ABSTRACT
The development of the otoliths in mice from
days’ gestation to 3 days after birth was studied in alcohol-, formalin-, and Bouin-fixed material.
The first sign of the otoliths in alcohol-fixed
day material was an apparent protein precipitate. Then a PAS-reacting substance appeared among the protein, together with biréfringent particles. Calcium salts were then deposited in their typical crystalline form and the otolith flattened so that at
days it consisted of a gelatinous layer surmounted by a layer of biréfringent calcareous granules with an organic matrix, which contained polysaccharide and phosphatase. Glycogen was also present.
After formalin fixation very little otolith material was seen until
days, and with Bouin very little until 2 days after birth; formalin preserved the calcium salts and granular matrix, Bouin the granular matrix only.
Comparison of the results of the various fixatives showed that the localization seen after alcohol fixation was unreliable, so that the position of the otolithic material when it first appeared, and hence the site of its secretion, were unknown.
The cells of the maculae showed phosphatase activity and probably took part in secreting substances; the secretory regions of the utriculus may also have done so.
The mechanism of concentration and precipitation of the calcium salts remains unknown, but it is suggested that the polysaccharide matrix may take part.
INTRODUCTION
There is still little known about the formation of the mammalian otolith. Work carried out on this problem many years ago (Nishio, 1926) left many questions unanswered. In the meantime considerable knowledge has been gained of the development of other calcified structures, namely, bones and teeth, and the present paper consists of a reinvestigation of the development of the otoliths of the house mouse (Mus musculus L.) in the light of this new knowledge of calcification.
The normal fully formed otoliths of the mouse consist of a flat mass of small prismatic calcareous crystals in an organic matrix. The calcium salts of the otoliths of all vertebrates are present as calcium carbonate in the form of aragonite. Hastings (1935), working with the salamander, Amblystoma tigrinum, found in addition 15·9 per cent, of calcium phosphate in an apatite form. The organic matrix is said to consist of protein. It forms an intimate part of each crystal and fills the spaces between them. According to Tenaglia (1925) it is of a semi-solid consistency and imparts a definite shape to the otolith. Embedded in it are the tips of the sensory hairs of the macula which the otolith overlies. An otolith is present in the sacculus and in the utriculus of each ear.
Questions for consideration in a study of the development of the otoliths include the following: which part of the wall of the labyrinth secretes the matrix, what is the mechanism of deposition of calcium salts on the matrix, and why are they deposited only on the matrix? There might be a local secretion of calcium salts by the labyrinth wall, a local absorption of water by the wall, or the matrix itself might actively take part in calcification.
Nishio (1926) studied the formation of the otoliths in various vertebrate groups and reviewed earlier work. Previous authors had generally reported droplets emanating from the surface of the macula and had considered these to be otolith material. Nishio, however, interpreted such droplets as fixation artifacts, and himself could not discover which parts of the labyrinth wall secreted the otolith material, though he suggested that the ‘secretory regions’ of the utriculus might play some part. These ‘secretory regions’ are certain highly vascular, pigmented regions (Text-fig. 1) of the utriculus, ampullae, and crus commune of the vertical canals which were described by Iwata (1924; and later Hazama, 1929), and to which a secretory function is attributed on histological grounds, although it is not known what substances are secreted. In the mouse Nishio found that calcium salts and matrix appeared suddenly in foetuses of 17–18 mm. length, the matrix forming part of each crystal. In an attempt to demonstrate local enrichment of the endolymph with calcium he fixed foetal mice by injection with a mixture of sodium oxalate and formalin. He found masses of calcium oxalate crystals lying between the otolith and the surface of the macula and interpreted this as evidence of local enrichment with calcium ions. This method of detecting calcium in solution is open to two serious objections, however. First, the solute could be significantly displaced by diffusion before the oxalate was precipitated, so that the localization would be inaccurate. Secondly, Nishio’s belief that no calcium phosphate or carbonate would be dissolved and reprecipitated as oxalate by the oxalate-formalin mixture is probably unjustified. These objections are unfortunately sufficiently serious to vitiate Nishio’s conclusions. Hence, neither the mechanism of secretion of the matrix nor of deposition of calcium salts can be considered established by Nishio’s work.
Distribution of the pigment of the secretory regions in the labyrinth of the mouse. Camera lucida drawing, seen from the medial side. × 25. AV A, anterior vertical ampulla; AVC, anterior vertical canal; CC, crus commune; DR, ductus reuniens (to cochlear duct); ED, endolymphatic duct (to endolymphatic sac); HA, horizontal ampulla; HC, horizontal canal; PVA, posterior vertical ampulla; PVC, posterior vertical canal; S, sacculus; U, utriculus. The cochlear duct is not shown.
