This is a preprint of an article published in The Journal of Comparative Neurology, 2000,

426:330-338.

ÓWiley-Liss 2000

Odorant Molecular Length: One Aspect of the Olfactory Code

Brett A. Johnson* and Michael Leon

Department of Neurobiology and Behavior, University of California,

Irvine, California 92697-4550

Number of text pages: 24

Number of figures: 5 total (2 color)

Number of tables: 0

Running title: Chemotopic coding of odorant length

Associate Editor: Joseph Price

Indexing terms: chemical senses; 2-deoxyglucose; fatty acids; mapping; odor; olfactory bulb

*Correspondence to: Brett A. Johnson

Department of Neurobiology and Behavior

University of California

2205 BioSci II

Irvine, CA 92697-4550

Telephone: (949) 824-7303

Fax: (949) 824-2447

E-mail: bajohnso@uci.edu

Grant sponsor: NIDCD; Grant number: DC03545.

Abstract

Organic acid odorants of differing carbon number produce systematically different spatial patterns of [¹⁴C]2-deoxyglucose uptake in the glomerular layer of the olfactory bulb. Because increasing carbon number correlates with progressive increases in several molecular features, including hydrophobicity, length, and volume, we determined which of these properties was most associated with systematic changes in the location of an anterior, dorsomedial module responding to fatty acids. We exposed groups of rats to two series of organic acids that each had the same number of carbons, but differed in their hydrocarbon structures. These straight-chained, branched, cyclic, and double-bonded molecules differed independently in hydrophobicity, length, and volume. The only molecular property that was strongly correlated with the location of the module was molecular length, suggesting that this molecular feature is the principal determinant of the chemotopic organization of glomeruli within the module. We also found that distinct hydrocarbon structures produced large differences in spatial patterns of 2-deoxyglucose uptake in posterior parts of the bulb. Even subtly distinct structural isomers evoked posterior responses that differed greatly. The odorant 2-methylbutyric acid evoked much greater uptake in the posterior bulb than did its structural isomer 3-methylbutyric acid (isovaleric acid). These data suggest that posterior portions of the bulb may encode specific steric features of odorant molecules and that some odorant features may have an inherent or acquired greater representation than do others.

Pure odorants of distinct chemical structure and distinctly perceived odor evoke different spatial patterns of activity in the olfactory bulb (Stewart et al., 1979; Jourdan et al., 1980; Royet et al., 1987). In the glomerular layer, these patterns are comprised of glomerular modules, each of which appears to respond to a particular odorant chemical feature (Johnson et al., 1998; Johnson and Leon, 2000). This odorant specificity likely reflects the binding specificity of the odorant receptors expressed by sensory neurons converging in their projections to the glomeruli (Ressler et al., 1994; Vassar et al., 1994; Mombaerts et al., 1996). Odorants that are closely related in chemical structure stimulate nearby glomeruli within a given module, suggesting a chemotopic spatial arrangement of glomeruli (Friedrich and Korsching, 1997; Johnson et al., 1998, 1999; Sachse et al., 1999; Tsuboi et al., 1999; Johnson and Leon, 2000). In rats, clustering has been observed for glomeruli responding to straight-chain aliphatic odorants possessing either the same number of carbons, but distinct functional groups (Johnson and Leon, 2000), or the same functional group, but incrementally different numbers of carbons (Johnson et al., 1998, 1999; Rubin and Katz, 1999).

The spatial arrangements of rostral glomeruli responding to straight-chained aliphatic acids and aldehydes are particularly well correlated with odorant carbon number. Increasing carbon number is correlated with increasingly rostral responses in the dorsolateral bulb (Johnson et al., 1999; Rubin and Katz, 1999), and with increasingly ventral responses in the dorsomedial bulb (Johnson et al., 1999). Mitral/tufted cell projection neurons in the dorsomedial rabbit bulb are highly tuned to respond to aldehydes and acids of a restricted carbon number (Mori et al., 1992; Imamura et al., 1992). Lateral inhibition between projection neurons associated with adjacent glomeruli apparently underlies this tuning (Yokoi et al., 1995), and the chemotopic arrangement of glomeruli responding to odorants of most similar carbon number may insure that tuning suppresses responses to the most similar odorant chemicals (Johnson et al., 1999).

