- Email: [email protected]
Collagen fibril aggregation-inhibitor from sea cucumber dermis
Matrix Biology 18 Ž1999. 569]578
Collagen fibril aggregation-inhibitor from sea cucumber dermis John A. Trotter a,U , Gillian Lyons-Levy a , Kazumi Chino a , Thomas J. Koob b, Douglas R. Keene c , Mark A.L. Atkinsond a
Department of Cell Biology and Physiology, Uni¨ ersity of New Mexico School of Medicine, Albuquerque, NM 87131, USA b Shriners Hospital for Children, Tampa, FL, USA c Shriners Hospital for Children, Portland, OR, USA d Department of Biochemistry, Uni¨ ersity of Texas Health Science Center, Tyler, TX, USA Received 25 June 1999; received in revised form 17 September 1999; accepted 21 September 1999
Abstract Collagen fibrils from the dermis of the sea cucumber Cucumaria frondosa are aggregated in vitro by the dermal glycoprotein stiparin ŽTrotter et al., 1996.. Under physiological ionic conditions stiparin appears to be both necessary and sufficient to cause fibrils to aggregate ŽTrotter et al., 1997.. We report here the initial biochemical and biophysical characterization of a sulfated glycoprotein from C. frondosa dermis that binds stiparin and inhibits its fibril-aggregating activity. This inhibitory glycoprotein, which has been named ‘stiparin-inhibitor,’ has the highest negative charge density of all the macromolecules extracted from the dermis. SDS-PAGE reveals three ; 31-kDa bands that stain with alcian blue but not with Coomassie blue. Analytical ultracentrifugation indicates a native molecular weight of 62 kDa. Transmission electron microscopy of rotary-shadowed molecules shows curved rods about 22 nm long. The glycoprotein does not bind collagen fibrils, but does bind stiparin with a 1:1 stoichiometry. The binding of stiparin-inhibitor to stiparin prevents the binding of stiparin to collagen fibrils. The carbohydrate moiety produced by papain-digestion of the glycoprotein retains all of its inhibitory activity. The carbohydrate moiety of the inhibitor is dominated by galactose and sulfate. Q 1999 Published by Elsevier Science B.V.rInternational Society of Matrix Biology. All rights reserved. Keywords: Collagen; Fibril; Aggregation; Inhibitor; Stiparin; Echinoderm
1. Introduction Animals in the phylum Echinodermata share the ability to regulate the tensile properties of their col-
Abbre¨ iations: CS: chondroitin sulfate; DMMB: dimethyl-methylene blue; EDTA: ethylenediamine-tetra-acetic acid; GuHCl: guanidine hydrochloride; MOPS: 3-w N-morpholinoxpropane-sulfonic acid; PTC: phenylthiocarbamyl; Tris: tris-hydroxymethyl-aminomethane; TCA: trichloroacetic acid; TFA: trifluoroacetic acid. U Corresponding author. Tel.: q1-505-272-5700; fax: q1-505272-6581. E-mail address: [email protected] ŽJ.A. Trotter.
lagenous tissues on a physiological time scale ŽMotokawa, 1984; Wilkie, 1984, 1996.. This phenomenon has not been reported in animals in other phyla. The biomechanical properties of echinoderm ‘mutable connective tissues’ ŽMCTs. are regulated by physiological control of the interactions between collagen fibrils ŽHidaka, 1983; Hidaka and Takahashi, 1983; Trotter and Koob, 1995; Trotter et al., 1996; Trotter and Chino, 1997; Koob et al., 1999.. The collagen fibrils of echinoderms are symmetrically spindle-shaped and relatively short Žup to approx. 1 mm in length. ŽMatsumura, 1973, 1974; Trotter and Koob, 1989; Thurmond and Trotter, 1994;
0945-053Xr99r$ - see front matter Q 1999 Published by Elsevier Science B.V.rInternational Society of Matrix Biology. All rights reserved. PII: S 0 9 4 5 - 0 5 3 X Ž 9 9 . 0 0 0 5 0 - 5
570
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
Trotter et al., 1994, 1998.. They have surface-bound proteoglycans which are associated with the same location in each D-period. The fibrils are stabilized internally by the same types of covalent crosslinks found in vertebrate fibrils ŽBailey et al., 1982; Eyre, et al., 1984; Trotter and Koob, 1989; Trotter et al., 1994.. Importantly, however, they lack permanent or covalent inter-fibrillar crosslinks ŽTrotter and Koob, 1989; Trotter et al., 1996.. The absence of permanent crosslinks between echinoderm collagen fibrils has made it possible to develop methods to isolate abundant quantities of intact fibrils in the native state ŽMatsumura, 1973, 1974; Trotter et al., 1996.. Native fibrils isolated from the dermis of the North Atlantic sea cucumber C. frondosa remain suspended in physiological salt solutions. However, addition of stiparin, which is the most abundant soluble glycoprotein in the dermis ŽTrotter et al., 1996., causes fibrils to aggregate. Stiparin-mediated fibril interactions are thought to be important both in the organization of the tissue and in the regulation of its viscoelastic properties ŽTrotter et al., 1996, 1997.. Because of the physiological reversibility of the stiffening reaction of the dermis, it was predicted that one or more soluble inhibitors of stiparin should also be present in the dermis. An in vitro assay for stiparin-inhibitory activity was used to identify macromolecules in dermal extracts that inhibit stiparin-dependent fibril-aggregation. This communication describes the purification and initial characterization of the first of such inhibitory molecules to have been identified.
centration of collagen in the fibril suspensions was estimated based on the colorimetric determination of hydroxyproline in hydrolysates ŽBergman and Loxly, 1963. and on our previous determination that hydroxyproline accounts for 7.7% of the mass of the collagen molecule ŽTrotter et al., 1995.. 2.3. Stiparin purification and chromatographic methods Stiparin was purified from the inner dermis as described in Trotter et al. Ž1996.. Ion-exchange and size-exclusion chromatography were performed as previously described ŽTrotter et al., 1996. using HiTrap Q, Mono-Q, Superdex-75 and 200, and Superose-6 columns ŽAmersham Pharmacia Biotech... 2.4. Amino acid analyses Amino acid analyses were performed on an Applied Biosystems Model 420A amino acid analyzer equipped with on-line PTC pre-column derivitization. Samples were subjected to vapor phase hydrolysis for 24 h at 1108C in 6 N HCl. Two separate hydrolysates were analyzed in triplicate and results are expressed as the averages of these determinations. For tyrosine sulfate determinations the protein sample and tyrosine sulfate standard were base hydrolyzed Ž4.2 M KOH. for 20 h at 1108C in vacuum. After hydrolysis base was neutralized with 4.2 M perchloric acid. The potassium perchlorate was removed by centrifugation. Twenty microliters were analyzed by the PicoTag method as described in Cohen and Strydom Ž1988..
2. Experimental
2.5. Analytical ultracentrifugation
2.1. Animals and tissue preparation
Equilibrium sedimentation measurements were made in a Beckman XL-A analytical ultracentrifuge. 100 ml samples of purified stiparin-inhibitor at a concentration of approximately 200 mgrml in 20 mM Tris]HCl ŽpH 8., 400 mM NaCl were centrifuged at 3000, 4000 and 4500 rev.rmin at 2778K in the An60T1 rotor. The cells were scanned at 280 nm. Each data point is the average of five readings taken at 0.001-cm intervals through the sample column. The sample was assumed to have reached equilibrium when multiple scans over a 4-h period did not change. The partial specific volume was calculated from the amino acid composition of the protein ŽZamyatnin, 1972., using standard values for the individual amino acids and known carbohydrate components ŽDurschlag, 1986.. Sedimentation velocity measurements were also made in the Beckman XL-A analytical ultracentrifuge. Samples were run in a two-sector cell using 400 ml of sample and 410 ml of reference buffer. Buffer conditions were the same as those used for
The collection of adult C. frondosa and preparation of dermal specimens were as described previously ŽTrotter et al., 1996.. Both the white inner dermis and the pigmented outer dermis were used, as detailed below. The tissues were stored at y708C until needed. 2.2. Preparation of collagen fibrils Collagen fibrils were isolated from pieces of inner dermis as described in Trotter et al. Ž1996.. Briefly, fibrils were extracted from minced tissue pieces that had been extensively pre-extracted in 3 M GuHCl, pH 6.0, at 48C. The GuHCl-extracted tissues were then sequentially extracted in deionized water; 4 mM EDTA, 0.1 M Tris]HCl ŽpH 7.8.; and deionized water. The second and subsequent water-extracts contained abundant quantities of ‘stripped’ fibrils, i.e. collagen fibrils that were free of non-covalently bound macromolecules ŽTrotter et al., 1995, 1996.. The con-
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
sedimentation equilibrium. The experiment was performed at 208C Ž2938K. at 35 000 rev.rmin in the An60Ti rotor. The cell was scanned at 280 nm and data points represent the average of five measurements made at 0.005-cm intervals. A total of 15 scans were made at 10-min intervals.
