Basal and Stimulated ExtracellularSerotonin Concentration in the Brainof Rats With Altered Serotonin UptakeLUZ ROMERO,1 BRANIMIR JERNEJ,2 NURIA BEL,1 LIPA CICIN-SAIN,2 ROSER CORTES,1

AND FRANCESC ARTIGAS1*1Department of Neurochemistry, Instituto de Investigaciones Biomedicas de Barcelona, CSIC, 08034 Barcelona, Spain

2Laboratory of Neurochemistry and Molecular Neurobiology, Ruder Boskovic Institute, Zagreb, Croatia

KEY WORDS dorsal raphe nucleus; hippocampus; microdialysis; platelets; rat sub-lines

ABSTRACT We examined the relationship between the density of serotonergic(5-hydroxytryptamine [5-HT]) uptake sites and extracellular 5-HT concentration in therat brain using microdialysis with two different models, lesions with 5,7-dihydroxytryp-tamine (50 µg in the dorsal raphe nucleus (DRN) 15 days before) and sublines of ratsgenetically selected displaying extreme values of platelet 5-HT uptake. Compared tocontrols, lesioned rats had a reduced cortical concentration of 5-hydroxyindoles (45%),unchanged basal extracellular 5-HT in the DRN and ventral hippocampus (VHPC), andreduced basal 5-hydroxyindoleacetic acid (5-HIAA) concentrations (46%, DRN; 22%,VHPC). Yet the perfusion of 100 mmol/L KCl or 1 µmol/L citalopram elevated dialysate5-HT significantly more in the DRN and VHPC of controls. In genetically selected rats,platelet 5-HT content and uptake were highly correlated (r2 5 0.9145). Baseline dialy-sate 5-HT (VHPC) was not different between high and low 5-HT rats and from normalWistar rats. However, KCl or citalopram perfusion increased dialysate 5-HT significantlymore in high 5-HT than in low 5-HT rats, and the former displayed a greater in vivotissue 5-HT recovery. Significant but small differences in the same direction were notedin [3H]citalopram binding in several brain areas, as measured autoradiographically.Thus, basal extracellular 5-HT (but not 5-HIAA) concentrations are largely independenton the density of serotonergic innervation and associated changes in uptake sites.However, marked differences emerge during axonal depolarization or reuptake blockade.The significance of these findings for the treatment of mood disorders in patients withneurological disorders is discussed. Synapse 28:313–321, 1998. r 1998 Wiley-Liss, Inc.

INTRODUCTION

The reuptake of serotonin (5-hydroxytryptamine [5-HT]) into nerve terminals takes place through a high-affinity sodium- and energy-dependent transporter pres-ent in nerve endings and cell bodies of serotonergicneurones (Fuxe et al., 1983; Cortes et al., 1988; Hrdinaet al., 1990). This process determines the concentrationof 5-HT in central nervous system (CNS) synapses andis therefore a key element in the control of the seroton-ergic activity. Several drugs of abuse, like cocaine andamphetamine, or compounds with therapeutic use, likeantidepressants, interact with the neuronal 5-HT trans-porter (Blakely et al., 1991; Hoffman et al., 1991; forreview see Jacobs and Azmitia, 1992; Hyttel 1994).

The basal extracellular 5-HT concentration in differ-ent regions of the rat brain, as measured with microdi-alysis, does not appear to be representative of thedensity of serotonergic innervation (Adell et al., 1991).

In rats lesioned with 6-hydroxydopamine at infancy, acondition resulting in a large serotonergic hyperinner-vation of the caudate-putamen, the baseline concentra-tion of 5-HT in striatal dialysates is unchanged com-pared to control rats (Jackson and Abercrombie, 1992).In contrast, regional dialysate 5-HIAA concentrationsparallel the amount of 5-HT and 5-HIAA in brain tissue(Adell et al., 1991). Lesions with 5,7-dihydroxytrypta-mine (5,7-DHT) have been used to assess the changes ofextracellular 5-HT concentration in relationship withthe density of innervation, with conflicting results(Daszuta et al., 1989; Kirby et al., 1995). Thus, the

Contract grant sponsors: Spanish Ministry of Health, 95/226; Ministry ofScience and Technology of the Republic of Croatia, 0108 0105; Generalitat deCatalunya (1995SGR-00445).

*Correspondence to: Dr. Francesc Artigas, Department of Neurochemistry,IIBB-CSIC, Jordi Girona 18-26, 08034 Barcelona, Spain. Email: fapnqi@cid.csic.es

Received 14 July 1997; accepted in revised form 30 September 1997.

SYNAPSE 28:313–321 (1998)

r 1998 WILEY-LISS, INC.

relationships between the density of serotonergic nerveendings, reuptake sites, and extracellular 5-HT concen-tration is unclear. A better understanding of theseaspects may help to improve the treatment of mooddisorders in patients with neurological illnesses associ-ated with (among others) serotonergic losses—for ex-ample, depressive symptoms in patients with Parkin-son’s disease (Cross, 1988; Chinaglia et al., 1993;Tejani-Butt et al., 1995).

To gain further insight into the role played by the5-HT transporter in the control of the extracellular(active) 5-HT concentrations in brain, we have exam-ined the changes induced by a partial lesion with5,7-DHT on the basal and potassium-stimulated extra-cellular 5-HT concentration in the dorsal raphe nucleus(DRN) (rich in serotonergic cell bodies) and hippocam-pus, which receives a very dense and structured seroton-ergic innervation (Oleskevich et al., 1991). The effectsof the local blockade of the 5-HT transporter have alsobeen examined in both areas. Furthermore, we haveassessed the effects of such manipulations in the VHPCof genetically selected rat sublines which exhibit ex-treme values of 5-HT uptake by blood platelets. As the5-HT transporter in platelets and CNS is encoded by asingle gene (Lesch et al., 1993), we reasoned that theseanimals might provide a physiological model to exam-ine the relationship between 5-HT uptake sites andextracellular 5-HT concentration in the absence ofchemically induced brain lesions.

