| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Anatomy and Cell Biology, Emory University School of Medicine, Atlanta, Georgia 30322-3030
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Scott, John W., Donna E. Shannon, Jeff Charpentier, Lisa M. Davis, and Craig Kaplan. Spatially organized response zones in rat olfactory epithelium. J. Neurophysiol. 77: 1950-1962, 1997. Electroolfactogram recordings were made with a four-electrode assembly from the olfactory epithelium overlying the endoturbinate bones facing the nasal septum. In this study we tested whether odors of different chemical structures produce maximal responses along longitudinally oriented regions following the olfactory receptor gene expression zones described in the literature. The distribution of responses along the dorsal-to-ventral direction of this epithelium (i.e., across the expression zones) was tested in two types of experiments. In one, four electrodes were fixed along the dorsal-to-ventral axis of one turbinate bone. In the other, four electrodes were placed in corresponding positions on four turbinate bones and moved together up toward the top of the bone. These experiments compared the odorants limonene and 
-terpinene, which are simple hydrocarbons, with carvone and menthone, which differ from the hydrocarbons by the presence of ketone groups. All responses were standardized to an amyl acetate or ethyl butyrate standard. The responses to limonene and -terpinene were often larger for the ventral electrodes. The responses to carvone and menthone were largest for the dorsal electrodes. Intermediate electrodes gave responses that were intermediate in amplitude for these odors. The possibility that direction of air flow caused the observed response distributions was directly tested in experiments with odor nozzles placed in two positions. The relatively larger dorsal responses to carvone and relatively larger ventral responses to limonene were present despite odor nozzle position. We conclude that the responses to this set of odors vary systematically in a fashion parallel to the four gene expression zones. The odorant property that governs this response distribution may be related to the presence of oxygen-containing functional groups. Certain odors evoked larger responses at the intermediate electrode sites than at other sites. Cineole was the best example of this effect. This observation shows that not all oxygen-containing functional groups produce the same effect. Although we cannot exclude other possible mechanisms, these three response gradients may be produced by the four receptor expression zones described for many of the putative olfactory receptor genes. Therefore many of the receptors in each zone may share common properties. It remains to be determined whether this zonal input is significant in central odor processing. However, the correlation of odor chemical properties with the structure of receptor molecules in each zone may provide significant leads to structure-function relationships in vertebrate olfaction.
Recent developments have inspired us to reinvestigate the question of distribution of sensitivity to odors on the olfactory epithelium. The particular inspiration for this investigation was the observation that many of the recently discovered olfactory receptor genes (Buck and Axel 1991 Male Sprague-Dawley rats (375-475 g) were anesthetized with ketamine (87 mg/kg), xylazine (13 mg/kg), and butorphanol (0.3 mg/kg). A single tracheal cannula was inserted for respiration. To give the widest exposure of the medial wall of the left endoturbinates, the right eye, skin over the nasal cavity, and zygomatic arch were removed. A large window of bone over the anterior part of the orbital surface, up to and including the base of the zygomatic arch, was removed. When the bone was thin enough to remove, a cotton pad soaked in Ringer solution was placed over the surgical site and the animal was killed. The head was elevated slightly and the animal was allowed to sit for a period of 20-30 min to allow blood flow to stop. In some later animals, those of experiment 1B, the animals were killed before the beginning of surgery. The use of freshly killed animals reduced bleeding and was necessary for maximum exposure of the olfactory epithelium. Previous results show that the responses in this preparation were nearly identical to those in live animals (Scott et al. 1996
The overall response amplitudes were smaller than the responses recorded by electrodes that penetrated from the lamina propria side of the epithelium in our previous papers (Ezeh et al. 1995 Experiment 1A
The comparison of carvone, limonene, and cineole was tested in eight animals with the electrodes arrayed along the anterior border of endoturbinate IV, as shown in Fig. 2A. In each of the five cases making up Fig. 2B, the response to carvone was largest at the more dorsal recording sites. The response to limonene was more nearly flat across the electrodes for the average curve. The limonene and carvone curves progressively diverged in four of the five cases. In analysis of variance the odor-by-position interaction was highly significant (P < 0.01) for the total figure and was highly significant (P < 0.01) for the comparison of carvone with limonene. The response to carvone was larger on electrode 1 and on electrode 4 (P < 0.001). The difference between the limonene responses on electrodes 1 and 4 was not significant (P > 0.05).
