Anatomical and physiological development of the human inner ear.

Hearing Research xxx (2016) 1e13

Contents lists available at ScienceDirect

Hearing Research journal homepage: www.elsevier.com/locate/heares

Review

Anatomical and physiological development of the human inner ear Rebecca Lim*, Alan M. Brichta School of Biomedical Sciences and Pharmacy, Faculty of Health and Medicine, The University of Newcastle, NSW, Australia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 October 2015 Received in revised form 20 January 2016 Accepted 12 February 2016 Available online xxx

We describe the development of the human inner ear with the invagination of the otic vesicle at 4 weeks gestation (WG), the growth of the semicircular canals from 5 WG, and the elongation and coiling of the cochlea at 10 WG. As the membranous labyrinth takes shape, there is a concomitant development of the sensory neuroepithelia and their associated structures within. This review details the growth and differentiation of the vestibular and auditory neuroepithelia, including synaptogenesis, the expression of stereocilia and kinocilia, and innervation of hair cells by afferent and efferent nerve fibres. Along with development of essential sensory structures we outline the formation of crucial accessory structures of the vestibular system e the cupula and otolithic membrane and otoconia as well as the three cochlea compartments and the tectorial membrane. Recent molecular studies have elaborated on classical anatomical studies to characterize the development of prosensory and sensory regions of the fetal human cochlea using the transcription factors, PAX2, MAF-B, SOX2, and SOX9. Further advances are being made with recent physiological studies that are beginning to describe when hair cells become functionally active during human gestation. © 2016 Published by Elsevier B.V.

Contents 1. 2. 3.

4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Morphological development of the inner ear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Development of vestibular organs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Development of the vestibular sensory neuroepithelia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Differentiation of vestibular hair cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1. Development of vestibular hair cell stereocilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2. Vestibular hair cell synaptogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3. Development of vestibular ganglion neurons (VGNs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.4. Development of accessory structures e cupulae and otoconial membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Development of the cochlea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Development of perilymphatic spaces, the scala vestibuli and scala tympani . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Development of scala media, Reissner's membrane and the stria vascularis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1. Development of the organ of Corti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2. Differentiation of auditory hair cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3. Development of auditory hair cell stereocilia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4. Auditory hair cell synaptogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.5. Spiral ganglion neurons (SGNs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

Abbreviations: WG, weeks gestation; VGNs, vestibular ganglion neurons; IHCs, inner hair cells; OHCs, outer hair cells; SGNs, spiral ganglion neurons; GER, greater epithelial ridge; LER, lesser epithelial ridge * Corresponding author. MSB 309 School of Biomedical Sciences and Pharmacy, The University of Newcastle, Callaghan, NSW 2308, Australia. E-mail address: [email protected] (R. Lim). http://dx.doi.org/10.1016/j.heares.2016.02.004 0378-5955/© 2016 Published by Elsevier B.V.

Please cite this article in press as: Lim, R., Brichta, A.M., Anatomical and physiological development of the human inner ear, Hearing Research (2016), http://dx.doi.org/10.1016/j.heares.2016.02.004

2

R. Lim, A.M. Brichta / Hearing Research xxx (2016) 1e13

1. Introduction The inner ear is a complex structure that comprises sensory organs that detect angular and linear accelerations (vestibular system) and sound (auditory system). To detect these sensations both organs use hair cell mechanoreceptors to convert or transduce mechanical movement (head motion or sound waves) into electrical signals. However, there are significant differences in the morphology, location, physiology, and innervation of hair cells of the vestibular and auditory systems. Much of our understanding of the development of these two inner ear systems has been derived from rodent models. Here, we review the more limited information on the morphological development of the inner ear, from rudimentary otic placode to fully formed semicircular canals, vestibule, and cochlea of the membranous labyrinth, in the human embryo and fetus, as well as recent advances describing physiological maturation.

swelling or ampulla containing the sensory organ or crista ampullares, associated with each semicircular canal, is evident. At this time a cleft also forms in the vestibule, partitioning the utricular macula from the saccular macula (Streeter, 1906). The cochlea elongates as a tubular structure from the ventral pouch and with continued growth, begins to rotate between 8 and 9 WG (Kim et al., 2011; Yasuda et al., 2007). By 10 WG, the cochlea has formed a full 2.5 turn coil (see Fig. 1 Streeter, 1906). Between 9 and 18 WG there is a three-fold increase in labyrinth length, after which there are no further length changes (Jeffery and Spoor, 2004). Between 17 and 19 WG, other indices of labyrinth size, such as canal and cochlea radius, have ceased to change and are comparable to adult form and size. In the following two weeks, bony ossification of the surrounding cartilage encapsulates the entire membranous labyrinth, forming the bony labyrinth (Jeffery and Spoor, 2004). 3. Development of vestibular organs

2. Morphological development of the inner ear Classic anatomical studies have described human inner ear development (Bast and Anson, 1949; Streeter, 1906). In the embryo the otic vesicle forms as an invagination of ectodermal cells at the level of rhombomere 5 by four weeks gestation (4 WG) (Bruska et al., 2009). At this stage of development, the otic vesicle has two pouches, one dorsal the other ventral. A projection extends from the dorsal pouch that will form the primordial endolymphatic duct (Bruska et al., 2009; Streeter, 1906). Between 4 and 5 WG, the dorsal pouch enlarges into a triangular shaped mass that will form the basis of all three semicircular canals (Streeter, 1906). As this triangular-shaped region develops there is concomitant resorption of the medial walls to form, in chronological order, the anterior, posterior, and horizontal semicircular canals (Streeter, 1906). At the same stage of development the vestibule, which will contain the future utricular and saccular maculae, also enlarges. An initial constriction between the immature saccular macula and the developing cochlea will eventually form the ductus reuniens (Streeter, 1906). Over the following two weeks of growth, there is continued remodeling of each semicircular canal and by ~7 WG, a

The general description above provides an outline of membranous labyrinth maturation from ~4 to 20 WG of human embryonic and fetal development. Since the vestibular system is ontogenetically and phylogenetically older than the auditory system, we will begin with the vestibular system. There are, however, only a limited number of studies describing growth and maturation of the vestibular apparatus in human tissue. 3.1. Development of the vestibular sensory neuroepithelia There are five vestibular sensory organs in each ear e three cristae, a utricular macula, and a saccular macula. The cristae develop from the ampullary walls where branches of the vestibular nerve enter (Bast and Anson, 1949). The ampullae enlarge as the canals form and expand, and by ~6 WG the cristae have differentiated as well-defined structures (Yokoh, 1971). At this stage, the utricular and saccular maculae are located in the vestibule, but are not as developmentally advanced. Individual vestibular organs (all three cristae and the utricle) are distinct and separate structures by 8 WG (Dechesne and Sans,

Fig. 1. Schematic representation of the development of the human inner ear from 4WG to 10WG. Initially, the inner ear develops from an otocyst at 4WG, with a dorsal pouch that will become the vestibular organs, and a protrusion that will form the endolymphatic sac. A ventral pouch becomes the cochlea. By 5WG, the semicircular canals are beginning to form and the medial walls undergo resorption. A constriction, the ductus reuniens develops between the saccule and cochlea. By 10WG, a partition forms between the utricle and saccule. The semicircular canals are complete, although they have not reached mature size. The ductus reuniens separates the vestibular organs from the cochlea that has reached a full-2.5 turns by 10WG. Length of inner ear: 9 mm at 4WG, 12 mm at 5WG, and 30 mm at 10WG Modified from Streeter (1906).



3

Fig. 2. Isolated, partial inner ear preparations of left side peripheral vestibular organs from human fetus (13 WG) and mature mouse (3 weeks postnatal). The human (A) and mouse (B) vestibular “triad,” consisting of a horizontal and anterior ampullae and their associated cristae, joined with the utricle and remnants of the VIIIth cranial nerve. Membranous tissue and overlying accessory structures (cupulae and otoconial membranes) have been removed for direct visualization of epithelial surfaces. Modified from Lim et al. (2014) according to permissions by Creative Commons (http://creativecommons.org/licenses/by/2.0).

1985). At this stage, the cristae are thick, short, and bulbous. Between 8 and 9 WG, however, there was a rapid lengthening and thinning of the cristae and they become the more familiar elongated crests. This increase in length is exemplified in the anterior crista where at 8 WG the length is ~150 mm, and a week later has elongated to ~300 mm (Dechesne and Sans, 1985). This rapid growth then slows between 9 and 12 WG, with a secondary rapid growth phase occurring between 12 and 14 WG when, the anterior cristae reaches ~55% of its adult size (Dechesne and Sans, 1985). At 7 WG the utricle is relatively thick, being more than 80% of its adult thickness. However, as it elongates, over the next week, the utricle decreases to 45% of mature thickness. Between 8 and 11 WG, there is a pause in development and the utricle remains the same thickness throughout this period. At this stage, otoconia are first seen associated with the utricular surface (Dechesne and Sans, 1985) discussed below. After this quiescent period, from 11 WG onwards, there is significant thinning of the utricle as it grows further in size. The decrease in thickness of the utricle coincides with a reduction in density of the supporting cell layer, from 2 to 3 layers to a monolayer by 13 WG. At this stage, the utricle is almost the thickness of an adult and the orientation in the vestibule is similar to that observed in the mature inner ear (Dechesne and Sans, 1985). There are few developmental studies beyond 13 WG that characterize the maturation of each vestibular apparatus (see Fig. 2 for size and orientation at 13WG relative to mouse aged 3 weeks postnatal). It is not known when the organs reach adult size, but it has been proposed this occurs when the bony labyrinth reaches maturity between 17 and 19 WG. Further descriptions of the vestibular sensory apparatus have been taken from reports describing adult tissue. Cristae of all three canals are similarly saddle-shaped and on average have a total surface area of 0.9 mm2 (Watanuki and Schuknecht, 1976). The utricular macula is oval in shape with the anterior pole projecting upward, while the saccular

