0306-4522/90 $3.00 + 0.00 Pergamon Press plc 0 1990 IBRO
Neuroscience Vol. 38, No. 3, pp. 629-641, 1990 Printed in Great Britain
LOCALIZATION OF NEUROPEPTIDE PRECURSORSYNTHESIZING NEURONS IN THE RAT OLFACTORY BULB: A HYBRIDIZATION HISTOCHEMICAL STUDY E. SENBA,* D. M. SIMMONSand L. W. SWAN.WN The Howard
Hughs Medical Institute, The Salk Institute for Biological Studies, La Jolla, CA 92037, U.S.A.
Abstract-The distribution of seven kinds of neuropeptide precursor mRNA-containing neurons was investigated in the rat main and accessory olfactory bulbs, where various peptides have previously been identified immunohistochemically, by means of in situ hybridization using [%]cRNA probes. In the glomeruiar layer, numerous preprothyrotropin-releasing hormone mRNA-expressing neurons, moderate numbers of preprosomatostatin and preproenkephalin A neurons, and a small number of preprocholecystokinin neurons were detected. In the external plexiform layer, numerous medium sized preprocholecystokinin and preprocorticotropin-releasing hormone neurons, and a small number of /I-preprotachykinin A neurons were observed. In addition, small preprovasoactive intestinal polypeptide and preprothyrotropin-releasing hormone neurons were evenly distributed in the external plexiform layer. Medium to large sized /I-preprotachykinin A neurons formed a thin layer in the mitral cell layer. In the granule cell layer, in addition to numerous small preproenkephalin A neurons, moderate numbers of small )?I-preprotachykinin A and pmprocorticotropin-releasing hormone neurons, and a small number of preprothyrotropin-releasing hormone neurons, were identified. Large sized preprosomatostatin neurons were located in the deep layer of the granule cell layer. The distribution patterns of these neurons, as a whole, confirmed previous studies based on immunohistochemistry, although peptide precursor mRNA-expressing neurons were far more numerous than those immunoreactive to the respective neuropeptides. Moreover, mRNA-expressing neurons were observed in areas where no immunoreactive neurons had been observed (e.g. preprovasoactive intestinal polypeptide and preprosomatostatin neurons in the mitral cell layer of the assessory olfactory bulb). The distribution patterns were generally similar in the main and accessory olfactory bulbs.
Because of the simple laminated structure of the olfactory bulb (OB), individual cell types are well characterized on both morphological and physiological grounds. ” For this reason, the OB has provided a useful model for investigating the functional significance of multiple neuropeptides within individual cells. Immunocytochemical studies have recently demonstrated that many neuroactive peptides are localized within intrinsic and output neurons of the main and accessory OB (see Refs 21 and 35 for a
*To whom correspondence should be addressed at: Department of Anatomy (II), Osaka University Medical School, 4 Nakanoshima Kitaku, Osaka, Japan. Abbreviations: AOB, accessory olfactory bulb; CCK-8, cholecystokinin-8; cDNA, complementary DNA; CRH, corticotropin-releasing hormone; cRNA, complementary RNA; DTT, dithiothreitol; EDTA, ethylene diaminetetra-acetate; ENK, enkephalin; EPL, external plexiform layer; GL, glomerular layer; GRL, granule cell layer; IPL, internal plexiform layer; IR, like immunoreactive/immunoreac&ity; Met-BNK, [Metlenkephalin; MCL, mitral cell layer: MGC. modified alomerular complex; MOB, main olfactory bulb; OB,- olfactory bulb; PHI, peptide histidine-leucine; pp, prepro; ppCCK, preprocholecystokinin; ppENK, proproenkephalin A; SP, substance P; SSC, sodium chloride, 0.15 M/sodium citrate, 0.015 M, pH = 7; TRH, thyrotropin-releasing hormone; VIP, vasoactive intestinal polypeptide.
review). These peptides may play important roles in the regulation of olfactory function and olfactory related behaviors, such as mating, maternal care, feeding, and many aspects of social behaviors.‘2,43*56 However, disagreements and uncertainties about the presence and distribution of neuropeptides still exist. One of the difficulties associated with immunocytochemistry is that it is sometimes very difficult to detect peptide-synthesizing cell somata for at least two possible reasons. Firstly, neuropeptides synthesized in the perikarya are actively transported to axon terminals, and their concentrations in the perikarya are in many cases too low to be detected. Thus, large amounts of colchicine are often required to visualize them. And secondly, big precursor molecules are sometimes not recognized by antibodies. The present study was undertaken to determine the localization and quantity of neuropeptide-synthesizing neurons in the rat OB under normal conditions, by means of the in situ hybridization method, which visualizes mRNA-containing cell bodies. Information about the distribution of neuropeptides in the accessory olfactory bulb (AOB) is still limited compared to that of the main olfactory bulb (MOB). Therefore, the distribution patterns of neuropeptides were compared in both compartments, in order to elucidate general principles.
