SEARCH

SEARCH BY CITATION

Keywords:

  • collagen;
  • heart;
  • cardiac;
  • valves;
  • myocardium

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Genetic mutations in minor fibrillar collagen types Va1 (ColVa1) and XIa1 (ColXI) have been identified in connective tissue disorders including Ehlers–Danlos syndrome and chondrodysplasias. ColVa1+/− and ColXIa1−/− mutant mice recapitulate these human disorders and show aberrations in collagen fiber organization in connective tissue of the skin, cornea, cartilage, and tendon. In the heart, fibrous networks of collagen fibers form throughout the ventricular myocardium and heart valves, and alterations in collagen fiber homeostasis are apparent in many forms of cardiac disease associated with myocardial dysfunction and valvular insufficiency. There is increasing evidence for cardiac dysfunction in connective tissue disorders, but the mechanisms have not been addressed. ColVa1+/− and ColXIa1−/− mutant mice were used to identify roles for ColVa1 and ColXIa1 in ventricular myocardial morphogenesis and heart valve development. These affected cardiac structures show a compensatory increase in type I collagen deposition, similar to that previously described in valvular and cardiomyopathic disease. Morphological cardiac defects associated with changes in collagen fiber homeostasis identified in ColVa1+/− and ColXIa1−/− mice provide an insight into previously unappreciated forms of cardiac dysfunction associated with connective tissue disorders. Developmental Dynamics 235:3295–3305, 2006. © 2006 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Collagen fibers are the main structural elements in the extracellular matrix (ECM), providing tensile strength and maintaining structural architecture of connective tissues that resist shear, tensile, or pressure forces (Burgeson and Hollister,1979; Linsenmayer et al.,1983; Fitch et al.,1984; Mendler et al.,1989; Bosman and Stamenkovic,2003). Type I collagen (ColI) is the major fibrillar collagen and is a prominent protein in connective tissue throughout the body (Bosman and Stamenkovic,2003). Quantitatively minor collagens, including types V (ColV) and XI (ColXI), associate with ColI and regulate collagen fibril diameter and number in vitro (Fichard et al.,1995; Bosman and Stamenkovic,2003). In vivo, ColV regulates collagen fibril initiation and assembly within extracellular matrices in the skin and cornea, and it has been postulated that ColXI has a similar regulatory role (Seegmiller and Monson,1982; Eyre and Wu,1987; Li et al.,1995; Birk,2001; Wenstrup et al.,2004). In the heart, collagens are the most abundant component of the cardiac ECM and contribute to normal cardiac function (Robinson et al.,1983; Weber et al.,1988).

The working myocardium is composed of myocytes and interstitial fibroblasts connected by an elaborate ECM network that includes collagen fibers, proteoglycans, and glycoproteins (Baudino et al.,2006). The matrix provides structural support and facilitates ventricular function in response to changes in mechanical forces within the myocardium during the cardiac cycle (Weber et al.,1988,1989; Eghbali et al.,1989). Dysregulation of collagen fibers is detrimental to myocardial function and increased deposition in the form of fibrosis has been reported in many models of myocardial hypertrophy and cardiac failure (Diez et al.,2005; Baudino et al.,2006). Fibrillar collagens are also prominent in the heart valve leaflets and supporting structures, where they crosslink with proteoglycans and glycosaminoglycans to maintain morphology, facilitate leaflet movement, and withstand sheer stress caused by continuous blood flow (Schoen,2005; Lincoln et al.,2006b). The collagen matrix is not static. During embryonic development and in response to changes in mechanical stimuli during all stages of life, the balance between collagen synthesis and degradation changes by ECM remodeling (Borg and Burgess,1993; Rabkin-Aikawa et al.,2005; Baudino et al.,2006; Lincoln et al.,2006b; Miner and Miller,2006). Histopathological alterations in collagen homeostasis are observed in forms of valvular and myocardial dysfunction and connective tissue disease (D'Armiento,2002; Schoen,2005; Wenstrup,2005; Baudino et al.,2006; Lincoln et al.,2006b). Therefore, an optimum balance of collagen fiber formation and organization is required for normal connective tissue function, and aberrations in collagen homeostasis are associated with tissue dysfunction.

