Cell Motility and the Cytoskeleton 39:159–165 (1998) Immunolocalization of Myosin Ib in the Hair Cell’s Hair Bundle Anne B. Metcalf* Laboratory of Sensory Neuroscience, Rockefeller University, New York, New York The hair bundle, the hair cell’s sensory organelle, transduces acoustical or vestibular stimulation into a change in membrane potential. The actin-based stereociliary processes of the hair bundle contain a number of myosin isoforms that may be important to the bundle’s function. One of these isoforms, myosin Ib, has been proposed to constitute an adaptation motor controling sensitivity of the hair bundle to mechanical displacement. To gain insight into myosin Ib’s function, its distribution within the hair bundle was examined. A polyclonal antibody was produced that recognizes a protein surface loop within the head domain of myosin Ib. This antibody was used to localize myosin Ib in the hair cell by indirectimmunofluorescence microscopy and indirect-immunoelectron microscopy. Within the hair bundle, myosin Ib immunoreactivity was located along the sides of the stereociliary actin core, concentrated in the distal two-thirds of the stereocilia. Within the hair cell soma, myosin Ib immunoreactivity was located throughout the cytoplasm exclusive of the cuticular plate. Cell Motil. Cytoskeleton 39:159–165, 1998. r 1998 Wiley-Liss, Inc. Key words: adaptation; auditory system; molecular motor; vestibular system INTRODUCTION The sensory cells of the ear are hair cells. These cells detect mechanical deflection of the hair bundle induced by acoustical or vestibular stimulation. The hair bundle is a beveled conical array of long cylindrical extensions from the cell’s apical surface, composed of 20–300 actin-based stereocilia and one tubulin-based kinocilium located at the bundle’s edge. (For a review of hair cell structure, see Jacobs and Hudspeth 1990.) Rows of stereocilia farther from the kinocilium are progressively about 500 nm shorter, which gives the top edge of the bundle its beveled appearance. Stereocilia are connected by tip links, extracellular linkages found only in the bundle’s plane of mechanical sensitivity and stretching from the tip of each stereocilium to a higher point on the side of its tallest neighbor. The top of each tip link terminates in a pucker in the plasma membrane backed by the insertional plaque, an electron-dense structure between the plasma membrane and the stereociliary actin core. The membrane of each stereocilium is connected to the parallel, cross-linked actin filaments within its core by the insertional plaque and by thin linkages located along its height. r 1998 Wiley-Liss, Inc. Each stereocilium has a uniform diameter along most of its length but narrows to 100 nm at a constriction called the pivot within 1 µm of its insertion into the soma. The cluster of about 50 actin filaments that pass through the stereociliary pivot descend a few nanometers and anchor each stereocilium in the cuticular plate, a thick, highly cross-linked meshwork of cortical actin analogous to a terminal web. The cuticular plate’s stiffness and concave surface stabilize the conical shape of the hair bundle. When the bundle is deflected toward the tallest stereocilia, the hair cell depolarizes. This leads to an increase in the rate of neurotransmission at the cell’s afferent synapses. Conversely, when the bundle is deflected toward the shortest stereocilia, the cell hyperpolarContract grant sponsor: National Institutes of Health; Contract grant number: DC00241 *Correspondence to: Anne B. Metcalf (current address): Department of Cell Biology and Neuroscience, University of Texas, Southwestern Medical Center, Dallas, TX 75235-9111 Received 6 August 1997; accepted 20 October 1997 160 Metcalf izes and the rate of neurotransmission decreases. When the hair bundle is held deflected at a constant position, the electrical response of the cell diminishes over time to a steady level; this process is termed adaptation. The gating-spring model of mechanoelectrical transduction and adaptation, formulated to explain current experimentation but not verified in detail, delineates how the hair bundle converts mechanical