Serotonin neurons grafted to the adult rat hippocampus. II. 5-HT release as studied by intracerebral microdialysis

Brain Research, 498 (1989) 323-332 Elsevier

323

BRES 14833

Serotonin neurons grafted to the adult rat hippocampus. II. 5-HT release as studied by intracerebral microdialysis A. Daszuta 2, P. Kal6n 1, R.E. Strecker 1"*, P. B r u n d i n ~ a n d A . B j 6 r k l u n d 1 1Department of Medical Cell Research, Section of Neurobiology, University of Lund, Lund (Sweden) and 2UnitO de Neurochimie, LNF4, C.N.R.S., Marseille (France) (Accepted 28 February 1989)

Key words: Neural grafting; Serotonin release; Intracerebral microdialysis; Hippocampus; Rat

Extracellular levels of serotonin (5-HT) and 5-hydroxyindoleacetic acid (5-HIAA) were monitored by microdialysis in the hippocampal formation previously denervated of its serotonergic input by an intraventricular injection of 5,7-dihydroxytryptamine (5,7-DHT), and in 5,7-DHT denervated hippocampi reinnervated by grafted fetal rat serotonin neurons. Two weeks after 5,7-DHT lesion, baseline 5-HT release was reduced to levels below detection, and KCI- and p-chloro-amphetamine-evoked release was reduced by 90-95%. In the chronically denervated hippocampus (3 months after lesion), baseline 5-HT release had recovered to near-normal levels, but KC1-and p-chloroamphetamine-evoked release remained severely impaired. Addition of the 5-HT re-uptake blocker indalpine to the perfusion medium induced a 5-6-fold increase in serotonin overflow in the normal hippocampus, while the serotonin overflow in the 5,7-DHT denervated hippocampus remained unaffected. The intrahippocampal fetal raphe transplants restored 5-HT release to near-normal levels, not only under baseline conditions but also in the presence of re-uptake blockade. Both KCI- and p-chloroamphetamine-induced release had recovered in the grafted hippocampus and the responses were even greater than those seen in normal animals. In both normal and grafted hippocampus addition of the sodium channel blocker tetrodotoxin reduced 5-HT overflow to the level seen in the denervated hippocampus. The new hippocampal serotonin innervation, established by the grafts, was markedly denser than normal, and the tissue 5-HT and 5-HIAA levels were 3-4-fold higher than normal in the grafted hippocampi. The 5-HIAA level in the perfusate collected from the grafted hippocampi showed a similar increase above normal, whereas 5-HT release was maintained within the normal range, both under baseline conditions and in the presence of re-uptake blockade. The results indicate that the grafted serotonergic raphe neurons are spontaneously active at the synaptic level, despite their ectopic location. The ability of the grafted neurons to maintain 5-HT release within the normal range suggests that local regulatory mechanisms at the terminal level can compensate for abnormalities in the graft-derived innervation density. INTRODUCTION

nections are functional24, and there are data to suggest that intracerebral or intraspinal raphe trans-

Previous studies have demonstrated that serotonergic neurons, obtained from the brainstem raphe region of rat fetuses, can survive grafting to the brain or spinal cord of adult recipients 5"8"1°-12"15"19'2°'23. The implanted n e u r o n s grow to form extensive serotonergic terminal networks in large areas of the surrounding host brain, and at least in the hippocampus and spinal cord the newly formed graft efferents have been shown to establish ultrastructurally normal synaptic contacts with elements in the host 4"6"22. Electrophysiological analyses, in slices in vitro, indicate that the serotonergic graft-host con-

plants can correct for lesion-induced behavioral deficits 19,22,26. In our previous study ~° we observed that fetal raphe transplants, implanted in the form of a cell suspension into the 5,7-dihydroxytryptamine (5, 7-DHT) denervated hippocampus, can establish a new serotonin (5-HT)-containing terminal network throughout the host hippocampal f o r m a t i o n , although the terminal density in several cases was markedly higher than normal. Since the grafted serotonergic n e u r o n s are electrophysiologically active and appear to possess at least some of their

* Present address: Hana Biologics Inc., Alameda, CA 94501, U.S.A. Correspondence: A. BjOrklund, Department of Medical Cell Research, Biskopsgatan 5, S-223 62 Lund, Sweden. 0006-8993/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

324 normal physiological properties 24'25"z9 it seems possible that the intrahippocampal raphe transplants could be able to modify or influence functional activity in the area of the host brain reached by the ingrowing efferent fibers. The present study was designed to investigate this possibility further using the intracerebral microdialysis technique to monitor ongoing serotonin release within the hippocampal formation. The changes in extracellular 5-HT and 5-hydroxyindoleacetic acid (5-HIAA) levels under baseline conditions and after pharmacological manipulations were studied after 5,7-DHT induced denervation and after reinnervation by fetal raphe cell suspension grafts. MATERIALS AND METHODS

