the tectorial membrane: mechanical properties and functions

The Tectorial Membrane: Mechanical Propertiesand Functions

Jonathan B. Sellon,1 Roozbeh Ghaffari,1 and Dennis M. Freeman1,2

1Research Laboratory of Electronics, MIT, Cambridge, Massachusetts 021392Department of Electrical Engineering and Computer Science, MIT, Cambridge, Massachusetts 02139

Correspondence: [email protected]; [email protected]; [email protected]

The tectorial membrane (TM) is widely believed to play a critical role in determining theremarkable sensitivity and frequency selectivity that are hallmarks of mammalian hearing.Recently developed mouse models of human hearing disorders have provided new insightsinto themolecular, nanomechanicalmechanisms that underlie resonance and travelingwaveproperties of the TM. Herein we review recent experimental and theoretical results detailingTM morphology, local poroelastic and electromechanical interactions, and global spread ofexcitation via TM traveling waves, with direct implications for cochlear mechanisms.

TECTORIAL MEMBRANE MORPHOLOGYAND COMPOSITION

The tectorial membrane (TM) is a highly hy-drated extracellular matrix that resides above

the hair bundles of mechanosensory hair cellsin the cochlea. Because of its strategic positionabove the bundles, the TM is believed to playa critical role in the stimulation of hair cells.Recently, genetic studies have confirmed theimportance of the TM in hearing, wherebymanipulations of genes targeting TM proteinshave resulted in significant hearing deficits.These mutations cause severe hearing deficits,even when the TM is nearly unchanged in itsphysical orientation and structural attachmentsto the sensory receptors.

The TM’s three main constituents are water(97%), glycosaminoglycans ([GAGs], e.g., uron-ic acid and keratan sulfate), and collagenous(collagen II, IX, and XI), and noncollagenous

proteins (α-tectorin, β-tectorin, CEACAM16,otogelin, and otogelin-like). Many of the TM’sstructural constituents are commonly found inconnective tissues (Freeman et al. 2003) with theexception of the glycoproteins. The TMs α- andβ-tectorin glycoproteins make up a large frac-tion of the TM’s dry weight, and are constituentsof the TM’s striated sheet matrix (Killick et al.1995; Legan et al. 2000), within which collagenfibrils are embedded and oriented in the radialdirection perpendicular to the cochlear spiral(Fig. 1A).

Recent results reported by Andrade et al.(2016) provide high-resolution images of theTM’s ultrastructure using immunogold label-ing and high-resolution freeze-fracture replicasused to investigate the finemolecular makeup ofTM molecules. Figure 1B shows parallel-orient-ed collagen bundles in the upper surface of theTM, extending from the spiral limbus towardthe TM marginal band. The most superficial

Editors: Guy P. Richardson and Christine PetitAdditional Perspectives on Function and Dysfunction of the Cochlea available at www.perspectivesinmedicine.org

Copyright © 2018 Cold Spring Harbor Laboratory Press; all rights reservedAdvanced Online Article. Cite this article as Cold Spring Harb Perspect Med doi: 10.1101/cshperspect.a033514

1

ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org

on April 13, 2022 - Published by Cold Spring Harbor Laboratory Press http://perspectivesinmedicine.cshlp.org/Downloaded from

mailto:[email protected]






http://www.perspectivesinmedicine.org



http://www.perspectivesinmedicine.org/site/misc/terms.xhtml

http://perspectivesinmedicine.cshlp.org/

A

B

D

F

E

C

BM

OHCIHC

TM

Covernet

Collagen fibrils

SL

SE

Wild-type Ceacam16 +/βgal Ceacam16 βgal/βgal

MB

CN

Figure 1. Tectorial membrane (TM) structure and morphology. (A) The TM is an extracellular matrix overlyingcochlear hair bundles comprised of 97% water by weight, a collagen matrix, and an embedded striated sheetcomprised of glycosaminoglycans and other noncollagenous proteins (α-, β-tectorin, and CEACAM16). (B) Ascanning electron microscope (SEM) image of the TM attached to the spiral limbus (SL) and above the sensoryepithelium (SE) (blue = limbal region; yellow = covernet [CN]; green = striated sheet; red =marginal band [MB];purple =Kimura’s membrane). Scale bar, 15 μm. (C) SEM image of the TM showing the top surface of thecovernet region. Scale bar, 10 μm. (D) A transmission electronmicroscope (TEM) image of the TM. Arrowheadsshow the location of the type II collagen fibrils interwoven by thin filaments. Scale bar, 100 nm. (E) High-resolution freeze-etching image of the TM formed by collagen fibrils cross-linked by filamentous materials. Scalebar, 50 nm. (Panels C–E from Andrade et al. 2016; reprinted, with permission, from Elsevier © 2016.) (F ) SEMimages of the TM within the cochlear partition in wild-type (left), Ceacam16 heterozygotes (middle), andCeacam16 homozygotes (right). IHC, Inner hair cell; OHC, outer hair cell; BM, basilar membrane. (Panel Ffrom Cheatham et al. 2014; reprinted, with permission, from the Society for Neuroscience © 2014.)

J.B. Sellon et al.

