Glycobiology Advance Access originally published online on March 24, 2004
Glycobiology 2004 14(7):621-633; doi:10.1093/glycob/cwh076
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Glycobiology vol. 14 no. 7 © Oxford University Press 2004; all rights reserved.
Glycosaminoglycans and proteoglycans in normal mitral valve leaflets and chordae: association with regions of tensile and compressive loading
2 Department of Bioengineering, MS 142, PO Box 1892, Rice University, Houston, TX 77005; 3 Department of Biomedical Engineering, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH 44195; and 4 Department of Vascular Biology, Hope Heart Institute, Seattle, WA 98104
Received on December 1, 2003; revised on March 2, 2004; accepted on March 5, 2004
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsReferences |
|---|
This study was designed to identify the specific proteoglycans and glycosaminoglycans (GAGs) in the leaflets and chordae of the mitral valve and to interpret their presence in relation to the tensile and compressive loads borne by these tissues. Leaflets and chordae from normal human mitral valves (n = 31, obtained at autopsy) were weighed and selected portions digested using proteinase K, hyaluronidase, and chondroitinases. After fluorescent derivatization, fluorophore-assisted carbohydrate electrophoresis was used to separate and quantify the derivatized saccharides specific for each GAG type. In addition, the lengths of the chondroitin/dermatan sulfate chains were determined. Proteoglycans were identified by western blotting. The regions of the valve that experience tension, such as the chordae and the central portion of the anterior leaflet, contained less water, less hyaluronan, and mainly iduronate and 4-sulfated N-acetylgalactosamine with chain lengths of 50–70 disaccharides. These GAGs are likely associated with the small proteoglycans decorin and biglycan, which were found in abundance in the tensile regions. The valve regions that experience compression, such as the posterior leaflet and the free edge of the anterior leaflet, contained significantly more water, hyaluronan, and glucuronate and 6-sulfated N-acetylgalactosamine with chain lengths of 80–90 disaccharides. These GAGs are likely components of water-binding versican aggregates, which were abundant in the compressive loading regions. The relative amounts and distributions of these GAGs are therefore consistent with the tensile and compressive loads that these tissues bear. Finally, the concentrations of total GAGs and many different chondroitin/dermatan sulfate subclasses were significantly decreased with advancing age.
Key words: compression / decorin / proteoglycans / tension / versican
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
The mitral valve is one of the most complex connective tissue structures in the entire body. It consists of two leaflets and numerous chordae tendineae. These chordae have a highly aligned collagenous core and a thin outer sheath of elastic fibers and endothelial cells. Both leaflets are laminated tissues containing a heavily collagenous layer on the ventricular side; a predominantly elastic layer on the atrial side; and an inner spongiosa layer containing abundant proteoglycans (PGs) and hyaluronan (HA). The relative thicknesses of these layers vary between the two leaflets and also within each leaflet from its attachment edge to its free edge (Kunzelman et al., 1993
).
The variability of the different leaflet layers, and hence the structural constituents within the mitral valve, are determined by the specific functional roles of the leaflets and chordae. The closed valve, in particular, maintains a balance of tensile and compressive loads, in which the chordae and the flat central region of the anterior leaflet are in tension, whereas the free edge of the anterior leaflet and most of the posterior leaflet are in appositional compression. Accordingly, the most collagenous components of the mitral apparatus are the chordae and the portion of the anterior leaflet between the annulus and the upper appositional border (Kunzelman et al., 1993). In the posterior leaflet and in the free edge of the anterior leaflet, the collagenenous layer is relatively thinner, whereas the PG-rich spongiosa is substantially thicker.
The wide diversity of glycosaminoglycans (GAGs) and their parent PGs exert considerable yet variable control over the physical properties of the extracellular matrix (Hardingham and Fosang, 1992; Junqueira and Montes, 1983). Over the past 50 years, exhaustive biochemical analyses have been conducted on the GAGs of heart valves of pigs (Castagnaro et al., 1997), cows (Bostrom et al., 1963; Deiss and Leon, 1955; Honda et al., 1976; Lowther et al., 1970; Meyer et al., 1969; Moretti and Whitehouse, 1963; von Figura et al., 1973), rodents (Colvee and Hurle, 1981; Hallgrimsson et al., 1970), and humans (Baig et al., 1978; Masuda, 1984,; Murata, 1981; Torii and Bashey, 1966; Torii et al., 1965). These studies have examined the influence of subject age on the synthesis and content of the classes of GAGs in all the heart valves. However, these previous studies agree only moderately about the general proportions of HA and the chondroitin/dermatan sulfates (CS/DS) in valve tissues. Although there is a consensus regarding the decrease of total GAG concentration with age, reports disagree on age-related changes in specific GAG classes. Despite the many studies of mitral valve leaflets, only one report describes the GAG classes in chordae (von Figura et al., 1973). Furthermore, the majority of previous studies have considered the mitral valve leaflets to be homogeneous, which clearly is not the case. One study did examine the regional biosynthesis and content of GAGs in mitral leaflets, but did not differentiate between GAG classes (Bostrom et al., 1963). There has also been only one group that attempted to link these GAG profiles with the presence of a particular PG (Toole and Lowther, 1968). In general, there is scant information regarding the influence of GAGs and PGs on the mechanical properties and function of the valve tissues.
Recent developments in carbohydrate analysis using fluorophore-assisted carbohydrate electrophoresis (FACE) (Calabro et al., 2000a,b) have enabled the study of GAGs in heart valves using novel and efficient methods. The resolution of the bands in FACE enables differentiation between 4- and 6-sulfation of N-acetylgalactosamine (galNAc) on iduronate (idoA)-based disaccharides. The quantification of terminal saccharides (primarily sulfated galNAc) on the same gels simplifies the calculation of the CS/DS chain lengths. Therefore this new technology can quickly provide characteristics about the GAG profile, chain lengths, and sulfation patterns that can provide clues to the identity of particular PGs within mitral valve tissues and might be used to postulate their mechanical contributions to mitral valve function.