macula has a dorsal extension projecting from the main body (Watanuki and Schuknecht, 1976). The average surface area of the utricular and saccular maculae is 3.6 mm2 and 2.2 mm2 respectively (Watanuki and Schuknecht, 1976). Various techniques have been used to estimate hair cell numbers in mature vestibular organs. Using surface scans, three studies indicate similar average number of hair cells in the cristae and range from 6700 to 8300 (Lopez et al., 2005; Rosenhall, 1972a; Watanuki and Schuknecht, 1976), and there were no differences between the horizontal, posterior, or anterior canals. However, one study showed there was a decrease in the number of type I and type II hair cells with advancing age (>80 years) (Lopez et al., 2005). Compared to the cristae, human utricular maculae have four times the number of hair cells. Total counts from adult utricular maculae ranged from 29,500 to 39,200 with a subset (between 2000 and 2900) forming the specialized central striola region (Rosenhall, 1972b). Similar estimates of hair cell numbers were reported by Watanuki (Watanuki and Schuknecht, 1976). In an accompanying study, Rosenhall estimated utricular hair cell counts from two fetal samples (14e23 WG) with an average of 32,900 in the utricle (Rosenhall,1972b). In general, saccular hair cell counts in the mature organs were almost half those estimated in the utricule, ranging from 16,000 to 21,300, with approximately 1500e1600 hair cells in the developing striola (Rosenhall, 1972b). Interestingly, the saccular maculae of the fetus had a higher average number of hair cells, 19,100 compared to the adult (Rosenhall, 1972b), suggesting the number of hair cells in the saccule decreases during development but increases in the utricle. 3.2. Differentiation of vestibular hair cells Differentiation of progenitor cells into sensory hair cells requires the expression and specific timing of transcription factors throughout development. A number of reviews have described in detail transcription factor expression in animal models (Raft and


4


Groves, 2015; Whitfield, 2015). However, few papers to date have described the expression of transcription factors in human embryonic or fetal tissue. Morphologically, differentiation may be described with respect to 1) chronotopic arrangement e the time that hair cells are present at a specific location in the sensory neuroepithelium; 2) cytodifferentiation e cell morphology, including the presence of stereocilia and kinocilium; 3) synaptic specializations e the pre- and post-synaptic densities become evident; and 4) innervation e establishing appropriate nerve fiber connections. Hair cell differentiation may also be signaled by the expression of appropriate transcription factors, such as PAX2, SOX2, and ATOH-1. There are two types of vestibular hair cells; type I and type II. In mature neuroepithelium, these hair cell types have distinct morphological differences. For example, type I hair cells are amphora-shaped with variability in the length of the constricted neck; and type II hair cells are more cylindrical in shape. In addition to their distinct morphologies, there is a significant difference in the way afferent nerve terminals contact each hair cell type. Type I hair cells are contacted by an enveloping cup-like or calyx nerve terminal, while type II hair cells are contacted by conventional button-like or bouton nerve terminals. There is a discrete organization of mature cochlea hair cells with a single row of inner hair cells (IHCs) and typically three rows of outer hair cells (OHCs). This organization is arranged in a tonotopic gradient that also reflects hair cell differentiation and development. However, in vestibular organs there is no similar chronotopic organization of hair cells or any evidence of a functional gradient, although there appear to be specialized central zones. Type I and type II hair cells are distributed throughout the neuroepithelia of the vestibular organs, with a higher proportion of type I hair cells localized to the central zones. Nevertheless, without a clear cellular segregation of vestibular hair cell types, differentiation into type I and type II hair cells based on specific location is not possible. Thus, alternative criteria have to be used, such as the presence of stereocilia and synaptic specializations as indicators of cytodifferentiation of vestibular hair cells. 3.2.1. Development of vestibular hair cell stereocilia In scanning electron microscope (SEM) and transmission electron microscope (TEM) studies, the earliest embryonic human vestibular neuroepithelium (7 WG) is described as undifferentiated, polystratified epithelial cells (Dechesne and Sans, 1985; Sans and Dechesne, 1985). However, it was shown even at this early stage, these undifferentiated cells have a single mini-kinocilium surrounded by many microvilli (Dechesne and Sans, 1985; Sans and Dechesne, 1985). These nascent hair bundles mark the initial stages of cytodifferentiation of hair cells and remain on the surface of most epithelial cells throughout this early stage of development. Some cells however undergo further differentiation, indicated by the clear presence of stereocilia at the apical pole and occurs at the same time as afferent nerve contacts at the basolateral surface of these cells (Sans and Dechesne, 1985). It should be noted that in mature vestibular hair cells numerous stereocilia emanate from the apical pole, in a staircase arrangement, from the shortest at one end to the longest stereocilia at the other. The longest stereocillia are adjacent to a single kinocilium located at the periphery of the hair bundle. The kinocilium is a specialized cilia composed of nine peripheral doublet microtubules that encircle two central single tubules. In vestibular hair cells, the position of the kinocilium coincides with directional sensitivity of the hair cell. At 8WG, stereocilia are apparent and cover the entire apical surface of each sensory hair cell. Hair bundles are already directionally polarized with the longer putative kinocilia located at the edge of the bundle (Dechesne and Sans, 1985).

In the subsequent two to three weeks there is a lengthening of stereocilia, particularly on the upper slopes of the cristae, while the stereocilia of hair cells located at the ampullary wall and base are similar to shorter, nascent hair bundles (Dechesne and Sans, 1985). Between 12 and 14 WG, hair bundle length increases dramatically but remains shorter than full adult size (Dechesne and Sans, 1985). Despite the lengthening of hair bundles, there are still many hair cells with short hair bundles. However, even these immature hair bundles display graded heights and were directionally polarized (Hoshino, 1982). The individual stereocilia of short hair bundles are thinner than those of long hair bundles but were distinctly longer than the microvilli expressed on the surface of supporting cells (Hoshino, 1982). In human saccular maculae, the microvilli on the surface of supporting cells are quite long and dense during fetal development but diminish in size and number in adults (Rosenhall and Engstrom, 1974). In the cristae and utricle the number of stereocilia expressed by hair cells aged 14 WG is between 80 and 90 stereocilia per hair cell (Hoshino, 1982) and in the posterior cristae at 16 WG there were between 76 and 102 stereocilia per hair cell (Rosenhall and Engstrom, 1974). The stereocilia of all vestibular organs were also seen to possess club-like endings during fetal development (Rosenhall, 1972a). In adult human vestibular neuroepithelium there are differences in the number of stereocilia expressed by type I and type II vestibular hair cells of the utricular maculae. Type I hair cells have more and thicker stereocilia than type II hair cells (n ¼ 70 versus 50; mean diameter ¼ 488 nm versus 373 nm, respectively) (Morita et al., 1997). Therefore, it is not clear whether the differences in stereocilia number between fetal and adult tissue are due to differences between hair cell types of the human cristae and utricular maculae, or if reduction in stereocilia number is a feature of development. An important characteristic of the peripheral vestibular system is directional selectivity as denoted by hair bundle orientation. In mature utricular and saccular maculae orientation of hair bundles fan out across the surface of the macular organs so that all hair bundles are oriented towards the striola. The striola is a centrally positioned, crescent-shaped band in the utricle and saccule. Associated with the striola region of the macular organs there is an obvious boundary line where the hair bundles abruptly reverse their direction. This is called the line of polarity reversal (LPR; Li et al., 2008). Currently there are no studies documenting the appearance of the LPR in fetal utricular or saccular maculae. However, whole mounts of utricular maculae aged between 14 and 23 WG, show hair bundles located at the periphery were oriented toward the striola (Rosenhall, 1972b) suggesting an LPR must be present. A comprehensive study of mature hair bundle length in striola of mouse utricle found there were significant differences in the length of stereocilia depending on the type of hair cell and the type of afferent innervation (calyx, bouton, or dimorphic afferents) (Li et al., 2008). Detailed investigation of hair bundle length in developing human or adult vestibular hair cells has not been done. However, hair bundles of the striola are similar in length to those hair cells located in the adjacent extrastriola region (Rosenhall and Engstrom, 1974) suggesting the variation seen in mice may not be present in humans, but this conclusion awaits further analysis. In contrast, at the periphery of the utricular and sacculae maculae, there were a higher proportion of hair cells with short stereocilia (Rosenhall and Engstrom, 1974). This parallels an increased density of type II vestibular hair cells along the edges of the utricular maculae (Rosenhall, 1972b). Taken together these peripherally located sensory cells with short hair bundles are likely to be type II hair cells and may represent an early form of hair cell. Indeed, there are low numbers of hair cells with short bundles found throughout



5

Fig. 3. Hair bundle markers in the developing vestibular neuroepithelium aged 12e14 WG. A. Phalloidin (green) labels stereocilia in the developing cristae ampullares aged 12 WG. B. Acetylated a-tubulin (red) labels kinocilium in the developing cristae ampullares aged 12 WG. C. Composite image of co-labeled stereocilia (A) and kinocilium (B). D. In 14 WG utricle (50 mm thick section) stereocilia were immunoreactive for phalloidin and kinocilium were labeled with a-acetylated tubulin. Hair bundles were of variable length, most likely due to processing technique. It is clear however that at 14 WG there is distinct immunolabeling of stereocilia and kinocilium and is consistent with other studies that show cell polarity is determined before 14 WG. Acetylated a-tubulin also labels microtubules in hair cells and supporting cells. Lim et al., unpublished results.

