0306-4522/91 $3.00 + 0.00 Pergamon Press pie © 1991 IBRO
Neuroscience Vol. 42, No. 2, pp. 365-377, 1991 Printed in Great Britain
MORPHOLOGICAL, NEUROCHEMICAL, A N D BEHAVIORAL STUDIES ON SEROTONERGIC DENERVATION A N D G R A F T - I N D U C E D REINNERVATION OF THE RAT HIPPOCAMPUS M. G. P. A. VAN LtnJTELAAR,*~ J. A. D. M. TONNAER,~ A. L. FRANKHUIJZEN,~"H. DIJKSTRA,t J. J. HAGAN~ and H. W. M. STEINBUSCH'~ 1"Department of Pharmacology, Free University, Amsterdam, The Netherlands :~Department of CNS Pharmacology, Organon Int. BV, Oss, The Netherlands Abstract--A procedure was developed to conduct simultaneously immunocytochemical and neurochemical studies on the serotonergic system in adjacent 300-#m-thick slices of rat hippocampus. This procedure was applied to correlate morphological (innervation pattern and density), neurochemical (5-hydroxytryptamine and 5-hydroxyindolacetic acid levels and [3H]5-hydroxytryptamine uptake and release) and behavioral (spatial learning) effects of neurotoxin-induced denervation and reinnervation by grafting fetal mesencephalic raphe cells. Intracerebroventricular injections of a low dose of 5,7-dihydroxytryptamine caused a discrete serotonergic denervation of the hippocampus. Eleven months after lesioning, 5-hydroxytryptamine and 5-hydroxyindolacetic acid levels and [3H]5-hydroxytryptamine uptake capacity were decreased by 50--60°/°. By this time, the residual fibers displayed an enhanced vulnerability towards K ÷-induced depolarization. Grafting of a fetal raphe cell suspension resulted in a reinnervation of the host hippocampus. The pattern of reinnervation was comparable to control innervation and the density was supranorrnal at the level of the graft. As observed semiquantitatively, the innervation density decreased with distance from the core of the graft. Neurochemical studies showed that the fibers were capable of synthesizing, metabolizing and releasing 5-hydroxytryptamine. The turnover of 5-hydroxytryptamine in both the denervated and the reinnervated hippocampus was comparable to that in control tissue. Previous behavioral testing of the denervated and of the denervated and implanted animals did not reveal any effect on spatial learning, either in an individual or in a social test paradigm. The latter data substantiate the notion that interference with the hippocampal serotonergic innervation does not hamper adequate spatial learning.
Long after developmental changes in the CNS are completed, serotonergic neurons still display a certain capacity for plasticity)'~'47'48 Thus, compensatory processes upon chemical lesioning of the adult rat brain have been described, such as sprouting of serotonergic fibers after partial lesioning by 5,7-dihydroxytryptamine (5,7-DHT) H,4s and increases in the turnover of 5-hydroxytryptamine (5-HT) in the denervated brain areas) 3 These compensation mechanisms are attenuated in the aged rat C N S ) 2,34,35It is thought that, amongst others, physical and humoral factors in the target areas of the serotonergic fibers in the mature brain are involved in the control of the neuronal plasticity. Hence, physiological and pathological conditions of the target tissue will influence the plasticity. The exact regulatory mechanisms are, however, far from understood. Denervation and subsequent implantion of fetal brain tissue provides an excellent tool to study the regulation of neuronal plasticity in adult and aged *To whom correspondence should be addressed. 5,7-DHT, 5,7-dihydroxytryptamine; 5HIAA, 5-hydroxyindolacetic acid; HPLC, high pressure liquid chromatography; 5-HT, 5-hydroxytryptamine; MR, mesencephalic raphe.
Abbreviations:
host brain. Thus, morphological studies have shown that fetal mesencephalic raphe (MR) cells survive for long periods and establish extensive networks upon grafting into a host b r a i n y Neurochemical studies have indicated changes in 5-HT and 5-hydroxyindolacetic acid (5-HIAA) levels following grafting into the brain. 4,2t Moreover, implants in the hippocampus display electrophysiological activity. 36,43 Behavioral studies reported the sequelae of grafting fetal cells into various brain areas) °,2°,25,44 In these studies analysis was made of either morphological features of grafted neurons, functional activity of the growing fibers or their possible involvement in the regulation of behavior. Our experiments aimed to study, within the same animal, the relationship between morphological (innervation pattern and density), neurocbemical (5-HT and 5-HIAA levels and [3H]5-HT uptake and release) and behavioral (spatial learning) parameters. The hippocampus was chosen as the target area for several reasons. Firstly, it has a high level of morphological organization, which allows accurate histological examination. Secondly, the hippocampus is an easily accessible structure and contamination in neurochemical analyses by non-hippocampal tissue can thus be readily avoided. Thirdly, the involvement
365
366
M. G. P. A. VAN LUIJTELAARet al.
o f the h i p p o c a m p u s in spatial learning has been well established 3°'31 a n d behavioral tests are available. 22 T h e serotonergic system was chosen as a selective depletion o f the serotonergic system can be achieved by intracerebroventricular 5,7-DHT injections. Moreover, o u t g r o w t h o f fetal M R cells implanted in the h i p p o c a m p u s will result in a dense i n n e r v a t i o n of the host tissue. Finally, the serotonergic h i p p o c a m p a l system has been implicated in m n e m o n i c functions. 2 EXPERIMENTAL PROCEDURES
Thirty-six male Wistar rats (Harlan CPB, Zeist, The Netherlands, 140-160g body weight) were randomly divided into four equal groups: control (C), sham-lesioned (ShL), lesioned (L) and lesioned + implanted (T). Behavioral testing started five months after surgery. Eleven months after surgery, the animals were killed for morphological and neurochemical analysis.
