Freeze fracture analysis of apical membranes in cochlear cultures: differences between basal and apical-coil outer hair cells and effects of neomycin

Previous studies have shown that exposure of cochlear cultures to the ototoxic aminoglycoside antibiotic neomycin leads to the rapid formation of numerous membrane filled protrusions on the apical surface of the hair cells but not on the surrounding supporting cells, and that hair cells in basal-coil cultures are more sensitive to these effects of neomycin than those in the distal end of apical-coil cultures. Freeze-fracture has been used to examine and compare the apical surfaces of hair cells and supporting cells in basal and apical-coil cultures in order to look for features that may explain the differential sensitivity of the various cell types to neomycin, and to characterize the membrane type that forms in response to neomycin and compare it with the normal apical membrane of the hair cell. The apical surface of the highly responsive basal-coil outer hair cells differs significantly from the apical surfaces of apical-coil outer hair cells and supporting cells in both regions of the cochlea in both surface area and the number and density of endocytotic vesicles associated with this surface. Basal-coil hair cells have an average of 120 +/- 39 vesicles per cell surface and a density of 3.5 +/- 0.89 vesicles per microns 2, whereas apical-coil hair cells have 14.8 +/- 15.8 vesicles/cell surface and density of 0.73 +/- 0.72 vesicles per microns 2. There are no significant differences in intramembrane particle (IMP) density on the apical surfaces of all the cell types examined, and qualitative observations of filipin-treated specimens indicate that cholesterol densities are also similar. The membrane that accumulates in response to neomycin treatment at the apical pole of the hair cell is IMP free, does not respond to filipin, and fractures in a manner that is indicative of a high content of unsaturated phospholipid in a fluid phase, and is therefore different in several respects from the normal apical surface of the hair cell. The results of this study suggest that apical surface associated endocytotic vesicle numbers may determine the differential sensitivity of apical and basal-coil hair cells to neomycin, and that neomycin may interfere with some aspect of phospholipid metabolism or membrane turnover in sensory hair cells.

Regenerative proliferation in inner ear sensory epithelia from adult guinea pigs and humans

Supporting cells in the vestibular sensory epithelia from the ears of mature guinea pigs and adult humans proliferate in vitro after treatments with aminoglycoside antibiotics that cause sensory hair cells to die. After 4 weeks in culture, the epithelia contained new cells with some characteristics of immature hair cells. These findings are in contrast to expectations based on previous studies, which had suggested that hair cell loss is irreversible in mammals. The loss of hair cells is responsible for hearing and balance deficits that affect millions of people.

Ultrastructural evidence for hair cell regeneration in the mammalian inner ear

It has long been thought that hair cell loss from the inner ears of mammals is irreversible. This report presents scanning electron micrographs and thin sections of the utricles from the inner ears of guinea pigs that show that, after hair cell loss caused by treatment with the aminoglycoside gentamicin, hair cells reappeared. Four weeks after the end of treatment, a large number of cells with immature hair bundles in multiple stages of development could be identified in the utricle. Thin sections showed that lost type 1 hair cells were replaced by cells with a morphology similar to that of type 2 hair cells. These results indicate an unexpected capacity for hair cell regeneration in vivo in the mature mammalian inner ear.

Membrane stains as an objective means to distinguish isolated inner and outer hair cells

The use of isolated cochlear outer and inner hair cells has become widespread. While the morphological features of these two cell types in general are sufficiently different to allow discrimination, there are situations where confusion can arise. Small outer hair cells, particularly when they are swollen or distorted, can take on an appearance suggestive of inner hair cells. We describe here two fluorescent membrane stains, 3,3'-dihexyloxacarbocyanine iodide and rhodamine B hexyl ester, as an objective means to distinguish between cochlear hair cell types. Both stains mark the subsurface cisternae of outer hair cells thereby delineating the cell outline, and the interior of the cell shows discrete structure. On the other hand, in inner hair cells, the outline of the cell is not resolved while the interior is diffusely fluorescent. Since the two probes have different excitation and emission wavelengths (fluorescein- and rhodamine-like, respectively), this staining procedure can even be used in the presence of another fluorescent marker (for example, a calcium-indicating dye) by appropriate choice of the membrane stain.