Pathological changes in the collagen matrix are a feature of many forms of connective tissue disease (Wenstrup,2005; Lincoln et al.,2006b). This change has been characterized in several mouse models with mutations in fibrillar collagens that recapitulate human connective tissue disorders (Andrikopoulos et al.,1995; Li et al.,1995; Aszodi et al.,2000; Wenstrup et al.,2000). Mutations in ColVa1 have been identified in over 40% of Ehlers–Danlos syndrome (EDS) cases (Schwarze et al.,2000; Wenstrup et al.,2000). ColVa1−/− mice are embryonic lethal at embryonic day (E) 10.5; however, ColVa1+/− mice are viable and exhibit features of EDS associated with structural weaknesses in connective tissues including skin, tendon, and cornea (Wenstrup et al.,2004,2006; Segev et al.,2006; Wenstrup, unpublished observations). The reduction of ColVa1 in these mice leads to an imbalance in collagen homeostasis indicated by abnormal sequestration of ColI and malformation of collagen scaffolds, collectively causing tissue dysfunction as previously reported in the skin (Wenstrup et al.,2004,2006). ColXIa1−/− mice die at birth and are a model for human disorders that resemble chondrodysplasias, including Stickler syndrome (Monson and Seegmiller,1981; Li et al.,1995). Chondrodysplasia disease is characterized by skeletal defects caused by abnormalities in the cartilage of limb, ribs, mandible, and trachea, resulting in abnormal ECM organization and tissue fragility (Seegmiller et al.,1971; Snead and Yates,1999). It has been suggested that the loss of ColXI results in aberrations in collagen fiber assembly and changes in cohesive properties leading to defects in affected tissues, similar to that described in ColVa1 heterozygous mutants (Monson and Seegmiller,1981; Li et al.,1995; Snead et al.,1996; Wenstrup et al.,2004). It appears from previous reports using these mutant mouse models that normal structure and function of connective tissues including skin, tendon, and cartilage is compromised in the absence or reduction of minor regulatory collagen types V and XI (Li et al.,1995; Wenstrup et al.,2004; Segev et al.,2006; Wenstrup, unpublished observations). These models provide insights into roles of type V and XI collagens in normal structure and function of skin, cornea, cartilage, and tendon, but roles in the heart have not been reported.

In the embryonic and adult heart, collagen-rich ECM is important for structural integrity and function of both the heart valves and ventricular myocardium (Jane-Lise et al.,2000; Lincoln et al.,2006b). The heart valve leaflets and supporting apparatus are composed of complex ECM consisting of highly organized collagen, elastin, and proteoglycan (Schoen,2005; Lincoln et al.,2006b). The distribution of specific collagen types in the valves is required for normal heart valve function, and dysregulation of ECM is associated with changes in valve structure leading to tissue dysfunction (Lincoln et al.,2006b). Several collagen types (I, II, III, V, VI, XI, XXIII) have been identified in distinct compartments of embryonic and adult heart valves and are associated with the differential functions of the valve leaflets vs. the supporting apparatus (Icardo and Colvee,1995; Hinton et al.,2006; Lincoln et al.,2006b). Specific collagens (I, III, V) also contribute to the fibrous ECM of the ventricular myocardium (Medugorac and Jacob,1983; Eghbali et al.,1989). These collagens are highly organized to provide a supporting fibrous scaffold able to transmit forces in the myocardium during the cardiac cycle (Baudino et al.,2006). Loss of collagen fiber regulation, in particular the accumulation of collagen in the myocardium, is detrimental to cardiac function and results in fibrosis in failing hearts (Baudino et al.,2006). There is increasing evidence to show that pathological alterations in collagen expression and organization in the ECM of the ventricular myocardium and heart valves is associated with a loss of structural morphology, tissue dysfunction, and degenerative heart disease (Schoen,2005; Baudino et al.,2006).

Cardiac dysfunction in connective tissue disease has not been fully investigated in human patient populations or mouse model systems. There is emerging evidence to show prevalence of cardiac malfunction in patients with EDS, Stickler syndrome, and other connective tissue disorders (Dolan et al.,1997; Ahmad et al.,2003; Schoen,2005; Hinton et al.,2006). However, the association of cardiac phenotypes with other previously reported connective tissue defects has not been examined. It is known from previous connective tissue disease studies in skin, tendon, and cartilage that defects in tissue integrity are associated with aberrant collagen fiber regulation and changes in tissue compliance (Li et al.,1995; Wenstrup et al.,2004; Wenstrup, unpublished observations). A thorough histological analysis of cardiac tissue in humans or mutant mice with connective tissue disorders is not yet available. In this study, ColVa1+/− and ColXIa1−/− mice were examined as models of connective tissue diseases to determine the effects of mutations in fibrillar collagens on the morphology of the heart. ColVa1 is expressed in the supporting structures of the heart valves and throughout the compact layer and trabeculae of the ventricular myocardium. In contrast, ColXIa1 is differentially expressed throughout the valve structures and the myocardial trabeculae, but not the compact layer. Hearts from ColVa1+/− mice are morphologically normal, yet there is a significant accumulation of ColI deposited in the myocardium of the ventricular chamber. In contrast the ventricular chamber of hearts from ColXIa1−/− mice are “round-ended,” which may be secondary to thickening of the heart valve leaflets with increased ColI expression. These findings highlight potential roles for ColVa1 and ColXIa1 in normal myocardial and heart valve function and identify previously unappreciated cardiac phenotypes in mouse models of connective tissue disorders.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

ColVa1, ColXIa1, and ColI Are Differentially Expressed in the Developing Heart Valves and Ventricular Myocardium During Late Stages of Embryogenesis