Experiment 1B
Because cineole produced distinctly different responses from the other odors, we looked for other related odors that might produce peak responses in the intermediate regions (i.e., at electrodes 2 or 3 with the orientation of Fig. 2A). Because cineole is bicyclic, we tried three other bicyclic compounds shown in Fig. 1. Figure 3 shows a comparison of these odors with the odors used in Fig. 2B for three animals. Pinene, a relatively simple compound, evoked a response distribution similar to that of limonene, although not as steep. This distribution had significant odor-by-electrode interactions with the fenchone (a ketone) and 7-oxabicycloheptane (an epoxide somewhat similar in structure to cineole) responses. All of the odor-by-position interactions in Fig. 3, A and B, are significant at P < 0.001.
Experiment 2
Two stimulus nozzle positions were compared (Fig. 4). One nozzle was at the same angle as in experiment I. This was called the ventral position. In some cases, the second stimulus tube was placed at the dorsal extremity of the epithelium directly above the midline, so that it was parallel to the midline epithelium. This was called the dorsal position. In other cases, the second stimulus tube was placed anterior to the recording electrodes at an angle of ~30° from the anterior-posterior (midline) axis and at ~30° from the dorsal-to-ventral axis. This was called the anterior position.
Experiment 3
The expression zones described by Vassar et al. (1993) Experiment 4
Although Fig. 2 shows similar distributions across two endoturbinate bones, these recordings were from different animals, and the resulting curves are of slightly different shapes. If the response zones have a longitudinal organization similar to that indicated in Fig. 1, they would be best demonstrated in a more extensive set of recordings from the same group of animals. To accomplish this, we recorded five cases in which the electrode array was oriented along the longitudinal axis of the olfactory epithelium (Fig. 6A). These were intended to parallel the expression zones as described for the neonatal rat by Vassar et al. (1993)
Response distribution and gene expression zones
The issue addressed in this work is whether there is a distribution of odor sensitivity in the rat olfactory epithelium that is consistent with the longitudinally oriented receptor gene expression zones. For the terpene odors that we tested, there is such a distribution. Carvone responses were larger than limonene responses in the most dorsal part of the epithelium, which followed the shape of the most dorsal of the expression zones. Limonene responses were larger than or equal to carvone responses in the most ventral and posterior part of the epithelium, in a region roughly corresponding to the most ventral expression zone. Cineole responses were always proportionately larger than limonene responses in the intermediate region of the epithelium. The mean response amplitudes illustrated in this figure follow those trends. Those cineole responses were usually larger in the intermediate region than in either the dorsal or ventral regions. Although the highly significant interactions from the analyses of variance do not do more than demonstrate that the distributions are different for different odors, the reproducibility of the mean curves and the regression analysis strongly agree with these descriptions of responses to those three odors. Thus the midline distributions of adult rat responses to carvone and cineole appear to correspond to the orientation of expression zones as described for neonatal rats (Vassar et al. 1993 Possible generalization to other odors
Although the strongest data are for the three odors described above, several other related odors had response distributions that were predictable from their structure. Pinene and Other potential explanations of the data
One area of concern in our previous paper was whether the observed response distributions could be produced by differential sorption of odors onto the epithelium, thus removing them from the air stream and leading to a gradient of stimulus concentration across the epithelium. This phenomenon is well established under certain conditions (Hahn et al. 1994 Comparisons with recent literature
Several recent papers (Kent et al. 1996 Implications for sensory processing
Finally, we can make a few comments on the relationship of these data to sensory processing. Because we have sampled only a small number of odor classes, many more data are necessary to specify the relationship between gene expression zones and olfactory sensations. The reliable relationship that we observe between odor response and recording position is not strong enough to account for olfactory experience. Carvone, menthone, and limonene oxide have detectably different odors to humans. We are also aware that individual receptor neurons even within a particular region of the epithelium may have different response properties (Sato et al. 1994
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
) are expressed in longitudinal, anterior-posterior-oriented expression zones (Nef et al. 1992; Ressler et al. 1993; Strotmann et al. 1992, 1994; Vassar et al. 1993). Subsequent papers (Buck 1996; Fülle et al. 1995; Ressler et al. 1994; Sullivan et al. 1996) explicitly mention four expression zones, although Strotmann et al. (1994) described only three zones. These zones do not have perfectly sharp boundaries, and there are a small number of genes that are not expressed in this longitudinal pattern (cf. Strotmann et al. 1994). Furthermore, the zonal expression maps are based on study of a relatively small number of genes. Nevertheless, the pattern is sufficiently striking that it is useful to ask whether some odor responses may be distributed in a similar pattern. Such a result would be an important clue to the receptor functions associated with the different genes, which to date have only been directly studied in one published report (Raming et al. 1993).