Increasing carbon number of straight-chained aliphatic compounds is associated with progressive increases in several molecular parameters, including molecular length, molecular volume, and hydrophobicity, any of which might affect the binding of an odorant ligand to its receptor. In the present study, we asked whether the chemotopic organization of glomeruli within the anterior, dorsomedial module of response to carboxylic acid odorants could be described more accurately in terms of only one of these molecular parameters. If only one parameter was correlated with the position of glomerular response, then this parameter might represent a primary odorant feature used to encode odor quality. The odorants we selected included structural isomers of five- and six-carbon fatty acids involving straight-chained, branched, cyclic, and double-bonded structures (Fig. 1). By virtue of possessing different hydrocarbon structures, these compounds vary independently in volume, length, hydrophobicity, and flexibility. To measure glomerular activity, we mapped [¹⁴C]2-deoxyglucose (2-DG) uptake across the entire glomerular layer (Johnson et al., 1999), which also allowed us to examine the effects of hydrocarbon structure on spatial representations throughout the entire bulb.

Materials and Methods

Odorant exposures

Odorants were purchased from Fisher Scientific (Acros Organics, Pittsburgh, PA). Purities listed by the manufacturer were as follows: valeric (pentanoic) acid, 99%; isovaleric (3-methylbutanoic) acid, 99%; (D,L)-2-methylbutyric (2-methylbutanoic) acid, 98%; cyclobutanecarboxylic acid, 98%; trans-2-pentenoic acid, 97%; caproic (hexanoic) acid, 98%; isocaproic (4-methylpentanoic) acid, 99%; tert-butylacetic (3,3-dimethylbutanoic) acid, 98%; cyclopentanecarboxylic acid, 98%; trans-3-hexenoic acid, 99%. All five-carbon odorants were presented at a vapor phase concentration of 8 parts per million, which is similar to the concentration used for our study of the effects of carbon number in straight-chained carboxylic acids (Johnson et al., 1999). All six-carbon odorants were presented at 4 ppm, which is the highest concentration obtainable for cyclopentane carboxylic without depriving rats of oxygen, and which is above the concentration of caproic acid that is necessary to detect a pattern of 2-DG uptake (Johnson et al., 1999). Concentrations of saturated vapors were calculated from vapor pressures, which were estimated from boiling points using the equation of Hass and Newton (1975). Desired vapor phase concentrations were achieved by diluting saturated vapors using ultra-zero grade air and a flow-dilution olfactometer. Dilutions were as follows: valeric acid, 1/22.4; isovaleric acid, 1/38.2; 2-methylbutyric acid, 1/38.2; cyclobutanecarboxylic acid, 1/12.3; trans-2-pentenoic acid, 1/8; caproic acid, 1/19.1; isocaproic acid, 1/21.7; tert-butylacetic acid, 1/47.4; cyclopentanecarboxylic acid, 1/7.9; trans-3-hexenoic acid, 1/13.2.

All procedures were approved by the University of California, Irvine Institutional Animal Care and Use Committee. Odorant exposures were conducted as described previously (Johnson et al., 1999). Separate gas washing bottles and Teflon tubing were used for each odorant to prevent cross-contamination between odorants, and the exposure chamber was thoroughly cleaned between rats. Exposures were initiated by injecting male rats (19-22 postnatal days) subcutaneously with 0.16 mCi/kg [¹⁴C]-2-DG (Sigma, St. Louis, MO), followed by a 45-minute odor exposure. The animals then were decapitated, and the brains were removed and frozen rapidly in isopentane at –45 °C. Sixty rats from twelve litters were used in the study, six rats for each of the ten odorants. No two rats from a given litter were exposed to the same odorant, and the order of odorant presentation was explicitly varied across different litters. Data from two rats, involving exposures to 2-methylbutyric acid and caproic acid, were eliminated from the study due to damage of the brain during sectioning and injection of insufficient quantities of radiolabel, respectively.

Mapping of 2-DG uptake

Uptake of radiolabel was mapped across the entire glomerular layer using our previously described method (Johnson et al., 1999). Briefly, this method involves analysis of radiolabel present in every sixth 20-µm coronal cryostat section. By using sets of radial grids designed to give samples at every 120 µm, measurements were taken at fixed angle increments from the glomerular layer as detected in adjacent cresyl violet-stained sections. Grids were chosen on the basis of the fractional distance between rostral-caudal anatomical landmarks detected in the Nissl-stained sections. Uptake was converted from grayscale readings (NIH IMAGE) into units of nCi/g tissue by reference to ¹⁴C-standards exposed to the same sheets of autoradiography film. Arrays of uptake were transformed by contraction or expansion to contain the same number of sections. Arrays from the two bulbs of each animal then were averaged, and relative uptake was expressed as z scores relative to the mean and standard deviation of uptake across the entire glomerular layer (Johnson et al., 1998; Johnson and Leon, 2000).

Contour charts were used to illustrate the distribution of uptake across the entire glomerular layer in arrays averaged across all animals exposed to the same odorant. The ventral-most measurement taken from each section is centered horizontally in the charts.

