Search

[Introduction] Page 2 of 20 Sensory proteins must relay structural signals from the sensory site over large distances to regulatory output domains. Phytochromes are a major family of red-light sensing kinases that control diverse cell ular functions in plants, bacteria, and fungi. 1-9 Bacterial phytochro mes consist of a photosensory core and a C-te rminal regulatory domain. 10,11 Structures of photosensory cores are reported in the resting state 12-18 and conformational responses to light activat ion have been proposed in the vicinity of the chromophore. 19-23 However, the structure of the signalling state and the mechanism of downstream signal re lay through the photosensory core remain elusive. Here, we report crystal and solution structures of the resting and active states of the photosensory core of the bacteriophytochrome from Deinococcus radiodurans . The structures reveal an open and closed form of the dimeric protein for the signalling an d resting state, respectively. This nanometre scale rearrangement is controlled by refolding of an evolutionarily conserved “tongue” , which is in contact with the chromophore. The findings reveal an unus ual mechanism where atomic scale conformational ch anges around the chromopho re are first amplified into an Ångström scale distance change in the tongue, and further grow into a nanometre scale conformational sign al. The structural mechanism is a blueprint for understanding how the sensor proteins connect to the cellular signalling network. peerReviewed

A combination of structural, computational and biophysical tools is used to characterize the bond between tip-link proteins protocadherin 15 and cadherin 23, which have an essential role in inner-ear mechanotransduction; the bond, involving an extended protein handshake, is found to be affected by deafness mutations and is mechanically strong enough to resist forces in hair cells, adding to our understanding of hair-cell sensory transduction and interactions among cadherins. Hair cells in the inner ear transform mechanical stimuli into electrical signals that are crucial for hearing and balance. These cells contain actin-rich stereocilia that are interconnected by filaments such as the tip link, which forms the mechanotransduction apparatus. The adhesion molecules protocadherin 15 and cadherin 23 form the tip link. In this study, David Corey et al. use a combination of structural and biophysical experiments to characterize the bond between cadherin 23 and protocadherin 15. The bond is mechanically strong enough to resist forces in hair cells and requires calcium to maintain stability. These results provide molecular insights into the mechanics of hair-cell sensory transduction. Hearing and balance use hair cells in the inner ear to transform mechanical stimuli into electrical signals1. Mechanical force from sound waves or head movements is conveyed to hair-cell transduction channels by tip links2,3, fine filaments formed by two atypical cadherins known as protocadherin 15 and cadherin 23 (refs 4, 5). These two proteins are involved in inherited deafness6,7,8,9,10 and feature long extracellular domains that interact tip-to-tip5,11 in a Ca2+-dependent manner. However, the molecular architecture of this complex is unknown. Here we combine crystallography, molecular dynamics simulations and binding experiments to characterize the protocadherin 15–cadherin 23 bond. We find a unique cadherin interaction mechanism, in which the two most amino-terminal cadherin repeats (extracellular cadherin repeats 1 and 2) of each protein interact to form an overlapped, antiparallel heterodimer. Simulations predict that this tip-link bond is mechanically strong enough to resist forces in hair cells. In addition, the complex is shown to become unstable in response to Ca2+ removal owing to increased flexure of Ca2+-free cadherin repeats. Finally, we use structures and biochemical measurements to study the molecular mechanisms by which deafness mutations disrupt tip-link function. Overall, our results shed light on the molecular mechanics of hair-cell sensory transduction and on new interaction mechanisms for cadherins, a large protein family implicated in tissue and organ morphogenesis12,13, neural connectivity14 and cancer15.

The Drosophila Piezo protein is shown to function in sensory neurons to transduce mechanical force in vivo. Many tissues are able to detect and respond to mechanical forces, and this mechanical sensitivity has been implicated in many biological processes and diseases, including touch, pain, deafness and hypertension. The conversion of mechanical force into biological signals, or 'mechanotransduction', is thought to involve specialized cation channels. In a pair of papers, Ardem Patapoutian and colleagues establish that the large transmembrane proteins of the 'Piezo' family — conserved from animals to plants and protozoa — are among the long-sought-after mechanically activated ion channels. Coste et al. show that the Drosophila melanogaster Piezo protein induces mechanically activated cationic currents in human embryonic kidney cells, establishing functional conservation. Comparison of the mechanically activated currents induced by mouse and fly Piezos reveals ion-channel activities with unique pore properties, suggesting that Piezos are bona fide ion channels. Kim et al. show that D. melanogaster Piezo is essential for sensing mechanical pain in fruitflies, giving the first demonstration that Piezos are physiologically relevant mechanosensors in vivo. Transduction of mechanical stimuli by receptor cells is essential for senses such as hearing, touch and pain1,2,3,4. Ion channels have a role in neuronal mechanotransduction in invertebrates1; however, functional conservation of these ion channels in mammalian mechanotransduction is not observed. For example, no mechanoreceptor potential C (NOMPC), a member of transient receptor potential (TRP) ion channel family, acts as a mechanotransducer in Drosophila melanogaster5 and Caenorhabditis elegans6,7; however, it has no orthologues in mammals. Degenerin/epithelial sodium channel (DEG/ENaC) family members are mechanotransducers in C. elegans8 and potentially in D. melanogaster9; however, a direct role of its mammalian homologues in sensing mechanical force has not been shown. Recently, Piezo1 (also known as Fam38a) and Piezo2 (also known as Fam38b) were identified as components of mechanically activated channels in mammals10. The Piezo family are evolutionarily conserved transmembrane proteins. It is unknown whether they function in mechanical sensing in vivo and, if they do, which mechanosensory modalities they mediate. Here we study the physiological role of the single Piezo member in D. melanogaster (Dmpiezo; also known as CG8486). Dmpiezo expression in human cells induces mechanically activated currents, similar to its mammalian counterparts11. Behavioural responses to noxious mechanical stimuli were severely reduced in Dmpiezo knockout larvae, whereas responses to another noxious stimulus or touch were not affected. Knocking down Dmpiezo in sensory neurons that mediate nociception and express the DEG/ENaC ion channel pickpocket (ppk) was sufficient to impair responses to noxious mechanical stimuli. Furthermore, expression of Dmpiezo in these same neurons rescued the phenotype of the constitutive Dmpiezo knockout larvae. Accordingly, electrophysiological recordings from ppk-positive neurons revealed a Dmpiezo-dependent, mechanically activated current. Finally, we found that Dmpiezo and ppk function in parallel pathways in ppk-positive cells, and that mechanical nociception is abolished in the absence of both channels. These data demonstrate the physiological relevance of the Piezo family in mechanotransduction in vivo, supporting a role of Piezo proteins in mechanosensory nociception.