Lesion and transplantation surgery Fourteen young adult female Sprague-Dawley rats (180-200 g) received an intraventricular injection of 5,7-dihydroxytryptamine (Sigma; 150 ~g 5,7-DHT free base dissolved in 20 ~1 ascorbate saline solution). They were pretreated with desmethylimipramine (DMI) (Ciba-Geigy; 25 mg/kg i.p. 30-45 min before the 5,7-DHT injection) in order to protect the noradrenaline system 7. Two weeks after the 5,7-DHT injection, 11 of the lesioned rats received unilateral grafts of mesencephalic raphe tissue from 14-day-old rat fetuses 1°. Three 1.5-/~1 aliquots of cell suspension were injected into the hippocampus at the following coordinates (tooth bar set at -3.3 mm): A, -6.0 mm from bregma; L, 5.0 mm; V, -6.0 mm, -4.5 mm and -3.0 mm from dura. The contralateral hippocampus remained as nongrafted lesioned control side. The 3 remaining 5,7-DHT lesioned rats did not receive transplants, and served as an acute lesion control group. Microdialysis At 3 months after transplantation, the 11 lesioned/ grafted rats and 13 age-matched normal control rats were taken for in vivo measurements of extracellular 5-HT and 5-HIAA using the intracerebral microdialysis technique 3°. The rats were separated into two groups: 6 lesioned/grafted rats and 5 normal rats were used to test the effects of KC! and p-chloroamphetamine (PCA) on 5-HT and 5-HIAA release, and 5 lesioned/grafted rats and 8 normal rats were

Fig. 1. Schematic illustration of the design of the dialysis probe and its placement in the hippocampal formation. The stars mark the approximate localization of the implanted fetal raphe transplants.

used to test the effects of 5-HT re-uptake blockade by indalpine. In the latter group the effects of KC1 and tetrodotoxin (TTX), added to the perfusion fluid, were studied in the presence of indalpine. The 3 non-grafted 5,7-DHT-lesioned rats were taken for dialysis at 2 weeks after treatment and served as an acute lesion group. Dialysis probes (Fig. 1) were made of 4-mm-iong loops of dialysis tubing (0.3 mm outer diameter) glued inside 0.4-mm stainless-steel cannulae. The probes were implanted stereotaxically, under chloral hydrate anesthesia, on the day preceding the dialysis experiment. They were positioned in the mid-ventral part of the hippocampal formation, as illustrated in Fig. 1 (A, -6.0; L, 5.0 mm; V, -6.5 mm). The probes were implanted bilaterally in the 5,7-DHT lesioned/ grafted animals and unilaterally in the normal control rats and in the acute 5,7-DHT lesioned rats. During the dialysis experiment the rats were kept anesthetized with halothane and their body temperature was maintained at +37 °C using incandescent light and a heating pad. The loop inlet cannula was connected to a microinfusion pump (Carnegie Medicin AB, Stockholm, Sweden) and continually

325 perfused at 2/~l/min with Ringer solution. Hippocampal perfusates were collected every 60 or 30 min in plastic Eppendorf microtubes containing 1.6 x 10-4 M L-cysteine and 1.5 pmol a-methyl serotonin as antioxidant and internal standard, respectively. The first 30 min perfusate was discarded as it typically contained excess amounts of 5-HT. The samples were stored on ice for up till 24 h before biochemical analysis. Probe placement was checked in Cresyl violet-stained frozen sections, and in some cases in sections processed for 5-HT immunohistochemistry l°. This histological control was performed in all rats except the 5 lesioned/grafted rats used for biochemical analysis of hippocampal 5-HT and 5H I A A content. Biochemical analysis

The concentrations of 5-HT and 5-HIAA in the dialysis perfusates were assayed using high-performance liquid chromatography (HPLC) with fluorimetric detection 16. Indoles were separated on a supelcosil LC 8 DB 3 /~m column protected by a supelguard LC 18 column using 26 mM phosphoric acid, 65 /~M sodium octanyl sulfonic acid, 1% acetonitrile and 3% tetrahydrofuran as mobile phase (pH 2.6; 1.2 ml/min flow rate). Tissue 5-HT and 5 - H I A A levels were assayed in 0.4 N perchloric acid extracts of the grafted and the contralateral lesioned hippocampus from 5 of the grafted rats, and from the hippocampus bilaterally in 5 normal control rats (Table Ill). Drugs

The Ringer solution contained 147 mM NaCI, 4 mM KC1, 2.3 mM CaCI 2, pH 6.0. In the experiment with Ca2+-free buffer, CaC12 was replaced by NaC1 and 0.1 mM EGTA was added. PCA (Sigma Chemical Co., St. Louis, U.S.A.) was administered at a dose of 2.5 mg/kg, i.p., dissolved in saline. For K+-induced depolarization, the KCI concentration of the Ringer solution was increased to 60 or 100 mM, while the NaCI concentration was reduced to 91 or 51 mM. The 5-HT re-uptake blocker indalpine (Pharmuka, Paris, France) was dissolved first in methanol (10 -2 M), and then in Ringer to a final concentration of 10-6 M. TFX, (10-6 M; Sigma), a selective blocker of sodium channels, was dissolved in Ringer containing 10-6 M indalpine.

