David P. Corey - US grants

Mechanical stimuli are transduced into neural signals by receptor cells of the auditory and vestibular systems, by muscle spindles and tendon organs, and by the myriad types of cutaneous and deep pressure receptors. In no case is the mechanism understood by which a mechanical stimulus generates an electrophysiological response. The studies proposed here are directed at the basic process of mechanical transduction by hair cells, the receptor cells of the auditory and vestibular systems. Hair cells are chosen because the physiological stimulus--deflection of a ciliary bundle--is well defined and can be reproducibly delivered, because much of the necessary background work has been done, and because the auditory and vestibular systems are more critical than other mechanical senses for our interaction with the environment, and more debilitating when damaged or diseased. Three sets of questions will be asked concerning the transduction mechanism. One deals with conformational changes in the membrane channel proteins that underlie transduction, and with the structural proteins that deliver the stimulus to the channel. The second set is directed at characterizing an adaptation process in hair cells that acts to reduce the effective stimulus during mechanical displacements. The third seeks to describe the pore region of the transduction channel in terms of the relative selectivity among permeant ions and the position of binding sites for small organic cations. By using combination of recently developed mechanical and electrophysiological techniques, these questions can begin to be answered at a molecular level. Single hair cells will be voltage clamped with the whole-cell patch-clamp method to record the current through tranduction channels. Their hair bundles will be directly stimulated with step displacements under visual observation, and the bundle motion will be optically measured. Solutions on both sides of the cell membrane will be controlled, and the concentration of drugs of ions that affect the transduction process will be rapidly changed. The understanding of the transduction mechanism gained may help to elucidate the mechanism of noise trauma in the auditory system, and the toxic effects of the aminoglycoside antibiotics on hair cells.

The studies proposed here continue our work on the biophysical mechanism of sensory transduction by hair cells. A great deal of physiological evidence acquired over the last ten years indicates that the transduction channels in vertebrate hair cells are directly activated by mechanical stimuli: that some sort of "gating spring" conveys the displacement of the hair bundle to the channels, and that the stress causes channels to open. Similarly, physiological evidence suggests that adaptation of the transduction channels, most likely by a mechanical adjustment of the gating spring attachment. In this study, we will investigate the structural correlates of these physiologically-defined elements. First, we will test whether the "tip links" extending between stereocilia, originally described by Pickles, are the gating springs. We will develop a method to destroy the mechanical sensitivity of single hair cells, and then find out, with scanning electron microscopy, whether the tip links are also destroyed. Second, we will test the hypothesis of Howard and Hudspeth that adaptation is mediated by the upper attachment of the point of each tip link moving along the side of the stereocilium. Bundles will be displaced by 1-2micros m, allowed to adapt, and fixed for transmission electron microscopy. The positions of the attachment points will be measured to see if their position corresponds to the expected adaptation. Further tests will be to cut the tip links to see if the attachment points move upwards, as the structural model predicts; to get the tip links and see if the bundle moves forward by a tenth of a micron, as the biophysical model predicts; and to rule out other, more macroscopic rearrangements during adaptation wit high-sensitivity video subtraction. Third, we will investigate the role of other structures associated with the stereocilia, to determine a particular what holds the stereocilia together at their hips. For this aim, we will first confirm with high-resolution video measurements that stereocilia pivot at their bases and touch (but slide) at their tips. Then we will sequentially cut each of the three linkages between stereocilia, and determine with transmission EM which are intact when the bundle remain together. This understanding of hair bundle structures associated with transduction may illuminate certain pathological conditions of the auditory system. In particular, strong evidence for the tip-links hypothesis would implicate these filaments in the temporary threshold shift caused by noise trauma. Knowledge of what holds the bundles together may help in understanding why they fall apart with noise trauma.

