CHAPTER 146 Cochlear Transduction and the Molecular Basis of Auditory Pathology

• Animal models have played a key role in the effort to elucidate the role of gene products in peripheral auditory physiology and the molecular basis of deafness.

• Congenital hearing loss generally is inherited, and the phenotypic spectrum produced by mutations in a particular gene can be attributed to the severity of the mutation, allelic heterogeneity, alternative splicing, and the contribution of modifier genes. In addition, some genes interact with the environment (e.g., noise, aminoglycoside antibiotics) to increase susceptibility to hearing loss.

• The physical characteristics of the cochlear partition (i.e., its mass and stiffness) vary from one end of the cochlea to the other, creating a basoapical gradient that results in a mechanical system with resonant properties that are continuously distributed and span the audiometric range of hearing from high to low frequency.

• The tectorial membrane, a gelatinous extracellular matrix overlying the organ of Corti, provides the necessary structure to convert basilar membrane displacement to radial shearing forces, which in turn influence hair bundle deflection and subsequent mechanoelectrical transduction. Defects in genes encoding extracellular matrix proteins that are components of the tectorial membrane (e.g., TECTA, COL11A2, COL9A1) result in hearing deficits.

• Cochlear amplification refers to a physiologically vulnerable mechanism that increases sensitivity to sound, enhances frequency selectivity, and broadens the dynamic operating range. Cochlear amplification is the result of unique voltage-dependent contractile properties of outer hair cells, which rely on the SLC26A5 gene that encodes the essential motor protein prestin.

• Numerous proteins that are localized to sensory hair cells are critical for the normal development and/or maintenance of the intricate structure and/or function of the hair bundle and when defective are responsible for various forms of deafness. These include motor proteins (e.g., myosin IIIa, VI, VIIa, XVa), cytoskeletal proteins (e.g., espin, γ-actin, TRIOBP, radixin, diaphanous 1), adhesion proteins (e.g., cadherin 23, protocadherin 15, usherin, Vlgr1), scaffolding proteins (e.g., harmonin, SANS, whirlin), and other necessary proteins of unknown function (e.g., TMC1, TMHS, TMIE).

• Hearing impairments categorized as “auditory neuropathy” include those associated with defective inner hair cell ribbon synapse function (e.g., those caused by mutations in the genes OTOF and VGLUT3), as well as neuropathies involving faulty conduction along auditory nerve fibers, which often occur when components of the Schwann cell myelin sheath are defective (e.g., MPZ, GJB1, PMP22).

• Development of the epithelial barrier forming the endolymphatic compartment within the cochlear duct requires the vital tight junction components claudin-14 and tricellulin. In addition, the gap junction proteins connexin 26, connexin 30, and connexin 31 (encoded by GJB2, GJB6, and GJB3, respectively) are essential for ion homeostasis, as are numerous ion channel subunits (e.g., KCNQ4, BSND, KCNE1, KCNQ1) and ion pumps and transporters (e.g., ATP6B1, SLC26A4 [PDS, pendrin]).

• Many other proteins, including transcription factors and regulators, enzymes, ligands and receptors, and proteins produced by mitochondrial genes, have been associated with human deafness. Moreover, many genes underlying deafness in mouse models have been identified for which human deafness has not yet been established, and vice versa.

• A thorough, comprehensive understanding of the role that distinctive proteins play in normal cochlear development and peripheral auditory function will require a continuing aggressive effort on the part of the scientific community, and the knowledge gained from the use of animal models will help guide clinical diagnoses and provide direction for appropriate treatment strategies.

The work on which this chapter is based was supported in part by NIH Grant NIDCD R01 #DC04566 and NEI R01 #EY016247.

Deafness, in all of its forms, affects an estimated 186 of every 100,000 newborns in the United States ³⁷² and increases in prevalence with advancing age. For example, by the age of 60 years, one of every three people in the United States experiences communication difficulties as a consequence of lost hearing, and by the age of 85 years, one half of the population has some degree of hearing disability. In most cases, peripheral auditory disease involves some form of cochlear abnormality that disrupts sensory transduction, and various etiologic factors have been recognized, including genetic abnormalities, trauma (acoustic and mechanical), ototoxic agents, bacterial and viral infections, and disturbances in other organ systems that influence cochlear physiology indirectly. This chapter describes the physiologic mechanisms underlying cochlear transduction at both the systemic and molecular levels. In addition, basic mechanisms of cochlear transduction are correlated with a subset of inner ear disorders that are representative of various classes of disease.

Significant breakthroughs in the identification of genes and gene products that serve as integral elements of cochlear transduction have been made beginning in the mid to late 1990s. Clearly, pathogenic mutations of many of these genes are implicated in various forms of deafness (Table 146-1). Although numerous cases of inherited deafness had been identified historically on the basis of phenotype, the etiopathogenic mechanisms responsible for deafness remained unknown before technologic advances allowed for the identification and cloning of genes associated with normal hearing. By comparing human genes with orthologous genes that are known to cause deafness in animals, the cell biology and disease associated with particular gene mutations can be studied, leading to an understanding of the role of gene products in sensory transduction. Although a variety of nonhuman species have played indispensable roles in the study of cochlear transduction and sensory pathology, in recent years the mouse model has emerged as particularly useful (Table 146-2; see also Table 146-1). This movement is fueled, in large part, by the highly successful, comprehensive investigation of the murine genome and the ever-expanding availability of tools that allow the targeting and manipulation of specific genes, technologic changes that greatly extend the capability of researchers to examine the role of gene products in the transduction process. However, although much has been learned about the molecular and genetic basis of cochlear function, or malfunction as the case may be, much work remains to be done in the ongoing effort to unravel the intricacies associated with cochlear transduction and deafness.

