Amelogenins: Multifaceted Proteins for Dental and Bone Formation and Repair

Hypermineralized tissues related to enamel are identified in early vertebrates, 450 millions years ago (Ma). The enamel matrix proteins (EMPs), amelogenin (AMEL), ameloblastin (AMBN) and enamelin (ENAM) being enamel specific, we can deduce that they were recruited early in vertebrate evolution. Molecular analyses support their presence by the end of Precambrian period ( > 600 Ma), i.e., prior vertebrates differentiated a mineralized skeleton. However, our knowledge of EMPs is currently limited to tetrapods, i.e., 360 Ma. AMEL was created from AMBN, itself resulting from a duplication of ENAM. EMP genes are therefore paralogs. The evolutionary analysis of AMEL highlights conserved and variable regions. The N- and C-terminal regions contain numerous residues that were unchanged during 360 Ma, which supports important functions. Only a few positions are known to play a role. Other positions are considered as being crucial because they are unchanged. Five of them are known to lead to a genetic disease, X-linked amelogenesis imperfecta (AIH1) when substituted. The largest AMEL region, encoded by exon 6, is less variable in mammals than in sauropsids and amphibians, but it accumulates numerous indels in all lineages. It was created through the repeat of PXQ triplets. Similar repeats occurred independently at the same locus in mammalian species, but their meaning is still obscure. The evolutionary analysis of AMEL allowed to validate functional residues and positions known to lead to AIH1 when changed. It also highlighted residues that could have interesting functions and predicts these conserved positions will lead to AIH1 when substituted.

The amelogenin genes encode the conserved amelogenin proteins, which are expressed at a high level during the secretory stage of amelogenesis by ameloblast cells in the developing tooth. Most of the amelogenin genes studied to date have 7 exons, however exons 8 and 9 have been identified in some rodents. Amelogenin genes are localized to X and Y chromosomes in human, bovine, and several other species, while in rodents, the single gene is X-chromosomal, and is autosomal in monotremes. Promoter and intron analyses of murine and bovine genes have revealed regulatory regions, and alternative splicing of the primary RNA transcript is a prominent feature in all species analyzed. The X-chromosomal amelogenin (Amelx) gene is nested within an intron of the Arhgap6 gene in human, mouse, bovine, dog and horse, generally in the opposite orientation. Amelogenin mRNA and protein have also been detected at low levels in dental pulp cells and in several nondental tissues, including brain, bone and stem cells. Future research is expected to address down-regulation at the end of the secretory stage of amelogenesis, mechanisms related to regulation of alternative RNA splicing and stability during development, and expression patterns and consequences of amelogenin expression in non-dental cell types.

Although others and we have earlier reported that the amelogenins play critical roles in proper enamel formation, the in vivo functions of each of the amelogenin isoforms have not been clearly defined because of their heterogeneity and complexity. This chapter will review the current information on the functions of amelogenin splice isoforms, full length amelogenin and leucine rich amelogenin peptide (LRAP), in enamel formation. We will also discuss the emerging evidence for the additional role of LRAP as a signaling molecule in mesenchymal cells of the periodontal tissues. We believe that further insights into the signaling pathway modulated by the multifunctional amelogenin proteins will lead to development of new therapeutic approaches to treat dental diseases and disorders.

The cells that form calcified tissues produce various matrix proteins that foster a favourable environment for the regulated and structured deposition of calcium and phosphate ions into a carbonated form of apatite mineral. Differing from collagen-based calcified tissues, the organic matrix of enamel is produced by epithelially-derived cells – the ameloblasts, and consists of two major classes of proteins, amelogenins (AMEL) and nonamelogenins [1]. The AMEL class comprises full-length proteins, truncated isoforms resulting from alternative mRNA splicing, and fragments generated by extracellular proteolytic processing [1]. Enamel is distinctive from other calcified tissues because its organic matrix must ultimately be almost totally removed for it to achieve its full mineralization status. Thus, amelogenesis involves both formative and degradative processes. Studies over the past few years have revealed unexpected potentials for AMEL beyond structuring and organizing mineral at the surface of teeth, in particular their apparent capacity to influence osteogenic events. The objective of this mini-review is to highlight some key features of AMEL as related to both processes, and briefly go over our efforts to introduce enamel protein transgenes in bone forming cells.

