University of Groningen SSRI augmentation strategies ...(5 centuries B.C.), these terms do not match the psychiatric classifications which are used nowadays. In the Diagnostic and

University of Groningen

SSRI augmentation strategiesCremers, Thomas Ivo Franciscus Hubert

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2002

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Cremers, T. I. F. H. (2002). SSRI augmentation strategies: pharmacodynamic and pharmacokineticconsiderations. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 11-05-2021

https://research.rug.nl/en/publications/ssri-augmentation-strategies(c5640d8b-1c46-450a-9bc2-c1945fc50590).html

SSRI augmentation strategies;pharmacodynamic and pharmacokinetic

considerations

Thomas Cremers

Rijksuniversiteit Groningen

SSRI augmentation strategies;pharmacodynamic and pharmacokinetic

considerations

Proefschrift

ter verkrijging van het doctoraat in deWiskunde en Natuurwetenschappenaan de Rijksuniversiteit Groningen

op gezag van deRector Magnificus

dr. F. Zwarts,

in het openbaar te verdedigen opvrijdag 13 september 2002

om 14:15 uur

door

Thomas Ivo Franciscus Hubert Cremersgeboren op 6 april 1972

te Heerlen

Promotores:Prof. Dr. H.V.WikströmProf. Dr. B.H.C. WesterinkProf. Dr. J.A. den Boer

Referent:Dr. F.J. Bosker

Boordelingscommissie:Prof. Dr. J. KorfProf. Dr. S. HjorthProf. Dr. J. Zaagsma

Paranimfen:Drs. S.C.L.M. CremersDrs. E.N. Spoelstra

The printing of this thesis was financially supported by:H. Lundbeck A/S

Printing: Grafisch Centrum Facilitair Bedrijf, University ofGroningen, The Netherlands

ISBN: 90-367-1585-7

Contents

Aim of this thesis 1

Chapter 1 The relevance of the serotonergic system in depression 3

Chapter 2 Clinical efficacy of antidepressants in depression 33

Chapter 3 Augmentation with a 5-HT1A, but not a 5-HT1B receptor 71antagonist critically depends on the dose of citalopram

Chapter 4 Desensitisation of serotonergic autoreceptors upon 93pharmacokinetically monitored chronic treatment withcitalopram

Chapter 5 Topical differences in serotonergic autoreceptor control; 111effects of chronic treatment with citalopram

Chapter 6 Acute and Chronic Effects of Citalopram on Postsynaptic 1335-Hydroxytryptamine1A(5-HT1A) Receptor MediatedFeedback: A Microdialysis Study in the Amygdala

Chapter 7 Is the beneficial antidepressant effect of co-administration 153of pindolol really due to somatodendritic autoreceptorantagonism ?

Chapter 8 Attenuation of SSRI-induced increases in extracellular 173brain 5-HT by benzodiazepines

Chapter 9 Augmentation of antidepressive effects of SSRIs by 185co-administration of 5-HT2C antagonists

Chapter 10 Summary and conclusions 201

Samenvatting voor de leek 207

Dankwoord 211

Aim

1

Aim of this thesis

Depression is a complex disorder and it is likely that many neurotransmitter systems areinvolved in its pathology. Yet, research has primarily focused on mono-aminergic systems forthe last decades. Through the years a large body of evidence has emerged for an involvementof monoamines (serotonin in particular) in depression. However, it may be somewhat naïve tobelieve that depression is merely based upon dysfunction of the serotonergic system, whichcan be cured by drugs that normalise an inappropriate functioning serotonergic system.In the first 2 chapters an introduction to depression and current insights in biochemistry andclinical relevance is given. Chapter 1 will focus on the classification of sub-forms ofdepression, as this illness presents itself as a cluster of symptoms, rather than a pathologicalentity. A closer look will be taken at the neurophysiology of the serotonergic system, includingthe function of auto- and heteroreceptors. Then, the evidence that has been gathered for a roleof serotonin in the neurobiology of depression will be discussed. Subsequently, the underlyingmechanism of the therapeutic effects of Selective Serotonin Re-uptake Inhibitors (SSRIs) willbe discussed. Specific attention will be paid to the desensitisation hypothesis and strategies toimprove therapy that can be derived from it. In chapter 2, the clinical consequences of thetherapeutic armentarium in depression will be discussed in light of its pharmacology,pharmacokinetics and short and long-term clinical efficacy.Although accumulating data is apparently indicating that there is relevance in enhancingfunctionality of the serotonergic system, still the underlying pharmacology is not fullyunderstood. The current thesis adresses this topic by evaluating the effects of chronictreatment with SSRIs in several brain areas, as well as comparing these effects to the acuteeffects that are observed when SSRIs are co-administered with serotonin autoreceptorantagonists. Furthermore, several clinical trials have focussed on the evaluation of thebeneficial effects of co-administration. These trials might have neglected severalpharmacodynamic as well as pharmacokinetic considerations in their design. The present thesiswill scrutenize these events in light of the relevance of poly-pharmacy and relevance of plasmaconcentrations. Finally an alternative method for augmentation of the effects of SSRIs will bepresented and evaluated with regard to its mechanism of action.

