[1] 17w J~iiirtrul ~~/ ' N m i r o ~i ~c ~m i . s ~r j ~

[1] 174w The present study addressed the possibility that disinhibition of serotonin (5-HT) autoreceptor-mediated negative feedback might potentiate the elevation of nerve terminal 5-HT output induced by selective 5-HT reuptake blockade. To this end, rats were given citalopram and the 5-HT autoreceptor-blocking agents (S)-UH-301 (SHT,,) and (-)-penbutolol (5-HTIA,,& and the effect on extracellular 5-HT in the ventral hippocampus was monitored by means of in vivo microdialysis. Citalopram (5 mg/kg, s.c.) approximately doubled the 5-HT output, a response that was markedly augmented by (S)-UH-301 (3 rng/kg, s.c.) and (-)-penbutolol (8 mg/kg, s.c.) and by combined treatment with (S)-UH-301 (3 rng/ kg, s.c.) plus (-)-penbutolol ( I p M , via the dialysis perfusion medium), but not by (-)-penbutolol (1 p M ) alone. These findings provide evidence that 5-HT, in particular 5-HT,,, autoreceptormediated negative feedback mechanisms are pivotal in determining the nerve terminal 5-HT output level after 5-HT reuptake inhibition. These findings have important implications for the interplay between different processes controlling 5-HT transmission in vivo and might possibly offer a lead toward novel, therapeutically exploitable principles.

[1] 31w The average absolute baseline output of 5-HT in the diffcrent treatment groups ranged between 8.0 and 14.6 fmol/ 20 pl of dialysate, consistent with previous findings (cf. Hjorth and Sharp, 1992).

[2] 67w HT,, receptor antagonist (S)-UH-301 (3 mg/kg, s.c.) significantly ( p = 0.038) enhanced the dialysate output of 5-HT when given 1 h after citalopram (Fig. 1B). The maximal increase amounted to -15 and 30 fmol/20 pl relative to citalopram and baseline (predrug) levels, respectively. Similar doses of (S)-UH-301 alone do not alter the 5-HT output in rat hippocampus compared with controls (Nomikos et al., 1992;author's unpublished data).

[3] 66w Local administration of (-)-penbutolol. As shown in Fig. lC, local administration of (-)-penbutolol (1 p M ) via the dialysis perfusion medium did not alter the 5-HT response to citalopram (5 mg/kg, s.c.). Previous studies have shown that this concentration and mode of application of (-)-penbutolol increase the 5-HT output by -40% when citalopram is instead concurrently applied via the perfusion fluid (Hjorth and Sharp, 1992).

[4] 122w Local (-)-penbutolol plus systemic (S)-UH-301 administration. The combination of systemic (S)-UH-301 (3 mg/ kg, s.c.) and local (-)-penbutolol (1 p M ) treatment resulted in a clear-cut potentiation ( p < 0.001) of the citalopram response (Fig. 1 D). The magnitude of this action compared well with the effect of systemic injection of (-)-penbutolol alone (8 mg/kg, s.c.) after citalopram treatment (cf. Fig. IA) and was significantly ( p = 0.004) more marked than that seen after systemic (S)-UH-301 only (cf. Fig. IB). Also, the response peak amplitude was higher after combined local (-)-penbutolol and systemic (S)-UH-30 1 treatment compared with systemic (S)-UH-301 alone (3 I .5 2 6.2 vs. 13.4 ? 5.0 fmol/20 pI; p = 0.03 by Mann-Whitney U test).

[1] 56w The present study shows that drugs with prominent 5-HT autoreceptor blocking properties can strongly potentiate the 5-HT output-promoting effect of the 5-HT reuptake inhibitor citalopram in rat hippocampus in vivo. Moreover, the data are consistent with the interpretation that both the so- matodendritic and the nerve terminal 5-HT autoreceptors play a significant role in this-potentiation action.

