Current views of odorant discrimination by the mammalian olfactory system suggest that the piriform cortex serves as a site of odor object synthesis. Given the enormous number of odorant feature combinations possible in nature, however, it seems unlikely that cortical synthetic receptive fields (RFs) are innate but rather require experience for their formation. The present experiment addressed two issues. First, we made a direct comparison of mitral/tufted cell and anterior piriform cortex (aPCX) neuron abilities to discriminate odorant mixtures from their components to further test whether aPCX neurons can treat collections of features different from the features themselves (synthetic coding). Second, we attempted to determine the minimum duration of experience necessary for formation of cortical synthetic RFs. Single-unit recordings were made from mitral/tufted cells and aPCX layer II/III neurons in urethan-anesthetized rats. Cross-habituation between novel binary mixtures and their novel components was used to determine odor discrimination abilities. The results suggest that after ≥50 s of experience with a binary mixture, aPCX neurons can discriminate the mixture from its components, whereas mitral/tufted cells cannot. However, when limited to 10 s of experience with the mixture, aPCX neurons appear similar to mitral/tufted cells and do not discriminate mixtures from components. These results suggest experience-dependent synthetic processing in aPCX and suggest an important role for perceptual learning in normal odor discrimination.
Current evidence supports the view that odorant discrimination is mediated peripherally by a large set of olfactory receptor proteins that recognize and bind not complete odorant molecules but rather specific molecular features (Mori et al. 1999). Thus individual odorants are represented combinatorially in initial stages of olfactory processing. Central olfactory structures then make at least two additions to this initial feature extraction critical for odorant discrimination, namely enhancement of contrast between odorant features and synthesis or binding of multiple features into odorant objects. Lateral and feedback inhibition and excitation combine to enhance signal-to-noise ratios in olfactory bulb output neuron (mitral/tufted cell) spike trains as well as enhance contrast between similar molecular features (Luo and Katz 2001; Yokoi et al. 1995).
In mammals, synthesis of odorant features into odorant objects has been hypothesized to occur, at least in part, through anatomical convergence of mitral cells conveying activity induced by multiple odorant features onto individual target cells in the piriform cortex (Granger and Lynch 1991; Haberly 2001; Zou et al. 2001). Individual mitral cells terminate in small patches within anterior piriform cortex (aPCX) (Buonviso et al. 1991; Ojima et al. 1984). Mitral cells receiving input from a phenotypically specific group of receptor neurons terminate in patches that overlap with patches from mitral cells conveying different receptor neuron input (Zou et al. 2001). This spatial convergence can be functionally enhanced by dynamic temporal synchrony in mitral cell spike trains in both vertebrates (Buonviso et al. 1992; Kashiwadani et al. 1999) and invertebrates (Christensen et al. 2000; Laurent et al. 2001; Stopfer et al. 1997).
Thus single piriform pyramidal cells may receive synchronous input from mitral cells conveying processed information from many different olfactory receptor types. Piriform cortical pyramidal cells, in turn, make extensive associational connections throughout the piriform cortex, back to the olfactory bulb as well as to other cortical structures (Haberly 2001; Johnson et al. 2000). The relatively diffuse afferent input combined with a broad, extensive intra-cortical association fiber system creates a highly combinatorial network, ideal for synthetic processing of complex feature ensembles (Haberly 2001).
However, given the wide range of odorants and odorant mixtures that animals can discriminate, it is unlikely that synthetic coding is due to innate hard-wiring but rather reflects an experience-dependent learning process that allows synthesis of novel co-occurring features into odorant objects. Both lateral olfactory tract (LOT) afferent synapses and association fiber synapses in the piriform cortex express activity-dependent plasticity (Hasselmo and Barkai 1995; Jung et al. 1990; Kanter and Haberly 1990; Litaudon et al. 1997; Roman et al. 1987; Saar et al. 2002; Stripling et al. 1991; Wilson 1998b). Thus repeated co-occurrence of synaptic activity evoked by specific combinations of odorant features could result in a functional synthesis of those features such that subsequent exposure to a partially degraded signal could still evoke a “complete” odor sensation and recognition (Barkai et al. 1994). A similar process may be involved in inferotemporal visual cortex in the formation of complex receptive fields for visual objects such as faces (Miyashita and Hayashi 2000; Rolls 2000; Tanaka 2000) and appears to occur within a few seconds of exposure to the visual object (Tovee et al. 1996).
