ECM in Health1. Main Components of the ECMThe extracellular matrix (ECM) iscomposed of various proteins that give rise to different structures andproperties of the ECM. The main components of the ECM are collagen,proteoglycans, laminin, and fibronectin. The composition of the ECM will differfrom organ to organ, giving rise to ECM that have distinct properties that areappropriate for the organ to function. 1.1 Collagen Collagen is one of the majorproteins of the ECM and is the most abundant in humans. It has been widelystudied with 28 different collagen types discovered.
1-4 Each type is composed ofhomotrimers or heterotrimers of left handed helical ? chains that are twistedto form a right handed triple helix structure.5There are over 40 distinct ? chains in humans as well as other proteins that containcollagen like domains.2, 5 The collagen superfamily isa large group of proteins that all contain the Gly-X-Y motif, where X and Y areusually either proline or hydroxyproline, of each ? chain is characteristic ofcollagens.5, 6 Despite the large amounts ofbulky proline, the right handed helical structure is stabilized by the smallglycine, interchain hydrogen bonds, and electrostatic interactions involvinglysine and aspartate.
6, 7 Collagens can be categorized accordingto the supramolecular structures that they form. There are fibrillar collagensthat form fibrous structures that can be found in tendons, cartilage, skin, andcornea.2, 3 Each collagen fiber is madeup of several types of collagen, depending on the tissue they are found. Collagentype I is the most abundant fibrillar collagen which is found in connectivetissues except the cartilage including skin, bone, tendon, and cornea.8All fibrillar collagens are firstproduced as precursors.
The ? chains are assembled together in the roughendoplasmic reticulum to form the triple helical structure. This is wherehydroxylation of proline and lysine as well as glycosylation takes place toinitiate the formation of the triple helical structure.9The procollagen is then brought to the golgi apparatus where it is prepared forcellular export. Processing of the procollagen happens either during or aftersecretion in the ECM.10-13 The C terminal propeptide iscleaved off by specific metalloproteinases. It has been shown that failure toremove the C terminal of the collagen leads to high solubility of collagen thatprevents it from forming fibrils.
14For collagen types I, II, and III, the N-propeptides are cleaved off while fortype V, XI, and other fibrillar collagens, the N-propeptides remain. This modifiesthe shape and diameter of the fibril without affecting fibril formation.4,14-16 The N-propeptides of type Vand XI collagens protrude out of the gaps between collagen molecules to preventlateral growth via steric hindrance and charge interactions.14, 15 Type V and XI collagens arecurrently believed to be responsible for nucleating and modulating fibrilformation of collagen. 14, 15 It has been shown that thedeletion of collagen V in mice leads to failure of fibril assembly despite itslow amounts in the total collagen content in most tissues.17Once the microfibrils are formed, thesemay bind with other microfibrils so that they will grow into larger fibers.
This process is mediated by other extracellular matrix proteins.18Small leucine rich proteoglycans (SLRPs) such as decorin and biglycan havecollagen binding motifs allowing them to modulate fiber growth, size,morphology, and content.4, 18,19 Another subfamily ofcollagen are fibril-associated collagens with interrupted helices (FACIT) thatdo not form fibrils themselves but are associated with the surface of collagenmicrofibrils.2Their primary function is to mediate the formation of higher order structurevia binding with other extracellular matrix proteins such as SLRPs andproteoglycans.15, 20 The supramolecular assemblyof collagen is further stabilized by lysyl oxidase, which leads to overallenhanced mechanical properties.
The N terminal and C terminal ends ofindividual collagen molecules are covalently cross linked by lysyl oxidase bothwithin and between microfibers20.Overall, these post translational modifications of collagen contributes to thegreat tensile strength of collagen.21Other than fibrillar and FACITcollagens, there are also network forming collagens such as type IV, VIII, andX. These are found in the basal lamina of basement membranes.2Collagen IV forms a tetramer through their 7S N-terminal domain. Each of thesecollagen IV molecules is bound to another collagen IV molecule via theirC-terminal NC1 domain of each ?-chain, forming a hexamer.
