Cloning and Characterization of Rat Density-Enhanced Phosphatase-1, a Protein Tyrosine Phosphatase Expressed by Vascular Cells
We have cloned from cultured vascular smooth muscle cells a protein tyrosine phosphatase, rat density-enhanced phosphatase-1 (rDEP-1), which is a probable rat homologue of DEP-1/HPTPη. rDEP-1 is encoded by an 8.7-kb transcript and is expressed as a 180- to 220-kD protein. The rDEP-1 gene is located on human chromosome 11 (region p11.2) and on mouse chromosome 2 (region 2E). The cDNA sequence predicts a transmembrane protein consisting of a single phosphatase catalytic domain in the intracellular region, a single transmembrane domain, and eight fibronectin type III repeats in the extracellular region (GenBank accession number U40790). In situ hybridization analysis demonstrates that rDEP-1 is widely expressed in vivo but that expression is highest in cells that form epithelioid monolayers. In cultured cells with epitheliod morphology, including endothelial cells and newborn smooth muscle cells, but not in fibroblast-like cells, rDEP-1 transcript levels are dramatically upregulated as population density increases. In vivo, quiescent endothelial cells in normal arteries express relatively high levels of rDEP-1. During repair of vascular injury, expression of rDEP-1 is downregulated in migrating and proliferating endothelial cells. In vivo, rDEP-1 transcript levels are present in very high levels in megakaryocytes, and circulating platelets have high levels of the rDEP-1 protein. In vitro, initiation of differentiation of the human megakaryoblastic cell line CHRF-288-11 with phorbol 12-myristate 13-acetate leads to a very strong upregulation of rDEP-1 transcripts. The deduced structure and the regulation of expression of rDEP-1 suggest that it may play a role in adhesion and/or signaling events involving cell-cell and cell-matrix contact.
Protein tyrosine kinases have been shown to play important roles in signal transduction pathways required for cell proliferation, adhesion, and migration in vascular SMCs. In the rat carotid artery, growth factors acting through receptors with tyrosine kinase activity have been shown to mediate the proliferative and migratory response of SMCs following balloon injury of the vessel wall. For instance, antibodies against basic fibroblast growth factor significantly inhibit the initial wave of SMC proliferation that occurs within 24 to 48 hours after vessel injury.1 In addition, antibodies against PDGF diminish the intimal accumulation of SMCs that occurs within the first week after vessel injury,2 apparently by blocking SMC migration into the intima.
In contrast to the information about the involvement of protein tyrosine kinases in modulating the behavior of vascular cells, there are almost no data about the role of PTPases in these processes. The PTPase family, like the protein tyrosine kinase family, can be subdivided into two main groups: the transmembrane receptor-like PTPases and the cytosolic nonreceptor PTPases.3 PTPases have now been implicated in many cellular processes, including proliferation, differentiation, transformation, and T-cell activation. In signal transduction pathways, they can provide either negative or positive regulatory signals, depending on the system. For instance, the cytosolic PTPase, Syp/SH-PTP2, has been shown to associate with growth factor receptors through its phosphotyrosine-binding SH2 (src homology-2) domains and to relay a positive mitogenic signal. In mammalian fibroblasts, the microinjection of a Syp antibody or a Syp-GST-SH2 fusion protein blocks DNA synthesis by 65% to 85% after insulin and epidermal growth factor addition.4 In NIH-3T3 cells, the expression of a catalytically inactive Syp phosphatase results in a significant reduction of the mitogenic response to insulin.5
Receptor-like PTPases contain a single transmembrane domain, one or two catalytic domains within the cytoplasmic region, and a wide variety of structural motifs in the extracellular region. These motifs include fibronectin type III, immunoglobulin, carbonic anhydrase, and MAM adhesive protein (meprin, A5-like motif) domains.3 The presence of the same structural motifs in adhesion molecules and growth factor receptors suggests that the PTPase activity of receptor-like PTPases might be regulated by soluble ligands or adhesive molecules present on neighboring cells or in the extracellular matrix. Since PTPases can play diverse and cell-specific regulatory roles, we have chosen to determine which PTPases are expressed by vascular cells and the roles that these PTPases play in regulating vascular cell behavior. In the present study, we report the cloning and characterization of a receptor-like PTPase, designated rDEP-1, which is expressed by various cell types, including vascular SMCs, endothelial cells, and platelets. The rDEP-1 cDNA encodes the putative rat homologue of the human PTPase DEP-1/HPTPη.6 7
Materials and Methods
SMC lines derived from the thoracic aortas of 12-day-old or 3-month-old Wistar-Kyoto rats (WKY-12 and WKY-3M, respectively) were obtained from Dr C. Giachelli, University of Washington, Seattle. SMC lines derived from the thoracic aortas of Sprague-Dawley rats were derived in our laboratory from 18-day-old and 3-month-old rats (SD-18 and SD-3M, respectively). Bovine aortic endothelial cells derived from the aortas of adult cows were obtained from Dr S. Schwartz, University of Washington, Seattle. The human megakaryoblastic cell line CHRF-288-11 was established in vitro and characterized by Fugman et al.8 The above cultures were maintained in DMEM supplemented with 10% bovine calf serum.