the developing neuroepithelium. Interestingly, human utricular hair cells with short stereocilia are also found throughout the neuroepithelium from patients aged >60 years (Taylor et al., 2015). This implies there may be a low level of hair cell differentiation and development even amongst aged humans. Recent work has also begun to characterize protein expression in hair bundles from developing human vestibular neuroepithelium. The stereocilia and kinocilium have been immunofluorescently labeled using specific antibodies, phalloidin and aacetylated tubulin respectively (Lim et al., unpublished results). At 12 WG in both cristae and utricule, the kinocilium express a-acetylated tubulin and stereocilia express F-actin (Fig. 3). At high magnification, there is a clear distinction of the stereocilia and kinocilia in a 14 WG utricle (Fig. 3). This image shows utricular hair bundles are relatively short. It has been observed in other species, utricular hair bundles are shorter than those of the crista. Hair bundle length in humans across the various neuroepithelia has yet to be quantified. 3.2.2. Vestibular hair cell synaptogenesis Hair cell innervation by afferent terminals requires the existence of cellular machinery for chemical communication or synaptic transmission. This includes specialized proteins expressed in both the presynaptic hair cell and the postsynaptic afferent terminal. Hair cells of the inner ear use a number of specialized organelles for synaptic transmission including the synaptic ‘ribbons’ (see Nouvian et al., 2006), which are thought to support up-regulated tonic release of neurotransmitter. Indeed, the morphological identification of hair cell synapses in TEM studies typically relies on the presence of presynaptic ribbons, identified as an electron dense structures surrounded by clear vesicles. Synaptic ribbons are diverse in number, shape, and size depending on species and stage of development (Moser et al., 2006). Mammalian vestibular hair cell ribbons can be spherical, ellipsoid, or have a bar or plate-like structure (Bagger-Sjoback and Gulley, 1979; Lysakowski and Goldberg, 1997; Moser et al., 2006). The number of clear vesicles surrounding the synaptic ribbon is dependent on its size and shape. The first evidence of nerve terminals at the base of the human vestibular neuroepithelium occurs at 7 WG, but it is not until 8e9 WG that these afferent nerve fibers contact the newly differentiating vestibular hair cells. Surprisingly, TEM evidence suggests even

at this early stage of development, there are synaptic specializations between hair cells and afferent nerve terminals. In hair cells, presynaptic densities and synaptic bodies are present at the basolateral surface while in afferent terminals, postsynaptic densities are also evident (Sans and Dechesne, 1985). Following initial afferent contact with hair cells, there appears to be a second wave of afferent innervation where nerve fibers of various thicknesses compete to contact hair cells and establish synaptic contacts (Sans and Dechesne, 1987). As the regions of contact increase between hair cells and afferent terminals there is an increase in the length of the pre- and post-synaptic densities. Concomitant with the increase in appositional contact was a proliferation in the number of synaptic specializations (Sans and Dechesne, 1987). In hair cells aged between 9 and 12 WG the length of synaptic bars increased 2e3-fold and the total number of clear vesicles surrounding them also increased (Sans and Dechesne, 1987). In some tissue, aged 11 WG, spherical ribbons were not tethered to presynaptic densities or active zones but appeared to ‘float’ within the cell soma although still adjacent to advancing nerve fibers (Sans and Dechesne, 1987). In addition to contacts between hair cells and afferent nerve fibers, efferent innervation of the vestibular periphery is also established early during fetal development (Sans and Dechesne, 1987). Terminals are considered efferent if they are vesiculated and appose postsynaptic specializations and persist throughout development into adulthood (Schrott-Fischer et al., 2007). TEM studies indicate, efferent fibers make contact with calyx terminals early in fetal development (11 WG) (Sans and Dechesne, 1987) Mature efferent terminals also make direct contact with type II hair cells and afferent axon fibers, however it is not known when these efferent contacts occur during human development. The fast neurotransmitter associated with efferent terminals in the inner ear is acetylcholine, however there have been no studies confirming cholinergic labeling in fetal vestibular epithelium. It has been shown that at 15 WG, the efferent neuropeptide, calcitonin gene related peptide (CGRP) was expressed in fibers of the vestibular ganglion but was not present in the peripheral vestibular organs (Yamashita et al., 1993). A unique morphological feature of the efferent terminal/hair cell synapse seen in the developing auditory system, and in the vestibular system of other species is the synaptoplasmic cistern


6


(Lioudyno et al., 2004). In guinea pigs, synaptoplasmic cisterns lie in the cytoplasm 12e15 nm from the hair cell membrane apposing efferent terminals (Bagger-Sjoback and Gulley, 1979). Since efferent synaptic contacts have yet to be confirmed in fetal type II hair cells it remains to be seen if synaptoplasmic cisterns are also present. There are few studies beyond 12 WG that characterize synaptic specializations in vestibular organs in humans. It is not until samples from humans >60 years of age where synapses in type I and type II hair cells have been described. In aged tissue (60e90 years), synaptic ribbons were observed in both vestibular hair cell types, with multiple ribbons apposing a single afferent bouton terminal in some type II hair cells (Taylor et al., 2015). Synaptic ribbons apposing calyceal terminals were not described. It was noted that in aged tissue, the calyceal terminals were thin and degenerating (Taylor et al., 2015). 3.2.3. Development of vestibular ganglion neurons (VGNs) As mentioned above, afferent innervation of differentiating hair cells occurs early, from 8 WG onward. These afferent nerve fibers originate from vestibular ganglion neurons (VGNs). The first studies to describe the afferent nerves of the vestibular ganglion were from tissue, aged 4 WG (Bruska et al., 2009). At this stage of development, horizontal sections through the embryo show the otic capsule lies adjacent to rhombomere 5 of the developing brain. The neurons of the vestibulocochlear ganglion, which have origins at rhombomere 4, are in close proximity to the otic capsule and form a distinct cluster of cells (Bruska et al., 2009). At 5 WG, the centrally directed nerve fibers from the differentiated VGNs extend toward the brainstem, while peripherally directed nerve fibres begin to develop (Bruska et al., 2009). The anterior and horizontal ampullary nerves are the first to enter vestibular neuroepithelia at 6 WG and by 7 WG the posterior ampullary nerve can be identified (Yokoh, 1971). Since the utricular and saccular maculae are not as developmentally advanced, their respective nerves are not as distinct at this stage (Yokoh, 1971, 1974). As mentioned above, at 7 WG a number of afferent nerve endings are found mostly at the base of the undifferentiated epithelial cells (Sans and Dechesne, 1985) although some nerve endings reach the cells’ upper poles (Sans and Dechesne, 1987). This implies afferent nerves are present in the putative neuroepithelium before hair cells begin differentiation. By 8e9 WG, there is an increase in the number of afferent fibers surrounding the base of sensory hair cells (Sans and Dechesne, 1985). The diameter of these fibers is variable. The increase in afferent fibers coincides with the number of afferent contacts at the base of hair cells and the differentiation of hair cell stereocilia. These hair cells are contacted by vesiculated afferent nerve fiber endings (Sans and Dechesne, 1985, 1987). In some instances, the nerve terminals appear to be calyceal in origin with growth cones beginning to encapsulate future type I hair cells (Sans and Dechesne, 1985). At 10 WG, the afferent terminals of VGNs still can not be clearly distinguished as either bouton terminals apposing type II hair cells or calyx terminals surrounding type I hair cells (Dechesne et al., 1987). Nerve terminal endings were characterized by neuron specific enolase (NSE) a marker of neuronal maturation expressed during periods of proliferation and differentiation. NSE was present in neuronal bodies and fibers of VGNs that penetrated the cristae and utricule. There is also some labeling of hair cells (Dechesne et al., 1987). NSE expression appears to coincide with synapse formation, suggesting at 10 WG, synaptogenesis is still ongoing. It is possible that NSE is expressed at the onset of synaptogenesis (8e9 WG), but this has yet to be confirmed in earlier fetal stages. Supporting this notion are immunolabelling studies that suggest expression of the calcium binding protein, calbindin. Expression of