To examine the expression patterns of ColVa1, ColXIa1, and ColI in the late fetal mouse heart, immunofluorescence (IF) or in situ hybridization was performed on tissue sections at E18.5. Expression of these collagens was specifically noted in remodeling heart valve leaflets and supporting structures (Fig. 1B,F,J), and the compact layer (Fig. 1C,G,K) and trabeculae of the ventricular myocardium (Fig. 1D,H,L). ColVa1 protein is present primarily along the ventricular surface (arrow, Fig. 1B) of the remodeling valve leaflets with notably less expression on the atrial surface (asterisk, Fig. 1B) and throughout the chordae tendineae (arrowhead, Fig. 1B). In the myocardium, ColVa1 expression is detected in both the compact layer and the developing trabeculae (arrows, Fig. 1C,D). ColXIa1 expression is more widely distributed throughout the valve leaflet, including the atrial and ventricular leaflet surfaces (arrow, asterisk, respectively, Fig. 1F) and supporting structures than ColVa1 (arrowhead, Fig. 1F). Notably in the ECM of the myocardium, ColXIa1 is not detectable in the compact layer (Fig. 1G), but is expressed in the trabeculae (arrows, Fig. 1H). ColI is expressed throughout the valve leaflets (arrows, asterisk, Fig. 1I,J), and low levels of expression are detected in the compact layer and trabeculae of the ventricular myocardium (arrows, Fig. 1K,L). Together these expression patterns demonstrate differential expression of ColXIa1 and ColI throughout the heart valves and supporting apparatus, in contrast to ColVa1 that is preferentially expressed in the ventricular aspect of the valve leaflet and supporting chordae tendineae. In the matrix of the ventricular myocardium, ColVa1 and ColI show overlapping expression in the compact layer and trabeculae, while ColXIa1 is only detectable in the developing trabeculae.

thumbnail image

Figure 1. ColVa1, ColXIa1, and ColI are differentially expressed in the heart valve structures and ventricular myocardium at embryonic day (E) 18.5. A–L: Immunohistochemistry (A–D, I–L) and in situ hybridization (E–H) of ColVa1 (A–D), ColXIa1 (E–H), and ColI (I–L) expression in heart at E18.5 (A,E,I). The developing heart valves (A,B,E,F,I,J), compact layer (C,G,K), and trabeculae (D,H,L) areas highlighted in A, E, and I are shown at higher magnifications in B–D, F–H, J–L. A–D, I–L: Immunoreactivity of ColV and ColI are shown by immunofluorescence (green) in contrast to TOPRO-3 staining that marks nuclei blue. B: ColVa1 is expressed toward the ventricular surface (arrow) as opposed to the atrial surface (asterisk) of the valve leaflet and throughout the chordae tendineae (arrowhead). F,J: ColXIa1 (purple colormetric staining) and ColI are similarly expressed throughout the developing mitral (mv) and tricuspid (tv) valve leaflets (asterisk, arrow) and chordae tendineae (arrowhead). C,D: ColVa1 is additionally expressed in the fibrous matrix of the compact myocardium and trabeculae (arrows). G,H: ColXIa1 expression is undetected in the compact layer, but expression is prominent in the matrix of the trabeculae (arrows). I–L: ColI is expressed in mv and tv leaflets (arrows), with low levels of expression throughout the ventricular chamber compartments (arrow). mv, mitral valve; tv, tricuspid valve.

Download figure to PowerPoint

Morphogenesis of the Atrioventricular Heart Valves and Ventricular Myocardium of ColXIa1−/− and ColVa1+/−;ColXIa1−/− Mice Is Abnormal

ColVa1+/− and ColXIa1−/− mutant mice have dysregulation of collagen fibril assembly in developing tendons, skin, and skeletal structures (Li et al.,1995; Wenstrup et al.,2004,2006). The effects of reduced ColVa1 or ColXIa1 on heart development was examined in ColVa1+/− and ColXIa1−/− mice. In addition, ColVa1+/−;ColXIa1−/− mice were also examined to determine whether the cardiac morphology is further compromised in mice in which both minor collagen types are reduced. ColVa1−/− mice die at E10.5 with indications of cardiac insufficiency, before formation of the four-chambered heart (Wenstrup et al.,2004). However, ColVa1+/− mice are viable and allow studies of the myocardium and heart valves. ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice die at birth, due to presumed respiratory failure (Li et al.,1995). Therefore, hearts from all mutant and wild-type (WT) individuals were harvested at E18.5. ECM composition, distribution, and cardiac morphology were evaluated and reported using Movat's Pentachrome stain in ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice.

Hearts from ColVa1+/− mice appear morphologically comparable to WT littermates (Fig. 1A–D) with normal-shaped ventricles, including a distinct apex (pound sign, Fig. 2A–D) and interventricular septum (IVS) that separates the ventricular chambers (asterisk, Fig. 2A,C). The remodeling heart valve leaflets of WT and ColVa1+/− animals are characteristically thin, elongated structures rich in proteoglycan as indicated by Alcian blue staining (arrows, Fig. 2A–D). In contrast, loss of ColXIa1 results in striking changes in the gross appearance of the ventricles (Fig. 2E). Specifically, the distinction of the apex in ColXIa1−/− animals is less well defined (pound sign, Fig. 2E), and the ventricles appear more rounded in shape than WT and ColVa1+/− mice (asterisk, Fig. 2E). These morphological changes are associated with a significant 1.4-fold increase in IVS thickness compared with WT littermates as determined by morphometric analysis (arrowhead, Fig. 2E). Similar to the morphology changes identified in the ventricles of ColXIa1−/− mice, ColVa1+/−;ColXIa1−/− mice also display a “round-end” apex (pound sign, Fig. 2G) and thickening of the IVS (1.5-fold increase over WT mice; asterisk, Fig. 2G). Notably there is significant thickening of the heart valve leaflets compared with WT littermates (1.3-fold; arrows, Fig. 2H). In summary, cardiac morphology in ColVa1+/− mice is comparable to WT; however, ColXIa1−/− mice display a loss of ventricular chamber apex definition along with significant thickening of the IVS. ColVa1+/−;ColXIa1−/− mice exhibit similar ventricular chamber phenotypes to ColXIa1−/− mice with additional significant thickening of the heart valve leaflets.

