Dynamic characterisation of the netrin-like domain of human type 1 procollagen C-proteinase enhancer and comparison to the N-terminal domain of tissue inhibitor of metalloproteinases (TIMP)

The backbone mobility of the C-terminal domain of procollagen C-proteinase enhancer (NTR PCOLCE1), part of a connective tissue glycoprotein, was determined using 15N NMR spectroscopy. NTR PCOLCE1 has been shown to be a netrin-like domain and adopts an OB-fold such as that found in the N-terminal domain of tissue inhibitors of metalloproteinases-1 (N-TIMP-1), N-TIMP-2, the laminin-binding domain of agrin and the C-terminal domain of complement protein C5. NMR relaxation dynamics of NTR PCOLCE1 highlight conformational flexibility in the N-terminus, strand A and the proximal CD loop. This region in N-TIMP is known to be essential for inhibitory activity against the matrix metalloproteinases and suggests that this region is of equal importance for NTR PCOLCE1, although the specific functional activity of the NTR PCOLCE1 domain is still unknown. Dynamics observed within the structural core of NTR PCOLCE1 that are not observed in N-TIMP molecules suggest that although the two domains have a similar architecture, the NTR PCOLCE1 domain will show different thermodynamic properties on binding and hence the target molecule could be somewhat different from that observed for the TIMPs. ModelFree order parameters show that NTR PCOLCE1 has more flexibility than both N-TIMP-1 and N-TIMP-2.     

Folding and activity of recombinant human procollagen C-proteinase enhancer.

Recombinant human procollagen C-proteinase enhancer (rPCPE) was expressed using a baculovirus system and purified to homogeneity using a three-step procedure including heparin affinity chromatography. Heparin binding was dependent on the C-terminal netrin-like domain. The recombinant protein was found to be active, increasing the activity of procollagen C-proteinase/bone morphogenetic protein-1 on type I procollagen in a manner comparable to the native protein. Enhancing activity was dependent on intact disulfide bonding within the protein. By circular dichroism, the observed secondary structure of rPCPE was consistent with the known three-dimensional structures of proteins containing homologous domains.     

The NTR module: domains of netrins, secreted frizzled related proteins, and type I procollagen C-proteinase enhancer protein are homologous with tissue inhibitors of metalloproteases.

Using homology search, structure prediction, and structural characterization methods we show that the C-terminal domains of (1) netrins, (2) complement proteins C3, C4, C5, (3) secreted frizzled-related proteins, and (4) type I procollagen C-proteinase enhancer proteins (PCOLCEs) are homologous with the N-terminal domains of (5) tissue inhibitors of metalloproteinases (TIMPs). The proteins harboring this netrin module (NTR module) fulfill diverse biological roles ranging from axon guidance, regulation of Wnt signaling, to the control of the activity of metalloproteases. With the exception of TIMPs, it is not known at present what role the NTR modules play in these processes. In view of the fact that the NTR modules of TIMPs are involved in the inhibition of matrixin-type metalloproteases and that the NTR module of PCOLCEs is involved in the control of the activity of the astacin-type metalloprotease BMP1, it seems possible that interaction with metzincins could be a shared property of NTR modules and could be critical for the biological roles of the host proteins.     

Structure and function of procollagen C-proteinase (mTolloid) domains determined by protease digestion, circular dichroism, binding to procollagen type I, and computer modeling

Procollagen C-proteinase-2 (pCP-2, mTld) is derived from the longest splicing variant of the gene encoding bone morphogenetic protein 1 (BMP-1). The variants have identical amino terminal signal peptides, prodomains and astacin-like protease domains. However, they differ in the length of their carboxy terminal part, which in pCP-2 has the composition CUB1, CUB2, EGF-like1, CUB3, EGF-like2, CUB4, CUB5, and C-tail. In the shorter form, pCP-1 (i.e., BMP-1), the sequence ends after the CUB3-domain. Using a combination of mutagenesis and structural approaches, we have investigated the structure and function of subfragments of pCP-2. The full-length latent recombinant enzyme and its N-terminally truncated form lacking the prodomain were tested for their enzymic activity. The intact protein showed only partial processing of procollagen type I, whereas the truncated form expressed enzymic activity indistinguishable from its native counterpart purified from chick embryo tendons. These results clearly demonstrated that the prodomain is required for the latency of the enzyme but not for its correct folding. Limited proteolysis of the recombinant protein with alpha-chymotrypsin produced four discrete fragments revealing the location of cleavage sites between the repetitive CUB/EGF domains. The results provide evidence that the CUB sequences form independently folded modules that are stabilized by two pairs of internal disulfide bridges. The modules are linked to each other by more flexible, hinge-like peptides. Solid-phase binding assays with isolated CUB domains and immobilized procollagen type I demonstrated that the first three but not the last two CUB domains specifically bound to the substrate. To define putative sites for CUB-CUB or CUB-substrate interactions, we generated molecular models for pCP-2 CUB domains. The models were obtained using as a template the structure of CUB domain in zona pellucida adhesion protein PSP-I/PSP-II from porcine sperm. The predicted conformations for homology models were, subsequently, confirmed by circular dichroism spectroscopy of polypeptide domains isolated following limited proteolysis with alpha-chymotrypsin.     

