High Time for Hair Cells: An Introduction to the Symposium on Sensory Hair Cells

Abstract

Sensory hair cells are highly specialized cells that form the basis for our senses of hearing, orientation to gravity, and perception of linear acceleration (head translation in space) and angular acceleration (head rotation). In many species of fish and aquatic amphibians, hair cells mediate perception of water movement through the lateral line system, and electroreceptors derived from hair cell precursors mediate electric field detection. In tunicates, cells of the mechanosensory coronal organ on the incurrent siphon meet the structural, functional, and developmental criteria to be described as hair cells, and they function to deflect large particles from entering the animal. The past two decades have witnessed significant breakthroughs in our understanding of hair cell biology and how their specialized structures influence their functions. This symposium combines the approaches of developmental biology, evolutionary biology, and physiology to share the gains of recent research in understanding hair cell function in different model systems. We brought together researchers working on sensory hair cells in organisms spanning the chordates in order to examine the depth and breadth of hair cell evolution. It is clear that these specialized cells serve a range of functions in different animals, due to evolutionary tinkering with a basic specialized cell type. This collection of papers will serve to mark the progress that has been made in this field and also stimulate the next wave of progress in this exciting field.

Introduction

The Integrative Biology of Sensory Hair Cells was a symposium given on Saturday, January 6, 2018 at the Society for Integrative and Comparative Biology Annual Meetings in San Francisco, CA. The symposium was organized by Duane R. McPherson and Billie J. Swalla in an effort to understand the evolution, development, physiology, and regeneration of hair cells in the chordates. We assembled an international group of outstanding speakers who work on hair cells from a variety of viewpoints in different systems in order to understand these special cells in an integrative way (Fig. 1). The unique mixture of speakers stimulated many questions and we look forward to watching the field unfold further in the coming years.

Fig. 1.

Members of the 2018 SICB symposium on Integrative Biology of Sensory Hair Cells, arranged as a hair bundle. Front row (L–R): Clare Baker, Olivia Berminham-McDonogh, Alberto Stolfi. Second row: Lucia Manni, Bernd Fritzsch, Ruth Anne Eatock, Billie Swalla. Third row: Bifeng Pan, Gerhard Schlosser, David He. Kinocilium: Duane McPherson.

Duane McPherson welcomed everyone and gave the first talk to provide everyone an overview to hair cells (McPherson 2018). He talked about the special structure and physiology of sensory hair cells and set the stage for the rest of the symposium. Sensory hair cells are mechanosensitive sensory cells that are distinguished by the presence of a highly organized bundle of stereocilia on the apical membrane of the cell. The stereocilia (modified microvilli) are connected to one another by tip links that are necessary for mechanotransduction (Müller 2008; Phillips et al. 2008; Schwander et al. 2010; Fettiplace and Kim 2014; Fettiplace 2017). At their basal ends, stereocilia are linked to the apical membrane through a narrow, hinge-like region (Goodyear et al. 2006; Schwander et al, 2010). The result of this elaborate construction is that the hair bundle is an array of stiff columns which swivels at its base as a single unit in response to mechanical perturbations applied at the tips. The final specialized cell, which includes mechanically activated ion channels for mechano-electrical transduction, is sensitive to displacements, at the tip, of as little as one nanometer (1 nm) (Crawford and Fettiplace 1985; Fettiplace and Kim 2014; Fettiplace 2017).

Hair cells using this same transduction mechanism have evolved into a multitude of sensory organs in different animals, including those that sense linear acceleration and orientation to gravity (the utricle and saccule), those that sense rotation of the head in all directions (the semicircular canal system), and those that sense pressure waves in the air, which we perceive as sound (the cochlea). In addition, fish and some amphibians have a lateral line system sensitive to local displacement of water, which is important in certain group behaviors, such as schooling in fish (Kasumyan 2003).

