Mouse AMACO,a kidney and skin basement membrane associated molecule that mediates RGD-dependent cell attachment

The VWA domain-containing extracellular matrix protein AMACO has not been extensively characterized and its function remains unknown. It has been proposed as a potential cancer marker and carries a rare O-glucosylation and O-fucosylation on its first EGF-like domain. AMACO is a basement membrane associated protein, however its exact localization has not been determined. Here we show by immunogold electron microscopy of mouse kidney and skin that AMACO does not occur within the basement membrane but rather subjacent to the basement membrane at its stromal surface. In skin, AMACO often colocalizes with triple-helical domains of collagen VII containing anchoring fibrils as they emerge from the basal lamina. However, the immunogold patterns for AMACO and the C-terminal end of collagen VII show discrete differences, indicating that AMACO and collagen VII do not colocalize at anchoring plaques. In contrast, the localization pattern of AMACO partially overlaps with that for collagen XVIII. In addition, mouse AMACO was shown to support β1 integrin-mediated adhesion of a keratinocyte-like cell line, HaCaT, and a fibroblast cell line, Wi26, in an RGD-dependent manner, most likely using an RGD-motif near the C-terminus of AMACO. However, the loss of cell adhesion to the C-terminal part of the human AMACO, due to the unique absence of an RGD sequence in the human protein, suggests that cell adhesion is not AMACO's major function.     

Identification of a disaccharide (Xyl-Glc) and a trisaccharide (Xyl2-Glc) O-glycosidically linked to a serine residue in the first epidermal growth factor-like domain of human factors VII and IX and protein Z and bovine protein Z

We have recently described a unique trisaccharide linked to a serine residue in the first epidermal growth factor-like domains of bovine blood coagulation factors VII (Ser-52) and IX (Ser-53) (Hase, S., Kawabata, S., Nishimura, H., Takeya, H., Sueyoshi, T., Miyata, T., Iwanaga, S., Takao, T., Shimonishi, Y., and Ikenaka, T. (1988) J. Biochem. (Tokyo) 104, 867-868). The sugar chain identified in these clotting factors consists of 1 mol of hexose (glucose (Glc] and 2 mol of pentose (xylose (Xyl]. We report here that human factors VII and IX and protein Z and bovine protein Z also contain such carbohydrate moieties linked to a serine residue at the same position found in bovine factors VII and IX. A glycopeptide derived from each of these proteins was subjected to amino acid sequence and component sugar analyses and fast atom bombardment mass spectrometric analysis. The results indicate that the glycopeptide derived from human factor IX contains 1 mol each of Glc and Xyl. The reducing end of this disaccharide was identified as Glc by analyzing the disaccharide generated by hydrazinolysis. In contrast, human factor VII and protein Z yielded two different glycopeptides which contained Glc and Xyl at molar ratios of 1:1 and 1:2, respectively, suggesting microheterogeneity of these O-linked sugar chains. Bovine protein Z glycopeptide contained 1 mol of Glc and 2 mol of Xyl. These sugar compositions were confirmed by analyses of the intact proteins. In relation to the trisaccharide sugar chain previously discovered in bovine factors VII and IX, these findings indicate the existence of a Xyl2-Glc-Ser and a Xyl-Glc-Ser structure in the first epidermal growth factor-like domains of human factors VII and IX and protein Z in addition to that of bovine protein Z. Whether these carbohydrate moieties contribute to the biological activities of these proteins is unknown.     

Identification and characterization of abeta1,3-glucosyltransferase that synthesizes the Glc-beta1,3-Fuc disaccharide on thrombospondin type 1 repeats

Thrombospondin type 1 repeats (TSRs) are biologically important domains of extracellular proteins. They are modified with a unique Glcbeta1,3Fucalpha1-O-linked disaccharide on either serine or threonine residues. Here we identify the putative glycosyltransferase, B3GTL, as the beta1,3-glucosyltransferase involved in the biosynthesis of this disaccharide. This enzyme is conserved from Caenorhabditis elegans to man and shares 28% sequence identity with Fringe, the beta1,3-N-acetylglucosaminyltransferase that modifies O-linked fucosyl residues in proteins containing epidermal growth factor-like domains, such as Notch. beta1,3-Glucosyltransferase glucosylates properly folded TSR-fucose but not fucosylated epidermal growth factor-like domain or the non-fucosylated modules. Specifically, the glucose is added in a beta1,3-linkage to the fucose in TSR. The activity profiles of beta1,3-glucosyltransferase and protein O-fucosyltransferase 2, the enzyme that carries out the first step in TSR O-fucosylation, superimpose in endoplasmic reticulum subfractions obtained by density gradient centrifugation. Both enzymes are soluble proteins that efficiently modify properly folded TSR modules. The identification of the beta1,3-glucosyltransferase gene allows us to manipulate the formation of the rare Glcbeta1,3Fucalpha1 structure to investigate its biological function.     