[2] 339w When given alone, a high systemic dose of citalopram approximately doubled the 5-HT output in the ventral hippocampus. For comparison, application of the compound ( 1 pA4) via the dialysis perfusion medium results in 5-HT levels five to six times the control (no drug) baseline (see, e.g., Hjorth and Sharp, 1992). Similar observations have been reported previously using other 5-HT reuptake blockers, like fluoxetine (Auerbach et al., 1989), chlorimipramine (Carboni and Di Chiara, 1989;Adell and Artigas, 1991), and sertraline (Invernizzi et al., 1991). Notably, such (systemic) treatment also increased the extracellular 5-HT content to a greater extent in the somatodendritic than in the nerve terminal parts ofthe serotoninergic projections (Adell and Artigas, 1991;Invernizzi et al., 1991). It is well known that elevated extracellular concentrations of 5-HT will suppress 5-HT neuronal firing, synthesis, and release. The relatively modest increase in nerve terminal 5-HT output elicited by systemically administered citalopram, chlorimipramine, and sertraline is thus likely accounted for by an increase in 5-HT autoreceptor-mediated negative feedback, which accompanies the elevation of endogenous 5-HT content, thereby offsetting a further rise in the output of the transmitter. What, then, is the relative importance of the somatodendritic versus the nerve terminal 5-HT autoreceptors in this action? The present data indicate that the former, 5-HTl, autoreceptor, sites are pivotal. As summarized in Table 1, systemic administration of the selective 5-HT,, receptor antagonist (S)-UH-301 (Bjork et al., 1991;Nomikos et al., 1992) or the mixed 5-HTl,/5-HTlB receptor blocker (-)- penbutolol (Hjorth and Sharp, 1992) markedly promotes the citalopram-induced increase in hippocampal 5-HT output, in doses known to reverse 5-HTl, autoreceptor-mediated suppression of 5-HT neuronal firing (Arborelius and Svensson, 199 I ;VanderMaelen and Braselton, 1992), synthesis (Bjork et al., 199 1 ; Hjorth, 1992), and release (Nomikos et al., 1992;Hjorth and Sharp, 1992;author's unpublished data). Furthermore, in a recent preliminary study, Invernizzi et al. (1991) reported that application of the unselective 5-HT receptor antagonist methiothepin into the dorsal raphe 5-HT somatodendritic area uncovers the nerve terminal 5-HT output-elevating action of an otherwise seemingly ineffective systemic dose of sertraline.

[3] 126w The nerve terminal 5-HTl, autoreceptors may also participate in the overall, negative feedback-induced dampening of the citalopram response. In this regard, it appears that 5-HT,, autoreceptor blockade becomes functionally significant only in the presence of 5-HT neuronal firing, i.e., when the 5-HTL,, autoreceptor influence is minimized. Thus, intrahippocampal application of (-)-penbutolol, at a concentration known to block the 5-HT release suppression caused by the 5-HTl, receptor agonist CP-93,129 (Hjorth and Sharp, 1992), fails to affect the 5-HT response to subcutaneous citalopram (Fig. IC) but elevates the 5-HT output when citalopram is instead applied via the dialysis perfusion fluid (Hjorth and Sharp, 1992). For comparison, (-)-penbutolol given via the dialysis medium significantly augmented the (S)-UH-301-induced potentiation of the 5-HT output response to subcutaneous citalopram (Fig. 1 D).

[4] 166w The abovementioned data are in line with the notion that 5-HT autoreceptors may receive little endogenous activation under normotonic conditions (Nomikos et al., 1992;Hjorth and Sharp, 1992) and provide further information on the interplay of 5-HT,, and 5-HTl, autoreceptors in the control of 5-HT output in vivo. In particular, the findings support the view that somatodendritic 5-HTl, autoreceptor activation overrides the effect of nerve terminal 5-HTl, autoreceptor blockade (Hjorth and Sharp, 1992). Facilitation of nerve impulse-mediated release of 5-HT, secondary to a chronic treatment-induced desensitization of somatodendritic and nerve terminal 5-HT autoreceptors, has been proposed to underlie the antidepressant action of 5-HT reuptake-blocking drugs (Blier et al., 1987; Chaput et al., 199 1). By analogy, it might be speculated that if the net rise in 5-HT availability is crucial for therapeutic efficacy, the combination of 5-HT reuptake inhibitors and 5-HT, particularly 5-HTl,, autoreceptor-blocking drugs would already on acute treatment elicit a n effect functionally comparable to that resulting from prolonged treatment with 5-HT reuptake inhibitors alone.