Surgery Under xylazine (25 mg/kg, s.c.) and ketamine (50 mg/kg, i.m.) anesthesia, 10/~g 5,7-DHT dissolved in 5/~1 saline with 0.02% ascorbic acid was injected into both lateral ventricles (0.8 mm posterior, 1.5 mm lateral to bregma, 3.2 mm below the surface of the overlying cortex). Animals were pretreated with desimipramine (25mg/kg, i.p.). Sham-lesion consisted of two 5-#1 vehicle injections. Fourteen days after the 5,7-DHT injections the T group was anesthetized as described above and received injections of 2/~1 from a cell suspension obtained from the MR region of El5 fetuses. 37 Approximately 100,000 viable cells, counted after incubation with Trypan Blue, were implanted bilaterally into the hippocampus (3.0 mm posterior, 1.8 mm lateral to bregma, 3.0mm below the surface of the overlying cortex). From each experimental group one animal was killed three months after surgery to serve as control for the immersion fixation that was to be used in the final experiments. Brains were fixed by perfusion fixation and processed as has been described elsewhere.45
Histochemical and neurochemical analysis Eleven months after surgery the animals were killed for combined neurochemical and histochemical analysis. Each experiment included one animal from each group. Upon decapitation, the hippocampal lobes were dissected on ice, cut into 300-/~m slices with a Mcllwain tissue chopper and divided into five levels according to the scheme in Fig. 1. From each level the first slice was processed for serotonin immunocytochemistry, the next two slices were processed for uptake and release studies and the remaining two slices (per level) were immediately frozen on dry ice, stored at - 7 0 ° C until HPLC analysis followed (details are given below). From C and ShL groups only the left hippocampal lobes were used; from the L and T groups both hippocampal lobes were used.
lmmunocytochemistry Immunocytochemistry was performed on the first three levels of C and ShL groups and on all five levels of the L and T groups. After 2 h of immersion fixation in 4% paraformaldehyde, 0.05% glutaraldehyde, 0.2% picric acid and 5% sucrose in 0. IM phosphate buffer (pH 7.4, 4°C), the slices were rinsed for 30m in in 5% sucrose in 0.1 M phosphate buffer. They were embedded in Tissue Tek OCT simultaneously, frozen with powdered dry ice and cut on a cryostat into 10-/tm sections at -18°C. Sections were collected on gelatin-chrome-alum coated slides and stored at - 2 0 ° C until further processing. The sections were pretreated for 30 min at room temperature with 0.3% peroxide
ICC
Re lease
H PI.C
Fig. l. Scheme showing the dissection of the hippocampal lobes. Five adjacent levels (1-5) with a dorsal-ventral orientation were each chopped into five 300-/t m s/ices. These slices were then used for immunocytoehemistry (ICC), uptake and release, and HPLC studies. in methanol to inhibit endogenous peroxidase activity. Immunocytochemical staining for 5-HT immunoreactivity was performed according to the peroxidase-antiperoxidase method a9 with a rabbit serotonin antiserum. 41 For details of the staining procedure see Ref. 46.
Uptake and release studies Slices from the first three levels were used for uptake and release studies. The two slices (per level) were incubated with radiolabeled serotonin and choline and were superfused as described previously. ~ In short, after a preincubation with Krebs-Ringer medium, the sets of slices were incubated during 30 min in medium containing 1/~Ci [3H]5-HT and 0.5/tCi [~4C]choline (final concentrations +__O.1 and 10/~M). After 45 min of superfusion, five successive 10 rain fractions were collected. Radioactivity remaining in the slices was extracted with 3 ml 0.1 N HCI. The total radioactivity in the superfusion fractions and extracts (Et) was taken as a measure for the uptake. Depolarization-induced calciumdependent transmitter release in excess of spontaneous efl~ux (effected by 10 min superfusion with medium containing 20 mM K +) was calculated as the percentage of the total radioactivity present in the tissue at the onset of the stimulation period. Intervial variance was accounted for by adjusting the [3H]5-HT content values to those of the II4C]acetylcholine which were previously shown not to be affected by 5,7-DHT lesion) 8
High pressure liquid chromatography analysis HPLC analysis was performed on the slices from levels 1 and 3. The slices (per level) were pooled and homogenized in 150/zl of cold 0.1 N HCI04, containing 12 ng of DHBA as internal standard. Upon centrifugation for 3 min at 8750 × g the pH of 100-~1 aliquots of these solutions was adjusted to 4 with 2 M NaAc, and 50-/~1 aliquots were subjected to HPLC for the separation of 5-HT and 5-HIAA. Chromatography was performed as described previously. ~a Protein content of the pellets was assessed according to Lowry's method. 19
Serotonergic denervation and graft-induced reinnervation of rat hippocampus
Statistical analysis of neurochemical data Data from the distinct levels within C, ShL and L rats, and for the latter from both hemispheres, were averaged, yielding one value per animal. Data from the distinct levels from both hemispheres within the T groups were treated as single values. All values were then converted to percentages of the corresponding control values and analysed by a one-way ANOVA, followed by a Duncan multiple range test. Data are presented as the mean _ S.E.M. Behavioral tests Spatial learning was tested in a social context and in an individual setting. Five months after surgery, C, L and T groups (n = 8) were submitted to the first test. The task, performed at one-day intervals in a 2 × 3 m cage, was reinforced both by food reward and social avoidance. At the end of the light period (9.30 a.m.) food was removed from the home cages. Three hours later the subject was placed in the center of a compartment (2 x 0.50 m) and allowed to make a choice between two identical boxes (0.15 × 0.30 m) located at opposite sides of the cage. Several visuospatial cues were present. When a box was entered, the entrance door was closed and the access door to the corresponding goal compartment behind it (2 × 2.5m) was opened. Correct responding was rewarded by 15 min of access to a food platform. Choice of the incorrect box was punished by exposure to one attack by a dominant colony male. After reaching criterion (three correct trials on successive days), the goal compartments were interchanged and two reinforced reversal training trials were given followed by nine test trials. Choice latency, number of trials to criterion, percentage of correct choices and the longest string of correct choices were taken as parameters of orientation behavior. Eight to nine months after surgery, spatial learning was tested in an individual context in a Morris water maze. 22 Latency to escape on the hidden platform was recorded during the training trials (four on day 1, eight trials on days 2 and 3, and four trials on day 4). At the end of the last training trial a 60 s transfer test was performed in which (i) the total number crossing the exact location previously occupied by the platform, and (ii) the total time spent searching within the training quadrant were taken as a measure of spatial bias. Two platform spatial discrimination training was conducted in a different experimental room. In this pool a rigid platform was placed (SW or NE) and a floating platform that was changed in position from trial to trial was used. Rats were given 10 trials on the first day of training and 20 trials on the subsequent days. Training sessions were separated by at least 48 h. Choice latency, number of correct choices and the longest string of correct choices were measured. RESULTS
Immunocytochemistry The preliminary processing of four animals to verify the extent of the lesion and to obtain a control for the adapted immunocytochemical procedure showed that the serotonergic innervation of the ShL hippocampus equals that of control animals (Fig. 2A, B). 5,7-DHT injection yielded hippocampal lobes that were almost devoid of serotonergic fibers (Fig. 2C). In the implanted dorsal hippocampus a dense reinnervation emerging from the grafted M R cells was observed throughout the hippocampus resembling the pattern of control innervation (Fig. 2D). The grafts were located bilaterally in the dorsal hippocampus (Fig. 2E). After immersion
367
fixation of the 300-#m slices as performed in the final experiments, staining comparable to that after perfusion fixation was achieved. Fibers were deafly and distinctly stained (Figs 2D and 3B). Some leakage of serotonin occurred from the cells, which was observed as a less sharp delineation of the borders of the serotonergic cells. Further presentation of the results is solely based on the material obtained from the combined experiments. It appeared impossible to discern the injection track macroscopically and only after immunohistochemical analysis were we able to locate the core of the implant. Eight hippocampal lobes showed neither surviving serotonergic cells nor their major outgrowth in the first level which would indicate the core of the implant. They were eliminated from statistical calculations. Two implanted rat groups that did not show any serotonergic fibers in their hippocampal lobes and one lesioned rat that appeared not denervated (normal fiber density) were eliminated from further analysis. Sham-lesioning did not affect the fiber distribution and density as compared to the control innervation pattern and density (Figs 3A and 4A). However, 5,7-DHT treatment resulted in a severely diminished fiber density throughout the hippocampal lobes (Figs 3A and 4A). This decreased fiber density was not only observed in the dorsal hippocampus, i.e. in the direct vicinity of the injection site, but also in the ventral hippocampus. Only in the most ventral level of the hippocampus were some serotonergic fibers encountered. U p o n implantation the fetal cells reinnervated the surrounding host tissue. Within the implant the fiber density appeared to be extremely high. Both the dentate gyrus (Fig. 3B) and the CA regions (Fig. 4B) received a dense innervation. The pattern of the fiber distribution in the host hippocampus, however, was comparable to the pattern in control animals. Thus, the pyramidal and granular cell layers in the A m m o n ' s horn and the dentate gyrus were only sparsely innervated, whereas a dense innervation of the other layers was found. Following the outgrowth over distances of 1500, 3000, 4500 and 6000 # m from the site of implantation (levels T2-T5; Figs 3B and 4B), the fiber density gradually decreased. However, even in the most distal part of the hippocampus (level T5; Figs 3B and 4B) the fiber density exceeded the density of the non-implanted lesioned hippocampus.