Previous work from our lab supports the view of experience-dependent odor discrimination by aPCX neurons. A cross-habituation paradigm has been used to assess odor discrimination by single neurons in the main olfactory bulb and aPCX. Previous work has demonstrated that a 50-s exposure to one odorant within a mitral cell's odorant receptive field produces cross-habituation to other similar and dissimilar odorants (Fletcher and Wilson 2002b; Wilson 1998a, 2000b), suggesting poor ability by these neurons to discriminate odorants within their receptive fields. In contrast, following the same 50-s exposure, aPCX neurons show significantly less cross-habituation, suggesting that they are able to discriminate between those odorants (Wilson 2000a,b). Muscarinic receptor blockade within the aPCX during the odorant exposure, however, prevents the enhanced odor discrimination by aPCX neurons, resulting strong cross-habituation between odorants within aPCX receptive fields similar to mitral cells (Wilson 2001a). Given that acetylcholine modulates piriform cortical synaptic plasticity (Hasselmo and Barkai 1995; Hasselmo et al. 1992; Patil et al. 1998), these latter results suggest an important role for plasticity in cortical odorant discrimination.
The present report further examines the role of experience in odor-coding properties of aPCX neurons. It was hypothesized that if synaptic plasticity was required for cortical synthesis of novel odorant features, then there should be a minimum duration of odorant exposure (familiarization) required for that plasticity to occur. Without sufficient exposure, aPCX neurons should function similar to feature detecting mitral/tufted cells. Specifically, it was hypothesized that odorant discrimination ability by aPCX neurons should increase as the duration of previous exposure to those odorants increases.
Male Long-Evans hooded rats (150 – 450 g), obtained from Harlan Lab Animals, were used as subjects. Animals were housed in polypropylene cages lined with wood chips. Food and water were available ad libitum. Lights were maintained on a 12:12 light: dark cycle with testing occurring during the light portion of the cycle.
The design and rationale behind this experiment are shown in Fig. 1 (intracellular data from Best and Wilson 2002). Familiarization with an odorant produces a rapid decrease in both odor-evoked spiking and odor-evoked postsynaptic potentials in the aPCX (Bouret and Sara 2002; McCollum et al. 1991; Wilson 1998a). While odor-evoked spiking is generally completely abolished, subthreshold postsynaptic activity is reduced but often partially maintained (Wilson 1998a,b). Previous work has demonstrated that after 50 s of exposure to a novel odorant, aPCX neurons can discriminate that odorant from similar odorants within their receptive field, whereas mitral/tufted cells cannot (Wilson 2000a,b). The present experiment examined aPCX single-unit discrimination between binary mixtures and their components after either 10 or 50 s of familiarization, using a cross-habituation paradigm. The 10-s time point was selected to provide sufficient self-habituation to allow use of the cross-habituation paradigm yet maintain a short exposure duration. The aPCX results are compared with discrimination of the same stimuli by mitral/tufted cells after 50-s exposure.
Recording and odorant stimulation
Details of single-unit recording and odor-response characterization techniques for mitral/tufted and layer II/III aPCX neurons have been reported in detail elsewhere (Wilson 1998a). Briefly, animals were anesthetized with urethan (1.5 g/kg) and placed in a stereotaxic apparatus for electrode placement. Animals were freely breathing with the respiratory cycle monitored through a piezoelectric device strapped to the chest. The single-unit nature of the recordings were verified by at least a 2-ms refractory period in interval histograms. Mitral/tufted cells were identified by antidromic stimulation of the LOT and layer II/III aPCX neurons were identified by LOT-evoked synaptic responses and/or histological confirmation.
Odorants were delivered with a flow-dilution olfactometer, with a constant, 1 liters per minute (LPM) flow of charcoal-filtered, humidified air presented 1–2 cm from the animal's nose. Saturated odor vapor was added to the clean airstream via computer-controlled solenoids to produce an approximate dilution of 1:10 of saturated vapor. Odor stimulus onset was triggered off the respiratory cycle to coincide with the transition from inhalation to exhalation, and test stimulus duration was 2 s. Given that urethan-anesthetized rats respire at ∼2 Hz, this stimulus duration corresponds to four inhalations. Stimuli delivered to produce familiarization were either 10 or 50 s in duration (∼20 and 100 inhalations, respectively). Odorants used were all novel to the animals, and particular odorants were only used for one experimental cell for each animal. Stimuli included molecularly dissimilar odorants, peppermint (McCormick), isoamyl acetate (Sigma), and eugenol (Sigma); molecularly similar odorants, pentane, heptane, and nonane (all from Sigma); and, to test the synthetic nature of responses, binary mixtures of odorants within the dissimilar or similar odorant set. To deliver odorants in binary mixtures, airflow was equally divided between the two odorants, thus total volume of odorant mixture was the same as for a single stimulus (see Wilson 2000a).