; Getchell 1974; Ottoson 1956). It has long been known that there is a reliable topography of odor responses on the epithelium of the salamander nose (Kauer and Moulton 1974; Mozell et al. 1987), although that topography has been questioned for the frog (Kent and Mozell 1992). Recent EOG (MacKay-Sim and Kesteven 1994) and voltage-sensitive dye (Kent et al. 1996; Youngentob et al. 1995) recordings of the rat olfactory epithelium have shown that there is also a differential distribution of odor response and that the distribution is consistent across animals. However, these investigations of the rat epithelium did not find response distributions that corresponded to the receptor gene expression zones.
; Scott et al. 1996). We observed that most odorants with certain oxygen-containing functional groups (especially ketones and aldehydes) tend to produce larger responses in the dorsomedial recess of the epithelium, whereas odors without these functional groups tend to produce larger responses in the ventral and lateral parts of the epithelium, near the base of the turbinate bones. These two regions occupy different gene expression zones. Recordings along the base of the endoturbinates further suggested a correspondence with the gene expression zones. From these observations, we hypothesized that simple hydrocarbons and ketone forms of the same compound would evoke different distributions of response. The relative amplitudes of these responses should vary with recordings from different epithelial regions and would follow a pattern like that of the gene expression zones.
; Vassar et al. 1993). Although those descriptions were made for neonatal rats and mice, they are also present in 21-day-old rats (R. Vassar and R. Axel, personal communication). We adopted a physiological preparation similar to that employed by Youngentob et al. (1995) for their voltage-sensitive dye studies and by Mackay-Sim and Kesteven (1994) for their EOG study. We used terpene odorants, particularly contrasting limonene and -terpinene, which do not have oxygen-containing functional groups, with carvone and menthone, both ketones. We chose these odorants because they are popular olfactory stimuli, because they are relatively rigid molecules, and because a related compound, cineole, evoked an interesting pattern of response in preliminary experiments.
).
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
). Several published EOG (Edwards et al. 1988; Wang et al. 1993) and voltage-sensitive dye (Kent et al. 1996; Youngentob et al. 1995) experiments have used a similar technique with freshly killed animals.
View larger version (37K):
[in a new window]
FIG. 1.
Organization of the olfactory receptor gene expression zones and of electrode placements. A: expression zones along the medial wall of the endoturbinate bones. This is not meant to be accurate in all details, because only a few genes have been explored in the literature. Roman numerals indicated the different endoturbinate bones. In the text, the term "anterior" means in a direction parallel to the expression zones on turbinate bones II-III, as shown by the arrow. "Dorsal" refers to a direction across those zones. OB, olfactory bulb. B: structures of the odorants used in this report. Most of these were terpene compounds. Note that carvone differs from limonene primarily in the presence of the ketone group. The double bond in -terpinene is placed differently than in limonene. The ketone group in menthone is placed differently than in carvone. The bicyclic compounds cineole, pinene, fenchone, and 7-oxabicycloheptane differ in the presence or placement of the oxygen atom. Cineole, limonene-oxide, and 7-oxabicycloheptane have an epoxy bond rather than a ketone.
. The leads from these electrodes were connected to a four-channel AC coupled amplifier (low-frequency time constant 0.1 Hz). An indifferent electrode was placed on the frontal bone overlying the left olfactory bulb. The outputs were displayed on an oscilloscope and also fed by an A-D converter (digitized at a rate of 53 Hz) to a computer that plotted the traces and computed the peak negative current relative to the baseline just before the stimulus. This peak voltage was printed by each plot and was stored in a computer file for later statistical analysis. The plots were inspected during data collection and before analysis for quality control.