Bone morphogenetic protein-1 (BMP-1). Identification of the minimal domain structure for procollagen C-proteinase activity

Bone morphogenetic protein-1 (BMP-1) is a shorter spliced variant of mammalian tolloid (mTld), both of which cleave the C-propeptides of type I procollagen during the synthesis of extracellular matrix collagen fibrils. The fact that BMP-1 and mTld both exhibit procollagen C-proteinase (PCP) activity and that BMP-1 is the smaller variant might indicate that BMP-1 comprises the minimal required sequences for PCP activity. BMP-1 comprises a metalloproteinase domain, three CUB domains, and an epidermal growth factor (EGF)-like domain, which is located between the second and third CUB (complement components C1r/C1s, the sea urchin protein Uegf, and BMP-1) domains. In this study we showed the following. 1) The CUB1 domain is required for secretion of the molecule. Domain swapping experiments, in which CUB1 and other CUB domains were interchanged, resulted in retention of the proteins by cells. Therefore, CUB1 and its location immediately adjacent to the metalloproteinase domain are essential for secretion of the protein. 2) Mutants lacking the EGF-like and CUB3 domains exhibited full C-proteinase activity. In contrast, mutants lacking the CUB2 domain were poor C-proteinases. 3) Further studies showed that Glu-483 on the beta4-beta5 loop of CUB2 is essential for C-proteinase activity of BMP-1. In conclusion, the study showed that the minimal domain structure for PCP activity is considerably shorter than expected and comprises the metalloproteinase domain and the CUB1 and CUB2 domains of BMP-1.     

Stabilization of RNA strands in protein synthesis by type I procollagen C-proteinase enhancer protein, a potential RNA-binding protein, in hepatic stellate cells.

Type I procollagen C-proteinase enhancer (PCPE) exists in hepatic stellate cells (HSCs) which can produce collagen. The deduced amino acid sequence of PCPE contains motifs specific for RNA-binding proteins. The effect of PCPE on the syntheses of collagen and noncollagenous protein was studied using an HSC clone derived from cirrhotic rat liver. When the cells were cultured in the presence of an antisense oligonucleotide (AS) against PCPE mRNA, the synthesis of noncollagenous protein as well as collagen was reduced compared to the cells cultured with addition of a nonsense oligonucleotide (NS). The extent of the reduction was similar in both syntheses. The total RNA content of the AS-treated cells and NS-treated cells did not differ. In the presence of actinomycin D, however, such total RNA content was decreased more rapidly in the AS-treated cells than in the NS-treated cells. PCPE may be involved in stabilization of RNA strands in noncollagenous protein synthesis as well as collagen synthesis.     

The 1.1 A crystal structure of human TGF-beta type II receptor ligand binding domain

Transforming growth factor beta (TGF-beta) is involved in a wide range of biological functions including development, carcinogenesis, and immune regulation. Here we report the 1.1 A resolution crystal structure of human TGF-beta type II receptor ectodomain (TBRII). The overall structure of TBRII is similar to that of activin type II receptor ectodomain (ActRII) and bone morphogenic protein receptor type IA (BRIA). It displays a three-finger toxin fold with fingers formed by the beta strand pairs beta1-beta2, beta3-beta4, and beta5-beta6. The first finger in the TBRII is significantly longer than in ActRII and BRIA and folds tightly between the second finger and the C terminus. Surface charge distributions and hydrophobic patches predict potential TBRII binding sites.     

