N.m.r. analyses of the histidine microenvironments in a human salivary proline-rich glycoprotein

The pKa's of the three histidine residues in a proline-rich glycoprotein from human parotid saliva (PRG) were determined by 360 MHz proton n.m.r. spectroscopy. The addition of calcium (0.64 mM) caused drops in the pKa's of all three histidines by approximately 0.25 units. When imidazole and cyclo)L-histidine-L-proline) were used as model compounds, corresponding concentrations of calcium had no effect on their pKa's. Also, the model compounds gave absolute pKa values in good agreement with similar chemical species reported in the literature. Exchange lifetime data and previously reported hydrogen----deuterium exchange experiments suggest that the PRG histidine N tau H protons are not involved in hydrogen-bonds. Collectively, these data imply that changes in PRG conformation occur upon the addition of calcium.     

Circular dichroism and fluorescence spectroscopic analyses of a proline-rich glycoprotein from human parotid saliva

A proline-rich glycoprotein (PRG) was isolated from human parotid saliva and examined by circular dichroism and fluorescence spectroscopy. Addition of guanidine hydrochloride to PRG labeled with an extrinsic dansyl probe had no effect on the fluorescence spectra's 511 nm lambda-max location. Thermodynamic calculations supported the contention that PRG has no significant tertiary structure. Circular dichroism results for PRG were simulated by computer and a secondary structure composed of 70% random coil and 30% beta-form conformation was predicted. Circular dichroism of PRG failed to detect either poly-L-proline type I or II structures. Deglycosylation of PRG had no measurable effect on the circular dichroism spectrum, indicating that the carbohydrate side chains had little influence on PRG secondary structure. Based upon mathematical calculations, beta-turns were predicted around three glycosylated Asn residues of PRG. These collective data suggest that PRG is composed of a disordered polypeptide chain with at least three of the N-linked Asn residues participating in some type of beta-turn.     

Use of the photoaffinity cross-linking agent N-hydroxysuccinimidyl-4-azidosalicylic acid to characterize salivary-glycoprotein-bacterial interactions.

The present study has utilized the iodinatable cross-linking agent N-hydroxysuccinimidyl-4-azidosalicylic acid (ASA) to examine the specific interaction between the proline-rich glycoprotein (PRG) of human parotid saliva and Streptococcus sanguis G9B. The binding of 125I-ASA-PRG to Streptococcus sanguis G9B displayed saturation kinetics, reversibility and was inhibited by unlabelled PRG. Inhibition studies with other glycoproteins and saccharides indicated that binding was mediated by a bacterial adhesin with specificity towards N-acetylneuraminic acid, galactose, and N-acetylgalactosamine. After cross-linking, the 125I-ASA-PRG-adhesin complex could be extracted with SDS and separated from uncoupled 125I-ASA-PRG by gel filtration on Sepharose CL-6B. Approx. 1% of the 125I-ASA-PRG was cross-linked to the bacterial surface. Examination of the 125I-ASA-PRG-adhesin complex by SDS/polyacrylamide-gel electrophoresis/fluorography on 5% -(w/v)-polyacrylamide gels revealed that PRG was bound to two bacterial components. These findings support our previous suggestion that human salivary glycoproteins can specifically interact with oral streptococci and that these interactions occur between the glycoprotein's carbohydrate units and lectin(s) on the bacterial cell surface.     

N.m.r. and computer-simulated conformational analyses of a nonapeptide found in a human salivary proline-rich glycoprotein

