Perlecan is essential for cartilage and cephalic development.

Perlecan, a large, multi-domain, heparan sulfate proteoglycan originally identified in basement membrane, interacts with extracellular matrix proteins, growth factors and receptors, and influences cellular signalling. Perlecan is present in a variety of basement membranes and in other extracellular matrix structures. We have disrupted the gene encoding perlecan (Hspg2) in mice. Approximately 40% of Hspg2-/- mice died at embryonic day (E) 10.5 with defective cephalic development. The remaining Hspg2-/- mice died just after birth with skeletal dysplasia characterized by micromelia with broad and bowed long bones, narrow thorax and craniofacial abnormalities. Only 6% of Hspg2-/- mice developed both exencephaly and chondrodysplasia. Hspg2-/- cartilage showed severe disorganization of the columnar structures of chondrocytes and defective endochondral ossification. Hspg2-/- cartilage matrix contained reduced and disorganized collagen fibrils and glycosaminoglycans, suggesting that perlecan has an important role in matrix structure. In Hspg2-/- cartilage, proliferation of chondrocytes was reduced and the prehypertrophic zone was diminished. The abnormal phenotypes of the Hspg2-/- skeleton are similar to those of thanatophoric dysplasia (TD) type I, which is caused by activating mutations in FGFR3 (refs 7, 8, 9), and to those of Fgfr3 gain-of-function mice. Our findings suggest that these molecules affect similar signalling pathways.

Dyssegmental dysplasia, Silverman-Handmaker type, is caused by functional null mutations of the perlecan gene.

Perlecan is a large heparan sulfate (HS) proteoglycan present in all basement membranes and in some other tissues such as cartilage, and is implicated in cell growth and differentiation. Mice lacking the perlecan gene (Hspg2) have a severe chondrodysplasia with dyssegmental ossification of the spine and show radiographic, clinical and chondro-osseous morphology similar to a lethal autosomal recessive disorder in humans termed dyssegmental dysplasia, Silverman-Handmaker type (DDSH; MIM 224410). Here we report a homozygous, 89-bp duplication in exon 34 of HSPG2 in a pair of siblings with DDSH born to consanguineous parents, and heterozygous point mutations in the 5' donor site of intron 52 and in the middle of exon 73 in a third, unrelated patient, causing skipping of the entire exons 52 and 73 of the HSPG2 transcript, respectively. These mutations are predicted to cause a frameshift, resulting in a truncated protein core. The cartilage matrix from these patients stained poorly with antibody specific for perlecan, but there was staining of intracellular inclusion bodies. Biochemically, truncated perlecan was not secreted by the patient fibroblasts, but was degraded to smaller fragments within the cells. Thus, DDSH is caused by a functional null mutation of HSPG2. Our findings demonstrate the critical role of perlecan in cartilage development.

Mammalian eyes and associated tissues contain molecules that are immunologically related to cartilage proteoglycan and link protein.

Monospecific antibodies to bovine nasal cartilage proteoglycan monomer and link protein were used to demonstrate that immunologically related molecules are present in the bovine eye and associated tissues. With immunofluorescence microscopy, reactions for both proteoglycan and link protein were observed in the sclera, the anterior uveal tract, and the endoneurium of the optic nerve of the central nervous system. Antibody to bovine nasal cartilage proteoglycan also reacted with some connective tissue sheaths of rectus muscle and the perineurium of the optic nerve of the central nervous system. Antibody to proteoglycan purified from rat brain cross-reacted with bovine nasal cartilage proteoglycan, indicating structural similarities between these proteoglycans. ELISA studies and crossed immunoelectrophoresis demonstrated that purified dermatan sulphate proteoglycans isolated from bovine sclera did not react with these antibodies but that the antibody to cartilage proteoglycan reacted with other molecules extracted from sclera. Two molecular species resembling bovine nasal link protein in size and reactivity with antibody were also demonstrated in scleral extracts: the larger molecule was more common. Antibody to link protein reacted with the media of arterial vessels demonstrating the localization of arterial link protein described earlier. Tissues that were unstained for either molecule included the connective tissue stroma of the iris, retina, vitreous body, cornea, and the remainder of the uveal tract. These observations clearly demonstrate that tissues other than cartilage contain molecules that are immunologically related to cartilage-derived proteoglycans and link proteins.