1). This represents the fraction of air saturated with odorant in the total volume of air passing the epithelium. The use of a single concentration allowed us to get repeated samples of several odor stimuli for statistical analysis. Recording time was limited because responses steadily declined in voltage over time. This decline is in part due to drying of the tissue, despite our attempts to keep it well humidified. We justified the use of a single high concentration by results in our previous experiments (Scott et al. 1996). There we found that odors producing larger responses in either of two regions of the epithelium did so over the whole concentration range for which they produced measurable responses.
(Fig. 2). Multiple stimulus series were presented with this one arrangement. Electrode 1 was always the dorsal electrode. This orientation was expected to span across most of the receptor gene expression zones. Published information about the expression zones is for neonatal animals. However, R. Vassar and R. Axel have shown (personal communication) very similar arrangements in rats up to 21 days old. For experiment 2, the recording electrodes were maintained in the same position on endoturbinate IV, but the position of the odor stimulus nozzle was varied to test the effect of air flow direction on the response distributions. For experiment 3, we repositioned the recording electrodes to test the responses along the dorsal border of endoturbinate IV (Fig. 5), a position where the expression zones curve into a different orientation than that seen on the other turbinate bones. For experiment 4, the four electrodes were oriented along the longitudinal axis of the olfactory epithelium with electrode 1 on endoturbinate IV, electrode 2 on endoturbinate III, electrode 3 on endoturbinate II, and electrode 4 on endoturbinate II (Fig. 6). In each animal we successively tested a set of seven such longitudinally arranged sites at intervals of ~0.5 mm across the dorsal-to-ventral extent of the four endoturbinate bones. Each set of electrode positions was marked on the photograph of the preparation for later analysis.
View larger version (31K):
[in a new window]
FIG. 2.
Experiment 1A. A: placement of electrodes for experiment 1. These electrodes were placed at approximately equal intervals along the anterior border of endoturbinates IV or II . B: average curves for 3 odors in 5 animals for the electrode positions along endoturbinate IV. There were 6-21 trials for each odor per animal (average 15). Error bars: 95% confidence intervals from analysis of variance. C: average responses for 3 animals with the use of the odors in A plus menthone and -terpinene. Also recorded from endoturbinate IV. There were 9-15 trials for each odor per animal (average 13). Note from Fig. 1 that carvone and menthone are similar in structure, whereas limonene and -terpinene are also similar in structure. Responses in A and B were standardized to ethyl butyrate. D: average responses for 4 animals (8-12 presentations of each odor per animal) for a similar set of recordings from endoturbinate II. These responses were standardized to amyl acetate. Note that the labeling of zones is schematic for purposes of orientation and is not meant to imply that a particular electrode was always centered in 1 of the expression zones.
View larger version (36K):
[in a new window]
FIG. 5.
Experiment 3. A: recording sites along the dorsal border of endoturbinate IV and the relation of these recording sites to the expression zones as described in the literature. B: data from 40 total presentations of each odor in three rats. The odor-by-position interaction was significant for the comparisons of the limonene response with the responses to either limonene or carvone (P < 0.01). Amyl acetate was the standardizing odor.
View larger version (36K):
[in a new window]
FIG. 6.
Experiment 4. A: the arrangement of electrodes laid parallel to the general orientation of the expression zones. The 1st recording for each session was always from the most ventral part of the olfactory epithelium. These were at the anterior border of each turbinate, corresponding to the row of numbers marked with an a. After 4 presentations of each odor, the electrodes were moved more dorsally by ~0.5 mm to the row of dots marked with a b. The electrodes were moved successively by the same increment to the final position at the row of dots marked with a g. B: average recordings from 1 animal for carvone and limonene. The 7 sets of data (a-g) correspond to the rows of simultaneous recordings. Inset: the 4 electrodes in each set. The standardizing odor was ethyl butyrate.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
; Scott et al. 1996). The responses in the opened preparation were sometimes quite large at the beginning (up to 12 mV), but often declined within first 10 min. This response decline may be due to slight mechanical damage or drying of the mucosa. It occurred in live animals with opened nasal cavities as well as in the freshly killed preparations (data not shown). Responses were maintained with a slow decline for periods >2 h in most cases. We looked at the data for several animals and determined that responses to all odors declined at the same rate (data not shown). Data were collected as long as responses were stable on all electrodes.
