Multiple myosin motors and mechanoelectrical transduction by hair cells.
Hair cells adapt to sustained stimuli; for example, hair cells in the bullfrog's sacculus can detect transient vertical accelerations of less than [10.sup.-5] g, despite a constant I g stimulus from gravity (Koyama et al., 1982). Gating-spring tension, and hence channel open probability, is controlled by adaptation motors, which likely contain myosin molecules (reviewed in Hudspeth and Gillespie, 1994; Gillespie, 1995). During an excitatory stimulus, when gating-spring tension is high, [Ca.sup.2+] entering through open transduction channels triggers the adaptation motors to slip down the cytoskeleton and reduce tension [ILLUSTRATION FOR FIGURE 1 OMITTED]. By contrast, during inhibitory bundle deflections that slacken the gating springs, the motors climb towards the apical ends of the stereocilia and restore tension. The adaptation motor thus acts as a negative-feedback control mechanism that ensures that gating-spring tension remains at its optimal level.
To prove that adaptation is carried out by myosin molecules and to identify the responsible isozyme, we have taken advantage of years of intensive study of the properties of skeletal-muscle myosin. We have applied general principles derived from muscle actomyosin to design of experiments for hair cells. The ATPase cycle of myosin is illustrated in a simplified form in Figure 2. Are the properties of skeletal-muscle myosin II likely to resemble those of a hair-cell myosin, even one of an unusual class? Recent results have suggested that this is so. Detailed kinetic characterization of two Acanthamoeba myosin-I isozymes indicates that ATPase hydrolysis utilizes the same mechanism as vertebrate myosin II. Furthermore, the rate and equilibrium constants defining the amoeba myosin-I cycle are nearly identical to those of vertebrate myosin II (Ostap and Pollard, 1996). These results give us confidence that the same will hold true for the myosin molecules of hair bundle.
To provide evidence that myosin mediates adaptation in hair cells, we dialyzed hair cells of the bullfrog's sacculus with tight-seal, whole-cell recording electrodes filled with compounds that should interfere with the ATPase cycle. In one series of experiments, we introduced into hair cells adenine nucleoside diphosphates, such as ADP and the metabolism-resistant analog ADP[Beta]S, expecting that they would promote the population of the diphosphate-bound state of myosin (Gillespie and Hudspeth, 1993). We saw the expected results: adaptation was blocked, but the transduction currents remained robust, as if the adaptation motors were arrested along the actin filaments, unable to climb or slip. In addition, channel open probability increased, as the number of myosin molecules in force-producing states increased. These results are entirely analogous to the effects of ADP on isometrically contracting muscle fibers, where ADP blocks both shortening and lengthening, as well as increasing isometric tension (Cooke and Pate, 1985).
In a second series of experiments, we filled hair cells with phosphate analogs, reasoning that this should lead to decreased actin-myosin interaction (Yamoah and Gillespie, 1996). Phosphate analogs such as vanadate, beryllium fluoride, and sulfate eliminated motor-force production, blocked climbing adaptation, and slowed slipping adaptation. The effects of phosphate analogs on slipping adaptation were quantitatively explained by a simple model that uses parameter values obtained from skeletal-muscle myosin.
The remarkable similarity of the effects of adenine nucleoside diphosphates and phosphate analogs on hair-cell adaptation with their effects on skeletal-muscle behavior lends strong support to the hypothesis that myosin molecules mediate adaptation. To date, no other model for adaptation fits more than a fraction of these and other data.
The evidence that myosin molecules probably mediate adaptation has triggered a search for isozymes of myosin expressed in hair cells. Early evidence that hair bundles contain myosin (Macartney et al., 1980) was later disputed (Drenckhahn et al., 1982), and may have been due to anti-actin-antibody contamination of anti-myosin polyclonal antisera (Gillespie et al., 1993). Because of the large number of myosin isozymes (Cheney et al., 1993) and lack of antibodies that recognize all myosin isozymes, we developed a photoaffinity-labeling approach that relies on the properties of myosin's ATPase cycle. In this method, purified hair bundles (Gillespie and Hudspeth, 1991) are incubated with a radioactive nucleotide, such as [[Alpha]-32P]UTP, and a trapping phosphate analog, such as vanadate. After the nucleotide is hydrolyzed and [P.sub.i] is released, vanadate binds to myosin and traps the nucleotide in a slowly dissociating state. Thorough washing eliminates nucleotides bound to other bundle proteins, then radioactive nucleotides are covalently crosslinked to myosin molecules by UV irradiation. We found that three myosin isozymes, of 120, 160, and 230 kD, were most consistently labeled in purified hair bundles (Gillespie et al., 1993). Although all three proteins have properties that suggest they are myosin molecules, the behavior of the 120-kD protein was most consistent with that of an adaptation motor. Labeling of the 120-kD protein was blocked by antagonists of myosin, including ADP, ADP[Beta]S, and NANTP, and the rank order of effectiveness of three trapping analogs (vanadate, beryllium fluoride, and aluminum fluoride) was identical to the order of stability of the corresponding myosin II-analog complexes in the presence of actin (Gillespie et al., 1993; Yamoah and Gillespie, 1996). The 120-kD protein bound calmodulin, which is the [Ca.sup.2+]binding mediator of adaptation (Walker and Hudspeth, 1996). The other photolabeled proteins shared some, but not all, of these characteristics. The data thus argue that bundles contain three myosin isozymes, and that the 120-kD myosin has the most appropriate behavior for an adaptation-motor myosin.