Identification and expression of a novel type I procollagen C-proteinase enhancer protein gene from the glaucoma candidate region on 3q21-q24

A novel human Type I procollagen C-proteinase enhancer protein-like gene, PCOLCE2, was identified by sequencing an EST in the primary open-angle glaucoma (POAG) region on 3q21. The total cDNA encoded a 415-amino-acid protein that has 43% identity to the Type I procollagen C-proteinase enhancer protein (PCOLCE1). PCOLCE2 contains two CUB domains, which are thought to be involved in protein-protein interactions, and an NTR module. PCOLCE2 message is expressed in the trabecular meshwork, lungs, heart, brain, liver, skeletal muscle, kidney, pancreas, and placenta as a 2-kb message. PCOLCE2, a 52-kDa protein, is expressed in the trabecular meshwork. A novel gene, PCOLCE2, has been identified and characterized. Based upon its homology with collagen-binding proteins, its expression in the trabecular meshwork, and its chromosome location, PCOLCE2 is a candidate gene for GLC1C. However, no coding sequence mutations were detected in PCOLCE2 in a POAG patient from the GLC1C family.     

Separation of the two domains of human mucus proteinase inhibitor: inhibitory activity is only located in the carboxy-terminal domain

Human mucus proteinase inhibitor is a two-domain protein which inactivates bovine trypsin and chymotrypsin, leukocyte elastase and cathepsin G. In order to localize the site(s) responsible for these inhibitory activities, the two domains were isolated after specific cleavage of the Asp49-Pro50 bond following mild acid treatment of the bronchial inhibitor. The carboxy-terminal domain was active against leukocyte elastase, trypsin and chymotrypsin whereas the amino-terminal domain, which contained a putative antitryptic active site, was devoid of activity. This implicates that, in the whole molecule, the inhibitory activity region is localized only in the carboxy-terminal domain.     

Identification of the minimal domain structure of bone morphogenetic protein-1 (BMP-1) for chordinase activity: chordinase activity is not enhanced by procollagen C-proteinase enhancer-1 (PCPE-1)

Bone morphogenetic protein 1 (BMP-1), which is a tolloid member of the astacin-like family of zinc metalloproteinases, is a highly effective procollagen C-proteinase (PCP) and chordinase. On the other hand, mammalian tolloid like-2 (mTLL-2) does not cleave chordin or procollagen; procollagen is cleaved by mTLL-2 in the presence of high levels of procollagen C-proteinase enhancer-1 (PCPE-1), for reasons that are unknown. We used these differences in activity between BMP-1 and mTLL-2 to narrow in on the domains in BMP-1 that specify PCP and chordinase activity. Using a domain swap approach, we showed that: 1) the metalloproteinase and CUB2 domains of BMP-1 are absolutely required for PCP activity; swaps with either of the corresponding domains in BMP-1 and mTLL-2 did not result in procollagen cleavage and 2) the proteinase domain of mTLL-2 can cleave chordin if coupled to the CUB1 domain of BMP-1. Therefore, the minimal structure for chordinase activity comprises a metalloproteinase domain (either from BMP-1 or from mTLL-2) and the CUB1 domain of BMP-1 (the CUB1 domain of mTLL-2 cannot substitute for the CUB1 domain of BMP-1). We showed that the minimal procollagen C-proteinase (BMP-1 lacking the EGF and CUB3 domain) was enhanced by PCPE-1 but not as well as BMP-1 retaining the CUB3 domain. Further studies showed that PCPE-1 had no effect on the ability of BMP-1 to cleave chordin. The data support a previously suggested mechanism of PCPE-1 whereby PCPE-1 interacts with procollagen, but in addition, the CUB3 domain of BMP-1 appears to augment the interaction.     

NMR solution structure of the receptor binding domain of human alpha(2)-macroglobulin

Human alpha(2)-macroglobulin-proteinase complexes bind to their receptor, the low density lipoprotein receptor-related protein (LRP), through a discrete 138-residue C-terminal receptor binding domain (RBD), which also binds to the beta-amyloid peptide. We have used NMR spectroscopy on recombinantly expressed uniformly (13)C/(15)N-labeled human RBD to determine its three-dimensional structure in solution. Human RBD is a sandwich of two antiparallel beta-sheets, one four-strand and one five-strand, and also contains one alpha-helix of 2.5 turns and an additional 1-turn helical region. The principal alpha-helix contains two lysine residues on the outer face that are known to be essential for receptor binding. A calcium binding site (K(d) approximately 11 mM) is present in the loop region at one end of the beta-sandwich. Calcium binding principally affects this loop region and does not significantly perturb the stable core structure of the domain. The structure and NMR assignments will enable us to examine in solution specific binding of RBD to domains of the receptor and to beta-amyloid peptide.     

NMR structure of a minimum activity domain of human parathyroid peptide hormone: structural origin of receptor activation.