thumbnail image

Figure 2. ColXIa1−/− and ColVa1+/−;ColXIa1−/− mutant mice show increased atrioventricular (AV) valve thickness and compromised ventricular chamber morphology. A–H: Movat's Pentachrome staining of sectioned wild-type (WT; A,B), ColVa1+/− (C,D), ColXIa1−/− (E,F), and ColVa1+/−;ColXIa1−/− (G,H) hearts at embryonic day (E) 18.5. The remodeled valve structures are easily distinguishable by the Alcian blue staining of the proteoglycans within this structure. B,D,F,H: Arrows indicate the septal mitral valve (mv) and mural tricuspid valve (tv) leaflets. E–H: ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice display increased thickness of AV valve leaflets and compromised ventricular chamber morphology indicated by changes in the position of the apex (pound sign), and an increased thickness in the intraventricular septum (asterisk), compared with ColVa1+/− (C,D) and control wild-type mice (A,B). IVS, intraventricular septum.

Download figure to PowerPoint

ColI Immunoreactivity Is Increased in Hearts From ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− Mutant Mice

Previous studies of ColVa1+/− and ColXIa1−/− mice demonstrated loss of collagen fiber organization and defects in the assembly of ColI in skin (Wenstrup et al.,2004,2006). ColI localization was determined by IF in hearts from ColVa1+/−, ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice (Fig. 3). Expression patterns were examined in the heart valve leaflets and myocardial compact layer and trabeculae in three independent embryos (Figs. 4, 5). In WT hearts, ColI is expressed in the great arteries and vessels (asterisk, Fig. 3A), throughout the atrial septum (pound sign, Fig. 3A), in the epicardium (concave arrow, Fig. 3A), in the remodeling valve structures (arrow, Fig. 3A), and in the ECM of the atrial and ventricular myocardium (arrowhead, Fig. 3A). Hearts from ColVa1+/− mice are morphologically normal; however, increased ColI deposition is evident in localized regions of the compact layer of the ventricular myocardium compared with WT animals (arrowhead, Fig. 3B). ColXIa1−/− mice show a marked increase in ColI within the thickened heart valves (arrow, Fig. 3C) and as localized deposits within the compact layer of the morphometrically abnormal myocardium (arrowhead, Fig. 3C). In hearts from ColVa1+/−;ColXIa1−/− mice, increased immunoreactivity of ColI is apparent in both the thickened heart valve structures (arrow, Fig. 3D) and the affected compact layer of the ventricular chamber compared with WT control mice (arrowhead, Fig. 3D).

thumbnail image

Figure 3. ColI expression is increased in the heart valves and ventricular chamber of ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− hearts. A–D: Immunofluorescent staining of ColI in wild-type (WT) (A), ColVa1+/− (B), ColXIa1−/− (C), and ColVa1+/−;ColXIa1−/− (D) hearts at embryonic day (E) 18.5. Expression is noted in the remodeled mitral (mv) and tricuspid (tv) valves, and the fibrous matrix of the ventricular myocardium (arrowhead).

Download figure to PowerPoint

thumbnail image

Figure 4. Atrioventricular and aortic valves of ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice show increased ColI expression. A–H: Immunofluorescent staining of ColI in the septal mitral valve (mv) leaflet (A,C,E,G) and aortic (Ao) valve leaflets (B,D,F,H) of wild-type (WT) and mutant mice at E18.5. C,D: ColVa1+/− mice show an increase in ColI expression throughout the septal mv and Ao leaflets compared with WT mice, with expression primarily in the core of the leaflet (arrow) (A,B). E–H: In comparison to ColVa1+/−, ColI expression is further expanded throughout the heart valves of ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice. I,J: The staining intensity of ColI immunoreactivity is shown for the mv (I) and Ao (J) valve leaflet of indicated genotypes.

Download figure to PowerPoint

thumbnail image

Figure 5. Increased ColI expression is observed in the ventricular myocardium and developing heart valves of ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice. A–H: Immunohistochemistry of ColI expression in the extracellular matrix (ECM) of the compact myocardium (A,C,E,F) and trabeculae (B,D,F,H) in ColVa1+/−, ColXIa1, and ColVa1+/−;ColXIa1−/− mice. ColI expression is increased throughout the matrix of the compact myocardium and trabeculae of ColVa1+/− (C,D), ColXIa1−/− (E,F), and ColVa1+/−:ColXIa1−/− mice (G,H) hearts, compared with wild-type (WT) hearts that show undetectable levels of ColI expression (A,B). I,J: The staining intensity of ColI immunoreactivity is shown for the compact myocardium (I) and trabeculae (J) of indicated genotypes. A, atrium; V, ventricle.