displacement to a change in ion-channel open probability [Howard and Hudspeth, 1987, 1988; Assad and Corey, 1992; Hudspeth and Gillespie, 1994]. Acoustical and vestibular stimulation causes deflection of the hair bundle and alters transduction channel open probability, resulting in a fast change in receptor current. It is proposed that the tip link, the extracellular linkage connecting the tip of each stereocilium to the insertional plaque on its tallest neighbor, is connected directly to transduction channels and functions as a gating spring. When the bundle is displaced, stereociliary tips slide with respect to one another, and the resulting change in tip-link tension is hypothesized to control gating. Additionally, when bundle deflection is prolonged, the receptor current adapts. This adaptation is thought to be achieved by a motor element, the adaptation motor, which is proposed to comprise a group of myosin molecules positioned at the insertional plaque. The adaptation motor slipping down and climbing up the stereociliary core sets the resting position of the insertional plaque and therefore the tension under which the tip link is held in the absence of hair bundle deflection. During adaptation, a disequilibrium in motor slipping and climbing alters the insertionalplaque position and the tip-link tension. Consistent with a myosin-based adaptation motor, adaptation is blocked by dialysis of hair cells with the nucleotide analog ADPbS as well as the phosphate analogues vanadate, sulfate, and beryllium fluoride [Gillespie and Hudspeth, 1993; Yamoah and Gillespie, 1996]. Biochemical and immunohistochemical analysis indicates that hair bundles contain several members of the myosin family, a large group of proteins that convert the binding and hydrolysis of ATP to force production or directional movement along actin filaments [Gillespie et al., 1993; Hasson et al., 1997]. Photocross-linking of frog hair-bundle proteins to vanadium-ion-trapped nucleotide under a variety of conditions identified three potential myosin isotypes [Gillespie et al., 1993]. One of these, a protein of molecular mass 120 kD, is recognized by the monoclonal antibody mT2, which was raised to bovine myosin Ib [Gillespie et al., 1993; Wagner et al., 1992; Reizes et al., 1994]. Subsequent molecular cloning of the frog myosin Ib (amphibian myosin Ib or AMIb) and epitope mapping of the mT2 binding site have confirmed the identification of the 120-kD hair-bundle myosin as myosin Ib [Metcalf et al., 1994; Solc et al., 1994; Metcalf, 1995]. Immunofluorescence experiments indicated that myosin Ib is located throughout the stereocilia of the hair bundle and concentrated near the bundle’s beveled top edge, in the vicinity of the tip links [Gillespie et al., 1993; Hasson et al., 1997]. To confirm and expand the study of myosin Ib’s location and function within the hair bundle, affinitypurified polyclonal antibody was produced as a reagent for both immunolocalization of myosin Ib in ultrastructural experiments and inhibition of myosin Ib’s mechanochemical cycle in functional experiments. Antiserum was raised to a peptide with a sequence identical to that of residues 485–505 of AMIb, a segment of the protein corresponding to a flexible, surface-exposed loop in the crystal structures of the head domain of myosin II [Rayment et al., 1993b; Fisher et al., 1995]. Modeling of the interaction of myosin II with F-actin suggested a potential interaction between this loop and the actin monomer one actin-helical turn away from the primary site of actomyosin interaction on the actin filament [Rayment et al., 1993a; Schroder et al., 1993]. This paper presents immunolocalization studies describing the subcellular distribution of myosin Ib in the hair cell of the bullfrog. MATERIALS AND METHODS Antibody Production The peptide HPHFVTHKLGDQKTRKVLGRDC, which corresponds to amino acids 485–505 of AMIb plus a terminal cysteine, was synthesized. The peptide was reduced by passage through an immobilized reductant column (Reduce-Imm; Pierce, Rockford, IL) following the manufacturer’s instructions and coupled to maleimideactivated keyhold-limpet hemocyanin (KLH) in 0.085 M sodium