3.01 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells G A Manley, Technische Universita¨t Mu¨nchen, Garching, Germany R Ladher, RIKEN Centre for Developmental Biology, Kobe, Japan ª 2008 Elsevier Inc. All rights reserved.
3.01.1 3.01.1.1 3.01.1.2 3.01.1.3 3.01.2 3.01.3 3.01.3.1 3.01.3.1.1 3.01.3.1.2 3.01.3.2 3.01.3.2.1 3.01.3.2.2 3.01.3.3 3.01.3.3.1 3.01.3.3.2 3.01.3.4 3.01.3.4.1 3.01.3.4.2 3.01.3.4.3 3.01.3.4.4 3.01.3.5 3.01.4 3.01.4.1 3.01.4.2 3.01.4.3 3.01.4.4 3.01.5 3.01.5.1 3.01.5.1.1 3.01.5.1.2 3.01.5.2 3.01.6 References
Introduction General Considerations The Concepts of Morphological Homology and Molecular Homology What Is a Ciliated Mechanoreceptor Cell? Phylogeny, Homology, and Homoplasy: The Historical Background of the Animal Groups Ciliated Mechanosensory Cells of Vertebrates and Their Relatives Mechanoreceptors in Tunicate Sea Squirts and Lancelets Mechanoreceptors in tunicate sea squirts Mechanoreceptors in lancelets Vertebrate Hair Cells The typical vertebrate hair cell Lateral-line hair cells of fish and Amphibia The Origin and History of Vestibular Systems in Craniates The structure and diversity of the vestibular organ Auditory-vestibular sensory epithelia of fish The Origin and Phylogeny of a Dedicated Auditory Epithelium The hearing organs of amphibians (frogs and toads and their relatives) The hearing organs of lepidosaurs (lizards and snakes) The hearing organs of archosaur groups (dinosaurs, crocodilians, birds) The hearing organs of mammals A Synopsis of Hair Cell Phylogeny in Vertebrates Ciliated Mechanoreceptors of Nonchordate Animals Cnidarian Hair Cells Mechanotransduction in Caenorhabditis elegans Ciliated Mechanoreceptors in Insects Mollusk Hair Cells Molecular Evolution of Mechanoreceptors Development of Ciliated Mechanoreceptors Extrinsic factors regulating mechanoreceptor development The role of transcription factors in mechanoreceptor development Summary of Development Conclusion
2 2 3 4 5 6 7 7 8 8 8 10 10 11 12 12 13 14 15 16 17 17 18 19 20 23 25 26 26 27 29 29 30
Glossary amniote The largest group of land vertebrates, all originally egg-laying (lepidosaur and archosaur reptiles, birds, mammals). apomorphy An evolutionarily new feature. Bilateria Animals with three germ layers (endoderm, mesoderm, and ectoderm) and that show
bilateral symmetry. All eumetazoans are Bilateria, except cnidarians and ctenophores. cladistics The formal science of the study of evolutionary relationships. deuterostomes Bilaterian animals in which during embryonic development the first opening, the
1
2 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
blastopore, becomes the anus. This includes the chordates as well as echinoderms and hemichordates. ectopic Outside its normal location. endothermy The ability to maintain body temperature independently of environmental sources of heat. Eumetazoa Animals with true tissues – most of the multicellular animals. homology The concept of sameness in evolution, where a structure can be traced back to a common origin. homoplasy The concept of independent evolution of structures in different groups. induction The effect of a tissue on the fate of another tissue, usually mediated by secreted signaling molecules. innervation The establishment of functional contacts between nerve fibers and other cells. neural/abneural The side of the auditory papilla where the nerve enters/the opposite side. paleontology The science of the study of extinct organisms.
phylogeny The study of the evolutionary relationships of organisms. plesiomorphy An evolutionarily old feature. protostomes In these bilaterians, the blastopore becomes the mouth. Protostomes can be subdivided into ecdysozoans (molting animals such as arthropods and nematodes) and lophotrochozoans (such as mollusks and annelids). The split of the deuterostomes and protostomes represents one of the first major bifurcations in the phylogeny of eumetazoans. signaling factor A secreted protein that binds to a cognate receptor to elicit a response. Its effect can be either cell autonomous (in which case signaling is autocrine) or nonautonomous (when signaling is paracrine) or both. synapse The contact point between a nerve cell and another cell that permits fast signaling. transcription factor A protein that binds DNA: it will bind to the promoter or enhancer of a particular gene to either activate or repress its transcription. This usually also requires binding of other transcription factors to different promoter or enhancer elements. Their function is cell autonomous.
The first vertebrate hair cells were probably located on the body surface. Such lateral-line organs specialized as a result of selection pressures to better analyze the stimuli from the environment. Hair cells in the inner ear organs of balance of fish also differentiated into different types and this became structurally obvious in the land vertebrates. Following the origin of a middle ear in lepidosaur, archosaur, and mammal ancestors, unique kinds of hair cell specializations arose that are recognizable along and across the respective epithelia. Structural features are accompanied by functional specializations that differ in the various groups, and in birds and mammals, there is a division of labor between two populations of highly specialized hair cells.
functions: detection of sound and of linear and rotational accelerations in the auditory and vestibular parts, respectively, of the inner ear. These sensory cells – the hair cells – are very similar to each other in both structure and function and they also strongly resemble the hair cells in the lateral-line system of fish and Amphibia. While there is little doubt that these sensory cells of vertebrates share the same phylogenetic origin, similar-looking sensory cells are found in quite different groups of organisms such as cnidarians, tunicates, and cephalopod mollusks (octopuses and relatives; reviews in Coffin, A. et al., 2004; Holland, L. Z., 2005). Other sensory cells in organisms such as insects, which also respond to various kinds of mechanical stimuli, do not strongly resemble hair cells but may still have a common evolutionary history. In view of the presence of similar sensory cells in such a wide variety of species, extreme caution is necessary before making any statements concerning the phylogeny of these cells. This chapter will have two main themes. Following an introduction to some of the important
3.01.1 Introduction 3.01.1.1
General Considerations
This chapter is concerned with the history and relationships of a special type of sensory cell found in modern land vertebrates that subserves dual
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
concepts, it describes first the hair cell types in the various groups of vertebrates and their relatives, with a view to providing an overview of their evolutionary history and of their structural and molecular variation that is linked to sometimes subtle differences in function. Second, it examines some examples of mechanosensory cells from nonchordate animals and seeks to answer the question as to whether these sensory cells have independently evolved to perform similar functions to those of hair cells, that is, whether there is a basis for postulating that they are derived from the same kind of cells in an unbroken descent from a common ancestor (i.e., are they homologous and synplesiomorphic in the cladistic sense?) or whether they are homoplastic or analogous and only resemble each other due to convergence. The criteria for determining common ancestry for morphological and molecular features are necessarily different and most conclusions must be preliminary in nature. The fossil history of soft-tissue structures such as sensory cells is virtually nonexistent, so any hypotheses based on the molecular constituents of sensory cells will probably never really be testable using comparisons of the morphological features of fossil organisms.
3.01.1.2 The Concepts of Morphological Homology and Molecular Homology Homology is both a simple and an elusive concept. At its heart is the idea, rooted in comparative anatomy, of similarity due to common descent, and some have called it the central concept for all of biology (Wake, D. B., 1994). Classically, similarity was measured using a number of criteria: the relative position of a structure with respect to other structures, their continuity through intermediate forms, their connectivity to adjacent structures, their similarity in structural detail and histology, and the correspondence of their developmental origins (Remane, A., 1952; Brigandt, I., 2003). Other, more inclusive and tolerant concepts of homology, such as West-Eberhard’s M. J. (2003) broad-sense homology, have not yet shown their usefulness in establishing ancestry and thus play no essential role in the discussion below. Homoplasy, that is, a similarity due to convergence, parallel evolution, and recurrence, needs to be identified; failure to recognize the characteristics of the homology concept can lead to fundamental misunderstandings of statements on evolutionary, historical commonality that are
3
made from within different disciplines. These distinctions have become important since, in the last century, biology branched out into different experimental fields that use widely different techniques and seek different kinds of biologically relevant information (Brigandt, I., 2003). With regard to the discussion of mechanosensory cells, the usages of the homology concept in comparative biology show some very important differences. Using the criteria defined above, comparative anatomists recognize structures in different species that are considered to correspond to each other and the inclusion of paleontological, comparative-anatomical evidence can provide powerful support for this notion. Homology in this context is a qualitative notion and structures or organs are either homologous or not. In molecular biology, however, the concept of homology is used in the context of the structure of genes and of the proteins they encode. The main criterion by which genes are considered to be homologous is the sequence of nucleic acids in the DNA or in their protein amino acid sequences. This concept is a quantitative one, in which it is said that two sequences are homologous to a certain degree (% sequence identity). At this level, the statement of homology permits comparisons between sequences with reference to their probable functions but only gives a statement of probability about the evolutionary origin of the sequences or about a potential common ancestor (Patterson, C., 1988; Brigandt, I., 2003). Developmental biologists interested in evolutionary questions are faced with the problem of attempting to reconcile these two usages of homology, since at the level of organismic development, the question of the genetic control of ontogeny and the commonality of evolutionary history overlap. It would of course be expected that a homologous structure would be derived during ontogeny by the actions of the same set of genes, expressed in the same order at the same location in the body’s bauplan. Since, however, genes can be involved in a wide variety of developmental processes and single genes may change their sequence during evolution and the gene may therefore change its function, the genetic background that leads to the building of homologous structures may in fact change over time. In addition, genes may become duplicated and diverge in their sequences. Orthologous genes are homologs that diverged as a result of a speciation event, whereas paralogous genes are homologs that diverged as a result of a gene duplication event (Hillis, D. M.,
4 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
1994). Here, the qualitative homology of comparative anatomy with its emphasis on common ancestry needs to be reconciled with the quantitative homology of molecular biology, a nontrivial problem indeed. For the purposes of the present chapter, concepts of evolutionary sameness need to be considered at different levels, especially those of the mechanosensory cells themselves, and also their integration into sensory organs in different animal groups. It is not sufficient to propose homology on the basis of the expression patterns of a few common genes (Adam, J. et al., 1998). A further distraction arises from the fact that frequently, homologous organs for hearing – or auditory organs – in terrestrial vertebrates are sought in invertebrate animals. As Budelmann B. U. (1992) pointed out, this is misguided, since all the invertebrate ancestors that come into question were aquatic and possessed sensory structures geared to respond to a variety of stimuli but not to classical auditory stimuli. To quote Budelmann B. U. (1992): ‘‘. . . any attempt to draw a direct evolutionary line of sound receptors or of hearing from any of the invertebrate groups, or via the hemi- and urochordates, to vertebrates must certainly fail. Also, we should not expect to find a direct homolog [his italics] precursor of the vertebrate hearing organ, or any other sense organ, in any of the invertebrate groups.’’ Thus the search in the invertebrate phyla for homologs of vertebrate hair cells must be restricted to the search for possibly homologous sensory cells and their underlying genetic developmental background.
3.01.1.3 What Is a Ciliated Mechanoreceptor Cell? Put briefly, the cells that are dealt with in this chapter are sensory elements that have a modified cilium (that arises from a basal body) that in some way is implicated in the cells’ ability to respond to mechanical stresses of the cell membrane. Typically, such cells harbor certain ion-channel complexes in the cell membrane that are anchored both to the cell’s cytoskeleton and outside the membrane to some nearby structure (Gillespie, P. G. and Walker, R. G., 2001). The channel can thus be mechanically stressed by changing the distance between the anchor points and this stress changes the open probability of the transduction (channel), increasing the flux of positive ions into the cell and thus depolarizing it.
As will become evident in this chapter, however, there is as yet no certainty that such mechanoreceptor cells of all animal species are derived from a common ancestral type, especially as mechanosensitive ionic channels are known even from bacteria and were thus most likely developed even before eukaryotic cells evolved. There are thus real problems in deciding evolutionary histories, especially when comparing on the one hand the ciliated mechanoreceptor cells of the chordates and their subgroup the vertebrates, which are specialized epithelial cells (hair cells, Figure 1) and thus belong to the group of secondary sensory cells, and on the other hand the ciliated mechanoreceptors of nonchordate groups that are, with few exceptions, specialized neurons with their own axon (Figure 1) and are thus primary sensory cells. In addition, these two kinds of cells show a fundamental difference in the involvement of the cilium in the sensory process. In the evolution of the vertebrate-type hair cell, the involvement of the cilium seems to have been steadily reduced until it may only be important in establishing the polarity of the cell’s sensory outer segment (that consists of large microvilli or stereovilli) and, in some systems, is lost entirely during ontogeny. Not only in the primary ciliated mechanoreceptors, by contrast, but also in secondary ciliated mechanoreceptors of cephalopod mollusks, the cilium is an integral part of the sensory response apparatus and in many cases acts without the involvement of stereovilli. In both kinds of receptor cells, the cilium is generally nonmotile, that is, the normal ability of the cilium to beat has been lost and is obviously not a requirement for the mechanosensory function. Instead, as in visual sensory cells of numerous animal groups (where the cilium has a vastly enlarged outer membrane segment based on a cilium), the cilium has been co-opted for another purpose. The original reason for the involvement of the cilium in chordate hair cells (and their predecessors, whatever they were) is unclear but may go as far back in evolution to ciliated single-celled organisms whose cilia are responsible for the sensitivity to touch. The cilium, the longest extension of the cell membrane, was perhaps used to anchor the mechanoreceptive channels. The cilium in the hair cell lineage (often called a kinocilium) later became supported by stereovilli, stiffened by an actin core, that form a symmetrical collar around the cilium. With the elongation of these stereovilli and their increase in numbers, the transduction channels
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
5
apparently at least partly shifted to their membranes. In some lineages, this made the cilium somewhat redundant. In many vertebrate hair cells, one of the main functions of the remaining (kino-)cilium is as a contact area to overlying tectorial structures. In endothermic vertebrates, there is a strong tendency toward reduction or loss of the kinocilium during ontogeny, but the basal body remains in its position and likely plays an important role on the orientation of the bundle of stereovilli during ontogeny. In Ecdysozoa, such as arthropods and nematode worms, the supporting function of stereovilli was taken over by a connection of the cilium to the animal’s cuticle. In these animals, therefore, all stereovilli have been lost and now only the modified cilium remains and carries the mechanotransductive channels (Thurm, U., 2001) (Figure 1).
3.01.2 Phylogeny, Homology, and Homoplasy: The Historical Background of the Animal Groups
Figure 1 A schematic drawing of the evolution of hair cell like mechanosensitive cells. The figure should be read from bottom to top. Beginning with a mechanoreceptor of a cnidarian that possesses a normal cilium (kinocilium) surrounded by a small number of stereovilli, the evolution of these mechanoreceptors can be traced along three parallel paths in which the stereovilli and kinocilium have different fates. On the left is shown the lineage of the Lophotrochozoa, including annelids and mollusks. In this group, the best-studied case is that of Octopus, which possesses both primary (left, lower figure) and secondary (left, upper figure) receptor cells. In both cases, the receptor cells have no stereovilli but large numbers of kinocilia. On the right of the figure, the lineage to the ecdysozoans is shown using the example of insects, where the receptor cells are primary and there are no stereovilli but one enlarged kinocilium. The central lineage leads to the vertebrates (here represented by a fish) via primitive chordate groups, here represented by an ascidian tunicate. In this lineage, all receptor cells are secondary and both kinocilium and stereovilli are preserved. The bundle in vertebrates shows a stepped height of the stereovilli across the axis of sensitivity, with the tallest stereovilli next to the offset kinocilium. In mammals and birds, the kinocilium tends to degenerate during ontogeny. An original drawing by Johanna Kraus, all rights reserved.
It will be necessary to refer to a wide variety of animal groups in this chapter, since cilia-bearing mechanoreceptive cells have been described from the group considered to be the most primitive of the Eumetazoa (multicellular animals with true tissues), that is, the Cnidaria or jellyfish. In addition to their stinging cells that use a cilium as the sensor element, these organisms have small sensory organs, statocysts, able to encode to the brain the direction of gravitational force. It is believed that the eumetazoans are monophyletic, that is, have a single common origin. Therefore it may be expected that all following groups have inherited the ability to form such cells. However, abundant evidence from other sensory modalities and from considerations of the complement of cell-membrane channels that multicellular animals have inherited from their singlecelled ancestors indicates that, although the genetic basis for mechanosensation exists in multiple cell types, it may only be expressed in a limited and varied subset. The question of homology at the level of cell types is thus not trivial. The eumetazoans are extremely diverse and we will be concentrating on those groups that have been reasonably well studied. All other groups will not be mentioned or even shown in diagrams. A cladistic (phylogenetic) presentation of the groups to be
6 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
discussed is as follows (an indent step to the right indicates a subordinate group): Common ancestor of eumetazoans Cnidaria (jellyfish) Common ancestor of bilaterally symmetrical eumetazoans Ecdysozoa Pseudocoelomata (e.g., nematode worms such as Caenorhabditis) Arthropoda (e.g., insects such as the fruit fly Drosophila) Lophotrochozoa Molluska (e.g., the cephalopod Octopus) Deuterostomia Chordata Tunicata (e.g., sea squirts) Acrania (lancelets) Craniata Agnatha (Ostracodermi, hagfish, lampreys) Vertebrata (animals with backbones) Chondrichthyes (cartilaginous fish) Osteichthyes (bony fish) Teleostei (modern ray-finned fish) Tetrapoda (land vertebrates) Amphibia Amniota Lepidosauria (lizards and snakes) Archosauria (birds and crocodilian) Mammalia (mammals) Evolutionary studies indicate clearly that the three major groups Ecdysozoa, Lophotrochozoa, and Deuterostomia diverged from one another very early and have long histories (more than 600 million years) of independent evolution behind them. In that time, their genomes have evolved completely without reference to each other (if we can exclude horizontal gene transfer via, e.g., common bacterial parasites). It might therefore be expected that their sensory substrates have diverged to a significant extent. In the first section of this chapter, we will discuss the chordates and attempt to trace the history of their structure and function through the ancestral fish to modern aquatic and terrestrial groups. We will describe types of hair cell from the group of the sea squirts or Tunicata, chordate relatives of vertebrates, which may provide insights into possible ancestral states of the hair cells of vertebrates. Since vertebrate chordates are endowed with hard skeletal elements
(cartilage or bone) that generally fossilize well, the fossil history of this group is extremely rich and as a result, the phylogenetic history is very well understood. In addition, the contours of some sensory structures are visible in some fossils. The history of hair cells in this group involves dealing initially with cells of the lateral-line system, then with vestibular, and finally with dedicated auditory sensory organs, whereby the functions of these organs may show considerable overlap. The second part of the chapter will compare ciliabased mechanosensory cells from other major groups, such as the nematode worms and insects from the group of the Ecdysozoa and the cephalopod mollusks from the group Lophotrochozoa. Here, the question will be discussed as to the extent to which the molecular–genetic information concerning a possible homology to vertebrate hair cells is able to provide clear answers to questions of common ancestry.