571
Chondroitin sulfate ŽCS. from bovine trachea ŽSigma product C8529; approx. 70% C-4-S and 30% C-6-S. was used as the standard. Concentrations were expressed as mg or mg equivalents ŽmEq. or mEq.. of CS. 2.8. Papain digestion of stiparin-inhibitor
2.6. Electron microscopy Electron microscopy and rotary shadowing of the isolated protein were carried out as described by Sakai and Keene Ž1994.. Briefly, solutions of the protein containing 0.2 M NH 4 HCO 3 were diluted in pure glycerol to a final concentration of 70% glycerol. Samples of 0.1 ml were sprayed onto freshly cleaved mica and dried in a Balzers BAE 250 evaporator at 10y6 torr. Shadowing and electron microscopy were then performed as described ŽSakai and Keene, 1994.. 2.7. Carbohydrate analyses Partial carbohydrate analysis was performed as follows. Four replicate 150-ml aliquots of the purified stiparin-inhibitor in 0.4 M NaCl, 20 mM Tris were added to ice-cold 95% ethanol, mixed and chilled in the y208C freezer for at least 1 h. The samples were then centrifuged at 16 000 = g for 15 min at 48C. The ethanol was aspirated and an additional 1.5 ml of 95% ice-cold ethanol was added to the precipitate, mixed, and held at y208C for several hours. The samples were again centrifuged as above. The precipitates were dried under vacuum. Two of the four replicates were hydrolyzed in 400 ml 2 M TFA Žnitrogen purged. at 1008C for 3 h followed by rapid drying under vacuum. To determine losses due to hydrolysis 200 ml of carbohydrate standards were hydrolyzed in parallel. The other two replicates were hydrolyzed in 400 ml 4 M HCl Žnitrogen purged. at 1008C for 5 h followed by rapid drying under vacuum. Standards Ž200 ml. were treated in parallel. The hydrolyzates were dissolved in 50 ml of water just prior to analysis on the Carbo-Pac column coupled to the pulsed amperometric detector on a Dionex system. At the start of each day, a freshly thawed standard was analyzed and used to normalize all subsequent analyses performed that day. Uronic acid was estimated by the method of Blumenkrantz and Asboe-Hansen Ž1973.; sialic acid was estimated by the method of Jourdian et al. Ž1971.; sulfate was estimated by the method of Dodgson Ž1961.; hexosamine was estimated by the method of Antonopoulos et al. Ž1964.; and protein was estimated using the micro-BCA protein assay kit purchased from Pierce Chemical ŽRockford, IL, USA.. Glycosaminoglycan concentrations were estimated using the dimethylmethylene blue assay ŽFarndale et al., 1986..
The protein moiety of the inhibitor was digested by papain ŽWorthington Biochemical. as follows. A sample containing 2 mEq. of stiparin-inhibitor was incubated with 0.03 units of papain in a total volume of 5 ml of 0.1 M Na-acetate ŽpH 5.5., containing 10 mM L-cysteine and 50 mM ethylenediamine tetra-acetic acid ŽEDTA. at 658C for 10 h. Digestion was terminated by the addition of ice-cold trichloroacetic acid ŽTCA. to a final concentration of 5%. Precipitated proteins were collected by low speed centrifugation, and the supernatant was fractionated on a Superdex 75 column in 0.4 M NaCl, 20 mM Tris]HCl ŽpH 7.8.. The column effluent was monitored for inhibitory activity. 2.9. Fibril aggregation assays Aggregation assays were carried out as previously described ŽTrotter et al., 1996. in 1 ml of solution, using buffers and salts as indicated. Stiparin-inhibitory activity was similarly assayed, with the following modifications. The minimal quantity of stiparin required to aggregate 60 mg of collagen fibrils was pre-determined and used in subsequent assays. On average, in 0.4 M NaCl ŽpH 7.8., approximately 0.25 mg of stiparin caused the aggregation of 60 mg of collagen fibrils. This is equivalent to a ratio of 3.3 mmol of stiparin per mole of collagen ŽTrotter et al., 1996.. Column fractions were analyzed by adding aliquots to stiparin and buffer in a volume of 1 ml. After gentle mixing, collagen fibrils were added and the tubes were gently mixed as described ŽTrotter et al., 1996.. 2.10. Affinity chromatography Affinity chromatography was performed using stripped collagen fibrils bound to an inert support matrix. To prepare a 1-ml fibril-affinity column, 1.6 ml of a slurry of Affi-Prep 10 support medium ŽBioRad. was washed under low vacuum with cold 10 mM MOPS ŽpH 7.0.. The washed support medium was rapidly transferred to a flask containing 20 mg of collagen fibrils in 20 ml of 10 mM MOPS ŽpH 7.0.. This mixture was evacuated for approximately 1 min, after which it was incubated overnight at 48C with gentle rocking. The binding of fibrils to the Affi-Prep matrix was stopped by the addition of 100 ml of 1 M
572
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
ethanolamine. The mixture was then centrifuged at 120 = g for 10 min and the supernatant was discarded. The pellet was washed three times with 30 ml of 0.5 M NaCl followed by two washes with phosphate-buffered saline ŽPBS.. The washed support was poured into a glass column in the same buffer and allowed to settle for 4 h. The column was then washed with 6 ml each of 0.4 M, 2 M and 0.4 M NaCl, all in 20 mM Tris]HCl ŽpH 7.8.. It was found in initial experiments that stiparin bound to this column in 0.4 M NaCl and was eluted from the column in 2 M NaCl. To assay the binding of stiparin in the presence of stiparin inhibitor, 200 ml Ž160 mg. of stiparin in 0.4 M NaCl, 20 mM Tris]HCl ŽpH 7.8. with and without stiparin inhibitor Ž600 mEq. CS; 10 mg protein. were applied to the column, followed by 3 ml of the same buffer and then 3 ml of 2 M NaCl, 20 mM Tris]HCl ŽpH 7.8.. 0.5 ml fractions of the column eluate were collected and assayed for protein using UV absorbance Ž280 nm., and for sulfated carbohydrate using the DMMB assay. 2.11. Polyacrylamide gel electrophoresis Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate ŽSDS-PAGE. was accomplished using 4]20% gradient gels, stained with alcian blue and Coomassie brilliant blue as described ŽTrotter et al., 1995.. The relative proportions of the three bands comprising the complete inhibitor were made by analyzing the laser-densitometer scans of alcian blue-stained gels with Peak Fit 4 ŽJandel Scientific, San Rafael, CA, USA..
Fig. 1. Fibril aggregation assay. Collagen fibrils are suspended in buffered salt solutions at a concentration of 60 mgrml of collagen. Oblique illumination shows diffuse light scattering of the suspended fibrils Žleft.. Addition of 0.25 mg of stiparin causes the fibrils to aggregate into a single clot Žright.. Addition of outer dermis extract both inhibits and reverses the action of stiparin.
column, which eluted in the position of the 440-kDa globular protein standard ŽFig. 2b.. SDS-PAGE of the active fractions from both anion exchange and gel filtration columns revealed a triplet of bands with apparent molecular masses of approximately 31 kDa ŽFig. 2a,b.. These bands were stained
3. Results 3.1. Identification and purification of stiparin-inhibitory acti¨ ity It was first noted that whereas crude aqueous extracts of inner dermis caused isolated collagen fibrils to aggregate, similar extracts of outer dermis did not, although both extracts contained stiparin, as shown by SDS-PAGE. This observation suggested that the outer dermis extract might contain one or more inhibitors of stiparin-dependent fibril aggregation. Indeed, the addition of outer dermis extracts to mixtures containing 60 mg of collagen fibrils and 0.25 mg of stiparin prevented stiparin-dependent aggregation ŽFig. 1.. All of the stiparin-inhibitory activity was associated with molecules that eluted from a strong anion exchanger at 2.2 M NaCl ŽFig. 2a.. No other macromolecules were eluted from the column at higher concentrations of NaCl. Similarly, stiparin-inhibitory activity was associated with a single peak from a gel filtration
Fig. 2. Chromatography of outer dermis extracts using Ža. a strong anion exchanger ŽHiTrap Q; Pharmacia. and Žb. a gel filtration column ŽSuperdex 200; Pharmacia.. Absorbance at 280 nm indicates protein concentrations; absorbance at 525 nm indicates DMMB-positive material in diluted column fractions. The extinction coefficient for chondroitin sulfate in this assay was 0.041 absorbance units per mg Žsee Section 2.7.. The arrowheads on the abscissas indicate the fractions that contained stiparin-inhibitory activity. The SDS-PAGE gels show the alcian blue staining patterns of the active fractions.
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
by alcian blue but not by Coomassie blue, a finding that is consistent with the high negative charge density of the molecule. The positions and relative amounts of the three bands were unchanged by the addition of dithiothreitol, indicating that they are not components of a disulfide-bonded complex. It was apparent from the SDS-PAGE patterns of individual fractions from gel filtration columns that the ratios of
573
the three bands were not constant, suggesting that the native molecule was not a heterotrimer containing constant molar ratios of the three bands. The relative amounts of each of the three bands was determined by quantitative densitometry of the gels of fractions taken at the beginning and end of the HiTrap Q peak ŽFig. 3a, gel lanes 1, 2. and at the beginning, middle, and end of the Superdex 200 peak ŽFig. 3b, gel lanes 1]3.. It was apparent that the slower migrating band Ži.e. the band with the highest apparent molecular mass. was most abundant at the beginning of each peak, whereas the faster migrating band was most abundant at the end, while the abundance of the middle band remained relatively constant. The composition of the inhibitory molecule was further analyzed by gel filtration chromatography in the presence of 6 M GuHCl, followed by quantitative analyses of SDS-PAGE gels of the fractions through the peak. The results indicated that the inhibitor contains three independent macromolecules that differ in molecular mass ŽFig. 3c.. They also showed that the relative positions of the three bands observed by SDS-PAGE corresponded to their relative molecular masses, i.e. the electrophoretic mobilities of the bands were inversely proportional to their molecular masses. 3.2. Physical properties of stiparin-inhibitor Sedimentation equilibrium studies in the analytical ultracentrifuge indicated that the molecular mass of the native molecule was approximately 62 kDa. Under native conditions, then, the inhibitor is probably a dimer. A dimer containing one molecule of the middle band and one molecule of either the upper or lower band would be consistent with the quantitative analyses of column fractions. Sedimentation velocity studies gave a sedimentation coefficient of 5.9 S, a diffusion coefficient of 1.003= 10y6 cm2 sy1 and a frictional ratio of 1.05, which corresponds to a prolate ellipsoid with an axial ratio of approximately 2.0. Transmission electron microscopy of the purified inhibitor following low angle-rotary shadowing ŽFig. 4B. revealed a 22-nm-long rod with some apparent flexibility Žbased on independent measurements by two individuals.. 3.3. Chemical analyses of stiparin-inhibitor
Fig. 3. Quantitative densitometry of SDS-PAGE gels of the active fractions from a HiTrap Q column Ža., and a Superdex 200 column Žb., both run under non-denaturing conditions. Also shown are the fractions of stiparin-inhibitor obtained from a Superdex 200 column run in the presence of 6 M GuHCl Žc.. The bars in each triplet indicate the relative amounts of the three bands observed after alcian blue staining. In all cases the slowest migrating band is on the left and the fastest migrating band is on the right.
