**7. The basement membrane supports valve endothelial cells and acts as a barrier between circulating blood and subendothelial components**

The basement membrane is a myriad of proteins, proteoglycans, and glycoproteins that not only supply a substrate to anchor the valve endothelial cells, but also has a large array of biological activities that regulate spatial organization, sequester growth factors, modulate angiogenesis and migration, and regulate the diffusion of nutrients through it towards the underlying valve interstitial cells [93]. The major constituents of the basement membrane are laminin, perlecan, collagen type IV and VIII, nidogen, and the glycoprotein SPARC (secreted protein, acidic and rich in cysteine). Each of these constituents play a role in the overall func‐ tion of the basement membrane. In addition, MMPs contribute to the biological activity that occurs within the basement membrane. Understanding basement membrane composition and behavior, during both healthy and diseased states of the aortic valve, may lead to a bet‐ ter understanding of calcific aortic valve disease.

Osteopontin is a phosphoprotein, meaning that it contains chemically bound phosphoric acid. Originally found in bone, it also contains the arginine-glycine-aspartate (RGD) motif more commonly attributed to fibronectin and is also a constituent of ECM in other tissues [82]. It is secreted by various tissues such as fibroblasts [82] and immune cells, including dendritic cells, macrophages, and neutrophils [83]. Osteopontin is known to interact with various surface receptors that make it a crucial player in bone remodeling [84], wound heal‐ ing, inflammation, and immune responses [83]. It is also known to be involved in vascular remodeling during endothelial injury [82]. Osteopontin is present in valves calcified as a re‐ sult of disease as well as in calcified bioprosthetic heart valves [85]. The calcification process of aortic valves closely resembles osteoblast differentiation in regards to expression of genes

Osteocalcin is a small, non-collagenous protein that is considered a late-stage marker for bone formation and is one of a small group of proteins that are osteoblast-specific [87,88]. It is present in general circulation [87] and its capacity for binding hydroxyapatite and calcium suggests that it is largely involved in mineral deposition [88], but it also has recently been shown to act in a hormone-like manner by enhancing insulin secretion [87]. Its traditional role as a product of bone indicates that valve calcification may actually be a result of active bone formation in the valve tissue [86]. This bone formation may be the result of VEGF se‐ cretion by endothelial cells during neoangiogenesis occurring in response to inflammation, as seen in rheumatic valve calcification [89]. Additionally, increased serum levels of osteo‐

In addition to the matrix proteins, matrix metalloproteinases (MMPs) and their inhibitors (TIMPs) are also found in heart valves. They assist in tissue development and remodeling and can be used as indicators of disease. It is also believed that the ECM degradation result‐ ing from MMP activity serves to release growth factors bound to ECM components and thus alter the microenvironment chemically, as well as structurally [91]. Calcified leaflets from stenotic valves have been shown to express levels of MMP-2 that are similar to those of nor‐ mal valves but express higher levels of MMP-3, MMP-9, and TIMP-1 [14]. MMP-1, produced by activated myofibroblasts and macrophages, is also prevalent in calcific aortic valve steno‐

**7. The basement membrane supports valve endothelial cells and acts as a**

The basement membrane is a myriad of proteins, proteoglycans, and glycoproteins that not only supply a substrate to anchor the valve endothelial cells, but also has a large array of biological activities that regulate spatial organization, sequester growth factors, modulate angiogenesis and migration, and regulate the diffusion of nutrients through it towards the underlying valve interstitial cells [93]. The major constituents of the basement membrane are laminin, perlecan, collagen type IV and VIII, nidogen, and the glycoprotein SPARC (secreted protein, acidic and rich in cysteine). Each of these constituents play a role in the overall func‐

**barrier between circulating blood and subendothelial components**

characteristic of bone formation, such as osteopontin and osteocalcin [86].

16 Calcific Aortic Valve Disease

calcin were shown to be indicative of aortic valve disease in patients [90].

sis and may be related to high TNF-α levels resulting from inflammation [92].