Degenerate primers corresponding to the PTPase catalytic domain sequences NDYINA and HCSAGI/V were used in a PCR to amplify a band of ≈500 bps from cDNA prepared by random priming total RNA from cultured rat aortic SMCs (WKY-12). The PCR product was cloned into pBluescriptII (Stratagene) and transformed into XL1-Blue Escherichia coli (Stratagene). Five different PTPase clones were obtained (L.G. Borges and D.F. Bowen-Pope, unpublished data, 1996). One of these clones, encoding the catalytic domain of rDEP-1, was used to screen an oligo-dT–primed cDNA library made from RNA extracted from SMCs derived from the aortas of 12-day-old rats. Of 6.72×105 plaques screened, 24 positive clones were picked. The clone with the largest insert, 6.4 kb, was used for sequencing by the dideoxy chain termination method9 using either 35S-labeled deoxynucleotides or fluorescence-labeled dideoxynucleotides. Both cDNA strands were sequenced. Reactions performed using fluorescence-labeled dideoxynucleotides were analyzed using an ABI automated DNA sequencer.
Northern Blot Analysis
Total RNA was isolated by the guanidinium isothiocyanate method.10 For Northern blot analysis, 10 to 15 μg of total RNA was loaded per lane. Hybridization was carried out at 42°C in 50% formamide, 0.75 mol/L sodium chloride, 0.05 mol/L Tris (pH 7.4), 1× Denhardt's solution, 2.1% SDS, 10% dextran sulfate, and 200 μg/mL heat-denatured salmon sperm DNA. Washes were performed at 65°C in 0.25% SDS and 0.3× SSC (45 mmol/L sodium chloride and 4.5 mmol/L sodium citrate, pH 7.0). The probes were radiolabeled with [α-32P]dCTP by DNA random priming (GIBCO-BRL). Blots were exposed to film and/or to phosphorimaging plates for analysis using a Molecular Dynamics phosphorimager at the Phosphorimager Facility of the Markey Molecular Medicine Center at the University of Washington. After quantification of the signal, the blots were stripped in boiling 0.5% SDS and reprobed with a 32P–end-labeled oligonucleotide (5′-GCGAGAGCGCCAGCTATCCTGAGG-3′), which hybridizes to the 28S ribosomal RNA. Blots were then reexposed to phosphorimaging plates, quantified, and normalized for RNA loading on the basis of the signal obtained with the 28S oligonucleotide probe.
DEP-1 was mapped to both human and mouse chromosomes using FISH. Briefly, an rDEP-1 cDNA probe containing a 6.4-kb insert was labeled with biotin-14-dATP by nick translation (GIBCO-BRL). Metaphase chromosome preparations from lymphocytes of a male C57BL/6J mouse and metaphase chromosome preparations from a normal human male synchronized by methotrexate were obtained using 75 mmol/L potassium chloride as a hypotonic buffer and methanol/acetic acid (3:1 [vol/vol]) as a fixative. Slides were denatured in the presence of 70% formamide at 70°C. The hybridization to mouse and human metaphase chromosomes was carried out as previously described by Edelhoff et al.11 After incubation with goat antibiotin antibody, slides were rinsed in posthybridization buffer (0.2 mol/L sodium phosphate, 0.15 mol/L sodium chloride, 0.1% Tween-20, and 0.15% BSA). A second incubation with a fluorescein-labeled anti-goat IgG and a rinse in posthybridization buffer followed. The chromosomes were banded using Hoechst 33258/actinomycin D staining and counterstained with propidium iodide. The chromosomes and hybridization signals were visualized by fluorescence microscopy, using a dual-pass filter.
Mapping to human chromosome 11 was confirmed by Southern blot hybridization of a rat cDNA rDEP-1 probe to a panel of somatic cell hybrids carrying different subsets of human chromosomes (BIOS Laboratories). The rDEP-1 probe used for the Southern blot analysis was a cDNA fragment 490-bp long between nucleotides 262 and 751 in the extracellular region of rDEP-1. A test panel blot containing human, hamster, and mouse genomic DNAs digested with EcoRI, HindIII, Pst I, and Taq I restriction enzymes was probed to find restriction fragment length polymorphisms that were clearly distinguishable between the human and rodent genomic DNAs. A somatic cell hybrid blot containing HindIII-digested DNA from 20 different somatic cell hybrids harboring subsets of human chromosomes was then probed with the rDEP-1 cDNA. The location of the rDEP-1 gene was determined using a discordance table provided with the blot. The prehybridization and hybridization of the Southern blots were carried out at 42°C in the same buffer used for the Northern blot analysis. Washes were performed at 65°C in 0.25% SDS and 0.5× SSC.
Expression of rDEP-1–GST Fusion Protein
A cDNA restriction fragment encoding the complete cytoplasmic domain sequence of rDEP-1 (nucleotides 2745 through 3868) was expressed in bacteria as a fusion protein with GST (Pharmacia). Briefly, a Bgl I–Afl III restriction fragment of the rDEP-1 cDNA was subcloned into the Sma I site of the pGEX 4T-3 vector, after repairing the 5′ and 3′ overhangs with Klenow DNA polymerase. After ligation, the construct was transformed into JM105 E. coli cells (Stratagene) and sequenced to confirm that the open reading frame had been maintained intact. The fusion protein was purified using glutathione-Sepharose beads with a modified Pharmacia protocol (available upon request).
