243[CrossRef][Medline] [Order article via Infotrieve]
34. Volpert, O. V., Zaichuk, T., Zhou, W., Reiher, F., Ferguson, T. A., Stuart, P. M., Amin, M., and Bouck, N. P. (2002) Nat. Med. 8, 349-357[CrossRef][Medline] [Order article via Infotrieve]
35. Abramson, L. P., Stellmach, V., Doll, J. A., Cornwell, M., Arensman, R. M., and Crawford, S. E. (2003) J. Pediatr. Surg. 38, 336-342[CrossRef][Medline] [Order article via Infotrieve]
36. Simonovic, M., Gettins, P. G., and Volz, K. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 11131-11135[Abstract/Free Full Text]
37. Becerra, S. P., Sagasti, A., Spinella, P., and Notario, V. (1995) J. Biol. Chem. 270, 25992-25999[Abstract/Free Full Text]
38. Alberdi, E., Hyde, C. C., and Becerra, S. P. (1998) Biochemistry 37, 10643-10652[CrossRef][Medline] [Order article via Infotrieve]
39. Meyer, C., Notari, L., and Becerra, S. P. (2002) J. Biol. Chem. 277, 45400-45407[Abstract/Free Full Text]
40. Alberdi, E., Aymerich, M. S., and Becerra, S. P. (1999) J. Biol. Chem. 274, 31605-31612[Abstract/Free Full Text]
41. Aymerich, M. S., Alberdi, E. M., Martinez, A., and Becerra, S. P. (2001) Investig. Ophthalmol. Vis. Sci. 42, 3287-3293[Abstract/Free Full Text]
42. Liu, H., Ren, J. G., Cooper, W. L., Hawkins, C. E., Cowan, M. R., and Tong, P. Y. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 6605-6610[Abstract/Free Full Text]
43. Filleur, S., Volz, K., Nelius, T., Mirochnik, Y., Huang, H., Zaichuk, T. A., Aymerich, M. S., Becerra, S. P., Yap, R., Veliceasa, D., Shroff, E. H., and Volpert, O. V. (2005) Cancer Res. 65, 5144-5152[Abstract/Free Full Text]
44. Bilak, M. M., Becerra, S. P., Vincent, A. M., Moss, B. H., Aymerich, M. S., and Kuncl, R. W. (2002) J. Neurosci. 22, 9378-9386[Abstract/Free Full Text]
45. Stellmach, V., Crawford, S. E., Zhou, W., and Bouck, N. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 2593-2597[Abstract/Free Full Text]
46. Yamagishi, S., Inagaki, Y., Amano, S., Okamoto, T., Takeuchi, M., and Makita, Z. (2002) Biochem. Biophys. Res. Commun. 296, 877-882[CrossRef][Medline] [Order article via Infotrieve]
47. Baladron, V., Ruiz-Hidalgo, M. J., Bonvini, E., Gubina, E., Notario, V., and Laborda, J. (2002) Biochem. Biophys. Res. Commun. 291, 193-204[CrossRef][Medline] [Order article via Infotrieve]
48. Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D. J. (1997) Nucleic Acids Res. 25, 3389-3402[Abstract/Free Full Text]
49. Jimenez-Atienzar, M., Cabanes, J., Gandia-Herrero, F., Escribano, J., Garcia-Carmona, F., and Perez-Gilabert, M. (2003) Lipids 38, 677-682[CrossRef][Medline] [Order article via Infotrieve]
50. Hubbard, T., Andrews, D., Caccamo, M., Cameron, G., Chen, Y., Clamp, M., Clarke, L., Coates, G., Cox, T., Cunningham, F., Curwen, V., Cutts, T., Down, T., Durbin, R., Fernandez-Suarez, X. M., Gilbert, J., Hammond, M., Herrero, J., Hotz, H., Howe, K., Iyer, V., Jekosch, K., Kahari, A., Kasprzyk, A., Keefe, D., Keenan, S., Kokocinsci, F., London, D., Longden, I., McVicker, G., Melsopp, C., Meidl, P., Potter, S., Proctor, G., Rae, M., Rios, D., Schuster, M., Searle, S., Severin, J., Slater, G., Smedley, D., Smith, J., Spooner, W., Stabenau, A., Stalker, J., Storey, R., Trevanion, S., UretaVidal, A., Vogel, J., White, S., Woodwark, C., and Birney, E. (2005) Nucleic Acids Res. 33, D447-D453[Abstract/Free Full Text]