2These two domains of collagen IV allow it to form a stable collagen networkthat separates the basal lamina from the interstitial stroma.22 Other extracellular matrixproteins such as laminin, nidogen, and perlecan can be found in the basallamina that strengthens this barrier to effectively maintain the organizationof the cells in the body.22, 23Although different types of collagen areable to build various types of supramolecular structures that form the basis ofthe architecture of the ECM, the contribution of other ECM proteins such asproteoglycans, laminins, and fibronectin cannot be ignored. They largelyinfluence the chemical and physical properties of the extracellular matrix suchas through their growth factor binding motifs and innate chemical properties.
Furthermore, they also serve as connectors between the cells and the ECM. 1.2 Proteoglycans Proteoglycans are characterized as proteinsthat have glycosaminoglycans (GAGs) covalently bonded to them. These GAGs arelong chains of negatively charged disaccharide repeats that can either beheparin sulphate, chondroitin/dermatan sulphate, hyaluronan, and keratinsulphate. Only hyaluronan is not sulphated. Due to the negative charge of theseGAGs, they are able to sequester water and cations, which gives them theirspace filling and lubrication functions.
24Proteoglycans are classified based on their structure and the distribution anddensity of their GAGs. They can be further classified based on where they arefound.25 For the purpose of thisreview, only transmembrane proteoglycans and those found in the pericellular andextracellular space will be discussed. There are thirteen genes that encodefor cell surface proteoglycans. Seven of these encode for transmembraneproteoglycans, which are four syndecans, CSPG4/NG2, betaglycan, and phosphocan.On the other hand, six of the genes encode for the GPI-anchored proteoglycanscalled glypicans. For the purpose of this review, only syndecans will bediscussed.
The other cell surface proteoglycans are reviewed by Iozzo & et.al..25 Syndecans have anintracellular domain, transmembrane domain, and ectodomain. The GAGs are foundattached to the ectodomain, which are usually heparan sulphates. Thisectodomain can be shed off through the action of MMPs. The ectodomain ofsyndecans is natively disordered, which allows it to interact with a widevariety of molecules giving it a broad range of biological functions. Some ofits functions involve binding to growth factors and morphogens, facilitatingexosome uptake, and being co-receptors of receptor tyrosine kinases.
One of the proteoglycans found inthe pericellular area or the basement membrane is perlecan. Perlecan is a largeheparan sulfate proteoglycan (HSPG) that has multiple domains, each withdifferent binding sites and functions. The heparan sulfate can bind to 3 siteson domain I and 1 site on domain V.26These heparan sulfates can bind to a variety of molecules such as growthfactors, growth factor receptors, collagen, and other ECM proteins.
In thebasement membrane, it binds to collagen IV, nidogen, and laminin, linking allof these proteins which further strengthens the basement lamina.22, 23,27 Proteoglycans found in theextracellular space are hyalectans and SLRPs. All hyalectans have the samestructure in which they have their hyaluronic acid binding N terminal and alectin binding C terminal with GAGs attached in between the N and C terminalends. Hyalectans are encoded by 4 distinct genes: aggrecan, versican, neurocan,and brevican.25 Aggrecan is found mostly inthe bone in the cartilage and brain and neurocan and brevican are found in thecentral nervous system.
On the other hand, versican is found in the ECM ofalmost any tissues and organs.28They can serve as molecular bridges between the cell surface and theextracellular matrix.25 Versican has been shown tobind to collagen type I and fibronectin, which are both substrates of integrins.29The binding of versican to fibronectin’s RGD motif leads to loss of celladhesion as it sequesters fibronectin from the cell’s integrins.28, 29 The largest family of proteoglycansare SLRPs with 18 distinct gene products with multiple splice variants andprocessed forms.