629
E. SENBAet al.
630 EXPERIMENTAL
PROCEDURES
Animals and tissue preparation
Adult male Sprague-Dawley rats weighing 25&3OOg (n = 4) were used. They were allowed free access to water and food (Teklad mouse/rat diet). For perfusion, animals were deeply anesthetized with chloral hydrate (0.1 ml of 35% solution/lOOg body weight) and perfused transcardially using a pH shift method described in detail elsewhere.16 Briefly, following 50ml of saline, ice-cold 4% paraformaldehyde in 0.1 M acetate buffer at pH 6.5 (200 ml), and then 4% paraformaldehyde plus 0.05% glutaraldehyde in 0.1 M borate buffer at pH 9.5 (400 ml), were perfused. The brains were quickly removed and were postfixed overnight at 4°C in 4% paraformaldehyde plus 10% sucrose in 0.1 M borate buffer at pH 9.5. The forebrains containing the OB were frozen with powdered dry ice, and 20-pm-thick serial sagittal sections were cut on a sliding microtome, and collected in ice-cold potassium phosphate-buffered saline. Ten series of sections were obtained from each animal. One series was always mounted and stained with 1% Thionin to provide a Nisslstained series for cytoarchitectonic analysis of the OB. The remaining series were processed for hybridization histochemistry, including control experiments. They were mounted or subbed, poly-L-lysine-coated slides as soon as possible after cutting. The mounted sections were allowed to air-dry for l(r15min and were stored overnight at room temperature under vacuum with desiccant. Preparation of probes
A total of seven different [3SS]complementary RNA ([‘5S]cRNA) probes complementary (anti-sense) to their respective mRNAs were used in this study. They were transcribed from linearized DNA templates using Riboprobe Gemini System kits purchased from Promega (Madison, WI). Linearized templates were purified by gel electrophoresis, precipitated in ethanol, reconstituted in Tris-EDTA buffer to a final concentration of 1 pg/pl, and stored at -20°C. I’sSIUTP (100&15OOCi/mM: New England Nuclear; Bostdn, MA) and unlabeled ‘ATPi GTP and CTP were used in the transcription reaction. Probe templates were cDNA inserts into pGem (Promega) or SP64 plasmids for RNA transcription as described in our previous papem.53,55Briefly, plasmids containing a full length complementary DNA (cDNA) insert (528 bp) corresponding to nucleotides 17-544 of the rat preprocholecystokinin (ppCCK) sequence (from J. E. Dixon, Purdue University; see Ref. 1l), and 935 bp fragment corresponding to nucleotides - 104-831 of rat preproenkephalin A (ppENK) cDNA (from S. L. Sabol, N. I. H.; see Ref. 63) and a 380 bp fragment including the entire vasoactive intestinal polypeptide (VIP)- and peptide histidineleucine (PHI-peptide coding regions (from R. H. Goodman, Tufts University; see Ref. 42) were linearized, and SP6 RNA polymerase was used to transcribe respective anti-sense RNA probes. A 380 bp fragment including the entire corticotropin-releasing hormone-41 (CRH-41) peptide coding region and part of the 3’ untranslated region of rat CRH mRNA (from K. E. Mayo, Northwestern University; see Refs 26 and 58) and a 508 pb fragment containing the entire coding region which corresponds to nucleotides - 74- + 434 of the rat fi-preprotachykinin A (/?-pptachykinin) sequence (from J. E. Krause, Washington University; see Ref. 31) were linearized and T7 RNA polymerase was used to transcribe ?S-labeled riboprobe complementary to preproCRH (ppCRH) and I-preprotachykinin (j?-ppT) mRNA, respectively. For preprosomatostatin, original cDNA (from J. F. Habener, Harvard Medical School; see Ref. 18’)was cut at XbaI and AvaI sites and subcloned into pGEMII1 and then linearized with Hind III, and T7 RNA polymerase was used to transcribe anti-sense probe. For preprothyrotropinreleasing hormone (ppTRH) SP64 plasmid containing a 1241 bp sequence of ppTRH cDNA (from R. H. Goodman,
Tufts University; see Ref. 50) was linearized with Hind III, and riboprobe was transcribed using SP6 polymerase. These probes were adjusted to concentrations of 5 x 106-10’ cpm/ ml, which provided the greatest signal-to-noise ratio. For control experiments, sense-strand probes were used, in substitution for the above described anti-sense probes. Hybridization histochemislry
The hybridization procedure used in the present study is based on that described by Angerer et al.’ and Simmons et aLS3 Before hybridization, vacuum-dried sections were treated with proteinase K for 30 min at 37”C, acetylated for 10 min at room temperature, and dehydrated in a graded series of ethanols (prehybridization step). About 80~1 of hybridization solution was applied to a coverslip, which was then placed over the sections mounted on the slide. Edges of the coverslip were sealed with liquid DPX mountant (Gallard and Schlessinger, NY). The slides were then incubated on a covered slide-warming tray at 55do”C for 16-20 h. The hybridization solution contained 50% formamide, 10% dextran sulfate, 0.3 M NaCl, 1 x Denhardt’s solution, 10mM Tris (pH 8.0), 1 mM EDTA, 500 pg/ml tRNA, 10 mM dithiothreitol (DTT), and the probe. After the hybrid&ion incubation,- the coverslips were removed in NaCl/Na citrate (SSC) solutions. The sections were digested with ribonucleak A ‘(0.02 mg/ml) for 30 min. at 37°C and rinsed in descending concentrations of SSC at room temperature. They were then incubated in 0.1 x SSC at 60°C for 30min, dehydrated in an ascending series of ethanols and dried under vacuum at room temperature for 30 min. All SSC rinse solutions following ribonuclease treatment contained 1 mM DTT. Autoradiography
The sections were de-lipidized in 95% ethanol (5 min), followed by absolute ethanol (3 x 5 min) xylene (10 min), fresh xylene (30 min), and absolute ethanol (3 x 5 min). They were then air-dried for at least 30 min, and dipped in Kodak autoradiography emulsion (NTB-2), diluted 1: 1 with distilled water. After an exposure time of one to three weeks, the slides were developed in Kodak D-19 (3.5 min at 14”C, fixed in Kodak Rapid Fix, rinsed for 1 h in running distilled water, and stained through the emulsion with 0.1% Thionin. Because of the digestion of RNAs by ribonuclease A treatment, only cell nuclei were stained. RESULTS
The seven probes described above labeled a variety of neurons in the rat OB. Background hybridization was low and similar to that seen when the sensestrand probes were used. Preprocholecystokinin
mRNA-containing
neurons
Main olfactory bulb. The largest collection of medium sized ppCCK mRNA-containing neurons (ppCCK neurons) (15-20 pm) was located within a border area between the glomerular layer (GL) and external plexiform layer (EPL), and in deeper lamina of the EPL, forming two distinct layers in the EPL (Fig. la,e). About twice as many neurons were observed in the outer layer as compared to the inner layer. A cluster of labeled cells was localized in the EPL of the border area between the MOB and AOB, the so-called modified glomerular complex (MGC)20.57(Fig. lb,c). Small sized neurons (S-10 pm) were also observed in the GL (Fig. Id) and in the granule cell layer (GRL).
Neutopeptide mRNAs in the rat olfactory bulb Accessory o~ctu~~ bulb. Almost no labeled cells were id&W in the AOB, except a few lightly labeled ones scattered in the mitral cell layer (MCL) (Fig. la,b). Preprocorticotropin-releusing hormone mRNA -confaining neuron Main olfactory bulb. ppCRH mRNA-~nt~ni~~ neurons (ppCRH neurons) were numerous and widely distributed in the MOB. in the EPL and MCL, hbe~ed neurons formed two distinct layers, extema1 and internal (including the MCL) {Fig. 2a,d). In contrast to the ppCCK mRNA containing neurons described above, about equal numbers of labeled cells were observed in each layer. Some of them were medium in size (IS-20 pm), while most of them were small (about IOgm) (Fig. 2d). A cluster of labeled cells was also located in the MGC (Fig. Zb). Most of these cells were medium in size (2zOpm). Small iabeled cells (IOgm) were scattered in the deeper two-thirds of the CRL (Fig. 2a,d). Accessory olfactory bulb. Medium sized labeled neurons (20 pm) were observed in the MCL (Fig. 2b). In addition, a few small celis (10 pm) were also scattered in the CRL (Fig. 2~).