51. Zimmermann, R., Strauss, J. G., Haemmerle, G., Schoiswohl, G., Birner-Gruenberger, R., Riederer, M., Lass, A., Neuberger, G., Eisenhaber, F., Hermetter, A., and Zechner, R. (2004) Science 306, 1383-1386[Abstract/Free Full Text]
52. Villena, J. A., Roy, S., Sarkadi-Nagy, E., Kim, K. H., and Sul, H. S. (2004) J. Biol. Chem. 279, 47066-47075[Abstract/Free Full Text]
53. Jenkins, C. M., Mancuso, D. J., Yan, W., Sims, H. F., Gibson, B., and Gross, R. W. (2004) J. Biol. Chem. 279, 48968-48975[Abstract/Free Full Text]
54. Hirschberg, H. J., Simons, J. W., Dekker, N., and Egmond, M. R. (2001) Eur. J. Biochem. 268, 5037-5044[Medline] [Order article via Infotrieve]
55. Chen, S., Samuel, W., Fariss, R. N., Duncan, T., Kutty, R. K., and Wiggert, B. (2003) J. Neurochem. 84, 972-981[CrossRef][Medline] [Order article via Infotrieve]
56. Jacob, M., Weech, P. K., and Salesse, C. (1996) J. Biol. Chem. 271, 19209-19218[Abstract/Free Full Text]
57. Baulande, S., Lasnier, F., Lucas, M., and Pairault, J. (2001) J. Biol. Chem. 276, 33336-33344[Abstract/Free Full Text]
58. Balsinde, J., Winstead, M. V., and Dennis, E. A. (2002) FEBS Lett. 531, 2-6[CrossRef][Medline] [Order article via Infotrieve]
59. Goetzl, E. J., and Tigyi, G. (2004) J. Cell. Biochem. 92, 867-868[CrossRef][Medline] [Order article via Infotrieve]
60. SanGiovanni, J. P., and Chew, E. Y. (2005) Prog. Retin. Eye Res. 24, 87-138[CrossRef][Medline] [Order article via Infotrieve]
61. Kim, H. Y., Akbar, M., and Kim, K. Y. (2001) J. Mol. Neurosci. 16, 223-227, 279-284[CrossRef][Medline] [Order article via Infotrieve]
62. Rose, D. P., and Connolly, J. M. (1999) Pharmacol. Ther. 83, 217-244[CrossRef][Medline] [Order article via Infotrieve]
63. Rose, D. P., and Connolly, J. M. (1999) Int. J. Oncol. 15, 1011-1015[Medline] [Order article via Infotrieve]
64. Mukherjee, P. K., Marcheselli, V. L., Serhan, C. N., and Bazan, N. G. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 8491-8496[Abstract/Free Full Text]
65. Bazan, N. G. (2005) Brain Pathol. 15, 159-166[Medline] [Order article via Infotrieve]
66. Lukiw, W. J., Cui, J. G., Marcheselli, V. L., Bodker, M., Botkjaer, A., Gotlinger, K., Serhan, C. N., and Bazan, N. G. (2005) J. Clin. Investig. 115, 2774-2783[CrossRef][Medline] [Order article via Infotrieve]
67. Bazan, N. G., Marcheselli, V. L., Hu, J., Finley, J., Bok, D., and Chandamuri, B. (2005) Investig. Ophthalmol. Vis. Sci. 46, 167
68. Imanishi, Y., Gerke, V., and Palczewski, K. (2004) J. Cell Biol. 166, 447-453[Abstract/Free Full Text]
69. Curcio, C. A., Millican, C. L., Bailey, T., and Kruth, H. S. (2001) Investig. Ophthalmol. Vis. Sci. 42, 265-274[Abstract/Free Full Text]
70. Rydel, T. J., Williams, J. M., Krieger, E., Moshiri, F., Stallings, W. C., Brown, S. M., Pershing, J. C., Purcell, J. P., and Alibhai, M. F. (2003) Biochemistry 42, 6696-6708[CrossRef][Medline] [Order article via Infotrieve]
71. Smirnova, E., Goldberg, E. B., Makarova, K. S., Lin, L., Brown, W. J., and Jackson, C. L. (2006) EMBO Rep. 7, 106-113[CrossRef][Medline] [Order article via Infotrieve]
72. Marcheselli, V. L., Bazan, N. G., Hu, J., Bok, D., Hardy, M., and Mukherjee, P. K. (2005) Neurotrophic Growth Factors Up-regulate Neuroprotectin D1 Synthesis and Prosurvival Signaling in Human Retinal Pigment Epithelial Cells. Program 148.9, Society for Neuroscience, Washington, D. C.