MATERIAL AND METHODSAnimals

Male Wistar rats (Iffa Credo, Lyon, France, andInterfauna, Sant Feliu de Codines, Spain) weighing280–320 g were used. Animals were kept in a controlledenvironment (12 h light-dark cycle and 22 6 2°C roomtemperature). Food and water were provided ad libi-tum. Animal care followed the European Union regula-tions (O.J. of E.C. L358/1 18/12/1986). Two groups ofgenetically selected male Wistar rats with extremevalues of platelet 5-HT concentration and VMAX ofplatelet 5-HT uptake were also used (Zgr; Wistar,Ruder Boskovic Institute, Zagreb, Croatia). Animalswere used 1 week after arrival. The procedure ofbreeding of the latter groups of rats has been describedin detail elsewhere (Jernej and Cicin-Sain, 1990).Briefly, for the initial generation, four pairs of normalWistar rats with high and low platelet serotonin concen-tration were selected and mated separately. Determina-tion of platelet 5-HT concentration was performed inoffspring when they reached a weight of 100–120 g.Males and females with extreme values of platelet 5-HT(high and low) were selected for mating. Usually, 8–16litters were generated per line, and, by use of thedescribed procedure, 24 generations of high and lowplatelet 5-HT sublines have been bred. Mean values ofplatelet 5-HT in these sublines diverged progressively

and reached stabilized values corresponding to approxi-mately 70% and 150% of mean values of the platelet5-HT concentration in the normal Wistar rat popula-tion. As platelet 5-HT concentrations were determinedby the VMAX of 5-HT uptake—the latter being undergenetic control (Jernej and Cicin-Sain, 1991)—addi-tional measures of VMAX were carried out from thenineteenth generation onwards. In the present study,animals from the twenty-second generation were used.

Drug treatments

5-HT was from RBI (Natick, MA). 5-HIAA and 5,7-DHT were from Sigma (St. Louis, MO). Desipramine,imipramine, and citalopram were kindly provided byCIBA-GEIGY S.A. (Barcelona, Spain) and LundbeckA/S, (Valby-Copenhangen, Denmark), respectively.[3H]citalopram (79.5 Ci/mmol) was from New EnglandNuclear (Boston, MA). [14C]5-HT (55 mCi/mmol) wasfrom Amersham (Buckinghamshire, UK). Other materi-als and reagents were from local commercial sources.Rats were lesioned by injection of 50 µg 5,7-DHTdissolved in 1% ascorbic acid (2 µL in total) in the DRN(stereotaxical coordinates in millimeters with respectto bregma and duramater were AP 27.8, DV 27.0, L 3.1[Paxinos and Watson, 1986]). Rats were pretreatedwith 10 mg/kg i.p. desipramine 30 min before adminis-tration of 5,7-DHT to prevent destruction of catechol-aminergic terminals. Control animals were injectedwith vehicle. Microdialysis experiments were con-ducted 2 weeks after the injection of the toxin.

Platelet 5-HT measures

Platelet-rich plasma was prepared from 1 mL ofvenous blood, obtained by jugular venipuncture underlight ether anesthesia, using a highly reproducibleprocedure which enables repetitive sampling of bloodfrom small laboratory animals. 5-HT was determinedby fluorimetry (Jernej and Cicin-Sain, 1988) and ex-pressed per 108 platelets. For the measure of theactivity of the 5-HT transporter, platelets (25–30 3 106,1 mL) were incubated with [14C]5-HT (final concentra-tion 1 µmol/L) at 37°C. After 30 s, the incubation wasterminated by the addition of ice-cold saline, andplatelets were isolated on glass fiber filters by rapidvacuum filtration. Filters were washed and the radioac-tivity counted. Blanks were carried out at 0°C usingexactly the same procedure. Uptake was expressed inmoles per number of platelets and minute.

Surgical and microdialysis procedures

Microdialysis experiments were performed as de-scribed (Adell and Artigas, 1991) in lesioned and sham-operated rats 2 weeks after 5,7-DHT lesion. Probeswere stereotaxically implanted in the DRN (AP 27.8,DV 27.5, L 3.1, in mm with respect to duramater andbregma with a vertical angle of 30°) and in ventral

314 L. ROMERO ET AL.

hippocampus (VHPC) (AP 25.8, DV 28.0, L 25.0)(Paxinos and Watson, 1986).

The length of the microdialysis membrane (Cupro-phan, Gambro, Sweden) exposed to brain tissue was 1.5mm in DRN and 4 mm in VHPC (O.D. 0.25 mm). Beforeimplantation, rats were anesthetized with sodium pen-tobarbital (60 mg/kg i.p.) and placed in a stereotaxicframe. Animals were allowed to recover from surgery indialysis cages (cubic, 40 cm side) for approximately 20h, and then probes were perfused with artificial cerebro-spinal fluid (CSF) (125 mmol/L NaCl, 2.5 mmol/L KCl,1.26 mmol/L CaCl2 and 1.18 mmol/L MgCl2) at 0.25µl/min. Under these conditions, the in vitro recovery ofprobes for 5-HT was 27% (1.5 mm) and 42% (4 mm).Sample collection started 60 min after the beginning ofperfusion. Usually four to six fractions were collected toobtain basal values. Successive 20 min (5 µL) dialysatesamples were collected before local infusion of KCl (100mmol/L) or addition of citalopram (1 µmol/L). DuringKCl infusion, the osmolarity of the artificial CSF wasmaintained by an isoosmotic reduction of NaCl. At theend of experiments, rats were killed, and their brainswere carefully removed and frozen until neurochemicaland autoradiographic analyses. The correct placementof the microdialysis probes was checked by infusingFast Green dye and inspecting the entire course of theprobe after cutting the brain at appropriate levels. Thedata from animals with probes outside the structures ofinterest were discarded.