Uptake and release experiments There was no difference measured between the mean tissue content of radiolabeled serotonin in slices from the control or the sham-lesioned group ( 1 0 0 _ 11% versus 99-1-8%; Fig. 5A). Slices from the lesioned group showed a statistically significant reduction in [3H]5-HT uptake capacity to 42 + 4% of control values. Slices from the most rostral level of the implanted hippocampal lobes (T 1), containing the
368
M. G. P. A. VAN LUIJTELAARet al. ,i
-%
Fig. 2. Serotonin immunoreactivity in the hippocampus (35-#m sections) three months after surgery. (A) Control innervation of the hilus. (B) Sham-lesioned hippocampus. (C) Lesioned hippocampus. Note the absence of serotonin immunoreactivity. (D) Implanted hippocampus. The very dense innervation of the molecular layer of the dentate gyrus is clearly visible. (E) An overview of a bilaterally grafted animal. The border of the grafts is indicated with arrowheads. Scale bars = 150/~m (A-D); 750 #m (E). grafted cells, showed a five-fold increase in uptake in comparison to slices from the lesioned animals (194 + 27% versus 42 ___4%; Fig. 5A). This rise in
uptake capacity decreased with the distance from the implanted cells. Uptake at level T2 (130 + 13%) and level T3 (110 ___20%) was significantly higher than
Serotonergic denervation and graft-induced reinnervation of rat hippocampus uptake in slices from lesioned animals but did not exceed the uptake capacity of the control animals (Fig. 5A). In all groups (C, ShL, L and T), K +-stimulation yielded a release of radiolabeled serotonin in surplus of the spontaneous effiux during standard superfusion conditions (Fig. 5B). Sham lesioning had no effect on the K+-stimulated [3H]5-HT release when compared to release in controls. Lesioning resulted in an increase of the mean K+-stimulated [3H]5-HT release to 206% of the control value. In the transplanted group the mean K +-stimulated release from levels T1, T2 and T3 was respectively 153_ 14%, 161 _+ 12% and 110_+ 11% of the unoperated control group.
High pressure liquid chromatography analysis The sham-lesioned rats showed no altered 5-HT levels when compared to control (96 _+ 6% versus 100_+ 18%; Fig. 6A). Injection of 2 0 # g 5,7-DHT into the lateral ventricles caused a reduction of the 5-HT levels to 50 _+ 6% of control. No differences were found between the left and right hippocampal lobes. The levels of the fight hippocampal lobes were used for further calculations. Implantation of fetal mesencephalic cells enhanced 5-HT levels to supranormal values (170+ 33%). When the overall measurements were divided into rostral (T1) and middle (T3) levels, the effect of the outgrowth of the fetal M R cells became evident. At the most rostral level (TI), 5-HT levels were doubled as compared to control values (199 _+ 34%) and four times as high as measured in the lesioned hippocampi (Fig. 6A). The 5-HT levels at the more distal level (T3; 93_+ 15%) were equal to those in the control animals (100 _+ 18%), but still twice as high as in the lesioned animals (50 _+ 6%; Fig. 6A). The same pattern of changes was observed for the 5-HIAA levels (Fig. 6B). Sham lesioning did not affect 5-HIAA levels (93_+5%), while treatment with 5,7-DHT reduced 5-HIAA levels to 47 _+ 9% of control values. The overall level of 5-HIAA in the transplanted group was 120 + 13% of control. From the implanted hippocampal lobes, level T1 showed the major increase in 5-HIAA content (150 + 16%), while slices from level T3 (85 -+ 10%) did not differ from control. When compared to the lesioned hippocampi, 5-HIAA levels in level T1 showed a three-fold increase and those in level T3 a two-fold increase (Fig. 6B). Ratios between 5-HIAA and 5-HT did not differ between the treatment conditions. The ratios in sham-lesioned, lesioned and lesioned + implanted lobes (97 _+ 5%, 109 _+ 24%, 85 + 4%) were comparable to the ratio in the control group (100 + 14%; Fig. 6C). 5-HT levels and [3H]5-HT accumulation in individual levels from control, sham-lesioned, lesioned and lesioned + implanted animals showed a correlation of r = 0.78 (P < 0.01, data not shown).