Amino acid analyses revealed a composition dominated by serine, glutamic acidrglutamine, and glycine ŽTable 1.. Comparison of amino acid data in the Swissprot database using the Propsearch program ŽHobohm et al., 1994. revealed no close fits to any known proteins. The high negative charge density of the inhibitor and its staining with alcian blue sug-
574
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
Fig. 4. Transmission electron microscopy after rotary shadowing of purified stiparin ŽA.; stiparin-inhibitor ŽB.; a mixture of 1 mg stiparin per 0.7 mEq. of stiparin-inhibitor Žestimated molar ratio: 0.2 mol stiparin per mol of stiparin-inhibitor . ŽC.; and a similar mixture containing 1 mg stiparin per 0.07 mEq. of stiparin inhibitor Žestimated molar ratio: 2 mol of stiparin per mol of stiparin-inhibitor . ŽD.. The molar ratios were estimated using 62 kDa as the mass of stiparin-inhibitor and 375 kDa as the mass of stiparin. In a, individual stiparin molecules are seen as long flexible rods. In b, individual stiparin-inhibitor molecules are seen as short, somewhat bent rods. In c, individual molecules of both stiparin and stiparin-inhibitor are seen. In d, no individual stiparin-inhibitor molecules are seen, and the stiparin molecules appear in small aggregates. Bar s 100 nm.
gested that it might contain conjugated sulfate residues. Indeed, sulfate analyses revealed 0.26 mg of sulfate per mEq. of DMMB-positive material Žsee Section 2.. No detectable tyrosine sulfate was found. In contrast, sulfate determinations of the carbohydrate residue produced by papain-digestion of the inhibitor showed essentially the same ratio of sulfate to DMMB-positive material Ž0.24 mgrmEq.., suggesting that the sulfate is mainly associated with carbohydrate ŽTable 2.. Sugar analyses of both the intact inhibitor and the carbohydrate residue produced by papain-digestion revealed mostly galactose, with small quantities of galactosamine, glucosamine, and fucose ŽTable 2.. Uronic acid and sialic acid were both undetectable. The molar ratio of sulfate to galactose was approximately 4. These results suggest that the carbohydrate moiety of the inhibitor may consist mainly of polygalactose sulfate. The carbohydrate residue produced by papain digestion of the inhibitor retained essentially all of the
inhibitory activity. Inhibition of 1 mg of stiparin required 0.2 mEq. of either the digested or undigested inhibitor. In contrast, neither chondroitin-4-sulfate nor the major glycosaminoglycan isolated from the dermis ŽTrotter et al., 1995. inhibited stiparin-dependent fibril aggregation at the highest concentration used, which was more than five times the required concentration of stiparin-inhibitor Žexpressed as equivalents of chondroitin sulfate .. 3.4. Association of stiparin-inhibitor with stiparin Centrifugation of isolated collagen fibrils in the presence of inhibitor failed to remove the inhibitor from the supernatant, indicating that it does not bind the fibrils with high affinity. To determine whether the inhibitor binds stiparin, the individual molecules and a mixture of the two containing a large excess of inhibitor were analyzed by sedimentation equilibrium analysis in the ultracentrifuge. The apparent molecu-
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578 Table 1 Amino acid composition of stiparin-inhibitor AA
Resr1000
Asx Ala Arg Cys Glx Gly His Ile Leu Lys Met Phe Pro Ser Thr Trp Tyr Val Oh-Pro Oh-Lys Total
83 67 25 ND 152 141 16 24 48 22 7 26 41 204 44 ND 40 61 0 0 1000
575
stiparin and stiparin-inhibitor, no interactions were detected when the two molecules were co-chromatographed on a Superose-6 column in the presence of either 0.4 M NaCl or artificial sea water ŽTrotter et al., 1996.. The elution volume of each molecule was unaffected by the presence of the other molecule. This finding indicated that the association between the molecules was too weak to be maintained under the conditions used for chromatography. The binding of stiparin to an affinity column containing stripped collagen fibrils was studied in the presence of different NaCl concentrations and in the presence and absence of stiparin inhibitor. Stiparin bound to the column in 0.4 M NaCl and was eluted in 2 M NaCl ŽFig. 5.. In the presence of stiparin-inhibitor, however, stiparin failed to bind to the fibrils in 0.4 M NaCl. 3.5. Relati¨ e abundance of stiparin-inhibitor in inner and outer dermis
lar mass of stiparin increased by an amount approximately equal to the apparent mass of the inhibitor Žfrom 356 000 to 411 000., suggesting that the two molecules form a 1:1 complex. The association of the inhibitor with stiparin was also shown by transmission electron microscopy of the two molecules mixed in different ratios ŽFig. 4C,D; see the figure caption for detailed description.. As the relative amount of stiparin was increased, the amount of free inhibitor was decreased. The stiparin molecules appeared to be somewhat less extended in the presence of the inhibitor ŽFig. 4C., although this could not be quantified. At high stiparinrinhibitor molar ratios Žgreater than f 2. the stiparin molecules formed small aggregates ŽFig. 4D.. Although both the ultracentrifuge and electron microscope studies indicate a saturable binding between
Stiparin-inhibitor is present in both inner and outer dermis. However, the relative concentration of stiparin-inhibitor in the outer dermis is approximately 200 times that in the inner dermis. Direct extracts of inner and outer dermis in SDS sample buffer followed by SDS-PAGE and quantitative densitometry indicated that the outer dermis contains 10.75 mEq. of stiparin-inhibitor per mg wet weight whereas the inner dermis contains only 0.05 mEq. per mg.
4. Discussion It was shown previously that extraction of the 375kDa glycoprotein stiparin from C. frondosa inner dermis led to the dissociation of the tissue into a suspension of intact collagen fibrils ŽTrotter et al., 1996.. Re-addition of crude stiparin caused a crude fibril suspension to form an aggregate. Purified sti-
Table 2 Monosaccharide composition of stiparin-inhibitor a pmolrmg of CS equivalent
Fuc GalN GlcN Gal Glc Man SO4 a
SI
PAP
Ratio PAPrSI
SI
Ratios to galactose PAP
55 17 13 606 112 60 2600
99 22 16 646 12 0 2400
1.8 1.3 1.2 1.1 0.1 0.0 0.9
0.09 0.03 0.02 1.00 0.18 0.10 4.3
0.15 0.03 0.02 1.00 0.02 0.00 3.7
SI, Stiparin-inhibitor; PAP, Carbohydrate moiety produced by complete papain-digestion of stiparin-inhibitor.
576
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
Fig. 5. Affinity chromatography of stiparin on a column containing immobilized collagen fibrils. The elution pattern of stiparin is shown by the solid line. A280 indicates protein absorbance at 280 nm. The leftmost arrow on the abscissa indicates the addition of sample in 0.4 M NaCl. The rightmost arrow indicates the change of the buffer to 2 M NaCl. Stiparin binds to the fibrils in 0.4 M NaCl and elutes in 2 M NaCl. The elution patterns of stiparin and stiparin-inhibitor after a mixture of the two molecules had been applied to the column are shown by the broken line ŽA280. and the dotted line ŽA525.. A525 indicates DMMB absorbance at 525 nm. The two arrows have the same significance. In the presence of stiparin-inhibitor, stiparin fails to bind to the fibrils.