PTPase Enzymatic Assays
Phosphatase assays using the tyrosine-phosphorylated peptide ENDYINASL as a substrate12 were performed in 50 mmol/L MES, 5% glycerol, 0.1% β-mercaptoethanol, 0.1% Triton X-100, and 0.1% fatty acid–free BSA. Sample aliquots were diluted in assay buffer to a final volume of 20 μL and preincubated at 30°C for 5 minutes in the presence or absence of effectors. The assays were started by adding 10 μL of substrate adjusted to 3 μmol/L phosphotyrosine. The reaction was terminated after 5 minutes by adding 320 μL of stop solution (10% activated charcoal [vol/vol], 1% Celite (Sigma) [wt/vol] in 0.9N hydrochloric acid, 0.1 mol/L pyrophosphate, and 2 mmol/L sodium phosphate). The samples were microfuged for 3 minutes at 14 000 rpm, and a 250-μL aliquot of the supernatant was counted. Assays using p-NPP were carried out with 10 mmol/L substrate for 5 minutes at 30°C in 25 mmol/L sodium acetate (pH 5.0), 20% glycerol, 1 mmol/L dithiothreitol, and 1 mmol/L EDTA. The p-nitrophenol produced was quantified by absorbance at 410 nm in 0.2 mol/L sodium hydroxide. For all assays, one unit of activity is defined as 1 nmol phosphate released per minute.
Preparation of Antisera
Two peptides corresponding to the sequences TSGKPTYKNITTEPWPVS (peptide 1) and GTEGQPGNKVFKTNPIQ (peptide 2) in the extracellular domain of rDEP-1 were used to immunize rabbits (HTI-Bioproducts). Eight copies of each peptide were synthesized on a branched lysine core (Research Genetics) to avoid the need for coupling to a carrier protein. The antisera were affinity-purified by passage through cyanogen bromide–activated Sepharose columns coupled to the peptide antigens. The bound antibodies were eluted with 0.1 mol/L glycine (pH 2.5) and immediately neutralized to pH 8.5 by the addition of 1 mol/L Tris (pH 9.2). Four different antisera were obtained: antisera 5319 and 5320 were raised against peptide 1, and antisera 5374 and 5375 were raised against peptide 2.
To prepare the protein samples for the Western blot analysis, cultured cells were put on ice, washed with PBS, scraped, and pelleted at 3000g for 5 minutes. The cell pellet was resuspended in 2× Laemmli sample buffer (0.1 mol/L Tris [pH 6.8], 2% SDS, and 4% glycerol). Protein concentration was determined by the Micro BCA assay (Pierce). Bromophenol blue and pyronin Y were then added to the samples to a final concentration of 0.002%, and they were boiled for 10 minutes. The lysates were cleared by centrifugation at 14 000 rpm for 20 minutes, and 30 μg of protein was loaded per lane. The proteins were separated by SDS-PAGE under reducing conditions and electrophoretically transferred to nitrocellulose membranes. After transfer, the membranes were incubated in blocking solution (150 mmol/L sodium chloride, 10 mmol/L Tris [pH 8.0], 0.05% Tween-20, and 7.5% nonfat dry milk) for 1 hour. Incubations with both the primary and secondary antibodies were performed for 1 hour in blocking solution. Membranes were washed four times with TBST (150 mmol/L sodium chloride, 10 mmol/L Tris [pH 8.0], and 0.05% Tween-20) between antibody incubations. The immunocomplexes were detected by enhanced chemiluminescence (Amersham).
In Situ Hybridization
Tissues from adult Sprague-Dawley rats (B & K Universal, Kent, Wash) were fixed by perfusion for 10 minutes in 4% paraformaldehyde/0.1 mol/L sodium phosphate buffer (pH 7.4). Tissues were removed, cut into fragments, placed into the same fixative overnight, embedded in paraffin, and cut into 6- to 8-μm sections. E15-E17 mouse embryos from time-mated Swiss-Webster mice (B & K Universal, Kent, Wash) were dissected free of the uterine wall and amniotic membranes and immersion-fixed in 4% paraformaldehyde/0.1 mol/L sodium phosphate buffer (pH 7.4) for 1 to 3 days. They were then placed into 70% ethanol until being embedded in paraffin and cut into 6-μm sections.
To prepare an rDEP-1 riboprobe, a 1774-bp PCR fragment between nucleotides 262 and 2035 of the available rDEP-1 sequence was subcloned into the EcoRI site of the pGEM-3ZF+ vector (Promega). The cDNA was cloned in both directions, such that the sense and antisense probes could both be generated from the same promoter. After linearization of the DNA template, the antisense and sense riboprobes were transcribed from the T7 promoter and labeled by the incorporation of [α-33P]UTP (NEN). The riboprobes were then hydrolyzed to an average size of 150 nucleotides. The in situ hybridization was performed in hybridization buffer (50% formamide, 2× SSPE, 1× Denhardt's solution, 10% dextran sulfate, 20 mmol/L Tris [pH 7.4], 5 mmol/L EDTA, and 100 μg/mL yeast tRNA) containing 50×106 cpm/mL of 33P-labeled rDEP-1 riboprobe. The sections were incubated overnight at 60°C, then washed for 10 minutes at 65°C in 50% formamide, 5 mmol/L EDTA, and 20 mmol/L Tris (pH 7.4), and rinsed in 2× SSC. The sections were exposed to 20 μg/mL RNase A in 10 mmol/L Tris (pH 7.1), 1 mmol/L EDTA, and 0.5 mol/L sodium chloride for 30 minutes at 37°C. They were then rinsed at room temperature in the same buffer, but without RNase A. A high-stringency wash at 65°C (as described above) was repeated, and the sections were washed for 30 minutes in 2× SSC with gentle agitation. The sections were then dehydrated, air-dried, and dipped in Kodak NTB2 emulsion. The emulsion-coated slides were stored in the dark for 6 weeks, developed in a Kodak D19 developer, and counterstained with hematoxylin and eosin.