Centroid determinations

The location of uptake across the anterior, dorsomedial module was estimated in each animal as a centroid, which has proven to be an effective statistical tool in many of our analyses of bulbar activity patterns (Johnson et al., 1998, 1999; Johnson and Leon, 2000). To define the area to be used for centroid analysis, we averaged z score standardized arrays across all animals in the study (Johnson et al., 1999). The module was clearly visible in contour charts of the grand average array. To set the boundaries of the module, we chose contiguous values that exceeded 0.5 in the grand average array. The analyzed region was irregularly shaped and contained a total of 78 measurements. The area spanned nine sections (sections 11-19) at its greatest rostral-caudal extent and 13 angle increments at its greatest dorsal-ventral extent. Centroids were calculated by using values standardized as a ratio of glomerular layer uptake to subependymal zone uptake. Uptake in the subependymal zone was measured in sections taken from a consistent portion of the bulb, as described previously (Johnson et al., 1999).

To establish correlations with molecular properties for straight-chained aliphatic acids of different carbon number, centroids were taken from our previous study, which used different boundaries for the same module due to different locations of uptake evoked by the smaller odorants in that study (Johnson et al., 1999). The statistical significance of differences in the locations of the centroids was determined by an analysis of variance (ANOVA) performed as described previously (Johnson and Leon, 2000).

Odorant molecular properties

Molecular length, molecular volume, and predicted hydrophobicity were calculated using Molecular Modeling Pro v.3 software (ChemSW Inc., Fairfield, CA). Prior to calculating the molecular length of each molecule, the bond between carbons 1 and 2 was fixed along the x-axis, and the bond between carbons 2 and 3 was oriented in the x-y plane. The length was then calculated along the x-axis. Values for hydrophobicity (log P) refer to the log of the octanol/water partition coefficient predicted from fragmental constants. Linear regression was used to assess the correlation between molecular properties and the dorsal-ventral coordinate of centroids within module A. Statistical significance and correlation coefficients were obtained by using the data analysis package of Microsoft Excel 98.

Results

Uptake within the anterior, dorsomedial fatty acid response module

All ten of the carboxylic acid odorants, regardless of hydrocarbon structure, stimulated a circumscribed region of the anterior, dorsomedial bulb (Fig. 2, large black arrows) that was shown previously to respond to straight-chain aliphatic acids ranging from 2 to 8 carbons (Johnson et al., 1999). The same area, which we term “module A,” also responded to pentanol and high concentrations of pentanal and methyl pentanoate in a previous study (Johnson and Leon, 2000). The corresponding region of the extreme dorsal and lateral bulb, “module a,” which likely receives projections from sensory neurons expressing the same odorant receptors (Johnson et al., 1999), also was activated by all ten of the acid odorants in this study (Fig. 2, small black arrows).

To determine whether the different carboxylic acid odorants optimally stimulated different glomeruli within module A, we calculated centroids of 2-DG uptake within the module. The relative locations of the centroids within the module are illustrated in Figure 3. Centers of ellipses in Figure 3 indicate the mean location across the different rats exposed to the same odorant, and the lengths and heights of the ellipses indicate ± standard errors of the mean along the two coordinates used to calculate the centroids. The centroids were found to differ significantly across odorants (ANOVA, P < 0.05, F[9,48] = 2.58).

In our previous study of straight-chain aliphatic acids, centroids also differed significantly across odorants containing different numbers of carbons (Johnson et al., 1999). As shown in Figure 4A, the dorsal-ventral coordinates of the centroids determined in that previous study were significantly correlated with odorant molecular length, molecular volume, and hydrophobicity, all of which are exactly correlated with carbon number in straight-chain compounds. The centroids shifted ventrally with increases in each of the three molecular properties (Fig. 4A). As shown in Figure 4B, the dorsal-ventral coordinates of the centroids calculated in the present study only were significantly (P < 0.005) and highly (r = 0.85) correlated with odorant molecular length. The centroids shifted ventrally with increasing length of the compounds. The positions of the centroids were not significantly correlated with either odorant molecular volume or odorant hydrophobicity (Fig. 4B), which vary independently from molecular length in these fatty acids that differ in hydrocarbon structure.

The dorsal-ventral coordinates of the centroids also were significantly correlated with molecular length (P < 0.01) when the five-carbon odorants were analyzed separately (not shown). The regression line from this analysis was very similar to that in Fig. 4B, as was the regression line obtained for the six-carbon odorants when they were analyzed separately. These findings suggest that the relationship between molecular length and the position of the module was not a result of the difference in concentration between the five- and six-carbon odorants (8 parts per million and 4 parts per million, respectively). There also was no significant correlation between amount of 2-DG uptake and the position of the centroid within either the five-carbon or the six-carbon acids (not shown).