parin had the same effect on purified fibrils, suggesting that stiparin alone was both necessary and sufficient to cause fibril aggregation in vitro. The aggregation of fibrils in vitro is clearly correlated with fibril]fibril interactions in vivo, but the exact physiological role played by stiparin is not yet certain ŽTrotter et al., 1997.. Other protein molecules have been isolated from the inner and outer dermis that cause dramatic stiffness changes in intact specimens of the inner dermis ŽTrotter and Koob, 1995; Koob et al., 1999.. Despite the uncertainty concerning its precise role in dermal regulation, it is clear that the interactions between stiparin and collagen fibrils are important determinants of the viscoelastic properties of the inner dermis. Because of the significance of stiparin]fibril interactions in the dermis of C. frondosa, it was important to determine whether the dermis contains molecules which are capable of modulating the activity of stiparin. Using as a bioassay the stiparin-dependent aggregation of suspended collagen fibrils, it was possible to identify inhibitory activity in extracts of both inner and outer dermis. Because the inhibitory activity was much more abundant in the outer dermis, that tissue was used as the source of crude extracts from which the inhibitory activity was purified. It is necessary to be cautious in interpreting the interactions that occur between large macromolecules, especially when one of them is a highly charged polyelectrolyte. Interactions of such macromolecules are dependent on the ionic strength, di-
electric constant, and specific ion composition of the bathing solution, including the pH. The interactions described in this communication were generally studied in 0.4 M NaCl buffered to pH 7.8]8.2 with either Tris or MOPS, but they were also confirmed in complete artificial sea water ŽASW.. The interactions detected in vitro may thus potentially occur in vivo as well. The binding of stiparin-inhibitor to stiparin appears to be specific. Other polyelectrolytes, including chondroitin sulfate and the highly sulfated major GAG found in the dermis ŽTrotter et al., 1995. failed to bind to stiparin or to inhibit stiparin-dependent fibril aggregation in vitro. Furthermore, the ultracentrifuge and electron microscope analyses indicate specific association between the two molecules. In addition the ultracentrifuge studies indicate a 1:1 stoichiometry. The fact that both components elute from a gel filtration column in the volume that would be predicted if there were no interactions between them shows that the association constant cannot be very high. Similarly, the association of stiparin with collagen fibrils cannot be very strong. If it were, it would not be possible to elute sufficient stiparin from pieces of dermis to cause their collagen fibrils to dissociate into a suspension ŽTrotter et al., 1996.. Thus, the stiparin]fibril and stiparin-inhibitor ]stiparin interactions are both weak. The stiparin]fibril interaction is apt to be weaker, however, since stiparin-inhibitor is able quantitatively to displace bound stiparin from a fibril-affinity column. These observations suggest that
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
even relatively low stiparin-inhibitor ]stiparin ratios in vivo could effectively modulate the effectiveness of stiparin and, as a consequence, the viscoelastic properties of the dermis. Testing these predictions will require quantitative determinations of both equilibrium and kinetic constants as well as detailed characterizations of binding partners. These studies are currently underway. The composition and structure of the native stiparin-inhibitor is only suggested by the present findings. It is clear that there are three different glycoprotein chains that differ by a few percent in their molecular masses. The molecular weight of the native molecule suggests that it contains only two of the three chains, held in association by non-covalent interactions. It is also clear that the native molecule appears to be a somewhat flexible rod approximately 22 nm long. There is no information concerning the organization of the two chains in the molecule. However, the results suggest that each of the two chains is able to bind to stiparin whereas each stiparin molecule has a single binding site for stiparin-inhibitor. The sedimentation equilibrium experiment showed that, in the presence of excess stiparin-inhibitor, each stiparin molecule bound a single stiparin-inhibitor molecule. On the other hand, the electron-microscopic data suggested that, in the presence of excess stiparin, two or more stiparin molecules formed small complexes with stiparin-inhibitor. Kinetic experiments are presently underway to determine the stoichiometry and valency of the interactions between stiparin and stiparin-inhibitor. The fact that all of the inhibitory activity is associated with the polygalactose-sulfate moiety of the molecule rather than with the protein raises the possibility that the sites on fibrils recognized by stiparin are GAG sites rather than protein sites. This would be consistent with the observation that stiparin does not bind to the triple helical region of C. frondosa collagen molecules Žunpublished observations..
Acknowledgements The authors gratefully acknowledge the help of Dr Andrzej Pastuszyn and the Protein Chemistry Core Facility of the UNM School of Medicine. This work was supported by research grants to JAT from the NSF, ONR, and DARPA; by the James Robert Montgomery Endowment to MALA; and by grant No. 8610 from The Shriners of North America to TJK. The electron microscope facility used by DRK is supported by the Shriners Hospital for Children and by the R. Blaine Bramble and Fred Meyer Charitable Trust foundations. The excellent technical assistance of Sara Tufa is gratefully acknowledged.
577
References Antonopoulos, C.A., Gerdesi, S., Szirmai, J.A., de Tyssonsk, E.R., 1964. Determination of glycosaminoglycans Žmucopolysaccharides. from tissues on the microgram scale. Biochim. Biophys. Acta. 83, 1]19. Bailey, A.J., Gathercole, L.J., Dlugosz, J., Keller, A., Voyle, C.A., 1982. Proposed resolution of the paradox of extensive crosslinking and low tensile strength of cuvierian tubule collagen from the sea-cucumber Holothuria forskali. Int. J. Biol. Macromol. 4, 329]334. Bergman, I., Loxly, R., 1963. Two improved and simplified methods for the spectrophotometric determination of hydroxyproline. Anal. Chem. 35, 1961]1965. Blumenkrantz, N., Asboe-Hansen, G., 1973. New method for quantitative determination of uronic acids. Anal. Biochem. 54, 484]489. Cohen, S.A., Strydom, D.J., 1988. Amino acid analysis utilizing phenylisothiocyanate derivatives. Anal. Biochem. 174, 1]16. Dodgson, K.S., 1961. Determination of inorganic sulfate in studies on the enzymic and non-enzymic hydrolysis of carbohydrates and other sulfate esters. Biochem. J. 78, 312]319. Durschlag, H., 1986. Specific volumes of biological macromolecules and some other molecules of biological interest. In: Hinz, H.J. ŽEd.., Thermodynamic Data for Biochemistry and Biotechnology. Springer-Verlag, Berlin, pp. 45]128. Eyre, D.R., Koob, T.J., Van Ness, K.P., 1984. Quantitation of hydroxypyridinium crosslinks in collagen by high-performance liquid chromatography. Anal. Biochem. 137, 380]388. Farndale, R.W., Buttle, D.J., Barrett, A.J., 1986. Improved quantitation and discrimination of sulfated glycosaminoglycans by use of dimethylmethylene blue. Biochim. Biophys. Acta. 883, 173]177. Hidaka, M., 1983. Effects of certain physico-chemical agents on the mechanical properties of the catch apparatus of the sea-urchin spine. J. Exp. Biol. 103, 15]29. Hidaka, M., Takahashi, K., 1983. Fine structure and mechanical properties of the catch apparatus of the sea-urchin spine, a collagenous connective with muscle-like holding capacity. J. Exp. Biol. 103, 1]14. Hobohm, U., Houthaeve, T., Sander, C., 1994. Amino acid analysis and protein database compositional search as a rapid and inexpensive method to identify proteins. Anal. Biochem. 222, 202]209. Jourdian, G.W., Dean, L., Roseman, S., 1971. The sialic acids: a periodate-resorcinol method for the quantitative estimation of the sialic acids and their glycosides. J. Biol. Chem. 246, 430]435. Koob, T.J., Koob-Emunds, M.M., Trotter, J.A., 1999. Cell-derived stiffening and plasticizing factors in sea cucumber Ž Cucumaria frondosa.. J. Exp. Biol. 202 Žin press.. Matsumura, T., 1973. Shape, size and amino acid composition of collagen fibril of the starfish Asterias amurensis. Comp. Biochem. Physiol. 44B, 1197]1205. Matsumura, T., 1974. Collagen fibrils of the sea-cucumber, Stichopus japonicus: purification and morphological study. Connect. Tissue Res. 2, 117]125. Motokawa, T., 1984. Connective tissue catch in echinoderms. Biol. Rev. 59, 255]270. Sakai, L.Y., Keene, D.R., 1994. Fibrillin: monomers and microfibrils. Methods Enzymol. 245, 29]54. Thurmond, F.A., Trotter, J.A., 1994. Native collagen fibrils from echinoderms are molecularly bipolar. J. Mol. Biol. 235, 73]79. Trotter, J.A., Chino, K., 1997. Regulation of cell-dependent viscosity in the dermis of the sea cucumber Actinopyga agassizi. Comp. Biochem. Physiol. 119A, 805]811. Trotter, J.A., Koob, T.J., 1989. Collagen and proteoglycan in a sea
578
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
urchin ligament with mutable mechanical properties. Cell Tissue Res. 258, 527]539. Trotter, J.A., Koob, T.J., 1995. Evidence that calcium-dependent cellular processes are involved in the stiffening response of holothurian dermis and that dermal cells contain an organic stiffening factor. J. Exp. Biol. 198, 1951]1961. Trotter, J.A., Thurmond, F.A., Koob, T.J., 1994. Molecular structure and functional morphology of echinoderm collagen fibrils. Cell Tissue Res. 275, 451]458. Trotter, J.A., Lyons-Levy, G., Thurmond, F.A., Koob, T.J., 1995. Covalent composition of collagen fibrils from the sea cucumber, Cucumaria frondosa, a tissue with mutable mechanical properties. Comp. Biochem. Physiol. 112A Ž3r4., 463]478. Trotter, J.A., Lyons-Levy, G., Luna, D., Koob, T.J., Keene, D.R., Atkinson, M.A.L., 1996. Stiparin: a glycoprotein from sea cu-
cumber dermis that aggregates collagen fibrils. Matrix Biol. 15, 99]110. Trotter, J.A., Salgado, J.P., Koob, T.J., 1997. Mineral content and salt-dependent viscosity in the dermis of the sea cucumber Cucumaria frondosa. Comp. Biochem. Physiol. 116A, 329]335. Trotter, J.A., Chapman, J.A., Kadler, K.E., Holmes, D.F., 1998. Growth of sea cucumber collagen fibrils occurs at the tips and centers in a coordinated manner. J. Mol. Biol. 284, 1417]1424. Wilkie, I.C., 1984. Variable tensility in echinoderm collagenous tissues: a review. Mar. Behav. Physiol. 11, 1]34. Wilkie, I.C., 1996. Mutable collagenous tissues: extracellular matrix as mechano-effector. In: Jangoux, M., Lawrence, J. ŽEds.., Echinoderm Studies. A.A. Balkema, Brookfield, pp. 61]102. Zamyatnin, A.A., 1972. Protein volume in solution. Prog. Biophys. Mol. Biol. 24, 107]123.