The amino acid sequence G(1)-P(2)-P(3)-P(4)-H(5)-P(6)-G(7)-K(8)-P(9) occurs twice in the proline-rich glycoprotein (PRG) found in human parotid saliva. As part of our efforts to elucidate the structure-function relationships of PRG, this nonapeptide sequence (PRG9) was synthesized for the purpose of conformational analyses by high-resolution proton n.m.r. spectroscopy and computer-modeling. The empirical n.m.r. spectrum differed from the simulated spectrum in that the overall chemical shift locations were displaced from their random coil positions and the five proline residues had non-degenerate C alpha H alpha protons. Other n.m.r. data indicated that no intramolecular hydrogen-bonding was present in the PRG. In conjunction with X-ray crystallographic data on a triproline-containing model compound (Kartha, g., Ashida, T. & Kakudo, M. (1974) Acta Cryst. B30, 1861-1866), four energy-minimized PRG9 structures were obtained. Two of the structures were energetically unfavorable, while the other two conformations were reasonable. The two most likely structures gave all prolines an S-type ring pucker, the P(2)-P(3)-P(4) sequence as a poly-L-proline II helix, the H(5) phi = -90.3 degrees, P(6) and P(9) with trans peptide bond orientation, G(7) in an extended state, and the K(8) phi = -93.2 degrees or -146.8 degrees for structures #1 and #2, respectively.     

Structure and bacterial receptor activity of a human salivary proline-rich glycoprotein

Using an overlay technique, we previously showed that the Gram-negative periodontal pathogen Fusobacterium nucleatum binds to a glycoprotein of Mr 89,000 (Prakobphol, A., Murray, P., and Fischer, S.J. (1987) Anal. Biochem. 164, 5-11) in the parotid saliva of some individuals. We now show that deglycosylation of the purified glycoprotein results in loss of receptor activity. Amino acid analysis of the protein core showed predominantly proline, glycine, and glutamic acid/glutamine, a characteristic of proline-rich glycoproteins (PRG). The amino terminus contained repeating sequences of Ser-Gln-Gly-Pro-Pro-Pro-Arg-Pro-Gly-Lys-Pro-Glu-Gly-Pro-Pro-Pro- Gln-Gly that had significant compositional and sequence homology to that encoded by exon 3 of the PRB3 gene. We analyzed the PRG oligosaccharides by a combination of mass spectrometry techniques and nuclear magnetic resonance spectroscopy. Twenty-seven highly fucosylated structures were identified. The most abundant was as follows (where Fuc is fucose). (formula; see text) To understand the structural basis of F. nucleatum binding, we screened glycolipids and neoglycolipids carrying carbohydrate structures related to those of the PRG for receptor activity; components with unsubstituted terminal lactosamine residues best supported adherence. Neoglycolipids constructed from PRG oligosaccharides were also receptors. Treatment with beta-galactosidase, but not alpha-fucosidase, abolished binding, suggesting that unsubstituted lactosamine units, including the 6-antenna of the major oligosaccharide, mediate F. nucleatum adherence.     

Investigation of cis/trans proline isomerism in a multiply occurring peptide fragment from human salivary proline-rich glycoprotein.

The solution-state conformations of eight proline-containing peptide fragments found in human salivary proline-rich glycoprotein (PRG) were investigated in 2 x distilled water (treated with metal ion chelating resin) using 13C-nuclear magnetic resonance (NMR) and circular dichroism (CD) spectroscopy. The peptide sequences and acronyms were as follows: PRG9-2 = NH2-G(1)-P(2)-CONH2, PRG9-3 = NH2-G(1)P(2)-P(3)-CONH2, PRG9-4 = NH2-G(1)-P(2)-P(3)-P(4)-CONH2, PRG9-5 = NH2-G(1)-P(2)-P(3)-P(4)-H(5)-CONH2, PRG9-6 = NH2-G(1)-P(2)-P(3)-P(4)-H(5)-P(6)-CONH2, PRG9-7 = NH2-G(1)-P(2)-P(3)-P(4)-H(5)-P(6)-G(7)-CONH2, PRG9-8 = NH2-G(1)-P(2)-P(3)-P(4)-H(5)-P(6)-G(7)-K(8)-CONH2 and PRG9-9 = NH2-G(1)-P(2)-P(3)-P(4)-H(5)-P(6)-G(7)-K(8)-P(9)-CONH2. Sequence-specific resonance assignments from the 13C-NMR spectra indicated that the trans proline isomer dominated the conformations of the peptides. CD results clearly showed the presence of the poly-L-proline II helix as the major conformation in PRG9-3----PRG9-5, supplemented by beta- and/or gamma-turns in PRG9-6----PRG9-9. These data suggest that in "metal free" water, native PRG could contain several small poly-L-proline II helices along with beta- and/or gamma-turns. Since proline is the major amino acid present in native PRG, these localized conformations may contribute to PRG's global conformation and act as a primary force in determining its biological activities.     