369
Behavioral tests During the orientation task in a social context, correct responding was always accompanied by eating. Consequently, admittance to the food platform may be considered as a salient reward. During the acquisition phase the number of trials to reach criteflon was equal in all groups. During the reversal testing the control males made their choice 282 s after introduction (logarithmic mean; 95% confidence limits: 108-733s); this choice was correct in 61 _+ 9.9% of the trials. The criterion was reached at trial 4.8 + 0.9 and the longest string of correct choices amounted to 4.4_+0.9 trials. Two-way A N O V A showed that treatment did not influence the individual choice latencies [F(2,18)=2.32, N.S.] whereas repeated testing did [F(8,144) = 2.89, P < 0.005]. Individual parameters of correct orientation (trials to criterion, total number of correct responses and the longest string of correct choices) were not affected by the treatments (one-way ANOVA; dr2,18; F = 0.44, F = 0.07 and F = 0.014; N.S.). Through the reversal trials no two treatment groups differed significantly with respect to the percentage of rats which responded correctly, suggesting a similar development of orientation reversal in all three groups tested. In the Morris water maze most rats adopted the normal adult swimming posture and showed no evidence of motoric difficulty. Testing of one rat was abandoned when it repeatedly jumped out of the pool during training. During acquisition, the escape latency decreased [F(23,575)= 4.39, P < 0.001] and although the groups x trial interaction reached significance [F(69,575) = 1.32, P < 0.05], post hoc tests revealed no consistent difference between the groups throughout the training. Analysis of the transfer test showed that after training the groups did not differ in both parameter measures (control values: 2.1 _+ 0.8 crossings and 26.4 + 2.8 s spent swimming in the training quadrant). During training on the twoplatform spatial discrimination task, lesioned and transplanted animals tended to perform worse than controls, but analysis failed to confirm any effect of treatments on either parameters (control values: 36.7 + 3.7 correct choices and 11.7 _+ 2.2 consecutive correct choices). DISCUSSION
The experimental design enables the simultaneous description of both morphological parameters, such as serotonergic fiber pattern and density, and neurochemical estimates, such as 5-HT and 5-HIAA levels and [3H]5-HT uptake and release, in hippocampal tissue from individual rats. Fresh, unfixed tissue is a prerequisite for neurochemical analyses such as HPLC and [3H]5-HT uptake and release experiments. However, a consequence of the use of fresh tissue is a reduced binding of serotonin to tissue components. Therefore, a reduced histological resolution might be expected. Nevertheless, no differences were observed
370
M . G . P . A . VAN LUIJTELAARel al.
!if
~_
IL_
Fig. 3(A).
Fig, 3(B). Fig. 3. Serotonin immunoreactivity in the dentate gyrus of the hippoeampal lobes as observed with dark-field microscopy. (A) Control (C), sham-lesioned (ShL), lesioned (L) and (B) lesioned + transplanted (T) rats. Sham lesioning did not affect the fiber density or distribution while 5,7-DHT treatment resulted in a severely diminished fiber density. The TI level contains the implant (borders are indicated with arrowheads). The levels are 1500 y m away from each other. Note the supranormal fiber density at level T1 and the decline in density with increasing distance to the implant. The fiber density at level T5 still exceeds the density of the lesioned hippocampal lobes• g, granular layer• Scale bar = 100/am. 371
372
M . G . P . A . VAN LUIJTELAARet al.
Fig. 4(A).
Fig. 4(B). Fig. 4. Serotonin immunoreactivity in the CA2-CA3 region of the hippocampal lobes as observed with dark-field microscopy. (A) Control (C), sham-lesioned (ShL), lesioned (L) and (B) lesioned + transplanted (T) rats. Sham lesioning did not affect the fiber density or distribution while 5,7-DHT treatment resulted in a severely diminished fiber density. Upon implantation the region is strongly reinnervated. Note the supranormal fiber density at level T1 and the decline in density with the increasing distance to the implant. The fiber density at T5 still exceeds the density of the lesioned hippocampal lobes, p, pyramidal cell layer. Scale bar = 100/~m. 373
374 o t_ c o o
M. G. P. A. VAN LUIJTELAARet
al.
250
250
A
A o L
200
200
o
150 0 -'a
I-" T I -r m
T
150
z~
.$
100' •litv
50 0
5
5
5
6
6
6
C
ShL
L
T1
T2
T3
0 --
.
o t_ to o
200
E
150
200
o
150-
=
100-
7
7
7
C
ShL
L
8 T1
T3
250
B
c o
.4
50' zl3
250 o ,-
100'
T
100
u
e t® -
50
50'
I I O
8 C
ShL
L
T1
T2
0
T3
Fig. 5. Tissue content (A: Et values) and K+-stimulated release (B: %P) of [3H]5-HT in slices from sham-lesioned (ShL), lesioned (L) and lesioned + transplated (T) hippocampal lobes compared to control values (100 + 11%). (A) Mean uptake in the slices from the first three levels of sham-lesioned and lesioned rats and the uptake in the slices from the levels T1, T2 and T3 of implanted rats is expressed as percentage of the mean uptake in slices from levels 1-3 in control rats. (B) K÷-stimulated [3H]5-HT release from the same slices as mentioned under A. The numbers in the bars represent the number of observations. ( ~ ) P < 0.05 when compared to control; (A) P < 0.05 and (A) P < 0.01 when compared to lesioned hippocampal lobes.