Fibulins: a versatile family of extracellular matrix proteins

Fibulins are a newly recognized family of extracellular matrix proteins. The five known members of the family share an elongated structure and many calcium-binding sites, owing to the presence of tandem arrays of epidermal growth factor-like domains. They have overlapping binding sites for several basement-membrane proteins, tropoelastin, fibrillin, fibronectin and proteoglycans, and they participate in diverse supramolecular structures. New insights into their biological roles are now emerging from studies of transgenic mice and of some inherited human diseases.     

The cysteine-rich domain of snake venom metalloproteinases is a ligand for von Willebrand factor A domains: role in substrate targeting

Snake venom metalloproteinases (SVMPs) are members of the Reprolysin family of metalloproteinases to which the ADAM (a disintegrin and metalloproteinase) proteins also belong. The disintegrin-like/cysteine-rich domains of the ADAMs have been implicated in their function. In the case of the SVMPs, we hypothesized that these domains could function to target the metalloproteinases to key extracellular matrix proteins or cell surface proteins. Initially we detected interaction of collagen XIV, a fibril-associated collagen with interrupted triple helices containing von Willebrand factor A (VWA) domains, with the PIII SVMP catrocollastatin. Next we investigated whether other VWA domain-containing matrix proteins could support the binding of PIII SVMPs. Using surface plasmon resonance, the PIII SVMP jararhagin and a recombinant cysteine-rich domain from a PIII SVMP were demonstrated to bind to collagen XIV, collagen XII, and matrilins 1, 3, and 4. Jararhagin was shown to cleave these proteins predominantly at sites localized at or near the VWA domains suggesting that it is the VWA domains to which the PIII SVMPs are binding via their cysteine-rich domain. In light of the fact that these extracellular matrix proteins function to stabilize matrix, targeting the SVMPs to these proteins followed by their specific cleavage could promote the destabilization of extracellular matrix and cell-matrix interactions and in the case of capillaries could contribute to their disruption and hemorrhage. Although there is only limited structural homology shared by the cysteine-rich domains of the PIII SVMPs and the ADAMs our results suggest an analogous function for the cysteine-rich domains in certain members of the expanded ADAM family of proteins to target them to VWA domain-containing proteins.     

Structure and chromosomal location of the human gene encoding cartilage matrix protein

Cartilage matrix protein (CMP) is a major component of the extracellular matrix of nonarticular cartilage. The structure and chromosomal location of the human gene encoding CMP was determined by molecular cloning analysis. We used a partial chicken CMP cDNA probe to isolate three overlapping human genomic clones. From one of these clones, a probe containing 2 human CMP exons was isolated and used to map the gene to chromosome 1p35 and to screen a human retina cDNA library. Two overlapping cDNA clones were isolated. The predicted protein sequence of 496 amino acids includes a 22-residue signal peptide and a 474-residue mature protein of Mr 51,344. The human CMP gene and polypeptide are strikingly similar to the chicken CMP gene and polypeptide. Human CMP is 79% identical to chicken CMP and contains two homologous domains separated by an epidermal growth factor-like domain. One potential N-glycosylation site is conserved between the two species. The human CMP gene spans 12 kilobase pairs with 8 exons and 7 introns which are similar in size to those of the chicken CMP gene. Both RNA splice junctions of intron G in the human and chicken CMP genes are nonconforming to the consensus splice sequences. This suggests that the CMP gene utilizes a new RNA splicing mechanism.     

A domain composed of epidermal growth factor-like structures of human thrombomodulin is essential for thrombin binding and for protein C activation

Thrombomodulin, an endothelial thrombin receptor, acts as a cofactor for the thrombin-catalyzed activation of anticoagulant protein C. The extracellular region of human thrombomodulin consists of three tentative domains, a NH2-terminal domain (D1), a domain involving six consecutive epidermal growth factor-like structures (D2), and an O-glycosylation-rich domain (D3). To identify the domain onto which thrombin binds, a series of recombinant proteins corresponding to the entire protein, D1, D2, D1 + D2, D1 + D2 + D3, and D2 + D3 were expressed in simian COS-1 cells. The proteins were partially purified by rabbit anti-thrombomodulin-F(ab')2-agarose chromatography. Western blotting analysis showed the expression of the respective recombinant proteins. All proteins involving D2, as well as D2 alone, had cofactor activity that allowed binding directly to thrombin, but D1 did not. The cofactor activity of the entire protein but not the mutants is increased in the presence of phospholipids and this is the only protein that binds to the phospholipid layer. These results indicate that the domain involving the epidermal growth factor-like structures of thrombomodulin is essential for thrombin binding and expression of the cofactor activity for protein C activation and that none of the extracellular domains interact with phospholipids.     

