Ž .Brain Research 760 1997 238–242

Research report

Chronic paroxetine desensitises 5-HT but not 5-HT autoreceptors in rat1D 1B

lateral geniculate nucleus

Colin Davidson 1, Jonathan A. Stamford )

( )Anaesthetics Unit Neurotransmission Laboratory , St. Bartholomew’s and the Royal London School of Medicine and Dentistry, Alexandra Wing, RoyalLondon Hospital, Whitechapel, London, E1 1BB, UK

Accepted 25 February 1997

Abstract

Ž .The present study examined the effect of chronic paroxetine 10 mgrkg p.o., 21 days on the 5-HT and 5-HT autoreceptors1B 1DŽcontrolling 5-HT efflux in slices of rat ventrolateral geniculate nucleus. Electrically stimulated 5-HT efflux 10 pulses, 200 Hz, 0.1 ms, 10

. Ž . ŽmA was measured using fast cyclic voltammetry. Peak 5-HT efflux was greater P-0.01 after chronic paroxetine 22.2"1.4 nM,. Ž . Ž . Ž .mean"S.E.M. than water 15.8"1.4 nM . 5-HT efflux was inhibited by CP 93129 1 nM–10 mM and sumatriptan 1 nM–1 mM ,

agonists at 5-HT and 5-HT receptors, respectively. Chronic paroxetine did not affect the sensitivity of the 5-HT autoreceptor but1B 1D 1BŽ .shifted the sumatriptan concentration-response curve to the right P-0.05 . These data suggest that chronic paroxetine increases evoked

5-HT efflux. This may be the result of desensitisation of 5-HT but not 5-HT autoreceptors.1D 1B

Ž .Keywords: Serotonin 5-HT ; Fast cyclic voltammetry; Lateral geniculate nucleus; Electrical stimulation; 5-HT ; 5-HT ; 5-HT uptake; Paroxetine1B 1D

1. Introduction

Depression is multifactorial and at least some of thesubclasses are thought to be due to hypofunction of the

Ž . w xserotonin 5-HT system 16 . The selective serotonin reup-Ž .take inhibitors SSRIs block 5-HT reuptake immediately

yet, like most antidepressants, do not exert their clinicaleffect for 2–3 weeks. This suggests that it is not theirreuptake blocking properties per se that induce the antide-pressant effect but another action dependent upon this. It

w xhas been postulated 2 that the SSRIs exert an antidepres-sant effect only after desensitisation of the inhibitory au-

w xtoreceptors on the 5-HT neuronal cell bodies 5,12 andw x5-HT terminals 17,18 induced by the SSRI-increased

5-HT tone at these autoreceptors.Until recently the rat somatodendritic and terminal

5-HT autoreceptors were thought to be of the 5-HT and1A

5-HT subtypes, respectively. However, the rat has now1Bw xbeen shown also to express the 5-HT subtype 10 .1D

Furthermore, these 5-HT receptors have been found to1Dw xact as inhibitory autoreceptors on 5-HT release 7,15,20 .

We have recently shown that electrically evoked 5-HT

) Ž .Corresponding author. Fax: q44 171 377-7126; e-mail: j.a.stam-ford@mds.qmw.ac.uk

1 Current address: School of Psychology, University of St. Andrews,St. Andrews, Fife, Scotland, KY16 9JU.

Ž .efflux in the ventrolateral geniculate nucleus vLGN isunder the control of both 5-HT and 5-HT autorecep-1B 1D

w xtors 8 . The purpose of this study was two-fold: firstly, todetermine whether SSRI-induced desensitisation occurredin the vLGN and, secondly, to establish whether desensiti-sation was exclusive to the 5-HT autoreceptor or also1B

involved 5-HT autoreceptors.1D

2. Materials and methods

Ž .Male Wistar rats Charles River, UK were housed, 5Žper cage, on a 12 h:12 h light:dark cycle lights on at 6

.am , with unrestricted access to water and standard ratŽ .chow. Groups of rats 120 g at start were dosed for 21Ž .days with paroxetine 10 mgrkg p.o., 14 rats or vehicle

Ž .water, 13 rats . The paroxetine dose was chosen on thew xbasis of its previously reported behavioural efficacy 13 .