2 Advanced Online Article. Cite this article as Cold Spring Harb Perspect Med doi: 10.1101/cshperspect.a033514

ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



fibrillar bundles (covernet) obliquely crossthe TM in the longitudinal direction (Fig. 1C).Collagen fibrils running through the TM arecross-linked (Fig. 1D) by thin filaments (Fig.1E). The nanopores in the TM are 10s of nano-meters in wild-type mice (Masaki et al. 2006;Sellon et al. 2014), and tend to vary acrossdifferent mutant mouse models, as shown inrecent studies detailing the Ceacam16 mutantmice (Fig. 1F) (Cheatham et al. 2014), whichdisplay spontaneous, stimulus-frequency, andtransiently evoked otoacoustic emissions withmarkedly larger amplitudes. The striated sheet,nanopores, and embedded collagen fibrils createa highly inhomogeneous and anisotropic TMmatrix that spans the entire length of the co-chlea. This anisotropy gives rise to TMmechan-ical properties that vary in the longitudinal andradial directions (Abnet and Freeman 2000; Guet al. 2005).

Mutant mouse models have revealed thestructural importance of the tectorins as partof the striated matrix (Legan et al. 2005; Russellet al. 2007). Legan and colleagues (2014) exam-ined three tectorinmutations, and found a num-ber of morphological changes compared withwild types, such as reduced limbal zone, absenceof striated-sheet matrix, disruption of collagenfibrils in the sulcal region, delamination ofKimura’s membrane, detachment of Hensen’sstripe, altered morphology of the covernet, andenlarged nanopores (Sellon et al. 2014). Further-more, recent studies of the developing TM revealthe importance of the tectorins for establish-ing collagen fibril alignment within the tissue(Goodyear et al. 2017). These results highlightthe importance of the tectorin proteins in con-trolling the nanoscale and macroscale structureof the TM.

NANOSCALE STRUCTURE AND MATERIALPROPERTIES

Point-stiffness studies developed by von Békésyand Zwislocki were the first to explore the role ofTM local mechanical properties (i.e., mass, stiff-ness) in stimulating the sensory hair bundlesof hair cells (von Békésy 1960; Zwislocki 1980;Zwislocki andCefaratti 1989).More recentmea-

surements in mice and guinea pig preparations(Freeman et al. 2003; Shoelson et al. 2004; Guetaet al. 2006;Masaki et al. 2006; Richter et al. 2007)have shown that TM point stiffness varies frombase to apex. However, there is considerablevariability across the studies (by as much as a100-fold), in part because the TM is anisotropic,and also because the static and dynamic meth-ods are different. It is now clear that the me-chanical and material properties of the TM arefrequency dependent and consistent with visco-elastic theory at the bulk level (Abnet and Free-man 2000; Gu et al. 2008), but it remains unclearhow the TM locally interacts with the hair bun-dles at audio frequencies (Ghaffari et al. 2007).

The dynamic material properties of the TMare determined not only by the matrix of mac-romolecules, but also by their interactions withinterstitial fluid. Forces of fluid origin depend onboth the viscosity of the fluid and the distancebetween macromolecules, which can be charac-terized by an effective pore size (Masaki et al.2006; Sellon et al. 2014). Because the TM con-tains a fixed charge and 97% of its mass is water(Freeman et al. 2003), it is hardly surprisingthat electrostatic and viscous properties of theTM are important. But what is more surprisingis that TM dynamic interactions at the sizescale of the hair bundles may be determinedby both electrokinetic and poroelastic local phe-nomena, as opposed to stiffness alone. Recentfindings show that the shear viscosity of theTM depends on porosity (Masaki et al. 2006;Ghaffari et al. 2013; Sellon et al. 2014), whichunderlies key differences in the cochlear pheno-types of TectaY1870C/+ and Tectb−/− mutants(Fig. 2A,B) (Sellon et al. 2014), and thus repre-sents a critical property of the TM.

The nanoscale porosity of TM in combina-tion with shear storage modulus and shear vis-cosity could contribute to local stimulation ofthe hair bundles. Hair cells are sensitive to shearforces that displace hair bundles along theiraxis of symmetry. Based on its strategic positionoverlying the hair bundles, the TM has beenwidely believed to deliver shear forces that stim-ulate hair cells at audio frequencies, consistentwith a second-order system. However, the pres-ence of fluid and pores in the TM suggests that

TM Mechanical Properties and Functions

Advanced Online Article. Cite this article as Cold Spring Harb Perspect Med doi: 10.1101/cshperspect.a033514 3

ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



the phase angle and magnitude of TM–hairbundle displacements are frequency-dependent,leading to more complex frequency-dependentinteractions than previously thought (Nia et al.2011). Application of nanoindentations at audiofrequencies causes a mechanical frequency re-sponse of the TM, which shows a prominentphase lead at middle frequencies (Fig. 2C). Atlow frequencies, the mechanical response ap-proaches an asymptotic value corresponding toequilibrium elastic response. At higher frequen-cies, the modulus approaches another asymp-

totic value and undergoes a change in phasethat peaks and decreases with frequency. Ac-cording to poroelastic theory, altering the sizeof the atomic force microscope (AFM) probe tipcauses the characteristic length over which fluidflows to change, thereby altering the couplingand peak frequency of the phase angle (Niaet al. 2015). Thus, the TM’s poroelastic responsemay be highly dependent on the size of the hairbundles. By altering the local phase of the hairbundles versus the displacement of the reticularlamina, the TM’s poroelastic propertiesmay also

