Elementary Response of Olfactory Receptor Neurons to Odorants
This paper describes experiments testing the response of frogs' olfactory neurons to odors, to see how they amplify the signal of an odor. This is compared to the visual amplification of single photons (through a phototransduction cascade). The two mechanisms are shown to be quite different at the level of individual neurons. The frog response is estimated for two odorants with very different response efficiencies, and the statistics of the results interpreted to give insight into the underlying process.
Questions about how to maintain olfactory sensitivity under the observed model are raised, and some theoretical explanations are proposed that involve synthesis across many neurons, but without related experiment.
Goals and Methods
This experiment aimed to probe the relation between responses of frog olfactory receptors to odor stimuli, and the intensity and duration of those stimuli. Working from knowledge of the specific proteins and cation channels involved in olfactory reception, the authors set out to develop a technique to stimulate isolated olfactory neurons of a frog, in Ringer solution (to stimulate a physical environment), while measuring resulting membrane currents.
This method assumed that there was some smallest quantum of response to odor detection, and that any observed response was some combination of these elementary (or 'unitary') units of response. This followed quantal analysis originally developed by Castillo and Katz in 1954. To do this required avoiding non-linear variations away from those simplest responses.
Olfactory adaptation was minimized: cells were known to adapt in the presence of Ca2+, leading to a non-linear dose-response relation. The experiment minimized this concentration, while confirming its impact on the results by repeating its tests in solutions with different concentrations of Ca2+ (replaced in low-Ca solutions with Mg2+) : tests were run with cells in regular Ringer solution, 20μM Ca2+, and 100nM Ca2+ solution.
Two odorants that ORNs were known to resopnd to very differently (acetophenone and cineole) were used to get a sense of response variation across different regimes of odorant binding and saturation. the dwell-time of the two odorants different by roughly a magnitude. The results were analyzed to see how long the receptor-odorant complexes lasted, and what could be understood about the activation of G proteins during the process. This was compared to the better known process of rhodopsin amplification cascade in which photoisomerized rhodopsin triggers a cascade of transducins before being shutoff by phosphorylation.
Concentration of odorants was varied both by using increasingly long pulses of low concentration, and by using increasingly intense pulses of fixed concentration odors, with similar results.
The unitary amplitude for a given response was measured as the transient peak of the response, assuming a Poisson distribution. Membrane currents were measured using the suction-pipette method. Where appropriate observations matched well to a least-squares fit to the Hill equation, which applies to such situations where cooperative binding is expected.
Quantal analysis and linear extrapolation from lower-concentration solutions allows the estimate of the unitary response in normal Ringer solution, even though the macroscopic response is nonlinear.
Results and Analysis
The unitary current stimulated by each and the overall response kinetics were similar for both odorants across five different cells. This unit of response varied primarily with the concentration of Ca2+ in the solution, and was consistent across different cells, though it is expected that every cell has different binding receptor.
A simple check for whether a quantal / unitary response is detected is the linearity of the relationship between concentration and response. If many different binding events start to overlap spatially, transduction effects would not be linear, so the observed response would be supralinear. And indeed in low-Ca2+ environments a linear response was seen, at both 100nM and 20μM Ca2+, up to a certain odorant concentration. The unitary current in Ringer solution is estimated to be roughly 0.03pA, 100 times smaller than previously measured by Menini, et al.
The response was analyzed for possible amplification cascades involved in converting those stimuli to signals from GTP-binding [G] proteins. The dwell-time of odorants was found to be quite small, under 1ms. Moreover the response time to odors was fast and steady: the linear increase in total response to a pulse of fixed molarity but different duration had a projected time-intercept of 0, indicating that some response was taking place very soon after the odor was released.
Each bound odorant seemed to have a low probability of activating a single G protein, which in turn has a low probability of activating more than one adenylyl cyclase molecule. This is projected from the overall linearity of response across a wide range of odor saturations. Even when all receptors were expected to be bound to cineole, the highly effective odorant, the total response amplitude continued to increase linearly with pulse duration.
These results were compared to known response patterns of optical neurons to light at low levels of stimulation. In contrast, rod phototransduction acts through photoisomerized rhodopsin activating a transducin cascade. This experiment suggests that such a mechanism may not exist for olfaction, and that the odorant-receptor complexes may generally be too short lived to trigger any mechanisms that might exist. Moreover, each receptor, once activated, did not seem to be inactivated for at least 100ms.
Some further understanding of the workings of receptors once they bind to an odorant are needed. The authors suggest that they may integrate responses when an odorant binds to them repeatedly without being phosphorylated. Moreover, the glomerulus in the olfactory bulb of the frog can integrate signals from all ORNs, perhaps separately integrating them for each receptor.
Unlike with photons, each odorant can go on to bind many times at the same site or at different sites. And unlike with an eye of fixed size, which is difficult to modify within the structure of its host, the surface area of olfactory epithelium can readily be expanded. This could have a proportional improvement in olfactory sensitivity.
- Measurement process: Quantal Components of the End-Plate Potential. Castillo and Katz, Journal of Physiology (1954)
- S. J. Kleene, Neuroscience 66, 1001 (1995)
Theoretical and practical relevance:
This is a study of low-level components of one olfactory system that suggest the sorts of sensory models which may be studied in further experiments. It provides a benchmark for similar work in other animals and at other levels in the frog.
The unitary response of individual ORNs was calculated to be 100x smaller than previously reported. (0.03 pA)