RESULTS Normal rats

Baseline 5-HT release averaged 0.014 pmol/30 min (Tables I and II). Addition of the selective 5-HT re-uptake blocker indalpine at 10-6 M to the perfusion fluid caused a 5-6-fold increase in 5-HT overflow, which was stable over time (Table I, Fig. 3A). Addition of depolarizing concentrations of KCI (60 or 100 mM) induced a dramatic increase in 5-HT overflow, both under baseline conditions (Table II, Fig. 2A) and in the presence of indalpine (Table I, Fig. 3A). This effect was substantially reduced when Ca 2+ was removed from the perfusion medium (Table I). Injection of the 5-HT-releasing drug PCA (2.5 mg/kg, i.p.) caused an immediate approximately 30-fold increase in 5-HT overflow which gradually declined over the subsequent hours (Fig. 2A). Addition of the sodium channel blocker, TTX, at 10-6 M caused a marked, approximately 75% inhibition of 5-HT overflow, as measured in the presence of indalpine (Fig. 3A). Baseline 5 - H I A A levels in the perfusate averaged 11-12 pmol/30 min (Tables I and II). Addition of indalpine (also in the presence of high K +concen-

TABLE I Release of 5-HT and 5-HIAA in normal hippocampus (n = 5-6) under baseline conditions, after addition of the 5-HT re-uptake blocker indalpine to the perfusion medium, and after addition of depolarizing concentrations of KCI, with and without Ca2÷ present in the perfusion medium

Values are means +_S.E.M.

Baseline releasea Indalpine (10 6M)b Indalpine (10-6M)c + KC160mM Indalpine (10-6M)c + KC160mM in Ca2÷ free buffer

5-HT (pmol/6Otd)

5-HIAA (pmol/60 gl)

0.015+0.004 0.072+0.017" 0.547+0.071"*

10.98+0.91 9.57+ 1.60 9.15+0.84

0.198+0.015"*'*** 6.72+0.72

a Mean of three 1-h (120pl) samples. b Mean of three 30-min (60pl) samples with indalpine. c Samplecollected during 30 min (60/fl). * Different from baseline value at P < 0.01. ** Different from indalpine alone at P < 0.01. *** Different from KCI in the presence of Ca2+ at P < 0.01 (Student's two-tailed t-test).

326 TABLE 1I Effect of KCI (added to the perfusion fluid at 100 mM) or p-chloroamphetamine (PCA) (injected i.p. at a dose of 2.5 mg/kg) on hippocampal 5-HT and 5-HIAA release in normal rats, in 5, 7-DHT lesioned rats 2 weeks after injection (acute lesion) or 3 months after injection (chronic lesion, contralateral hippocampus in the grafted animals), and in the lesioned and grafted hippocampus

Values give means + S.E.M. (pmol/60/~l perfusate); ND, value below detection (< 0.005 pmol/60/~l perfusate). Group

Normal (n = 5)

Acute 5, 7-DHTlesion (n = 3)

Chronic 5, 7-DHT lesion (n = 6)

5, 7-DHT + graft (n = 6)

5-HT Baseline releasea KCl-evoked releaseb PCA-evoked releasec

0.014 _+0.052 0.529 _+0.049*'** 0.347 + 0.054*'**

ND 0.037 _+0.015 0.016 + 0.007

0.016 + 0.005 0.100 + 0.009** 0.035 _+0.009**

0.016 _+0.005 0.936 + 0.166"** 0.525 + 0.098 . . . .

5-HIAA Baseline releasea KCl-evoked releaseb PCA-evoked releasec

11.85 + 2.05* 6.20 + 1.05"** 9.60 + 1.51"**

0.23 ± 0.081 0 0.18 + 0.055

0.80 + 0.16 0.53 + 0.15"* 1.50 + 0.38**

51.32 ___12.9" 21.05 + 4.59 . . . . 31.13 + 7.9*'**

Mean of four 1-h (120ktl) samples. b One 30-min (60/~1) sample. c Mean of six 30-min samples taken after PCA injection. * Difference from chronic 5,7-DHT-lesioned hippocampus at P < 0.01. ** Difference from baseline value at P < 0.05 or 0.01. Student's (two-tailed) paired t-test.

trations) did not cause any significant change (Table I). A d d i t i o n of 100 m M KC1, in the absence of indalpine, was accompanied by a significant decrease in the 5 - H I A A level, and a similar effect was seen after P C A injection (Table II, Fig. 2B). TI'X, which caused a substantial decrease in 5-HT overflow, had no effect on the extracellular 5 - H I A A level (Fig. 3B). Effect o f 5, 7-DHT-induced denervation