Parathyroid hormone (PTH) which increases osteoblast numbers and bone formation by activating bone-lining cells to osteoblasts plays an important role in calcium and phosphate homeostasis and bone remodeling by activating PTH receptors. To determine the structural origin of a minimum activity domain of hPTH, we initiated a detailed structural determination of the hPTH(H14) in aqueous solution using NMR spectroscopy. Circular dichroism and NMR data demonstrated that hPTH(H14) maintains a typical helical conformation in both membrane-mimicking environments and 30% TFE solution. The solution structure clearly showed that the residues from Ser(3) to Leu(11) of hPTH(H14) formed a stable helical structure, especially having charged side-chains oriented in opposite directions relative to one another for optimum interaction with the receptor molecule.     

1H and 15N NMR resonance assignments and secondary structure of titin type I domains

Titin/connectin is a giant muscle protein with a highly modular architecture consisting ofmultiple repeats of two sequence motifs, named type I and type II. Type I modules have beensuggested to be intracellular members of the fibronectin type III (Fn3) domain family. Alongthe titin sequence they are exclusively present in the region of the molecule located in thesarcomere A-band. This region has been shown to interact with myosin and C-protein. Oneof the most noticeable features of type I modules is that they are particularly rich insemiconserved prolines, since these residues account for about 8% of their sequence. We havedetermined the secondary structure of a representative type I domain (A71) by 15N and 1HNMR. We show that the type I domains of titin have the Fn3 fold as proposed, consisting ofa three- and a four-stranded -sheet. When the two sheets are placed on top of each other toform the -sandwich characteristic of the Fn3 fold, 8 out of 10 prolines are found on the sameside of the molecule and form an exposed hydrophobic patch. This suggests that thesemiconserved prolines might be relevant for the function of type I modules, providing asurface for binding to other A-band proteins. The secondary structure of A71 was structurallyaligned to other extracellular Fn3 modules of known 3D structure. The alignment shows thattitin type I modules have closest similarity to the first Fn3 domain of Drosophila neuroglian.     

Differential ubiquitin binding of the UBA domains from human c-Cbl and Cbl-b: NMR structural and biochemical insights

The Cbl proteins, RING-type E3 ubiquitin ligases, are responsible for ubiquitinating the activated tyrosine kinases and targeting them for degradation. Both c-Cbl and Cbl-b have a UBA (ubiquitin-associated) domain at their C-terminal ends, and these two UBA domains share a high sequence similarity (75%). However, only the UBA from Cbl-b, but not from c-Cbl, can bind ubiquitin (Ub). To understand the mechanism by which the UBA domains specifically interact with Ub with different affinities, we determined the solution NMR structures of these two UBA domains, cUBA from human c-Cbl and UBAb from Cbl-b. Their structures show that these two UBA domains share the same fold, a compact three-helix bundle, highly resembling the typical UBA fold. Chemical shift perturbation experiments reveal that the helix-1 and loop-1 of UBAb form a predominately hydrophobic surface for Ub binding. By comparing the Ub-interacting surface on UBAb and its counterpart on cUBA, we find that the hydrophobic patch on cUBA is interrupted by a negatively charged residue Glu12. Fluorescence titration data show that the Ala12Glu mutant of UBAb completely loses the ability to bind Ub, whereas the mutation disrupting the dimerization has no significant effect on Ub binding. This study provides structural and biochemical insights into the Ub binding specificities of the Cbl UBA domains, in which the hydrophobic surface distribution on the first helix plays crucial roles in their differential affinities for Ub binding. That is, the amino acid residue diversity in the helix-1 region, but not the dimerization, determines the abilities of various UBA domains binding with Ub.     

Crystal structure of the human carboxypeptidase N (kininase I) catalytic domain

Human carboxypeptidase N (CPN), a member of the CPN/E subfamily of "regulatory" metallo-carboxypeptidases, is an extracellular glycoprotein synthesized in the liver and secreted into the blood, where it controls the activity of vasoactive peptide hormones, growth factors and cytokines by specifically removing C-terminal basic residues. Normally, CPN circulates in blood plasma as a hetero-tetramer consisting of two 83 kDa (CPN2) domains each flanked by a 48 to 55 kDa catalytic (CPN1) domain. We have prepared and crystallized the recombinant C-terminally truncated catalytic domain of human CPN1, and have determined and refined its 2.1 A crystal structure. The structural analysis reveals that CPN1 has a pear-like shape, consisting of a 319 residue N-terminal catalytic domain and an abutting, cylindrically shaped 79 residue C-terminal beta-sandwich transthyretin (TT) domain, more resembling CPD-2 than CPM. Like these other CPN/E members, two surface loops surrounding the active-site groove restrict access to the catalytic center, offering an explanation for why some larger protein carboxypeptidase inhibitors do not inhibit CPN. Modeling of the Pro-Phe-Arg C-terminal end of the natural substrate bradykinin into the active site shows that the S1' pocket of CPN1 might better accommodate P1'-Lys than Arg residues, in agreement with CPN's preference for cleaving off C-terminal Lys residues. Three Thr residues at the distal TT edge of CPN1 are O-linked to N-acetyl glucosamine sugars; equivalent sites in the membrane-anchored CPM are occupied by basic residues probably involved in membrane interaction. In tetrameric CPN, each CPN1 subunit might interact with the central leucine-rich repeat tandem of the cognate CPN2 subunit via a unique hydrophobic surface patch wrapping around the catalytic domain-TT interface, exposing the two active centers.     

