Aoba, Takaaki

Current Understanding of Structure, Properties, and Function of Amelogenin

Takaaki AOBA and Hisao YAGISHITA

Department of Pathology,
The Nippon Dental University School of Dentistry at Tokyo,
1-9-20 Fujimi, Chiyoda-ku, Tokyo 102, Japan

Since it is known that amelogenin and its related proteins are highly conserved during vertebrate evolution, much interest has been directed to their functional significance in enamel mineralization. It is now widely accepted that amelogenins function in controlling the size and organization of the developing enamel crystals, giving rise to a highly ordered array of the largest biological apatite crystals. A recent in vitro work using anti-sense oligonucleotide strategies provided additional evidence for the postulated function of amelogenin in enamel mineralization1). In
this report, we give a brief view of the current understanding of the amelogenin.
What we have known about amelogenin are : its primary structures, location of the genes at X- and Y-chromosomes, secretion of multiple amelogenins by alternative splicing, heterogeneity stemmed from their post-secretory proteolytic processing, their solubility and aggregative properties, functional significance of the conserved hydrophilic segment at the C-terminus, and a linkage of disturbance of amelogenin degradation to tooth anomalies. The parent amelogenin displays high adsorption onto apatite crystals and inhibition of apatite crystal growth. It appears that those functional properties are not associated with partial sequences in the amelogenin
macromolecules but are determined by the whole molecular structure2). At present, however, interpretation of the structure-linked function of amelogenin is hindered by our limited knowledge of their molecular structure. In the last decade, conformational studies using various techniques, e.g., Fourier transform infrared spectroscopy, Raman spectroscopy, and x-ray diffraction, suggested that the amelogenin may have ｧ-turn and ｧ-sheet elements. Our recent studies using circular dichroism3) showed that the structure of the whole protein consists of discrete folding units. It was tentatively assigned that the N-terminal domain contains ｧ-sheet structures, while the spectral characteristics of the C-terminal domain are similar to those of a random coil conformation. It should be pointed out that the putative folding units appear to work in a cooperative manner to realize the postulated function of the protein in enamel mineralization.
Significant advancements made in the recent amelogenin studies are: progresses in determination of the amelogenin gene structures, scrutiny of amelogenin expression during the vertebrate evolution and mammalian tooth development, establishment of recombinant mouse-amelogenin expression in E. coli, and partial success of in vitro anti-sense oligonucleotide strategies. With knowledge about the unique amelogenin genes at the X- and Y-chromosomes, human X- and Y-amelogenins have been successfully used for sex determination of the skeletons of Tsar Nicholas II and his family4). One of the difficulties encountered in the past studies on amelogenins was that isolation and purification of a large amount of discrete amelogenins was hampered by the complexity of amelogenin polypeptides and their aggregative properties in the extracellular matrix. A defined recombinant amelogenin is now available using the E. coli expression system5). This success paves a way toward site-directed in vitro mutagenesis. It is also becoming realistic to investigate the function of amelogenins using transgenic animal models.
What we do not yet know about amelogenins are: their tertiary and oligomeric structures, biological　significance of the structural motifs, mechanism of amelogenin aggregation and its significance, interaction between amelogenin and non-amelogenins, and mechanistic basis of enamel tissue organization. It is still not clear whether the crystallization of amelogenin for X-ray crystallographic analysis could be or may not be achieved due to inherent properties of the molecule. High resolution nuclear magnetic resonance (NMR) spectroscopy is one of the methods conventionally used to study conformational, dynamic aspects of protein structure in solution and in the solid state. However, application of NMR to the amelogenin work is still limited so that there is a paucity of information about the assignment of NMR signals of the amelogenin, except for several aromatic protons in the protein6).
It was demonstrated that amelogenin proteins exhibit unusual reversible aggregation properties; the aggregate can be dissociated into discrete monomers which, depending on surrounding conditions, spontaneously reassemble into aggregate structures. Since protein aggregation phenomena remain formidable for biochemical or molecular biological approaches, only a little is known about the mechanism of amelogenin aggregation and its significance in enamel mineralization.Several techniques have proven to or would be useful in studies on amelogenin aggregation, namely, atomic force microscopy (AFM), small-angle X-ray scattering using synchrotron radiation, low-angle laser light scattering photometry in conjunction with electrophoresis or chromatography, solid-phase NMR, microcalorimetry so on. AFM imaging permits the direct visualization of structures of the protein or aggregates in a hydrated environment. Recently, Fincham et al.8) demonstrated by both AFM and TEM, that the recombinant mouse amelogenin on mica or glass surface consisted of spherical aggregate structures of about 18 nm diameter. As the authors discussed, these structures may be brought about and stabilized by intermolecular hydrophobic interactions between 100-200 protein molecules.
An intriguing hypothesis concerning the structural motif and aggregation of amelogenin is that there is a structural property of the amelogenin protein which promotes a spontaneous self-assembly process. As Williamson indicated in a recent review9), repetitive proline-rich sequences or multiple tandem repeats may generate long open extended structures so that the proline-rich motif can facilitate protein-protein interactions. Proline has a large flat hydrophobic surface and therefore binds well to other flat hydrophobic surfaces such as aromatic rings. The regularly spaced proline residues are presumably important in maintaining an extended chain conformation and also in guiding associative processes. It is also of interest that many proline-rich sequences contain a large number of glutamine residues, forming common unifying motifs. As an extreme example, bovine amelogenin contains (QPX)20 motifs in the central region. The postulated function of the Pro and Glu-rich motifs for the amelogenin was substantiated, at least in part, by the previous NMR studies, in which we examined porcine amelogenin and its degraded fragments6). The 13 kD fragment, containing (QPX)13 but lacking the hydrophobic and hydrophilic segments, respectively, at the N- and C-termini, showed very sharpened proton signals in solution, supporting that the fragment has an open, extended structure. In contrast, the corresponding signals became broad in the presence of the N-terminal segment, which contains 6 Tyr and 2 Trp residues, suggesting involvement of the presumed hydrophobic-hydrophobic interactions between proline and aromatic residues. This interpretation is consistent with the observation that in the presence of high concentrations of urea, the NMR signals became sharp again.
A curiosity was given to the diversity in the Pro-rich regions amongmammalian amelogenins, despite their high homology at both the N- and C-termini. In this connection, it could be said that the homologous segments may provide specific functions for regulation of the protein-crystal interaction as supposed previously10), while the exact number or sequence of the Pro and Glu-rich repeats makes little difference to the relatively non-specific protein　protein association as assumed above.
Prior to conclude this report, it is notable that differential solubilities among the parent amelogenin and its degraded products have an important role in enamel mineralization11). The amelogenin is generally characterized as hydrophobic in nature on the basis of the high contents of proline and other hydrophobic residues. However, we found that the (QPX)n-containing fragment, produced in situ after cleavages of the N- and C-terminal segment of amelogenin, is concentrated in the enamel fluid12). In connection with such high solubility of Pro-rich polypeptide, Williamson indicated that the property of proline as a very good hydrogen bond acceptor (possibly because the electron-donating potential of the methylene group attached to the amide nitrogen causes the amide carbonyl to be electron-rich) leads to confusion as to whether proline should be classed as a hydrophobic or a hydrophilic residue.
Further understanding of the structure, property, and function of amelogenin will certainly lead to elucidation of the mechanism of enamel crystallization and tissue organization in the near future.

