Skip to main content
NCBI home page
Chemical Senses logoLink to Chemical Senses
. 2014 Jul 23;39(7):601–616. doi: 10.1093/chemse/bju033

The Regeneration of P2 Olfactory Sensory Neurons Is Selectively Impaired Following Methyl Bromide Lesion

Eric H Holbrook 1,2,*, Carrie L Iwema 3,*, Carolyn E Peluso 4, James E Schwob 2,3,✉
PMCID: PMC4133598  PMID: 25056730

Abstract

The capacity of the peripheral olfactory system to recover after injury has not been thoroughly explored. P2-IRES-tauLacZ mice were exposed to methyl bromide, which causes epithelial damage and kills 90% of the P2 neurons. With subsequent neuronal regeneration, P2 neurons recover within their usual territory to equal control numbers by 1 month but then decline sharply to roughly 40% of control by 3 months. At this time, the P2 projection onto the olfactory bulb is erroneous in several respects. Instead of converging onto 1 or 2 glomeruli per surface, small collections of P2 axons innervate multiple glomeruli at roughly the same position in the bulb as in controls. Within these glomeruli, the P2 axons are aggregated near the edge, whereas the remainder of the glomerulus contains olfactory marker protein (+), non-P2 axons, violating the one receptor–one glomerulus rule normally observed. The aggregates are denser than found in control P2-innervated glomeruli, suggesting that the P2 axons may not be synaptically connected. Based on published literature and other data, we hypothesize that P2 neurons lose out in an activity-based competition for synaptic territory within the glomeruli and are not maintained at control numbers due to a lack of trophic support from the bulb.

Key words: axonal regeneration, olfactory bulb, olfactory epithelium, olfactory receptor, olfactory sensory neuron, olfactotoxin

Introduction

The mammalian olfactory system continually generates new sensory neurons throughout adult life as well as on a much larger scale following injury to the olfactory epithelium (OE), olfactory bulb (OB), and/or olfactory nerve (Hinds and Hinds 1976a, 1976b; Graziadei and Graziadei 1979a, 1979b; reviewed by Graziadei and Monti Graziadei 1985; Costanzo 1991; Schwob et al. 1995). In the setting of the limited neuronal replacement observed in the unlesioned epithelium, a large population of olfactory sensory neurons (OSNs) of like type (i.e., expressing the same olfactory receptor) are intact, in contact with the OB, and consequently provide a scaffold along which newly generated axons can choose to track. The limited extent of ongoing turnover contrasts with the wholesale loss and replacement of neurons after either category of olfactory lesion: nerve transection (Harding and Wright 1979; Monti Graziadei et al. 1980; Yee and Costanzo 1995; Koster and Costanzo 1996) or peripheral injury by exposure to olfactotoxins (Harding et al. 1978; Nadi et al. 1981; Kawano and Margolis 1982; Kream and Margolis 1984; Burd 1993; Schwob et al. 1995, 1999; Iwema and Schwob 2003). Reconstitution of functional capacity following damage and the wave of neurogenesis that it elicits requires that a number of steps in the regenerative process be executed properly. The OE must first generate a sufficient number of OSNs in both the correct location and of the correct type to reconstitute prelesion status. Next, axons from the newly generated OSNs must project to their appropriate targets in the OB. Finally, complete recovery following injury requires the restoration of functional synaptic connections between OSN axons and the dendrites of mitral/tufted cells within the glomeruli of the OB.

Judging by global measures of OSN number and maturation, the OE recovers nearly perfectly following direct injury caused by inhalation of the olfactotoxin, methyl bromide (MeBr) (Schwob et al. 1995). In normal animals, OR-defined types of OSNs are located in discrete medial-to-lateral bands or swathes throughout the anteroposterior extent of the OE (Strotmann et al. 1992, 1996, 2000; Ressler et al. 1993; Vassar et al. 1993; Strotmann, Wanner, Helfrich, Beck, and Breer 1994; Strotmann, Wanner, Helfrich, Beck, Meinken, et al. 1994; Mombaerts et al. 1996; Iwema et al. 2004; Miyamichi et al. 2005). We demonstrated that OSNs expressing the same OR subtype have been regenerated with the same distribution across the OE in terms of circumferential distribution, total number of OR-expressing OSNs, and location along the anteroposterior axis within 3 months (Iwema et al. 2004). Accordingly, restoration to near normal in the periphery of the olfactory system occurs without substantial evidence of distortion despite the profound and comprehensive nature of the initial injury.

Functional recovery of the system also depends on the axonal connections that form between the reconstituted epithelium and the bulb. In the unlesioned animal, the convergence of olfactory axons onto glomeruli reflects the identity of the OR that is expressed by each OR-defined set of neurons. Thus, OSNs expressing a particular OR project onto a small number of glomeruli on the medial and lateral surfaces of each OB (Ressler et al. 1994; Vassar et al. 1994; Mombaerts et al. 1996; Wang et al. 1998), which provides the anatomical basis for transforming chemical signals into firing patterns across neural space at this first stage of the pathway for processing odorant stimuli. As a consequence, disruption of axonal projections to the glomeruli likely distorts the proper detection and discrimination of odorants. Whether or not receptotopic mapping remains distorted despite the reconstitution of the olfactory periphery remains subject to debate, as summarized below.

On the one hand, selective, but transient ablation of the P2 population of OSNs by genetic means culminates in the accurate reinnervation of the correct, P2-targeted glomeruli in the bulb after this neuronal population is restored (Gogos et al. 2000). On the other hand, receptotopy is disrupted after other forms of injury. For example, following recovery from nerve transection axons extended by the P2-expressing OSNs do fasciculate, but they do not reinnervate the same prelesion glomeruli and project widely across the glomerular layer (Costanzo 2000). Because nerve transection causes a physical injury to the usual route taken by olfactory axons and seems to disrupt the formation of connections at a rhinotopic level (Christensen et al. 2001), this paradigm may represent a particularly invidious form of injury that damages glia and/or other elements of the nerve and thereby abrogates the mechanisms that subserve accurate axon targeting during piecemeal turnover. Some mistargeting of the regenerated P2 projection has been reported when systemic injection of the compound dichlobenil is used to damage the epithelium. However, dichlobenil causes irreversible destruction of the dorsomedial part of the OE (Vedin et al. 2004; Xie et al. 2013); while not P2 territory, permanent destruction of the dorsomedial lining may have effects that also interfere with axonal targeting from ventrolateral epithelium. In addition, there is much left unknown regarding the recovery of this specific population of neurons after epithelial lesion.

Inhalation of MeBr produces a widespread but largely reversible lesion of the OE, and epithelial recovery is accompanied by a remarkable degree of reinnervation of the OB (Schwob et al. 1999). The extent to which specific connectivity is restored remains as yet incompletely defined. The bulb is fully reinnervated when lesions of moderate severity are produced (Schwob et al. 1999). However, abnormalities are observed after reinnervation in animals that were exposed to MeBr while food-restricted. In these animals, the severity of the lesion is exacerbated, causing a permanent reduction in OSN number via the replacement of olfactory by respiratory epithelium in a substantial fraction of epithelial area. With the enhanced lesion, the anterior half of the OB is preferentially reinnervated, whereas the posterior half is hypoinnervated and the glomeruli near the posterior margin remain totally denervated, suggesting that either the OSNs that normally target the posterior glomeruli are selectively absent from the reconstituted population of neurons after lesion or that the mechanisms responsible for correct targeting fail with lesions of such severity. A variety of correlative data, including the analysis of OR expression patterns after recovery from MeBr lesion, suggest that a failure of axon guidance is the relevant mechanism.