In rats with extreme values of platelet 5-HT uptake,the in vivo recovery of 5-HT by hippocampal tissue wasestimated by infusing 100 nmol/L 5-HT through themicrodialysis probes and quantifying the 5-HT concen-tration at the outlet of the probe (Bel et al., 1997). Asthe extracellular concentration of 5-HT in hippocampusis approximately 1 nmol/L (Adell et al., 1991), thecontribution of endogenous 5-HT to the measure of invivo recovery of 5-HT by brain tissue is negligible. Theselective serotonin reuptake inhibitors (SSRIs) citalo-pram, fluoxetine, and paroxetine dose-dependently in-hibit in vivo 5-HT recovery in forebrain to a maximumof approximately 20% of the infused concentration, thusindicating that most 5-HT infused through the dialysisprobe is taken up by a SSRI-sensitive high affinityuptake system by the adjacent tissue (Bel et al., 1997).Thus, changes of the efficacy of the 5-HT reuptake canbe reliably estimated by this procedure.

Chromatographic analysis

5-HT and 5-HIAA were analyzed in brain dialysatesby a modification of a HPLC method previously de-scribed (Adell and Artigas, 1991). The composition ofHPLC eluant was as follows: 0.15 M NaH2PO4, 1.3mmol/L octyl sodium sulphate, 0.2 mmol/L EDTA (pH2.8 adjusted with phosphoric acid) plus 27% methanol.5-HT was separated on a 3 µm ODS 2 column (7.5cm 3 0.46 cm) (Beckman, Fullerton, CA) and detected

amperometrically (10.6 V) with a Hewlett-Packard(Palo Alto, CA) 1049 detector having an absolute detec-tion limit of 0.5–1 fmol per sample. Retention time for5-HT was usually between 3 and 4 min. The concentra-tion of 5-hydroxyindoles in brain was determined afterultrasonic homogeneization of frozen tissue in 0.4 mol/lperchloric acid containing 0.1% ascorbic acid and 0.1%EDTA. Brain homogenates were centrifugated (12,000g30 min) and supernatants were analyzed by HPLC.

Autoradiography

Coronal sections (14 µm thick) were cut from frozenbrain tissue of some rats treated with saline and5,7-DHT and from all genetically selected rats at vari-ous anterior-posterior levels. Pieces of frontal cortexwere used for the analysis of 5-HT and 5-HIAA intissue. The autoradiographic analysis of the 5-HT trans-porter density was conducted using [3H]citaloprambinding (D’Amato et al., 1987) in 14 different brainareas. Sections were preincubated with 50 mmol/LTris-HCl, 120 mmol/L NaCl, 5 mmol/L KCl, pH 7.4, for15 min. Incubation with 1 nmol/L [3H]citalopram wascarried out for 1 h at room temperature in the samebuffer to label 5-HT uptake sites (D’Amato et al., 1987).Following incubation, the sections were washed twice(15 min each) with cold (4°C) buffer and were finallyrinsed with cold distilled water. Nonspecific bindingwas assessed in adjacent sections in presence of 1µmol/L imipramine. Sections were then dried overnightin a cold chamber and exposed to a tritium-sensitivefilm (Hyperfilm 3H; Amersham) for 15 days. The density(femtomoles/gram of tissue) of cortical [3H]citaloprambinding was measured with an image analysis system(Imaging Research Inc., Ontario, Canada) using [3H]Mi-croscales (Amersham) standards in each film.

Data analysis

Microdialysis results are expressed as femtomoles/fraction. In vivo tissue 5-HT recovery (expressed aspercentage) was calculated as 100 2 Coutlet, the latterbeing the concentration of 5-HT at the probe outletafter infusing 100 nM 5-HT by reverse dialysis. Statisti-cal analysis of raw data has been performed usingANOVA for independent or repeated measures, fol-lowed by t-tests where appropriate. Data are expressedas means 6 SEM. Statistical significance has been setat the 95% confidence level (two-tailed).

RESULTS5,7-DHT–lesioned rats

The lesion of dorsal raphe serotonergic neurons with5,7-DHT reduced the concentration of 5-hydroxyindoles(sum of 5-HT 1 5-HIAA) by 55% in frontal cortex(P , 0.001, Student’s t-test). The reduction was some-what more marked for 5-HIAA (approximately 60%)(Fig. 1). Lesioned rats exhibited also a marked reduc-

3155-HT UPTAKE AND EXTRACELLULAR CONCENTRATION

tion of [3H]citalopram binding in the DRN and mostforebrain areas, but this was less homogeneous thanthat observed when 5,7-DHT was injected i.c.v. (e.g.,Bel et al., 1997). Figure 2 shows the autoradiograms athippocampal and midbrain levels of sham-operated(Fig. 2A,B) and 5,7-DHT–lesioned rats (Fig. 2C,D),respectively.

Despite the reduction of 5-HT uptake sites and tissue5-HT content, basal dialysate 5-HT values did notsignificantly differ between sham-operated and le-sioned animals in the two areas examined, VHPC andDRN (Table I). However, 5,7-DHT–lesioned rats exhib-

ited reduced baseline 5-HIAA concentrations (46% and22% of controls in the DRN and VHPC, respectively)(Table I).