Frequency (Hz)1 10 100 10000.10.01

π/2

π/3

π/6

0

D = 1.5 μmD = 15 μm

φ (r

ad)

fpeak

PEG moleculeTM pores

PEG excluded

Small PEGs(8–100 kDa)

Medium PEGs(200–400 kDa)

Large PEGs(600–900 kDa)

Incr

ease

in w

ave

spee

d (%

)

0

15

30

45

60

75

PEG molecular weight (kDa)9008 100 200 300 400 600

2215 28 32 40 486

PEG radius of gyration (nm)

PEGexcluded

PEGpermeates

= Wild-type

TectaY1870C/+

Wild-type +Tectb−/−

= Tectb−/−

= TectaY1870C/+

B

CA

+

OHCmembrane RC

OHC

xHB

xRLxRL xTM xS

xB

fByRLyBM

yRL OHCs

Levermodel

+

xTM, φTMTM

OHCφHB

IMET

IMET

VOHCyOHC

yOHC

ResonantTM

OHCmotility

METkTM+– –

D

Dynamic AFM probe

TMD

kB

1

Figure 2. Role of tectorial membrane (TM) porosity in controlling local and global mechanical properties.(A) Polyethylene glycols (PEGs) with varying molecular masses (8 kDa to 900 kDa) added to artificial endo-lymph surrounding TectaY1870C/+, Tectb−/−, and wild-type TMs cause changes in wave speed, implying differ-ences in the nanoporous structure of these tissues. (B) Schematic illustration of the interaction between PEGmolecules and the nanoporous structure of the TM. (PanelsA andB reprinted, with permission, from Sellon et al.2014.) (C) Schematic drawing and phase versus frequency response highlighting differences in phase of stimu-lation for hair bundle and stereocilia-sized probes. Poroelasticity theory predicts that the size of a probe (or hairbundle) in contact with a porous tissue would affect its phase of stimulation. (D) Depiction of how local me-chanical properties of the TM can influence hair bundle mechanics and feedback loops in the cochlea. Displace-ment of the basilar membrane ([BM] yBM) displaces the reticular lamina (xRL), which then deflects outer hair cell(OHC) hair bundles with a given TMdisplacement (xTM) and an angle ϕ. This angle ϕ depends on the frequency-dependent interactions of hair bundles with the poroelastic TM and the magnitude depends on the hair bundlesstiffness kB and theTMstiffness kTM.Deflection of hair bundles causes current toflow throughmechanoelectricaltransduction channels (IMET), charging OHC capacitance, which applies a voltage on the OHC membrane(VOHC) and ultimately triggers OHCmotility with given displacement (yOHC). AFM, Atomic force microscope.

J.B. Sellon et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



alter timing and polarity of cochlear feedbackpaths (Fig. 2C), and this phase lead may helpovercome the variety of phase lags that are pres-ent throughout the cochlea and thus contributeto the sensitivity of mammalian hearing.

FIXED CHARGE AND ELECTROKINETICS

In addition to glycoproteins and collagen, bio-chemical studies have long established that the

TM contains highly charged GAGs (Steel 1983;Ghaffari et al. 2013). These fixed charge groupsare fully ionized at physiological pH and neu-tralized at acidic pHs. Quantifying the densityof fixed charge (cf ) in the TM is important forunderstanding the mechanical properties ofthe TM (Fig. 3A). Traditional electrical record-ing techniques that require micropipettes havebeen largely unsuccessful in measuring TM cfbecause producing stable voltages across the

100

KCI concentration (mmol/L)

027Time (min)

0

40 102101

102

Frequency (Hz)

Dis

plac

emen

t pk-

pk (

nm)

Pha

se (

radi

ans)

103 40–π/2

–π/4

0

102

Frequency (Hz)103

54 200

TM

TM

TM

Electric field lines

AE

AE

AEvariant

VD = VT – VB

VD (

mV

)

VB

Externalpressure

External pressuresresisted by:

Mechanical forces

Electrosatic repulsionof fixed charge

+

++

–

–

–

VT

0

–5

3 21 mM 87 32 43 174 21 43 32 87

Slope = –1

Stimulus: 15 kV/m

A

B C

F G

D

E

Figure 3. Localized electrokinetic properties of the tectorial membrane (TM). (A) Schematic drawing of the TMshowing negative fixed charge groups attached to a network of collagen fibers (represented by mechanicalsprings). (B) Schematic drawing of the microaperture chamber showing isolated TM segment over a circularmicroaperture. The microaperture creates a fluid path from the overlying artificial endolymph (AE) bath to theunderlying microfluidics channel (test bath) perfused with variants of artificial endolymph having variable KClconcentrations. (C) The potential difference between the two baths was recorded over time with Ag/AgClelectrodes placed in contact with two baths. (D) Voltage was plotted as a function of the test bath concentrationwith best fit estimates (solid line) yielding the fixed charge concentration of the TM (−7.1 mmol/L; n= 5 TMpreparations). (E) The microaperture setup in panel B was used to deliver electric fields to the TM with a pair ofAg/AgCl stimulating electrodes. (F) Electrically evoked displacements of the TM were nanoscale in amplitudeand decreased with increasing stimulus frequency (40–1000 Hz). (G) Phase angle of TM displacements as afunction of stimulus frequency remained flat around -π/2. (PanelsA–G reprinted, with permission, fromGhaffariet al. 2013.)