2.1. Preparation of brain slices

Twenty-four hours after the last dose, rats were sacri-ficed by cervical dislocation. A single 350 mm brain sliceof the vLGN from the left hemisphere was taken from

Žeach rat under ice-cold artificial cerebrospinal fluid ACSF,composition in mM: NaCl 124, KCl 2.0, KH PO 1.25,2 4

Ž .MgSO .7H O 2.0, NaHCO 25, CaCl 2.0, and q -glu-4 2 3 2

0006-8993r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0006-8993 97 00289-8

( )C. DaÕidson, J.A. StamfordrBrain Research 760 1997 238–242 239

. Ž .cose 11.0 and placed in an oxygenated 95%O r5%CO2 2Ž .brain slice chamber and superfused with ACSF 31–328C

at 2.0 mlrmin. After 1 h of equilibration, the voltammetricŽ . Žcarbon fibre: 8=40 mm tip and stimulating bipolar

.tungsten, 100–150 mm seperation electrodes were placedin the vLGN at a depth of around 80–100 mm. The tips ofthe electrodes were positioned so as to form an equilateraltriangle with the voltammetric electrode. The workingelectrode–stimulating electrode array was positioned in thevLGN, just ventral to the intergeniculate leaflet.

2.2. Electrical stimulation

Ž5-HT efflux was evoked by electrical stimulation 10=

.0.1 ms pulses, 200 Hz, 10 mA, 95 ms train durationapplied every 5 min to an implanted bipolar tungstenstimulating electrode using standard Neurologe modules.

w xThese are pseudo-one-pulse stimulations 23 since thestimulation event is essentially over before the released5-HT has time to act on its own autoreceptors. On eachstimulation, two variables were measured: 5-HT efflux and

Ž .uptake see below .

2.3. Voltammetric detection of 5-HT efflux and uptake

w x5-HT was measured by fast cyclic voltammetry 24 .ŽBriefly a triangular voltage waveform q1.4 to y1.0 V

.versus AgrAgCl, 480 Vrs is applied to the carbonelectrode. At around 570 mV, 5-HT is oxidised, causing anincreased current to flow which, by calibrating the carbon

Fig. 1. Typical 5-HT efflux profiles in vLGN slices from paroxetine-Ž . Ž .treated upper trace and control rats lower trace after local electrical

Ž .stimulation 10 pulses, 200 Hz, 0.1 ms, 10 mA . 5-HT efflux wasmeasured as the peak extracellular 5-HT concentration attained during the

Ž .stimulation. The ‘half-time t ’ of 5-HT uptake was the time taken for1r2

the extracellular 5-HT concentration to fall to half of the maximumconcentration attained following stimulation.

electrode in a known quantity of 5-HT, could be convertedto a 5-HT concentration. The voltammetric scan was ap-plied twice per second to allow accurate measurement ofefflux and reuptake events. A sample and hold circuit

Ž .monitored the current 5-HT level at 570 mV. This signalwas fed into a chart recorder and a CED 1401 ArDconverter. Data was captured using CED ‘Chart’ andanalysed using CED ‘Signal Averager’ software. We havepreviously shown that 5-HT was the only monoamine

w xcontributing to the signal in the vLGN 8 .5-HT efflux was taken as the peak extracellular 5-HT

concentration attained following stimulation. On cessationof stimulation, 5-HT was removed from the extracellular

Ž .space by uptake. The ‘half-time t ’ of 5-HT uptake1r2Žthe time for the extracellular 5-HT concentration to fall to

.half of the peak attained was used as a measure of 5-HTuptake. Fig. 1 shows representative 5-HT efflux and reup-take profiles in slices from control and paroxetine-treatedrats and illustrates how 5-HT efflux and uptake t were1r2

measured.

2.4. Experimental protocol

Cumulative agonist concentration response curvesŽ .CRCs were obtained for the 5-HT agonist CP 931291Bw x w x14 and the 5-HT agonist sumatriptan 11 to test autore-1D

ceptor sensitivity. The slice was stimulated every 5 minuntil 5 successive stable release events were obtained.Subsequent 5-HT efflux values were expressed as a per-centage of the mean 5-HT efflux during this ‘control’period. The agonist was then added in increasing concen-

Ž .trations for 25 min 5 stimulations per concentration toensure plateau was reached.