PROJECT SUMMARY This work is designed to understand how proteins encoded by two human deafness genes?CDH23 and PCDH15?assemble to form the mechanosensory apparatus of hair cells in the auditory and vestibular systems. Each hair cell has a bundle of actin-based stereocilia arranged with increasing heights; each stereocilium of a cell extends a filamentous ?tip link? to the next taller stereocilium. Movement of the bundle tightens tip links; they in turn pull open force-gated ion channels that open to depolarize the cell. Thus tip links are at the heart of the inner ear?s function?to turn sound and head movement into neural signals. Each tip link is composed of CDH23 and PCDH15 proteins arranged in an antiparallel hetero-tetrameric filament so as to create two parallel strands with a slight helical twist. This unusual arrangement raises new questions: Why did the tip link evolve to have two strands rather than one? Why is it composed of two cadherin proteins that meet in the middle, rather than one protein that spans the distance between stereocilia? How strong is the bond between cadherins, and how does loud noise break it? Work in the previous project period led to an understanding of the bond at the atomic level, achieved by solving the X-ray crystal structure of the two cadherins where they join. Steered molecular simulations then predicted the mechanism of unbinding and predicted that it would depend steeply on time. But these predictions must be confirmed with biophysical measurements of single-protein unbinding. In this project, we will carry out single-molecule force spectroscopy measurements, to directly measure the unbinding and rebinding of the CDH23-PCDH15 bond. We will first explore the mechanical properties of a single-stranded tip link bond. We will then engineer proteins to contain two CDH23 N-terminal fragments or two PCDH15 N-terminals, to measure the mechanical properties of a double-stranded tip link bond. In each case, we will measure how the unbinding depends on force and on calcium concentration, to understand tip- link rupture with loud sound and in the ionic environment of the inner ear. Finally, we will explore how deafness-causing mutations in CDH23 and PCDH15 compromise the integrity of the tip link. We hypothesize that a double-stranded tip link is far stronger than a single-stranded one for the same reason that trapeze artists hold with two hands: if the cadherins of one strand unbind, they can rapidly rebind because they are held in close proximity, so the overall unbinding rate is greatly slowed. We further hypothesize that hair cells evolved a CDH23-PCDH15 bond that slips by a nanometer or so before release, which in turn makes the unbinding rate force-dependent. A bond that can completely break at very high forces acts as a mechanical circuit breaker, protecting other elements of the mechanotransduction apparatus.

This core serves to develop and make accessible technology for stochastic optical reconstruction microscopy (STORM), a combination of selective chromophore activation and computational reconstruction that has resolution in the 20-30 nm range. We begin with an instrument that was almost fully constructed, donated to the Core by the Corey Lab, but we needed to complete construction and develop a use-friendly interface appropriate for general use by a core user base. An additional goal was to continue to develop the instrument to reflect technological advances in STORM in the field.

Project Summary This project will develop a gene therapy strategy to treat hearing loss associated with Usher Syndrome type 3A (USH3A) in human. This will involve optimizing viral vectors in mouse models of USH3A, testing the best vectors for efficacy and cochlear toxicity in non-human primates and in human hair cells, and testing sequence variants in the human USH3A gene (CLRN1) for pathogenicity. At the completion of the project, we will be able to request a pre-IND meeting with the Food and Drug Administration to determine specific requirements for preclinical testing. Aim 1 is to test different AAV vectors for gene delivery to cochlear hair cells in mouse. Three new AAV- related vectors have shown promise in transducing cochlear hair cells, and we will test them all to identify the best. We will then test AAV-mediated restoration of hair cell function in the Clrn1-/- and N48K mouse models. Finally, we will test toxicity to hearing of AAV-mediated Clrn1 expression. Aim 2 is to test the efficacy of novel AAV vectors in human hair cells in vitro, and to test, in non-human primates, the efficiency, toxicity and immunogenicity of these vectors administered by round window membrane injection. We will test the two best vectors from Aim 1, expressing GFP to assess the degree of viral transduction of hair cells. We will also describe toxicity and systemic distribution of GFP- and CLRN1- encoding AAV vectors. An immune response might interfere with subsequent AAV injections or could lead to toxicity in the inner ear, so we will determine whether AAV injection into the ear induces neutralizing antibody or T-cell response. In Aim 3 we will characterize human pre-existing immunity against applicable serotypes in perilymph, using perilymph collected during surgery for other auditory disorders. We will also determine whether there is a correlation between antibody titers in perilymph and serum. Many variants have been identified in the CLRN1 gene but not all are known to be pathogenic. To determine which variants in CLRN1 in USH3A patients are pathogenic, we will set up a complementation assay using the Clrn1-/- mouse, use AAV to express Clrn1 with specific mutations equivalent to human variants of unknown significance, and assess degree of functional rescue with electrophysiology and electron microscopy.