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



TM has proven difficult (Steel 1983). Such mea-surements are complicated by the fact that theTM lacks an insulating cell membrane layer,thus making it difficult for the tip of a micro-electrode to be topologically inside the TM. Toovercome the challenges associated with micro-electrode techniques, Ghaffari et al. (2013) de-veloped a technique whereby isolated TM seg-ments were mounted in a two-compartmentchamber containing a microaperture (15 μm ra-dius) connecting the two compartments. Thepotential difference between the two-fluid mi-crochannel and the top bath arises primarilyfrom Donnan potentials that form betweeneach bath and the TM, because of the presenceof a fixed charge (Fig. 3B). If the two baths haveidentical compositions, the Donnan potentialbetween the TM and each bath is identical, sothe net potential difference between baths iszero. If one bath has a lower ionic concentration,the magnitude of the Donnan potential increas-es, so the potential difference between baths de-viates from zero in a manner that depends on cfwithin the TM (Fig. 3C) consistent withDonnantheory (Fig. 3D) (Ghaffari et al. 2013).

The presence of fixed charge in the TM sug-gests the possibility that electrical stimuli mightgenerate amechanical response. The applicationof oscillating electric fields at audio frequencies(1–1000 Hz) was found to generate displace-ments of the TM in a microaperture chamber(Fig. 3E). TM displacements had peak ampli-tudes at positions on the undersurface of theTM locally above the microaperture. Displace-ment amplitudes dropped significantly withdistance away from themicroaperture, increasedwith electric field strength, and decreased as afunction of stimulus frequency, consistent withviscous dominated interactions (Fig. 3F,G).

TM electrokinetics raises the intriguing no-tion that the TM may not behave as a purelymechanical structure locally but, instead, showsfrequency-dependent electromechanical prop-erties across multiple rows of hair cells. Thiselectrically evoked motion may be driven bylocal interactions with the hair bundles of outerhair cells (OHCs), which are inserted into theTM. It is thus important to note that OHC bun-dles have been shown to generate mechanical

forces in response to deflections (Jia and He2005; Ashmore 2008). Such forces could con-tribute to the local motions of the TM, or maybe enhanced by the mechanoelectrical trans-duction currents, which could exert electroki-netic forces locally near the surface of the TM(Ghaffari et al. 2013).

RADIAL FIBRILLAR STRUCTURE AND WAVEMECHANICS

In addition to its local porosity and electrome-chanical properties, the radial fibrillar structureof the TM may also play an important role incochlear mechanics. Several studies have exam-ined how the radial fibrillar structure of theTM might contribute to its mechanical proper-ties. For example, to apply shear motions in thedirections relevant to hair bundle mechanics,magnetic beads (Abnet and Freeman 2000) andmicrofabricated shearing probes (Gu et al. 2005)have been used. Both of these techniques showthat the TM is a viscoelastic tissue that can spa-tially couple motions in both the radial and lon-gitudinal directions. Furthermore, these studieshave shown that the TM is stiffer in the radialdirection (along the fibrils) than in the longitu-dinal direction. In addition to shear probes, TMmaterial properties have been determined frompoint indentationswithAFMcantilevers (Shoel-son et al. 2004; Gueta et al. 2006; Richter et al.2007). Recently, AFM cantilevers were used todetermine viscoelastic properties of the TM ataudio frequencies (Gavara and Chadwick 2010).The techniques described thus far have charac-terized the local material properties of theTM relevant to hair bundle mechanics. Howev-er, the organ of Corti is longitudinally coupledby the TM, suggesting that bulk measurementsof the TM are important. Recently, Lee et al.(2016) showed significant radial motion of theTM in vivo, which is exaggerated by α-tectorinmutation, TectaC1509G/C1509G, consistent withthe observation of a reduced limbal attachmentin this mutant (Xia et al. 2010). These resultsshow the importance of the TM’s bulk, radialstructure in mechanical stimulation of OHCfor proper maintenance of cochlear turning.To further examine how the radial fibrillar struc-

J.B. Sellon et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



ture of theTMaffects theTMsglobalmechanicalproperties, JB Sellon, R Ghaffari, DM Freeman,et al. (in prep.) recently explored traveling wavesin Col11a2−/− mutant mice. Transgenic micelacking Col11a2 have 40–50 dB threshold eleva-tion measured at the brainstem (McGuirt et al.1999), which is thought to be attributable to adisruption in the TM’s collagen fibrils and itsradial fibrillar structure. Compressional wavesin the radial direction of TMs excised fromwild-type and Col11a2−/− mutant mice propa-gate between the limbal andmarginal zones (Fig.4A,B).Col11a2−/−mice have increasedphase ac-cumulation in the radial direction (along the col-lagen fibrils) compared with wild-type TMs. In

addition to phase accumulation in the radial di-rection, Col11a2−/− mice have increased phaseaccumulation in the longitudinal direction, andthus reduced longitudinal wave speed. The radialfibrillar structure of the TM is thus an importantphysical feature that controls both longitudinaland radial coupling phenomena.