Table 146-1 Selected Genes Underlying Various Forms of Deafness in Humans

Table 146-2 Selected Genes Underlying Murine Deafness for Which Human Deafness Has Not Been Determined

Approximately 68% of all cases of congenital hearing loss are inherited, and of these, at least 70% are nonsyndromic—that is, auditory disability is the only manifestation of the disorder.³⁷² Nonsyndromic hearing losses are grouped according to their inheritance mode: DFNA refers to autosomal dominant cases, DFNB to autosomal recessive, and DFN to X-linked or mitochondrial cases; a unique numeral following each acronym has been used to identify each chromosomal locus discovered in chronologic order. As a consequence of a major effort to identify and classify inherited hearing disabilities, it is clear that the nonsyndromic basis of inherited deafness is highly heterogeneous. Many loci known to produce autosomal dominant forms of deafness when mutated have been genetically mapped, although only a subset of the genes responsible for autosomal dominant forms have been identified and cloned. Autosomal recessive forms of inherited hearing loss account for most nonsyndromic abnormalities (approximately 80%). It has become increasingly clear, however, that different mutations in a given gene may be responsible for differences in the severity of hearing loss, different patterns of inheritance, and susceptibility to environmental factors such as noise or aminoglycoside antibiotics, as well as the involvement of other organ systems. For example, mutations in a number of genes (e.g., COL11A2, ESPN, GJB2, GJB3, GJB6, MYO6, MYO7A, PJVK, TECTA, TMC1) are responsible for both autosomal dominant and autosomal recessive forms of nonsyndromic deafness. Moreover, mutations in several genes (e.g., CDH23, COL11A2, GJB2, GJB3, GJB6, MYH9, MYO7A, MYO15A, PCDH15, SLC26A4, USH1C, WFS1, WHRN) are responsible for both nonsyndromic and syndromic forms of deafness (see Table 146-1). The phenotypic spectrum produced by mutations in a gene has been attributed to the severity of the mutation, allelic heterogeneity, alternative splicing, and the contribution of modifier genes.^60,^136,^171,^254,^336,^362,⁴⁸⁶

This chapter reviews basic mechanisms underlying sensory transduction associated with hearing and integrates contemporary models of transduction with the latest acquired information regarding the molecular structure and organization of key inner ear systems. Because of the sheer number of factors now known to be necessary for normal inner ear development, maintenance, and function, a comprehensive accounting is not possible here—not all such factors are addressed. Of note, the stated biologic role of most of the proteins discussed here remains speculative, and substantial work is needed to determine their precise role in the inner ear, as well as the nature of their interactions with other proteins. Excellent reviews that are relevant to the subject of this chapter are listed in the Bibliographic References.^29,^61,^170,^187,^215,^236,^310,^320,^443,^453,^585,^588,⁶³¹

Passive Cochlear Mechanics

The cochlea, and the organ of Corti in particular, constitute an evolutionary marvel of biologic engineering as intricate, microscaled mechanical systems. Although the hallmark of peripheral auditory function in mammals may be the evolutionary adaptation of a relatively simple, passive resonant system into an active one that consumes energy and efficiently detects and amplifies vibratory energy, it is useful to note that the fundamental nature of cochlear mechanics is evident even in cadavers, as so elegantly demonstrated by Nobel prize winner Georg von Békésy.⁵⁸⁶ Because cochlear mechanics at this level of function are independent of other factors—neither requiring nor consuming adenosine triphosphate (ATP), for example—the sound-driven motion of the organ of Corti observed by von Békésy and others is commonly referred to as passive.

The primary elements of passive cochlear transduction are hydromechanical in character, and the integrated contributions of many structural elements are known to determine the resonant character of cadaveric and therefore mechanically passive cochleae (Fig. 146-1). The cochlear partition is composed of an epithelial sheet known as the tympanic cover layer that supports the fibroelastic basilar membrane, itself the central structure of the end organ both functionally and anatomically. This partition supports, in turn, both the sensory and nonsensory supporting epithelium. The fluid-filled major chambers of the cochlea—specifically, the scala vestibuli and scala tympani, which are filled with perilymph, and the scala media, which is filled with endolymph—also play an important role in transduction. These structural elements, together with the tectorial membrane (an extracellular gelatinous structure secreted by nonsensory inner ear epithelial cells) that extends from the spiral limbus radially over the apical surface of the organ of Corti, comprise the principal elements that give rise to passive mechanical transduction. Although Reissner’s membrane separates scala media from scala vestibuli and therefore plays an important homeostatic role in the maintenance of cochlear electrochemistry, it does not seem to play a significant role as an element of passive cochlear mechanics.

Figure 146-1. A, Midmodiolar section through a human cochlea showing the basal (lower), middle, and apical (upper) turns. The modiolus is the central core of the cochlea and houses spiral ganglion neurons. The cochlear duct, otherwise known as the scala media (SM), contains endolymph and is separated from the perilymphatic fluid chambers, scala tympani (ST), and scala vestibuli (SV) by the cochlear partition and Reissner’s membrane, respectively. At the cochlear apex, the fluids of the ST and SV mix through an opening called the helicotrema. B, Higher-magnification image of a modiolar section through the basal turn of the cochlea of a Rhesus monkey. The organ of Corti rests on the cochlear partition, a structure that includes the basilar membrane (BM) and a layer of mesothelial cells lining the BM and facing the ST. The stria vascularis (StV) and spiral ligament (SLig) form the lateral boundary of SM. C, Cross section of the organ of Corti from the middle turn of a cat cochlea. The triangular-shaped tunnel of Corti (TC), formed by the inner and outer pillar cells, separates the inner hair cell (IHC) from the outer hair cells (OHCs). The tectorial membrane (TM) can be seen pulled away from the sensory cells. D, Representation of the organ of Corti, showing the one row of IHCs and three rows of OHCs. Note the fluid spaces surrounding the OHCs, and that the tallest stereocilia of OHCs are embedded in the tectorial membrane, whereas the stereocilia of IHCs do not reach the tectorial membrane.

Rights were not granted to include this figure in electronic media. Please refer to the printed book.

(A, Courtesy of Dr. William B. Warr; B, from Engstrom H, Angelborg C. Morphology of the walls of the cochlear duct. In: Zwicker E, Terhardt E, eds. Facts and Models in Hearing. New York: Springer-Verlag; 1974:3; C, from Engstrom H, Ades HW, Andersson A. Structural Pattern of the Organ of Corti. Baltimore: Williams & Wilkins; 1966; D, from Dallos P, Fakler B. Prestin, a new type of motor protein. Nat Rev Mol Cell Biol. 2002;3:104.)