Download figure to PowerPoint

At higher magnification, ColI expression is observed in the “core” of the mitral valve leaflet of WT mice, with undetectable levels noted on the atrial (A) and ventricular (V) leaflet surface (arrow and arrowhead, Fig. 4A). In comparison to the mitral valve, lower levels of ColI are detected in the aortic heart valve leaflets (Fig. 4B). ColI IF was quantified as a measure of collagen deposition in mitral and aortic valve leaflets of mutant mice relative to WT. Although morphologically indistinguishable from WT, heart valves in ColVa1+/− mice show increased immunoreactivity of ColI in the mitral (3-fold) and aortic (2.5-fold) valves (Fig. 4,C,D,I,J). Expression in the mitral valve is expanded throughout the valve leaflet (arrow, Fig. 4C), and increased ColI immunoreactivity is observed as punctate staining throughout the aortic valve leaflets (arrows, Fig. 4D). The expanded expression of ColI seen in the mitral valve of ColVa1+/− mice is also seen throughout the thickened valve leaflets of ColXIa1−/− mice, with a fourfold increase in immunoreactivity levels (Fig. 4E,I). Expression in the aortic valve of ColXIa1−/− mice is also increased (fivefold), with immunoreactivity observed throughout the valve leaflet and annulus structure (arrow and arrowhead, respectively, Fig. 4F,J). In ColVa1+/−;ColXIa1−/− mice, ColI immunoreactivity in the mitral valve leaflet is increased fourfold, similar to that observed in ColXIa1−/− mice (arrow, Fig. 4G,I). However, in the aortic valve leaflet, ColI immunoreactivity is greater in ColVa1+/−;ColXIa1−/−mice, with a sevenfold increase compared with WT (Fig. 4H,J).

Immunoreactivity and deposition of ColI was similarly examined in the compact layer and trabecular compartments of the ventricular myocardium of mutant mice. WT mice show low levels of ColI expression throughout the ECM of the myocardium (Fig. 5A,B). ColVa1+/− mice display increased levels of ColI (7-fold) throughout the compact layer (Fig. 5C,I) of the myocardium; however, immunoreactivity was unchanged in the trabeculae (Fig. 5D,J). Increased ColI expression (5-fold) is detected uniformly throughout the compact layer of ColXIa1−/− hearts (Fig. 5E,I). However, this increase is significantly less than that observed in ColVa1+/− mice. In the trabeculae of ColXIa1−/− mice, ColI immunoreactivity is increased throughout by 3-fold, although there are notable focal areas of ColI deposition (Fig. 3F,J). Likewise, a 3-fold increase in ColI is noted in the trabeculae of ColVa1+/−;ColXIa1−/− mice. However, a much greater fold increase in ColI immunoreactivity is observed in the compact layer (8-fold), comparable with that seen in ColVa1+/− mice. While ColVa1+/− mice show a significant increase in ColI in the compact layer of the ventricular myocardium, it is not accompanied by obvious defects in chamber morphogenesis. The greatest increase of ColI immunoreactivity in ColXIa1−/− mice was evident in the mitral valve leaflets. Interestingly, ColVa1+/−;ColXIa1−/− mice show significant increases in ColI deposition in both these affected cardiac structures compared with WT; however, there is no significant accumulative increase compared with single-mutant mice. Together, these studies have identified requirements for collagens Va1 and XIa1 in morphogenesis of the ventricular myocardium and heart valves and show that type I collagen deposition is increased in these cardiac structures in mice with reduced ColVa1 and/or ColXIa1.

ColIII Immunoreactivity Is Increased in Hearts From ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− Mutant Mice

Previous studies have described the importance of ColI to ColIII ratios for mechanical properties of cardiac tissue (Badenhorst et al.,2003). Therefore, the fold change of immunoreactivity for ColIII was measured in the mitral valve leaflet and myocardium of ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice over wild-type controls. ColIII expression was detected at low levels in the heart valves and myocardium of wild-type mice (arrow, Fig. 6A,B), although expression was observed in the epicardium and great vessels (data not shown). ColIII immunoreactivity was significantly increased in ColVa1+/− mice, both in the heart valves (13-fold increase) and myocardium (4-fold increase) compared with wild-type (Fig. 6C,D,I,J). There was less of a significant increase in ColIII immunoreactivity in ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice. In the mitral valve of these mice, immunoreactivity was significantly increased compared with wild-type (arrows, Fig. 6E,G,I), whereas there was no significant difference in the myocardium (arrows, Fig. 6F,H,J). These data show a large increase in ColIII immunoreactivity in cardiac tissue from ColVa1+/− mice and a smaller increase in valve tissue of ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice. Therefore, these findings suggest that the different relative changes in ColIII and ColI immunoreactivity observed between genotypes (Figs. 5, 6) lead to disrupted ColI to ColIII ratios for mice with altered expression levels of ColVa1 and/or ColXIa1.