Marked differences in dialysate 5-HT concentrationsbetween both experimental groups were noted after theperfusion of KCl (100 mmol/L) and citalopram (1 µmol/L)by reverse dialysis. KCl infusion caused an approxi-mately sixfold increment of dialysate 5-HT in theVHPC and DRN of control rats (P , 0.001, repeatedmeasures ANOVA) (Fig. 3). In contrast, perfusion ofKCl elevated 5-HT by 40–50% in hippocampus andinduced a fourfold increase in the DRN of lesioned rats.A two-way ANOVA analysis of dialysate 5-HT valuesrevealed significant effects of the KCl infusion(P , 0.0001) and KCl 3 pretreatment (lesion) interac-tion (P , 0.01) in the latter region. Similarly, the effectsof KCl infusion were significantly different in the VHPCof sham-operated and lesioned rats (P , 0.0001, effectof the treatment; P , 0.001, treatment 3 lesion interac-tion).

The infusion of 1 µmol/L citalopram increased theconcentration of 5-HT in dialysates from the DRN morethan in VHPC of control rats (see fractions 10–14 inFig. 3). Mean concentrations in the DRN attainedapproximately 50 fmol/fraction after the infusion ofcitalopram and approximately 20 fmol/fraction in VHPC.The infusion of citalopram resulted in a much lowerincrement of dialysate 5-HT in either region of 5,7-DHT–lesioned rats (Fig. 3). Similarly to controls, the 5-HTincrements were lower in VHPC. A two-way ANOVArevealed a significant effect of the citalopram infusion(P , 0.0001) and of the lesion 3 citalopram interaction(P , 0.005) in both brain areas.

Genetically selected rats

High 5-HT and low 5-HT rats displayed significantlydifferent values of platelet 5-HT content and of platelet5-HT uptake (P , 0.01, Student’s t-test) (Fig. 4). Bothvariables were highly correlated (r2 5 0.9145, P ,0.0001) (see inset in Fig. 4). Despite these markeddifferences in platelet 5-HT content, the 5-HT concentra-tion in tissue from frontal cortex did not differ signifi-cantly among controls, low 5-HT, and high 5-HT rats(Table II).

Fig. 1. Concentration of 5-HT and 5-HIAA in frontal cortex of ratsinjected with 50 µg 5,7-DHT or an equivalent volume of vehicle (sham)in the dorsal raphe nucleus 14 days before. Asterisks representstatistically significant differences between groups. The sum of 5-HTand 5-HIAA concentrations was also significantly lower in lesionedrats (3.97 6 0.23 vs. 1.82 6 0.41 nmol/g). Data are means 6 SEM often controls and nine lesioned rats.

Fig. 2. Autoradiograms showing the labeling of 5-HT uptake sitesby [3H]citalopram in coronal sections at midbrain (B,D) and hippocam-pal (A,C) levels of a sham-operated (A,B) and 5,7-DHT–treated (C,D)rats. Note the overall reduction of labeling, including the dorsal raphenucleus but not the median raphe nucleus. Bar: 2 mm.

TABLE I. Effect of the lesion with 5,7-DHT on baseline 5-HT and5-HIAA dialysate values1

Baseline 5-HT Baseline 5-HIAABrain region (fmol/fraction) (pmol/fraction)

Dorsal raphe nucleusSham-operated 8.9 6 1.3(5) 14.9 6 2.1 (5)5,7-DHT–lesioned 6.3 6 2.2(5) 6.9 6 2.5* (5)

Ventral hippocampusSham-operated 6.0 6 1.1(5) 6.5 6 1.0 (5)5,7-DHT–lesioned 8.6 6 2.2(4) 1.4 6 0.6**(4)

1Data are means 6 SEM of the number of animals shown in brackets.*P , 0.04 with respect to sham-operated animals.**P , 0.004 with respect to sham-operated animals.

316 L. ROMERO ET AL.

Basal dialysate 5-HT values in VHPC did not differsignificantly among control Wistar (3.5 6 0.7 fmol/fraction, N 5 6), high 5-HT (5.3 6 1.8 fmol/fraction,N 5 5), and low 5-HT rats (2.4 6 0.8 fmol/fraction,N 5 7). Two-way ANOVA revealed a significantly differ-ent effect of KCl infusion in the three groups of rats(P , 0.0001, treatment effect; P , 0.0001 treatment 3group interaction). The effect of KCl was minimal in low5-HT rats (see fraction 5 in Fig. 5). The infusion of 1µmol/L citalopram by reverse dialysis elicited alsomarkedly different changes in dialysate 5-HT in high5-HT and low 5-HT rats, with smaller increments in thelatter group. A two-way ANOVA of the data revealedsignificant effects of the citalopram infusion (P , 0.0001)and of the citalopram 3 group interaction (P , 0.001).

At the end of experiments shown in Figure 5, thedialysis probes were disconnected from the perfusionpump, and the rats remained overnight in the dialysiscages. On the following day, probes were perfused againwith artificial CSF. After 1 h, the perfusion fluid wasreplaced by one supplemented with 100 nmol/L 5-HT.Four 20 min fractions were collected and analyzed. Themean of the 5-HT concentration in these fractions wascalculated and used as a single uptake value per each

Fig. 3. Dialysate 5-HT values in dorsal raphe nucleus and ventralhippocampus of sham-operated (open circles) and 5,7-DHT–treated(solid circles) rats. Bars in the upper part of both panels represent thetime of perfusion of KCl (one fraction) and 1 µmol/L citalopram (sixfractions). Points are means 1 SEM of five rats/group except for thehippocampus of lesioned rats (N 5 4). See text for statistical details.