~ "E o v .o ,.° ~_ iI < t
to
ShL
L
TI
T3
250 200150100 T
50 7
-r-
between staining of serotonin-immunoreactive fibers upon immersion fixation and perfusion fixation of the hippocampus. 5,7-DHT treatment of the animals resulted in a marked decrease in serotonergic innervation density as observed histologically. U p o n partial denervation, sprouting of serotonergic fibers has been reported. H.48 In the present study, no indication of such a regeneration was found, as even 11 months after depletion, the hippocampus is still devoid of serotonergic fibers. The decrease in fiber density was paralleled by changes in 5-HT and 5-HIAA levels and in [3H]5-HT uptake capacity. Similar neurochemical observations have been reported after neonatal or adult administration of 5,6- or 5,7-DHT. 9'27'32'42 Although both the histological and neurochemical results point to a marked denervation, a discrepancy was apparent between the absolute values obtained by both approaches. Lesioned hippocampi stained for serotonin immunoreactivity showed hardly any remaining fibers, whereas neurochemical data from both H P L C analysis and uptake experiments indicated a reduction to only 45-50% of control values. At the concentration of [3H]5-HT used in this study,
C
0
C
8 ShL
L
T1
T3
Fig. 6. 5-HT levels (A), 5-HIAA levels (B) and 5-HIAA/5HT ratios (C) in slices from sham-lesioned (ShL), lesioned (L) and lesioned + transplanted (T) hippocampal lobes compared to control values (100 + 18%) as determined by HPLC analysis. 5-HT and 5-HIAA levels were determined in the first and the third levels from control, sham-lesioned (ShL) and lesioned (L) rats and in the levels T1 and T3 from transplanted rats. For ShL and L rats the mean values from the two levels are presented. The numbers in the bars represent the number of observations. (~r) P < 0.05 and (~') P < 0.01 when compared to control; (A) P < 0.05 and (/k) P
Neuroscience Vol. 42, No. 2, pp. 365-377, 1991 Printed in Great Britain
MORPHOLOGICAL, NEUROCHEMICAL, A N D BEHAVIORAL STUDIES ON SEROTONERGIC DENERVATION A N D G R A F T - I N D U C E D REINNERVATION OF THE RAT HIPPOCAMPUS M. G. P. A. VAN LtnJTELAAR,*~ J. A. D. M. TONNAER,~ A. L. FRANKHUIJZEN,~"H. DIJKSTRA,t J. J. HAGAN~ and H. W. M. STEINBUSCH'~ 1"Department of Pharmacology, Free University, Amsterdam, The Netherlands :~Department of CNS Pharmacology, Organon Int. BV, Oss, The Netherlands Abstract--A procedure was developed to conduct simultaneously immunocytochemical and neurochemical studies on the serotonergic system in adjacent 300-#m-thick slices of rat hippocampus. This procedure was applied to correlate morphological (innervation pattern and density), neurochemical (5-hydroxytryptamine and 5-hydroxyindolacetic acid levels and [3H]5-hydroxytryptamine uptake and release) and behavioral (spatial learning) effects of neurotoxin-induced denervation and reinnervation by grafting fetal mesencephalic raphe cells. Intracerebroventricular injections of a low dose of 5,7-dihydroxytryptamine caused a discrete serotonergic denervation of the hippocampus. Eleven months after lesioning, 5-hydroxytryptamine and 5-hydroxyindolacetic acid levels and [3H]5-hydroxytryptamine uptake capacity were decreased by 50--60°/°. By this time, the residual fibers displayed an enhanced vulnerability towards K ÷-induced depolarization. Grafting of a fetal raphe cell suspension resulted in a reinnervation of the host hippocampus. The pattern of reinnervation was comparable to control innervation and the density was supranorrnal at the level of the graft. As observed semiquantitatively, the innervation density decreased with distance from the core of the graft. Neurochemical studies showed that the fibers were capable of synthesizing, metabolizing and releasing 5-hydroxytryptamine. The turnover of 5-hydroxytryptamine in both the denervated and the reinnervated hippocampus was comparable to that in control tissue. Previous behavioral testing of the denervated and of the denervated and implanted animals did not reveal any effect on spatial learning, either in an individual or in a social test paradigm. The latter data substantiate the notion that interference with the hippocampal serotonergic innervation does not hamper adequate spatial learning.
Long after developmental changes in the CNS are completed, serotonergic neurons still display a certain capacity for plasticity)'~'47'48 Thus, compensatory processes upon chemical lesioning of the adult rat brain have been described, such as sprouting of serotonergic fibers after partial lesioning by 5,7-dihydroxytryptamine (5,7-DHT) H,4s and increases in the turnover of 5-hydroxytryptamine (5-HT) in the denervated brain areas) 3 These compensation mechanisms are attenuated in the aged rat C N S ) 2,34,35It is thought that, amongst others, physical and humoral factors in the target areas of the serotonergic fibers in the mature brain are involved in the control of the neuronal plasticity. Hence, physiological and pathological conditions of the target tissue will influence the plasticity. The exact regulatory mechanisms are, however, far from understood. Denervation and subsequent implantion of fetal brain tissue provides an excellent tool to study the regulation of neuronal plasticity in adult and aged *To whom correspondence should be addressed. 5,7-DHT, 5,7-dihydroxytryptamine; 5HIAA, 5-hydroxyindolacetic acid; HPLC, high pressure liquid chromatography; 5-HT, 5-hydroxytryptamine; MR, mesencephalic raphe.
Abbreviations:
host brain. Thus, morphological studies have shown that fetal mesencephalic raphe (MR) cells survive for long periods and establish extensive networks upon grafting into a host b r a i n y Neurochemical studies have indicated changes in 5-HT and 5-hydroxyindolacetic acid (5-HIAA) levels following grafting into the brain. 4,2t Moreover, implants in the hippocampus display electrophysiological activity. 36,43 Behavioral studies reported the sequelae of grafting fetal cells into various brain areas) °,2°,25,44 In these studies analysis was made of either morphological features of grafted neurons, functional activity of the growing fibers or their possible involvement in the regulation of behavior. Our experiments aimed to study, within the same animal, the relationship between morphological (innervation pattern and density), neurocbemical (5-HT and 5-HIAA levels and [3H]5-HT uptake and release) and behavioral (spatial learning) parameters. The hippocampus was chosen as the target area for several reasons. Firstly, it has a high level of morphological organization, which allows accurate histological examination. Secondly, the hippocampus is an easily accessible structure and contamination in neurochemical analyses by non-hippocampal tissue can thus be readily avoided. Thirdly, the involvement
365
366
M. G. P. A. VAN LUIJTELAARet al.
o f the h i p p o c a m p u s in spatial learning has been well established 3°'31 a n d behavioral tests are available. 22 T h e serotonergic system was chosen as a selective depletion o f the serotonergic system can be achieved by intracerebroventricular 5,7-DHT injections. Moreover, o u t g r o w t h o f fetal M R cells implanted in the h i p p o c a m p u s will result in a dense i n n e r v a t i o n of the host tissue. Finally, the serotonergic h i p p o c a m p a l system has been implicated in m n e m o n i c functions. 2 EXPERIMENTAL PROCEDURES
Thirty-six male Wistar rats (Harlan CPB, Zeist, The Netherlands, 140-160g body weight) were randomly divided into four equal groups: control (C), sham-lesioned (ShL), lesioned (L) and lesioned + implanted (T). Behavioral testing started five months after surgery. Eleven months after surgery, the animals were killed for morphological and neurochemical analysis.