3.01.3 Ciliated Mechanosensory Cells of Vertebrates and Their Relatives A growing body of literature has demonstrated considerable heterogeneity in morphology and physiology of vertebrate hair cells in different taxa and even within different end organs of the same species. Electroreceptive sensory cells of vertebrates (which will not be discussed here) were also derived from hair cells ( Jørgensen, J., 1989). Since they are no longer mechanoreceptive, their cilium has been lost and any stereovilli present are no longer stiffened. Several lines of evidence support the presence of hair cells in the very earliest vertebrates, and in vertebrate ancestors, the early chordates. Such evidence includes the presence of hair cells in all craniates (Coffin, A. et al., 2004) and in other chordates. Secondary sensory cells with one cilium and numerous stereovilli occur in tunicates (e.g., Bone, Q. and Ryan, K. P., 1978). Some of these hair cells in tunicates have symmetrical and some even have bilaterally symmetrical bundles (as in vertebrates), making it very likely that vertebrates inherited their basic hair cell pattern from their invertebrate ancestors (Shimeld, S. M. and Holland, P. W. H., 2000; Burighel, P. et al., 2003). These animals, represented today by the tunicates and lancelets, may also have inherited their hair cells from even simpler ancestors going right back to the earliest eumetazoans, the cnidarians.
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
3.01.3.1 Mechanoreceptors in Tunicate Sea Squirts and Lancelets Chordates are comprised of three groups: Tunicata (Urochordata, sea squirts, salps, and appendicularians), Acrania (Cephalochordata, lancelets), and Vertebrata (see, e.g., Hennig, W., 1983). There is no doubt concerning the close relationship of the three groups in the phylum Chordata and thus their relevance to questions concerning the phylogeny of vertebrate hair cells. However, the consistent association of vertebrate hair cells and their innervation with ectodermal placodes and neural crest cells, structures that do not occur in nonvertebrate chordates and for which no homologs have been firmly identified, makes considerations of sensory cell homologies difficult. 3.01.3.1.1 Mechanoreceptors in tunicate sea squirts
To illustrate the variety of mechanosensory cells in sea squirts, we will discuss recent data from ascidian tunicates (Burighel, P. et al., 2003; Manni, L. et al., 2004). These animals are filter feeders that are sessile as adults but have freely swimming larvae. As filter feeders, they
7
use large areas of ciliated cells over the gill-slit supports to draw water through the filtering apparatus. Numerous mechanosensory cells are found, and many are presumably involved in monitoring such water currents. Most of the superficial sensory cells of tunicates are of the primary receptor cell type, carrying their own axon. However, a recently described coronal organ in the oral siphon of ascidians contains secondary mechanoreceptive cells that bear strong resemblances to vertebrate hair cells (Burighel, P. et al., 2003; Manni, L. et al., 2004). This similarity is of great interest in view of the fact that the atrial chamber in these animals bears some ontogenetic and gene-expression similarities to the otic placodes of vertebrates with which hair cells are associated (Burighel, P. et al., 2003). In the coronal organ of these tunicates, there are about 2000 mechanosensory cells which are secondary sensory cells that make contact to two different kinds of nerve fibers, both afferent and efferent. These hair cells are often arranged in rows of large numbers of cells. In addition, the stereovillar bundles (90–95 stereovilli) of the cells sometimes surround the cilium concentrically, but in other cases, the bundle is highly asymmetrical, with the cilium on one side (Figure 2; Burighel, P. et al., 2003).
Cilium
Stereovillus
Striated rootlets
Centriole
Tight junction
Glycocalyx Multivesicular body Golgi Nucleus
Rough endoplasmic reticulum
Mitochondrion
Efferent fiber Basal lamina Afferent fiber Figure 2 A schematic drawing of a hair cell accompanied by supporting cells in the coronal organ of the ascidian tunicate Botryllus schlosseri. Reproduced from Burighel, P., Lane, N. J., Fabio, G., Stefano, T., Zaniolo, G., Carnevali, M. D. C., and Manni, L. 2003. Novel, secondary sensory cell organ in ascidians: in search of the ancestor of the vertebrate latetal line. J. Comp. Neurol. 462, 236–249, with permission from John Wiley and Sons.
8 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
The stereovilli of these bundles are either of the same height or graded in height (Manni, L. et al., 2004). In vertebrates, the stereovilli are always arranged such that there is a staircase of height up to the tallest and these are the closest to the cilium or basal body. The suggestion that the above hair cells are homologous to vertebrate hair cells, even to the vertebrate lateral line (Burighel, P. et al., 2003) has, however, been criticized, since none of the tunicate cells has all the characteristics of a vertebrate hair cell and there is a high degree of variation among ectodermal sensory cells within and between different species (review in Holland, L. Z., 2005). Thus, similarities between the ectodermal sensory cells of any one species of tunicate and craniate hair cells may possibly represent convergent evolution rather than homology. However, Shimeld S. M. and Holland P. W. H. (2000) concluded that since the atrium of ascidians develops from a pair of ectodermal invaginations that are topographically similar to the sensory placodes of vertebrates and since the atrial primordia also express members of the Pax2/5/8 gene family as in placodes, these factors point to a possible homology between the atrial primordia of ascidia and the otic placodes of vertebrates. This strengthens the view that the hair cells in both the adult structures are also homologous (Manni, L. et al., 2004). Obviously, further studies of the morphology and genetic control of ontogeny of these cells and their surrounding structures are necessary to enable firmer conclusions to be reached. Mechanoreceptors in lancelets The lancelets (often known as amphioxus and placed in a group known as cephalochordates, of which the best known are in the genus Branchiostoma) are small marine chordates that are most closely related to the vertebrates. Many primary, but also secondary sensory cells with cilia have been described in this group (review in Lacalli, T. C., 2004). Some are scattered in the rostral epithelium of the juvenile and classified as probable chemoreceptors. Others are in the spines of the mouth of the larvae. These cells bear a cilium, are grouped in clusters, and seem to be mechanoreceptors in synaptic contact with peripheral neurons. However, they do not possess a bundle of stereovilli. Lacalli T. C. (2004) concludes ‘‘There are a number of such [secondary receptor] cells in amphioxus, but none is an obvious precursor of any of the secondary sense cells of vertebrates, e.g., [hair cells of] neuromasts, taste buds, or solitary chemoreceptors’’. Thus with respect to sensory cells that are possible homologs of vertebrate hair cells, the lancelets show less promise than the sea squirts.
3.01.3.2
Vertebrate Hair Cells
3.01.3.2.1
The typical vertebrate hair cell The hair cells of vertebrates have a typical structure that makes them immediately recognizable (Figure 3). This very strongly suggests that this kind of cell is plesiomorphic to vertebrates and that
3.01.3.1.2
Figure 3 A schematic drawing of a typical vertebrate hair cell. This secondary receptor cell is innervated by both efferent (here drawn as larger synapses) and afferent nerve fibers. The cell bundle is polarized by the staircase structure of the rows of stereovilli that are joined by tip links. The bundle is also bilaterally symmetrical, based on the eccentric position of the kinocilium. An original drawing by Johanna Kraus, all rights reserved.
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
the first hair cells in vertebrate evolution would have resembled many modern ones. At their apical pole, each hair cell has a sensory hair bundle that is the sensory outer segment responsible for mechanoelectrical transduction (for reviews see Hudspeth, A. J., 1985; 1989). The bundle consists of a true cilium (in this context often referred to as a kinocilium and only lost during development in mammal and bird hearing organs and in electroreceptors) and a (generally shorter) bundle of large, stiff microvilli known as stereovilli (or, confusingly, stereocilia). The cilium is placed eccentrically on the cell surface, and the stereovilli show a gradient in their height, with the tallest next to the cilium. This gradient in height means that the bundle is bilaterally symmetrical and thus polarized in a particular direction. Although we do not intend to review the detailed structure of such hair cells, we need to mention those features shown to have been the subject of natural selection during vertebrate evolution and thus differ between the hair cells seen in different groups, in different sensory organs, or even within one organ (reviewed in Lewis, E. R. et al., 1985). The stereovillar component of the bundle consists of a number, between about 30 and 300, of rigid stereovilli that may have different thicknesses. The stereovilli are anchored not only in the cuticular plate (an actin matrix) of the hair cell’s apex, but also to each other via various connections at different heights up the bundle. The connections between the stereovilli along their shafts help maintain their integrity, whereas the uppermost connections that form between the tip of the next lower stereovillus and the shaft of the taller neighbor (so-called tip links, Pickles, J. O. et al., 1984) play a crucial role in sensory transduction (Hudspeth, A. J., 1989). A very small (a few nanometers) shift of the stereovilli toward the kinocilium puts tension on the mechanotransduction channels and increases their opening probability. This permits positive ions to flow into the cell, depolarizing it, and leading to an increased release of transmitter at the afferent synapse (Hudspeth, A. J., 1989). The cell body of a hair cell can vary considerably in shape and size, from cylindrical cells with a height of over 70 mm, down to cup-shaped cells with a height of less than 4 mm. To a certain degree, the height of the cell body correlates – at least within the same auditory organ – with the bundle height, short bundles being found on shorter hair cells. The bundle height plays a role in determining the cell’s frequency response (shorter bundles correlate with
9
higher frequencies), and such higher frequencies are apparently better processed by smaller cell membranes. Under tectorial cupulae, bundle heights can exceed 130 mm (Lewis, E. R. et al., 1985). The archaic state of vertebrate hair cells sees them connected synaptically to both afferent and efferent nerve fibers and this condition is maintained in almost all hair cell epithelia. The main exceptions are found in the cochleae of mammals and birds, where a clear specialization of the innervation is found, with a strong reduction or loss of the afferent innervation but increase in dominance of the efferent innervation of abneurally lying hair cells (Spoendlin, H., 1978; Fischer, F. P., 1994). In some areas of amniote auditory organs, efferent innervation is also completely lost (Manley, G. A., 2004; Vater, M. et al., 2004). Many features – visible and submicroscopic – of hair cells can change their stimulus–response properties. The response (depending on the sensory organ) will depend on whether the stereovillar bundle is connected to a gelatinous tectorial structure or not and whether this tectorium contains calcareous otoliths or not. It will also depend on the size, number, and thickness of the stereovilli, which together determine the bundle stiffness; stiffer bundles respond to higher stimulus frequencies. It will be noted below, using some examples, that hair cell micromechanics can be changed in a graduated fashion to produce frequency-response patterns that vary gradually over space (generally along a hearing organ, e.g., a phenomenon known as a tonotopic organization). In many auditory organs (see below), the frequency response of the hair cells is largely determined by the stiffness of the bundle and the mass of the tectorial membrane. In lizards, the evolution of auditory organs was largely independent in different families and in some families, the tectorial mass over many of the hair cell bundles has been lost, but this loss of mass has been compensated for by an increase in bundle height and a reduction in stereovillar number of the bundles (see below and Manley, G. A., 2002; 2004). Of course, the type and number of transduction and other membrane channels in the outer and inner segments of the cell also play an important role in determining the cells’ response characteristics. In some cases, as in turtle and bird auditory organs (see below), changes in the number, structure, and kinetics of membrane channels have been shown to strongly influence the gradient of the tonotopic organization (review in Ko¨ppl, C. et al., 2000). Not surprisingly, variability in all of these features has produced a great variety of
10 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
morphological types and corresponding physiological responses. The following sections describe some of this diversity in different animal groups and different sensory epithelia. 3.01.3.2.2 Lateral-line hair cells of fish and Amphibia
Canals in dermal bone indicating the presence of an extensive lateral-line system have been described in the earliest of craniate fossils, in the jawless fish or ostracoderms (Carroll, R. L., 1987). Although there is no way to be sure what the sensory cells looked like, the structure of these canals, their distribution on the body (seven pairs in modern fish), and their innervation patterns indicate that they are homologous to the lateral lines of fish. Here, vertebrate-typical hair cells occur in groups that almost always have their stereovillar bundles embedded in sail-like tectorial masses called cupulae. Except in agnathans, both afferent and efferent fibers (the latter absent in agnathans, Go¨rner, P., 1987) innervate typical hair cells. There is evidence that the earliest hair cells occurred in groups associated with supporting cells in so-called neuromasts that were scattered in pits over the surface of the body, face, and mouth (Denison, R. H., 1966). This complexity was reduced very early by concentrating most or all of the neuromasts in a small number of canals with openings to the outside, most canals occurring on the animal’s head. Such neuromast systems are found in all groups of modern fish and in the aquatic stages of amphibians, although in lampreys, slowly swimming, bottom-living fish and in amphibians, the neuromasts tend to lie in superficial pits rather than be sunken in canals (Young, J. Z., 1981). The lateral-line system is supposed to have given rise to electrosensory cells in at least two separate evolutionary events (Go¨rner, P., 1987; Gibbs, M. A., 2004) and may also even be related to other secondary receptor cell systems such as taste cells (Smith, H. M., 1960). The origin of the lateral-line system as such must be sought in the earliest craniates, since only these have the dorsolateral placodal tissue from which this system (with few exceptions in amphibians) arises during ontogeny (Gibbs, M. A., 2004). In a typical neuromast, the hair cells are arranged such that their bilaterally symmetrical stereovillar bundles show either of two orientations that are 180 opposed to each other. Thus these two groups of cells have their sensitive axes pointing in exactly opposite directions. Thus when one group of cells is stimulated, the other is not or even
suppressed. Indeed, since these hair cells tend to be spontaneously active (indicating that at rest, their transduction channels have a significant opening probability), one population of hair cells would increase, the other decreasing its activity in response to the same stimulus. Neuromast hair cells respond with great sensitivity to displacement of the water over them and these responses provide information about the animal’s own movements, the movements of other nearby organisms, and the presence of structures in the water (reviews in Flock, A., 1967; Russell, I. J., 1976; Sand, O., 1984; Bleckmann, H., 1986). There appears to be several stages in the evolution of lateral-line patterns and development in fish and amphibians (Pichon, F. and Ghysen, A., 2004). Although hair cells in lateral-line neuromast systems tend to look similar to each other, there are few detailed studies of their fine structure and, given their different locations on the body, in canals, or on the surface, it would be surprising if specializations had not occurred. It is possible that the lateralline mechanoreceptors of primitive hagfish resemble an early evolutionary stage of hair cell development, since they lack true stereovilli, their cilium being surrounded by microvilli and the cells lack polarization. It is, however, also possible that these cells in hagfish are degenerate. Lamprey lateral-line cells resemble typical hair cells but usually the neuromasts lack a cupula (review in Gibbs, M. A., 2004). In one study of the sensitivity of lateral-line hair cells to gentamicin that can destroy hair cells in hearing systems, hair cells in canal neuromasts were shown to be sensitive to gentamicin, whereas superficial hair cells were not (Song, J. et al., 1995), indicating some kind of specialization. Since gentamicin acts through its ability to interact with the transduction channels, this suggests that the channels differ in some way between different groups of lateral-line hair cells, as has been noted in vestibular systems of vertebrates (Coffin, A. et al., 2004). In addition, structurally differentiated types of hair cells that may correspond to different hair cell types found in vestibular organs have been demonstrated in a study of the lateral line of the ratfish (Ekstro¨m von Lubitz, D. K. J., 1981). 3.01.3.3 The Origin and History of Vestibular Systems in Craniates A primitive vestibular organ was present in the earliest craniates, the ostracoderms (Carroll, R. L., 1987). Modern hagfish have one semicircular canal and lampreys have two (the anterior and posterior lateral
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
canals), as did the ostracoderms. It is possible that some ostracoderms had three or more, but the orientation of those canals did not allow for their preservation as fossils (Carroll, R. L., 1987). In addition to the sensory epithelia in these canals, agnathans only have a single additional area of sensory hair cells, the macula communis. The structural pattern typical for the vestibular organ of vertebrates, with three semicircular canals and two large spaces, the sacculus and utriculus, each with its own macula that can be supplemented by additional sensory organs in pouches (e.g., the lagenar macula) that arise from these spaces, obviously developed early in the evolution of the gnathostome fish (Figure 4; for details of the variety of organs and references, see Lewis, E. R. et al., 1985). It appears that the sensory epithelium of the macula communis has been repeatedly and perhaps independently subdivided during the long evolutionary history of gnathostome vertebrates (Fritzsch, B. et al., 2002). 3.01.3.3.1 The structure and diversity of the vestibular organ
In the semicircular canals, hair cells are restricted to so-called cristae in the terminal ampullae, where
Figure 4 A schematic drawing of a typical vestibular organ of a vertebrate, here from the Australian bobtail skink Tiliqua rugosa. The vestibular system is seen from the animal’s right side. Dorsal is up and the view is from lateral. The large globe in the center is the sacculus. The three semicircular canals can be seen, as can some of their ampullae. The lagena lies ventrally and is partially obscured by the sacculus. In the lagena, the finger processes of the nerve to the lagenar macula can be seen (right), as can the fan-shaped nerve to the basilar papilla. An original drawing by Johanna Kraus, all rights reserved.