phosphate, 0.9 M sodium chloride, 0.05 M EDTA, and 0.02% sodium azide at pH 7.2. Maleimide-activated KLH is KLH coupled to the heterobifunctional cross-linker sulfo succinimidyl 4-[N-maleimidomethyl]cyclohexane1-carboxylate (Pierce). Every 2 weeks for 18 weeks, rabbits were bled and immediately injected with 100 µg peptide-coupled KLH in Freund’s adjuvant, 0.041 M sodium phosphate, and 0.45 M sodium chloride at pH 7.2. Antipeptide antibody was purified from serum by affinity chromatography on a protein-A affinity column, followed by affinity chromatography on a peptide-affinity column. Antibody was applied to a prewashed protein-A column by three gravity-flow passages of serum. The column was washed with 10 mM Tris at pH 8.2, and the bound proteins were eluted with 100 mM sodium citrate at pH 3.0. After neutralization to pH 8.0 with NaOH, the pooled protein-containing eluant fractions were applied to washed peptide columns, which were produced by coupling immunizing peptide to a sulfo-link matrix Immunolocalization of Myosin Ib in Hair Cell’s Hair Bundle (Pierce) following the manufacturer’s instructions. Peptide-specific antibody was eluted from the peptideaffinity column and neutralized as described above for the protein-A affinity column. The specificity of the affinity-purified polyclonal antibody, called I4, was analyzed by assessing its reactivity with SDS solubilized protein from saccular maculae in immunoblot analysis and comparing this reactivity to that of the monoclonal antibody mT2. The saccular macula is the sensory epithelium of the sacculus, one of the vestibular organs. Saccular-macula protein was resolved on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels and transferred to PVDF in 10 mM CAPS at pH 11. The membrane-bound protein was detected by first blocking with 5% nonfat dried milk in phosphate-buffered saline (PBS) and then incubating sequentially with primary antibody followed by goatantirabbit IgG antibody conjugated to alkaline phosphatase or, in the case of mT2, goat-antimouse IgG antibody conjugated to alkaline phosphatase. After each antibody incubation, the membranes were washed with PBS. Chemiluminescent detection was performed using CDPstar as phosphatase substrate (Tropix, Bedford, MA). Immunofluorescence Experiments Immunofluorescence experiments were performed on both epithelial sheets and isolated hair cells from the frog’s sacculus. To produce isolated epithelia, saccular maculae were dissected in standard saline solution (108 mM NaCl, 2 mM KCl, 4 mM CaCl2, 3 mM D-glucose, and 5 mM HEPES at pH 7.25) and digested for 1 hr in 500 µg/ml adipocyte collagenase (Sigma type II) in standard saline solution (Sigma, St. Louis, MO). The epithelium was peeled from the basement membrane with an eyelash and attached to coverslips coated with Cell-tak (Collaborative Biomedical Products, Bedford, MA). The otolithic membrane was removed from the hair bundles with forceps after the epithelial sheet was attached to the coverslip. Hair cells were isolated from saccular maculae by a previously described procedure, with some modifications [Lumpkin and Hudspeth, 1995]. Maculae were dissected into saline solution lacking added Ca21, then incubated for 12 min in 1 mM EGTA, 1 mM MgCl2, and 25 µg/ml bacterial protease (Sigma type XXIV) in the above saline solution. After transfer to fresh saline solution lacking added Ca21, the procedure was finished by removing the otolithic membrane with forceps and brushing hair cells from the epithelium onto coverslips coated with concanavalin A, using an eyelash. Both preparations were processed similarly. The tissue or cells were fixed with 1.3 M formaldehyde in PBS for 15–120 min. The specimen was rinsed 5–6 times with PBS, then permeabilized and blocked for 30–120 161 min with PBS containing 12.5% normal goat serum, 1.5 M NaCl, and 0.16% Triton X-100. Samples were exposed sequentially to primary and secondary antibodies diluted in blocking solution containing 0.024% Tween-20 and lacking Triton X-100. Each antibody incubation lasted 60–120 min and was followed by PBS rinses. RhodamineTRITC conjugated goat-antirabbit IgG antibody was used as a secondary