Despite (or perhaps because of) their central role in this wide variety of sense modalities, there have been few symposia devoted to a broad discussion of hair cells. To address that deficiency, we organized the symposium from which this collection of papers was produced, doing so with the design to bring together physiologists and developmental biologists who approach the biology of hair cells from different directions and who would otherwise be unlikely to cross paths at a scientific meeting.

Where are hair cells found?

Many animals have ciliated or microvilli-based mechanoreceptors, statocysts, or other gravity- and acceleration-sensing organs, but hair cells have only been found in chordates (Coffin et al. 2004). Developmentally, they derive from cranial placodes, which are specialized thickenings of ectoderm in the head region, and the otic placodes are the precursors of the entire inner ear of vertebrates (Schlosser 2006; Riley 2013; Schlosser 2015; Schlosser 2018). After its initial appearance as an ectodermal thickening, the otic placode invaginates to form a cup-like structure, and as the edges of the cup converge and fuse, an otic vesicle or otocyst is formed, completely isolated from the outside environment.

Long before observable morphogenetic movements occur, the cells that will give rise to all of the cranial placodes can be identified by their expression of distinctive transcription factors and the subsequently expressed genes for downstream effectors. This expression occurs while the cells still dwell in a region of relatively undifferentiated ectoderm that lies anterior and lateral to the neural plate ectoderm during the interval prior to neurulation. Cells in this region, which is known as the preplacodal ectoderm, express the genes Six1 and Eya1, whose gene products form a transcription factor complex. Six1 is the transcription factor and Eya1 is a protein tyrosine phosphatase (Baker and Bronner-Fraser 2001; Riley 2013). Subsequent specification of cells to produce an otic placode requires fibroblast growth factor 3 (Fgf3) in mice and in chicks (Riley 2013). Further differentiation of the otic placode requires expression of the genes Pax2 and Pax8. Another transcription factor, Sox9, is necessary for the normal invagination of the otic placode to form the otocyst (Barrionuevo et al. 2008; Whitfield 2015).

Differentiation of placodal cells to form sensory hair cells occurs after formation of the otocyst, and depends strongly on the gene Atoh1 (formerly known as Math1). This gene encodes a basic helix-loop-helix (bHLH) transcription factor, and deletion or knock-down of Atoh1 abolishes hair cell development (Bermingham et al. 1999; Fritzsch et al. 2011). Conversely, ectopic expression of Atoh1 can induce differentiation and formation of functional hair cells where none would normally occur (Kelly et al. 2012).

To summarize, our understanding of the molecular biology of hair cell development, while far from complete, contains enough detail to allow interesting comparisons across the vertebrates (mouse, chick, zebrafish, etc.). Another intriguing possibility is that different sensory modalities—in particular, those that originate with cranial placodes—may share common developmental pathways. That is the topic of Gerhard Schlosser’s 2018 paper.

How did hair cells evolve? Where did they originate?

Since hair cells originate in placodes, any ancestral animal with hair cells would be expected to have placodes. Examination of extant vertebrates provides few clues in this search, because all vertebrates, including hagfish, have a functional vestibular system based on hair cells and derived from cranial placodes. But what about our chordate cousins? Both tunicates and cephalochordates develop a neural plate after gastrulation, and that neural plate develops into a hollow, dorsal nerve cord homologous to the vertebrate spinal cord. The presence of a neural plate therefore suggests the possibility of a preplacodal region that could give rise to vertebrate-style placodes and hair cells. Molecular evidence indicates that, indeed, non-vertebrate chordates develop homologs of neural crest tissue and neurogenic placodes (Gasparini et al. 2013; Holland 2013; Abitua et al. 2015; Stolfi et al. 2015). In cephalochordates (commonly known as lancelets or amphioxus, most belonging to the genus Branchiostoma), cells derived from the neural plate border (the region where neural crest appears in vertebrates) express transcription factors characteristic of neural crest, such as Msx, Snail, and Pax 3/7 (Stolfi et al. 2015). Some of these cells become pigment cells and contain melanin, similar to vertebrate melanocytes, while others migrate and differentiate into peripheral sensory neurons, which in vertebrates are derived from neural crest (Stolfi et al. 2015). Also in lancelets, the developing pre-oral organ, known as Hatscheck’s Pit, expresses transcription factors typical of neurogenic placodes, such as Eya and Six 1/2, and Hatschek’s Pit is a candidate homolog to the adenohypophysial placode of vertebrates, which goes on to form the anterior pituitary (Holland 2013; Abitua et al. 2015).