Collagen fibril aggregation-inhibitor from sea cucumber dermis John A. Trotter a,U , Gillian Lyons-Levy a , Kazumi Chino a , Thomas J. Koob b, Douglas R. Keene c , Mark A.L. Atkinsond a
Department of Cell Biology and Physiology, Uni¨ ersity of New Mexico School of Medicine, Albuquerque, NM 87131, USA b Shriners Hospital for Children, Tampa, FL, USA c Shriners Hospital for Children, Portland, OR, USA d Department of Biochemistry, Uni¨ ersity of Texas Health Science Center, Tyler, TX, USA Received 25 June 1999; received in revised form 17 September 1999; accepted 21 September 1999
Abstract Collagen fibrils from the dermis of the sea cucumber Cucumaria frondosa are aggregated in vitro by the dermal glycoprotein stiparin ŽTrotter et al., 1996.. Under physiological ionic conditions stiparin appears to be both necessary and sufficient to cause fibrils to aggregate ŽTrotter et al., 1997.. We report here the initial biochemical and biophysical characterization of a sulfated glycoprotein from C. frondosa dermis that binds stiparin and inhibits its fibril-aggregating activity. This inhibitory glycoprotein, which has been named ‘stiparin-inhibitor,’ has the highest negative charge density of all the macromolecules extracted from the dermis. SDS-PAGE reveals three ; 31-kDa bands that stain with alcian blue but not with Coomassie blue. Analytical ultracentrifugation indicates a native molecular weight of 62 kDa. Transmission electron microscopy of rotary-shadowed molecules shows curved rods about 22 nm long. The glycoprotein does not bind collagen fibrils, but does bind stiparin with a 1:1 stoichiometry. The binding of stiparin-inhibitor to stiparin prevents the binding of stiparin to collagen fibrils. The carbohydrate moiety produced by papain-digestion of the glycoprotein retains all of its inhibitory activity. The carbohydrate moiety of the inhibitor is dominated by galactose and sulfate. Q 1999 Published by Elsevier Science B.V.rInternational Society of Matrix Biology. All rights reserved. Keywords: Collagen; Fibril; Aggregation; Inhibitor; Stiparin; Echinoderm
1. Introduction Animals in the phylum Echinodermata share the ability to regulate the tensile properties of their col-
Abbre¨ iations: CS: chondroitin sulfate; DMMB: dimethyl-methylene blue; EDTA: ethylenediamine-tetra-acetic acid; GuHCl: guanidine hydrochloride; MOPS: 3-w N-morpholinoxpropane-sulfonic acid; PTC: phenylthiocarbamyl; Tris: tris-hydroxymethyl-aminomethane; TCA: trichloroacetic acid; TFA: trifluoroacetic acid. U Corresponding author. Tel.: q1-505-272-5700; fax: q1-505272-6581. E-mail address: [email protected] ŽJ.A. Trotter.
lagenous tissues on a physiological time scale ŽMotokawa, 1984; Wilkie, 1984, 1996.. This phenomenon has not been reported in animals in other phyla. The biomechanical properties of echinoderm ‘mutable connective tissues’ ŽMCTs. are regulated by physiological control of the interactions between collagen fibrils ŽHidaka, 1983; Hidaka and Takahashi, 1983; Trotter and Koob, 1995; Trotter et al., 1996; Trotter and Chino, 1997; Koob et al., 1999.. The collagen fibrils of echinoderms are symmetrically spindle-shaped and relatively short Žup to approx. 1 mm in length. ŽMatsumura, 1973, 1974; Trotter and Koob, 1989; Thurmond and Trotter, 1994;
0945-053Xr99r$ - see front matter Q 1999 Published by Elsevier Science B.V.rInternational Society of Matrix Biology. All rights reserved. PII: S 0 9 4 5 - 0 5 3 X Ž 9 9 . 0 0 0 5 0 - 5
570
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
Trotter et al., 1994, 1998.. They have surface-bound proteoglycans which are associated with the same location in each D-period. The fibrils are stabilized internally by the same types of covalent crosslinks found in vertebrate fibrils ŽBailey et al., 1982; Eyre, et al., 1984; Trotter and Koob, 1989; Trotter et al., 1994.. Importantly, however, they lack permanent or covalent inter-fibrillar crosslinks ŽTrotter and Koob, 1989; Trotter et al., 1996.. The absence of permanent crosslinks between echinoderm collagen fibrils has made it possible to develop methods to isolate abundant quantities of intact fibrils in the native state ŽMatsumura, 1973, 1974; Trotter et al., 1996.. Native fibrils isolated from the dermis of the North Atlantic sea cucumber C. frondosa remain suspended in physiological salt solutions. However, addition of stiparin, which is the most abundant soluble glycoprotein in the dermis ŽTrotter et al., 1996., causes fibrils to aggregate. Stiparin-mediated fibril interactions are thought to be important both in the organization of the tissue and in the regulation of its viscoelastic properties ŽTrotter et al., 1996, 1997.. Because of the physiological reversibility of the stiffening reaction of the dermis, it was predicted that one or more soluble inhibitors of stiparin should also be present in the dermis. An in vitro assay for stiparin-inhibitory activity was used to identify macromolecules in dermal extracts that inhibit stiparin-dependent fibril-aggregation. This communication describes the purification and initial characterization of the first of such inhibitory molecules to have been identified.
centration of collagen in the fibril suspensions was estimated based on the colorimetric determination of hydroxyproline in hydrolysates ŽBergman and Loxly, 1963. and on our previous determination that hydroxyproline accounts for 7.7% of the mass of the collagen molecule ŽTrotter et al., 1995.. 2.3. Stiparin purification and chromatographic methods Stiparin was purified from the inner dermis as described in Trotter et al. Ž1996.. Ion-exchange and size-exclusion chromatography were performed as previously described ŽTrotter et al., 1996. using HiTrap Q, Mono-Q, Superdex-75 and 200, and Superose-6 columns ŽAmersham Pharmacia Biotech... 2.4. Amino acid analyses Amino acid analyses were performed on an Applied Biosystems Model 420A amino acid analyzer equipped with on-line PTC pre-column derivitization. Samples were subjected to vapor phase hydrolysis for 24 h at 1108C in 6 N HCl. Two separate hydrolysates were analyzed in triplicate and results are expressed as the averages of these determinations. For tyrosine sulfate determinations the protein sample and tyrosine sulfate standard were base hydrolyzed Ž4.2 M KOH. for 20 h at 1108C in vacuum. After hydrolysis base was neutralized with 4.2 M perchloric acid. The potassium perchlorate was removed by centrifugation. Twenty microliters were analyzed by the PicoTag method as described in Cohen and Strydom Ž1988..
2. Experimental
2.5. Analytical ultracentrifugation
2.1. Animals and tissue preparation
Equilibrium sedimentation measurements were made in a Beckman XL-A analytical ultracentrifuge. 100 ml samples of purified stiparin-inhibitor at a concentration of approximately 200 mgrml in 20 mM Tris]HCl ŽpH 8., 400 mM NaCl were centrifuged at 3000, 4000 and 4500 rev.rmin at 2778K in the An60T1 rotor. The cells were scanned at 280 nm. Each data point is the average of five readings taken at 0.001-cm intervals through the sample column. The sample was assumed to have reached equilibrium when multiple scans over a 4-h period did not change. The partial specific volume was calculated from the amino acid composition of the protein ŽZamyatnin, 1972., using standard values for the individual amino acids and known carbohydrate components ŽDurschlag, 1986.. Sedimentation velocity measurements were also made in the Beckman XL-A analytical ultracentrifuge. Samples were run in a two-sector cell using 400 ml of sample and 410 ml of reference buffer. Buffer conditions were the same as those used for
The collection of adult C. frondosa and preparation of dermal specimens were as described previously ŽTrotter et al., 1996.. Both the white inner dermis and the pigmented outer dermis were used, as detailed below. The tissues were stored at y708C until needed. 2.2. Preparation of collagen fibrils Collagen fibrils were isolated from pieces of inner dermis as described in Trotter et al. Ž1996.. Briefly, fibrils were extracted from minced tissue pieces that had been extensively pre-extracted in 3 M GuHCl, pH 6.0, at 48C. The GuHCl-extracted tissues were then sequentially extracted in deionized water; 4 mM EDTA, 0.1 M Tris]HCl ŽpH 7.8.; and deionized water. The second and subsequent water-extracts contained abundant quantities of ‘stripped’ fibrils, i.e. collagen fibrils that were free of non-covalently bound macromolecules ŽTrotter et al., 1995, 1996.. The con-
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
sedimentation equilibrium. The experiment was performed at 208C Ž2938K. at 35 000 rev.rmin in the An60Ti rotor. The cell was scanned at 280 nm and data points represent the average of five measurements made at 0.005-cm intervals. A total of 15 scans were made at 10-min intervals.