This study was designed with several objectives. The first was to determine if the fine structure and chain length of GAGs present in mitral valve tissues are dependent on the predominant type of tissue loading. For this purpose, mitral valves from normal human autopsy subjects were cut into regions of compressive and tensile loading, and their GAG characteristics were measured using FACE. The second objective was to determine if the resulting load-related GAG profiles are dependent on subject demographics (age, body surface area, gender, and race). The third objective was to compare the measured GAG profiles, chain lengths, and sulfation patterns with published data characterizing different PGs to identify the main types of PGs contained within the different loading regions of the valve. Last, western blotting was used to confirm the identity of these PGs in valve tissues.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Mitral valve anatomy
The mitral valve contains two leaflets that attach to the junction between the left atrium and left ventricle (Figure 1). The free edges and ventricular surfaces of the leaflets are connected to the papillary muscles of the left ventricle by numerous chordae tendineae. Mitral valve closure is maintained by apposition of large portions of the anterior and posterior leaflets, the tension on the chordae following contraction of the papillary muscles and the strength of the pressurized anterior leaflet. Mitral valves from 35 normal autopsy subjects were dissected into anterior leaflets (free edge and center), posterior leaflets, and chordae tendineae. Thirty of these valves were used for the entire study; five were used to measure water concentration only.
|
Water and total GAGs
The hydration and GAG composition of mitral valve tissues were straightforwardly examined by measuring the concentrations of water and total GAGs. Not surprisingly, the water content followed GAG content between different regions of the valve (Figure 2). The concentrations of both water and total GAGs were greatest in the two compressive regions of the valve (posterior leaflet and free edge of the anterior leaflet), significantly less in the center of the anterior leaflet (p < 0.001 for water, p = 0.002 for total GAGs), and lowest in the chordae (p < 0.001 for both variables). Chordal type was also a significant factor, in that the basal chordae that insert directly underneath the leaflets contained less water content than the marginal chordae that insert into the leaflet free edges (73% versus 77%, p = 0.003).
|
ANOVA p < 0.01 versus leaflet groups.GAG compositional analysis
GAG composition was measured from the solubilized tissue samples by enzymatic cleavage of the GAG chains into disaccharides, followed by fluorescent derivatization and FACE. This procedure resulted in the quantitation of
DiHA, glucuronate (glcA)-based disaccharides containing either 4-, 6-, and unsulfated galNAc, idoA-based disaccharides containing either 4- or 6-sulfated galNAc, di- and trisulfated CS/DS, and the terminal N-acetylhexosamines (Figure 3). (In this study, the designation of a GAG as containing iduronate or glucuronate was based on the susceptibility to chondroitinase ACII. CS/DS is meant to refer to bands that could contain either iduronate- or glucuronate-based disaccharides. This nomenclature is consistent with the designation of an intact GAG chain as CS if the chain contains exclusively glucuronate and DS if the chain contains any iduronate at all.) In the posterior leaflet and the free edge of the anterior leaflet (compressive regions), the proportions of DiHA and total CS/DS were approximately equivalent (Figure 4). In the chordae and in the center of the anterior leaflet (tensile regions), there was approximately three times as much CS/DS as DiHA.
|
|
|
The minimum percentage of glcA (Figure 5) was determined from the relative band intensities after two different enzyme treatments: (1) chondroitinase ABC combined with hyaluronidase SD (HABC) and (2) chondroitinase ACII (ACII). Because the ACII treatment cleaves bonds adjacent to glcA but not idoA, the ACII treatment measures the minimum glcA content, which was then divided by the HABC product (glcA + idoA) to determine the percentage value. In the first set of columns in Figure 5 (top), the minimum percentage of glcA content in HA was approximately 100%, as expected. These data indicated that the ACII and HABC treatments were capable of liberating equivalent amounts of
DiHA. (Because we determined that the chondroitinase ACII enzyme performs equally well as hyaluronidase SD in the digestion of hyaluronan, we have since switched from our HABC mixture to an ACII/ABC mixture (3 µl each) for complete digestion of hyaluronan and CS/DS. This alternative offers substantial economic savings.) The second set of columns (top) shows that the total CS/DS disaccharide bands from valve tissues contained a mixture of glcA and idoA. The minimum glcA present in the posterior leaflet and free edge of the anterior leaflet was higher than in the chordae and the center of the anterior leaflet (analysis of variance [ANOVA] p < 0.001 between groups). Likewise, disaccharides appearing in the 6-sulfated or 4-sulfated band positions (Figure 5, middle) could have contained either glcA or idoA. The 6-sulfated position bands contained glcA almost exclusively, whereas the 4-sulfated position bands were largely idoA. Both the unsulfated and the oversulfated CS/DS bands (Figure 5, bottom) contained 75%–100% glcA.
|
Additional data about the specific GAG disaccharide species were then calculated by subtracting the ACII data from the HABC data to estimate the maximum idoA contents. The ratios between different GAGs and their sulfation patterns were also calculated. In both the posterior leaflet and the free edge of the anterior leaflet, the most common GAG classes were found to be
DiHA (47%–50% of total GAG disaccharides) and glucuronate with 6- sulfated galNAc (glcA-6S-galNAc) (Table I, Figures 6 and 7). The two compressive regions contained similar GAG profiles, although the free edge of the anterior leaflet had significantly more glcA-6S-galNAc (p = 0.01) and unsulfated CS/DS (Di0S, p = 0.01) and slightly more DiHA (p = 0.08) than the posterior leaflet (Figure 6). In contrast, the central region of the anterior leaflets contained mainly iduronate with 4-sulfated galNAc (idoA-4S-galNAc, 36% of total GAGs), less glcA relative to idoA (p < 0.001), and greater amounts of 4-sulfation overall (p < 0.001) as compared to the other leaflet samples (Table I, Figures 6 and 7). Quite notably, this central portion of the anterior leaflet contained significantly less glcA-6S-galNAc (p < 0.001) and DiHA (p < 0.001) than the adjacent anterior leaflet free edge. The most predominant GAG in chordae was also idoA-4S-galNAc (40% of total GAGs). Chordae contained the greatest proportions of idoA relative to glcA and of 4-sulfation relative to 6-sulfation, as compared with the compressive leaflet regions (Table I, p < 0.001).
|
|
The FACE gels also demonstrated bands corresponding to 6- and 4-sulfated galNAcs (the nonreducing termini of the CS/DS chains, with a 4S:6S ratio of 5:1) for all the valve samples. It has been demonstrated in aggrecan that the termini of CS/DS chains consist of 85%–90% nonreducing galNAcs and 10%–15% nonreducing disaccharides (Plaas et al., 1997
). Therefore to identify nonreducing disaccharides that would run in the Di6S and Di4S band positions, another set of samples were digested with HABC, then treated with mercuric acetate to cleave the uronate from the hexosamine. This treatment showed that there were no detectable nonreducing disaccharides; after the reducible Di6S and Di4S bands had migrated to the sulfated galNAc band positions, only residual 2-aminoacridone hydrochloride remained underneath the Di6S and Di4S bands.