Latent transforming growth factor β-binding protein-3 and fibulin-1C interact with the extracellular domain of the heparin-binding EGF-like growth factor precursor

Background  

The membrane-bound cell-surface precursor and soluble forms of heparin-binding epidermal growth factor-like growth factor (HB-EGF) contribute to many cellular developmental processes. The widespread occurrence of HB-EGF in cell and tissue types has led to observations of its role in such cellular and tissue events as tumor formation, cell migration, extracellular matrix formation, wound healing, and cell adherence. Several studies have reported the involvement of such extracellular matrix proteins as latent transforming growth factor β-binding protein, TGF-β, and fibulin-1 in some of these processes. To determine whether HB-EGF interacts with extracellular matrix proteins we used the extracellular domain of proHB-EGF in a yeast two-hybrid system to screen a monkey kidney cDNA library. cDNA clones containing nucleotide sequences encoding domains of two proteins were obtained and their derived amino acid sequences were evaluated.     

Interaction of the cysteine-rich domain of snake venom metalloproteinases with the A1 domain of von Willebrand factor promotes site-specific proteolysis of von Willebrand factor and inhibition of von Willebrand factor-mediated platelet aggregation

Snake venom metalloproteinases (SVMPs) have recently been shown to interact with proteins containing von Willebrand factor A (VWA) domains, including the extracellular matrix proteins collagen XII, collagen XIV, matrilins 1, 3 and 4, and von Willebrand factor (VWF) via their cysteine-rich domain. We extended those studies using surface plasmon resonance to investigate the interaction of SVMPs with VWF, and demonstrated that jararhagin, a PIII SVMP containing a metalloproteinase domain followed by disintegrin-like and cysteine-rich domains, catrocollastatin C, a disintegrin-like/cysteine-rich protein, and the recombinant cysteine-rich domain of atrolysin A (A/C) all interacted with immobilized VWF in a dose-dependent fashion. Binding of VWF in solution to immobilized A/C was inhibited by ristocetin and preincubation of platelets with A/C abolished ristocetin/VWF-induced platelet aggregation, indicating that the interaction of A/C with VWF is mediated by the VWA1 domain. Jararhagin cleaved VWF at sites adjacent to the VWA1 domain, whereas atrolysin C, a SVMP lacking the cysteine-rich domain, cleaved VWF at dispersed sites. A/C and catrocollastatin C completely inhibited the digestion of VWF by jararhagin, demonstrating that the specific interaction of jararhagin with VWF via the VWA1 domain is necessary for VWF proteolysis. In summary, we localized the binding site of PIII SVMPs in VWF to the A1 domain. This suggests additional mechanisms by which SVMPs may interfere with the adhesion of platelets at the site of envenoming. Thus, specific interaction of cysteine-rich domain-containing SVMPs with VWF may function to promote the hemorrhage caused by SVMP proteolysis of capillary basements and surrounding stromal extracellular matrix.     

Backbone dynamics of a cbEGF domain pair in the presence of calcium

Calcium binding (cb) epidermal growth factor-like (EGF) domains are found in a wide variety of extracellular proteins with diverse functions. In several proteins, including the fibrillins (1 and 2), the low-density lipoprotein receptor, the Notch receptor and related molecules, these domains are organised as multiple tandem repeats. The functional importance of calcium-binding by EGF domains has been underscored by the identification of missense mutations associated with defective calcium-binding, which have been linked to human diseases. Here, we present (15)N backbone relaxation data for a pair of cbEGF domains from fibrillin-1, the defective protein in the Marfan syndrome. The data were best fit using a symmetric top model, confirming the extended conformation of the cbEGF domain pair. Our data demonstrate that calcium plays a key role in stabilising the rigidity of the domain pair on the pico- to millisecond time-scale. Strikingly, the most dynamically stable region of the construct is centred about the domain interface. These results provide important insight into the properties of intact fibrillin-1, the consequences of Marfan syndrome causing mutations, and the ultrastructure of fibrillins and other extracellular matrix proteins.     