11
there is a cupula covering all hair cell bundles. In the sacculus and utriculus, there are – often large – sensory epithelia known as maculae and here, the hair cells are covered by an otolithic membrane. Other sensory epithelia (such as the macula neglecta and the macula lagenae) are generally small and carry otolithic membranes. All of these organs consist of sometimes very large hair cell epithelia that sit on a firm substrate (Lewis, E. R. et al., 1985). The cristae are sensory epithelia that are stimulated by head rotation that causes a displacement of the cupula due to the inertia of the endolymphatic fluid in the semicircular canals. All hair cell stereovillar bundles in a crista are oriented in the same direction but those of different cristae may be oppositely oriented. In contrast to the relative uniformity of the cristae, great variety exists in the number, size, shape, and degree of specialization of the sensory maculae of the vestibular systems of different vertebrates (Lewis, E. R. et al., 1985). The specializations are obviously not confined to macroscopic structural changes, since changes in the electrical properties of hair cells have also been found. Both structural and membrane properties of the hair cells will exert powerful influences on their sensory response properties. Vestibular sensory epithelia in amniote vertebrates, both cristae and maculae with otoliths in their covering membranes, tend to show two types of hair cells that are differently distributed throughout the epithelia (see Lewis, E. R. et al., 1985 for references). The more primitive kind of hair cell is the so-called type II. It is a mostly cylindrical hair cell that makes small synapses with afferent and, less frequently, efferent nerve fibers. Type I hair cells are more globular or flask-shaped, with a narrow neck and widened apex above it, and are a characteristic of amniote vestibular systems. There is some evidence that, ontogenetically, type I hair cells begin life as type II forms (Masetto, S. and Carreia, M. J., 1997). Most of the type I hair cell is surrounded by a large nerve terminal, the calyx, that may also enclose other hair cells. Efferent fibers synapse on this calyx and not on the hair cell. Especially in mammals, intermediate cell forms are known. Most physiological studies of these two hair cell types have been carried out in mammals, resulting in the tentative conclusion that type I hair cells respond to stimulation faster, with increased gain for small stimuli and an expanded dynamic range, whereas type II hair cells may be more sensitive. The huge synaptic calyx of type I hair cells apparently contains reciprocal
12 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
synapses with the hair cell that are responsible for a feedback process within the calyx that can produce large responses rapidly (Devau, G., 2000). Different hair cell types are not distributed uniformly throughout vestibular epithelia, but their patterns are species-specific (Lewis, E. R. et al., 1985). In general, type I hair cells tend to be concentrated along the center of the epithelium. Especially in otolith organs, a centrally lying strip of hair cells known as the striola tends to contain mainly type I, whereas type II hair cells are more numerous toward the periphery. Within these two types of hair cells and in nonamniotes, which show little evidence of such pronounced hair cell specializations, there is still a great variety in hair cell form due, for example, to the size, shape, and height of the bundles and the precise arrangement of stereovilli in the bundles. 3.01.3.3.2 Auditory-vestibular sensory epithelia of fish
In both cartilaginous and bony fish, hair cells of diverse vestibular maculae mediate the senses not only of rotational and linear acceleration, but also of hearing (for refs see Popper, A. N. and Fay, R. R., 1999; Ladich, F. and Popper, A. N., 2004). Hearing apparently coexists with vestibular responses from end organs that may or may not be covered by an otolithic membrane and, in fish as a whole, there is a great variety in size and structure of the various end organs between and even within systematic groups. The vestibular maculae display a variety of different hair cell orientation patterns, which are believed to contribute to the abilities of those fish to localize sounds (Fay, R. R. and Edds-Walton, P. L., 1997; Lu, Z. and Popper, A. N., 2001). In addition, in some fish groups, hearing sensitivity is improved by additional structures such as the Weberian ossicles to the swim bladder. Although fish vestibular hair cells show no overt evidence of specialized cell types, such cells do have different arrangements of the kinocilium and stereovilli in their bundles (for references see Coffin, A. et al., 2004). The columnar shape of the hair cells resembles the type earlier defined in amniotes as type II. However, biochemical studies using a calcium-binding protein and antibiotic agents revealed the presence of two types of hair cells with a differential distribution in the organs. Hair cells in different areas of fish maculae also show clear differences in their physiological response patterns. This suggests that during vertebrate evolution, hair cell specialization in terms of cell-membrane channels
and other proteins, for example, began before obvious morphological specialization. For agnathan fish, auditory function has not been localized to one of the various vestibular epithelia, and auditory function is poorly understood in cartilaginous fish (where the macula neglecta, lagena, and utricle may all play a role). However, it is well understood in bony fish (Ladich, F. and Popper, A. N., 2004). Whereas in these animals the utricular macula, with few exceptions, is relatively uniform, the saccular and (in fewer species) the lagenar macula vary greatly in size. There is evidence that this variety is linked to a function of these epithelia as both vestibular and auditory organs. The function of hearing is based on the sound in water moving the entire fish, whereby the denser otolithic material tends to lag in phase and therefore stimulate the underlying hair cells. In some groups, the frequency response goes up remarkably high, to several kilohertz, and the frequency responses may be spread systematically on the sensory epithelium. The latter implies a highly organized epithelium, with the properties of individual hair cells (presumably both structural and biochemical) varying systematically from place to place. 3.01.3.4 The Origin and Phylogeny of a Dedicated Auditory Epithelium The dedicated auditory receptor of land vertebrates, the first epithelium in vertebrate evolution that responded exclusively to sound, is called the basilar papilla. It developed, like many vestibular epithelia, as an out-pocketing from the saccular macula. There is evidence that such an epithelium is present in the coelocanth Latimeria (Fritzsch, B., 1992) and thus by inference in the sister group of the ancestors of land vertebrates. It is, however, uncertain whether this was a unique event for all lobe-finned fish and their close relatives or whether Latimeria ancestors developed this epithelium independently of the true ancestors of land vertebrates. One of the difficulties with the assumption of an originating event in the common ancestor of lobe-finned fish and land vertebrates is the situation encountered in modern amphibians, where neither of the two auditory epithelia is conclusively considered to be homologous to the amniote basilar papilla. Thus either the ancestors of the Lissamphibia lost the basilar papilla and later developed their own two auditory organs or the Latimeria epithelium does not indicate that the basilar papilla of amniotes can be traced back to the
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
lobe-finned fish. Using the second possibility, this would mean that the amniote basilar papilla arose in the ancestors of the stem reptiles. 3.01.3.4.1 The hearing organs of amphibians (frogs and toads and their relatives)
In an extensive review of amphibian hearing mechanisms, Smotherman M. and Narins P., (2004) state that ‘‘Differences in the physiology of hearing can be seen within every branch of the amphibian phylogenetic tree, and range from such gross morphological differences as the absence of a middle ear down to subtle differences in hair cell stereovillar architecture.’’ Considering this great diversity, we will ignore the middle ear and morphology of the endolymphatic system (that show important differences to the amniotes, Smotherman, M. and Narins, P., 2004), and restrict our discussion here to the specializations of the hair cells in the hearing organs of amphibians. As noted above, modern amphibians, which are considered to be a monophyletic group, in addition to the retention of some acoustic sensitivity of the sacculus as in most fish, usually have two hearing epithelia, known as the basilar and the amphibian papillae. In contrast to the amphibian papilla, the basilar papilla is not found in all groups of Amphibia. The amphibian papilla is not considered to be homologous to the amniote basilar papilla, whereas there is considerable discussion as to whether the basilar papillae of amphibians and amniotes are homologous (for a review see Smotherman, M. and Narins, P., 2004). If the basilar papilla is a homolog, it has lost many of the structural (e.g., the freely moving basilar membrane) and functional characteristics of the amniote basilar papilla (i.e., the hair cell specialization and tonotopic organization). The amphibian papilla probably originated as an outpocket near the duct joining the utriculus and the ventrally lying sacculus. It varies greatly in its structure throughout the amphibians. In some amphibian lineages, the amphibian papilla, with its tectorial membrane, is elongated and equipped with many more hair cells. At the distal end of the papilla there is a so-called contact membrane that permits pressure waves to leave the surrounding space, enabling sensitive responses to sound. The most primitive amphibian papilla is probably that seen in urodeles and the most primitive anurans and is a small patch of sensory hair cells located on the dorsomedial wall of the recess. More advanced anuran amphibians have a caudal extension to the sensory epithelium, the
13
length of which correlates with an increase in the frequency-response range processed by this papilla (Smotherman, M. and Narins, P., 2004). The basilar papilla of amphibians is found in all frogs, some salamanders, and a minority of caecilians. It contains a very small patch of 50 or so hair cells located beneath a thin tectorial membrane, the latter being stretched across the recess of the basilar papilla. Behind the sensory epithelium lies a contact membrane separating the recess from the periotic space. The basilar papilla is anatomically very uniform in all amphibian groups, but its exact location varies between the groups. In strong contrast to the hair cell array of the amphibian papilla (and indeed to hearing epithelia of amniotes), all the hair cells of the basilar papilla of the amphibians respond to a single, narrow frequency range that is species-specific. The basilar papilla thus appears to be a simple resonant structure (Meenderink, S. W. et al., 2005) and it would not be expected that the hair cells would differ from one another in any pronounced way. The amphibian papilla, by contrast, shows specializations in different amphibian groups in the presence or absence of electrically tuned hair cells, in the characteristics of the hair cell bundles and in the hair cell orientation patterns. Electrical tuning of hair cells, known from the fish and amphibian sacculus, probably arose fairly early in the evolution of hair cells and provides an important mechanism for frequency selectivity at least for frequencies below about 1 kHz (review in Fettiplace, R., 1987; Manley, G. A., 2000). It is found in about half of the hair cells of the amphibian papilla and in a small percentage of hair cells from the basilar papilla, although the latter only show this resonance in a different frequency range to that of their acoustic responses. In electrically tuned hair cells of the amphibian papilla in the anuran Rana pipiens, the hair cell response frequencies are organized into a tonotopic spatial array in which the best response frequency of any given hair cell is partly determined by the number and the kinetics of particular membrane channels that vary systematically along the epithelium (Smotherman, M. and Narins, P., 2004). The electrical properties of the saccular and amphibian papilla hair cells are fundamentally similar, but in the amphibian papilla, additional structural gradients in the hair cells influence the range of best frequencies along the epithelium. Hair cell orientation patterns in the anuran amphibian papilla and basilar papilla are diverse. Most basilar papilla hair cells are polarized in the same direction, as are hair cells of the amphibian
14 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
papilla in the Caudata and Gymnophiona. In the more derived anuran amphibians, however, the amphibian papilla exhibits a much more complex array of hair cell orientation patterns, but no functional correlates have yet been reported (Smotherman, M. and Narins, P., 2004). 3.01.3.4.2 The hearing organs of lepidosaurs (lizards and snakes)
The phylogeny of the lepidosaur hearing epithelium has recently been extensively reviewed and will thus only be summarized here (Manley, G. A., 2000; 2002; 2004). The snake auditory papilla is very simple in its structure and resembles that of primitive lepidosaurs (tuataras) and the presumably primitive hearing organs of turtles. It consists of an elongated array of up to 1000 hair cells that are all oriented in the same direction (Miller, M. R., 1978; Wever, E. G., 1978) and whose response frequencies, which reach up to about 1 kHz, vary systematically from low frequencies apically to the highest frequencies basally. Although in the turtle papilla, the tonotopic arrangement of frequency responses is known to be due to both hair cell stereovillar bundle structure and electrical tuning (see Manley, G. A., 1990 for references), there are no recordings from hair cells or from the auditory nerve of snakes. Considerably more information is available on specializations of hair cells in lizards (reviews in Wever, E. G., 1978; Miller, M. R., 1992; Manley, G. A., 2004). In these animals, the auditory epithelium is unique among amniotes in that it is always divided along its length into at least two hair cell arrays in which the hair cells belong to two different types which respond to different frequency ranges. In one of these areas, the hair cells (for which Miller, M. R., 1992 used the term UDT or unidirectional type to describe their unique morphology) are almost always oriented in one direction (abneurally), receive both afferent and efferent nerve fibers, are always covered by a tectorial membrane, and respond to frequencies below 1 kHz. Although in this respect they resemble the entire turtle papilla, there is no evidence for a pronounced electrical tuning of these lizard hair cells. The other hair cell array consists of hair cells (Miller, M. R., 1992 referred to these as BDT or bidirectional type) which are generally in groups that have their bundles oppositely oriented (Figure 5) and have only an afferent innervation. Their response frequencies lie higher than 1 kHz and generally up to 5 kHz, in some cases higher. These hair cells may or may not be covered by a tectorial membrane, which
Figure 5 A schematic drawing of hair cell orientation pattern in lizards. In this figure, two hair cells are shown from the high-frequency area of a lizard papilla to illustrate the opposite orientation of the stereovillar bundles. In the species illustrated here (a lacertid lizard), the kinocilia are bulbed at their tips. Such high-frequency hair cells are micromechanically tuned and receive only an afferent innervation. An original drawing by Johanna Kraus, all rights reserved.
can take different forms. The response properties of these two hair cell populations differ not only in their frequency ranges but also in their intensity function patterns and their ability to phase lock to sinusoidal sound stimuli (Ko¨ppl, C. and Manley, G. A., 1990; Manley, G. A. et al., 1990; Eatock, R. A. et al., 1991; Manley, G. A., 2004). It is likely that the plesiomorphic lizard basilar papilla contained three hair cell areas, one central area of UDT hair cells flanked at both ends by areas of BDT hair cells that were mirror images of each other (Manley, G. A., 2000; 2004). In some of the various families of lizards, one of these areas was lost (presumably because, being mirror images, one of them was redundant). In different lizard lineages, either the apical or the basal area was lost (Manley, G. A., 2002); in snakes, both areas were lost. Since the BDT areas respond to higher frequencies, primitive lizard papillae have a curious tonotopicity in which the cell response frequencies increase toward the two ends of the papilla. Where the apical high-frequency
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
area was lost, the tonotopicity resembles that of most other amniotes, increasing steadily toward the basal end of the papilla (Manley, G. A., 2000). Where the basal BDT area was lost, as in geckos, the tonotopicity is, as expected, reversed, a unique situation among amniotes (Manley, G. A. et al., 1999). It thus appears as if the specializations of hair cells in lizards vary between the plesiomorphic (for stem reptiles) UDT area, which in its properties continues in the long tradition established in the saccular macula. The BDT hair cells are almost unique among amniotes in their total loss of the efferent innervation (some specialized regions of some bat cochleae show this phenomenon). The BDT hair cell area(s) that are an apomorphy of lizards consist of hair cells that do not use electrical tuning, but instead perfected the mechanism of frequency selectivity known as micromechanical tuning. In such cells, the frequency selectivity results from the precise mechanical properties of the hair cell bundles (such as the bundle stiffness) and of the structures that surround them, such as the mass of the tectorial membrane (Figure 6). Precise and location-specific variations in these properties, often in groups of hair cells that are coupled, are the basis for the tonotopic organization. Knowledge of these gradients makes it possible to model the frequency-response limits and gradient (e.g., Manley, G. A. et al., 1989). Since the basilar membrane in lizards is not tuned to different frequencies along its length (see Manley, G. A., 1990 for references), all of the frequency selectivity must originate in the hair cells and, if present, their local tectorial structure. 3.01.3.4.3 The hearing organs of archosaur groups (dinosaurs, crocodilians, birds)
The evolution of the auditory epithelium of archosaurs has recently been reviewed (Gleich, O. et al., 2004) and will therefore only be discussed here with reference to hair cell specializations. A comparison of the inner ears of the group of crocodilians with those of modern birds and an examination of recent fossil findings of casts of the inner ears of dinosaur groups have made it possible to speculate that the archosaur inner ear originated early in their lineage and is an apomorphy of the entire group. The hearing epithelium is elongated, up to 11 mm in length. It differs from that of lepidosaurs on the one hand and mammals on the other (both groups in which hair cells tend to be arranged in rows along the papilla) in that the hair cells and supporting cells form irregular
15
Figure 6 A schematic drawing of four hair cells to illustrate the principal structural changes underlying micromechanical tuning in the high-frequency hair cell areas of lizard auditory papillae. The hair cell on the left has no tectorial structure and this is associated with tall stereovillar bundles. The three cells to its right each carry a tectorial structure, which is shown highly schematically as if isolated, but could be part of a continuous tectorial membrane or of a local salletal unit. The mass of the tectorial structure and the bundle height vary systematically along the papilla and are responsible for the gradual change in the preferred response frequency of the hair cells. In this case, the frequency would rise from left to right hair cells. An original drawing by Johanna Kraus, all rights reserved.
mosaics. In addition, the width of the papilla can be large and there may be up to 50 hair cells in any transverse section of the wider, apical end of the epithelium. This contrasts with the 4–6 hair cells seen across the width of the papillae of lizards and mammals. Thus, although on average the avian and crocodilian papilla may be shorter than that of mammals, the number of hair cells can easily be as large or larger. In addition to the above, the archosaur epithelium is not all arrayed over the freely moving basilar membrane; instead, a substantial proportion of hair cells lies over the neural limbus, a firm and thick structure. It is thus apparent that the tectorial membrane, which is thick and covers all hair cells, surrounding their bundles in dome-like, individual concavities, plays a critical important role in hair cell stimulation (Manley, G. A., 1995).
16 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
The archosaur auditory papilla is also characterized by substantial gradients in different planes and in different structural and molecular features. It is presumed that together, these micromechanical and electrical features create the tonotopic organization of the papilla (Gleich, O. et al., 2004). There are clear morphological gradients, such as in the mass of the tectorial membrane, the height and width of the hair cells, the height of their stereovillar bundles, and the number of stereovilli per bundle and the pattern of innervation (Figure 7). At the extremes, apical, neurally lying hair cells are very tall, their bundles have relatively few stereovilli that are tall and covered by a thick tectorial membrane. They receive both afferent and efferent innervation. At the other extreme, basal, abneural hair cells are much shorter, their bundles have 5–6 times as many stereovilli that are short and covered by a thinner tectorial membrane. These cells are only efferently innervated (Fischer, F.
Figure 7 A schematic drawing of hair cell specializations in birds. These two hair cells illustrate the extremes of the structural differences shown by hair cells across and along the avian auditory papilla. On the left is a tall hair cell, which is both afferently and efferently innervated. The cuticular plate tends to go deeply into the cell. Such tall hair cells are typical for apical and neural hair cell areas. On the right is shown a short hair cell that is only efferently innervated. The apical surface of this cell is tilted somewhat toward the observer in order to more clearly show the broad, toothbrush-like, stereovillar bundle structure. Such short hair cells are typical of basal and abneural hair cell areas. An original drawing by Johanna Kraus, all rights reserved.
P., 1994; Gleich, O. et al., 2004). In crocodilians, there tends to be a more sudden change from the one hair cell type to the other across the papilla. In addition and parallel to these morphological gradients, there are systematic changes in the molecular constituents of the cell membrane in various positions on the papilla, especially in those ion channels that are involved in electrical tuning (see, e.g., Rosenblatt, K. P. et al., 1997; review in Ko¨ppl, C. et al., 2000). The number and kinetics of KCa channels, which vary in their calcium and voltage sensitivities, are differentially distributed along the papilla, strongly suggesting that they also contribute to the tonotopic frequency arrangement. Alternative splicing can potentially generate thousands of ionchannel phenotypes that, combined with variations in calcium sensitivities, modulation by subunits, and differing channel numbers per cell, would easily account for a large range of frequency responses. The avian auditory papilla provides one of the most interesting examples of genetic flexibility that creates a wide spectrum of functional variations within one type of cell. The gradients described and the patterns of bundle orientation produce an exactly tonotopically organized auditory epithelium, with the lowest frequencies apically and the highest basally. There is also evidence that the tall hair cells lying over the neural limbus are more sensitive than hair cells lying further abneurally, with, in the starling, a change of 6 B per hair cell across the epithelium (see Gleich, O. et al., 2004 for references). The expected change of hair cell sensitivity due to rotation of the bundle toward the middle of the papilla is compatible with these data, as is the fact that abneurally lying hair cells are not even afferently innervated. It is possible that abneural, short hair cells function as amplifiers in a similar way to the outer hair cells of mammals.