-terpinene and menthone odors, structurally similar to limonene and carvone, respectively. The difference between the limonene curves for Fig. 2, B and C, illustrates that there was some individual variation. This variation may arise in part from variations in electrode placement, but at this point it is not completely explained. The parallel carvone and menthone distributions and parallel limonene and terpinene distributions were apparent in all individual cases. The differences between carvone and limonene were also apparent in all individual cases. The responses to limonene and terpinene were greater on electrode 4 than electrode 1 (P < 0.01). The responses to carvone and menthone were larger on electrode 1 than electrode 4 (P < 0.01).
(Fig. 2D). There the response to carvone was larger on electrode 1 than electrode 4 (P < 0.01) and the response to limonene was larger on electrode 4 than electrode 1 (P < 0.01).
View larger version (15K):
[in a new window]
FIG. 3.
Experiment 1B. Comparison of responses to bicyclic and epoxy compounds. A: odor-by-position interaction is significant for the comparisons of limonene responses with either carvone or cineole responses (P < 0.01). B: interaction is significant for the comparison of pinene responses with either fenchone or 7-oxabicycloheptane responses (P < 0.01). Data from 3 rats with 4-6 presentations of each odor per animal (A) and from 2-6 presentations of each odor per animal (B). Responses were standardized to amyl acetate.
View larger version (34K):
[in a new window]
FIG. 4.
Experiment 2. Comparison of responses under different conditions of air flow. A: odor stimulation and recording setup. The electrodes were placed along endoturbinate IV as in Fig. 2. The electrodes and nozzles remained in the same position during the experiment, but the odor source was switched between the 2 nozzles after each odor in the series had been tested twice. The standard odor source used in all the other figures was the ventral nozzle illustrated in this figure. The anterior odor nozzle was placed rostral to the recording array, as shown at right. This anterior nozzle was at an angle of ~30° to the vertical dimension, as represented by the position of the dorsal nozzle, and ~30° to the plane of the epithelium, indicated by the dotted arc. B: odor source was switched between the ventral and dorsal nozzles in 2 animals. There were 4-12 presentations of each odor per animal per condition (average 7.6). C: odor source was switched between the ventral nozzle and an anterior nozzle directly anterior to the electrode array in 2 animals (8 and 6 presentations of each odor per animal per condition). This anterior nozzle was at an angle of 30° relative to the sagittal plane. The tips of the nozzles were 2 cm from the tissue. Amyl acetate was the standardizing odor.
are not entirely straight, but curve upward at the back of turbinate IV. Therefore the response distribution in the back of this region should be different from that tested in Fig. 2. Figure 5 shows a test of this hypothesis. There is a large difference between the limonene and carvone response distributions and a smaller, but significant (P < 0.001) difference between the limonene and cineole response distributions. The space along the dorsal border of turbinate IV is substantially smaller than that along the anterior border. As a consequence, the placement of electrodes relative to expression zones is even less certain than in experiment 1. Nevertheless, the general agreement with the findings of experiment 1 is strong.
. The expectation was that, with a properly oriented array, all four electrodes should show a greater response to carvone if they were in a dorsal position. Conversely, all four should show a greater response to limonene if they were in a ventral position. Figure 6B shows this result in the recordings from a single animal.
View larger version (39K):
[in a new window]
FIG. 7.
Combined polynomial regression analysis from 4 animals in experiment 4. A: regression curves and individual points for comparisons across turbinate bones. Data from the 4 animals were pooled in these calculations. Error bars: 95% confidence intervals for the regression line. Limonene and carvone are illustrated because they showed the most and least scatter, respectively, around the regression curves. All regression curves were 2nd order. B: curves for the odors comparing the 4 animals and pooling data from the 4 turbinate bones. One set of points for each odor (of 196 points) was removed before the calculations of A and B, because an anomalous response to the standard odor inflated all of the data values for 1 electrode. C: regression curves for 6 odors from experiment 2. The curves for carvone, limonene, and limonene-oxide are based on 4 animals, whereas the others are based on 3 animals. The standardizing odor was ethyl butyrate.