Structural variability of the sub-surface cisternae in intact, isolated outer hair cells shown by fluorescent labelling of intracellular membranes and freeze-fracture

The intracellular membrane systems in intact, isolated outer hair cells were visualised using the fluorescent membrane probe 3,3'-dihexyloxacarbocyanine iodide (DiOC6) and by freeze-fracture, and f-actin distribution was examined with rhodamine-phalloidin. DiOC6 stained the sub-surface cisternal membranes in the lateral wall and revealed a membrane system running in the centre of the cell from the nucleus to the sub-cuticular region. In optical sections of the lateral wall of fluorescently labelled cells, obtained by scanning laser confocal microscopy, the sub-surface membrane appeared as a fenestrated sheet or a fine network of tubules. Freeze-fracture replicas of rapidly-frozen, unfixed outer hair cells also showed the sub-surface membrane as a fenestrated sheet in some cells or as a network of tubules in others. These combined studies indicate that the interruptions within the cisternal membranes as seen in normal thin sections of outer hair cells are not fixation artefacts but may reflect the dynamic and plastic properties of this membrane system. Double staining of cells with rhodamine-phalloidin and DiOC6 showed substantial co-localisation of intracellular membranes and f-actin. The results suggest there may be a continuous, dynamic endoplasmic reticulum system, forming a core in the centre of the cell, broadening in the subcuticular region and extending down the lateral wall, that may have a role in the turnover and distribution of cytoskeletal assemblies within the outer hair cell.

Structural features of the lateral walls in mammalian cochlear outer hair cells

Freeze-fracture, freeze-etching and thin sections have been used to determine features of the structural organisation of the lateral walls in cochlear outer hair cells. The presence of an organised meshwork of filaments in the lateral cortex of the cell is confirmed in intact unfixed cells. This meshwork showed morphological features similar to the cytoskeletal lattice. The lateral plasma membrane is shown to be protein-rich and to contain cholesterol. The membranes of the subplasmalemmal lateral cisternae contain much less protein, and little cholesterol as judged by their responses to filipin and tomatin. These findings indicate differences in the physical properties of the two membrane systems. On the fracture faces of the plasma membrane there is a high density of intramembrane particles and this particle population is heterogeneous. Some particles show morphological features consistent with those of transmembrane channels. Regularly spaced pillars crossing the space between the plasma and cisternal membranes were identified both in thin sections and in freeze-etched preparations, but neither the plasma nor cisternal membrane fracture faces showed any feature corresponding directly to the pillar. This suggests the pillars do not insert directly into either membrane. Freeze-fracture and freeze-etching of unfixed cells indicated that the pillar is indirectly associated with the cytoplasmic surface of the plasma membrane, and, at its inner end, linked to the cortical cytoskeletal lattice on the outer surface of the cisternal membrane.

Assessment of ultrastructure in isolated cochlear hair cells using a procedure for rapid freezing before freeze-fracture and deep-etching

Separated cochlear outer hair cells and isolated strips of organ of Corti containing hair cells and supporting cells have been rapidly frozen before freeze-fracture and deep-etching by immersion of samples sandwiched between two copper plates into liquid nitrogen-cooled propane: isopentane. Assessment of this procedure has shown that no significant freezing damage occurs. The ultrastructure of the hair cells revealed by freeze-fracture of these non-chemically fixed preparations was generally very similar to that seen in fixed material. This indicates that the processing of cochlear tissue normally used for electron microscopy produces few obvious structural artefacts. It also demonstrated that procedures for isolating cochlear hair cells generally do not affect cell structure significantly. However, some isolated hair cells did show abnormalities within the membranes of the lateral cisternae. Such membrane alterations, which would not be identified by light microscopy, occurred to a variable extent but were more commonly present after prolonged periods in maintenance medium. Deep-etching of the preparations to examine extracellular features around stereocilia revealed clearly lateral cross-links between stereocilia. However, tip-links could not be positively identified in either unfixed or prefixed preparations.

Endocochlear potential generation is associated with intercellular communication in the stria vascularis: structural analysis in the viable dominant spotting mouse mutant

Deafness in the viable dominant spotting mouse mutant is due to a primary defect of the stria vascularis which results in absence of the positive endocochlear potential in scala media. Endocochlear potentials were measured and the structure of stria vascularis of mutants with potentials close to zero was compared with that in normal littermate controls by use of morphometric methods. The stria vascularis was significantly thinner in mutants. Marginal cells were not significantly different from controls in terms of volume density or intramembrane particle density but the network density of tight junctions was significantly reduced in the mutants. A virtual absence of gap junctions between basal cells and marginal or intermediate cells was observed, but intramembrane particle density and junctional complexes between adjacent basal cells were not different from controls. The volume density of basal cells was significantly greater in mutants. Intermediate cells accounted for a significantly smaller volume density of the stria vascularis in mutants and had a lower density of intramembrane particles than controls. Melanocytes were not identified in the stria vascularis of mutants. These results suggest that communication between marginal, intermediate and basal cells might be important to the normal function of the stria vascularis.