The incubation of laminin, collagen IV, and heparan sulfate proteoglycan at 35 degrees C yields basement membrane-like structures.

Three basement membrane components, laminin, collagen IV, and heparan sulfate proteoglycan, were mixed and incubated at 35 degrees C for 1 h, during which a precipitate formed. Centrifugation yielded a pellet which was fixed in either potassium permanganate for ultrastructural studies, or in formaldehyde for Lowicryl embedding and immunolabeling with protein A-gold or anti-rabbit immunoglobulin-gold. Three types of structures were observed and called types A, B, and C. Type B consisted of 30-50-nm-wide strips that were dispersed or associated into a honeycomb-like pattern, but showed no similarity with basement membranes. Immunolabeling revealed that type B strips only contained heparan sulfate proteoglycan. The structure was attributed to self-assembly of this proteoglycan. Type A consisted of irregular strands of material that usually accumulated into semisolid groups. Like basement membrane, the strands contained laminin, collagen IV, and heparan sulfate proteoglycan, and, at high magnification, they appeared as a three-dimensional network of cord-like elements whose thickness averaged approximately 3 nm. But, unlike the neatly layered basement membranes, the type A strands were arranged in a random, disorderly manner. Type C structures were convoluted sheets composed of a uniform, dense, central layer which exhibited a few extensions on both surfaces and was similar in appearance and thickness to the lamina densa of basement membranes. Immunolabeling showed that laminin, collagen IV, and proteoglycan were colocalized in the type C sheets. At high magnification, the sheets appeared as a three-dimensional network of cords averaging approximately 3 nm. Hence, the organization, composition, and ultrastructure of type C sheets made them similar to the lamina densa of authentic basement membranes.

Glomerular basement membrane proteoglycans are derived from a large precursor.

The basement membrane heparan sulfate proteoglycan produced by the Englebreth-Holm-Swarm (EHS) tumor and by glomeruli were compared by immunological methods. Antibodies to the EHS proteoglycan immunoprecipitated a single precursor protein (Mr = 400,000) from [35S]methionine-pulsed glomeruli, the same size produced by EHS cells. These antibodies detected both heparan sulfate proteoglycans and glycoproteins in extracts of unlabeled glomeruli and glomerular basement membrane. The proteoglycans contained core proteins of varying size (Mr = 150,000 to 400,000) with a Mr = 250,000 species being predominant. The glycoproteins are fragments of the core protein which lack heparan sulfate side chains. Antibodies to glomerular basement membrane proteoglycan immunoprecipitated the precursor protein (Mr = 400,000) synthesized by EHS cells and also reacted with most of the proteolytic fragments of the EHS proteoglycan. This antibody did not, however, react with the P44 fragment, a peptide situated at one end of the EHS proteoglycan core protein. These data suggest that the glomerular basement membrane proteoglycan is synthesized from a large precursor protein which undergoes specific proteolytic processing.

Gene structure of chick lumican and identification of the first exon.

Three overlapping genomic clones to chick lumican were isolated and then characterized using restriction enzyme analyses, Southern blot analyses with cDNA probes, and by DNA sequencing. The results showed chick lumican gene to consist of 3 exons with a 2.9-kb first intron and a 4.2-kb second intron. Transcription initiation sites, identified by S1 nuclease experiments using genomic fragments containing exon 1 and by primer extension analysis of RNA, indicated the first exon to be 303 b. Two TATA sequences were 31 and 49 bases upstream of the first exon. The first exon contained all 5' untranslated sequence. The second exon was 896 b and contains 20 b of untranslated sequence, and codes for the start methionine to the end of the 10th leucine rich repeat. The third exon is 880 b and codes for the remainder of the core protein, and 724 b of untranslated 3' sequence. A 1-kb genomic fragment containing a portion of exon 1 and upstream sequence in a luciferase reporter sector showed specific promotor activity in the forward, but not the reverse direction when transfected into corneal fibroblasts. These results show the chick lumican gene to consist of three exons, and that regulatory elements are present within 1 kb upstream of the first exon.