calbindin may signal the time when vestibular hair cells are capable of releasing neurotransmitter since calcium is essential for this process. There was strong calbindin immunoreactivity in VGNs, afferent fibers, as well as the majority of sensory hair cells of both the cristae and utricule (Dechesne et al., 1987). Some calbindin immunoreactive hair cells had a classic type I hair cell amphora shape, however these were not quantified or further characterized (Dechesne et al., 1987). At later stages of development there are several studies that have characterized the expression of structural proteins in nerve fibers. Briefly, laminin, typically associated with basement membranes, is expressed in soma and fibers of VGNs from 15 WG and therefore suggests a role in neural outgrowth (Yamashita et al., 1993). Expression of the ion pump, NaþeKþ-ATPase is first observed in a subset of VGNs from 16 WG onwards (Yamashita et al., 1993). Over the following 3 weeks, immunoreactivity increases so that a majority of VGN somas were NaþeKþ-ATPase positive by 19 WG, making this a useful marker to help identify VGNs (Yamashita et al., 1993). The cytoskeletal protein actin, was also labelled between the ages of 14e18 WG and found to be expressed in nerve fibers but not the soma of VGNs from 18 WG (Anniko et al., 1987). Immunolabelling was also observed using antibodies against neurofilament, an axonal cytoskeletal protein that helps determine axon diameter. Neurofilament was strongly expressed in VGN soma and fibers that traversed the epithelium toward cristae and utricule from 14 WG, and by 18 WG, large immunoreactive regions within the neuroepithelia have formed. These are presumably terminal endings, but no calyceal terminals were observed (Thornell et al., 1987). The studies above describe VGN innervation of peripheral end organs and the expression of various structural proteins. However, there are no estimates of the number of VGNs that are present in the developing inner ear. Counts from tissue specimens aged between 2 and 88 years old report approximately 25,812 VGNs in the human vestibular ganglion (Park et al., 2001) while another study reported a maximum count of 20,600 VGNs in specimens aged between 9 weeks and 91 years (Richter, 1980). The number of VGNs during the course of development has yet to be determined. Although various aspects of vestibular hair cell differentiation such as ciliogenesis, synaptogenesis, and innervation, have been partly characterized, it remains unknown at what stage of fetal development vestibular hair cells acquire type I and type II hair cell morphological and physiological characteristics. Recent preliminary evidence suggests the acquisition of voltage-gated conductances in fetal vestibular hair cells occurs between 11 and 18 WG. For example, between 11 and 14 WG, voltage activated conductances from all hair cells were similar to those responses evoked from type II hair cells recorded in mice (Lim et al., 2014). These currents typically exhibited small inward currents at hyperpolarized membrane potentials and large outward conductances with depolarization. The peak outward conductances increased in magnitude with age, but even by 18 WG did not reach the same amplitude as currents obtained from isolated adult human vestibular hair cells (Oghalai et al., 1998). However, it wasn't until 15 WG that some cells began to exhibit the low-voltage activated conductance, Gk,l, unique to type I hair cells (Lim et al., 2014). These recordings established a time point in fetal development that vestibular hair cells become differentiated enough to begin expression of physiological characteristics observed in mature hair cells. Interestingly, this chronological time-point coincided with first recordings of calyx terminals that showed the capacity of the afferents to discharge action potentials (see Fig. 4). The ability of calyx afferents to transmit sensory signals to the central vestibular system by 15 WG is consistent with the first reports of human fetal



7

Fig. 4. Whole-cell voltage-activated currents from type I, type II vestibular hair cells, and calyx terminals from human neuroepithelium. Voltage-activated currents from human type II vestibular hair cells aged 12 WG (A) and 17 WG (B) shows outward currents increase in amplitude with age. C. The first recording from a type I vestibular hair cell was at 15 WG. This example shows small low-voltage activated GK,L conductance (arrow), a droop in the steady state current (circle), as well as a collapsing tail currents (asterisk). Inset shows collapsing tail currents in response to voltage steps from 40 mV to þ20 mV on an expanded timescale. D. At 15WG, a calyx terminal is capable of generating a single action potential after injection of a 20 pA depolarizing current step. Inset right shows infrared DIC image of a calyx terminal appears as a hollow ring. Inset left shows a fluorescent halo after intracellular filling the terminal with Alexa-594. Modified from Lim et al. (2014) according to permissions by Creative Commons (http://creativecommons.org/licenses/by/2.0).

vestibular reflexes recorded at 19 WG. The presence of vestibular reflexes at this stage of development suggests hair cells, afferent fibers, and vestibular neurons are connected and form a functional pathway prior to mid-gestation (Humphrey, 1964). 3.2.4. Development of accessory structures e cupulae and otoconial membranes Mature cristae stereocilia project into the gelatinous mass of the cupula. Those of the maculae organs project into the otoconial membrane, a gelatinous matrix, over which lies a layer of calcium carbonate crystals, the otoconia. There is little information regarding the development of these vital accessory structures in human fetal tissue. In mature animals, the cupula is a gelatinous partition or diaphragm that overlies the cristae and is tethered peripherally at the roof and walls of the ampulla. Even in well-prepared tissue the cupula is a difficult structure to preserve and characterize anatomically. Nevertheless, in preparations of the cristae aged 8e9 WG, the cupula can be seen as an amorphous substance deposited on the neuroepithelial surface (Dechesne and Sans, 1985; Sans and Dechesne, 1985). In adults however, the cupula of the cristae is fibrillar in appearance (Rosenhall and Engstrom, 1974). From 7 WG, the otoconial membrane begins to differentiate and spindle-shaped crystals containing calcium, 1 mm in length, are present on the anterior surface of the utricule (Wright and Hubbard, 1982). One week later otoconia are also observed in the saccule (Wright and Hubbard, 1982). By 8e9 WG, otoconia are present near the tips of newly developing hair bundles (Sans and Dechesne, 1985). From 10 WG, a thick layer of otoconia forms but with small groups of immature dumbbell-shaped otoconia also

present (Dechesne and Sans, 1985). There appear to be differences between saccular and utricular otoconial membranes. At 11 WG the surface of the saccular membrane is covered by rudimentary otoconia, cylindrical or rhombohedral in shape, whereas utricular membrane is covered by an amorphous matrix and needle-like otoconia (Sanchez-Fernandez and Rivera-Pomar, 1984). Over the following weeks, both types of otoconial membrane form a wavy network matrix containing spherical spongelike bodies. Otoconia lie adjacent to these spherical bodies or contiguous to the needle-like structures and are larger in size compared to those seen at 11WG (Dechesne and Sans, 1985; Sanchez-Fernandez and Rivera-Pomar, 1984). The gelatinous layers of the otoconial membranes covered almost all the maculae and separated the otoconia from the neuroepithelia by 12WG (Wright and Hubbard, 1982). There was also an increase in the calcium concentration of otoconia with gestational age (Wright and Hubbard, 1982). By 22e24 WG, the otoconial membrane is near maturity with a honeycomb appearance at peripheral borders with otoconia of various shapes (ovoid, rhombohedral, and cylindrical) and ranging in size from ~2 to 3 mm (Sanchez-Fernandez and Rivera-Pomar, 1984). In summary, we have a detailed chronology of morphological development in peripheral vestibular organs, however we know virtually nothing about the molecular signaling proteins involved in this maturation process. For example, no transcription factors have been characterized in human embryonic or fetal vestibular tissue. This remains a substantial obstacle to our understanding of human peripheral vestibular development. Identifying the chronology of transcription factor expression would be essential for developing regenerative therapies in human vestibular tissue.


8


4. Development of the cochlea The cochlea is a complex three-dimensional structure that has completed its formation of 2.5 coiled turns by 10 WG. At this stage of development, only the immature cochlea duct is patent when observed in modiolar cross-section. Reviewing the development of the human cochlea, first we describe the formation of the perilymphatic compartments, followed by Reissner's membrane and stria vascularis to familiarize readers with the gross morphological changes that occur during human fetal development. This description is followed by a timeline of differentiation of auditory hair cells and their innervation, which occurs approximately two gestational weeks later than comparable structures in the vestibular system. 4.1. Development of perilymphatic spaces, the scala vestibuli and scala tympani Differentiation of cochlear hair cells occurs relatively early in development (12 WG; described below), prior to maturation of the three cochlea compartments, the scala media, scala vestibuli, and scala tympani. Indeed, scala tympani and scala vestibuli are only beginning to form as irregular perilymphatic spaces as inner hair cells (IHCs) begin to differentiate in the developing cochlea duct (Kim et al., 2011). The formation of the perilymphatic spaces occurs by vacuolization of mesenchymal tissue surrounding the cochlea duct. This process continues over the subsequent four to five weeks, so that by 16e17 WG the perilymphatic spaces are close to mature size. The scala vestibuli exhibits more mature characteristics than the scala tympani at this stage with the formation of a mesothelial lining that separates the scala lumen from the periotic reticulum (Kim et al., 2011; Streeter, 1917). And by 16 WG, the scala vestibuli and scala tympani have elongated along the length of the cochlea duct and at the apex becomes continuous, thereby forming the helicotrema (Streeter, 1917). 4.2. Development of scala media, Reissner's membrane and the stria vascularis The sensory neuroepithelium that will become the organ of Corti develops from the primordial cochlear duct. Reissner's membrane is an integral structural component of the cochlea. It is a delicate fibrous partition, lined on either side by simple squamous epithelium spanning between the spiral limbus medially and the lateral border of the stria vascularis. It has a crucial role separating the scala media containing endolymph from the scala vestibuli containing perilymph. At 11 WG, Reissner's membrane is formed by layers of cuboidal cells lined by mesenchyme (Lavigne-Rebillard and Bagger-Sjoback, 1992). During the next two weeks, two distinct cellular layers constitute Reissner's membrane e an epithelial layer lining the scala media and a mesenchymal layer lining the scala vestibuli (Lavigne-Rebillard and Bagger-Sjoback, 1992). By 14e15 WG Reissner's membrane has separated from the cochlea duct epithelium, and this results in the compartmentalization of the scala media. By 20e22 WG scala media reaches maturity (Kim et al., 2011). As Reissner's membrane develops, the stria vascularis, the endolymph-producing cells of the cochlea, also begin to mature. Stria development is critical since mechanosensory transduction is not only dependent on hair cell function but also on the secretion of endolymphatic fluid. The stria vascularis is layered epithelial structure that develops from the lateral wall of the cochlea duct. The three cell types of the stria vascularis; marginal, intermediate (melanocytes), and basal cells develop at different chronological times. The marginal cells lining the cochlea duct are the first cells to develop and are present