Chain lengths
The number of disaccharides in the CS/DS chains was estimated from the ratio of the total amount of the CS/DS disaccharides to the total amount of saccharides at the nonreducing termini of the chains. The average lengths of the CS/DS chains were 82 ± 24 and 83 ± 28 disaccharides in the posterior leaflet and the free edge of the anterior leaflet, respectively. Chain length in the center of the anterior leaflet was slightly less (p = 0.081) at 69 ± 19 disaccharides. The shortest chain lengths were found in chordae, with an average of 56 ± 27 disaccharides (p < 0.001 versus posterior leaflet and free edge of anterior leaflet). Regions of compression therefore had longer chain lengths than regions of tension.
Influence of subject demographics
The mitral valves used in this study were discarded from autopsies of normal humans who had no known or suspected history of any cardiac, cardiovascular (i.e., hypertension), or valvular dysfunction. This subject population was 46.4 ± 15.3 years old, 71% male/26% female (1 gender unknown), and 60% Caucasian/40% black. There were no significant differences in age between different genders or different races. Moreover, there were identical distributions of race and gender above and below the mean age. Likewise, the mean body surface area was identical for subgroups younger and older than the mean age and for difference races.
Because the age of the normal human autopsy subjects has been previously reported to influence the concentration of GAGs in heart valves (Baig et al., 1978; Murata, 1981; Torii et al., 1965), the individual GAG class concentrations were analyzed with respect to subject demographics. Subject age had a considerable influence on the content of the individual GAG classes in the tensile loading regions. Significant negative correlations were found between subject age and the concentration of DiHA, glcA-6S-galNAc, glucuronate with 4-sulfated galNAc (glcA-4S-galNAc), idoA-4S-galNAc, total GAGs, and water (correlation coefficients and p-values in Table II, Figure 8). When the proportions of the different GAG classes were calculated relative to the total GAG quantity as measured by FACE, however, the relative proportions of idoA-4S-galNAc and idoA-4S-galNAc were positively correlated with age in the tensile loading regions. The proportion of DiHA in tensile regions decreased slightly with age.
|
|
In compressive tissues (the posterior leaflet and the free edge of the anterior leaflet analyzed together), there were significant negative correlations between age and the concentration of glcA-6S-galNAc and glcA-4S-galNAc (Table II). Age was correlated somewhat with the concentration of total GAGs and the concentration of water. The length of the CS/DS chains was significantly reduced with age in the compressive tissues, and was slightly reduced with age in chordae tendineae.
Subject gender was a significant factor in the content of individual GAG classes in the tensile loading regions, particularly in chordae. The concentration of DiHA in tensile regions was 27% less in females than in males (p = 0.05). The concentration of idoA-4S-galNAc and total GAGs, respectively, were 35% and 28% greater in females (p < 0.01 for both). Furthermore the proportions of DiHA and glcA-6S-galNAc in the tensile regions were 25% less (p = 0.009) and 17% less (p = 0.034) in females. The only significant gender difference in the compressive loading regions was a 16% lower proportion of glcA-6S-galNAc in females (p = 0.032).
The influences of race and body surface area were also examined. Black subjects had an 89% greater concentration and 47% greater relative proportion of glcA-4S-galNAc in the posterior leaflet (p = 0.013 for both), and a 20% greater concentration of idoA-4S-galNAc in their chordae (p = 0.05). Caucasian subjects had a 42% greater concentration (p = 0.045) and a 25% greater proportion (p = 0.049) of glcA-6S-galNAc in the anterior leaflet free edge. There were no significant correlations between any biochemical variable and the subject body surface area.
PG measurements
Immunoblotting was used to confirm the identity of guanidine HCl–extracted, purified PG core proteins from different regions of the mitral valve. The core proteins for decorin (43 kDa doublet) and biglycan (43 kDa) were found in blots from all valve tissues. Versican core protein was also present in all valve tissues as lightly visible V0 (
450 kDa) and abundant V1 (400 kDa) isoforms, in a 3:1 V1:V0 ratio. Because all the specimens from each valve were run together on the same gel, trends between regions of compressive and tensile loading were examined by normalizing band intensities from each PG to the band intensity from the anterior leaflet free edge for each individual valve (Figure 9). These ratios of band intensities generally agreed with our predicted findings based on the GAG results. Decorin was most abundant in the center of the anterior leaflet (tensile), present in lower quantities in the free edge of the anterior leaflet (compressive) and least abundant in the chordae tendineae (tensile) and posterior leaflet (compressive). Biglycan was again most abundant in the center of the anterior leaflet, less abundant in the chordae tendineae and anterior leaflet free edge, and least abundant in the posterior leaflet.
|
An ANOVA found a statistical difference between valve regions for the decorin and biglycan (both p < 0.001). In contrast, versican was present in greater quantities in the posterior leaflet and free edge of the anterior leaflet (compressive regions) with lower quantities in the chordae tendineae and center of the anterior leaflet (tensile regions), although this trend was not a significant difference, likely due to the small sample size. The decorin and biglycan samples were analyzed eight times to establish intraobserver variability as 3.9% for band quantitation, 5.5% for lane-lane repeatability from the same precipitated volume, and 9.7% for repeatability in separately precipitated samples from the same extraction volume. There were no statistical differences between data from different gel runs, precipitations, and GelPro analyses (paired t-tests, two-tailed p > 0.05). Due to the large volume of sample required to visualize the versican bands, the versican samples were not run in duplicate.
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
This study showed that the components of the mitral valve apparatus that experience predominantly tensile loads, such as the central portion of the anterior leaflet and the chordae, contain relatively less water and less hyaluronan. The most abundant GAG was iduronate with 4-sulfated galNAc, and the average chain length was 50–70 disaccharides. In the regions that experience predominantly compressive loads, such as the leaflet free edges and most of the posterior leaflet, there were longer GAG chains and elevated concentrations of water,
DiHA, and glucuronate with 6-sulfated galNAc. These data are supported by immunoblots that confirm the presence of versican, decorin, and biglycan in these valve tissues.