For studies of rDEP-1 expression after deendothelialization of the vessel wall, the aortas of male Sprague-Dawley rats (400 g, 3 to 4 months old, Bantin & Kingman, Edmonds, Wash) were denuded with a 2F balloon catheter as previously described.13 Eight days after ballooning, the animals were killed, and the aortas were removed, cut open longitudinally, and pinned out flat on polytetrafluoroethlylene (Teflon) cards. En face preparations and processing of the aortas for in situ hybridization were performed as described by Lindner and Reidy.14
Sequence and Deduced Structure of rDEP-1
We used PCR and PTPase-degenerate primers to amplify five different partial PTPase cDNA sequences from cultured rat vascular SMCs (WKY-12) (L.G. Borges and D.F. Bowen-Pope, unpublished data, 1996). One of these cDNAs, rDEP-1, was then radiolabeled and used to screen a random-primed cDNA library made from RNA isolated from cultured SMCs (WKY-12) derived from the aortas of 12-day-old rats. Since Northern blot analysis had shown that the rDEP-1 probe hybridized to a large transcript (8.7 kb), the cDNA used to make the cDNA library was size-selected to enrich for cDNAs larger than 4 kb. Our initial screen of 6.7×105 plaques yielded 20 positive clones. The largest cDNA insert (6.4 kb) was used as a template for DNA sequencing. Sequence analysis revealed that the methionine encoded by nucleotides 147 to 149 is flanked by nucleotides in good agreement with the Kozak initiation consensus sequence.15 The open reading frame initiated at this methionine would encode a translation product of 1216 amino acids. A 20–amino acid hydrophobic stretch, which begins 19 amino acids downstream from the initiator methionine, probably functions as a signal peptide. After cleavage of the putative signal peptide, the predicted mature rDEP-1 protein is organized as a receptor-like PTPase, with eight fibronectin type III repeats in the extracellular compartment, a single transmembrane domain (amino acids 851 to 873), and a single catalytic domain (amino acids 939 to 1175) in the intracellular compartment (Fig 1⇓). In addition to the catalytic domain, the intracellular sequence includes potential phosphorylation sites for protein kinase A (amino acid 1127), Ca2+-calmodulin–dependent protein kinase II (amino acid 1127), and glycogen synthase kinase III (amino acid 1112).
The cDNA sequence of a probable human homologue (DEP-1/HPTPη) has been published by two groups.6 16 Using different criteria to define the fibronectin type III repeats, Ostman et al6 concluded that there were 8 repeats, and Honda et al16 concluded that there were 10. Both groups agree on the identification of the eight N-terminal repeats, but Honda et al define two additional repeats proximal to the transmembrane region. These last two repeats lack several conserved amino acids, including the invariant tyrosine and tryptophan within the fibronectin type III core (corresponding to positions 22 and 68 within the 10th repeat of fibronectin) and residues within the conserved sequence LXPG.17 18 Since rDEP-1 also lacks several of these conserved residues, we have chosen to define eight fibronectin type III domains. In rDEP-1, the third fibronectin type III repeat appears to be truncated at the N-terminal end. However, the conserved tyrosine and tryptophan residues are present as well as the leucine and proline of the LXPG loop. Sequence comparison between the rDEP-1 (rat) sequence and the DEP-1/HPTPη (human) sequence reveals that there is good sequence alignment only after the second fibronectin type III repeat. At the N-terminus of this repeat, the sequences diverge. In this region, the human protein contains ≈120 more amino acids than the rat (rDEP-1) protein. These differences could either be due to true sequence divergence or to alternative splicing of the RNA messages.
On the basis of the deduced amino acid sequence, the mature rDEP-1 protein backbone would be predicted to have a molecular mass of 134 kD. Since there are 33 potential N-linked glycosylation sites spread throughout the putative extracellular region, the molecular mass of the mature protein is likely to be considerably larger. We prepared four rabbit antisera against two nonoverlapping peptides in the extracellular domain and used these antisera to detect the mature rDEP-1 protein on Western blots. In Western blots, all four affinity-purified antisera recognized an rDEP-1 pGEX fusion protein containing the two peptide epitopes (data not shown). In cell lysates from rat SMCs and platelets, all anti–rDEP-1 antisera recognize two bands of ≈180 and 220 kD (Fig 2⇓). These two forms of the rDEP-1 protein might be the result of alternative splicing of the RNA or differences in posttranslational processing. As observed for the rDEP-1 transcript levels, the rDEP-1 protein levels are significantly higher in cultures of newborn SMCs than adult SMCs.
We mapped the rDEP-1 gene on mouse and human metaphase chromosomes by FISH using a 6.4-kb rat cDNA probe. Of 84 mouse cells examined, 46 (54.8%) showed signals on both chromatids of one or both chromosomes 2 at region E, most likely E2, and 7 (8.3%) showed signals on one chromatid of one or both chromosomes at this locus. There was a second, much weaker, site of hybridization to chromosome 1 at region E. Three cells showed signals on both chromatids, and 7 (8.3%) showed signals on a single chromatid of one or both chromosomes 1 at region E. Of 117 human cells examined, 15 (12.8%) showed signals on both chromatids of one or both chromosomes 11 at band p11.2, and 24 (20.5%) showed signals on a single chromatid of one or both chromosomes at this locus. There was a second, much weaker, site of hybridization to chromosome 12 at band q13. Three cells (2.6%) showed signals on both chromatids, and 2 (1.7%) showed signals on a single chromatid of one or both chromosomes at this locus.