Neural basis for regulation of cochlear blood flow: peptidergic and adrenergic innervation of the spiral modiolar artery of the guinea pig

The spiral modiolar artery is the terminal artery in the cochlea, and as such is expected to play a major role in the control of cochlear blood flow. In this study, we examined the distribution of adrenergic and peptidergic nerve fibres on the spiral modiolar artery of the guinea pig using histofluorescence and immunofluorescence techniques. The spiral modiolar artery was dissected from the modiolus so that the entire length of the vessel and its branches, could be observed. Noradrenaline was identified using the glyoxylic acid histofluorescence technique. The presence of the vasoactive peptides substance P, calcitonin gene-related peptide (CGRP), neuropeptide Y (NPY) and vasoactive intestinal polypeptide (VIP), was investigated using antibodies against these peptides. Each putative transmitter tested yielded labelled nerve fibres throughout the length of the spiral modiolar artery and its branches. Double-labelling experiments confirmed that CGRP and substance P are contained in the same fibres but that VIP and substance P appear to be contained in different populations of fibres. These results establish that nerve fibres containing vasoactive peptides and noradrenaline supply the spiral modiolar artery and suggest that they are involved in the regulation of cochlear blood flow.

Acoustic distortion products can be used to monitor the effects of chronic gentamicin treatment

Acoustic distortion has been measured from the ear canals of guinea pigs treated with the aminoglycoside antibiotic gentamicin. The correlation between pre- and post-treatment levels of distortion can be related to the extent of outer hair cell (OHC) pathology induced by gentamicin. All animals that were treated and show functional changes also show changes in OHC morphology. Changes in distortion properties are evident before clear changes in surface hair cell morphology are seen. In the early stages of intoxication, thin sections of the organ of Corti are necessary to reveal the underlying structural changes accompanying functional impairment. Where OHCs have been affected by gentamicin treatment, the greatest change in 2f1-f2 level is for distortion generated by widely spaced stimuli. With closely spaced stimuli, 2f1-f2 may be unchanged in level, but higher order distortion components are depressed. Functional evidence of mild OHC disruption is only revealed by stimulus levels at or below 60 dB SPL. With higher sound levels, apparently normal levels of distortion can be generated by ears in which there is clear evidence of OHC pathology. In a single, untreated animal with an extensive basal turn lesion, some 2f1-f2 distortion was generated when the stimulus frequency region lacked OHCs.

The effect of gentamicin-induced hair cell loss on the tight junctions of the reticular lamina

Freeze fracture has been used to examine junctional complexes within the gentamicin damaged guinea pig organ of Corti during the period of structural reorganisation associated with hair cell loss. When examined up to 28 days after a regime of chronic gentamicin administration, no significant alterations to the pattern or nature of the complex tight junctions associated with the hair cell apex could be detected. In contrast, marked abnormalities of the junctions between adjacent supporting cells, and evidence of formation of new gap junctions on the lateral membranes of Deiter's cell was seen. The use of lanthanum as an electron dense tracer in thin sections also showed that no breakdown of permeability barriers occurred in the reticular lamina. These results support the concept of a controlled response by the organ of Corti, with conservation of the tight junction of the hair cell and maintenance of permeability barriers. Early changes to the membranes of the stereocilia and lateral cisternae of the outer hair cells were also identified. These may indicate direct effects of gentamicin upon membrane structure.

The vessels of the stria vascularis: quantitative comparison of three rodent species

The stria vascularis (SV) was quantitatively compared in three species commonly used in auditory research: guinea pig, mouse and gerbil. Measurements were obtained for surface area, cross-sectional area, length, width and thickness of SV. Surface area and length were proportional to the overall size of the cochlea in each species, but there was no significant difference between species in mean cross-sectional area. In guinea pig and mouse, there was no significant difference in thickness (endolymphatic surface to spiral ligament) and a similar pattern was observed for width (Reissner's membrane to spiral prominence): the width of SV increased from the apical end to a point 80% of the distance from the apex, then decreased to the basal end of SV. The thickness of gerbil SV was significantly less (P less than 0.001) and there was less of a gradient in width as compared to guinea pig and mouse. The vessels of SV were compared in terms of vascular density (vessels per unit area), rbc density (red blood cells per unit area), R/V (rbc density/vascular density), inter-vessel spacing and vessel diameter. Highly significant (P less than 0.001) differences between species were found in vascular density, RBC density and vessel diameter, but there were no differences between species for R/V or inter-vessel spacing. The results of this study may reflect differences in the metabolic requirements of SV among different species.