Fig. 4. Platelet 5-HT content and uptake in rats geneticallyselected (high 5-HT rats, solid bars, N 5 5; low 5-HT, open bars,N 5 7) used in microdialysis experiments. Bars are means 1 SEM.Asterisks denote significant differences. Inset: Correlation betweenthe individual values of platelet 5-HT content and uptake (r 5 0.9145,P , 0.0001).

TABLE II. Tissue 5-HT and 5-HIAA concentrations in frontal cortexof genetically selected rats1

5-HT (nmol/g) 5-HIAA (nmol/g)

Controls 1.71 6 0.08 (6) 1.19 6 0.10 (6)High 5-HT 1.75 6 0.09 (5) 1.19 6 0.06 (5)Low 5-HT 1.89 6 0.09 (7) 1.45 6 0.12 (7)1Data are means 6 SEM of the number of animals shown in brackets. Nonsignifi-cant differences between groups (one-way ANOVA).

Fig. 5. Baseline and stimulated dialysate 5-HT values in low 5-HTrats (open squares), high 5-HT rats (solid squares), and control Wistarrats (open circles). The periods of infusion of KCl and 1 µmol/Lcitalopram are shown by black bars in the upper part of the graph. Seetext for statistical analysis of the data.

3175-HT UPTAKE AND EXTRACELLULAR CONCENTRATION

rat. High 5-HT rats displayed a significantly higherrecovery of 5-HT (55.4 6 2.4%; N 5 5) than low 5-HTrats (41.3 6 3.7%; N 5 7) or controls (34.2 6 4.1%;N 5 6) (P , 0.004 one-way ANOVA followed by signifi-cant post hoc t-test) (Fig. 6).

The differences in the function of the 5-HT trans-porter between the two genetically selected strains ofrats were accompanied by small differences in thedensity of [3H]citalopram binding in 10 out of the 14brain regions examined, as assessed autoradiographi-cally (Fig. 7). A two-way ANOVA analysis of the datarevealed a significant effect of the region (P , 0.0001)and of the group 3 region interaction (P , 0.0001).

DISCUSSION

The concentration of 5-HT in the brain extracellularspace depends on different factors, such as the firingrate of serotonergic neurones (Sharp et al., 1990), theactivity of monoamine oxidase (Sleight et al., 1988;Celada and Artigas, 1993), and the function of theantidepressant-sensitive 5-HT uptake system (for re-view see Fuller, 1994; Artigas et al., 1996). Previousstudies indicated that, in absence of pharmacologicalinterventions, the extracellular concentration of 5-HTin different areas of the rat brain is confined within arelatively narrow range and does not reflect that pres-ent in brain tissue (Adell and Artigas, 1991; Jacksonand Abercrombie, 1992). In experiments involving de-struction of the serotonergic system with 5,7-DHT,Daszuta et al. (1989) reported a very severe reduction ofbasal dialysate 5-HT values in the VHPC of lesionedrats, whereas Kirby et al. (1995) found unaltereddialysate 5-HT values even after an almost complete(.90%) lesion of the striatal 5-HT system.

The present work was intended to clarify thesediscrepancies and to examine the relationship betweenthe density of 5-HT uptake sites in brain and theextracellular concentration, using two different models.First, we used rats with a partial denervation of theserotonergic system to mimic the moderate losses of5-HT uptake sites observed in certain neurologicalillnesses, such as Parkinson’s disease (Chinaglia et al.,1993). Secondly, we used sublines of rats geneticallyselected on the basis of their extreme activity of the5-HT uptake system in platelets. As the same geneencodes the 5-HT transporter in platelets and brain(Lesch et al., 1993), we hypothesized that geneticallyselected animals could also exhibit parallel differencesof the 5-HT transporter in brain. This might enable theexamination of the above relationship in absence ofadaptative changes to the chemical lesion (e.g., 5,7-DHT–induced reactive gliosis [Frankfurt et al., 1991]).

The data obtained with lesioned rats are consistentwith the notion that the concentration of 5-HT in theinterstitial brain space has little relationship to thatpresent in tissue during steady-state conditions. Thecomparable baseline 5-HT concentration in lesionedand control rats is in agreement with the data of Kirbyet al. (1995) and may be likely explained by the fact thatthe partial destruction of serotonergic nerve endingsresults in a parallel reduction of release and reuptake.However, dialysate 5-HIAA was markedly reduced in5,7-DHT–lesioned rats, in parallel with tissue 5-HTand 5-HIAA changes. The more pronounced reductionof dialysate 5-HIAA in VHPC (22% vs. 46% of controlsin the DRN) accords with the greater survival of cell

Fig. 6. Bar graph showing the in vivo recovery of exogenous 5-HT(100 nmol/L, added through the dialysis probe) by hippocampal tissue.High 5-HT rats had a significantly greater recovery than low 5-HTrats or controls.

Fig. 7. Bar graph showing the density of labeling of 5-HT uptakesites by [3H]citalopram in coronal brain sections. The identification ofthe different areas and nuclei is as follows: 1, dorsal raphe nucleus; 2,median raphe nucleus; 3, superior colliculus; 4, entorhinal cortex; 5–9,frontoparietal cortex at different depths (corresponding approximatelyto layers I, II–IV, Va, Vb, and bottom of layer VI—supracallosalatria—respectively [Blue et al., 1988; Sur et al., 1996]); 10, caudateputamen; 11, olfactory tubercle; 12, amygdala; 13, CA3 field; 14,dentate gyrus. See text for statistical analysis. Data are means 1 SEMfrom six controls, five high 5-HT and seven low 5-HT rats.

318 L. ROMERO ET AL.

bodies after 5,7-DHT treatment observed previously(Gobbi et al., 1990). In the present experiments, thetoxin was injected in the DRN. This resulted in a moreheterogeneous and variable loss of uptake sites thanwhen the drug is injected ic.v. (e.g., Gobbi et al., 1990;Bel et al., 1997), thus suggesting that both proceduresmay not be equivalent in terms of reduction of seroton-ergic activity in DRN-innervated areas.