3.01.3.4.4 The hearing organs of mammals
With the exception of the egg-laying monotreme mammals, the auditory epithelium of mammals is coiled helically with the neural side on the inside of the helix. The hair cells are generally strictly arranged in rows, with a single row of inner hair cells and, except at the apex in some groups, three rows of outer hair cells (Figure 8). This specific arrangement, which is common to both placental and marsupial mammals, likely arose after their separation from the monotreme mammals that have
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
17
This clear and profound specialization into two hair cell types that has been described in both birds and mammals is the clearest case of extreme differences between hair cell groups that has been observed in vertebrates. It exists at the same time as the general presence of gradients in both hair cell types along and across the papilla that determine, for example, their frequency responses and thus the pattern of the tonotopic organization. 3.01.3.5 A Synopsis of Hair Cell Phylogeny in Vertebrates
Figure 8 A schematic drawing of hair cell specializations in mammals. The two distinct hair cell populations of the mammalian cochlea are illustrated in this figure. On the left is a typically bulbous inner hair cell that is innervated by a large number of afferent fibers, which in turn receive synapses from a few efferent fibers. Its bundle is straight and made up of only a few rows of stereovilli. The outer hair cell on the right is typically cylindrical and mainly, but not exclusively, innervated by large efferent endings. Its bundle has a W shape, where the angle between the long arms of the W can vary with position along the cochlea. An original drawing by Johanna Kraus, all rights reserved.
more rows of both inner and outer hair cells (for references see Vater, M. et al., 2004). Inner hair cells are somewhat globular in shape, enclosed by supporting cells, and innervated by large numbers (15–30) of afferent nerve fibers that themselves (but seldom the hair cells) are efferently innervated. The hair bundle of inner hair cells contains only a few rows of stereovilli, and the bundle stands parallel to the long axis of the epithelium. Outer hair cells are more columnar in shape and their basal areas lie free in the fluid of Corti’s organ. Their bundles contain more stereovilli and are in general arranged in a W-shape, with the open side of the W roughly pointing toward the inner hair cells. Outer hair cells, although numerically about 80% of all cochlear hair cells, are only innervated by 5–10% of the afferent nerve fibers, most of their basal synaptic area being contacted by large efferent endings (Spoendlin, H., 1978). There is much evidence that, in convergent evolution to the archosaur papilla, outer hair cells have a reduced afferent innervation because their main function is within the hearing organ, amplifying weak sounds. The resulting stimulus is then detected and transmitted to the brain via the inner hair cells.
It is likely that the first vertebrate hair cell organs were lateral lines that contained a single type of hair cell. With differentiation of the lateral line into different types of organs, hair cell specialization set in as a result of selection pressures to better analyze the sensory information offered by the environment. Similarly, the analysis of stimuli by semicircular canal cristae and vestibular maculae depended not only on canal dimensions and the characteristics of the otolithic masses overlying the maculae, but also on hair cell differentiation. During the evolution of fish, the fundamental specialization of vestibular hair cells into two types began with minor structural and biochemical changes, but this became much more pronounced in the land vertebrates, where the two types are easily recognizable via overt structural features. Although it is likely that the hearing organs of land vertebrates began by utilizing only type II hair cells, unique kinds of hair cell specializations arose in lepidosaur, archosaur, and mammal ancestors that are recognizable along and across the respective epithelia. These generally show functional specializations at least as profound as their differing structural features to the extent that in birds and mammals at least, the functions of stimulus detection and stimulus amplification have been separated by a division of labor between two populations of highly specialized hair cells.
3.01.4 Ciliated Mechanoreceptors of Nonchordate Animals Ciliated mechanoreceptors are found even in the earliest of Eumetazoa; however, the consideration of their homology with vertebrate auditory hair cells is not trivial. As we have already mentioned, considerations of homology can be made using comparative anatomy, which are qualitative (a structure
18 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
is either homologous or not), or can be made using molecular biology using sequence comparisons that are quantitative. As data exist on both anatomy and molecular biology, both aspects need to be considered. In this section we primarily focus on describing the anatomy of ciliated mechanoreceptors from a sampling of nonchordate groups, starting from the most primitive, the cnidarians, and progressing to ecdysozoans as exemplified by the nematode and dipteran insects and the lophotrochozoan cephalopod mollusks. We describe the structural characteristics of the ciliated mechanoreceptor found in these groups and discuss the molecular correlates that are known. While it is possible that an archetype for the ciliated mechanoreceptor is present (Vinnikov, Y. A., 1982), its significance must be placed within the context of the phylogeny (Adoutte, A. et al., 2000). In most cases the ciliated sensory neuron is of the primary type, that is, it can be thought of as a bipolar neuron, with its own axon. Secondary ciliated sensory neurons, seen in coelid cephalopod mollusks and in vertebrates, are modified epithelial cells and do not have axons. 3.01.4.1
Cnidarian Hair Cells
Cnidarians, the group containing hydra, jellyfish, and sea anemones, are basal eumetazoans positioned before the radiation of metazoans into deuterostomes and protostomes. This location highlights the importance of the cnidarians in any evolutionary discussion. Cnidarians do possess ciliated mechanoreceptors, responsible for both touch sensation and gravity sensation; they also show morphological variation depending not only on species but also on the function within an individual. In cnidarians, ciliated mechanoreceptors are found in specialized orientation organs, and also serve a combined chemosensory and mechanosensory role in the various types of stinging cell, the nematocysts. The cnidocyte is a unique and defining feature of the phylum Cnidaria, playing a role in prey detection, capture, and defense (Ruppert, E. E. et al., 2004; Figure 9). A cnidocyte itself houses a large fluidfilled sac, a cnidae which itself contains a large inverted tubule. Upon appropriate stimulation, the tubule is rapidly everted and penetrates the prey, in many cases concomitantly releasing toxin that causes stinging or paralysis. Three types of cnidocytes have been identified; two of them, spirocysts and pthychocysts, are only found in the class Anthozoa (corals, sea anemones, sea ferns, and sea pens); indeed
Figure 9 A schematic drawing of a nematocyte of the Cnidaria. These cells are connected synaptically at their base to nerve cells and/or epithelial muscle cells. Within the cell is the nematocyst, a capsule containing a thread that can be rapidly discharged into the prey. An original drawing by Johanna Kraus, all rights reserved.
pthychocysts are only found in the epidermis of the tube anemone. The third type of cnidocyte is the nematocyst, found throughout the cnidarian phylum and, for our purposes, the most interesting. These are associated with supporting cells. The nematocyst itself is part of a type I sensory neuron and bears a sensory cilium. In anthozoans (sea anemones and corals), the cilium is motile (and called a ciliary cone) whereas in hydrozoans and
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
scyphozoans it is nonmotile (and called a cnidocil) (Peteya, D. J., 1975; Holstein, T. and Hausmann, K., 1988). The ciliary cone consists of a long cilium with a typical 9 2 (these are the peripheral doublets of A and B microtubules) þ2 (a central pair of microtubules) configuration of microtubules, a common feature of sensory cilia (Vinnikov, Y. A., 1982). The cilium is surrounded by large diameter stereovilli of shorter length than the cnidocil itself, which are in turn surrounded by a palisade of shorter, smaller diameter stereovilli from the adjacent cells, the supporting cells (Peteya, D. J., 1975). The numbers of these larger and small stereovilli can vary; however, all seem to be stiffened by actin filaments (Wood, R. L. and Novak, P. L., 1982; Stidwill, R. P. et al., 1988). The resemblance of the sensory apparatus of the nematocyst to the hair cells of vertebrates suggests a mechanosensory function of these cilia, and indeed some data indicate that this is the case. Using a loosepatch, cell-associated recording method in slices of anemone tentacles, Mire and Watson measured the electrophysiological changes of the hair bundles of the nematocyst under mechanical stimulation (Mire, P. and Watson, G. M., 1997; Watson, G. M. and Mire, P., 1999). When the hair bundles were deflected using puffs of seawater, changes in membrane potential were recorded. Some cells were hyperpolarized in a graded manner, while others depolarized, depending on the direction of the deflection. By plotting the response amplitude as a function of the deflection magnitude, it became clear that the negative response is smaller and saturates relatively easily, whereas the positive response is larger and saturates more gradually, a characteristic of other mechanotransducers such as hair cells of the vertebrate inner ear (Hudspeth, A. J. and Corey, D. P., 1977). Indeed the presence of tip links in the nematocyst hair bundle may suggest that a process common to the vertebrate hair cell and involving a mechanically gated ion channel is responsible for mechanoreception (Watson, G. M. et al., 1997). Ciliated mechanoreceptors do not occur just in nematocysts; hydrozoan, scyphozoan, and cubozoan species also have a gravity-sensing organ known as the statocyst (Horridge, G. A., 1969). As shown by statocyst ablations in the hydrozoan Aequorea, this organ is responsible for the righting behavior of the animal (Singla, C. L., 1975). There is a huge variety of statocysts in the cnidarians, even within those of the same class; however, they all have a cell, the lithocyte, which contains a large concretion of some calcium salt (Horridge, G. A., 1969). The weighted
19
lithocyte is deflected when the position or equilibrium of the animal is altered. Ciliated sensory cells that surround the concreted cell register this deflection. These ciliated cells are type I sensory cells, that is, cells that have a distal cilium and a proximal axon (Horridge, G. A., 1969). There are three types of ciliated sensory cells found in the cnidarian statocysts that have been examined. Leptomedusae typically have a single kinocilium and lack associated stereovilli. In Narcomedusae and Tracymedusae the sensory cells bear a kinocilium that is surrounded by a ring of stereovilli (Singla, C. L., 1983). The third type of ciliated sensory cell is found in a subset of statocysts in the basal cushion and sensory papilla. The kinocilia are extremely long and stiff with a well-developed rootlet system extending into the cytoplasm. While the cytomechanics of nematocyst hair bundle function are beginning to be characterized, little work has been performed on the cytomechanical function of the cnidarian statocyst mechanoreceptor. Although it is likely that a mechanism consistent with the gated-channel model also functions in the statocyst hair cells of cnidarians, until the electrophysiology is investigated and until the molecular candidates are known, any proposed mechanism remains speculative. 3.01.4.2 Mechanotransduction in Caenorhabditis elegans The nematode worm C. elegans has been used extensively as a model organism for developmental biology. The lineage of each individual cell has been traced and, when combined with the genetic tractability of the animal, has led to an understanding of the interactions that specify these lineages. C. elegans hermaphrodites are known to have 302 neurons, and a number of them are responsible for different mechanosensory behaviors: the gentle body touch, the harsh body touch, the nose touch, and a response to texture (Ernstrom, G. G. and Chalfie, M., 2002). As the neurons that mediate both the harsh and gentle body touch are not ciliated and detection appears to be mediated by varicosities, these will not be considered further (Chalfie, M. and Thomson, J. N., 1982; Chalfie, M. et al., 1985). There are some behaviors in C. elegans that are mediated by ciliated mechanosensory neurons. One is the nose touch avoidance, which causes the worm to move back if it bumps its anterior pole (Kaplan, J. M. and Horvitz, H. R., 1993). This behavior is mediated by
20 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
a number of sensory neurons, named ASH, FLP, and OLQ. The ASH neurons are mono-ciliated type I sensory neurons. Perkins, L. A. et al., (1986) describe three segments to the cilium. The proximal zone, found at the base of the cilium, consists of 9 2 microtubules attached to the membrane by Y-shaped links arranged around a central cylinder. This cylinder contains a core of a variable number of singlet microtubules. Further distally, in the transition zone, there is no central cylinder, although the odd singlet microtubule can be seen. The peripheral microtubules spread apart, causing an apparent loss in the doublet structure and indeed the B subfibers are gradually lost toward the distal segment, the end of the microtubule. This, in effect, causes the cilium to flare, becoming thicker distally. The other neurons of this pathway, the FLP and OLQ neurons, rather resemble the CEP and ADE neurons which are described below. The CEP and ADE neurons mediate a behavior related to texture sensation that indicates food (Sawin, E. R. et al., 2000). When the worm encounters a bacterial lawn on an agar substrate, it slows down, presumably to dwell for longer so it can ingest more food. This behavior can be mimicked using sepharose beads and so it is likely that this slowing behavior is due to mechanoreception and not chemosensation. Indeed these cells express the worm homolog of the mechanotransduction channel found in Drosophila sensory bristles, NompC (Walker, R. G. et al., 2000). The neurons themselves (i.e., the CEP, ADE, FLP, and OLQ) are mono-ciliated type I sensory neurons. The transition zone in their cilia resembles that of the ASH cilium, with 9 2 peripheral microtubules arranged around a central core (Perkins, L. A. et al., 1986). However, some of these doublets terminate just distal to the transition zone. There is some variety in the characteristics of the distal segments of these neurons. For example, in the CEP cilia, the microtubules are interspersed with an electrondense material and mold them into irregular rods. In OLQ neurons, the cores of the microtubules making up the doublets are filled. Both the CEP and OLQ neurons are anchored into the cuticle of the worm using a small nubbin, a varicosity on one side of the cilium, at the base of the transducing region. An interesting property of the texture sensation behavior is its modulation depending on whether the animal is satiated or starved. This depends on serotonin and has provided some insights into the mechanisms by which neuromodulation can impact on mechanosensation (Sanyal, S. et al., 2004). These
kinds of studies highlight the strength of C. elegans for studies into neuromodulation. Synaptic contacts have been found on the CEP neurons from the neurons mediating the gentle touch (Chalfie, M. et al., 1985). It is thought that encountering bacteria may register as a gentle touch (causing the animal to stop); by direct neurological input, the gentle touch response is abrogated when the animals encounter bacteria. While C. elegans is a good model for the modulation of mechanosensation and the mechanisms by which other modalities can impact mechanoreception, there is still no clear mechanism as to how the mechanosensory transduction channels respond to vibrations, and indeed none that conforms to the gated-spring model proposed for hair cell stereovilli. One conjectured model could be similar to that proposed for Drosophila mechanotransduction (see below). Here NompC on the sensory neuron is attached to NompA anchored on a supporting cell. While NompC is found on mechanosensitive CEP neurons, a C. elegans NompA homolog has not yet been characterized.
3.01.4.3 Insects
Ciliated Mechanoreceptors in
Insects have three basic types of ciliated mechanoreceptor: the bristle organ, the campaniform sensillum, and the chordotonal organ (Keil, T. A., 1997). All of them share common features, although they have become elaborated for their particular functions. The basic structure of what has been termed the organule is a central type I ciliated sensory neuron innervating a structure formed by three support cells (Lawrence, P. A., 1966). The cilium of the sensory neuron can be separated into two distinct domains. The cilium itself starts just above the distalmost centriole in the neuron and is initially of the 9 2 þ 0 type. As the cilium extends distally, it may flare out, increasing in diameter and in the number of microtubules. At its most distal point the cilium contains a highly ordered cytoskeletal complex, the tubular body. This is made of a few to many thousands of short microtubules tightly packed into an electron-dense structure (Keil, T. A., 1997). This is attached to the membrane via various connectors, which is in turn attached to the dendritic cap, an extracellular matrix (ECM) structure secreted by the membrane of the sheath cell (also known as the theocogen) using fine connectors (Thurm, U. and Ku¨ppers, J., 1980).
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
As well as the sheath cell and the ciliated sensory neuron itself, two other cell types form the mechanosensory apparatus. The trichogen (shaft) cell forms the hair, which is most obvious in the bristle organ. The tormogen (socket) cell forms the joint membrane at the base of the hair. Both cells retract from the cuticle later in development to form the outer sensillum lymph cavity. This cavity is isolated from the rest of the subcuticular cavity by a seal formed by the periphery of the tormogen cell. The sensillum lymph is high in Kþ and low in Naþ ions (Thurm, U. and Ku¨ppers, J. 1980; Gru¨nert, U. and Gnatzy, W., 1987). The supporting cells, actively transporting potassium ions from the hemolymph, maintain this ionic balance. By analogy with vertebrate hair cells, it is probable that the movement of the cilium causes an opening of cation channels allowing a rapid flood of Kþ into the neuron and leading to neuronal depolarization. As we have already discussed in relation to the cnidarian nematocyst, many mechanoreceptors show a directional sensitivity; movement of the cilium in the opposite direction inhibits receptor potential. Insect mechanoreceptors, at least those found on the bristle organ, also show the same profile (Corfas, G. and Dudai, Y., 1990). While many mechanoreceptors can be incredibly sensitive, the response decreases if deflection is sustained, that is, the cells show adaptation (Walker, R. G. et al., 2000). Models have been proposed that draw similarities to the gating-spring model proposed for vertebrate hair cell function (Gillespie, P. G. and Walker, R. G., 2001). This model requires that an ion channel be anchored under tension between the ciliary cytoskeleton and the dendritic cap. Deflection causes a certain number of channels to open; further deflection would cause more channels to open. Indeed adaptation is proposed to use a similar mechanism to the vertebrate hair cell, where the channel on the cilium would slide along the cytoskeletal tracks, in one direction passively and in the other direction using a molecular motor, thereby altering the tension on the channel. The relevant structural features of insect mechanoreceptors can be clearly seen in the bristle mechanoreceptors. Here the dendritic cap links the cilium of the sensory neuron to the hair embedded in the cuticle of the insect. Another mechanoreceptor, the campaniform sensillum, is quite similar to the bristle organ and is responsive to deformation of the cuticle. Rather than connecting to a hair, the ciliated dendrite makes a connection with a button on the cuticle, the dome cuticle. The organization of this structure closely
21
mirrors that of the bristle organ with perhaps some alteration in the function of the trichogen cell. The structure of the chordotonal (also known as the scolopidial) cell is, at first sight, somewhat divergent; however, it still follows the organule plan of insect mechanoreceptors, even though the morphology and arrangement of the cells are somewhat different. The chordotonal organ is used as an internal pressure sensor; however, it is also the mechanoreceptor responsible for hearing in many insects. In contrast to the rather solitary bristles and sensilla, the chordotonal units (known as scolopidia) are organized into arrays in the various regions of the insect. The structure of a typical insect bristle mechanoreceptor is illustrated in Figure 10. The main feature of the scolopidium is the scolapale space, a fluid-filled, spindle-shaped space (Todi, S. V. et al., 2004; Yack, J. E., 2004). This space encloses the ciliated dendrite of the type I sensory neuron, the cilium of which is suspended between two surfaces where it responds to vertical deflections and deformation. The cilium itself is encased in the dendritic cap (more correctly called a dendritic sheath) made by the theocogen (sheath) cell. Just before entering the sheath, the cilia show a dilation containing an
Figure 10 A schematic drawing of the structure of a typical superficial insect mechanoreceptor, in this case a bristle organ. The cytological plan of chordotonal organs is similar to that shown here. Mechanosensation is thought to depend on the relationship between the cilium and its dendritic cap. An original drawing by Johanna Kraus, all rights reserved.