Although the division of the mitral valve into regions that experience tensile and compressive loading certainly simplifies its very complex loading regime, a rationale for this segregation is provided by the valve's mechanical and microstructural heterogeneity. For example, the elastic moduli of the center region and of the chordae tendineae are higher than those of the anterior leaflet free edge and the posterior leaflet (Clark, 1973; Kunzelman and Cochran, 1992; May-Newman and Yin, 1995). In addition, both the central portion of the anterior leaflet (adjacent to the annulus) (Cochran et al., 1991) and the chordae (Millington-Sanders et al., 1998) contain collagen fibers that are highly oriented in the main loading direction, whereas the posterior leaflet and the free edge of the anterior leaflet have less collagen fiber alignment (Kunzelman, 1991) and a thicker GAG-rich spongiosa (Kunzelman et al., 1993).
Our findings agree well with previously published data, despite the fact that many of these previous studies did not differentiate between the 6-sulfated and 4-sulfated forms of galNAc. The GAG profiles of the compressive loading regions (predominantly DiHA) match the published data from adult humans (Baig et al., 1978; Murata, 1981; Torii et al., 1965), although the data here contain a greater proportion of glcA-6S-galNAc than reported previously, likely due to our sectioning of the tissue into different loading regions. The profiles found in the tensile regions (dominated by idoA-4S-galNAc) resemble the reported profile of chordae tendineae (von Figura et al., 1973), although that study did not differentiate between sulfation patterns. The findings of this present study show better agreement overall with previous studies in human adults (Baig et al., 1978; Murata, 1981; Torii et al., 1965) than with studies of valves from young animals (Castagnaro et al., 1997; von Figura et al., 1973). These similarities show that FACE, a much more sensitive and convenient method, can be used in place of earlier techniques.
The correlations between GAG content and subject age confirmed or expanded on previous reports. The total concentration of GAGs in mitral valves decreased with age (Baig et al., 1978; Moretti and Whitehouse, 1963; Murata, 1981; Torii et al., 1965), as did the concentrations in the many GAG subclasses. When the GAG class concentrations were converted into proportions of the total GAG content, the proportion of idoA in the tensile regions of the mitral valve increased with age, as previously noted for heart valves overall (Moretti and Whitehouse, 1963; Murata, 1981; Torii et al., 1965). In the compressive regions, the reduction of glcA with age also confirms other reports (Moretti and Whitehouse, 1963; Torii et al., 1965). Although prior studies found that the proportion of HA either increased (Murata, 1981) or decreased with age (Moretti and Whitehouse, 1963; Torii et al., 1965), this study found no significant correlation between HA proportion and age in any tissue component of the normal mitral valve. Furthermore the contents of the GAG classes were not correlated with the subject body surface area, but gender and race had a mild effect. Although the relevance of these demographic influences is unknown, it is notable that the incidences of certain valvular cardiac diseases are also associated with age, gender, and race.
To hypothesize the type of PGs found in the tensile and compressive loading regions of the mitral valve, the GAG class concentrations and fine structure characteristics measured from mitral valves were compared with the reported characteristics of PGs found in cardiovascular soft tissues. The compressive loading regions contained high concentrations of water, DiHA, and glcA, an average chain length of 83 disaccharides, and an average glcA-6S-galNAc/glcA-4S-galNAc ratio of 2.7, which are similar to the characteristics of the PG versican. Versican, also known as PG-M, is a large PG that is synthesized by arterial smooth muscle cells (Chang et al., 1983; Schonherr et al., 1991), and has been found in aorta (Yao et al., 1994), skin (Zimmermann et al., 1994), and malignant tumors (Isogai et al., 1996). Versican consists of 15–20 CS chains (Schonherr et al., 1991) branching off a 430–460-kDa core protein (Chang et al., 1983; Schonherr et al., 1991; Yao et al., 1994). Versican CS chains are reportedly 43–45 kDa (85–100 disaccharides) in length (Chang et al., 1983) and composed of 60%–80% glcA (Schonherr et al., 1991) with a glcA-6S-galNAc/glcA-4S-galNAc ratio of at least 2.0 (Chang et al., 1983; Schonherr et al., 1991). Furthermore the versican GAG chains create macromolecular domains encompassing large volumes of solvent (Hardingham and Fosang, 1992), and versican itself aggregates with HA (Chang et al., 1983; Schonherr et al., 1991; Zimmermann and Ruoslahti, 1989). These macromolecular constituents are ideally suited to distribute changes in compressive load as a consequence of repetitive changes in strain.
The tensile regions, in contrast, contained a high concentration of idoA, an overall 6S/4S ratio of 0.34, and an average chain length of 58 disaccharides, which are more indicative of the PGs decorin and biglycan. These two small PGs contain one and two GAG chains, respectively, of 25–40 kDa in length (50–80 disaccharides) (Fisher et al., 1989; Oldberg et al., 1996; Toole and Lowther, 1968; Wight et al., 1991). These PGs contain 70% idoA with an overall 6S/4S ratio of 0.35 for idoA and glcA combined (Chang et al., 1983; Roughley and White, 1992). DS and, more specifically, decorin have been frequently found in tissues with an abundance of type I collagen, such as tendon (Junqueira and Montes, 1983; Toole and Lowther, 1968) and pericardium (Simionescu et al., 1988). Decorin binds tightly to type I collagen fibers and, based on the tensile forces, governs fibril formation and fibril diameter. Biglycan is similar to decorin but contains two GAG chains and may also have functional roles in highly collagenous tissues (Ameye et al., 2002). Furthermore, a small collagen-associated DS PG that was likely decorin was found in bovine valve tissues by Toole and Lowther (1965). The data from this study are also supported by the work of Daniel and Mills (1988), in which cells grown from the tensile region of rabbit flexor tendon produced small PGs that were only 48% susceptible to enzymatic digestion by chondroitinase AC, indicating a high proportion of idoA. Cells grown from the compressive region of the tendon, in contrast, produced large PGs that were wholly digestible by both chondroitinases ABC and AC, indicating that the GAGs present uniformly contained glcA. These data are very similar to our results. In addition, the concentration of collagen in heart valve tissues increases with age (Trnavsky et al., 1965), as does the proportion of idoA-4S-galNAc measured here, suggesting that these GAGs may be components of collagen-associated decorin and biglycan.