Gentamicin alters membrane structure as shown by freeze-fracture of liposomes

Freeze-fracture has been used to examine the effects of gentamicin on membrane structure in liposomes of different anionic phospholipids combined with a neutral phospholipid, phosphatidylcholine. The molar ratios of neutral: anionic lipid were 1:1 (high anionic lipid ratio) and 4:1 (low anionic lipid) and the liposomes were incubated with 0.1 mM (low) and 1 mM (high) gentamicin. With the anionic phospholipid phosphatidylinositol bisphosphate, an identifiable disruption of the membrane bilayer was observed as well as aggregation of liposomes leading to membrane fusion. These effects occurred both at low gentamicin concentration and low anionic lipid content of the liposomes; these responses were not inhibited by 1 mM Ca2+. With the other anionic lipids tested (phosphatidylserine, phosphatidylinositol and phosphatidylinositol monophosphate), only aggregation and fusion of liposomes was observed and this effect only occurred at high gentamicin concentration and high anionic lipid content. Further, 1 mM Ca2+ inhibited the responses of these other anionic lipids to gentamicin. The results demonstrate the unique character of the interaction between gentamicin and phosphatidylinositol bisphosphate and provide further support for the hypothesis that a specific binding to this lipid is a key step in the ototoxic action of aminoglycoside antibiotics. They also suggest that such an interaction in vivo might cause alterations to the structure and properties of cell membranes in the inner ear.

Characteristics of the membrane of the stereocilia and cell apex in cochlear hair cells

Freeze-fracture has been used to examine the membrane of the cell apex and of the stereocilia in cochlear hair cells. The apical (non-stereociliary) membrane of inner hair cells (IHCs) exhibited a lower density of intramembrane particles (IMP) than that of the outer hair cells (OHCs) but in both cell types the apical membrane responded to the effects of filipin. The distribution of IMP and of filipin-induced membrane deformations was uniform over the apical membranes in both IHC and OHC, thus, providing no evidence for local membrane differentiation on the non-stereociliary part of the hair cell apex. The stereociliary membranes of IHC and of OHC differed not only in the density of IMP, but also in their responses to filipin and to tomatin. IHC stereocilia responded intensely to both agents. OHC stereocilia showed a significantly lower density of filipin-induced lesions and appeared almost unaffected by tomatin. This suggests that the OHC stereocilial membrane may be structurally specialized. The membrane at the apical end of stereocilia appeared to be differentiated from the membrane of the stereociliary shaft. The tip region was free of the usual IMP and showed no filipin-induced lesions. The differentiation at the apical end was also apparent in samples which have been rapidly frozen without prior chemical fixation or cryoprotection, showing that the particle-free area was not an artefact induced by glutaraldehyde fixation. Close examination of the membrane at the apical-most tip of the stereocilium revealed the presence of a small number of large particles of 10.5-11.0 nm diameter. The occurrence of membrane differentiation localized to the tip of the stereocilium may be consistent with the suggestion that transduction channels in hair cells are situated at this point.

Analysis of structural changes in the stria vascularis following chronic gentamicin treatment

Changes in the stria vascularis following chronic gentamicin treatment have been examined using quantitative methods. Albino guinea pigs were given gentamicin at 100 mg.kg-1.day-1 subcutaneously for 10 days. Comparisons were made of strial tissue from the treated animals sacrificed either 1 h or 4 weeks following the last injection with that from saline-injected controls. Strial width (spiral prominence to Reissner's membrane) and marginal cell (MC) number across the stria were determined from scanning electron micrographs. Strial thickness (endolymphatic surface to spiral ligament) and the volume fractions of the strial components (MCs, intermediate cells (ICs), basal cells (BCs) and capillaries) were derived from thin sections. Qualitative changes to both MCs and ICs were apparent 1 h after the last injection. At four weeks post-treatment, there was a small, but statistically significant, decrease in the number of marginal cells and a highly significant decrease in strial thickness. This was almost entirely due to a highly significant decrease in the volume fraction of MCs (i.e. shrinkage). The volume fraction of ICs was increased but this could be accounted for by MC shrinkage; after allowing for the reduction in strial thickness, the volume occupied by ICs was unchanged. Thus, following chronic gentamicin treatment, the stria is affected but significant progressive and permanent structural effects are confined to the marginal cells.