Interactions of basement membrane components.

The binding of laminin, type IV collagen, and heparan sulfate proteoglycan to each other was assessed. Laminin binds preferentially to native type IV (basement membrane) collagen over other collagens. A fragment of laminin (Mr 600 000) containing the three short chains (Mr 200 000) but lacking the long chain (Mr 400 000) showed the same affinity for type IV collagen as the intact protein. The heparan sulfate proteoglycan binds well to laminin and to type IV collagen. These studies show that laminin, type IV collagen and heparan sulfate proteoglycan interact with each other. Such interactions in situ may determine the structure of basement membranes.

Altered steady-state mRNA levels of basement membrane proteins in diabetic mouse kidneys and thromboxane synthase inhibition.

We examined steady-state levels of mRNA encoding type IV collagen, B1 chain of laminin, and the basement membrane heparan sulfate proteoglycan in the kidney cortex of a mouse model (KKAy) of non-insulin-dependent diabetes. mRNAs encoding laminin B1 and the proteoglycan were unchanged in kidneys taken from diabetic mice with demonstrable basement membrane thickening. mRNA levels for type IV collagen, in contrast, were significantly elevated (2-fold) in diabetic mice concurrent with but not preceding morphologically thickened basement membranes. There was a negative correlation between a ratio of proteoglycan/type IV collagen and levels of albuminuria in the diabetic mice. No correlation was noted with laminin. We also examined the effects of inhibiting the synthesis of thromboxane, a potent vasoconstrictor, on the steady-state levels of type IV collagen in the diabetic mice. Inhibition of thromboxane stopped the progression of albuminuria and prevented an increase in type IV collagen mRNA levels. We conclude that basement membrane thickening in diabetes, a hallmark of diabetic nephropathy, is partly a consequence of an unbalanced increase in the production of type IV collagen. The relative decrease in proteoglycan production may contribute to chronic albuminuria.

Alterations in the basement membrane (heparan sulfate) proteoglycan in diabetic mice.

We have grown the EHS (Engelbreth-Holm, Swarm) tumor in normal and genetically diabetic mice (db/db) and measured some components of basement membrane produced in the tumor. These studies showed similar amounts of total protein in control and diabetic tissue and similar patterns of proteins on SDS gel electrophoresis of extracts of the tissue. Laminin, a basement membrane specific glycoprotein utilized as an attachment factor by epithelial cells, was present in increased amounts in diabetic tissue. In contrast, the amount of BM-1 (heparan sulfate) proteoglycan was reduced. Less 35S-sulfate was incorporated into this proteoglycan, and the proteoglycan, but not its component glycosaminoglycans, was heterogeneous in size. The data indicate that either the synthesis of proteoglycan was decreased or its degradation was increased in diabetic tissue. Since the heparan sulfate proteoglycan serves to block the passage of anionic macromolecules through the basement membrane, decreased levels could account for the increased porosity of diabetic basement membrane. Compensatory synthesis of the basement membrane components to restore normal permeability could account for the thickened basement membranes observed in diabetes.

Response of diabetic basement membrane--producing cells to glucose and insulin.

The effect of insulin and glucose on the synthesis of basement membrane components was studied in organ cultures of a basement membrane-producing tumor grown in diabetic and normal mice. Tumor tissue grown in diabetic mice produced more protein and basement membrane-specific proteoglycan in response to insulin than tissue grown in normal mice. Addition of high levels of glucose to the culture medium did not alter insulin-stimulated protein synthesis by diabetic or normal tissue but dampened insulin-stimulated production of proteoglycan. These data suggest that basement membrane-producing cells in diabetic hosts may be hypersensitive to insulin and that stimulation of protein production by insulin may play some role in the in situ hypertrophy of basement membranes.

Localization of binding sites for laminin, heparan sulfate proteoglycan and fibronectin on basement membrane (type IV) collagen.