Just how animals are able to use the Earth's magnetic field for navigation is the subject of much current interest. It is known that the UV-A/blue light photoreceptor cryptochrome (Cry) mediates the light-dependent magnetic sense in Drosophila. Steven Reppert and colleagues now use genetic manipulation to show that two cryptochromes from the monarch butterfly, Drosophila-like Cry1 and vertebrate-like Cry2, can restore magnetic sensation in cry-deficient Drosophila. This suggests that both types of Cry have the capacity to sense magnetic fields in the migratory monarch butterfly via a mechanism that remains unclear — and that light sensitivity is involved in some way. Animals use the Earth's magnetic field for orientation but the biophysical basis of this is unclear. The light-dependent magnetic sense of Drosophila melanogaster was recently shown to be mediated by the cryptochrome (Cry) photoreceptor; here, using a transgenic approach, the type 1 and 2 Cry of the monarch butterfly are shown to both function in the magnetoreception system of Drosophila, and probably use an unconventional photochemical mechanism. Understanding the biophysical basis of animal magnetoreception has been one of the greatest challenges in sensory biology. Recently it was discovered that the light-dependent magnetic sense of Drosophila melanogaster is mediated by the ultraviolet (UV)-A/blue light photoreceptor cryptochrome (Cry)1. Here we show, using a transgenic approach, that the photoreceptive, Drosophila-like type 1 Cry and the transcriptionally repressive, vertebrate-like type 2 Cry of the monarch butterfly (Danaus plexippus) can both function in the magnetoreception system of Drosophila and require UV-A/blue light (wavelength below 420 nm) to do so. The lack of magnetic responses for both Cry types at wavelengths above 420 nm does not fit the widely held view that tryptophan triad-generated radical pairs mediate the ability of Cry to sense a magnetic field. We bolster this assessment by using a mutant form of Drosophila and monarch type 1 Cry and confirm that the tryptophan triad pathway is not crucial in magnetic transduction. Together, these results suggest that animal Crys mediate light-dependent magnetoreception through an unconventional photochemical mechanism. This work emphasizes the utility of Drosophila transgenesis for elucidating the precise mechanisms of Cry-mediated magnetosensitivity in insects and also in vertebrates such as migrating birds.

An investigation into how poikilothermic organisms are able to maintain highly effective mechanical responses finds that a protein called Spam forms an extracellular shield that guards mechanosensory neurons from environmental insult. Spam also enables cells to resist physically and chemically induced deformation. Maintaining cell shape and tone is crucial for the function and survival of cells and tissues. Mechanotransduction relies on the transformation of minuscule mechanical forces into high-fidelity electrical responses1,2,3. When mechanoreceptors are stimulated, mechanically sensitive cation channels open and produce an inward transduction current that depolarizes the cell. For this process to operate effectively, the transduction machinery has to retain integrity and remain unfailingly independent of environmental changes. This is particularly challenging for poikilothermic organisms, where changes in temperature in the environment may impact the function of mechanoreceptor neurons. Thus, we wondered how insects whose habitat might quickly vary over several tens of degrees of temperature manage to maintain highly effective mechanical senses. We screened for Drosophila mutants with defective mechanical responses at elevated ambient temperatures, and identified a gene, spam, whose role is to protect the mechanosensory organ from massive cellular deformation caused by heat-induced osmotic imbalance. Here we show that Spam protein forms an extracellular shield that guards mechanosensory neurons from environmental insult. Remarkably, heterologously expressed Spam protein also endowed other cells with superb defence against physically and chemically induced deformation. We studied the mechanical impact of Spam coating and show that spam-coated cells are up to ten times stiffer than uncoated controls. Together, these results help explain how poikilothermic organisms preserve the architecture of critical cells during environmental stress, and illustrate an elegant and simple solution to such challenge.

TRP channels are the vanguard of our sensory systems, responding to temperature, touch, pain, osmolarity, pheromones, taste and other stimuli. But their role is much broader than classical sensory transduction. They are an ancient sensory apparatus for the cell, not just the multicellular organism, and they have been adapted to respond to all manner of stimuli, from both within and outside the cell.

Hearing sensitivity in mammals is enhanced by more than 40 dB (that is, 100-fold) by mechanical amplification thought to be generated by one class of cochlear sensory cells, the outer hair cells. In addition to the mechano-electrical transduction required for auditory sensation, mammalian outer hair cells also perform electromechanical transduction, whereby transmembrane voltage drives cellular length changes at audio frequencies in vitro. This electromotility is thought to arise through voltage-gated conformational changes in a membrane protein, and prestin has been proposed as this molecular motor. Here we show that targeted deletion of prestin in mice results in loss of outer hair cell electromotility in vitro and a 40-60 dB loss of cochlear sensitivity in vivo, without disruption of mechano-electrical transduction in outer hair cells. In heterozygotes, electromotility is halved and there is a twofold (about 6 dB) increase in cochlear thresholds. These results suggest that prestin is indeed the motor protein, that there is a simple and direct coupling between electromotility and cochlear amplification, and that there is no need to invoke additional active processes to explain cochlear sensitivity in the mammalian ear.

