Skip Navigation


Glycobiology Advance Access originally published online on March 16, 2007
Glycobiology 2007 17(7):784-794; doi:10.1093/glycob/cwm031
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
17/7/784    most recent
cwm031v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Sano, K.
Right arrow Articles by Ogawa, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sano, K.
Right arrow Articles by Ogawa, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org

Changes in glycosylation of vitronectin modulate multimerization and collagen binding during liver regeneration

Kotone Sano2, Kimie Asanuma-Date2, Fumio Arisaka3, Shunji Hattori4 and Haruko Ogawa1,2

2 Graduate school of Humanities and Sciences and The Glycoscience Institute, Ochanomizu University, 2-1-1 Otsuka, Bunkyo-ku, Tokyo, 112-8610 Japan
3 Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama, 226-8502, Japan
4 Nippi Research Institute of Biomatrix, 1-1, Senju-midoricho, Adachi-ku, Tokyo 120-8601, Japan

1 To whom correspondence should be addressed; Tel: 03-5978-5343; Fax: 03-5978-5343; E-mail: ogawa.haruko{at}ocha.ac.jp

Received on February 20, 2007; revised on February 20, 2007; accepted on March 9, 2007


   Abstract
Top
 Abstract
Introduction
 Results
 Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
Elucidating the mechanisms and factors regulating multimerization is biologically important in order to modulate the biological activities of functional proteins, especially adhesive proteins in the extracellular matrix (ECM). Vitronectin (VN) is a multifunctional glycoprotein present in plasma and ECM. Linkage of cellular adhesion and fibrinolysis by VN plays an essential role during tissue remodeling. Our previous study determined that the collagen-binding activity of VN was markedly enhanced with the decreased glycosylation during liver regeneration. This study demonstrated how alternations of glycans modulate the biological activity of VN. Human and rat VNs were used because of their similarities in structure and activities. The binding affinity of human VN to immobilized collagen was shown to be higher at pH 4.5 than at 7.5, at 37 °C than at 4 °C. Sedimentation velocity studies indicated that the greater the multimerization of human VN, the better it bound to collagen. The results indicate that the collagen binding of VN was modulated through its multimerization. Stepwise trimming of glycan with various exoglycosidases increased both the multimer size and the collagen binding of human VN, indicating that they are modulated by changes in glycosylation. The multimer sizes of VN purified from plasma of partially hepatectomized (PH) rats and sham-operated (SH) rats increased by about 45 and 31%, respectively, compared with those of nonoperated (NO) rats. In accordance with this, PH-VN exhibited remarkably enhanced collagen binding than SH-VN and NO-VN on surface plasmon resonance. In the PH rat sera, the multimer VN was increased in both amount and size compared with those in SH- and NO-sera. The results demonstrate that glycan alterations during tissue remodeling induce increased multimerization state to enhance the biological activity of VN.

Key words: vitronectin / multimerization / oligosaccharide / partial hepatectomy / collagen binding / multivalent effect


   Introduction
Top
 Abstract
 Introduction
Results
 Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
The extracellular matrix (ECM) surrounding cells plays an important role in many biological and pathological processes, including inflammation, tissue remodeling, and invasion by cancer cells (DeClerck et al. 2004Go). The ECM is composed of many kinds of adhesive glycoproteins that regulate cellular signaling, motility, and proliferation. During tissue remodeling, the glycoconjugates synthesized by cells are different from those of normal tissues owing to the changes in the expression of many proteins that are responsible for glycan synthesis, such as glycosyl transferases, glycosidases, and transporters. The nature of the link between such glycan changes and the process of tissue remodeling is still unclear.

One of the components of ECM, vitronectin (VN), is a multifunctional glycoprotein that originates mainly in hepatocytes and circulates in the blood stream at high concentrations (0.2 mg/mL in humans) (Schvartz et al. 1999Go). VN binds to various biological ligands and it plays a key role in tissue remodeling by regulating cell adhesion and cellular motility through binding to various types of integrins on the cell surface, and by connecting the cell behavior to the pericellular proteolysis through binding with type-1 plasminogen activator inhibitor (PAI-1), urokinase-type plasminogen activator, and urokinase receptor (Preissner and Seiffert 1998Go; Schvartz et al. 1999Go). VN also regulates blood systems related to protease cascades, such as cell lysis by complements, blood coagulation, and fibrinolysis (Preissner 1991Go; Seiffert 1997Go) through interaction with heparin and thrombin–antithrombin III complexes (Podack et al. 1977Go; McKeown-Longo and Panetti 1996Go).

Tissue VN is an active multimeric form believed to interact with various matrix ligands including proteoglycans and collagen (Preissner and Seiffert 1998Go), whereas most VN in plasma is an inactive monomer form and does not bind to these ligands (Izumi et al. 1989Go). VN acquires binding activities through a conformational transition of the native inactive form to an active form after treatment with urea, heating, or in the presence of certain ligands, such as heparin or membrane attack complex in vivo (Gebb et al. 1986Go; Yatohgo et al. 1988Go). That the activation of VN is accompanied by transition to a multimer from the inactive monomer has been shown by ultracentrifugation and gel-filtration analyses (Zhuang, Blackburn, et al. 1996Go; Zhuang, Li, et al. 1996Go). Understanding the mechanism by which VN functions in various biological events has been difficult, however, due to its conformational lability and the overlapping of most of its ligand-binding activities with those of other matrix molecules.

Recently, mice with a genetic deletion of VN have shown altered responses to tissue injury, i.e., increased wound fibrinolysis and decreased angiogenesis, that resulted in delayed wound healing (Jang et al. 2000Go). This indicates the importance of VN as a multifunctional glycoprotein in tissue remodeling after injury by binding with matrix molecules and cell surface receptors. However, the binding activity of VN to matrix ligands remains ambiguous in that the collagen-binding activity of VN has been observed to be slight under physiological conditions (Gebb et al. 1986Go). In our preceding studies, the binding activities of human and porcine VN to collagen and sulfatide were found to be affected by changes in the glycosylation of VNs in vitro (Yoneda et al. 1998Go). In particular, collagen binding was markedly enhanced by the absence of N-glycans covalently linked to VNs.