These proteins have a relatively short protein core with acentral region dominated by leucine rich repeats (LRRs). They are expressed inall extracellular matrices even in development suggesting their involvement indirecting organ size and shape during embryonic development and homeostasis. Decorinand biglycan are SLRPs that have collagen binding motifs and regulate collagenfiber assembly along with other proteoglycans. Overall, proteoglycans vary in formand structure that confer different functions in the ECM. They are integral inthe maintenance of a healthy ECM without which would lead to 1.3 Laminin Laminins are trimeric glycoproteinsconsisting of an ?, ?, and ? chains that are often found in the basal lamina orsome mesenchymal compartments.4 The twelve mammalian ?, ?,and ? chains can theoretically create 60 unique laminins but only 16combinations have been observed so far.
23, 30 The ? chains vary in sizefrom 200 and 400 kDa while ? and ? chains have sizes from 120 to 200 kDa. Atrimer can then have a size varying from 400 to 800 kDa.30With a rotary shadowing electron microscopy, laminins look like cross-shapedmolecules.23, 31,32 The three chains form an?-helical coiled coil structure that forms the long arm of the cross while thethree short arms are composed of one chain each.
23At the end of the long arm are 5 laminin G-like (LG) domains from the ? chainthat serve as attachment sites for the cell. Integrins, dystroglycan, Lutheranglycoprotein, or sulfated glycolipids bind to these LG domains.30At the end of each short arm are laminin N-terminal (LN) domains that areimportant for laminin polymerization and basement membrane assembly.
23Laminins have cell type specificfunctions such as adhesion, differentiation, migration, phenotype maintenance,and apoptotic resistance.30Through binding of integrins, laminins are able to create a dynamic linkbetween the cell and the ECM.30Different heterotrimeric laminins will have different integrin heterodimersbinding partners.30This allows the induction of signaling pathways and organization ofintracellular cystoskeleton.30, 33 Although the binding oflaminins to integrin is primarily mediated by the LG domains on the ? chain,the ? and ? chains can also influence binding. It was found that ?2 containinglaminins have higher affinity to integrins ?3?1 and ?7X2?1 when compared to ?2containing laminins in vitro. The difference in binding affinity was attributedto the 20 amino acid residues on the coiled coil domain of the long arm. 34Furthermore, it was found that the glutamic acid residues found on the Cterminal end of ?1 and ?2 chains are necessary for integrin binding.
35The binding of laminin to the cellsurface is essential in laminin polymerization.23Laminin polymerization is mediated by the LN domains found at the end of eachshort arm. The short arms form a ternary node with the short arms of otherlaminin molecules. However, it is still unknown whether the laminin-typeepidermal growth factor-like (LE) domains found next to the LN domains play arole in the formation of the ternary node since LN domains cannot be producedwithout the LE domains.23Laminins have a unique role in basement membrane assembly and they initiatethis through laminin polymerization. Genetic ablation of ?1 or ?1 chains proved to be embryonically lethaldue to the resultant failure of heterotrimeric formation and basement membraneassembly.36, 37 On the other hand, geneticablation of other basement membrane components led to defects later indevelopment and did not prevent basement membrane formation.
38-42Laminins interact with other componentsof the basement membrane such as nidogen, perlecan, and collagen IV.23, 30 Collagen IV in the basementmembrane is seen as the maturation of the basement membrane that is essentialfor structural stability later in development.23, 40 The exact mechanism howlaminins bind to collagen IV is still debated upon. Initial studies indicatedthat nidogen binds to laminin through the LE domains of the ?1 chain andcollagen IV, thus serving as an intermediary between the two networks found inthe basement membrane. However, studies have shown that nidogen might not bethe major bridge in connecting laminins and collagen IV.23It was then found that the interaction between laminins and collagen IV wasmediated by heparan sulfates.
43Perlecan was thought to mediate this function. However, it was found thatgenetic ablation of perlecan in mice still had the collagen IV intact.23, 40 It is then postulated thatagrin, another pericellular HSPG, serves as a compensating candidate. In thismodel, both perlecan and agrin would bind to the nidogen containing lamininnetwork and to the collagen IV’s 7S and NC1 domains.44, 45Laminins are essential in connecting thecell to the ECM through its interactions with both cell surface receptors andother components of the ECM. While collagen, proteoglycans, and hyaluronic acidcomprise the major structural component of the ECM, laminins are one of themolecules that bridge the interaction gap between the cells and the ECM.
