Mutations in the Pejvakin (Pjvk) gene cause autosomal recessive hearing loss DFNB59 with audiological features of auditory neuropathy spectrum disorder (ANSD) or cochlear dysfunction. The precise mechanisms underlying the variable clinical phenotypes of DFNB59 remain unclear. Here, we demonstrate that mice with conditional ablation of the Pjvk gene in all sensory hair cells or only in outer hair cells (OHCs) show similar auditory phenotypes with early-onset profound hearing loss. By contrast, loss of Pjvk in adult OHCs causes a slowly progressive hearing loss associated with OHC degeneration and delayed loss of inner hair cells (IHCs), indicating a primary role for pejvakin in regulating OHC function and survival. Consistent with this model, synaptic transmission at the IHC ribbon synapse is largely unaffected in sirtaki mice that carry a C-terminal deletion mutation in Pjvk. Using the C-terminal domain of pejvakin as bait, we identified in a cochlear cDNA library ROCK2, an effector for the small GTPase Rho, and the scaffold protein IQGAP1, involved in modulating actin dynamics. Both ROCK2 and IQGAP1 associate via their coiled-coil domains with pejvakin. We conclude that pejvakin is required to sustain OHC activity and survival in a cell-autonomous manner likely involving regulation of Rho signaling.

Mutations in the Pejvakin (PJVK) gene are thought to cause auditory neuropathy and hearing loss of cochlear origin by affecting noise-induced peroxisome proliferation in auditory hair cells and neurons. Here we demonstrate that loss of pejvakin in hair cells, but not in neurons, causes profound hearing loss and outer hair cell degeneration in mice. Pejvakin binds to and colocalizes with the rootlet component TRIOBP at the base of stereocilia in injectoporated hair cells, a pattern that is disrupted by deafness-associated PJVK mutations. Hair cells of pejvakin-deficient mice develop normal rootlets, but hair bundle morphology and mechanotransduction are affected before the onset of hearing. Some mechanotransducing shorter row stereocilia are missing, whereas the remaining ones exhibit overextended tips and a greater variability in height and width. Unlike previous studies of Pjvk alleles with neuronal dysfunction, our findings reveal a cell-autonomous role of pejvakin in maintaining stereocilia architecture that is critical for hair cell function.SIGNIFICANCE STATEMENT Two missense mutations in the Pejvakin (PJVK or DFNB59) gene were first identified in patients with audiological hallmarks of auditory neuropathy spectrum disorder, whereas all other PJVK alleles cause hearing loss of cochlear origin. These findings suggest that complex pathogenetic mechanisms underlie human deafness DFNB59. In contrast to recent studies, we demonstrate that pejvakin in auditory neurons is not essential for normal hearing in mice. Moreover, pejvakin localizes to stereociliary rootlets in hair cells and is required for stereocilia maintenance and mechanosensory function of the hair bundle. Delineating the site of the lesion and the mechanisms underlying DFNB59 will allow clinicians to predict the efficacy of different therapeutic approaches, such as determining compatibility for cochlear implants.

An electrospun nonwoven matrix of polyamide nanofibers was employed as a new model for the capillary basement membrane at the blood-brain barrier (BBB). The basement membrane separates astrocytes from endothelial cells and is associated with growth factors, such as fibroblast growth factor-2 (FGF-2). FGF-2 is produced by astrocytes and induces specialized functions in endothelial cells, but also has actions on astrocytes. To investigate potential autocrine actions of FGF-2 at the BBB, astrocytes were cultured on unmodified nanofibers or nanofibers covalently modified with FGF-2. The former assumed an in vivo-like stellate morphology that was enhanced in the presence of cross-linked FGF-2. Furthermore, astrocyte monolayers established on unmodified nanofibers were more permissive for neurite outgrowth when cultured with an overlay of neurons than similar monolayers established on standard tissue culture surfaces, while astrocytes cultured on FGF-2-modifed nanofibers were yet more permissive. The observed differences were due in part to progressively increasing amounts of FGF-2 secreted by the astrocytes into the medium; hence FGF-2 increases its own expression in astrocytes to modulate astrocyte-neuron interactions. Soluble FGF-2 was unable to replicate the effects of cross-linked FGF-2. Nanofibers alone up-regulated FGF-2, albeit to a lesser extent than nanofibers covalently modified with FGF-2. These results underscore the importance of both surface topography and growth factor presentation on cellular function. Moreover, these results indicate that FGF-2-modified nanofibrillar scaffolds may demonstrate utility in tissue engineering applications for replacement and regeneration of lost tissue following central nervous system (CNS) injury or disease.