We separately found that activation of rat VN occurs during liver regeneration after partial hepatectomy. An increase in the collagen-binding activity synchronized with the glycan changes in vivo (Uchibori-Iwaki et al. 2000Go). The molecular mass of VN purified from partially hepatectomized rats (PH-VN) at 24 h had shrunk to 65 kDa compared with the 68–69 kDa of VNs from nonoperated (NO) and sham-operated (SH). The reduction of molecular mass was attributed to the decrease in carbohydrate concentration of PH-VN, which was two-thirds of SH-VN and one-third of NO-VN. This was accompanied by changes in N-glycosylation, sialylation, and isoelectric point (pI), which shifted to 6 from 4 of NO-VN and slightly above 4 of SH-VN rats, while the amino acid composition did not change significantly among the three VNs. Related to the glycosylation changes, PH-VN exhibited three times higher binding to type I collagen than NO-VN and about two times that of SH-VN by enzyme-linked immunosorbent assay (ELISA) (Nakashima et al. 1992Go). There are three, two, and four potential N-glycosylation sites for human, porcine, and rat VNs, respectively, and the potential sites of human and porcine VNs were shown to be fully glycosylated in normal animals (Zheng et al. 1995Go). Because the positions of the glycosylation sites and the presence of complex-type sialylated N-linked glycans are highly conserved among mammalian VNs (Kitagaki-Ogawa et al. 1990Go; Nakashima et al. 1992Go), the effects of glycosylation on the biological activity may be common among mammalian VNs.

How does glycosylation activate the multifunctional glycoprotein VN during liver regeneration? Recently, changes in glycosylation have been found to regulate various biological events through conformational effects or signaling functions. The present study attempted to elucidate how changes in glycans modulate the biological activity of VN during liver regeneration. The findings provide a novel insight into the significance of protein glycosylation in regulating the biological activities expressed through the multimerization process.


   Results
Top
 Abstract
 Introduction
 Results
Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
In this study, investigation of the activation mechanism was performed mainly using human VN due to the similarities between human and rat VNs and because rat VN cannot be obtained in sufficient quantities to perform the various ultracentrifugal analyses throughout. This study simultaneously provided several lines of evidence to demonstrate the common properties between human and rat VNs in the relationship between collagen-binding activity and glycosylation.

Factors affecting binding of VN to collagen
Human VN bound to type I collagen in a concentration-dependent manner, and the binding affinity of VN was much higher at 37 °C than 4 °C throughout the pH range examined (Figure 1A and B). As shown in Figure 1B, the collagen binding showed a sharp maximum at pH 4.5 in the range of pH 3.5–8.5. As shown in Figure 1C, the binding of VN to collagen at pH 7.5 was sensitive to ionic strength, indicating that the interaction at pH 7.5 depends largely on electrostatic interactions. At pH 4.5, which is close to the isoelectric point of human VN, pI = 4, the binding was maximum around the physiological ionic strength, suggesting that other binding forces, such as hydrophobic interaction, in addition to electrostatic interaction were involved in the binding at this pH.


Figure 1
View larger version (9K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Effect of concentration, temperature, pH, ionic strength, and aggregation of collagen on the collagen binding of VN. Microtiter wells were coated with type I collagen and incubated with purified human VN under various conditions. The bound VN was measured by ELISA using sheep anti-human VN IgG and HRP-conjugated rabbit anti-sheep IgG as described in the text. (A) Concentration dependency of VN at 37 °C (bullet) and 4 °C ({Delta}). (B) Effect of pH at 37 °C (bullet) and 4 °C ({Delta}). (C) Effect of ionic strength at pH 4.5 ({blacksquare}) and pH 7.5 ({circ}). (D) Effect of alkaline treatment of collagen. Microtiter wells were coated with type I collagen that had been alkaline-treated to dissociate it to monomeric collagen (solid line) or untreated to remain in aggregated form (dashed line) and incubated with human VN at 37 °C. The binding activity of human VN was measured by ELISA at various pH. The data represent the average of four experiments.

 
Collagen monomers aggregate under neutral pH at more than 20 °C, but alkaline treatment of collagen changes the charge of the peptide to form a stable monomer at neutral pH at room temperature (Nakashima et al. 1992Go). The effect of collagen aggregation on the binding of VN is shown in Figure 1D. The binding of VN to alkaline-treated collagen was diminished at neutral pH, indicating that the aggregation of collagen is essential for interaction with VN.

Differential s-value of VN under various temperatures and pH
Activation of VN has been shown to be accompanied by transition to a multimer from the inactive monomer (Gebb et al. 1986Go; Izumi et al. 1989Go). Therefore, we measured the relationship between the multimer size of VN and collagen binding under various conditions by analytical ultracentrifugation. As shown in Figure 2A, the sedimentation velocity indicated that the s-values of the peak position in c(s), distribution funciton of sedimentation coefficients, becomes larger as the temperature increases, s20,w = 17 S (4 °C), s20,w = 18 S (20 °C), and s20,w = 21 S (37 °C). Monomer human VN was reported to have a s20,w of 4.1 S (20 °C) (Zhuang, Blackburn, et al. 1996Go). VN at 37 °C, a better temperature for collagen binding than 4 °C (Figure 1A and B), exhibits a larger s-value for the multimer than at 4 °C.


Figure 2
View larger version (9K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Distribution of s-values of VN changes by temperature, pH, and deglycosylation. (A) Solutions of human VN in PBS containing 5 mM EDTA (A280 = 0.2) were centrifuged for sedimentation velocity analysis at 4 °C (solid line), 20 °C (dashed line), or 37 °C (dotted line) at pH 7.5 as described in the text. (B–D) The effects of enzymatic deglycosylation of VN on s-value. Human VN treated with neuraminidase in citrate-phosphate buffer (B and C, solid line), PNGase F in PBS–EDTA (D, solid line), or controls incubated without enzyme (dashed line) were centrifuged at 20 °C at pH 4.5 (B), and pH 7.5 (C, D). The time-dependent moving boundaries recorded at 280 nm were fitted to the numerical solution of the Lamm equation by using the program Sedfit in order to obtain the distribution function of the apparent sedimentation coefficient s, c(s).

 
In our preceding studies, the collagen-binding activity of VN was enhanced by treatment with neuraminidase or peptide N-glycosidase (PNGase F) in vitro (Yoneda et al. 1998Go). As shown in Figure 2B–D, a new component peak of larger s values appeared in the c(s) profiles of neuraminidase or PNGase F treated VNs when compared with those of control VN. At pH 4.5, neuraminidase-treated VN contained molecular species with a high s-value of 22 S (Figure 2B), suggesting a clear effect of glycan structure on the multimerization of VN. As shown in Figure 2B and C, neuraminidase-treated VN formed multimers of larger s-value, 52 S, at pH 4.5 than at pH 7.5. Together with the results shown in Figure 1, these results indicate that the collagen-binding activity correlates with the degree of multimerization of VN, i.e., the interaction between VN and collagen must be affected by the multimerization of VN.