Distribution of the pigment of the secretory regions in the labyrinth of the mouse. Camera lucida drawing, seen from the medial side. × 25. AV A, anterior vertical ampulla; AVC, anterior vertical canal; CC, crus commune; DR, ductus reuniens (to cochlear duct); ED, endolymphatic duct (to endolymphatic sac); HA, horizontal ampulla; HC, horizontal canal; PVA, posterior vertical ampulla; PVC, posterior vertical canal; S, sacculus; U, utriculus. The cochlear duct is not shown.
The present work on the development of the normal mouse otolith was under-taken in conjunction with a study of the development of hereditary absence of otoliths in mice (Lyon, 1951,1953). The development of the ears of the mutant mice will be considered in a later paper. Since Nishio’s work it has been learned that in the formation and calcification of bone the presence of phosphatase, of the polysaccharide bone matrix, and of glycogen are all essential, as well as of a suitable concentration of inorganic salts (Dallemagne, 1950). Staining techniques for phosphatase and polysaccharide were therefore used in the present work, as well as birefringence methods intended to detect the earliest appearance of calcium salts.
In addition, the cupulae and the tectorial membrane, the covering membranes of the cristae and organ of Corti respectively, were studied, since these mem-branes bear the same relationship to the sensory areas to which they belong as the otolith does to the macula.
MATERIALS AND METHODS
Timing embryos, dissection, fixation, and microtomy
Mouse embryos of days’ gestation and young mice from birth to 4 days old were studied. They came from a stock segregating for the recessive mutant pallid, which when homozygous produces absence of otoliths. All the animals considered in this paper, however, were heterozygotes and therefore had normal ears. (Homozygous embryos of more than
days’ gestation can be recognized by the absence of pigment from their eyes.)
The stage of gestation of foetuses was determined by daily inspection of mated females for vaginal plugs and was checked by the external features (Grüneberg, 1943). Post-natal stages were timed from birth, inspection for births being made once daily.
To obtain foetuses, pregnant females were killed with ether and the young were removed and separated from their membranes in Ringer’s solution. They were then decapitated and only the heads were fixed. For stages after birth the animals were anaesthetized or killed with ether and decapitated. The ears and surrounding parts of the skull were dissected out and fixed by immersion.
Bouin’s fluid, Susa, 80 per cent, alcohol, and 4 per cent, formaldehyde in 0·9 per cent, sodium chloride neutralized with sodium hydroxide were used as fixatives. Specimens fixed in Bouin or Susa were decalcified by the fixative. The other specimens were not decalcified since at these early stages paraffinembedded material can easily be sectioned without decalcification. All specimens were embedded in paraffin and sectioned transversely at 7 μ; some of the postnatal material was previously infiltrated with 1 per cent, celloidin in methyl benzoate, to give additional support to the tissues.
Staining and histochemical methods
(a) Alkaline phosphatase
Gomori’s (1939) staining method for alkaline phosphatase was used on specimens fixed in ice-cold 80 per cent, alcohol. The sections were incubated with sodium glycerophosphate for 6 hours and the calcium phosphate precipitated by the enzyme was made visible by substitution with cobalt sulphide. Preformed calcium phosphate, present in the bones of the skull, had first to be removed; this was done by taking the sections down to water, decalcifying in Lorch’s (1947) solution for 15 minutes and washing thoroughly.
It was then necessary to ensure that the decalcification had been complete and that subjection to the low pH of the decalcifying solution had not inactivated the enzyme. Slides treated with Lorch’s solution and then incubated in a control medium served to ensure the completeness of decalcification. Confirmation that prior decalcification did not reduce the phosphatase activity was obtained by comparing undecalcified slides incubated in substrate medium with similar slides incubated in control medium and decalcified slides incubated in substrate medium. No appreciable inhibition of phosphatase activity was detected.
(b) Polysaccharides
The staining method used to demonstrate polysaccharides was the McManus (1948) variant of the Hotchkiss-McManus (‘PAS’) method, for the oxidation of 1-2 glycol groups with periodic acid, and staining of the aldehydes formed with Schiff’s reagent. Specimens fixed in formalin, alcohol, Bouin, and Susa were stained by this method. Delafield’s haematoxylin and light green both proved suitable as counterstains.
Control slides were placed in distilled water instead of periodic acid. Examination of these confirmed that the sections contained no substances which could rec.olourize Schiff’s reagent without previous oxidation. In order to determine what part of the response to PAS could be attributed to glycogen, some slides were incubated at 40° C. for 15 minutes with a strong solution of human saliva, and then rinsed, before being placed in the periodic acid. Those PAS-reacting substances which were removed by the salivary amylase were considered to be glycogens.
(c) Calcium salts
Staining the inorganic part of the developing otoliths proved difficult and various methods were tried.