Uptake in posterior bulb regions

In addition to evoking 2-DG uptake in subtly different locations within module A, the carboxylic acids of different hydrocarbon structure evoked uptake in the posterior portions of the lateral and medial olfactory bulb, and the posterior patterns were unique to each odorant (Fig. 2).

Valeric acid evoked a pattern of activity that was similar to that observed in previous studies (Johnson et al., 1999; Johnson and Leon, 2000). As in our most recent study (Johnson and Leon, 2000), the posterior response involved two modules (d and e) in the lateral aspect of the bulb and two modules (D and E) in the medial aspect of the bulb (Fig. 2). The two modules are perhaps more apparent in sections from individual bulbs (Fig. 5) than they are in the average activity maps. 2-Methylbutyric acid evoked greater uptake in the posterolateral and posteromedial aspects of the bulb than did valeric acid (Fig. 2). In average maps, the rostral portions of the posterior regions activated by 2-methylbutyric acid overlapped with modules e and E evoked by valeric acid (Fig. 2, large white arrows). Module e typically extended across only three or four sections in rats stimulated with valeric acid (Fig. 5). In contrast, the uptake evoked by 2-methylbutyric acid in this region often involved as many as 12 sections, and the foci in the more caudal sections appeared to shift dorsally, which is consistent with the presence of a distinct module (Fig. 5). Modules d and D that were evoked by valeric acid (Fig. 2, small white arrows) were not evoked to the same extent by 2-methylbutyric acid.

In contrast to 2-methylbutyric acid, the branch variant 3-methylbutyric acid (isovaleric acid) evoked much less uptake in the posterior portions (modules e and E) of the lateral and medial aspects of the bulb (Fig. 2, large white arrows). The difference between 2-methylbutyric acid and isovaleric acid existed despite the fact that the two odorants were presented at the same vapor phase concentration and despite the fact that they stimulated module A to a similar extent (Fig.2, large black arrows).

The posterior pattern evoked by cyclobutanecarboxylic acid was different from those evoked by any of the structural isomers of valeric acid. In individual bulbs from all animals studied, the pattern involved smaller foci of intense uptake that typically were separated from each other by unlabeled glomeruli (Fig. 5). Subtle variation across animals in the exact location and number of these smaller foci yielded the appearance of a broader, more patchy pattern in the posterior portions of the average map than was seen for valeric and 2-methylbutyric acids (Fig. 2). Cyclopentanecarboxylic acid and trans-3-hexenoic acid also stimulated multiple, small foci of 2‑DG uptake distributed across the glomerular layer (Fig. 5). Despite the more scattered distributions of uptake, the posterior patterns were reliably different for these two odorants (Fig. 2, 5).

Like isovaleric acid, isocaproic acid stimulated very little uptake in the posterior bulb (Fig. 2). When viewed in individual sections or when uptake was standardized relative to uptake in the subependymal zone, trans-2-pentenoic acid, caproic acid, and tert-butylacetic acid also evoked lesser uptake in the posterior bulb (data not shown). When standardized as z scores such as used in Figure 2, the fact that posterior uptake was lower for these odorants is not as obvious. In individual bulbs, these patterns appeared to involve scattered, small foci of uptake that were less numerous and less intense than the foci evoked by cyclobutanecarboxylic acid, cyclopentanecarboxylic acid, or trans-3-hexenoic acid (not shown).

Discussion

We have found that the positions of activity within the anterior, dorsomedial carboxylic acid response module (module A) correlate simply with the molecular length of organic acid odorants that differ profoundly in hydrocarbon structure. This finding explains the previously noted correlation between location of activity and carbon number in straight-chained, saturated aliphatic acids (Johnson et al., 1999). The chemotopic organization of glomeruli within a module in relation to a single molecular attribute of the effective odorants suggests that the olfactory system may use spatial information to perform a true chemical analysis of odorous stimuli. This analysis would presumably involve tuning of projection neuron responses to odorant molecular length by way of lateral inhibition at the glomerular and/or external plexiform layers (Yokoi et al., 1995).

Glomerular responses likely reflect the specificity of olfactory sensory neurons projecting to the glomeruli, and these specificities are likely determined by the odorant receptors expressed by the sensory neurons. Our data therefore suggest that progressively more ventral glomeruli within module A are likely to receive projections from sensory neurons expressing receptors recognizing progressively longer odorant molecules, perhaps by virtue of possessing longer or deeper hydrophobic binding surfaces.

Our analyses involved calculations of centroids of activity within a module that contained a total of 78 distinct measurements of 2-DG uptake. Previous results indicated that a single odorant (valeric acid) could stimulate as many as 35 glomeruli within the module (Johnson et al., 1999). Centroids yield only crude measurements of the balance of activity across the module (Johnson and Leon, 2000). Thus, the locations of individual active glomeruli within module A, as well as the relative levels of glomerular activity, could vary to some degree across animals and yet still result in the significant correlation found here.