Because the focus of this chapter is on cochlear transduction, we will only remind the reader that sound waves collected in the external ear canal produce vibrations of the tympanic membrane, which in turn vibrate middle ear ossicles (Fig. 146-2A). It is, of course, the vibratory motion of the stapedial footplate that transmits the mechanical energy of the ossicular chain directly through the oval window of the cochlea, effectively delivering sound pressure waves to the scala vestibuli and translating mechanical motion into pressure waves that propagate through the virtually incompressible cochlear fluids at a velocity approximating 1.5 km/sec. At this rate of propagation, the spread of the pressure wave throughout the volume of the cochlea is nearly instantaneous. Because the bony walls of the cochlea are rigid and its fluid contents incompressible, the higher pressure (or lower, depending on the direction of motion of the stapedial footplate) in scala vestibuli relative to scala tympani produces a pressure differential across the cochlear partition that creates intracochlear forces that set the partition into motion (see Fig. 146-2D). As described in more detail in a later section, the physical makeup of the organ of Corti, and of the basilar membrane in particular, establishes a space-frequency map that determines the limits of hearing and the capacity of the subject to resolve frequency differences among vibrations that contain energy in the audible range.

Figure 146-2. A, Schematic illustration of the inner ear with the cochlea uncoiled revealing the basilar membrane (BM) and a traveling wave at one instant in time in response to a pure tone. The eardrum (tympanic membrane) and middle ear ossicles also are shown in relation to the semicircular canals. The oval window opens into scala vestibuli (SV) and the round window to scala tympani (ST). B, Drawing of the inner ear with a segment of a cochlear turn extracted to expose the scalae. C, Diagram of a section through the cochlea showing the location of inner and outer hair cells (IHC and OHCs, respectively) in relation to the BM and TM and the direction of wave propagation. SM, scala media. D, The ratio of pressures measured within the SV and ST to the pressure at the eardrum produced in response to tone bursts are plotted as a function of stimulus frequency; responses were recorded from a cat. The difference between SV and ST pressures is responsible for displacement of the basilar membrane. E, BM displacements produced in a cadaveric human cochlea in response to 200 Hz at four separate points in time. The envelope of the traveling wave also is indicated. F, Envelopes of traveling waves measured along the BM in a cadaveric human cochlea in response to four stimulus frequencies, showing the relationship between the location of peak displacement and frequency (higher frequencies produce peaks at locations progressively closer to the base).

(A, From Dallos P, Fakler B. Nat Rev Md Cell Biol. 2002;3:104; B, from Holme RH, Steel KP. Genes involved in deafness. Curr Opin Genet Dev. 1999;9:309; C, adapted from Ashmore JF, Kolston PJ. Hair cell based amplification in the cochlea. Curr Opin Neurobiol. 1994;4:503; D, from Nedzelnitsky V. Sound pressures in the basal turn of the cat cochlea. J Acoust Soc Am. 1980;68:1676; E and F, from von Bekesy G. Experiments in Hearing. New York: McGraw-Hill; 1960.)

Physical Dimensions

At the basal end of the cochlea near the round window, the width of the basilar membrane, as measured between the tympanic lip of the spiral limbus and the basilar crest of the spiral ligament, is narrow relative to dimensions representing more apical locations (Fig. 146-3). In fact, the width of the basilar membrane progressively increases along a base-to-apex gradient in mammals, such that in humans, for example (see Fig. 146-3 E), the width at the base is approximately 100 µm and approaches 500 µm near the apex.⁶⁰⁴ In association with graded basilar membrane width, the number of mesothelial cells, or tympanic border cells, lining the basilar membrane increases along the basoapical length of the end organ, as does the length of outer hair cells (OHCs) and associated stereocilia. Conversely, the thickness of the basilar membrane and the density of radial filaments embedded in the homogenous ground matrix of the partition decrease progressively from base to apex (see Fig. 146-3C). It is notable that radial filaments are composed primarily of collagen type II,^132,^545,⁵⁵⁸ whereas the ground matrix of the partition itself is composed primarily of fibronectin, tenascin, proteoglycans containing keratin sulfate and chondroitin sulfate, and emilin-2 (elastin microfibril interface-located protein-2).^21,^375,^475,^539,⁶¹¹ Consequently, the basilar membrane and organ of Corti complex is stiffer and less massive at the base than at the apex (see Fig. 146-3F). The result is an end organ with resonant properties that are continuously distributed from base to apex that supports a mechanical system that separates, or filters, complex fluid waves into sinusoidal constituents. As a result, high-frequency acoustic events are preferentially transduced in the base (see Fig. 146-3G), because it is stiff and less massive. The inverse is true in the case of low-frequency acoustic events.

Figure 146-3. A, Photograph of the cochlea of a guinea pig after the outer bony shell has been thinned. B, Scanning electron micrograph (SEM) of a chinchilla cochlea after the bone on one side has been removed. The bony modiolus (M) is shown in the center, and the fluid-filled scala tympani (ST) and scala vestibuli (SV) are indicated in the basal turn. The helicotrema (H) is shown at the apex, and the arrows indicate the osseous spiral lamina to which the cochlear partition is attached at its medial extent. The round window (RW) and the stapes (S) are indicated, and the footplate of the stapes has been pulled away slightly, exposing the oval window (OW). C, Schematic illustration of the variation in thickness of the basilar membrane (dark blue area) and the associated mesothelial layer (light blue area) and the width of the basilar membrane (horizontal axis) at six locations along the cochlear spiral of the cat; specified locations are relative to the base. The demarcation between the arcuate and pectinate zones is indicated by the left arrow at each location; the arcuate zone (pars tecta) extends from the lip of the osseous spiral lamina to the foot of the outer pillar cell (OPC) and the pectinate zone from the OPC to the basilar crest of the spiral ligament (right arrows). D, Scaled representation of the ribbon-like nature of the human basilar membrane indicating its width relative to its length. E, Quantitative relationship between the width of the basilar membrane (BM) and the distance from the base for the human,⁶⁰⁴ cat,⁶⁴ guinea pig,¹⁵⁴ and gerbil.⁴⁴⁵ F, Estimates of the stiffness of the basilar membrane in the guinea pig²⁰⁵ and gerbil¹³⁹ as a function of location. G, Maps of the characteristic frequency (CF) of a cochlear location (i.e., peak of the envelope of the traveling wave) as a function of position along the basilar membrane for several species. Note that the lengths of the basilar membrane for each species can be estimated by the distance that each curve occupies along the x-axis. Frequency-position maps were based on a formulation developed by Greenwood¹⁹⁶ using empirical fits to data for the species shown here, except for rat, which was developed by Muller.³⁷⁴