Some stains which gave good staining of the calcium of the developing bone present in the sections failed to stain the inorganic part of the otoliths. Such stains included alizarin red S, gallamine blue (Stock, 1949), and cobalt sulphide. Recourse was therefore made to the less specific and less sensitive stains, Delafield’s haematoxylin and the von Kossa calcium stain. Both these gave a reaction, and the reactions could be prevented by previous decalcification. (In the case of haematoxylin this is in contrast to the findings of Cameron (1930) who stated that the staining of calcified tissues with haematoxylin was not prevented by decalcification. In the present work calcium deposits in bone and otoliths were stained a deep royal blue by haematoxylin, a colour easily distinguished from the light blue which might be seen after decalcification.) Alcohol- and formalinfixed material was used for these tests.
For a more sensitive detection of calcium salts some slides of alcohol-fixed sections were taken from xylol to absolute alcohol, to eosin in 95 per cent, alcohol, then returned via absolute alcohol to xylol and mounted. This process was intended to minimize the loss of salts by solution during staining. The calcareous matter could then be seen, unstained, either by normal light as highly refractile crystals or, with the polarizing microscope, as biréfringent particles.
(d) Ear morphology
Delafield’s haematoxylin and eosin (H & E), Masson’s haematoxylin-ponceau-light green trichrome stain, and Mallory’s connective tissue stain were all used in studying the general morphology of the ear.
RESULTS
The first and main section will describe the development of the otoliths as seen after the various types of fixation. The development and staining reactions of the walls of the sacculus and utriculus will also be described, since changes in the walls may throw light on the origin of the otolithic material. A final section will deal briefly with the appearance of the cupulae and tectorial membrane. Table 1 shows the material available.
The number of specimens, fixed and stained in various ways, which were available at each stage of development

Observations on the developing otoliths will be described by stages beginning with the earliest available, i.e. days’ gestation. At this stage all the main parts of the inner ear had been formed and the sensory epithelia of the otolith organs, ampullae, and cochlea had been differentiated from the surrounding wall, though not of course fully differentiated. The secretory regions of the utriculus and canals were not yet pigmented but already had a rich vascular supply. The lateral wall of the sacculus, on the other hand, had a poor blood supply.
Of the covering membranes of the sensory epithelia the cupulae were already formed, and the tectorial membrane was also present. In alcohol-fixed specimens there was also something representing the developing otolith in the lumen of the sacculus and utriculus. In formalin- and Bouin- or Susa-fixed specimens nothing was yet visible, making it clear that these fixatives did not preserve this part of the developing otolith. In fact, alcohol fixation showed the otolith to consist of at least three parts: a precipitate, probably fluid or gelatinous in life; granules of calcium salts; and an organic matrix forming part of each calcium granule. Neither formalin nor the acid mixtures preserved the gelatinous layer and both of these at various stages of development also failed to preserve other parts of the developing otolith. Thus, alcohol fixation seemed to give the fullest picture of otolith development and therefore in the following account alcohol-fixed specimens will be described in most detail and animals fixed with other solutions will be compared with alcohol-fixed ones.
Development of the otoliths
(a)
days’ gestation
At this stage, in alcohol-fixed specimens, there was present in the lumen of both the sacculus and the utriculus something appearing to be a fixation precipitate of a protein (Text-fig. 2). That is, the bulk of this substance presented no very definite form, either filamentous or granular, but merely looked like a loose network. In the midst of it, however, a few small pieces of membranous material were visible, particularly in the utriculus.
Transverse section through the membranous labyrinth of a -day mouse embryo. Camera lucida drawing. × 85. AV A, anterior vertical ampulla; A VC, anterior vertical canal; C, cupula; CD, cochlear duct; HC, horizontal canal; OC, otic capsule; OP, otolith precipitate; S, sacculus; U, utriculus.
Transverse section through the membranous labyrinth of a -day mouse embryo. Camera lucida drawing. × 85. AV A, anterior vertical ampulla; A VC, anterior vertical canal; C, cupula; CD, cochlear duct; HC, horizontal canal; OC, otic capsule; OP, otolith precipitate; S, sacculus; U, utriculus.
When stained with Delafield’s haematoxylin and eosin the whole of this preotolithic material took up the pink eosin stain. With PAS, however, the protein precipitate was negative and took up only the counterstain, light green, whereas the membranous specks were PAS-positive and stained a pink to mauve colour. In addition, strongly PAS-positive droplets were scattered all over the protein precipitate, and were also found to a small extent free in the lumen of the sacculus (Plate 1, fig. C). These droplets could be removed by salivary amylase and were therefore considered to be glycogen.
The animals of one litter were slightly more advanced than the rest and had more PAS-positive membranous material. In addition, two of them had a few small PAS-positive granules in the utricular precipitate, lying along a PASpositive membrane. These were considered to be the first sign of the developing otolith granules.
Birefringence was tested in eosin-stained slides of one animal from one of the younger litters and two from the older litter. In the younger animal no birefringence was found either in the lumen or walls of the sacculus or utriculus. In both the older ones the utricular precipitate showed birefringence in the form of a cloud of minute pin-points of weak birefringence lying in the precipitate itself. Each pin-point was much smaller than an otolith granule and nothing resembling otolith granules was seen. Slides which had been through the PAS routine, and hence had been treated with acid solutions, showed no birefringence.