631
~epr~fhyr4tr~pin-re~e~~ng hormone ~~A-~o~#~~ing neurons Main o~c~Qry bulb. The MOB of the rat was rimmed by ppTRH mRNA-containing neurons, wh&h were den&y packed between ~#rn~li in the CL These neurons were small (about IOgm) and clearly demarcated the borders of the glomeruli (Fig. Ja,d,e). Small labeled cells were also observed in the EPL (Fig. 3a,d,e), in w&h these ne~ons were evenly distributed. Even less n~erou$ smaff sized neurons (10ym) were scattered in the GRL (Fig. 3a,d). dccessory o~c~~ry bulb. Small clusters of labekd cehs (10 pm) were observed surroundjng glomeruti in the GL, and labeled cells of similar size were distributed in the deeper half of the MCL and also in the GRL (Fig. 3b,c).
Ma& of$acctozy bulb. Most of the ppVIP mRNAcontaining neurons (ppVIP neurons) were confined to the EPL. These neurons were small to medium in
Fig. 1. Photomicrographs of dipped autoradiographs to show hybridization signals for ppCCK mRNA. c is a higher magnification of MGC in b. e is a bright-field photomicrograph of the same area as d. Scale bars = 500 pm (a); 250 pm fb); IOOpm &-ef.
632
E. SENB.4ei al.
size (IO-15 pm) and evenly distributed in the layer, forming a bundle of labeled neurons (Fig. 4a). Small sized labeled neurons (about 10fi.m) were occasionally observed in the CL and GRL (Fig. 4a). Accessory ol$uctory bulb. Lightly labeled neurons (10-15 urn) were evenly scattered in the MCL (Fig. 4b,c). Preproson~at~st~t~nmRNA -containing neuron Main olfactory bulb. Two kinds of preprosomatostatin (ppsomatostatin mRNA-containing neurons (ppsomatostatin neurons) were observed in the MOB. One consisted of small neurons (about 10 pm), which dist~buted around the glomeruii in the GL (Fig. 5a,d). The other consisted of densely labeled large neurons (Z@-30pm) located in the deepest part of the GRL (Fig. 5a,b); similar neurons were occasionally found in the CL (Fig. 5a). Accessory olfactory bulb. Labeled cells of small to medium size (IO-20 Frn) were concentrated in the MCL (Fig. 5a-c). ,lI-Preprotachykinin A mRNA -containing neurons Main olfactory bulb. A row of densely labeled ~-pptachykinin mRNA-containing neurons (B-
pptachykinin neurons) was observed in the MCL (Fig. 6a,c). These neurons were medium to large in size (15-30 pm). Some medium sized labeled neurons (about 1.5pm) were also scattered in the EPL (Fig. 6a,c). Lightly labeled small neurons (IOgm) were distributed in the deeper layer of the GRL (Fig. 6a). Accessory olfactory bulb. Densely labeled small cells (10pm) were packed in the GRL (Fig. 6a,b). Medium sized labeled neurons (15 pm) were also observed in the MCL (Fig. 6a,b). These neurons
tended to be located in the posterior half of this layer. Preproenkephalin A mRNA -containing neurons Main olfactory bulb. Densely packed preproenkephafin A (ppENK) mRNA-containing neurons were observed in the GRL (Fig. 6d,e). Labeling was more intense in the superficial one-third of the GRL, including the internal plexiform layer (IPL) and MCL, in which these neurons seemed to surround mitral cells (Fig. 6f). Small labeled cells (5-10pm) were distributed around glomeruli in the GL (Fig. 6d,f). Accessory oifactory bulb. Small labeled cells (about 10 fern) were diffusely distributed in the GRL and IPL
Fig. 2. Dark-field photomicrographs of dipped autoradiographs to show hybridization signals for ppCRH mRNA. Scale bars = 500 pm (a); 100pm (b-d).
Neuropeptide mRNAs in the rat olfactory bulb
(Fig. 6d,e). A few of them were also observed in the GL. The distributions of these peptide precursor mRNAs, which were quite similar in the OR sections from four animals, are thematically summarized in Fig. 7.
DISUNION
This report is the first compilation of results based on irr situ hybridization for the mRNAs of various
633
neuropeptide precursor molecules in the rat OB. In comparing the present results with earlier reports based on immun~yt~he~st~, evidence for the greater sensitivity of the in sifu hybridization method was provided, in view of the fact that peptide precursor mRNA-containing neurons were far more numerous than neurons reported to be immunoreactive for peptide end-products, although it is not known whether all of these mRNA expressing neurons synthesize end-products. Moreover, IabeIed cells were also observed in areas in which no immunorea~tive neurons have been reported.
Fig. 3. Dark-field photomicrographsshowing hybridization signals for ppTRH mRNA. c is a higher magnification of the squared area in b. Scale bars = 500~rn (a); 250 pm (b,d); 100pm (c,e).
E. SENBA ff al.
634
Preprocholecystokinin
The content of cholecystokinin-like immunoreactivity in the rat OB is relatively high, and is about one-half to one-third that measured in the isocortex, which shows the highest levels in the CNS.’ The majority of the cholecystokinin-&like immunoreact~ve (CCK-I-IR) neurons in the EPL have been identified as middle tufted cells. In addition, external and internal tufted cells have also been shown to be immunoreactive for CCK-8.” It is generally accepted that middle and internal tufted cells are projection neurons, along with mitral cells. Indeed, these CCK-8-IR tufted cells have been shown to project to anterior parts of the bulbar target regions.39 In addition to these projection neurons, some intrinsic neurons, such as periglomerular cells, superficial short axon cells in the GL, and deep short axon cells in the GRL, have also been shown to be CCK-8-IR.37~38~5’The present study confirmed the presence of these neurons in the OB. It has been reported in previous morphological studies that there are no tufted cells in the AOB.35 The
present finding that few, if any, ppCCK neurons can be identified in the AOB, in contrast to the abundance of ppCCK containing tufted cells in the MOB, appears to support this previous suggestion. Prepr~~~rt~c~tropin-releasing hormone
In contrast to other neuropeptides, the content of CRH-IR in the OB has been reported to be relatively low.54 Few CRH-IR neurons were observed in the OB in the previous immunohistochemical distribution studies in the rat brain.40,47The latter authors stated that CRH-IR structures in the OB were less intense than in other brain regions. The present study, however, revealed a widespread distribution of ppCRH neurons in the OB, confirming a recent immunohistochemical study combined with in situ hyb~di~t~on histochemistry,25 which suggested that most of the ppCRH neurons in the MCL and EPL are mitral cells or tufted cells, and that a subset of periglomerular and granule cells also express CRH. Our preliminary immunohistochemical study (data not shown) has shown that CRH-IR neurons in the MGC are
Fig. 4. Dark-field photomicrographs showing hybridization signals for ppVIP mRNA. c is a higher magnification of AOB in b. d is a higher magnification of the EPL of the MOB. Scale bars = 500 #&rn(a): 250pm (b); 1OOpm (c,d).