Chapter 1

2

The relevance of the serotonergic system in depression

3

Chapter 1

The relevance of the serotonergic system in

depression

Chapter 1

4

Contents

1. Classification of Depression

2. Neurophysiology of the serotonergic system:

2.1 Anatomy and physiology

2.2 Serotonergic receptors

2.3 Autoreceptors/Heteroreceptors

3. Neurobiology

3.1 Neurobiology of depression

3.2 Mechanism of action of SSRIs

3.3 Serotonergic augmentation strategies


5

1. Classification of depression

Depression presents itself as a diffuse cluster of psychiatric symptoms rather than strictpathological entities, which complicates their diagnosis considerably. Diagnostic criteria, likeDSM III, IIIr and IV, may be of help to psychiatrists and physicians but high co-morbidity,different sub-forms and incomplete symptomatology still complicates diagnosis and therapy.Attempts to classify psychiatric disorders have been made for centuries. Whereas terms likemelancholia (depression) and mania (manic episodes) were already present in ancient history(5 centuries B.C.), these terms do not match the psychiatric classifications which are usednowadays. In the Diagnostic and Statistical Manual of Mental Disorders IV (DSM), severalsubdivisions are used to classify patients into a subform of the disorder (adapted from (Franceset al. 1994)). Table 1 illustrates the complexity of the classification of depression. Given thehigh complexity of subforms, it would be expected that the concurrent neurobiologicalprocesses underlying them would be equally divers. Consequently, therapeutical interventionin these illnesses would be equally complex. However as will be shown in the followingchapters this is not entirely true.

Chapter 1

6

Major Depression

The disorder is characterised by feelings of depressed mood or loss ofgeneral interest for at least 2 weeks. In addition, at least five items from a 9-item feature scale should be present. This list contains items like in-, orhypersomnia, a marked decrease of general interest or pleasure, fatigue,psychomotor agitation or retardation and suicidal thoughts.

Dysthymic Disorder

Dysthymic disorder might be considered as the less severe form of majordepressive disorder. The differentiation between the two disorders isparticularly difficult since most symptoms are similar. The disorder ischaracterised by at least 2 years of depressed mood, accompanied byadditional depressive symptoms that do not meet the criteria for a majordepressive disorder

Depressive disorder(not otherwiseclassified)

In this classification, disorders are classified which do not meet with criteriaof other disorders like major depressive disorder, dysthymic disorder,adjustment disorders with depressed mood and/or anxiety.

Bipolar I and IIDisorder

This disorders are characterised by the presence of manic periodsalternating with periods of depression. Whereas in type I, manic periodprevail, type II bipolar disorder is characterised by the main presence ofdepressive period.

Cyclothymic disorder

Cyclothymic disorder, like dysthymic disorder might be considered as theless severe form of Bipolar disorders. This illness is characterised by at least2 years of alternating periods of hypomanic and depressive periods,however, these periods do not meet the criteria for manic periods or majordepressive periods.

Bipolar disorder (nototherwise classified)

This disorder classifies disorders with bipolar features, but which can not beclassified as such.

Mood Disorder Due toa General MedicalCondition

General medical conditions have been known to be able to cause severebehavioral changes in patients. The mood disorder which patients areclassified in is characterised by a prominent and persistent disturbance inmood that is judged to be a direct physiological consequence of a generalmedical condition.

Substance inducedMood disorder

This disorder is characterised by a prominent and persistent disturbance inmood that is a direct consequence of drugs (like drugs of abuse, as well astherapeutic drugs) or toxin exposure.

Mood Disorder (nototherwhiles classified)

In this classification, patients which can not be classified according to thecoding of any of the described mood disorders even the “not otherwiseclassified”, are included.

Table 1. Classification of depressive disorders according to DSM IV (Frances et al. 1994).