LONGITUDINAL COUPLING VIATRAVELING WAVES

In addition to radial motion, the TM has beenshown to propagate longitudinally propagatingwaves in computational models (Meaud andGrosh 2010; Sellon et al. 2017), in vitro (Ghaffari

A

Vibrating support

Longitudinal (l)

Stationary support

TM

Rad

ial (

r)

Wild-type

–60

–50

–40

–30

–20

–10

0

2.5 kHz3.0 kHz4.0 kHz

2.5 kHz3.0 kHz4.0 kHz

Pha

se (

deg)

Col11a2–/–

Col11a2–/–

Wild-type

17–20 kHz

Pha

se (

deg)

Longitudinal distance (μm)

17–20 kHz

B

C D

0 20 40 60 80

0 50 100 150–250

–200

–150

–100

–50

0

Vib

ratin

g su

ppor

t

Longitudinal (l)

Stationary support

TM

Rad

ial (

r)

Figure 4. Compressional radial waves and longitudinal traveling waves in Col11a2 mutant tectorial membranes(TMs). (A) Wave chamber setup depicting stimulation of compressional waves along the radial direction of theTM. (B) Phase accumulation ofmotion along the TM’s radial fibrillar structure. Deletion ofCol11a2 increases theamount of phase accumulation along the radial direction and thus could impact outer hair cell (OHC)–inner haircell (IHC) coupling mechanisms. (C) Wave chamber setup depicting stimulation of longitudinal traveling wavesalong the TM. (D) Deletion of Col11a2 increases the amount of phase accumulation along the longitudinaldirection and thus reduces TM traveling wave speed. Reduced TMwave speedmay causemismatches with basilarmembrane (BM) wave speed and thus affect the sensitivity of hearing in Col11a2−/− mice.



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



et al. 2007; Jones et al. 2013; Sellon et al. 2014,2015), and in vivo (Lee et al. 2015). Similar tothe role of waves in Col11a2−/− mice, TM trav-eling waves could explain the differences in co-chlear function between both the Tectb−/− andTectaY1870C/+ mutants and the wild types.TectaY1870C/+ mutants, in particular, have nor-mal hair bundles, TM attachments, and basilarmembrane (BM) sensitivity response, but showreduced neural sensitivity (Russell et al. 2007),which cannot be explained by changes in TMpoint stiffness alone. Recentmeasurements (Sel-lon et al. 2014) at audio frequencies show thatTM shear viscosity is significantly lower inTectaY1870C/+ TMs than in wild-type TMs, ow-ing to larger nanopores (Fig. 5A,B). ReducingTM shear viscosity (via increased pore size) re-duced wave transmission losses, which, in turn,

facilitated TM waves to propagate with largerdecay constants in TectaY1870C/+ mutants.

Tectb−/−mutant mice, which lack the β-tec-torin glycoproteins, show a striking phenotypethat differs from TectaY1870C/+ and Col11a2−/−

mutants. They have reduced sensitivity andsharper BM tuning (Russell et al. 2007). Thesedifferences in sensitivity and tuning cannot beexplained by changes in cochlear amplification.However, this combination of hearing abnor-malities was found to be consistent with mea-sured changes in TM wave decay constants.Ghaffari et al. (2010) showed that TM waves inTectb−/−mice have smaller decay constants by afactor of two, suggesting that differences in TMwaves may underlie differences in longitudinalcoupling. These findings together show that theTM is not a purely elastic structure, but rather it

Marginal

Limbal

B C

A

5

150

600

2

10

Dec

ay c

onst

ant σ

(μm

)

σ ~ 110 μm

σ ~ 230 μm

σ ~ 230 μm

σ ~ 230 μm

Dec

ay c

onst

ant σ

(μm

)

Spe

ed V

(m

/S)

Spe

ed V

(m

/S)

50

10Frequency (kHz)

20

5 10Frequency (kHz)

Mouse - 40 h PM

Human - 48 h PM

2010

4

3

5

6

7

8= Wild-type

Wild-type

= Tectb–/–

Tectb–/–

TectaY1870C/+

= TectaY1870C/+

15Frequency (kHz)

Human

20

10

100

150

200

250

300

350

400

15Frequency (kHz)

20

Figure 5. Spread of excitation via tectorial membrane (TM) traveling waves in wild-type mice, mutant mousemodels, and humans. (A) Snapshots of TMsmounted in the wave chamber with an exponentially decaying wavepasted on the image for wild-type, Tectb−/−, TectaY1870C/+, and human TMs. The decaying wave highlights thespatial extent of TMwaves at a given frequency. (B) Pooled datasets showing TMwave speed (top) andwave decayconstants (bottom) measured across wild-type, TectaY1870C/+, and Tectb−/− TMs. (C) Pooled datasets for TMwave speed (top) and wave decay constants (bottom) for human TMs compared with mouse TMs (tested 48 hpostmortem [PM]). (Panels adapted from data in Sellon et al. 2014 and Farrahi et al. 2016.)