22 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
electron-dense oval body. The cilium itself shows the same axoneme pattern, 9 peripheral doublets of microtubules without any central filaments (9 2 þ 0). Proximally there are extensive ciliary rootlets that run far into the cell body of the neuron, indeed some pass the nucleus. Scolopidia that are responsible for hearing are typically innervated by two or, sometimes, three type I ciliated sensory neurons (Hallberg, E. and Hansson, B. S., 1999). The significance of this is unclear. There is yet more variation at the distal end of the scolopidium; in one type of scolopidium, known as amphinematic, the cilium is attached to the distal cuticle through the elongated dendritic sheath connected to what can be considered to be a modified tubular body of the cilium. The other type of scolopidium is the mononematic. Here, the cilium is indirectly connected to the distal integument through a cuticular cap (provided by the dendritic sheath) that is in turn associated with one or two distally located attachment cells (Yack, J. E., 2004). In both types of scolopidium are massive bundles of actin cables, the scolopale rod, that give the scolopale cell (which can be considered to be a modified theocogen or sheath cell) its stiffness and afford some tension on the ciliated dendrite. These rods have also been used to suggest a mechanism of mechanotransduction based on a string on a bow model (Todi, S. V. et al., 2004; 2005). As mentioned earlier, the scolopidia are used in some insects as auditory mechanotransducers. The array of scolopidia that constitute the hearing organ is called, in flies, Johnston’s organ (Boekhoff-Falk, G., 2005). In the fruit fly Drosophila and in the mosquito, this is found as a specialized region in the joints between basal segments of the antennae. However, this is not the case in all insects. In the bladder grasshopper, the main hearing organ is found in the first abdominal segment, with five smaller organs on abdominal segments A2–A6, and in the hawkmoth the hearing organ is found located just lateral to the tongue (van Staaden, M. J. and Ro¨mer, H., 1998; Go¨pfert, M. C. and Wasserthal, L. T., 1999). An interesting aside here, and one outside the scope of this chapter, is the further diversity of hearing organs within insects. In many insects, where sound plays a very important role in their lives (e.g., for mating, prey detection, and predator avoidance), as in some moths and true crickets, the animal uses a tympanal hearing organ (Stumpner, A. and von Helversen, D. 2001; Yack, J. E., 2004). This is characterized externally by a cuticular-derived tympanal membrane,
which is deflected by sound. The tympanum is associated internally with an air sac derived from a trachea. The scolopidia are attached to the air sac or to the tympanal membrane. Drosophila is used as the major insect model organism and the ability to identify molecules by directed genetic screens makes this preeminent in defining the components of the transduction system involved in mechanosensation. Two such screens have been carried out. Kernan M. et al. (1994) took extracellular recordings of bristle organs of mutagenized larvae, finding a number of genes through which mechanosensation was affected. Eberl D. F. et al. (1997) used a property of Drosophila courtship behavior for their screen. Male Drosophila will vigorously court when they hear a component of the mating sound (known as the pulse sound) produced when the wings are rubbed together (BennetClark, H. C. and Ewing, A. W., 1970). In fact this sound is so effective that when presented with a recording of it, male Drosophila will court other males (von Schilcher, F., 1976). This provided a very simple behavioral screen. In total, Eberl D. F. et al. (1997) identified 15 mutants in which auditory mechanosensation was perturbed. Many of these genes have now been identified. Possibly the most intriguing discovery, and one that indirectly led to the identification of a vertebrate mechanotransducing channel, is the identification of the No mechanoreceptor potential C gene or NompC (Walker, R. G. et al., 2000). Flies that are mutant for NompC show a complete absence of mechanoreceptor potential in bristle organs. The product of the NompC gene is a part of the TRP (transient receptor potential) gene family. This provoked a great deal of interest, as the TRP family had been shown to mediate a large number of sensory processes such as vision, chemosensation, thermosensation, and osmosensation (Montell, C., 2005). Furthermore, this family was represented in numerous species from different phyla, ranging from nematodes (C. elegans) to vertebrates. The vertebrate homolog of NompC is in the TrpN subgroup and seems to form part of the mechanotransduction complex in fish (Sidi, S. et al., 2003). However, the only vertebrates where a TrpN/NompC has been found are fish and frogs; TrpN/NompC has not been found in birds or mammals and indeed it appears to have been deleted in the mammalian and chicken genome (Boekhoff-Falk, G., 2005; Gillespie, P. G. et al., 2005). TrpN1 is expressed at the tip of the microtubule-based kinocilium of the frog hair cells (Shin, J. B. et al., 2005), whereas TrpA1 is found at the tip of the actin-based stereovilli (Corey, D. P. et al., 2004). As we
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
have already described, mammal and, to a significant extent, avian hair cells do not have a kinocilium, relying instead on just the stereovilli. Indeed it is possible to correlate the presence of TrpA1 and TrpN1/NompC with the presence of stereovilli and kinocilia in the ciliated mechanoreceptors. Unex-pectedly, the absence of an inner ear phenotype of TrpA1 mutants (despite an absence of auditory response in other forms of ablation in the mouse) suggests that there may be another, as yet uncharacterized participant in the mechanoreceptor transduction channel of vertebrate hair cells (Kwan, K. Y. et al., 2006). A cautionary note: while NompC mutants show no mechanoreceptor potential in the bristle organs, auditory function is more or less normal. This indicates that other Trp channels operate in Johnston’s organ, the auditory organs of the fly. Indeed two TrpV homologs can be detected, and both are localized to the proximal part of the ciliated dendrite (Kim, J. et al., 2003; Gong, Z. et al., 2004). Interestingly, the localization of TrpV channels is mediated by heterotetramerization with another unknown channel that is more localized. Mutants for either Drosophila TrpV channel show mislocalization of the other channel to the ciliated dendrite (Gong, Z. et al., 2004). Thus far, TrpV homologs in vertebrates have not been characterized. Other molecules that are important in vertebrate mechanoreception have also been shown to play a role in Drosophila mechanosensation. Mutations, in mammals, of the atypical myosin VIIa result in stereovilli that are incorrectly positioned and disarrayed, and indeed this is the cause of the human deafness syndrome Usher 1B (Hasson, T. et al., 1995; Weil, D. et al., 1995). In flies that are mutant for the myosin VIIa homolog crinkled, the scolopale cells are disorganized, perhaps due to a loss of the apical attachment of this cell (Todi, S. V. et al., 2005). Another molecule that appears in both mammalian and Drosophila hearing organs is prestin, a member of the solute carrier-26 (SLC26) family (Weber, T. et al., 2003). The SLC26 family has diverse functions; in particular they are anion exchangers. In mammals, prestin is found in outer hair cells of the cochlea; however, it does not act as an anion transporter. Instead it acts as a motor driving cochlear amplification of sound signals (Zheng, J. et al., 2000). Cochlear amplification via prestin (as contrasted with cochlear amplification by an active bundle mechanism; Manley, G. A., 2001) is a process that occurs exclusively in mammals (Ge´le´oc, G. S. and Holt, J. R., 2003). Its presence in the auditory structure of the fly is
23
therefore perhaps unexpected and until the function of Drosophila prestin is elucidated, statements about the relationship between the fly and vertebrate prestin homologs will be speculative. This brings us to consider the larger issue of homology between the insect scolopidium and the vertebrate inner ear hair cell. Morphologically, the difference is simple; flies uses the deflection and deformation of a ciliated dendrite made of microtubules for mechanosensation, while vertebrates use actin-based stereovilli. Despite the presence of homologous molecules, mechanistically there must be differences between both modes of mechanosensation.
3.01.4.4
Mollusk Hair Cells
Thus far we have discussed nonchordate mechanoreceptors in cnidarians, a basal metazoan, and in nematodes and insects, both examples of Ecdysozoa (moulting animals). This leaves the lophotrochozoans. Ecdysozoa and Lophotrochozoa are protostomes; however, the two lineages diverged soon after the protostomes/deuterostomes split. Thus, an evolutionary perspective needs an example from the Lophotrochozoa. Mollusks are the largest group in this taxon and are also the most diverse. They range from chitons (Polyplacophora), snails and slugs (Gastropoda), bivalved animals such as oysters (Bivalvia) to octopuses, squids, and Nautilus (Cephalopoda). We will focus on the ciliated mechanoreceptors of cephalopods, which are perhaps the most advanced of invertebrate mechanoreceptors. Cephalopods have a very well developed nervous system, and indeed remarkable sensory organs that are comparable in complexity to those found in vertebrates. It is the statocyst that attracts our attention. This organ helps maintain equilibrium, sensing linear (gravitational) and rotational acceleration (Budelmann, B. U. et al., 1997). The statocyst is essentially a closed cavity that is lined with a complicated arrangement of specialized epithelia dotted with ciliated mechanoreceptors. Statocysts of simpler cephalopods (Nautilus) are completely lined with sensory epithelia and half-filled with a number of calciferous granular statoconia and with endolymph (Neumeister, H. and Budelmann, B. U., 1997). There are two types of mechanoreceptor in the Nautilus statocyst; type A is found in the ventral half, the hemisphere of the statocyst where the statoconia are found. type B mechanoreceptors are found in the dorsal hemisphere, in the region containing the endolymph.
24 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
The mechanoreceptors found in Nautilus are primary sensory neurons, that is, an axon leaves the basal part of the cell, joining with a complicated nerve plexus underneath the statocyst epithelium. A type A mechanoreceptor cell typically carries between 10 and 15 kinocilia, arranged in a single row, which is inclined toward the cell surface. The kinocilia are of the typical 9 2 þ 2 configuration. The microtubules connect to the basal body, which in turn is part of an extensive basal root network. Running parallel to the kinociliary row and right beside each kinocilium is a row of shorter microvilli (Neumeister, H. and Budelmann, B. U., 1997). The rows of kinocilium and microvilli are not centered on the cells, but are polarized, displaced to the periphery on the cell. Not only the bundles but also the basal root network is polarized and points in the same direction as the kinocilium. Two interesting points are that the long axis of the kinociliary row is horizontal to the equator of the statocyst and that the polarization of each type A cell (that is to say the side of the cell to which the row is located) is the same for neighboring cells, suggesting a more global mechanism for polarizing type A mechanoreceptors in Nautilus. The association of these cells with the statoconia suggests a role in the sensation of gravity and other forms of linear acceleration. A type B mechanoreceptor cell has 8–10 kinocilia, and as in the type A, these are placed on one side of the cell. Unlike the type A cells, however, the kinocilia are not arranged in a row but have an almost haphazard arrangement, although it is reported that some seem to have a circular configuration. There are again shorter microvilli on these cells, but they are more numerous than those found in the type A mechanoreceptors. These mechanoreceptors, and their association with the endolymph, are thought to be responsible for sensing angular acceleration (Neumeister, H. and Budelmann, B. U., 1997). The statocysts of Coleoid cephalopods are more complicated than those of Nautilus (and in turn other mollusks). There are two main types, the Octopoda type and a Decapoda type which share some features. Contained in the statocyst is the gravity-sensing organs containing a macula/statolith system that is quite separate from the angular acceleration sensing organ that consists of a crista/cupula system. In the Octopoda type, the crista/cupula system is subdivided into nine segments, arranged so they can detect movements in the three spatial dimensions. The Decapoda type is irregularly shaped, as cartilaginous lobes protruding into the cyst direct the
flow of the endolymph. It has three maculae, with a statolith attached to one and statoconia to the other two, arranged at right angles to each other. The crista/cupula system of the Decapoda is subdivided into four segments, arranged into three dimensions. The mechanosensory cells found in these statocysts resemble the type A mechanoreceptors; however, while Nautilus has between 10 and 15 kinocilia per cell, the Octopoda- and Decapoda-type mechanoreceptors have between 50 and 150 kinocilia per cell (Figure 11). Cells similar to the type B mechanoreceptors in Nautilus are not found in these cephalopods (Budelmann, B. U. et al., 1973). Again all
Figure 11 A schematic drawing of a cephalopod hair cell carrying a large number of kinocilia that are flanked by very much smaller microvilli. The ciliary bundles are displaced to one side of the cell, and the cilia themselves are angled toward the surface of the cell. An original drawing by Johanna Kraus, all rights reserved.
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
kinocilia are of the same length, arranged in a row along the long axis of the cell. All are inclined toward the surface of the cell, showing the same kind of polarization as found in Nautilus. The kinocilia are also associated with microvilli, found in parallel rows either side and next to the kinociliary row. These are not stereovilli as found in vertebrates, and there seem to be no tip links associating the microvilli to the kinocilia. Instead, each kinocilium is associated with the other by membrane junctions (Budelmann, B. U. et al., 1973). The significance of this is still unclear. At this point it is worth mentioning that other mollusks, that is, most gastropods and bivalves, have mechanoreceptors that are similar to the Nautilus type B mechanoreceptor (a random or circular arrangement of kinocilia); however, they are unpolarized. As such, the lack of polarization suggests that they are involved only in gravity sensation (McKee, A. E. and Wiederhold, M. L., 1974; Moir, A. J., 1977). Polarization of the basal feet with respect to the kinocilium is thought to correlate with the deflection of the kinocilia. Deflection in the direction of the basal feet causes maximum depolarization while deflection away from the basal foot causes hyperpolarization (Budelmann, B. U. and Williamson, R., 1994). This is important for detecting the directions of linear and angular acceleration. One important difference between coelid cephalopods and other mollusks is that the mechanoreceptors found in the statocyst are predominantly secondary sensory cells (Stephens, P. R. and Young, J. Z., 1982; Budelmann, B. U. et al., 1987). One exception is in the crista, which is a ridge of cells. In the middle of the ridge are 2 or 4 rows of larger hair cells and 2–4 less regularly arranged hair cells on either side. In the octopuses (the coelids that have been studied) the hair cells on one side of the ridge are primary, while on the other side, they are secondary. These secondary cells are large or small and are synaptically connected to two types of afferent neurons. No other animal has both primary and secondary sensory cells in the same epithelium. The ionic character of the cephalopod hair cell is almost identical to that of vertebrates. This may indicate that a similar mechanoreceptor channel system acts, but as yet the molecular correlates for mechanosensation in mollusks are unknown. We would expect a Trp receptor to form a part of a presumptive spring-gated channel, similar to that found in vertebrates.