Laminins belong to a family of heterotrimeric glycoproteins composed of combinations of α, β, and γ chains that form a cross-like structure averaging between 400-900 kDa in size [94,95]. Laminins play an integral role in the formation of the supportive ECM network. The unique cross-like shape allows laminin molecules to bind with neighboring laminins and ECM via the three short chains, and use the long alpha chain as a cell anchoring site [96]. In addition to their structural contributions to the basement membrane, laminins are essential for proper biological activity. These glycoproteins have been shown to promote cell adhe‐ sion, migration, differentiation, and maintenance of cellular phenotype [94,97,98]. Dysfunc‐ tion in laminin expression has been linked to diseases with improper tissue formation such as muscular dystrophy, epidermolysis bullosa, and various nephritic syndromes [94,99].

Although laminin is not as ubiquitous as collagen, this basement membrane component has been highly investigated as an ECM substrate for *in vitro* cultures. However, this glycopro‐ tein may influence valve cell types differently. *In vivo* and *in vitro* studies have shown that laminin interacts with endothelial and epithelial cells, and can help maintain physiological functionality of the cells [97,100,101]. However, VICs cultured on laminin have been found to support high quantities of calcific nodule formation in the presence of TGF-β, when com‐ pared to subendothelial ECM components collagen type I and fibronectin [19,102]. The vari‐ ous regions of laminin protein have been reported to mediate specific cell responses. The Gdomains of laminin α chains are associated with heparin binding and cell adhesion, whereas regions along the laminin β chains promote cell differentiation [98,100,101]. The peptide se‐ quence YIGSR from the laminin β-1 chain has been shown to promote endothelial cell adhe‐ sion and proliferation, however, it also influences other cell types including smooth muscle and tumor cells [98,101]. VICs cultured on YIGSR were also shown to promote calcific nod‐ ule formation, although less than those seeded on fibronectin derived RGDS peptides. How‐ ever, when the 67-kDa laminin cell receptor was blocked, the YIGSR seeded VIC cultures significantly increased in nodule formation and gene expression for various myogenic and osteogenic markers, suggesting that disruption in laminin binding may be linked to valve calcification [103]. IKVAV, another peptide sequence derived from the laminin α1-chain, has been linked to promoting angiogenesis, cell migration and spreading [97,98,104]. Though most work with this peptide has been done with endothelial and tumor cells, its ability to promote angiogenesis may also be a future area of interest in studying how angiogenesis mediates valve tissue calcification. Furthermore, laminin influence on cell activity varies be‐ tween cell types, and may promote VIC activation and tissue calcification in diseased states.

Perlecan (Pln) is one of the more abundant heparan sulfate proteoglycans and is found in several tissues including in the endochondral barrier in bones [105]; however, it is primarily localized in vascular basement membranes. It has a major role in regulating the develop‐ ment of blood vessels, the heart, cartilage, and the nervous system. Physiologically, perlecan plays a prominent role in regulating cellular proliferation, differentiation, organization, and mediating inflammation [106]. Perlecan derives its functionality from five protein subdo‐ mains which share their sequence homology with several other proteins [107]. Domain 1 contains an SEA (Sperm protein, Enteokinase, Agrin) module and three SGD (Ser-Gly-Asp) tripeptide sequences to which three heparan sulfate (HS) glycosaminoglycans attach. These HS can bind and sequester several important growth factors for determining endothelial quiescence in a process known as matricrine signaling. The SEA section is unique to perle‐ can, and it has no known function other than to influence the O-linked glycosylation of the SGD domain. Interestingly, it has been shown that several factors that determine the activity of these sugar chains vary greatly by the cell source that is producing them [108]. These fac‐ tors can include the ratio of heparan sulfate to chondroitin sulfate, the length of the chains, and the sulfation level of the chains which all affect how the chains modulate the bioactivity of nearby growth factors. Domain II contains 4 low-density lipoprotein receptor sequences and one immunoglobulin-like repeat. Domain III contains three laminin-like domain mod‐ ules and eight epidermal growth factor-like repeats. Domain IV, the largest domain, con‐ tains many N-CAM-like Ig repeats. Domain V has been demonstrated to be the major cellbinding domain of perlecan due to the laminin and agrin homologies that it contains. Domain V can also be glyocosylated, which can contribute along with domain I to the matri‐ crine signaling capabilities of perlecan, which could potentially contribute to the develop‐ ment of CAVD [109].