Effects of glycosidase-trimming on multimerization and collagen-binding of VN
Various glycans on VN were trimmed stepwise with exoglycosidase to elucidate the effect of alterations of glycans on collagen binding, as indicated in Figure 3A. The involvement of disulfide linkage in the multimerization of VN was estimated from sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) under reducing and nonreducing conditions using 5–20% gradient gel. From the intensities of bands (Figure 3B), the relative amounts of monomer (65 and 75 kDa) and multimer VNs (top of the gel) were calculated by using Scion Image. As shown in Figure 3C, the intensities of multimer bands were remarkably increased by neuraminidase treatment. The results indicate that formation of multimer VN by intermolecular disulfide bonds was gradually increased with stepwise glycan trimming. Supportingly, as shown in Figure 3D, the amount of sediment from glycosidase-trimmed VNs was gradually increased with the glycan trimming at pH 4.5, but the sediment was solubilized by adding 0.1 M NaOH to pH 11, irrespective of the glycan trimming of VN in the five tubes, showing that the sedimentation is caused by secondary aggregation by ionic interaction due to the charges of the protein surface. These results indicate that VN aggregation was enhanced as the stepwise glycan trimming proceeded. It suggests that the glycans of intact VNs contribute to the suppression of multimer formation and the deglycosylation of VN leads to promotion of multimerization.


Figure 3
View larger version (23K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. Multimerization and sedimentation of VN changes by glycosidase-trimming. (A), Cleavage sites of enzymatic deglycosylation in the structure of the major N-glycan of human VN. (B) Glycosidase-trimmed human VNs (5 µg/lane) were loaded on each lane of 5–20% polyacrylamide gel and run for SDS–PAGE in the presence or absence of 2-mercaptoethanol (2ME). The gel was stained with Coomassie brilliant blue (CBB). (C) Relative amount of multimer (white bar) to monomer (black bar) under nonreducing conditions was calculated from the band intensities using the software program Scion Image, and expressed by taking the intensities of the monomer band of each VN under reducing condition as 100%. (D) Sedimentation of glycosidase-trimmed VNs at pH 4.5. Right, pH 11. U, Untreated VN; C, control VN incubated without enzyme; N, VN treated with neuraminidase; G, VN treated with neuraminidase and ß-galactosidase; H, VN treated with neuraminidase, ß-galactosidase and ß-N-acetylhexosaminidase; NG, VN treated with PNGase F.

 
As shown in Figure 4, the increase in multimerization of VN by glycan trimming accompanied the enhancement of average molecular weights as measured by sedimentation equilibrium. The multimer size of human VN had increased by 1.5-fold on total deglycosylation of intact VN (Figure 4B). These results may indicate that the enzymatic deglycosylation of VN leads to the promotion of multimer formation in both size and amount.


Figure 4
View larger version (30K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. Molecular weights of glycosidase-trimmed VNs by sedimentation equilibrium. Sedimentation equilibrium analysis of deglycosylated human VNs was performed at a concentration of 0.3 mg/mL in PBS containing 5 mM EDTA at 20 °C. Data were collected after equilibrium had been established at rotor speeds of 3000 and 5000 rpm by absorbance at 280 nm, and the data were analyzed using program, "nonlin" to calculate the averages molecular weights. (A) Radial distribution of absorbance at equilibrium at 5000 rpm for VNs. Individual absorbance readings are shown by open circles, with solid lines through the data representing the fit to a single species model. The residuals for the fit are shown in the upper panel. (B) The calculated weight average molecular weights of VNs. U, Untreated VN; C, control VN incubated without enzyme; N, VN treated with neuraminidase; G, VN treated with neuraminidase and ß-galactosidase; H, VN treated with neuraminidase, ß-galactosidase and ß-N-acetylhexosaminidase; and NG, VN treated with PNGase F.

 
As shown in Figure 5, the binding of VN to collagen was found to gradually increase with stepwise glycan trimming. The increase due to treatments with neuraminidase and PNGase F was especially remarkable. The observation indicates that the collagen binding activity correlated with the extent of multimerization of VN and that the changes in glycosylation altered the collagen-binding activity of VN through a multivalent effect.


Figure 5
View larger version (20K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5. The collagen-binding activities of glycosidase-trimmed VN. Microtiter wells were coated with type I collagen and incubated with glycosidase-trimmed human VNs. The amounts of bound VNs were measured with ELISA at pH 7.5 at 4 °C, as described in the text. The data are averages of four experiments. U, Untreated VN ({lozenge}); C, control VN incubated without enzyme ({diamondsuit}); N, VN treated with neuraminidase ({square}); G, VN treated with neuraminidase and ß-galactosidase ({blacktriangleup}); H, VN treated with neuraminidase, ß-galactosidase and ß-N-acetylhexosaminidase (*); and NG, VN treated with PNGase F ({circ}). The data of untreated VN and control VN completely overlapped each other.

 
Multimerization of NO-, SH-, and PH-VNs
VNs were purified from rat plasma before and after partial hepatectomy, and the multimerization degree of the purified VN was analyzed. As shown in Figure 6A, SDS–PAGE under reducing and nonreducing conditions showed that multimer VNs were completely dissociated by 2-mercaptoethanol, indicating that multimers are cross-linked by disulfide linkage. The intensities of each band under the nonreducing condition were quantified by using Scion image (Figure 6B). The amount of multimers increased more in PH-VN (91%) than in NO (79%) and SH-VNs (82%).


Figure 6
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 6. Amount of rat VN multimer changes after partial hepatectomy (A and B) and due to enzymatic deglycosylation (C and D). (A) NO-, SH-, and PH-VN (each 5 µg) were loaded on each lane of 7.5% polyacrylamide gel and run for SDS–PAGE in the presence or absence of 2-mercaptoethanol (2ME). The gels were stained with CBB. (C) Neuraminidase-, PNGase F-treated or untreated VNs (each 4 µg) were subjected to SDS–PAGE under the same condition as in (A). (B) and (D) The relative amount of multimer (white bar) to monomer (black bar) was calculated from the intensities of each band in (A) and (C), respectively, under nonreducing conditions as described in Fig. 3, and expressed by taking the intensities of monomer band of each VN under reducing condition as 100%. Abbreviation used in (C) and (D): N, VN treated with neuraminidase; and NG, VN treated with PNGase F; C, control VN incubated without enzyme; and U, Untreated VN.