Both mitral/tufted and aPCX neuron responses to odorants were quantified as the difference in number of spikes evoked during the 2-s stimulus compared with a 2-s prestimulus period. An individual cell was tested with either molecularly similar or dissimilar odorants and at least one of their binary mixtures. The experimental protocol consisted of determining the cell's responses to the test odorants twice (including both single components and their mixtures) with at least a 60-s inter-stimulus interval. Then a binary mixture familiarization stimulus was presented lasting either 10 or 50 s. Within 30 s after the termination of the familiarization stimulus, the responses to test stimuli were reassessed to examine the magnitude of habituation and cross-habituation between the mixtures and their components. Finally, test odorants were reapplied 2–15 min later to confirm at least partial recovery of habituated responses. Response magnitudes to test stimuli were expressed as a percent of prefamiliarization magnitudes.
For aPCX neurons, comparisons were made between odorants and familiarization duration using a three-way ANOVA (odorant class × stimulus duration × component/mixture odorant) and post hoc comparisons. For mitral/tufted cells, levels of self-versus cross-habituation were compared with a paired t-test.
Data from a total of 27 mitral tufted cells from 13 animals and 77 aPCX neurons from 44 animals are included in the analyses. To ensure minimal previous experience with the odorants, specific odorants were used for only one experiment/animal, and in many cases, data from only one cell/animal were collected.
While even simple odorants are believed to be treated by the peripheral olfactory system as being composed of multiple features, and thus requiring synthesis for object recognition, the present experiment utilized odorant mixtures to more directly test issues of synthetic processing. Most aPCX neurons responding to novel single odorants also responded to novel binary mixtures of those components (Fig. 2). Although stimuli were not equated for intensity, responses to novel mixtures included both mixture addition with the response magnitude to the mixture at or above the algebraic summation of the response to the components, and mixture suppression with the response magnitude to the mixture below the algebraic summation of the response to the components. Mixture suppression occurred most commonly with no difference in probability of addition and suppression between mitral/tufted cells and aPCX neurons—72% of mitral/tufted cells showed mixture suppression and 28% of mitral/tufted cells showed mixture addition. The proportions were identical for the aPCX.
Figure 3A shows a representative example of a mitral/tufted cell single-unit response to a novel binary mixture of molecularly dissimilar odorants and one of its components. Exposure to the binary mixture for 50 s (not including initial 2-s test stimuli) produced substantial habituation in the response to the mixture as well as dramatic cross-habituation to the component. The response of an aPCX neuron to the same novel binary mixture and one of its components is shown in Fig. 3C. In contrast to the mitral/tufted cell, after 50 s of exposure to the binary mixture the aPCX neuron shows minimal cross-habituation to the component despite marked habituation to the mixture.
Figure 3B shows a representative aPCX single-unit response to a novel binary mixture and one of its components before and after only 10 s of mixture exposure. The 10-s exposure period (not including initial 2-s test stimuli) was sufficient to produce marked self-habituation to the mixture but also resulted in marked cross-habituation to the component. This strong cross-habituation is in contrast to the amount of cross-habituation seen in aPCX after 50 s of exposure (Fig. 3C) but is similar to that observed in mitral/tufted cells (Fig. 3A).
Figure 4 shows mean data for mitral/tufted cell self- and cross-habituation after 50 s of exposure to a binary mixture. There was no difference in cross-habituation between similar and dissimilar odorants, so data from these two odorant classes were combined. There was no significant difference between the levels of self- and cross-habituation between binary mixtures and their components in mitral/tufted cells [paired t-test, t(26) = 1.29, N.S.]. Thus habituation of a mitral/tufted cell to a binary mixture produced comparable levels of cross-habituation to the mixture components, suggesting an inability of these cells to make this discrimination.