Rotary shadowing electron microscopy was used to examine complexes formed by incubating combinations of the basement membrane components: type IV collagen, laminin, large heparan sulfate proteoglycan and fibronectin. Complexes were analyzed by length measurement from the globular (COOH) domain of type IV collagen, and by examination of the four arms of laminin and the two arms of fibronectin. Type IV collagen was found to contain binding sites for laminin, heparan sulfate proteoglycan and fibronectin. With laminin the most frequent site was centered approximately 81 nm from the carboxy end of type IV collagen. Less frequent sites appeared to be present at approximately 216 nm and approximately 291 nm, although this was not apparent when the sites were expressed as a fraction of the length of type IV collagen to which they were bound. For heparan sulfate proteoglycan the most frequent site occurred at approximately 206 nm with a less frequent site at approximately 82 nm. For fibronectin, a single site was present at approximately 205 nm. Laminin bound to type IV collagen through its short arms, particularly through the end of the lateral short arms and to heparan sulfate proteoglycan mainly through the end of its long arm. Fibronectin bound to type IV collagen through the free end region of its arms. Using a computer graphics program, the primary laminin binding sites of two adjacent type IV collagen molecules were found to align in the "polygonal" model of type IV collagen, whereas with the "open network" model, a wide meshed matrix is predicted. It is proposed that basement membrane may consist of a lattice of type IV collagen coated with laminin, heparan sulfate proteoglycan and fibronectin.

Molecular cloning and relative tissue expression of keratocan and mimecan in embryonic quail cornea.

We have cloned and sequenced the cDNAs for quail cornea keratan sulfate proteoglycan core proteins, keratocan and mimecan. The deduced quail keratocan protein contains a single conservative amino acid difference from the chick sequence, whereas quail mimecan protein contains a 58 amino acid-long avian-unique sequence that shares no homology with mammalian mimecan. Ribonuclease protection assay of Day 16 embryonic quail tissues reveals that keratocan and lumican are expressed at highest levels in cornea, whereas mimecan mRNA is expressed at a much lower level. Keratocan is expressed only in quail cornea, whereas mimecan is expressed in many different tissues as four transcripts of different sizes. Both lumican and mimecan are expressed at lowest levels in brain, liver and sternum.

Molecular cloning and relative tissue expression of decorin and lumican in embryonic quail cornea.

We have cloned and sequenced the cDNAs for quail cornea proteoglycan core proteins, decorin and lumican. Comparison of deduced amino acid sequences shows that two of five amino acid differences in the mature protein between quail and chick decorin, and two of three for lumican, are non-conservative. Ribonuclease protection assay of Day 16 embryonic quail tissues reveals that decorin and lumican are most highly expressed in cornea, and that both are also highly expressed at approximately equal levels in most other tissues. Decorin is highly expressed in sclera and sternum, whereas lumican is expressed in these tissues, as well as in liver, at very low levels. Both decorin and lumican are expressed at lowest levels in brain.

Tissue fixation with diimidoesters as an alternative to aldehydes. II. Cytochemical and biochemical studies of rat liver fixed with dimethylsuberimidate.

Rat liver fixed with dimethylsuberimidate (DMS) was studied to investigate the use of diimidoesters as dixatives for light and electron microscopic cytochemistry. Paraffin sections of liver fixed with DMS at pH 9.5 were weakly stained with the ninhydrin-Schiff procedure, indicating extensive reaction of NH3+ groups with the fixative. Nuclei were strongly strained by the Feulgen procedure, with no background [corrected] reaction. In contrast, glutaraldehyde fixation resulted in a significant background reaction in the cytoplasm and nuclei in controls for the Schiff-based stains. DMS-fixed liver stained intensely for glycogen with the Periodic acid-Schiff procedure, and biochemical analysis of glycogen retention and extractability indicated that DMS retained considerably more glycogen in sections than glutaraldehyde. DMS-fixed liver incubated for thiamine pyrophosphatase activity revealed reaction product in ER cisternae, Goli saccules and bile canaliculi. Peroxisomes were strongly reactive for catalase activity after incubation in diaminobenzidine medium, and reaction product of glucose-6-phosphatase activity was considerably greater following DMS fixation than after glutaraldehyde. Biochemical studies revealed up to twice as musch residual activity of glucose-6-phosphatase after DMS fixation. These results suggest that DMS may be useful as a primary fixative for certain cytochemical procedures.