For example, a subset of OSNs that are labeled by antibody against heat shock protein 70 and normally innervate a single glomerulus per lateral and medial surface of the bulb also mistarget subtly during reinnervation; in this case, the axons reinnervate roughly the same areas of the OB as expected under normal circumstances, but they do not converge to single glomeruli (McMillan Carr et al. 2004). In addition, subsets of OSNs that project to specialized “necklace” glomeruli located at the posterior edge of the OB regenerate in the OE and reinnervate the OB after MeBr lesion (Ring et al. 1997) although the projections are mistargeted to nearby glomeruli away from the edge of the glomerular sheet (Schwob JE, unpublished observations). Likewise, some degree of mistargeting is observed when optical recordings of glomerular activation are used to assess functional recovery by comparing lesioned-recovered OB either to the prelesion pattern of activation or to the contralateral unlesioned side (when naris-plugged animals are exposed to MeBr gas) (Cheung et al. 2014). Thus, following recovery from MeBr exposure, the mechanisms responsible for targeting OR-expressing OSN axons into the appropriate general area of the OB apparently remain intact unless the lesion results in widespread respiratory metaplasia; but, these data suggest that the guidance mechanisms fail to steer the axons to the correct glomeruli from this point onward.

It is important to note that the foregoing data carry the proviso that these anonymous molecular markers are reliable surrogates for the expression of specific ORs, which is an untested assumption. As well, the physiological data cannot be assigned to particular set(s) of OR-defined axons and, thus, are indirect measures of errors in the reestablishement of the olfactory map after the epithelium has reconstituted.

The experiments described here extend previous results by addressing whether the axons of OSNs known to express the same OR, namely the P2-OR, fasciculate and reinnervate the same glomerular targets in the OBs of MeBr-lesioned animals, a lesion model that has been subject to substantial investigation both peripherally and centrally. We determined that in the mouse OE, the regenerated P2 population at 3-month postlesion is equivalent to control in terms of circumferential and anteroposterior distribution patterns. However, the number of P2-expressing OSNs declines from a peak at 1-month postlesion to end up significantly reduced by comparison with control; moreover, the axons from these mice project inaccurately to the OB despite the 3-month recovery period. Although the axons are in the correct region of the OB, there is a lack of convergence onto the expected 1 or 2 glomeruli per surface of the bulb, and where glomeruli contain P2 axons, they generally form dense aggregates at the periphery of those glomeruli, which result resembles the errors in P2 targeting that have been observed with a different lesion-regeneration model (St John and Key 2003). Our data suggest that mistargeting onto the OB, in turn, may compromise both the size and status of the P2 population and potentially the functional recovery of the system.

Materials and methods

Animals

Male homozygous P2-IRES-tauLacZ (P2-ITL) transgenic mice were used for all experiments, bred in house from founder animals obtained from Dr Peter Mombaerts. Initially received on a mixed B6/129 background, the line was made congenic by breeding with C57BL/6J mice. All animals were maintained on ad libitum chow and water in a heat and humidity controlled vivarium. The animal work was conducted at SUNY Upstate Medical University and Tufts University School of Medicine, and all animal use protocols were approved by the respective Committee for Humane Use of Animals.

MeBr lesion

Awake 3-month-old mice were placed in a wire enclosure measuring 15×15×15cm centered in a 30×30×30cm Plexiglas box and exposed for 8h to MeBr gas (Matheson Gas Products) at 180 ppm in purified air at a flow rate of 10L/min as described (Schwob et al. 1995). MeBr-lesioned mice were cardiac perfused at various time points after lesion as described below.

Tissue preparation

MeBr-lesioned and age-matched–unlesioned P2-ITL mice were deeply anesthetized with sodium pentobarbital or by injection of an anesthetic cocktail consisting of ketamine, acepromazine, and xylazine and then perfused intracardially with phosphate buffered saline (PBS) and 4% paraformaldehyde (Fisher Scientific) in 0.075M sodium phosphate buffer, pH 7.2. The nose and brain were stripped of soft tissue, the heavy bones of the skull and nose, postfixed for 1h, and decalcified in saturated ethylenediaminetetraacetic acid on a shaker at 4 °C for 1 week. After blocking, the tissue was cryoprotected in 30% sucrose made in PBS and then frozen in O.C.T. compound (Miles Inc.). OE was sectioned on a cryostat (Reichert-Jung 2800 Frigocut) at 5 µm, while sections through the OB beginning from the anteriormost appearance of glomeruli were cut at 12 µm, and collected onto Fisherbrand Plus slides (Fisher Scientific) and stored at −20 °C.

X-gal reaction

When reacting whole mounts of the tissue with X-gal, the nose from control and post-MeBr-lesioned P2-ITL mice were split into 3 parts (septum and right and left turbinates) and separated from the OBs following fixation and decalcification. All whole-mount tissue and mounted sections were exposed to the X-gal working solution consisting of 0.01% deoxycholic acid, 0.02% Nonidet P-40, 20mM K-ferrocyanide, 20mM K-ferricyanide, 2mM magnesium chloride, 1mg/mL X-gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside, Roche Molecular Biochemicals), and PBS (pH 7.2) at 37 °C until signal was detected.

Immunohistochemistry

Antibody labeling with rabbit anti-β-galactosidase (anti-β-gal; 1:4000, ICN Pharmaceuticals), goat anti-olfactory marker protein (OMP) (1:15 000, Wako Chemicals USA), mouse anti-βIII tubulin (Tuj1 1:800, Covance), and rabbit anti-PGP9.5 (1:4000, Cedarlane) was performed using methods described previously (Schwob et al. 1992). In brief, sections through the nasal tissues from control and P2-ITL mice were incubated for 90min, whereas whole mounts of the septal mucosa were incubated for 5 days in primary antibody solution then followed sequentially by the appropriate biotinylated secondary antibody (1:50, Jackson Immunoresearch Laboratories, Inc.) and avidin:biotinlyated horseradish peroxidase complex (Vector Laboratories). Incubation times for these steps were lengthened for whole-mount staining to 1 day each. Labeled cells were visualized with diaminobenzidine tetrahydrochloride (DAB, Sigma–Aldrich), and sections were counterstained with either hematoxylin or eosin. In cases of double labeling with anti-β-gal and neuronal markers anti-Tuj1, anti-PGP9.5, or anti-OMP, staining was performed in series with anti-β-gal visualized with DAB first, then incubating sections with an antineuronal antibody visualized with either Slate Gray or VIP (Vector Laboratories). Sections were coverslipped with DPX mounting media (Sigma–Aldrich). Septal mucosa samples were submerged in 10% glycerine overnight and flattened by weights onto Fisherbrand Plus slides for 3 days. Whole mounts were then cleared and coverslipped with DPX as routine for stained sections and allowed to dry with weights applied to prevent air accumulation below the coverslip.