[5] 93w In conclusion, the present data provide evidence that 5-HT,, and SHT,, autoreceptor-mediated negative feedback mechanisms are important in limiting the increase in nerve terminal output of endogenous 5-HT after 5-HT reuptake inhibition. In particular, blockade of the 5-HTl, autoreceptors strongly potentiated the nerve terminal 5-HT outputpromoting action of the 5-HT reuptake inhibitor citalopram. These findings have considerable implications for the understanding of the interplay between different processes controlling 5-HT transmission in vivo and might possibly offer a lead toward the development of novel, therapeutically exploitable principles, e.g., in the treatment of hyposerotoninergic conditions.

[6] 64w The skillful technical assistance of Jessica Schjott is gratefully acknowledged. I also wish to thank Prof. Uli Hacksell (University of Uppsala) and the pharmaceutical companies indicated in Materials and Methods for their generous gifts of drugs used in the study. This study was financially supported by grants from the Soderberg, Stohne, Gamla Tjanarinnor, Lundbeck, and Krapperup Foundations and by the Swedish MRC (grant 07486).

[1] 77w Male Sprague-Dawley rats (weighing 300-370 g; AB Laboratorietjanst, Stockholm, Sweden) were used in the study. The animals were kept five per cage under controlled temperature (22-25"C), humidity (60-65%), and 14/10-h light/ dark cycle (lights on 5 a.m.) conditions for at least 10 days before being used in the experiments. Food and tap water were allowed ad libitum. All animal procedures used in the present experiments were approved by the Animal Ethics Committee at the University of Goteborg.

[2] 204w The rats were implanted, under chloral hydrate anesthesia (400 mg/kg, i.p.; supplementary dosing given as needed to maintain adequate surgical anesthesia throughout the experiment, -80-1 00 mg/kg/h), with U-shaped microdialysis probes equipped with a Cuprophane-type membrane (folded tip length, 3 mm), prepared in our laboratory as described elsewhere (Tao and Hjorth, 1992). After the rat was mounted in a standard stereotaxic frame (Kopf), a hole was drilled in the skull bone, the dura was removed, and a microdialysis probe was implanted into the ventral hippocampus [probe tip at A-P -4.8 mm, L 4.6 mm, and D-V -8.5 mm, from bregma and dura surface, respectively (Paxinos and Watson, 1982)l. Artificial CSF (for composition, see Tao and Hjorth, 1992) was perfused through the probe at a constant rate (1.2 pl/min), and dialysates were collected every 20 rnin after probe implantation. The experiments were commenced when the levels of 5-HT had stabilized ( 5 10% change between consecutive samples); this occurred typically 2-3 h after probe implantation. Citalopram, (S)-UH-30 I , and (-)-penbutolol were dissolved in 0.9% NaCl and administered subcutaneously in the neck region in a volume of 1 ml/kg. In some experiments, (-)-penbutolol was instead applied to the microdialysis sampling area via the perfusion medium.

[3] 64w The dialysate contents of 5-HT were determined by means of standard HPLC with electrochemical detection [mobile phase, 0.126 A4 NaH,PO,, 0.85 m M EDTA, 13% (vol/vol) methanol, and 0.01 m M sodium n-octyl sulfate, pH 4.0 with phosphoric acid], with the detector potential set at 0.65 V. The sensitivity of the assay was 1 fmol of 5-HT/20 pl of dialysate (S/N ratio = 3).