antibody. Immunoelectron Microscopy Experiments Samples for electron microscopy were labeled with antibody prior to fixation with glutaraldehyde and processing for electron microscopy. Two separate formaldehyde fixation and permeabilization protocols were used prior to antibody labeling. In most experiments, isolated maculae were fixed for 30 min with 1.3 M formaldehyde and 1.3 µM phalloidin in PBS. After rinsing, the tissue was permeabilized and blocked for 90 min with PBS containing 12.5% normal goat serum, 1.5 M NaCl, 0.16% Triton X-100, and 1.3 µM phalloidin. Saccular maculae were then incubated overnight at 4°C in primary antibody in the same solution lacking phalloidin. After rinsing with PBS, samples were incubated for 4 hr with goat-antirabbit IgG antibody coupled to colloidal gold. This procedure removed most or all of the plasma membrane and gave good antibody access to the cytoplasm. Some samples were fixed and permeabilized in an hypo-osmotic solution, followed by antibody incubations in a milder detergent. This procedure led to a better preservation of permeabilized membranes. Samples were fixed and permeabilized for 30 min with 1.3 M formaldehyde, 0.1% Triton X-100, 1.3 mM phalloidin, and 50 mM phosphate buffer at pH 6.3, then blocked and permeabilized for 30 min with 12.5% goat serum, 1.5 mM MgCl2, 0.1% Triton X-100, and 20 mM Pipes at pH 6.8. Samples were then incubated with antibody as described above, except that 0.024% Tween-20 was used for the antibody incubations instead of 0.16% Triton X-100. Samples from both labeling protocols were transfered to 100 mM cacodylate buffer and fixed overnight at room temperature in 100 mM glutaraldehyde, 500 mM formaldehyde, 3 mM tannic acid, and 100 mM cacodylate buffer at pH 7.4. Saccular maculae were subsequently fixed for 1 hr at 4°C in 50 mM OsO4 and 100 mM cadodylate buffer at pH 7.4. Samples were then dehydrated and stained at room temperature. Dehydration was accomplished by immersion in progressively increasing concentrations of ethanol to 70%. At this point, the samples were stained by incubation in 0.5% uranyl acetate, 70% ethanol for 60 min. Then dehydration was completed by immersion in progressively increasing concentrations of ethanol to 100%. The 100% ethanol incubation was extended to 60 min and was followed by two incubations in 100% propylene oxide for 60 min 162 Metcalf Fig. 1. Immunoblot analysis of saccular macula protein labeled with antibody I4 (A,B). Each lane contains SDS-solubilized protein from 0.75 macula separated on a 12% acrylamide SDS-PAGE gel. mT2 (A) and I4 (B) recognize the 120-kD myosin Ib in saccular-maculae protein. The molecular weight markers between the two blots show 144, 87, 44.1, 32.7, and 17.7 kD. C–N: Indirect-immunofluorescence localization of myosin Ib in hair cells. Pairs of light microscopic images obtained with a Zeiss Axiovert-135, 363 1.4 NA lens are shown. I4 labeling of hair cells visualized with a rhodamine-TRITC conjugated secondary is shown in the left panel of each pair. The right panel contains the corresponding differential interference contrast (DIC) image. C,D: Apical surface of a saccular macula and were acquired with a charge-coupled-device camera. E–L: Views of hair bundles from isolated hair cells. M,N: Whole cell images of an isolated hair cell. E–N: Acquired using a confocal microscope with an 363, 1.4 numerical aperture (NA) lens. Scale bars 5 5 µm. each. Saccular maculae were embedded by rotating for 24 hr in 50% propylene oxide, 50% plastic mixture comprising 45.5% Embed-812 resin (Electron Microscopy Sciences), 27.7% dodecenyl succinic anhydride, 25.6% nadic methyl anhydride, and 1.2% 2,4,6-tri(dimethylaminomethyl)phenol. After incubation in 100% plastic mixture for 24 hr, samples were transfered to fresh plastic mixture and cured at 50°C for 2 days. tissue known to contain at least nine other myosin isoforms [Solc et al., 1994]. Preimmune sera did not bind any proteins in the saccular macula (data not shown). In initial immunofluorescence experiments, I4 labeled hair bundles more strongly and consistently than did mT2. Therefore, I4 was used exclusively for the rest of the study. Myosin Ib’s