The position of the pre-otolithic material in these preparations was very different from that of the fully differentiated otolith. In the sacculus, as Plate 1, fig. C shows, the mass lay near the lateral wall, opposite the macula over which the otolith finally lies, and wisps of material appeared to be in contact with the wall. In the utriculus, where the final position of the otolith is ventral, the preotolithic mass lay dorsolaterally, near the mouths of the ampullae of the anterior vertical and horizontal canal, and some of it was actually inside the ampullae rather than the utriculus. Again there were wisps in contact with the dorsolateral wall of the utriculus, and in fact some material appeared flattened against the wall.
In formalin-fixed specimens none of the developing otolith was preserved, and in three Bouin-fixed animals there were merely small pieces of PAS-positive membranous material in each sacculus and utriculus. In the sacculus they lay near the lateral wall, and in the utriculus near the macula. When stained with Masson’s trichrome and with Mallory’s stain they took up the light green and aniline blue respectively.
In alcohol- and formalin-fixed specimens all the cells of the lateral and dorsal walls of the sacculus, up to the edges of the macula, were rich in glycogen, in the superficial parts of the cells (Plate 1, fig. C). Some of it appeared to be escaping into the lumen of the sacculus, but this may have been a fixation artefact. The connective tissue applied to the lateral wall was also rich in glycogen. The utricular wall, on the other hand, contained only a little glycogen in the dorsolateral region.
Only five specimens showed any pigmentation of the secretory regions. It took the form of a few separate melanoblasts in the connective tissue of the dorsal and ventro-lateral utricular walls, a long way from the final position of the pigment layer.
(b)
days’ gestation
The general appearance of the developing otoliths of alcohol-fixed animals resembled that at days (Plate 1, fig. D). Diffuse masses resembling precipitated protein still floated freely in the lumina of the sacculus and the utriculus. More glycogen was present, however, and also more PAS-positive amylase-resistant material. In some specimens this material was of no definite form, in others it looked membranous, and in some a few small granules were present, lying along a PAS-positive membrane.
Three eosin-stained specimens were examined with the polarizing microscope. In one, birefringence was absent from the inner ear. In the other two it was present in the pre-otolithic material, as at days, in the form of a cloud of pin-points, much smaller than otolith granules. The birefringence was stronger than at
days, however, and the points were larger and much more numerous. In addition, biréfringent material now lined the entire dorsal wall of the utriculus, the dorsolateral region of the saccular wall and the more dorsal part of the macula, and even the walls of the anterior and lateral ampullae and the basal coil of the cochlea.
Six animals were tested for the presence of alkaline phosphatase. Activity was present diffusely in the precipitate of both sacculus and utriculus (Plate 1, fig. A). In addition, the few granules present in the utriculus stained heavily. No attempt was made to assess accurately the strength of this phosphatase activity. In two specimens, however, a series of slides were incubated for varying lengths of time. Activity was detected in the perichondrium of the skull cartilages after 1 hour’s incubation, and in calcifying cartilage, chorioid plexus, and brain tissue after 2 hours. It was not until after 5 hours’ incubation that the pre-otolith showed cobalt sulphide staining. Thus, the activity in this material was not great.
The apparent position of the precipitated pre-otoliths had changed slightly by comparison with day specimens. Laterally, in the sacculus, and dorsally, in the utriculus, the material was still in contact with the walls. The mass was somewhat greater, however, and extended farther, medially or ventrally, towards the final position of the otolith.
In formalin-fixed specimens a little of the developing otolith was now preserved. There were a few filamentous specks (Plate 1, fig. B), which were PAS-positive, stained green with Masson’s, an indeterminate colour with haematoxylin and eosin, and were negative to the von Kossa calcium stain. They were scattered through the lumina of the sacculus and utriculus. Only one specimen was Susa-fixed; it showed a few small PAS-positive specks lying near the maculae.
The lateral wall of the sacculus was still rich in glycogen at days, but in the utricular wall very little remained. Phosphatase activity was also present in the walls of the sacculus and utriculus, but in the maculae and not in the lateral or dorsal walls. The nuclei, cell surfaces, and hairs themselves of the hair cells of the maculae were all stained black by cobalt sulphide after incubation in substrate for 6 hours. Thus, as in the developing otolith, the activity was weak.
Pigment had now spread over the dorsal wall of the utriculus and a small area of the ventrolateral wall near the mouth of the ampulla of the horizontal canal, but the ramifications of the individual melanoblasts could still easily be made out. They were still not closely applied to the epithelial lining of the utriculus, but were scattered in the loose connective tissue.