Ameloblastin (AMBN) and enamelin (ENAM) are two of the 23 secretory calcium-binding phosphoprotein (SCPP) genes on human chromosome 4q. They are members of the proline and glutamine rich group. Both ameloblastin and enamelin function specifically during tooth formation, and probably both are only critical for dental enamel formation. Both genes have been shown to degenerate in mammals that have lost the ability to make dental enamel during evolution. In mice that cannot express functional ameloblastin or enamelin during tooth formation the mineralization front along the secretory portion of the ameloblast plasma membrane fails to generate enamel mineral ribbons. A non-functional mineral crust forms over dentin that readily abrades away when the teeth erupt into function. Ameloblasts undergo pathological changes, such as mineralization of the cell layer, disorganization, and apoptosis. In this chapter we review what is known about ameloblastin structure and function and conclude that enamelin and ameloblastin are both critical for the formation of enamel mineral ribbons during the secretory stage of amelogenesis and that their primary function is at the mineralization front.

As identified using in vitro methodologies, the enamel matrix proteins (EMPs) bind a class of proteins called lysosomal associated membrane proteins (LAMPs), whose members are LAMP1, LAMP2 and CD63/LAMP3. LAMPs are transmembrane proteins present in the plasma membrane, and all three directly interact with the AP-3 complex to initiate receptor-mediated endocytosis. In yeast, AP-3 associated endocytosis is clathrin-independent, while in higher organisms the requirement of clathrin is unclear. AP-3 is a protein complex with four subunits referred to as β, δ, μ and σ Three other adaptor protein complexes (AP-1, AP-2 and AP-4), each having four similar but unique subunits, are also involved with endocytosis and are either clathrin-dependent (AP-1 and AP-2) or clathrin-independent (AP-4). Only AP-3 interacts directly with plasma membrane-bound LAMPs. It is likely that endocytosis of the degraded EMPs plays a significant role in enamel maturation, and if so, AP-3 dysfunction may result in enamel abnormalities. We have initiated in vivo studies to better define the dental phenotype in a strain of mice null for the β subunit of AP-3 (coded by the AP3B1 gene), which results in the loss of AP-3 function. The driving hypothesis of future studies is that “failure of normal ameloblast endocytosis of the degraded EMPs during enamel maturation results in functionally inferior enamel more prone to dental disease”. The question we expect to answer from these mice studies is whether the receptor-mediated AP-3 endocytotic pathway is a prominent feature of amelogenesis. This chapter discusses the potential role of AP-3 mediated endocytosis in amelogenesis and describes animal experimental models that could be used to examine such a relationship.

For decades, the amelogenins which comprise 90% of the developing extracellular enamel matrix proteins were known to be exclusively expressed in ectodermal enamel and play a major role in its structural organization and biomineralization. In more recent years other roles for amelogenin have been discovered. This manuscript focuses on amelogenins possible roles during regeneration of tooth supporting tissues, in normal long bone, and during development of the embryonic craniofacial complex; A single recombinant human amelogenin protein (rHAM+), induced in-vivo regeneration of all tooth-supporting (periodontal) tissues after creation of experimental periodontitis. Amelogenin recruited (directly or indirectly) mesenchymal stem cells during periodontal regeneration. Amelogenin beads applied on developing embryonic mesenchyme and tooth germ recruited mesenchymal cells. Amelogenin was expressed in normal and regenerating alveolar bone cells (osteocytes, osteoblasts and osteoclasts), periodontal ligament, cementum and bone marrow stromal cells. Amelogenin was also expressed in long bone, periosteum, articular cartilage and differentially in ephyphysial growth plate cells. Amelogenin expression was highest in areas of high bone turnover and remodeling in normal and regenerating tissues. Spatio-temporal expression of amelogenin during embryonic craniofacial development is dynamic. These results and accumulating reports suggest that amelogenin participates in regulation of some developmental processes, acts as a signaling molecule affecting both normal and regenerating tissues, and is involved in maintaining the balance between bone formation and resorption. The pattern of amelogenin expression and the induction of periodontal regeneration, possibly offer future treatment to some widespread periodontal, bone, ligament and joint diseases, which are of major clinical, scientific and financial challenges.