Specialisations of the lateral membrane of inner hair cells

Freeze-fracture exposes on the lateral membrane of inner hair cell clusters of large particles organised in regular square array ('plaques'). Any one cell possesses a number of plaques. The centre-to-centre spacing of the particles forming the plaques is approximately 14 nm and at high resolution each particle appears to consist of 4 or 5 subunits. The plaque regions were resistant to the membrane deforming activity of filipin and consequently could be identified as localised membrane thickenings in thin sections of filipin-treated tissue; subsequently similar specialisations were identified in thin sections of normal tissue. From the thin section and freeze-fracture studies it is shown that the plaques are not junctional or synaptic specialisations and are not the sites of anchoring of some cytoplasmic component to the membrane. As the individual elements composing each plaque span the membrane and have a subunit structure, it is possible that the plaques represent the sites of transmembrane channels. It is estimated that there may be up to 20,000 such plaque-forming, channel-like elements on each inner hair cell.

Preparation of the mammalian organ of Corti for scanning electron microscopy

The effect of various standard tissue preparation procedures for scanning microscopy on the morphology of the mammalian organ of Corti has been examined. Comparison was made of (i) different fixatives: glutaraldehyde (GA) alone, osmium (Os) alone, GA and Os together and GA followed by Os; (ii) freeze-drying (FD) versus critical point-drying (CPD); (iii) ligand binding of osmium with thiocarbohydrazide (TOTO procedure) versus sputter coating to provide a conductive coat. Results were assessed in terms of known structural parameters in fresh tissue. It is shown: (i) that double fixation, either GA, Os together or GA followed by Os, consistently produced less tissue distortion than either fixative alone, (ii) that FD produced less shrinkage of the tissue than CPD, and (iii) that the TOTO procedure appeared to rigidify the tissue and enhanced preservation of the morphology.

Cholesterol distribution in cells of the stria vascularis of the mammalian cochlea and some effects of ototoxic diuretics

The distribution of cholesterol in cells of the stria vascularis of guinea pigs and gerbils has been investigated at the ultrastructural level by incubation of tissue in filipin, followed by freeze-fracture. Verification of results has been sought by using tomatin. It is shown that in the cell body region of the marginal cells, the apical and lateral membranes reacted intensely with both agents, but the membranes of the basal processes of the marginal cells did not respond significantly to either filipin or tomatin. On basal cell membranes, filipin-cholesterol complexes were present at a high density, even within the strands of the tight-junctional network of these cells and occasionally within the gap-junctional areas also. Complexes were present on intermediate cell membranes at a lower density than on other plasma membranes that showed a positive response. Tissue from animals that had received an ototoxic diuretic, either ethacrynic acid or furosemide, was characterized by the appearance of membrane regions with closely clustered filipin complexes, suggesting some change in cell membrane structure. At an early stage following diuretic administration, such clusters were particularly noticeable on the membranes of intermediate cells. As intercellular spaces enlarged in response to the effects of diuretics, vesicles released into the extracellular spaces appeared to be cholesterol-enriched. The results are discussed in relation to known features of the structure and function of cells in the normal stria vascularis and of the changes that follow from acute diuretic ototoxicity.

Outer hair cell loss and supporting cell expansion following chronic gentamicin treatment

A sequence of changes in the organ of Corti associated with the destruction of outer hair cells (OHCs) and their replacement by supporting cells following chronic gentamicin treatment has been examined using thin-sections and SEM. The progression of change of OHCs was matched by concomitant expansion of adjacent supporting cells. Hair cells ruptured in the lateral membrane. The apical fragment was retained in the reticular lamina and became surrounded basally by the expanded supporting cells. No large breaches at the surface of the organ of Corti were formed. Rather, it appeared that the tight junctions around the hair cell were maintained until junctions were established between newly adjacent supporting cells in the space once occupied by the hair cell body. Only then was the OHC apex disrupted and the debris released into the sub-tectorial space. Some features of the OHC degeneration process were reminiscent of the controlled, cellular self-destruction phenomenon of apoptosis. The results suggest the possibility that the processes of hair cell loss and replacement may be controlled enabling maintenance of permeability barriers during structural reorganisation.