(c)
days’ gestation
The four alcohol-fixed animals studied were at rather different stages of development but in all a complete change in the appearance of the developing otolith had taken place since the -day stage. There was now a more or less flat layer of protein precipitate surmounted by a layer of refractile calcareous granules of definite otolith shape. Each granule had an organic matrix, staining pink to mauve with PAS, and when the calcium salts were removed by acid treatment this matrix retained the shape of granules, though of a much smaller size. The protein or gelatinous layer had a somewhat more solid appearance than formerly; it was on the whole negative to PAS, but may have been positive at some points. One specimen was stained with haematoxylin and eosin; here the otolith granules stained heavily with haematoxylin, an indication of the presence of calcium salts, and the underlying gelatinous layer stained pink with eosin. In all specimens the calcium salts were highly biréfringent (Plate 2, figs. I and J).
The edges of the developing otoliths still had the appearance that the whole had had at days. There was a diffuse PAS-negative precipitate intermingled with a PAS-positive amylase-resistant substance and with droplets of glycogen. (Glycogen was now absent from the regions where the calcareous granules had formed.) There was also a cloud of pin-point birefringence, in contrast to the granular form of the biréfringent material in the more fully developed region.
The differences among the four alcohol-fixed specimens concerned chiefly the amount of calcareous material deposited and the general shape of the developing otolith. As the amount of calcium salts increased, from the apparently youngest to the apparently oldest specimen, the otolith changed in form from its original globular shape to the final flat lamellar form. In all four specimens the edges remained diffuse.
The apparent position of the developing otoliths had also changed greatly since days. The main bulk of each was now lying over and parallel to the macula to which it belonged, although neither otolith was yet as close to the macula as when fully developed. Although the main bulk of the otolith had moved, however, there were still wisps of material, including biréfringent pinpoints, near the walls, especially in the utriculus. There was much similar material, also biréfringent, in the anterior or lateral ampullae, giving the appearance of having been displaced from the utriculus. In all four specimens, however, there were a few wisps and pin-points of birefringence which did not look displaced, lying in contact with the pigmented regions of the utricular walls, both dorsally and ventrally (Plate 1, fig. E; Plate 2, fig. I).
The formalin-fixed developing otolith also changed greatly in appearance between and
days. At
days there was a continuous layer of otolith matrix preserved in both the sacculus and the utriculus (Plate 1, fig. F). In contrast to the alcohol-fixed specimens, however, no calcium was detected either with haematoxylin or with the von Kossa stain. Another difference was that the organic material present was not separable into two layers, and although the layer present presumably corresponded to the granular matrix of the alcoholfixed specimens it did not show the shape of the individual granules. The reaction to PAS was much stronger than in alcohol-fixed specimens; Masson’s and Mallory’s stains gave green and blue colours respectively. After formalin fixation the otoliths always lay near and parallel to the maculae.
Bouin still failed to preserve any part of the developing otolith.
In the walls of the sacculus and utriculus histological differentiation was leading to flattening of the cells of the lateral saccular and dorsal utricular walls, so that the glycogen of the lateral saccular wall was now less obviously superficial. Further development had occurred in the pigmentation of the utricular wall, and there was now a continuous layer of pigment, still not very closely applied to the wall. A new feature in the sacculus of the alcohol-fixed specimens was a PAS-positive membrane lining the lateral wall. Glycogen droplets were seen along it.
(d)
days’ gestation
No radical change in the appearance of the alcoholfixed otoliths took place between and
days’ gestation. More calcium salts were deposited and more organic matrix, so that after decalcification the matrix now retained the shape of full-size granules. The details of the shape of the fully developed otolith were beginning to be apparent; thus the thin central region of the fully developed utricular otolith could be detected in
-day foetuses (Plate 2, fig. G). At the edges of the otolith there was still the diffuse material seen at
days, and wisps and biréfringent specks were still seen in contact with the pigmented regions of the utricular wall.
The staining reactions to PAS of the gelatinous and granular components of the otolith matrix remained unchanged since days: the gelatinous layer was doubtfully negative and the granular matrix positive. Glycogen was no longer present in the matrix, though there was a little in the lumen of the sacculus near the lateral wall. After staining with Delafield’s haematoxylin and eosin the gelatinous layer took up the eosin, and the calcium salts gave a dark blue stain. If the salts were removed by acid treatment, however, the granular matrix did not stain with either haematoxylin or eosin; it remained completely colourless and very difficult to see. When stained for alkaline phosphatase activity the granular matrix appeared uniformly greyish. The gelatinous layer, on the other hand, showed variable staining; it was in some places black, in others greyish, and in others again negative, having taken up only the eosin counterstain. In control slides the granular matrix remained completely unstained by the eosin counterstain and almost invisible. In undecalcified control slides the calcium salts also were unstained by the cobalt sulphide and were seen simply as refractile crystals.
The otolith granules were strongly biréfringent in all specimens, and birefringence was absent from acid-treated slides. In one specimen, stained with alcoholic eosin without being taken down to water, pin-point birefringence was also still present in the material near the utricular wall. In two other specimens, the slides of which had been passed through water, such birefringence was missing, but it is not certain that biréfringent material had not been dissolved by the watery solutions.