J.B. Sellon et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



has important wave properties that account fordifferences in cochlear tuning phenotypes ofTectaY1870C/+ and Tectb−/− mutant mice.

IMPLICATIONS FOR HEARINGMECHANISMS

TMwave results inmutant mousemodels raisedthe intriguing possibility that neural frequencytuning in humansmay be sharper attributable inpart to differences in TM longitudinal couplingand wave mechanics. To determine whetherTM wave properties underlie sharper tuning in

humans (Shera and Charaziak 2018), Farrahiet al. (2016) conducted the first direct measure-ments on freshly excised human TMs. Their invitro results showed that spatial decay constantsfor both mice and humans were on the order of150 µm at 20 kHz (Fig. 5C). This distance is ameasure of longitudinal spread of excitation,which has been shown to correlate with spectralspread of excitation. In mice, the spatial spreadof ∼150 μm corresponds to a frequency spreadof 1.6 kHz and therefore a quality of tuningQ10dB of ∼10 at 20 kHz. In Tectb knockoutmice, the decay constant of TMwaves is approx-

MouseD = 1

.25

mm

Octave

ApexHuman

Human

BaseQ10dB ~ 38+/–13

2 mm

10

2

6

3

9 13

3

Bes

t pla

ce (

mm

)B

est p

lace

(m

m)

4

20 30

Q10dB12.5

Tectb

Place-frequency map

WT

8.3

A B

D = 5 mm

Best frequency (kHz)

Figure 6.Tectorialmembrane (TM)wave properties correlatewith frequency tuning via a place-frequencymap ofthe cochlea in wild-type mice, mutant mouse models, and humans. (A) Place-frequency maps of wild-type andmutant mice (top) compared with humans (bottom), showing the relationship between the spread of the TMwave and the frequency spread. Sharpness of frequency tuning (Q10dB) was larger in human TMs (Q10dB ∼ 38)compared with Tectb−/− (Q10dB ∼ 12.5) and wild-type mice (Q10dB ∼ 8.3). (Panel A [top] reprinted, withpermission, from Sellon et al. 2014.) (B) Schematic drawings of cochlear spirals in human and mouse. Thespatial extents of octave intervals as calculated from the cochlear maps of Greenwood (1990) and Muller et al.(2005) are indicated with colored lines along the spiral. WT, wild type. (Panel B reprinted, with permission, fromFarrahi et al. 2016.)



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



imately halved, leading to a Q10dB value that isdoubled. This sharpened tuning, predicted fromthemechanical spread of excitationmeasured inan isolated TM, correlates well with sharpenedtuning in auditory nerve measurements of thesemutant mice. Although TM waves spread exci-tations over similar distances in mice andhumans (Fig. 6A,B), these distances span signif-icantly different ranges of frequencies. Thisdifference has important implications for ourability to separate sounds by their frequencycontent, suggesting that the spread of excitationvia TM traveling waves is broader in mice thanin humans. It is therefore important to charac-terize the spread of excitation in terms of therange of frequencies over which the excitationis spread. In addition to measurements on iso-lated TMs, Dewey et al. (2018) show that reduc-tion in hair bundle stiffness sharpens cochleartuning by decreasing the spread of excitationthrough the TM traveling wave in vivo. Thus,these results have shown that the spatial extentof TM waves strongly correlates with cochleartuning in humans, mice, and likely other mam-mals, suggesting that the TM is essential forcontributing to cochlear sensitivity and frequen-cy selectivity.

CONCLUSIONS

The findings presented in this review show thatthe TM is a highly porous, hydrated, and dy-namic structure. The poroelastic, electrokinetic,andwave properties of theTMare frequency andplace dependent and thereby may be essentialfor cochlear sensitivity and frequency selectivity.Future research directions in imaging, nanoscalemechanics, and electrokinetic measurementswill help further advance our understanding ofTM local, radial, and longitudinal cochlear in-teractions, with direct implications for cochlearmechanisms.

REFERENCES�Reference is also in this collection.

Abnet CC, Freeman DM. 2000. Deformations of the isolatedmouse tectorialmembrane produced byoscillatory forces.Hear Res 144: 29–46.

Andrade LR, Salles FT, Grati M, Manor U, Kachar B. 2016.Tectorins crosslink type II collagen fibrils and connect thetectorial membrane to the spiral limbus. J Struct Biol 194:139–146.

Ashmore J. 2008. Cochlear outer hair cell motility. PhysiolRev 88: 173–210.

Cheatham MA, Goodyear RJ, Homma K, Legan PK, Kor-chagina J, Naskar S, Siegel JH, Dallos P, Zheng J, Rich-ardson GP. 2014. Loss of the tectorial membrane proteinCEACAM16 enhances spontaneous, stimulus-frequency,and transiently evoked otoacoustic emissions. J Neurosci34: 10325–10338.

Dewey JB, Xia A, Müller U, Belyantseva IA, Applegate BE,Oghalai JS. 2018. Mammalian auditory hair cell bundlestiffness affects frequency tuning by increasing couplingalong the length of the cochlea. Cell Rep 23: 2915–2927.

Farrahi S, Ghaffari R, Sellon JB, Nakajima HH, FreemanDM. 2016. Tectorial membrane traveling waves underliesharp auditory tuning in humans.Biophys J 111: 921–924.