location within the hair cell was first investigated with light-microscopy. Observation of the apical surface of the sensory epithelium using a compound microscope allowed views of the three major actin networks of the hair cell: the hair bundle, the cuticular plate, and the actin ring at the adhering belt. Experiments with isolated sensory epithelia indicated that myosin Ib immunoreactivity was present in hair bundles and absent from, or present at a very low level in, cuticular plates (Fig. 1C–D). Although myosin Ib immunoreactivity was present in the region of the adhering belt, it was not enriched there with respect to the surrounding cytoplasm (Fig. 1C–D). Confocal images of isolated hair cells afford an excellent view of hair bundles. Myosin Ib immunoreactivity was present in the stereocilia but not in the kinocilium of hair bundles (Fig. 1E–L). Within stereocilia, myosin Ib RESULTS To determine myosin Ib’s distribution within the hair bundle, a polyclonal antibody was produced, affinitypurified, and used in immunolocalization studies. Antibodies were raised in rabbit to a peptide corresponding to amino acids 485–505 of AMIb. The peptide sequence is quite specific for myosin Ib; the antibody, therefore, is not expected to recognize other proteins, including other myosins. The specificity of the polyclonal antibody I4 was assessed by analyzing its reactivity with frog saccularmaculae proteins by immunoblot analysis and comparing this reactivity with that of the monoclonal antibody mT2 [Wagner et al., 1992]. Both mT2 (Fig. 1A) and I4 (Fig. 1B) recognized only myosin Ib in the saccular macula, a Immunolocalization of Myosin Ib in Hair Cell’s Hair Bundle immunoreactivity was concentrated in the top two-thirds of the bundle. In some cells, myosin Ib immunoreactivity was concentrated in just the top third of the bundle (Fig. 1E–H). The image plane in Figure 1E–K is close to the surface of the cell and passes mostly or entirely above the cuticular plate. The immunoreactivity seen in the vicinity of the cuticular plate is within the cytoplasm surrounding the cuticular plate. Full-cell images of isolated hair cells showed myosin Ib immunoreactivity throughout the cytoplasm but absent from the nucleus (Fig. 1M–N). These images confirmed a dearth of myosin Ib immunoreactivity in the cuticular plate and in the region immediately above the stereociliary pivots. Hair cells were not labeled by an irrelevant primary, preimmune serum, or secondary antibody alone (data not shown). Immunoreactivity could be eliminated by preincubation of antibody with an excess of the immunizing peptide (data not shown). Myosin Ib’s localization within the stereocilia was investigated at higher resolution by immunoelectron microscopy. Within stereocilia, myosin Ib immunoreactivity was strictly limited to the edge of the actin core (Fig. 2A–E). Immunoreactivity was scattered along the length of the stereocilia, with more labeling found in the top two-thirds than in the bottom third of the bundle. The low level of labeling seen in the bottom third of stereocilia was significant as hair cells were not labeled in this or other regions by secondary antibody alone (Fig. 2F). No myosin Ib immunoreactivity was seen at the pivots, in the kinocilium, or in the cuticular plate. This pattern of labeling is entirely consistent with the results described above from light microscopy. Very little labeling was observed at the tips of stereocilia, where the actin filaments terminate. Some immunoreactivity was seen in a position where insertional plaques might be located, about 100–500 nm above the next shorter stereocilium along the beveled edge (Fig. 2A–C). However, the clusters of gold seen at these positions were not larger, nor were they observed more frequently than those seen elsewhere along the sides of stereocilia. The two fixation protocols yielded qualitatively equivalent results but a uniformly lower degree of labeling was found when more membrane was preserved. DISCUSSION Immunolabeling with the affinity-purified antibody I4 indicates that myosin Ib is located along the sides of the stereociliary actin core and at positions that might correspond to