The position of both otoliths was now more or less the final one. Both were lying parallel and close to the maculae, although still not clearly in contact with the tips of the sensory hairs.
After formalin fixation the utricular otolith now gave a reaction for calcium with both haematoxylin (Plate 2, fig. H) and von Kossa, and when the calcium salts were removed by acid the matrix showed the granular form. In the sacculus the calcium salts were still not preserved and the matrix remained shapeless. However, when stained with haematoxylin and eosin it took on a light blue colour, in contrast to the complete failure to stain of the alcohol-fixed granular matrix.
Histological differentiation was continuing in the walls of the labyrinth. The cells of the lateral saccular wall had become flatter, but glycogen was still present. The cells of the dorsal and ventro-lateral regions of the utriculus had also become flatter, and a continuous layer of pigment was now closely applied to the wall. As in the day specimens the surface of the cells of the maculae of the sacculus and utriculus showed phosphatase activity. This activity was weak in both, and weaker in the sacculus than in the utriculus.
(e)
days’ gestation
Only three formalin-fixed specimens were studied. These differed from -day formalin animals in that calcium salts were now preserved in the sacculus as well as the utriculus.
(f) Newborn
In the three alcohol-fixed ears the chief change that had taken place in the otolith since the -day stage was growth; both the maculae and the overlying otoliths had extended in area. The thickness of the layer of calcium salts had increased, whereas the gelatinous layer had become thinner, so that it was now no more than a membrane. The edges of the otolith were no longer diffuse.
The staining reactions of the parts of the otolith to haematoxylin and eosin were unchanged. When stained for phosphatase activity the granular matrix appeared greyish, and the gelatinous layer in some places slightly greyish, but otherwise negative. In the walls of the cavities no phosphatase activity could now be detected.
(g) 1 and 2 days
In 1-day-old Bouinand Susa-fixed animals the first sign of preservation of otolith material was noticed, and at 2 days a definite layer of granular matrix was preserved. The calcium salts and gelatinous layer were still not preserved.
(h) 3 days
The two ears of one animal only were fixed in alcohol; one was stained with PAS and the other for phosphatase. No phosphatase activity was now present in either the gelatinous or the granular part of the matrix. The gelatinous layer took up the eosin counterstain and the granular matrix remained unstained. With PAS the granular matrix stained rose pink, while the gelatinous layer was doubtful. In the walls no phosphatase activity remained but in the lateral saccular wall some glycogen was still present.
One formalin-fixed ear was stained with the von Kossa calcium stain. Both otoliths gave a strong reaction, but otherwise there was no new feature.
The staining reactions of the granular matrix now preserved by Susa fixation were similar to those seen after formalin; with haematoxylin and eosin a blue colour, and with Masson’s trichrome green. A strong positive reaction was obtained with the PAS stain.
Development of the cupulae and tectorial membrane
For the cupulae and the tectorial membrane, fixation with formalin or Bouin’s fluid seemed to give a better picture of the state of affairs than alcohol fixation.
After all three fixatives the cupulae were already visible at days’ gestation. After formalin or Bouin fixation they were found in the final normal position closely overlying the sensory epithelium of the cristae. After alcohol fixation they were displaced, towards the dorsal wall in the anterior and posterior ampullae, and either laterally or medially in the lateral ampulla. The cupulae stained green with Masson’s stain, were PAS-positive, and stained black with cobalt sulphide after some hours’ incubation for phosphatase. The cells of the cristae contained glycogen but no phosphatase. Other diffuse precipitated material was also present in the ampullae after alcohol fixation but appeared irrelevant, some probably being displaced from the utriculus, and some probably artefact. No significant change in the cupulae was noted after
days.
The tectorial membrane was also visible at days after all three types of fixation. After Bouin fixation it appeared as a thin membrane lying in its final position, closely applied to the epithelium in the region of the future organ of Corti. After formalin fixation the appearance was similar, but in addition the cochlea was filled with a free precipitate with the same staining reactions as the tectorial membrane itself. After fixation with alcohol, on the other hand, the tectorial membrane was always moved from its normal position and displaced towards the vestibular membrane. In formalin- and Bouin-fixed material the tectorial membrane appeared to increase in thickness from
to
days’ gestation, and by
days the free precipitated material had disappeared from the cochlea of formalin preparations. Like the cupulae the tectorial membrane was PAS-positive and stained green with Masson’s stain, but, unlike the cupulae, it gave no reaction for phosphatase.
Glycogen was present in the cells of the future vestibular membrane from to
days’ gestation, but there was none in the future stria vascularis. No phosphatase reaction was given by any part of the cochlear duct.
DISCUSSION
In attempting to formulate some hypothesis about the formation of the otoliths one must first consider the nature of the various substances, then their probable sites of origin, and then one may attempt to deduce chemical mechanisms involved in the deposition of these calcareous masses in the lumen of the sacculus and utriculus.
