Pathway: Crosslinking of collagen fibrils
Reactions in pathway: Crosslinking of collagen fibrils :
Crosslinking of collagen fibrils
After removal of the N- and C-procollagen propeptides, fibrillar collagen molecules aggregate into microfibrillar arrays, stabilized by covalent intermolecular cross-links. These depend on the oxidative deamination of specific lysine or hydroxylysine residues in the telopeptide region by lysyl oxidase (LOX) with the subsequent spontaneous formation of covalent intermolecular cross-links (Pinnell & Martin 1968, Siegel et al. 1970, 1974, Maki 2009, Nishioka et al. 2012). Hydroxylysine is formed intracellularly by lysine hydroxylases (LH). There are different forms of LH responsible for hydroxylation of helical and telopeptide lysines (Royce & Barnes 1985, Knott et al.1997, Takaluoma et al. 2007, Myllyla 2007). The chemistry of the cross-links formed depends on whether lysines or hydroxylysines are present in the telopeptides (Barnes et al. 1974), which depends on the proportion of collagen lysines post-translationally converted to hydroxylysine by LH. The lysine pathway predominates in adult skin, cornea and sclera while the hydroxylysine pathway occurs primarily in bone, cartilage, ligament, tendons, embryonic skin and most connective tissues (Eyre 1987, Eyre & Wu 2005, Eyre et al. 2008). Oxidative deamination of lysine or hydroxylysine residues by LOX generates the allysine and hydroxyallysine aldehydes respectively. These can spontaneously react with either another aldehyde to form an aldol condensation product (intramolecular cross-link), or with an unmodified lysine or hydroxylysine residue to form intermolecular cross-links.
The pathway of cross-linking is regulated primarily by the hydroxylation pattern of telopeptide and triple-helix domain lysine residues. When lysine residues are the source of aldehydes formed by lysyl oxidase the allysine cross-linking pathway leads to the formation of aldimine cross-links (Eyre & Wu 2005). These are stable at physiological conditions but readily cleaved at acid pH or elevated temperature. When hydroxylysine residues are the source of aldehydes formed by lysyl oxidase the hydroxyallysine cross-linking pathway leads to the formation of more stable ketoimine cross-links.
Telopeptide lysine residues can be converted by LOX to allysine, which can react with a helical hydroxylysine residue forming the lysine aldehyde aldimine cross-link dehydro hydroxylysino norleucine (deHHLNL) (Bailey & Peach 1968, Eyre et al. 2008). If the telopeptide residue is hydroxylysine, the hydroxyallysine formed by LOX can react with a helical hydroxylysine forming the Schiff base, which spontaneously undergoes an Amadori rearrangement resulting in the ketoimine cross link hydroxylysino 5 ketonorleucine (HLKNL). This stable cross-link is formed in tissues where telopeptide residues are predominanly hydroxylated, such as foetal bone and cartilage, accounting for the relative insolubility of collagen from these tissues (Bailey et al. 1998). In bone, telopeptide hydroxyallysines can react with the epsilon-amino group of a helical lysine (Robins & Bailey 1975). The resulting Schiff base undergoes Amadori rearrangement to form lysino-hydroxynorleucine (LHNL). An alternative mechanism of maturation of ketoimine cross-links has been reported in cartilage leading to the formation of arginoline (Eyre et al. 2010).
These divalent crosslinks greatly diminish as connective tissues mature, due to further spontaneous reactions (Bailey & Shimokomaki 1971, Robins & Bailey 1973) with neighbouring peptides that result in tri- and tetrafunctional cross-links. In mature tissues collagen cross-links are predominantly trivalent. The most common are pyridinoline or 3-hydroxypyridinium cross-links, namely hydroxylysyl-pyridinoline (HL-Pyr) and lysyl-pyridinoline (L-Pyr) cross-links (Eyre 1987, Ogawa et al. 1982, Fujimoto et al. 1978). HL-Pyr is formed from three hydroxylysine residues, HLKNL plus a further hydroxyallysine. It predominates in highly hydroxylated collagens such as type II collagen in cartilage. L-Pyr is formed from two hydroxylysines and a lysine, LKNL plus a further hydroxyallysine, found mostly in calcified tissues (Bailey et al. 1998). Trivalent collagen cross-links can also form as pyrroles, either Lysyl-Pyrrole (L-Pyrrole) or hydroxylysyl-pyrrole (HL-Pyrrole), respectively formed when LKNL or HLKNL react with allysine (Scott et al. 1981, Kuypers et al. 1992). A further three-way crosslink can form when DeH-HLNL reacts with histidine to form histidino-hydroxylysinonorleucine (HHL), found in skin and cornea (Yamauchi et al. 1987, 1996). This can react with an additional lysine to form the tetrafunctional cross-link histidinohydroxymerodesmosine (Reiser et al. 1992, Yamauchi et al. 1996).