Freeman D, Masaki K, McAllister A, Wei J, Weiss T. 2003.Static material properties of the tectorial membrane:A summary. Hear Res 180: 11–27.

GavaraN, Chadwick RS. 2010. Noncontactmicrorheology atacoustic frequencies using frequency-modulated atomicforce microscopy. Nat Methods 7: 650–654.

Ghaffari R, Aranyosi AJ, Freeman DM. 2007. Longitudinallypropagating traveling waves of the mammalian tectorialmembrane. Proc Natl Acad Sci 104: 16510–16515.

Ghaffari R, Aranyosi AJ, Richardson GP, Freeman DM.2010. Tectorial membrane traveling waves underlie ab-normal hearing in Tectb mutants. Nat Commun 1: 96.

Ghaffari R, Page SL, Farrahi S, Sellon JB, FreemanDM. 2013.Electrokinetic properties of the mammalian tectorialmembrane. Proc Natl Acad Sci 110: 4279–4284.

Goodyear RJ, Lu X, Deans MR, Richardson GP. 2017.A tectorin-based matrix and planar cell polarity genesare required for normal collagen-fibril orientation in thedeveloping tectorial membrane. Development 144: 3978–3989.

Greenwood DD. 1990. A cochlear frequency-position func-tion for several species—29 years later. J Acoust Soc Am87: 2592–2605.

Gu JW, Aranyosi AJ, Hemmert W, Freeman DM. 2005.Measuring mechanical properties of the isolated tectorialmembrane at audio frequencies with a microfabricatedprobe. In Abstracts of the Twenty-Eighth Midwinter Re-searchMeeting of Association for Research in Otolaryngol-ogy. New Orleans, LA.

Gu JW, Hemmert W, Freeman DM, Aranyosi AJ. 2008. Fre-quency-dependent shear impedance of the tectorialmembrane. Biophys J 95: 2529–2538.

Gueta R, Barlam D, Shneck RZ, Rousso I. 2006. Measure-ment of the mechanical properties of isolated tectorialmembrane using atomic force microscopy. Proc NatlAcad Sci 103: 14790–14795.

Jia S, He DZ. 2005. Motility-associated hair-bundle motionin mammalian outer hair cells. Nat Neurosci 8: 1028–1034.

Jones GP, Lukashkina VA, Russell IJ, Elliott SJ, LukashkinAN. 2013. Frequency-dependent properties of the tecto-

J.B. Sellon et al.


ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



rial membrane facilitate energy transmission and ampli-fication in the cochlea. Biophys J 104: 1357–1366.

Killick R, Legan PK, Malenczak C, Richardson GP. 1995.Molecular cloning of chick β-tectorin, an extracellularmatrix molecule of the inner ear. J Cell Biol 129: 535–547.

Lee HY, Raphael PD, Park J, Ellerbee AK, Applegate BE,Oghalai JS. 2015. Noninvasive in vivo imaging revealsdifferences between tectorial membrane and basilarmembrane traveling waves in the mouse cochlea. ProcNatl Acad Sci 112: 3128–3133.

Lee HY, Raphael PD, Xia A, Kim J, Grillet N, Applegate BE,Ellerbee Bowden AK, Oghalai JS. 2016. Two-dimensionalcochlear micromechanics measured in vivo demonstrateradial tuning within the mouse organ of corti. J Neurosci36: 8160–8173.

Legan PK, Lukashkina VA, Goodyear RJ, Kössl M, Russell IJ,Richardson GP. 2000. A targeted deletion in α-tectorinreveals that the tectorial membrane is required for thegain and timing of cochlear feedback. Neuron 28: 273–285.

Legan PK, Lukashkina VA, Goodyear RJ, Lukashkin AN,Verhoeven K, Camp GV, Russell IJ, Richardson GP.2005. A deafness mutation isolates a second role for thetectorial membrane in hearing. Nat Neurosci 8: 1035–1042.

Legan PK, Goodyear RJ, Morín M, Mencia A, Pollard H,Olavarrieta L, Korchagina J, Modamio-Hoybjor S, MayoF, Moreno F, et al. 2014. Three deaf mice: Mousemodels for tecta-based human hereditary deafness revealdomain-specific structural phenotypes in the tectorialmembrane. Hum Mol Gen 23: 2551–2568.

Masaki K, Weiss TF, Freeman DM. 2006. Poroelastic bulkproperties of the tectorial membrane measured withosmotic stress. Biophys J 91: 2356–2370.

McGuirt WT, Prasad SD, Griffith AJ, Kunst HP, Green GE,Shpargel KB, Runge C, Huybrechts C, Mueller RF, LynchE, et al. 1999. Mutations in COL11A2 cause non-syn-dromic hearing loss (DFNA13). Nat Genet 23: 413–419.

Meaud J, Grosh K. 2010. The effect of tectorial membraneand basilar membrane longitudinal coupling in cochlearmechanics. J Acoust Soc Am 127: 1411–1421.

Muller M, von Hunerbein K, Hoidis S, Smolders JW. 2005.A physiological place–frequency map of the cochlea inCBA/J mouse. Hear Res 202: 63–73.