The greater increments of dialysate 5-HT in sham-operated rats after KCl application is consistent withthe view that differences in serotonergic transmissionbetween normal and lesioned rats appear during axo-nal depolarization. Likewise, the blockade of the 5-HTuptake by the local application of the SSRI citalopramresulted in very moderate increments of dialysate 5-HTin lesioned rats (compared to controls) in the two areasexamined, the DRN and VHPC. Because dialysate5-HT is representative of the equilibrium betweenrelease and reuptake, inhibition of the latter resulted inlower increments of the 5-HT concentration in lesionedrats. An opposite situation (i.e., hyperinnervation of thestriatum by 5-HT terminals in dopamine-depleted ratsat infancy) also results in unchanged baseline 5-HTvalues but greater increments after blockade of the5-HT reuptake (Jackson andAbercrombie, 1992). Hence,it appears that changes in the brain extracellularfraction of 5-HT take place only after the disruption ofthe balance between release and reuptake (e.g., byaxonal depolarization or inactivation of the 5-HT trans-porter), despite the existence of an extensive lesion ofthe serotonergic system. The greater 5-HT increment inthe DRN (compared to VHPC) after citalopram infusionin lesioned rats may be related to the lesser impact of5,7-DHT on serotonergic cell bodies (Gobbi et al., 1990)and to the preferential effect of SSRIs in the midbrainraphe nuclei vs. forebrain areas after local and systemicadministration (for review see Artigas et al., 1996).

There is a loss of serotonergic innervation in Parkin-sonian patients (Cross, 1988; Chinaglia et al., 1993)and a large comorbidity of Parkinson’s disease withdepression (Cummings, 1992). The present observa-tions suggest that SSRIs may have a limited antidepres-sant efficacy in such patients, as smaller increments ofthe 5-HT concentration in the brain interstitial spacewould be attained after their administration. In fact,conflicting data have been reported on the usefulness ofthe SSRI fluoxetine in treating depressive symptoms inParkinsonian patients (Caley and Friedman, 1992;Steur, 1993; Monastruc et al., 1995).

Genetically selected rats displaying extreme valuesof platelet 5-HT uptake showed also differences in thebrain 5-HT reuptake. As with lesioned rats, baseline5-HT dialysate values did not differ between groups.However, axonal depolarization with KCl and localblockade of the uptake with 1 µmol/L citalopram re-vealed the existence of differences that may be inter-preted as resulting from changes in the density or

efficacy of the 5-HT transporter. On account of the twomain opposite factors (release and reuptake) that con-trol extracellular 5-HT concentrations, the observeddifferences could conceivably be also caused by anenhanced 5-HT release in high 5-HT rats. To test suchpossibility, we conducted an additional experimentinvolving the in vivo measure of the tissue 5-HT uptakein the same animals. This procedure is based on themethod developed by Justice and coworkers to estimatethe actual extracellular concentration of dopamine inbrain (for review see Justice, 1993).

Rats with higher platelet 5-HT uptake also had asignificantly greater in vivo 5-HT recovery by hippocam-pal tissue, as compared to the low 5-HT group, thussupporting that the activity of the 5-HT transporterwas enhanced in the brains of rats of the former group.However, whereas the 5-HT elevations induced by theinfusion of KCl and citalopram were similar in controlsand the high 5-HT group (Fig. 5), the in vivo 5-HTrecovery of controls was comparable to that of low 5-HTrats (Fig. 6). This may indicate the existence of adap-tive changes of serotonergic brain function other thanreuptake. A method for the estimation of the 5-HTtransporter mRNA in rat platelets has been recentlydeveloped (Hranilovic et al., 1996) which will enable theexamination of whether parallel changes of the 5-HTtransporter transcript also occur in brain.

To assess whether the differences in tissue 5-HTrecovery between the different sublines of rats werecaused by changes of the 5-HT transporter density, weconducted an autoradiographic study, measuring the[3H]citalopram binding in several midbrain and fore-brain structures. The autoradiographic data indicatethe presence of significant differences in the density of5-HT uptake sites. Yet these were small and were notpresent in all brain regions examined. Thus, it remainsto be established whether the observed differences in5-HT uptake, when assessed by microdialysis, aresolely accounted for by a greater expression of uptakesites in some brain areas of high 5-HT rats. The veryhigh selectivity of citalopram for the 5-HT transporter(Blakely et al., 1991; Hoffman et al., 1991; Hyttel, 1994)and the characteristics of the autoradiographic assay(D’Amato et al., 1987) make it unlikely that other siteswere labeled. Moreover, [3H]citalopram binding paral-lels the density of serotonergic terminals in conditionsof hypo-, normo- and hyperinnervation (Descarries etal., 1995). Together, these observations indicate thatonly 5-HT uptake sites were detected using this radiola-beled ligand. One possible, although speculative, expla-nation for the discrepancy between the extent of theneurochemical and autoradiographic differences wouldbe that only a part of the total number of sites detectedwith [3H]citalopram are functional (i.e., anchored in thenerve-ending membrane). The rest might occur as areserve cytoplasmic pool or in the process of transportfrom cell bodies to axonal varicosities. Indeed, the vast

3195-HT UPTAKE AND EXTRACELLULAR CONCENTRATION

arborization of the 5-HT system (e.g., .106 varicosities/mm3 in hippocampus [Oleskevich et al., 1991]) is likelyto require the synthesis and transport of large quanti-ties of 5-HT transporter protein which may account inpart for the total density of uptake sites detected inbrain with tritiated ligands. In such a case, smalldifferences in the total number of uptake sites detectedmay result in important functional differences. Alterna-tively, the neuronal 5-HT transporter in these ratsublines might undergo a different type of regulationthan that in platelets, leading in the end to an in-creased affinity without marked changes in density.The present autoradiographic data cannot clarify thispoint as only one ligand concentration was used.