2.5. Drugs and dosing

Paroxetine, sumatriptan and CP 93129 were gifts fromSmithKline Beecham Pharmaceuticals. Paroxetine was dis-solved in distilled water at 10 mgrml and rats received 0.1mlr100 g body weightrday p.o. Control rats received anequivalent volume of water. CP 93129 and sumatriptanwere initially dissolved in distilled water to make stocksolutions of 10y2 M. Further stock solutions were madeby serial dilution in ACSF. All agonists were added to thebrain slices via the superfusate.

2.6. Statistical analysis

Rat body weights, 5-HT efflux and 5-HT reuptakeŽ .half-time t were compared using a Student’s t-test.1r2

Agonist CRCs were compared using a 2-way ANOVA. Allvalues quoted are means"S.E.M. Graphpad Instat andGraphpad Prism software was used for analysis.

3. Results

Ž .Treatment of rats with paroxetine 10 mgrkg p.o. for21 days had no effect on the body weight of the animals at

( )C. DaÕidson, J.A. StamfordrBrain Research 760 1997 238–242240

Table 1The effect of chronic paroxetine on weight, evoked 5-HT efflux and 5-HTreuptake

Water ParoxetineŽ . Ž .ns13 rats ns14 rats

Ž .Rat weight g 323"33 317"28aŽ .Evoked 5-HT efflux nM 15.8"1.4 22.2"1.4

Ž .5-HT reuptake half-time ms 1570"86 1852"113

a Ž .All values are means"S.E.M. P -0.05 Student’s t-test .

time of sacrifice compared with water-treated controlsŽ .Table 1 . In both groups, electrical stimulation in the

Ž .vLGN 10 pulses, 0.1 ms, 10 mA, 200 Hz evoked a rapid5-HT efflux. On cessation of stimulation, the extracellular5-HT concentration fell. Electrically evoked 5-HT efflux

Ž .was significantly greater P-0.01, t-test in the paroxe-tine-treated group when compared to control. Fig. 1 showsa representative pair of traces whilst Table 1 contains thegroup data. There was no difference found in the reuptaket between the paroxetine- and water-treated groups1r2Ž .Table 1 .

In slices that were incubated in ACSF and repeatedlystimulated without addition of sumatriptan or CP 93129,5-HT efflux was constant over the course of at least 2.5 h.Stimulated 5-HT efflux after 2.5 h was 106"6% and

X Ž114"13% of the values at ts0 means"S.E.M, ns3,.paroxetine and water groups, respectively .

Fig. 2 shows the effect of chronic paroxetine treatmentupon 5-HT autoreceptor sensitivity in the vLGN. The1B

5-HT agonist CP 93129 caused a concentration-depen-1B

dent inhibition of 5-HT efflux in both the water- andparoxetine-treated groups. Calculated CP 93129 EC val-50

ues were 277 nM and 143 nM in the water- and paroxe-tine-treated groups, respectively. The two curves were not

Ž .significantly different 2-way ANOVA .Fig. 3 shows the effect of chronic paroxetine treatment

upon 5-HT autoreceptor sensitivity. The 5-HT agonist1D 1D

ŽFig. 2. Effects of CP 93129 on 5-HT efflux as a percent of pre-drug.values in vLGN slices taken from rats treated for 21 days with paroxe-

Ž . Ž .tine 10 mgrkg p.o.: ` or water v . Values are means"S.E.M.,. Ž .ns7 . The two curves were not significantly different 2-way ANOVA .

ŽFig. 3. Effects of sumatriptan on 5-HT efflux as a percent of pre-drug.values in vLGN slices taken from rats treated for 21 days with paroxe-Ž . Ž .tine 10 mgrkg p.o.: ` or water v . Values are means"S.E.M,

. Žns3r4 . The two curves were significantly different P -0.05, 2-way.ANOVA .

sumatriptan also caused a concentration-dependent inhibi-tion of vLGN 5-HT efflux in both the water- and paroxe-tine-treated groups. The two agonist concentration-re-

Žsponse curves were significantly different P-0.05, 2-way.ANOVA and there was a more than 10-fold higher EC50

Žvalue for sumatriptan in the paroxetine-treated group 443. Ž .nM compared with the water-dosed controls 35 nM .

