Abstract
Our strategy was to examine each of the three loci capable of contributing to the observed complexity 0 of mouse amelogenin proteins recovered from forming enamel: the genome (gene); the transcription apparatus (messenger RNA); and the translation apparatus (proteins). Our approach was based on recombinant DNA technology and a complementary DNA (cDNA) clone, pMa5-5, specific to the predominant mouse amelogenin protein. An "artificial ameloblast" was engineered based on pMa5-5 and the resulting synthetic products compared to those from authentic ameloblasts. First, the genome probably is not responsible for amelogenin complexity: Southern analysis indicates that the amelogenin gene exists as a single copy in either differentiated dental tissue or germ line tissue. Thus, ectomesenchymal-derived instructive signals for ameloblast differentiation do not lead to re-arrangement or amplification of the amelogenin gene. Next, using nucleic acid hybridization techniques, we examined messenger RNA from mouse ameloblasts. Northern analysis of authentic mRNA from mouse ameloblasts, with either the intact or 3'-end of pMa 5-5 used as the reporter molecule, indicates that only one size class of mRNA was detectable. We conclude that at the sensitivity of this assay there is no evidence for multiple mRNAs. Last, "artificial ameloblasts" were engineered so that the translation apparatus could be examined as a source of amelogenin complexity. Capped, artificial mRNAs were constructed to the pMa 5-5 template and used to program the synthesis of amelogenin polypeptides by translation in a cell-free system. When the resulting total translation products were immunoprecipitated with the rabbit anti-mouse amelogenin antibody, we observed multiple polypeptides, suggesting that the utilization of alternative start sites may also contribute to the observed complexity of amelogenin proteins, at least for artificial mRNAs translated in vitro.