We confirmed the FISH mapping of rDEP-1 to human chromosome 11 by Southern blot analysis of genomic DNA extracted from a panel of rodent/human somatic cell hybrids (Fig 3⇓). Each somatic cell hybrid carries a different subset of the human chromosomes. We detected a hybridization signal to human genomic DNA on a single cell line, which carried human chromosomes 11 and 5. Since none of 13 different cell hybrids carrying the complete or part of human chromosome 5 produced a hybridization signal, this confirms that the rDEP-1 gene is located on human chromosome 11. The source of the much weaker hybridization signals detected by FISH is not clear. The signal from human chromosome 12q13 was not detected in the somatic cell hybrid analysis. It is possible that they represent the locations of related PTPase genes.
Characterization of the Enzymatic Activity of rDEP-1
In order to perform a preliminary analysis of the enzymatic activity of the putative PTPase domain of rDEP-1, we expressed the complete cytoplasmic domain of rDEP-1 in bacteria as a fusion protein with GST and purified it on glutathione-Sepharose. The enzymatic activity of the GST–rDEP-1 cytoplasmic domain fusion protein was evaluated in vitro using two artificial substrates, p-NPP and the tyrosine-phosphorylated peptide ENDYINASL. Both of these substrates have been widely used by others in characterizing PTPase activities. After establishing the limits of linearity for both the enzyme concentration and reaction time, the specific activity of rDEP-1 against both substrates was determined in six replicate experiments using two different preparations of the fusion protein. The specific activities of rDEP-1 toward ENDYINASL and p-NPP are 4500 and 0.568 U/mg, respectively. Considering that the Vmax of the enzyme was probably not reached under the assay conditions used, the Vmax of rDEP-1 toward ENDYINASL is at least threefold to fourfold higher than those reported for the transmembrane PTPases CD-45 and RPTPα.12
To further characterize the enzymatic activity of rDEP-1, we evaluated the effect of various agents on the catalytic activity of the fusion protein. As shown in the Table⇓, the PTPase activity of rDEP-1 is completely inhibited by micromolar concentrations of sodium vanadate and sodium molybdate, which are well-established inhibitors of PTPases, and by millimolar concentrations of sodium fluoride. Other strong inhibitors include zinc acetate (1 mmol/L), heparin (1 to 10 μg/mL), and poly–Glu-Tyr (1 to 10 μmol/L), a random copolymer of glutamate and tyrosine at a ratio of 1 to 4. Tetramisole (an inhibitor of alkaline phosphatases) and okadaic acid (a potent inhibitor of type 1 and 2A Ser-Thr phosphatases) were largely without effect on the catalytic activity of rDEP-1 fusion protein. No strong activators of the rDEP-1 fusion protein were found in the present study.
Modulation of rDEP-1 Expression With Cell Density
Given the presence of fibronectin type III motifs in rDEP-1, we evaluated the relationship between rDEP-1 transcript level and population density and cell morphology. rDEP-1 transcript levels are 10.5-fold higher in confluent than in sparse cultures of newborn Wistar-Kyoto SMCs (WKY-12) (Fig 4⇓). In another newborn SMC line derived from the Sprague-Dawley strain of rats, rDEP-1 transcript levels are 5-fold higher in confluent than in sparse cultures. By contrast, adult SMC lines derived from both Wistar-Kyoto (WKY-3M) and Sprague-Dawley (SD-3M) rats have low levels of rDEP-1 transcripts at both population densities.
Even though both newborn and adult lines were derived from the same blood vessel (aorta) using the same protocol, they have very different morphologies and behavior in culture. The adult SMCs are spindle-shaped fibroblast-like cells and grow in culture with extensive overlapping. The newborn SMCs are flattened epithelioid cells and grow in culture as a monolayer. These differences in morphology are not unique to these particular isolates of vascular SMCs, since the same differences are seen in many other isolates of vascular SMCs.19 20 Since the newborn SMC cultures closely resemble vascular endothelial cells, we evaluated rDEP-1 transcript levels in cultured bovine endothelial cells. We used bovine cells for this one cell type because well-characterized rat endothelial cells from large vessels were not available. There is a ninefold upregulation of rDEP-1 transcript levels as cell density is increased from sparse to confluent (Fig 4⇑).
We have also used the newborn and the adult SMC lines to determine whether they express and modulate the rDEP-1 protein levels at different population densities (Fig 2⇑). We found that adult SMCs express low levels of rDEP-1 protein at all population densities. In contrast, the newborn cells express higher levels of the rDEP-1 protein than do the adult cells, and these levels do not change significantly when the population density is increased.
rDEP-1 Expression In Vivo
We evaluated rDEP-1 expression in vivo both by Northern blot analysis and in situ hybridization. Northern blot analysis reveals that rDEP-1 is expressed at various levels in several different tissues in vivo (Fig 5⇓). rDEP-1 transcripts are particularly high in the cerebellum, brain cortex, and kidney cortex and somewhat less abundant in spleen and lung. Interestingly, rDEP-1 is expressed at very low levels in all of the muscle tissues studied, which included heart, diaphragm, and skeletal muscle. We used in situ hybridization to further localize the rDEP-1 transcripts within each organ system. Although rDEP-1 transcripts are expressed by many cell types in vivo, it does appear that the highest expression levels are present in epithelial cells. For example, seminal vesicle (Fig 6A⇓ and 6B) and prostate (data not shown) epithelia express rDEP-1 at very high levels. In contrast, the connective tissue stroma surrounding the seminal vesicle epithelia express very little rDEP-1. rDEP-1 expression in vivo is not high in all epithelial cells. For example, distinct epithelia in the kidney express different levels of rDEP-1. Epithelial cells of the proximal tubules and the visceral and parietal epithelia of the glomerulus express relatively little rDEP-1, whereas distal tubule epithelia express rDEP-1 at much higher levels (Fig 7F⇓).