Neuropeptide mRNAs in the rat olfactory bufb medium sized, triangular in shape and extend a few straight dendrites in the EPL, suggesting that they are tufted cells. Freprothyrotrop~n-rele~~~g
hormone
Previous neurochemical studies have revealed an abundance of TRH-IR in the OB that corresponds to one-quarter of that detected in the hypothalamus and about twice the average amounts detected in other brain regions. 33 The same authors presumed that TRH in the OB was derived from intrinsic neurons.34 TRH receptor has been shown to be concentrated in the OB, especially in the AOB.36 The AOB is one of the regions richest in TRH receptors in the brain. These findings, taken together, suggest that TRH plays very implant roles in the olfactory system. An abundance of TRH-IR neurons in the OB has been reported in the previous immunohistochemical study,6o and was confirmed in the present study. The most conspicuous feature for ppTRH neurons in the OB is their concentration in the GL. These neurons may correspond to periglomerular cells based on their cell size and location. A majority of the periglomeru-
635
lar cells appears to express ppTRH in the present study, while TRH-IR neurons were estimated to constitute only lo-15% of all the periglomerular cells in the previous immunohistochemical study.@’ A considerable number of ppTRH neurons were also observed in the EPL. These cells were previously identified as superficial short axon cells because they are distributed in the superficial part of the EPL.W However, in the present study, ppTRH neurons were evenly distributed in the EPL, suggesting that at least some of them may correspond to Van Gehuchten cells on the basis of their distribution pattem.49 A small number of ppTRH neurons were observed in the GRL. These neurons were also observed in the previous immunohistochemical study, and are considered to be deep short axon cells.@ Preprovasoactive intestinal polypeptide
The content of VIP-IR in the OB is 6.7pmol/g wet weight, which corresponds to one-fifth of that found in the hypothalamus.‘3 Some 6O-lOOpmol/g wet weight VIP was measured in the isocortex of the rat.
Fig. 5. Dark-field photomicrographs showing hybridization signals for ppSS mRNA. c is a higher magnification of AOB in a. d is a higher magnification of the GL of the MOB. Medium sized labeled cells in the GL are indicated with arrows in a. Scale bars = 500 pm (a); 250pm (b); 100 pm (c,d).
635
E. SENBAtv ai
Most of the ppVIP neurons were confined to the EPL, where they were evenly distributed. This confirmed a previous ~mmunoh~st~hem~a~ study, which reported numerous small sized VIP-IR cells in the EPL.” These neurons were considered to be Van
Gehuchten cells because their pcrikarya, and processes were confined to the EPL. 35.49Some ppVIP neurons, though much less neurons. were also oh-
served in the GL, MCL, IPL, and GRL. These neurons were detected by immunobistochemistry and
identified as superftcial and deep short axon cells on the basis of cell size, location, and arborjzation
of
processes.‘5 The present results confirmed alt of the findings of the previous immunohistochemical
studies, except for
the lightly labeled cells identified in the MCL of the
Fig. 6. Dark-field photomicrographs showing hybridization signals for fl-pptachykinin (a--c) and ppENK (d-4) mRNAs. b and e are higher magnifications of AOB in a and c, respectively. Scale bars = 500gm fa,d); 250 pm (b& 100pm. (c.f).
Neuropeptide mRNAs in the rat olfactory bulb AOB. Although no immunohistochemical reports are available regarding these neurons, they may be interneurons, like Van Gehuchten cells, in the MOB. Preprosomatostatin The content of somatostatin-IR in the OB corresponds to one-fifth of the levels found in the hypothalamus.7*28 The present study identified two types of ppsomatostatin neurons in the MOB. One of them consists of extremely large neurons generally located deep in the GRL. in contrast, the other type consists of small neurons confined to the GL. The former cells may correspond to deep short axon cells described in previous immunohistochemical studies on the hamster,’ guinea-pig,3s and rat.37 Small somatostatinIR neurons clustering around glomeruli seem to correspond to periglomerular cells when their cell size and density are taken into account. Some medium sized cells scattered in the same Iayer may correspond to superficial short axon cells. These findings confirmed the recent immunohistochemical study by Seroogy et al.‘? The most conspicuous feature about the ppsomatostatin neurons in the OB is their occurrence in the MCL of the AOB, which has never been detected in previous immunohistochemical studies. These neurons are medium sized and seem to have well developed dendritic processes. Further immuno-
histochemical studies are needed to characterize these neurons. 8 - Preprotachykinin A Previous radioimmunoassay studies reported concentrations of substance P (SP)-IR that corresponded to one-tenth of that found in the hypothalamus.8,27 SP and neurokinin A are major end-products of ~-pptachykinin.3i B-pptachykinin neurons in the deep layer of the EPL and MCL may correspond to internal tufted and mitral cells. In a previous immunohist~hemi~l study, no SP-IR neurons were observed in the rat MOB.3 However, colchicine treatment revealed the presence of SP-IR neurons in the MCL of the MOB,” which was confirmed by the present and previous in situ hybridization studies. 62 On the other hand, the EPL contained only a few /?-pptachykinin neurons, which were shown to be SP-IR in a previous immunohistochemical study using colchicine pretreatment.37 The size and mo~hology of these neurons suggested that they are external and middle tufted cells. This distribution pattern is quite different from that in the hamster, where many periglomerular and external tufted cells were shown to be SP-IR.9,32SP-IR granule cells were observed in the GRL of the MOB,37 in addition to densely packed similar neurons in the GRL of the AOB.3,37 SP-IR neurons were also observed in the posterior half of the MCL in the AOB.37
AOB
.. . -,MGc__:~------.-
._.._ ____,___... PFCCH ”,_. I:.?., : : ..__,_, ., - ., .-.&*; ... __..c-:-i:.. &* “_i ...(
Neuroscience Vol. 38, No. 3, pp. 629-641, 1990 Printed in Great Britain
LOCALIZATION OF NEUROPEPTIDE PRECURSORSYNTHESIZING NEURONS IN THE RAT OLFACTORY BULB: A HYBRIDIZATION HISTOCHEMICAL STUDY E. SENBA,* D. M. SIMMONSand L. W. SWAN.WN The Howard
Hughs Medical Institute, The Salk Institute for Biological Studies, La Jolla, CA 92037, U.S.A.