by 11e12 WG. This is followed closely by the distinct intermediate cells that are present within the epithelium by 12e13 WG (LavigneRebillard and Bagger-Sjoback, 1992). It should be noted that immunolabelling of intermediate cells using melan-A, a melanocyte marker, showed labeling of melanocytes only in the basal turn at 12 WG (Locher et al., 2015). However, there was increased melan-A expression in the periotic mesenchyme of the mid-turn at 14 WG, with penetration into the lateral wall of all turns by 16 WG (Locher et al., 2015). By 17 WG, intermediate cells appear to be notched (Bibas et al., 2000), accepting cytoplasmic projections from the overlying marginal cells. Basal cells however do not emerge as a distinct epithelial layer until 18 WG (Bibas et al., 2000) or later (20e22 WG) (Lavigne-Rebillard and Bagger-Sjoback, 1992). These results suggest potassium-rich endolymph production, necessary for the generation of the essential endocochlear potential, does not occur in the scala media until late in second trimester. This timeline of development of the stria vascularis and endolymph production is consistent with findings that hearing in the neonate occurs at ~24 WG, when auditory startle reflexes are first observed in response to vibroacoustic stimulation (Birnholz and Benacerraf, 1983). 4.2.1. Development of the organ of Corti During the period in which the cochlea elongates and rotates to form 2.5 turns, an immature cochlea duct begins to differentiate in conjunction with the development of the greater epithelial ridge (GER) and lesser epithelial ridge (LER) (Lavigne-Rebillard and BaggerSjoback, 1992). The region between the GER and LER is the site of the future organ of Corti. Studies of the cochlea duct surface show an undifferentiated sensory epithelium at 9 WG (Lavigne-Rebillard and Pujol, 1987) comprising of cells with microvilli emanating from their apical surfaces (Pujol and Lavigne-Rebillard, 1985). As differentiation advances, at 9e10 WG, stereocilia begin to replace the microvilli and lengthen. They are round in cross-section and are initially of similar height (Pujol and Lavigne-Rebillard, 1995). Simultaneously, the tectorial membrane is formed at this time (Lavigne-Rebillard and Pujol, 1987). In mature tissue the tectorial membrane is composed of collagen fibers embedded within a striated-sheet matrix that runs the length of the cochlea (Richardson et al., 2008). In fetal cochlea there are differences in development between basal and apical regions of the tectorial membrane. The basal region of tectorial membrane at 9 WG is a network of fibrils attached to the undifferentiated epithelial surface, while apical regions appear as amorphous substance with superficial fibrils laying on the surface (Lavigne-Rebillard and Pujol, 1987; Sanchez Fernandez et al., 1983). At 11 WG, microvilli are in close contact with the developing tectorial membrane and by 16e17 WG it appears as a compact structure with its characteristic arch, spanning from the spiral limbus to OHCs (Sanchez Fernandez et al., 1983). 4.2.2. Differentiation of auditory hair cells A period of rapid development from 11 WG and includes: initial vacuolization of mesenchyme to form the perilymphatic spaces; development of the stria vascularis; and the differentiation of hair cells results in significant changes in the primordial cochlea neuroepithelium. By 11e12 WG, some nascent hair cells of the putative cochlea epithelium have a well-defined cuticular plate and stereocilia (Sanchez Fernandez et al., 1983) and by 12 WG form a single row of IHCs along the entire length of the cochlea, from basal to apical regions (Pujol and Lavigne-Rebillard, 1985). SEM studies of the epithelium suggest outer hair cells (OHCs) were not obviously differentiated at this stage, however, when viewed in cross-section, using TEM, the differentiating cochlea epithelium showed a single row of IHCs from base to apex, while three and sometimes four



rows of OHCs were observed in basal regions of the cochlea at 12 WG (Pujol and Lavigne-Rebillard, 1985). Two weeks later, three rows of OHCs are observed at the mid-turn region of the cochlea (Lavigne-Rebillard and Pujol, 1990). The 1e2 week delay in differentiation of human OHCs is consistent with that observed in other mammalian species including cats, rats, guinea pigs, and hamsters that show a maturation of IHCs prior to OHCs (Pujol et al., 1980). Also consistent across species is the maturation of basal regions of the cochlea that precedes apical regions. This suggests IHC and OHC cellular and regional maturation of cochlea hair cells is evolutionarily conserved. Although studies describe the differentiation of human auditory hair cells, we still do not know when in fetal development the full complement of IHCs and OHCs occurs along the length of the human cochlea. Total hair cell counts are an important basis for comparison, particularly for studies that describe the effects of ototoxicity, noise-related damage, or age-related hearing loss. Arguably, the maximum total hair cell count would occur at the late fetal or newborn stages of development. In a comprehensive study, cochlear hair cell estimates were made from nine fetal samples (gestational ages not stated) and were compared with counts from forty-four adult specimens, aged between 20 and 79 years (Wright et al., 1987). In these samples, counts of IHCs and OHCs were expressed as a proportion of cochlea length. For example, in fetal samples, the density of OHCs at the base of the cochlea was ~300/mm and at the apex was ~500/mm. There was a significant decrease in OHC density with age, so that >70 years of age, the density of OHCs at the base was ~65/mm and at the apex, was ~230/mm (Wright et al., 1987). For IHCs in fetal samples, the density was ~80/mm at the base of the cochlea and ~110/mm at the apex. In aged samples (>70 years), the density of IHCs decreased to ~30/mm at the base and ~90/mm at the apex (Wright et al., 1987). These data show there are significant losses with age in the density of IHCs and OHCs at the base, but only OHCs at the apex (Wright et al., 1987). Interestingly there were some anomalies in some samples, there were irregular rows of OHCs, and additional IHCs located medial to the typical single row of IHCs. These extra hair cells have also been observed in the developing rodent cochlea but not in mature rodent cochlea. The observation

9

of additional OHCs and IHCs in fetal cochlea (Pujol and LavigneRebillard, 1985) and the continued presence of these hair cells in mature human cochlea suggests hair cells may not undergo further refinement by apoptosis with age. Rather, once a hair cell differentiates, it remains within the neuroepithelium throughout life. 4.2.3. Development of auditory hair cell stereocilia Stereocilia projecting from amongst microvilli was first observed at 9 WG. However, the development of stereocilia in human cochlea hair cells was described in detail from tissue aged 14e22 WG. At 14 WG, a single row of IHCs and 3e4 rows of OHCs showed clusters of stereocilia (Igarashi, 1980). These hair bundles were distinct in shape and increase in height, from medial (modiolar) to lateral direction. In mature adult tissue, IHCs have a rectilinear arrangement of hair bundles and OHCs have ‘V’ or ‘W’ shaped hair bundles with a staircase arrangement in the height of stereocilia (Comis et al., 1990). Early in the second trimester, hair bundles of IHCs have a ‘U” shaped configuration with shorter medial stereocilia and longer lateral stereocilia, and a centrally located kinocilium (Igarashi, 1980). In contrast, the stereocilia of OHCs were short in length and converged at the top without a specific arrangement, suggesting a less mature stereocilia formation (Igarashi, 1980). At this stage of development there are fine-thread like connections among stereocilia that appear to resemble tip links (see Fig. 5) (Igarashi, 1980). Interestingly, tip links were first identified in guinea pigs (Pickles et al., 1984) and were described in the human cochlea shortly thereafter (Rhys Evans et al., 1985). These results suggest that with the presence of tip links by 14WG, hair cells may have the capacity for transduction relatively early in fetal development. It is not until 22 WG that IHCs and OHCs exhibit mature patterns of stereociliary arrangements (Igarashi, 1980). For OHCs there were a total of 120e150 stereocilia arranged in 5e6 rows. At this stage, both IHCs and OHCs possessed a single kinocilium, which are not present in adult auditory hair cells (Igarashi, 1980). The kinocilium project from a bulbous cytoplasmic structure on the apical surface of the hair cell. With age these bulbous cytoplasmic structures decrease in size and were absent by 28 WG. This corresponds with a

Fig. 5. Tip links are present in cochlea hair cells by 14 WG. A. Thin thread-like connections (red arrows) were observed in inner hair cells by 14 WG but not identified as tip-links at this stage (Igarashi, 1980). The first description of tip-links in human cochlea hair cells was made five years later (Rhys Evans et al., 1985). B. In adult cochlea, outer hair cells exhibit tip links, both horizontal (arrows) and vertical (arrowheads) (Modified from Igarashi, 1980; Rhys Evans et al., 1985).


10


loss of some kinocilia from hair cells at ~24 WG (Fujimoto et al., 1981). 4.2.4. Auditory hair cell synaptogenesis The emergence of synapses is a critical feature of hair cell development and maturation. Auditory hair cells, like vestibular hair cells, have specialized ribbon synapses, presumably for the tonic release of glutamate neurotransmitter. Presynaptic specializations are first observed at 11e12 WG as synaptic ribbons or asymmetric thickenings at the base of IHCs and OHCs, at sites apposing afferent nerve connections (innervation discussed below in Spiral Ganglion Neurons) (Pujol and Lavigne-Rebillard, 1985, 1995). At this stage of development, highly vesiculated nerve terminals appose IHCs, and suggests early efferent innervation of the cochlea neuroepithelium (Pujol and Lavigne-Rebillard, 1985). In contrast, no efferent connections were made with OHCs. At 14 WG, the number of synaptic contacts made with IHCs increases (LavigneRebillard and Pujol, 1988). Synaptic ribbons in IHCs were localized to the hair cell basolateral surface and apposed to afferent nerve terminals that were typically swollen and devoid of cytoplasmic content (Lavigne-Rebillard and Pujol, 1988). In OHCs, multiple synaptic ribbons were observed to appose single afferent contacts (Lavigne-Rebillard and Pujol, 1988). The only synaptic contacts of OHCs at this stage of development are with afferent nerve terminals. It is not until 20e22 WG that efferent contacts are made with the base of OHCs, forming axo-somatic synapses (Lavigne-Rebillard and Pujol, 1988; Pujol and Lavigne-Rebillard, 1995). The late innervation of OHCs by the medial olivocochlear system and early innervation of IHCs by the lateral olivocochlear system is consistent with efferent innervation in other species (Simmons, 2002). These SEM and TEM studies have classified differentiated hair cells by one of three different criteria: 1) their location in the neuroepithelium as a single row of IHCs or three rows of OHCs; 2) the expression of hair cell associated structures including hair bundles; and 3) the presence of synaptic specializations. Typically, these reports have characterized morphological changes as a means of identifying sensory hair cells. However, more recent molecular studies have broadened the study of hair cell differentiation by studying transcription factors that are expressed during the development of prosensory domains. Research has shown a differential expression of a number of transcription factors with placode formation and cell fate specification roles in the developing human cochlea epithelium (Kelley, 2006; Locher et al., 2013; Pechriggl et al., 2015). One of the first transcription factors to be