Epithelial cells are not the only cells that express rDEP-1 transcripts at very high levels in vivo. rDEP-1 transcripts are very abundant in megakaryocytes in all locations at which this cell type is found, including adult rat spleen (Fig 6C⇑), bone marrow, and embryonic mouse liver (data not shown). High expression of rDEP-1 transcripts is also seen in the marginal zone of the spleen (data not shown). We have not identified the cell type(s) expressing rDEP-1 in the splenic marginal zone, but neither they nor the megakaryocytes have an epithelioid morphology. rDEP-1 expression is detected in several other tissues, including the liver, lung, placenta, and central nervous system (data not shown).
rDEP-1 was cloned from vascular SMCs, and we were thus particularly interested in determining its pattern of expression in the cardiovascular system. Aortic medial SMCs express very low to undetectable levels of rDEP-1, at least as measured by in situ hybridization (Fig 7A⇑). However, rDEP-1 expression is readily detected in vascular endothelial cells (Fig 7A and 7B⇑⇑). rDEP-1 is expressed in endothelial cells of all muscular arteries including the aorta (Fig 7A and 7B⇑⇑), large arteries in the spleen (Fig 7D⇑), and central arterioles and arteries in the white pulp of the spleen (Fig 7C and 7G⇑⇑). Expression of rDEP-1 in endothelial cells is only detectable in vessels composed of both SMCs and endothelium. In capillaries composed of only endothelial cells, we have not been able to detect any rDEP-1 expression by in situ hybridization. For example, in the kidney glomerulus, rDEP-1 expression is not observed in the endothelial cells of the capillary network (Fig 7F⇑). In contrast, endothelial cells lining the central arteriole of the spleen, which contains a single SMC in the section shown in Fig 7C⇑, are heavily labeled.
rDEP-1 Expression in Response to Vascular Injury
We have observed that rDEP-1 expression in cultured endothelial cells is regulated in parallel with population density (Fig 4⇑). To determine whether expression is regulated in vivo, we evaluated rDEP-1 expression by rat aortic endothelial cells as they migrate and proliferate in vivo to cover a segment of aorta deendothelialized using a balloon catheter. In this model, endothelium from adjacent nondenuded regions (including intercostal arteries) advances as a monolayer to cover the denuded region. Cross sections of vessels reveal only a thin monolayer of endothelium at a single plane of section and are not well suited for evaluation of endothelial cell phenotype as a function of position relative to the leading edge. To facilitate this, we opened the vessels longitudinally, pinned them out flat to expose the vessel surface, and evaluated expression en face by in situ hybridization as described by Lindner and Reidy.14 We observed that rDEP-1 is strongly expressed in the endothelial monolayer, which has formed by 8 days after injury. However, rDEP-1 expression is not uniform over the whole monolayer, and in particular, rDEP-1 expression frequently appears to be downregulated in migrating and proliferating endothelial cells at the advancing leading edge of the endothelial monolayer (Fig 7E⇑).
Expression of rDEP-1 During Megakaryopoiesis and in Platelets
The high level of rDEP-1 expression in fetal and adult megakaryocytes detected by in situ hybridization led us to analyze rDEP-1 expression during megakaryocyte differentiation in vitro, where changes in transcript levels can be quantified. We used the human megakaryoblastic cell line CHRF-288-11,8 which grows in suspension (or very loosely adherent) until induced to differentiate by exposure to PMA. Within minutes after PMA treatment, the cells begin to adhere to the culture dish and begin to upregulate expression of transcripts for proteins that will eventually be present in platelets. rDEP-1 transcripts are detected within 4 hours after PMA treatment, and a high level of expression is maintained for at least 24 hours (Fig 8⇓). Although megakaryoblast cell lines do not complete the process of cytoplasmic fragmentation and platelet formation in vitro, we have found that circulating rat platelets contain very high levels of rDEP-1 protein detected by Western blot analysis (Fig 2⇑).
We have identified and initially characterized a rat PTPase sequence, rDEP-1, which is the putative homologue of the human PTPase DEP-1/HPTPη.6 16 The rDEP-1 cDNA encodes a receptor-like PTPase with a single catalytic domain in the intracellular region, a single transmembrane domain, and eight fibronectin type III repeats in the extracellular region. In addition to inferences from the deduced structure of rDEP-1, our data indicate that rDEP-1 may be involved in cell-cell or cell-matrix interactions. rDEP-1 transcript levels are strongly upregulated by population density in cultures of newborn SMCs and bovine aortic endothelial cells. Both of these cell types form an epithelioid monolayer in culture. In vivo, as evaluated by in situ hybridization, rDEP-1 tends to be expressed at higher levels in epithelial structures than in connective tissues. Although rDEP-1 may be involved in establishing/maintaining epithelial monolayers, we have not been able to establish a simple correlation between rDEP-1 expression and a specific physiological function(s) for rDEP-1 in epithelium.