Occurrence of Phialophora radicicola var. graminicola and Gaeumannomyces graminis var. tritici on roots of wheat in field crops

Assessments of Phialophora radicicola var. graminicola (PRG) and Gaeumannomyces graminis var. tritici (GGT) were made by culturing and by direct microscopic examination of pieces of seminal roots from 16 winter wheat crops grown in different cropping sequences and with different phosphate manuring. PRG occurred on all wheat crops, but was abundant only on wheat after grass, where it seemed to delay the onset of damaging take-all by 1 yr. Delayed occurrence of take-all by phosphate fertiliser was not related to differences in populations of PRG. Wheat grown in ‘take-all decline’ soils had only small amounts of PRG, indicating that the development and the decline of take-all epidemics may be influenced by different biological control mechanisms; breaking sequences of wheat crops by 1 yr grass leys might harness the advantages of both mechanisms.     

Immunolocalisation and expression of proteoglycan 4 (cartilage superficial zone proteoglycan) in tendon.

Cartilage superficial zone protein/proteoglycan (SZP) or proteoglycan 4 (PRG4), has been demonstrated to have the potential for several distinct biological functions including cytoprotection, lubrication and matrix binding. In the present study, we have examined both the immunolocalisation and the mRNA expression pattern of PRG4 in tissue harvested from the compressed and tensional regions of young and mature bovine tendons. Immunohistochemical analyses, utilizing monoclonal antibody 3-A-4 which recognizes a conformational-dependent epitope on native PRG4, demonstrated that PRG4 is present predominantly at the surface of fibrocartilaginous regions of tendon, with the intensity of immunoreactivity in this region increasing with age. RT-PCR analyses revealed that the expression of PRG4 mRNA can be modulated by exposure to cytokines and growth factors. In addition, analyses of human pathological tendon revealed that PRG4 may also be expressed as an alternatively spliced form lacking exons which encode part of the N-terminal matrix-binding and cell-proliferative domain; however, it remains to be determined whether such splice variants are a feature of human tendon, regardless of disease state. Taken together, these data indicate that PRG4 may play an important cytoprotective role by preventing cellular adhesion to the tendon surface as well as providing lubrication during normal tendon function, in a manner complimentary to cartilage PRG4. Structural modifications to SZP, together with a reduction in synthesis during tendon inflammation with injury and disease may account for the formation of tendon adhesions and contribute to the overall dysfunction of the tissue.     

Proteins of aleurone grains isolated from cotyledons ofLupinus luteus L.

Lupin aleurone grains from dry cotyledons contain about 87% of globulin including 57% of proteins solubilized by NaCl solution and 30% by NaOH extractions, and also about 12% of albumin. Activity of acid protease with the optimum at pH 3.6-3.8 was associated with albumin. Albumin was separated by gel filtration on Ultrogel AcA 22 into 3 fractions with mol. m. 179 000, 113 000, and below 100 000, respectively. The component with mol. m. 113 000 prevailed and comprised about 60% of albumin. Globulin was composed mainly of two major components with mol. m. above 364 000 and about 280 000. These correspond to α-and β-conglutin fractions, respectively. γ-conglutin minor globulin fraction of lupin aleurone grains was composed of about 66% glycoprotein. In contrast, about 82% of albumin lack glycosylation. Essentially, globulins in lupin aleurone grains are glycoprotein.     

Interaction between calcofluor white and carbohydrates of alpha 1-acid glycoprotein.