thumbnail image

Figure 6. Increased ColIII immunoreactivity is observed in the heart valves of ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice and in the myocardium of ColVa1+/− mice. A–H: Immunohistochemistry of ColIII expression in the ECM of the mitral valve leaflet (A,C,E,G) and myocardium (B,D,F,H) in wild-type (A,B), ColVa1+/− (C,D), ColXIa1 (E,F), and ColVa1+/−;ColXIa1−/− (G,H) mice. Type III collagen expression is increased in the heart valve leaflets of ColVa1+/− (C), ColXIa1−/− (E), and ColVa1+/−:ColXIa1−/− mice (G) hearts, and the myocardium of ColVa1+/− mice (D) compared with WT hearts that show low levels of ColIII expression (A,B). I,J: The staining intensity of ColIII immunoreactivity is shown for the mitral valve (I) and myocardium (J) of indicated genotypes. A, atrium; V, ventricle.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

In connective tissues of the skin, cartilage, and tendon, highly organized networks of fibrillar collagens are a major component of the ECM (Bosman and Stamenkovic,2003). Aberrations in the assembly, composition, and organization of these collagenous scaffolds are common in many forms of connective tissue disorders and are detrimental to the structure and function of affected tissues. However, cardiac phenotypes in mouse models of such disorders have previously been unappreciated. This study has identified defects in heart valve development and ventricular chamber morphogenesis in mice that recapitulate EDS and chondrodysplasia diseases. Histologically, hearts from viable ColVa1+/− mice are apparently normal, in contrast to striking changes in the ventricular myocardium of ColXIa1−/− mice. Compared with WT and ColVa1+/− mice, hearts from ColXIa1−/− mice display a loss of apex definition and an increase in IVS thickness, leading to an overall spherical shape of the ventricles. Thickening of the heart valve leaflets was also observed. Defects in the shape of the ventricular myocardium persist in ColVa1+/−;ColXIa1−/− mice; however, there is further significant thickening of the heart valve leaflets compared with ColXIa1−/− and WT mice. Collectively, these data identify previously unreported roles for ColVa1 and ColXIa1 in contributing to the fibrous matrix required for normal heart development.

At the molecular level, hearts from ColVa1+/− mice exhibit a significant increase in ColI immunoreactivity as a measure of collagen deposition, notably in the compact layer of the myocardium compared with WT controls. This finding is in contrast to the prominent increase in ColI deposition in the thickened valve leaflets of ColXIa1−/− mice. Increased ColI deposition in ColVa1+/−;ColXIa1−/− hearts is observed in both the myocardium and valves, although quantitative levels are not significantly different from single-mutant mice. The ratio of ColI to ColIII has been previously shown to be important for determining the mechanical integrity of cardiac tissue (Badenhorst et al.,2003). In general, higher levels of ColI compared with ColIII is associated with tissue stiffness, whereas greater levels of ColIII promote tissue elasticity (Iimoto et al.,1988; Kato et al.,1995; Badenhorst et al.,2003). In this study, the relative immunoreactivity levels of ColI and ColIII were different in the cardiac tissues of ColVa1+/−, ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice, which would lead to altered ratios of ColI to ColIII, depending on the genotype. Notably, highest levels of ColIII relative to ColI were observed in ColVa1+/− mice, whereas increased ColI relative to ColIII was observed in ColXIa1−/− mice. The observed alterations in ColI to ColIII ratios may be indicative of compromised mechanical compliance of cardiac tissue in these mutant mice relative to the heterozygous loss of ColVa1 or the complete loss of ColXIa1. These differences in ECM structure and function may underlie various types of connective tissue disease.

In the ventricular myocardium, the fibrous network of ColI and other minor collagens, including types V and XI, maintains tissue structure, contributes to the visco-elastic properties of the myocardium, and transmits forces during the cardiac cycle (Jane-Lise et al.,2000; Baudino et al.,2006). During cardiac failure, interstitial ColI deposition is increased, the ECM organization is diminished, and often the ventricle appears spherical in shape, a feature associated with ventricular remodeling (Mann and Spinale,1998; Jane-Lise et al.,2000; Diez et al.,2005; Mann,2005; Miner and Miller,2006). The resulting myocardial fibrosis is a hallmark feature of cardiac disease, resulting from a variety of primary lesions (Jane-Lise et al.,2000). Cardiac fibrosis affects the mechanical compliance of the myocardium and further contributes to compromised heart function (Miner and Miller,2006). This study has demonstrated a significant increase in ColI in the compact layer of the ventricular myocardium in hearts from ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice. Increased deposition of ColI in the heart of these mutant mice is likely compensatory, due to reduced amounts of ColVa1 or ColXIa1. It has been previously shown in the skin that excess ColI forms abnormal aggregates through unregulated self-assembly. Therefore, it is likely that both abnormal and normal collagen fibers are present in the cardiac tissue of ColVa1 and ColXIa1 mutant mice (Eyre and Wu,1987; Wenstrup et al.,2004,2006). Higher levels of ColI compensation in the compact layer were observed in hearts from ColVa1+/− mice, compared with ColXIa1−/− hearts. This finding may be explained by the expression of ColVa1, but not ColXIa1, in the compact myocardium. Because ColXIa1 is not expressed in this region, the morphological defects observed in the ventricular myocardium of ColXIa1−/− mice are likely secondary to valve malformations (see below). The complete loss of ColVa1 in late fetal stages of ventricular myocardial morphogenesis cannot be assessed due to premature embryonic lethality of null mice (Wenstrup et al.,2004). However, even the heterozygous loss of ColVa1 fibers in the ventricular myocardium leads to abnormal collagen fiber distribution, consistent with increased ColI fiber diameter and unregulated assembly observed by ultrastructural analysis in other connective tissue systems (Wenstrup et al.,2004). Defining the mechanisms of myocardial fibrosis in these mutant mice will contribute to our understanding of this process and, in addition, may provide insights into potential cardiac dysfunction in ColVa1-related connective tissue disorders.