rDEP-1 transcript levels are strongly upregulated in epithelioid cells as population density increases. However, the rDEP-1 protein levels do not differ significantly between low- and high-density conditions. For instance, in the newborn SMCs (WKY-12 cells), we have observed differences as high as 21-fold between transcript levels in low- and high-density cultures with only very small differences in protein levels (zero to twofold difference). One possible explanation for this disparity would be that rDEP-1 protein is degraded more rapidly in confluent cultures and that this shortened half-life balances the increased rate of synthesis driven by the higher transcript levels. The rate of internalization and degradation of transmembrane receptors with tyrosine kinase activity is characteristically increased when the receptors are occupied by ligand, and in at least some cases, this decrease in protein level is accompanied by upregulation of transcript level.21 If the receptor-like rDEP-1 protein were behaving like a tyrosine kinase receptor, this would indicate that it was more likely to be bound to a ligand at higher cell density, consistent with a role for rDEP-1 in mediating some aspect of density-dependent regulation of cell function. Evidence for the participation of PTPases in cell-cell or cell-matrix adhesive interactions has been accumulating. Homophilic interactions have been demonstrated for the PTPases RPTPμ22 23 and RPTPκ.24 More recently, it was shown that another PTPase, RPTPβ, is capable of interacting heterophilically with the extracellular matrix molecule tenascin25 and with the axon membrane protein contactin.26
Another indication that rDEP-1 might be involved in cell adhesion mechanisms comes from the observation that both rDEP-1 and its human homologue HPTPη/DEP-1 contain the tripeptide sequence EVT in their extracellular domains. This same amino acid sequence is present in the T-cadherin adhesion molecule, replacing the HAV tripeptide that is present in other cadherins.27 A glutamic acid residue is also present four residues upstream from the EVT sequence in the rDEP-1, DEP-1/HPTPη, and T-cadherin molecules. The HAV tripeptide is conserved in N-, P-, and E-cadherins and in the L-CAM molecule and has been shown to be involved in determining the binding specificity of these molecules.28 Peptides containing this sequence block embryo compaction, a process known to require cadherin function.28 Mutations in amino acids flanking the HAV sequence alter the binding specificity of E-cadherin.29 The EVT sequence in rDEP-1 could thus play a role in the interaction between rDEP-1 and its ligand/counter-receptor. Since rDEP-1 and T-cadherin are both highly expressed in the brain, it is also possible that they could interact with each other via the EVT sequences.
The rDEP-1 gene is located on human chromosome 11 (band p11.2) (see also Honda et al16 ) and on mouse chromosome 2 (region 2E). The mapping of rDEP-1 to these loci confirms a region of genetic conservation between human chromosome 11p and mouse chromosome 2.30 Loss of heterozygosity or deletion of sequences in the short arm of human chromosome 11 have been detected in various tumors of epithelial origin, including breast cancer,31 transitional cell carcinoma of the bladder,32 hepatoblastoma,33 and hepatocellular carcinoma.34 Since rDEP-1 shows the strongest level of expression in epithelial tissues and since it encodes a PTPase activity that could potentially be involved in some aspects of cell adhesion and signaling, it will be important to determine whether rDEP-1 has a role in epithelial oncogenesis.
In situ hybridization studies of vascular tissue detect the highest level of rDEP-1 expression in the endothelial cells. To evaluate rDEP-1 expression in endothelial cells in vivo during response of the vessel wall to balloon injury, we injured the rat aorta. After ballooning of the aorta, the regenerating endothelium from the branching intercostal arteries advances as a coherent monolayer onto the uninjured region area of the aorta. By examining the surface of the aorta en face by in situ hybridization, we can easily evaluate rDEP-1 expression as a function of position relative to the leading edges of the endothelium as it advances through the injured area. rDEP-1 expression is frequently downregulated at the leading edge of the endothelium, where cells are least crowded and are most actively replicating. Behind this zone, where the endothelial cells have stopped replicating and became more tightly packed, rDEP-1 expression returns to the level seen in uninjured endothelium. This pattern seems analogous to the dependence on population density observed with cultured bovine endothelial cells and is consistent with the possibility that rDEP-1 is involved in regulating adhesive interactions and/or contact-dependent mitogenic signaling pathways in endothelial cells. In any of these cases, the use of rDEP-1 agonists/antagonists would provide a potential method to modulate endothelial cell behavior in vivo.
The platelet is the third cell type most extensively investigated as a component in the response of the vessel to injury. Most of the proteins that function in platelets are synthesized by the polyploid megakaryocyte progenitors, which fragment to generate the platelets. Our in situ hybridization survey indicated that megakaryocytes are one of the richest sources of rDEP-1 transcript in vivo. In culture, the human megakaryoblastic cell line CHRF-288-11 can be induced to undergo partial differentiation by treatment with PMA. Upon PMA treatment, CHRF-288-11 cells rapidly attach to the culture dish and undergo morphological and DNA ploidy changes similar to those seen during megakaryocytic differentiation in vivo.8 These changes are accompanied by the appearance of rDEP-1 transcripts. It is possible that rDEP-1 plays some role in the megakaryocyte differentiation program or that rDEP-1 is expressed during this period because it needs to be provided for subsequent function in the mature biosynthetically inactive platelet. In favor of the latter possibility, Western blot analysis indicates that rDEP-1 protein is very abundant in circulating platelets. Platelets are also very rich in tyrosine kinases, which have been shown to be part of signal transduction pathways required for platelet activation and aggregation.35 36 PTPase activity is also present in platelets, with ≈80% of it associated with membranes.37 Studies with the PTPase inhibitor vanadate have indicated that blocking platelet PTPases induces Tyr phosphorylation and platelet activation,35 36 thus suggesting that PTPases might play critical roles in platelet function. Given the very high expression level of rDEP-1 in circulating platelets, we postulate that rDEP-1 may play a role in such signaling pathways.
Further insight into the role of rDEP-1 in cell types evaluated above will depend on obtaining biochemical information about the cytoplasmic components with which rDEP-1 interacts and on identifying the ligand/counter-receptor, which interacts with the extracellular domain of rDEP-1. These studies are in progress.