In this study of the mouse otolith the alcohol-fixed material has shown that the fully developed otolith consists of at least three parts: (a) the gelatinous layer, (b) the granular matrix, and (c) the inorganic salts.
The gelatinous layer had the appearance of a protein precipitate, and was not preserved by formalin, which is not a protein precipitant. It was therefore believed to consist of a globular protein and probably to be at least semifluid in life. No other definite properties, such as the presence of polysaccharides or phosphatase, were detected.
The granular matrix gave a positive reaction to the PAS stain and may therefore be presumed to contain a polysaccharide. Since it showed phosphatase activity protein was presumably present also. After formalin fixation this component of the otoliths was still present, but its staining reactions were rather different. This could mean either that the formalin had reacted with the granular matrix in such a way as to change its staining properties, or that some component of it had been preserved by one fixative and not by the other. It does not seem possible to decide between these alternatives, but it is well known that formalin can change the staining properties of proteins, so as to make them stain more strongly with basic dyes (Baker, 1945). Similarly, fixation with Bouin or Susa showed that the granular matrix became resistant to acid fixation at about 2 days after birth, and again one cannot say whether this was due to the secretion of a new component of the matrix, or to a change in the properties of substances secreted previously. However, the effect seen in pallid mice (described by Lyon, 1955), in which the otoliths fail to calcify, suggests that at least three layers of matrix are formed: at -
days,
days, and at 1-3 days after birth.
Recent work by Belanger (1953) has shown by means of an autoradiographic study of the uptake of S35, and by metachromasia after toluidine blue staining, that the otolithic membrane may be assumed to include a sulphur-containing mucopolysaccharide. It is reasonable to suppose that this substance was the polysaccharide seen in the granular matrix in the present work.
The calcium salts were detected by their birefringence. This was first seen as minute pin-points, and only later as granules of a definite shape. It seems likely that this pin-point material was not yet solid in life, but had been precipitated by the fixative, so that its localization may have shown the points of precipitation rather than the location of calcium salts in life. The appearance of birefringence coincided with the appearance of the mucopolysaccharide and increased in amount in parallel with it.
Very little evidence was obtained about the parts of the labyrinth wall which secreted the various parts of the otolith. In the alcohol-fixed material at the earliest stages the otolith material was lying near the lateral wall of the sacculus and the dorsal wall of the utriculus, in both cases the wall opposite the macula. However, it was clear from observation of the tectorial membrane and cupulae that alcohol fixation caused displacement of the epithelial coverings of the labyrinth, and in fact in some specimens otolith material had obviously been displaced by the alcohol from the utriculus into the ampullae. Hence, the localization seen after alcohol fixation must be considered unreliable. After formalin or Bouin fixation very little otolith material was seen until a stage in which even after alcohol fixation the newly formed otolith was lying near the macula; so the position of first appearance of the developing otolith remains unknown. Its apparent migration across the lumina with age in alcohol-fixed material could be due to a change with time in its resistance to displacement. However, in the - and
-day alcohol specimens, some of the wisps of otolithic material and pin-points of birefringence touching the wall of the secretory regions of the utriculus presented a convincing appearance, and could be believed to have been there in life. Thus it is possible that some at least of the material taking part in the formation of the otolith was formed by the secretory regions of the utriculus wall. It seems more probable, however, that most of it was secreted by the maculae; the presence of phosphatase in the surface of these cells suggests some secretory function. In the sacculus the accumulation of glycogen in the lateral wall might indicate some histogenetic activity, but could also merely be associated with the poor blood supply of this region.
As it is not certain which cells secrete the various parts of the otolith, little can be said about the mechanism of its deposition. The widespread birefringence at days suggested that at this stage the endolymph had been rich in ionic calcium, which was precipitated by the fixative to form biréfringent particles. What has not been brought out by the present work is the mechanism by which this enrichment occurred. It is not possible to say whether the calcium was actively secreted in an ionic form, or was split off in the lumen of the inner ear from some previously unionized compound, or whether, on the other hand, the ionic calcium concentration was raised by local absorption of water from the endolymph. The substances present, mucopolysaccharide, glycogen, phosphatase, and salts, are similar to those concerned in the calcification of bone and cartilage. In bone it is suggested that the glycogen is used to form hexose phosphates, from which the phosphate is liberated by alkaline phosphatase and fixed to the protein matrix, while calcium ions may be fixed to the mucopolysaccharide. There is then a chemical change in the organic matrix and the calcium and phosphate ions are liberated and precipitated as bone salt (Pritchard, 1952). In otolith formation the appearance and development of the mucopolysaccharide ran parallel to that of the calcium salts and the picture suggested that the mucopolysaccharide was important in calcification. On the other hand, the phosphatase activity was weak and it is doubtful whether it had any role in calcification, particularly since the salts of the otolith are chiefly calcium carbonate. Phosphatase has also been found during the laying down of calcium carbonate by invertebrates (Wagge, 1951) but here again its role was not clear. In the present work phosphatase was also present in the cupulae, which do not calcify, making it seem more probable that its function in cupulae and otoliths was to do with fibrillar protein secretion (Bradfield, 1949).