Quantitative analysis

Numbers of anti-β-gal-labeled OSNs were determined in control mice (n = 4) and at 5 days and 3 months after MeBr exposure (n = 5 for each time point) by counting immunolabeled cells at ×400 magnification in 5-µm-thick sections of the OE taken every 200 µm between the anterior tip of endoturbinate II anteriorly and the anterior tip of the bulb posteriorly. The number of sections between these landmarks varied slightly among the groups of animals due to minor differences in the angle of sectioning relative to true coronal; an average of 45 sections were counted per animal. In order to estimate total number of P2 neurons, counts of labeled profiles with nuclei were extrapolated to include the intervening tissue using the following formula: [(r1 + r2)/2t × u], where r1 is raw cell count for section 1, r2 for section 2, t is section thickness (5 µm), and u is uncounted distance between counted sections (195 µm). The count was corrected for double-counting error using Abercrombie’s correction [corrected count = raw count × (section thickness/{section thickness + cell diameter})] (Abercrombie 1946) and a nuclear diameter of 5 µm, which has the effect of excluding small fragments of staining that could not be reliably identified as neuronal; that value was not significantly different among control tissues and tissues harvested at 5 days and 3 months after MeBr. Averages and the standard error of the mean were calculated. For calculations of P2-expressing OSNs in septal whole-mount tissue, digitized mosaic images were analyzed using IPLab imaging software (Biovision Technologies). P2-expressing OSNs were counted by eye and a sum total was calculated for each survival time (n = 4 except at the 30-day time point where n = 3). Averages and the standard error of the mean were calculated and subjected to a 1-way analysis of variance (P < 0.0001; F = 8.569). Significant differences (P < 0.05) between the control and each time point postlesion were determined using Dunnett’s multiple comparison test.

Image analysis

Digital images of anti-β-gal-stained serial sections through the OBs were collected using a Hamamatsu color chilled 3CCD camera (Hamamatsu USA) mounted on a Nikon Microphot-SA microscope with a ×4 objective lens. Individual tiles of septal whole-mount mucosa were combined into composite mosaics using IPLab Spectrum image analysis (Scanalytics, Inc.), Autopano Pro (Kolor), and Adobe Photoshop (Adobe Systems, Inc.) software. Images of the septal whole-mount surface were taken at high power using a homemade positioning goniometer, in order to capture and illustrate the 3D features of the P2 OSNs; in this case, the 3D effect was obtained by capturing 2 images with a 3-mm shim placed under the right and left edges of the slide, individually and respectively, while maintaining focus on objects in the center of the field.

Results

Effect of MeBr exposure on number and distribution of P2 OSNs

Exposure of P2-ITL mice to MeBr at 180 ppm damages the ventrolateral OE, including the P2 territory (Mombaerts et al. 1996), to an extent similar to that previously observed in outbred Sprague-Dawley, F344, and hooded rats (Schwob et al. 1995; Youngentob and Schwob 1997; Jang et al. 2003), in that neurons and sustentacular cells are destroyed in the vast majority of the epithelium in this region producing a dramatic reduction in overall epithelial thickness in these areas. In contrast to the circumferential damage caused by MeBr in rats, the dorsomedial region of P2-ITL mice made congenic with the C57BL/6J strain is less severely damaged than ventrolateral parts, and many or even most olfactory neurons and sustentacular cells are spared here.

In order to establish a baseline against which to assess recovery, we counted the number of P2 OSNs that survived MeBr exposure at 5-day postlesion (n = 5) in a series of coronal sections evenly spaced throughout the anteroposterior extent of the OE that were subjected to anti-β-gal immunostaining and counterstaining with eosin (Figure 1A,C) and compared their distribution and number to control, unlesioned 3-month-old P2-ITL mice (n = 4) (Figures 1 and 2). P2-expressing OSNs were readily identified by the deposition of orange-brown chromagen in their somata, axons, and dendrites. All labeled cells were counted on both the right and left sides of each section, and the total cell counts were corrected for unexamined tissue using the correction factors and for double counting using the Abercrombie correction formula as indicated in the Materials and methods (Abercrombie 1946). The average size of the P2 OSN population in 3-month-old control mice (taking both sides into account) is roughly 10 200 OSNs.

Figure 1.

Figure 1

Open in a new tab

The vast majority of P2 neurons are destroyed by MeBr inhalation. P2 receptor-expressing neurons in mouse OE labeled with anti-β-gal antibody visualized with DAB. (A, C, E) Photomicrographs showing P2 receptor expression in control epithelium (CI-46M). (B, D, F) Photomicrographs from 5-day post-MeBr-lesioned animal (CI-99). At 5 days after lesion, β-gal (+) neurons are absent. After DAB labeling for P2 neurons, the sections were immunostained for either neuron-specific tubulin (Tuj1, E) or neural cell adhesion molecule (F), visualized by Slate Grey substrate (Vector), in order to assay the status of the population of OSNs as a whole. (F) At 5 days after MeBr, marker (+) neurons are present, but dendrites are poorly developed and P2 expression is not evident. Scale bars: A and B = 1mm, C and D = 200 µm, and E and F = 25 µm. Arrowhead = basal lamina.

Figure 2.

Figure 2

Open in a new tab

The distribution of β-gal (+) P2 OSNs across the anteroposterior (AP) extent of the OE at 5-day post-MeBr and in age-matched control mice. For each group, the number of P2 OSN profiles (y axis) was counted, binned at the location relative to the full AP extent of the epithelium, and presented as an average ± standard error of the mean (SEM) (x axis).

In contrast to the control mice, anti-β-gal-stained OSNs are absent at 5-day postexposure from areas that have been lesioned. As is typically observed with MeBr lesion in either mice or rats, some very small patches of the ventrolateral epithelium, including some within the P2 territory, are spared the effects of the exposure. Thus, the counts of P2 neurons across the epithelium are markedly reduced but not zero, and the average number of P2 OSNs at 5-day postlesion is 1650 or roughly 11% of control numbers (Figure 2), the vast majority of which are found in areas of the epithelium that had been spared the effects of the exposure. The distribution along the anteroposterior axis parallels, but its height is blunted by comparison, with the control tissue (Figure 2). Sequential double labeling with anti-βIII tubulin (Tuj1) or anti-PGP9.5 was used to assess the overall status of the lesioned epithelium (Figure 1E,F). At this time point, very few if any β-gal-labeled neurons are seen in areas that have been lesioned and are undergoing recovery despite the regeneration of a relatively substantial layer of marker-identified immature OSNs by this time (Figure 1F). These data are consistent with a delayed onset of P2 expression relative to the “birth” of the olfactory neurons as suggested by earlier experiments (Iwema and Schwob 2003).

The time course of P2-OR recovery after lesion

In order to acquire a more global assessment of the neuron recovery process following MeBr lesioning, whole septal mucosae from P2-ITL mice were harvested and stained at various time points during recovery. The septal tissue was labeled with anti-β-gal antibody, which was visualized with DAB, and a digital photographic mosaic of the mucosa was captured at relatively high magnification and assembled. For purposes of illustrating the distribution across the epithelium, each stained neuron in the whole-mount images was highlighted by segmentation above background and painted for ease of visualization, quantification, and comparison (Figure 3A). From a nadir at 5-day postlesion, there is a gradual repopulation of P2 OSNs with increasing survival time in a spatial pattern along the septum similar to the control tissue. The total number of P2 OSNs returns to near normal by 30-day postlesion. However, the number then trends below control at 2 months (Figure 3B). Although the number of P2 OSNs at 30- and 60-day postlesion is not statistically different compared with control tissue, there is a noticeable decline in total number of P2 OSNs by 90-day postlesion (Figure 5). It is worth noting that all of the newly generated P2 neurons bear a dendrite, and a process-less stained soma is never observed (Figure 3C). These data are consistent with the delay between neuron birth and detectable expression of the P2-ITL gene suggested above.

Figure 3.