View this table:
TABLE 1.
Coefficients of determination
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
). The correspondence for limonene is less strong, but it often evokes larger responses in the most ventral two electrode sites than elsewhere. This interpretation seems compelling despite the differences in animal ages and the fact that we do not have direct visualization of the expression zones for these EOG experiments.
). The dorsomedial recess corresponds to the dorsal zone in Fig. 1 of the present paper, and the ventrolateral recess probably includes the two ventral zones of Fig. 1 (cf. Ressler et al. 1993; Vassar et al. 1993). Because we did not have reliable access to the intermediate regions in the previous experiments, we did not have information to indicate the difference between the cineole and limonene response distributions. Note that the present data agree with that previous report in showing that the cineole response is consistently larger in the ventral part of the epithelium. Taken together, the two EOG studies are a strong argument for a relationship between the putative gene expression zones and differential sensitivity to at least this set of odors.
), although it should also be taken as a caution against a premature conclusion that the response distribution results exclusively from the zonal distribution as described by the available probes. One possibility is that future expression mapping may demonstrate other zones slightly out of register with the ones currently known. Such a family of zones would be easier to reconcile with the gradual transitions suggested in our data.
). These data could be explained by a small number of receptor cells in the dorsal zones being strongly responsive to the ketones (such as carvone) but not responsive to the hydrocarbons (such as limonene). Alternatively, many of the receptors expressed in a zone might have some common properties, so that most cells of the dorsal zone would be more responsive to ketones than to hydrocarbons. We have previously argued for the latter interpretation because strong stimuli like the ones used here would almost completely block antidromic activation of the receptor cell population (Ezeh et al. 1995). The final resolution of this question will require extensive sampling of single receptor cells, something that is beyond the scope of this study.
), there is not a strong sequence similarity of genes expressed in the same zone. Singer and Shepherd (1994) have recently modeled the olfactory receptors, suggesting that the binding pocket for odors lies among the transmembrane helices of the protein in a position analogous to that of the 
-adrenergic receptor. Singer and Shepherd have recently found, in analysis of 21 mouse receptor sequences from Sullivan et al. (1996), that there is a significant correlation between the residues occurring in position 430, which Singer and Shepherd believe to be part of the binding site, and the expression zones. In particular, Singer and Shepherd (personal communication) find a high incidence of histidine in this site in the most dorsal zone but no histidine in this position in other zones. Indeed, histidine 430 has been predicted to interact selectively with oxygen-containing groups (Singer et al. 1995). Assuming that the same principles apply to the rat, this may be a very fruitful area of study.
-terpinene evoked larger responses ventrally than dorsally. Fenchone and menthone had larger responses dorsally. This supports our previous generalization that terpene ketones evoke their largest responses in the dorsal zones. We had previously extended that generalization to other groups of odorants, including alkanes, cyclic alkanes, and aromatic compounds. We also reported that some other compounds with oxygen-containing functional groups (some aldehydes and some epoxides) evoked larger responses in the dorsal region than the corresponding hydrocarbons, although alcohols were equivocal (Scott et al. 1996). We have not yet extensively tested those here, but some data support the previous generalizations. Limonene oxide was more effective dorsally in experiment 4, with some tendency toward a peak response in the intermediate zone. In the few cases tested, methyl benzoate evoked larger dorsal responses than did toluene, and cyclooctanone evoked larger dorsal responses than cyclooctane (data not shown). These comparisons were similar to those we had previously reported. Therefore we predict that the observations on comparison of the dorsomedial and lateral recesses of our previous paper will extend to the most dorsal and most ventral expression zones wherever they are tested.
reported distributions of responses in the olfactory bulb that are consistent with these data and suggested that the carbonyl group was an important determining factor in their data. The fact that ketones produce a greater proportion of dorsal over ventral responses than similar aldehydes or esters (Scott et al. 1996) suggests that more than the presence or absence of the carbonyl group is involved. This point is also obvious in our present data, in which all responses were standardized to esters. On the other hand, Scott et al. (1996) also found that aromatic esters and an aldehyde had much more pronounced localization than the aliphatic esters. In this report, the fact that limonene oxide had a greater tendency for dorsal responses also indicates that a more general property is involved. One possibility is the degree of electronegativity of the functional group, as suggested by Sato et al. (1994).