ACKNOWLEDGEMENTS
The author is grateful to Miss E. Mavor for histological assistance, to Mr. E. D. Roberts for drawing Text-figure 2, and to Mr. D. Pinkney and Mr. A. Graham for photography.
REFERENCES
EXPLANATION OF PLATES
Plate 1
Fig. A. Inner ear of -day mouse embryo. Alcohol-fixed. Stained for phosphatase. × 55. The cupula, the otolith precipitate in sacculus and utriculus, and the surface of the cells of the maculae have reacted for phosphatase.
Fig. B. Formalin-fixed -day embryo. PAS, light green, × 55. Only small specks of otolith material are preserved in sacculus and utriculus.
Fig. C. Alcohol-fixed -day embryo. PAS, light green, × 230. Glycogen (dark) is visible in connective tissue, cartilage, lateral wall of the sacculus, and the otolith precipitate (specks only). The bulk of the otolith precipitate is negative to PAS (light).
Fig. D. Utriculus of alcohol-fixed -day embryo. PAS, light green, amylase-treated. × 230. Glycogen having been removed, the PAS-reacting filaments in the otolith precipitate scarcely stand out from the background.
Fig. E. Utriculus of alcohol-fixed -day embryo. PAS, amylase-treated. × 230. Granular matrix is present but its PAS staining is not sufficiently dark to stand out. Note wisps of matrix in contact with the pigmented regions of the wall.
Fig. F. Utriculus of formalin-fixed -day embryo. PAS, haematoxylin. × 230. The PAS staining of the otolith granular matrix is much darker, but the gelatinous layer and wall wisps are missing.
Fig. A. Inner ear of -day mouse embryo. Alcohol-fixed. Stained for phosphatase. × 55. The cupula, the otolith precipitate in sacculus and utriculus, and the surface of the cells of the maculae have reacted for phosphatase.
Fig. B. Formalin-fixed -day embryo. PAS, light green, × 55. Only small specks of otolith material are preserved in sacculus and utriculus.
Fig. C. Alcohol-fixed -day embryo. PAS, light green, × 230. Glycogen (dark) is visible in connective tissue, cartilage, lateral wall of the sacculus, and the otolith precipitate (specks only). The bulk of the otolith precipitate is negative to PAS (light).
Fig. D. Utriculus of alcohol-fixed -day embryo. PAS, light green, amylase-treated. × 230. Glycogen having been removed, the PAS-reacting filaments in the otolith precipitate scarcely stand out from the background.
Fig. E. Utriculus of alcohol-fixed -day embryo. PAS, amylase-treated. × 230. Granular matrix is present but its PAS staining is not sufficiently dark to stand out. Note wisps of matrix in contact with the pigmented regions of the wall.
Fig. F. Utriculus of formalin-fixed -day embryo. PAS, haematoxylin. × 230. The PAS staining of the otolith granular matrix is much darker, but the gelatinous layer and wall wisps are missing.
Plate 2
Fig. G. Utriculus of alcohol-fixed -day embryo. Phosphatase. × 250. The otolith granules and cells of the macula have reacted for phosphatase.
Fig. H. Utriculus of formalin-fixed -day embryo. Haematoxylin and eosin. × 250. No gelatinous layer of matrix is visible; the calcium salts have reacted strongly with haematoxylin.
Fig. I. Utriculus of alcohol-fixed -day embryo. Seen by polarized light. × 210. The calcium granules are strongly biréfringent; in addition there are pin-points of birefringence among the wisps of matrix near the pigmented regions of the wall.
Fig. J. Sacculus of alcohol-fixed -day embryo. Seen by polarized light. × 210. Birefringence is visible in the calcium granules and (a few specks only) in the diffuse matrix at the edge of the otolith.
Fig. G. Utriculus of alcohol-fixed -day embryo. Phosphatase. × 250. The otolith granules and cells of the macula have reacted for phosphatase.
Fig. H. Utriculus of formalin-fixed -day embryo. Haematoxylin and eosin. × 250. No gelatinous layer of matrix is visible; the calcium salts have reacted strongly with haematoxylin.
Fig. I. Utriculus of alcohol-fixed -day embryo. Seen by polarized light. × 210. The calcium granules are strongly biréfringent; in addition there are pin-points of birefringence among the wisps of matrix near the pigmented regions of the wall.
Fig. J. Sacculus of alcohol-fixed -day embryo. Seen by polarized light. × 210. Birefringence is visible in the calcium granules and (a few specks only) in the diffuse matrix at the edge of the otolith.