We have used selective antibodies to identify and localize three myosin isozymes in saccular hair bundles. Two of these isozymes, myosin VI and myosin VIIa, have been shown, by genetic methods, to play essential roles in hair cells (Avraham et al., 1995; Gibson et al., 1995). Using an anti-myosin-VI antiserum, we showed that an unusually high-mass form of this isozyme is found exclusively in hair bundles (T. Hasson, P. G. Gillespie, and D. P. Corey, unpubl. data). This 160-kD form probably corresponds to the 160-kD photoaffinity-labeled protein. The location of myosin VI in bullfrog hair cells is complex. A large amount of myosin VI was found in the cytoplasm, and it can readily diffuse out of permeabilized hair cells. Another fraction was apparently tightly bound within the cuticular plate. In hair bundles, myosin VI was associated with basal tapers of a fraction of stereocilia (only 10%-25%). In other bundles, usually small bundles from newly formed hair cells, myosin VI was found throughout the bundle. We suspect that myosin VI may be playing a structural role in hair bundles.
Myosin VIIa was abundant in purified hair bundles, and it comigrated with the 230-kD photoaffinity-labeled protein (T. Hasson, P. G. Gillespie, and D. P. Corey, unpubl. data). In frog bundles, myosin VIIa was largely concentrated in a band 1-2 [[micro]meter] in height just above the basal tapers. Because one class of interstereociliary linkages is found in that location, we suspect that myosin VIIa serves as the intracellular anchor for these basal linkages. This localization is entirely consistent with its distributed localization in cochlear hair bundles (Hasson et al., 1995), where this class of linkages appears to be found along the length of the stereociliary opposition.
Finally, myosin I[Beta] appears to be the 120-kD photo-affinity-labeled bundle protein. Myosin I[Beta] has been definitively localized to hair bundles by using two antibodies. The first antibody, mT2, is selective for this isozyme over all other known myosin-I isozymes (Metcalf, 1996). In immunoblots, mT2 identified a 120-kD protein present at 100-200 molecules per stereocilium (Gillespie et al., 1993). A polyclonal antiserum raised against the tail of cloned myosin I[Beta] also detected, on immunoblots, a 120-kD protein of similar abundance (T. Hasson, P. G. Gillespie, and Corey, D. P., unpubl. data). Both antibodies are selective, but additional confirmation of the responsible isozyme derives from separation, on SDS-PAGE, of the 120-kD protein from myosin la (unpubl. data), the only other myosin-I isozyme known to be expressed in hair-cell-containing tissues (Solc et al., 1994).
Both antibodies labeled stereociliary tips, as expected for an adaptation-motor myosin (Gillespie et al., 1993; T. Hasson, P. G. Gillespie, and D. P. Corey, unpubl. data). But only a fraction of tips were labeled, and a substantial amount of myosin was found below the stereociliary tips; this observation is consistent with a recent demonstration that the transduction apparatus is highly dynamic and can reform after the tip links are broken (Zhao et al., 1996). In fixed cells observed with anti-myosin-I[Beta] antibodies, myosin molecules may be immobilized while carrying transduction channels and tip links to their proper locations.
Proof that myosin I[Beta] is indeed the adaptation motor will come from three classes of experiments, all in progress. Because the adaptation motor is thought to reside in the insertional plaque found at the upper end of a tip link (Hudspeth and Gillespie, 1994), immunolocalization of myosin I[Beta] at the plaque by immunoelectron microscopy will provide one level of proof. More convincing would be the inhibition of adaptation by isozyme-selective inhibitors of myosin I[Beta], such as inhibitory antibodies or selective peptides. Most definitive would be introduction of a mutated myosin gene into the genome of an animal with its myosin-I[Beta] gene deleted; if the properties of the reintroduced myosin gene were sufficiently distinctive, adaptation would be affected in a predictable manner.
This work was supported by the NIH (R01 DC02368 and P60 DC00979). P.G.G. is a Pew Scholar in the Biomedical Sciences.