In tunicates there is similar evidence for a homolog to the adenohypophysial placode, including evidence that the resulting neurons express genes for the synthesis of gonadotropin-releasing hormone (GnRH), one of the important hormones released from the anterior pituitary (Abitua et al. 2015). Thus, the ground material for neurogenic cranial placodes, which in vertebrates includes the otic placode, exists in both cephalochordates and tunicates (Burighel et al. 2011).

While evidence for an otic placode homolog has not been shown in amphioxus, there is anatomical and physiological evidence for mechanosensory cells on the tentacles and incurrent siphon velum of some tunicates, including Botryllus schlosseri, which share many features with vertebrate hair cells (Caicci et al. 2007; Burighel et al. 2011; Rigon et al. 2013). Furthermore, there is molecular evidence for a transcription factor network in the embryos of these tunicates that shares significant features with the corresponding network in otic placodes found in vertebrates. This transcriptional gene network includes the tunicate homologs of Six1, Eya1, and FoxI (Gasparini et al. 2013). The paper by Manni et al. (2018) further explores the hypothesis that mechanosensory cells in the tunicate coronal organ are homologous to vertebrate hair cells.

If we knew how it was put together in the first place, we might be able to fix it

Not only are hair cells necessary for sound transduction in the mammalian cochlea, but also they are fundamentally important to the development and maintenance of the organ of Corti, which contains the epithelium on which those hair cells rest. The organ of Corti is much more complex in structure than the auditory epithelia of non-mammalian vertebrates, and its function requires a complex coordination of inner hair cells, which do the work of sensory transduction, along with outer hair cells, which function to amplify the mechanical stimulus applied to inner hair cells. There is also a complex arrangement of supporting cells, and the inner and outer hair cells are divided into discrete, non-overlapping cellular districts.

As mentioned above, the transcription factor Atoh1 is necessary for normal differentiation of hair cells. Without Atoh1, hair cells do not develop and in addition, the organ of Corti is grossly underdeveloped (Pan et al. 2011). When the animal (typically a mouse) is in a post-natal stage, destruction of hair cells by ototoxic drugs not only eliminates the hair cells but also causes the complex organ of Corti to regress to the state of a simple, flat epithelium (Taylor et al. 2012). These results indicate that hair cells have a necessary instructive role in the development of the organ of Corti, and in addition serve a necessary trophic role in the maintenance of a healthy organ of Corti (Fritzsch et al. 2011). The paper by Jahan et al. (2018) explores this topic in depth and offers ideas for future exploration in the field of cochlear regeneration the restoration of hearing to people who suffer from neurosensory deafness.

What else can evolve with hair cells?

It is well-known that fish have a lateral line system which is highly sensitive to water movement near the animal. Like the inner ear, this system uses hair cells for mechanotransduction and the clusters of hair cells and supporting cells, together termed neuromasts, derive from cranial sensory placodes. However, it is less well-known that a sensory system responsive to weak electric fields exists alongside the mechanosensory system in the lateral lines of a wide variety of fish and amphibians (Kalmijn 1988; Tricas 2008; Baker et al. 2013; Piotrowski and Baker 2014). That brings forward the question: are these electrosensory cells related to hair cells, or are they simply borrowing the convenient real estate of the lateral lines? In other words, do hair cells and electrosensory cells share a common developmental and evolutionary origin?

