Scientists discover tiny protein region that shapes tooth enamel

Tooth enamel looks simple from the outside and consists of a hard, smooth surface built to withstand years of wear. On the other hand, under the microscope, mammalian enamel is arranged into tightly ordered prisms and inter-prisms, and those patterns are laid down by elongated cells called ameloblasts as enamel forms. Scientists have long known that ameloblast shape is closely tied to enamel architecture. What has been less clear is the molecular cue that helps these cells become so precisely polarized in the first place.

Now, Professor Janet Moradian-Oldak from the University of Southern California, USA, and her team focused on a very small region of ameloblastin (Ambn), the second most abundant extracellular matrix protein in developing enamel. This study was published in Volume 18 of the International Journal of Oral Science on August 20, 2026. The region forms an amphipathic helix, or AH motif, that can bind to cell membranes. It is also evolutionarily conserved; within the 11-amino-acid segment targeted in the study, nine residues were identical across mouse, pig, and human Ambn. That made the motif a compelling place to look for a mechanism that links the enamel matrix to the cells building it.

Using CRISPR-Cas9, the team generated mice in which the hydrophobic residues from Lys76 to Pro86 of the AH motif were deleted. Before turning to the animals, the researchers checked that the mutation had not simply destroyed Ambn as a protein. Recombinant mutant Ambn could still self-assemble, although less uniformly than the wild-type protein. Its ability to interact with ameloblast-lineage cells, however, was markedly reduced. In other words, the mutation selectively weakened the cell-binding function the team wanted to test. "Using this model, we reinforce the concept of multifunctionality of Ambn, with a selective disruption of Ambn-ameloblast interactions," said Prof. Moradian-Oldak.

Homozygous mutant mice formed enamel that reached essentially normal thickness, but the material was different in quality. Micro-computed X-Ray tomography showed delayed secretory and maturation stages, slower densification and a final mature enamel density of about 70% of the wild-type plateau. Scanning electron microscopy revealed a rough, 'sandpaper-like' surface and a striking loss of the usual rod-interrod organization. The enamel was there. Its internal architecture was not. That split between thickness and quality is important. The expression levels of the major enamel matrix genes AmelX and Enam remained normal in the mutants, which helps explain why the overall amount of enamel could still be laid down. The AH motif appears to do something more specific. It helps organize the cell polarity and matrix patterning needed to turn deposited enamel into a densely mineralized, prismatic material.

Ameloblasts in mutant mice were 19%-23% shorter, with disturbed Golgi positioning and mislocalized polarity markers such as Pard3 and claudin-1. Ambn also lost its normal localization along the distal membrane and Tomes' processes, which were rudimentary in the mutants. Signaling changes accompanied these polarity defects. Beta-catenin shifted into the nucleus, p-Smad2/3 showed increased nuclear localization, and RhoA signal intensity was reduced, pointing to possible involvement of Wnt, TGF-beta, and RhoA-ROCK pathways. Heterozygous mice also showed disrupted prism-interprism architecture, membrane interaction, and cell polarity despite normal enamel mineral density, suggesting that the AH motif has a specific functional role beyond simple protein dosage.

The research also has a direct link to hereditary enamel disease. Variants in Ambn are associated with amelogenesis imperfecta, and a previously reported truncation within the AH-motif region has been linked to contrasting clinical phenotypes and inheritance patterns.

Our new mouse model gives researchers a powerful way to uncover how defects in this tiny region of Ambn disrupt normal enamel development. Although our findings do not yet translate into a treatment, they highlight promising biological targets for future approaches to preventing or repairing enamel defects."  

Professor Janet Moradian-Oldak, University of Southern California, USA