571
Chondroitin sulfate ŽCS. from bovine trachea ŽSigma product C8529; approx. 70% C-4-S and 30% C-6-S. was used as the standard. Concentrations were expressed as mg or mg equivalents ŽmEq. or mEq.. of CS. 2.8. Papain digestion of stiparin-inhibitor
2.6. Electron microscopy Electron microscopy and rotary shadowing of the isolated protein were carried out as described by Sakai and Keene Ž1994.. Briefly, solutions of the protein containing 0.2 M NH 4 HCO 3 were diluted in pure glycerol to a final concentration of 70% glycerol. Samples of 0.1 ml were sprayed onto freshly cleaved mica and dried in a Balzers BAE 250 evaporator at 10y6 torr. Shadowing and electron microscopy were then performed as described ŽSakai and Keene, 1994.. 2.7. Carbohydrate analyses Partial carbohydrate analysis was performed as follows. Four replicate 150-ml aliquots of the purified stiparin-inhibitor in 0.4 M NaCl, 20 mM Tris were added to ice-cold 95% ethanol, mixed and chilled in the y208C freezer for at least 1 h. The samples were then centrifuged at 16 000 = g for 15 min at 48C. The ethanol was aspirated and an additional 1.5 ml of 95% ice-cold ethanol was added to the precipitate, mixed, and held at y208C for several hours. The samples were again centrifuged as above. The precipitates were dried under vacuum. Two of the four replicates were hydrolyzed in 400 ml 2 M TFA Žnitrogen purged. at 1008C for 3 h followed by rapid drying under vacuum. To determine losses due to hydrolysis 200 ml of carbohydrate standards were hydrolyzed in parallel. The other two replicates were hydrolyzed in 400 ml 4 M HCl Žnitrogen purged. at 1008C for 5 h followed by rapid drying under vacuum. Standards Ž200 ml. were treated in parallel. The hydrolyzates were dissolved in 50 ml of water just prior to analysis on the Carbo-Pac column coupled to the pulsed amperometric detector on a Dionex system. At the start of each day, a freshly thawed standard was analyzed and used to normalize all subsequent analyses performed that day. Uronic acid was estimated by the method of Blumenkrantz and Asboe-Hansen Ž1973.; sialic acid was estimated by the method of Jourdian et al. Ž1971.; sulfate was estimated by the method of Dodgson Ž1961.; hexosamine was estimated by the method of Antonopoulos et al. Ž1964.; and protein was estimated using the micro-BCA protein assay kit purchased from Pierce Chemical ŽRockford, IL, USA.. Glycosaminoglycan concentrations were estimated using the dimethylmethylene blue assay ŽFarndale et al., 1986..
The protein moiety of the inhibitor was digested by papain ŽWorthington Biochemical. as follows. A sample containing 2 mEq. of stiparin-inhibitor was incubated with 0.03 units of papain in a total volume of 5 ml of 0.1 M Na-acetate ŽpH 5.5., containing 10 mM L-cysteine and 50 mM ethylenediamine tetra-acetic acid ŽEDTA. at 658C for 10 h. Digestion was terminated by the addition of ice-cold trichloroacetic acid ŽTCA. to a final concentration of 5%. Precipitated proteins were collected by low speed centrifugation, and the supernatant was fractionated on a Superdex 75 column in 0.4 M NaCl, 20 mM Tris]HCl ŽpH 7.8.. The column effluent was monitored for inhibitory activity. 2.9. Fibril aggregation assays Aggregation assays were carried out as previously described ŽTrotter et al., 1996. in 1 ml of solution, using buffers and salts as indicated. Stiparin-inhibitory activity was similarly assayed, with the following modifications. The minimal quantity of stiparin required to aggregate 60 mg of collagen fibrils was pre-determined and used in subsequent assays. On average, in 0.4 M NaCl ŽpH 7.8., approximately 0.25 mg of stiparin caused the aggregation of 60 mg of collagen fibrils. This is equivalent to a ratio of 3.3 mmol of stiparin per mole of collagen ŽTrotter et al., 1996.. Column fractions were analyzed by adding aliquots to stiparin and buffer in a volume of 1 ml. After gentle mixing, collagen fibrils were added and the tubes were gently mixed as described ŽTrotter et al., 1996.. 2.10. Affinity chromatography Affinity chromatography was performed using stripped collagen fibrils bound to an inert support matrix. To prepare a 1-ml fibril-affinity column, 1.6 ml of a slurry of Affi-Prep 10 support medium ŽBioRad. was washed under low vacuum with cold 10 mM MOPS ŽpH 7.0.. The washed support medium was rapidly transferred to a flask containing 20 mg of collagen fibrils in 20 ml of 10 mM MOPS ŽpH 7.0.. This mixture was evacuated for approximately 1 min, after which it was incubated overnight at 48C with gentle rocking. The binding of fibrils to the Affi-Prep matrix was stopped by the addition of 100 ml of 1 M
572
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
ethanolamine. The mixture was then centrifuged at 120 = g for 10 min and the supernatant was discarded. The pellet was washed three times with 30 ml of 0.5 M NaCl followed by two washes with phosphate-buffered saline ŽPBS.. The washed support was poured into a glass column in the same buffer and allowed to settle for 4 h. The column was then washed with 6 ml each of 0.4 M, 2 M and 0.4 M NaCl, all in 20 mM Tris]HCl ŽpH 7.8.. It was found in initial experiments that stiparin bound to this column in 0.4 M NaCl and was eluted from the column in 2 M NaCl. To assay the binding of stiparin in the presence of stiparin inhibitor, 200 ml Ž160 mg. of stiparin in 0.4 M NaCl, 20 mM Tris]HCl ŽpH 7.8. with and without stiparin inhibitor Ž600 mEq. CS; 10 mg protein. were applied to the column, followed by 3 ml of the same buffer and then 3 ml of 2 M NaCl, 20 mM Tris]HCl ŽpH 7.8.. 0.5 ml fractions of the column eluate were collected and assayed for protein using UV absorbance Ž280 nm., and for sulfated carbohydrate using the DMMB assay. 2.11. Polyacrylamide gel electrophoresis Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate ŽSDS-PAGE. was accomplished using 4]20% gradient gels, stained with alcian blue and Coomassie brilliant blue as described ŽTrotter et al., 1995.. The relative proportions of the three bands comprising the complete inhibitor were made by analyzing the laser-densitometer scans of alcian blue-stained gels with Peak Fit 4 ŽJandel Scientific, San Rafael, CA, USA..
Fig. 1. Fibril aggregation assay. Collagen fibrils are suspended in buffered salt solutions at a concentration of 60 mgrml of collagen. Oblique illumination shows diffuse light scattering of the suspended fibrils Žleft.. Addition of 0.25 mg of stiparin causes the fibrils to aggregate into a single clot Žright.. Addition of outer dermis extract both inhibits and reverses the action of stiparin.
column, which eluted in the position of the 440-kDa globular protein standard ŽFig. 2b.. SDS-PAGE of the active fractions from both anion exchange and gel filtration columns revealed a triplet of bands with apparent molecular masses of approximately 31 kDa ŽFig. 2a,b.. These bands were stained
3. Results 3.1. Identification and purification of stiparin-inhibitory acti¨ ity It was first noted that whereas crude aqueous extracts of inner dermis caused isolated collagen fibrils to aggregate, similar extracts of outer dermis did not, although both extracts contained stiparin, as shown by SDS-PAGE. This observation suggested that the outer dermis extract might contain one or more inhibitors of stiparin-dependent fibril aggregation. Indeed, the addition of outer dermis extracts to mixtures containing 60 mg of collagen fibrils and 0.25 mg of stiparin prevented stiparin-dependent aggregation ŽFig. 1.. All of the stiparin-inhibitory activity was associated with molecules that eluted from a strong anion exchanger at 2.2 M NaCl ŽFig. 2a.. No other macromolecules were eluted from the column at higher concentrations of NaCl. Similarly, stiparin-inhibitory activity was associated with a single peak from a gel filtration
Fig. 2. Chromatography of outer dermis extracts using Ža. a strong anion exchanger ŽHiTrap Q; Pharmacia. and Žb. a gel filtration column ŽSuperdex 200; Pharmacia.. Absorbance at 280 nm indicates protein concentrations; absorbance at 525 nm indicates DMMB-positive material in diluted column fractions. The extinction coefficient for chondroitin sulfate in this assay was 0.041 absorbance units per mg Žsee Section 2.7.. The arrowheads on the abscissas indicate the fractions that contained stiparin-inhibitory activity. The SDS-PAGE gels show the alcian blue staining patterns of the active fractions.
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
by alcian blue but not by Coomassie blue, a finding that is consistent with the high negative charge density of the molecule. The positions and relative amounts of the three bands were unchanged by the addition of dithiothreitol, indicating that they are not components of a disulfide-bonded complex. It was apparent from the SDS-PAGE patterns of individual fractions from gel filtration columns that the ratios of
573
the three bands were not constant, suggesting that the native molecule was not a heterotrimer containing constant molar ratios of the three bands. The relative amounts of each of the three bands was determined by quantitative densitometry of the gels of fractions taken at the beginning and end of the HiTrap Q peak ŽFig. 3a, gel lanes 1, 2. and at the beginning, middle, and end of the Superdex 200 peak ŽFig. 3b, gel lanes 1]3.. It was apparent that the slower migrating band Ži.e. the band with the highest apparent molecular mass. was most abundant at the beginning of each peak, whereas the faster migrating band was most abundant at the end, while the abundance of the middle band remained relatively constant. The composition of the inhibitory molecule was further analyzed by gel filtration chromatography in the presence of 6 M GuHCl, followed by quantitative analyses of SDS-PAGE gels of the fractions through the peak. The results indicated that the inhibitor contains three independent macromolecules that differ in molecular mass ŽFig. 3c.. They also showed that the relative positions of the three bands observed by SDS-PAGE corresponded to their relative molecular masses, i.e. the electrophoretic mobilities of the bands were inversely proportional to their molecular masses. 3.2. Physical properties of stiparin-inhibitor Sedimentation equilibrium studies in the analytical ultracentrifuge indicated that the molecular mass of the native molecule was approximately 62 kDa. Under native conditions, then, the inhibitor is probably a dimer. A dimer containing one molecule of the middle band and one molecule of either the upper or lower band would be consistent with the quantitative analyses of column fractions. Sedimentation velocity studies gave a sedimentation coefficient of 5.9 S, a diffusion coefficient of 1.003= 10y6 cm2 sy1 and a frictional ratio of 1.05, which corresponds to a prolate ellipsoid with an axial ratio of approximately 2.0. Transmission electron microscopy of the purified inhibitor following low angle-rotary shadowing ŽFig. 4B. revealed a 22-nm-long rod with some apparent flexibility Žbased on independent measurements by two individuals.. 3.3. Chemical analyses of stiparin-inhibitor
Fig. 3. Quantitative densitometry of SDS-PAGE gels of the active fractions from a HiTrap Q column Ža., and a Superdex 200 column Žb., both run under non-denaturing conditions. Also shown are the fractions of stiparin-inhibitor obtained from a Superdex 200 column run in the presence of 6 M GuHCl Žc.. The bars in each triplet indicate the relative amounts of the three bands observed after alcian blue staining. In all cases the slowest migrating band is on the left and the fastest migrating band is on the right.