Ca2+-dependent interface formation in fibrillin-1

The calcium-binding epidermal growth factor-like (cbEGF) domain is a common structural motif in extracellular and transmembrane proteins. K(d) values for Ca2+ vary from the millimolar to nanomolar range; however the molecular basis for this variation is poorly understood. We have measured K(d) values for six fibrillin-1 cbEGF domains, each preceded by a transforming growth factor beta-binding protein-like (TB) domain. Using NMR and titration with chromophoric chelators, we found that K(d) values varied by five orders of magnitude. Interdomain hydrophobic contacts between TB-cbEGF domains were studied by site-directed mutagenesis and could be correlated directly with Ca2+ affinity. Furthermore, in TB-cbEGF pairs that displayed high-affinity binding, NMR studies showed that TB-cbEGF interface formation was strongly Ca2+-dependent. We suggest that Ca2+ affinity is a measure of interface formation in both homologous and heterologous cbEGF domain pairs, thus providing a measure of flexibility in proteins with multiple cbEGF domains. These data highlight the versatile role of the cbEGF domain in fine tuning the regional flexibility of proteins and provide new constraints for the organization of fibrillin-1 within 10-12-nm microfibrils of the extracellular matrix.     

The two N-glycans present on bovine Pofut1 are differently involved in its solubility and activity

O-Fucosylation is a post-translational glycosylation in which an O-fucose is covalently attached to the hydroxyl group of a specific serine or threonine residue. This modification occurs within the consensus sequence C2X(4-5)(S/T)C3 present on epidermal growth factor-like repeats of several proteins, including the Notch receptors and their ligands. The enzyme responsible for the addition of O-fucose to epidermal growth factor-like repeats is protein O-fucosyltransferase 1. Protein O-fucosyltransferase 1-mediated O-fucosylation is essential in Notch signaling, folding and targeting to the cell surface. Here, we studied the expression pattern of protein O-fucosyltransferase 1 in cattle and showed that the active enzyme is present in all tissues examined from embryo and adult as a glycoprotein with two N-glycans. By comparing protein O-fucosyltransferase 1 sequences available in databases, we observed that mammalian protein O-fucosyltransferase 1 enzymes possess two putative N-glycosylation sites, and that only the first is conserved among bilaterians. To gain more insight regarding the significance of N-glycans on protein O-fucosyltransferase 1, we substituted, by site-directed mutagenesis, bovine protein O-fucosyltransferase 1 N65, N163 or both, with L or Q. We demonstrated that the loss of N-glycan on N163 caused a slight decrease in protein O-fucosyltransferase 1 activity. In contrast, glycosylation of N65 was crucial for protein O-fucosyltransferase 1 functionality. Loss of glycosylation at N65 resulted in aggregation of protein O-fucosyltransferase 1, suggesting that N-glycosylation at this site is essential for proper folding of the enzyme.     

SREC-II,a new member of the scavenger receptor type F family,trans-interacts with SREC-I through its extracellular domain

The scavenger receptor expressed by endothelial cells (SREC) with an extremely large cytoplasmic domain, was originally identified in a human endothelial cell line. In this study, we have cloned a second isoform named SREC-II and shown that there is a heterophilic interaction between SREC-I and -II at their extracellular domains. The cDNA for murine SREC-II encodes an 834-amino acid protein with 35% homology to SREC-I. Similar to SREC-I, SREC-II contains multiple epidermal growth factor-like repeats in its extracellular domain. However, in contrast to SREC-I, SREC-II had little activity to internalize modified low density lipoproteins (LDL). A Northern blot analysis revealed a tissue expression pattern of SREC-II similar to that of SREC-I with predominant expression in human heart, lung, ovary, and placenta. Mouse fibroblast L cells with no tendency to associate showed noticeable aggregation when SREC-I was overexpressed in these cells, whereas overexpression of SREC-II caused only slight aggregation. Remarkably, intense aggregation was observed when SREC-I-expressing cells were mixed with those expressing SREC-II. Deletion of almost all of the cytoplasmic receptor domain had no effect on the receptor expression and cell aggregation, indicating that solely the extracellular domain is involved in cell aggregation. The association of SREC-I and -II was effectively suppressed by the presence of scavenger receptor ligands such as acetylated LDL and oxidized LDL. These findings suggest that SREC-I and -II show weak cell-cell interaction by their extracellular domains (termed homophilic trans-interaction) but display strong heterophilic trans-interaction through the extracellular epidermal growth factor-like repeat domains.     