 
Whether the relationship between the glycosylation change and multimerization that was observed for human VN holds for rat VN or not, the effects of enzymatic deglycosylation of NO-VN on the multimerization were examined. As shown in Figure 6C and D, both neuraminidase- and PNGase F-treated VNs showed an increase in the ratio of multimers to monomers compared with the control VN on SDS–PAGE, similar to that observed for human VN (Figure 3). Therefore, the conclusion that the glycosylation change causes the increases of multimer VN and collagen binding, which were obtained by using human VN, was shown to be a common attribute of rat and human VNs.

The weight average molecular weights of multimer sizes of NO-, SH-, and PH-VNs were measured by sedimentation equilibrium. As shown in Figure 7, the weight average molecular weights of PH-, and SH-VNs were 420 and 380 kDa, respectively, which were increases of 45 and 31%, respectively, compared with that of NO-VN, 290 kDa. The degrees of polymerization of PH-, SH-, and NO-VNs were calculated to be 7.0, 6.1, and 4.2-mer, respectively, on the basis of the molecular weights of each VN monomer, 68.8, 61.9, and 59.6 kDa, respectively, that were calculated from the cDNA sequence and sugar composition (Yoneda et al. 1998Go). The multimerization states of VNs indicate that the size and number of multimers are increased in PH-VN, which would be attributable to the changes in glycosylation according to the results of glycosidase-trimming of VN (Figures 3 and 4).


Figure 7
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 7. Molecular weights of NO-, SH-, and PH-VN by sedimentation equilibrium. Sedimentation equilibrium analysis of NO-, SH-, and PH-VNs was performed at 20°C. Data were collected and analyzed as described in Fig. 4. (A–C) Show the radial distribution of absorbance at equilibrium at 5,000 rpm for NO- (A), SH- (B), and PH-VN (C). Individual absorbance readings are shown by open circles, with solid lines through the data representing the fit to a single species model. The residuals for the fit are shown in the upper panels. (D) shows the calculated weight average molecular weight (bars) and degree of polymerization of VNs (diamonds). NO, NO-VN; SH, SH-VN; and PH, PH-VN.

 
SPR measurements of collagen-binding activity of NO-, SH-, and PH-VNs
Previously, we reported that ELISA showed that the collagen-binding activity of VN increased three-fold with the glycan change after partial hepatectomy (Yoneda et al. 1998Go). In this study, the interaction of NO-, SH-, and PH-VN with type I collagen was quantitatively analyzed by SPR. As shown in Figure 8A, the resonances were monitored when NO-, SH-, or PH-VN was placed in contact with immobilized collagen. The amount of type I collagen immobilized on one flow cell of a sensor chip was 4312 RU. As shown in Figure 8A–C, the binding of collagen to each VN was concentration-dependent, and the dissociation was very slow. Among the three VNs, PH-VN bound best to type I collagen. Under these conditions, VN is in a multimer form (Figure 7), and multivalent–multivalent interaction with type I collagen cannot be analyzed to obtain kinetic parameters. However, when injected into a single flow cell at the same concentration, the amount of each bound VN on the sensor chip correlates with the total affinity of VN to the collagen immobilized on the sensor chip. To calculate the bound VN (femtomole/square millimeters), the maximum responses monitored at the end of the injection phase were divided by the molecular weight of each VN monomer, as shown in Figure 8D. The relative affinity per monomer of PH-VN is remarkably high compared with NO- and SH-VNs, especially at the lower concentrations. The results support the relative binding activities of VNs to type I collagen suggested by ELISA (Yoneda et al. 1998Go) and indicate that the binding activity of PH-VN, which is much enhanced compared with NO- and SH-VNs, is due to a multivalent effect.


Figure 8
View larger version (38K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 8. Interaction between NO, SH-, or PH-VNs and type I collagen by SPR. Type I collagen was immobilized on a CM5 sensor chip as described in the text. NO- (A), SH- (B), and PH-VN (C) in PBS were injected onto the sensorchip at concentrations of 20, 40, 60, 80, or 100 nM at a flow rate of 20 µL/min at 20°C. The response is expressed as the change of resonance units induced by the binding of each VN to the collagen-immobilized flow cell, which was corrected for bulk effect by subtracting the change on the BSA-immobilized reference cell. Sensorgrams of NO- (A), SH- (B), and PH-VN (C). (D) Amounts per each VN monomer (femtomole/square millimeters) bound at various concentrations.

 
Elution profile of NO-, SH-, and PH-VNs in sera on gel-filtration chromatography
Most VN in plasma has been reported to be an inactive monomer form, and a small portion of plasma VN is in an active multimer form (Izumi et al. 1989Go). During the purification in this study, the originally active molecules were removed by the first heparin affinity chromatography, and then the originally inactive VN in serum was activated with urea during the purification procedure to be purified by the second heparin affinity chromatography. Therefore, the sizes and amounts of VNs in native NO-, SH-, and PH- sera were analyzed by gel-filtration high-performance liquid chromatography (HPLC) in combination with ELISA detection to elucidate the state of the VN multimers in each serum. As shown in Figure 9A, immunological detection of the eluted fractions using anti-VN immunoglobulins G (IgGs) indicated that most VN was eluted in the major protein peak that corresponds to a monomer VN. The major protein peak of each serum was detected at A280 at the elution time of 11.3 min in HPLC chromatograms, which corresponds to a molecular weight of about 65 kDa. The difference in the three sera was found in the elution positions of multimer VN at the elution time of 6–11 min, corresponding to more than 150 kDa. The amount of multimer VN was increased in the PH serum compared with those in NO and SH sera. A peak of significantly larger molecular weight was eluted at 6.5–7 min in only PH serum (Figure 9A). There was little difference between the elution patterns of NO and SH sera, except that the VN eluted at 10–11 min was a little lower in SH serum.


Figure 9
View larger version (39K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 9. Gel-filtration HPLC of serum VN from NO, SH, and PH rats. (A–C), rat sera were diluted to one-fourth, and a 50 µL aliquot was directly applied to a Shodex Protein KW-804 column (8 mm x 300 mm) and eluted with PBS containing 5 mM EDTA. The eluted fractions were analyzed by ELISA to detect VNs. The elution profile was monitored by absorbance at 280 nm for protein, as indicated by dashed lines and the amount of rat VNs were detected by ELISA, as indicated by solid lines. Samples are: NO (A), SH (B), and PH (C) rat sera. (D–F), Each serum was applied to a heparin-Sepharose column to separate it into the heparin-bound (bound and eluted with 0.5 M NaCl) and unbound (pass-through) fractions. Aliquots (50 µL) of the heparin-bound fraction, which contains active VNs, and heparin-unbound fraction, which contains inactive VNs, were applied separately to the HPLC column. Active VNs (dashed-and-dotted lines) and inactive VNs (dotted lines) in the heparin-unbound fractions were detected by ELISA. Samples are fractions of NO- (D), SH- (E), and PH- (F) rat sera. Arrows indicate the elution position of molecular weight standards: 670 kDa (thyroglobulin), 158 kDa (bovine gamma globulin), 44 kDa (chicken ovalbumin), 17 kDa (equine myoglobin), and 1.35 kDa (vitamin B-12).