The organization and stratification of defined ECM within the heart valve leaflets and supporting structures begins during embryonic development, continues postnatally, and is required for efficient valve function throughout life (Schoen,2005; Hinton et al.,2006; Lincoln et al.,2006b). Malformed or malfunctioning diseased valves display aberrations in collagen fiber composition and organization (Rabkin-Aikawa et al.,2005; Lincoln et al.,2006b). Mitral valve prolapse (MVP) is a genetic or acquired valvular connective tissue disorder, where the valve leaflet is abnormally thickened or myxomatous with increased ECM, including ColI (Rabkin-Aikawa et al.,2005; Schoen,2005). The ECM is characteristically unorganized, causing tissue weakness, and often leading to progressive heart failure (Rabkin-Aikawa et al.,2005; Schoen,2005; Hinton et al.,2006). There is increasing evidence showing a correlation between connective tissue disorders and MVP, notably in patients with EDS and Stickler syndrome; however, the valve pathogenesis and ECM organization has not been assessed (Dolan et al.,1997; Snead and Yates,1999; Ahmad et al.,2003; Seve et al.,2005; McDonnell et al.,2006). ColXIa1−/− mice display similarly thickened valve leaflets that express high levels of ColI compared with ColVa1+/− mice. It is anticipated, based on previous reports of matrix defects in chondrodysplasic models, that the ECM of these valves is lacking organization (Seegmiller and Monson,1982; Li et al.,1995). Collectively, this study has shown that heart valves from mice that show features of EDS and chondrodysplasias are thickened, with increased ColI that resemble phenotypes of diseased valves. Therefore, understanding the mechanisms for changes in ECM homeostasis and organization should contribute to our understanding of valve disease in human connective tissue disorders.

Based on findings in skin, cornea, cartilage, and tendons from mutant mice models with structurally abnormal collagen fibers, it has been proposed that minor collagen types V and XI play regulatory roles in fiber formation and assembly in these tissues (Seegmiller and Monson,1982; Andrikopoulos et al.,1995; Li et al.,1995; Wenstrup et al.,2004; Segev et al.,2006; Wenstrup, unpublished observations). ColVa1 and ColXa1 are expressed in the heart valves and myocardium, yet their roles in collagen fibril formation and regulation have not been defined (Yoshioka et al.,1995; Jane-Lise et al.,2000; MacKenna et al.,2000). This study has demonstrated that changes in ECM homeostasis, as indicated by increased ColI immunoreactivity, occur in the myocardium and heart valves of mutant mice with reduced ColVa1 or ColXIa1. It is proposed that this increase in type I collagen deposition is a compensatory mechanism due to the reduction of ColVa1 and ColXIa1 fiber formation. These changes in matrix homeostasis in the myocardium and heart valves of ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice, respectively, resemble ECM pathology and are consistent with previously described ECM dynamics in myocardial fibrosis and myxomatous MVP. The utilization of ColI within the fibrous matrix of affected cardiac structures in mutant mice has yet to be determined, but it is suggested that ECM organization in the myocardium and heart valves is similarly diminished, as previously described in the skin and cartilage. Should ColV and ColXI play similar roles in collagen fiber assembly in the heart as they do in the skin, cartilage, and tendon, it is likely that the reduction in ColV or ColXI and increase in ColI in the ventricular myocardium and heart valves will lead to changes in tissue compliance and ultimately dysfunction (Seegmiller and Monson,1982; Andrikopoulos et al.,1995; Li et al.,1995; Wenstrup et al.,2004; Segev et al.,2006; Wenstrup, unpublished observations). Connective tissue disorders have been associated with vascular abnormalities, and it is possible that altered collagen ratios within the vasculature could also affect compliance and resistance of the heart (Wenstrup et al.,2004). However, implications of collagen dysregulation and dysfunction in the heart itself have not been fully appreciated. The pathogenesis of myocardial and heart valve ECM abnormalities in connective tissue disease in relation to aberrant collagen fiber dynamics and matrix disorganization has yet to be determined. This study demonstrates that loss of specific minor fibrillar collagen components can lead to overall changes in collagen deposition that may be related to cardiac dysfunction associated with connective tissue disorders.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

