Modeling enamel matrix secretion in mammalian teeth

The most mineralized tissue of the mammalian body is tooth enamel. Especially in species with thick enamel, three-dimensional (3D) tomography data has shown that the distribution of enamel varies across the occlusal surface of the tooth crown. Differences in enamel thickness among species and within the tooth crown have been used to examine taxonomic affiliations, life history, and functional properties of teeth. Before becoming fully mineralized, enamel matrix is secreted on the top of a dentine template, and it remains to be explored how matrix thickness is spatially regulated. To provide a predictive framework to examine enamel distribution, we introduce a computational model of enamel matrix secretion that maps the dentine topography to the enamel surface topography. Starting from empirical enamel-dentine junctions, enamel matrix deposition is modeled as a diffusion-limited free boundary problem. Using laboratory microCT and synchrotron tomographic data of pig molars that have markedly different dentine and enamel surface topographies, we show how diffusion-limited matrix deposition accounts for both the process of matrix secretion and the final enamel distribution. Simulations reveal how concave and convex dentine features have distinct effects on enamel surface, thereby explaining why the enamel surface is not a straightforward extrapolation of the dentine template. Human and orangutan molar simulations show that even subtle variation in dentine topography can be mapped to the enamel surface features. Mechanistic models of extracellular matrix deposition can be used to predict occlusal morphologies of teeth.Author summary: Teeth of most mammals are covered by a layer of highly mineralized enamel that cannot be replaced or repaired. The enamel layer is not uniform over the underlying dentine, and spatial regulation of enamel formation is critical for making a functional tooth. To explore which kind of mechanisms could underlie the complex patterns of enamel distribution, we present a computational model. Starting from tomography-imaged teeth from which enamel has been digitally removed, enamel is restored on dentine surfaces by simulating diffusion-limited secretion of enamel matrix. Our simulations show how the combination of subtle features of dentine and diffusion-limited secretion of the enamel matrix can substantially increase the complexity of the enamel surface. We propose that the strength of the diffusion-limited process is a key factor in determining enamel distribution among mammalian species.