 
After serum proteins were separated into heparin-bound and unbound fractions by heparin affinity chromatography, each fraction was applied to gel-filtration HPLC. As shown in Figure 9B, active VNs in the heparin-bound fraction of NO and SH sera were eluted in multimer form at around 9.2 min, which corresponds to around 670 kDa, about 10-mer of VN. The active VN in PH serum was eluted faster by 0.3 min than those of NO and SH-VNs, suggesting that the multimer size of VN is larger in PH serum than in NO and SH sera, whereas the inactive VNs of the heparin-unbound fraction were mainly monomers in all three sera (Figure 9B). The ratio of active VN to total VN in each serum was estimated to be 25, 21, and 21% for PH, SH, and NO sera, respectively, from the integrated peak areas in Figure 9A. These results indicate that active multimer VNs are a larger size and higher numbers in PH plasma than in SH and NO sera.


   Discussion
Top
 Abstract
 Introduction
 Results
 Discussion
Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
This study clarified the molecular mechanism by which ECM glycoprotein is activated by changes in glycans by analyzing VNs purified from PH-, SH-, and NO-rats in comparison with the deglycosylated VNs. PH-VN exhibited increased amounts and sizes of multimers compared with NO- and SH-VNs (Figures 6, 7, and 9) and markedly enhanced collagen binding (Figure 8). These changes in the multimerization and activity of PH-VN were reproduced by deglycosylation of VN, especially desialylation (Figures 3–6), and the relationship between glycosylation and the multimerization state was elucidated. Together with our previous observation that the carbohydrate concentration of PH-VN decreased to one-third of that of NO-VN and that the oligosaccharide structure changed from those of SH- and NO-VNs (Uchibori-Iwaki et al. 2000Go), the enhanced collagen binding of PH-VN is, therefore, attributable to a mechanism very similar to that caused by enzymatic deglycosylation.

VN multimers cross-linked via intermolecular disulfide bonds were shown to be increased in PH-VN compared with NO-VN (Figures 8 and 9) or on enzymatic deglycosylation of VN (Figure 3B). The formation of cross-linked multimers of VN was attributable to free sulfhydryl residues, Cys411 and Cys453, in the hemopexin I and II domains (Gibson et al. 1999Go), respectively. According to previous studies on activation of VN, native VN acquires binding activities to various biological ligands through a conformational transition and subsequent multimerization in the presence of heparin, thrombin-AT-III, membrane attack complex, or due to chemical treatments (McKeown-Longo and Panetti 1996Go; Preissner 1991Go). It was demonstrated that the decrease in multimer VN was correlated with the increase in ionic strength by gel filtration (Zhuang, Li, et al. 1996Go), suggesting that multimerization proceeds with the help of ionic interaction. The changes in glycosylation of VN after partial hepatectomy may contribute electrostatically and sterically to the multimerization process to enhance VN activity because it included both changes of sialylation and other intrinsic structures (Nakashima et al. 1992Go).

The interaction between VN and collagen is known to be sensitive to ionic strength (Gebb et al. 1986Go; Izumi et al. 1988Go), and VN had been considered to be inactive with collagen under physiological saline concentrations (Izumi et al. 1988Go). This study indicated that VN in a changed glycosylation state binds well to collagen under the physiological concentration, and that binding was enhanced by the multivalent effect caused by increased multimerization (Figures 3–6). Besides the multimerization of VN, electrostatic interaction is changed by glycosylation, especially sialylation. As shown in Figure 1C, the collagen binding of human VN was less affected by ionic strength at pH 4.5 than at pH 7.5, because pH 4.5 is close to the pI of mammalian VN. The pI of PH-VN changed to 6 from the 4 of NO-VN and slightly above 4 of SH-VN. The elevated pI of PH-VN is attributable mainly to the reduced sialylation (Yoneda et al. 1998Go), which would reduce the contribution of ionic interaction to the collagen at neutral pH. In this way, the change in glycosylation, especially sialylation alters the binding force between VN and collagen.

On the other hand, the effect of collagen aggregation on the interaction is notable, too. Collagen is known to aggregate at neutral pH and room temperature, but alkali-treated collagen aggregates at acid pH (Hattori et al. 1999Go). The binding of multimer VN to alkali-treated collagen was significantly reduced at neutral pH compared with binding to untreated collagen, but it recovered at pH 4 (Figure 1D). Thus, VN-collagen interaction is significantly affected by aggregation of collagen as well as the multimerization of VN, indicating that multivalencies of both VN and collagen are required for the interaction.

There have been several cases reported in which the oligosaccharide plays an important role in the interaction of the protein molecule. For example, the N-type glycan is necessary to form an assembly complex between the {alpha} and ß subunits of integrins (Zheng et al. 1994Go). Particularly, the oligosaccharides of integrin {alpha}5ß1 are involved in adhesion of several cells to fibronectin, suggesting that a specific glycan structure is required (Pretzlaff et al. 2000Go; Nadanaka et al. 2001Go; Isaji et al. 2004Go). On the other hand, examples in which the multimerization relates to the function of the molecule are also reported. For example, integrin, described earlier, is known to be activated following clustering (Li et al. 2003Go). However, very few examples have proven that the glycosylation is related to functional multimerization of the molecule in ECM. As far as we know, VN is the first case of it.

Human and rat VN have 73% sequence identity with respect to the common domain organization and ligand binding activities. All three glycosylation sites of human VN (Asn 86, 169, and 242) are conserved in rat VN, which has one additional glylcosylation site (Asn 96). Therefore, the effects of glycosylation change on the activity of human VN are expected to be similar to those of rat VN, besides the possibility that an additional effect might be exhibited by glycosylation at Asn 96. This study showed that the change in glycosylation frequently increased the ratio and size of multimers in human and rat VNs and demonstrated the mechanism of how VN shows increased collagen binding due to glycan changes. We also observed that collagen binding increased by neuraminidase or PNGase F treatments for porcine VN (A Yoneda et al., unpublished results), which suggests that this property is common to mammalian VNs, in spite of the fact that the number of potential N-glycosylation sites is different among these VNs.