The simple correlation between molecular length and the position of response in module A suggests that this module may not discriminate between odorants of similar length that differ in branch structure or carbon number. It is, therefore, of great interest that fatty acids of even slightly different hydrocarbon structure evoked very different patterns of uptake in the posterior portions of the olfactory bulb. Small differences in carbon number also produce large differences in posterior patterns of 2-DG uptake (Johnson et al., 1999). The independent variation in the patterns of response in module A and in the posterior regions of the bulb provides further evidence that distinct chemical features of odorant molecules are encoded in distinct spatial locations within the bulb (Johnson et al., 1998; Johnson and Leon, 2000). A given odorant is found to be unique only through considering the combination of its distinct molecular features, which are reflected in the relative activity of distinct glomerular response modules (Johnson et al., 1998; Johnson and Leon, 2000).

The differences between the structural isomers, valeric acid, isovaleric acid, and 2-methylbutyric acid, involved differences in total amounts of 2-DG uptake as well as differences in the relative stimulation of nearby modules (e.g., d/D and e/E). The remaining organic acids evoked more complex patterns of uptake that differed from one another despite overlapping in their distributions. The differences in the posterolateral and posteromedial patterns evoked by cyclobutanecarboxylic acid and valeric acid in the present study were greater than the differences between valeric acid and other odorants possessing the same five-carbon, straight-chained structure, but distinct functional groups (Johnson and Leon, 2000). In our study on odorant functional groups, differences in the posterior portions of the lateral and medial aspects of the bulb primarily involved different subsets of glomerular modules that were located near one another (Johnson and Leon, 2000). The profound dependence of the posterior pattern on the hydrocarbon structure of carboxylic acid odorants, together with the greater similarity across odorants differing only in functional groups, suggests that glomeruli in these regions may be specific for very particular steric features of the hydrocarbon portion of the molecules (i.e., the precise arrangements of hydrocarbon hydrogen atoms in space).

Trans-3-hexenoic acid and the cyclic carboxylic acids in the present study stimulated numerous, small glomerular foci in the posterior portions of the bulb. These foci appeared to be separated by unlabeled glomeruli. In our previous studies, responses to small aliphatic odorants typically involved the stimulation of larger, uninterrupted modules that consisted of clusters of nearby glomeruli, similar to module A in the present study (Johnson et al., 1998, 1999; Johnson and Leon, 2000). The modules were interpreted as involving glomeruli of related specificity clustered together to produce local networks for lateral inhibition. In this model, the orderly arrangement of glomeruli within each module insures the tuning of projection neurons to odorant chemical features that would be difficult to distinguish on the basis of the activity of a single type of olfactory sensory neuron (Yokoi et al., 1995). The posterior responses to valeric acid and 2-methylbutyric acid in the present study also involved clusters of neighboring glomeruli, suggesting that lateral inhibition may be used to tune responses in the posterior bulb, as well. In contrast, the stimulation by cyclic and double-bonded odorants of isolated glomeruli with the intervening presence of inactive glomeruli would yield a final output that is more difficult to predict in terms of lateral inhibitory networks.

Straight-chained carboxylic acids are more flexible than the corresponding cyclic and double-bonded odorants. It is possible that the activation of posterior clusters of neighboring glomeruli by straight-chained and simple branched odorants represents recognition of various conformations of the flexible odorants by related odorant receptors associated with adjacent glomeruli. The comparative rigidity of the cyclic and double-bonded odorants may preclude binding to such related receptors, thereby resulting in punctate activity patterns involving isolated glomeruli. In addition, the cyclic compounds and double-bonded short chain fatty acids in the trans-configuration apparently are rare in nature (Mann et al., 1994). Because of the low probability of encountering such odorants in nature, clustered responses to distinguish these odorants from other, similar odorants may not have been selected over evolutionary time.