Structural abnormalities in the stria vascularis following chronic gentamicin treatment

Freeze-fracture and thin-sectioning have been used to examine the stria vascularis of albino guinea pigs chronically treated with gentamicin. Immediately following the end of treatment, most marginal cells showed lipid bodies in the cell body region and freeze-fracture revealed alterations to the marginal cell plasma membrane. Intermediate cells also showed peculiarities including a dilation of the rough endoplasmic reticulum. A co-incidence was noted in the location in the cochlea in which effects in the stria and in outer hair cells occurred. At 4 weeks post-treatment, the stria was significantly thinner than normal and appeared less structurally complex. A minority of marginal cells degenerated. Some morphological features associated with degeneration resembled those of apoptosis, a process of controlled, cellular self-destruction. There were also indications of turnover of gap-junctions throughout the post-treatment period examined. The results indicate significant ototoxic effects of gentamicin occur in the stria and that changes to plasma membranes are one of the initial alterations.

Gap junctions in the stria vascularis and effects of ethacrynic acid

Gap junctions in the stria vascularis of guinea pigs were studied using freeze-fracture. Nearly all junctions were associated with basal cells. They were present between basal cells and spiral ligament cells, adjacent basal cells, basal and marginal cells and basal and intermediate cells. Following administration of ethacrynic acid, gap junction morphology altered. There was a statistically significant decrease in the centre-to-centre spacing of gap junction subunits and the subunits became regularly packed. Such changes were distinct before any other gross morphological change in the stria had occurred. These morphological alterations suggest that physiological uncoupling of stria cells may occur in response to the effects of ethacrynic acid.

A tubulo-cisternal endoplasmic reticulum system in the potassium transporting marginal cells of the stria vascularis and effects of the ototoxic diuretic ethacrynic acid

Sections of metal impregnated tissue and freeze-fracture have been used to examine intracellular membrane systems in marginal cells of the stria vascularis in mammalian cochleae. A continuous network of elements of the smooth endoplasmic reticulum was revealed. Notable features of this system were a series of flattened cisternae just inside and parallel with the lateral plasma membrane in continuity with an apical network of tubules, cisternae and sheets oriented in parallel with the luminal membrane. The whole system was closely associated with mitochondria. These characteristics suggest that the potassium transporting marginal cells possess a tubulo-cisternal endoplasmic reticulum (TER) like that found in many sodium transporting epithelial cells. The lateral elements of the TER dilated, appearing like vacuoles, and opened to the lateral extracellular space in response to the effects of ethacrynic acid. This diuretic impairs ion transport in the stria vascularis. It is suggested that the TER in marginal cells is involved in the transport of ions and fluid from the cell to the intercellular space when ion balance is disturbed and may play a role in cell volume regulation.

Ultrastructural and electrophysiological studies of acute ototoxic effects of furosemide

Previous studies of the effects of ethacrynic acid on the inner ear following intraperitoneal injection of the diuretic have shown a progression of reversible changes occurring in the stria vascularis. The time course of these changes approximately parallels alterations in endolymphatic potential (EP). In this report, some preliminary findings concerning the effects of furosemide after intraperitoneal injection of 80 mg/kg are described. EP declined over a longer time course than that recorded with intravenous injection. Cochlear microphonic (CM) and compound action potential (CAP) also declined but to differing degrees. In the stria vascularis a progression of changes was apparent. In general, the changes were similar to those observed following ethacrynic acid intoxication and affected marginal cells, intermediate cells and strial capillaries. The upper basal turn of the cochlea was affected first and the damage spread apically. In the organ of Corti, stereocilia on the outermost row of outer hair cells were disorganized. This was apparent in approximately the same region as initial strial effects and was only observed when strial derangement was quite marked.

Ultrastructure in the stria vascularis of the guinea pig following intraperitoneal injection of ethacrynic acid

Following intraperitoneal injection of ethacrynic acid, progressive, mainly reversible changes occurred in the stria vascularis, affecting all three cell types and the capillary basal laminae. Both marginal and intermediate cells showed abnormalities early, at a time when EP was just beginning to decline. Progressive changes in marginal cells culminated in apical bulging followed by recession of the swelling and stretching of the cells concomitant with gross interstitial oedema. Marginal cell mitochondria showed damage and the transcellular tubule system was dilated. Intermediate cells also showed a progression of changes, culminating in a marked, but reversible, shrinkage. The time of appearance of severe strial derangement correlated with the time of maximal depression of EP. The ability of the stria to regain rapidly an almost normal morphology appeared to be due partly to the distribution and orientation of microtubules in marginal and intermediate cells preventing major disruption of stria vascularis architecture.