Taste is the sensory system devoted primarily to a quality check of food to be ingested. Although aided by smell and visual inspection, the final recognition and selection relies on chemoreceptive events in the mouth. Emotional states of acute pleasure or displeasure guide the selection and contribute much to our quality of life. Membrane proteins that serve as receptors for the transduction of taste have for a long time remained elusive. But screening the mass of genome sequence data that have recently become available has provided a new means to identify key receptors for bitter and sweet taste. Molecular biology has also identified receptors for salty, sour and umami taste.

Adaptation to odorants begins at the level of sensory receptor cells, presumably through modulation of their transduction machinery. The olfactory signal transduction involves the activation of the adenylyl cyclase/cyclic AMP second messenger system which leads to the sequential opening of cAMP-gated channels and Ca2+-activated chloride ion channels. Several reports of results obtained from in vitro preparations describe the possible molecular mechanisms involved in odorant adaptation; namely, ordorant receptor phosphorylation, activation of phosphodiesterase, and ion channel regulation. However, it is still unknown whether these putative mechanisms work in the intact olfactory receptor cell. Here we investigate the nature of the adaptational mechanism in intact olfactory cells by using a combination of odorant stimulation and caged cAMP photolysis which produces current responses that bypass the early stages of signal transduction (involving the receptor, G protein and adenylyl cyclase). Odorant- and cAMP-induced responses showed the same adaptation in a Ca2+-dependent manner, indicating that adaptation occurs entirely downstream of the cyclase. Moreover, we show that phosphodiesterase activity remains constant during adaptation and that an affinity change of the cAMP-gated channel for ligands accounts well for our results. We conclude that the principal mechanism underlying odorant adaptation is actually a modulation of the cAMP-gated channel by Ca2+ feedback.

Cellular responses to mechanical forces are crucial in embryonic development and adult physiology, and are involved in numerous diseases, including atherosclerosis, hypertension, osteoporosis, muscular dystrophy, myopathies and cancer. These responses are mediated by load-bearing subcellular structures, such as the plasma membrane, cell-adhesion complexes and the cytoskeleton. Recent work has demonstrated that these structures are dynamic, undergoing assembly, disassembly and movement, even when ostensibly stable. An emerging insight is that transduction of forces into biochemical signals occurs within the context of these processes. This framework helps to explain how forces of varying strengths or dynamic characteristics regulate distinct signalling pathways.

The sense of smell is highly evolved in mammals, allowing discrimination between a vast number of odorants, with detection thresholds as low as 10(-17) M (ref. 1). Although several features of mammalian olfactory transduction have been revealed by biochemical and molecular biological studies, the odorant-induced membrane current has remained elusive. In amphibians this current is mediated by cyclic-nucleotide-gated channels, which depolarize the cell by Na+ and Ca+ influx and consequent Cl- efflux through Ca(2+)-dependent Cl- channels. The Cl- current may be absent in mammals, however, because its proposed role is linked to the aquatic habitat of amphibians. Here we show that the transduction current in rat olfactory receptor cells is initiated by cyclic-nucleotide-gated channels. The Cl- current is also present and endows the transduction current with a steep sigmoidal dependence on cyclic AMP concentration in both rat and in an amphibian, indicating a new function for the Cl- channel: nonlinear amplification of the transduction signal, whereby suprathreshold responses are boosted relative to basal transduction noise.

OLFACTORY transduction is thought to be mediated by a membrane-bound receptor protein initiating a multistep reaction cascade which ultimately leads to a depolarizing generator current1,2. There is considerable evidence for the involvement of adenylate cyclase in vertebrate olfactory transduction3–6, and some data indicate that phospholipase C may have a central role in insect olfaction7. However, one must show that odorants not only stimulate enzyme activity but also induce changes in concentrations of relevant second messengers. One important criterion for a candidate second messenger of chemo-electrical transduction is that its formation must precede the onset of the odorant-induced membrane permeability changes which proceed on a subsecond time-scale8. Here we report an odorant-induced, transient accumulation of cyclic AMP in isolated olfactory cilia from rats, and the generation of inositol trisphosphate in antennal preparations from insects, both of which show subsecond time courses that are sufficiently rapid to mediate the odorant-regulated permeability of olfactory receptor cells.

Sensory neurons innervating the skin encode the familiar sensations of temperature, touch and pain. An explosion of progress has revealed unanticipated cellular and molecular complexity in these senses. It is now clear that perception of a single stimulus, such as heat, requires several transduction mechanisms. Conversely, a given protein may contribute to multiple senses, such as heat and touch. Recent studies have also led to the surprising insight that skin cells might transduce temperature and touch. To break the code underlying somatosensation, we must therefore understand how the skin's sensory functions are divided among signalling molecules and cell types.

Throughout the nervous system, moment-to-moment communication relies on postsynaptic receptors to detect neurotransmitters and change the membrane potential. For the Cys-loop superfamily of receptors, recent structural data have catalysed a leap in our understanding of the three steps of chemical-to-electrical transduction: neurotransmitter binding, communication between the binding site and the barrier to ions, and opening and closing of the barrier. The emerging insights might be expected to explain how mutations of receptors cause neurological disease, but the opposite is generally true. Namely, analyses of disease-causing mutations have clarified receptor structure-function relationships as well as mechanisms governing the postsynaptic response.