25
3.01.5 Molecular Evolution of Mechanoreceptors In the preceding sections we described cellular relatedness between vertebrate and nonvertebrate ciliated mechanoreceptors and we find that these cells share certain characteristics that make them able to respond to mechanical stimuli. However, one important question remains: are ciliated mechanoreceptors in different species monophyletic or the result of convergent evolution? To fully answer this, the developmental program, or ontogeny, of mechanoreceptors from different species needs to be examined. In this section, the agents that transmit this information, the genes, are discussed. Only by understanding the molecular core of mechanoreceptor development can we be informed regarding a monophyletic or polyphyletic origin. In a typical developmental/differentiation paradigm, an external signal – emitted, for example, by an adjacent group of cells – elicits a response in a target cell, a process known as induction. Induction usually results in the activation of a protein that acts on DNA. Such proteins, called transcription factors, bind stretches of DNA, known as cis-regulatory elements, and activate the genes controlled by these regions. These genes form a hierarchical organization of fate choices in the form of a transcription factor cascade. This becomes important when the consensus view of molecular evolution is taken into consideration. At the molecular level, developmental programs evolved as conserved genetic regulatory pathway are subtly modified, usually at the level of cis-regulatory elements and by equally subtle changes in the activity of a particular protein. Thus to extend the presumed cytological homology to the molecular level, an appreciation of the conservation of gene networks, in particular transcription factor cascades, is important. At this point we introduce an important caveat; developmental biology uses a limited number of model systems, comparative biology (by definition) needs far more. More specifically, much is known about the molecular networks that control the development of mechanoreceptors in vertebrates and in Drosophila; however, much less is known about the genetic cascades involved in the development of the other species considered. Therefore, any discussion about molecular comparisons makes large assumptions in the light of a somewhat limited data set. Notwithstanding, by identifying homologous genes from the phyla discussed, it is possible to be
26 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
informed on the validity of these assumptions, and indeed many genes involved in the development of vertebrate or fly mechanoreceptors can be found in cnidarians, nematodes, and cephalopods. 3.01.5.1 Development of Ciliated Mechanoreceptors Ciliated mechanoreceptors are almost exclusively restricted to the ectodermal lineage of all animals. This highlights that the development of the mechanoreceptors is constrained by the molecular history of the cell. Indeed, mechanoreceptor cytodifferentiation can be thought of as the end product of a gradual restriction in cell fate, with each step being represented by a progression in the genetic transcription factor cascade. In vertebrates, the ciliated mechanoreceptors form from a specialized, thickened region of epithelium, the sensory placode (Baker, C. V. and BronnerFraser, M., 2001). For the purpose of this review we discuss the vertebrate inner ear that forms from the otic placode. The otic placode is specified from ectoderm already committed to adopt a sensory fate. This pre-placodal domain is specified by surrounding tissues and refers to a territory of the vertebrate embryo that will give rise to not only the otic placode, but also other sense organs, peripheral neurons of the head and the lens. The pre-placodal domain is marked by families of transcription factors showing homology to a number of genes originally cloned in Drosophila: distalless (dlx) gene, the sine oculis (six) gene, and eyeless (eya) family (Schlosser, G., 2005). In flies these genes also play a role in epithelial patterning and outgrowth; however, it is not clear whether the role of these genes in the fly is to establish a field of competence or whether they participate more directly in the genetic cascade that leads directly to the formation of the mechanoreceptor. 3.01.5.1.1 Extrinsic factors regulating mechanoreceptor development
The early development of mechanoreceptors depends on the action of extrinsic factors. These act on competent ectoderm and are usually signaling molecules acting on cells to direct them to adopt a ciliated mechanoreceptor fate. The extrinsic mechanisms are not very well conserved between different phyla, and indeed there is considerable variability even within vertebrates. Thus their individual descriptions are not so pertinent to this discussion and will not be dealt with in depth. In
vertebrates, a specific area of the pre-placodal region is acted upon by localized signaling interactions that direct the formation of the otic placode within this territory (Groves, A. K. and BronnerFraser, M., 2000; Ladher, R. K. et al., 2000; Kil, S. H. et al., 2005). This early phase of signaling is mediated, in vertebrates, by members of the fibroblast growth factor (FGF) family in combination with the wingless (WNT) family and induces pre-placodal ectoderm to adopt an otic fate (Schlosser, G., 2006). In Drosophila, the role of this early signaling in the development of the mechanoreceptors is not clear; Niwa N. et al. (2004) suggest the existence of a pre-proneural state that depends on the action of decapentaplegic (the fly homolog of Bmp2/4) and wingless (wnt). The induced vertebrate inner ear invaginates to form an otocyst within the mesenchyme of the head. During this process extrinsic signals once again specify different territories within the inner ear. The otocyst is embedded in head mesoderm and flanked by the neural tube and endoderm, tissues that to varying extents exert some influence onto the otocyst (Wu, D. K. and Oh, S. H., 1996; Wu, D. K. et al., 1998; Riccomagno, M. M. et al., 2002; Bok, J. et al., 2005; Riccomagno, M. M. et al., 2005). In amniotes, where the studies have been performed, the hindbrain and the notochord provide the otocyst with dorsal and ventral identities, respectively. Recent reports described the antagonistic effect of Wnt3a, emanating from the hindbrain, and sonic hedgehog, from the notochord, directing development of the vestibular and auditory portions of the otocyst (Riccomagno, M. M. et al., 2002; 2005). As has already been described, the morphology of hair cells is different in the two parts of the inner ear, leading to the speculation that the action of these extrinsic factors has a direct bearing on the differentiation of the hair cell subtypes. Finally, there is a role for extrinsic signals in the differentiation of the hair cells of the cochlea. Montcouquiol M. and Kelley M. W. (2004) showed that without mesenchyme, pattern and differentiation within the cochlea are perturbed. Hair cells form from regions of the otocyst known as sensory patches. The sensory patches express a number of extrinsic factors that have been shown to play a role in mechanoreceptor development. In vertebrates, the sensory patches are specified by Notch and its ligands and by BMP molecules; however, both groups of molecules have additional functions later in development. As already mentioned, in flies, sensory
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
patches are though to be specified by the combined action of decapentaplegic and wingless (Niwa, N. et al., 2004). Subsequent differentiation is triggered by the Notch ligand Delta (Schweisguth, F., 1995). It should be noted that in both vertebrates and insects, the mechanisms that specify exact mechanoreceptor identity do so in parallel to the differentiation of the sensory patches. In vertebrates, Bmp7 is expressed very early in the inner ear (Oh, S. H. et al., 1996), with its close family member Bmp4 appearing soon after. In chickens, Bmp4 is expressed in all the sensory patches, and in mouse and zebrafish it is expressed in most (Oh, S. H. et al., 1996; Morsli, H. et al., 1998; Mowbray, C. et al., 2001). While the overexpression of BMP7 has not been studied, BMP4 has been investigated. These data support a role for BMP4 as a permissive factor, allowing the differentiation of the inner ear hair cells; ectopic application of BMP4 increases the number of hair cells that are found, and application of a BMP4 antagonist decreases their number (Chang, W. et al., 1999; Gerlach, L. M. et al., 2000; Gerlach-Bank, L. M. et al., 2004; Li, H. et al., 2005). However in all cases, the experiments were performed too late to address any role in the formation of the sensory patches themselves. Furthermore, these experiments have only been performed in the chicken; and while BMP molecules are expressed in the inner ear, and more pertinently for us in the sensory patches, their function has not been addressed in other vertebrates. The notch family of signaling molecules comprises the Notch receptors and its ligands, Delta and Jagged/Serrate. Signaling between these molecules occurs between neighboring cells and in the plane of the cell. More detail on the signaling pathway is provided in several excellent reviews (Fekete, D. M. and Wu, D. K., 2002). Notch–DSL (for Delta, Serrate, and Lag-2, the C. elegans homolog of Delta) signaling plays two roles in the development of ciliated mechanosensors; the first is in the specification of the sensory patches, imparting competence onto particular regions of the otocyst so that they can adopt a sensory fate. This property is mediated by the interaction of Notch with Jagged1 (also known as Serrate); Notch overexpression results in the production of supernumerary sensory patches, while knockouts of the Jagged1 gene result in the loss of hair cells (Daudet, N. and Lewis, J., 2005; Brooker, R. et al., 2006). The second role is within the sensory patch itself. It appears that Notch signaling is important in the fate choice between hair cells and the associated
27
accessory cells, the supporting cells. The role of Notch in the binary fate choice between hair cells and supporting cells was demonstrated by work in fish, chicken, and mouse. Loss- and gain-of-function analysis using Notch, its ligands, or its downstream effectors has shown that sensory cells expressing one of the Notch ligands cause the activation of Notch in adjacent cells. Hair cell fate is inhibited in these adjacent cells which then become supporting cells. In contrast to sensory patch specification, this effect, termed lateral inhibition, is mediated by the interaction of Notch with Delta (Brooker, R. et al., 2006). Drosophila mechanoreceptors develop from patches of epidermis known as sensory organ precursors (SOPs). The development of the SOPs themselves is subject to patterning cues acting on the epidermis: dpp (the homolog of Bmp/2/4) and wingless act in concert to specify the SOP, together with ecdysone, the hormone that regulates the timing of metamorphosis of the Drosophila larva (Niwa, N. et al., 2004). In addition, epidermal growth factor (EGF) signaling specifies some of this group of cells (Okabe, M. and Okano, H., 1997). This has parallels with the development of vertebrate hair cells: EGF is required for the survival and proliferation of mouse hair cells (Doetzlhofer, A. et al., 2004). The concerted action of dpp, wingless, and EGF causes a group of cells within the Drosophila ectoderm to become specified as proneural. Lateral inhibition mediated by an interaction between Notch and Delta then specifies one of the proneural cells as the sensory precursor cell (Schweisguth, F., 1995). 3.01.5.1.2 The role of transcription factors in mechanoreceptor development
The action of extrinsic factors, that is, signaling molecules not acting cell-autonomously, has been described in the preceding section. These must provoke an intracellular response that affects the changes wrought on the cell by extracellular events. The intracellular response is invariably mediated by transcription factors that bind to specific DNA sequences, activating particular genes that lead to a particular developmental outcome. The transcription factor cascade involved in mechanoreceptor development is dynamic and the same transcription factors recur within different contexts and with slightly different functions as the mechanoreceptor becomes more and more differentiated. Near the top of the hierarchy of transcription factors regulating mechanoreceptor development is the paired-box homeodomain-containing protein
28 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
Pax2. In vertebrates, Pax2 is one of the first genes expressed in the inner ear primordia. In most phyla, the expression of Pax2, or its ancestral ortholog Pax2/ 5/8, prefigures the development of ciliated mechanoreceptors. The Pax2/5/8 ancestor PaxB can be found in cnidarians (Sun, H. et al., 1997; Gro¨ger, H. et al., 2000; Miller, D. J. et al., 2000). Expression analysis in hydrozoa localized PaxB to the developing sensory neurons of the tentacles (Gro¨ger, H. et al., 2000) while in cubozoans, PaxB can be found in the statocyst (Kozmik, Z. et al., 2003). In Drosophila, the Pax2 homolog is found in mechanoreceptors, and mutation results in a loss of the bristle organs; other mechanoreceptor subtypes were not investigated (Czerny, T. et al., 1997; Kavaler, J. et al., 1999). Mollusk Pax2/5/8 has also been described and is expressed in both the chemosensory and mechanosensory cells (O’Brien, E. K., and Degnan, B. M., 2003). Finally, either Pax2 itself or the Pax2/5/8 gene can be found in deuterostome lineages, for example, in echinoderms (Czerny, T. et al., 1997), cephalochordates (Kozmik, Z. et al., 1999), and vertebrates (Krauss, S. et al., 1991; Rinkwitz-Brandt, S. et al., 1996; Torres, M. et al., 1996; Heller, N. and Brandli, A. W., 1997; Hutson, M. R. et al., 1999; Hidalgo-Sanchez, M. et al., 2000; McCauley, D. W. and Bronner-Fraser, M., 2002). It is tempting to suppose that the Pax2 gene plays a fundamental role in establishing a mechanosensory propensity; however, as with all generalizations, caution must be applied. In vertebrates, Pax2 operates upstream of Notch to specify an area of ectoderm competent to become mechanosensory (Groves, A. K. and Bonner-Fraser, M., 2000). These cells then, through subsequent interactions, become definitively mechanosensory. In Drosophila, however, the role of Pax2 is downstream of Notch and specifies the mechanosensory lineage (Kavaler, J. et al., 1999). However, this may reflect a difference between vertebrates and Drosophila; vertebrates use a placodal sense organ, arranging the ciliated mechanosensory receptors in specialized structures, whereas in Drosophila this is not the case. Analysis of Pax2 mutants in the mouse and zebrafish indicate that the initial development of the inner ear is normal, due to a degree of redundancy with the Pax8 gene, a close relative of Pax2 (Hans, S. et al., 2004). In mice mutant for Pax2, however, the auditory portion of the inner ear fails to form (Favor, J. et al., 1996; Torres, M. et al., 1996). The formation of mechanoreceptors is apparently normal in the rest of the ear, as there are still cells positive for markers of early mechanoreceptors (Burton, Q. et al.,
2004). Zebrafish have two Pax2 genes, and mutations in one of them actually results in an increased number of hair cells by disrupting the expression of Delta (Riley, B. B. et al., 1999). Both sets of data do not emphatically rule out a role for Pax2 in the direct specification of ciliated mechanoreceptors; the mouse study relied on molecular markers for assaying hair cells, without observing the final morphology of the hair cells. One cannot rule out the possibility that the hair cells had merely arrested at an earlier, Pax2-independent step of their development. In the zebrafish study, only one of the two Pax2 genes was mutated. When both Pax2a and Pax2b are mutated, hair cell number is reduced (Whitfield, T. T. et al., 2002). These data are similar to those from Drosophila; when Pax2 is lost, the external sensory organs of the fly are lost (Kavaler, J. et al., 1999). Another transcription factor that plays an important role in the specification of the ciliated mechanoreceptors is the transcription factor Atoh1. Atoh1 is one of the four vertebrate orthologs of the Drosophila gene atonal. Genetic ablation of Atoh1 (formerly known as Math1) blocks the formation of the hair cells without any impact on other cell types (such as supporting cells) or on the overall morphology of the inner ear (Bermingham, N. A. et al., 1999). Mutation of the Drosophila atonal indicates its requirement for differentiation of the scolopidia in the fly; mutants of atonal completely lack these cells as well as some other sensory cell types. The functional similarity between the fly atonal and vertebrate Atoh1 (Atoh1 can rescue a fly atonal mutant and vice versa; Ben-Arie, N. et al., 2000; Wang, V. Y. et al., 2002) lends support to a homology between fly and vertebrate ciliated mechanoreceptors. However, closer inspection reveals important functional differences between Drosophila and vertebrates. These differences, reviewed by Hassan B. A. and Bellen H. J. (2000), highlight important mechanistic changes between the fly and vertebrates. In the fly, atonal functions in cell selection within the SOP and mediates the elaboration of chordotonal lineage identity, that is, proneural and lineage identities are coupled. In mouse, Atoh1 functions in postmitotic cells, directing only terminal differentiation of the hair cell (Chen, P. et al., 2002). Indeed while Atoh1 mutant mice do not have hair cells, the prosensory domain (the sensory patches) are unaffected. In vertebrates, the proneural and lineage identity function of Atoh1 has become uncoupled. This is not because of a structural change in the molecule itself; Atoh1 will rescue flies mutant for atonal, and atonal completely
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
rescues the Atoh1 knockout mouse (Ben-Arie, N. et al., 2000; Wang, V. Y. et al., 2002). Furthermore, transfection of Atoh1 into mammalian postnatal inner ears results in the formation of extra hair cells (Kawamoto, K. et al., 2003; Woods, C. et al., 2004). Seemingly Atoh1 is both necessary and sufficient for the development of the vertebrate ciliate mechanoreceptor. Atoh1 regulation is still unclear and thus is subject to some speculation. An ablation of the Sox2 gene in the otocyst has been shown to also abolish the development of the hair cells (Kiernan, A. E. et al., 2005). Sox2 affects the development of the sensory patches rather than the differentiation of the hair cells themselves and epistatic studies have shown that Sox2 functions upstream of Atoh1 (Kiernan, A. E. et al., 2005). Atoh1 is also under negative regulation from the downstream effectors of notch–DSL signaling, the HES genes, genes that are the vertebrate orthologs of the Drosophila enhancer of split genes (Zine, A. et al., 2001). Again this is not so surprising given the effect Notch has on the determination of the hair cells. In Drosophila, atonal expression is regulated by a cascade of transcription factors. It is regulated by a zinc-finger protein called senseless and there are contributions from genes such as pannier and u-shaped as well as distalless and homothorax (Quan, X. J. et al., 2004). While the vertebrate homolog of senseless, Gfi1, is involved in hair cell function, its relationship with Atoh1 is not clear (Jafar-Nejad, H. and Bellen, H. J., 2004). Gfi1 has itself been mutated in mice and these animals also show hair cell abnormalities (Wallis, D. et al., 2003). In the inner ear, Gfi1 has been identified as a downstream response of the transcription factor, Pou4f3, the causative mutation of the autosomal human hereditary deafness gene DFNA15. The relationship with Atoh1 has not been elucidated, however, and it is still unclear whether Atoh1 functions upstream, downstream, or in parallel with Pou4f3. Hertzano R. et al. (2004) showed that in mice mutant for Pou4f3, Atoh1 expression was at 50% of its normal levels. Gfi1 functions downstream of Atoh1 in the development of secretory cells in the mouse intestine, suggesting that there may also be a similar relationship within the inner ear (Shroyer, N. F. et al., 2005). Of the other genes suggested to play a role in the regulation of Drosophila atonal, homologs for pannier (GATA), u-shaped (FOG), distalless (dlx), and homothorax (Meis) can all be found in vertebrates and the expression can be detected in the inner ear. However, the detailed molecular relationships of these genes with each other and with atonal/Atoh1, either in the fly or in vertebrates, are still to be elucidated.
3.01.5.2
29
Summary of Development
Focusing on mechanoreceptor development in flies and in vertebrates allows important genes to be defined. Orthologs for these genes can also be found in all of the organisms we described. As discussed, Pax2 orthologs can be found in every metazoan phylum including Porifera, the sponges (Hadrys, T. et al., 2005). Atonal, involved in the development of mechanoreceptors in both vertebrates and insects, is also found in cnidarians, nematodes, and mollusks (Ladher, R., Tarui, H., and Agata, K. unpublished data). Notch and Delta, involved in fate choice within the sensory precursors in the fly and in the development of the sensory patches in vertebrates, can also be found in cnidarians, nematodes, and mollusks (Ladher, R., Tarui, H., and Agata, K. unpublished data). The list goes on, with genes involved in mechanoreceptor development in vertebrates being found as far back as cnidarians. This suggests a conservation of the molecular modules that could direct development of the ciliated mechanoreceptors. The could is important though, as in other organisms the involvement of these genes in mechanoreceptor development can only be postulated but has not yet been unequivocally shown. Indeed Fritzsch B. et al. (2006) have suggested that the presence of homologous genes suggests a deep homology describing a basic cellular property (such as the ability to rapidly proliferate), not really describing a derived structure, and that any attempt to describe further homologies is likely to fail. In fact many homologous genes are not exclusively mechanosensory but are found in other contexts during ontogeny. Further resolution of the homology question will require more data, particularly from nonmodel species. At this stage, there is no doubt that ciliated mechanoreceptors share important molecular, structural, and cytological similarities. However, as well as these obvious similarities, ontogenetic differences become very important to characterize, enabling us to understand where and how the ciliated mechanoreceptor has arisen and how evolution has modified the ancestral prototype in the different lineages.
3.01.6 Conclusion There is a considerable amount of information on the structure, function, and development of ciliated mechanosensory cells in vertebrates and invertebrate
30 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells
animals. These provide tantalizing suggestions regarding their possible homology and make it possible to sketch the evolution of these cells through the major lines of multicellular animals. It is, however, still not possible to be sure whether all these cell types, in every organism, are derived from a common ancestor or whether the strong similarities they show are at least in some cases the result of developmental and physiological constraints. As so often in science, more data, especially from the less-studied invertebrates such as mollusks, will resolve this question.
Acknowledgments Work in the authors’ labs is supported by grants from the Deutsche Forschungsgemeinschaft (GAM) and from MEXT administered through RIKEN and the MEXT Leading Projects (RKL).
References Adam, J., Myat, A., LeRoux, I., Eddison, M., Henrique, D., and Lewis, J. 1998. Cell fate choices and the expression of notch, delta and serrate homologs in the chick inner ear: parallel with Drosophila sense-organ development. Development 125, 4645–4654. Adoutte, A., Balavoine, G., Lartillot, N., Lespinet, O., Prud’homme, B., and de Rosa, R. 2000. The new animal phylogeny: reliability and implications. Proc. Natl. Acad. Sci. U. S. A. 97, 4453–4456. Baker, C. V. and Bronner-Fraser, M. 2001. Vertebrate cranial placodes. I. Embryonic induction. Dev. Biol. 232, 1–61. Ben-Arie, N., Hassan, B. A., Bermingham, N. A., Malicki, D. M., Armstrong, D., Matzuk, M., Bellen, H. J., and Zoghbi, H. Y. 2000. Functional conservation of atonal and Math1 in the CNS and PNS. Development 127, 1039–1048. Bennet-Clark, H. C. and Ewing, A. W. 1970. The love song of the fruit fly. Sci. Am. 223, 85–90. Bermingham, N. A., Hassan, B. A., Price, S. D., Vollrath, M. A., Ben-Arie, N., Eatock, R. A., Bellen, H. J., Lysakowski, A., and Zoghbi, H. Y. 1999. Math1: an essential gene for the generation of inner ear hair cells. Science 284, 1837–1841. Bleckmann, H. 1986. The Role of the Lateral Line in Fish Behaviour. In: The Behaviour of Teleost Fishes (ed. T. J. Pitcher), pp. 178–202. Sydney. Boekhoff-Falk, G. 2005. Hearing in Drosophila: development of Johnston’s organ and emerging parallels to vertebrate ear development. Dev. Dyn. 232, 550–558. Bok, J., Bronner-Fraser, M., and Wu, D. K. 2005. Role of the hindbrain in dorsoventral but not anteroposterior axial specification of the inner ear. Development 132, 2115–24. Bone, Q. and Ryan, K. P. 1978. Cupular sense organs in Ciona (Tunicata: Ascidiacea). J. Zool. (Lond.) 186, 417–429. Brigandt, I. 2003. Homology in comparative, molecular, and evolutionary developmental biology: the radiation of a concept. J. Exp. Zool. B. Mol. Dev. Evol. 299, 9–17. Brooker, R., Hozumi, K., and Lewis, J. 2006. Notch ligands with contrasting functions: Jagged1 and Delta1 in the mouse inner ear. Development. 133, 1277–86.