Another mechanism which could be involved in the cross-linking of collagen IV networks is the sulfilimine bond (Vanacore et al. 2009), catalyzed by peroxidasin, an enzyme found in basement membrane (Bhave 2012).
To improve clarity inter-chain cross-linking is represented here for Collagen type I only. Although the formation of each type of cross-link is represented here as an independent event, the partial and random nature of lysine hydroxylation and subsequent lysyl oxidation means that any combination of these cross-linking events could occur within the same collagen fibril .
The pathway of cross-linking is regulated primarily by the hydroxylation pattern of telopeptide and triple-helix domain lysine residues. When lysine residues are the source of aldehydes formed by lysyl oxidase the allysine cross-linking pathway leads to the formation of aldimine cross-links (Eyre & Wu 2005). These are stable at physiological conditions but readily cleaved at acid pH or elevated temperature. When hydroxylysine residues are the source of aldehydes formed by lysyl oxidase the hydroxyallysine cross-linking pathway leads to the formation of more stable ketoimine cross-links.
Telopeptide lysine residues can be converted by LOX to allysine, which can react with a helical hydroxylysine residue forming the lysine aldehyde aldimine cross-link dehydro hydroxylysino norleucine (deHHLNL) (Bailey & Peach 1968, Eyre et al. 2008). If the telopeptide residue is hydroxylysine, the hydroxyallysine formed by LOX can react with a helical hydroxylysine forming the Schiff base, which spontaneously undergoes an Amadori rearrangement resulting in the ketoimine cross link hydroxylysino 5 ketonorleucine (HLKNL). This stable cross-link is formed in tissues where telopeptide residues are predominanly hydroxylated, such as foetal bone and cartilage, accounting for the relative insolubility of collagen from these tissues (Bailey et al. 1998). In bone, telopeptide hydroxyallysines can react with the epsilon-amino group of a helical lysine (Robins & Bailey 1975). The resulting Schiff base undergoes Amadori rearrangement to form lysino-hydroxynorleucine (LHNL). An alternative mechanism of maturation of ketoimine cross-links has been reported in cartilage leading to the formation of arginoline (Eyre et al. 2010).
These divalent crosslinks greatly diminish as connective tissues mature, due to further spontaneous reactions (Bailey & Shimokomaki 1971, Robins & Bailey 1973) with neighbouring peptides that result in tri- and tetrafunctional cross-links. In mature tissues collagen cross-links are predominantly trivalent. The most common are pyridinoline or 3-hydroxypyridinium cross-links, namely hydroxylysyl-pyridinoline (HL-Pyr) and lysyl-pyridinoline (L-Pyr) cross-links (Eyre 1987, Ogawa et al. 1982, Fujimoto et al. 1978). HL-Pyr is formed from three hydroxylysine residues, HLKNL plus a further hydroxyallysine. It predominates in highly hydroxylated collagens such as type II collagen in cartilage. L-Pyr is formed from two hydroxylysines and a lysine, LKNL plus a further hydroxyallysine, found mostly in calcified tissues (Bailey et al. 1998). Trivalent collagen cross-links can also form as pyrroles, either Lysyl-Pyrrole (L-Pyrrole) or hydroxylysyl-pyrrole (HL-Pyrrole), respectively formed when LKNL or HLKNL react with allysine (Scott et al. 1981, Kuypers et al. 1992). A further three-way crosslink can form when DeH-HLNL reacts with histidine to form histidino-hydroxylysinonorleucine (HHL), found in skin and cornea (Yamauchi et al. 1987, 1996). This can react with an additional lysine to form the tetrafunctional cross-link histidinohydroxymerodesmosine (Reiser et al. 1992, Yamauchi et al. 1996).
