Abstract
Auditory bulla cavitation defects are a cause of otitis media, but the normal cellular pattern of bulla mesenchyme regression and its failure are not well understood. In mice, neural-crest-derived mesenchyme occupies the bulla from embryonic day 17.5 (E17.5) to postnatal day 11 (P11) and then regresses to form the adult air-filled bulla cavity. We report that bulla mesenchyme is bordered by a single layer of non-ciliated epithelium characterized by interdigitating cells with desmosome cell junctions and a basal lamina, and by Bpifa1 gene expression and laminin staining of the basal lamina. At P11-P12, the mesenchyme shrinks: mesenchyme-associated epithelium shortens, and mesenchymal cells and extracellular matrix collagen fibrils condense, culminating in the formation of cochlea promontory mucosa bordered by compact non-ciliated epithelial cells. FBXO11 is a candidate disease gene in human chronic otitis media with effusion and we report that a bulla cavitation defect initiates the pathogenesis of otitis media in the established mouse model Jeff (Fbxo11Jf/+ ). Persistent mesenchyme in Fbxo11Jf/+ bullae has limited mesenchymal cell condensation, fibrosis and hyperplasia of the mesenchyme-associated epithelium. Subsequent modification forms fibrous adhesions that link the mucosa and the tympanic membrane, and this is accompanied by dystrophic mineralization and accumulation of serous effusion in the bulla cavity. Mouse models of bulla cavitation defects are important because their study in humans is limited to post-mortem samples. This work indicates new diagnostic criteria for this otitis media aetiology in humans, and the prospects of studying the molecular mechanisms of murine bulla cavitation in organ culture.