Amino acid analyses revealed a composition dominated by serine, glutamic acidrglutamine, and glycine ŽTable 1.. Comparison of amino acid data in the Swissprot database using the Propsearch program ŽHobohm et al., 1994. revealed no close fits to any known proteins. The high negative charge density of the inhibitor and its staining with alcian blue sug-
574
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
Fig. 4. Transmission electron microscopy after rotary shadowing of purified stiparin ŽA.; stiparin-inhibitor ŽB.; a mixture of 1 mg stiparin per 0.7 mEq. of stiparin-inhibitor Žestimated molar ratio: 0.2 mol stiparin per mol of stiparin-inhibitor . ŽC.; and a similar mixture containing 1 mg stiparin per 0.07 mEq. of stiparin inhibitor Žestimated molar ratio: 2 mol of stiparin per mol of stiparin-inhibitor . ŽD.. The molar ratios were estimated using 62 kDa as the mass of stiparin-inhibitor and 375 kDa as the mass of stiparin. In a, individual stiparin molecules are seen as long flexible rods. In b, individual stiparin-inhibitor molecules are seen as short, somewhat bent rods. In c, individual molecules of both stiparin and stiparin-inhibitor are seen. In d, no individual stiparin-inhibitor molecules are seen, and the stiparin molecules appear in small aggregates. Bar s 100 nm.
gested that it might contain conjugated sulfate residues. Indeed, sulfate analyses revealed 0.26 mg of sulfate per mEq. of DMMB-positive material Žsee Section 2.. No detectable tyrosine sulfate was found. In contrast, sulfate determinations of the carbohydrate residue produced by papain-digestion of the inhibitor showed essentially the same ratio of sulfate to DMMB-positive material Ž0.24 mgrmEq.., suggesting that the sulfate is mainly associated with carbohydrate ŽTable 2.. Sugar analyses of both the intact inhibitor and the carbohydrate residue produced by papain-digestion revealed mostly galactose, with small quantities of galactosamine, glucosamine, and fucose ŽTable 2.. Uronic acid and sialic acid were both undetectable. The molar ratio of sulfate to galactose was approximately 4. These results suggest that the carbohydrate moiety of the inhibitor may consist mainly of polygalactose sulfate. The carbohydrate residue produced by papain digestion of the inhibitor retained essentially all of the
inhibitory activity. Inhibition of 1 mg of stiparin required 0.2 mEq. of either the digested or undigested inhibitor. In contrast, neither chondroitin-4-sulfate nor the major glycosaminoglycan isolated from the dermis ŽTrotter et al., 1995. inhibited stiparin-dependent fibril aggregation at the highest concentration used, which was more than five times the required concentration of stiparin-inhibitor Žexpressed as equivalents of chondroitin sulfate .. 3.4. Association of stiparin-inhibitor with stiparin Centrifugation of isolated collagen fibrils in the presence of inhibitor failed to remove the inhibitor from the supernatant, indicating that it does not bind the fibrils with high affinity. To determine whether the inhibitor binds stiparin, the individual molecules and a mixture of the two containing a large excess of inhibitor were analyzed by sedimentation equilibrium analysis in the ultracentrifuge. The apparent molecu-
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578 Table 1 Amino acid composition of stiparin-inhibitor AA
Resr1000
Asx Ala Arg Cys Glx Gly His Ile Leu Lys Met Phe Pro Ser Thr Trp Tyr Val Oh-Pro Oh-Lys Total
83 67 25 ND 152 141 16 24 48 22 7 26 41 204 44 ND 40 61 0 0 1000
575
stiparin and stiparin-inhibitor, no interactions were detected when the two molecules were co-chromatographed on a Superose-6 column in the presence of either 0.4 M NaCl or artificial sea water ŽTrotter et al., 1996.. The elution volume of each molecule was unaffected by the presence of the other molecule. This finding indicated that the association between the molecules was too weak to be maintained under the conditions used for chromatography. The binding of stiparin to an affinity column containing stripped collagen fibrils was studied in the presence of different NaCl concentrations and in the presence and absence of stiparin inhibitor. Stiparin bound to the column in 0.4 M NaCl and was eluted in 2 M NaCl ŽFig. 5.. In the presence of stiparin-inhibitor, however, stiparin failed to bind to the fibrils in 0.4 M NaCl. 3.5. Relati¨ e abundance of stiparin-inhibitor in inner and outer dermis
lar mass of stiparin increased by an amount approximately equal to the apparent mass of the inhibitor Žfrom 356 000 to 411 000., suggesting that the two molecules form a 1:1 complex. The association of the inhibitor with stiparin was also shown by transmission electron microscopy of the two molecules mixed in different ratios ŽFig. 4C,D; see the figure caption for detailed description.. As the relative amount of stiparin was increased, the amount of free inhibitor was decreased. The stiparin molecules appeared to be somewhat less extended in the presence of the inhibitor ŽFig. 4C., although this could not be quantified. At high stiparinrinhibitor molar ratios Žgreater than f 2. the stiparin molecules formed small aggregates ŽFig. 4D.. Although both the ultracentrifuge and electron microscope studies indicate a saturable binding between
Stiparin-inhibitor is present in both inner and outer dermis. However, the relative concentration of stiparin-inhibitor in the outer dermis is approximately 200 times that in the inner dermis. Direct extracts of inner and outer dermis in SDS sample buffer followed by SDS-PAGE and quantitative densitometry indicated that the outer dermis contains 10.75 mEq. of stiparin-inhibitor per mg wet weight whereas the inner dermis contains only 0.05 mEq. per mg.
4. Discussion It was shown previously that extraction of the 375kDa glycoprotein stiparin from C. frondosa inner dermis led to the dissociation of the tissue into a suspension of intact collagen fibrils ŽTrotter et al., 1996.. Re-addition of crude stiparin caused a crude fibril suspension to form an aggregate. Purified sti-
Table 2 Monosaccharide composition of stiparin-inhibitor a pmolrmg of CS equivalent
Fuc GalN GlcN Gal Glc Man SO4 a
SI
PAP
Ratio PAPrSI
SI
Ratios to galactose PAP
55 17 13 606 112 60 2600
99 22 16 646 12 0 2400
1.8 1.3 1.2 1.1 0.1 0.0 0.9
0.09 0.03 0.02 1.00 0.18 0.10 4.3
0.15 0.03 0.02 1.00 0.02 0.00 3.7
SI, Stiparin-inhibitor; PAP, Carbohydrate moiety produced by complete papain-digestion of stiparin-inhibitor.
576
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
Fig. 5. Affinity chromatography of stiparin on a column containing immobilized collagen fibrils. The elution pattern of stiparin is shown by the solid line. A280 indicates protein absorbance at 280 nm. The leftmost arrow on the abscissa indicates the addition of sample in 0.4 M NaCl. The rightmost arrow indicates the change of the buffer to 2 M NaCl. Stiparin binds to the fibrils in 0.4 M NaCl and elutes in 2 M NaCl. The elution patterns of stiparin and stiparin-inhibitor after a mixture of the two molecules had been applied to the column are shown by the broken line ŽA280. and the dotted line ŽA525.. A525 indicates DMMB absorbance at 525 nm. The two arrows have the same significance. In the presence of stiparin-inhibitor, stiparin fails to bind to the fibrils.