[4] 61w The drugs used were chloral hydrate (Sigma), 1-[3-(dimeth y1amino)prop yl]-1 -(p-fluorophenyl)-5-phthalancar-bonitrile HCI (citalopram; courtesy of Drs. J. Arnt and J. Hyttel, H. Lundbeck A/S, Copenhagen-Valby, Denmark), (-)-1 -t-butylamino-3-(2-cyclopentylphenoxy)-propan-2-0~ sulfate [( -)-penbutolol; courtesy of Hoechst AB, Sweden], and S-(-)5-fluoro-8-hydroxy-2-(di-n-propylamino)tetralin HCI [(S)-UH-301 (Bjork et al., 1991); courtesy of Prof. U. Hacksell, Department of Organic Pharmaceutical Chemistry, University of Uppsala Biomedical Center, Uppsala, Sweden].

[1] 210w Drugs that block the reuptake of serotonia (5-HT) are effective in the treatment of affective illness (Asberg et al., 1986). The therapeutic mechanism of action of such agents is considered to relate to a net enhancement of the 5-HT neurotransrnission (Blier et al., 1987;Chaput et al., 1991). As the reuptake process is a major inactivation mode for neuronally released 5-HT, a primary consequence ofits inhibition is an increase in the extracellular concentrations of the amine. However, the increase in 5-HT availability fol-lowing systemic administration of 5-HT reuptake inhibitors, e.g., citalopram, is balanced by concurrent reductions in 5-HT neuronal firing (Chaput et al., 1986), synthesis (Carlsson and Lindqvist, 1978), and release (Moret, 1985). Autoradiographic and in vitro radioligand binding studies have demonstrated the presence of the 5-HT reuptake carrier not only in the axon terminals, but as well-and in very high density-in the cell bodies of ascending 5-HT neurons (Hrdina et al., 1990). Also, the 5-HT autoreceptor populations involved in the feedback control of 5-HT neuronal activity in the rat are found both in the nerve terminal (5-HT,,) and in the cell somatodendritic regions (5-HTIA). It thus appears likely that the 5-HT transmission-enhancing effect of 5-HT reuptake blockade is self-limiting, owing to the concomitant activation of the 5-HT,, and 5-HT,B autoreceptors.

[2] 79w Based on the above considerations, it was reasoned that a disinhibition of 5-HT autoreceptor-mediated negative feedback might potentiate the 5-HT availability-increasing effect of 5-HT reuptake blockers. The present study was camed out to address this possibility, using (S)-UH-30 1, (-)-penbutolol, and citalopram as examples of agents that block 5-HT,, autoreceptors, SHT,,,,, autoreceptors, and the reuptake of 5-HT, respectively [Hyttel, 1982;Bjork et al., 1991;Hjorth and Sharp, 1992 (this latter reference has been presented in abstract form; Hjorth et al., 1991)].

[3] 87w The results are expressed as absolute differences in the 5-HT output (femtomoles per 20 p1 of dialysate) versus baseline, predrug levels. The summated absolute changes between 60 and 180 min after citalopram administration (using the value at t = 60 min as the starting point) were used as estimates of overall drug effect on the citalopram response. Statistical comparisons were made on these data, using oneway ANOVA followed by Fisher's protected least significant difference test (Macintosh; StatView 4.0). Probability levels of ~5 % were considered statistically significant.

[4] 83w Systemic administration of (-)-penbutolol. (-)-Penbuto-lo1 (8 mg/kg, s.c.), given I h after the citalopram injection, further elevated the 5-HT output: The absolute difference from citalopram and predrug baseline levels amounted to roughly 20 and 35 fmol/20 pl, respectively, that is, about twice the level after citalopram alone ( p = 0.001) or four to six times higher than the initial baseline (predrug) level (Fig. 1A). This dose of (-)-penbutolol does not change the 5-HT output under control (noncitalopram) conditions (Hjorth and Sharp, 1992).