Extracellular signalling by the purine nucleotide ATP has long been associated with sensory function. In the periphery, ATP mediates nociception, mechanosensitivity, thermal sensitivity and O2 chemosensitivity. These processes share a common mechanism that involves the release of ATP to excite afferent fibres via activation of ionotropic P2X and/or metabotropic P2Y receptors. Chemosensors located in the brainstem are crucial for the maintenance of physiological levels of blood gases through the regulation of breathing. Here we show that an increase in pCO2 in the arterial blood triggers the immediate release of ATP from three chemosensitive regions located on the ventral surface of the medulla oblongata. Blockade of ATP receptors at these sites diminishes the chemosensory control of breathing, suggesting that ATP release constitutes a key step in central chemosensory transduction. These new data suggest that ATP, a phylogenetically ancient, unique and simple molecule, has been widely used in the evolution of afferent systems to mediate distinct forms of sensory transduction not only in the periphery but also within the central nervous system.

The mechanical environment crucially influences many cell functions. However, it remains largely mysterious how mechanical stimuli are transmitted into biochemical signals. Src is known to regulate the integrin-cytoskeleton interaction, which is essential for the transduction of mechanical stimuli. Using fluorescent resonance energy transfer (FRET), here we develop a genetically encoded Src reporter that enables the imaging and quantification of spatio-temporal activation of Src in live cells. We introduced a local mechanical stimulation to human umbilical vein endothelial cells (HUVECs) by applying laser-tweezer traction on fibronectin-coated beads adhering to the cells. Using the Src reporter, we observed a rapid distal Src activation and a slower directional wave propagation of Src activation along the plasma membrane. This wave propagated away from the stimulation site with a speed (mean +/- s.e.m.) of 18.1 +/- 1.7 nm s(-1). This force-induced directional and long-range activation of Src was abolished by the disruption of actin filaments or microtubules. Our reporter has thus made it possible to monitor mechanotransduction in live cells with spatio-temporal characterization. We find that the transmission of mechanically induced Src activation is a dynamic process that directs signals via the cytoskeleton to spatial destinations.

Natural Caenorhabditis elegans isolates exhibit either social or solitary feeding on bacteria. We show here that social feeding is induced by nociceptive neurons that detect adverse or stressful conditions. Ablation of the nociceptive neurons ASH and ADL transforms social animals into solitary feeders. Social feeding is probably due to the sensation of noxious chemicals by ASH and ADL neurons; it requires the genes ocr-2 and osm-9, which encode TRP-related transduction channels, and odr-4 and odr-8, which are required to localize sensory chemoreceptors to cilia. Other sensory neurons may suppress social feeding, as social feeding in ocr-2 and odr-4 mutants is restored by mutations in osm-3, a gene required for the development of 26 ciliated sensory neurons. Our data suggest a model for regulation of social feeding by opposing sensory inputs: aversive inputs to nociceptive neurons promote social feeding, whereas antagonistic inputs from neurons that express osm-3 inhibit aggregation.

Gap junction channels are structurally distinct from other ion channels in that they comprise two hemichannels which interact head-to-head to form an aqueous channel between cells. Intercellular voltage differences together with increased intracellular concentrations of H+ and Ca2+ cause closure of these normally patent channels. The relative sensitivity to voltage varies with the subunit (connexin) composition of the channels. The third of four transmembrane-spanning regions (M3) in connexins has been proposed to form the channel lining, and a global 'tilting' of the hemichannel subunits has been correlated with channel closure. But specific components involved in transduction of channel gating events have not been identified in either gap junctions or other ion channel classes (however, see model in ref. 5). We have examined a strictly conserved proline centrally located in M2 of connexin proteins. Mutation of this proline (Pro 87) in connexin 26 causes a reversal in the voltage-gating response when the mutant hemichannel is paired with wild-type connexin 26 in the Xenopus oocyte system. This suggests that the unique properties associated with this residue are critical to the transduction of voltage gating in these channels.

Sour taste is initiated by protons acting at receptor proteins or channels. In vertebrates, transduction of this taste quality involves several parallel pathways1,2,3,4,5. Here we examine the effects of sour stimuli on taste cells in slices of vallate papilla from rat. From a subset of cells, we identified a hyperpolarization-activated current that was enhanced by sour stimulation at the taste pore. This current resembled Ih found in neurons and cardio-myocytes6,7, a current carried by members of the family of hyperpolarization-activated and cyclic-nucleotide-gated (HCN) channels8,9,10,11,12,13. We show by in situ hybridization and immunohistochemistry that HCN1 and HCN4 are expressed in a subset of taste cells. By contrast, gustducin, the G-protein involved in bitter and sweet taste14, is not expressed in these cells. Lowering extracellular pH causes a dose-dependent flattening of the activation curve of HCN channels and a shift in the voltage of half-maximal activation to more positive voltages. Our results indicate that HCN channels are gated by extracellular protons and may act as receptors for sour taste.

CALCIUM is involved in the adaptation of vertebrate photoreceptors to light1,2 and may have a similar role in invertebrate phototransduction3,4. But the molecular mechanisms mediating this stimulus-dependent regulation are not well understood in any G protein-coupled transduction system. We have developed a preparation of isolated Drosophila photoreceptors that has allowed us to carry out an electrophysiological characterization of the light-activated response in these sensory neurons using patch-clamp techniques. We report here that extracellular calcium entering through the light-activated conductance is a key regulator of both the activation and deactivation phases of the phototransduction cascade, and that inaC mutant photoreceptors5 are specifically defective in the calcium-dependent deactivation mechanism. These data suggest that the light-dependent calcium influx inactivates this cascade through a biochemical pathway that requires the inaC gene product, and that this mechanism represents a molecular basis for stimulus-dependent regulation of visual transduction in Drosophila photoreceptors.

The ability to detect taste stimuli results from the activation of taste receptors located in taste-bud cells. There are several gustatory transduction mechanisms, involving membrane receptors, guanine-nucleotide-binding proteins (G proteins), second messengers and ion channels1, but genes encoding taste receptors have not yet been identified. Here we identify a complementary DNA that encodes a receptor for sour tastes.