expressed in the inner ear is PAX, which signals cochlea development including sensory cell differentiation and ganglion cell survival (Favor et al., 1996). In PAX2/PAX8 double knock-out mice the ear does not mature beyond the otocyst stage, there is no sensory cell differentiation, and no innervation of the otocyst by developing afferent fibres (Bouchard et al., 2010). During early human fetal development (8 WG) PAX2 is expressed in SGNs and at the site of the future organ of Corti and by 10 WG, both IHCs and OHCs at the basal end of the cochlea express PAX2 (Pechriggl et al., 2015). Expression of PAX2 in basal regions of the cochlea preceded expression at apical regions and showed only IHCs were PAX2positive at 10 WG (Pechriggl et al., 2015). Thus, PAX2 expression has a tonotopic and IHC to OHC gradient of expression as previously observed. It has been shown there is a sequential expression of transcription factors, with PAX preceding SOX expression (discussed below) in the inner ear (Christophorou et al., 2010). The transcription factor SOX2, is known to be crucial in the development of the prosensory domain in mouse cochlea, and SOX2 mutations in humans result in sensorineural hearing loss (Dabdoub et al., 2008). In developing human cochlea aged 12 WG, SOX2 expression is found in the prosensory domain of the epithelium. Adjacent to this prosensory region is Kolliker's organ, a group of non-neuronal support cells expressed only transiently during a critical period of auditory development and was SOX2 negative (Locher et al., 2013). Labeling with the hair cell specific marker myosin VIIa, showed immunoreactivity at the basal turn between SOX2 positive prosensory domain and the SOX2 negative Kolliker's organ, suggesting the presence of putative IHCs (Fig. 6; Locher et al., 2013). While TEM showed morphological features consistent with differentiation of OHCs at 12 WG (Pujol and Lavigne-Rebillard, 1985), myosin VIIa was localized to the single row of IHCs and was not expressed in presumed OHC rows. With time there is increased myosin VIIa expression in IHCs and this coincides with a decrease in the expression of SOX9, a transcription factor important for signaling the initial stages of cell differentiation (Locher et al., 2013). These studies highlight the need for more information on the expression of signaling molecules such as transcription factors that are essential if we are to understand the complex process underlying hair cell differentiation in humans. 4.2.5. Spiral ganglion neurons (SGNs) Differentiation of hair cells is a critical component of the developmental process e similarly innervation of hair cells by SGN afferents is also vital. SGNs are the bipolar primary afferent neurons with peripheral and central projection fibers that transmit sensory

Fig. 6. Transcription factor SOX9 expression and myosin VIIa labeling during development of the human organ of Corti aged 14 WG. A. The apical turn of a 14 WG organ of Corti, shows SOX9 expression (green) in the prosensory domain/organ of Corti (denoted by curly bracket). A single inner hair cell (IHC) is present as shown by myosin VIIa labeling (red). B. At mid-turn, SOX9 (green) and myosin VIIa (red) are expressed in the prosensory domain, labeling a single outer hair cell (OHC) and a single IHC, respectively. C. At the base there is reduced SOX9 expression (green) and increased myosin VIIa immunoreactivity (red). Myosin VIIa labeling shows a single row of IHCs and three rows of OHCs (O1, O2, O3). Abbr: KO e Kolliker’ organ. Cell nuclei in all panels are labeled with DAPI (blue). Scale bar: 20 mm. Modified from Locher et al., (2013)) permissions by Creative Commons (http:// creativecommons.org/licenses/by/2.0). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)



11

Table 1 Summary of major developmental milestones in the development of the human inner ear from 4 to 24 weeks gestation (WG). Weeks gestation

Labyrinth development

4 5

6 7

8

Vestibular organs

Formation of otic vesicle Elongation of otocyst Primordial SCC formation Ductus reuniens forms

Ampulla becomes evident Utricle and saccule separated by cleft Cochlea begins to elongate

9

10

Full 2.5 turns of cochlea complete

11

12

13 14

15

18 19 20 21 22

Crista is defined structure Vestibular nerve enters undifferentiated epithelium Otoconia present in utricle Utricle and 3 cristae are distinct Stereocillia cover apical surface of HCs. Hair bundles are unidirectionally polarized Nerves contact base hair cells Synapses present Innervation increases e terminals not distinct as bouton or calyx HCs express type II voltage activated conductances Efferent fibers make contact with calyx terminals Synaptic specializations increase in length a-acetylated tubulin and F-actin expressed in cristae and utricle

SGNs express PAX2 and MAF-B

Stereocilia replace microvilli on undifferentiated cells

b-III tubulin expressed in SGN fibres Tectorial membrane begins to form IHCs along length of cochlea express PAX2 Only basal OHCs express PAX2 Synapses present Reissner's membrane begins to develop

SOX2 expressed in prosensory region Single row IHCs from base to apex 3e4 rows OHCs at base of cochlea IHCs express myosin VIIa Scala tympani and scala vestibuli begin to form

Tip links present between stereocilia 3 rows OHC at mid-turn of cochlea Scala tympani and scala vestibuli reach anatomical maturity HCs express type I voltage activated Peripherin-ir fibres with spiral organization appose OHCs conductances Reissner's membrane separates from cochlea duct Calyx terminals fire single action potential VGNs express CGRP

16

17

Cochlea

In next 2 weeks, SCC and cochlea radius reach adult size By 19WG, 3-fold increase in labyrinth Vestibular reflexes are first elicited length Scala media reaches maturity Bony ossification labyrinth begins

of

membranous Otoconial membrane reaches maturity

23 24

information from hair cells to the CNS. It is still unclear whether SGN nerve innervation of precursor hair cells is necessary for determining their fate. Although anatomical studies show innervation precedes hair cell differentiation and suggests a link, current evidence does not support innervation as a trigger. Early development of SGNs has been observed from 8 WG using markers against the transcription factor MAF B, a HOX B gene regulator (Pechriggl et al., 2015). At this early stage of development, MAF B labeling is present in most cells of the spiral ganglion (Pechriggl et al., 2015). However, a mature spiral ganglion is composed of two cell types; type I SGNs that innervate IHCs and type II SGNs that innervate OHCs. As yet there are no specific SGN markers that could distinguish between the two types and help identify the fate of SGNs during the next stage of development. In mature tissue microtubule b-III tubulin has been used as a general indicator of all nerve fibers and the intermediate filament protein peripherin, is a specific marker for type II SGN fibers. A recent study has described innervation of the human inner ear from 8 WG using these two markers. At 8 WG the fetal form of the spiral

Mature stereociliary arrangements are observed Spiral afferents express peripherin and radial efferents cross Tunnel of Corti Auditory startle reflex is first elicited

ganglion lies adjacent to the GER. Nerve fibers of the SGN begin to project centrally and express both b-III tubulin and peripherin, suggesting both type I and type II SGNs have early central fibres (Pechriggl et al., 2015). In contrast, nerve fibers that project peripherally penetrated into the epithelium reaching the GER and expressed the presynaptic marker, synaptophysin (Pechriggl et al., 2015). This is puzzling, since the presence of synaptophysin and its close association with synaptic vesicles would suggest presynaptic activity, not usually associated with a postsynaptic afferents. It should be noted, however, synaptophysin immunoreactivity has also been observed in the very specialized afferent calyx terminals that contact type I vestibular hair cells. Therefore synaptophysin in an afferent e hair cell synaptic contact is not inconceivable (Scarfone et al., 1988) and therefore can not be dismissed in the developing cochlea. Alternatively, although less likely, is that these early nerve fibers are efferent but there is no evidence to support this assertion. By 9 WG, b-III tubulin is expressed in SGNs along the length of the cochlea, with peripherally projecting fibers found below the