The amelogenin gene (AMELX) was the first extracellular enamel matrix coding gene to be discovered. Mutations in AMELX cause a variety of changes in the amelogenin protein that cause abnormal enamel formation. The resulting clinical phenotypes are diverse and are referred to as amelogenesis imperfecta (AI). To date, fourteen allelic AMELX mutations have been described that purportedly result in markedly different expression and types of amelogenins. Interestingly, the different protein domains affected by distinct AMELX gene mutations result in unique and functionally altered amelogenin proteins that are associated with several rather distinct AI phenotypes. The AMELX mutations and associated phenotypes fall generally into three categories. Mutations (e.g. signal peptide mutations) causing a total of loss of amelogenin protein are associated with a predominantly hypoplastic phenotype (though mineralization defects can also occur). Missense mutations affecting the Nterminal region, especially those causing changes in the putative lectin-binding domain and Tyrosine Rich Amelogenin Protein region that contains several critical protease cleavage sites, result in a predominantly hypomineralization / hypomaturation AI phenotype with enamel that is discolored and has retained amelogenin. Mutations causing loss of the amelogenin C terminus result in a phenotype characterized by hypoplasia. The phenotype – genotype correlations indicate there are important functional domains of the amelogenin molecule that are critical for the development of normal enamel structure, composition and thickness and that perturbations in these different domains can result in different clinical phenotypes.

MMP20 (matrix metalloproteinase-20; enamelysin) and KLK4 (Kallikrein-4) are the two major, and perhaps only, proteinases present in developing dental enamel. MMP20 cleaves enamel resident proteins during the early developmental stages when the enamel grows in thickness and KLK4 cleaves enamel proteins prior to their absorption by the ameloblasts that sit atop full thickness enamel. KLK4 is expressed when the enamel starts to harden into its final mature form. Mutations in humans that eliminate the function of either MMP20 or KLK4 cause non-syndromic enamel malformations termed amelogenesis imperfecta. Homozygous deletion of either of these proteinases from the mouse genome will also result in severely malformed dental enamel. With the exception of malformed enamel, no other phenotype is yet known to occur when either of these proteinases are rendered functionless in humans or mice. Therefore, both MMP20 and KLK4 are essential for enamel formation. This chapter will provide historical insights into the discovery of these proteinases and will provide insights into the mechanistic function of how these proteinases act to direct enamel development.

In addition to its critical structural function during enamel biomineralization amelogenin protein also exhibits distinct signaling activities in a variety of in vitro and in vivo experimental systems. We present computational and biophysical data to demonstrate that amelogenin belongs to the class of Intrinsically Disordered Proteins (IDPs). IDPs lack a well-defined 3D structure under native conditions and are typically flexible, extended, and have little secondary structure in vitro in the absence of partners. This chapter considers details of the conformation, secondary structure, and degree of disorder in amelogenin. We propose that the occurrence of “conformational plasticity” in amelogenin explains its multi-functionality, contributes to its strong tendency to self-assemble, and its potential to specifically interact with different targets.

Significant advances have been made in understanding the role of amelogenin and other matrix proteins in the regulation of enamel formation, although the mechanism for this process is not fully understood. However, it is apparent that there are several guiding principles associated with the formation of mineralized tissues, although the mechanism by which enamel mineral forms differs from other mineralized tissues in several distinct aspects. This review briefly describes some of the similarities and difference between enamel, bone and dentin formation and highlights recent in vitro and in vivo studies from our laboratories that show that amelogenin regulates enamel mineral formation using a strategy that appears to be universally utilized in the development of mineralized tissues. More specifically, the presented studies show that full-length amelogenin has the capacity to stabilize the formation of amorphous calcium phosphate (ACP) and guide its organization into linear needle-like structures that subsequently fuse and organize into parallel bundles of apatitic crystals in vitro. Importantly, these in vitro findings are consistent with recent in vivo data presented that show that mineral morphology and organization in enamel are established by the organic matrix prior to its crystallization. These new findings should aid in the development of novel approaches for mineralized tissue regeneration and repair.

The amelogenin (AMEL) gene intron-exon structure, order and nucleotide sequence has been well preserved during evolutionary development. The variations of splicing of the nuclear pre-mRNA to produce the mRNA and subsequently translated proteins, depend on the sequence lengths of both the introns and exons, the presence of possible cryptic splice sites in large exons, the intrinsic “strengths” of the splice boundary sequences in interaction with the spliceosome components and the intervention of a variety of tissue and species specific accessory factors. A large number of splice isoforms has indeed been found; 16 alternative protein isoforms have been reported in mouse amelogenin, but the actual splicing patterns do differ in different animals. The major higher mass isoforms are used in building the amelogenin ECM and directing the organization and mineralization of the enamel. The question raised here is the role of the smaller isoforms? Their role appears to be largely as participants in regulatory processes during development of the enamel organ and in odontogenesis. However, it is becoming clear that amelogenin expression is not restricted to the enamel organ. Amelogenin peptides have been found in such diverse organs as brain, eye, cartilage and bone during embryogenesis. The knockout of amelogenin gene is not lethal, and these non-odontogenic tissues do form and appear to function normally, suggesting that the normal in vivo effects of the smaller splice isoforms may be subtle, modifying developmental rates, but it is evident that individual small amelogenin protein isoforms, added exogenously in vivo and in culture can markedly alter phenotypic expression in non-odontogenic tissues, as well as induce various forms of tissue repair in teeth. Thus, the study of the AMEL gene small isoforms and their mechanisms of action are fields worthy of extended study.