Interactions between the fluorescent probe, calcofluor white, and human serum albumin (HSA) and alpha 1-acid glycoprotein (orosomucoid) are compared. The two proteins have comparable isoelectric points, but alpha 1-acid glycoprotein is highly glycosylated (40% of glycans by weight), while the serum albumin is not. Binding of calcofluor to the proteins induces an increase in both the fluorescence anisotropy and the fluorescence intensity of the fluorophore. Also, we found that the calcofluor exhibits a fluorescence emission with a maximum located at 432, 415 or 445 nm, respectively, in the absence of proteins, in the presence of HSA, and in the presence of alpha 1-acid glycoprotein. The stoichiometries of the calcofluor-serum albumin and calcofluor-alpha 1-acid glycoprotein complexes are 2:1 and 1:1, respectively. The association constants are 0.04 and 0.15 microM-1, respectively. The calcofluor does not interact with Lens culinaris agglutinin (LCA), although the protein has a hydrophobic site. Nevertheless, one cannot exclude that the binding of the fluorophore to the HSA is nonspecific. Our results, when compared with those obtained with calcofluor dissolved in the hydrophobic solvent isobutanol, and with the fluorescent probe, potassium 6-(p-toluidino)-2-naphthalenesulfonate (TNS), bound to alpha 1-acid glycoprotein, indicate that the emission of calcofluor bound to HSA occurs from a hydrophobic state, while that of calcofluor bound to alpha 1-acid glycoprotein occurs from a hydrophilic state. The fluorescence intensity of calcofluor decreases in the presence of carbohydrates isolated from alpha 1-acid glycoprotein, while it increases in the presence of alpha 1-cellulose. Thus, calcofluor interacts mainly with the glycan moiety of alpha 1-acid glycoprotein, and its fluorescence is sensitive to the secondary structure of the glycans.     

Interaction between calcofluor white and carbohydrates of alpha 1-acid glycoprotein.

Interactions between the fluorescent probe, calcofluor white, and human serum albumin (HSA) and alpha 1-acid glycoprotein (orosomucoid) are compared. The two proteins have comparable isoelectric points, but alpha 1-acid glycoprotein is highly glycosylated (40% of glycans by weight), while the serum albumin is not. Binding of calcofluor to the proteins induces an increase in both the fluorescence anisotropy and the fluorescence intensity of the fluorophore. Also, we found that the calcofluor exhibits a fluorescence emission with a maximum located at 432, 415 or 445 nm, respectively, in the absence of proteins, in the presence of HSA, and in the presence of alpha 1-acid glycoprotein. The stoichiometries of the calcofluor-serum albumin and calcofluor-alpha 1-acid glycoprotein complexes are 2:1 and 1:1, respectively. The association constants are 0.04 and 0.15 microM-1, respectively. The calcofluor does not interact with Lens culinaris agglutinin (LCA), although the protein has a hydrophobic site. Nevertheless, one cannot exclude that the binding of the fluorophore to the HSA is nonspecific. Our results, when compared with those obtained with calcofluor dissolved in the hydrophobic solvent isobutanol, and with the fluorescent probe, potassium 6-(p-toluidino)-2-naphthalenesulfonate (TNS), bound to alpha 1-acid glycoprotein, indicate that the emission of calcofluor bound to HSA occurs from a hydrophobic state, while that of calcofluor bound to alpha 1-acid glycoprotein occurs from a hydrophilic state. The fluorescence intensity of calcofluor decreases in the presence of carbohydrates isolated from alpha 1-acid glycoprotein, while it increases in the presence of alpha 1-cellulose. Thus, calcofluor interacts mainly with the glycan moiety of alpha 1-acid glycoprotein, and its fluorescence is sensitive to the secondary structure of the glycans.     

Morphological study of rat thymus after tumoral, non-tumoral antigens and thiamine pyrophosphate administration.

The study was performed on 60 Wistar white rats sacrificed 24 hours and 7 days after administration of glycoprotein 8 (extracted from Jensen rat tumors), glycoprotein 5 (from normal rat blood), bovine albumin, tuberculin (PPD), and a complex of strychnine, picrotoxin and TPP administered alone or together with the above-mentioned antigens. In order to study thymus morphological changes, methods of quantitative morphology and current histological techniques were used. The thymus responds morphologically to the antigenic administration - (glycoprotein 8, bovine albumin and PPD) - by cortico-medullary ratio modification, expressed by decrease of cortical surface and increase of medullary one, thymocytes agglomeration in the perivascular sheaths and their deliverance in the blood flow. Under other conditions (strychnine, picrotoxin and TPP administration) thymus responds only by thymocytes agglomeration in the perivascular sheaths. Glycoprotein 5 does not induce any modifications in thymus structure.     