Photoreceptors detect light through a seven-helix receptor (rhodopsin) and heterotrimeric G protein (transducin) coupled to a cyclic GMP phosphodiesterase1,2. Similar pathways are used to amplify responses to hormones, taste and smell3,4,5. The amplification of phototransduction is reduced by a fall in cytoplasmic Ca2+ (refs 6, 7, 8, 9, 10), but it is not known how the deactivation of rhodopsin and transducin influence this response and hence the extent and duration of phosphodiesterase activity11,12,13,14. Here we investigate this by recording the electrical response to flashes of light in truncated rod photoreceptors10. By removing ATP to block the deactivation of rhodopsin by phosphorylation15, we show that this reaction limits the amplitude of the response and begins within 3.2 s of a flash in a solution containing 1 μM Ca2+, falling to 0.9 s in a zero-Ca2+ solution. In contrast, the activation and amplitude of the response were unaffected when transducin deactivation by GTP hydrolysis was blocked by replacing GTP with its non-hydrolysable analogue GTP-γS11, demonstrating that there is little GTP hydrolysis occurring over the period in which photoexcited rhodopsin is quenched. The rapid deactivation of rhodopsin is therefore a Ca2+-sensitive step controlling the amplitude of the light response, whereas transducin deactivation is slower and controls recovery.

GENETIC deafness is common, affecting about 1 in 2,000 births1. Many of these show primary abnormalities of the sensory neuro-epithelia of the inner ear, as do several hearing-impaired mouse mutants, suggesting that genes involved in sensory transduction could be affected. Here we report the identification of one such gene, the mouseshaker-1(shl) gene. Shaker-1 homozygotes show hyperactivity, head-tossing and circling due to vestibular dysfunction, together with typical neuroepithelial-type cochlear defects involving dysfunction and progressive degeneration of the organ of Corti2–7. The shl gene encodes an unconventional myosin molecule of the type VII family. Three mutations are described, two mis-sense mutations and a splice acceptor site mutation, all in the region encoding the myosin head. The myosin type VII molecule encoded byshl is the first molecule to be identified that is known, by virtue of its mutations, to be involved in auditory transduction.

Mechanically controlled ion channels — transduction channels — are a key feature of the cells that detect sound, touch and movement. In fruitfly ears, the channels belong to a very familiar group of proteins.

It is thought that the sensitivity of mammalian hearing depends on amplification of the incoming sound within the cochlea by a select population of sensory cells, the outer hair cells. It has been suggested that these cells sense displacements and feedback forces which enhance the basilar membrane motion by reducing the inherent damping of the cochlear partition. In support of this hypothesis, outer hair cells show membrane-potential-induced length changes at acoustic rates. This process has been termed 'reverse transduction'. For amplification, the forces should be large enough to move the basilar membrane. Using a displacement-sensitive interferometer, we tested this hypothesis in an isolated cochlea while stimulating the outer hair cells with current passed across the partition. We show here that the cochlear partition distorts under the action of electrically driven hair cell length changes and produces place-specific vibration of the basilar membrane of a magnitude comparable to that observed near auditory threshold (about 1 nm). Such measurements supply direct evidence that cochlear amplification arises from the properties of the outer hair cell population.

Odorant stimulation leads to a depolarization of olfactory receptor neurons. A mechanism underlying this transduction, which occurs in the sensory cilia, involves a G-protein-mediated increase in adenylyl cyclase activity, and therefore a rise in internal cyclic AMP and consequent opening of a cAMP-gated cation channel on the plasma membrane. Another mechanism, not as well established, involves the opening of an inositol trisphosphate-activated cation channel on the plasma membrane as a result of phospholipase C activity. In both cases, an influx of cations is thought to generate the depolarizing receptor potential. We now report, however, that the mechanism is actually more complex. The odorant-induced current appears to contain an inward chloride component also, which is triggered by calcium influx through the cation-selective channel. This newly found chloride component can be as large as the cationic component. The co-existence of cationic and chloride components in the odorant response, possibly unique among sensory transduction mechanisms, may serve to reduce variations in the transduction current resulting from changes in external ionic concentrations around the olfactory cilia. Our finding can explain the long-standing puzzle of why removal of most mucosal cations still does not diminish the amplitude of the olfactory receptor cell response.

Many G-protein-coupled receptors are only transiently active because an inactivation process stops the receptor from activating G protein molecules. Although this inactivation has been investigated in vitro, the real kinetics of the process can only be obtained from intact cells. Here we describe a method for measuring the inactivation of rhodopsin in intact photoreceptors and the application of this method to the ultraviolet rhodopsin of Limulus median eye. The results show that the inactivation process is very rapid (less than 150 ms) and occurs well before the peak of the receptor potential. We have also investigated whether the inactivation process can itself be modulated. Our results show that light-adaptation accelerates inactivation by about 10-fold, providing evidence that G-protein-mediated transduction can be modulated at this first stage.

Depolarizing bipolar cells (DBCs) of the retina are the only neurons in the vertebrate central nervous system known to be hyperpolarized by the neurotransmitter glutamate. Both glutamate and its analogue L-2-amino-4-phosphonobutyrate (APB) hyperpolarize DBCs by decreasing membrane conductance. Furthermore, glutamate responses in DBCs slowly decrease during whole-cell recording, suggesting that the response involves a second messenger system. Here we report that intracellular cyclic GMP or GTP activates a membrane conductance that is suppressed by APB, resulting in an enhanced APB response. In the presence of GTP-gamma-S, APB causes an irreversible suppression of the conductance. Inhibitors of G-protein activation or phosphodiesterase activity decrease the APB response. Thus, the DBC glutamate receptor seems to close ion channels by increasing the rate of cGMP hydrolysis by a G protein-mediated process that is strikingly similar to light transduction in photoreceptors.

The molecular cloning of all the major components in the transduction pathway described in Otto Loewi's famous experiments on isolated frog hearts' is now complete with the publication of a letter by Kubo et al. on page 802 of this issue. Loewi cannulated two hearts and connected the effluent from heart A to the input of heart B. He then stimulated the vagus nerve to heart A; in agreement with previous experiments, heart A ceased beating for several minutes.

