Hearing, Hair Cells and the Cytoskeleton

The tectorial membrane actually contacts the top of the OHC and this membrane plays a vital role as mutants in a protein unique to this tissue (tectorin) cause deafness.   The Pillar cells have a contractile function and are assumed to somehow fine tune the system.  Not much detail is currently available on the role of pillar cells.  As the tectorial membrane (TM) oscillates in response to a sound, the OHC themselves respond by changing length (see below), this then moves the TM increasing the deflective current of endolymph over the IHC.  The shape of the cochlea means that there is a frequency selection along its length.

The exact details are far from clear at this stage, but the following is probably correct.  At the tip of each stereocilia, there is a fine filament that connects to a Ca²⁺ channel.  when the stereocilia are moved to the right by a sound source, this filament pulls the channel open.  This signal is then send to the brain via the attendant nerve cells.  The channel is now pulled down the cilium and closes during adaptation. A myosin (possibly myosin 1b?) moves the displaced channel component back up the length of the cilium in recovery.  

An inbuilt Amplifier 

The exquisite sensitivity of the ear is explained by an inbuilt amplifier.  Trains of oscillations are detected and amplified  OHC shape change is ultra rapid and acts to amplify the vibrations this is accomplished by an amazing protein called "prestin" whose shape is voltage dependent. Voltages applied across the OHC by nerve cells change to shape of prestin and as this protein is so abundant in the OHC the whole cell changes shape! Each OHC amplifies a bit of the signal so that all sounds become amplified. The amplified sounds are then detected by the IHC and messages are send to the brain.  

Human Genetics and the Ear

An comprehensive review on the molecular genetics of hearing loss is available on the web (Petit et al, 2001). Usher’s disease is an autosomal recessive retinitis pigmentosa that leads to both blindness and deafness, it is relatively common affecting about 1 in 23,000 (in the USA).  A much higher number (1 in 2000) are genetically deaf and 70% of these are non-syndromic, that is without other symptoms. The mouse equivalent gene mutations cause the Shaker1 phenotype these mice are deaf and cannot keep their balance, but are not blind.  The gene responsible for the condition is the Myosin VIIa gene.  It may turn out that the difference between Shaker1 mice and Usher’s sufferers is the type of mutation in the Myosin VIIa gene but another possibility is that Myosin VIIa is present in the photoreceptor cells while it is not in mice.  A different mutation in human Myosin VI also causes a non-syndromic deafness.  Many different mutations of the myosin VIIa are known, and these produce a spectrum of severity of the disease in humans like Duchenne’e and Becker’s musclular dystrophies.  The ear is a very complex structure being able to detect sound over 8 octaves (mice 3.5 and whales 10!).

The structure of the stereocilium and hair cells

It is important to note that unlike most cell type, auditory hair cells are not typically replaced..  Although some other mammalian hair cells can be stimulated to regenerate.  More than 25% of people of 65 year olds have significant hearing difficulty, and this is nearly 50% by the age of 80!  Infections and disease can cause “nerve deafness” a misnomer for conditions that result in the destruction of hair cells.  Note that the tips of each cilium is connected by a fine filamentous “tip link”.

Actin Binding Proteins of the hair cell.

Hair cells are remarkable in that they contain three very distinct actin filament assemblies right next to each other.  Each has its own actin binding proteins associated within it (Drenckhahn et al, 1991).  The stereocilia are microvillar -like projections supported by bundled actin filaments all with the same polarity, with their barbed ends at the tip and their pointed ends towards the cell body.  These bundles are held together by the ABP fimbrin.  No fimbrin is seen in other areas of the hair cell.  

Fimbrin. (plastin) A 67kd monomeric actin filament bundling protein which has two CH actin binding motifs that are similar to the actin binding region of dystrophin, but has no rod-like domain. Fimbrin has two EF hands (domains which typically bind calcium) but as no calcium sensitivity has been demonstrated.  

Epsin.  Another actin filament bundling protein. (30kDa) 

Villin.  It has been shown that the stereocilia are devoid of villin, a bundling protein found in similar structures in the brush border.  Fimbrin replaces villin in these structures possibly as villin has calcium activated severing activity whereas fimbrin does not.  Calcium increases when the stereocilia are activated so this would lead to the collapse of these structures. 