In the acutely 5,7-DHT-lesioned animals, where the dialysis was p e r f o r m e d 2 weeks after 5 , 7 - D H T injection, the 5-HT levels in the baseline perfusates were below the limit of detection (i.e. <0.005 pmol/60 ~!), which represents less than a third of the normal level, and the 5 - H I A A level was less than 2% of normal (Table II). In the chronically denervated h i p p o c a m p u s (i.e. non-grafted contralateral side in the grafted animals), 3 months after 5 , 7 - D H T injection, baseline 5-HT overflow was similar to normal, whereas the 5 - H I A A levels were reduced by about 93% (Table II, Fig. 2). A s discussed further below, this indicates a progressive compensatory change in 5-HT release over time in the 5 , 7 - D H T d e n e r v a t e d hippocampus. In contrast to the baseline overflow, KC1- and P C A - e v o k e d 5 - H T release was substantially reduced, by about 8 0 - 9 5 % of normal, both in the

acute and the chronic lesioned animals (Table II, Fig. 2A). Similarly, the addition of indalpine to the perfusion m e d i u m , which caused a 5 - 6 - f o l d increase in 5-HT release in the normal hippocampus, had virtually no effect in the chronically d e n e r v a t e d hippocampus (Fig. 3A). Thus, in the presence of indalpine, 5-HT overflow was r e d u c e d by 85% in the chronically d e n e r v a t e d h i p p o c a m p u s , and the addition of KC1 (60 m M ) or T T X (10 -6 M) had no effect (open circles in Fig. 3A). The very low 5 - H I A A levels r e m a i n e d unchanged throughout these various manipulations (open circles in Fig. 3B). Grafted h i p p o c a m p u s

In the grafted hippocampi (3 m o n t h s after lesion and grafting) baseline 5-HT overflow was similar to normal. As shown in Fig. 2 A both KC1- and P C A - e v o k e d 5-HT release was restored in the grafted hippocampi, and the responses s e e m e d to be even greater than those seen in the n o r m a l animals. 5 - H I A A oveflow was about 4-fold higher than normal in the experiment illustrated in Table II and Fig. 2, and about 2-fold higher than n o r m a l in the experiment shown in Fig. 3. As in the normal hippocampus, 5 - H I A A output was m a r k e d l y reduced in the presence of KCI and after P C A injection (Fig. 2B).

327

In the presence of indalpine steady-state 5-HT release from grafted hippocampus was indistinguishable from that seen in the normal hippocampus, and it was 3-4 times higher than that seen on the contralateral lesioned control side (Fig. 3A). As in normal animals (triangles and dotted line in Fig. 3A), addition of depolarizing concentrations of KCI (60 mM) caused a sharp, 4-5 fold, increase in 5-HT release in the grafted hippocampi (filled circles), and TTX (10 -6 M) reduced 5-HT release to the level seen

Effect of KCI and p-chtoroamphetamine

in the lesioned hippocampi. The 5-HIAA levels remained unchanged throughout these manipulations (Fig. 3B). Microscopic analysis demonstrated surviving grafts in all animals. The dialysis probes were located in the mid-temporal portion of the hippocampal formation, as illustrated in Figs. 1 and 4, extending through both the dentate gyrus and the CA 1 area. The graft tissue occurred as one or several tissue masses within the choroidal fissure, medially, within the lateral ventricle, laterally, or within the hippocampal fissure (i.e. between the dentate gyrus and the hippocampus). Thus, in all cases the grafts were located at a distance from the dialysis probe.

1.00 5,7-DHT÷graft 5,7-DHT a,..-,h N o r m a l

o---o

Effect of KCI and TTX during

t

5-HT uptake btockade

T

0.60.

=~

__ 0.60:=].

0.sc~

~.

~0.40,

tJ,t /

0.20. 0-

1

®

~ 5,7-DHT+graft o--~, 5,7-DHT

• ~:

2

3

4 "~ KCI

5

6

~ ..d, Normal

0.~ 0.20 ~

"""A .......... J.

..&-

0.ro-

7

8

9

10

0-

f PCA

Time (hrs)

•

...

=&

T

®

Base

1

2 KCL

3

/4 TTX

5

Indalpine Time (hrs) 70

I

I

~

~

i

I

~-~

5,7-DHT +graft 5,7:-DHT

~ 60_~

25-

"c

~_ 50Q.

:z 15-

~.~...i i..-~. ;...~, ~ ......... . ,i.

o_ 10-

<~ 20<£

®

A...-,L N o r m a l

30-

0-

~--o5 , 7 - D H T

20-

~- 40-

~

5,7-OHT+graft

i

<< ., - -. . . ,. - -

~ ,_ _ _ ~

, - ~4¢-' -, ~_~_~

, - ..rr.-~---6.-.6_ , , . . ". ? '. - ~

1

3

& 5 ~, KC[

6

2

?