promote neutrophil binding [115,118]. Collagen IV networks are highly adhesive to all cells types except erythrocytes [115,119]. Furthermore, cell binding has been found to be en‐ hanced in the presence of various ECM molecules such as perlecan, SPARC, and von Wille‐ brand factor (vWF) [115,118,119]. Interestingly, COL IV also has numerous anti-angiogenic domains that are activated after MMP degradation at the non-collagenous (NC) 1 domain, thereby limiting angiogenesis or migration of endothelial and tumor cells [115]. The changes in COL IV bioactivity depending on the domain region and cleavage state can greatly affect the functionality of surrounding cells. Dysfunctional COL IV expression or mutations in the heterotrimer formation have been found to be extremely detrimental and cause matrix dis‐ orders such as Goodpasture's syndrome or Alport syndrome [94,115]. Therefore, additional studies should be done to investigate how the highly bioactive COL IV meshworks may

Extracellular Matrix Organization, Structure, and Function

http://dx.doi.org/10.5772/52842

19

COL VIII has also been found to play a network forming role, maintaining the sheet-like structure ECM, while sequestering various integrin binding sites and growth factors. COL VIII is smaller than COL IV, and can form tetrahedral and hexagonal assemblies [117,118]. Though work on COL VIII in regards to valve tissues has been limited, vascular basement membrane studies have found that COL VIII plays a large role in interacting with subendo‐ thelial cells such as smooth muscle cells and fibroblasts. In vitro, COL VIII promotes fibro‐ blast proliferation and migration [114]. Furthermore, COL VIII may be linked to atherogenesis, a pathology similar to CAVD, as its expression in cells is upregulated during vessel injury [114,120]. This collagen has even been found to interact with elastic fibers in liver tissues, suggesting it may have a bridging function between the basement membrane components and subendothelial ECM [118]. Therefore, COL VIII could play an integral role in mediating valve interstitial and endothelial cell communication. Recent studies have found after enzyme cleavage at the NC1 domain, the resulting C-terminal fragment known as vastatin will prevent endothelial cell proliferation and induce cell apoptosis [120]. While some work has investigated using vastatin as an anti-angiogenic agent, further studies are needed to elucidate how it may affect the functionality of surrounding cells and ECM, espe‐

Similar to perlecan, nidogen is a 150 kDa glycoprotein that has sequence homologies with other basement membrane proteins. It consists of two amino (G1, G2) and one carboxyl (G3) terminal globular domains that are connected by a rod domain composed primarily of endo‐ thelial growth factor repeats [121]. Nidogen binds collagen type IV, perlecan, and laminin. This binding contributes to the hypothesis that nidogen is important in basement membrane assembly; although some recent animal studies have demonstrated that nidogen may not be necessary for basement membrane formation [121]. The role of nidogen in CAVD is unex‐ plored, but it may play a role in maintaining valvular basement membrane functionality by

SPARC positive neovascularisation is a documented histological change in CAVD [122]. Se‐ creted protein acidic and rich in cysteine (SPARC), also known as osteonectin, is a small basement membrane protein. It interacts with cells, binds to other members of the basement membrane, growth factors, various proteases, and is found in newly developing neovessels.

promote the onset of calcification in valve tissues.

regulating infiltration of inflammatory agents [93].

cially in older valve tissue.

Matricrine signaling occurs when the ECM modulates cell behavior by controlling the local levels of growth factor concentrations by sequestering or releasing them when the underly‐ ing matrix is intact or degraded, respectively [109]. Proteoglycans, like perlecan, and their GAG chains are the major sites for matricrine signaling due to their heparan sulfate and chondroitin sulfate chains electrostatically binding free growth factors. Their role in the pathology of CAVD is widely unexplored despite their presence in normal valves and their increased production in diseased valves [67]. It is known that PGs and GAGs play an inte‐ gral role in the progression of atherosclerosis via sequestering of inflammatory molecules and lipids [110–113] and mediating angiogenesis into the vessel supplying an entry way for additional inflammatory entities. Both of these factors are seen histologically in CAVD, but their role is merely speculative at this moment.