12


prosensory region at the basal turn by 10 WG. During this period peripherally projecting nerve fibers also began to express peripherin immunoreactivity (Locher et al., 2013; Pechriggl et al., 2015). These results support the potential development of type I and type II SGNs and their innervation of the cochlea epithelium. TEM evidence also suggests that some nerve endings have established contact with the base of individual cells by 10 WG (Pujol and Lavigne-Rebillard, 1985). Over the following two weeks, these afferent nerve terminals proliferate and appose newly differentiated IHCs and future OHCs, with IHCs receiving more numerous contacts (Locher et al., 2013; Pujol and Lavigne-Rebillard, 1985). During mouse embryonic development there is a proliferation of type I SGN fibers that innervate both IHCs and OHCs, and in time are eliminated or “pruned” to form appropriate connections (Huang et al., 2007, 2012). It has yet to be determined if innervation of human auditory hair cells follows a similar pattern of pruning during innervation. Co-labelling of b-III tubulin and peripherin at 12 WG showed a number of fibers with dual expression penetrated the epithelium more laterally at the site of future OHCs, and supports the contention that innervation of future OHCs occurs prior to observable morphological differentiation (Locher et al., 2013). The dual expression of b-III tubulin and peripherin in nerve fibers increased with development, so that by 14e15 WG, these coexpressing fibers contacted nearly all IHCs and OHCs of the basal and mid turns (Locher et al., 2013). Further characterization of peripherin immunoreactivity apposing OHCs showed that by 15 WG, peripherin positive fibers had assumed a spiral organization. The number of peripherin immunoreactive fibers innervating IHCs was reduced by 18 WG and by 20 WG had all but disappeared. Therefore at this time a mature phenotype of peripherin-positive fibers innervating OHCs exclusively has been established (Locher et al., 2013; Pechriggl et al., 2015). At 22 WG, spiral afferent fibers expressing peripherin, and radial efferent fibers cross the tunnel of Corti to contact OHCs (Hoshino, 1990). However, as described above, synaptic specializations between OHCs and efferent fibers were not observed until 20e22 WG (Lavigne-Rebillard and Pujol, 1988; Pujol and Lavigne-Rebillard, 1995). Finally, for fast conduction of sound impulses to the central nervous system, myelination of auditory nerve fibers is necessary. Myelination begins at 20e22 WG (0.02e0.1 mm) and continues throughout the rest of development, so that by 28 WG myelin has increased to reach a mature thickness of ~0.3 mm and remains constant for the rest of development (Ray et al., 2005). 5. Conclusions The majority of studies investigating the development of human inner ear have concentrated on the gross anatomical changes. Recent work, however, has broadened this approach to include cell and molecular levels of enquiry and begun to focus on transcription factors and protein expression. Together with recent progress in electrophysiological techniques, these studies are helping to construct a more complete picture of the developmental process in the human inner ear (see Table 1 for an overview of anatomical, molecular, and physiological developmental time-points). This new information will not only improve our current understanding of human inner ear development, but will also help to improve regenerative technologies and enhance prosthetic devices for hearing and balance. A significant study recently differentiated human embryonic stem cells into otic epithelial progenitor cells (OEPs) and otic neural progenitor cells (ONPs) as potential replacement hair cells and spiral ganglion cells, respectively (Chen et al., 2012). ONPs transferred into gerbils that had type I SGNs previously damaged or destroyed showed these transplanted cells were

able to develop both peripheral and central projections. Importantly, there were significant improvements in auditory brainstem function compared to control non-transplanted gerbils (Chen et al., 2012). Given recent advances using human cochlea and vestibular tissue (Lim et al., 2014; Locher et al., 2013; Taylor et al., 2015), similar stem cell studies could be used to examine the regenerative and restorative capacities of human stem cells transplanted back into mature inner ear tissue. Acknowledgments This research was funded by the Garnett Passe and Rodney Williams Memorial Foundation and the National Health and Medical Research Council of Australia (Grants 1022717, 1048232). References Anniko, M., Thornell, L.E., Virtanen, I., 1987. Cytoskeletal organization of the human inner ear. Acta Otolaryngol. Suppl. 437, 5e76. Bagger-Sjoback, D., Gulley, R.L., 1979. Synaptic structures in the type II hair cell in the vestibular system of the guinea pig. A freeze-fracture and TEM study. Acta Otolaryngol. 88, 401e411. Bast, T., Anson, B., 1949. In: Thomas, Charles C. (Ed.), The Temporal Bone and the Ear. Springfield, Illinois, USA. Bibas, A., Liang, J., Michaels, L., Wright, A., 2000. The development of the stria vascularis in the human foetus. Clin. Otolaryngol. Allied Sci. 25, 126e129. Birnholz, J.C., Benacerraf, B.R., 1983. The development of human fetal hearing. Science 222, 516e518. Bouchard, M., de Caprona, D., Busslinger, M., Xu, P., Fritzsch, B., 2010. Pax2 and Pax8 cooperate in mouse inner ear morphogenesis and innervation. BMC Dev. Biol. 10, 89. Bruska, M., Ulatowska-Blaszyk, K., Weglowski, M., Wozniak, W., Piotrowski, A., 2009. Differentiation of the facio-vestibulocochlear ganglionic complex in human embryos of developmental stages 13-15. Folia Morphol. Warsz. 68, 167e173. Chen, W., Jongkamonwiwat, N., Abbas, L., Eshtan, S.J., Johnson, S.L., Kuhn, S., Milo, M., Thurlow, J.K., Andrews, P.W., Marcotti, W., Moore, H.D., Rivolta, M.N., 2012. Restoration of auditory evoked responses by human ES-cell-derived otic progenitors. Nature 490, 278e282. Christophorou, N.A., Mende, M., Lleras-Forero, L., Grocott, T., Streit, A., 2010. Pax2 coordinates epithelial morphogenesis and cell fate in the inner ear. Dev. Biol. 345, 180e190. Comis, S.D., Osborne, M.P., O'Connell, J., Johnson, A.P., 1990. The importance of early fixation in preservation of human cochlear and vestibular sensory hair bundles. Acta Otolaryngol. 109, 361e368. Dabdoub, A., Puligilla, C., Jones, J.M., Fritzsch, B., Cheah, K.S., Pevny, L.H., Kelley, M.W., 2008. Sox2 signaling in prosensory domain specification and subsequent hair cell differentiation in the developing cochlea. Proc. Natl. Acad. Sci. U. S. A. 105, 18396e18401. Dechesne, C.J., Sans, A., 1985. Development of vestibular receptor surfaces in human fetuses. Am. J. Otolaryngol. 6, 378e387. Dechesne, C.J., Escudero, P., Lamande, N., Thomasset, M., Sans, A., 1987. Immunohistochemical identification of neuron-specific enolase and calbindin in the vestibular receptors of human fetuses. Acta Otolaryngol. Suppl. 436, 69e75. Favor, J., Sandulache, R., Neuhauser-Klaus, A., Pretsch, W., Chatterjee, B., Senft, E., Wurst, W., Blanquet, V., Grimes, P., Sporle, R., Schughart, K., 1996. The mouse Pax2(1Neu) mutation is identical to a human PAX2 mutation in a family with renal-coloboma syndrome and results in developmental defects of the brain, ear, eye, and kidney. Proc. Natl. Acad. Sci. U. S. A. 93, 13870e13875. Fujimoto, S., Yamamoto, K., Hayabuchi, I., Yoshizuka, M., 1981. Scanning and transmission electron microscope studies on the organ of Corti and stria vascularis in human fetal cochlear ducts. Arch. Histol. Jpn. 44, 223e235. Hoshino, T., 1982. Scanning electron microscopic observation of the foetal labyrinthine vestibule. Acta Otolaryngol. 93, 349e354. Hoshino, T., 1990. Scanning electron microscopy of nerve fibers in human fetal cochlea. J. Electron Microsc. Tech. 15, 104e114. Huang, L.C., Thorne, P.R., Housley, G.D., Montgomery, J.M., 2007. Spatiotemporal definition of neurite outgrowth, refinement and retraction in the developing mouse cochlea. Development 134, 2925e2933. Huang, L.C., Barclay, M., Lee, K., Peter, S., Housley, G.D., Thorne, P.R., Montgomery, J.M., 2012. Synaptic profiles during neurite extension, refinement and retraction in the developing cochlea. Neural Dev. 7, 38. Humphrey, T., 1964. Some correlations between the appearance of human fetal reflexes and the development of the nervous system. In: Schade, D.P.S.J.P. (Ed.), Growth and Maturation of the Brain. Elsevier, Amsterdam, pp. 93e135. Igarashi, Y., 1980. Cochlea of the human fetus: a scanning electron microscope study. Arch. Histol. Jpn. 43, 195e209. Jeffery, N., Spoor, F., 2004. Prenatal growth and development of the modern human labyrinth. J. Anat. 204, 71e92. Kelley, M.W., 2006. Regulation of cell fate in the sensory epithelia of the inner ear.