Another mechanism which could be involved in the cross-linking of collagen IV networks is the sulfilimine bond (Vanacore et al. 2009), catalyzed by peroxidasin, an enzyme found in basement membrane (Bhave 2012).
To improve clarity inter-chain cross-linking is represented here for Collagen type I only. Although the formation of each type of cross-link is represented here as an independent event, the partial and random nature of lysine hydroxylation and subsequent lysyl oxidation means that any combination of these cross-linking events could occur within the same collagen fibril .
Collagen is a family of at least 29 structural proteins derived from over 40 human genes (Myllyharju & Kivirikko 2004). It is the main component of connective tissue, and the most abundant protein in mammals making up about 25% to 35% of whole-body protein content. A defining feature of collagens is the formation of trimeric left-handed polyproline II-type helical collagenous regions. The packing within these regions is made possible by the presence of the smallest amino acid, glycine, at every third residue, resulting in a repeating motif Gly-X-Y where X is often proline (Pro) and Y often 4-hydroxyproline (4Hyp). Gly-Pro-Hyp is the most common triplet in collagen (Ramshaw et al. 1998). Collagen peptide chains also have non-collagenous domains, with collagen subclasses having common chain structures. Collagen fibrils are mostly found in fibrous tissues such as tendon, ligament and skin. Other forms of collagen are abundant in cornea, cartilage, bone, blood vessels, the gut, and intervertebral disc. In muscle tissue, collagen is a major component of the endomysium, constituting up to 6% of muscle mass. Gelatin, used in food and industry, is collagen that has been irreversibly hydrolyzed.
On the basis of their fibre architecture in tissues, the genetically distinct collagens have been divided into subgroups. Group 1 collagens have uninterrupted triple-helical domains of about 300 nm, forming large extracellular fibrils. They are referred to as the fibril-forming collagens, consisting of collagens types I, II, III, V, XI, XXIV and XXVII. Group 2 collagens are types IV and VII, which have extended triple helices (>350 nm) with imperfections in the Gly-X-Y repeat sequences. Group 3 are the short-chain collagens. These have two subgroups. Group 3A have continuous triple-helical domains (type VI, VIII and X). Group 3B have interrupted triple-helical domains, referred to as the fibril-associated collagens with interrupted triple helices (FACIT collagens, Shaw & Olsen 1991). FACITs include collagen IX, XII, XIV, XVI, XIX, XX, XXI, XXII and XXVI plus the transmembrane collagens (XIII, XVII, XXIII and XXV) and the multiple triple helix domains and interruptions (Multiplexin) collagens XV and XVIII (Myllyharju & Kivirikko 2004). The non-collagenous domains of collagens have regulatory functions; several are biologically active when cleaved from the main peptide chain. Fibrillar collagen peptides all have a large triple helical domain (COL1) bordered by N and C terminal extensions, called the N- and C-propeptides, which are cleaved prior to formation of the collagen fibril. The intact form is referred to as a collagen propeptide, not procollagen, which is used to refer to the trimeric triple-helical precursor of collagen before the propeptides are removed. The C-propeptide, also called the NC1 domain, directs chain association during assembly of the procollagen molecule from its three constituent alpha chains (Hulmes 2002).
Fibril forming collagens are the most familiar and best studied subgroup. Collagen fibres are aggregates or bundles of collagen fibrils, which are themselves polymers of tropocollagen complexes, each consisting of three polypeptide chains known as alpha chains. Tropocollagens are considered the subunit of larger collagen structures. They are approximately 300 nm long and 1.5 nm in diameter, with a left-handed triple-helical structure, which becomes twisted into a right-handed coiled-coil 'super helix' in the collagen fibril. Tropocollagens in the extracellular space polymerize spontaneously with regularly staggered ends (Hulmes 2002). In fibrillar collagens the molecules are staggered by about 67 nm, a unit known as D that changes depending upon the hydration state. Each D-period contains slightly more than four collagen molecules so that every D-period repeat of the microfibril has a region containing five molecules in cross-section, called the 'overlap', and a region containing only four molecules, called the 'gap'. The triple-helices are arranged in a hexagonal or quasi-hexagonal array in cross-section, in both the gap and overlap regions (Orgel et al. 2006). Collagen molecules cross-link covalently to each other via lysine and hydroxylysine side chains. These cross-links are unusual, occuring only in collagen and elastin, a related protein.