Surgery Under xylazine (25 mg/kg, s.c.) and ketamine (50 mg/kg, i.m.) anesthesia, 10/~g 5,7-DHT dissolved in 5/~1 saline with 0.02% ascorbic acid was injected into both lateral ventricles (0.8 mm posterior, 1.5 mm lateral to bregma, 3.2 mm below the surface of the overlying cortex). Animals were pretreated with desimipramine (25mg/kg, i.p.). Sham-lesion consisted of two 5-#1 vehicle injections. Fourteen days after the 5,7-DHT injections the T group was anesthetized as described above and received injections of 2/~1 from a cell suspension obtained from the MR region of El5 fetuses. 37 Approximately 100,000 viable cells, counted after incubation with Trypan Blue, were implanted bilaterally into the hippocampus (3.0 mm posterior, 1.8 mm lateral to bregma, 3.0mm below the surface of the overlying cortex). From each experimental group one animal was killed three months after surgery to serve as control for the immersion fixation that was to be used in the final experiments. Brains were fixed by perfusion fixation and processed as has been described elsewhere.45
Histochemical and neurochemical analysis Eleven months after surgery the animals were killed for combined neurochemical and histochemical analysis. Each experiment included one animal from each group. Upon decapitation, the hippocampal lobes were dissected on ice, cut into 300-/~m slices with a Mcllwain tissue chopper and divided into five levels according to the scheme in Fig. 1. From each level the first slice was processed for serotonin immunocytochemistry, the next two slices were processed for uptake and release studies and the remaining two slices (per level) were immediately frozen on dry ice, stored at - 7 0 ° C until HPLC analysis followed (details are given below). From C and ShL groups only the left hippocampal lobes were used; from the L and T groups both hippocampal lobes were used.
lmmunocytochemistry Immunocytochemistry was performed on the first three levels of C and ShL groups and on all five levels of the L and T groups. After 2 h of immersion fixation in 4% paraformaldehyde, 0.05% glutaraldehyde, 0.2% picric acid and 5% sucrose in 0. IM phosphate buffer (pH 7.4, 4°C), the slices were rinsed for 30m in in 5% sucrose in 0.1 M phosphate buffer. They were embedded in Tissue Tek OCT simultaneously, frozen with powdered dry ice and cut on a cryostat into 10-/tm sections at -18°C. Sections were collected on gelatin-chrome-alum coated slides and stored at - 2 0 ° C until further processing. The sections were pretreated for 30 min at room temperature with 0.3% peroxide
ICC
Re lease
H PI.C
Fig. l. Scheme showing the dissection of the hippocampal lobes. Five adjacent levels (1-5) with a dorsal-ventral orientation were each chopped into five 300-/t m s/ices. These slices were then used for immunocytoehemistry (ICC), uptake and release, and HPLC studies. in methanol to inhibit endogenous peroxidase activity. Immunocytochemical staining for 5-HT immunoreactivity was performed according to the peroxidase-antiperoxidase method a9 with a rabbit serotonin antiserum. 41 For details of the staining procedure see Ref. 46.
Uptake and release studies Slices from the first three levels were used for uptake and release studies. The two slices (per level) were incubated with radiolabeled serotonin and choline and were superfused as described previously. ~ In short, after a preincubation with Krebs-Ringer medium, the sets of slices were incubated during 30 min in medium containing 1/~Ci [3H]5-HT and 0.5/tCi [~4C]choline (final concentrations +__O.1 and 10/~M). After 45 min of superfusion, five successive 10 rain fractions were collected. Radioactivity remaining in the slices was extracted with 3 ml 0.1 N HCI. The total radioactivity in the superfusion fractions and extracts (Et) was taken as a measure for the uptake. Depolarization-induced calciumdependent transmitter release in excess of spontaneous efl~ux (effected by 10 min superfusion with medium containing 20 mM K +) was calculated as the percentage of the total radioactivity present in the tissue at the onset of the stimulation period. Intervial variance was accounted for by adjusting the [3H]5-HT content values to those of the II4C]acetylcholine which were previously shown not to be affected by 5,7-DHT lesion) 8
High pressure liquid chromatography analysis HPLC analysis was performed on the slices from levels 1 and 3. The slices (per level) were pooled and homogenized in 150/zl of cold 0.1 N HCI04, containing 12 ng of DHBA as internal standard. Upon centrifugation for 3 min at 8750 × g the pH of 100-~1 aliquots of these solutions was adjusted to 4 with 2 M NaAc, and 50-/~1 aliquots were subjected to HPLC for the separation of 5-HT and 5-HIAA. Chromatography was performed as described previously. ~a Protein content of the pellets was assessed according to Lowry's method. 19
Serotonergic denervation and graft-induced reinnervation of rat hippocampus
Statistical analysis of neurochemical data Data from the distinct levels within C, ShL and L rats, and for the latter from both hemispheres, were averaged, yielding one value per animal. Data from the distinct levels from both hemispheres within the T groups were treated as single values. All values were then converted to percentages of the corresponding control values and analysed by a one-way ANOVA, followed by a Duncan multiple range test. Data are presented as the mean _ S.E.M. Behavioral tests Spatial learning was tested in a social context and in an individual setting. Five months after surgery, C, L and T groups (n = 8) were submitted to the first test. The task, performed at one-day intervals in a 2 × 3 m cage, was reinforced both by food reward and social avoidance. At the end of the light period (9.30 a.m.) food was removed from the home cages. Three hours later the subject was placed in the center of a compartment (2 x 0.50 m) and allowed to make a choice between two identical boxes (0.15 × 0.30 m) located at opposite sides of the cage. Several visuospatial cues were present. When a box was entered, the entrance door was closed and the access door to the corresponding goal compartment behind it (2 × 2.5m) was opened. Correct responding was rewarded by 15 min of access to a food platform. Choice of the incorrect box was punished by exposure to one attack by a dominant colony male. After reaching criterion (three correct trials on successive days), the goal compartments were interchanged and two reinforced reversal training trials were given followed by nine test trials. Choice latency, number of trials to criterion, percentage of correct choices and the longest string of correct choices were taken as parameters of orientation behavior. Eight to nine months after surgery, spatial learning was tested in an individual context in a Morris water maze. 22 Latency to escape on the hidden platform was recorded during the training trials (four on day 1, eight trials on days 2 and 3, and four trials on day 4). At the end of the last training trial a 60 s transfer test was performed in which (i) the total number crossing the exact location previously occupied by the platform, and (ii) the total time spent searching within the training quadrant were taken as a measure of spatial bias. Two platform spatial discrimination training was conducted in a different experimental room. In this pool a rigid platform was placed (SW or NE) and a floating platform that was changed in position from trial to trial was used. Rats were given 10 trials on the first day of training and 20 trials on the subsequent days. Training sessions were separated by at least 48 h. Choice latency, number of correct choices and the longest string of correct choices were measured. RESULTS
Immunocytochemistry The preliminary processing of four animals to verify the extent of the lesion and to obtain a control for the adapted immunocytochemical procedure showed that the serotonergic innervation of the ShL hippocampus equals that of control animals (Fig. 2A, B). 5,7-DHT injection yielded hippocampal lobes that were almost devoid of serotonergic fibers (Fig. 2C). In the implanted dorsal hippocampus a dense reinnervation emerging from the grafted M R cells was observed throughout the hippocampus resembling the pattern of control innervation (Fig. 2D). The grafts were located bilaterally in the dorsal hippocampus (Fig. 2E). After immersion
367
fixation of the 300-#m slices as performed in the final experiments, staining comparable to that after perfusion fixation was achieved. Fibers were deafly and distinctly stained (Figs 2D and 3B). Some leakage of serotonin occurred from the cells, which was observed as a less sharp delineation of the borders of the serotonergic cells. Further presentation of the results is solely based on the material obtained from the combined experiments. It appeared impossible to discern the injection track macroscopically and only after immunohistochemical analysis were we able to locate the core of the implant. Eight hippocampal lobes showed neither surviving serotonergic cells nor their major outgrowth in the first level which would indicate the core of the implant. They were eliminated from statistical calculations. Two implanted rat groups that did not show any serotonergic fibers in their hippocampal lobes and one lesioned rat that appeared not denervated (normal fiber density) were eliminated from further analysis. Sham-lesioning did not affect the fiber distribution and density as compared to the control innervation pattern and density (Figs 3A and 4A). However, 5,7-DHT treatment resulted in a severely diminished fiber density throughout the hippocampal lobes (Figs 3A and 4A). This decreased fiber density was not only observed in the dorsal hippocampus, i.e. in the direct vicinity of the injection site, but also in the ventral hippocampus. Only in the most ventral level of the hippocampus were some serotonergic fibers encountered. U p o n implantation the fetal cells reinnervated the surrounding host tissue. Within the implant the fiber density appeared to be extremely high. Both the dentate gyrus (Fig. 3B) and the CA regions (Fig. 4B) received a dense innervation. The pattern of the fiber distribution in the host hippocampus, however, was comparable to the pattern in control animals. Thus, the pyramidal and granular cell layers in the A m m o n ' s horn and the dentate gyrus were only sparsely innervated, whereas a dense innervation of the other layers was found. Following the outgrowth over distances of 1500, 3000, 4500 and 6000 # m from the site of implantation (levels T2-T5; Figs 3B and 4B), the fiber density gradually decreased. However, even in the most distal part of the hippocampus (level T5; Figs 3B and 4B) the fiber density exceeded the density of the non-implanted lesioned hippocampus.