7

2. The Serotonergic system

2.1 Anatomy and physiology

More than a century ago serotonin (5-hydroxytryptamine) was discovered in blood. Thecompound was named after the vasoconstrictive properties it exerted on blood vessels (“sero”is plasma and “tonin” is constriction). Much later serotonin was also found to be present in thebrain, and subsequently hypothesised to be a neurotransmitter. Since its discovery in CNS,serotonin has become the target of neuropsychopharmacology.

Neuroanatomy

Most serotonin cell bodies are located in clusters of cells which are found near the midline ofthe midbrain, pons and brain stem. Using fluorescence histochemistry methods, Dahlstrom andFuxe (1964) have described nine serotonergic nuclei, called B1 to B9. Recent studies usingimmunocytochemistry have shown additional serotonergic cells in the area postrema, caudalpart of the locus coeruleus and around the interpenducular nucleus (Figure 1.).

Figure 1. Neuroanatomy of the serotonergic system in rat-brain.

The caudal positioned nuclei (B1 and B3), project mainly to the medulla and the spinal cord.The more rostrally positioned nuclei (B7 to B9 (median raphe (MRN), centralis superior anddorsal raphe (DRN) respectively) are thought to innervate ascending structures, while theintermediate located nuclei innervate ascending as well as descending structures (Figure 1.).The majority of innervation of the ascending structures is derived from the dorsal raphe andthe median raphe, which innervate for instance the cortex, limbic system, basal ganglia andhypothalamus. Although terminal structures are mostly “cross innervated” by these nuclei,topographical differentiations do exist. Whereas the dorsal raphe nucleus preferentially

Chapter 1

8

innervates structures like the cortex, striatum and globus pallidus, median raphe innervation ismost predominant in medial septum and dorsal hippocampus (McQuade and Sharp 1997).Several brain structures like ventral hippocampus are innervated by DRN as well as MRN(McQuade and Sharp 1997). In addition, some evidence exists for reciprocal connectionsbetween these two nuclei.

Physiology

Although high amounts of serotonin are present in the body, serotonin has to be synthesisedby the brain itself, as it can not cross the blood brain barrier under normal conditions (Sharmaet al. 1990; Sharma and Dey 1987; Sharma and Dey 1986).Plasma tryptophan, which is mainly derived from diet, is actively transported into the brain.Competition exists for this active transport between several aminoacids like aromaticaminoacids (tyrosine and phenylalanine), branched chain aminoacids (leucine, isoleucine andvaline), and others (methionine and histidine). As a consequence, the amount of tryptophaneavailable in the brain not only depends on the plasma concentration of tryptophan, but also onother amino-acids (Daniel et al. 1976).Tryptophane is converted into 5-hydroxytryptophan (5-HTP) by tryptophan hydroxylase, anenzyme which uses tetrahydrobiopterin as cofactor and which is located in serotonergicterminals as well as perikarial cytoplasm. Hydroxylation of tryptophan is the rate limiting stepin the synthesis of serotonin (Mockus et al. 1998).Upon formation of 5-hydroxytryptophan, the molecule is decarboxylised to form 5-hydroxytryptamine (serotonin) by an enzyme called aromatic acid decarboxylase (AADC),which is not regarded to be the rate limiting in the formation of serotonin (figure 2).

Figure 2. Biosynthesis and metabolism of serotonin

N

NH2

OHO

OH

N

NH2

OHO

N

NH2OH

N

OHOH

O

TryptophanHydroxylase

Amino acidDecarboxylase

MonoamineOxidase

AldehydeDehydrogenase

Tryptophan 5-hydroxytryptophan

Serotonin(5-hydroxytryptamine) 5-HIAA

(5-hydroxy-indole-acetic-acid)


9

Catabolism of serotonin mainly occurs by monoamine oxidase to form 5-hydroxyindoleacetaldehyde. Oxidation of this intermediate by aldehyde dehydrogenase forms5-hydroxyindoleaceticacid (5-HIAA), which is the predominant metabolite of serotonin in thebrain (figure 2). Similar to the formation of 5-HIAA, another metabolite, 5-hydroxytryptophol, can also be formed by reduction after monoamine oxidation of serotonin.

Once released from the neuron, the action of serotonin is mainly terminated by re-uptake ofthe molecule in the neuron. This re-uptake mechanism comprises a plasma membrane carrier,which is capable of transporting serotonin in both directions, depending on the concentrationgradient of the transmitter.