Antibody mapping and tissue localization of globular and cysteine-rich regions of perlecan domain III.

Perlecan is the best-characterized basement membrane heparan sulfate proteoglycan. It has a large (approximately 400 KD) core protein consisting of five distinct domains. Domain III, a centrally located domain, contains three globular domains separated by cysteine-rich epidermal growth factor (EGF)-like repeats. Domain III has overall homology with the N-terminus of the laminin alpha 1-chain. The aim of this study was to map a library of nine rat monoclonal antibodies (MAbs) against murine perlecan core protein, using recombinant whole Domain III and defined subdomains of Domain III. ELISA and Western blotting showed that six of the nine MAbs recognized Domain III of perlecan, three of them mapping to globular Subdomain IIIc, and the other three recognized epitopes within the cysteine-rich regions. All six MAbs stained every basement membrane of several mouse organs as well as some connective tissues, including cartilage. Therefore, several distinct epitopes of perlecan Domain III are present in most if not all basement membranes and are not obscured by intermolecular interactions. These precisely mapped antibodies may therefore be useful in understanding the function of perlecan and its core protein.

High glucose downregulates glucose transport activity in retinal capillary pericytes but not endothelial cells.

PURPOSE: To characterize the properties of the glucose transporters of bovine retinal capillary endothelial cells and pericytes and to determine the effects of increased glucose concentrations on glucose transport activity. METHODS: Primary cultures of bovine retinal capillary endothelial cells and pericytes were exposed to low and high glucose concentrations, and immunoblot analysis, 14C-3-O-methylglucose transport activity, and cytochalasin B binding assays were used to characterize the glucose transporters. RESULTS: GLUT1, but not GLUT3 or GLUT4 transporter isoforms, was present in plasma membranes isolated from each cell type. The EC50 for glucose transport was similar in endothelial cells and pericytes (3.94 to 0.48 mM versus 2.24 to 0.69 mM) and was consistent with the EC50 previously reported for GLUT1 transporters on other cells, as was the observation that insulin did not acutely stimulate glucose transport in either cell type. The Vmax for glucose transport was greater in pericytes than endothelial cells (71 to 25 versus 14.5 to 0.8 pmol/10 s/g DNA). Exposure of pericytes to 20 mM glucose for 8 days decreased the initial maximal rate of glucose transport by 30%, compared to pericytes cultured in 5 mM glucose (187 to 7 versus 133 to 9 fmol/20 s/g DNA, P < 0.01), but had no effect on glucose transport activity in endothelial cells. Culture in high glucose decreased the apparent amount of immunoreactive pericyte plasma membrane GLUT1 in immunoblots (0.611 to 0.055 versus 1.0 relative density units), decreased the binding of 3H-cytochalasin B to pericyte plasma membranes, and decreased the mRNA level for GLUT1 in pericytes by 25%. CONCLUSIONS: High-glucose concentrations downregulate glucose transport activity and GLUT 1 content in retinal capillary pericytes but not in endothelial cells. This effect occurred at a pretranslational level. The selective effects of high-glucose concentrations on retinal capillary pericytes in culture might be related to the selective effects of hyperglycemia on these cells in vivo.

Human corneal stroma contains three distinct collagens.

Acetic acid-pepsin extracts of normal human corneal stromas were shown by carboxymethyl cellulose-column chromatography, sodium dodecyl sulfate-slab gel electrophoresis, highly purified collagenase, trypsin sensitivity, and cyanogen bromide peptide mapping to contain types I, III, and IV collagen, with type I by far the predominant species. This degree of collagen heterogeneity in normal human corneal stroma has not been reported previously and may be important in the understanding of wound healing and disease states.

Increased biosynthesis of specific glycoconjugates in rat corneal epithelium following treatment with vitamin A.

The biosynthesis of corneal epithelial proteins and glycoconjugates and their response to vitamin A were measured in vitamin A-deficient and normal rats. The synthesis of specific high-molecular-weight epithelial glycoconjugates was directly related to the levels of vitamin A administered. Protein synthesis, however, remained unaltered. Since vitamin A is known to reverse the keratinization process, the vitamin A-regulated high-molecular-weight glycoconjugates may play a role in modulating the expression of the keratinizing phenotype.