Nature 381, 796-800 (1996). THE Nature cover relating to this letter, in the 27 June issue, did not do justice to the original image. The image has now been re-scanned to emphasise the significant features (right), and the new and improved version will be available on reprints of the letter. In addition, credit should have been given in the caption to Dr Luis Ruiz-Avila.

NEURONS and other cells, such as those of epithelia, accumulate particular proteins in spatially discrete domains of the plasma membrane. This enrichment is probably important for localization of function, but it is not clear how it is accomplished. One proposal for epithelial cells is that proteins contain targeting signals which guide preferential accumulation in basal or apical membranes1. The growth-cone membrane of a neuron serves as a specialized transduction system, which helps to convert cues from its environment into regulated growth. Because it can be physically separated from the cell soma, it has been possible to show that the growth-cone membrane contains a restricted set of total cellular proteins2, although, to our knowledge, no proteins are limited to that structure. One of the most prominent proteins in the growth-cone membrane is GAP-43 (refs 3–5; for reviews see refs 6 and 7). Basi et al. have suggested that the N-terminus of GAP-43 might be important for the binding of GAP-43 to the growth-cone membrane8. Skene and Virag9 recently found that the cysteines in the N-terminus are fatty-acylated and that this post-translational modification correlates with membrane-binding ability. We investigated the binding of GAP-43 to the growth-cone membrane by mutational analysis and by laser-scanning confocal microscopy of fusion proteins that included regions of GAP-43 and chloram-phenicol acetyltransferase (CAT). We found that a short stretch of the GAP-43 N-terminus suffices to direct accumulation in growth-cone membranes, especially in the filopodia. This supports a previous proposal8,9 for the importance of this region of GAP-43 in determining the membrane distribution of GAP-43.

The senses of hearing and equilibrium depend on sensory receptors called hair cells which can detect motions of atomic dimensions and respond more than 100,000 times a second. Biophysical studies suggest that mechanical forces control the opening and closing of transduction channels by acting through elastic components in each hair cell's mechanoreceptive hair bundle. Other ion channels, as well as the mechanical and hydrodynamic properties of hair bundles, tune individual hair cells to particular frequencies of stimulation.

THE involvement of calcium in the transduction process in vertebrate photoreceptors, in particular rods, has been postulated by Yoshikami and Hagins1, who suggest that this cation is sequestered in the membranous disks of the outer segment and released on stimulation by light. Recently, evidence supporting this hypothesis was obtained by Fishman et al.2 who, using an electron microscopic technique3–5 for the localisation of calcium, showed precipitate within the disks of frog rod outer segments. However, they did not explore the possibility of a light-induced decrease of calcium within the disks, which would have further supported the hypothesis. We report here evidence for the involvement of calcium in the phototransduction process in a unique vertebrate photoreceptive system.

Like retinal rods, cone photoreceptors contain cyclic GMP and light-activated phosphodiesterase1–3. The cGMP phosphodiesterase cascade is thought to mediate phototransduction in rods. Biochemical assays of nucleotide content in cone-dominant retinas, however, have failed to demonstrate light-induced changes in cGMP1. Changes in cyclic AMP following light exposure have been reported, leading to the suggestion that in cone phototransduction cAMP assumes a role analogous to that played by cGMP in rods4. Cyclic GMP introduced from tight-seal pipettes into isolated cones of the larval tiger salamander, Ambystoma tigrinum, rapidly increases light-modulated membrane current more than 10-fold. In cones, as in rods, cGMP also causes an approximately 10-fold increase in photocurrent duration and a 5- to 10-fold increase in light-sensitivity. Cyclic AMP has no effect on cone photocurrents under the same conditions. Because cGMP has similar effects on photocurrent magnitude and kinetics in both rods and cones, we conclude that cGMP plays corresponding roles in transduction in both vertebrate photoreceptor classes.

Light-induced activation of cyclic GMP phosphodiesterase (PDE) in the outer segments of vertebrate rod photoreceptors has been suggested as a step in the transduction process linking the absorption of light by rhodopsin with a hyperpolarization of the plasma membrane1–3. Activation of PDE is mediated by a ‘GTP-binding protein’ (G). As a result of interaction with a rhodopsin photoproduct (possibly metarhodopsin II380), this GTP-binding protein exchanges a previously bound GDP for a GTP4,5. This ‘GTP-charged’ protein (GGTP) is thought to activate PDE through an interaction in which an inhibitory protein is released from PDE6–8. Assays of PDE activity in vitro have demonstrated that the hydrolysis of cyclic GMP occurs within milliseconds of light onset9, which suggests that this process is rapid enough to be an intermediate step in transduction. To test this idea electrophysiologically, we have injected purified GTP-binding protein that was binding a hydrolysis-resistant analogue of GTP, guanylyl imidodiphosphate (p(NH)ppG), termed Gp(NH)ppG, as well as partially purified PDE, into toad rod outer segments while recording membrane voltage. We report here that injection of either of these biochemically active proteins induces a reversible hyperpolarization of the rod membrane, while injection of inactive proteins seems to have no effect on membrane voltage.