4. Discussion

Chronic paroxetine treatment caused an increase inevoked 5-HT efflux in the vLGN slices. This result isconsistent with the theory that antidepressants, including

w xthe SSRIs, tend to increase 5-HT neurotransmission 2 .This result also agrees with other, similar studies whichfound an increase in evoked 5-HT efflux after chronic

w x5-HT reuptake blockade 17,18,21 .This increase in evoked 5-HT efflux appeared to be

independent of any change in the 5-HT reuptake site sincewe found no difference in the reuptake t in the vLGN1r2

after chronic paroxetine. This result is in agreement withw xO’Connor and Kruk 18 who found no difference in the

5-HT reuptake half-time in the dorsal raphe nucleus or thesuprachiasmatic nucleus after 21 days fluoxetine. Thesedata are also consistent with some binding studies e.g.

w xCheetham et al. 6 who found no change in corticalw3 x Ž .H paroxetine binding K or B after 14 days treat-d max

ment with various SSRIs and tricyclic antidepressants al-w xthough others e.g. Pineyro et al. 19 have found a decrease

Ž .in 5-HT reuptake sites B in cortical membranes aftermax

21 days paroxetine treatment.It should, however, be remembered that the current

method measures functional 5-HT uptake and does notdistinguish between different uptake sites. For instance,there may be more than a single uptake process for 5-HTw x22 and our data do not discount the possibility of down-

( )C. DaÕidson, J.A. StamfordrBrain Research 760 1997 238–242 241

regulation of the neuronal uptake carrier. As the datapresented here relate primarily to the 5-HT reuptake site inthe extracellular space, any changes in 5-HT reuptake hereneed not necessarily reflect changes in synaptic reuptakesites.

Elevated 5-HT efflux could be due to increased synthe-sis. However, most reports find that SSRIs decrease 5-HTsynthesis and turnover after both acute and chronic dosingw x1,9 . Another possibility is, therefore, that there is a deficitin the autoreceptor control of 5-HT efflux.

There was no change in response to the 5-HT agonist1B

CP 93129 after chronic paroxetine treatment. This impliesthat there is no desensitisation of the 5-HT autoreceptor1B

in the vLGN. This finding is, at first sight, at odds withmost previous reports. For instance, O’Connor and Krukw x18 found a decreased response to RU 24969 in the rat

w xSCN after 21 days of fluoxetine, and Moret and Briley 17found a decreased response to LSD in the rat hypothala-mus following chronic citalopram. On the other hand,

w xBosker et al. 3 found no difference in 5-HT autorecep-1B

tor sensitivity in the hippocampus after chronic fluvoxam-ine treatment.

There was a shift to the right in the concentration-re-sponse relation to the 5-HT agonist sumatriptan after1D

chronic paroxetine treatment. This suggests that there hasbeen a desensitisation of the 5-HT autoreceptor control-1D

ling 5-HT efflux in the vLGN. Sumatriptan, however, hasonly modest selectivity for 5-HT over 5-HT receptors1D 1Bw x11 and thus, on its own, the decreased response tosumatriptan implies only a desensitisation of a 5-HT -1Br D

like receptor. However, we have previously shown thatsumatriptan can be antagonised by the 5-HT antago-1DrB

nist GR 127935 but not by the selective 5-HT blocker1Bw xisamoltane 8 . Furthermore, as the response to the highly

selective 5-HT agonist CP 93129 was unaltered by1B

chronic paroxetine, the sumatriptan response strongly im-plies a desensitisation of the 5-HT autoreceptor.1D

Two factors may help to explain these disparities withsome of the literature. Firstly, many previous studies didnot use drugs selective for the 5-HT over the 5-HT1B 1D

receptor since it was not recognised at the time that the ratpossessed the 5-HT receptor. Consequently, decreased1D

responses to agonists in 5-HT terminal regions have beentacitly assumed to be a consequence of 5-HT autorecep-1B

tor desensitisation but may additionally reflect a popula-tion of 5-HT sites. Secondly, there are variations in the1D

relative densities of 5-HT and 5-HT receptors in dif-1B 1Dw xferent 5-HT terminal regions 4 . Thus it is possible that

the response to chronic SSRI treatment would be depen-dent on the relative proportions of 5-HT and 5-HT1B 1D

receptors present. The vLGN has a relatively high densityw xof 5-HT receptors 4 . It should also be remembered that1D

the effects demonstrated in this study were recorded after 3weeks of drug treatment. It would be interesting to look atearlier time points to record the dynamics of the phe-nomenon.