Chapter 1

10

2.2 5-HT receptors

Due to the rapid developments in the field of molecular biology, the amount of serotonergicreceptors has expanded significantly in the last decades. Although a large quantity of thesereceptors are known, pharmacologists are trying to evaluate their exact function in the CNS.Based on structural and functional properties, seven types of serotonergic receptors have beenclassified until now. As can be observed in table 3, the amount of subtypes of receptors isconsiderable. Table 3 summarises some characteristics of these receptors.

Family Type G-

protein

Effector

pathway

Putative agonist Putative antagonist

1 5-HT1A Gi/G0 cAMP ↓,

Opening K

channels,

closing Ca

channels

(+)-8-OHDPAT,

Flesinoxan

Way 100635

1 5-HT1B Gi cAMP ↓ Species

dependent

Species dependent

1 5-HT1D Gi cAMP ↓ PNU 109,291 Ketanserin

1 5-HT1E Gi cAMP ↓ 5-CT -

1 5-HT1F Gi/Go? cAMP ↓ LY 344864 LY 334370

2 5-HT2A Gq IP3 DOI MDL 100907

2 5-HT2B - IP3 BW 723C86 LY 266097

2 5-HT2C - IP3 Ro-60-0175 SB 242084

3 5-HT3 None Ion channel 2-Me-5-HT Ondansetron

4 5-HT4 Gs cAMP ↑ Cisapride RS 39604

5 5-HT5A/B - - - -

6 5-HT6 Gs cAMP ↑ - Ro-04-6790

7 5-HT7 Gs cAMP ↑ 5-CT SB-258719

Table 3 5-HT receptor classification and it selective ligands (adapted from Kennett. 1998)


11

5-HT1 subtypes

5-HT1A

In common with all other serotonin receptors (except for 5-HT3), the 5-HT1A receptor is amember of the 7 transmembrane G-protein coupled receptor (7-TM GPCR) superfamily.Upon activation of the receptor, cAMP formation decreases, potassium channels are openedor calcium channels are closed. The last two actions hyperpolarise the membrane of theneuron, thereby decreasing their susceptibility to depolarise. Overall, the 5-HT1A receptor isregarded as an inhibitory receptor. It is located in the raphe nuclei, but is also highly abundantin terminal areas such as prefrontal cortex and hippocampus.

5-HT1B

Similar to the 5-HT1A receptor, the 5-HT1B receptor is a 7 transmembrane receptor anddecreases cAMP formation upon activation. Since the nomenclature of 5-HT1B and 5-HT1D

receptors has changed, numerous miscommunications have occurred. Table 4 shows presentand old nomenclature of 5-HT1B and 5-HT1D receptors.

Species Present name Old nameHuman h-5-HT1B 5-HT1Dβ

Human 5-HT1D 5-HT1Dα

Guinea Pig GP-5-HT1B 5-HT1Dβ

Guinea Pig 5-HT1D 5-HT1Dα

Rat r-5-HT1B 5-HT1B

Rat 5-HT1D 5-HT1Dα

Table 4 Current and former nomenclature of 5-HT1B/D receptors

5-HT1B receptors differ between species. Although the 5-HT1B receptors of human, guinea pig,calf and several other species are very similar with respect to their structure as well as affinitiesfor ligands, the rat 5-HT1B receptor, although slightly different of composition, differsdrastically in its affinity for ligands. However, these receptors are thought to be specieshomologues with respect to their function in the brain (Wurch et al. 1997). The receptor islocated in terminal areas like the ventral hippocampus.

5-HT1D

As mentioned above, this receptor, which upon activation decreases cAMP formation, is alsosubject to frequent miscommunications due to a change in nomenclature (Table 2). However,in opposite to the 5-HT1B receptor, this receptor is a species homologue with respect to itscomposition, its functionality as well as its affinity for ligands. This receptor is found interminal areas like the cortex as well as in the raphe nuclei (Wurch et al. 1997; Sprouse et al.1997).

Chapter 1

12

5-HT1E

Similar to the other receptors of the 5-HT1 receptor family, 5-HT1E receptors are negativelycoupled to the formation of cAMP. Currently no selective ligands have been reported and itsfunction is unknown. Human brain binding studies have reported that 5-HT1E receptors(representing up to 60 % of 5-HT1 binding) are concentrated in the caudate putamen withlower levels in the amygdala, frontal cortex and globus pallidus (Hoyer et al. 1994).