) that responses to several odors showed little variation along the anterior-to-posterior direction in either the dorsal or lateral parts of the epithelium. The present paper demonstrates that response differences (particularly those between limonene, carvone, and cineole responses) follow the distribution along the medial wall of the endoturbinate bones that has been described for expression of putative olfactory receptor genes in the neonatal rat (Vassar et al. 1993). Our observations are merely correlations with the expression zones, and more data will be necessary to demonstrate a causal relationship. Nevertheless, the correspondence seen here and in our previous papers (Ezeh et al. 1995; Scott et al. 1996) encourages us to believe that these distributions are related to the expression zones.
; Kent et al. 1996; Mozell et al. 1987). There are several arguments against this suggestion. At the flow rate and angles used, there may be minimal exposure of the odor in the air stream to the epithelial surface. Differential sorption might account for the difference between carvone, which is highly sorbed, and limonene, which is not highly sorbed, but a more complicated mechanism would have to be invoked for the cineole response distribution. The fact that the odor effects follow the particular course of the putative receptor gene expression zones would require peculiar assumptions about air flow turbulence in the cavity. Visual inspection of the rise times of the carvone and limonene responses (data not shown) does not support a consistent relationship with air flow direction as would be expected from the sorption hypothesis. The strongest statement is the fact that odor-by-position interactions in the same direction were present with three different air flow directions. These considerations lead us to believe that the response distributions reflect a mechanism that is intrinsic to the epithelium rather than one imposed by air flow.
), and in comparisons between the dorsal recess and the spaces between the bases of the turbinate bones (Ezeh et al. 1995). This is in agreement with a number of demonstrations of this effect in the frog (reviewed in Hahn et al. 1994). Sorption is purposely minimized in the present series of experiments, but it is likely to contribute strongly to sensory processing.
) and mechanisms that internalize and break down odorants have been described (Lazard et al. 1991). There is some evidence for spatial segregation of odorant binding proteins (Rama Krishna et al. 1995). The action of either of these activities could influence the odor responses observed here if they acted selectively on odorants with particular chemical properties. Until the specifics of the distributions of these materials are better known, we must be guarded in the tentative conclusion that we have observed a property of odor molecules that is bound by receptors of particular expression zones.
; Mackay-Sim and Kesteven 1994; Youngentob and Kent 1995; Youngentob et al. 1995) investigated the distribution of odor responses along the midline of the nose and did not observe the strong longitudinal orientation of responses reported here. We believe that this difference from the present report results from the particular odors sampled in each of the studies. We chose to test odorants that we expected to produce response gradients across the zones on the basis of the presence of certain functional groups tested in our previous paper (Scott et al. 1996). We contrasted compounds that differed primarily in this property. With the exception of Figs. 6 and 7, we did not make comparisons along the longitudinal axis of the expected gene expression zones. The other authors were attempting to sample odorant properties more generally. Their odor stimuli differed in several characteristics. Unfortunately, they did not use more than one of the odorants in our series, so direct comparisons with our contrasting responses with limonene and carvone are not possible. It is possible that some of the response gradients that the other authors observed are related to some of the nonzonal distributions reported in the literature (Strotmann et al. 1994).
). Our previous recordings with homologous series (Scott et al. 1996) did not explore a sufficiently large range of chain length to test for regional differences with this variable. This would be an important variable to study in the future. It seems possible that one of the reasons that other epithelial recordings have not produced the dorsal-to-ventral response gradients like those we observe is the fact that they have mixed several molecular properties, including chain length and different chemical classes, in their stimulus series (Kent et al. 1996; Mackay-Sim and Kesteven 1994; Youngentob et al. 1995). EOG and voltage-sensitive dyes are appropriate techniques for study of epithelial properties given the fact that all such investigations have reported systematic regional variations. The large numbers of genes expressed in the epithelium and the low probability of finding any individual gene expressed in a particular cell (Buck 1996) mean that principles learned about the classes of odors that are effective in stimulating receptor cells in different regions of the epithelium may be very useful in selecting stimuli for study of individual cells.