Spectrin.  A long anti-parallel dimer protein with a CH actin binding motif.  Unusually for spectrin it is found in the gel-like region of the cuticular plate region.  Spectrin is almost always associated with actin under the plasma-membrane with actin (like dystrophin).   

a-actinin. Yet another actin bundling protein, but unlike fimbrin, a-actinin can bundle anti-parallel filaments.  The bundles in the zonula adherens junction formed between the hair cell and the support cell form a band of anti-parallel filaments around the cell.  

Tropomyosin.  This side-binding and stabilizing protein is found in all actin regions of the hair cells except the stereocilia bundles.  In muscle cells tropomyosin regulates the interaction of actin with myosin.

Diaphanous.  Mutants of this gene product lead to an autosomal dominant profound deafness (Lynch et al, 1997).  The gene product is located in the stereocilia, but it is not yet known if it binds actin directly.  It is known that the Drosophila homolog is a ligand for profilin, a small actin binding protein that under certain circumstance encourages the formation of actin filaments.  It is also a ligand for Rho, a G-protein that regulates actin assembly.  It is presumed that Diaphanous regulates the formation of the stereocilia, but because this protein is also expressed in mature hair cells it probably is important in their maintenance. 

Actin.  Only the b-isoform of actin is expressed in the stereocilia (Hofer et al, 1997).  Usually two actin iso-types are expressed in non-muscle cells (b & g), it is thought that in the stereocilia the b form is restricted as it is known that fimbrin has a higher affinity to b-actin than g-actin.

Cahderins  Cadherins are adhesion plaque proteins forming a link between the cell surface and the cytoskeleton.  Recently, it has come to light that mutations in a member of the cadherin family causes symptoms similar to the "waltzer" mouse model (Di Palma et al, 2001; Alagramam et al, 2001).  

Myosins of the Hair Cell.

The figure opposite represents the distribution of the known isoforms of myosin in hair cells.  This picture has come from both immuno- fluorescence and immuno-electron microscopy (Hasson & Mooseker, 1997; Liang et al, 1999)

Myosin Ib.  Is located  throughout the stereocilia and concentrated at the bundles beveled top edge especially at the tip links (see below). In the stereocilia generally, myosin Ib is at the sides.  In the cell body staining is diffuse but excluded from the cuticular plate (Metcalf, 1998).

Myosin II.  There appears not to be any myosin II in the actual hair cell, even at the zonula adherens.  However, myosin II is present in the support cell in the zonula adherens on its side (Drenckhahn et al, 1991).  

MyosinV.  Present in nerve cells associated with the hair cells only, none in the hair cells.  It is not known if dilute mice have hearing difficulties as they have so many other problems (such as death).  

Myosin VI.  