Figure 3

Open in a new tab

The P2 OSN population approaches control numbers by 1 month after MeBr lesion but then declines. (A) The distribution of P2 OSNs was assessed across the epithelial sheet in whole mounts of the septal mucosa by immunostaining with anti-β-gal. After assembly of the digital mosaic, the image was segmented on the basis of staining density and painted for ease of visualization. P2 neurons were counted to generate the numbers plotted in (B). For purposes of illustration, a cartoon of the pattern of P2 expression across the septal sheet in an unlesioned control mouse and at 5-, 7-, 10-, and 30-day post–MeBr lesion was generated. (B) Total numbers of P2 OSNs summed across the septal epithelial sheet from controls and at various time points after MeBr lesion (average ± SEM, n = 4 for each time point except at 30 days where n = 3). The number of P2 OSNs increases with time after injury until 30 days where it peaks near control levels and then declines. Asterisk indicates statistical significance (P < 0.05) compared with control. (C) Paired photomicrographs of a single field of P2 OSNs from a septal epithelium whole mount 7 days after MeBr lesion for purposes of 3D imaging of newly born and labeled neurons. The photos were obtained after tilting the slide first one way and then the other, respectively. As a consequence, a 3D image of the field can be obtained by visually converging the 2 micrographs. The first new β-gal (+), P2 neurons emerge between 5 and 7 days after lesion, even though a population of new neurons is established by 5 days. Even at first appearance, the β-gal (+) OSNs have a well-developed dendrite, suggesting that OR expression lags by 1–2 days the onset of neuronal differentiation.

Figure 5.

Figure 5

Open in a new tab

The distribution of β-gal (+) P2 OSNs across the anteroposterior (AP) extent of the OE at 3-month post-MeBr and in age-matched control mice. For each group, the number of P2 OSN profiles (y axis) was counted, binned at the location relative to the full AP extent of the epithelium, and presented as an average ± SEM (x axis).

A more complete analysis of P2 OSN numbers after recovery from MeBr lesioning was performed by counting anti-β-gal-stained P2 neurons from coronal sections of control and 3-month recovery tissue according to the same protocol as used with the 5-day post-MeBr analysis above. By 3 months, the OE approaches normal thickness with the presence of scattered P2 neurons within the expected zonal pattern (Figure 4A–H). P2 OSNs from lesioned and control animals are identical in morphology with dendrites extending to knobs at the apical surface (Figure 4E,F). Average counts of P2-expressing OSNs stained with the anti-β-gal antibody demonstrate a similar peak in numbers in the midportion of the coronal series of the 3-month recovered and control sections as was also seen with the 5-day postlesion animals (Figure 5). The total average number of P2 OSNs after 3-month postlesion (n = 3) is 6400 compared with 17 000 in age-matched controls. This decrease in total P2 OSNs, occurring well after complete recovery from MeBr lesioning, does not appear to be related to an overall decline in mature olfactory neurons. Sections stained with anti-OMP to identify mature olfactory neurons from both the 3-month recovery and control animals demonstrate equal thickness of the layer of OMP (+) OSNs in both lesioned and control animals (Figure 4G,H). Measurements of epithelial thickness taken from the midportion of the septum from control and 3-month recovery animals are not significantly different (data not shown) indicating that P2 OSNs are reduced in number relative to other receptor types.

Figure 4.

Figure 4

Open in a new tab

The number of P2-expressing OSNs at 3 months of recovery is less than age-matched control, but the epithelium overall has recovered with respect to thickness and the population of OMP (+) OSNs. (A, C, E) Photomicrographs showing P2 expression in control epithelium (CI-118). (B, D, F) Photomicrographs from 3-month post-MeBr-lesioned animal (CI-108). The territory occupied by P2 (+) OSNs is demarcated by the dashed lines. Note that the territories are equivalent in the control (A) and lesioned-recovered (B) mice. (G, H) Sections from control and 3-month post-MeBr mice labeled with anti-OMP antibody demonstrate the return of normal or near-normal OE thickness and numbers of mature OSNs. Scale bars: A and B = 1mm, C and D = 200 µm, and E–H = 50 µm. Arrowhead = basal lamina.

P2 OSN axonal projections after lesion

As a first order estimate of whether the axons of P2 OSNs project to the appropriate glomerular location following recovery from MeBr lesion, X-gal stained whole-mount bulbs were examined in both 3-month post-MeBr-lesioned and control animals (Figure 6). In control mice, the pattern that we observe is highly consistent with the most recent published descriptions of the P2 OSN innervation (Mombaerts et al. 1996; Royal and Key 1999; Schaefer et al. 2001; Costanzo and Kobayashi 2010): when viewed as a whole mount, P2 OSNs project onto 1–4 discrete spots on the medial surface of the OB, of a size and appearance that has been interpreted as glomeruli, and an additional 1–4 glomeruli situated more anteriorly on the lateral surface (Figure 6A). Although the example illustrates a single lateral glomerulus, all of the control animals exhibited more than one P2 glomerulus on at least one surface of one OB (Table 1), a feature that was not emphasized in the original description (Mombaerts et al. 1996) but that is consistent with other published analyses of P2-ITL mice (Royal and Key 1999; Schaefer et al. 2001; Costanzo and Kobayashi 2010). Examination of serially sectioned tissue determined that in some cases, 2 P2 glomeruli are observed in the same section and are often contiguous, but in many cases, there is a distance of 25–150 µm along the anteroposterior axis that intervenes between labeled glomeruli. All P2 glomeruli in control animals are in the equivalent regions of medial and lateral OB from animal to animal even though the number and size of glomeruli are variable within the target zones. Some P2 glomeruli are quite small (~30 µm), whereas others are quite large (~128 µm); it is possible that some of the larger labeled glomeruli are actually 2 adjacent glomeruli. Nonetheless, all X-gal-labeled P2 axons in control subjects project precisely to glomeruli and no stray fibers were observed.

Figure 6.

Figure 6

Open in a new tab

β-gal (+), P2 olfactory axons fail to converge normally compared with age-matched control mice as shown in whole mounts of the OB stained with X-gal. (A) Convergence of P2 axons onto a single glomerulus (arrow) in the lateral surface of the right OB from a control mouse (CI-130M). (B, C) There is a lack of convergence of P2 axons onto 1 or even 2 glomeruli in the lateral surface of the right OB in 3-month post-MeBr-lesioned animals (B: CI-133M; C: CI-134M). Instead, numerous small terminal axonal aggregates are scattered across the surface of the bulb surrounding the location where the P2 glomerulus is normally located. Scale bars = 200 µm.

Table 1.

Number of P2 glomeruli and extent (in μm) of each glomerulus or collection of aberrantly projecting P2 axons

Subject Lateral Medial
Right OB Left OB Right OB Left OB
No. of glomeruli Extent No. of glomeruli Extent No. of glomeruli Extent No. of glomeruli Extent
Control
 CI-117M 2 82a, 82a 2 96a, 70a 2 82, 82 2 94a, 128a
 CI-118M 2 46, 70 2 43, 36 1 94 3 36, 30, 58
 CI-119M 1 48 3 70a, 70a, 24 2 106, 40 2 70, 60
 CI-120M 2 48, 94 2 94a, 96a 2 70, 70 3 70a, 24a, 82
Three-month post-MeBr
 CI-106M Axons 502 Axons 222 Axons 922 Axons 292
 CI-108M Axons 96 Axons 432 Axons 152 Axons 362
 CI-109M Axons 292 Axons 292 Axons 300 Axons 292
 CI-136M Axons 444 Axons 228 Axons 300 Axons 228
Open in a new tab

aDirectly adjacent glomeruli.

Conversely, in 3-month post-MeBr subjects, the P2 axons do not show the same convergence into glomeruli as controls despite targeting the appropriate medial and lateral regions of the OB (Figure 6B,C). Instead, the P2 axons project through the olfactory nerve layer to the generally correct region of the OB (either medial or lateral) at which point the axons appear to randomly project into numerous glomeruli. Some of the P2 axons fasciculate, but a proportion of them appears to either veer off or remain solitary. Initial outgrowth of P2 axons at earlier time points postlesion was not assessed for this study. In one animal examined 3 months after MeBr exposure, the P2 innervation pattern differs somewhat. In this case, the P2 axons are denser and attempt to converge onto 2 glomeruli on the lateral surface of the OB (Figure 6B), but many other smaller fascicles appear to terminate in multiple glomeruli.