Because nothing is known on the activity of trypto-phan hydroxylase and monoamine oxidase—key en-zymes in the control of serotonergic function—in theserat sublines, it may be that some of the observedchanges (e.g., differences during depolarization withKCl) may be unrelated to the 5-HT transporter. Indeed,it is likely that other adaptative changes might haveoccurred in these rats through inbreeding. Yet the dataobtained with two different experimental proceduresfor the estimation of 5-HT uptake in vivo (local infusionof an SSRI and recovery of exogenous 5-HT by braintissue) appear to support the existence of differences inthe activity of the 5-HT transporter in brain.

Taken together, the present findings confirm andextend previous observations indicating that the basalextracellular concentration of 5-HT in areas rich in cellbodies and nerve terminals (DRN and VHPC, respec-tively) is, to a large extent, independent of the densityof serotonergic elements. However, during periods ofenhanced neuronal activity or after the pharmacologi-cal blockade of the 5-HT transporter with antidepres-sant drugs, the extracellular 5-HT concentration wouldbe markedly dependent on the density of serotonergicnerve terminals. The data obtained in unlesioned,genetically selected rats is consistent with this view. Inaddition, the parallelism between the function of the5-HT transporter in CNS and the VMAX of 5-HT uptakein platelets is a further element in support of the use ofplatelets to examine certain aspects of the 5-HT–mediated transmission. It is hoped that the futureavailability of knockout mice lacking the 5-HT trans-porter may help to further clarify these aspects. In themeantime, the genetically selected animals used in thisstudy represent a useful model to study key aspects ofthe regulation of serotonergic neurotransmission.

ACKNOWLEDGMENTS

This work was supported by grants from the SpanishMinistry of Health (Fondo de Investigacion Sanitaria95/266) and the Ministry of Science and Technology ofthe Republic of Croatia (0108 0105). The Department ofNeurochemistry is recipient of a grant from the Gener-

alitat de Catalunya (1995SGR-00445). Luz Romero isthe recipient of a predoctoral fellowship from the CIRIT(Generalitat de Catalunya). Thanks are given to CIBA-GEIGY and Lundbeck A/S for the generous supply ofdrugs. The excellent technical assistance of LeticiaCampa is gratefully acknowledged.

REFERENCES

Adell, A., and Artigas, F. (1991) Differential effects of clomipraminegiven locally or systemically on extracellular 5-hydroxytryptaminein raphe nuclei and frontal cortex. An in vivo microdialysis study.Naunyn Schmiedebergs Arch. Pharmacol., 343:237–244.

Adell, A., Carceller, A., and Artigas, F. (1991) Regional distribution ofextracellular 5-hydroxytryptamine and 5-hydroxyindoleacetic acidin the brain of freely moving rats. J. Neurochem., 56:709–712.

Artigas, F., Romero, L., de Montigny, C., and Blier, P. (1996) Accelera-tion of the effect of selected antidepressant drugs in major depres-sion by 5-HT1A antagonists. Trends Neurosci., 19:378–383.

Bel, N., Figueras, G., Sunol, C., Vilaro, T., Mengod, G., and Artigas, F.(1997) Inhibition of a glial 5-HT transporter by antidepressants.Eur. J. Neurosci., 9:1728–1738.

Blakely, R.D., Berson, H.E., Fremeau, R.T., Caron, M.G., Peek, M.M.,Prince, H.K., and Bradley, C.C. (1991) Cloning and expression of afunctional serotonin transporter from rat brain. Nature, 354:66–70.

Blue, M.E., Yagaloff, K.A., Mamounas, L.A., Hartig, P.R., and Molliver,M.E. (1988) Correspondence between 5-HT receptors and serotoner-gic axons in rat neocortex. Brain Res., 453:315–328.

Caley, C.F., and Friedman, J.H. (1992) Does fluoxetine exacerbateParkinson’s disease. J. Clin. Psychiatry, 53:278–282.

Celada, P., and Artigas, F. (1993) Monoamine oxidase inhibitorsincrease preferentially extracellular 5-hydroxytryptamine in themidbrain raphe nuclei. A brain microdialysis study in the awake rat.Naunyn Schmiedebergs Arch. Pharmacol., 347:583–590.

Chinaglia, G., Landwehrmeyer, B., Probst, A., and Palacios, J.M.(1993) Serotoninergic terminal transporters are differentially af-fected in Parkinson’s disease and progressive supranuclear palsy:An autoradiographic study with [3H]citalopram. Neuroscience, 54:691–699.

Cortes, R., Soriano, E., Pazos, A., Probst, A., and Palacios, J.M. (1988)Autoradiography of antidepressant binding sites in the humanbrain: Localization using [3H]imipramine and [3H]paroxetine. Neu-roscience, 27:473–496.

Cross, A.J. (1988) Serotonin in neurodegenerative disorders. In:Neuronal Serotonin. N.N. Osborne, and M. Hamon, eds. John Wiley& Sons Ltd., New York, pp. 231–253.

Cummings, J.L. (1992) Depression and Parkinson’s disease: A review.Am. J. Psychiatry, 149:443–454.

D’Amato, R.J., Largent, B.L., Snowman, A.M., and Snyder, S.H. (1987)Selective labeling of serotonin uptake sites in rat brain by [3H]citalo-pram contrasted to labeling of multiple sites by [3H]imipramine. J.Pharmacol. Exp. Ther., 242:364–371.