Molecular polymorphism of lumican during corneal development.

PURPOSE: To evaluate the expression of lumican and decorin, the major proteoglycans of the adult corneal stroma, during the acquisition of corneal transparency in developing chick embryos. METHODS: mRNA levels of decorin and lumican were measured in total RNA extracted from corneas of days 9 to 18 of development by Northern blot analysis using a 32P-labeled cDNA clone to each proteoglycan. The synthesis lumican and decorin precursor proteins were determined by biosynthetically radiolabeling corneas from day 7 to 18 chick embryos with 35S-methionine, and then using antibodies specific for lumican and decorin core proteins to precipitate the radiolabeled precursor proteins. The accumulation of lumican and decorin was determined by fractionating extracts of day 7 to 18 embryonic corneas by DEAE chromatography into glycoprotein and proteoglycan fractions, and then analyzing each fraction by Western blot using antibodies to lumican and decorin. RESULTS: Lumican and decorin mRNA increased from day 9 to day 18, with respect to beta-actin. The rate of decorin precursor protein synthesis remained relatively low and constant throughout development, but lumican precursor protein synthesis increased dramatically between days 7 and 9 of embryonic development, to a value 80-fold higher than that of decorin, and then decreased exponentially through day 18. Lumican with polylactosamine (nonsulfated keratan sulfate) side chains was detected in extracts of corneas as early as day 7 of embryonic development, and continued to accumulate within the cornea through day 18. Decorin and lumican with sulfated glycosaminoglycan side chains (ie, proteoglycans), however, were not detected in corneal extracts until day 15, when transparency starts to increase, and then accumulated considerably within the cornea by day 18. CONCLUSIONS: The results of these studies suggest that decorin and lumican expression are independently regulated during the period of acquisition of corneal transparency. The switch in production of the polylactosamine form of lumican to the proteoglycan form of lumican at the onset of increasing corneal transparency suggests that the sulfation of lumican may be important for the development of corneal transparency.

Basement membrane synthesis by human corneal epithelial cells in vitro.

PURPOSE: Collagen gels may prove to be potential carriers for transplantation of cultured corneal epithelial cells. The purpose of this study was to evaluate the suitability of collagen gels in comparison with corneal stromal blocks as the substrate to support the growth of human corneal epithelial cells in culture and the synthesis and deposition of the basement membrane components by these cells. METHODS: Corneal epithelial sheets, freed from the culture dishes using Dispase II (Boehringer Mannheim, Indianapolis, IN), were cultured on corneal stromal blocks. Deposition of laminin, type IV collagen, type VII collagen, and perlecan (heparan sulfate proteoglycan) were evaluated immunohistochemically after 4 days, 7 days, 2 weeks, and 3 weeks. Human limbal explant cultures were established on collagen gels prepared from bovine type I collagen with or without addition of cultured human corneal fibroblasts. After 1, 2, 3, and 4 weeks, the deposition of the basement membrane components was evaluated immunohistochemically. RESULTS: Corneal epithelial cells, cultured on corneal stromal blocks as well as on collagen gels with or without fibroblasts, deposited laminin, type IV collagen, perlecan, and type VII collagen at the interface of the cells and the substrates. However, different substrates differentially influenced the temporal pattern of the deposition of various basement membrane components. On the stromal blocks, deposition of laminin, type IV collagen, and perlecan by the epithelial cells was evident at 1 week. Type VII collagen was detected at 2 weeks. On the collagen gels with fibroblasts, deposition of laminin, type IV collagen and perlecan was detectable at 1 week. In the epithelial cultures on the collagen gels without fibroblasts, only perlecan was detectable at 1 week. At 2 weeks, all of the basement membrane components, including type VII collagen were detectable on the collagen gels, either with or without fibroblasts. CONCLUSION: Human corneal epithelium cultured on collagen gels or on corneal stromal blocks can synthesize and deposit basement membrane components, including laminin, type IV collagen, type VII collagen, and perlecan within 2 weeks in culture. Therefore, collagen gels may serve as potential carriers for human corneal epithelial transplantation.