We assessed the projection onto the OB and the innervation of glomeruli by the axons of P2 OSNs using the same set of mice analyzed above for epithelial recovery. Serial sections throughout the anteroposterior extent of 3-month post-MeBr and age-matched control OB were immunostained/DAB visualized with the anti-β-gal antibody and counterstained with hematoxylin, and the number and extent of P2 glomeruli were noted (Table 1). As mentioned, it was not uncommon to detect more than one P2 glomerulus on at least one surface of the OB in control tissue (Figure 7A,C). The illustrated example from the lateral surface of the OB (Figure 7A 1–5) demonstrates the existence of possibly 3 glomeruli, 2 of which are in the same coronal/anteroposterior location. One of the glomeruli is very small in diameter and may be an example of a microglomerulus (Lipscomb et al. 2002) (Figure 7A 4). Thus, the number and appearance of glomeruli in which P2 axons are observed is much more variable than previously described although others have also noted the existence of multiple P2 glomeruli (Royal and Key 1999; Costanzo and Kobayashi 2010). Nonetheless, in all examples of control tissue, there is no doubt that the majority if not all of the P2 axons project to and fill distinct glomeruli.

Figure 7.

Figure 7

Open in a new tab

Examination of coronal sections through the OB at 3 months after injury confirms the whole-mount findings that P2 axons fail to converge normally in MeBr-lesioned-recovered mice. Sections were stained with anti-β-gal antibody, visualized with DAB, and counterstained with hematoxylin. (A 1–5, C 1–6) Sections through the location of the P2-innervated glomeruli in the lateral bulb and medial bulb, respectively, of a control mouse. (B 1–7, D 1–6) Sections of lateral and medial OB, respectively, of a mouse lesioned with MeBr 3 months prior to harvest. The distances are from the most anterior point in the glomerular layer in which P2 axons are found. Arrows indicate glomeruli that are innervated by P2 axons. (A, C) The examples of age-matched controls were chosen to illustrate the most extreme examples of multiple and microglomeruli in the uninjured bulb; despite being somewhat atypical examples, nonetheless, the glomerular neuropil is filled with P2 axons in its entirety. Individual P2 glomeruli in each series are numbered as *1, *2, and so on. (B, D) In contrast, P2 axons in the lesioned animals occupy only a fraction of the neuropil in those glomeruli that they enter. In (A5), n = nerve cell layer, g = glomerular layer, and gc = granule cell layer. Scale bars = 100 µm.

In contrast, the distribution of P2 axons across the OB of 3-month postlesion mice is markedly different from controls. In parallel with the results obtained with the whole mounts (Figure 6B,C), analysis of the serial sections indicate that P2 axons do not project to specific glomeruli in the OBs on either the lateral (Figure 7B 1–7) or medial (Figure 7D 1–6) surfaces of lesioned mice. On the contrary, although numerous P2 axons are evident, most of the labeled fibers do not fasciculate but instead project independently into multiple glomeruli, which canvass a greater spread of the bulb than in controls, although situated in roughly the correct area. None of the glomeruli innervated by P2 fibers in the lesioned-recovered mice at 3-month post-MeBr exposure resembled the P2 glomeruli observed in control animals, that is, none of them receive a dense projection due to the convergence of a large contingent of P2 axons. The spread of P2 axons across multiple glomeruli was estimated for each surface of each bulb by measuring the anteroposterior extent of P2 axons in the glomerular layer. For the lesioned-recovered animals, the range of spread was 40–480 µm, with the majority of the projections covering well above 200 µm of anteroposterior distance (Table 1). These values far exceed the corresponding values for anteroposterior extent of P2 glomeruli in the age-matched control animals. In those glomeruli invaded by the marked fibers, the P2 axons occupy only a small portion of the glomerular neuropil, suggesting that axons of other OR subtypes occupy the rest of the glomerular neuropil. Indeed, OB sections that are double labeled for P2-derived axons (anti-β-gal antibody in brown) and mature OSN axons (anti-OMP in violet) emphasize the presence of other olfactory axons in the glomeruli partially innervated by P2 fibers postrecovery. In control tissue, P2 labeling and OMP labeling overlap completely (Figure 8A). However, in glomeruli of lesioned tissue, the portion not innervated by P2 axons contain OMP (+) staining presumably from axons of OR subtypes other than P2 (Figure 8B,C). The P2 axonal endings from lesioned-recovered mice are limited to the margins of the glomeruli where they collect in small aggregates that are notably denser than the wholly innervated glomeruli of control animals (Figure 8C). This is in contrast to the more diffuse innervation pattern seen in control mice and suggests an altered relationship with the dendritic tufts in the glomeruli and possibly a lack of synapse formation with the usual synaptic targets.

Figure 8.

Figure 8

Open in a new tab

In lesioned-recovered mice, P2 axons and non-P2 axons intermingle within glomeruli. Sections were stained with anti-β-gal antibody visualized with DAB (brown) and then subsequently stained with anti-OMP antibody visualized with VIP substrate (Vector) (purple). (A) In a P2 glomerulus of a control mouse, P2 axons completely fill the glomerulus and coincide with OMP staining. (B, C) In MeBr-lesion-recovered mice, glomeruli are only partially filled with P2 axons, which share the neuropil with other, non-P2, OMP (+) axons. Insets in (A) and (C) are higher magnification views near the respective glomerular margins demonstrating the enhanced density of staining in the setting of the lesioned-recovered mouse, suggesting that postsynaptic elements do not intermingle with them. Scale bar (white) in (A) = 25 µm.

Discussion

The experiments described here document the severe depletion of the P2 OSN population acutely following MeBr exposure, the time course and consequences of the recovery of P2 OSNs after lesion with respect to their number and distribution, and the P2 axonal projections in lesioned-recovered mice. From a nadir of roughly 10% of control at 5 days after MeBr, P2 numbers increase as the epithelium regenerates OSNs. In whole mounts of the septal mucosa, the numbers of P2 OSNs peak at 30-day postlesion but then decline and counts that sample the extent of the epithelium at 3 months of recovery indicate that P2 number falls to about 35% of age-matched controls by that time. It is worth noting that the β-gal-labeled neurons bear a well-developed dendrite from their earliest reappearance that extends near to, or reaches, the surface of the epithelium. From these data, it appears that expression of the P2 locus is delayed relative to the birth of the P2 OSNs, as was indicated by other analyses (Iwema and Schwob 2003). In addition, we have directly demonstrated that 5-ethynyl-2′-deoxyuridine-labeled P2 OSNs first appear in substantial numbers 3 days after the administration of the thymidine analog, which is several days after the onset of neuronal differentiation as monitored by neural cell adhesion molecule expression (Hewitt JC, Schwob JE, unpublished results). The distribution of the newly born P2 OSNs following regeneration is indistinguishable from age-matched control mice with reference to the location of the boundaries of P2 expression across the coronal plane and along the anteroposterior axis of the epithelium. Finally, at 3 months after epithelial injury, axons of P2 OSNs project onto the appropriate region of the OB, but glomerular targeting lacks precision, in that multiple glomeruli are innervated by P2 axons. Each of these results will be discussed in turn.