The macromolecular structures of collagen are diverse. Several group 3 collagens associate with larger collagen fibers, serving as molecular bridges which stabilize the organization of the extracellular matrix. Type IV collagen is arranged in an interlacing network within the dermal-epidermal junction and vascular basement membranes. Type VI collagen forms distinct microfibrils called beaded filaments. Type VII collagen forms anchoring fibrils. Type VIII and X collagens form hexagonal networks. Type XVII collagen is a component of hemidesmosomes where it is complexed wtih alpha6Beta4 integrin, plectin, and laminin-332 (de Pereda et al. 2009). Type XXIX collagen has been recently reported to be a putative epidermal collagen with highest expression in suprabasal layers (Soderhall et al. 2007). Collagen fibrils/aggregates arranged in varying combinations and concentrations in different tissues provide specific tissue properties. In bone, collagen triple helices lie in a parallel, staggered array with 40 nm gaps between the ends of the tropocollagen subunits, which probably serve as nucleation sites for the deposition of crystals of the mineral component, hydroxyapatite (Ca10(PO4)6(OH)2) with some phosphate. Collagen structure affects cell-cell and cell-matrix communication, tissue construction in growth and repair, and is changed in development and disease (Sweeney et al. 2006, Twardowski et al. 2007). A single collagen fibril can be heterogeneous along its axis, with significantly different mechanical properties in the gap and overlap regions, correlating with the different molecular organizations in these regions (Minary-Jolandan & Yu 2009).
Fibril forming collagens are the most familiar and best studied subgroup. Collagen fibres are aggregates or bundles of collagen fibrils, which are themselves polymers of tropocollagen complexes, each consisting of three polypeptide chains known as alpha chains. Tropocollagens are considered the subunit of larger collagen structures. They are approximately 300 nm long and 1.5 nm in diameter, with a left-handed triple-helical structure, which becomes twisted into a right-handed coiled-coil 'super helix' in the collagen fibril. Tropocollagens in the extracellular space polymerize spontaneously with regularly staggered ends (Hulmes 2002). In fibrillar collagens the molecules are staggered by about 67 nm, a unit known as D that changes depending upon the hydration state. Each D-period contains slightly more than four collagen molecules so that every D-period repeat of the microfibril has a region containing five molecules in cross-section, called the 'overlap', and a region containing only four molecules, called the 'gap'. The triple-helices are arranged in a hexagonal or quasi-hexagonal array in cross-section, in both the gap and overlap regions (Orgel et al. 2006). Collagen molecules cross-link covalently to each other via lysine and hydroxylysine side chains. These cross-links are unusual, occuring only in collagen and elastin, a related protein.
The macromolecular structures of collagen are diverse. Several group 3 collagens associate with larger collagen fibers, serving as molecular bridges which stabilize the organization of the extracellular matrix. Type IV collagen is arranged in an interlacing network within the dermal-epidermal junction and vascular basement membranes. Type VI collagen forms distinct microfibrils called beaded filaments. Type VII collagen forms anchoring fibrils. Type VIII and X collagens form hexagonal networks. Type XVII collagen is a component of hemidesmosomes where it is complexed wtih alpha6Beta4 integrin, plectin, and laminin-332 (de Pereda et al. 2009). Type XXIX collagen has been recently reported to be a putative epidermal collagen with highest expression in suprabasal layers (Soderhall et al. 2007). Collagen fibrils/aggregates arranged in varying combinations and concentrations in different tissues provide specific tissue properties. In bone, collagen triple helices lie in a parallel, staggered array with 40 nm gaps between the ends of the tropocollagen subunits, which probably serve as nucleation sites for the deposition of crystals of the mineral component, hydroxyapatite (Ca10(PO4)6(OH)2) with some phosphate. Collagen structure affects cell-cell and cell-matrix communication, tissue construction in growth and repair, and is changed in development and disease (Sweeney et al. 2006, Twardowski et al. 2007). A single collagen fibril can be heterogeneous along its axis, with significantly different mechanical properties in the gap and overlap regions, correlating with the different molecular organizations in these regions (Minary-Jolandan & Yu 2009).
