Specific interaction of SPARC with endothelial cells is mediated through a carboxyl-terminal sequence containing a calcium-binding EF hand.

SPARC is a secreted, Ca(2+)-binding protein that modulates cell shape and gene expression (Sage, E.H., and Bornstein, P. (1991) J. Biol. Chem. 266, 14831-14834). In the present study we questioned whether SPARC interacted with an endothelial cell surface receptor. The binding of 125I-SPARC to bovine aortic endothelial cells was dependent on Ca2+ and was sensitive to small changes in extracellular pH; maximal binding occurred at pH 7.1. Scatchard analysis indicated approximately 2.3 x 10(7) binding sites/cell with an apparent KI of 1.1 nM. The interaction was diminished specifically by competition with synthetic peptides corresponding to amino acids 54-73 (SPARC 54-73) and 254-273 (SPARC254-273). The binding of 125I-SPARC254-273, a sequence containing a Ca(2+)-binding EF-hand, was saturated within 45 min at a concentration of 5 microM; Scatchard analysis indicated 4.2 x 10(7) sites/cell and a KI of 2.4 nM. Iodinated proteins from plasma membranes were affinity-chromatographed on SPARC254-273; several proteins with apparent masses ranging from 153 to 100 kDa (unreduced) or from 153 to 122 kDa (reduced) were eluted with the soluble peptide. These proteins represent candidates for a SPARC receptor(s) that mediates the biological activity of this protein on endothelial cells.

Age-related and site-specific adaptation of the arterial endothelial cytoskeleton during atherogenesis.

The authors probed the vascular endothelial cell cytoskeleton in strains of pigeons that are atherosclerosis-susceptible and disease-resistant, namely, the White Carneau and Show Racer pigeons. Endothelial cell actin and myosin were localized with the use of affinity-purified antibodies in conjunction with indirect immunofluorescence microscopy. The endothelial cell cytoskeleton was characterized in a site-specific and time-dependent manner by examination of arterial segments from each strain of pigeons. Anti-actin and anti-myosin fluorescence staining patterns of endothelial cells lining the ascending aorta, aortic arch, and thoracic aorta from the White Carneau and Show Racer pigeons sacrificed at 1 and 12 months of age were compared and analyzed. In the Show Racer, irrespective of arterial site or chronologic age, endothelial cell cytoskeletal organization is similar. Actin and myosin fluorescence is brightest at the cortex, where endothelial cells meet their neighbors. There is also an amorphous (diffuse) fluorescence throughout the cytoplasm. In addition to the diffuse and cortical cytoskeletal fluorescence in the endothelial cells of the Show Racers, the White Carneau also possess a unique cytoskeletal array of linear fluorescence, ie, the endothelial cell ridge. At 1 month of age, anti-actin staining of endothelial cell ridges averages 28.5 mu in length in the ascending aorta, 28.0 mu in the aortic arch, and 40.0 mu in the thoracic aorta. At the same time, anti-myosin fluorescence extends past both ends of the anti-actin-stained endothelial cell ridge fluorescence. In the ascending aorta, anti-myosin labeling of endothelial cell ridges is 3.5 times longer than anti-actin staining. This staining is absent in the aortic arch, whereas the thoracic aorta possesses endothelial cell ridges that extend over the entire length of the vessel segment. At 12 months of age, actin-stained endothelial cell ridges increase 1.6- and 1.4-fold in the ascending aorta and aortic arch, respectively. The thoracic aorta possesses endothelial cell ridges that cover its entire length. At 12 months of age, the length of myosin-stained endothelial cell ridges does not increase in the ascending or thoracic aorta. In contrast, the aortic arch expresses endothelial cell ridges that exceed 150 mu in length. It is proposed that the endothelial cell ridge assembles from cytoskeletal components as a focal endothelial cell response to injury, perhaps promoting endothelial cell adhesion to the underlying basal lamina through a transmembrane linkage.

Substratum-induced stress fiber assembly in vascular endothelial cells during spreading in vitro.

We tested whether aortic endothelial cell (EC)-synthesized substrata, which modulate smooth muscle cell proliferation and EC motility following injury, could influence EC actin cytoskeleton and spreading in vitro. A partial characterization of the substrata indicates that the substratum prepared by deoxycholic acid extraction (DOC-derived substratum) is enriched with fibronectin and type IV collagen. Substratum prepared by removal of the intact monolayer with 20 mM EGTA in PBS (EGTA-derived substratum) contains fibronectin and heparan sulfate proteoglycan, but no type IV collagen. Morphometric analyses were performed on fixed and cytoskeletal antibody treated EC in order to quantitate the extent of spreading and stress fiber (SF) assembly. Compared to plastic, the DOC-derived substratum, a collagenase-treated DOC-derived substratum (CT-DOC-derived substratum) and the EGTA-derived substratum promote EC spreading 2.3-, 2.9- and 1.7-fold, respectively. In addition, there are 4.2-, 4.1- and 2.0-fold more SF on DOC-, CT-DOC- and EGTA-derived substrata, respectively, when compared to plastic. Subcellular fractionation and immunoprecipitation of cytoskeletal proteins from metabolically labeled EC were performed prior to electrophoresis and fluorography. The DOC-derived substratum increases immunoprecipitable actin and myosin 3- to 4.5-fold in both fractions compared to the EGTA-derived substratum and plastic. Collagenase treatment of the DOC-derived substratum partially inhibits this increase. Cycloheximide treatment prevents the rise in soluble actin and myosin as well as causing a reduction in SF number by 1/2 on the DOC-derived substratum and 2/3 on CT-DOC-derived substratum. We propose that fibronectin-collagen interactions are, in part, responsible for inducing endothelial synthesis of cytoskeletal proteins required for SF assembly. This substratum-induced actin-cytoskeletal reorganization facilitates EC spreading in vitro.