(A, From Wasterstrom SA. Accumulation of drugs on inner ear melanin. Therapeutic and ototoxic mechanisms. Scand Audiol Suppl. 1984;23:1; B, from Harrison RV, Hunter-Duvar IM. An anatomical tour of the cochlea. In: Jahn AF, Santos-Sacchi J, eds. Physiology of the Ear. New York: Raven Press; 1988:160; C, from Cabezudo LM. The ultrastructure of the basilar membrane in the cat. Acta Otolaryngol. 1978;86:160; D, from Wever EG. Theory of Hearing. New York: Dover Publications; 1949.)

Another physical characteristic of the cochlea that exerts influence over the mechanics of transduction is the relative volume of perilymphatic fluid compartments that are larger at the base than at the apex (see Fig. 146-1). Consequently, there is a clear difference in fluid-mass loading of basilar membrane motion along the length of the end organ. The magnitude of the intracochlear fluid load is thought to dominate longitudinal coupling along the length of the cochlea. Of note, very little coupling occurs otherwise between adjacent regions of the basilar membrane—beyond, that is, simple stretching and compression forces.^204,^394,^395,⁵⁸³ At rest, therefore, the basilar membrane seems to be under very little tension, and without direct mechanical coupling, the traveling wave cannot propagate further apically after reaching its resonant peak, or place. Moreover, the viscous friction between the basilar membrane and cochlear fluids produces a damping effect that limits its displacement.

The Traveling Wave

The hydromechanical consequences of introducing a pressure pulse, or wave, into perilymphatic fluids is a traveling wave that propagates along the organ of Corti from the base to the apex (see Fig. 146-2). The traveling wave propagates basoapically, because the stiffness of the cochlear partition decreases longitudinally from the base to the apex (see Fig. 146-3F) as its mass increases. The motion and direction of the wave that travels along the basilar membrane are independent of the location of the vibratory source. von Békésy⁵⁸⁶ observed this feature of cochlear mechanics by introducing a pressure pulse through an artificial window carved into the bony labyrinth near the cochlear apex. Traveling waves created under these conditions are essentially indistinguishable from traveling waves triggered by ossicular vibration. The basoapical propagation of the traveling wave is an outcome of the end organ’s physical makeup, with the stiffer and less massive elements forming the base of the organ of Corti being set into motion by an instantaneous pressure difference between the major scalae before the more compliant and massive sections of the apical cochlea. The resultant traveling wave reflects the more or less independent, yet orderly, sequential movement of loosely coupled basilar membrane segments that form a frequency decomposition system generally thought of as the space-frequency or tonotopic map (see Fig. 146-3G).

As the traveling wave propagates toward the apex, traveling initially at a high rate of speed, the velocity of the wave decreases, with production of progressively shorter wavelengths for any given input frequency as a consequence (see Fig. 146-2E). As the amplitude of the propagating wave steadily increases as it passes toward the apex, partition displacement reaches a maximum value at its characteristic place—where intrinsic resonance meets, and nearly matches, the vibratory frequency of the triggering acoustic event (see Fig. 146-2F). As pointed out previously, the stiffness of the basilar membrane decreases exponentially with distance from the stapes,^205,^394,^395,⁵⁸⁶ so it is not surprising that the resonant frequency also decreases exponentially as a function of distance from the stapes, because the characteristic place is determined by the stiffness of the cochlear partition. Thus, the aforementioned space-frequency, or tonotopic, map is created, whereby the peak in vibration amplitude is located at the base of the cochlea for high frequencies and at progressively apical regions for lower frequencies.

Just as stiffness dominates the physics of the traveling wave as it travels toward its characteristic place, basilar membrane motion at and just above the characteristic place is dominated by the mass of the system. It is important to point out that the actual peak of the traveling wave occurs at a location just basal to the true resonant place, characterized by a match between intrinsic resonance and the triggering vibratory energy. At the place of peak displacement, the traveling wave slows to an essential standstill (i.e., it stalls) before remaining energy is completely dissipated, and the partition is completely and rapidly restored to its resting position slightly apical to the resonant place. Near the characteristic place for a given stimulus frequency, the pressure wave returns to the basal end of the cochlea through the fluids of the scala tympani, where the flexible round window membrane is displaced in an equal but opposite direction to that of the volume displacement of the stapes footplate.

At the apical end of the cochlea, the endolymphatic space terminates just short of the bony labyrinthine wall and maintains a firm physical boundary between endolymph and perilymph, allowing the fluids in the scala vestibuli and scala tympani to communicate directly by way of a small duct called the helicotrema (see Fig. 146-3B). The helicotrema acts as an acoustic shunt across the cochlear partition that reduces the pressure difference between the scalae produced by very-low-frequency stimulation.⁹⁶ The size of the helicotrema is thought to determine the low frequency cutoff of the system; that is, large openings shunt pressure waves extending to higher frequencies than those associated with small helicotrema. The shunting effect of helicotrema reduces the pressure differential across the cochlear partition, and it is this aspect of cochlear design that is thought to decrease damage to the cochlea produced by intense low-frequency pressure fluctuations.⁴⁰³

Active Cochlear Mechanics

Von Békésy recognized more than half a century ago that the envelope of the traveling wave measured in cochleae from human cadavers is broadly tuned—so broadly tuned as to be inconsistent with a living person’s ability to discriminate closely spaced frequencies. Numerous groups of investigators have since confirmed the findings of von Békésy and have shown with clarity that sharp tuning and high sensitivity are lost in physiologically compromised cochleae or cochleae from cadavers.^432,⁴⁹⁴ Only after mechanical measurements were made in vivo with nonhuman species did it become clear that the peak of the traveling wave in living animals is, in fact, very sharply tuned at low levels of stimulation and exhibits nonlinear growth as sound levels increase.^274,^314,^433,^454,⁴⁹⁴ The seminal study of Rhode,⁴³³ which showed the amplitude of basilar membrane vibration relative to stapes displacement to be as much as two to three orders of magnitude greater under low-stimulus-level conditions than at higher levels in normal animals, led most directly to the contemporary understanding of cochlear mechanics as a nonlinear phenomenon (i.e., high sensitivity to changes in stimulus pressure at low levels, marked compression in the midlevel range, and linear growth at high levels) (Fig. 146-4).