Lectin binding to gp60 decreases specific albumin binding and transport in pulmonary artery endothelial monolayers.

The effect of albumin binding to cultured bovine pulmonary artery endothelial cell (BPAEC) monolayers on the transendothelial flux of 125I-labelled bovine serum albumin (BSA) was examined to determine its possible role on albumin transcytosis. The transport of 125I-BSA tracer across BPAEC grown on gelatin- and fibronectin-coated filters (0.8 microns pore diam.) was affected by the presence of unlabelled BSA in the medium in that transendothelial 125I-BSA permeability decreased, reaching a 40% reduction at BSA concentrations equal to or greater than 5 mg/ml. BSA binding to BPAEC monolayers was saturated at concentration of 10 mg/ml with an apparent binding affinity of 6 x 10(-7) M. In contrast, gelatin added to the medium altered neither 125I-BSA binding nor transport. Several lectins were tested for their ability to inhibit 125I-BSA binding and transport. One lectin, Ricinus communis (RCA), reduced 125I-BSA binding by 70% and transport by 40%. Other lectins, Ulex europaeus, Triticum vulgare, and Glycine max decreased neither 125I-BSA binding nor transport. The reduction of 125I-BSA transport by RCA was not observed in the presence of saturating levels of BSA, indicating that RCA influenced only the albumin-dependent component of transport. RCA, but not other lectins, precipitated a 60 kDa plasmalemmal glycoprotein from cell lysates of surface radioiodinated BPAEC monolayers. This 60 kDa glycoprotein appears to be the equivalent of gp60 identified previously as an albumin binding glycoprotein in rat microvascular endothelium. In summary, approximately 40% of albumin transport across BPAEC monolayers is dependent on albumin binding. This component of albumin transport is inhibited by 80% by the binding of RCA to gp60. These results suggest that binding of albumin to gp60 on pulmonary artery endothelial cell membrane is a critical determinant of transendothelial albumin flux involving mechanisms such as plasmalemmal vesicular transcytosis.     

A novel p53-inducible apoptogenic gene,PRG3, encodes a homologue of the apoptosis-inducing factor (AIF)

The p53 tumor suppressor protein induces cell cycle arrest or apoptosis in response to cellular stresses. We have identified PRG3 (p53-responsive gene 3), which is induced specifically under p53-dependent apoptotic conditions in human colon cancer cells, and encodes a novel polypeptide of 373 amino acids with a predicted molecular mass of 40.5 kDa. PRG3 has significant homology to bacterial oxidoreductases and the apoptosis-inducing factor, AIF, and the gene was assigned to chromosome 10q21.3-q22.1. Expression of PRG3 was induced by the activation of endogenous p53 and it contains a p53-responsive element. Unlike AIF, PRG3 localizes in the cytoplasm and its ectopic expression induces apoptosis. An amino-terminal deletion mutant of PRG3 that lacks a putative oxidoreductase activity retains its apoptotic activity, suggesting that the oxidoreductase activity is dispensable for the apoptotic function of PRG3. The PRG3 gene is thus a novel p53 target gene in a p53-dependent apoptosis pathway.     

Effect of lipids and proteins on the viscosity of gastric mucus glycoprotein

The effect of associated lipids and covalently bound fatty acids, and the contribution of serum albumin and secretory IgA to the viscosity of dog gastric mucus glycoprotein was investigated. Using a cone/plate viscometer at shear rates between 1.15 - 230s -1, it was found that extraction of associated lipids from the glycoprotein lead to 80-85% decrease in the viscosity. Further loss (39%) in viscosity of the delipidated glycoprotein occurred following removal of covalently bound fatty acids. Reassociation of the delipidated glycoprotein with its neutral lipids increased the viscosity 3-fold, a 2.5-fold increase was obtained with glycolipids, and 2-fold with phospholipids. Preincubation of purified mucus glycoprotein with albumin or IgA resulted in the increase in viscosity. This increase in viscosity was proportional to albumin concentration up to 10%, and to IgA concentration up to 5%. The results show that interaction of lipids and proteins with mucus glycoprotein contributes significantly to the viscosity of gastric mucus.     