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				I. Jorge, P. Navarro, P. Martinez-Acedo, E. Nunez, H. Serrano, A. Alfranca, J. M. Redondo, and J. Vazquez Statistical Model to Analyze Quantitative Proteomics Data Obtained by 18O/16O Labeling and Linear Ion Trap Mass Spectrometry: Application to the Study of Vascular Endothelial Growth Factor-induced Angiogenesis in Endothelial Cells Mol. Cell. Proteomics, May 1, 2009; 8(5): 1130 - 1149. [Abstract] [Full Text] [PDF]

				A. Bernard, J. Gao-Li, C.-A. Franco, T. Bouceba, A. Huet, and Z. Li Laminin Receptor Involvement in the Anti-angiogenic Activity of Pigment Epithelium-derived Factor J. Biol. Chem., April 17, 2009; 284(16): 10480 - 10490. [Abstract] [Full Text] [PDF]

				P. C. Kienesberger, M. Oberer, A. Lass, and R. Zechner Mammalian patatin domain containing proteins: a family with diverse lipolytic activities involved in multiple biological functions J. Lipid Res., April 1, 2009; 50(Supplement): S63 - S68. [Abstract] [Full Text] [PDF]

				Y. Mirochnik, A. Aurora, F. T. Schulze-Hoepfner, A. Deabes, V. Shifrin, R. Beckmann, C. Polsky, and O. V. Volpert Short Pigment Epithelial-Derived Factor-Derived Peptide Inhibits Angiogenesis and Tumor Growth Clin. Cancer Res., March 1, 2009; 15(5): 1655 - 1663. [Abstract] [Full Text] [PDF]

				T.-C. Ho, S.-L. Chen, Y.-C. Yang, T.-H. Lo, J.-W. Hsieh, H.-C. Cheng, and Y.-P. Tsao Cytosolic phospholipase A2-{alpha} is an early apoptotic activator in PEDF-induced endothelial cell apoptosis Am J Physiol Cell Physiol, February 1, 2009; 296(2): C273 - C284. [Abstract] [Full Text] [PDF]

				R. Zechner, P. C. Kienesberger, G. Haemmerle, R. Zimmermann, and A. Lass Adipose triglyceride lipase and the lipolytic catabolism of cellular fat stores J. Lipid Res., January 1, 2009; 50(1): 3 - 21. [Abstract] [Full Text] [PDF]

				P. Wang, E. Smit, M. C G J Brouwers, G. H Goossens, C. J H van der Kallen, M. M J van Greevenbroek, and E. C M Mariman Plasma pigment epithelium-derived factor is positively associated with obesity in Caucasian subjects, in particular with the visceral fat depot Eur. J. Endocrinol., December 1, 2008; 159(6): 713 - 718. [Abstract] [Full Text] [PDF]

				Y. Murakami, Y. Ikeda, Y. Yonemitsu, M. Onimaru, K. Nakagawa, R.-i. Kohno, M. Miyazaki, T. Hisatomi, M. Nakamura, T. Yabe, et al. Inhibition of Nuclear Translocation of Apoptosis-Inducing Factor Is an Essential Mechanism of the Neuroprotective Activity of Pigment Epithelium-Derived Factor in a Rat Model of Retinal Degeneration Am. J. Pathol., November 1, 2008; 173(5): 1326 - 1338. [Abstract] [Full Text] [PDF]

				M. Prentki and S. R. M. Madiraju Glycerolipid Metabolism and Signaling in Health and Disease Endocr. Rev., October 1, 2008; 29(6): 647 - 676. [Abstract] [Full Text] [PDF]

				T.-C. Ho, S.-L. Chen, Y.-C. Yang, C.-L. Liao, H.-C. Cheng, and Y.-P. Tsao PEDF induces p53-mediated apoptosis through PPAR gamma signaling in human umbilical vein endothelial cells Cardiovasc Res, November 1, 2007; 76(2): 213 - 223. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 281/49/38022    most recent M600353200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Notari, L. Articles by Becerra, S. P. Search for Related Content PubMed PubMed Citation Articles by Notari, L. Articles by Becerra, S. P. Social Bookmarking                      What's this?