Figure 146-4. A, Schematic representation of traveling waves propagating along the basilar membrane (BM) in response to a pure tone. When the cochlear amplifier is functioning, outer hair cells (OHCs) within a restricted region of the traveling wave peak exert a force that enhances BM motion relative to that observed during passive mechanics alone for low-level stimuli. B, Similar to A, except note that the peak of the traveling wave produced with cochlear amplification is much narrower than the peak produced during passive mechanics alone, which is important for frequency discrimination. Also note that during active amplification, the peak of the traveling wave occurs slightly more basal than that produced during passive mechanics. C, Schematic relationship between stimulus level and BM displacement. When OHCs are normal, stimulus level increments of less than 40 dB sound pressure level (SPL) produce a linear increase in BM displacement. Between approximately 40 and 80 dB SPL, OHC responses saturate, producing a compressed response and nonlinear growth. Above 80 dB SPL, growth becomes linear again. When OHCs are damaged, displacements occur only to higher level stimuli and exhibit linear growth. D, The relationship between stimulus frequency (measured in octaves relative to the characteristic frequency [oct re CF] at that location) and the level required to elicit a criterion or “threshold” response at a given point along the BM (e.g., tuning curves). When the cochlear amplifier is operational, thresholds may be 40 to 60 dB lower and frequency selectivity greatly enhanced compared with a passive cochlea.

(A, From Ashmore JF, Kolston PJ. Hair cell based amplification in the cochlea. Curr Opin Neurobiol. 1994;4:503. B, from Gummer AW, Preyer S. Cochlear amplification and its pathology: emphasis on the role of the tectorial membrane. Ear Nose Throat J. 1997;76:151.)

Active mechanics associated with basilar membrane amplification is a metabolically labile, energy-consuming process that is, consequently, physiologically vulnerable. It is of particular importance to recognize that the active mechanical event underlying amplification is highly localized. That is, the principle of active mechanics applies to a circumscribed portion of the cochlear partition that is limited to the immediate vicinity of the characteristic place for a given stimulus frequency (see Fig. 146-4A and B). At frequencies removed from the characteristic place, the properties of cochlear partition vibration in a living animal follow linear rules of growth, as in cadaverous tissue (Fig. 146-5).¹⁷⁹ Another important point is that the traveling wave peaks at increasingly apical regions of the cochlear partition with increasing stimulus level, shifting the cochlear place-frequency map toward lower frequencies at any given place.

Figure 146-5. Measurements of basilar membrane (BM) motion in a healthy chinchilla cochlea at one location near the base in response to tone bursts. A, Measurement of BM velocity in response to tone bursts of different frequencies (abscissa) at various levels (parameter). The characteristic frequency (CF) at this location is 9 kHz. SPL, sound pressure level. B, Tuning curves (i.e., stimulus levels required to generate a criterion response as a function of stimulus frequency) are shown for a criterion BM velocity of 0.1 mm/sec and for a criterion displacement amplitude of 1.77 nm. Data were taken from A. A tuning curve from an auditory nerve fiber that innervates this BM location is also shown. C, BM velocities as a function of stimulus level for various frequencies. A line showing a linear relationship between velocity and level also is shown. D, The ratio between BM velocity and stapes velocity (i.e., gain) as a function of stimulus frequency. Note the high gain (i.e., amplification) at low stimulus levels near the CF location that is reduced as a stimulus level increases.

(A and B, From Ruggero MA, Rich MC. Application of a commercially-manufactured Doppler-shift laser velocimeter to the measurement of basilar-membrane vibration. Hear Res. 1991;51:215; C, from Ruggero MA. Responses to sound of the basilar membrane of the mammalian cochlea. Curr Opin Neurobiol. 1992;2:449; D, from Ruggero MA, Robles L, Rich NC, Recio A. Basilar membrane responses to two-tone and broadband stimuli. Philos Trans R Soc Lond B Biol Sci. 1992;336. The figure was prepared by Geisler CD. From Sound to Synapse: Physiology of the Mammalian Ear. New York: Oxford University Press; 1998.)