Exons 8 and 9 are two novel exons found in several transcripts of rodent amelogenin gene. These transcripts result from alternative splicing, in which exon 7 is replaced by exons 8 and 9, and they add a unique 3' terminus to amelogenin isoforms. Therefore, the alternatively spliced amelogenins end either at exon 7, encoding a single aspartic acid, or at exons 8 and 9, coding for 14 amino acids. So far, a total of seven alternatively spliced amelogenin mRNAs ending at exons 8 and 9 have been identified in rodents. The formation of these two exons results from a duplication of a DNA region containing exon 5, followed by its translocation dowstream exon 7. This event occurred during evolution of the rodent lineage at a period that postdates the divergence of the squirrel lineage, around 50 millions years ago. Therefore, exon 8 is homolog to exon 5. The amelogenins carrying the region encoded by exons 8 and 9 locate not only in ameloblasts and the enamel matrix but also in odontoblasts and dentin. The replacement of exon 7 by exons 8 and 9 leads to an additional hydrophilic domain, the function of which is not known, but may in part be related to the uniquely different rod and interrod structure seen in rodent enamel as compared to human enamel. The function of the amelogenins with exons 8 and 9 terminal peptides has been preliminarily investigated. They were found to enhance the proliferation of mesenchymal cells.

Two low molecular weight amelogenin gene splice products (A+4 and A-4), used for pulp capping after surgical pulp exposure in the rat first maxillary molars, have been shown to induce the formation of a reparative dentinal bridge. A-4 can induce mineralization of the atubular orthodentinal type to totally fill the root pulp canal with a homogeneous mineralized structure. The two molecules recruit latent adult pulp progenitors to form actively dividing cells, which differentiated into cells bearing an osteoblast-odontoblast phenotype to produce an extracellular matrix implemented in reparative mineralization. We conclude that pulp capping with bioactive molecules will direct new approaches in biodentistry.

The developing tooth is a well-recognized model for epithelial-mesenchyme organ formation. It is also a model for biomineralization, in which the formation of several unique bioceramic tissues, dentine, enamel and cementum are formed by the influence exerted by the protein matrix on their mineral habit. Enamel is the only ectodermal derived tissue to biomineralize and demonstrates several informative strategies to meet the requirement to produce a tissue that last a lifetime in a wet, bacterial laden environment subjected to the stresses associated with food ingestion. Enamel formation is dependent upon protein self-assembly to fabricate a precursor protein matrix to direct the mineral phase. Cell to matrix interactions are used to continuously monitor matrix fabrication and the mineral conversion process. We demonstrate the use of an artificial matrix created by nanotechology in which peptide amphiphiles undergo self-assembly into nanofibers that support cell proliferation and whose surfaces can be engineered to present biological signals to direct ameloblast differentiation. This biologically inspired technology may be used for regeneration of enamel tissue, offering a biomimetic alternative to traditional dental materials.

Amelogenin is the most abundant protein of the enamel organic matrix and is a structural protein indispensable for enamel formation. A commercial product Emdogain, consisting largely of alternatively spliced and processed amelogenins, could induce new bone, cementum and periodontal ligament formation in the jaws. Using mouse embryonic stem (ES) cells as a model system, we demonstrate that one of the amelogenin splicing isoforms, Leucine-rich Amelogenin Peptide (LRAP) activates the canonical Wnt signaling pathway to induce osteogenic differentiation of mouse ES cells through the concerted regulation of Wnt agonists and antagonists. Furthermore, LRAP stimulates bone regeneration in a rat femoral critical size defect model.

The goal of regenerative periodontal therapy is the reconstitution of the lost periodontal structures (i.e. the new formation of root cementum, periodontal ligament and alveolar bone). Findings from basic research have pointed to the important role of an enamel matrix protein derivative (EMD) in periodontal wound healing. Histological results from acute and chronic animal models and from human case reports have provided evidence that treatment with EMD may indeed promote periodontal regeneration and positively influences wound healing in intrabony, class II furcation and recession type defects. Recent long-term randomized controlled clinical studies have demonstrated that the improvements obtained following surgery with EMD can be maintained on a long term basis. The aim of the present chapter is to provide an overwiev on the clinical applications for EMD in regenerative periodontal therapy.