Myosin VIIa. Found throughout the length of the sterocilia in normal structures.  In mutants the sterocilia are very much shorter, although their number and position seem unaffected.  Myosin VIIA is present in both inner and outer hair cells, although expression is greater in the former. Eudy & Sumegi, 1999; Maniak, 2001).  Patients suffer combined hearing and eye function impairments.  In the eye, the condition gets progressively worse, but the course of the disease is dependent on the nature of the mutation. Loss of myosin VIIa function leads to the failure of cilia to form.  Usher syndrome is usually inherited as a recessive trait meaning that as long as an individual has a correct copy of the gene, then he/she will be OK.  However, this is a mutation that is dominant and that is found within the coiled-coil domain.  Such domains in other proteins and in fact other myosins are dimerization domains and so it is suspected that a “bad” copy may disrupt the function of good copies of the protein by interference with their structure.  Usher patient show abnormal nasal ciliary cells, sperm as well as photorecpetor cell damage  (see below).  Mutations in the mouse homologue give rise to the shaker-1 phenotype, which is characterized by hyperactivity, head-tossing and circling behaviour due to vestibular and cochlear  dysfunction.  A number of possible explanations have been suggested for the involvement of myosin VIIa in the ear and eye.  The protein is localised to the cilium connecting the outer-segment to the inner segment (the cell body).  The waist formed by the cilium may be to allow two very different membrane domains to coexist in the same cell, in which case, the function of myosin VII may well be to facilitate this by actively removing inappropriate proteins.  It has also been suggested that the function is to transport opsin to the outer segment from the inner segment where it is made.  A recent study in Dictyostelium finds that this organism (despite neither hearing nor seeing) expressed a myosin VII homolog (A number of possible explanations have been suggested for the involvement of myosin VIIa in the ear and eye.  The protein is localised to the cilium connecting the outer-segment to the inner segment (the cell body).  The waist formed by the cilium may be to allow two very different membrane domains to coexist in the same cell, in which case, the function of myosin VII may well be to facilitate this by actively removing inappropriate proteins.  It has also been suggested that the function is to transport opsin to the outer segment from the inner segment where it is made.  A recent study in Dictyostelium finds that this organism (despite neither hearing nor seeing) expressed a myosin VII homolog (Titus, 1999).  When this gene was knocked out the amoeba could not phagocytose.  This amoeba expresses abundance of myosin family members particularly the myosin I group.  Many of these have produced phenotype including a reduction in phagocytosis but none as marked as this myosin VII knockout.  So it seems that the function of this particular myosin in Dictyostelium is phagocytosis.  There has been an almost uncanny conservation in the cytoskeletal proteins right across the phyla from amoeba to humans and this infers that the function of myosin VII in humans may also be phagocytosis.  This makes sense in the biological context of the ear (lots of phagocytosis at the roots of the cilia) and in the eye.  The outer segment of the eye is continually shed, so as to turnover the structures damaged by the constant exposure to light. Yet another function that Myosin VII appears to have is a role in the dynamics of cell adhesion (Tuxworth et al, 2001).  Again using the Dictyostelium mutant (lacking the functional myosin VII homolog), were discovered to be deficient in cell adhesion.  Myosin VII in these cells is normally localised to the leading edge in motile cells and to the tips of extending filopods.  Myosin VII in stereo cilia is located at the tip link connection and so may be playing a similar role in stabilizing the connection through the cell membrane to the tip link in concert with the cadherin family member described earlier. 

Myosin XV.  This gene was discovered very recently by screening the progeny of an X-ray irradiated mouse (Probst et al, 1998; Wang et al, 1998).  The mouse is profoundly deaf and the stereociliary bundles are extremely short (like the Ushers mouse).  This has been described as the shaker-2 phenotype.  Genetic mapping of the mutation showed eventually that the gene was that encoding a novel myosin (myosin XV) most closely related to myosin IX (in the head region at least).  Not much is known about myosin XV apart from its sequence.  More recently, it has been shown that mutations in the human homolog produces the recessive congenital nonsyndromic deafness DFNB3, originally identified in Bengkala, Bali. (Wang et al, 1998).  Although the function is not known an alternative Myosin XV has been detected with a large N-terminal extension, larger even than that found in Myosin III (nina C).  Immunofluorecence work suggests that the protein is found both in the cuticular plate and in the stereocilia (Liang et al, 1999). 

Lateral link- cytoskeleton  connection,Vezatin   This is a trans-membranous protein identified by the yeast two-hybrid system using the FERM domain of myosin VII as bate.  Vezatin therefore binds myosin VII by the FERM region and also to the cadherin/a-catenein complex. 

Sensory cell specific cadherins family member bind to a lateral-link across the extracellular space.  This link may also bind vezatin although there is no information at present. The C-terminus of cadherin binds a complex of cytoskeletal proteins so that forces are transmitted from one cell to the other via a lateral link. The attachment of Myosin VII to the complex may allow the lateral link to be transported in the plane of the membrane.

Hope for the future?

Verhoeven, K., Van Laer, L., Kirschhofer, K., Legan, P. K., Hughes, D. C., Schatteman, I., Verstreken, M., Van Hauwe, P., Coucke, P., Chen, A., Smith, R. J. H., Somers, T., Offeciers, E. F., Van de Heyning, P., Richardson, G. P., Wachtler, F., Kinberling, W. J., Willems, P. J., Govaerts, P. J. & Van Camp, G. (1998) Mutations in the human a-tectorin gene cause autosomal dominant non-syndromic hearing impairment. Nature Genetics. 19, 60-62.

Zheng, J., Shen, W., He, D. Z. Z., Long, K. B., Madison, L. D. & Dallos, P. (2000) Prestin is the motor protein of cochlear outer hair cells. Nature. 405, 149-155.