8

9

o-

,

10

tPCA

Time (hrs)

Fig. 2. Effect of KCI (added to the perfusate at a concentration of 100 mM during the sample period marked by the hatched bar) and PCA (injected i.p. at 2.5 mg/kg) on the levels of 5-HT (A) and 5-HIAA (B) in the perfusate. Baseline samples were collected every 60 min (120 #1), and during KCI and PCA application every 30 min (60 #1). The values are means + S.E.M. Differences from pre-drug baseline: *P < 0.05; **P < 0.01. Student's (two-tailed) paired t-test.

5-

®

o r------1

Base

~ - ~ _ ~ _ -b-- ,.o- - ~ =

i

1

,

i

2

i

KcM

i

3

~ - .o. _ ~ _ .o..--.o r

1

4

i

,

5

i

.Jr...x

Indalpine Time (hrs)

Fig. 3. Effect of KCI (60 mM) and T r x (1/xM) in the presence of the serotonin re-uptake blocker indalpine (1 #M) on the extracellular levels of 5-HT (A) and 5-HIAA (B). The drugs were applied to the perfusion fluid during the periods indicated by the hatched bars. Values are means + S.E.M. Differences from predrug baseline: *P < 0.05; **P < 0.01. Student's (two-tailed) paired t-test.

I

329 Serotonin i m m u n o h i s t o c h e m i s t r y , p e r f o r m e d in some selected cases, revealed a m a r k e d hyperinnervat±on p a t t e r n in the tissue surrounding the p r o b e (Figs. 4 and 5). In the contralateral, non-grafted h i p p o c a m p u s only few, scattered 5 - H T - i m m u n o r e active fibers remained. Correlation of 5-HT

release to tissue i n d o l e levels

In 5 of the lesioned and grafted rats and 5 normal controls from the e x p e r i m e n t illustrated in Table II and Fig. 2 p o s t - m o r t e m tissue 5 - H T and 5 - H I A A levels were assayed in the tissue surrounding the p r o b e , corresponding to the ventral two-thirds of the h i p p o c a m p a l formation. In Table III, the resultant tissue concentration values are c o m p a r e d with the levels of 5-HT and 5 - H I A A r e c o v e r e d in the perfusates from the same animals, u n d e r baseline and KCI stimulated conditions. Consistent with previous findings m the m e a n tissue 5-HT and 5 - H I A A levels in the grafted hippocampi w e r e on the average 3 - 4 - f o l d higher than normal. As shown in our previous report, these supranormal 5-HT concentrations reflect a higher than normal density of serotoninergic terminals f o r m e d by the grafted neurons in the host hippocampal formation. Baseline 5-HT overflow was nevertheless similar in the dialysates from the grafted and n o r m a l (as well as in the chronically lesioned), h i p p o c a m p i , whereas the 5 - H I A A overflow was about 5 times higher in the dialysates from

Fig. 5. Section 0.5-1 mm rostral to the probe track, showing a high density of 5-HT-immunoreactive fibers in the host dentate gyrus in the area close to the transplant. In this specimen the transplant (T) occurred as a number of small 5-HT-rich cell clusters attached to the alveus (a). lgb, lateral geniculate body. Bar = 0.5 mm.

TABLE III Basal and KCl-evoked 5-HT and 5-HIAA release in grafted, lesioned and normal hippocampi compared to post-mortem hippocampal 5-HT and 5-H1AA levels measured in the tissue surrounding the dialysis probe

Values are means + S.E.M. Figures within brackets give the concentration in the perfusate (pmol/pl) as a percentage of the concentration in the surrounding tissue (pmol/mg). KC1 was added to the perfusion fluid at a concentration of 100 mM. Normal" (n = 5)

5, 7-DHT (n = 5)

5, 7-DHT-graft (n = 5)

5-HT

5-HIAA

5-HT

5-HIAA

5-HT

5-HIAA

Baseline release (pmol/601d perfusate)

0.014 _+0.005 (0.009%)

11.9 ± 2.1 (11%)

0.020 + 0.005 (0.20%)

0.854+ 0.17 (9.5%)

0.016± 0.006 (0.003%)

57.5 + 14.6 (14%)

KCl-evoked release (pmol/60~ul perfusate)

0.498 _+0.004 (0.3%)

6.20 _+ 1.1 (5.6%)

0.109_+0.004 (1.1%)

0.53_+0.15 (5.9%)

1.01+0.18 (0.2%)

22.4+5.4 (5.6%)

2.52 _+0.36

1.84 _+0.48

0.17_+0.038

0.15_+0.03

8.00+2.0

6.76+1.83

Tissue content (pmol/mg w. wt.)

The baseline values from normal hippocampus is based on 4 animals.