The extracellular matrix is a component of all mammalian tissues, a network consisting largely of the fibrous proteins collagen, elastin and associated-microfibrils, fibronectin and laminins embedded in a viscoelastic gel of anionic proteoglycan polymers. It performs many functions in addition to its structural role; as a major component of the cellular microenvironment it influences cell behaviours such as proliferation, adhesion and migration, and regulates cell differentiation and death (Hynes 2009).
ECM composition is highly heterogeneous and dynamic, being constantly remodeled (Frantz et al. 2010) and modulated, largely by matrix metalloproteinases (MMPs) and growth factors that bind to the ECM influencing the synthesis, crosslinking and degradation of ECM components (Hynes 2009). ECM remodeling is involved in the regulation of cell differentiation processes such as the establishment and maintenance of stem cell niches, branching morphogenesis, angiogenesis, bone remodeling, and wound repair. Redundant mechanisms modulate the expression and function of ECM modifying enzymes. Abnormal ECM dynamics can lead to deregulated cell proliferation and invasion, failure of cell death, and loss of cell differentiation, resulting in congenital defects and pathological processes including tissue fibrosis and cancer.
Collagen is the most abundant fibrous protein within the ECM constituting up to 30% of total protein in multicellular animals. Collagen provides tensile strength. It associates with elastic fibres, composed of elastin and fibrillin microfibrils, which give tissues the ability to recover after stretching. Other ECM proteins such as fibronectin, laminins, and matricellular proteins participate as connectors or linking proteins (Daley et al. 2008).
Chondroitin sulfate, dermatan sulfate and keratan sulfate proteoglycans are structural components associated with collagen fibrils (Scott & Haigh 1985; Scott & Orford 1981), serving to tether the fibril to the surrounding matrix. Decorin belongs to the small leucine-rich repeat proteoglycan family (SLRPs) which also includes biglycan, fibromodulin, lumican and asporin. All appear to be involved in collagen fibril formation and matrix assembly (Ameye & Young 2002).
ECM proteins such as osteonectin (SPARC), osteopontin and thrombospondins -1 and -2, collectively referred to as matricellular proteins (reviewed in Mosher & Adams 2012) appear to modulate cell-matrix interactions. In general they induce de-adhesion, characterized by disruption of focal adhesions and a reorganization of actin stress fibers (Bornstein 2009). Thrombospondin (TS)-1 and -2 bind MMP2. The resulting complex is endocytosed by the low-density lipoprotein receptor-related protein (LRP), clearing MMP2 from the ECM (Yang et al. 2001).
Osteopontin (SPP1, bone sialoprotein-1) interacts with collagen and fibronectin (Mukherjee et al. 1995). It also contains several cell adhesive domains that interact with integrins and CD44.
Aggrecan is the predominant ECM proteoglycan in cartilage (Hardingham & Fosang 1992). Its relatives include versican, neurocan and brevican (Iozzo 1998). In articular cartilage the major non-fibrous macromolecules are aggrecan, hyaluronan and hyaluronan and proteoglycan link protein 1 (HAPLN1). The high negative charge density of these molecules leads to the binding of large amounts of water (Bruckner 2006). Hyaluronan is bound by several large proteoglycans proteoglycans belonging to the hyalectan family that form high-molecular weight aggregates (Roughley 2006), accounting for the turgid nature of cartilage.
The most significant enzymes in ECM remodeling are the Matrix Metalloproteinase (MMP) and A disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) families (Cawston & Young 2010). Other notable ECM degrading enzymes include plasmin and cathepsin G. Many ECM proteinases are initially present as precursors, activated by proteolytic processing. MMP precursors include an amino prodomain which masks the catalytic Zn-binding motif (Page-McCawet al. 2007). This can be removed by other proteinases, often other MMPs. ECM proteinases can be inactivated by degradation, or blocked by inhibitors. Some of these inhibitors, including alpha2-macroglobulin, alpha1-proteinase inhibitor, and alpha1-chymotrypsin can inhibit a large variety of proteinases (Woessner & Nagase 2000). The tissue inhibitors of metalloproteinases (TIMPs) are potent MMP inhibitors (Brew & Nagase 2010).