Figure 5 shows mean data for aPCX neuron self- and cross-habituation after 10 s (n = 18 cells for similar odorants and 19 cells for dissimilar mixtures) and 50 s (n = 20 cells for similar odorants and 20 cells for dissimilar mixtures) of exposure to a binary mixture. There was a main effect of odorant class with alkanes showing greater habituation than combinations of peppermint, isoamyl acetate or eugenol [F(1,146) = 12.46, P < 0.01] and a significant main effect of stimulus duration with 10 s of exposure producing slightly less habituation than 50 s of exposure [F(1,146) = 4.22, P < 0.05]. Importantly, there was a significant interaction between stimulus duration and component/mixture odorant variables [F (1,146) = 9.58, P < 0.01], suggesting that the duration of exposure to the mixture stimulus affected the magnitude of cross-habituation to the components. This was expressed for both mixtures of similar and dissimilar odorants [nonsignificant 3-way interaction, F(1,146) = 0.16].
Combined, these results show that both mitral/tufted cells exposed to an odorant mixture for 50 s and aPCX neurons exposed to a mixture for 10 s display significant cross-habituation to the components of that mixture, whereas aPCX neurons exposed to a binary odorant for 50 s showed significantly less cross-habituation to the components. Thus given sufficient (>10 s, <50 s) exposure to a novel binary mixture, aPCX neurons can discriminate that mixture from its components while mitral/tufted cells cannot.
Analyses of individual cell responses
An examination of discrimination performance of individual cells revealed that for aPCX neurons exposed to the binary mixture for 50 s, 15 of 39 cells (38.5%) had mean responses to the mixture components that were >50% of preexposure levels. In contrast, for aPCX neurons exposed to the binary mixture for 10 s, only 4 of 38 cells (10.5%) had mean responses to the mixture components that were >50% of preexposure levels. This difference is significant [χ2(1) = 8.21, P < 0.01] and supports the preceding data suggesting enhanced discrimination by aPCX neurons of mixtures from components as the duration of exposure increases.
In a post hoc analysis, the individual data were further divided according to the nature of the cell's initial response to the mixture, i.e., whether the cell demonstrated mixture suppression or mixture addition. As noted in the preceding text, in this data set as a whole, 72% of aPCX neurons showed mixture suppression and 28% showed mixture addition relative to their responses to the components during initial response mapping. Cells showing mixture addition did not differ from cells showing mixture suppression in terms of the amount of cross-habituation they expressed after 10 s of mixture exposure. Thus 1 of 12 cells (8%) showing mixture addition had mean responses to the mixture components after 10 s of mixture exposure that were >50% of preexposure levels, while 3 of 25 cells (12%) showing mixture suppression had mean responses to the mixture components >50% of preexposure levels. This suggests that most cells, regardless of the nature of their original response to the mixture, were poor at discriminating the mixture from its components after only 10 s of exposure.
In contrast, while 50 s of mixture exposure improved discrimination by all cells compared with 10-s exposure, cells showing mixture addition were substantially better at discriminating the mixture from its components after a 50-s mixture exposure than cells showing mixture suppression. Of the 39 cells tested in the 50-s mixture exposure (collapsed across similar and dissimilar mixtures), 9 showed mixture addition and 30 showed mixture suppression. Of these cells, 7 of the 9 cells (78%) showing mixture addition had mean responses to the mixture components after 50 s of mixture exposure that were >50% of preexposure levels, whereas 8 of 30 cells (27%) showing mixture suppression had mean responses to the mixture components >50% of preexposure levels. This suggests that cells showing mixture addition to novel odor mixtures were better able to learn to discriminate those mixtures from their components after 50 s of exposure than cells showing mixture suppression, although it should be emphasized that these analyses are based on low numbers of cells per group.
The present results demonstrate that <1 min (∼100 inhalations) of familiarization with a previously novel odorant mixture is sufficient for the piriform cortex to subsequently discriminate that mixture from its components. However, 10 s of exposure to a novel mixture is insufficient as demonstrated by significant cross-habituation between the mixture and its components by aPCX neurons. Mitral/tufted cells in contrast show cross-habituation between the mixture and its components even after 50 s of exposure. These results are consistent with the hypothesis that mitral/tufted cells function as feature detectors with receptive fields that include odorants containing the feature to which that cell is responsive. Thus habituation to that feature reduces mitral/tufted cell responses throughout their receptive fields. This is similar to the conclusions reached by Giraudet et al. (2002) in their analysis of rat mitral cell temporal responses to binary odorant mixtures. In addition, the results are consistent with the hypothesis that aPCX neurons, given sufficient time for synaptic plasticity to occur, have synthetic receptive fields, where each odorant within the receptive field is treated as a relatively independent odorant object. With insufficient exposure, or with the effects of that exposure pharmacologically disrupted (Wilson 2001a), aPCX neurons function as simple coincidence detectors of multiple feature input and show strong cross-habituation to odorants within their receptive fields as feature input is reduced.