330 the grafted hippocampi. KCl-evoked 5-HT release was on the average twice as high in the grafted hippocampi, compared to normal, and the 5-HIAA level was about 3.5 times higher than normal in the presence of high KCI. In Table III the figures within brackets give the concentration in the perfusate (pmol/pl) as a percentage of the concentration in the surrounding tissue (expressed as pmol/mg). Interestingly, the concentration of 5-HIAA in the perfusate represented a similar fraction of the surrounding tissue 5-HIAA levels in both normal, lesioned and grafted hippocampi, i.e. about 10-15% under baseline conditions and 5 - 6 % under KCl-stimulated conditions. By contrast, the 5-HT concentration recovered in the perfusate under baseline conditions represented a 3-fold higher fraction of the total tissue level in the normal hippocampus (0.009%) as compared to the grafted hippocampi (0.003%), and this fraction was 22 times higher than normal in the 5,7-DHT denervated hippocampi (0.20%). DISCUSSION Previous studies ~6-18 have shown that the steadystate extracellular levels of 5-HT, which are reached 6-8 h after implantation of the dialysis probe, are neuronally derived. Several lines of evidence indicate that changes in the extracellular levels of 5-HT, established after the initial stabilization period, correlate closely with changes in synaptic activity of the serotonergic neurons innervating the area, whereas the extracellular levels of the principal serotonin metabolite, 5-HIAA, primarily reflect intraneuronal metabolism of the transmitter ~6 18. Moreover, addition of a serotonin re-uptake blocker may be used to increase the very low baseline extracellular 5-HT levels with a maintained close association between changes in synaptic activity and 5-HT overflow 16"17. Consistent with previous observations in the striatum ~6 depolarizing concentrations of KCl (60 or 100 mM in the perfusion fluid), or systemic injection of the 5-HT releasing drug PCA, caused massive release of 5-HT in the normal hippocampus, and these responses were almost totally abolished in hippocampi previously denervated by an intraventricular 5,7-DHT injection. Similarly, in the pres-

ence of the serotonin re-uptake blocker indalpine, 5-HT overflow increased about 5-6 fold in the normal hippocampus but it was virtually unchanged in the 5,7-DHT denervated specimens. The grafted hippocampi exhibited 5-HT release characteristics that were similar to normal. 5-HT release was restored not only under baseline conditions, but also in the presence of re-uptake blockade and depolarizing concentrations of KCI. The ability of T I ' X to reduce 5-HT overflow to the level seen in the denervated hippocampus suggests that the release was impulse-dependent. In agreement with previous electrophysiological studies 25'29, therefore, the present data indicate that the grafted serotonergic raphe neurons are spontaneously active despite their ectopic location. In our previous study 1° we observed that intrahippocampal fetal raphe transplants, similar to the ones used here, establish a new 5-HT-containing innervation throughout the previously denervated host hippocampus. In most cases, this innervation had a markedly higher than normal terminal density in the areas surrounding the graft. Consistent with these results, the post-mortem hippocampal 5-HT tissue levels in the grafted animals studied here were on the average 3-fold higher than normal in the tissue surrounding the dialysis probe. Nevertheless, the extracellular 5-HT levels collected in the dialysis perfusates were close to normal in the grafted hippocampi, both under baseline conditions and in the presence of indalpine. This is in contrast to 5-HIAA which was markedly elevated in the perfusates collected from the grafted hippocampi. In all 3 conditions studied (normal, 5,7-DHT lesioned and 5,7-DHT-lesioned plus grafted hippocampi) the 5H I A A concentration in the perfusate represented a fairly constant fraction of the 5-HIAA concentration present in the surrounding tissue (10-15% under baseline conditions and about 5% under K+-stimu lation). This is consistent with the view that 5-HIAA output represents passive, unregulated diffusion of an intraneuronally formed metabolite. 5-HT overflow, on the other hand, appeared to be subjected to adaptive regulatory mechanisms at the terminal level. In the 5,7-DHT-denervated hippocampi, where the tissue 5-HT level was reduced by 90-95%, the extracellular 5-HT level had dropped to below the level of detection (i.e. <0.005 pmol/30

331 min) at 2 weeks after lesion, but it had recovered to near-normal levels at 3 months. Similar to what has previously been described after lesions in dopaminergic and noradrenergic pathways in the brain 1-3" 13,~4, this suggests that the residual 5-HT terminals develop a compensatory hyperactivity in the chronically denervated target, making it possible to maintain a near-normal extracellular 5-HT level in the partially denervated area. Such increased transmitter release probably reflects a combination of increased neuronal firing rate, increased transmitter synthesis and release, and reduced elimination of the transmitter by terminal re-uptake 13'28. The present results show, however, that the spared serotonin terminals were unable to respond normally to activating stimuli, such as K+-induced depolarization and PCA. The impaired 5-HT release was also evident when the 5-HT re-uptake mechanism was blocked by indalpine. The results obtained in the grafted hippocampi indicate an adaptive change in the opposite direction. Here, normal baseline extracellular 5-HT levels were maintained both in the presence and in the absence of indalpine, despite a substantial hyperinnervation in the area surrounding the probe. This suggests that the 5-HT release rate (i.e. release per terminal) was reduced, compared to normal, in the graft-reinnervated hippocampus. This is consistent with our previous observation 1° that the 5-HT turnover rate (measured as the tissue 5-HIAA:5-HT ratio) was increased by about 25% above normal in the poorly reinnervated hippocampi, and reduced below normal in the long-term hyperinnervation specimens. Similar adaptive changes have previously been observed by Ponzio and Jonsson 21 and