VN is known to be related to wound healing and to the progression of hepatic fibrosis and cirrhosis (Inuzuka et al. 1992Go; Kobayashi et al. 1994Go). The collagen binding of VN must be significantly charged for these biological and pathological processes to occur because the collagen binding of VN provides a scaffold for the progression of these events. Increased depositions of VN have been observed in areas of hepatic injury and necrosis, whereas little VN immunoreactivity is detectable in most normal livers. Although the targeted deletion of VN in mice resulted in normal development and normal coagulation parameters (Zheng et al. 1995Go), the VN-null mice showed increased fibrinolysis and decreased angiogenesis on tissue injury, causing delayed wound healing (Jang et al. 2000Go). These observations indicate the importance of VN in tissue remodeling by binding with matrix molecules such as collagen, PAI-1, urokinase-type plasminogen activator, urokinase receptor, and integrin that are involved in the remodeling process. It well coincides with the observations that VN binds with fibrinolytic factors such as fibrinogen, thrombin, and PAI-1 (Preissner and Seiffert 1998Go), and that expression of VN increases in acute phases (Seiffert 1997Go). We propose that alteration of glycosylation is very significant in modulation of the biological activity of VN during these tissue-remodeling processes. Modulation of glycans on VN could contribute to development of a strategy to regulate matrix deposition and fibrinolysis after fibrosis.


   Materials and methods
Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
Conflict of interest statement
 Acknowledgments
 References
 
Materials
Rat and human VNs were purified from each plasma by two-step heparin affinity chromatography before and after urea treatment as described previously (Yatohgo et al. 1988Go). Neuraminidase (Vibrio cholerae) and recombinant peptide-N4-(N-acetyl-ß-D-glucosaminyl) asparagine amidase from Fravobacterium meningosepticum, (PNGase F) were purchased from Roche Diagnostics (Mannheim, Germany). ß-galactosidase (jack bean) and ß-N-acetylhexosaminidase (jack bean) were purchased from Seikagaku Corporation (Tokyo, Japan). Unconjugated rabbit anti-human VN IgGs and horseradish peroxidase (HRP)-conjugated sheep anti-rabbit IgGs were purchased from The Binding Site (Birmingham, UK). Type I collagen from calfskin and other reagents of special grade were from Wako (Osaka, Japan). Alkali-treated type I collagen, a monomeric form that does not aggregate even at neutral pH and 37°C, was prepared as described previously (Hattori et al. 1999Go). Psathyrella velutina lectin was prepared in our laboratory as previously described (Ueda et al. 1999Go). HRP-labeled Con A was purchased from Seikagaku Corporation.

Animals
Male Wistar rats aged 5 weeks (weighing about 110 g; Nihon Clea, Tokyo, Japan) were maintained at a constant temperature (23.5 °C) with 12 h each light (6:00–18:00) and darkness. Two-thirds partial hepatectomy was performed under diethyl ether anesthesia as described previously (Higgins and Anderson 1931Go). Sham-operated rats were anesthetized, and their livers were completely exposed outside the peritoneum and manipulated but not resected.

Glycosidase digestion of VN
Enzymatic deglycosylation of VN was performed as described previously (Uchibori-Iwaki et al. 2000Go). For desialylation, VN (32 µg/200 µL) was dialyzed against 50 mM acetate buffer containing 4 mM CaCl2 (pH 5.5) and treated with neuraminidase (0.001 U) at 37 °C overnight. For stepwise glycan trimming, VN was treated with neuraminidase alone, neuraminidase and ß-galactosidase, or neuraminidase, ß-galactosidase, and ß-N-acetylhexosaminidase in 50 mM acetate buffer containing 4 mM CaCl2 (pH 5.5). For de-N-glycosylation, VN (32 µg/200 µL) was dialyzed against 10 mM phosphate buffer (pH 7.5) containing 0.13 M NaCl and 5 mM ethylenediaminetetraacetic acid (EDTA) [phosphate-buffered saline (PBS)–EDTA], and digested with PNGase F (0.2 U) at 37 °C for 48 h. The treatments of VN were performed without the enzymes for controls. The treated VNs were dialyzed against 10 mM Tris-buffer (pH 7.5) containing 0.14 M NaCl [tris-buffered saline (TBS)]. Desialylation or de-N-glycosylation was ascertained by both the behavior on SDS–PAGE and the loss of reactivity with biotinylated P. velutina lectin for desialylation and ahexosaminylation (Ueda et al. 2002), or HRP-Con A on western blotting.

Analytical ultracentrifugation
Sedimentation velocity and equilibrium measurements were performed with a Beckman Optima XL-A analytical ultracentrifuge (Beckman Instruments, Palo Alto, CA) in a 8-hole An50Ti rotor with standard double-sector centerpieces and quartz windows at 4, 20, or 37°C. Concentration profiles of samples were monitored by absorbance at 280 nm. Sedimentation velocity experiments were performed using VN solutions of 0.2 mg/mL in PBS–EDTA (pH 7.5) or citrate-phosphate buffer (pH 4.5, 7.5) at a rotor speed of 40 000 or 45 000 rpm. Data acquisition was performed without time intervals with about 50 scans collected per run. The data were analyzed using a software program, Sedfit (Schuck 1998Go), to obtain the distribution function of the apparent sedimentation coefficient, s; c(s). Sedimentation equilibrium experiments were performed using 0.3 mg/mL VN in PBS–EDTA at 20 °C at rotor speeds of 3000 and 5000 rpm. Scans were recorded every 2 h, and the equilibrium of the system was judged by the superimposition of the last three scans. The two datasets were globally fitted to a single species model to determine the weight average molecular weight using the "nonlin" program preinstalled by the manufacturer. The partial specific volumes, 0.6768 cm3/g, for VN were based on the amino acid and sugar compositions of the glycoprotein.

Assays of binding of VN to immobilized type I collagen
Collagen-binding activities were studied by ELISA essentially according to the method reported previously (Gebb et al. 1986Go; Yoneda et al. 1998Go) at 4 °C or 37 °C. Briefly, the wells of microtiter plates (Immulon 1, Dynatech Laboratories Inc., Chantilly, VA) were coated with 100 µL aliquots of a solution of type I collagen (10 µg/mL in 0.1 M carbonate buffer, pH 9.0) for 3 h. The wells were washed and blocked with 0.5% skim milk in TBS (pH 7.5) for 2 h. Various concentrations of human or rat VN in TBS (pH 7.5 or 8.5) or 10 mM citrate-phosphate buffer (pH 3.5, 4.5, 5.5, or 6.5) containing 0.14 M NaCl (50 µL) were added to each well, followed by incubation for 2 h. For measurement of the binding dependency on ionic strength, VN was dissolved in 10 mM citrate-phosphate buffer (pH 4.5 or 7.5), each containing 0–2.24 M NaCl and used for the binding assay at 4 °C. To assess the effect of aggregation of collagen, alkali-treated monomeric collagen was immobilized on a microtiter plates and measured for VN binding at 37 °C.

