Transduction of sound is dependent upon the organisation of the cytoskeletal structures that form and support the stereociliary bundle of cochlear hair cells. Freeze-fracture followed by deep etching reveals the organisation of the hair cell cytoskeleton and indicates how mechanical support for stereocilia is provided. This can offer explanations of why mutations in certain genes that are widely expressed in the body have effects only upon the functioning of auditory hair cells to cause non-syndromic deafness.

Introduction

The sensory cells of the hearing and balance systems in the inner ear are called hair cells. They derive their name from the organised bundle of projections from their apical surface. The hair bundle is formed of specialised microvilli, termed stereocilia, arranged in rows that increase in height in one particular direction across the apical surface of the cell (fig. 1a) [1]. Deflections of the stereocilia towards and away from the longest row opens and closes, respectively, ion channels, modulating a potassium current flowing through the hair cells from the potassium-rich fluid (endolymph) that bathes the apical surface to a sodium-rich, potassium poor fluid (perilymph) that bathes the baso-lateral surface. The current modulation as stereocilia are deflected leads to changes in hair cell membrane potentials stimulating responses. At the threshold of hearing, the stereociliary deflection that stimulates the cochlear hair cells is less that 0.5 nm (4-5A) [2]. This remarkable sensitivity is dependent upon mechanical properties derived from the organisation of cytoskeletal elements at the apical end of the hair cells and interactions with those of the non sensory supporting cells that surround each hair cell.
Stereocilia are formed of closely packed filaments of actin the supported on a platform, the "cuticular plate" in the apical cytoplasm of the hair cell (fig. 1b). The cuticular plate, which appears relatively unstructured in thin sections (fig. 1b), is composed of actin with a variety of actin-interacting proteins [3]. More recently, analysis of the mutations in a number of different genes that cause hereditary deafness has identified several other cytoskeletal proteins that are essential to the formation and maintenance of the transduction apparatus [4].

Identifying how these proteins are organised and interact structurally is crucial to understanding how hair cells work and how they go wrong.

Freeze-fracture and Deep-etching ­Procedures

Deep-etching of tissue following freeze-fracture provides a means to visualise the internal structures of cells at high resolution. Freeze-fracture alone exposes membrane leaflets in face view and thus the membrane-faces of the cell and of organelles; cytoplasmic structures are not usually revealed because they do not deflect the fracture plane to impart topographic contrast. Sublimation of ice from the fractured surface ("etching") exposes structure otherwise obscured. To enable etching, non-volatile cryoprotective agents such as gly­cerol must be avoided during sample preparation, necessitating the use of rapid-freezing procedures to reduce ice-crystal formation. Unfixed, "fresh" tissue could therefore be examined but on sublimation of ice in such samples, salts and other solutes in the cytoplasm remain and settle out onto the cell structures obscuring detail. It is therefore more usual to fix samples in glutaraldehyde before applying a rapid freezing procedure so that the tissue can be rinsed with water or water-methanol to wash out soluble material prior to freezing.
Early studies of the deep-etching of the sensory epithelia of the inner ear of chickens [5] revealed the potential of this technique for examination of the organisation of the structural elements of the hair bundle. In that work, tissue was frozen by slam-freezing against a cooled copper block. We have found that that procedure can compress the hair bundle as the tissue contacts with the block. We use plunge freezing into propane-isopentane cooled in liquid nitrogen of tissue sandwiched between two thin copper plates. Fracture is subsequently performed by separating the sandwich inside the freeze-fracture device. Following etching at -90 °C for 15 minutes, the exposed surface is rotary shadowed with platinum-carbon from an angle of 22°.

Deep-etching of Hair Cells

Freeze-fracture, deep etching of hair cells reveals the structural elements in and around stereocilia and in the apical cytoplasm of the hair cells, and their relationship to plasma membrane structures, exposed by fracture. The actin filaments of the stereocilia and extracellular crosslinks between adjacent stereocilia are exposed (fig. 2a). The stereociliary actin filaments can be seen to cross-link to the surrounding plasma membrane, especially at the point of entry into the hair cell body where the actin filaments are linked to the apical plasma membrane. The cuticular plate appears as a dense meshwork of cross-linked filaments in the apical cytoplasm of the cell (fig. 2b). Within the cuticular plate, the rootlets of the stereocilia are crosslinked to the filamentous network of the cuticular plate (fig. 2c). The cutic­ular plate itself is crosslinked to the under side of the apical plasma membrane and to the lateral plasma membrane along the length of the intercellular junction between the hair cell and the adjacent supporting cells (fig. 2d). On the supporting cell side of the junction, cytoskeletal elements of the size of actin filaments are also linked to the cell-cell junction and to other cytoskeletal elements including microtubules in the supporting cell cytoplasm. This organisation suggests that the stereocilia are rigidly supported upon the cuticular plate which in turn is "clamped" to the plasma membrane all along the intercellular junction and is thereby supported by the cytoskeleton of the supporting cell. The supporting cell, unlike the hair cell, rests on the underlying extracellular matrix of the sensory epithelium. Consequently, through this arrangement, the hair cell stereocilia are provided with a rigid support that will contribute to their extreme sensitivity to small vibrations induced by sound.
This interpretation has provided an explanation for why mutations in certain genes that encode proteins associated with the tight junction between cells that are widely expressed in the body cause only deafness (non-syndromic deafness). Disruption or loss of such proteins, including claudin 14 [6] and tricellulin [7], the protein present at the point where the a cell contacts two neighbours (i.e. where three cells meet), may adversely affect the cross-linking of the cuticular plate to the junctional region of the hair cell and thus the rigidity of the stereociliary support, compromising sensitivity to stimulating deflections [7].

Future Work

It is now important to attempt to identify the location of all the different proteins known to be associated with the transduction apparatus within the cytoskeletal network. One way is to attempt to apply the techniques of immunogold labelling of freeze-fracture replicas [8] to the deep-etched material. This is not simple and will be time consuming. An alternative that we are investigating is to use high resolution scanning electron microscopy of tissue de-membranated in detergent to expose the cytoskeleton prior to fixation, followed by immunogold labelling and identification to the gold-label in the SEM by back-scatter detection. This procedure derives from studies of de-membranated cells in culture, that have been immunogold labelled and then rapidly frozen and etched (without prior fracture) [9]. Labelling gold particles are easily discerned. The initial results from our SEM studies suggest the feasibility of this approach.

References:
[1] Forge A., Wright A., Br Med Bull 63, 5-24 (2002)
[2] Dallos P., in The Cochlea, Dallos P. Popper AN, Fay RR, Eds., Springer, 1-22 (1196)
[3] Slepecky N., in The Cochlea, Dallos P., Popper AN, Fay RR, Eds., Springer, 23-40 (1996)
[4] El-Amraoui A., Petit C., J Cell Sci 118, 4593-4603 (2005)
[5] Hirokawa N., Tilney LG, J Cell Biol 95, 249-61 (1982)
[6] Ben-Yosef T., et al., Hum Mol Genet 12, 2049-61 (2003)
[7] Riazuddin S., et al., Am J Hum Genet 79, 1040-51 (2006)
[8] Fujimoto K., J Cell Sci 108, 3443-3449 (1995)
[9] Shao D., et al., Cell Motil Cytoskeleton 63, 395-414 (2006)