those of the insertional plaque. Myosin Ib may therefore contribute to the lateral linkages seen between the actin core and the stereociliary plasma membrane [Hirokawa and Tilney, 1982]. The function of 163 Fig. 2. Immunoelectron-microscopic localization of myosin Ib in stereocilia. Transmission electron micrographs of hair bundles subjected to pre-embedding, indirect immunolabeling of myosin Ib with the polyclonal serum I4. Images are of sections in the plane of mechanical sensitivity. The kinocilia are located to the right in the images shown. The beveled edge of the bundle is in or above the upper left corner of the panels. Arrowheads indicate the positions of gold particles. Myosin Ib immunoreactivity is located along the sides of the actin core of stereocilia (D,E) and in positions that might correspond to those of the insertional plaque (A–C). No labeling was seen when primary antibody was omitted (F). The experiments shown in B,C, and F were fixed and permeabilized by the second procedure, which resulted in the preservation of more membrane (see Methods). The overall distribution of the label was similar, using either permeabilization procedure, although the panels presented only show labeling at positions that might correspond to those of the insertional plaque. Scale bar 5 200 nm. 164 Metcalf these lateral linkages is unclear, although they may contribute to the mechanical integrity of the stereocilium. The gating-spring model of mechanoelectrical transduction and adaptation proposes a specific function for an unconventional myosin located at the insertional plaque [Howard and Hudspeth, 1987, 1988; Assad and Corey, 1992; Hudspeth and Gillespie, 1994]. Results from mechanical and electrophysiological studies of transduction and adaptation have been used to formulate a quantitative model that predicts that the adaptation motor is formed by an accumulation of myosin molecules at the insertional plaque; this model predicts that 50 or more myosin molecules constitute the adaptation motor [Howard and Hudspeth, 1987, 1988; Assad and Corey, 1992; Hudspeth and Gillespie, 1994]. In the current study, the permeabilization conditions required to allow antibody access to the cytoplasm removed the tip links between adjacent stereocilia, precluding direct identification of the insertional plaques. Myosin Ib immunoreactivity was found at positions that might correspond to those of the insertional plaque, but there was no compelling increase in myosin Ib immunoreactivity at these positions when compared to other points along the sides of the stereocilium. However, two technical effects related to epitope masking could potentially reduce the sensitivity of myosin Ib immunolabeling at the position of the insertional plaque compared to elsewhere. First, the accessibility of the I4 epitope may depend on the myosin Ib molecule’s interaction with actin filaments. Structural models of myosin II interaction with actin suggest that the I4 binding site could be located in close apposition to actin when myosin is tightly bound to actin [Rayment et al., 1993a; Schroder et al., 1993]. In the absence of ATP and in the presence of actin, myosin enters the rigor state where it is tightly bound to actin. The insertional plaque may hold the myosin molecules of the adaptation motor in register with the actin filaments, making these myosin molecules more likely to bind tightly to actin in our permeabilized, fixed preparation compared to the myosin molecules scattered elsewhere along the sides of stereocilia. The sensitivity of myosin immunolabeling at the insertional plaque could also be reduced as a result of dense packing in the cluster of myosin Ib molecules, which is predicted by the adaptation-motor model. If this prediction is correct, the close packing of myosin Ib molecules could potentially restrict access to the I4 epitope on the side of the myosin head in some molecules in the cluster. However, although these epitope masking effects could reduce the sensitivity of AMIb immunolabeling at the insertional plaque, there is no proof that such effects are operative in the experiments reported here. Previous studies with antibodies that