). Recent results on the projection pattern of olfactory sensory neurons (Ressler et al. 1994; Vassar et al. 1994) agree with neurophysiological (Imamura et al. 1992; Mori and Yoshihara 1995) and other (Guthrie et al. 1993; Lancet et al. 1981) results in arguing for a convergence of axons from sensory neurons with like properties onto small numbers of glomeruli. This convergence is doubtless very important in arranging the sensory information for processing in the olfactory bulb.
; Jastreboff et al. 1984; Pedersen et al. 1986; Schoenfeld et al. 1994) indicates that the bulb must have an input organization that represents the odor chemistry differences expressed in the epitheliuim in some systematic way. This is already indicated by recordings from rabbit mitral cells (Mori and Yoshihara 1995). The long mitral cell basal dendrites span a distance that may correspond to a substantial portion of the input from one of the expression zones (Orona et al. 1984). The intrabulbar association described by Schoenfeld et al. (1985) clearly allows for interactions of part of the bulb receiving inputs from different expression zones. It would be important to contrast the function of these systems in analyzing olfactory information.
| |
ACKNOWLEDGEMENTS |
|---|
The authors thank F. H. Schmidt for aid with instrumentation and data analysis.
This work was supported by National Institute of Deafness and Other Communications Disorders Grant DC-00113.
| |
FOOTNOTES |
|---|
Address reprint requests to J. W. Scott.
Received 27 June 1996; accepted in final form 10 December 1996.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  G. C. Yang, P. W. Scherer, K. Zhao, and M. M. Mozell Numerical Modeling of Odorant Uptake in the Rat Nasal Cavity Chem Senses, March 1, 2007; 32(3): 273 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. Scott, H. P. Acevedo, and L. Sherrill Effects of Concentration and Sniff Flow Rate on the Rat Electroolfactogram Chem Senses, July 1, 2006; 31(6): 581 - 593. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. Scott Sniffing and Spatiotemporal Coding in Olfaction Chem Senses, February 1, 2006; 31(2): 119 - 130. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Youngentob, P. F. Kent, and F. L. Margolis OMP Gene Deletion Results in an Alteration in Odorant-Induced Mucosal Activity Patterns J Neurophysiol, December 1, 2003; 90(6): 3864 - 3873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.A. Chaput EOG Responses in Anesthetized Freely Breathing Rats Chem Senses, December 1, 2000; 25(6): 695 - 701. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P.E. Scott-Johnson, D. Blakley, and J.W. Scott Effects of Air Flow on Rat Electroolfactogram Chem Senses, December 1, 2000; 25(6): 761 - 768. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ma and G. M. Shepherd Functional mosaic organization of mouse olfactory receptor neurons PNAS, October 23, 2000; (2000) 220301797. [Abstract] [Full Text]
|
|
|
|
 |
|
 J. W. Scott, T. Brierley, and F. H. Schmidt Chemical Determinants of the Rat Electro-Olfactogram J. Neurosci., June 15, 2000; 20(12): 4721 - 4731. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. Scott and T. Brierley A Functional Map in Rat Olfactory Epithelium Chem Senses, December 1, 1999; 24(6): 679 - 690. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. Friedrich and S. I. Korsching Chemotopic, Combinatorial, and Noncombinatorial Odorant Representations in the Olfactory Bulb Revealed Using a Voltage-Sensitive Axon Tracer J. Neurosci., December 1, 1998; 18(23): 9977 - 9988. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. J. Dreyer The area code hypothesis revisited: Olfactory receptors and other related transmembrane receptors may function as the last digits in a cell surface code for assembling embryos PNAS, August 4, 1998; 95(16): 9072 - 9077. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. C. Bozza and J. S. Kauer Odorant Response Properties of Convergent Olfactory Receptor Neurons J. Neurosci., June 15, 1998; 18(12): 4560 - 4569. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ma and G. M. Shepherd Functional mosaic organization of mouse olfactory receptor neurons PNAS, November 7, 2000; 97(23): 12869 - 12874. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