Proteome analysis of glandular parotid and submandibular-sublingual saliva in comparison to whole human saliva by two-dimensional gel electrophoresis

The secretions of the salivary parotid and submandibular-sublingual (SMSL) glands constitute the main part of whole human saliva (WS) in which proline-rich proteins (PRPs) and mucins represent dominant groups. Although proteome analysis had been performed on WS, no identification of PRPs or mucins by 2-DE and MS was achieved in WS and no comprehensive analysis of both glandular secretions is available so far. The aim of this study was to compare the protein map of WS to parotid and SMSL secretions for the display of PRPs and mucins. WS and glandular secretions were subjected to 2-DE and spots were analyzed by MALDI-MS. New components identified in WS were cyclophilin-B and prolyl-4-hydroxylase. Also acidic and basic PRPs as well as the proline-rich glycoprotein (PRG) could now be mapped in WS. Acidic PRPs were found equally in parotid and SMSL secretions, whereas basic PRPs and PRG were found primarily in parotid secretion. Salivary mucin MUC7 was identified in SMSL secretion. Thus, the more abundant proteins of WS can be explained mainly by mixed contributions of parotid and SMSL secretions with only few components remaining that may be derived from local sources in the oral cavity.     

C1q and C3bi binding activities of the components of a plasmin-treated human immunoglobulin preparation

Each of the three major components isolated from a commercial plasmin-treated human immunoglobulin preparation, namely, the plasmin-resistant 7S IgG fraction (PRG), Fab fragment and Fc fragment, was tested before and after heat treatment for binding C1q and fixing C3bi. In unheated state, only PRG was found to bind C1q, whereas none bound C3bi. The binding of C1q by PRG was enhanced by heat treatment which also conferred the activity of binding C3bi to PRG and to Fc fractions, From these results, anticomplementary activity of unheated PRG fraction seems to be due mainly to the complement activation via the classical pathway, whereas the activation by the heat-treated Fc fragment might be via an alternative pathway.     

Plasticity-related Gene 5 Promotes Spine Formation in Murine Hippocampal Neurons

The transmembrane protein plasticity-related genes 3 and 5 (PRG3 and PRG5) increase filopodial formation in various cell lines, independently of Cdc42. However, information on the effects of PRG5 during neuronal development is sparse. Here, we present several lines of evidence for the involvement of PRG5 in the genesis and stabilization of dendritic spines. First, PRG5 was strongly expressed during mouse brain development from embryonic day 14 (E14), peaked around the time of birth, and remained stable at least until early adult stages (i.e. P30). Second, on a subcellular level, PRG5 expression shifted from an equal distribution along all neurites toward accumulation only along dendrites during hippocampal development in vitro. Third, overexpression of PRG5 in immature hippocampal neurons induced formation of spine-like structures ahead of time. Proper amino acid sequences in the extracellular domains (D1 to D3) of PRG5 were a prerequisite for trafficking and induction of spine-like structures, as shown by mutation analysis. Fourth, at stages when spines are present, knockdown of PRG5 reduced the number but not the length of protrusions. This was accompanied by a decrease in the number of excitatory synapses and, consequently, by a reduction of miniature excitatory postsynaptic current frequencies, although miniature excitatory postsynaptic current amplitudes remained similar. In turn, overexpressing PRG5 in mature neurons not only increased Homer-positive spine numbers but also augmented spine head diameters. Mechanistically, PRG5 interacts with phosphorylated phosphatidylinositols, phospholipids involved in dendritic spine formation by different lipid-protein assays. Taken together, our data propose that PRG5 promotes spine formation.