Collagens in the basement membrane can form lateral, axial, and linear connections with surrounding ECM. Of the basement membrane collagens, collagen IV (COL IV) is the most abundant and essential for network formation. Only found in basement membrane tissues, COL IV molecules are approximately 400 kDa, and composed of two α1, and one α<sup>2</sup> [115– 117]. COL IV proteins have many biologically active domains that can influence specific cel‐ lular responses, as well as have specific affinities to other molecules such as BMP-4, fibro‐ nectin, Von Hippel Lindau protein, and factor IX [115,117]. Mapping of COL IV protein reveals 3 major integrin motifs that are located in strategic regions to promote cell activity or protein degradation when activated [115]. During angiogenesis and tumor invasion, COL IV is degraded by MMP-2 and MMP-9 enzymes to allow for cell migration and infiltration into the matrix. Studies have found that the cleavage sites also overlap with many integrin bind‐ ing domains such as α1β1, resulting in the availability of αvβ3 integrin binding sites known to

promote neutrophil binding [115,118]. Collagen IV networks are highly adhesive to all cells types except erythrocytes [115,119]. Furthermore, cell binding has been found to be en‐ hanced in the presence of various ECM molecules such as perlecan, SPARC, and von Wille‐ brand factor (vWF) [115,118,119]. Interestingly, COL IV also has numerous anti-angiogenic domains that are activated after MMP degradation at the non-collagenous (NC) 1 domain, thereby limiting angiogenesis or migration of endothelial and tumor cells [115]. The changes in COL IV bioactivity depending on the domain region and cleavage state can greatly affect the functionality of surrounding cells. Dysfunctional COL IV expression or mutations in the heterotrimer formation have been found to be extremely detrimental and cause matrix dis‐ orders such as Goodpasture's syndrome or Alport syndrome [94,115]. Therefore, additional studies should be done to investigate how the highly bioactive COL IV meshworks may promote the onset of calcification in valve tissues.

mediating inflammation [106]. Perlecan derives its functionality from five protein subdo‐ mains which share their sequence homology with several other proteins [107]. Domain 1 contains an SEA (Sperm protein, Enteokinase, Agrin) module and three SGD (Ser-Gly-Asp) tripeptide sequences to which three heparan sulfate (HS) glycosaminoglycans attach. These HS can bind and sequester several important growth factors for determining endothelial quiescence in a process known as matricrine signaling. The SEA section is unique to perle‐ can, and it has no known function other than to influence the O-linked glycosylation of the SGD domain. Interestingly, it has been shown that several factors that determine the activity of these sugar chains vary greatly by the cell source that is producing them [108]. These fac‐ tors can include the ratio of heparan sulfate to chondroitin sulfate, the length of the chains, and the sulfation level of the chains which all affect how the chains modulate the bioactivity of nearby growth factors. Domain II contains 4 low-density lipoprotein receptor sequences and one immunoglobulin-like repeat. Domain III contains three laminin-like domain mod‐ ules and eight epidermal growth factor-like repeats. Domain IV, the largest domain, con‐ tains many N-CAM-like Ig repeats. Domain V has been demonstrated to be the major cellbinding domain of perlecan due to the laminin and agrin homologies that it contains. Domain V can also be glyocosylated, which can contribute along with domain I to the matri‐ crine signaling capabilities of perlecan, which could potentially contribute to the develop‐