REFERENCES 1 Acheson, A.L. and Zigmond, M.J., Short and long term changes in tyrosine hydroxylase activity in rat brain after subtotal destruction of central noradrenergic pathways, J. Neurosci., 1 (1981) 493-504. 2 Acheson, A.L., Zigmond, M.J. and Stricker, E.M., Compensatory increase in tyrosine hydroxylase activity in rat brain after intraventricular injections of 6-hydroxydopamine, Science, 207 (1980) 537-540. 3 Agid, Y., Javoy, E and Glowinski, J., Hyperactivity of remaining dopaminergic neurones after partial destruction of the nigro-striatai dopaminergic system in the rat, Nature New Biol., 245 (1973) 150-151. 4 Andersson, K.J., Holets, V.R., Mazur, P.C., Lasher, R.

Bjfrklund and Wiklund 9 after regeneration of serotonergic neurons following neurotoxin-induced axotomy. In these studies reduced serotonin turnover and release was regionally restricted, i.e. confined to those areas which had become hyperinnervated by the regenerating axons, which is consistent with the idea that local autoregulation at the terminal level is important for maintenance of tonic transmitter release. In conclusion, the present results show that grafted fetal serotonergic raphe neurons are spontaneously active and can restore 5-HT release in a previously denervated target. Furthermore, the present observations provide compelling evidence that local regulatory mechanisms at the terminal level are operating in the graft-reinnervated target area, and that this regional regulation of transmitter release can compensate within each terminal field for abnormalities in the graft-derived innervation density. In line with the observations of Strecker et al. 27 on the autoregulation of dopamine release from intrastriatal nigral transplants, such compensatory adaptive regulation at the terminal level should provide an effective mechanism to maintain tonic baseline transmitter release from intracerebrally grafted monoaminergic neurons within the normal range. ACKNOWLEDGEMENTS We thank Birgit Haraldsson and Gertrude Stridsberg for excellent technical assistance and Ragnar M~rtensson for expert photographic work. The study was supported by grants from the Swedish MRC (04X-3874), the National Institutes of Health (NS-06701) and the Segerfalk and Crafoord foundations.

and Cotman, C.W., Immunocytochemical localization of serotonin in the rat dentate gyrus following raphe transplants, Brain Research, 369 (1986) 21-28. 5 Azmitia, E.C., Perlow, M.J.; Brennan, M.J. and Lauder, J.M., Fetal raphe and hippocampal transplants into adult and aged CB57BL/6N mice: a preliminary immunocytochemical study, Brain Res. Bull., 7 (1981) 703-710. 6 Beebe, B.K., M611g~rd, K., Bj6rklund, A. and Stenevi, U., Ultrastructural evidence of synaptogenesis in the adult rat dentate gyrus from brainstem implants, Brain Research, 167 (1979) 391-395. 7 Bj6rklund, A., Baumgarten, H.-G. and Rensch, A., 5,7-Dihydroxytryptamine: improvement of its selectivity for serotonin neurons in the CNS by pretreatment with desipramin, J. Neurochem., 24 (1975) 833-835.