The high levels of 2-DG uptake evoked by certain odorants (e.g., 2-methylbutyric acid) in comparison to very closely related compounds (e.g., isovaleric acid) suggests that some odorants may have a greater bulbar representation than do others. A large number of high affinity receptors for biologically relevant odorants may have evolved so that these odorants might be more easily detected and discriminated from similar compounds. Biologically significant sources of organic acids include urine (Chalmers et al., 1974; Singer et al., 1997), skin surfaces (Nicolaides, 1974; Kanda et al., 1990; Zeng et al., 1991), and various plants (Higa and Fuyama, 1993). Valeric, isovaleric, caproic, isocaproic, and 2-methylbutyric acids are excreted in feces, often through the catabolism of amino acids by anaerobic bacteria in the intestines or cecum (Allison, 1978; Høverstad et al., 1985, 1986). The relative concentration of the different acids is a function of both diet and bacterial strain (Høverstad et al., 1985, 1986; Høverstad and Midtvedt, 1986; Murase et al., 1995; Padilha et al., 1995). Rat pups learn to become attracted to the odorants produced within their dams’ ceca, a chemical cue that varies with maternal diet (Leon, 1983). Early odor preferences are correlated with increased 2-DG uptake in glomeruli responding to the learned odorants (Coopersmith and Leon, 1984; Johnson and Leon, 1996), including those produced by the mother (Sullivan et al., 1990). Therefore, the greater uptake evoked by particular organic acid odorants may be a function of both natural selection of receptors for these relevant odorants and enhancement of glomerular responses through prior odorant exposure.

The relatively low activity evoked by isovaleric acid in comparison to either valeric acid or 2‑methybutyric acid may indicate that rats are less sensitive to isovaleric acid than to the other two structural isomers. This prediction remains to be tested. Certain strains of mice (Wysocki et al., 1977) and certain humans (Amoore, 1967) are hyposmic to isovaleric acid. In both control and hyposmic mouse strains, isovaleric acid stimulated robust 2-DG uptake in the dorsomedial bulb region (module A), but statistical analyses suggested a difference in glomerular activation patterns between the strains (Sicard et al., 1989). These differences were attributed to a more patchy distribution of uptake within module A of the hyposmic strain (Sicard et al., 1989). Similar to our present results, little uptake was evoked by isovaleric acid in posterolateral or posteromedial bulb regions for either strain of mice (Royet et al., 1987; Sicard et al., 1989). Because these regions were not subjected to a separate statistical analysis in the mouse studies, it remains possible that different levels of activity in these posterior bulb regions might correlate with sensitivity to isovaleric acid.

The present report represents one of a series of our studies exploring the relationships between odorant chemical structure and spatial distributions of olfactory bulb responses. Previous reports showed systematic correlations of spatial locations of response with hydrocarbon features of aliphatic esters (Johnson et al., 1998), with carbon number in aliphatic acids (Johnson et al., 1999), with different functional groups (Johnson and Leon, 2000), and with odorant concentration (Johnson and Leon, 2000). Mapping 2-DG uptake across the glomerular layer continues to demonstrate that each odorant evokes activity in multiple bulbar locations, each separate location apparently signaling some unique aspect (feature) of odorant chemistry. The simple correlation between organic acid molecular length and location of response within a glomerular module indicates that using this approach to determine which odorant features are analyzed and represented in different parts of the glomerular layer will likely reveal important aspects of the neural coding of odorant chemistry.

Acknowledgements

We thank Zhe Xu, Edna E. Hingco, and Zsuzsanna B. Najbauerné for excellent technical assistance, and Dr. Cynthia C. Woo for helpful discussions.

LITERATURE CITED

Allison MJ. 1978. Production of branched-chain volatile fatty acids by certain anaerobic bacteria. Appl Environ Microbiol 35:872-877.

Amoore JE. 1967. Specific anosmia: a clue to the olfactory code. Nature 214:1095-1098.

Chalmers RA, Bickle S, Watts RWE. 1974. A method for the determination of volatile organic acids in aqueous solutions and urine, and the results obtained in propionic acidaemia, beta-methylcrotonylglycinuria, and methylmalonic aciduria. Clin Chim Acta 52:31-41.

Coopersmith RM, Leon M. 1984. Enhanced neural response to familiar olfactory cues. Science 225:849-851.

Friedrich RW, Korsching SI. 1997. Combinatorial and chemotopic odorant coding in the zebrafish olfactory bulb visualized by optical imaging. Neuron 18:737-752.

Hass HB, Newton RF. 1975. Correction of boiling points to standard pressure. In: Weast RC, editor. Handbook of chemistry and physics, 56th ed. Cleveland: CRC Press. p D176-D177.

Higa I, Fuyama Y. 1993. Genetics of food preference in Drosophila sechellia. Genetica 88:129-136.

Høverstad T, Midtvedt T. 1986. Short-chain fatty acids in germfree mice and rats. J Nutrition 116:1772-1776.

Høverstad T, Midtvedt T, Bøhmer, T. 1985. Short-chain fatty acids in intestinal content of germfree mice monocontaminated with Escherichia coli or Clostridium difficile. Scand J Gastroenterol 20:373-380.

Høverstad T, Carlstedt-Duke B, Lingaas E, Norin E, Saxerholt H, Steinbakk M, Midtvedt T. 1986. Influence of oral intake of seven different antibiotics on faecal short-chain fatty acid excretion in healthy subjects. Scand J Gastroenterol 21:997-1003.