Purification of biologically active SPARC expressed in Saccharomyces cerevisiae.

SPARC (secreted protein, acidic and rich in cysteine) is a secreted, Ca+2-binding glycoprotein that modulates interactions between cells and their immediate extracellular matrix. Traditional sources of SPARC have been mammalian bone, platelets, a basement membrane tumor, and cultured cells; most if not all preparations, however, contain platelet-derived growth factor and one or more serum proteins that bind specifically to purified SPARC. To avoid these contaminants, as well as the toxic lipid moiety associated with endotoxin, we expressed recombinant wild-type and a mutated murine SPARC in two strains of Saccharomyces cerevisiae: one strain was transfected with an expression vector encoding a proprietory signal peptide that directed the secretion of the recombinant protein. Recombinant SPARC was also purified from cell lysates of a different, nonreverting strain of S. cerevisiae that was optimized for large-scale fermentation runs. A mutant murine SPARC lacking the single glycosylation site was also expressed following substitution of Asn98 with Asp98 in the wild-type sequence. Purification of SPARC was achieved by copper-affinity and hydrophobic-interaction chromatography. Both the wild-type and the glycosylation-defective recombinant proteins exhibited high levels of activity in two bioassays with endothelial cells: inhibition of cell spreading/disruption of actin microfilaments and competition for the binding of nonrecombinant 125I-labeled SPARC to the cell surface. The availability of biologically active, recombinant SPARC will facilitate investigation of the structural and functional properties of this protein, which is expressed at high levels in healing wounds, atherosclerotic plaque, and several cancers and diseases of connective tissue.

Aortic endothelial and smooth muscle histamine metabolism. Relationship to aortic 125I-albumin accumulation in experimental diabetes.

We studied rat aortic endothelial and smooth muscle cell de novo histamine synthesis mediated by histidine decarboxylase (HD) and the effects of its inhibition by alpha-hydrazinohistidine on the intracellular histamine content and intraaortic albumin accumulation in streptozotocin-induced diabetes. Diabetes was induced by a single jugular vein injection of streptozotocin (60 mg/kg, pH 4.5, ether anesthesia), with animals held 4 weeks following the overt manifestation of diabetes. Additional diabetic and nondiabetic rats received alpha-hydrazinohistidine (25 mg/kg, i.p. every 12 hours) during the last week; this had no effect on the severity of diabetes in any animal receiving streptozotocin. Data indicate that the aortic endothelial (EC) HD activity was increased more than 130% in the untreated diabetic group but was similar to control values in the diabetic group receiving alpha-hydrazinohistidine; similarily, the EC histamine content from diabetic aortas increased 127% over control values, but in EC from diabetic animals receiving alpha-hydrazinohistidine it was comparable to control values. Similar trends were observed for the subjacent aortic smooth muscle. In untreated diabetic animals the aortic 125I-albumin mass transfer rate was increased 60% over control values, while in diabetic animals receiving alpha-hydrazinohistidine the 125I-albumin mass transfer rate was essentially identical to controls. These data indicate that in streptozotocin diabetes there is an expansion of the inducible aortic histamine pool, and that this expansion is intimately related to the increased aortic albumin accumulation.

Inhibition of endothelial cell proliferation by SPARC is mediated through a Ca(2+)-binding EF-hand sequence.

SPARC (secreted protein, acidic and rich in cysteine, also known as osteonectin and BM-40) is a metal-binding glycoprotein secreted by a variety of cultured cells and characteristic of tissues undergoing morphogenesis, remodeling, and repair. Recently it has been shown that SPARC inhibits the progression of the endothelial cell cycle in mid-G1, and that a synthetic peptide (amino acids 54-73 of secreted murine SPARC, peptide 2.1) from a cationic, disulfide-bonded region was in part responsible for the growth-suppressing activity [Funk and Sage (1991): Proc Natl Acad Sci USA 88:2648-2652]. Moreover, SPARC was shown to interact directly with bovine aortic endothelial (BAE) cells through a C-terminal EF-hand sequence comprising a high-affinity Ca(2+)-binding site of SPARC and represented by a synthetic peptide (amino acids 254-273) termed 4.2 [Yost and Sage (1993): J Biol Chem 268:25790-25796]. In this study we show that peptide 4.2 is a more potent inhibitor of DNA synthesis that acts cooperatively with peptide 2.1 to diminish the incorporation of [3H]-thymidine by both BAE and bovine capillary endothelial (BCE) cells. At concentrations of 0.019-0.26 mM peptide 4.2, thymidine incorporation by BAE cells was decreased incrementally, relative to control values, from approximately 100 to 10%. Although somewhat less responsive, BCE cells exhibited a dose-responsive decrement in thymidine incorporation, with a maximal inhibition of 55% at 0.39 mM. The inhibitory effect of peptide 4.2 was essentially independent of heparin and basic fibroblast growth factor and was blocked by anti-SPARC peptide 4.2 IgG, but not by antibodies specific for other domains of SPARC. To identify residues that were necessary for inhibition of DNA synthesis, we introduced single amino acid substitutions into synthetic peptide 4.2 and tested their activities and cell-surface binding characteristics on endothelial cells. Two peptides displayed null to diminished effects in the bioassays that were concentration-dependent: peptide 4.2 K, containing an Asp258 --> Lys substitution, and peptide 4.2 AA, in which the two disulfide-bonded Cys (positions 255 and 271) were changed to Ala residues. Peptide 4.2 K, which failed to fulfill the EF-hand consensus formula, exhibited an anomalous fluorescence emission spectrum, in comparison with the wild-type 4.2 sequence, that was indicative of a compromised affinity for Ca2(+).(ABSTRACT TRUNCATED AT 400 WORDS)