Uptake and release experiments There was no difference measured between the mean tissue content of radiolabeled serotonin in slices from the control or the sham-lesioned group ( 1 0 0 _ 11% versus 99-1-8%; Fig. 5A). Slices from the lesioned group showed a statistically significant reduction in [3H]5-HT uptake capacity to 42 + 4% of control values. Slices from the most rostral level of the implanted hippocampal lobes (T 1), containing the
368
M. G. P. A. VAN LUIJTELAARet al. ,i
-%
Fig. 2. Serotonin immunoreactivity in the hippocampus (35-#m sections) three months after surgery. (A) Control innervation of the hilus. (B) Sham-lesioned hippocampus. (C) Lesioned hippocampus. Note the absence of serotonin immunoreactivity. (D) Implanted hippocampus. The very dense innervation of the molecular layer of the dentate gyrus is clearly visible. (E) An overview of a bilaterally grafted animal. The border of the grafts is indicated with arrowheads. Scale bars = 150/~m (A-D); 750 #m (E). grafted cells, showed a five-fold increase in uptake in comparison to slices from the lesioned animals (194 + 27% versus 42 ___4%; Fig. 5A). This rise in
uptake capacity decreased with the distance from the implanted cells. Uptake at level T2 (130 + 13%) and level T3 (110 ___20%) was significantly higher than
Serotonergic denervation and graft-induced reinnervation of rat hippocampus uptake in slices from lesioned animals but did not exceed the uptake capacity of the control animals (Fig. 5A). In all groups (C, ShL, L and T), K +-stimulation yielded a release of radiolabeled serotonin in surplus of the spontaneous effiux during standard superfusion conditions (Fig. 5B). Sham lesioning had no effect on the K+-stimulated [3H]5-HT release when compared to release in controls. Lesioning resulted in an increase of the mean K+-stimulated [3H]5-HT release to 206% of the control value. In the transplanted group the mean K +-stimulated release from levels T1, T2 and T3 was respectively 153_ 14%, 161 _+ 12% and 110_+ 11% of the unoperated control group.
High pressure liquid chromatography analysis The sham-lesioned rats showed no altered 5-HT levels when compared to control (96 _+ 6% versus 100_+ 18%; Fig. 6A). Injection of 2 0 # g 5,7-DHT into the lateral ventricles caused a reduction of the 5-HT levels to 50 _+ 6% of control. No differences were found between the left and right hippocampal lobes. The levels of the fight hippocampal lobes were used for further calculations. Implantation of fetal mesencephalic cells enhanced 5-HT levels to supranormal values (170+ 33%). When the overall measurements were divided into rostral (T1) and middle (T3) levels, the effect of the outgrowth of the fetal M R cells became evident. At the most rostral level (TI), 5-HT levels were doubled as compared to control values (199 _+ 34%) and four times as high as measured in the lesioned hippocampi (Fig. 6A). The 5-HT levels at the more distal level (T3; 93_+ 15%) were equal to those in the control animals (100 _+ 18%), but still twice as high as in the lesioned animals (50 _+ 6%; Fig. 6A). The same pattern of changes was observed for the 5-HIAA levels (Fig. 6B). Sham lesioning did not affect 5-HIAA levels (93_+5%), while treatment with 5,7-DHT reduced 5-HIAA levels to 47 _+ 9% of control values. The overall level of 5-HIAA in the transplanted group was 120 + 13% of control. From the implanted hippocampal lobes, level T1 showed the major increase in 5-HIAA content (150 + 16%), while slices from level T3 (85 -+ 10%) did not differ from control. When compared to the lesioned hippocampi, 5-HIAA levels in level T1 showed a three-fold increase and those in level T3 a two-fold increase (Fig. 6B). Ratios between 5-HIAA and 5-HT did not differ between the treatment conditions. The ratios in sham-lesioned, lesioned and lesioned + implanted lobes (97 _+ 5%, 109 _+ 24%, 85 + 4%) were comparable to the ratio in the control group (100 + 14%; Fig. 6C). 5-HT levels and [3H]5-HT accumulation in individual levels from control, sham-lesioned, lesioned and lesioned + implanted animals showed a correlation of r = 0.78 (P < 0.01, data not shown).