The hypothesized differential abundances of these PGs in tensile versus compressive loading regions of the mitral valve was partially confirmed by our immunoblotting of valve extracts. It was evident that versican, decorin, and biglycan were present in some quantity throughout the valves, which agreed with the earlier results showing that all regions had a wide distribution of GAGs. There were significant regional differences, however, in the relative amounts of decorin and biglycan, and versican was slightly more abundant in the compressive loading regions.
The difference in the GAG profiles and relative PG composition of the compressive as compared to tensile loading regions of the valve reflects the structural and functional differences of these regions. The long, multiple glcA-6S-galNAc chains on the PG versican extend away from the core protein to minimize electrostatic interactions. The resulting large hydrodynamic volumes are ideal to respond to variable compressive loads and to withstand high pulsatile forces (Wight et al., 1991). Versican would thus allow the loosely layered appositional surfaces of the mitral valve to reversibly buffer the considerable impact and shearing deformations that occur during valve opening and closing. The core protein of versican also contains a HA-binding region (LeBaron et al., 1992; Zimmermann and Ruoslahti, 1989) that anchors the PGs on long strands of hyaluronan and thereby retains them in the tissue. The combined versican-HA aggregate has been suggested to perform a lubrication and antiadhesive role in tissues by preventing ligand–receptor interactions between cells and matrix (Lemire et al., 1996; Yamagata et al., 1989). This lubrication could also aid in viscous dissipation for the impact forces during valve closure (von Figura et al., 1973). In contrast, several characteristics of the GAG profile suggest the presence of decorin and biglycan in the more solid tensile loading structures of the valve. Decorin in particular is abundant in connective tissues with high concentrations of type I collagen (Bianco et al., 1990), in which it stabilizes and orients collagen fibrils (Hedbom and Heinegard, 1989; Toole and Lowther, 1968). Decorin therefore may play a role in the overall strength of high load–bearing tissues. Biglycan, which typically does not colocalize with decorin, also contributes to collagen fibril diameter and tissue strength (Ameye et al., 2002).
In conclusion, these data on GAGs and PGs from human mitral valves indicate that the regions in compression are rich in versican and contain abundant glucuronate and 6-sulfated galNAc, with little iduronate. Regions in tension are rich in decorin and biglycan and demonstrate the opposite pattern, with abundant iduronate (and almost exclusively 4-sulfated galNAc) but less glucuronate. The compositional patterns of mitral valve GAGs and PGs thus provide new insight into the roles of these molecules in load-bearing tissues.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Normal mitral valves (n = 35) were obtained at autopsy from persons who died of noncardiac-related causes and were harvested after less than 24 h in cold storage. Our laboratory has determined that valves kept at 4°C retain their matrix-based mechanical properties for at least 5 days (Patel, 2003
). We therefore believe that the matrix in these cadaveric valves was not measurably degraded during the postmortem period. All chemicals were obtained from either Sigma (St. Louis, MO) or Fisher (Pittsburgh, PA), except for glycerol, guanidine HCl, and proteinase-K (Invitrogen, Carlsbad, CA); hyaluronidase SD, chondroitinases ABC and ACII, and 2-B-1 antibody for versican (Seikagaku America, Falmouth, MA); Q-Sepharose Fast Flow beads and peroxidase-linked secondary antibodies (Amersham, Uppsala, Sweden); Triton X-100 (Roche, Indianapolis, IA); and 2-aminoacridone HCl (Molecular Probes, Eugene, OR). Monosaccharide electrophoresis running buffer and the preformed monosaccharide gels were purchased from Glyko (Novato, CA). The polyclonal antibodies LF136 for decorin and LF51 for biglycan were generously provided by Larry Fisher at the National Institute of Dental and Craniofacial Research, NIH.
Sample preparation
Mitral valves were prepared by first removing the chordae then sectioning the valve into anterior and posterior leaflets (Figure 1). The anterior leaflets were sectioned into the central region and the free edge. The chordae were trimmed from the leaflets at their point of attachment, and divided into four groups: anterior basal, anterior marginal, posterior basal, and posterior marginal. The wet weights of the leaflet samples and chordal samples (chordae were pooled into the four groups) were determined. The water content was calculated after lyophilizing the samples for 16 h and then weighing again to obtain the dry weight (Figure 2). The dried samples (typically 20–200 mg each) were stored in microcentrifuge tubes at –20°C prior to analysis.
GAG analyses
For subsequent analysis, each sample was minced with fine scissors and 1 ml of 100 mM ammonium acetate buffer (pH 7) added. A 100-µl aliquot of proteinase-K solution (10 mg/ml) was added to each sample followed by incubation for 16 h at 60°C. The completely solubilized digests were then heated at 100°C to inactivate the proteinase-K. An aliquot (100 µl) of each digest was taken to measure the quantity of uronic acid to estimate total GAG content (Figure 2) (Blumenkrantz and Asboe-Hansen, 1973) and hence to determine the volume required for the FACE analysis.
The samples were analyzed by FACE to quantify the different classes of GAGs (Calabro et al., 2000b). Two aliquots containing 5 µg uronic acid were diluted to 100 µl in buffer and then incubated with either (1) 2 µl hyaluronidase SD + 3 µl chondroitinase ABC (each 10 mU/µl, termed HABC), or (2) 3 µl chondroitinase ACII (10 mU/µl, termed ACII) for 3 h at 37°C. The HABC treatment provides a measure of the minimum glcA content, allowing the maximum idoA contents to be estimated by subtracting the ACII data from the HABC data. One HABC digested aliquot was further treated with mercuric acetate to cleave the unsaturated hexuronate from the hexosamines and to determine the identity and quantity of any nonreducing terminal disaccharides. After digestion, the samples were dried, fluorotagged, mixed with glycerol and an internal standard (see later description), and electrophoresed on a monosaccharide gel as previously described (Calabro et al., 2000a,b). The gel bands were imaged and analyzed using Gel-Pro (Media Cybernetics, Silver Spring, MD).
Known serially diluted quantities of a 2-sulfated chondroitin sulfate disaccharide (not found in valves) were added to individual samples to provide an internal calibration curve for fluorescence intensity (Figure 3). Enzyme digestion products were identified by correspondence to bands in a disaccharide standard lane. The quantity of each GAG was determined from the integrated optical density of the band(s), the calibration curve, and the aliquot volume used for the enzyme digest (Calabro et al., 2000a,b).