Budelmann, B. U. 1992. Hearing in Nonarthropod Invertebrates. In: The Evolutionary Biology of Hearing (eds. R. R. Fay, A. N. Popper, and D. B. Webster), pp. 141–155. Springer. Budelmann, B. U. and Williamson, R. 1994. Directional sensitivity of hair cell afferents in the Octopus statocyst. J. Exp. Biol. 187, 245–259. Budelmann, B. U., Barber, V. C., and West, S. 1973. Scanning electron microscopical studies of the arrangements and numbers of hair cells in the statocysts of Octopus vulgaris, Sepia officinalis and Loligo vulgaris. Brain Res. 56, 25–41. Budelmann, B. U., Sachie, M., and Staudigl, M. 1987. The angular acceleration receptor system of the statocyst of Octopus vulgaris. Philos. Trans. R. Soc. Lond. B. 315, 305–343. Budelmann, B. U., Schipp, R., and von Boletzky, S. 1997. Cephalopoda. In: Mollusca II (eds. F. W. Harrison and A. J. Kohn), pp. 119–414. Wiley-Liss. Burighel, P., Lane, N. J., Fabio, G., Stefano, T., Zaniolo, G., Carnevali, M. D. C., and Manni, L. 2003. Novel, secondary sensory cell organ in ascidians: in search of the ancestor of the vertebrate lateral line. J. Comp. Neurol. 461, 236–249. Burton, Q., Cole, L. K., Mulheisen, M., Chang, W., and Wu, D. K. 2004. The role of Pax2 in mouse inner ear development. Dev. Biol. 272, 161–75. Carroll, R. L. 1987. Vertebrate Palaeontology and Evolution. Freeman. Chalfie, M. and Thomson, J. N. 1982. Structural and functional diversity in the neuronal microtubules of Caenorhabditis elegans. J. Cell Biol. 93, 15–23. Chalfie, M., Sulston, J. E., White, J. G., Southgate, E., Thomson, J. N., and Brenner, S. 1985. The neural circuit for touch sensitivity in Caenorhabditis elegans. J. Neurosci. 5, 956–964. Chang, W., Nunes, F. D., De Jesus-Escobar, J. M., Harland, R., and Wu, D. K. 1999. Ectopic noggin blocks sensory and nonsensory organ morphogenesis in the chicken inner ear. Dev. Biol. 216, 369–381. Chen, P., Johnson, J. E., Zoghbi, H. Y., and Segil, N. 2002. The role of Math1 in inner ear development: uncoupling the establishment of the sensory primordium from hair cell fate determination. Development 129, 2495–2505. Coffin, A., Kelley, M., Manley, G. A., and Popper, A. N. 2004. Evolution of Sensory Hair Cells. In: Evolution of the Vertebrate Auditory System. (eds. G. A. Manley, A. Popper, and R. R. Fay), pp. 55–94. Springer. Corey, D. P., Garcia-An˜overos, J., Holt, J. R., Kwan, K. Y., Lin, S.-Y., Vollrath, M. A., Amalfitano, A., Cheung, E. L.-M., Derfler, B. H., Duggan, A., Ge´le´oc, G. S. G., Gray, P. A., Hoffman, M. P., Rehm, H. L., Tamasauskas, D., and Zhang, D.-S. 2004. TRPA1 is a candidate for the mechanosensitive transduction channel of vertebrate hair cells. Nature 432, 723–730. Corfas, G. and Dudai, Y. 1990. Adaptation and fatigue of a mechanosensory neuron in wild-type Drosophila and in memory mutants. J. Neurosci. 10, 491–499. Czerny, T., Bouchard, M., Kozmik, Z., and Busslinger, M. 1997. The characterization of novel Pax genes of the sea urchin and Drosophila reveal an ancient evolutionary origin of the Pax2/5/8 subfamily. Mech. Dev. 67, 179–92. Daudet, N. and Lewis, J. 2005. Two contrasting roles for Notch activity in chick inner ear development: specification of prosensory patches and lateral inhibition of hair-cell differentiation. Development 132, 541–551. Denison, R. H. 1966. The origin of the lateral-line sensory system. Am. Zool. 6, 369–370. Devau, G. 2000. Glycine-induced calcium concentration changes in vestibular type I sensory cells. Hear. Res. 140, 126–136.
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells Doetzlhofer, A., White, P. M., Johnson, J. E., Segil, N., and Groves, A. K. 2004. In vitro growth and differentiation of mammalian sensory hair cell progenitors: a requirement for EGF and periotic mesenchyme. Dev. Biol. 272, 432–447. Eatock, R. A., Weiss, T. F., and Otto, K. L. 1991. Dependence of discharge rate on sound pressure level in cochlear nerve fibers of the alligator lizard: implications for cochlear mechanisms. J. Neurophysiol. 65, 1580–97. Eberl, D. F., Duyk, G. M., and Perrimon, N. 1997. A genetic screen for mutations that disrupt an auditory response in Drosophila melanogaster. Proc. Natl. Acad. Sci. U. S. A. 94, 14837–14842. Ekstro¨m von Lubitz, D. K. J. 1981. Ultrastructure of the lateralline sense organs of the ratfish, Chimaera monstrosa. Cell Tissue Res. 215, 651–665. Ernstrom, G. G. and Chalfie, M. 2002. Genetics of sensory mechanotransduction. Annu. Rev. Genet. 36, 411–453. Favor, J., Sandulache, R., Neuhauser-Klaus, A., Pretsch, W., Chatterjee, B., Senft, E., Wurst, W., Blanquet, V., Grimes, P., Sporle, R., and 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, 13870–13875. Fay, R. R. and Edds-Walton, P. L. 1997. Directional response properties of saccular afferents of the toadfish, Opsanus tau. Hear. Res. 111, 1–21. Fekete, D. M. and Wu, D. K. 2002. Revisiting cell fate specification in the inner ear. Curr. Opin. Neurobiol. 12, 35–42. Fettiplace, R. 1987. Electrical tuning of hair cells in the inner ear. Trends Neurosci. 10, 421–425. Fischer, F. P. 1994. General pattern and morphological specializations of the avian cochlea. Scanning Microsc. 8, 351–364. Flock, A. 1967. Ultrastructure and Function in the Lateral Line Organs. In: Lateral Line Detectors, (ed. P. Cahn), pp. 163–197. Indiana University Press. Fritzsch, B. 1992. The Water-to-Land Transition: Evolution of the Tetrapod Basilar Papilla, Middle Ear, and Auditory Nuclei. In: The Evolutionary Biology of Hearing. (eds. D. B. Webster, R. R. Fay, and A. N. Popper), pp. 351–375. Springer. Fritzsch, B., Beisel, K. W., Jones, K., Farinas, I., Maklad, A., Lee, J., and Reichardt L. F. 2002. Development and evolution of inner ear sensory epithelia and their innervation. J. Neurobiol. 53, 143–56. Fritzsch, B., Pauley, S., and Beisel, K. W. 2006. Cells, molecules and morphogenesis: the making of the vertebrate ear. Brain Res. Epub ahead of print, PMID: 16643865. Ge´le´oc, G. S. and Holt, J. R. 2003. Auditory amplification: outer hair cells pres the issue. Trends Neurosci. 26, 115–117. Gerlach, L. M., Hutson, M. R., Germiller, J. A., Nguyen-Luu, D., Victor, J. C., and Barald, K. F. 2000. Addition of the BMP4 antagonist, noggin, disrupts avian inner ear development. Development 127, 45–54. Gerlach-Bank, L. M., Cleveland, A. R., and Barald, K. F. 2004. DAN directs endolymphatic sac and duct outgrowth in the avian inner ear. Dev. Dyn. 229, 219–230. Gibbs, M. A. 2004. Lateral line receptors: where do they come from developmentally and where is our research going? Brain Behav. Evol. 64, 163–181. Gillespie, P. G. and Walker, R. G. 2001. Molecular basis of mechanosensory transduction. Nature 413, 194–202. Gillespie, P. G., Dumont, R. A., and Kachar, B. 2005. Have we found the tip link, transduction channel, and gating spring of the hair cell? Curr. Opin. Neurobiol. 15, 389–396. Gleich, O., Fischer, F. P., Ko¨ppl, C., and Manley, G. A. 2004. Hearing Organ Evolution and Specialization: Archosaurs. In: Evolution of the Vertebrate Auditory System,
31
(eds. G. A. Manley, A. Popper, and R. R. Fay), pp. 224–255. Springer. Gong, Z., Son, W., Chung, Y. D., Kim, J., Shin, D. W., McClung, C. A., Lee, Y., Lee, H. W., Chang, D. J., Kaang, B. K., Cho, H., Oh, U., Hirsh, J., Kernan, M. J., and Kim, C. 2004. Two interdependent TRPV channel subunits, inactive and Nanchung, mediate hearing in Drosophila. J. Neurosci. 24, 9059–9066. Go¨pfert, M. C. and Wasserthal, L. T. 1999. Auditory sensory cells in hawkmoths: identification, physiology and structure. J. Exp. Biol. 202, 1579–1587. Go¨rner, P. 1987. Lateral Line System. In: Encyclopedia of Neuroscience, I (ed. G. Adelman), pp. 567–569. Birkha¨user. Gro¨ger, H., Callaerts, P., Gehring, W. J., and Schmid, V. 2000. Characterization and expression analysis of an ancestortype Pax gene in the hydrozoan jellyfish Podocoryne carnea. Mech. Dev. 94, pp. 157–169. Groves, A. K. and Bronner-Fraser, M. 2000. Competence, specification and commitment in otic placode induction. Development 127, 3489–3499. Gru¨nert, U. and Gnatzy, W. 1987. Kþ and Caþþ in the receptor lymph of arthropod cuticular mechanoreceptors. J. Comp. Physiol. A. 161, 329–333. Hadrys, T., DeSalle, R., Sagasser, S., Fischer, N., and Schierwater, B. 2005. The Trichoplax PaxB gene: a putative Proto-PaxA/B/C gene predating the origin of nerve and sensory cells. Mol. Biol. Evol. 22, 1569–1578. Hallberg, E. and Hansson, B. S. 1999. Arthropod sensilla: morphology and phylogenetic considerations. Microsc. Res. Tech. 47, 428–439. Hans, S., Liu, D., and Westerfield, M. 2004. Pax8 and Pax2a function synergistically in otic specification, downstream of the Foxi1 and Dlx3b transcription factors. Development 131, 5091–102. Hassan, B. A. and Bellen, H. J. 2000. Doing the MATH: is the mouse a good model for fly development? Genes Dev. 14, 1852–1865. Hasson, T., Heintzelman, M. B., Santos-Sacchi, J., Corey, D. P., and Mooseker, M. S. 1995. Expression in cochlea and retina of myosin VIIa, the gene product defective in Usher syndrome type 1B. Proc. Natl. Acad. Sci. U. S. A. 92, 9815–9819. Heller, N. and Brandli, A. W. 1997. Xenopus Pax-2 displays multiple splice forms during embryogenesis and pronephric kidney development. Mech. Dev. 69, 83–104. Hennig, W. 1983. Stammesgeschichte der Chordaten. Verlag Paul Parey. Hertzano, R., Montcouquiol, M., Rashi-Elkeles, S., Elkon, R., Yucel, R., Frankel, W. N., Rechavi, G., Moroy, T., Friedman, T. B., Kelley, M. W., and Avraham, K. B. 2004. Transcription profiling of inner ears from Pou4f3(ddl/ddl) identifies Gfi1 as a target of the Pou4f3 deafness gene. Hum. Mol. Genet. 13, 2143–2153. Hidalgo-Sanchez, M., Alvarado-Mallart, R., and Alvarez, I. S. 2000. Pax2, Otx2, Gbx2 and Fgf8 expression in early otic vesicle development. Mech. Dev. 95, 225–229. Hillis, D. M. 1994. Homology in Molecular Biology. In: Homology: The Hierarchical Basis of Comparative Biology (ed. B. K. Hall), pp. 339–368. Academic Press. Holland, L. Z. 2005. Non-neural ectoderm is really neural: evolution of developmental patterning mechanisms in the non-neural ectoderm of chordates and the problem of sensory cell homologies. J. Exp. Zool. B. Mol. Dev. Evol. 304, 304–323. Holstein, T. and Hausmann, K. 1988. The Cnidocil Apparatus of Hydrozoans: A Progenitor of Higher Metazoan Mechanoreceptors. In: The Biology of Nematocysts (ed. D. A. Hessinger and H. M. Lenhoff), pp. 53–73. Academic Press.
32 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells Horridge, G. A. 1969. Statocysts of medusae and evolution of stereocilia. Tissue Cell 1, 341–353. Hudspeth, A. J. 1985. The cellular basis of hearing: the biophysics of hair cells. Science 230, 745–752. Hudspeth, A. J. 1989. How the ear’s works work. Nature 341, 397–404. Hudspeth, A. J. and Corey, D. P. 1977. Sensitivity, polarity, and conductance change in the response of vertebrate hair cells to controlled mechanical stimuli. Proc. Natl. Acad. Sci. U. S. A. 74, 2407–2411. Hutson, M. R, Lewis, J. E, Nguyen-Luu, D., Lindberg, K. H., and Barald, K. F. 1999. Expression of Pax2 and patterning of the chick inner ear. J. Neurocytol. 28, 795–807. Jafar-Nejad, H. and Bellen, H. J. 2004. Gfi/Pag-3/senseless zinc finger proteins: a unifying theme? Mol. Cell Biol. 24: 8803–8812. Jørgensen, J. 1989. Evolution of Octavolateralis Sensory Cells. In: The Mechanosensory Lateral Line: Neurobiology and Evolution (eds. S. Coombs, P. Go¨rner, and H. Mu¨nz), pp. 115–145. Springer. Kaplan, J. M. and Horvitz, H. R. 1993. A dual mechanosensory and chemosensory neuron in Caenorhabditis elegans. Proc. Natl. Acad. Sci. U. S. A. 90, 2227–2231. Kavaler, J., Fu, W, Duan, H., Noll, M., and Posakony, J. W. 1999. An essential role for the Drosophila Pax2 homolog in the differentiation of adult sensory organs. Development. 126, 2261–72. Kawamoto, K., Ishimoto, S., Minoda, R., Brough, D. E., and Raphael, Y. 2003. Math1 gene transfer generates new cochlear hair cells in mature guinea pigs in vivo. J. Neurosci. 23, 4395–4400. Keil, T. A. 1997. Functional morphology of insect mechanoreceptors. Microsc. Res. Tech. 39, 506–531. Kernan, M., Cowan, D., and Zuker, C. 1994. Genetic dissection of mechanosensory transduction: mechanoreceptiondefective mutations of Drosophila. Neuron 12, 1195–1206. Kiernan, A. E., Pelling, A. L., Leung, K. K., Tang, A. S., Bell, D. M., Tease, C., Lovell-Badge, R., Steel, K. P., and Cheah, K. S. 2005. Sox2 is required for sensory organ development in the mammalian inner ear. Nature 434, 1031–1035. Kil, S. H., Streit, A., Brown, S. T., Agrawal, N., Collazo, A., Zile, M. H., and Groves, A. K. 2005. Distinct roles for hindbrain and paraxial mesoderm in the induction and patterning of the inner ear revealed by a study of vitamin-Adeficient quail. Dev. Biol. 285, 252–271. Kim, J., Chung, Y. D., Park, D. Y., Choi, S., Shin, D. W., Soh, H., Lee, H. W., Son, W., Yim, J., Park, C. S., Kernan, M. J., and Kim, C. 2003. A TRPV family ion channel required for hearing in Drosophila.. Nature 424, 81–84. Ko¨ppl, C. and Manley, G. A. 1990. Peripheral auditory processing in the bobtail lizard Tiliqua rugosa. III. Patterns of spontaneous and tone-evoked nerve-fibre activity. J. Comp. Physiol. A. 167, 113–127. Ko¨ppl, C., Manley, G. A., and Konishi, M. 2000. Auditory processing in birds. Curr. Opin. Neurobiol. 10, 474–481. Kozmik, Z., Daube, M., Frei, E., Norman, B., Kos, L., Dishaw, L. J., Noll, M., and Piatigorsky, J. 2003. Role of Pax genes in eye evolution: a cnidarian PaxB gene uniting Pax2 and Pax6 functions. Dev. Cell 5, 773–85. Kozmik, Z., Holland, N. D., Kalousova, A., Paces, J., Schubert, M., and Holland, L. Z. 1999. Charcterization of an amphioxus paired box gene, Amphi Pax 2/5/8: developmental expression patterns in optic support cells, nephridium, thyroid-like structures and pharyngeal gill slits, but not in the midbrain-hindbrain boundary region. Development 126, 1295–1304. Krauss, S., Johansen, T., Korzh, V., and Fjose, A. 1991. Expression pattern of zebrafish pax genes suggests a role in early brain regionalization. Nature 353, 267–270.
