Olfactory transduction is thought to be initiated by the binding of odorants to specific receptor proteins in the cilia of olfactory receptor cells. The mechanism by which odorant binding could initiate membrane depolarization is unknown, but the recent discovery of an odorant-stimulated adenylate cyclase in purified olfactory cilia suggests that cyclic AMP may serve as an intracellular messenger for olfactory transduction. If so, then there might be a conductance in the ciliary plasma membrane which is controlled by cAMP. Here we report that excised patches of ciliary plasma membrane, obtained from dissociated receptor cells, contain a conductance which is gated directly by cAMP. This conductance resembles the cyclic GMP-gated conductance that mediates phototransduction in rod and cone outer segments, but differs in that it is activated by both cAMP and cGMP. Our data provide a mechanistic basis by which an odorant-stimulated increase in cyclic nucleotide concentration could lead to an increase in membrane conductance and therefore, to membrane depolarization. These data suggest a remarkable similarity between the mechanisms of olfactory and visual transduction and indicate considerable conservation of sensory transduction mechanisms.

Cyclic nucleotide-gated ion channels in olfactory sensory neurons (OSNs) are hypothesized to play a critical role in olfaction. However, it has not been demonstrated that the cAMP signaling is required for olfactory-based behavioral responses, and the contributions of specific adenylyl cyclases to olfaction have not been defined. Here, we report the presence of adenylyl cyclases 2, 3, and 4 in olfactory cilia. To evaluate the role of AC3 in olfactory responses, we disrupted the gene for AC3 in mice. Interestingly, electroolfactogram (EOG) responses stimulated by either cAMP- or inositol 1,4,5-triphosphate- (IP3-) inducing odorants were completely ablated in AC3 mutants, despite the presence of AC2 and AC4 in olfactory cilia. Furthermore, AC3 mutants failed several olfaction-based behavioral tests, indicating that AC3 and cAMP signaling are critical for olfactory-dependent behavior.

We have used gene targeting to examine the role of the G alpha subunit, G(olf), in olfactory signal transduction. Mice homozygous for a null mutation in G(olf) show a striking reduction in the electrophysiological response of primary olfactory sensory neurons to a wide variety of odors. Despite this profound diminution in response to odors, the topographic map of primary sensory projections to the olfactory bulb remains unaltered in G(olf) mutants. Greater than 75% of the G(olf) mutant mice are unable to nurse and die within 2 days after birth. Rare surviving homozygotes mate and are fertile, but mutant females exhibit inadequate maternal behaviors. Surviving homozygous mutant mice also exhibit hyperactive behaviors. These behavioral phenotypes, taken together with the patterns of G(olf) expression, suggest that G(olf) is required for olfactory signal transduction and may also function as an essential signaling molecule more centrally in the brain.

Olfactory neurons transduce the binding of odorants into membrane depolarization. Two intracellular messengers, cyclic AMP (cAMP) and inositol trisphosphate (IP3), are thought to mediate this process, with cAMP generating responses to some odorants and IP3 mediating responses to others. cAMP causes membrane depolarization by activating a cation-selective cyclic nucleotide-gated (CNG) channel. We created a mutant "knockout" mouse lacking functional olfactory CNG channels to assess the roles of different second messenger pathways in olfactory transduction. Using an electrophysiological assay, we find that excitatory responses to both cAMP- and IP3-producing odorants are undetectable in knockout mice. Our results provide direct evidence that the CNG channel subserves excitatory olfactory signal transduction, and further suggest that cAMP is the sole second messenger mediating this process.

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 neurons expressing the same odorant receptor converge to a small number of glomeruli in the olfactory bulb. In turn, mitral and tufted cells receive and relay this information to higher cortical regions. In other sensory systems, correlated neuronal activity is thought to refine synaptic connections during development. We asked whether the pattern of connections between olfactory sensory axons and mitral cell dendrites is affected when odor-evoked signaling is eliminated in mice lacking functional olfactory cyclic nucleotide-gated (CNG) channels. We demonstrate that olfactory sensory axons converge normally in the CNG channel mutant background. We further show that the pruning of mitral cell dendrites, although slowed during development, is ultimately unperturbed in mutant animals. Thus, the olfactory CNG channel-and by inference correlated neural activity--is not required for generating synaptic specificity in the olfactory bulb.