The experience-induced increase in odorant discrimination by aPCX neurons could be due to plasticity within the olfactory bulb and/or plasticity within aPCX. Given the short duration of exposure required for these effects and the nature of the effects, we propose that changes at the receptor sheet are not involved.
Rapid changes within the olfactory bulb that could contribute to the observed simple exposure-induced enhancement of odor discrimination by aPCX neurons include fine tuning of individual glomerular (Spors and Grinvald 2002) or mitral/tufted cell receptive fields (Buonviso and Chaput 2000; Fletcher and Wilson 2002b), changes in synchrony or spatio-temporal patterning of mitral/tufted cell ensembles (Christensen et al. 2000; Friedrich and Laurent 2001; Grajski and Freeman 1989; Laurent et al. 2001; Ravel et al. 2003; Stopfer and Laurent 1999), and/or changes in descending cortical feedback to the olfactory bulb (Gray and Skinner 1988; Kay and Freeman 1998).
Odorant exposure produces an odorant-specific spatial pattern of glomerular activation (Guthrie et al. 1993; Johnson et al. 1998; Rubin and Katz 1999), presumably reflecting the unique combination of olfactory receptor neurons activated by a particular odorant. Recent functional imaging data in rodents suggest that this spatial pattern is dynamic, decreasing in spatial extent and/or increasing in focus over the duration of an extended odorant stimulus (Spors and Grinvald 2002). Single-unit recordings suggest a similar dynamic shaping of rat mitral/tufted cell receptive fields induced by simple odorant exposure (Fletcher and Wilson 2002b). In invertebrates, olfactory lobe projection neurons undergo a similar dynamic re-organization of ensemble activity over the course of prolonged or repeated odorant stimulation, with spike timing becoming more precise with increasing odorant familiarity (Stopfer and Laurent 1999). These changes in olfactory bulb circuit function may reflect experience-induced changes in synaptic efficacy between receptor neurons and projection neurons and/or interneurons and projection neurons.
In vertebrates, changes in olfactory bulb output patterns may also reflect plasticity in descending cortical-olfactory bulb projections (Gray and Skinner 1988; Kay and Freeman 1998; Potter and Chorover 1976). Similar to thalamocortical sensory systems, the vertebrate olfactory bulb receives a massive feedback projection from the olfactory cortex. As odorants and feature combinations are processed by the piriform cortex (see following text) cortical feedback, largely targeted at inhibitory granule cells in the olfactory bulb, could help further enhance processing of odorant features at the bulb level, similar to the role of cortico-thalamic projections in other sensory systems (Ghazanfar et al. 2001; Murphy et al. 1999).
These potential adjustments in olfactory bulb odorant feature coding and temporal synchrony of olfactory bulb output could make a direct impact on stimulus discrimination even assuming the cortex functions as a simple coincidence detector. However, we hypothesize that these experience-dependent changes in activity of cortical afferents, in addition to enhancing odorant identification during the stimulus, also promote synaptic plasticity within cortical association fibers that allow a more permanent record of odorant experience (Haberly 2001). In fact, disruption of the normal cholinergic modulation of association fiber synaptic plasticity (Hasselmo and Barkai 1995; Hasselmo et al. 1992; Patil et al. 1998) by scopolamine application limited to the aPCX alone, is sufficient to disrupt the normally enhanced odorant discrimination by aPCX neurons compared with mitral/tufted cells (Wilson 2001a). Thus changes occurring within the piriform cortex, or originating within the piriform cortex and projected back to the olfactory bulb, appear to be critical for enhanced odor discrimination and synthetic coding in the aPCX.