P2 OSNs in lesioned OE reappear in a normal spatial pattern but eventually their number falls below normal

That newly generated P2 OSNs are distributed across the epithelium of lesioned-recovered mice in a pattern that closely resembles age-matched controls is highly consistent with our previous in situ hybridization study of OR expression in the lesioned and 3-month-recovered epithelium of rats (Iwema et al 2004). The results of that examination, in which the OE on one side of the nose was lesioned by MeBr and the other was protected by naris occlusion during exposure, indicated that the boundaries that demarcate zones of OR expression are recapitulated after regeneration. The present results extend our prior observations by assessing what we know to be a single OR, as opposed to the possibility that several ORs might be marked by any single OR in situ probe. Thus, the current results also support the previously proposed hypothesis that local signals, which may derive from the lamina propria (which is not itself damaged in any detectable manner by MeBr exposure), constrain OR choice to the set of spatially appropriate ORs during the regeneration that follows extensive epithelial injury (Chen et al. 2004; Iwema et al 2004; Peluso et al. 2012).

The initial recovery and then substantial decline in the number of P2 OSNs that we observe at later times after lesion has not previously been described, although we have noted an initial modest overproduction and then winnowing of the population of OSNs as a whole at the end of the first month following MeBr lesion in rats (Schwob et al. 1995). Selective ablation of the P2 neurons via a genetic approach followed by cessation of toxin expression resulted in apparently full restoration of the population (Gogos et al. 2000). In the prior report describing P2 regeneration after dichlobenil lesion of the OE, errors in glomerular targeting were reduced at later survivals, although not eliminated (St John and Key 2003); that finding is not inconsistent with the decline in P2 OSNs that we observe, but a reduction in mistargeting is certainly an indirect measure of the P2 population at best. It is important to note that we observe no overall diminishment in mature olfactory neuron number in the lesioned-regenerated OE at the 3-month time point, given that overall epithelial height and the thickness of the zone occupied by OMP (+) region are equivalent in the control mice versus the 3-month recovery mice. The overall equivalency of the total population of OSNs and the initial overproduction of P2 OSNs followed by their decline suggest that the reduced number of P2 OSNs is not a general failure of neurogenesis but is more likely a selective effect on the stabilization of that population. In contrast to the P2 population, regeneration of multiple other groups of neurons, identified by in situ hybridization with cRNA probes, is much more complete when they stabilize; at 3 months, differences in number between MeBr-exposed and naris-closed sides of the epithelium are minor, and I7-probe (+) OSNs are even slightly increased on the MeBr-lesioned-recovered side at 3 months after toxin exposure (Iwema et al. 2004).

Newly generated P2 OSNs in lesioned-recovered OE project erroneously to the OB

Errors in targeting characterize the projection of the P2 OSNs onto the OB at 3 months of recovery following MeBr lesion. Even though the majority of control P2-ITL mice exhibit multiple P2 glomeruli at one or more of the 4 target surfaces of the bulbs as shown here and elsewhere (Royal and Key 1999; Costanzo and Kobayashi 2010), the number of glomeruli contacted by P2 OSNs in the lesioned-recovered mice is markedly increased and extends over a much greater distance along the anteroposterior axis of the bulb. The reinnervated glomeruli are centered near their usual target in the bulb, as shown by the whole mounts reacted with X-gal. The aberrations in the arrangement of the P2 fibers—their restriction to the edges of the glomeruli where they coalesce into tight aggregates that are markedly more condensed than in control or the small clusters of stray fibers—suggest that the axons fail to capture dendritic territory within these glomeruli, which are dominated by other, non-P2, OMP (+) axons. A similar conclusion was reached by St John and Key (2003) in their analysis of postlesion reinnervation of the bulb. In contrast to the permanent damage to the dorsomedial epithelium that accompanies dichlobenil-induced injury, that part of the epithelium is minimally damaged with MeBr lesion ruling out that feature as explanation for the mistargeting noted previously.

A similar mixing of fiber types has been observed elsewhere. Costanzo and Kobayashi (2010) described glomeruli that are partially innervated by P2 fibers in young and aged nonlesioned mice. In young mice, these aberrant projections are presumably pruned over time as the animals gain olfactory experience because the frequency of partially filled glomeruli is rare in middle-aged mice. However, the return of aberrant projections in elderly mice implies an underlying error in targeting or loss of guidance cues related to aging.

Why do P2 OSNs fail to recapture territory during the reinnervation of the bulb?

The results presented here and the previous report of mistargeting of P2 OSNs during recovery from epithelial lesion stand in stark contrast to the apparent specificity of reinnervation when only P2 OSNs die and are regenerated (Gogos et al. 2000). P2 OSNs are at an apparent disadvantage when the entire population of neurons of all receptor types is almost completely destroyed and then replaced. The excellent recovery of the neuronal population as a whole in the territory in which P2 neurons reside contrasts with the decline in the number of P2 OSNs from 1 to 3 months after lesion. Moreover, that disadvantage is also evident in the domination of glomerular neuropil by non-P2 axons in those several glomeruli that receive P2 innervation.

Multiple lines of evidence suggest that the comparative disadvantage of the P2 subpopulation may reflect an activity-dependent competition among OSNs for capturing glomerular synaptic territory. First, activity-based competition has been posited during the initial establishment of the glomerular map during development; evidence for this interpretation includes results from OCNC1 heterozygous knockout mice in which OCNC1-null neurons disappear from many (but not all) glomeruli during early postnatal life (Zhao and Reed 2001). Moreover, direct manipulation of odorant exposure has an impact on the sorting of axons within the glomerular layer of the bulb (Zou et al. 2004). Second, P2 OSNs appear to be insensitive to odorant stimuli in a standard laboratory environment, based on the scanty accumulation of S100A5 and Lrrc3b in P2-innervated glomeruli (Bennett et al. 2010). Third, exposure to the odorants geraniol, hexyl-acetate, 1-octanol, heptanal, and 2-hexanone is ineffective at increasing c-fos expression in P2 OSNs, in contrast to M72-expressing and I7-expressing OSNs; in both of the later populations, c-fos accumulation is observed with known ligands and with a broad array of odorants (Holbrook EH, Schwob JE, unpublished data). Fourth, P2 OSNs that are silenced by knockout of the OCNC1 channel are at no disadvantage by comparison with OCNC1-expressing P2 OSNs (Zheng et al. 2000); when the demonstration of P2 insensitivity is also taken into account, the comingling of silenced and OCNC1-expressing fibers might best be explained by the absence of differential activation between the two (particularly because silenced M72 OSNs segregate from channel-complemented M72 neurons under similar conditions; Zheng et al. 2000). Fifth, in mice genetically engineered to express tetanus toxin in all mature OSNs, synaptic vessel release is inhibited and glomerular targeting of receptor-specific OSNs proceeds normally (Yu et al. 2004). However, when the toxin is expressed exclusively within P2 OSNs, multiple targets are innervated and eventually the P2 axons disappear (Yu et al. 2004).

An alternative explanation for mistargeting and the subsequent decline in P2 neuron number, namely that a lack of odorant stimulation directly kills newly born P2 neurons when the population needs to be reconstituted en toto, is less likely because reinnervation is accurate when virtually all P2 neurons are killed by diphtheria toxin expression and then regenerated (Gogos et al. 2000).

Why is targeting so accurate when only P2 neurons die, but so aberrant, including the intermingling of P2 and non-P2 axons within individual glomeruli when all types of OSNs are lesioned by either MeBr or methimazole? In addition to the guidance mechanisms identified during the initial establishment of the glomerular map (Takeuchi and Sakano 2014), which may or may not persist into adulthood, the accuracy observed when the near-total but selective death of P2 OSNs is followed by regeneration may be fostered or even subserved by several mechanisms. First, there may be a preference for available synaptic space by the newly generated and reinnervating neurons, which is suggested by the preferential filling of more rostral glomeruli leaving posterior glomeruli hypo- or uninnervated when overall epithelial regeneration is incomplete (Schwob et al. 1999; McMillan Carr et al. 2004). Second, non-P2 axons may be inclined to track away from the empty P2 glomerulus and into adjacent glomeruli, for example, by following axons of like-neurons expressing the same, non-P2 receptor (Feinstein and Mombaerts 2004).