Imamura K, Mataga N, Mori K. 1992. Coding of odor molecules by mitral/tufted cells in rabbit olfactory bulb. I. Aliphatic compounds. J Neurophysiol 68:1986-2002.

Johnson BA, Leon M. 1996. Spatial distribution of [¹⁴C]2-deoxyglucose uptake in the glomerular layer of the rat olfactory bulb following early olfactory preference learning. J Comp Neurol 376:557-566.

Johnson BA, Leon M. 2000. Modular representations of odorants in the glomerular layer of the rat olfactory bulb and the effects of stimulus concentration. J Comp Neurol, 409:495-509.

Johnson BA, Woo CC, Leon M. 1998. Spatial coding of odorant features in the glomerular layer of the rat olfactory bulb. J Comp Neurol 393:457-471.

Johnson BA, Woo CC, Hingco EE, Pham KL, Leon M. 1999. Multidimensional chemotopic responses to n-aliphatic acid odorants in the rat olfactory bulb. J Comp Neurol 409:529-548.

Jourdan F, Duveau A, Astic L, Holley A. 1980. Spatial distribution of [¹⁴C]2-deoxyglucose uptake in the olfactory bulbs of rats stimulated with two different odours. Brain Res 188:139-154.

Kanda F, Yagi E, Fukuda M, Nakajima K, Ohta T, Nakata O. 1990. Elucidation of chemical compounds responsible for foot malodour. Brit J Dermatol 122:771-776.

Leon M. 1983. Chemical communication in mother-young interactions. In: Vandenbergh J, editor. Pheromones and mammalian communication. New York: Academic Press.p 39-77.

Mann J, Davidson RS, Hobbs JB, Banthorpe DV, Harborne JB. 1994. Natural products: their chemistry and biological significance. Essex, England: Longman.

Mombaerts P, Wang F, Dulac C, Chao SK, Nemes A, Mendelsohn M, Edmonson J, Axel R. 1996. Visualizing an olfactory sensory map. Cell 87:675-686.

Mori K, Mataga N, Imamura K. 1992. Differential specificities of single mitral cells in rabbit olfactory bulb for a homologous series of fatty acid odor molecules. J Neurophysiol 67:786-789.

Murase M, Kimura Y, Nagata Y. 1995. Determination of portal short-chain fatty acids in rats fed various dietary fibers by capillary gas chromatography. J Chromatogr 664:415-420.

Nicolaides N. 1974. Skin lipids: their biochemical uniqueness. Science 186:19-26.

Padilha MT, Licois D, Gidenne T, Carrâe B, Fonty G. 1995. Relationships between microflora and caecal fermentation in rabbits before and after weaning. Reprod Nutrition Devel 35:375-386.

Ressler KJ, Sullivan SL, Buck LB. 1994. Information coding in the olfactory system: evidence for a stereotyped and highly organized epitope map in the olfactory bulb. Cell 79:1245-1255.

Royet JP, Sicard G, Souchier C, Jourdan F. 1987. Specificity of spatial patterns of glomerular activation in the mouse olfactory bulb: computer-assisted image analysis of 2-deoxyglucose autoradiograms. Brain Res 417:1-11.

Rubin BD, Katz LC. 1999. Optical imaging of odorant representations in the mammalian olfactory bulb. Neuron 23:499-511.

Sachse S, Rappert A, Galizia CG. 1999. The spatial representation of chemical structures in the antennal lobe of honeybees: steps towards the olfactory code. Eur J Neurosci 11:3970-3982.

Sicard G, Royet J-P, Jourdan F. 1989. A comparative study of 2-deoxyglucose patterns of glomerular activation in the olfactory bulbs of C57 BL/6J and AKR/J mice. Brain Res 481:325-334.

Singer AG, Beauchamp GK, Yamazaki K. 1997. Volatile signals of the major histocompatibility complex in male mouse urine. Proc Natl Acad Sci USA 94:2210-2214.

Stewart WB, Kauer JS, Shepherd GM. 1979. Functional organization of rat olfactory bulb analysed by the 2-deoxyglucose method. J Comp Neurol 185:715-734.

Sullivan RM, Wilson DA, Wong R, Correa A, Leon M. 1990. Modified behavioral and olfactory bulb responses to maternal odors in preweanling rats. Dev Brain Res 53:243-247.

Tsuboi A, Yoshihara S, Yamazaki N, Kasai H, Asai-Tsuboi H, Komatsu M, Serizawa S, Ishii T, Matsuda Y, Nagawa F, Sakano H. 1999. Olfactory neurons expressing closely linked and homologous odorant receptor genes tend to project their axons to neighboring glomeruli on the olfactory bulb. J Neurosci 19:8409-8418.