Crystal structure of human lectin-like, oxidized low-density lipoprotein receptor 1 ligand binding domain and its ligand recognition mode to OxLDL

Lectin-like, oxidized low-density lipoprotein (LDL) receptor 1, LOX-1, is the major receptor for oxidized LDL (OxLDL) in endothelial cells. We have determined the crystal structure of the ligand binding domain of LOX-1, with a short stalk region connecting the domain to the membrane-spanning region, as a homodimer linked by an interchain disulfide bond. In vivo assays with LOX-1 mutants revealed that the "basic spine," consisting of linearly aligned arginine residues spanning over the dimer surface, is responsible for ligand binding. Single amino acid substitution in the dimer interface caused a severe reduction in LOX-1 binding activity, suggesting that the correct dimer arrangement is crucial for binding to OxLDL. Based on the LDL model structure, possible binding modes of LOX-1 to OxLDL are proposed.     

The A and B fragments of normal type I procollagen have a similar thermal stability to proteinase digestion but are selectively destabilized by structural mutations

Previous studies demonstrated that the thermal stability of the procollagen triple helix can be assayed by digesting the protein for short periods with high concentrations of trypsin and chymotrypsin. Here we cleaved human type I procollagen or collagen with vertebrate collagenase to generate A fragments from the three-quarter amino termini and B fragments from the one-quarter carboxy termini of the molecules. The thermal stabilities of the fragments were then assayed by rapid trypsin/chymotrypsin digestion. Both fragments were resistant up to 36 degrees C and completely degraded between 37 degrees C and 39 degrees C. In subsequent experiments the same assay was carried out with type I procollagens synthesized by fibroblasts from two patients with lethal variants of osteogenesis imperfecta. With one, the A fragments were selectively destabilized, an observation consistent with previous data indicating that the mutation in the patient produced a deletion of 84 amino acids from the middle of the alpha 1(I) chain. With procollagen synthesized by fibroblasts from the second patient the B fragments were selectively destabilized, an observation consistent with preliminary data indicating a mutation that alters the primary structure of the carboxy-terminal region of the alpha 1(I) chain. Therefore, the procedures described here present a simple and direct method for locating mutations that destabilize the collagen triple helix.     

Identification of residues of the IFNAR1 chain of the type I human interferon receptor critical for ligand binding and biological activity

The antiviral and antiproliferative activities of human type I interferons (IFNs) are mediated by two transmembrane receptor subunits, IFNAR1 and IFNAR2. To elucidate the role of IFNAR1 in IFN binding and the establishment of biological activity, specific residues of IFNAR1 were mutated. Residues (62)FSSLKLNVY(70) of the S5-S6 loop of the N-terminal subdomain of IFNAR1 and tryptophan-129 of the second subdomain of IFNAR1 were shown to be crucial for IFN-alpha binding and signaling and establishment of biological activity. Mutagenesis of peptide (278)LRV in the third subdomain shows that these residues are critical for IFN-alpha-induced biological activity but not for ligand binding. These data, together with the sequence homology of IFNAR1 with cytokine receptors of known structure and the recently resolved NMR structure of IFNAR2, led to the establishment of a three-dimensional model of the human IFN-alpha/IFNAR1/IFNAR2 complex. This model predicts that following binding of IFN to IFNAR1 and IFNAR2 the receptor complex assumes a "closed form", in which the N-terminal domain of IFNAR1 acts as a lid, resulting in the activation of intracellular kinases. Differences in the primary sequence of individual IFN-alpha subtypes and resulting differences in binding affinity, duration of ligand/receptor association, or both would explain differences in intracellular signal intensities and biological activity observed for individual IFN-alpha subtypes.