Abstract-The distribution of seven kinds of neuropeptide precursor mRNA-containing neurons was investigated in the rat main and accessory olfactory bulbs, where various peptides have previously been identified immunohistochemically, by means of in situ hybridization using [%]cRNA probes. In the glomeruiar layer, numerous preprothyrotropin-releasing hormone mRNA-expressing neurons, moderate numbers of preprosomatostatin and preproenkephalin A neurons, and a small number of preprocholecystokinin neurons were detected. In the external plexiform layer, numerous medium sized preprocholecystokinin and preprocorticotropin-releasing hormone neurons, and a small number of /I-preprotachykinin A neurons were observed. In addition, small preprovasoactive intestinal polypeptide and preprothyrotropin-releasing hormone neurons were evenly distributed in the external plexiform layer. Medium to large sized /I-preprotachykinin A neurons formed a thin layer in the mitral cell layer. In the granule cell layer, in addition to numerous small preproenkephalin A neurons, moderate numbers of small )?I-preprotachykinin A and pmprocorticotropin-releasing hormone neurons, and a small number of preprothyrotropin-releasing hormone neurons, were identified. Large sized preprosomatostatin neurons were located in the deep layer of the granule cell layer. The distribution patterns of these neurons, as a whole, confirmed previous studies based on immunohistochemistry, although peptide precursor mRNA-expressing neurons were far more numerous than those immunoreactive to the respective neuropeptides. Moreover, mRNA-expressing neurons were observed in areas where no immunoreactive neurons had been observed (e.g. preprovasoactive intestinal polypeptide and preprosomatostatin neurons in the mitral cell layer of the assessory olfactory bulb). The distribution patterns were generally similar in the main and accessory olfactory bulbs.
Because of the simple laminated structure of the olfactory bulb (OB), individual cell types are well characterized on both morphological and physiological grounds. ” For this reason, the OB has provided a useful model for investigating the functional significance of multiple neuropeptides within individual cells. Immunocytochemical studies have recently demonstrated that many neuroactive peptides are localized within intrinsic and output neurons of the main and accessory OB (see Refs 21 and 35 for a
*To whom correspondence should be addressed at: Department of Anatomy (II), Osaka University Medical School, 4 Nakanoshima Kitaku, Osaka, Japan. Abbreviations: AOB, accessory olfactory bulb; CCK-8, cholecystokinin-8; cDNA, complementary DNA; CRH, corticotropin-releasing hormone; cRNA, complementary RNA; DTT, dithiothreitol; EDTA, ethylene diaminetetra-acetate; ENK, enkephalin; EPL, external plexiform layer; GL, glomerular layer; GRL, granule cell layer; IPL, internal plexiform layer; IR, like immunoreactive/immunoreac&ity; Met-BNK, [Metlenkephalin; MCL, mitral cell layer: MGC. modified alomerular complex; MOB, main olfactory bulb; OB,- olfactory bulb; PHI, peptide histidine-leucine; pp, prepro; ppCCK, preprocholecystokinin; ppENK, proproenkephalin A; SP, substance P; SSC, sodium chloride, 0.15 M/sodium citrate, 0.015 M, pH = 7; TRH, thyrotropin-releasing hormone; VIP, vasoactive intestinal polypeptide.
review). These peptides may play important roles in the regulation of olfactory function and olfactory related behaviors, such as mating, maternal care, feeding, and many aspects of social behaviors.‘2,43*56 However, disagreements and uncertainties about the presence and distribution of neuropeptides still exist. One of the difficulties associated with immunocytochemistry is that it is sometimes very difficult to detect peptide-synthesizing cell somata for at least two possible reasons. Firstly, neuropeptides synthesized in the perikarya are actively transported to axon terminals, and their concentrations in the perikarya are in many cases too low to be detected. Thus, large amounts of colchicine are often required to visualize them. And secondly, big precursor molecules are sometimes not recognized by antibodies. The present study was undertaken to determine the localization and quantity of neuropeptide-synthesizing neurons in the rat OB under normal conditions, by means of the in situ hybridization method, which visualizes mRNA-containing cell bodies. Information about the distribution of neuropeptides in the accessory olfactory bulb (AOB) is still limited compared to that of the main olfactory bulb (MOB). Therefore, the distribution patterns of neuropeptides were compared in both compartments, in order to elucidate general principles.