All aminoglycoside antibiotics in clinical use are ototoxic in that they can cause permanent malfunctioning of the cochlea and vestibule1–3. It is generally assumed that these ototoxic antibiotics primarily affect the sensory hair cells in the inner ear and that degeneration of secretory and neural elements comes later2–5. Most experimental work on aminoglycoside antibiotic ototoxicity has concentrated on the irreversible effects on inner ear functioning and hair cell morphology2,3. Little is known about their mode of action on the hair cells. Interference with hair cell protein synthesis, in analogy with the antimicrobial mechanism, has been ruled out as the primary cause of the high selective6 toxicity of the aminoglycosides3,5. Recent evidence indicates that these antibiotics may interfere with membrane phos-pholipid metabolism7. Morphological studies have demonstrated that early on in the ototoxic process the membrane of the sensory hairs is already damaged4. The aminoglycoside antibiotics are reported to have a direct and reversible8,9 effect on hair cell functioning in organs of the acoustico-lateralis system of lower vertebrates8–11. Hence, the primary site of action is probably the hair cell membrane. We have now therefore examined reversible effects of ototoxic antibiotics on single fibre afferent nerve activity and extracellular receptor potentials in the lateral-line organ of Xenopus laevis. Our results show that the antibiotics, in concentrations similar to those measured in the cochlear fluids3, have a dual action on the hair cells. They not only increase the spontaneous afferent nerve activity, probably by an effect on the hair cell membrane, but also severely impair the mechano-electric transduction process, resulting in a large phase lag of the receptor potentials. The latter effect, which is antagonised by Ca2+ may be due to interference of the antibiotics with the mechanical properties of the sensory hairs.

THE semicircular canal is considered classically to be a sensor of angular head motion during animal navigation; however, the functional significance of the canal's morphology and innervation for transduction and encoding of specific head trajectories is unclear. Neural responses from single afferent fibres innervating the crista of isolated elasmobranch labyrinths were used by Groen et al. to determine linear system coefficients of an overdamped pendulum equation, considered to reflect directly the dynamics of the motion of the cupula–endolymphatic fluid in the ampulla1. Similar studies on other preparations revealed certain neural response characteristics that were at variance with simple pendulum models, and these differences were ascribed variously to differences in hair cell physiology and afferent or efferent innervation properties, whose effects were superimposed on the fundamental fluid pendulum dynamics2–4. But in addition to these hypotheses based on cellular influences, the topological organisation of the receptor could result in specific afferent response characteristics deriving from different anatomical regions within the crista. As evidence for the latter, we report diverse classes of afferent responses, describable by distinctly different linear system equations, that are correlated with the position of separate bundles innervating specific areas of the crista of the guitarfish, Rhinobatos productus (Fig. 1).

A central aspect of neuronal function is how each nerve cell translated synaptic input into a sequence of action potentials that carry information along the axon, coded as spike frequency. When transduction from a graded depolarizing input to spikes is studied by injecting a depolarizing current, there is often a remarkably long delay to the first action potential, both in mammalian and molluscan neurons. Here, I report that the delayed excitation in rat hippocampal neurons is due to a slowly inactivating potassium current, ID. ID co-exists with other voltage-gated K+ currents, including a fast A current and a slow delayed rectifier current. As ID activates in the subthreshold range, and takes tens of seconds to recover from inactivation, it enables the cell to integrate separate depolarizing inputs over long times. ID also makes the encoding properties of the cell exceedingly sensitive to the prevailing membrane potential.

ONE of the many metabolic functions attributed to adenosine 3′:5′-cyclic monophosphate (cyclic AMP) is that of a mediator in the responses of living organisms to light stimulation. It has been associated with dopaminergic synaptic activities and also with the primary visual process in the retina. In the rod outer segments of the frog, Rana pipiens, cyclic AMP concentrations are diminished by illumination. Regulation is through the cyclic nucleotide phosphodiesterase1. In addition, an adenyl cyclase has been isolated from retina which is activated by the catecholamine dopamine; dopamine is thought to be an important neurotransmitter in the retina2. The exact relationship between the initial transduction of the light signal and the attenuation in the levels of cyclic AMP is unknown. One reason for this is the complexity of the tissue and the difficulty of isolating clean fractions. I describe here a simpler system in which levels of cyclic AMP decrease in response to light stimulation much as in retina and where all the transduction occurs within a single cell.

The development of the nervous system includes the formation of specific neuronal connections. Some insight into the mechanisms by which these connections are made may be obtained by determining the order in which the elements of a developing system begin to function and examining any changes that may occur in the system shortly after it begins functioning. I have investigated the lateral line of the South African clawed frog, Xenopus laevis and I have classified the afferent system into three elements. First, the transduction mechanism, which includes all components that couple the water vibrations to the release of transmitter from the hair cell. Second, the afferent synapse, at which there is transmitter release and subsequent postsynaptic conductance changes. And third, the afferent axon. I report here experiments suggesting that these elements become functional in the following order: first, the axon can conduct action potentials; second, transmission is established; third, the transduction mechanism can modulate transmitter release. This same sequence has been reported in a system related to the lateral line, the auditory pathway of mammals.

DURING the past few years a great deal of work has been done on the physiology of crustacean limb proprioceptors. The study reported here involved ultrastructural differences between functionally distinct cells in the propodite–dactylopodite (PD) chordotonal organ of Cancer pagurus, and was based on serial transverse sections. Final details will be published elsewhere.

The small size and fragility of primate photoreceptors have impeded study of their neural signals by single cell recording. Although physiological information is meagre, psychophysical experiments indicate remarkable performance, including single photon detection by human rods1. What electrical change results from absorption of a quantum? How do the photoreceptors themselves contribute to the sensitivity and temporal resolution of the primate visual system? We report here a study of transduction in single rods of the cynomolgus monkey. From results of colour matching tests2 and microspectrophotometry3 the light receptors of the cynomolgus monkey are thought to be similar to those of man. We have determined the size and shape of the quantal response, and related parameters of the transduction mechanism. These measurements provide a physiological basis for the single photon detection, saturation, and limited temporal resolution characteristic of rod behaviour in psycho-physical experiments. Spectral sensitivity measurements over the entire visible region extend previous determinations of the spectral absorption of rhodopsin, and allow assessment of the factors that determine human scotopic sensitivity.