After washing the plates with the same buffer three times, the amounts of VN bound to immobilized collagen were detected with unconjugated rabbit anti-human VN IgG and HRP-conjugated sheep anti-rabbit IgG at room temperature (15 °C). The reaction was developed using O-phenylenediamine (2.3 mg/ml) in 0.1 M phosphate-citrate buffer (pH 5.0) containing 0.007% H2O2, stopped by addition of 4 N H2SO4, and measured at 490 nm using a microplate reader, model 680 (Bio-Rad, Hercules, CA).

Quantification of interactions between VN and immobilized collagen by surface plasmon resonance
The interaction between VNs and type I collagen was analyzed by surface plasmon resonance (SPR) using a BIAcore2000 SPR apparatus (BIAcore AB, Uppsala, Sweden). After equilibration with N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid) (HEPES)-buffered saline, type I collagen (15 µg/mL) in sodium acetate buffer (pH 4.3) was injected onto a CM5 sensor chip (BIAcore) preactivated with the amine coupling kit, and then the remaining N-hydroxysuccinimide esters were blocked by the addition of 1.0 M ethanolamine hydrochloride (pH 8.0). Each step was performed for 14 min at a flow rate of 10 µL/min. The reference flow cell was prepared with bovine serum albumin (BSA) as a ligand.

To measure the binding curves, various concentrations of rat NO-, SH-, or PH-VN in PBS (pH 7.5) were separately injected onto a collagen-immobilized chip with a contact time of 720 s, a dissociation time of 150 s, and a flowrate of 20 µL/min at 25 °C. The sensor chip was regenerated with 0.1 M HCl for 60 s. The bound amounts were expressed in BIAcore resonance units (RU, 1000; RU = 1 ng/mm2), and the changes in resonance units induced by binding of VN to the collagen-immobilized flow cell were corrected by bulk effect by subtracting the changes on the BSA-immobilized reference cell.

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis
SDS–PAGE was carried out using a 7.5 or 5–20% gradient polyacrylamide gel (ATTO, e-PAGEL, Tokyo, Japan) according to the method of Laemmli (1970)Go under reducing or nonreducing conditions.

Gel-filtration HPLC and immunodetection of VN by ELISA
Sera of NO, SH, and PH rats or the pass-through and bound fractions from heparin-Sepharose were analyzed by gel-filtration HPLC on a Shodex Protein KW-804 column (8 mm x 300 mm, Showa Denko, Tokyo, Japan). Rat sera were diluted four-fold with the elution buffer (PBS–EDTA), and a 50 µL aliquot of each sample was applied to the HPLC and eluted at a flow rate of 1 mL/min at room temperature. The eluted fraction was monitored by absorbance at 280 nm and collected every 15 or 30 s by a fraction collector. The amounts of VN in each fraction were detected by ELISA at 4 °C. The wells of microtiter plates (Immulon 1, Dynatech Laboratories Inc.) were coated with 100 µL aliquots of the eluted fractions from HPLC at 4 °C for 3 h, and the procedures were performed as described in Assays of binding of VN to immobilized type I collagen.


   Conflict of interest statement
Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Conflict of interest statement
Acknowledgments
 References
 
None declared.


   Acknowledgments
Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
References
 
This work was supported in part by Hayashi Memorial Foundation for Female Natural Scientists (KS) and Grants-in-Aid for Scientific Research (C) 14580622 and 17570109 (HO) from the Japan Society for the Promotion of Science and Grants-in-aid for Scientific Research on Priority Areas 15040209 and 17046004 (HO) from the Ministry of Education, Culture, Sports, Science, and Technology. We thank our laboratory members for helping the operations and K. Ono for editing the English.


   Abbreviations
 
2ME, 2-mercaptoethanol; BSA, bovine serum albumin; CBB, Coomassie brilliant blue; ECM, extracellular matrix; EDTA, ethylenediamine tetraacetic acid; ELISA, enzyme-linked immunosorbent assay; HEPES, N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid); HPLC, high-performance liquid chromatography; HRP, horseradish peroxidase; NO, nonoperated; PAI-1, plasminogen activator inhibitor-1; PBS, phosphate-buffered saline; PH, partially hepatectomized; PNGase F, peptide N-glycosidase F; SDS-PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; SH, sham-operated; SPR, surface plasmon resonance; TBS, tris-buffered saline.


   References
Top
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 Conflict of interest statement
 Acknowledgments
 References
 
DeClerck YA, Mercurio AM, Stack MS, Chapman HA, Zutter MM, Muschel RJ, Raz A, Matrisian LM, Sloane BF, Noel A, et al. Proteases, extracellular matrix, and cancer: a workshop of the path B study section. Am J Pathol (2004) 164:1131–1139.[Abstract/Free Full Text]

Gebb C, Hayman EG, Engvall E, Ruoslahti E. Interaction of vitronectin with collagen. J Biol Chem (1986) 261:16698–16703.[Abstract/Free Full Text]

Gibson AD, Lamerdin JA, Zhuang P, Baburaj K, Serpersu EH, Peterson CB. Orientation of heparin-binding sites in native vitronectin. Analyses of ligand binding to the primary glycosaminoglycan-binding site indicate that putative secondary sites are not functional. J Biol Chem (1999) 274:6432–6442.[Abstract/Free Full Text]

Hattori S, Adachi E, Ebihara T, Shirai T, Someki I, Irie S. Alkali-treated collagen retained the triple helical conformation and the ligand activity for the cell adhesion via alpha2beta1 integrin. J Biochem. (Tokyo) (1999) 125:676–684.[Abstract/Free Full Text]

Higgins GM, Anderson RM. Experimental pathology of the liver; Restoration of the liver of the white rat following partial surgical removal. Arch Pathol (1931) 12:186–202.[Web of Science]

Inuzuka S, Ueno T, Torimura T, Tamaki S, Sakata R, Sata M, Yoshida H, Tanikawa K. Vitronectin in liver disorders: biochemical and immunohistochemical studies. Hepatology. (1992) 15:629–636.