R. Lim, A.M. Brichta / Hearing Research xxx (2016) 1e13 Nat. Rev. Neurosci. 7, 837e849. Kim, J.H., Rodriguez-Vazquez, J.F., Verdugo-Lopez, S., Cho, K.H., Murakami, G., Cho, B.H., 2011. Early fetal development of the human cochlea. Anat. Rec. Hob. 294, 996e1002. Lavigne-Rebillard, M., Pujol, R., 1987. Surface aspects of the developing human organ of Corti. Acta Otolaryngol. Suppl. 436, 43e50. Lavigne-Rebillard, M., Pujol, R., 1988. Hair cell innervation in the fetal human cochlea. Acta Otolaryngol. 105, 398e402. Lavigne-Rebillard, M., Pujol, R., 1990. Auditory hair cells in human fetuses: synaptogenesis and ciliogenesis. J. Electron Microsc. Tech. 15, 115e122. Lavigne-Rebillard, M., Bagger-Sjoback, D., 1992. Development of the human stria vascularis. Hear Res. 64, 39e51. Li, A., Xue, J., Peterson, E.H., 2008. Architecture of the mouse utricle: macular organization and hair bundle heights. J. Neurophysiol. 99, 718e733. Lim, R., Drury, H.R., Camp, A.J., Tadros, M.A., Callister, R.J., Brichta, A.M., 2014. Preliminary characterization of voltage-activated whole-cell currents in developing human vestibular hair cells and calyx afferent terminals. J. Assoc. Res. Otolaryngol. 15, 755e766. Lioudyno, M., Hiel, H., Kong, J.H., Katz, E., Waldman, E., Parameshwaran-Iyer, S., Glowatzki, E., Fuchs, P.A., 2004. A “synaptoplasmic cistern” mediates rapid inhibition of cochlear hair cells. J. Neurosci. 24, 11160e11164. Locher, H., Frijns, J.H., van Iperen, L., de Groot, J.C., Huisman, M.A., Chuva de Sousa Lopes, S.M., 2013. Neurosensory development and cell fate determination in the human cochlea. Neural Dev. 8, 20. Locher, H., de Groot, J.C., van Iperen, L., Huisman, M.A., Frijns, J.H., Chuva de Sousa Lopes, S.M., 2015. Development of the stria vascularis and potassium regulation in the human fetal cochlea: Insights into hereditary sensorineural hearing loss. Dev. Neurobiol. 75, 1219e1240. Lopez, I., Ishiyama, G., Tang, Y., Tokita, J., Baloh, R.W., Ishiyama, A., 2005. Regional estimates of hair cells and supporting cells in the human crista ampullaris. J. Neurosci. Res. 82, 421e431. Lysakowski, A., Goldberg, J.M., 1997. A regional ultrastructural analysis of the cellular and synaptic architecture in the chinchilla cristae ampullares. J. Comp. Neurol. 389, 419e443. Morita, I., Komatsuzaki, A., Tatsuoka, H., 1997. The morphological differences of stereocilia and cuticular plates between type-I and type-II hair cells of human vestibular sensory epithelia. ORL J. Otorhinolaryngol. Relat. Spec. 59, 193e197. Moser, T., Brandt, A., Lysakowski, A., 2006. Hair cell ribbon synapses. Cell Tissue Res. 326, 347e359. Nouvian, R., Beutner, D., Parsons, T.D., Moser, T., 2006. Structure and function of the hair cell ribbon synapse. J. Membr. Biol. 209, 153e165. Oghalai, J.S., Holt, J.R., Nakagawa, T., Jung, T.M., Coker, N.J., Jenkins, H.A., Eatock, R.A., Brownell, W.E., 1998. Ionic currents and electromotility in inner ear hair cells from humans. J. Neurophysiol. 79, 2235e2239. Park, J.J., Tang, Y., Lopez, I., Ishiyama, A., 2001. Unbiased estimation of human vestibular ganglion neurons. Ann. N. Y. Acad. Sci. 942, 475e478. Pechriggl, E.J., Bitsche, M., Glueckert, R., Rask-Andersen, H., Blumer, M.J., SchrottFischer, A., Fritsch, H., 2015. Development of the innervation of the human inner ear. Dev. Neurobiol. 75, 683e702. Pickles, J.O., Comis, S.D., Osborne, M.P., 1984. Cross-links between stereocilia in the guinea pig organ of Corti, and their possible relation to sensory transduction. Hear Res. 15, 103e112. Pujol, R., Lavigne-Rebillard, M., 1985. Early stages of innervation and sensory cell differentiation in the human fetal organ of Corti. Acta Otolaryngol. Suppl. 423, 43e50. Pujol, R., Lavigne-Rebillard, M., 1995. Sensory and neural structures in the developing human cochlea. Int. J. Pediatr. Otorhinolaryngol. 32 (Suppl. l), S177eS182. Pujol, R., Carlier, E., Lenoir, M., 1980. Ontogenetic approach to inner and outer hair cell function. Hear Res. 2, 423e430. Raft, S., Groves, A.K., 2015. Segregating neural and mechanosensory fates in the developing ear: patterning, signaling, and transcriptional control. Cell Tissue Res. 359, 315e332.

13

Ray, B., Roy, T.S., Wadhwa, S., Roy, K.K., 2005. Development of the human fetal cochlear nerve: a morphometric study. Hear Res. 202, 74e86. Rhys Evans, P.H., Comis, S.D., Osborne, M.P., Pickles, J.O., Jeffries, D.J., 1985. Crosslinks between stereocilia in the human organ of Corti. J. Laryngol. Otol. 99, 11e19. Richardson, G.P., Lukashkin, A.N., Russell, I.J., 2008. The tectorial membrane: one slice of a complex cochlear sandwich. Curr. Opin. Otolaryngol. Head. Neck Surg. 16, 458e464. Richter, E., 1980. Quantitative study of human Scarpa's ganglion and vestibular sensory epithelia. Acta Otolaryngol. 90, 199e208. Rosenhall, U., 1972a. Mapping of the cristae ampullares in man. Ann. Otol. Rhinol. Laryngol. 81, 882e889. Rosenhall, U., 1972b. Vestibular macular mapping in man. Ann. Otol. Rhinol. Laryngol. 81, 339e351. Rosenhall, U., Engstrom, B., 1974. Surface structures of the human vestibular sensory regions. Acta Otolaryngol. Suppl. 319, 3e18. Sanchez Fernandez, J.M., Rivera, J.M., Macias, J.A., 1983. Early aspects of human cochlea development and tectorial membrane histogenesis. Acta Otolaryngol. 95, 460e469. Sanchez-Fernandez, J.M., Rivera-Pomar, J.M., 1984. A scanning electron microscopy study on human otoconia genesis. Acta Otolaryngol. 97, 479e488. Sans, A., Dechesne, C., 1985. Early development of vestibular receptors in human embryos. An electron microscopic study. Acta Otolaryngol. Suppl. 423, 51e58. Sans, A., Dechesne, C.J., 1987. Afferent nerve ending development and synaptogenesis in the vestibular epithelium of human fetuses. Hear Res. 28, 65e72. Scarfone, E., Dememes, D., Jahn, R., De Camilli, P., Sans, A., 1988. Secretory function of the vestibular nerve calyx suggested by presence of vesicles, synapsin I, and synaptophysin. J. Neurosci. 8, 4640e4645. Schrott-Fischer, A., Kammen-Jolly, K., Scholtz, A., Rask-Andersen, H., Glueckert, R., Eybalin, M., 2007. Efferent neurotransmitters in the human cochlea and vestibule. Acta Otolaryngol. 127, 13e19. Simmons, D.D., 2002. Development of the inner ear efferent system across vertebrate species. J. Neurobiol. 53, 228e250. Streeter, G.L., 1906. On the development of the membranous labyrinth and the acoustic and facialis nerves in the human embryo. Am. J. Anat. 6, 139e165. Streeter, G.L., 1917. The development of the scala tympani, scala vestibuli and perioticular cistern in the human embryo. Am. J. Anat. 21, 299e320. Taylor, R.R., Jagger, D.J., Saeed, S.R., Axon, P., Donnelly, N., Tysome, J., Moffatt, D., Irving, R., Monksfield, P., Coulson, C., Freeman, S.R., Lloyd, S.K., Forge, A., 2015. Characterizing human vestibular sensory epithelia for experimental studies: new hair bundles on old tissue and implications for therapeutic interventions in ageing. Neurobiol. Aging 36, 2068e2084. Thornell, L.E., Anniko, M., Virtanen, I., 1987. Cytoskeletal organization of the human inner ear. I. Expression of intermediate filaments in vestibular organs. Acta OtoLaryngologica Suppl. 437, 5e27. Watanuki, K., Schuknecht, H.F., 1976. A morphological study of human vestibular sensory epithelia. Arch. Otolaryngol. 102, 853e858. Whitfield, T.T., 2015. Development of the inner ear. Curr. Opin. Genet. Dev. 32, 112e118. Wright, A., Davis, A., Bredberg, G., Ulehlova, L., Spencer, H., 1987. Hair cell distributions in the normal human cochlea. Acta Otolaryngol. Suppl. 444, 1e48. Wright, C.G., Hubbard, D.G., 1982. SEM observations on development of human otoconia during the first trimester of gestation. Acta Otolaryngol. 94, 7e18. Yamashita, H., Bagger-Sjoback, D., Sekitani, T., 1993. The expression of maturation factors in the vestibular ganglia in the human fetus. Acta Otolaryngol. Suppl. 503, 166e169. Yasuda, M., Yamada, S., Uwabe, C., Shiota, K., Yasuda, Y., 2007. Three-dimensional analysis of inner ear development in human embryos. Anat. Sci. Int. 82, 156e163. Yokoh, Y., 1971. Early formation of nerve fibers in human otocyst. Acta Anat. 80, 99e106. Yokoh, Y., 1974. On sensory cells in the human otocyst. Acta Anat. (Basel) 87, 72e76.


Development of the inner ear.

Development of the innervation of the human inner ear.

Sp8 regulates inner ear development.

Immunoelectron microscopy of the human inner ear.

Immunohistochemical techniques for the human inner ear.

Middle-ear and inner-ear contribution to bone conduction in chinchilla: The development of Carhart's notch.

Patterns of epithelial cell death during early development of the human inner ear.

S-100 protein in human inner ear.

The effects of vascular occlusion on the human inner ear.

Inner ear ossification and mineralization kinetics in human embryonic development - microtomographic and histomorphological study.

Anatomical correlates of functional recovery in the avian inner ear following aminoglycoside ototoxicity.

Expression of the heparan sulfate 6-O-endosulfatases, Sulf1 and Sulf2, in the avian and mammalian inner ear suggests a role for sulfation during inner ear development.

[Formation of the inner ear lymphs. Permeability of inner ear membranes (author's transl)].

Interplay of proliferation and proapoptotic and antiapoptotic factors is revealed in the early human inner ear development.

The windows of the inner ear.

Signaling and Transcription Factors during Inner Ear Development: The Generation of Hair Cells and Otic Neurons.

Evolution and development of hair cell polarity and efferent function in the inner ear.

ABO(H) group substances in human inner ear fluid.

Cochlin in autoimmune inner ear disease: is the search for an inner ear autoantigen over?

Human fetal inner ear involvement in congenital cytomegalovirus infection.

Production and role of inner ear fluid.

Gm(1) factor in human inner ear fluid.

Early steps in inner ear development: induction and morphogenesis of the otic placode.

Tricellular Tight Junctions in the Inner Ear.