4 1.4 Fibronectin Fibronectin is a multidomain proteinthat interacts with the various ECM components described previously.46, 47 Similar to laminin, itconnects the cell to the ECM.4 It is encoded by a single gene,but it has 20 isoforms in humans as a result of alternative splicing of themRNA.46, 47 Its amino acid sequenceconsists of repetitive domains that are categorized either as type I, II, orIII. Alternative splicing occurs at EIIIA, EIIIB, and variable regions of thegene.46 Similar to collagen, it forms afibrillar network in the extracellular matrix.46Fibronectin naturally exists as a dimer outside the cell, mediated by the two cysteinedisulfide bonds.
This dimerization is crucial for fibrillar assembly offibronectin. Fibronectins lacking these cysteines resulted in the secretion ofmonomers that did not form fibrils.46, 48 Fibronectin matrix assemblyis mediated by selective binding to ?5?1 integrins through an RGD binding motifand a synergy site on the fibronectin molecule.46Both of these sites are required for the initiation of fibril assembly.
46, 49 Through these integrins, thecompact and soluble secreted fibronectin is unfolded revealing cryptic bindingsites for other fibronectin molecules.46, 50,51 Using anti integrin or antifibronectin antibodies have been shown to prevent fibronectin fibril formation.46, 52,53 Fibronectin binding inducesintegrin clustering that provides local high concentrations of fibronectin atthe cell surface. This phenomenon promotes fibronectin-fibronectin interactionsthrough the N terminal assembly domains of each molecule.46 Once it is tethered on the cell surface by integrins, theactin cytoskeleton can pull onto fibronectin molecules to change itsconformation.46, 47 This will affect the Cterminal regions of fibronectin, revealing cryptic binding sites forfibronectin, heparan sulfates, heparin, collagen, and other ECM proteins.46, 47 This allows it to interactwith proteoglycans such as syndecans and perlecan that aid in matrix assemblyas well.
54-58 It is through strongnoncovalent protein-protein interactions that the fibronectin network maturesand becomes insoluble. Although it is mainly through thefibronectin-fibronectin interactions, other ECM proteins may mediate lateralinteractions between fibrils.46This stabilizes the relatively weak binding sites at individual sites.
However,the turnover of the fibronectin matrix is still largely unexplored.46 Due to fibronectin’s multiple binding sites for other ECMproteins, it has been implicated in various functions including a role incollagen type I assembly. It is postulated that fibronectin provides directionfor collagen I assembly.46It has also been shown that without fibronectin, collagen fibrils do notaccumulate.59, 60 Furthermore, in developingtissues, fibronectin has been shown to be present before collagen.46The interactions between fibronectin and collagen are regulated by mechanicalforces.61 Moreover, structuralanalysis revealed cooperative binding between collagen and fibronectin.
62 Additionally, studies haveshown that collagen can also enhance fibronectin assembly.46, 63,64 2. Function of ECM Thecomplex ECM has several functions and can influence biochemical and biophysicalprocesses in the cell simultaneously. While the ECM has been largely regardedas just a scaffold that provides structure for the cells, it has several otherfunctions that largely impact the cell phenotype. The ECM can serve as bindingsites, controlling the adhesion and movement of cells. This is emphasized inthe complex structure and composition of the basement membrane that serves as abarrier between epithelial cells and the interstitial stroma.
In addition tostructural integrity and anchorage, the ECM components have several bindingsites for growth factors, controlling their release and presentation to targetcells. This is especially important in morphogenesis as it establishesmorphogen gradients. Finally, the ECM transmits mechanical signals to thecells, which activates several intracellular signalling pathways andcytoskeletal machinery. Indeed, the ECM serves several functions and here wereview the function of the ECM in the context of development and maintenance ofthe stem cell niche.