Isaji T, Gu J, Nishiuchi R, Zhao Y, Takahashi M, Miyoshi E, Honke K, Sekiguchi K, Taniguchi N. Introduction of bisecting GlcNAc into integrin alpha5beta1 reduces ligand binding and down-regulates cell adhesion and cell migration. J Biol Chem (2004) 279:19747–19754.[Abstract/Free Full Text]

Izumi M, Shimo-Oka T, Morishita N, Ii I, Hayashi M. Identification of the collagen-binding domain of vitronectin using monoclonal antibodies. Cell Struct Funct (1988) 13:217–225.[Web of Science][Medline]

Izumi M, Yamada KM, Hayashi M. Vitronectin exists in two structurally and functionally distinct forms in human plasma. Biochim Biophys Acta (1989) 990:101–108.[Medline]

Jang YC, Tsou R, Gibran NS, Isik FF. Vitronectin deficiency is associated with increased wound fibrinolysis and decreased microvascular angiogenesis in mice. Surgery (2000) 127:696–704.[CrossRef][Web of Science][Medline]

Kitagaki-Ogawa H, Yatohgo T, Izumi M, Hayashi M, Kashiwagi H, Matsumoto I, Seno N. Diversities in animal vitronectins. Differences in molecular weight, immunoreactivity and carbohydrate chains. Biochim Biophys Acta (1990) 1033:49–56.[Medline]

Kobayashi J, Yamada S, Kawasaki H. Distribution of vitronectin in plasma and liver tissue: relationship to chronic liver disease. Hepatology (1994) 20:1412–1417.[CrossRef][Web of Science][Medline]

Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (1970) 227:680–685.[CrossRef][Medline]

Li R, Mitra N, Gratkowski H, Vilaire G, Litvinov R, Nagasami C, Weisel JW, Lear JD, DeGrado WF, Bennett JS. Activation of integrin alphaIIbbeta3 by modulation of transmembrane helix associations. Science (2003) 300:795–798.[Abstract/Free Full Text]

McKeown-Longo PJ, Panetti TS. Structure and Function of Vitronectin. Trends Glycosci Glycotechnol (1996) 8:327–340.

Nadanaka S, Sato C, Kitajima K, Katagiri K, Irie S, Yamagata T. Occurrence of oligosialic acids on integrin alpha 5 subunit and their involvement in cell adhesion to fibronectin. J Biol Chem (2001) 276:33657–33664.[Abstract/Free Full Text]

Nakashima N, Miyazaki K, Ishikawa M, Yatohgo T, Ogawa H, Uchibori H, Matsumoto I, Seno N, Hayashi M. Vitronectin diversity in evolution but uniformity in ligand binding and size of the core polypeptide. Biochim. Biophys Acta (1992) 1120:1–10.[CrossRef][Medline]

Podack ER, Kolb WP, Muller-Eberhard HJ. The SC5b-7 complex: formation, isolation, properties, and subunit composition. J Immunol (1977) 119:2024–2029.[Abstract/Free Full Text]

Preissner KT. Structure and biological role of vitronectin. Annu. Rev. Cell Biol (1991) 7:275–310.[CrossRef][Web of Science][Medline]

Preissner KT, Seiffert D. Role of vitronectin and its receptors in haemostasis and vascular remodeling. Thromb Res (1998) 89:1–21.[CrossRef][Web of Science][Medline]

Pretzlaff RK, Xue VW, Rowin ME. Sialidase treatment exposes the beta1-integrin active ligand binding site on HL60 cells and increases binding to fibronectin. Cell Adhes Commun (2000) 7:491–500.[Web of Science][Medline]

Schuck P. Sedimentation analysis of noninteracting and self-associating solutes using numerical solutions to the Lamm equation. Biophys J (1998) 75:1503–1512.[Web of Science][Medline]

Schvartz I, Seger D, Shaltiel S. Vitronectin. Int J Biochem Cell Biol (1999) 31:539–544.[CrossRef][Web of Science][Medline]

Seiffert D. Constitutive and regulated expression of vitronectin. Histol Histopathol (1997) 12:787–797.[Web of Science][Medline]

Uchibori-Iwaki H, Yoneda A, Oda-Tamai S, Kato S, Akamatsu N, Otsuka M, Murase K, Kojima K, Suzuki R, Maeya Y, et al. The changes in glycosylation after partial hepatectomy enhance collagen binding of vitronectin in plasma. Glycobiology (2000) 10:865–874.[Abstract/Free Full Text]

Ueda H, Kojima K, Saitoh T, Ogawa H. Interaction of a lectin from Psathyrella velutina mushroom with N-acetylneuraminic acid. FEBS Lett (1999) 448:75–80.[CrossRef][Web of Science][Medline]

Yatohgo T, Izumi M, Kashiwagi H, Hayashi M. Novel purification of vitronectin from human plasma by heparin affinity chromatography. Cell Struct Funct (1988) 13:281–292.[CrossRef][Web of Science][Medline]

Yoneda A, Ogawa H, Kojima K, Matsumoto I. Characterization of the ligand binding activities of vitronectin: interaction of vitronectin with lipids and identification of the binding domains for various ligands using recombinant domains. Biochemistry (1998) 37:6351–6360.[CrossRef][Medline]

Zheng M, Fang H, Hakomori S. Functional role of N-glycosylation in alpha 5 beta 1 integrin receptor. De-N-glycosylation induces dissociation or altered association of alpha 5 and beta 1 subunits and concomitant loss of fibronectin binding activity. J Biol Chem (1994) 269:12325–12331.[Abstract/Free Full Text]

Zheng X, Saunders TL, Camper SA, Samuelson LC, Ginsburg D. Vitronectin is not essential for normal mammalian development and fertility. Proc Natl Acad Sci USA (1995) 92:12426–12430.[Abstract/Free Full Text]

Zhuang P, Blackburn MN, Peterson CB. Characterization of the denaturation and renaturation of human plasma vitronectin. I. Biophysical characterization of protein unfolding and multimerization. J Biol Chem (1996) 271:14323–14332.[Abstract/Free Full Text]

Zhuang P, Li H, Williams JG, Wagner NV, Seiffert D, Peterson CB. Characterization of the denaturation and renaturation of human plasma vitronectin. II. Investigation into the mechanism of formation of multimers. J Biol Chem (1996) 271:14333–14343.[Abstract/Free Full Text]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
17/7/784    most recent
cwm031v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Sano, K.
Right arrow Articles by Ogawa, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sano, K.
Right arrow Articles by Ogawa, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?