Physical Chemistry

In vertebrates, hearing and the sense of balance are initiated in the inner ear by specialized mechano-sensory cells, the hair cells. Hair cells mediate transduction of sound-pressure waves (hearing) or head accelerations (balance) into electrical signals that then propagate along nervous pathways to the brain. The inner ear's structures are immersed in a viscous fluid that should heavily damp sound-evoked mechanical vibrations. The ear, however, behaves as a highly-tuned resonator. Uniquely among sensory receptors, hair cells amplify their inputs by actively producing mechanical work that compensates viscous friction, thereby enhancing the sensitivity and sharpening the frequency selectivity of hearing. Our research at the interface between physics and biology aims at shedding light on the amplificatory process that shapes the sensation of sounds at the periphery of the auditory system.

Hair cells are each endowed with a mechano-sensory organelle, the hair bundle, that projects from the cell's apical surface into the surrounding fluid (Fig. 1). 

fig1: A: View of a single hair cell extracted from the saccule of the bullfrog (Rana catesbeiana). B: A scanning electron micrograph of a hair bundle. About 60 cylindrical processes, the stereocilia, are arranged in rows of increasing height. (sources : A : AJ Hudspeth, the Rockefeller University, New York ; B : P Gillespie, OHSU, Portland)

In the laboratory, we use flexible microfibers to measure the mechanical properties of the hair bundle. Our experiments demonstrate that this mechano-sensory antenna behaves as a sort of micro-muscle that can actively set a stimulus fiber under tension and even oscillate spontaneously (movie 1).

Movie 1: Spontaneous hair-bundle oscillations. Two hair bundles from the sensory epithelium of the bullfrog's saccule are viewed from the top. Only the tips of the tallest stereocilia are seen. The oscillation characteristics vary from cell to cell within a frequency range of 5-180 Hz and amplitudes of 20-100 nm. The video is in real time. The field of view spans 3 µm.

The hair cell can harness spontaneous hair-bundle oscillations to amplify its responsiveness to sinusoidal stimuli (Martin, 2008 ; Martin et Hudspeth, 1999). This hair-bundle amplifier offers double benefit for auditory detection: it enlarges the range of sound intensities that can be heard by amplifying only the weakest sounds and sharpens frequency selectivity by filtering the input to the hair cell. The ear is sensitive to sounds within a range of 20-20,000 Hz in humans. Our results led us to propose that sensitivity to different frequencies might result from the operation of an assembly of active oscillators with characteristic frequencies distributed within the auditory range.

Calcium controls the occurrence of spontaneous oscillations and the kinetics of active hair-bundle movements (Tinevez et al, 2007 ; voir Fig. 2A). In addition, active hair-bundle movements are associated with an unusual mechanical property (Fig. 2B). The stiffness of a hair bundle can indeed vary with bundle position and even be negative within a limited range of positions!

fig 2: A: From quiescence to spontaneous oscillations with Ca2+ iontophoresis. When the Ca2+ concentration is raised above a threshold value, we observe a bifurcation (Hopf bifurcation) towards an oscillatory behavior whose frequency increases with Ca2+. B: Force-displacement relation of an oscillatory hair bundle. The hair bundle displays a region of “negative stiffness” in the central region of the curve. The hair bundle cannot rest stably at the corresponding positions.

By combining experimental observations with simulations, we have built a theoretical description of the various manifestations of active hair-bundle motility in different species, including mammals (Martin et al, 2003 ; Nadrowski et al., 2004; Tinevez et al., 2007; Martin, 2008). In this model, active hair-bundle movements are powered by molecular motors of the myosin type.

To complement our research at the cellular level, we study in vitro the mechanical properties of the acto-myosin system, both at the level of a single acto-myosin crossbridge and of a few tens of motor molecules. Our experimental set-up combines optical tweezers, fluorescence microscopy and photorelease of molecules that can regulate the mechanical activity of myosin (video 2).

Movie 2: Gliding assay of actin filaments (fluorescent) deposited on a substrate densely coated with myosin molecules (not visible). When ATP (adenosine tri-phosphate) is released by pulses of ultraviolet light, the actin filaments are set in motion by the molecular motors and glide at an average speed of a few microns per second. The video is in real time.

Present and future research will clarify the effects of factors that influence the properties of the hair-bundle amplifier. These includes the viscosity of the fluid in which the hair bundles are immersed and mechanical coupling between cells of similar characteristic frequencies. In addition, we will seek correlates at the single hair-cell level of well-known psycho-acoustical phenomena, such as auditory illusions or masking of a sound by another. In parallel, we study the aptitude of a motor assembly to produce spontaneous oscillations in vitro.