Kwan, K. Y., Allchorne, A. J., Vollrath, M. A., Christensen, A. P., Zhang, D. S., Woolf, C. J., and Corey, D. P. 2006. TRPA1 contributes to cold, mechanical, and chemical nociception but is not essential for hair-cell transduction. Neuron. 50, 277–89. Lacalli, T. C. 2004. Sensory systems in amphioxus: a window on the ancestral chordate condition. Brain Behav. Evol. 64, 148–162. Ladher, R. K., Anakwe, K. U., Gurney, A. L., Schoenwolf, G. C., and Francis-West, P. H. 2000. Identification of synergistic signals initiating inner ear development. Science 290, 1965–1967. Ladich, F. and Popper, A. N. 2004. Parallel Evolution in Fish Hearing Organs. In: Evolution of the Vertebrate Auditory System (eds. G. A. Manley, A. Popper, and R. R. Fay), pp. 95–127. Springer. Lawrence, P. A. 1966. Development and determination of hairs and bristles in the milkweed bug, Oncopeltus fasciatus (Lygaeidae, Hemiptera). J. Cell Sci. 1, 475–498. Lewis, E. R., Leverenz, E. L., and Bialek, W. S. 1985. The Vertebrate Inner Ear. CRC Press. Li, H., Corrales, C. E., Wang, Z., Zhao, Y., Wang, Y., Liu, H., and Heller, S. 2005. BMP4 signaling is involved in the generation of inner ear sensory epithelia. BMC Dev. Biol. 5, 16. Lu, Z. and Popper, A. N. 2001. Neural response directionality correlates of hair cell orientation in a teleost fish. J. Comp. Physiol. A. 187, 453–465. Manley, G. A. 1990. Peripheral Hearing Mechanisms in Reptiles and Birds. Springer. Manley, G. A. 1995. The Avian Hearing Organ: A Status Report. In: Advances in Hearing Research (eds. G. A. Manley, G. M. Klump, C. Ko¨ppl, H. Fastl, and H. Oeckinghaus), pp. 219–229. World Scientific Publishers. Manley, G. A. 2000. The Hearing Organs of Lizards. In: Comparative Hearing: Birds and Reptiles (eds. R. Dooling, A. N. Popper, and R. R. Fay), pp. 139–196. Springer Handbook of Auditory Research. Manley, G. A. 2001. Evidence for an active process and a cochlear amplifier in non-mammals. J. Neurophysiol. 86, 541–549. Manley, G. A. 2002. Evolution of structure and function of the hearing organ of lizards. J. Neurobiol. 53, 202–211. Manley, G. A. 2004. The Lizard Basilar Papilla and Its Evolution. In: Evolution of the Vertebrate Auditory System (eds. G. A. Manley, A. Popper, and R. R. Fay), pp. 200–223. Springer. Manley, G. A., Ko¨ppl, C., and Sneary, M. 1999. Reversed tonotopic map of the basilar papilla in Gekko gecko. Hear. Res. 131, 107–116. Manley, G. A., Ko¨ppl, C., and Yates, G. K. 1989. Micromechanical Basis of High-Frequency Tuning in the Bobtail Lizard. In: Mechanics of Hearing, (eds. J. P. Wilson and D. Kemp), pp. 143–150. Plenum. Manley, G. A., Yates, G. K., Ko¨ppl, C., and Johnstone, B. M. 1990. Peripheral auditory processing in the bobtail lizard Tiliqua rugosa. IV. Phase locking of auditory-nerve fibres. J. Comp. Physiol. A. 167, 129–138. Manni, L., Caicci, F., Gasparini, F., Zaniolo, G., and Burighel, P. 2004. Hair cells in ascidians and the evolution of lateral line placodes. Evol. Dev. 6, 379–381. Masetto, S. and Carreia, M. J. 1997. Ionic currents in regenerating avian vestibular hair cells. Int. J. Dev. Neurosci. 15, 387–399. McCauley, D. W. and Bronner-Fraser, M. 2002. Conservation of Pax gene expression in ectodermal placodes of the lamprey. Gene 287, 129–139. McKee, A. E. and Wiederhold, M. L. 1974. Aplysia statocyst receptor cells: fine structure. Brain Res. 81, 310–313. Meenderink, S. W., Narins, P. M., and van Dijk, P. 2005. Detailed f(1) f(2) area study of distortion product otoacoustic
Phylogeny and Evolution of Ciliated Mechanoreceptor Cells emissions in the frog. J. Assoc. Res. Otolaryngol. 6, 3713–3726. Miller, M. R. 1978. Scanning electron microscope studies of the papilla basilaris of some turtles and snakes. Am. J. Anat. 151, 409–435. Miller, M. R. 1992. The Evolutionary Implications of the Structural Variations in the Auditory Papilla of Lizards. In: The Evolutionary Biology of Hearing (eds. R. R. Fay, A. N. Popper, and D. B. Webster), pp. 463–487. Springer. Miller, D. J., Hayward, D. C., Reece-Hoyes, J. S., Scholten, I., Catmull, J., Gehring, W. J., Callaerts, P., Larsen, J. E., and Ball, E. E. 2000. Pax gene diversity in the basal cnidarian Acropora millepora (Cnidaria, Anthozoa): implications for the evolution of the Pax gene family. Proc. Natl. Acad. Sci. U. S. A. 97, 4475–4480. Mire, P. and Watson, G. M. 1997. Mechanotransduction of hair bundles arising from multicellular complexes in anemones. Hear. Res. 113, 224–234. Moir, A. J. 1977. Ultrastructural studies on the ciliated receptors of the long tentacles of the giant scallop, Placopecten magellanicus (gmelin). Cell. Tissue. Res. 184, 367–380. Montcouquiol, M. and Kelley, M. W. 2004. Planar and vertical signals control cellular differentiation and patterning in the mammalian cochlea. J. Neurosci. 23, 9469–78. Montell, C. 2005. The TRP superfamily of cation channels. Science STKE 2005, re3. Morsli, H., Choo, D., Ryan, A., Johnson, R., and Wu, D. K. 1998. Development of the mouse inner ear and origin of its sensory organs. J. Neurosci. 18, 3327–3335. Mowbray, C., Hammerschmidt, M., and Whitfield, T. T. 2001. Expression of BMP signalling pathway members in the developing zebrafish inner ear and lateral line. Mech. Dev. 108, 179–184. Neumeister, H. and Budelmann, B. U. 1997. Structure and function of the Nautilus statocyst. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 352, 1565–1588. Niwa, N., Hiromi, Y., and Okabe, M. 2004. A conserved developmental program for sensory organ formation in Drosophila melanogaster. Nat. Genet. 36, 293–297. O’Brien, E. K. and Degnan, B. M. 2003. Expression of Pax258 in the gastropod statocyst: insights into the antiquity of metazoan geosensory organs. Evol. Dev. 5, 572–578. Oh, S. H., Johnson, R., and Wu, D. K. 1996. Differential expression of bone morphogenetic proteins in the developing vestibular and auditory sensory organs. J. Neurosci. 16, 6463–6475. Okabe, M. and Okano, H. 1997. Two-step induction of chordotonal organ precursors in Drosophila embryogenesis. Development 124, 1045–1053. Patterson, C. 1988. Homology in classical and molecular biology. Mol. Ecol. Evol. 5, 603–625. Perkins, L. A., Hedgecock, E. M., Thomson, J. N., and Culotti, J. G. 1986. Mutant sensory cilia in the nematode Caenorhabditis elegans. Dev. Biol. 117, 456–487. Peteya, D. J. 1975. The ciliary-cone sensory cell of anemones and cerianthids. Tissue Cell 7, 243–252. Pichon, F. and Ghysen, A. 2004. Evolution of posterior latera line development in fish and amphibians. Evol. Dev. 6, 187–193. Pickles, J. O., Comis, S. D., and Osborne, M. P. 1984. Crosslinks between stereocilia in the guinea pig organ of Corti and their possible relation to sensory transduction. Hear. Res. 15, 103–112. Popper, A. N. and Fay, R. R. 1999. The Auditory Periphery in Fishes. In: Comparative Hearing: Fish and Amphibians (eds. R. R. Fay and A. N. Popper), pp. 43–100. Springer. Quan, X. J., Denayer, T., Yan, J., Jafar-Nejad, H., Philippi, A., Lichtarge, O., Vleminckx, K., and Hassan, B. A. 2004. Evolution of neural precursor selection: functional divergence of proneural proteins. Development 131, 1679–1689.
33
Remane, A. 1952. Die Grundlagen des natu¨rlichen Systems, der vergleichenden Anatomie und der Phylogenetik. I. Theoretische Morphologie und Systematik. Geest und Portig. Riccomagno, M. M., Martinu, L., Mulheisen, M., Wu, D. K., and Epstein, D. J. 2002. Specification of the mammalian cochlea is dependent on Sonic hedgehog. Genes Dev. 16, 2365–2378. Riccomagno, M. M., Takada, S., and Epstein, D. J. 2005. Wntdependent regulation of inner ear morphogenesis is balanced by the opposing and supporting roles of Shh. Genes Dev. 19, 1612–1623. Riley, B. B., Chiang, M., Farmer, L., and Heck, R. 1999. The deltaA gene of zebrafish mediates lateral inhibition of hair cells in the inner ear and is regulated by pax2.1. Development. 126, 5669–5678. Rinkwitz-Brandt, S., Arnold, H. H., and Bober, E. 1996. Regionalized expression of Nkx5-1, Nkx5-2, Pax2 and sek genes during mouse inner ear development. Hear. Res. 99, 129–138. Rosenblatt, K. P., Sun, Z.-P., Heller, S., and Hudspeth, A. J. 1997. Distribution of Ca2þ-activated Kþ channel isoforms along the tonotopic gradient of the chicken’s cochlea. Neuron, 19, 1061–1075. Ruppert, E. E., Fox, R. S., and Barnes, R. D. 2004. Invertebrate zoology: a functional evolutionary approach. Belmont. Russell, I. J. 1976. Amphibian Lateral Line Receptors. In: Frog Neurobiology (eds. R. Llinas and W. Precht), pp. 513–550. Springer. Sand, O. 1984. Lateral Line Systems. In: Comparative Physiology of Sensory Systems, (eds. L. Bolish, R. Keynes, and S. H. P. Madrell), pp. 3–32. Cambridge University Press. Sanyal, S., Wintle, R. F., Kindt, K. S., Nuttley, W. M., Arvan, R., Fitzmaurice, P., Bigras, E., Merz, D. C., Hebert, T. E., van der Kooy, D., Schafer, W. R., Culotti, J. G., and Van Tol, H. H. 2004. Dopamine modulates the plasticity of mechanosensory responses in Caenorhabditis elegans. EMBO J. 23, 473–482. Sawin, E. R., Ranganathan, R., and Horvitz, H. R. 2000. C. elegans locomotory rate is modulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron 26, 619–631. Schlosser, G. 2005. Evolutionary origins of vertebrate placodes: insights from developmental studies and from comparisons with other deuterostomes. J. Exp. Zool. B. Mol. Dev. Evol. 304, 347–399. Schlosser, G. 2006. Induction and specification of cranial placodes. Dev. Biol. (in press). Schweisguth, F. 1995. Suppressor of hairless is required for signal reception during lateral inhibition in the Drosophila pupal notum. Development 121, 1875–1884. Shimeld, S. M. and Holland, P. W. H. 2000. Vertebrate innovations. Proc. Nat. Acad. Sci. 97, 4449–4452. Shin, J. B., Adams, D., Paukert, M., Siba, M., Sidi, S., Levin, M., Gillespie, P. G., and Grunder, S. 2005. Xenopus TRPN1 (NOMPC) localizes to microtubule-based cilia in epithelial cells, including inner-ear hair cells. Proc. Natl. Acad. Sci. U. S. A. 102, 12572–12577. Shroyer, N. F., Wallis, D., Venken, K. J., Bellen, H. J., and Zoghbi, H. Y. 2005. Gfi1 functions downstream of Math1 to control intestinal secretory cell subtype allocation and differentiation. Genes Dev. 19, 2412–2417. Sidi, S., Friedrich, R. W., and Nicolson, T. 2003. NompC TRP channel required for vertebrate sensory hair cell mechanotransduction. Science 301, 96–99. Singla, C. L. 1975. Statocysts of hydromedusae. Cell Tissue Res. 158, 391–407. Singla, C. L. 1983. Fine structure of the sensory receptors of Aglantha digitale (Hydromedusae: Trachylina). Cell Tissue Res. 231, 415–425.
34 Phylogeny and Evolution of Ciliated Mechanoreceptor Cells Smith, H. M. 1960. Evolution of Chordate Structure. Reinhart and Winston. Smotherman, M. and Narins, P. 2004. Evolution of the Amphibian Ear. In: Evolution of the Vertebrate Auditory System (eds. G. A. Manley, A. Popper, and R. R. Fay), pp. 164–199. Springer. Song, J., Yan, H. Y., and Popper, A. N. 1995. Damage and recovery of hair cells in fish canal (but not superficial) neuromasts after gentamicin exposure. Hear. Res. 91, 63–71. Spoendlin, H. 1978. The Afferent Innervation of the Cochlea. In: Evoked Electrical Activity in the Auditory Nervous System (eds. R. F. Naunton and C. Fernandez), pp. 3–19. Academic Press. Stephens, P. R. and Young, J. Z. 1982. The statocyst of the squid, Loligo. J. Zool. 197, 241–266. Stidwill, R. P., Honegger, T. G., and Tardent, P. 1988. Polymerized Actin in the Apical Region of Hydra Nematocytes. In: The Biology of Nematocysts (eds. D. A. Hessinger and H. M. Lenhoff), pp. 567–574. Academic Press. Stumpner, A. and von Helversen, D. 2001. Evolution and function of auditory systems in insects. Naturwissenschaften 88, 159–170. Sun, H., Rodin, A., Zhou, Y., Dickinson, D. P., Harper, D. E., Hewett-Emmett, D., and Li, W.-H. 1997. Evolution of paired domains: isolation and sequencing of jellyfish and hydra Pax genes related to Pax-5 and Pax-6. Proc. Natl. Acad. Sci. U. S. A. 94, 5156–5161. Thurm, U. 2001. Mechanosensorik. In: Neurowissenschaften (eds. J. Dudel, R. Menzel, and R. F. Schmidt), pp. 333–353. Springer. Thurm, U. and Ku¨ppers, J. 1980. Epithelial Physiology of Insect Sensilla. In: Insect Biology in the Future (eds. M. Locke and D. S. Smith), pp. 735–758. Academic Press. Todi, S. V., Franke, J. D., Kiehart, D. P., and Eberl, D. F. 2005. Myosin VIIA defects, which underlie the Usher 1B syndrome in humans, lead to deafness in Drosophila. Curr. Biol. 15, 862–868. Todi, S. V., Sharma, Y., and Eberl, D. F. 2004. Anatomical and molecular design of the Drosophila antenna as a flagellar auditory organ. Microsc. Res. Tech. 63, 388–399. Torres, M., Gomez-Pardo, E., and Gruss, P. 1996. Pax2 contributes to inner ear patterning and optic nerve trajectory. Development. 122, 3381–91. van Staaden, M. J. and Ro¨mer, H. 1998. Evolutionary transition from stretch to hearing organs in ancient grasshoppers. Nature 394, 773–776. Vater, M., Meng, J., and Fox, R. C. 2004. Hearing Organ Evolution and Specialization: Early and Later Mammals. In: Evolution of the Vertebrate Auditory System (eds. G. A. Manley, A. Popper, and R. R. Fay), pp. 256–288. Springer. Vinnikov, Y. A. 1982. Evolution of receptor cells. Cytological, membranous and molecular levels. Mol. Biol. Biochem. Biophys. 34, 1–141. von Schilcher, F. 1976. The role of auditory stimuli in the courtship of Drosophila melanogaster. Anim. Behav. 24, 18–26. Wake, D. B. 1994. Comparative terminology. Science 265, 268–269. Walker, R. G., Willingham, A. T., and Zuker, C. S. 2000. A Drosophila mechanosensory transduction channel. Science 287, 2229–2234.
Wallis, D., Hamblen, M., Zhou, Y., Venken, K. J., Schumacher, A., Grimes, H. L., Zoghbi, H. Y., Orkin, S. H., and Bellen, H. J. 2003. The zinc finger transcription factor Gfi1, implicated in lymphomagenesis, is required for inner ear hair cell differentiation and survival. Development 130, 221–232. Wang, V. Y., Hassan, B. A., Bellen, H. J., and Zoghbi, H. Y. 2002. Drosophila atonal fully rescues the phenotype of Math1 null mice: new functions evolve in new cellular contexts. Curr. Biol. 12, 1611–1616. Watson, G. M. and Mire, P. 1999. A comparison of hair bundle mechanoreceptors in sea anemones and vertebrate systems. Curr. Top. Dev. Biol. 43, 51–84. Watson, G. M., Mire, P., and Hudson, R. R. 1997. Hair bundles of sea anemones as a model system for vertebrate hair bundles. Hear. Res. 107, 53–66. Weber, T., Gopfert, M. C., Winter, H., Zimmermann, U., Kohler, H., Meier, A., Hendrich, O., Rohbock, K., Robert, D., and Knipper, M. 2003. Expression of prestin-homologous solute carrier (SLC26) in auditory organs of nonmammalian vertebrates and insects. Proc. Natl. Acad. Sci. U. S. A. 100, 7690–7695. Weil, D., Blanchard, S., Kaplan, J., Guilford, P., Gibson, F., Walsh, J., Mburu, P., Varela, A., Levilliers, J., and Weston, M. D. 1995. Defective myosin VIIA gene responsible for Usher syndrome type 1B. Nature 374, 60–61. West-Eberhard, M. J. 2003. Developmental Plasticity and Evolution. Oxford University Press. Wever, E. G. 1978. The Reptile Ear, Princeton Univ Press. Whitfield, T. T., Riley, B. B., Chiang, M. Y., and Phillips, B. 2002. Development of the zebrafish inner ear. Dev. Dyn. 223, 427–58. Wood, R. L. and Novak, P. L. 1982. The anchoring of nematocysts and nematocytes in the tentacles of Hydra. J. Ultrastruct. Res. 81, 104–116. Woods, C., Montcouquiol, M., and Kelley, M. W. 2004. Math1 regulates development of the sensory epithelium in the mammalian cochlea. Nat. Neurosci. 7, 1310–1318. Wu, D. K. and Oh, S. H. 1996. Sensory organ generation in the chick inner ear. J. Neurosci. 16, 6454–6462. Wu, D. K., Nunes, F. D., and Choo, D. 1998. Axial specification for sensory organs versus non-sensory structures of the chicken inner ear. Development 125, 11–20. Yack, J. E. 2004. The structure and function of auditory chordotonal organs in insects. Microsc. Res. Tech. 63, 315–337. Young, J. Z. 1981. The Life of Vertebrates. Oxford. Zheng, J., Shen, W., He, D. Z., Long, K. B., Madison, L. D., and Dallos, P. 2000. Prestin is the motor protein of cochlear outer hair cells. Nature 405, 149–155. Zine, A., Aubert, A., Qiu, J., Therianos, S., Guillemot, F., Kageyama, R., and de Ribaupierre, F. 2001. Hes1 and Hes5 activities are required for the normal development of the hair cells in the mammalian inner ear. J. Neurosci. 21, 4712–4720.
Relevant Website http://www.wzw.tu-muenchen.de – Wissenschaftszentrum Weihenstephan fu¨r Erna¨hrung, Landnutzung und Umwelt.