Daszuta, A., Kalen, P., Strecker, R.E., Brundin, P., and Bjorklund, A.(1989) Serotonin neurons grafted to the adult rat hippocampus. II.5-HT release as studied by intracerebral microdialysis. Brain Res.,498:323–332.

Descarries, L., Soucy, J.P., Lafaille, F., Mrini, A., and Tanguay, R.(1995) Evaluation of three transporter ligands as quantitativemarkers of serotonin innervation density in rat brain. Synapse,21:131–139.

Frankfurt, M., O’Callaghan, J., and Beaudet, A. (1991) 5,7-dihydroxy-tryptamine injections increase glial fibrillary acidic protein in thehypothalamus of adult rats. Brain Res., 549:138–140.

Fuller, R.W. (1994) Minireview: Uptake inhibitors increase extracellu-lar serotonin concentration measured by brain microdialysis. LifeSci., 55:163–167.

Fuxe, K., Calza, L., Benfenati, F., Zini, I., and Agnati, L.F. (1983)Quantitative autoradiographic localization of [3H]imipramine bind-ing sites in the brain of the rat: Relationship to ascending 5-hydroxytryptamine neuron systems. Proc. Natl. Acad. Sci. U.S.A.,80:3836–3840.

Gobbi, M., Cervo, L., Taddei, C., and Mennini, T. (1990) Autoradio-graphic localization of [3H]paroxetine specific binding in the ratbrain. Neurochem. Int., 16:247–251.

Hoffman, B.J., Mezey, E., and Brownstein, M. (1991) Cloning of aserotonin transporter affected by antidepressants. Science, 249:1303–1306.

320 L. ROMERO ET AL.

Hranilovic, D., Lesch, K.P., Ugarkovic, D., Cicin-Sain, L., and Jernej,B. (1996) Identification of serotonin transporter messenger-RNA inrat platelets. J. Neural Transm., 103:957–965.

Hrdina, P.D., Foy, B., Hepner,A., and Summers, R.J. (1990) Antidepres-sant binding sites in brain: Autoradiographic comparison of [3H]par-oxetine and [3H]imipramine, localization and relationship to seroto-nin transporter. J. Pharmacol. Exp. Ther., 252:410–418.

Hyttel, J. (1994) Pharmacological characterization of selective seroto-nin reuptake inhibitors (SSRIs). Int. Clin. Psychopharmacol., 9:19–26.

Jackson, D., and Abercrombie, E.D. (1992) In vivo neurochemicalevaluation of striatal serotonergic hyperinnervation in rats depletedof dopamine at infancy. J. Neurochem., 58:890–897.

Jacobs, B.L., and Azmitia, E.C. (1992) Structure and function of thebrain serotonin system, Physiol. Rev., 72:165–229.

Jernej, B., and Cicin-Sain, L. (1988) A simple and reliable method formonitoring platelet serotonin level in rats. Life Sci., 43:1663–1670.

Jernej, B., and Cicin-Sain, L. (1990) Platelet serotonin level in rats isunder genetic control. Psychiatry Res., 32:167–174.

Jernej, B., and Cicin-Sain, L. (1991) Genetic alterations of serotoninuptake kinetics in rat platelets. Biol. Psychiatry, 29:601S.

Justice, J.B. (1993) Quantitative microdialysis of neurotransmitters.J. Neurosci. Methods, 48:263–276.

Kirby, L.G., Kreiss, D.S., Singh, A., and Lucki, I. (1995) Effect ofdestruction of serotonin neurons on basal and fenfluramine-inducedserotonin release in striatum. Synapse, 20:99–105.

Lesch, K.P., Wolozin, B.L., Murphy, D.L., and Riederer, P. (1993)Primary structure of the human platelet serotonin uptake site—

identity with the brain serotonin transporter. J. Neurochem., 60:2319–2322.

Monastruc, J.L., Fabre, N., Blin, O., Senard, J.M., Rascol, O., andRascol, A. (1995) Does fluoxetine aggravate Parkinson’s disease? Apilot prospective study. Mov. Disord., 10:355–357.

Oleskevich, S., Descarries, L., Watkins, K.C., Seguela, P., and Das-zuta, A. (1991) Ultrastructural features of the serotonin innervationin adult rat hippocampus: An immunocytochemical description insingle and serial thin sections. Neuroscience, 42:777–791.

Paxinos, G., and Watson, C. (1986) The Rat Brain in StereotaxicCoordinates. Academic Press, Sydney.

Sharp, T., Bramwell, S.R., and Grahame-Smith, D.G. (1990) Release ofendogenous 5-hydroxytryptamine in rat ventral hippocampus evokedby electrical stimulation of the dorsal raphe nucleus as detected bymicrodialysis: Sensitivity to tetrodotoxin, calcium and calciumantagonists. Neuroscience, 39:629–637.

Sleight, A.J., Marsden, C.A., Martin, K.F., and Palfreyman, M.G.(1988) Relationship between extracellular 5-hydroxytryptamineand behaviour following monoamine oxidase inhibition and L-tryptophan. Br. J. Pharmacol., 93:303–310.

Steur, E.N. (1993) Increase of Parkinson disability after fluoxetinemedication. Neurology, 443:311–313.

Sur, C., Betz, H., and Schloss, P. (1996) Immunocytochemical detectionof the serotonin transporter in rat brain. Neuroscience, 73:217–231.

Tejani-Butt, S.M., Yang, J.X., and Pawlyk, A.C. (1995) Altered seroto-nin transporter sites in Alzheimer’s disease raphe and hippocam-pus. Neuroreport, 6:1207–1210.

3215-HT UPTAKE AND EXTRACELLULAR CONCENTRATION