The resulting GAG profiles were used to calculate several factors specific to particular PGs. The proportions of DiHA to total CS/DS, as well as the ratio of ACII to HABC digestion products, assessed the regional balance of GAG classes and glcA/idoA (Figures 4– 5). The ratios of 6-sulfation to 4-sulfation of the galNAcs were calculated for glcA- and idoA-containing disaccharides together and separately (Table I). The ratio of glcA to idoA was also calculated. Finally, the number average lengths of the CS/DS chains were estimated by calculating on a lane-by-lane basis the total amount of the CS/DS disaccharides (except the 2S internal standard) and dividing by the total amount of saccharides at the nonreducing termini of the chains (Figure 3).
The measured GAG class contents from each sample were divided by the sample dry weight to estimate the tissue concentrations (Figure 6). The proportions of the different GAG classes were then calculated relative to the total GAG quantity as measured by FACE (Figure 7). Statistical comparisons between GAG class concentrations, proportions, or chain lengths in the different valve regions were calculated using a repeated measures ANOVA, followed by Tukey tests as needed, with two-tailed significance accepted at p < 0.01 (reduced due to numerous statistical comparisons between regions). Two-group comparisons (different genders and race) were conducted using t-tests with two-tailed significance accepted at p < 0.05. Correlation coefficients between the GAG class concentrations/proportions and subject age and body surface area (Figure 8) were calculated using linear regression with significance accepted at p < 0.05.
PG extraction and purification
Five fresh mitral valves were dissected into regional samples as described, lyophilized overnight, and reweighed to obtain dry weight. Each sample was minced with fine scissors in a 2-ml centrifuge tube and then agitated in extraction buffer overnight at 4°C (4 M guanidine HCl, 0.5% 3- [(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate, 0.05 M ammonium acetate, 0.01 M ethylenediamine tetra-acetic acid (EDTA), 0.1 M 6-aminohexanoic acid, 0.08% benzamidine HCl, 10 mM N-ethyl maleimide, 1 mM phenylmethylsulfonyl fluoride (PMSF); 1 ml per 25 mg tissue dry weight). After extraction, the samples were centrifuged (13,000 rpm) and the supernatant dialyzed four times against 7 M urea buffer (containing 2 mM EDTA, 0.05 M Tris, 0.5% Triton X-100, pH 7.5) to remove the guanidine. After dialysis, volumes of extract solutions containing equivalent proportions of starting dry mass were mixed with Q-Sepharose beads, and the beads were rinsed with 40 column volumes of 7 M urea buffer containing 0.25 M NaCl. The bound purified PGs were eluted with six column volumes of 7 M urea buffer containing 3 M NaCl.
SD–PAGE and western blotting
Equivalent aliquots of the purified PG samples were mixed with water to a final volume of 300 µl and precipitated by adding 1 ml 95% ethanol/1.3% potassium acetate and incubating at –20°C for 2 h. The precipitate was suspended in 20 µl of enzyme digest solution (containing 2.5 mU/µl chondroitinase ABC, 0.01% bovine serum albumin (BSA), 0.05 M Tris, 3 mM Na acetate, 8 mM 6-aminohexanoic acid, 0.42 mM benzamidine HCl, and 0.08 mM PMSF) and incubated at 37°C for 3 h. Samples were then mixed with an equivalent volume of sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) sample buffer containing 5% ß-mercaptoethanol, boiled for 5 min, and run on a 4%–12% SDS–PAGE gel at constant 180 V. The gel was transferred to a 0.2-µm nitrocellulose membrane at 100 V (1 h for decorin and biglycan, 3 h for versican). The membrane was blocked in Tris-buffered saline (TBS) with 0.1% Tween-20 and 2% BSA overnight at 4°C, then treated with primary antisera to decorin, biglycan, or versican (1:6000 dilution in TBS/Tween containing 2% fetal bovine serum) overnight at 4°C. After four washes in TBS/Tween, the membrane was treated with horseradish peroxidase–linked secondary antibodies (1:20,000 dilution in 2% BSA) for 2 h at room temperature, then washed six times more. Proteins were detected using chemiluminescent exposure to radiographic film. PG bands were identified by comparison with positive controls and quantified using densitometry. Each PG band (decorin, biglycan, or versican) was normalized by the content of the corresponding band from the anterior leaflet free edge sample run on the same gel (Figure 9).
|
Acknowledgements |
|---|
We express heartfelt gratitude to the late Dr. Hans Kresse, University of Münster, for his encouraging discussions and experimental advice. We also thank Christina Chan in the Wight laboratory for experimental instruction, and Jamie Hulbert in the Vesely laboratory for assistance in the validation of chondroitinase ACII as a substitute for hyaluronidase SD. Funding for this research was provided by the Mareb Foundation, the Research Programs Council of the Cleveland Clinic Foundation Lerner Research Institute, an NIH National Research Service Award Postdoctoral Fellowship, and the Whitaker Foundation.
|
Footnotes |
|---|
1 To whom correspondence should be addressed; e-mail: grande{at}rice.edu
|
Abbreviations |
|---|
ACII, chondroitinase ACII; ANOVA, analysis of variance; BSA, bovine serum albumin; CS, chondroitin sulfate; DS, dermatan sulfate; EDTA, ethylenediamine tetra-acetic acid; FACE, fluorophore-assisted carbohydrate electrophoresis; GAG, glycosaminoglycan; HA, hyaluronan; HABC, hyaluronidase SD combined with chondroitinase ABC; PG, proteoglycan; PMSF, phenylmethylsulfonyl fluoride; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; TBS, Tris-buffered saline
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Ameye, L., Aria, D., Jepsen, K., Oldberg, A., Xu, T., and Young, M.F. (2002) Abnormal collagen fibrils in tendons of biglycan/fibromodulin-deficient mice lead to gait impairment, ectopic calcification, and osteoarthritis. FASEB J., 16, 673–680.
Baig, M., Daicoff, G., and Ayoub, E. (1978) Comparative studies of the acid mucopolysaccharide composition of rheumatic and normal heart valves in man. Circ. Res., 42, 271–275.