Ampullary electroreceptors are composed of electrosensory and supporting cells at the base of a mucus-filled canal. These organs respond to weak, cathodal (exterior-negative) electric fields and are widespread in chondrichthyans (cartilaginous fish), sarcopterygians (coelacanths, lungfish, amphibians, and caecilians), and actinopterygians such as paddlefish, sturgeons, and bichirs (Jorgensen 2005; Wueringer et al. 2012; Baker et al. 2013). Epidermal electrosensitive end bud organs, which are also cathodal-sensitive, are found in lampreys (Ronan and Bodznick 1986). In situ hybridization for placode-associated transcription factors (Eya1-4, Six1, Six2, and Six4) has confirmed that electrosensory ampullary organs derive from lateral-line placodes in the paddlefish, Polyodon spathula (Modrell et al. 2011; Baker et al. 2013).

Teleost fish have lateral line organs but generally lack electroreceptors; however, they have evolved secondarily in several taxa, such as gymnotids and mormyrids, which use electroreception primarily for orientation and communication. The electroreceptors in these fish are anodal (outside-positive) and have a tuberous morphology, meaning that the canal between the electroreceptors and the environment is plugged by epidermal cells rather than mucus. Some catfish (Siluridae) have ampullary-type organs with a mucus plug but these are also anodal-sensitive, indicating their relationship to other teleost electroreceptors (Baker et al. 2013).

The next question is, are lateral line hair cells and electroreceptors directly linked by evolution and development? That they can be separated is shown by the loss of electroreceptors in most teleost fish, while their combined presence in more primitive fish suggests that they are linked (Piotrowski and Baker 2014). Clare Baker and Melinda Modrell, in their paper, employ molecular evidence for genes in developmental pathways and for genes with physiological function to argue convincingly that electroreceptors in the paddlefish are indeed closely related to hair cells (Baker and Modrell 2018).

Synaptic specializations in the vestibular system

The vestibular hair cells of amniotes (reptiles, birds, and mammals) develop in two distinct ways. Type II hair cells have a synaptic morphology which is similar to cochlear and lateral line hair cells: there is a specialized presynaptic active zone, or ribbon synapse, at each afferent synapse on the basolateral side of the cell, and a similar-sized specialization containing neurotransmitter receptors (the postsynaptic density) on the postsynaptic, primary afferent membrane. The requirements for continuous transmitter release and for rapid, precise changes in the rate of transmitter release have led to anatomical specializations at the ribbon synapse, although the basic principle of calcium-dependent exocytosis there is not fundamentally different what transpires at an ordinary synapse in the brain (Wichmann and Moser 2015).

The synapse between type I vestibular hair cells and their primary afferent partners has an important additional difference: the postsynaptic membrane forms a calyx that surrounds most of the basolateral surface of the hair cell. The functional significance of this arrangement has long been a puzzle, but is beginning to yield its secrets to physiologists. Voltage-clamp studies reveal that type I cells have a higher density of low-voltage activated potassium channels, which would lower the membrane time constant and enhance fidelity to high frequency stimuli. In addition, some calyceal synapses showed very brief synaptic delays, suggesting a non-quantal form of synaptic transmission (Songer and Eatock 2013). The paper by Ruth Anne Eatock explores this unusual synapse in detail and provides new insights into its functional mechanisms (Eatock 2018).

Funding

This work was supported by a grant from the National Science Foundation [IOS Award No. 1809860]; a grant from the National Institute on Deafness and Other Communication Disorders of the National Institutes of Health [Award No. R13DC017092]; the Company of Biologists; the Society for Integrative and Comparative Biology through the following divisions: Division for Evolutionary Developmental Biology and Division for Neurobiology, Neuroethology, and Sensory Biology; and a generous contribution from the American Microscopical Society. The content of this work is solely the responsibility of the authors and does not necessarily reflect the official views of the National Institutes of Health.