The cellular mechanism responsible for the decline in P2 number, as the population loses out in a competition for synaptic space because the P2 OSNs lack sensory stimulation, cannot be identified at present. However, we hypothesize that the deprivation of trophic support from the bulb, which is known to abbreviate the life span of OSNs (Schwob et al. 1992), is the ultimate consequence when the P2 neurons lose that competition. The interrelationship of axonal convergence and overall number of receptor-specific OSNs cannot be overlooked. An underrepresentation of P2 OSNs may result in decreased convergence and inability to maintain a glomerular domain, a concept termed interdependence (Ebrahimi and Chess 2000), and would in effect potentiate the decline in overall P2 number.

Funding

This work was supported by the National Institutes of Health [R01 DC000467 to J.E.S. and K08 DC008109 to E.H.H.].

Acknowledgements

The authors thank Dr P. Mombaerts for his gift of P2-ITL breeder mice, Dr D. Kurtz for his advice on statistical analysis, and Dr S. Youngentob for use of equipment. We also thank members of the Schwob lab for their comments on the manuscript.

References

  1. Abercrombie M. 1946. Estimation of nuclear population from microtome sections. Anat Rec. 94:239–247 [DOI] [PubMed] [Google Scholar]
  2. Bennett MK, Kulaga HM, Reed RR. 2010. Odor-evoked gene regulation and visualization in olfactory receptor neurons. Mol Cell Neurosci. 43(4):353–362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Burd GD. 1993. Morphological study of the effects of intranasal zinc sulfate irrigation on the mouse olfactory epithelium and olfactory bulb. Microsc Res Tech. 24(3):195–213 [DOI] [PubMed] [Google Scholar]
  4. Chen X, Fang H, Schwob JE. 2004. Multipotency of purified, transplanted globose basal cells in olfactory epithelium. J Comp Neurol 469(4):457–474. [DOI] [PubMed] [Google Scholar]
  5. Cheung MC, Jang W, Schwob JE, Wachowiak M. 2014. Functional recovery of odor representations in regenerated sensory inputs to the olfactory bulb. Front Neural Circuits. 7:207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Christensen MD, Holbrook EH, Costanzo RM, Schwob JE. 2001. Rhinotopy is disrupted during the re-innervation of the olfactory bulb that follows transection of the olfactory nerve. Chem Senses. 26(4):359–369 [DOI] [PubMed] [Google Scholar]
  7. Costanzo RM. 1991. Regeneration of olfactory receptor cells. Ciba Found Symp. 160:233–242; discussion 243. [DOI] [PubMed] [Google Scholar]
  8. Costanzo RM. 2000. Rewiring the olfactory bulb: changes in odor maps following recovery from nerve transection. Chem Senses. 25(2):199–205 [DOI] [PubMed] [Google Scholar]
  9. Costanzo RM, Kobayashi M. 2010. Age-related changes in p2 odorant receptor mapping in the olfactory bulb. Chem Senses. 35(5):417–426 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ebrahimi FA, Chess A. 2000. Olfactory neurons are interdependent in maintaining axonal projections. Curr Biol. 10(4):219–222 [DOI] [PubMed] [Google Scholar]
  11. Feinstein P, Mombaerts P. 2004. A contextual model for axonal sorting into glomeruli in the mouse olfactory system. Cell. 117(6):817–831 [DOI] [PubMed] [Google Scholar]
  12. Gogos JA, Osborne J, Nemes A, Mendelsohn M, Axel R. 2000. Genetic ablation and restoration of the olfactory topographic map. Cell. 103(4):609–620 [DOI] [PubMed] [Google Scholar]
  13. Graziadei GA, Graziadei PP. 1979b. Neurogenesis and neuron regeneration in the olfactory system of mammals. II. Degeneration and reconstitution of the olfactory sensory neurons after axotomy. J Neurocytol. 8(2):197–213 [DOI] [PubMed] [Google Scholar]
  14. Graziadei PP, Graziadei GA. 1979a. Neurogenesis and neuron regeneration in the olfactory system of mammals. I. Morphological aspects of differentiation and structural organization of the olfactory sensory neurons. J Neurocytol. 8(1):1–18 [DOI] [PubMed] [Google Scholar]
  15. Graziadei PP, Monti Graziadei GA. 1985. Neurogenesis and plasticity of the olfactory sensory neurons. Ann N Y Acad Sci. 457:127–142 [DOI] [PubMed] [Google Scholar]
  16. Harding JW, Getchell TV, Margolis FL. 1978. Denervation of the primary olfactory pathway in mice. V. Long-term effect of intranasal ZnSO4 irrigation on behavior, biochemistry and morphology. Brain Res. 140(2):271–285 [DOI] [PubMed] [Google Scholar]
  17. Harding JW, Wright JW. 1979. Reversible effects of olfactory nerve section on behavior and biochemistry in mice. Brain Res Bull. 4(1):17–22 [DOI] [PubMed] [Google Scholar]
  18. Hinds JW, Hinds PL. 1976a. Synapse formation in the mouse olfactory bulb. I. Quantitative studies. J Comp Neurol. 169(1):15–40 [DOI] [PubMed] [Google Scholar]
  19. Hinds JW, Hinds PL. 1976b. Synapse formation in the mouse olfactory bulb. II. Morphogenesis. J Comp Neurol. 169(1):41–61 [DOI] [PubMed] [Google Scholar]
  20. Iwema CL, Fang H, Kurtz DB, Youngentob SL, Schwob JE. 2004. Odorant receptor expression patterns are restored in lesion-recovered rat olfactory epithelium. J Neurosci. 24(2):356–369 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Iwema CL, Schwob JE. 2003. Odorant receptor expression as a function of neuronal maturity in the adult rodent olfactory system. J Comp Neurol. 459(3):209–222 [DOI] [PubMed] [Google Scholar]
  22. Jang W, Youngentob SL, Schwob JE. 2003. Globose basal cells are required for reconstitution of olfactory epithelium after methyl bromide lesion. J Comp Neurol. 460(1):123–140 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kawano T, Margolis FL. 1982. Transsynaptic regulation of olfactory bulb catecholamines in mice and rats. J Neurochem. 39(2):342–348 [DOI] [PubMed] [Google Scholar]
  24. Koster NL, Costanzo RM. 1996. Electrophysiological characterization of the olfactory bulb during recovery from sensory deafferentation. Brain Res. 724(1):117–120 [DOI] [PubMed] [Google Scholar]
  25. Kream RM, Margolis FL. 1984. Olfactory marker protein: turnover and transport in normal and regenerating neurons. J Neurosci. 4(3):868–879 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lipscomb BW, Treloar HB, Greer CA. 2002. Novel microglomerular structures in the olfactory bulb of mice. J Neurosci. 22(3):766–774 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. McMillan Carr V, Ring G, Youngentob SL, Schwob JE, Farbman AI. 2004. Altered epithelial density and expansion of bulbar projections of a discrete HSP70 immunoreactive subpopulation of rat olfactory receptor neurons in reconstituting olfactory epithelium following exposure to methyl bromide. J Comp Neurol. 469(4):475–493 [DOI] [PubMed] [Google Scholar]