These data naturally beg the questions: Why are theretwo subtypes of 5-HT autoreceptor in this terminal regionand why is one desensitised but not the other? This is hardto answer but one possibility would be that the receptorshave different perisynaptic locations. For example, the5-HT autoreceptor might be found in the synapse and1B

Ž .thus frequently exposed to high many micromolar con-centrations of 5-HT. In this case the 5-HT autoreceptor1B

would then be physiologically desensitised and an increasein 5-HT tone following uptake blockade might have nofurther desensitising effect. Conversely, the 5-HT au-1D

toreceptor may be located outside the synapse and thusŽ .normally be exposed to low nanomolar 5-HT concentra-

tions. In such a case the 5-HT receptor might then be1D

desensitised by the elevated extracellular 5-HT concentra-tions that follow paroxetine treatment. It may also bepossible that the 5-HT and 5-HT receptors are coupled1B 1D

to different second messengers and paroxetine may differ-entially modulate the transduction cascades. These possi-bilities need further investigation.

In summary, we found that 21 days paroxetine treat-ment caused an increase in evoked 5-HT efflux in thevLGN. This increase was independent of any change in the5-HT reuptake site but may be related to a paroxetine-in-duced desensitisation of the 5-HT autoreceptor: there1D

was a decreased sensitivity in evoked 5-HT efflux tosumatriptan. The 5-HT autoreceptor sensitivity to CP1B

93129 was unchanged.

Acknowledgements

This research was principally funded by SmithKlineBeecham Pharmaceuticals. We also thank the Nuffieldfoundation for support.

References

w x1 A.A. Badawy, C.J. Morgan, Effects of acute paroxetine administra-tion on tryptophan metabolism and disposition in the rat, Br. J.

Ž .Pharmacol. 102 1991 429–433.w x2 P. Blier, C. De Montigny, Current advances and trends in the

Ž .treatment of depression, Trends Pharmacol. Sci. 15 1994 220–226.w x3 F.J. Bosker, A.A. Klompmakers, H.G. Westenberg, Effects of single

and repeated oral administration of fluvoxamine on extracellularserotonin in the median raphe nucleus and dorsal hippocampus of

Ž .the rat, Neuropharmacol. 34 1995 501–508.w x4 A.T. Bruinvels, J.M. Palacios, D. Hoyer, Autoradiographic charac-

terisation and localisation of 5-HT compared to 5-HT binding1D 1B

sites in rat brain, Naunyn-Schmiedeberg’s Arch. Pharmacol. 347Ž .1993 569–582.

w x5 Y. Chaput, C. De Montigny, P. Blier, Effects of a selective 5-HTreuptake blocker, citalopram, on the sensitivity of 5-HT autorecep-tors: Electrophysiological studies in the rat brain, Naunyn-Schmiede-

Ž .berg’s Arch. Pharmacol. 334 1986 342–348.w x6 S.C. Cheetham, J.A. Viggers, N.A. Slater, D.J. Heal, W.R. Buckett,

w3 xH Paroxetine binding in rat frontal cortex strongly correlates withw3 xH 5-HT uptake: Effect of administration of various antidepressant

Ž .treatments, Neuropharmacology 32 1993 737–743.

( )C. DaÕidson, J.A. StamfordrBrain Research 760 1997 238–242242

w x7 C. Davidson, J.A. Stamford, The effect of paroxetine on 5-HT effluxin the rat dorsal raphe nucleus is potentiated by both 5-HT and1A

Ž .5-HT receptor antagonists, Br. J. Pharmacol. 114 1995 1107–1BrD

1109.w x8 C. Davidson, J.A. Stamford, Serotonin efflux in the rat ventral

lateral geniculate nucleus measured by fast cyclic voltammetry isunder 5-HT and 5-HT autoreceptor control, Neuropharmacology1B 1D