Avraham, K. B., T. Hasson, K. P. Steel, D. M. Kingsley, L. B. Russell, M. S. Mooseker, N. G. Copeland, and N. A. Jenkins. 1995. The mouse Snell's waltzer deafness gene encodes an unconventional myosin required for structural integrity of inner ear hair cells. Nature Genetics 11: 369-375.
Cheney, R. E., M. A. Riley, and M. S. Mooseker. 1993. Phylogenetic analysis of the myosin superfamily. Cell Motil. Cytoskeleton 24: 215-223.
Cooke, R., and E. Pate. 1985. The effects of ADP and phosphate on the contraction of muscle fibers. Biophys. J. 48: 789-798.
Drenckhahn, D., J. Kellner, H. G. Mannherz, U. Groschel-Stewart, J. Kendrick-Jones, and J. Scholey. 1982. Absence of myosin-like immunoreactivity in stereocilia of cochlear hair cells. Nature 300: 531-532.
Gibson, F., J. Walsh, P. Mburu, A. Varela, K. A. Brown, M. Antonio, K. W. Beisel, K. P. Steel, and S. D. M. Brown. 1995. A type VII myosin encoded by the mouse deafness gene shaker-1. Nature 374: 62-64.
Gillespie, P. G. 1995. Molecular machinery of auditory and vestibular transduction. Curr. Opin. Neurobiol. 5: 449-455.
Gillespie, P. G., and A. J. Hudspeth. 1991. High-purity isolation of bullfrog hair bundles and subcellular and topological localization of constituent proteins. J. Cell Biol. 112: 625-640.
Gillespie, P. G., and A. J. Hudspeth. 1993. Adenine nucleoside diphosphate block adaptation of mechanoelectrical transduction in hair cells. Proc. Natl. Acad. Sci. USA 90: 2710-2714.
Gillespie, P. G., M. C. Wagner, and A. J. Hudspeth. 1993. Identification of a 120-kD hair-bundle myosin I located near stereociliary tips. Neuron 11: 581-594.
Hasson, T., M. B. Heintzelman, J. Santos-Sacchi, D. P. Corey, and M. S. Mooseker. 1995. Expression in cochlea and retina of myosin VIIa, the gene product defective in Usher syndrome type 1B. Proc. Natl. Acad. Sci. USA 92:9815-9819.
Hudspeth, A. J. 1989. How the ear's works work. Nature 341: 397-404.
Hudspeth, A. J. 1992. Hair-bundle mechanics and a model for mechanoelectrical transduction by hair cells. Pp. 357-370 in Sensory Transduction, D. P. Corey and S. Roper, eds. Rockefeller University Press, New York.
Hudspeth, A. J., and P. G. Gillespie. 1994. Pulling springs to tune transduction: adaptation by hair cells. Neuron 12: 1-9.
Koyama, H., E. R. Lewis, E. L. Leverenz, and R. A. Baird. 1982. Acute seismic sensitivity in the bullfrog ear. Brain Res. 250: 168-172.
Macartney, J. C., S. D. Comis, and J. O. Pickles. 1980. Is myosin in the cochlea a basis for active motility? Nature 288: 491-492.
Metcalf, A. B. 1996. Molecular characterization of amphibian myosin I[Beta], a candidate for the hair bundle's adaptation motor. Ph.D. thesis, University of Texas Southwestern Medical Center, Dallas, TX.
Ostap, E. M., and T. D. Pollard. 1996. Biochemical kinetic characterization of Acanthamoeba myosin-I ATPase. J. Cell Biol. 132: 1053-1060.
Solc, C. K., B. H. Derfler, G. M. Duyk, and D. P. Corey. 1994. Molecular cloning of myosins from bullfrog saccular macula: a candidate for the hair cell adaptation motor. Auditory Neurosci. 1: 63-75.
Walker, R. G., and A. J. Hudspeth. 1996. Calmodulin controls adaptation of mechanoelectrical transduction by hair cells of the bull-frog's sacculus. Proc. Natl. Acad. Sci. USA 93: 2203-2207.
Yamoah, E. N., and P. G. Gillespie. 1996. Phosphate analogs block hair-cell adaptation by inhibiting adaptation-motor force production. Neuron 17: 523-533.
Zhao, Y.-d., E. N. Yamoah, and P. G. Gillespie. 1996. The hair cell's tip links rapidly regenerate. Proc. Natl. Acad. Sci. USA, 93: 15469-15474.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||The Future of Aquatic Research in Space: Neurobiology, Cellular and Molecular Biology|
|Author:||Gillespie, Peter G.|
|Publication:||The Biological Bulletin|
|Date:||Feb 1, 1997|
|Previous Article:||Myosin drives retrograde F-actin flow in neuronal growth cones.|
|Next Article:||Choreography of the squid's "nuptial dance."|