369
Behavioral tests During the orientation task in a social context, correct responding was always accompanied by eating. Consequently, admittance to the food platform may be considered as a salient reward. During the acquisition phase the number of trials to reach criteflon was equal in all groups. During the reversal testing the control males made their choice 282 s after introduction (logarithmic mean; 95% confidence limits: 108-733s); this choice was correct in 61 _+ 9.9% of the trials. The criterion was reached at trial 4.8 + 0.9 and the longest string of correct choices amounted to 4.4_+0.9 trials. Two-way A N O V A showed that treatment did not influence the individual choice latencies [F(2,18)=2.32, N.S.] whereas repeated testing did [F(8,144) = 2.89, P < 0.005]. Individual parameters of correct orientation (trials to criterion, total number of correct responses and the longest string of correct choices) were not affected by the treatments (one-way ANOVA; dr2,18; F = 0.44, F = 0.07 and F = 0.014; N.S.). Through the reversal trials no two treatment groups differed significantly with respect to the percentage of rats which responded correctly, suggesting a similar development of orientation reversal in all three groups tested. In the Morris water maze most rats adopted the normal adult swimming posture and showed no evidence of motoric difficulty. Testing of one rat was abandoned when it repeatedly jumped out of the pool during training. During acquisition, the escape latency decreased [F(23,575)= 4.39, P < 0.001] and although the groups x trial interaction reached significance [F(69,575) = 1.32, P < 0.05], post hoc tests revealed no consistent difference between the groups throughout the training. Analysis of the transfer test showed that after training the groups did not differ in both parameter measures (control values: 2.1 _+ 0.8 crossings and 26.4 + 2.8 s spent swimming in the training quadrant). During training on the twoplatform spatial discrimination task, lesioned and transplanted animals tended to perform worse than controls, but analysis failed to confirm any effect of treatments on either parameters (control values: 36.7 + 3.7 correct choices and 11.7 _+ 2.2 consecutive correct choices). DISCUSSION
The experimental design enables the simultaneous description of both morphological parameters, such as serotonergic fiber pattern and density, and neurochemical estimates, such as 5-HT and 5-HIAA levels and [3H]5-HT uptake and release, in hippocampal tissue from individual rats. Fresh, unfixed tissue is a prerequisite for neurochemical analyses such as HPLC and [3H]5-HT uptake and release experiments. However, a consequence of the use of fresh tissue is a reduced binding of serotonin to tissue components. Therefore, a reduced histological resolution might be expected. Nevertheless, no differences were observed
370
M . G . P . A . VAN LUIJTELAARel al.
!if
~_
IL_
Fig. 3(A).
Fig, 3(B). Fig. 3. Serotonin immunoreactivity in the dentate gyrus of the hippoeampal lobes as observed with dark-field microscopy. (A) Control (C), sham-lesioned (ShL), lesioned (L) and (B) lesioned + transplanted (T) rats. Sham lesioning did not affect the fiber density or distribution while 5,7-DHT treatment resulted in a severely diminished fiber density. The TI level contains the implant (borders are indicated with arrowheads). The levels are 1500 y m away from each other. Note the supranormal fiber density at level T1 and the decline in density with increasing distance to the implant. The fiber density at level T5 still exceeds the density of the lesioned hippocampal lobes• g, granular layer• Scale bar = 100/am. 371
372
M . G . P . A . VAN LUIJTELAARet al.
Fig. 4(A).
Fig. 4(B). Fig. 4. Serotonin immunoreactivity in the CA2-CA3 region of the hippocampal lobes as observed with dark-field microscopy. (A) Control (C), sham-lesioned (ShL), lesioned (L) and (B) lesioned + transplanted (T) rats. Sham lesioning did not affect the fiber density or distribution while 5,7-DHT treatment resulted in a severely diminished fiber density. Upon implantation the region is strongly reinnervated. Note the supranormal fiber density at level T1 and the decline in density with the increasing distance to the implant. The fiber density at T5 still exceeds the density of the lesioned hippocampal lobes, p, pyramidal cell layer. Scale bar = 100/~m. 373
374 o t_ c o o
M. G. P. A. VAN LUIJTELAARet
al.
250
250
A
A o L
200
200
o
150 0 -'a
I-" T I -r m
T
150
z~
.$
100' •litv
50 0
5
5
5
6
6
6
C
ShL
L
T1
T2
T3
0 --
.
o t_ to o
200
E
150
200
o
150-
=
100-
7
7
7
C
ShL
L
8 T1
T3
250
B
c o
.4
50' zl3
250 o ,-
100'
T
100
u
e t® -
50
50'
I I O
8 C
ShL
L
T1
T2
0
T3
Fig. 5. Tissue content (A: Et values) and K+-stimulated release (B: %P) of [3H]5-HT in slices from sham-lesioned (ShL), lesioned (L) and lesioned + transplated (T) hippocampal lobes compared to control values (100 + 11%). (A) Mean uptake in the slices from the first three levels of sham-lesioned and lesioned rats and the uptake in the slices from the levels T1, T2 and T3 of implanted rats is expressed as percentage of the mean uptake in slices from levels 1-3 in control rats. (B) K÷-stimulated [3H]5-HT release from the same slices as mentioned under A. The numbers in the bars represent the number of observations. ( ~ ) P < 0.05 when compared to control; (A) P < 0.05 and (A) P < 0.01 when compared to lesioned hippocampal lobes.
~ "E o v .o ,.° ~_ iI < t
to
ShL
L
TI
T3
250 200150100 T
50 7
-r-
between staining of serotonin-immunoreactive fibers upon immersion fixation and perfusion fixation of the hippocampus. 5,7-DHT treatment of the animals resulted in a marked decrease in serotonergic innervation density as observed histologically. U p o n partial denervation, sprouting of serotonergic fibers has been reported. H.48 In the present study, no indication of such a regeneration was found, as even 11 months after depletion, the hippocampus is still devoid of serotonergic fibers. The decrease in fiber density was paralleled by changes in 5-HT and 5-HIAA levels and in [3H]5-HT uptake capacity. Similar neurochemical observations have been reported after neonatal or adult administration of 5,6- or 5,7-DHT. 9'27'32'42 Although both the histological and neurochemical results point to a marked denervation, a discrepancy was apparent between the absolute values obtained by both approaches. Lesioned hippocampi stained for serotonin immunoreactivity showed hardly any remaining fibers, whereas neurochemical data from both H P L C analysis and uptake experiments indicated a reduction to only 45-50% of control values. At the concentration of [3H]5-HT used in this study,
C
0
C
8 ShL
L
T1
T3
Fig. 6. 5-HT levels (A), 5-HIAA levels (B) and 5-HIAA/5HT ratios (C) in slices from sham-lesioned (ShL), lesioned (L) and lesioned + transplanted (T) hippocampal lobes compared to control values (100 + 18%) as determined by HPLC analysis. 5-HT and 5-HIAA levels were determined in the first and the third levels from control, sham-lesioned (ShL) and lesioned (L) rats and in the levels T1 and T3 from transplanted rats. For ShL and L rats the mean values from the two levels are presented. The numbers in the bars represent the number of observations. (~r) P < 0.05 and (~') P < 0.01 when compared to control; (A) P < 0.05 and (/k) P