Ž .35 1997 1627–1634.w x9 L.M. Hall, G.M. Anderson, D.J. Cohen, Acute and chronic effects of

fluoxetine and haloperidol on mouse serotonin and norepinephrineŽ .turnover, Life Sci. 57 1995 791–801.

w x10 M.W. Hamblin, R.W. McGuffin, M.A. Metcalf, D.M. Dorsa, K.M.Merchant, Distinct 5-HT and 5-HT serotonin receptors in rat:1B 1D

Structural and pharmacological comparison of the two cloned recep-Ž .tors, Mol. Cell Neurosci. 3 1992 578–587.

w x11 D. Hoyer, D.E. Clarke, J.R. Fozard, P.R. Hartig, G.R. Martin, E.J.Mylecharane, P.R. Saxena, P. Humphrey, VII. International Unionof Pharmacology classification of receptors for 5-hydroxytryptamineŽ . Ž .serotonin , Pharmacol. Rev. 46 1994 157–203.

w x12 R. Invernizzi, M. Bramante, R. Samanin, Chronic treatment withcitalopram facilitates the effect of a challenge dose on corticalserotonin output: Role of presynaptic 5-HT receptors, Eur. J.1A

Ž .Pharmacol. 260 1994 243–246.w x13 G.A. Kennett, S. Lightowler, V. De Biasi, N.C. Stevens, M.D.

Wood, I.F. Tulloch, T.P. Blackburn, Effect of chronic administrationof selective 5-hydroxytryptamine and noradrenaline uptake in-hibitors on a putative index of 5-HT2Cr2B receptor function,

Ž .Neuropharmacology 33 1994 1581–1588.w x14 B.K. Koe, J.A. Nielson, J.E. Macor, J. Heym, Biochemical and

behavioural studies of the 5-HT receptor agonist CP 94,253, Drug1BŽ .Dev. Res. 26 1992 241–250.

w x15 N. Limberger, R. Deicher, K. Starke, Species differences in presy-

naptic serotonin autoreceptors: Mainly 5-HT but possibly in addi-1B

tion 5-HT in the rat, 5-HT in the rabbit and guinea-pig brain1D 1DŽ .cortex, Naunyn-Schmiedeberg’s Arch. Pharmacol. 343 1991 333–

364.w x16 H.Y. Meltzer, M.T. Lowy, The serotonin hypothesis of depression,

Ž .in: H. Meltzer Ed. , Psychopharmacology: The Third Generation ofProgress, Raven Press, New York, 1987, pp. 513–526.

w x17 C. Moret, M. Briley, Serotonin autoreceptor subsensitivity and an-Ž .tidepressant activity, Eur. J. Pharmacol. 180 1990 351–356.

w x18 J.J. O’Connor, Z.L. Kruk, Effects of 21 days treatment with fluoxe-tine on stimulated endogenous 5-hydroxytryptamine overflow in therat dorsal raphe and suprachiasmatic nucleus studied using fast

Ž .cyclic voltammetry in vitro, Brain Res. 640 1994 328–335.w x19 G. Pineyro, P. Blier, T. Dennis, C. De Montigny, Desensitization of˜

the neuronal 5-HT carrier following its long-term blockade, J.Ž .Neurosci. 14 1994 3036–3047.

w x20 G. Pineyro, C. De Montigny, P. Blier, 5-HT receptors regulate˜ 1D

5-HT release in the rat raphe nuclei – In vivo voltammetry and inŽ .vitro superfusion studies, Neuropsychopharmacology 13 1995

249–260.w x21 A.A. Schoups, W.P. De Potter, Species dependence of adaptations at

the pre- and postsynaptic serotonergic receptors following long-termŽ .antidepressant drug treatment, Biochem. Pharmacol. 37 1988

4451–4460.w x22 E.G. Shaskan, S.H. Snyder, Kinetics of serotonin accumulation into

slices from rat brain: Relationship to catecholamine uptake, J.Ž .Pharmacol. Exp. Ther. 175 1970 404–418.

w x23 E.A. Singer, Transmitter release from brain slices elicited by singlepulses: A powerful method to study presynaptic mechanisms, Trends

Ž .Pharmacol. Sci. 9 1988 274–276.w x24 J.A. Stamford, Fast cyclic voltammetry: Monitoring transmitter re-

Ž .lease in real time, J. Neurosci. Methods 34 1990 67–72.