629
E. SENBAet al.
630 EXPERIMENTAL
PROCEDURES
Animals and tissue preparation
Adult male Sprague-Dawley rats weighing 25&3OOg (n = 4) were used. They were allowed free access to water and food (Teklad mouse/rat diet). For perfusion, animals were deeply anesthetized with chloral hydrate (0.1 ml of 35% solution/lOOg body weight) and perfused transcardially using a pH shift method described in detail elsewhere.16 Briefly, following 50ml of saline, ice-cold 4% paraformaldehyde in 0.1 M acetate buffer at pH 6.5 (200 ml), and then 4% paraformaldehyde plus 0.05% glutaraldehyde in 0.1 M borate buffer at pH 9.5 (400 ml), were perfused. The brains were quickly removed and were postfixed overnight at 4°C in 4% paraformaldehyde plus 10% sucrose in 0.1 M borate buffer at pH 9.5. The forebrains containing the OB were frozen with powdered dry ice, and 20-pm-thick serial sagittal sections were cut on a sliding microtome, and collected in ice-cold potassium phosphate-buffered saline. Ten series of sections were obtained from each animal. One series was always mounted and stained with 1% Thionin to provide a Nisslstained series for cytoarchitectonic analysis of the OB. The remaining series were processed for hybridization histochemistry, including control experiments. They were mounted or subbed, poly-L-lysine-coated slides as soon as possible after cutting. The mounted sections were allowed to air-dry for l(r15min and were stored overnight at room temperature under vacuum with desiccant. Preparation of probes
A total of seven different [3SS]complementary RNA ([‘5S]cRNA) probes complementary (anti-sense) to their respective mRNAs were used in this study. They were transcribed from linearized DNA templates using Riboprobe Gemini System kits purchased from Promega (Madison, WI). Linearized templates were purified by gel electrophoresis, precipitated in ethanol, reconstituted in Tris-EDTA buffer to a final concentration of 1 pg/pl, and stored at -20°C. I’sSIUTP (100&15OOCi/mM: New England Nuclear; Bostdn, MA) and unlabeled ‘ATPi GTP and CTP were used in the transcription reaction. Probe templates were cDNA inserts into pGem (Promega) or SP64 plasmids for RNA transcription as described in our previous papem.53,55Briefly, plasmids containing a full length complementary DNA (cDNA) insert (528 bp) corresponding to nucleotides 17-544 of the rat preprocholecystokinin (ppCCK) sequence (from J. E. Dixon, Purdue University; see Ref. 1l), and 935 bp fragment corresponding to nucleotides - 104-831 of rat preproenkephalin A (ppENK) cDNA (from S. L. Sabol, N. I. H.; see Ref. 63) and a 380 bp fragment including the entire vasoactive intestinal polypeptide (VIP)- and peptide histidineleucine (PHI-peptide coding regions (from R. H. Goodman, Tufts University; see Ref. 42) were linearized, and SP6 RNA polymerase was used to transcribe respective anti-sense RNA probes. A 380 bp fragment including the entire corticotropin-releasing hormone-41 (CRH-41) peptide coding region and part of the 3’ untranslated region of rat CRH mRNA (from K. E. Mayo, Northwestern University; see Refs 26 and 58) and a 508 pb fragment containing the entire coding region which corresponds to nucleotides - 74- + 434 of the rat fi-preprotachykinin A (/?-pptachykinin) sequence (from J. E. Krause, Washington University; see Ref. 31) were linearized and T7 RNA polymerase was used to transcribe ?S-labeled riboprobe complementary to preproCRH (ppCRH) and I-preprotachykinin (j?-ppT) mRNA, respectively. For preprosomatostatin, original cDNA (from J. F. Habener, Harvard Medical School; see Ref. 18’)was cut at XbaI and AvaI sites and subcloned into pGEMII1 and then linearized with Hind III, and T7 RNA polymerase was used to transcribe anti-sense probe. For preprothyrotropinreleasing hormone (ppTRH) SP64 plasmid containing a 1241 bp sequence of ppTRH cDNA (from R. H. Goodman,
Tufts University; see Ref. 50) was linearized with Hind III, and riboprobe was transcribed using SP6 polymerase. These probes were adjusted to concentrations of 5 x 106-10’ cpm/ ml, which provided the greatest signal-to-noise ratio. For control experiments, sense-strand probes were used, in substitution for the above described anti-sense probes. Hybridization histochemislry
The hybridization procedure used in the present study is based on that described by Angerer et al.’ and Simmons et aLS3 Before hybridization, vacuum-dried sections were treated with proteinase K for 30 min at 37”C, acetylated for 10 min at room temperature, and dehydrated in a graded series of ethanols (prehybridization step). About 80~1 of hybridization solution was applied to a coverslip, which was then placed over the sections mounted on the slide. Edges of the coverslip were sealed with liquid DPX mountant (Gallard and Schlessinger, NY). The slides were then incubated on a covered slide-warming tray at 55do”C for 16-20 h. The hybridization solution contained 50% formamide, 10% dextran sulfate, 0.3 M NaCl, 1 x Denhardt’s solution, 10mM Tris (pH 8.0), 1 mM EDTA, 500 pg/ml tRNA, 10 mM dithiothreitol (DTT), and the probe. After the hybrid&ion incubation,- the coverslips were removed in NaCl/Na citrate (SSC) solutions. The sections were digested with ribonucleak A ‘(0.02 mg/ml) for 30 min. at 37°C and rinsed in descending concentrations of SSC at room temperature. They were then incubated in 0.1 x SSC at 60°C for 30min, dehydrated in an ascending series of ethanols and dried under vacuum at room temperature for 30 min. All SSC rinse solutions following ribonuclease treatment contained 1 mM DTT. Autoradiography
The sections were de-lipidized in 95% ethanol (5 min), followed by absolute ethanol (3 x 5 min) xylene (10 min), fresh xylene (30 min), and absolute ethanol (3 x 5 min). They were then air-dried for at least 30 min, and dipped in Kodak autoradiography emulsion (NTB-2), diluted 1: 1 with distilled water. After an exposure time of one to three weeks, the slides were developed in Kodak D-19 (3.5 min at 14”C, fixed in Kodak Rapid Fix, rinsed for 1 h in running distilled water, and stained through the emulsion with 0.1% Thionin. Because of the digestion of RNAs by ribonuclease A treatment, only cell nuclei were stained. RESULTS
The seven probes described above labeled a variety of neurons in the rat OB. Background hybridization was low and similar to that seen when the sensestrand probes were used. Preprocholecystokinin
mRNA-containing
neurons
Main olfactory bulb. The largest collection of medium sized ppCCK mRNA-containing neurons (ppCCK neurons) (15-20 pm) was located within a border area between the glomerular layer (GL) and external plexiform layer (EPL), and in deeper lamina of the EPL, forming two distinct layers in the EPL (Fig. la,e). About twice as many neurons were observed in the outer layer as compared to the inner layer. A cluster of labeled cells was localized in the EPL of the border area between the MOB and AOB, the so-called modified glomerular complex (MGC)20.57(Fig. lb,c). Small sized neurons (S-10 pm) were also observed in the GL (Fig. Id) and in the granule cell layer (GRL).
Neutopeptide mRNAs in the rat olfactory bulb Accessory o~ctu~~ bulb. Almost no labeled cells were id&W in the AOB, except a few lightly labeled ones scattered in the mitral cell layer (MCL) (Fig. la,b). Preprocorticotropin-releusing hormone mRNA -confaining neuron Main olfactory bulb. ppCRH mRNA-~nt~ni~~ neurons (ppCRH neurons) were numerous and widely distributed in the MOB. in the EPL and MCL, hbe~ed neurons formed two distinct layers, extema1 and internal (including the MCL) {Fig. 2a,d). In contrast to the ppCCK mRNA containing neurons described above, about equal numbers of labeled cells were observed in each layer. Some of them were medium in size (IS-20 pm), while most of them were small (about IOgm) (Fig. 2d). A cluster of labeled cells was also located in the MGC (Fig. Zb). Most of these cells were medium in size (2zOpm). Small iabeled cells (IOgm) were scattered in the deeper two-thirds of the CRL (Fig. 2a,d). Accessory olfactory bulb. Medium sized labeled neurons (20 pm) were observed in the MCL (Fig. 2b). In addition, a few small celis (10 pm) were also scattered in the CRL (Fig. 2~).
631
~epr~fhyr4tr~pin-re~e~~ng hormone ~~A-~o~#~~ing neurons Main o~c~Qry bulb. The MOB of the rat was rimmed by ppTRH mRNA-containing neurons, wh&h were den&y packed between ~#rn~li in the CL These neurons were small (about IOgm) and clearly demarcated the borders of the glomeruli (Fig. Ja,d,e). Small labeled cells were also observed in the EPL (Fig. 3a,d,e), in w&h these ne~ons were evenly distributed. Even less n~erou$ smaff sized neurons (10ym) were scattered in the GRL (Fig. 3a,d). dccessory o~c~~ry bulb. Small clusters of labekd cehs (10 pm) were observed surroundjng glomeruti in the GL, and labeled cells of similar size were distributed in the deeper half of the MCL and also in the GRL (Fig. 3b,c).
Ma& of$acctozy bulb. Most of the ppVIP mRNAcontaining neurons (ppVIP neurons) were confined to the EPL. These neurons were small to medium in
Fig. 1. Photomicrographs of dipped autoradiographs to show hybridization signals for ppCCK mRNA. c is a higher magnification of MGC in b. e is a bright-field photomicrograph of the same area as d. Scale bars = 500 pm (a); 250 pm fb); IOOpm &-ef.
632
E. SENB.4ei al.
size (IO-15 pm) and evenly distributed in the layer, forming a bundle of labeled neurons (Fig. 4a). Small sized labeled neurons (about 10fi.m) were occasionally observed in the CL and GRL (Fig. 4a). Accessory ol$uctory bulb. Lightly labeled neurons (10-15 urn) were evenly scattered in the MCL (Fig. 4b,c). Preproson~at~st~t~nmRNA -containing neuron Main olfactory bulb. Two kinds of preprosomatostatin (ppsomatostatin mRNA-containing neurons (ppsomatostatin neurons) were observed in the MOB. One consisted of small neurons (about 10 pm), which dist~buted around the glomeruii in the GL (Fig. 5a,d). The other consisted of densely labeled large neurons (Z@-30pm) located in the deepest part of the GRL (Fig. 5a,b); similar neurons were occasionally found in the CL (Fig. 5a). Accessory olfactory bulb. Labeled cells of small to medium size (IO-20 Frn) were concentrated in the MCL (Fig. 5a-c). ,lI-Preprotachykinin A mRNA -containing neurons Main olfactory bulb. A row of densely labeled ~-pptachykinin mRNA-containing neurons (B-
pptachykinin neurons) was observed in the MCL (Fig. 6a,c). These neurons were medium to large in size (15-30 pm). Some medium sized labeled neurons (about 1.5pm) were also scattered in the EPL (Fig. 6a,c). Lightly labeled small neurons (IOgm) were distributed in the deeper layer of the GRL (Fig. 6a). Accessory olfactory bulb. Densely labeled small cells (10pm) were packed in the GRL (Fig. 6a,b). Medium sized labeled neurons (15 pm) were also observed in the MCL (Fig. 6a,b). These neurons
tended to be located in the posterior half of this layer. Preproenkephalin A mRNA -containing neurons Main olfactory bulb. Densely packed preproenkephafin A (ppENK) mRNA-containing neurons were observed in the GRL (Fig. 6d,e). Labeling was more intense in the superficial one-third of the GRL, including the internal plexiform layer (IPL) and MCL, in which these neurons seemed to surround mitral cells (Fig. 6f). Small labeled cells (5-10pm) were distributed around glomeruli in the GL (Fig. 6d,f). Accessory oifactory bulb. Small labeled cells (about 10 fern) were diffusely distributed in the GRL and IPL
Fig. 2. Dark-field photomicrographs of dipped autoradiographs to show hybridization signals for ppCRH mRNA. Scale bars = 500 pm (a); 100pm (b-d).
Neuropeptide mRNAs in the rat olfactory bulb
(Fig. 6d,e). A few of them were also observed in the GL. The distributions of these peptide precursor mRNAs, which were quite similar in the OR sections from four animals, are thematically summarized in Fig. 7.
DISUNION
This report is the first compilation of results based on irr situ hybridization for the mRNAs of various
633
neuropeptide precursor molecules in the rat OB. In comparing the present results with earlier reports based on immun~yt~he~st~, evidence for the greater sensitivity of the in sifu hybridization method was provided, in view of the fact that peptide precursor mRNA-containing neurons were far more numerous than neurons reported to be immunoreactive for peptide end-products, although it is not known whether all of these mRNA expressing neurons synthesize end-products. Moreover, IabeIed cells were also observed in areas in which no immunorea~tive neurons have been reported.
Fig. 3. Dark-field photomicrographsshowing hybridization signals for ppTRH mRNA. c is a higher magnification of the squared area in b. Scale bars = 500~rn (a); 250 pm (b,d); 100pm (c,e).
E. SENBA ff al.
634
Preprocholecystokinin
The content of cholecystokinin-like immunoreactivity in the rat OB is relatively high, and is about one-half to one-third that measured in the isocortex, which shows the highest levels in the CNS.’ The majority of the cholecystokinin-&like immunoreact~ve (CCK-I-IR) neurons in the EPL have been identified as middle tufted cells. In addition, external and internal tufted cells have also been shown to be immunoreactive for CCK-8.” It is generally accepted that middle and internal tufted cells are projection neurons, along with mitral cells. Indeed, these CCK-8-IR tufted cells have been shown to project to anterior parts of the bulbar target regions.39 In addition to these projection neurons, some intrinsic neurons, such as periglomerular cells, superficial short axon cells in the GL, and deep short axon cells in the GRL, have also been shown to be CCK-8-IR.37~38~5’The present study confirmed the presence of these neurons in the OB. It has been reported in previous morphological studies that there are no tufted cells in the AOB.35 The
present finding that few, if any, ppCCK neurons can be identified in the AOB, in contrast to the abundance of ppCCK containing tufted cells in the MOB, appears to support this previous suggestion. Prepr~~~rt~c~tropin-releasing hormone
In contrast to other neuropeptides, the content of CRH-IR in the OB has been reported to be relatively low.54 Few CRH-IR neurons were observed in the OB in the previous immunohistochemical distribution studies in the rat brain.40,47The latter authors stated that CRH-IR structures in the OB were less intense than in other brain regions. The present study, however, revealed a widespread distribution of ppCRH neurons in the OB, confirming a recent immunohistochemical study combined with in situ hyb~di~t~on histochemistry,25 which suggested that most of the ppCRH neurons in the MCL and EPL are mitral cells or tufted cells, and that a subset of periglomerular and granule cells also express CRH. Our preliminary immunohistochemical study (data not shown) has shown that CRH-IR neurons in the MGC are
Fig. 4. Dark-field photomicrographs showing hybridization signals for ppVIP mRNA. c is a higher magnification of AOB in b. d is a higher magnification of the EPL of the MOB. Scale bars = 500 #&rn(a): 250pm (b); 1OOpm (c,d).