Bianco, P., Fisher, L., Young, M., Termine, J., and Robey, P. (1990) Expression and localization of the two small proteoglycans biglycan and decorin in developing human skeletal and non-skeletal tissues. J. Histochem. Cytochem., 38, 1549–1563.[Abstract]
Blumenkrantz, N. and Asboe-Hansen, G. (1973) New method for quantitative determination of uronic acids. Anal. Biochem., 54, 484–489.[CrossRef][Web of Science][Medline]
Bostrom, H., Moretti, A., and Whitehouse, M. (1963) Studies on the biochemistry of heart valve I. On the biosynthesis of mucopolysaccharides in bovine heart valves. Biochim. Biophys. Acta, 74, 213–221.[Medline]
Calabro, A., Benavides, M., Tammi, M., Hascall, V., and Midura, R. (2000a) Microanalysis of enzyme digests of hyaluronan and chondroitin/dermatan sulfate by fluorophore-assisted carbohydrate electrophoresis (FACE). Glycobiology, 10, 273–281.
Calabro, A., Hascall, V., and Midura, R. (2000b) Adaptation of FACE methodology for microanalysis of total hyaluronan and chondroitin sulfate composition from cartilage. Glycobiology, 10, 283–293.
Castagnaro, M., Amedeo, S., Bertolotto, A., Manzardo, E., Riccio, A., and Guarda, F. (1997) Morphological and biochemical investigations of mitral valve endocardiosis in pigs. Res. Vet. Sci., 62, 121–125.[CrossRef][Web of Science][Medline]
Chang, Y., Yanagishita, M., Hascall, V.C., and Wight, T.N. (1983) Proteoglycans synthesized by smooth muscle cells derived from monkey (Macaca nemestrina) aorta. J. Biol. Chem., 258, 5679–5688.
Clark, R. (1973) Stress-strain characteristics of fresh and frozen human aortic and mitral leaflets and chordae tendineae. Implications for clinical use. J. Thorac. Cardiovasc. Surg., 66, 202–208.[Web of Science][Medline]
Cochran, R.P., Kunzelman, K.S., Chuong, C.J., Sacks, M.S., and Eberhart, R.C. (1991) Nondestructive analysis of mitral valve collagen fiber orientation. ASAIO Trans., 37, M447–M448.[Medline]
Colvee, E. and Hurle, J.M. (1981) Maturation of the extracellular material of the semilunar heart values in the mouse. A histochemical analysis of collagen and mucopolysaccharides. Anat. Embryol., 162, 343–352.[CrossRef][Medline]
Daniel, J.C. and Mills, D.K. (1988) Proteoglycan synthesis by cells cultured from regions of the rabbit flexor tendon. Connect. Tissue. Res., 17, 215–230.[Web of Science][Medline]
Deiss, W. and Leon, A. (1955) Mucopolysaccharides of heart valve mucoprotein. J. Biol. Chem., 215, 685–689.
Fisher, L.W., Termine, J.D., and Young, M.F. (1989) Deduced protein sequence of bone small proteoglycan I (biglycan) shows homology with proteoglycan II (decorin) and several nonconnective tissue proteins in a variety of species. J. Biol. Chem., 264, 4571–4576.
Hallgrimsson, J., Friberg, U., and Burke, J.F. (1970) Changes with age in the metabolism of sulfated glycosaminoglycans in guinea pig cardiac valves. Exp. Mol. Pathol., 12, 70–83.[CrossRef][Web of Science][Medline]
Hardingham, T.E. and Fosang, A.J. (1992) Proteoglycans: many forms and many functions. FASEB J., 6, 861–870.[Abstract]
Hedbom, E. and Heinegard, D. (1989) Interaction of a 59-kDa connective tissue matrix protein with collagen I and collagen II. J. Biol. Chem., 264, 6898–6905.
Honda, A., Kanke, Y., and Mori, Y. (1976) Proteoglycan complexes from bovine heart valve. Fractionation by density-gradient centrifugation and gel filtration under dissociative conditions. J. Biochem., 79, 17–25.
Isogai, Z., Shinomura, T., Yamakawa, N., Takeuchi, J., Tsuji, T., Heinegard, D., and Kimata, K. (1996) 2B1 antigen characteristically expressed on extracellular matrices of human malignant tumors is a large chondroitin sulfate proteoglycan, PG-M/versican. Cancer Res., 56, 3902–3908.
Junqueira, L.C. and Montes, G.S. (1983) Biology of collagen-proteoglycan interaction. Arch. Histol. Jpn., 46, 589–629.[Medline]
Kunzelman, K.S. (1991) Engineering analysis of mitral valve structure and function. PhD diss., Biomedical Engineering, University of Texas Southwestern Medical Center, Dallas, TX.
Kunzelman, K.S. and Cochran, R.P. (1992) Stress/strain characteristics of porcine mitral valve tissue: parallel versus perpendicular collagen orientation. J. Card. Surg., 7, 71–78.[Web of Science][Medline]
Kunzelman, K.S., Cochran, R.P., Murphree, S.S., Ring, W.S., Verrier, E.D., and Eberhart, R.C. (1993) Differential collagen distribution in the mitral valve and its influence on biomechanical behaviour. J. Heart Valve Dis., 2, 236–244.[Medline]
LeBaron, R.G., Zimmermann, D.R., and Ruoslahti, E. (1992) Hyaluronate binding properties of versican. J. Biol. Chem., 267, 10003–10010.
Lemire, J.M., Potter-Perigo, S., Hall, K.L., Wight, T.N., and Schwartz, S.M. (1996) Distinct rat aortic smooth muscle cells differ in versican/PG-M expression. Arterioscler. Thromb. Vasc. Biol., 16, 821–829.