Nia HT, Han L, Li Y, Ortiz C, Grodzinsky A. 2011. Poro-elasticity of cartilage at the nanoscale. Biophys J 101:2304–2313.

NiaHT, Han L, Bozchalooi IS, Roughley P, Youcef-ToumiK,Grodzinsky AJ, Ortiz C. 2015. Aggrecan nanoscale solid–fluid interactions are a primary determinant of cartilagedynamicmechanical properties.ACSNano 9: 2614–2625.

Richter CP, Emadi G, Getnick G, Quesnel A, Dallos P. 2007.Tectorial membrane stiffness gradients. Biophys J 93:2265–2276.

Russell IJ, Legan PK, Lukashkina VA, Lukashkin AN, Good-year RJ, RichardsonGP. 2007. Sharpened cochlear tuningin a mouse with a genetically modified tectorial mem-brane. Nat Neurosci 10: 215–223.

Sellon JB, Ghaffari R, Farrahi S, Richardson GP, FreemanDM. 2014. Porosity controls spread of excitation in tecto-rial membrane traveling waves. Biophys J 106: 1406–1413.

Sellon JB, Farrahi S, Ghaffari R, Freeman DM. 2015. Longi-tudinal spread of mechanical excitation through tectorialmembrane traveling waves. Proc Natl Acad Sci 112:12968–12973.

Sellon JB, Ghaffari R, Freeman DM. 2017. Geometric re-quirements for tectorial membrane traveling waves inthe presence of cochlear loads. Biophys J 112: 1059–1062.

� Shera CA, Charaziak KK. 2018. Cochlear frequency tuningand otoacoustic emissions. Cold Spring Harb PerspectMed doi: 10.1101/cshperspect.a033498.

Shoelson B, Dimitriadis EK, Cai H, Kachar B, Chadwick RS.2004. Evidence and implications of inhomogeneity intectorial membrane elasticity. Biophys J 87: 2768–2777.

Steel KP. 1983. Donnan equilibrium in the tectorial mem-brane. Hear Res 12: 265–272.

von Békésy G. 1960. Experiments in hearing. McGraw-Hill,New York.

Xia A, Gao SS, Yuan T, Osborn A, Bress A, Pfister M, Mari-cich SM, Pereira FA, Oghalai JS. 2010. Deficient forwardtransduction and enhanced reverse transduction in theα tectorin C1509G human hearing loss mutation. DisModel Mech 3: 209–223.

Zwislocki JJ. 1980. Five decades of research on cochlearmechanics. J Acoust Soc Am 67: 1679–1685.

Zwislocki JJ, Cefaratti LK. 1989. Tectorial membrane. II:Stiffness measurements in vivo. Hear Res 42: 211–227.



ww

w.p

ersp

ecti

vesi

nm

edic

ine.

org



published online October 22, 2018Cold Spring Harb Perspect Med Jonathan B. Sellon, Roozbeh Ghaffari and Dennis M. Freeman The Tectorial Membrane: Mechanical Properties and Functions

Subject Collection Function and Dysfunction of the Cochlea

Etiologies and MechanismsHidden Hearing Loss: A Disorder with Multiple

et al.David C. Kohrman, Guoqiang Wan, Luis Cassinotti,

Convergent Extension to Planar PolarityDevelopment and Patterning of the Cochlea: From

Mireille Montcouquiol and Matthew W. Kelley

DysfunctionHair Cell Afferent Synapses: Function and

Mustapha, et al.Stuart L. Johnson, Saaid Safieddine, Mirna

FunctionHair-Bundle Links: Genetics as the Gateway to

Guy P. Richardson and Christine Petit

Active Biomechanics of Sensory Hair BundlesDolores Bozovic

StrategiesOtotoxicity: Mechanisms and Otoprotective Aminoglycoside- and Cisplatin-Induced

Corné J. Kros and Peter S. Steyger

and FunctionsThe Tectorial Membrane: Mechanical Properties

M. FreemanJonathan B. Sellon, Roozbeh Ghaffari and Dennis

Inner Ear Hair CellsFunction and Dysfunction of TMC Channels in

David P. Corey, Nurunisa Akyuz and Jeffrey R. Holt

The Epidemiology of DeafnessAbraham M. Sheffield and Richard J.H. Smith

Cochlear Gene TherapyLawrence Lustig and Omar Akil

Toward the Optical Cochlear Implant

MoserTobias Dombrowski, Vladan Rankovic and Tobias

Age-Related Hearing LossMichael R. Bowl and Sally J. Dawson

Outer Hair Cells and ElectromotilityJonathan Ashmore

FunctionDevelopment and Maintenance of Cochlear Inner Ear Connexin Channels: Roles in

Fabio Mammano

Sensory Cells of the Inner EarInteractions between Macrophages and the

Mark E. WarcholCochleaA Functional Perspective on the Evolution of the

Christine Köppl and Geoffrey A. Manley

http://perspectivesinmedicine.cshlp.org/cgi/collection/ For additional articles in this collection, see

Copyright © 2018 Cold Spring Harbor Laboratory Press; all rights reserved


http://perspectivesinmedicine.cshlp.org/cgi/collection/

http://perspectivesinmedicine.cshlp.org/cgi/collection/


the tectorial membrane: mechanical properties and functions

Documents