ECM composition is highly heterogeneous and dynamic, being constantly remodeled (Frantz et al. 2010) and modulated, largely by matrix metalloproteinases (MMPs) and growth factors that bind to the ECM influencing the synthesis, crosslinking and degradation of ECM components (Hynes 2009). ECM remodeling is involved in the regulation of cell differentiation processes such as the establishment and maintenance of stem cell niches, branching morphogenesis, angiogenesis, bone remodeling, and wound repair. Redundant mechanisms modulate the expression and function of ECM modifying enzymes. Abnormal ECM dynamics can lead to deregulated cell proliferation and invasion, failure of cell death, and loss of cell differentiation, resulting in congenital defects and pathological processes including tissue fibrosis and cancer.
Collagen is the most abundant fibrous protein within the ECM constituting up to 30% of total protein in multicellular animals. Collagen provides tensile strength. It associates with elastic fibres, composed of elastin and fibrillin microfibrils, which give tissues the ability to recover after stretching. Other ECM proteins such as fibronectin, laminins, and matricellular proteins participate as connectors or linking proteins (Daley et al. 2008).
Chondroitin sulfate, dermatan sulfate and keratan sulfate proteoglycans are structural components associated with collagen fibrils (Scott & Haigh 1985; Scott & Orford 1981), serving to tether the fibril to the surrounding matrix. Decorin belongs to the small leucine-rich repeat proteoglycan family (SLRPs) which also includes biglycan, fibromodulin, lumican and asporin. All appear to be involved in collagen fibril formation and matrix assembly (Ameye & Young 2002).
ECM proteins such as osteonectin (SPARC), osteopontin and thrombospondins -1 and -2, collectively referred to as matricellular proteins (reviewed in Mosher & Adams 2012) appear to modulate cell-matrix interactions. In general they induce de-adhesion, characterized by disruption of focal adhesions and a reorganization of actin stress fibers (Bornstein 2009). Thrombospondin (TS)-1 and -2 bind MMP2. The resulting complex is endocytosed by the low-density lipoprotein receptor-related protein (LRP), clearing MMP2 from the ECM (Yang et al. 2001).
Osteopontin (SPP1, bone sialoprotein-1) interacts with collagen and fibronectin (Mukherjee et al. 1995). It also contains several cell adhesive domains that interact with integrins and CD44.
Aggrecan is the predominant ECM proteoglycan in cartilage (Hardingham & Fosang 1992). Its relatives include versican, neurocan and brevican (Iozzo 1998). In articular cartilage the major non-fibrous macromolecules are aggrecan, hyaluronan and hyaluronan and proteoglycan link protein 1 (HAPLN1). The high negative charge density of these molecules leads to the binding of large amounts of water (Bruckner 2006). Hyaluronan is bound by several large proteoglycans proteoglycans belonging to the hyalectan family that form high-molecular weight aggregates (Roughley 2006), accounting for the turgid nature of cartilage.
The most significant enzymes in ECM remodeling are the Matrix Metalloproteinase (MMP) and A disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) families (Cawston & Young 2010). Other notable ECM degrading enzymes include plasmin and cathepsin G. Many ECM proteinases are initially present as precursors, activated by proteolytic processing. MMP precursors include an amino prodomain which masks the catalytic Zn-binding motif (Page-McCawet al. 2007). This can be removed by other proteinases, often other MMPs. ECM proteinases can be inactivated by degradation, or blocked by inhibitors. Some of these inhibitors, including alpha2-macroglobulin, alpha1-proteinase inhibitor, and alpha1-chymotrypsin can inhibit a large variety of proteinases (Woessner & Nagase 2000). The tissue inhibitors of metalloproteinases (TIMPs) are potent MMP inhibitors (Brew & Nagase 2010).