ColVa1+/−, ColXIa1−/−, and ColVa1+/−;ColXIa1−/− mice

Gene targeting and genotyping procedures for generating ColVa1+/− mice have been previously described (Wenstrup et al.,2004). Mice deficient for ColXIa1 were a kind gift from Dr. R. Seegmiller. Genotyping of ColXIa1−/− mice was carried out as previously reported with reported modifications (Li et al.,1995). Genomic DNA was analyzed for a 1 nucleotide (nt) deletion in ColXIa1 using TaqMan single nt polymorphism assay (Applied Biosystems, Foster City, CA) and an ABI PRISM sequence detection system using primers 5′-AGTGTGGAGAAGAAAACTGTGACAA-3′ and 5′-CCATGATTCCATTGGTATCGACTATCG-3′. ColVa1+/− and ColXIa1+/− heterozygous mice were intercrossed to generate cohorts of embryos with mixed genotypes at expected Mendelian ratios (n = 12 litters). All animal procedures were approved and performed in accordance with institutional guidelines. Timed matings were established with the morning of a copulation plug defined as E0.5. Due to the premature death of ColXIa1−/− and ColVa1+/−;ColXIa1−/− mice at birth, embryos were harvested at E18.5. At the time of dissection, embryonic yolk sacs were collected for genotyping analysis, and embryos were dissected in 1× phosphate buffered saline (PBS), fixed whole overnight in 4% paraformaldehyde/PBS, and subjected to cytochemistry, immunohistochemistry, or in situ hybridization.

Immunostaining

Hearts were dissected from fixed E18.5 embryos and processed for paraffin embedding as previously described (Lincoln et al.,2004). Eight-micrometer sections were cut, mounted onto superfrost slides (Fisher), and subjected to immunocytochemistry or immunohistochemistry. After deparaffinization and hydration through a graded ethanol series (100%, 95%, 75%, 50%, 25%), slides were subjected to either Movat's Pentachrome cytochemistry or immunohistochemistry for ColVa1, ColI, or ColIII. Pentachrome staining was performed as previously described (Jones,2001), visualized with an Olympus BX60 microscope, and captured using Advanced SPOT image software. To minimize nonspecific binding of the rabbit polyclonal antibodies directed against mouse ColI, ColIII, and ColVa1, tissue sections were incubated with blocking solution (1% bovine serum albumin, 0.1% cold water fish skin gelatin, 0.1% Tween-20, 0.05% NaN3/PBS) for 1 hr at room temperature. ColI (R&D Systems, 1:500), Col III (Cosmo Bio. Co. Ltd., Japan), and ColVa1 ((Wenstrup et al.,2004, #1), 1:800) antibodies were diluted in blocking solution and 150 μl was incubated on tissue sections using Coverwell chamber slides (Grace Biolabs) overnight at 4°C. After incubation, sections were washed in 1× PBS and incubated with Alexa-goat anti-rabbit-488 secondary antibody (Molecular probes, 1:100/PBS) and TO-PRO-3 (Molecular probes, 1:1,000) for 1 hr at room temperature. Sections were washed thoroughly in 1× PBS and mounted in hard set Vectashield (Vector Labs). Fluorescent images were analyzed using the Nikon PCM2000 confocal microscope and Simple PCI software. For each experimental set, fluorescent images were captured in parallel using identical confocal lasers and constant PMT filters and integration levels. Fluorescent staining intensity after immunohistochemistry of ColI and ColIII was measured in six microscopic fields from three independent embryos and the analyzed based on pixel brightness using Image J software (NIH, version 1.24). For each independent set of experiments, immunoreactivity was measured in mutant mice and recorded as a fold change over WT littermates. Statistical significance of observed differences was carried out using two-tailed Student's t-test and compared with WT animals (P < 0.01).

In Situ Hybridization

The mouse ColXIa1 sequence (Genbank accession no. NM_007729) was amplified from cDNA generated from E18.5 flexor digitorum longus tendon cultures by reverse transcriptase polymerase chain reaction using specific primers for amplification: forward 5′-ATTTTTATGAATACAAAGAATAT-3; and reverse, 5′-CCCTCATTCCATCATCGCCAGGG-3′ and resulting fragments were confirmed by sequencing. The 756 base pair (bp) fragment was subcloned into pGEM-T vector (Promega) and digoxigenin UTP-labeled antisense riboprobe was synthesized with Sp6 polymerase from plasmid linearized with SacII. In situ hybridization of cryosectioned tissue was carried out as previously described (Lincoln et al.,2006a), and images were captured using Olympus BX60 microscope and Advanced SPOT image software.

Morphometric Analysis

The thickness of the IVS and atrioventricular (AV) valve leaflets was measured on photomicrographs using Image J software. The fold change in IVS thickness was calculated from the average measurements from four independent experiments, in which 12 tissue sections were analyzed for each genotype. AV valve thickness of each heart was determined from an average of measurements taken at the base, middle, and tip of the tricuspid and mitral valve septal leaflets. An overall average was calculated from 16 different heart sections from 4 independent individuals. Statistical significance of observed differences was determined using two-tailed Student's t-test of values for mutants vs. wild-type animals (P < 0.05).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Hanna Osinska, Michelle Sargent, and members of the Division of Molecular Cardiovascular Biology for technical support. J.L. was funded by an AHA Ohio Valley Affiliate Postdoctoral Fellowship Award with additional funding from (K.E.Y.) SCCOR Pediatric Heart Development and Disease Award.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES