- Email: [email protected]
Olfactory disturbance induced by deafferentation of serotonergic fibers in the olfactory bulb
Neuroscience Vol. 61, No. 4, pp. 733-738, 1994
Pergamon
0306-4522(94)00224-X
Letter OLFACTORY DEAFFERENTATION
Elsevier ScienceLtd Copyright 0 1994IBRO Printed in Great Britain. All rights reserved 0306-4522194$7.00t 0.00
to neuroscience
DISTURBANCE INDUCED BY OF SEROTONERGIC FIBERS IN THE OLFACTORY BULB
T. MORIIZUMI,*t T. TSUKATANIJ H. SAKASHITA$ and T. MIWA$ *Department of Anatomy and $Department of Otorhinolaryngology, School of Medicine, Kanazawa University, Kanazawa 920, Japan
remain to k elucidated. In the present study, using 5,7-dihydroxytryptamine (5,7-DHT), a specific neurotoxin for 5-HT, we examined whether or not olfactory bulb is one of the major forebrain targets of the ascending serotonin patbway.‘3q’6 According to deafferentation of the bulbar 5-HT fibers induces physiological studies:J neurons of the olfactory bulb olfactory abnormality in rats. The conditioned rats who learned to avoid cyclowere found to reduce their spontaneousdischarge rates heximide were injected with 5,7-DHT into the ascendby electrophoretically applied serotonin. However, roles of the bulbar serotonin in the sease of smell ing 5-HT fiber pathway to deafferentate the 5-HT fibers in the bulb. Then olfactory function was examremain unanswered. In the present study, using 5,7dihydroxytryptamine, a specificneurotoxin for serotonin, ined by capability to discriminate cycloheximide solwe found that the conditioned rats who learned to avoid ution from water. Olfactory function of the control a repellent by olfaction lost ability of discrimination by group remained unchanged throughout the postoperdeafiereatation of the bulbar serotonergic fibers. Such ative period, while olfactory function of the 5,7olfactory dysfunction did not occur in the early stage DHT-injected group varied from rat to rat. From our (three days after injection of the toxin) when the retrospective study, we found a close relationship serotooergic fibers disappeared in the bulb, but devel- between the amount of the bulbar 5-HT fibers and oped a few weeks later. Interestingly, histological olfaction. According to the degree of deafferentation examination revealed marked shrinkage of the bulbar of the 5-HT fibers, we divided the 5,7-DHT-injected glomerulus which is a major termination site of the rats into two groups; those with partial deafferentabulhopetal serotonergic fibers, and also a synaptic site tion and those with complete deafferentation. Figure of olfactory receptor cells and bulbar output neurons. I shows the results of the discrimination test of the The results indicate that depletion of the serotonergic three groups. Olfactory function of the partially fibers in the olfactory bulb causes glomerular atrophy deafferentated group did not differ from that of the and olfactory disturbance in the rat. control group, while olfactory function of the comThe olfactory bulb receives monoaminergic inputs pletely deafferentated group significantly deteriorated from the brain stem.“~16~‘y These monoamines include two to four weeks after toxin injection. The onset of two well-known chemical substances, norepinephrine loss of olfaction varied from rat to rat. Now we can (NA) and serotonin (5-HT). The bulbar NA fibers are say for sure that if the olfactory bulb is completely distributed from the external plexiform layer to the deprived of 5-HT, anosmia develops within four granule cell layer,” and have been reported to be weeks after toxin injection. To know whether the related to formation of olfactory memory.6,” The toxin produces reversible or irreversible olfactory bulbar S-HT fibers are localized mainly in the abnormality, we continued the discrimination test in glomerular layer. I6 Because of their glomerular some rats which have lost olfaction within four weeks location, the 5-HT fibers are speculated to influence after toxin injection, and found no recovery of olfacon synaptic integration, but their roles in olfaction tion up to six weeks. Immunohistochemistry was done to assess nonspecific effects of 5,7-DHT on the bulbar NA fibers, In tTo whom correspondence should be addressed. half of the 5,7-DHT-injected anosmic rats, the toxin Abbmiurions: DBH, dopamine b-hydroxylase; 5,7-DHT, did not affect the hulbar NA fibers (Fig. 2d), despite 5,7-dihydroxytryptamine; S-HT, serotonin; NA, noretotal lack of bulbar 5-HT fibers (Fig. 2b). In the pinephrine; PB, phosphate buffer; PBS, phosphatebuffered saline: TH, tyrosine hydroxylase remaining anosmic rats, the bulbar NA fibers were The serotooergic neurons of the brain stem project
widely throughout the central nervous system, and the
733
734
T. Morizumi
0 Days
after
3
7
injections
14 of
21 5,7-DHT
28 or
vehicle
Fig. 1. Correct responses of the three groups with controls (O-O, n = lo), partial deafferentation of the bulbar S-HT fibers (O-0, n = 9) and complete deafferentation of the bulbar S-HT fibers (A-A, n = 11). The completely deafferentated group shows loss of ability of discrimination 14-28 days after injection of 5,7-DHT. Mean percentages and standard errors of correct responses of the control group are 97 + 2 (three days), 97 f 2 (seven days), 99 i 1 (14 days), 98 + 1 (21 days) and 98 & 1 (28 days). Those values of the partially deafferentated group are 98 + 1, 93 _+2, 96 + 2, 90 + 2 and 94 + 2. Those values of the completely deafferentated group are 97 + 1, 83 + 7, 76 + 6, 59 k 3, 52 f 3. [+I = infusion of 5,7-DHT or vehicle. *Significantly different (P < 0.01) from the values obtained in the control group as determined by Scheffi’s multiple comparison test. The experiments were carried out on adult male Sprague-Dawley rats (250-350 g). All surgical manipulations were done under general anesthesia with sodium pentobarbital (SOmg/kg, i.p.). (I) Conditioning. Since umlateral bulbectomy did not affect odor aversion learning as described later, rats (n = 30) were subjected to removal of unilateral olfactory bulb to assess relationships between deafferentation of the 5-HT fibers, olfactory function and morphological changes by limiting examined olfactory system on one side. In particular, since the neurotoxic effects of 5,7-DHT on S-HT fibers varied in terms of deafferentation of the bulbar 5-HT fibers, the procedure of unilateral bulbectomy has a considerable advantage to precisely examine the effects of depletion of the bulbar 5-HT on olfaction. The unilaterally bulbectomized rats were trained to avoid cycloheximide, which has a strong repellent action to mdents.‘2.‘R They were deprived of water for two to three days
prior to each discrimination test. Each rat was placed in a cage. Two bottles were offered to each rat. One contained 0.01% cycloheximide solution, while another water. Ten trials of discrimination were usually performed at one time. In the 10 trials, the position of the bottle containing cycloheximide solution was made right, left, left, right, left, left, right, right, right and left relative to the position of the bottle containing water. Observations were carried out in an air-conditioned room. When the rat drank water, the response was interpreted as a correct response. When the rat drank cycloheximide solution, the response was regarded as a wrong response. The number of correct responses was divided by the number of total responses, and the percentage of correct responses was calculated. Since the taste of cycloheximide solution is so disgusting, the rats remember the smell of the solution when they drink it, and thereafter learn to avoid it by olfaction. Thus the olfactory conditioned reaction is established. Rats drank cycloheximide solution two to four times in the first 10 trials. The correct responses were increased in the second 10 trials. Usually 30-40 trials were necessary to condition rats, which is quite similar in normal rats without unilateral bulbectomy. Thus all unilaterally bulbectomized rats were successfully conditloned to avoid cycloheximide and gave 100% correct responses in the last 10 trials, indicating that unilateral
el al.
bulbectomy did not affect odor aversion learning. The finding that no differences in odor aversion learning could be detected between normal and unilaterally bulbectomized control rats may be explained by the phenomenon that olfactory information learned by one side of the olfactory system is transferred to another side through olfactory commissural pathways.‘4,‘5 (2) Lesioning. The bulbopetal 5-HT fibers originate from the dorsal and median raphe nuclei, and pass through the medial forebrain bundle.“,16 Thus, to deafferentate the 5-HT fibers in the remaining olfactory bulb, the conditioned rats (n = 20) were injected with 5,7-DHT (Sigma, 4 or 8 pg in 4 ~1) into the rostra1 (A: 2.0, L: 1.5, D: -8.Omm) and caudal (A: -2.0, L: 2.0, D: - 8.5 mm) parts of the medial forebrain bundle ipsilateral to the remaining bulb. For control (n = IO), vehicle (4~1 of isotonic saline containing 0.1% ascorbic acid) was applied to the same region. (3) Olfactory function. After infusion of 5,7-DHT or vehicle into the medial forebrain bundle, the conditioned rats were submitted to odor aversion behavior to examine their olfactory function to discriminate cycloheximide solution from water. Behavioral tasks were performed at three. seven, 14, 21 and 28 days after injection of 5,7-DHT or vehicle. Several rats from the control and 5,7-DHT-injected groups were continued to the discrimination test up to 42 days. To confirm soundness of the discrimination test as a method to examine olfactory function, we removed the remaining olfactory bulb in the unilaterally bulbectomized, conditioned rats (n = 5). All rats with bilateral bulbectomy gave 50% correct responses, indicating loss of olfaction. (4) Grouping. After completing the discrimination test, the rats were deeply anesthetized with sodium pentobarbital @Omg/kg, i.p.), and perfused with a fixative. Deafferentation of the 5-HT fibers in each olfactory bulb and specificity of 5,7-DHT was evaluated by immunohistochemistry as described in Fig. 2. Since most of bulbar 5-HT-immunoreactive fibers are located in the glomerular layer and have distinct varicosities,‘” the number of glomerular axonal varicosities with 5-HT immunoreactivity was counted to estimate the effects of 5,7-DHT on the bulbar 5-HT fibers. Based on the degree of 5-HT depletion at each olfactory bulb, the 5,7-DHT-injected rats were divided into two groups; those with partial deafferentation of the S-HT fibers and those with complete deafferentation of the 5-HT fibers. The bulbs of the partially deafferentated group contained intact, S-HT-immunoreactive fibers of 49.4% (33.3-70.1%) of control value. The bulbs of the completely deafferentated group were almost free of _5-HT-immunoreactive fibers (2.9%, O-9.5% of control value). Thus the rats were divided into three groups; controls, those with partial deafferentation, and those with complete deafferentation. The presence of the anosmic rats apparent decrease of the bulbar NA fibers seems to exclude involvement of NA in olfactory disturbance induced by deafferentation of the bulbar S-HT fibers. The olfactory bulb contains intrinsic dopaminergic neurons. Immunohistochemistry revealed that the bulb of the 5,7-DHT-injected anosmic rats had normal to fairly decreased number of dopaminergic neurons depending on glomeruli (Figs 2f,g). In total, the bulb of these anosmic rats apparently contained fewer dopaminergic neurons than that of the control rats. This is explained well by the slightly
reduced.
without
phenomenon
that
dopamine
cline by odor deprivation
levels
significantly
de-
or deafferentation of peripheral olfactory afferents. ‘L’ Histological examination revealed atrophy of the olfactory bulb of the 5,7-DHT-injected anosmic rats
Role of serotonin in olfaction
735
Fig. 2. Monaamin~ (S-HT, NA and dopamine) in the olfactory bulb of control rat with normal ofaction {ac,e) and 5.7-DHT-injected rat with olfactory disturbance (b,d,f,g). To visualize the monoamines, immunohistochemistry was done using antibodies to S-HT, dopamine j9-hydroxylase (DBH, a NA-synthesizing enzyme) and tyrosine hydroxylase (TH, a synthetic enzyme for both NA and dopamine). Since TH-immunoreactive neurons and fibers in the bulbar glomerular layer lack DBH immunoreactivity, glomerular TH-positive neuronal elements are regarded as dopaminergic. 5-HT- and do~mine~on~ining fibers are rich in the glometular layer, while NA-containing fibers are localized in the mitral and granule cell layers, and the external and internal plexiform layers.-S-HT immunoreactlve fibers are abundant in control bulb (a), but absent in 5.7-DHT-iniected bulb (b). No significant differences between control (c) and 5,7-DHT-injected (d) bulbs are found ii DBH immu~oreactiv~ fibers in the granule cell layer. (e) Note the numerous TH-immunoreactive neurons and fibers in control bulb. Neuronal elements with TH ~munoreactivity in S,?-DHT-inj~ted bulbs show no decrease (f) or marked decrease (g) in their glomeruli. G, glomerular layer. Magnification, x 167. The rats were perfused through the heart with TOO-I50 ml phosphate-buffered saline (PBS), followed by 400 ml of a mixture of 4% Graformaldehyde, 0.1% glutaraldehyde and 0.2% picric acid in phosphate buffer (PB). The olfactory epithelium, olfactory bulb and other brain were removed, and postfixed overnight with a fixative containing 4% paraformaldehyde and 0.2% pi& acid in PB. The olfactory bulb was cut into three u&es (rostrai, middle and caudal p&s) with a knife. After 30-48 h of incubation in 20% sucrose. the middle pait of the bulb and the brain were cut on a freezing microtome (20 ~1m) m the frontal plane. The sections were incubated for 1h with 5% normal goat serum, fallowed by an overnight incubation with rabbit primary antibodies (Eugene) against 5-HT, DBH and TH at a dilution of l:S,OOO in PBS with 0.3% Triton-X100. An avidin-biotin-~roxidase complex system was used to detect the primary antibodies, with 3.3’-diamino~azidine as a chromogen.
(Fig. 3a). The olfactory bulb consists of well-defined layers, which makes it easy to explore correlation between mo~holo~ and function. Marked changes occurred in the glomerular and olfactory nerve layers with shrinkage of glomeruli and loss of olfactory
nerve bundles (Figs 3a,c). Electron microscopy unveiled that the atrophic glomerulus lacked peripheral afferent terminals mostly without apparent changes of postsynaptic dendritic elements of mitral and tufted cells (Fig. 3e). Consequently, glomerular
Fig, 3. NIorphoiogy of the olfactory bulb and the olfactory epithehum from control rat with normal olfaction (a,b,d and f) and S,%DHT-injected rat with olfactory disturbance (a,c,e,g and h). (a) Note the reduced size of 5,7-DHT-injected bulb (left) compared to the size of control bulb (right). Nis&stain (b and cf normal structure with the olfactory nerve (0). glomerular (G) and external plexiform (E) layers is well preserved in control rat (b). Shrinkage of glomeruli and disappearance of olfactory nerve bundles are apparent in S,?-DHT-injected bulb (cf. Toiuidine Blue stain. (d and e) Electron m~cro~rapbs of glomeruh. Gtomeruius of coutol bulb (dd)shows an intricate framework mixed with offactory afferent terminals of dark appearance and postsynaptic dendrites of clear appearance. Olfactory terminal elements are lost from glomerulus of 5,7-DHT-injected bulb (e) with prominent decrease of synaptic contacts. An asterisk indicates a reactive astrogiial process. (f-h) Oifactory epitheliums immunohistochemically stained by an antibody ta protein gene product 9.5 (PGP 9.5) recagnizing olfactory receptor cells.9~20 Dlfactory epithelium of control rat (f) contains several layers of receptor cells with PGP 9.5 immunoreactivity, Olfactory epitheliums of 5,7-DHT-injected rat (g,h) are decreased in thickness with reduced or no cells with PGP 95 imm~no~ac~~vity. An arrowhead points to basal lamina of the o~actory epjthel~um. Ma~i~~tions: a, x I1.S;b and c, x 16% d and e, x 3,850: f-h, x 167. Sections (20 pm) from the middle bulbar part of the control and 5,?-DHT-injected groups were slaiued by Nissi. The rostra1 buibar part was processed for both light and electron microscopic observations as follows. Specimens were postfixed in 1%osmium t&oxide, dehydrated in ascending series of ethanol, en bloc stained with 1% uranyl acetate in absolute ethanol, processed with propylene oxide and embedded in Epon. For light microscopy, semithin sections (1 pm) were cut from the Epon blocks and stained with Toluidine Blue. These semithin sections were very useful to examine buibar morphology in detail. For electron microscopy, ultrathin sections were brie& stained with lead citrate and observed with &e&on microscope. olfactory epitheliums from the control rats with normal olfaction and 5,~-~~T-inj~~~~ rats with olfactory disturbance were embedded in paraffm. Paraffin sections (5pm) were cut, maunted on giass slides, deparaffinized with xylene and processed for immunohistochemistry with a rabbit anti-PGP 9.5 serum (Ultraclone) at a dihtion of 1:2,oM)to detect olfactory receptor cells. Subsequent procedures were made according to the same protocol described in Fig 2.
737
Role of serotonin in olfaction
synaptic density greatly decreased. The granule cell layer and both the external and internal plexiform layers were atrophic. These morphological changes prompted us to examine olfactory receptor cells of the 5,7-DHT-injected anosmic rats. Immunohistochemistry revealed that receptor cells of the anosmic rats were reduced or lost in parallel with marked atrophy of the olfactory epithelium (Figs 3g,h). lntrabulbar 5,7-DHT injection was carried out to see if the locally applied toxin produces similar effects to those mentioned above. The conditioned rats (n = 9) were bilaterally injected with 5,7-DHT (4 pg in 4~1) into the bulbs. Injection was applied to five different bulbar sites to cover the whole bulb, and reached a total volume of 2-2.5~1 per bulb. The control rats (n = 4) received the same amount of vehicle (isotonic saline containing 0.1% ascorbic acid) into both bulbs. Among the 5,7-DHT-injected rats, three rats lost olfaction with 5&60% correct responses two to four weeks after toxin injection. Immunohistochemistry and histological examinations of these anosmic rats showed that deafferentation of the bulbar 5-HT fibers was complete on both sides with morphological changes similar to those described previously. Olfactory function of the remaining rats injected with 5,7-DHT did not differ from that of the control rats with 9&100% correct responses. because the toxin failed to make effective deafferentation of the bulbar 5-HT fibers on either one side or on both sides. 5,7-DHT, when administered intraventricularly or intracerebrally, causes rapid degeneration of the 5HT fiber pathway within a few days.4.‘o To explore early effects of the toxin on the bulb, we examined the lesioned rats (n = 6) injected with 5,7-DHT into the medial forebrain bundle in the same manner as used in our experiment, and confirmed that in four rats, the toxin did produce complete deafferentation of the
bulbar 5-HT fibers, as early as three days after injection. Of particular interest is the finding that the 5,7-DHT-injected rats with complete deafferentation of the bulbar 5-HT fibers showed no olfactory abnormality three days after toxin injection when those fibers were destroyed. Taken together, it is concluded that long-term as well as complete depletion of the bulbar 5-HT is essential to induce olfactory dysfunction in rats. Recently,
Okado
et ~1.” have reported
an interest-
ing finding, in that removal of S-HT fibers inhibits synaptogenesis in the spinal cord. Olfactory receptor cells have a unique property of undergoing continual degeneration and replacement during life and thus, continual synaptogenesis is ongoing in the bulbar glomerulus.* Therefore, it is reasonable to consider that because of preferential glomerular localization of the bulbar 5-HT fibers, deafferentation of those fibers might have disturbed synaptogenesis at the glomerulus, followed by retrograde degeneration of receptor cells. Another possibility is that 5-HT itself or a chemical regulated by 5-HT, primarily acts as a trophic factor in survival of receptor cells which were severely damaged in the S.7-DHT-injected anosmic rats. The present study provides not only a useful animal model of anosmia, but a basis of understanding the centrally controlled mechanism of olfaction by a specific neurotransmi tter Ackno&dgemenu-We thank Drs S. Nakamura, M. Kuda, Y. Kitao, S. Okoydma, J. Kawano and M. Furukawa for helpful discussions, Drs H. Kimura and I. Tohyama for useful information about immunohistochemicalprocedures,
and Mr S. Nakatani and MSA. Masubara for their technical assistance. This work was supported by a Fund for the Medical Treatment of the Elderly (School of Medicine, Kanazawa University), a Grant provided by the Ichiro Kanehara Foundation and a Grant-in-aid for Scientific Research, Ministry of Education, Science and Culture of Japan.
REFERENCES
1, Baker H., Kawano T., Albert V., Joh T. H., Reis D. J. and Margolis F. L. (1984) Olfactory bulb dopamine neurons survive deafferentation-induced loss of tyrosine hydroxylase. Neuroscience 11, 605-615. 2. Baker H., Kawano T., Margolis F. L. and Joh T. H. (1983) Transneuronal regulation of tyrosine hydroxylase expression in olfactory bulb of mouse and rat. J. Neurosci. 3, 69-78. 3. Baumgarten R., Bloom F. E., Oliver A. P. and Salmoiraghi G. C. (1963) Response of individual olfactory nerve cells to microelectrophoretically administered chemical substances. P’gers Arch. ges. PhyJioL 277, 125-140. 4. BjBrklund A., Nobin A. and Stenevi U. (1973) The use of neurotoxic dihydroxytryptamines as tools for morphological studies and localized lesioning of central indolamine neurons. Z. Zellfarsch. mikrosk. Amt. 145, 479-501. 5. Bloom F. E., Costa E. and Salmoiraghi G. C. (1964) Analysis of individual rabbit olfactory bulb neuron responses to the microelectrophoresis of acetylcholine, norepinephrine and serotonin synergists and antagonists. J. Pharmac. exp. Ther. 146, 16-23. 6. Brennan P., Kaba H. and Keveme E. B. (1990) Otfactory recognition: a simple memory system. Science 250, 1223-1226. 7. Brunjcs P. C., Smith-Crafts L. and MaCarty R. (1985) Unilateral odor deprivation: effects on the development of olfactory bulb catecholamines and behavior. Devl Brain Rex 22, 1-6. 8, Greer C. A. (1991) Structural organization of the olfactory system. In Smell and Taste in Heulrh and &sense (eds Getchell T. V,, Doty R. L., Bartoshuk L. M. and Snow J. B, Jr), pp. 65-81. Raven Press, New York. 9. Iwanaga T., Han H., Kanazawa H. and Fujita T. (1992) lmmunohistochemical localization of protein gene product 9.5 (PGP 9.5) in sensory paraneurons of the rat. Efimed.Res. 13,225-230, 10. Jonsson G. (1983)Chemical lesioning techniques: monoamine neurotoxins. In Handbook of ChemicuI Neuroanatarny (eds Bjarklund A. and Hiikfelt T.), Vol. 1, pp. 463-507. Elsevier, Amsterdam. I 1. Keverne E. B. and de la Riva C. (1982) Pheromones in mice: reciprocal interaction between the nose and brain, Nature 296, 148-150.
738
T. Moriizumi
et al
12. Krmura Y., Miwa T., Furukawa M. and Umeda R. (1991) Effects of topical application of sterords on olfactory disturbance in mice. Chem. Senses 16, 2977302. 13. Kohler C. and Steinbusch H. (1982) Identification of serotonin and non-serotonin-containing neurons of the mid-brain raphe projecting to the entorhinal area and the hippocampal formation. A combined immunohistochemical and fluorescent retrograde tracing study in the rat brain. Neuroscience 7, 951-975. 14. Kucharski D. and Hall W. G. (1987) New routes to early memories. Science 238, 786788. 15. Leon M. (1987) Plasticity of olfactory output circuits related to early olfactory learning. TrenlLFNeurosci. 10,434438. 16. McLean J. H. and Shipley M. T. (1987) Serotonergic afferents to the rat olfactory bulb: I. Origins and laminar specificity of serotonergic inputs in the adult rat. J. Neurosci. 7, 3016~3028. 17. Okado N., Cheng L., Tanatsugu Y., Hamada S. and IIamaguchi K. (1993) Synaptic loss following removal of serotonergic fibers in newly hatched and adult chickens. J. Neurobiol. 24, 687498. 18. Omura K. and Takagi S. F. (1961) On the mechanism of the repellent action of Naramycin to rats. Gunma J. Med. Sci. 10, 211-221. 19. Shipley M. T., Halloran F. J. and de la Torre J. (1985) Surprisingly rich projection from locus coeruleus to the olfactory bulb in the rat. Bruin Res. 329, 294-299. 20. Taniguchi K., Saito H., Okamura M. and Ogawa K. (1993) lmmunohistochemical demonstration of protein gene product 9.5 (PGP 9.5) in the primary olfactory system of the rat. Neurosci. f&f. 156, 24426. (Amptpd
25 April
1994)
Pergamon
0306-4522(94)00224-X
Letter OLFACTORY DEAFFERENTATION
Elsevier ScienceLtd Copyright 0 1994IBRO Printed in Great Britain. All rights reserved 0306-4522194$7.00t 0.00
to neuroscience
DISTURBANCE INDUCED BY OF SEROTONERGIC FIBERS IN THE OLFACTORY BULB
T. MORIIZUMI,*t T. TSUKATANIJ H. SAKASHITA$ and T. MIWA$ *Department of Anatomy and $Department of Otorhinolaryngology, School of Medicine, Kanazawa University, Kanazawa 920, Japan
remain to k elucidated. In the present study, using 5,7-dihydroxytryptamine (5,7-DHT), a specific neurotoxin for 5-HT, we examined whether or not olfactory bulb is one of the major forebrain targets of the ascending serotonin patbway.‘3q’6 According to deafferentation of the bulbar 5-HT fibers induces physiological studies:J neurons of the olfactory bulb olfactory abnormality in rats. The conditioned rats who learned to avoid cyclowere found to reduce their spontaneousdischarge rates heximide were injected with 5,7-DHT into the ascendby electrophoretically applied serotonin. However, roles of the bulbar serotonin in the sease of smell ing 5-HT fiber pathway to deafferentate the 5-HT fibers in the bulb. Then olfactory function was examremain unanswered. In the present study, using 5,7dihydroxytryptamine, a specificneurotoxin for serotonin, ined by capability to discriminate cycloheximide solwe found that the conditioned rats who learned to avoid ution from water. Olfactory function of the control a repellent by olfaction lost ability of discrimination by group remained unchanged throughout the postoperdeafiereatation of the bulbar serotonergic fibers. Such ative period, while olfactory function of the 5,7olfactory dysfunction did not occur in the early stage DHT-injected group varied from rat to rat. From our (three days after injection of the toxin) when the retrospective study, we found a close relationship serotooergic fibers disappeared in the bulb, but devel- between the amount of the bulbar 5-HT fibers and oped a few weeks later. Interestingly, histological olfaction. According to the degree of deafferentation examination revealed marked shrinkage of the bulbar of the 5-HT fibers, we divided the 5,7-DHT-injected glomerulus which is a major termination site of the rats into two groups; those with partial deafferentabulhopetal serotonergic fibers, and also a synaptic site tion and those with complete deafferentation. Figure of olfactory receptor cells and bulbar output neurons. I shows the results of the discrimination test of the The results indicate that depletion of the serotonergic three groups. Olfactory function of the partially fibers in the olfactory bulb causes glomerular atrophy deafferentated group did not differ from that of the and olfactory disturbance in the rat. control group, while olfactory function of the comThe olfactory bulb receives monoaminergic inputs pletely deafferentated group significantly deteriorated from the brain stem.“~16~‘y These monoamines include two to four weeks after toxin injection. The onset of two well-known chemical substances, norepinephrine loss of olfaction varied from rat to rat. Now we can (NA) and serotonin (5-HT). The bulbar NA fibers are say for sure that if the olfactory bulb is completely distributed from the external plexiform layer to the deprived of 5-HT, anosmia develops within four granule cell layer,” and have been reported to be weeks after toxin injection. To know whether the related to formation of olfactory memory.6,” The toxin produces reversible or irreversible olfactory bulbar S-HT fibers are localized mainly in the abnormality, we continued the discrimination test in glomerular layer. I6 Because of their glomerular some rats which have lost olfaction within four weeks location, the 5-HT fibers are speculated to influence after toxin injection, and found no recovery of olfacon synaptic integration, but their roles in olfaction tion up to six weeks. Immunohistochemistry was done to assess nonspecific effects of 5,7-DHT on the bulbar NA fibers, In tTo whom correspondence should be addressed. half of the 5,7-DHT-injected anosmic rats, the toxin Abbmiurions: DBH, dopamine b-hydroxylase; 5,7-DHT, did not affect the hulbar NA fibers (Fig. 2d), despite 5,7-dihydroxytryptamine; S-HT, serotonin; NA, noretotal lack of bulbar 5-HT fibers (Fig. 2b). In the pinephrine; PB, phosphate buffer; PBS, phosphatebuffered saline: TH, tyrosine hydroxylase remaining anosmic rats, the bulbar NA fibers were The serotooergic neurons of the brain stem project
widely throughout the central nervous system, and the
733
734
T. Morizumi
0 Days
after
3
7
injections
14 of
21 5,7-DHT
28 or
vehicle
Fig. 1. Correct responses of the three groups with controls (O-O, n = lo), partial deafferentation of the bulbar S-HT fibers (O-0, n = 9) and complete deafferentation of the bulbar S-HT fibers (A-A, n = 11). The completely deafferentated group shows loss of ability of discrimination 14-28 days after injection of 5,7-DHT. Mean percentages and standard errors of correct responses of the control group are 97 + 2 (three days), 97 f 2 (seven days), 99 i 1 (14 days), 98 + 1 (21 days) and 98 & 1 (28 days). Those values of the partially deafferentated group are 98 + 1, 93 _+2, 96 + 2, 90 + 2 and 94 + 2. Those values of the completely deafferentated group are 97 + 1, 83 + 7, 76 + 6, 59 k 3, 52 f 3. [+I = infusion of 5,7-DHT or vehicle. *Significantly different (P < 0.01) from the values obtained in the control group as determined by Scheffi’s multiple comparison test. The experiments were carried out on adult male Sprague-Dawley rats (250-350 g). All surgical manipulations were done under general anesthesia with sodium pentobarbital (SOmg/kg, i.p.). (I) Conditioning. Since umlateral bulbectomy did not affect odor aversion learning as described later, rats (n = 30) were subjected to removal of unilateral olfactory bulb to assess relationships between deafferentation of the 5-HT fibers, olfactory function and morphological changes by limiting examined olfactory system on one side. In particular, since the neurotoxic effects of 5,7-DHT on S-HT fibers varied in terms of deafferentation of the bulbar 5-HT fibers, the procedure of unilateral bulbectomy has a considerable advantage to precisely examine the effects of depletion of the bulbar 5-HT on olfaction. The unilaterally bulbectomized rats were trained to avoid cycloheximide, which has a strong repellent action to mdents.‘2.‘R They were deprived of water for two to three days
prior to each discrimination test. Each rat was placed in a cage. Two bottles were offered to each rat. One contained 0.01% cycloheximide solution, while another water. Ten trials of discrimination were usually performed at one time. In the 10 trials, the position of the bottle containing cycloheximide solution was made right, left, left, right, left, left, right, right, right and left relative to the position of the bottle containing water. Observations were carried out in an air-conditioned room. When the rat drank water, the response was interpreted as a correct response. When the rat drank cycloheximide solution, the response was regarded as a wrong response. The number of correct responses was divided by the number of total responses, and the percentage of correct responses was calculated. Since the taste of cycloheximide solution is so disgusting, the rats remember the smell of the solution when they drink it, and thereafter learn to avoid it by olfaction. Thus the olfactory conditioned reaction is established. Rats drank cycloheximide solution two to four times in the first 10 trials. The correct responses were increased in the second 10 trials. Usually 30-40 trials were necessary to condition rats, which is quite similar in normal rats without unilateral bulbectomy. Thus all unilaterally bulbectomized rats were successfully conditloned to avoid cycloheximide and gave 100% correct responses in the last 10 trials, indicating that unilateral
el al.
bulbectomy did not affect odor aversion learning. The finding that no differences in odor aversion learning could be detected between normal and unilaterally bulbectomized control rats may be explained by the phenomenon that olfactory information learned by one side of the olfactory system is transferred to another side through olfactory commissural pathways.‘4,‘5 (2) Lesioning. The bulbopetal 5-HT fibers originate from the dorsal and median raphe nuclei, and pass through the medial forebrain bundle.“,16 Thus, to deafferentate the 5-HT fibers in the remaining olfactory bulb, the conditioned rats (n = 20) were injected with 5,7-DHT (Sigma, 4 or 8 pg in 4 ~1) into the rostra1 (A: 2.0, L: 1.5, D: -8.Omm) and caudal (A: -2.0, L: 2.0, D: - 8.5 mm) parts of the medial forebrain bundle ipsilateral to the remaining bulb. For control (n = IO), vehicle (4~1 of isotonic saline containing 0.1% ascorbic acid) was applied to the same region. (3) Olfactory function. After infusion of 5,7-DHT or vehicle into the medial forebrain bundle, the conditioned rats were submitted to odor aversion behavior to examine their olfactory function to discriminate cycloheximide solution from water. Behavioral tasks were performed at three. seven, 14, 21 and 28 days after injection of 5,7-DHT or vehicle. Several rats from the control and 5,7-DHT-injected groups were continued to the discrimination test up to 42 days. To confirm soundness of the discrimination test as a method to examine olfactory function, we removed the remaining olfactory bulb in the unilaterally bulbectomized, conditioned rats (n = 5). All rats with bilateral bulbectomy gave 50% correct responses, indicating loss of olfaction. (4) Grouping. After completing the discrimination test, the rats were deeply anesthetized with sodium pentobarbital @Omg/kg, i.p.), and perfused with a fixative. Deafferentation of the 5-HT fibers in each olfactory bulb and specificity of 5,7-DHT was evaluated by immunohistochemistry as described in Fig. 2. Since most of bulbar 5-HT-immunoreactive fibers are located in the glomerular layer and have distinct varicosities,‘” the number of glomerular axonal varicosities with 5-HT immunoreactivity was counted to estimate the effects of 5,7-DHT on the bulbar 5-HT fibers. Based on the degree of 5-HT depletion at each olfactory bulb, the 5,7-DHT-injected rats were divided into two groups; those with partial deafferentation of the S-HT fibers and those with complete deafferentation of the 5-HT fibers. The bulbs of the partially deafferentated group contained intact, S-HT-immunoreactive fibers of 49.4% (33.3-70.1%) of control value. The bulbs of the completely deafferentated group were almost free of _5-HT-immunoreactive fibers (2.9%, O-9.5% of control value). Thus the rats were divided into three groups; controls, those with partial deafferentation, and those with complete deafferentation. The presence of the anosmic rats apparent decrease of the bulbar NA fibers seems to exclude involvement of NA in olfactory disturbance induced by deafferentation of the bulbar S-HT fibers. The olfactory bulb contains intrinsic dopaminergic neurons. Immunohistochemistry revealed that the bulb of the 5,7-DHT-injected anosmic rats had normal to fairly decreased number of dopaminergic neurons depending on glomeruli (Figs 2f,g). In total, the bulb of these anosmic rats apparently contained fewer dopaminergic neurons than that of the control rats. This is explained well by the slightly
reduced.
without
phenomenon
that
dopamine
cline by odor deprivation
levels
significantly
de-
or deafferentation of peripheral olfactory afferents. ‘L’ Histological examination revealed atrophy of the olfactory bulb of the 5,7-DHT-injected anosmic rats
Role of serotonin in olfaction
735
Fig. 2. Monaamin~ (S-HT, NA and dopamine) in the olfactory bulb of control rat with normal ofaction {ac,e) and 5.7-DHT-injected rat with olfactory disturbance (b,d,f,g). To visualize the monoamines, immunohistochemistry was done using antibodies to S-HT, dopamine j9-hydroxylase (DBH, a NA-synthesizing enzyme) and tyrosine hydroxylase (TH, a synthetic enzyme for both NA and dopamine). Since TH-immunoreactive neurons and fibers in the bulbar glomerular layer lack DBH immunoreactivity, glomerular TH-positive neuronal elements are regarded as dopaminergic. 5-HT- and do~mine~on~ining fibers are rich in the glometular layer, while NA-containing fibers are localized in the mitral and granule cell layers, and the external and internal plexiform layers.-S-HT immunoreactlve fibers are abundant in control bulb (a), but absent in 5.7-DHT-iniected bulb (b). No significant differences between control (c) and 5,7-DHT-injected (d) bulbs are found ii DBH immu~oreactiv~ fibers in the granule cell layer. (e) Note the numerous TH-immunoreactive neurons and fibers in control bulb. Neuronal elements with TH ~munoreactivity in S,?-DHT-inj~ted bulbs show no decrease (f) or marked decrease (g) in their glomeruli. G, glomerular layer. Magnification, x 167. The rats were perfused through the heart with TOO-I50 ml phosphate-buffered saline (PBS), followed by 400 ml of a mixture of 4% Graformaldehyde, 0.1% glutaraldehyde and 0.2% picric acid in phosphate buffer (PB). The olfactory epithelium, olfactory bulb and other brain were removed, and postfixed overnight with a fixative containing 4% paraformaldehyde and 0.2% pi& acid in PB. The olfactory bulb was cut into three u&es (rostrai, middle and caudal p&s) with a knife. After 30-48 h of incubation in 20% sucrose. the middle pait of the bulb and the brain were cut on a freezing microtome (20 ~1m) m the frontal plane. The sections were incubated for 1h with 5% normal goat serum, fallowed by an overnight incubation with rabbit primary antibodies (Eugene) against 5-HT, DBH and TH at a dilution of l:S,OOO in PBS with 0.3% Triton-X100. An avidin-biotin-~roxidase complex system was used to detect the primary antibodies, with 3.3’-diamino~azidine as a chromogen.
(Fig. 3a). The olfactory bulb consists of well-defined layers, which makes it easy to explore correlation between mo~holo~ and function. Marked changes occurred in the glomerular and olfactory nerve layers with shrinkage of glomeruli and loss of olfactory
nerve bundles (Figs 3a,c). Electron microscopy unveiled that the atrophic glomerulus lacked peripheral afferent terminals mostly without apparent changes of postsynaptic dendritic elements of mitral and tufted cells (Fig. 3e). Consequently, glomerular
Fig, 3. NIorphoiogy of the olfactory bulb and the olfactory epithehum from control rat with normal olfaction (a,b,d and f) and S,%DHT-injected rat with olfactory disturbance (a,c,e,g and h). (a) Note the reduced size of 5,7-DHT-injected bulb (left) compared to the size of control bulb (right). Nis&stain (b and cf normal structure with the olfactory nerve (0). glomerular (G) and external plexiform (E) layers is well preserved in control rat (b). Shrinkage of glomeruli and disappearance of olfactory nerve bundles are apparent in S,?-DHT-injected bulb (cf. Toiuidine Blue stain. (d and e) Electron m~cro~rapbs of glomeruh. Gtomeruius of coutol bulb (dd)shows an intricate framework mixed with offactory afferent terminals of dark appearance and postsynaptic dendrites of clear appearance. Olfactory terminal elements are lost from glomerulus of 5,7-DHT-injected bulb (e) with prominent decrease of synaptic contacts. An asterisk indicates a reactive astrogiial process. (f-h) Oifactory epitheliums immunohistochemically stained by an antibody ta protein gene product 9.5 (PGP 9.5) recagnizing olfactory receptor cells.9~20 Dlfactory epithelium of control rat (f) contains several layers of receptor cells with PGP 9.5 immunoreactivity, Olfactory epitheliums of 5,7-DHT-injected rat (g,h) are decreased in thickness with reduced or no cells with PGP 95 imm~no~ac~~vity. An arrowhead points to basal lamina of the o~actory epjthel~um. Ma~i~~tions: a, x I1.S;b and c, x 16% d and e, x 3,850: f-h, x 167. Sections (20 pm) from the middle bulbar part of the control and 5,?-DHT-injected groups were slaiued by Nissi. The rostra1 buibar part was processed for both light and electron microscopic observations as follows. Specimens were postfixed in 1%osmium t&oxide, dehydrated in ascending series of ethanol, en bloc stained with 1% uranyl acetate in absolute ethanol, processed with propylene oxide and embedded in Epon. For light microscopy, semithin sections (1 pm) were cut from the Epon blocks and stained with Toluidine Blue. These semithin sections were very useful to examine buibar morphology in detail. For electron microscopy, ultrathin sections were brie& stained with lead citrate and observed with &e&on microscope. olfactory epitheliums from the control rats with normal olfaction and 5,~-~~T-inj~~~~ rats with olfactory disturbance were embedded in paraffm. Paraffin sections (5pm) were cut, maunted on giass slides, deparaffinized with xylene and processed for immunohistochemistry with a rabbit anti-PGP 9.5 serum (Ultraclone) at a dihtion of 1:2,oM)to detect olfactory receptor cells. Subsequent procedures were made according to the same protocol described in Fig 2.
737
Role of serotonin in olfaction
synaptic density greatly decreased. The granule cell layer and both the external and internal plexiform layers were atrophic. These morphological changes prompted us to examine olfactory receptor cells of the 5,7-DHT-injected anosmic rats. Immunohistochemistry revealed that receptor cells of the anosmic rats were reduced or lost in parallel with marked atrophy of the olfactory epithelium (Figs 3g,h). lntrabulbar 5,7-DHT injection was carried out to see if the locally applied toxin produces similar effects to those mentioned above. The conditioned rats (n = 9) were bilaterally injected with 5,7-DHT (4 pg in 4~1) into the bulbs. Injection was applied to five different bulbar sites to cover the whole bulb, and reached a total volume of 2-2.5~1 per bulb. The control rats (n = 4) received the same amount of vehicle (isotonic saline containing 0.1% ascorbic acid) into both bulbs. Among the 5,7-DHT-injected rats, three rats lost olfaction with 5&60% correct responses two to four weeks after toxin injection. Immunohistochemistry and histological examinations of these anosmic rats showed that deafferentation of the bulbar 5-HT fibers was complete on both sides with morphological changes similar to those described previously. Olfactory function of the remaining rats injected with 5,7-DHT did not differ from that of the control rats with 9&100% correct responses. because the toxin failed to make effective deafferentation of the bulbar 5-HT fibers on either one side or on both sides. 5,7-DHT, when administered intraventricularly or intracerebrally, causes rapid degeneration of the 5HT fiber pathway within a few days.4.‘o To explore early effects of the toxin on the bulb, we examined the lesioned rats (n = 6) injected with 5,7-DHT into the medial forebrain bundle in the same manner as used in our experiment, and confirmed that in four rats, the toxin did produce complete deafferentation of the
bulbar 5-HT fibers, as early as three days after injection. Of particular interest is the finding that the 5,7-DHT-injected rats with complete deafferentation of the bulbar 5-HT fibers showed no olfactory abnormality three days after toxin injection when those fibers were destroyed. Taken together, it is concluded that long-term as well as complete depletion of the bulbar 5-HT is essential to induce olfactory dysfunction in rats. Recently,
Okado
et ~1.” have reported
an interest-
ing finding, in that removal of S-HT fibers inhibits synaptogenesis in the spinal cord. Olfactory receptor cells have a unique property of undergoing continual degeneration and replacement during life and thus, continual synaptogenesis is ongoing in the bulbar glomerulus.* Therefore, it is reasonable to consider that because of preferential glomerular localization of the bulbar 5-HT fibers, deafferentation of those fibers might have disturbed synaptogenesis at the glomerulus, followed by retrograde degeneration of receptor cells. Another possibility is that 5-HT itself or a chemical regulated by 5-HT, primarily acts as a trophic factor in survival of receptor cells which were severely damaged in the S.7-DHT-injected anosmic rats. The present study provides not only a useful animal model of anosmia, but a basis of understanding the centrally controlled mechanism of olfaction by a specific neurotransmi tter Ackno&dgemenu-We thank Drs S. Nakamura, M. Kuda, Y. Kitao, S. Okoydma, J. Kawano and M. Furukawa for helpful discussions, Drs H. Kimura and I. Tohyama for useful information about immunohistochemicalprocedures,
and Mr S. Nakatani and MSA. Masubara for their technical assistance. This work was supported by a Fund for the Medical Treatment of the Elderly (School of Medicine, Kanazawa University), a Grant provided by the Ichiro Kanehara Foundation and a Grant-in-aid for Scientific Research, Ministry of Education, Science and Culture of Japan.
REFERENCES
1, Baker H., Kawano T., Albert V., Joh T. H., Reis D. J. and Margolis F. L. (1984) Olfactory bulb dopamine neurons survive deafferentation-induced loss of tyrosine hydroxylase. Neuroscience 11, 605-615. 2. Baker H., Kawano T., Margolis F. L. and Joh T. H. (1983) Transneuronal regulation of tyrosine hydroxylase expression in olfactory bulb of mouse and rat. J. Neurosci. 3, 69-78. 3. Baumgarten R., Bloom F. E., Oliver A. P. and Salmoiraghi G. C. (1963) Response of individual olfactory nerve cells to microelectrophoretically administered chemical substances. P’gers Arch. ges. PhyJioL 277, 125-140. 4. BjBrklund A., Nobin A. and Stenevi U. (1973) The use of neurotoxic dihydroxytryptamines as tools for morphological studies and localized lesioning of central indolamine neurons. Z. Zellfarsch. mikrosk. Amt. 145, 479-501. 5. Bloom F. E., Costa E. and Salmoiraghi G. C. (1964) Analysis of individual rabbit olfactory bulb neuron responses to the microelectrophoresis of acetylcholine, norepinephrine and serotonin synergists and antagonists. J. Pharmac. exp. Ther. 146, 16-23. 6. Brennan P., Kaba H. and Keveme E. B. (1990) Otfactory recognition: a simple memory system. Science 250, 1223-1226. 7. Brunjcs P. C., Smith-Crafts L. and MaCarty R. (1985) Unilateral odor deprivation: effects on the development of olfactory bulb catecholamines and behavior. Devl Brain Rex 22, 1-6. 8, Greer C. A. (1991) Structural organization of the olfactory system. In Smell and Taste in Heulrh and &sense (eds Getchell T. V,, Doty R. L., Bartoshuk L. M. and Snow J. B, Jr), pp. 65-81. Raven Press, New York. 9. Iwanaga T., Han H., Kanazawa H. and Fujita T. (1992) lmmunohistochemical localization of protein gene product 9.5 (PGP 9.5) in sensory paraneurons of the rat. Efimed.Res. 13,225-230, 10. Jonsson G. (1983)Chemical lesioning techniques: monoamine neurotoxins. In Handbook of ChemicuI Neuroanatarny (eds Bjarklund A. and Hiikfelt T.), Vol. 1, pp. 463-507. Elsevier, Amsterdam. I 1. Keverne E. B. and de la Riva C. (1982) Pheromones in mice: reciprocal interaction between the nose and brain, Nature 296, 148-150.
738
T. Moriizumi
et al
12. Krmura Y., Miwa T., Furukawa M. and Umeda R. (1991) Effects of topical application of sterords on olfactory disturbance in mice. Chem. Senses 16, 2977302. 13. Kohler C. and Steinbusch H. (1982) Identification of serotonin and non-serotonin-containing neurons of the mid-brain raphe projecting to the entorhinal area and the hippocampal formation. A combined immunohistochemical and fluorescent retrograde tracing study in the rat brain. Neuroscience 7, 951-975. 14. Kucharski D. and Hall W. G. (1987) New routes to early memories. Science 238, 786788. 15. Leon M. (1987) Plasticity of olfactory output circuits related to early olfactory learning. TrenlLFNeurosci. 10,434438. 16. McLean J. H. and Shipley M. T. (1987) Serotonergic afferents to the rat olfactory bulb: I. Origins and laminar specificity of serotonergic inputs in the adult rat. J. Neurosci. 7, 3016~3028. 17. Okado N., Cheng L., Tanatsugu Y., Hamada S. and IIamaguchi K. (1993) Synaptic loss following removal of serotonergic fibers in newly hatched and adult chickens. J. Neurobiol. 24, 687498. 18. Omura K. and Takagi S. F. (1961) On the mechanism of the repellent action of Naramycin to rats. Gunma J. Med. Sci. 10, 211-221. 19. Shipley M. T., Halloran F. J. and de la Torre J. (1985) Surprisingly rich projection from locus coeruleus to the olfactory bulb in the rat. Bruin Res. 329, 294-299. 20. Taniguchi K., Saito H., Okamura M. and Ogawa K. (1993) lmmunohistochemical demonstration of protein gene product 9.5 (PGP 9.5) in the primary olfactory system of the rat. Neurosci. f&f. 156, 24426. (Amptpd
25 April
1994)