parin had the same effect on purified fibrils, suggesting that stiparin alone was both necessary and sufficient to cause fibril aggregation in vitro. The aggregation of fibrils in vitro is clearly correlated with fibril]fibril interactions in vivo, but the exact physiological role played by stiparin is not yet certain ŽTrotter et al., 1997.. Other protein molecules have been isolated from the inner and outer dermis that cause dramatic stiffness changes in intact specimens of the inner dermis ŽTrotter and Koob, 1995; Koob et al., 1999.. Despite the uncertainty concerning its precise role in dermal regulation, it is clear that the interactions between stiparin and collagen fibrils are important determinants of the viscoelastic properties of the inner dermis. Because of the significance of stiparin]fibril interactions in the dermis of C. frondosa, it was important to determine whether the dermis contains molecules which are capable of modulating the activity of stiparin. Using as a bioassay the stiparin-dependent aggregation of suspended collagen fibrils, it was possible to identify inhibitory activity in extracts of both inner and outer dermis. Because the inhibitory activity was much more abundant in the outer dermis, that tissue was used as the source of crude extracts from which the inhibitory activity was purified. It is necessary to be cautious in interpreting the interactions that occur between large macromolecules, especially when one of them is a highly charged polyelectrolyte. Interactions of such macromolecules are dependent on the ionic strength, di-
electric constant, and specific ion composition of the bathing solution, including the pH. The interactions described in this communication were generally studied in 0.4 M NaCl buffered to pH 7.8]8.2 with either Tris or MOPS, but they were also confirmed in complete artificial sea water ŽASW.. The interactions detected in vitro may thus potentially occur in vivo as well. The binding of stiparin-inhibitor to stiparin appears to be specific. Other polyelectrolytes, including chondroitin sulfate and the highly sulfated major GAG found in the dermis ŽTrotter et al., 1995. failed to bind to stiparin or to inhibit stiparin-dependent fibril aggregation in vitro. Furthermore, the ultracentrifuge and electron microscope analyses indicate specific association between the two molecules. In addition the ultracentrifuge studies indicate a 1:1 stoichiometry. The fact that both components elute from a gel filtration column in the volume that would be predicted if there were no interactions between them shows that the association constant cannot be very high. Similarly, the association of stiparin with collagen fibrils cannot be very strong. If it were, it would not be possible to elute sufficient stiparin from pieces of dermis to cause their collagen fibrils to dissociate into a suspension ŽTrotter et al., 1996.. Thus, the stiparin]fibril and stiparin-inhibitor ]stiparin interactions are both weak. The stiparin]fibril interaction is apt to be weaker, however, since stiparin-inhibitor is able quantitatively to displace bound stiparin from a fibril-affinity column. These observations suggest that
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
even relatively low stiparin-inhibitor ]stiparin ratios in vivo could effectively modulate the effectiveness of stiparin and, as a consequence, the viscoelastic properties of the dermis. Testing these predictions will require quantitative determinations of both equilibrium and kinetic constants as well as detailed characterizations of binding partners. These studies are currently underway. The composition and structure of the native stiparin-inhibitor is only suggested by the present findings. It is clear that there are three different glycoprotein chains that differ by a few percent in their molecular masses. The molecular weight of the native molecule suggests that it contains only two of the three chains, held in association by non-covalent interactions. It is also clear that the native molecule appears to be a somewhat flexible rod approximately 22 nm long. There is no information concerning the organization of the two chains in the molecule. However, the results suggest that each of the two chains is able to bind to stiparin whereas each stiparin molecule has a single binding site for stiparin-inhibitor. The sedimentation equilibrium experiment showed that, in the presence of excess stiparin-inhibitor, each stiparin molecule bound a single stiparin-inhibitor molecule. On the other hand, the electron-microscopic data suggested that, in the presence of excess stiparin, two or more stiparin molecules formed small complexes with stiparin-inhibitor. Kinetic experiments are presently underway to determine the stoichiometry and valency of the interactions between stiparin and stiparin-inhibitor. The fact that all of the inhibitory activity is associated with the polygalactose-sulfate moiety of the molecule rather than with the protein raises the possibility that the sites on fibrils recognized by stiparin are GAG sites rather than protein sites. This would be consistent with the observation that stiparin does not bind to the triple helical region of C. frondosa collagen molecules Žunpublished observations..
Acknowledgements The authors gratefully acknowledge the help of Dr Andrzej Pastuszyn and the Protein Chemistry Core Facility of the UNM School of Medicine. This work was supported by research grants to JAT from the NSF, ONR, and DARPA; by the James Robert Montgomery Endowment to MALA; and by grant No. 8610 from The Shriners of North America to TJK. The electron microscope facility used by DRK is supported by the Shriners Hospital for Children and by the R. Blaine Bramble and Fred Meyer Charitable Trust foundations. The excellent technical assistance of Sara Tufa is gratefully acknowledged.
577
References Antonopoulos, C.A., Gerdesi, S., Szirmai, J.A., de Tyssonsk, E.R., 1964. Determination of glycosaminoglycans Žmucopolysaccharides. from tissues on the microgram scale. Biochim. Biophys. Acta. 83, 1]19. Bailey, A.J., Gathercole, L.J., Dlugosz, J., Keller, A., Voyle, C.A., 1982. Proposed resolution of the paradox of extensive crosslinking and low tensile strength of cuvierian tubule collagen from the sea-cucumber Holothuria forskali. Int. J. Biol. Macromol. 4, 329]334. Bergman, I., Loxly, R., 1963. Two improved and simplified methods for the spectrophotometric determination of hydroxyproline. Anal. Chem. 35, 1961]1965. Blumenkrantz, N., Asboe-Hansen, G., 1973. New method for quantitative determination of uronic acids. Anal. Biochem. 54, 484]489. Cohen, S.A., Strydom, D.J., 1988. Amino acid analysis utilizing phenylisothiocyanate derivatives. Anal. Biochem. 174, 1]16. Dodgson, K.S., 1961. Determination of inorganic sulfate in studies on the enzymic and non-enzymic hydrolysis of carbohydrates and other sulfate esters. Biochem. J. 78, 312]319. Durschlag, H., 1986. Specific volumes of biological macromolecules and some other molecules of biological interest. In: Hinz, H.J. ŽEd.., Thermodynamic Data for Biochemistry and Biotechnology. Springer-Verlag, Berlin, pp. 45]128. Eyre, D.R., Koob, T.J., Van Ness, K.P., 1984. Quantitation of hydroxypyridinium crosslinks in collagen by high-performance liquid chromatography. Anal. Biochem. 137, 380]388. Farndale, R.W., Buttle, D.J., Barrett, A.J., 1986. Improved quantitation and discrimination of sulfated glycosaminoglycans by use of dimethylmethylene blue. Biochim. Biophys. Acta. 883, 173]177. Hidaka, M., 1983. Effects of certain physico-chemical agents on the mechanical properties of the catch apparatus of the sea-urchin spine. J. Exp. Biol. 103, 15]29. Hidaka, M., Takahashi, K., 1983. Fine structure and mechanical properties of the catch apparatus of the sea-urchin spine, a collagenous connective with muscle-like holding capacity. J. Exp. Biol. 103, 1]14. Hobohm, U., Houthaeve, T., Sander, C., 1994. Amino acid analysis and protein database compositional search as a rapid and inexpensive method to identify proteins. Anal. Biochem. 222, 202]209. Jourdian, G.W., Dean, L., Roseman, S., 1971. The sialic acids: a periodate-resorcinol method for the quantitative estimation of the sialic acids and their glycosides. J. Biol. Chem. 246, 430]435. Koob, T.J., Koob-Emunds, M.M., Trotter, J.A., 1999. Cell-derived stiffening and plasticizing factors in sea cucumber Ž Cucumaria frondosa.. J. Exp. Biol. 202 Žin press.. Matsumura, T., 1973. Shape, size and amino acid composition of collagen fibril of the starfish Asterias amurensis. Comp. Biochem. Physiol. 44B, 1197]1205. Matsumura, T., 1974. Collagen fibrils of the sea-cucumber, Stichopus japonicus: purification and morphological study. Connect. Tissue Res. 2, 117]125. Motokawa, T., 1984. Connective tissue catch in echinoderms. Biol. Rev. 59, 255]270. Sakai, L.Y., Keene, D.R., 1994. Fibrillin: monomers and microfibrils. Methods Enzymol. 245, 29]54. Thurmond, F.A., Trotter, J.A., 1994. Native collagen fibrils from echinoderms are molecularly bipolar. J. Mol. Biol. 235, 73]79. Trotter, J.A., Chino, K., 1997. Regulation of cell-dependent viscosity in the dermis of the sea cucumber Actinopyga agassizi. Comp. Biochem. Physiol. 119A, 805]811. Trotter, J.A., Koob, T.J., 1989. Collagen and proteoglycan in a sea
578
J.A. Trotter et al. r Matrix Biology 18 (1999) 569]578
urchin ligament with mutable mechanical properties. Cell Tissue Res. 258, 527]539. Trotter, J.A., Koob, T.J., 1995. Evidence that calcium-dependent cellular processes are involved in the stiffening response of holothurian dermis and that dermal cells contain an organic stiffening factor. J. Exp. Biol. 198, 1951]1961. Trotter, J.A., Thurmond, F.A., Koob, T.J., 1994. Molecular structure and functional morphology of echinoderm collagen fibrils. Cell Tissue Res. 275, 451]458. Trotter, J.A., Lyons-Levy, G., Thurmond, F.A., Koob, T.J., 1995. Covalent composition of collagen fibrils from the sea cucumber, Cucumaria frondosa, a tissue with mutable mechanical properties. Comp. Biochem. Physiol. 112A Ž3r4., 463]478. Trotter, J.A., Lyons-Levy, G., Luna, D., Koob, T.J., Keene, D.R., Atkinson, M.A.L., 1996. Stiparin: a glycoprotein from sea cu-
cumber dermis that aggregates collagen fibrils. Matrix Biol. 15, 99]110. Trotter, J.A., Salgado, J.P., Koob, T.J., 1997. Mineral content and salt-dependent viscosity in the dermis of the sea cucumber Cucumaria frondosa. Comp. Biochem. Physiol. 116A, 329]335. Trotter, J.A., Chapman, J.A., Kadler, K.E., Holmes, D.F., 1998. Growth of sea cucumber collagen fibrils occurs at the tips and centers in a coordinated manner. J. Mol. Biol. 284, 1417]1424. Wilkie, I.C., 1984. Variable tensility in echinoderm collagenous tissues: a review. Mar. Behav. Physiol. 11, 1]34. Wilkie, I.C., 1996. Mutable collagenous tissues: extracellular matrix as mechano-effector. In: Jangoux, M., Lawrence, J. ŽEds.., Echinoderm Studies. A.A. Balkema, Brookfield, pp. 61]102. Zamyatnin, A.A., 1972. Protein volume in solution. Prog. Biophys. Mol. Biol. 24, 107]123.