bind to other regions of the myosin Ib molecule gave immunohisto- chemical results with significant differences compared to those observed with I4. However, in addition to epitope masking, antigenic cross-reactivity, and nonspecific reactivity are also potential hazards in immunolocalization studies, and all these factors can result in different staining patterns between antibodies. Therefore, it is essential to compare the staining patterns of a number of antibodies to the same molecule. The monoclonal antibody mT2 binds amino acids 813–904 within the tail domain of AMIb [Wagner et al., 1993; Metcalf, 1995]. By immunofluorescence localization, mT2 immunoreactivity was found scattered throughout the bundle as well as concentrated near the bevel [Gillespie et al., 1993]. In contrast, I4 immunoreactivity was also scattered within the bundle but was not concentrated near the bevel. The position of mT2 immunoreactivity with respect to the insertional plaque has not been reported at higher resolution. The affinity-purified polyclonal antibody rafMIb is directed against a fusion protein containing amino acids 899–1028 of AMIb, which comprises the molecule’s tail domain between the mT2 binding site and the C-terminus [Hasson et al., 1997]. The distribution of rafMIb immunoreactivity was investigated using both light and electron microscopy. Within stereocilia, it was found concentrated in the distal third, a pattern similar to that observed with I4. However, in contrast with I4 staining, rafMIb immunoreactivity was seen in two other locations within stereocilia. First, rafMIb immunoreactivity, but never I4 immunoreactivity, was seen at the pivots. Second, immunofluorescence experiments showed intense rafMIb immunoreactivity at the tips of the stereocilia, and EM studies indicated that this concentration of immunoreactivity was located at the very top of the stereocilium, above the ends of the actin filaments. The function of a population of myosin Ib molecules clustered in this position is unclear, although it might represent motor molecules no longer associated with actin or motor molecules which have re-associated with internal filaments at the very top of the stereocilium after dissociating from their physiological tracks at the outer edge. If this population is not associated with actin, conformational masking of the I4 epitope is unlikely, making it unclear why I4 would not detect an AMIb population in this position. On the other hand, partial degradation of the myosin Ib in this location could account for the differences in the pattern of immunoreactivity if the head is degraded more quickly than the tail. On the sides of the stereocilium, some clustering of rafMIb was seen at positions that might correspond to those of the insertional plaque, although this staining was not as prominent as that at the very tips of stereocilia and was not responsible for the increased immunoreactivity seen at the tips by light microscopy. Immunolocalization of Myosin Ib in Hair Cell’s Hair Bundle An additional difference between the staining pattern of the two antibodies is observed in the cell body. The immunoreactivity of both I4 and rafMIb immunoreactivity was seen throughout the cytoplasm exclusive of the nucleus and the cuticular plate. However, rafMIb immunoreactivity was concentrated between the cuticular plate and the actin ring of the adhering belt, and electron microscopic investigation indicated this immunoreactivity was clustered on vesicles. Although I4 immunoreactivity was also seen on vesicles in this region, there was no net increase in I4 immunoreactivity between the cuticular plate and the adhering belt when compared to the rest of the cytoplasm. Although significant questions remain as to the exact pattern and density of the subcellular distribution of myosin Ib, the results presented here and in the other works described above are consistent with a myosin Ib-based adaptation motor. Further investigation of myosin Ib’s role in adaptation would be aided by an inhibitory antibody. I4 binding to myosin Ib might be expected to inhibit myosin function. If this is the case, I4 could be a useful reagent for functional inhibition studies of myosin Ib both in vitro and in vivo. ACKNOWLEDGMENTS I thank Dr. A.J. Hudspeth for support and encouragement throughout this project. Thanks to Dr. E.A. Lumpkin for help in acquiring confocal images and to Drs. S. Burlacu, A.J. Hudspeth, J.F. Hunt, H. Krämer, and E.A. Lumpkin for critically reading the manuscript. This work was supported by grant DC00241 from the National Institutes of Health (to A.J.H.). REFERENCES Assad, J.A., and Corey D.P. (1992): An active motor model for adaptation by vertebrate hair cells. J. Neurosci. 