Role of Outer Hair Cells in Active Mechanics

Many early studies indicated that OHCs play a fundamental role in active cochlear mechanics, as indicated by estimates of diminished cochlear sensitivity and frequency selectivity and loss of nonlinear operating properties when OHCs are missing from the end organ (see Fig. 146-4). The identification of specializations that distinguish inner hair cells (IHCs) from OHCs of the mammalian cochlea offered early clues that the contribution of OHCs to the transduction process is unique (Fig. 146-6). For example, only approximately 5% of all auditory nerve fibers innervate OHCs,^371,^525,⁵²⁴ yet there are at least three times as many OHCs as IHCs (Fig. 146-7). These findings suggested that the transmission of sensory information to the central nervous system (CNS) probably is not the primary function of OHCs. In studies of acute preparations, investigators noted that the cochlea, and OHCs in particular, are highly vulnerable to hypoxia or other forms of insult.^76,^148,^293,^432,^433,^448,^449,⁴⁹⁴ In animal models of chronic damage, OHCs were found to be more susceptible to noise-induced or aminoglycoside-induced damage than IHCs, and when OHCs were lost, with IHCs appearing unaffected, cochlear sensitivity was significantly reduced, as was frequency selectivity, and input-output curves acquired a more linear character in the region of OHC damage.^77,^97,^101,^149,^208,^278,^279,³²⁹ ^,450 ^,465 In addition, other cochlear nonlinearities, such as two-tone suppression (i.e., the reduction in the response to one tone elicited by the presence of another) and distortion tones (i.e., the appearance of additional tones not present in the stimulus produced by the interaction of two or more primary tones, such as 2f₁-f₂ or f₂-f₁, where f₁ and f₂ represent the primary tones, with f₁ less than f₂), are either lost or diminished when OHCs are traumatized.¹⁰² ^,480 The discovery of sound-evoked otoacoustic emissions, sounds that are generated by the end organ during transduction and are measured in the external ear canal,²⁷⁰ along with spontaneous otoacoustic emissions,⁶⁰⁸ and their dependence on OHCs provided additional evidence that an active, OHC-driven mechanism operates within the cochlea, which is capable of producing energy and propagating it in the reverse direction. The influence of OHCs on cochlear mechanics was further suggested on the basis of observations made after stimulating the crossed olivocochlear bundle (OCB), a brainstem pathway populated by large, myelinated fibers that arise among neurons located in the medial region of the superior olivary complex that cross the brainstem at the floor of the fourth ventricle and terminate in calyx-like presynaptic terminals on OHCs.^200,^287,^427,^428,^518,⁵⁹⁷ ^,598 Specifically, OCB activation altered cochlear output,^118,^155,^156,^176,⁶⁰⁵ as well as otoacoustic emissions^373,⁵⁰⁵ and basilar membrane displacement,³⁷⁶ supporting the notion that OHCs directly influence cochlear mechanics. Further support for the view that OHCs play an essential role in active mechanics comes from the work of Mammano and Ashmore,³⁴⁹ who demonstrated that outer hair cell motility induced by electrical stimulation of the excised cochlea displaced both the reticular lamina and the basilar membrane.

Figure 146-6. A, Drawing of an outer hair cell (OHC) showing the cylindrical soma; the nucleus located near the basal end of the cell; the cuticular plate at the apex of the cell where the rootlets of the stereocilia insert; the subsurface cisternae, which are stacks of smooth endoplasmic reticulum that line the lateral sides of the cell; and Hensen’s body, which is a continuation of the endoplasmic reticulum, located beneath the cuticular plate. This structure probably functions as a Golgi apparatus. B, Transmission electron microscopy of the lateral membrane of a guinea pig OHC, showing the plasma membrane (pm), the cortical lattice (cl) just beneath the membrane, the lateral cisterns (lc), and mitochondria (m) just inside the cisterns. Scale bar is 200 nm. C, Diagram of the OHC lateral membrane, where the inner and outer leaflets of the plasma membrane have been separated revealing a high density of membrane particles not present in IHCs, which presumably are oligomers of the motor protein prestin. The cortical lattice is composed of many subdomains of parallel actin filaments oriented circumferentially around the cell that are cross-linked by spectrin. The actin filaments are bound to the plasma membrane by pillar molecules (composition unknown), and the subsurface cisternae lie just inside the cortical lattice. D, Illustration of the expanded apical surfaces of the inner and outer pillar cells (IP and OP, respectively) in relation to the apical surfaces of OHCs and the phalangeal processes of Deiters cells (D1, D2, D3) forming the reticular lamina. E, Diagram of the orientation of several subdomains of membrane particles (left) and an array of particles that undergo conformational changes that alter particle packing density associated with a change in transmembrane voltage. During depolarization, tighter packing results in shortening of the subdomain, as well as the hair cell itself, and during hyperpolarization, elongation occurs.

Rights were not granted to include this figure in electronic media. Please refer to the printed book.

(A, from Lim DJ. Functional structure of the organ of Corti: a review. Hear Res. 1986;22:117; B, from Holley MC. Outer hair cell motility. In: Dallos P, Popper AN, Fay RR, eds. The Cochlea. New York: Springer-Verlag; 1996:386; C, adapted from Oghalai, et al. J Neurosci. 1998;18:48; D, from Slepecky NB. Structure of the mammalian cochlea. In: Dallos P, Popper AN, Fay RR, eds. The Cochlea. New York: Springer-Verlag; 1996:44; E, from Frolenkov GI, Atzori M, Kalinec F, Mammano F, Kachar B. The membrane-based mechanism of cell motility in cochlear outer hair cells. Mol Biol Cell. 1998;9:1961.)

Figure 146-7. A, Scanning electron micrograph of the surface of the organ of Corti of a chinchilla showing the three rows of outer hair cells (OH₁, OH₂, OH₃), one row of inner hair cells (IH), inner and outer pillar cells (IP and OP, respectively), inner phalangeal cell (IPH), Deiters cells (D₁, D₂, D₃), and Hensen cells (H). B, Schematic illustration of the structure of the organ of Corti: 1, basilar membrane; 2, Hensen cells; 3, outer phalangeal cells; 4, nerve endings; 5, outer hair cells; 6, outer spiral fibers; 7, outer pillar cells; 8, tunnel of Corti; 9, inner pillar cells; 10, inner phalangeal cells; 11, border cells; 12, tectorial membrane; 13, type I spiral ganglion cell (SGC); 14, type II SGC; 15, bony spiral lamina; 16, spiral blood vessel; 17, spindle cells; 18, auditory nerve fibers; 19, radial fiber.

(A, From Lim DJ. Functional structure of the organ of Corti: a review. Hear Res. 1986;22:117; B, from Kiang NY: Peripheral neural processing of auditory information. In: Darian-Smith I, ed. Handbook of Physiology. New York: Oxford; 1984:639.)