THE process of transduction in vision is that mechanism whereby absorption of light by a photoreceptor cell in the retina results in a polarisation of the receptor cell membrane, excitation of the cell, and the eventual generation of a neural impulse to the brain from higher order neurones. It is already well known that the initial and primary process in vision is the absorption of photons by 1l-cis retinal and concomitant isomerisation of the visual pigment to all-trans retinal. We have therefore studied events resulting from this photoisomerisation, events which provide a link between the absorption of light and the excitation of the receptor cell.

Photoreceptor cells in both vertebrates and invertebrates respond to a flash of light with a slow graded change in membrane potential, which is generally depolarising in invertebrates and hyperpolarising in vertebrates. Although some of the early photochemical and biochemical stages of the transduction process have been elucidated in both cases, these reactions are fast compared with the time course of the electrical response, which is typically hundreds of milliseconds long. To explain this slow response in the eye of Limulus, Fuortes and Hodgkin proposed a mechanism in which the light signal passes through a cascade of simple low pass filters. The model was later defined more specifically in terms of a chain of chemical reactions linked together through products and reactants, and has been used with small modifications and different numbers of stages to account for the behaviour of various vertebrate and invertebrate photoreceptors including the fly compound eye. I have now obtained evidence that phototransduction in the fly in small signal conditions involves underdamped resonance behaviour and a significant pure time delay, neither of which can be accounted for by the conventional cascade model.

Movements of the stiff sensory hairs, the stereocilia, which extend outwards from the apical surface of the hair cells in the auditory organ, are thought to have a key role in the process of transduction of sound energy into electrical impulses. Each stereocilium is supported by a paracrystalline axial bundle of actin filaments, which have identical polarities and are extensively cross-linked1–4. Recent immunohistochemical observations on whole mounts of the guinea pig organ of Corti indicated myosin-like immunoreactivity in association with hair cell stereocilia5. Considering the steric situation, it is, however, difficult to imagine any functional interaction between the bundled actin filaments and the assumed myosin molecules. Here we report that in semi-thin, transverse sections of quick-frozen, freeze-dried and plastic-embedded guinea pig organ of Corti, myosin-like immunostaining was restricted to the apical cytoplasm of hair cells and was not detected along the stereocilia. With respect to the immunohistochemical distribution of actin, myosin and the muscular Z-line protein, α-actinin, the apex of hair cells strongly resembles the intestinal epithelial brush border6–7.

ELECTRIC potentials generated by the inner ear in response to sound stimulation were first recorded 40 yr ago1. Studies of these extracellular potentials have led to hypotheses about the role of intracellular potentials in sensory transduction in receptor (hair) cells2,3. It is difficult, however, to infer these intracellular potentials from extracellular measurements alone4–7; measurements of intracellular potentials from hair cells are clearly desirable. Although intracellular potentials in other hair-cell receptors have been reported recently8,9, the few previous attempts to record such potentials in the cochlea have not yielded significant results10. We report here intracellular responses from single hair cells and supporting cells in the cochlea of the alligator lizard Gerrhonotus multicarinatus.

FOR a wide variety of photoreceptor cells, it has been shown that light alters the conductance of the cell membrane, giving rise to the voltage response of the cell1–4. It is generally accepted that the primary event in the photo-excitation process is a light-induced isomerisation of the photopigment chromophore. The mechanisms linking this isomerisation to the conductance change remain unknown. It has been suggested recently that the transduction process may depend on oxidative metabolism. Intracellular recordings from photoreceptors of the honeybee (A pis)5, horseshoe crab (Limulus)5 and barnacle (Balanus)—observed recently by one of the authors (A.M.) in collaboration with R. Lantz—have shown that the light-induced conductance increase during illumination can be reversibly blocked by anoxia. Neither the site nor the mechanism of the action of anoxia on the transduction process is known. The light-induced spectral transitions of the photopigments are known, however, to take place under a variety of conditions including anoxia6–10.

IONIC calcium is important in many membrane functions and the possibility of it being involved in the cochlear transduction processes is supported by the scanty evidence available1. In the analogous lateral line organ of submammalian species, the mechanosensitivity of the hair cells is related directly to the external calcium concentration at their specialised receptor poles2. There is also evidence of interaction with monovalent cation effects2,3. In mammals, the cochlear duct is filled with endolymph, the unusual constitution of which (in the rat d.c. potential + 88 mV, K+ 148 mM, Na+ 0.84 mM) is seemingly maintained entirely by active transport mechanisms4 and is generally conceded to be necessary for normal transduction. We report here that the cochlear endolymph calcium occurs almost entirely in the non-complexed state at a concentration of 3×10−5 M in the rat. This is substantially below the usual extracellular concentration, and must influence the function of the receptor pole membranes and might well affect the role of calcium in the transduction process in the cochlea. The total magnesium concentration we measured was 1×10−5 M.

In taste chemoreception, cyclic adenosine monophosphate (cAMP) appears to be one of the intracellular messengers coupling reception of stimulus to the generation of the response. The recent finding that sweet agents cause a GTP-dependent generation of cAMP poses the question of how this cytosolic messenger acts at the membrane of taste receptor cells. We have shown that cAMP causes a substantial depolarization in these cells. Here we show with whole-cell recordings and inside-out membrane patches that the depolarization caused by cAMP is accounted for by the action of cAMP-dependent protein kinase, which inactivates potassium channels predominantly of 44 pS conductance. Thus, intracellular signalling of the gustatory cells differs from that of olfactory and photoreceptor cells, where cyclic nucleotides control unspecific channels by binding to them rather than by inducing their phosphorylation.

THE function of the two populations of sensory cells in the mammalian inner ear is not well understood. Anatomical evidence indicates that the inner hair cells (IHCs) and the outer hair cells (OHCs) play separate roles in the transduction of acoustic stimuli1. Furthermore, there have been numerous proposals attributing different roles to the two hair cell populations in the production of the various cochlear potentials2–5. On the other hand, theoretical considerations and the interpretation of data from several experiments have led to suggestions of different types of interaction between the OHCs and IHCs6–8.