Neuropeptide mRNAs in the rat olfactory bufb medium sized, triangular in shape and extend a few straight dendrites in the EPL, suggesting that they are tufted cells. Freprothyrotrop~n-rele~~~g
hormone
Previous neurochemical studies have revealed an abundance of TRH-IR in the OB that corresponds to one-quarter of that detected in the hypothalamus and about twice the average amounts detected in other brain regions. 33 The same authors presumed that TRH in the OB was derived from intrinsic neurons.34 TRH receptor has been shown to be concentrated in the OB, especially in the AOB.36 The AOB is one of the regions richest in TRH receptors in the brain. These findings, taken together, suggest that TRH plays very implant roles in the olfactory system. An abundance of TRH-IR neurons in the OB has been reported in the previous immunohistochemical study,6o and was confirmed in the present study. The most conspicuous feature for ppTRH neurons in the OB is their concentration in the GL. These neurons may correspond to periglomerular cells based on their cell size and location. A majority of the periglomeru-
635
lar cells appears to express ppTRH in the present study, while TRH-IR neurons were estimated to constitute only lo-15% of all the periglomerular cells in the previous immunohistochemical study.@’ A considerable number of ppTRH neurons were also observed in the EPL. These cells were previously identified as superficial short axon cells because they are distributed in the superficial part of the EPL.W However, in the present study, ppTRH neurons were evenly distributed in the EPL, suggesting that at least some of them may correspond to Van Gehuchten cells on the basis of their distribution pattem.49 A small number of ppTRH neurons were observed in the GRL. These neurons were also observed in the previous immunohistochemical study, and are considered to be deep short axon cells.@ Preprovasoactive intestinal polypeptide
The content of VIP-IR in the OB is 6.7pmol/g wet weight, which corresponds to one-fifth of that found in the hypothalamus.‘3 Some 6O-lOOpmol/g wet weight VIP was measured in the isocortex of the rat.
Fig. 5. Dark-field photomicrographs showing hybridization signals for ppSS mRNA. c is a higher magnification of AOB in a. d is a higher magnification of the GL of the MOB. Medium sized labeled cells in the GL are indicated with arrows in a. Scale bars = 500 pm (a); 250pm (b); 100 pm (c,d).
635
E. SENBAtv ai
Most of the ppVIP neurons were confined to the EPL, where they were evenly distributed. This confirmed a previous ~mmunoh~st~hem~a~ study, which reported numerous small sized VIP-IR cells in the EPL.” These neurons were considered to be Van
Gehuchten cells because their pcrikarya, and processes were confined to the EPL. 35.49Some ppVIP neurons, though much less neurons. were also oh-
served in the GL, MCL, IPL, and GRL. These neurons were detected by immunobistochemistry and
identified as superftcial and deep short axon cells on the basis of cell size, location, and arborjzation
of
processes.‘5 The present results confirmed alt of the findings of the previous immunohistochemical
studies, except for
the lightly labeled cells identified in the MCL of the
Fig. 6. Dark-field photomicrographs showing hybridization signals for fl-pptachykinin (a--c) and ppENK (d-4) mRNAs. b and e are higher magnifications of AOB in a and c, respectively. Scale bars = 500gm fa,d); 250 pm (b& 100pm. (c.f).
Neuropeptide mRNAs in the rat olfactory bulb AOB. Although no immunohistochemical reports are available regarding these neurons, they may be interneurons, like Van Gehuchten cells, in the MOB. Preprosomatostatin The content of somatostatin-IR in the OB corresponds to one-fifth of the levels found in the hypothalamus.7*28 The present study identified two types of ppsomatostatin neurons in the MOB. One of them consists of extremely large neurons generally located deep in the GRL. in contrast, the other type consists of small neurons confined to the GL. The former cells may correspond to deep short axon cells described in previous immunohistochemical studies on the hamster,’ guinea-pig,3s and rat.37 Small somatostatinIR neurons clustering around glomeruli seem to correspond to periglomerular cells when their cell size and density are taken into account. Some medium sized cells scattered in the same Iayer may correspond to superficial short axon cells. These findings confirmed the recent immunohistochemical study by Seroogy et al.‘? The most conspicuous feature about the ppsomatostatin neurons in the OB is their occurrence in the MCL of the AOB, which has never been detected in previous immunohistochemical studies. These neurons are medium sized and seem to have well developed dendritic processes. Further immuno-
histochemical studies are needed to characterize these neurons. 8 - Preprotachykinin A Previous radioimmunoassay studies reported concentrations of substance P (SP)-IR that corresponded to one-tenth of that found in the hypothalamus.8,27 SP and neurokinin A are major end-products of ~-pptachykinin.3i B-pptachykinin neurons in the deep layer of the EPL and MCL may correspond to internal tufted and mitral cells. In a previous immunohist~hemi~l study, no SP-IR neurons were observed in the rat MOB.3 However, colchicine treatment revealed the presence of SP-IR neurons in the MCL of the MOB,” which was confirmed by the present and previous in situ hybridization studies. 62 On the other hand, the EPL contained only a few /?-pptachykinin neurons, which were shown to be SP-IR in a previous immunohistochemical study using colchicine pretreatment.37 The size and mo~hology of these neurons suggested that they are external and middle tufted cells. This distribution pattern is quite different from that in the hamster, where many periglomerular and external tufted cells were shown to be SP-IR.9,32SP-IR granule cells were observed in the GRL of the MOB,37 in addition to densely packed similar neurons in the GRL of the AOB.3,37 SP-IR neurons were also observed in the posterior half of the MCL in the AOB.37
AOB
.. . -,MGc__:~------.-
._.._ ____,___... PFCCH ”,_. I:.?., : : ..__,_, ., - ., .-.&*; ... __..c-:-i:.. &* “_i ...(