An odor-stimulated adenylate cyclase [ATP pyrophosphate-lyase (cyclizing), EC 4.6.1.1] is thought to mediate olfactory transduction in vertebrates. However, it is not known whether the adenylate cyclase serves this function for all odorants or for only certain classes of odorants. To investigate this question, we have compared the abilities of 35 odorants to stimulate the adenylate cyclase and to elicit an electrophysiological response. We report a strong positive correlation between the magnitude of adenylate cyclase stimulation and the summated electrical response of the olfactory epithelium (electro-olfactogram) evoked by individual odorants. We also show that the adenylate cyclase stimulator forskolin equally attenuates the electro-olfactogram response for all odorants tested. These data provide evidence that the adenylate cyclase mediates transduction for a wide variety of odorants.Olfactory receptor cells respond to odors by an increase in membrane conductance, leading to membrane depolarization and the generation of action potentials (reviewed in refs. 1 and 2). The mechanism(s) by which odors regulate the membrane conductance is not known with certainty; however, a considerable body of evidence indicates that cyclic AMP serves as an intracellular messenger in this process. For example, olfactory cilia contain an odor-stimulated adenylate cyclase (3)(4)(5). In addition, we have described a cyclic nucleotide-gated conductance that could translate the odorinduced rise in cyclic AMP concentration into a conductance increase (6). The involvement of the cyclic AMP pathway is further suggested by the effects of pharmacological agents on the electrical responses to odors, measured with transepithelial voltage and current-recordings (7)(8)(9).A key question concerning the cyclic AMP mechanism is whether it mediates transduction for all odorants or only certain classes of odorants. We (Union Beach, NJ). Forskolin and 1,9-dideoxyforskolin were purchased from Calbiochem.The EOG was measured from excised bullfrog olfactory epithelia, which were mounted in a perfusion chamber. Frogs were decapitated and pithed, and the dorsal surface of the nasal cavity was soaked in Ringer's solution for ca. 2 hr at room temperature. The 2-hr soak in Ringer's solution was found to give EOG responses of larger amplitudes. The olfactory epithelium was then dissected away from the remaining portion of the skull and mounted, apical (ciliated) surface up, in a Plexiglas recording chamber (15), and the edges were sealed with vacuum grease. The apical surface was continuously superfused with Ringer's solution via a plug, which sealed the chamber from above. The plug had a central port for delivery of fresh Ringer's solution (flow rate 2.0 ml/min) and a concentric array of peripheral exit ports to maintain an approximately radial flow pattern. The exposed area of the epithelium was 12 mm2, which utilized most ofthe dissected tissue. The Ringer's solution contained 100 mM NaCl, 2.5 mM KCl, 10 mM NaHCO3, 1 mM CaCl2, 1 mM MgSO4, an...

SUMMARY1. Flash photolysis of caged cyclic nucleotides was used to examine the contribution of the ciliary cyclic nucleotide-gated conductance to olfactory transduction in the tiger salamander. Brief illumination of solitary olfactory receptor cells loaded with 100 /aM caged cyclic AMP caused a large inward current (peak amplitude 355 + 200 pA; mean + S.D. for eleven cells) under whole-cell voltage clamp at -50 mV.2. The photolysis response was initiated after a latency of 4-12 ms, whereas an odorant response of identical amplitude had a latency of several hundred milliseconds. The amplitudes of both responses exhibited almost identical voltage dependence between -50 and +25 mV, with both reversing near 0 mV. The time courses of the falling phases of odorant and photolysis responses also exhibited similar voltage dependence, both being prolonged at positive voltages.3. Photolysis of caged cyclic GMP activated a current similar in amplitude and time course to that produced by photolysis of caged cyclic AMP.4. When the flash was spatially limited to the cilia, the amplitude and duration of the photolysis response increased linearly with the length of the cilia illuminated (for cilia not longer than 30-40 ,um) while the latency remained constant at 4-12 ms. The increase in duration was described semi-quantitatively by a model which incorporated diffusion and saturable hydrolysis of cyclic AMP. When the flash was limited to the soma or proximal dendrite, the response latency was proportional to the square of the distance between the illuminated region and the cilia.5. Dialysis of cells with 500 #M cyclic AMP from a whole-cell electrode under voltage clamp activated a large transient inward current. Simultaneous suction electrode recording showed that this current originated almost entirely from the ciliary membrane. The density of cyclic nucleotide-gated channels was estimated to be 800-fold higher in the cilia than in the soma.6. Summation of simultaneous odorant and photolysis responses was non-linear, the flash-induced current being enhanced during a small odorant response and attenuated during a large odorant response. Summation of two photolysis responses was similarly non-linear. The data were consistent with odorant stimuli and cyclic AMP both activating a common cyclic nucleotide-gated conductance with a Hill NIS 1155 G. LOWE AND G. H. GOLD coefficient, n, of 2-04-4. For n = 2-5, the basal cyclic AMP concentration was estimated to be less than 20 % of the K1, which predicts a basal current of 5-8 pA, less than 2 % of the maximum.7. No effect on membrane current was observed during dialysis of cells with up to 100 /tM inositol 1,4,5-trisphosphate (1P3) or photolysis of cells loaded with 100 fM caged 'P3. 8. The above results are consistent with the hypothesis that olfactory transduction is mediated by an odorant-induced increase in cyclic nucleotide concentration which depolarizes the cell by acting directly on the ciliary cyclic nucleotide-gated conductance.