References

1) Diekwisch T, David S, Bringas P Jr, Santos V, Slavkin HC (1993) Antisense inhibition of AMEL translation demonstrates supramolecular controls for enamel HAP crystal growth during embryonic mouse molar development. Development 117:471-482.
2) Aoba T, Moreno EC, Kresak M, Tanabe T (1989) Possible roles of partial sequences at N- and C-termini of amelogenin in protein-enamel mineral interaction. J Dent Res 68:1331-1336
3) Goto Y, Kogure E, Takagi T, Aimoto S, Aoba T (1993) Molecular conformation of porcine amelogenin in solution :Three folding units at the N-terminal, central, and C-terminal regions. J Biochem 113:55-60.
4) Kingman S (1994) British DNA sleuths identify skeletons of Tsar, family. J NIH6: 40-42.
5) Simmer JP, Lau EC, Hu C-C, Aoba T, Lacey M, Nelson D, Thiemann F, Zeichner- David M, Snead ML, Slavkin H, Fincham A : Expression, isolation and characterization of mouse amelogenin in Escherichia coli. Calcif Tissue Int 54:312-319, 1994.
6) Aoba T, Kawano K, Moreno EC (1990) Molecular conformation of porcine amelogenins and its significance in protein-mineral interaction: 1H-NMR photo CIDNP study. J Biol Buccale 18:189-194.
7) Nikiforuk G and Simmons NS (1965) Purification and properties of enamel protein. J Dent Res 44:229-248.
8) Fincham AG, Moradian-Oldak J, Simmer JP, Sarte P, Lau EC, Diekwisch T, Slavkin HC (1994) Self-assembly of a recombinant amelogenin protein generates supramolecular structures. J Struct Biol 112:103-109.
9) Williamson, MP : The structure and function of proline-rich regions in proteins. Biochem J 297:249-260.
10) Aoba T, Tanabe T, Moreno EC (l987) Function of amelogenins in porcine enamel mineralization during the secretory stage of amelogenesis. Adv Dent Res 1:252- 260.
11) Shimizu M and Fukae M (l983) Enamel Proteins. In: Mechanisms of Tooth Enamel Formation, S.Suga, Ed., Tokyo: Quintessence Pub Co.,pp.125-141.
12) Aoba T and Moreno EC (1989) Mechanism of amelogenetic mineralization in minipig secretory enamel. In: Fearnhead RW (ed) Tooth Enamel V, Florence Pub., Yokohama, pp 163-167.

{　No Figures }