Fibrillin-1, a calcium binding protein of extracellular matrix

Fibrillin-1 is a large extracellular matrix glycoprotein which assembles to form 10-12 nm microfibrils in extracellular matrix. Mutations in the human fibrillin-1 gene (FBN-1) cause the connective tissue disease Marfan syndrome and related disorders, which are characterised by defects in the skeletal, cardiovascular and ocular systems of the body. Fibrillin-1 has a striking modular organisation which is dominated by multiple tandem repeats of the calcium binding epidermal growth factor-like (cbEGF) domain. This review focuses on recent studies which have investigated the structural and functional role of calcium binding to cbEGF domains in fibrillin-1 and 10-12 nm microfibrils.     

1H, 13C and 15N resonance assignments for the fibrillin-1 EGF2-EGF3-hybrid1-cbEGF1 four-domain fragment

Fibrillins are large extracellular glycoproteins that form the principal component of microfibrils. These perform a vital structural function in the extracellular matrix of many tissues. Fibrillins have also been implicated in mediating a number of protein–protein interactions, some of which may be significant in regulating growth factors such as transforming growth factor β. Here we present the backbone and side-chain ¹H, ¹³C and ¹⁵N assignments for a 19 kDa protein fragment derived from the N-terminus of human fibrillin-1, encompassing four domains in total. These domains include the second and third epidermal growth factor-like (EGF) domains, the first hybrid domain (hyb1), and the first calcium-binding EGF domain of fibrillin-1. This region of fibrillin-1 is of particular interest as the hyb1 domain has been suggested to play a role in microfibril assembly, as well as several other protein–protein interactions.     

Protein interaction analysis of ST14 domains and their point and deletion mutants

ST14 (suppression of tumorigenicity 14) is a transmembrane serine protease that contains a serine protease catalytic (SP) domain, an SEA domain, two complement subcomponent C1r/s (CUB) domains, and four low density lipoprotein receptor class A domains. Glutathione S-transferase fusion proteins with SP, CUB, and low density lipoprotein receptor domains and their corresponding mutants were generated to analyze protein interactions with these domains. Modified glutathione S-transferase pull-down assays demonstrated the interaction between the SP domain and hepatocyte growth factor activator inhibitor-1. With the same method, a CUB domain-interacting protein was isolated and turned out to be the transmembrane protein with epidermal growth factor-like and two follistatin-like domains 1 (TMEFF1). Quantitative real time PCR revealed that the expression of the TMEFF1 gene was dependent on the transfection of the ST14 gene in the RKO cell line. Our results also suggested that ST14 and TMEFF1 were co-expressed in the human breast cancer cell line MCF7, human placenta, kidney, and liver tissues. Interestingly, these two genes were co-up-regulated in kidney tumors versus normal tissues, consistent with our results that showed the dependence of TMEFF1 expression on ST14 in RKO cells. Finally, homology modeling studies suggested that TMEFF1 might form a complex with ST14 by an interaction between epidermal growth factor and CUB domains.     

Evidence for the existence of O-linked sugar chains consisting of glucose and xylose in bovine thrombospondin.

We have recently discovered unusual sugar chains [xylose-glucose and (xylose)2-glucose] linked to a serine residue in the first epidermal growth factor (EGF)-like domains of human and bovine coagulation factors VII, IX, and protein Z. The sequence surrounding this serine residue has a common -Cys-X-Ser-X-Pro-Cys- structure. Since one (residues 533-538) of the three EGF-like domains found in human thrombospondin contains the conserved sequence, we examined the presence of such O-linked sugar chains in bovine thrombospondin (bTSP) and its 210-kDa fragment. Component sugar analysis after pyridylamination (PA) of the acid hydrolysates of the S-aminoethylated proteins revealed that the proteins contain glucose (Glc) and xylose (Xyl). The oligosaccharide moieties released from intact bTSP by hydrazinolysis followed by pyridylamination were separated into two PA-oligosaccharides by high performance liquid chromatography (HPLC). Component sugar analysis of these PA-oligosaccharides indicated that they consist of Glc and Xyl in molar ratios of 1:1 and 1:2 (or 1:3). The reducing ends of both PA-sugar chains were found to be PA-Glc, as judged from the retention time of the HPLC peak of their hydrolysates. The presence of these PA-sugar chains in bTSP was confirmed by HPLC mapping with two different columns, using standard PA-di- or PA-trisaccharide derived from coagulation factors. From these results, we concluded that bTSP contains O-linked sugar chains consisting of Glc and Xyl in one of its three EGF-like domains.