As outlined in detail elsewhere (Haberly 2001; Hasselmo et al. 1990; Wilson 2001b), we propose that the synthesis of co-occurring odorant features occurs largely through plasticity of association fiber synapses. Several observations support this locus. First, association fibers show more robust associative synaptic plasticity (e.g., long-term potentiation) than do LOT synapses (Kanter and Haberly 1990; Roman et al. 1987; Saar et al. 2002; Stripling et al. 1991). In fact, LOT synapses show a marked, homosynaptic depression after either prolonged (50 s) odor exposure in vivo or tetanic stimulation of LOT fibers (30 –50 s) in vitro (Best and Wilson 2002). This LOT synaptic depression is hypothesized to underlie cortical adaptation to odors (Best and Wilson 2002; Wilson 1998b). However, subthreshold odor-evoked postsynaptic potentials are often observed beyond this time window (Fig. 1) (Best and Wilson 2002; Wilson 1998b) and may represent maintained association fiber activity, which could then allow for plasticity underlying the changes in discrimination observed here. Cholinergic modulation of synaptic potentials (Hasselmo and Barkai 1995; Hasselmo et al. 1992; Linster et al. 1999; Patil et al. 1998) and neuronal adaptation (Barkai and Hasselmo 1994) could limit spread of the induced synaptic plasticity, further enhancing specificity of the stored pattern.
A second reason for focusing on association fibers is that although afferent fibers conveying information about activity from different olfactory receptors converge within the aPCX, presumably onto single neurons (Zou et al. 2001), cortical association fibers show much more widely distributed projections throughout and even beyond the piriform cortex (Datiche et al. 1996; Johnson et al. 2000). In fact, odor-evoked spatial patterns of c-fos labeling more closely match the widespread association fiber circuitry than the patchy afferent terminations (Illig and Haberly 2003). Thus the potential for feature convergence and object synthesis via association fibers appears greater than for cortical afferents. Again, this is similar to thalamocortical systems such as vision, where association fibers are believed to play a larger role in feature synthesis and object completion than cortical afferents (e.g., Chance et al. 1999; Crist et al. 2001).
Strengthening of association fiber synapses based on temporal convergence of co-occurring odorant features could allow synthetic coding of familiar odorants. Synthetic coding of odors as unique objects should enhance discrimination of similar objects as well as enhance recognition of those objects even if input is partially degraded (Barkai et al. 1994; Haberly 2001; Hasselmo et al. 1992) as demonstrated here.
It is proposed that the rapid change in cortical odorant processing shown here and its underlying neural plasticity are a fundamental, critical feature of even basic odorant discrimination. That is, implicit olfactory perceptual learning is required for odor discrimination, perhaps particularly for similar odorants. In humans, a reliance of odor discrimination on memory is evidenced by the impaired olfactory discrimination associated with memory disorders in humans (e.g., Mair et al. 1980). In fact, patient HM, whose bilateral temporal lobe resection included olfactory cortex, showed severe impairment in discrimination of equal intensity odorants (Eichenbaum et al. 1983). In rats, discrimination of similar odorants can be enhanced by previous experience with those odorants (Fletcher and Wilson 2002a; Linster et al. 2002). For example, naïve rats do not discriminate ethyl esters differing by a single methyl group but can make this discrimination 24 h after exposure to the odorants (Fletcher and Wilson 2002a). The cortical plasticity described here may underlie this behavioral olfactory perceptual learning. In fact, both enhanced aPCX odorant discrimination (Wilson 2001a) and olfactory perceptual learning (Fletcher and Wilson 2002a) are disrupted by scopolamine.
These results also suggest that experience could enhance mixture analysis and, similarly, discrimination of a target odorant against an odorous background. In fact, human data suggest that mixture analysis—identification of components within an odor mixture— can be enhanced by past experience, although the ability to correctly identify components of mixtures rapidly decreases as mixtures exceed three to four components regardless of experience (Livermore and Laing 1996). This limited mixture analysis may reflect an upper limit on cortical isolation of simultaneous, independent odor object representations as well as a predisposition of the olfactory system toward synthetic processing. Further evidence of a strong synthetic component to odor processing comes from recent work showing that odors experienced together can acquire the perceptual characteristics of each other very rapidly (Stevenson 2001).
Discrimination of odorants by neurons in the aPCX is enhanced by brief (<50 s) experience with those odorants. Without sufficient experience, aPCX neurons are no better than mitral/tufted cells at discriminating mixtures from their components. These results are consistent with a feature detecting role for mitral/tufted cells and an experience-dependent, synthetic processing role for aPCX neurons in olfactory processing. Together with previous findings, the results are also consistent with the view that implicit memory (perceptual learning) based on cortical plasticity is necessary for odor discrimination (Wilson and Stevenson 2003) and that simple spatial convergence and temporal coincidence of cortical afferents is insufficient to account for discrimination of complex odorants.
This work was supported by National Institute on Deafness and Other Communication Disorders Grant DC-03906 and a grant from the Oklahoma Center for the Advancement of Science and Technology to D. A.Wilson.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 by the American Physiological Society