Lowther, D.A., Preston, B.N., and Meyer, F.A. (1970) Isolation and properties of chondroitin sulphates from bovine heart valves. Biochem. J., 118, 595–601.[Web of Science][Medline]
Masuda, H. (1984) The preparation of heparan sulfate from the mitral valve of the human heart. Int. J. Biochem., 16, 99–103.[CrossRef][Web of Science][Medline]
May-Newman, K. and Yin, F.C. (1995) Biaxial mechanical behavior of excised porcine mitral valve leaflets. Am. J. Physiol., 269, H1319–H1327.[Web of Science][Medline]
Meyer, F.A., Preston, B.N., and Lowther, D.A. (1969) Isolation and properties of hyaluronic acid from bovine heart valves. Biochem. J., 113, 559–563.[Web of Science][Medline]
Millington-Sanders, C., Meir, A., Lawrence, L., and Stolinski, C. (1998) Structure of chordae tendineae in the left ventricle of the human heart. J. Anat., 192, 573–581.[CrossRef][Web of Science][Medline]
Moretti, A. and Whitehouse, M. (1963) Changes in the mucopolysaccharide composition of bovine heart valves with age. Biochem. J., 87, 396–402.[Medline]
Murata, K. (1981) Acid glycosaminoglycans in human heart valves. J. Mol. Cell. Cardiol., 13, 281–292.[CrossRef][Web of Science][Medline]
Oldberg, A., Antonsson, P., Moses, J., and Fransson, L.A. (1996) Amino-terminal deletions in the decorin core protein leads to the biosynthesis of proteoglycans with shorter glycosaminoglycan chains. FEBS Lett., 386, 29–32.[CrossRef][Web of Science][Medline]
Patel, J. (2003) Effect of absolute specimen size on the tensile properties of porcine aortic valve tissues. Master's thesis, Biomedical Engineering, Case Western Reserve University, Cleveland, OH.
Plaas, A., Wong-Palms, S., Roughley, P., Midura, R., and Hascall, V. (1997) Chemical and immunological assay of the nonreducing terminal residues of chondroitin sulfate from human aggrecan. J. Biol. Chem., 272, 20603–20610.
Roughley, P.J. and White, R.J. (1992) The dermatan sulfate proteoglycans of the adult human meniscus. J. Orthop. Res., 10, 631–637.[CrossRef][Web of Science][Medline]
Schonherr, E., Jarvelainen, H.T., Sandell, L.J., and Wight, T.N. (1991) Effects of platelet-derived growth factor and transforming growth factor-beta 1 on the synthesis of a large versican-like chondroitin sulfate proteoglycan by arterial smooth muscle cells. J. Biol. Chem., 266, 17640–17647.
Simionescu, D., Alper, R., and Kefalides, N.A. (1988) Partial characterization of a low molecular weight proteoglycan isolated from bovine parietal pericardium. Biochem. Biophys. Res. Commun., 151, 480–486.[CrossRef][Web of Science][Medline]
Toole, B.P. and Lowther, D.A. (1965) The isolation of a dermatan sulphate-protein complex from bovine heart valves. Biochim. Biophys. Acta, 101, 364–366.[Medline]
Toole, B.P. and Lowther, D.A. (1968) Dermatan sulfate-protein: isolation from and interaction with collagen. Arch. Biochem. Biophys., 128, 567–578.[CrossRef][Web of Science][Medline]
Torii, S. and Bashey, R. (1966) High content of hyaluronic acid in normal human heart valves. Nature, 209, 506–507.[Medline]
Torii, S., Bashey, R., and Nakao, K. (1965) Acid mucopolysaccharide composition of human-heart valve. Biochim. Biophys. Acta, 101, 285–291.[Medline]
Trnavsky, K., Kopecky, S., Trnavska, Z., and Cebecauer, L. (1965) Influence of age on biochemical composition of the mitral valve connective tissue. Gerontologia, 11, 169–178.[Web of Science][Medline]
von Figura, K., Born, W., and Buddecke, E. (1973) 14C and 35S-SO4 metabolism of acid glycosaminoglycans in bovine heart valves. Angiologica, 10, 129–141.[Web of Science][Medline]
Wight, T.N., Heinegard, D., and Hascall, V.C. (1991) Proteoglycans: structure and function, in Hay, E.D. (Ed.), Cell biology of extracellular matrix. Plenum Press, New York, pp. 45–78.
Yamagata, M., Suzuki, S., Akiyama, S.K., Yamada, K.M., and Kimata, K. (1989) Regulation of cell-substrate adhesion by proteoglycans immobilized on extracellular substrates. J. Biol. Chem., 264, 8012–8018.
Yao, L.Y., Moody, C., Schonherr, E., Wight, T.N., and Sandell, L.J. (1994) Identification of the proteoglycan versican in aorta and smooth muscle cells by DNA sequence analysis, in situ hybridization and immunohistochemistry. Matrix Biol., 14, 213–225.[CrossRef][Web of Science][Medline]
Zimmermann, D.R. and Ruoslahti, E. (1989) Multiple domains of the large fibroblast proteoglycan, versican. EMBO J., 8, 2975–2981.[Web of Science][Medline]
Zimmermann, D.R., Dours-Zimmermann, M.T., Schubert, M., and Bruckner-Tuderman, L. (1994) Versican is expressed in the proliferating zone in the epidermis and in association with the elastic network of the dermis. J. Cell Biol., 124, 817–825.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Krishnamurthy, D. B. Ennis, A. Itoh, W. Bothe, J. C. Swanson, M. Karlsson, E. Kuhl, D. C. Miller, and N. B. Ingels Jr. Material properties of the ovine mitral valve anterior leaflet in vivo from inverse finite element analysis Am J Physiol Heart Circ Physiol, September 1, 2008; 295(3): H1141 - H1149. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Ferdous, V. M. Wei, R. Iozzo, M. Hook, and K. J. Grande-Allen Decorin-transforming Growth Factor- Interaction Regulates Matrix Organization and Mechanical Characteristics of Three-dimensional Collagen Matrices J. Biol. Chem., December 7, 2007; 282(49): 35887 - 35898. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.J. Grande-Allen, N. Osman, M.L. Ballinger, H. Dadlani, S. Marasco, and P.J. Little Glycosaminoglycan synthesis and structure as targets for the prevention of calcific aortic valve disease Cardiovasc Res, October 1, 2007; 76(1): 19 - 28. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Timek, D. T. Lai, P. Dagum, D. Liang, G. T. Daughters, N. B. Ingels Jr, and D. C. Miller Mitral Leaflet Remodeling in Dilated Cardiomyopathy Circulation, July 4, 2006; 114(1_suppl): I-518 - I-523. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. C. Caira, S. R. Stock, T. G. Gleason, E. C. McGee, J. Huang, R. O. Bonow, T. C. Spelsberg, P. M. McCarthy, S. H. Rahimtoola, and N. M. Rajamannan Human Degenerative Valve Disease Is Associated With Up-Regulation of Low-Density Lipoprotein Receptor-Related Protein 5 Receptor-Mediated Bone Formation J. Am. Coll. Cardiol., April 18, 2006; 47(8): 1707 - 1712. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||