Selected Abbreviations and Acronyms
|FISH||=||fluorescence in situ hybridization|
|PCR||=||polymerase chain reaction|
|PMA||=||phorbol 12-myristate 13-acetate|
|PTPase||=||protein tyrosine phosphatase|
|SD-18, SD-3M||=||cell lines derived from 18-day-old and 3-month-old Sprague-Dawley rats|
|SMC||=||smooth muscle cell|
|WKY-12, WKY-3M||=||cell lines derived from 12-day-old and 3-month-old Wistar-Kyoto rats|
This study was supported by National Institutes of Health grant HL-18645. Dr Borges was supported by the fellowships “Cieˆncia” (BD/451/90-ID) and “PRAXIS XXI” (BD/3850/94) administrated by the Portuguese government through the Junta Nacional de Investigac¸a˜o Cienti´fica e Tecnolo´gica, Lisbon, Portugal. We would like to thank Monica Vuong and Joyce Murphy for their valuable technical assistance and Dr Charles Murry for fruitful discussions during the course of this project.
- Received February 26, 1996.
- Accepted June 3, 1996.
Lindner V, Reidy MA. Proliferation of smooth muscle cells after vascular injury is inhibited by an antibody against basic fibroblast growth factor. Proc Natl Acad Sci U S A. 1991;88:3739-3743.
Xiao S, Rose DW, Sasaoka T, Maegawa H, Burke TR Jr, Roller PP, Shoelson SE, Olefsky JM. Syp (SH-PTP2) is a positive mediator of growth factor-stimulated mitogenic signal transduction. J Biol Chem. 1994;269:21244-21248.
Milarski KL, Saltiel AR. Expression of catalytically inactive Syp phosphatase in 3T3 cells blocks stimulation of mitogen-activated protein kinase by insulin. J Biol Chem. 1994;269:21239-21243.
Ostman A, Yang Q, Tonks NK. Expression of DEP-1, a receptor-like protein-tyrosine-phosphatase, is enhanced with increasing cell density. Proc Natl Acad Sci U S A. 1994;91:9680-9684.
Fugman DA, Witte DP, Jones CL, Aronow BJ, Lieberman MA. In vitro establishment and characterization of a human megakaryoblastic cell line. Blood. 1990;75:1252-1261.
Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A. 1977;74:5463-5467.
Edelhoff S, Ayer DE, Zervos AS, Steingrimsson E, Jenkins NA, Copeland NG, Eisenman RN, Brent R, Disteche CM. Mapping of two genes encoding members of a distinct subfamily of MAX interacting proteins: MAD to human chromosome 2 and mouse chromosome 6, and MXI1 to human chromosome 10 and mouse chromosome 19. Oncogene. 1994;9:665-668.
Lindner V, Reidy MA. Expression of basic fibroblast growth factor and its receptor by smooth muscle cells and endothelium in injured rat arteries: an en face study. Circ Res. 1993;73:589-595.
Kozak M. The scanning model for translation: an update. J Cell Biol. 1989;108:229-241.
Honda H, Inazawa J, Nishida J, Yazaki Y, Hirai H. Molecular cloning, characterization, and chromosomal localization of a novel protein-tyrosine phosphatase, HPTPη. Blood. 1994;84:4186-4194.
Leahy DJ, Hendrickson WA, Aukhil I, Erickson HP. Structure of a fibronectin type III domain from tenascin phased by MAD analysis of the selenomethionyl protein. Science. 1992;258:987-991.
Majesky MW, Giachelli CM, Reidy MA, Schwartz SM. Rat carotid neointimal smooth muscle cells reexpress a developmentally regulated mRNA phenotype during repair of arterial injury. Circ Res. 1992;71:759-768.
Earp HS, Austin KS, Blaisdell J, Rubin RA, Nelson KG, Lee LW, Grisham JW. Epidermal growth factor (EGF) stimulates EGF receptor synthesis. J Biol Chem. 1986;261:4777-4780.
Brady Kalnay SM, Flint AJ, Tonks NK. Homophilic binding of PTP mu, a receptor-type protein tyrosine phosphatase, can mediate cell-cell aggregation. J Cell Biol. 1993;122:961-972.
Gebbink MF, Zondag GC, Wubbolts RW, Beijersbergen RL, van Etten I, Moolenaar WH. Cell-cell adhesion mediated by a receptor-like protein tyrosine phosphatase. J Biol Chem. 1993;268:16101-16104.
Sap J, Jiang YP, Friedlander D, Grumet M, Schlessinger J. Receptor tyrosine phosphatase R-PTP-kappa mediates homophilic binding. Mol Cell Biol. 1994;14:1-9.
Barnea G, Grumet M, Milev P, Silvennoinen O, Levy JB, Sap J, Schlessinger J. Receptor tyrosine phosphatase beta is expressed in the form of proteoglycan and binds to the extracellular matrix protein tenascin. J Biol Chem. 1994;269:14349-14352.
Peles E, Nativ M, Campbell PL, Sakurai T, Martinez R, Lev S, Clary DO, Schilling J, Barnea G, Plowman GD, Grumet M, Schlessinger J. The carbonic anhydrase domain of receptor tyrosine phosphatase β is a functional ligand for the axonal cell recognition molecule contactin. Cell. 1995;82:251-260.
Copeland NG, Jenkins NA, Gilbert DJ, Eppig JT, Maltais LJ, Miller JC, Dietrich WF, Weaver A, Lincoln SE, Steen RG, Stein LD, Nadeau JH, Lander ES. A genetic linkage map of the mouse: current applications and future prospects. Science. 1993;262:57-66. Comments.
Ali IU, Lidereau R, Theillet C, Callahan R. Reduction to homozygosity of genes on chromosome 11 in human breast neoplasia. Science. 1987;238:185-188.