Matricrine signaling occurs when the ECM modulates cell behavior by controlling the local levels of growth factor concentrations by sequestering or releasing them when the underly‐ ing matrix is intact or degraded, respectively [109]. Proteoglycans, like perlecan, and their GAG chains are the major sites for matricrine signaling due to their heparan sulfate and chondroitin sulfate chains electrostatically binding free growth factors. Their role in the pathology of CAVD is widely unexplored despite their presence in normal valves and their increased production in diseased valves [67]. It is known that PGs and GAGs play an inte‐ gral role in the progression of atherosclerosis via sequestering of inflammatory molecules and lipids [110–113] and mediating angiogenesis into the vessel supplying an entry way for additional inflammatory entities. Both of these factors are seen histologically in CAVD, but

Collagens in the basement membrane can form lateral, axial, and linear connections with surrounding ECM. Of the basement membrane collagens, collagen IV (COL IV) is the most abundant and essential for network formation. Only found in basement membrane tissues, COL IV molecules are approximately 400 kDa, and composed of two α1, and one α<sup>2</sup> [115– 117]. COL IV proteins have many biologically active domains that can influence specific cel‐ lular responses, as well as have specific affinities to other molecules such as BMP-4, fibro‐ nectin, Von Hippel Lindau protein, and factor IX [115,117]. Mapping of COL IV protein reveals 3 major integrin motifs that are located in strategic regions to promote cell activity or protein degradation when activated [115]. During angiogenesis and tumor invasion, COL IV is degraded by MMP-2 and MMP-9 enzymes to allow for cell migration and infiltration into the matrix. Studies have found that the cleavage sites also overlap with many integrin bind‐ ing domains such as α1β1, resulting in the availability of αvβ3 integrin binding sites known to

ment of CAVD [109].

18 Calcific Aortic Valve Disease

their role is merely speculative at this moment.

COL VIII has also been found to play a network forming role, maintaining the sheet-like structure ECM, while sequestering various integrin binding sites and growth factors. COL VIII is smaller than COL IV, and can form tetrahedral and hexagonal assemblies [117,118]. Though work on COL VIII in regards to valve tissues has been limited, vascular basement membrane studies have found that COL VIII plays a large role in interacting with subendo‐ thelial cells such as smooth muscle cells and fibroblasts. In vitro, COL VIII promotes fibro‐ blast proliferation and migration [114]. Furthermore, COL VIII may be linked to atherogenesis, a pathology similar to CAVD, as its expression in cells is upregulated during vessel injury [114,120]. This collagen has even been found to interact with elastic fibers in liver tissues, suggesting it may have a bridging function between the basement membrane components and subendothelial ECM [118]. Therefore, COL VIII could play an integral role in mediating valve interstitial and endothelial cell communication. Recent studies have found after enzyme cleavage at the NC1 domain, the resulting C-terminal fragment known as vastatin will prevent endothelial cell proliferation and induce cell apoptosis [120]. While some work has investigated using vastatin as an anti-angiogenic agent, further studies are needed to elucidate how it may affect the functionality of surrounding cells and ECM, espe‐ cially in older valve tissue.

Similar to perlecan, nidogen is a 150 kDa glycoprotein that has sequence homologies with other basement membrane proteins. It consists of two amino (G1, G2) and one carboxyl (G3) terminal globular domains that are connected by a rod domain composed primarily of endo‐ thelial growth factor repeats [121]. Nidogen binds collagen type IV, perlecan, and laminin. This binding contributes to the hypothesis that nidogen is important in basement membrane assembly; although some recent animal studies have demonstrated that nidogen may not be necessary for basement membrane formation [121]. The role of nidogen in CAVD is unex‐ plored, but it may play a role in maintaining valvular basement membrane functionality by regulating infiltration of inflammatory agents [93].

SPARC positive neovascularisation is a documented histological change in CAVD [122]. Se‐ creted protein acidic and rich in cysteine (SPARC), also known as osteonectin, is a small basement membrane protein. It interacts with cells, binds to other members of the basement membrane, growth factors, various proteases, and is found in newly developing neovessels. Intact SPARC protein inhibits cellular proliferation and has anti-angiogenic activity in vitro [123]. However, enzymatic degradation of SPARC can release matricryptic fragments with the KGHK motif that may induce angiogenic activity both *in vitro* and *in vivo* [124]. SPARC has been observed lining blood vessels in early to mid stage calcified valves suggesting the presence of a fully formed basement membrane lining these vessels [125]. However, the presence of the other constituents of the basement membrane is merely speculative at this point as the studies investigating their presence during CAVD have not been completed.