Vassar R, Chao SK, Sitcheran R, Nuñez JM, Vosshall LB, Axel R. 1994. Topographic organization of sensory projections to the olfactory bulb. Cell 79:981-991.

Wysocki CJ, Whitney G, Tucker D. 1977. Specific anosmia in the laboratory mouse. Behav Genet 7:171-188.

Yokoi M, Mori K, Nakanishi S. 1995. Refinement of odor molecule tuning by dendrodendritic synaptic inhibition in the olfactory bulb. Proc Natl Acad Sci USA 92:3371-3375.

Zeng X-N, Leyden JJ, Lawley HJ, Sawano K, Nohara I, Preti G. 1991. Analysis of characteristic odors from human male axillae. J Chem Ecol 17:1469-1492.

Figure legends

Fig. 1. Organic acid odorants included in the study. The compounds shared a carboxylic acid functional group, but differed in hydrocarbon structure. Brackets indicate structural isomers, which possess the same molecular formula. Structures are shown as ball-and-stick diagrams indicate the bonding between carbon atoms. For clarity, hydrogen atoms within the hydrocarbon portions of the molecules are not shown. The asterisk indicates the chiral carbon of 2‑methylbutyric acid.

Fig. 2. Maps of 2-deoxyglucose (2-DG) uptake across the glomerular layer. Uptake is shown in the form of color-coded contour charts oriented as shown at top, left. Each color-coded chart represents the average uptake across all rats exposed to the same odorant. The orientation diagram at top, center shows the location of modules defined in our previous study (Johnson and Leon, 2000). Large black arrows in each contour chart indicate the uptake corresponding to module A in each exposure condition. Small black arrows indicate activity within module a, the lateral equivalent of module A (Johnson and Leon, 2000). Large white arrows indicate the locations of modules e (lateral) and E (medial) previously defined in our analysis of responses to valeric acid (Johnson and Leon, 2000). Small white arrows indicate the locations of modules d (lateral) and D (medial) in the maps of uptake evoked by valeric acid.

Fig. 3. Centroids of uptake within module A. The left panel shows an average contour chart of z score-standardized arrays from all rats in the present study. Darker gray shading indicates values that exceeded 0.5 in this grand average array. A contiguous field of such values was located in the region of module A, and this area (outlined in white) was used for the analysis of centroids. Centroids were calculated in terms of a rostral-caudal coordinate (section number) and a dorsal-ventral coordinate (measurement number) for each individual rat in the study. The mean locations of these centroids across all animals exposed to a given odorant are indicated by centers of ellipses in the right panel. The widths and heights of the ellipses denote ± the standard error of the mean of the rostral-caudal and dorsal-ventral coordinates, respectively. The centroids were all located within a small area indicated by the white rectangle in the left panel. Analysis of variance demonstrated that centroids differed across odorant conditions.

Fig. 4. Correlations between the dorsal-ventral locations of module A and different molecular properties of carboxylic acid odorants. Relative dorsal-ventral coordinates of the centroids within module A were plotted as a function of odorant molecular length, molecular volume, and hydrophobicity (the log of the predicted partition coefficient between octanol and water). A: Values for dorsal-ventral coordinates of centroids were taken from our previous study of n-aliphatic acids differing in carbon number, where module A was termed “Field 2” (Johnson et al., 1999). Statistically significant correlations were found for all molecular properties. B: Only molecular length was significantly correlated with the dorsal-ventral coordinates of centroids in module A of the present study. In the two studies, the borders of the module were defined differently on the basis of the responses obtained in each study. For this reason, the y-axes of the plots in A and B are not directly comparable. n.s., not significant.

Fig. 5. Uptake in streams of individual pseudocolor-enhanced autoradiography sections from the posterior, lateral bulb. Representative patterns of 2-deoxyglucose (2-DG) uptake evoked by selected odorants are shown to illustrate the different types of distributions encountered in the study. The area represented is outlined in red in the upper left panel. Each section is 20 µm in thickness. Adjacent sections are separated by 100 µm. The same region of the bulb is shown for each odorant. For a bulb of modal size (Johnson et al., 1999), this region would extend for 1.56 mm, starting at 1.56 mm from the rostral pole of the bulb. For the bulbs illustrated here, the distances shown ranged from 1.34 to 1.46 mm, depending on the animal. Converging lines within each section stream indicate foci of uptake judged to align in consecutive sections. Arrows denote foci of uptake judged to occur in only one section. Foci corresponding to modules d and e from our previous study (Johnson and Leon, 2000) are labeled in the section stream from the valeric acid-exposed animal.