Prestin (SLC26A5) Is Essential for Outer Hair Cell Motility and Active Transduction

The finding that OHCs contract or elongate at very high rates (up to at least 70 kHz) along the length of their longitudinal axes when depolarized or hyperpolarized, respectively,^30,^62,^63,^98,^99,^168,^177,^261,^478,⁶³⁰ led to the suggestion that OHCs may amplify the displacement of the basilar membrane if forces generated by their motility are produced on a cycle-by-cycle basis and precisely timed relative to passive mechanical events (see Fig. 146-6E). Somatic length changes of up to 5% of the total length of OHCs are determined directly by the magnitude of the transmembrane voltage and not by ionic currents. In addition, the voltage-motility relationship observed among OHCs is highly nonlinear (Fig. 146-8B). Unlike in IHCs, the lateral wall of OHCs is densely packed with particles that are integral components of the plasma membrane,^165,^202,^264,⁴⁷⁰ and these particles are thought to be oligomers of the protein recently identified as the molecular motor underlying fast somatic motility (see Fig. 146-6C and E).

Figure 146-8. A, Schematic representation of the membrane topology of prestin (SLC26A5). Prestin consists of 744 amino acids and is predicted to have 12 transmembrane domains and intracellular amino-terminal (NH₂) and carboxyl-terminal (COOH) ends. Circles and diamonds represent positively and negatively charged amino acids, respectively. B, Change in the length of an isolated outer hair cell (OHC) as a function of voltage stepped from a holding potential of −68.4 mV. The inset is an illustration of single prestin molecules located in the plasma membrane that can toggle between two conformational states, depending on transmembrane voltage. The larger the depolarization, the greater likelihood that the shorter state will be adopted, and vice versa. The nonlinear capacitance associated with length changes may be due to the translocation of anions, such as Cl⁻ (yellow circles), that move closer to the extracellular side during hyperpolarization and toward the intracellular side during depolarization. C, Targeted disruption of prestin results in severely elevated auditory brainstem response (ABR) thresholds in homozygous mice, and somatic motility is absent in isolated OHCs (D).

(A, From Dallos P, Fakler B. Prestin, a new type of motor protein. Nat Rev Mol Cell Biol. 2002;3:104; Reprinted by permission from Macmillan Publishers Ltd. B, from Santos-Sacchi J. On the frequency limit and phase of outer hair cell motility: effects of the membrane filter. J Neurosci. 1992;12:1906; inset, prepared by M. Walsh; C, and D, adapted from Liberman MC, Gao J, He DZ, et al. Prestin is required for electromotility of the outer hair cell and for the cochlear amplifier. Nature. 2002;419:300.)

The discovery that OHC motility is accompanied by gating currents that reflect a nonlinear capacitance led to the suggestion that it was the movement of charged particles within the domain of the plasma membrane that operates as the motive force of contractility.^31,⁴⁷⁶ The voltage-sensitive motor protein underlying the process was identified in 2000 and given the name prestin (SLC26A5). Prestin (Fig. 146-8A) is a modified anion-transporter protein belonging to the solute carrier 26 (SLC) family of proteins.⁶³³ Antibodies to prestin have been localized in OHC membranes, and developmental changes in the expression of prestin immunoreactivity follow a temporal profile that overlaps the development of electromotility, providing further support for the view that prestin is the motor protein.⁴² Although the precise location of the voltage sensor remains unknown, and although it may be an integral component of the prestin molecule itself, findings from recent studies suggest that intracellular Cl⁻ ions act as extrinsic charged voltage sensors (see Fig. 146-8B, inset). In this model, Cl⁻ ions occupy a position in the intracellular domain of the membrane-bound prestin protein and move along an inward electrical gradient away from the extracellular domain during depolarization. This charge movement produces, in turn, a conformational change in the protein that reduces its surface area and shortens the cell; the opposite action occurs during hyperpolarization, when Cl⁻ ions move toward the extracellular space within the domain of the protein (although the anion is never translocated to the extracellular space), causing an increase in surface area and lengthening of the cell.^100,³⁹¹ Salicylates, which are known to cause temporary hearing loss when administered at high doses,⁵³⁵ are permeable to the plasma membrane and act as competitive antagonists of Cl⁻, thereby reducing electromotility and, presumably, diminishing sensitivity.³⁹¹ The role of prestin in transduction was studied directly in a mouse strain in which prestin was deleted. The knockout mouse exhibited a loss of OHC electromotility in homozygous mutant animals (see Fig. 146-8D), a 40- to 60-dB loss in sensitivity, and loss of OHCs in the base of the cochlea (see Fig. 146-8C).³²⁸ In addition, the frequency selectivity of individual auditory nerve fibers is diminished in knockout mice.⁶⁸ Although the length and stiffness of OHCs are reduced in the knockout mouse, these findings collectively suggest that active transduction is a direct product of OHC somatic motility. Confirmation of prestin’s role in cochlear amplification was recently established in a study of a “knock-in” mouse model, in which the protein’s function was diminished but OHCs exhibited normal length and stiffness.¹⁰³ Natural mutations of the prestin gene have been identified in humans, and the resulting condition is transmitted as a recessive, nonsyndromic form of congenital deafness. In this case, however, hearing loss is in the severe to profound category (i.e., on the order of 90 dB or greater) and exists as a bilateral form of deafness among persons who are homozygous for the trait. The degree of hearing loss is more variable among heterozygotes, ranging from normal in one case to mild (40 dB) and even moderate to profound (60 dB) in others, suggesting the semidominance of the abnormal allele through haploinsufficiency.³³⁷

Conversion of Basilar Membrane Displacement to Radial Shearing Forces

All existing models of transduction center on the idea that displacement of the basilar membrane results in a radial shearing motion between the reticular lamina and the tectorial membrane, a motion that serves as the mechanical trigger of transduction currents. The reticular lamina consists of the interdigitating heads of the inner and outer pillar cells, the apical surfaces of the phalangeal processes of Deiters cells, and the apical surfaces of the hair cells through which the stereocilia project (see Fig. 146-6D). The apical surfaces of the supporting cells are flattened and are composed of microtubules and actin microfilaments, lending support to the idea that the reticular lamina is a relatively rigid structure that moves in concert with basilar membrane motion. The tectorial membrane is firmly attached to the spiral limbus and overlies the reticular lamina, with its tip or marginal zone connected to Hensen cells. The tallest row of stereocilia protrude from the apical surface of each outer hair cell and embed within the inferior surface of the tectorial membrane along the entire length of the cochlear spiral.^288,³³³ Bundles of radially oriented filaments, consisting of type II, V, IX, and XI collagens, are embedded within the gelatinous matrix of the mammalian tectorial membrane (see Fig. 146-10A and B).^209,^361,^444,^514,^515,^545,⁵⁴⁶ The presence of collagen enhances the rigidity of the structure, making it less compressible and less elastic, particularly in the radial dimension.^3,^199,⁴⁴⁶ These features support a system in which the shearing motion between the reticular lamina and the tectorial membrane causes stereocilia to bend in the direction of the modiolus or the spiral limbus, depending on whether the basilar membrane is displaced toward the scala tympani or scala vestibuli, respectively.