332 8 Bj6rklund, A., Stenevi, U. and Svendgaard, N.-A., Growth of transplanted monoaminergic neurons into the adult hippocampus along the perforant path, Nature (Lond.), 262 (1976) 787-790. 9 Bj6rklund, A. and Wiklund, L., Mechanisms of regrowth of the bulbospinal serotonin system following 5,6-dihydroxytryptamine-induced axotomy. I. Biochemical correlates, Brain Research, 191 (1980) 109-127. 10 Daszuta, A., Strecker, R.B., Brundin, P. and Bj6rklund, A., Serotonin neurons grafted to the adult rat hippocampus. I. Time course of growth as studied by immunohistochemistry and biochemistry, Brain Research, 458 (1988) 1-19. 11 Foster, G.A., Schultzberg, M., Gage, F.H., Bj6rklund, A., H6kfelt, T., Nornes, H., Cuello, A.C., Verhofstad, A.A.J. and Visser, T.J., Transmitter expression and morphological development of embryonic medullary and mesencephalic raph6 neurones after transplantation to the adult rat central nervous system. I. Grafts to the spinal cord, Exp. Brain Res., 60 (1985) 427-444. 12 Foster, G.A., Schultzberg, M., Gage, EH., Bj6rklund, A., H6kfelt, T., Cuello, A.C., Verhofstad, A.A.J., Visser, T.J. and Emson, P.C., Transmitter expression and morphological development of embryonic medullary and mesencephalic raph6 neurones after transplantation to the adult rat central nervous system. II. Grafts to the hippocampus, Exp. Brain Res., 70 (1988) 225-241. 13 Gage, F.H., Bj6rklund, A. and Stenevi, U., Local regulation of compensatory noradrenergic hyperactivity in the partially denervated hippocampus, Nature (Lond.), 303 (1983) 819-821. 14 Hefti, F., Melamed, E. and Wurtman, R.J., Partial lesions of the dopaminergic nigrostriatal system in rat brain: biochemical characterization, Brain Research, 195 (1980) 123-137. 15 Holets, V.R. and Cotman, C.W., Postnatal development of the serotonin innervation of the hippocampus and dentate gyrus following raphe implants, J. Comp. Neurol., 226 (1984) 157-176. 16 Kal6n, P., Strecker, R.E., Rosengren, E. and Bj6rklund, A., Endogenous release of neuronal serotonin and 5HIAA in the caudate-putamen of. the rat as revealed by intracerebral dialysis coupled to high performance liquid chromatography with fluorimetric detection, J. Neurochem., 51 (1988) 1422-1435. 17 Kal6n, P., Strecker, R.E., Rosengren, E. and Bj6rklund, A., Regulation of striatal serotonin release by the lateral habenula-dorsal raphe pathway in the rat as demonstrated by in vivo microdialysis: role of excitatory amino acids and GABA, Brain Research, in press. 18 Kal6n, P., Rosengren, E., Lindvall, O. and Bj6rklund, A., Hippocampal noradrenaline and serotonin release over 24

hours as measured by the dialysis technique in freely moving rats: correlation to behavioral activity state, effect of handling and tail-pinch, Eur. J. Neurosci., in press. 19 Luine, V.N., Renner, K.J., Frankfurt, M. and Azmitia, E.C., Facilitated sexual behavior reversed and serotonin restored by raphe nuclei transplanted into denervated hypothalamus, Science, 226 (1984) 1436-1439. 20 Nygren, L.-G., Olson, L. and Seiger, A., Monoaminergic reinnervation of the transected spinal cord by homologous fetal brain grafts, Brain Research, 129 (1977) 227-235. 21 Ponzio, E and Jonsson, G., Effects of neonatal 5, 7-dihydroxytryptamine treatment on the development of serotonin neurons and their transmitter metabolism, Dev. Neurosci., 1 (1978) 80-89. 22 Privat, A., Mansour, H., Rajaofetra, N. and Geffard, M., Intraspinal transplants of serotonergic neurons in the adult rat, Brain Res. Bull., 22 (1989) 123-129. 23 Privat, A., Mansour, H., Pavy, A., Geffard, M. and Sandillon, E, Transplantation of dissociated foetal serotonin neurons into the transected spinal cord of adult rats, Neurosci. Lett., 66 (1986) 61-66. 24 Segal, M., Interactions between grafted serotonin neurons and adult host rat hippocampus, Ann. N.Y.. Acad. Sci., 495 (1987) 284-294. 25 Segal, M. and Azmitia, E.C., Fetal raphe neurons grafted into the hippocampus develop normal adult physiological properties, Brain Research, 364 (1986) 162-166. 26 Segal, M., Azmitia, E., Bj6rklund, A., Greenberger, V. and Richter-Levin, G., Physiology of graft-host interactions in the rat hippocampus, Prog. Brain Res., 78 (1988) 95-101. 27 Strecker, R.E., Sharp, T., Brundin, P., Zetterstr6m, T., Ungerstedt, U. and Bj6rklund, A., Autoregulation of dopamine release and metabolism by intrastriatal nigral grafts as revealed by intracerebral dialysis, Neuroscience, 22 (1987) 167-178. 28 Stricker, E.M. and Zigmond, M.J., Brain monoamines, homeostasis and adaptive behavior. In V.B. Mountcastle, EE. Bloom and S.K. Geiger (Eds.), Handbook of Physiology, Sec. L Vol. IV: Intrinsic Regulatory Systems of the Brain, American Physiological Society, Bethesda, MD, 1986, pp. 677-700. 29 Trulson, M.E., Hosseini, A. and Trulson, T.J., Serotonin neuron transplants: electrophysiological unit activity of intrahippocampal raphe grafts in freely moving cats, Brain Res. Bull., 17 (1986) 461-468. 30 Ungerstedt, U., Measurement of neurotransmitter release by intracranial dialysis. In C.A. Marsden (Ed.), Measurements of Neurotransmitter release in vivo. IBRO Handbook Series: Methods in Neuroscience, Vol. 6, Wiley, New York, 1984, pp. 81-107.