[4] Vesely I, Lozon A. Natural preload of aortic valve leaflet components during gluta‐ raldehyde fixation: effects on tissue mechanics. *Journal of Biomechanics*. 1993 Feb;26(2):

Extracellular Matrix Organization, Structure, and Function

http://dx.doi.org/10.5772/52842

21

[5] Schoen FJ. Aortic valve structure-function correlations: role of elastic fibers no longer a stretch of the imagination. *Journal of Heart Valve Disease*. 1997 Jan;6(1):1–6.

[6] Scott MJ, Vesely I. Aortic valve cusp microstructure: the role of elastin. *Annals of*

[7] Filion RJ, Ellis CG. A finite difference model of O2 transport in aortic valve cusps: importance of intrinsic microcirculation. *American Journal of Physiology. Heart and Cir‐*

[8] Tseng H, Grande-Allen KJ. Elastic fibers in the aortic valve spongiosa: A fresh per‐ spective on its structure and role in overall tissue function. *Acta Biomaterialia*. 2011

[9] Missirlis YF, Chong M. Aortic valve mechanics–Part I: material properties of natural

[10] Sauren AAHJ, van Hout MC, van Steenhoven AA, Veldpaus FE, Janssen JD. The me‐ chanical properties of porcine aortic valve tissues. *Journal of Biomechanics*. 1983 Jan;

[11] Lee JM, Courtman DW, Boughner DR. The glutaraldehyde-stabilized porcine aortic valve xenograft. I. Tensile viscoelastic properties of the fresh leaflet material. *Journal*

[12] Hinton RB, Lincoln J, Deutsch GH, Osinska H, Manning PB, Benson DW, et al. Ex‐ tracellular matrix remodeling and organization in developing and diseased aortic

[13] Chen WLK, Simmons CA. Lessons from (patho)physiological tissue stiffness and their implications for drug screening, drug delivery and regenerative medicine. *Ad‐*

[14] Fondard O, Detaint D, Iung B, Choqueux C, Adle-Biassette H, Jarraya M, et al. Ex‐ tracellular matrix remodelling in human aortic valve disease: the role of matrix met‐ alloproteinases and their tissue inhibitors. *European Heart Journal*. 2005 Jul;26(13):

[15] Edep ME, Shirani J, Wolf P, Brown DL. Matrix metalloproteinase expression in non‐

[16] Aikawa E, Aikawa M, Libby P, Figueiredo J-L, Rusanescu G, Iwamoto Y, et al. Arteri‐ al and aortic valve calcification abolished by elastolytic cathepsin S deficiency in

porcine aortic valves. *Journal of Bioengineering*. 1978;2(3-4):287.

*of Biomedical Materials Research*. 1984 Jan;18(1):61–77.

valves. *Circulation Research*. 2006 Jun 9;98(11):1431–8.

*vanced Drug Delivery Reviews*. 2011 Apr 30;63(4-5):269–76.

rheumatic aortic stenosis. *Cardiovascular Pathology*. 9(5):281–6.

chronic renal disease. *Circulation*. 2009 Apr 7;119(13):1785–94.

*Thoracic Surgery*. 1995;60(Fig 2):S391–S394.

*culatory Physiology*. 2003 Nov;285(5):H2099–104.

121–31.

Jan 19;7(5):2101–8.

16(5):327–37.

1333–41.

#### **8. Summary**

In conclusion, the last several years have witnessed significant acceleration in the number of studies characterizing specific types of extracellular matrix in heart valves, although there is still much to be learned. The basement membrane of heart valves, and its role in regulating valvular endothelial cell function, are particularly understudied. The broad scope of cell-ma‐ trix and matrix-matrix interactions within heart valves, and how these are regulated by the local, dynamic signaling environment, is another subject that merits further investigation. We expect that insights gained from these research endeavors will lead to novel treatments for valve diseases in the future.