In contrast with OHCs, the stereocilia of IHCs do not appear to firmly contact the tectorial membrane (see Fig. 146-1D). Therefore, during basilar membrane vibration, the relevant mechanical stimulus to IHC stereocilia is thought to be the flow of endolymph within the subtectorial space, the narrow channel between the reticular lamina and tectorial membrane into which the stereocilia protrude.³⁸⁷ Fluid coupling is then thought to lead directly to the displacement of IHC stereocilia in proportion to basilar membrane velocity at low frequencies (less than 500 Hz) and to basilar membrane displacement at higher frequencies; sustained displacements of the basilar membrane, or very-low-frequency displacements, do not effectively bend IHC stereocilia.⁴⁰³

Radial Displacement Patterns of the Basilar Membrane

The dynamics of basilar membrane displacement at the characteristic frequency place are more complex when examined radially. Inner and outer pillar cells that form the fluid-filled tunnel of Corti are packed with bundles of microtubules and actin microfilaments arranged in parallel that are cross-linked and are thought to produce structural rigidity within the cochlear partition.^25,^140,²⁴⁴ The foot of the inner pillar cell lies near the bony spiral lamina, whereas the foot of the outer pillar cell lies over the basilar membrane and is not supported by bone (Fig. 146-9). Therefore, when the basilar membrane is displaced in a passive cochlea, movement occurs maximally near the foot of the outer pillar cell, with the basilar membrane pivoting at the foot of the inner pillar cell.¹⁰⁴ When OHCs contract, the reticular lamina pivots at the apex of the tunnel of Corti, and the basilar membrane and reticular lamina are pulled together, enhancing overall basilar membrane displacement.^168,^349,^350,³⁸³ Furthermore, unlike in passive mechanical events, in which the full width of the basilar membrane moves uniformly (synchronously), during active mechanics, the magnitude and phase of displacement vary in the radial dimension.^83,^383,^388,^395,^396,^431,^495,^563,^609,⁶²² Although differences among studies from different laboratories are significant regarding the precise pattern of basilar membrane movement in the radial direction after displacement, many workers suggest that the position of greatest motion is in the domain of the OHC, and that motion in the OHC region is out of phase relative to the foot of the outer pillar cell. This finding has significant implications in the ongoing effort to understand the relationship between inner and outer hair cell responses during the transduction process.

Figure 146-9. Model of basilar membrane (BM) vibration across the radial extent of the membrane. A, As the BM moves toward scala vestibuli, at the point of maximum BM velocity, the shearing motion between the BM and tectorial membrane (TM) causes the stereocilia to bend in the excitatory direction, producing depolarization and contraction or shortening of the outer hair cells (OHCs), pulling the reticular lamina (RL) toward the BM. IHC, inner hair cell; IPL, inner pillar cell; OPC, outer pillar cell; OSL, osseus spiral lamina. B, As the BM moves toward the scala tympani, and at the point of maximum BM velocity, hair bundles are deflected in the inhibitory direction, causing hyperpolarization and elongation of the OHCs. Dotted lines indicate the positions of the BM and RL at rest. Note that the large red dot along the BM located at the foot of the OPC represents a pivot point, which allows the displacement of the BM beneath the tunnel of Corti to be out of phase with the displacement that occurs under the OHCs.

(Redrawn from Nilsen KE, Russell IJ. The spatial and temporal representation of a tone on the guinea pig basilar membrane. Proc Natl Acad Sci U S A. 2000;97:11751.)

Tectorial Membrane Pathophysiology

The tectorial membrane is composed of two primary matrix elements: a group of radially oriented collagen fibrils and a striated sheet matrix that is made up of tightly packed, small-diameter filaments that are interlinked by cross-bridges to give it a striated appearance.^209,^288,^300,^333,⁵⁵⁹ In addition to the variety of collagens making up the core structure, a collection of glycoproteins—α-tectorin, β-tectorin, and otogelin—also are constituent elements of the tectorial membrane.^78,^193,^285,^319,⁴⁴⁴ Otogelin, a glycoprotein that is found in essentially all of the acellular membranes of the inner ear, is associated with the large collagen fibrils, whereas α-tectorin and β-tectorin are components of the striated sheet matrix.

The mass and stiffness of the tectorial membrane are important variables that together determine the resonant properties of the structure that, in turn, affects its displacement pattern during acoustic stimulation. Much like dimensional changes that have been measured along the length of the basilar membrane, the cross-sectional area of the tectorial membrane has been shown to increase from the base to the apex as indicated by morphometric measurements of fully hydrated tissue specimens (Fig. 146-10C).⁴⁴⁵ Moreover, base-to-apex stiffness gradients in both the transverse and radial axes of the tectorial membrane have been observed,^199,⁴⁴⁶ and they are similar to those reported for basilar membrane stiffness gradients (see Fig. 146-10C).^139,²⁰⁵ Taken together, basoapical gradients in the physical properties of the tectorial membrane suggest the possibility that the structure may support longitudinally propagating traveling waves,¹⁸¹ giving rise to a second cochlear frequency-place map. Some investigators have suggested that the resonant frequency of the tectorial membrane lies approximately one-half octave below that of the basilar membrane’s characteristic or resonant frequency at any given position along the cochlear spiral.^17,^18,^203,⁶⁴¹ However, until the opportunity to study mutant mice in which the tectorial membrane is completely detached from the spiral limbus and the surface of the organ of Corti³¹⁷ presented itself, direct evidence regarding the precise role of the tectorial membrane as an element of transduction was largely unavailable.