  28. Miyamichi K, Serizawa S, Kimura HM, Sakano H. 2005. Continuous and overlapping expression domains of odorant receptor genes in the olfactory epithelium determine the dorsal/ventral positioning of glomeruli in the olfactory bulb. J Neurosci. 25(14):3586–3592 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mombaerts P, Wang F, Dulac C, Chao SK, Nemes A, Mendelsohn M, Edmondson J, Axel R. 1996. Visualizing an olfactory sensory map. Cell. 87(4):675–686 [DOI] [PubMed] [Google Scholar]
  30. Monti Graziadei GA, Karlan MS, Bernstein JJ, Graziadei PP. 1980. Reinnervation of the olfactory bulb after section of the olfactory nerve in monkey (Saimiri sciureus). Brain Res. 189(2):343–354 [DOI] [PubMed] [Google Scholar]
  31. Nadi NS, Head R, Grillo M, Hempstead J, Grannot-Reisfeld N, Margolis FL. 1981. Chemical deafferentation of the olfactory bulb: plasticity of the levels of tyrosine hydroxylase, dopamine and norepinephrine. Brain Res. 213(2):365–377 [DOI] [PubMed] [Google Scholar]
  32. Peluso CE, Jang W, Drager UC, Schwob JE. 2012. Differential expression of components of the retinoic acid signaling pathway in the adult mouse olfactory epithelium. Exp Neurol. 214(1):25–36 [DOI] [PMC free article] [PubMed]
  33. Ressler KJ, Sullivan SL, Buck LB. 1993. A zonal organization of odorant receptor gene expression in the olfactory epithelium. Cell. 73(3):597–609 [DOI] [PubMed] [Google Scholar]
  34. 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(7):1245–1255 [DOI] [PubMed] [Google Scholar]
  35. Ring G, Mezza RC, Schwob JE. 1997. Immunohistochemical identification of discrete subsets of rat olfactory neurons and the glomeruli that they innervate. J Comp Neurol. 388(3):415–434 [PubMed] [Google Scholar]
  36. Royal SJ, Key B. 1999. Development of P2 olfactory glomeruli in P2-internal ribosome entry site-tau-LacZ transgenic mice. J Neurosci. 19(22):9856–9864 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Schaefer ML, Finger TE, Restrepo D. 2001. Variability of position of the P2 glomerulus within a map of the mouse olfactory bulb. J Comp Neurol. 436(3):351–362 [PubMed] [Google Scholar]
  38. Schwob JE, Szumowski KE, Stasky AA. 1992. Olfactory sensory neurons are trophically dependent on the olfactory bulb for their prolonged survival. J Neurosci. 12(10):3896–3919 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Schwob JE, Youngentob SL, Mezza RC. 1995. Reconstitution of the rat olfactory epithelium after methyl bromide-induced lesion. J Comp Neurol. 359(1):15–37 [DOI] [PubMed] [Google Scholar]
  40. Schwob JE, Youngentob SL, Ring G, Iwema CL, Mezza RC. 1999. Reinnervation of the rat olfactory bulb after methyl bromide-induced lesion: timing and extent of reinnervation. J Comp Neurol. 412(3):439–457 [DOI] [PubMed] [Google Scholar]
  41. St John JA, Key B. 2003. Axon mis-targeting in the olfactory bulb during regeneration of olfactory neuroepithelium. Chem Senses. 28(9):773–779 [DOI] [PubMed] [Google Scholar]
  42. Strotmann J, Conzelmann S, Beck A, Feinstein P, Breer H, Mombaerts P. 2000. Local permutations in the glomerular array of the mouse olfactory bulb. J Neurosci. 20(18):6927–6938 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Strotmann J, Konzelmann S, Breer H. 1996. Laminar segregation of odorant receptor expression in the olfactory epithelium. Cell Tissue Res. 284(3):347–354 [DOI] [PubMed] [Google Scholar]
  44. Strotmann J, Wanner I, Helfrich T, Beck A, Breer H. 1994. Rostro-caudal patterning of receptor-expressing olfactory neurones in the rat nasal cavity. Cell Tissue Res. 278(1):11–20 [DOI] [PubMed] [Google Scholar]
  45. Strotmann J, Wanner I, Helfrich T, Beck A, Meinken C, Kubick S, Breer H. 1994. Olfactory neurones expressing distinct odorant receptor subtypes are spatially segregated in the nasal neuroepithelium. Cell Tissue Res. 276(3):429–438 [DOI] [PubMed] [Google Scholar]
  46. Strotmann J, Wanner I, Krieger J, Raming K, Breer H. 1992. Expression of odorant receptors in spatially restricted subsets of chemosensory neurones. Neuroreport. 3(12):1053–1056 [DOI] [PubMed] [Google Scholar]
  47. Takeuchi H, Sakano H. 2014. Neural map formation in the mouse olfactory system. Cell Mol Life Sci. PMID: 24638094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Vassar R, Chao SK, Sitcheran R, Nuñez JM, Vosshall LB, Axel R. 1994. Topographic organization of sensory projections to the olfactory bulb. Cell. 79(6):981–991 [DOI] [PubMed] [Google Scholar]
  49. Vassar R, Ngai J, Axel R. 1993. Spatial segregation of odorant receptor expression in the mammalian olfactory epithelium. Cell. 74(2):309–318 [DOI] [PubMed] [Google Scholar]
  50. Vedin V, Slotnick B, Berghard A. 2004. Zonal ablation of the olfactory sensory neuroepithelium of the mouse: effects on odorant detection. Eur J Neurosci. 20(7):1858–1864 [DOI] [PubMed] [Google Scholar]
  51. Wang F, Nemes A, Mendelsohn M, Axel R. 1998. Odorant receptors govern the formation of a precise topographic map. Cell. 93(1):47–60 [DOI] [PubMed] [Google Scholar]
  52. Xie F, Fang C, Schnittke N, Schwob JE, Ding X. 2013. Mechanisms of permanent loss of olfactory receptor neurons induced by the herbicide 2,6-dichlorobenzonitrile: effects on stem cells and noninvolvement of acute induction of the inflammatory cytokine IL-6. Toxicol Appl Pharmacol. 272(3):598–607 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Yee KK, Costanzo RM. 1995. Restoration of olfactory mediated behavior after olfactory bulb deafferentation. Physiol Behav. 58(5):959–968 [DOI] [PubMed] [Google Scholar]
  54. Youngentob SL, Schwob JE. 1997. Changes in odorant quality perception following methyl bromide induced lesions of the olfactory epithelium. Chem Senses. 22:830–831 [DOI] [PubMed] [Google Scholar]
  55. Yu CR, Power J, Barnea G, O’Donnell S, Brown HE, Osborne J, Axel R, Gogos JA. 2004. Spontaneous neural activity is required for the establishment and maintenance of the olfactory sensory map. Neuron. 42(4):553–566 [DOI] [PubMed] [Google Scholar]
  56. Zhao H, Reed RR. 2001. X inactivation of the OCNC1 channel gene reveals a role for activity-dependent competition in the olfactory system. Cell. 104(5):651–660 [DOI] [PubMed] [Google Scholar]
  57. Zheng C, Feinstein P, Bozza T, Rodriguez I, Mombaerts P. 2000. Peripheral olfactory projections are differentially affected in mice deficient in a cyclic nucleotide-gated channel subunit. Neuron. 26(1):81–91 [DOI] [PubMed] [Google Scholar]
  58. Zou DJ, Feinstein P, Rivers AL, Mathews GA, Kim A, Greer CA, Mombaerts P, Firestein S. 2004. Postnatal refinement of peripheral olfactory projections. Science. 304(5679):1976–1979 [DOI] [PubMed] [Google Scholar]

Articles from Chemical Senses are provided here courtesy of Oxford University Press

RESOURCES