12:3291–3309. Fisher, A.J., Smith, C.A., Thoden, J., Smith, R., Sutoh, K., Holden, H.M., and Rayment, I. (1995): Structural studies of myosin: Nucleotide complexes: A revised model for the molecular basis of muscle contraction. Biophys. J. 68, 19s–28s. Gillespie, P.G., and Hudspeth, A.J. (1993): Adenine nucleotide diphosphates block adaptation of mechanoelectrical transduction in hair cells. Proc. Natl. Acad. Sci. U.S.A. 90:2710–2714. Gillespie, P.G., Wagner, M.C., and Hudspeth, A.J. (1993): Identification of a 120 kD hair-bundle myosin located near stereociliary tips. Neuron 11:581–594. 165 Hasson, T., Gillespie, P.G., Garcia, J.A., MacDonald, R.B., Zhao, Y.-D., Yea, A.G., Mooseker, M.S., and Corey, D.P. (1997): J. Cell Biol. 137:1287–1307. Hirokawa, N., and Tilney, L.G. (1982): Interactions between actin filaments and between actin filaments and membranes in quick-frozen and deeply etched hair cells of the chick ear. J. Cell Biol. 95:249–261. Howard, J., and Hudspeth A.J. (1987): Mechanical relaxation of the hair bundle mediates adaptation in mechanoelectrical transduction by the bullfrog’s saccular hair cell. Proc. Natl. Acad. Sci. U.S.A. 84:3064–3068. Howard, J., and Hudspeth A.J. (1988): Compliance of the hair bundle associated with the gating of mechanoelectrical transduction channels in the bullfrog’s saccular hair cell. Neuron 1:189–199. Hudspeth, A.J., and Gillespie, P.G. (1994): Pulling springs to tune transduction: adaptation by hair cells. Neuron 12:1–9. Jacobs, R.A., and Hudspeth, A.J. (1990): Ultrastructural correlates of mechanoelectrical transduction in hair cells of the bullfrog’s internal ear. Cold Spring Harb. Symp. Quant. Biol. 55:547–561. Lumpkin, E.A., and Hudspeth A.J. (1995): Detection of Ca21 entry through mechanosensitive channels localizes the site of mechanoelectrical transduction in hair cells. Proc. Natl. Acad. Sci. U.S.A. 92:1506–1509. Metcalf, A.B. (1995): Molecular characterization of amphibian myosin Ib, a candidate for the hair bundle’s adaptation motor. Thesis, University of Texas Southwestern Medical Center, Dallas, TX. Metcalf, A.B., Chelliah, Y., and Hudspeth, A.J. (1994): Molecular cloning of a myosin Ib isozyme that may mediate adaptation by hair cells of the bullfrog’s internal ear. Proc. Natl. Acad. Sci. U.S.A. 91:11821–11825. Rayment, I., Holden, H.M., Whittaker, M., Yohn, C.B., Lorenz, M., Holmes, K.C., and Milligan, R.A. (1993a): Structure of the actin–myosin complex and its implications for muscle contraction. Science 261:58–65. Rayment, I., Rypniewski, W.R., Schmidt-Base, K., Smith, R., Tomchick, P.R., Benning, M.M., Winkelmann, D.A., Wesenberg, G., and Holden, H.M. (1993b): Three dimensional structure of myosin subfragment-1: A molecular motor. Science 261:50–58. Reizes, O., Barylko, B., Li, C., Südhof, T.C., and Albanesi, J.P. (1994): Domain structure of a mammalian myosin Ib. Proc. Natl. Acad. Sci. U.S.A. 91:6349–6353. Schroder, R.R., Manstein, D.J., Jahn, W., Holden, H., Rayment, I., Holmes, K.C., and Spudich, J.A. (1993): Three-dimensional atomic model of F-actin decorated with Dictyostelium myosin S1. Nature 364:171–174. Solc, C.K., Derfler, B.H., Duyk, G.M., and Corey D.P. (1994): Molecular cloning of myosins from the bullfrog saccular macula: A candidate for the hair cell adaptation motor. Aud. Neurosci. 1:63–75. Wagner, M.C., Barylko, B., and Albanesi, J.P. (1992): Tissue distribution and subcellular localization of mammalian myosin I. J. Cell Biol. 119:163–170. Yamoah, E.N., and Gillespie, P.G. (1996): Phosphate analogs block adaptation in hair cells by inhibiting adaptation-motor force production. Neuron 17:523–533.