5-HT1F

The 5-HT1F receptor has close homology with the 5-HT1E receptor and is also negativelylinked to adenylyl cyclase. mRNA is abundant in the dorsal raphe, hippocampus and cortex ofthe rat (Hoyer et al. 1994). Although its exact function is unknown, it might be relevant in thetreatment of migraine, as the selective agonist LY 334370 was claimed to block the effects oftrigeminal nerve stimulation (Phebus et al. 1996)

5-HT2 subtypesThese receptors, comprising subdivisions into 2A,B and C, have in common that theirsignalling pathway activates phosphoinisitide metabolism, which mobilises Calcium andactivates protein kinase C. Centrally, the 5-HT2A receptor is mainly found in cortex, claustrumand basal ganglia (Hoyer et al. 1994). In man, 5-HT2B receptors are mostly locatedperipherally, with high concentrations in the placenta, lung, liver, kidney, heart, intestines andstomach. Some presence of 5-HT2B receptors has been observed in amygdala, septum,hypothalamus and cerebellum, and stimulation of 5-HT2B receptors in rodents has beenreported to cause moderate anxiolysis (Kennett 1996; Duxon et al. 1997; Kennett. 1993). Incontrast to 5-HT2B receptors, 5-HT2C receptors, formerly called 5-HT1C receptors, are mainlyfound in the central nervous system. High levels of 5-HT2C receptors have been observed inthe chorioid plexus, cerebral cortex, hippocampus, striatum, and substantia nigra of rats aswell as humans (Barnes and Sharp. 1999; Hoyer et al. 1994). Whereas 5-HT2C receptoragonists (mCPP) have been shown to be anxiogenic, antagonists behave as anxiolytics(Kennett 1994; Burns et al. 1997; Abrahamowski et al. 1995). In addition, 5-HT2C

responsiveness might be decreased upon chronic treatment with antidepressants, indicatingpossible relevance in antidepressant treatment (Kennett 1993)

5-HT3 subtypesAs indicated in table 3, the 5-HT3 receptor is the only ligand gated ion channel in the serotoninreceptor family. It enhances depolarisation by increasing permeability to cations. The receptoris highly abundant in the gastric system, whereas in the CNS it is most abundant in the areapostrema, nucleus tractus solitarius, substantia gelatinosa and nuclei of the lower brain stem.Centrally active 5-HT3 antagonists have been reported to behave as anxiolytics (Hoyer et al.1994).


13

5-HT4 subtypes5-HT4 receptors are positively coupled to adenylyl cyclase. The 5-HT4 receptor has beenfound centrally as well as peripherally. It has been shown to be present in the nigrostriatal andmesolimbic systems of the brain of several species among which rat and human (Barnes andSharp. 1999). 5-HT4 agonists have been observed to enhance cognitive performance in rats,while 5-HT4 antagonists are reported to act as anxiolytics in animal models.

5-HT5 subtypesThe 5-HT5 receptor is probably the least understood receptor of the 5-HT class. Whereasmolecular biological research has provided evidence for 5-HT5A and 5-HT5B subtypes, theirpharmacology is largely unknown. The presence of the 5-HT5A receptor has been observed inhippocampus, hypothalamus, cerebral cortex, thalamus, pons, striatum, raphe and medulla. No5-HT5A receptors have been observed in peripheral tissue (Barnes and Sharp. 1999).Conversely, 5-HT5B receptors have been observed in peripheral tissue like heart, kidney andlungs, though no or low levels of the receptors were found in brain tissue (Barnes and Sharp.1999). Research focused on the pharmacology of these receptors could not elucidate secondmessenger cascade related to 5-HT5 receptors, as no effects were observed on IP3 nor oncAMP formation (Barnes and Sharp. 1999).

5-HT6 subtypesThese receptors appear to be largely confined to the CNS, though some have been observed inperipheral structures like stomach and adrenal gland as well. High levels of 5-HT6 mRNA havebeen observed in the caudate nucleus, olfactory tubercle, nucleus accumbens andhippocampus. Their activation has been shown to enhance cAMP formation. The relevance inanxiety and depression of this receptor could not be unequivocally observed using 5-HT6

knockout mice. However, the animals tended to show enhanced anxiety in behavioralparadigms (Barnes and Sharp. 1999).

5-HT7 subtypesThough being the newest child of molecular biology the 5-HT7 receptor has already gained alot of interest. Four splice variants have been found (5-HT7(a) to 5-HT7(d) ) in humans as well asrats. However, their exact pharmacological relevance has not been elucidated yet. Thisreceptor is highly abundant in the SCN raphe nuclei, thalamus, hypothalamus andhippocampus. Somewhat lower levels have been observed in regions like the cerebral cortexand amygdala. Although no data exist on the relevance of 5-HT7 agonists and antagonistsregarding their implication in anxiety and depression, the observation that 5-HT7 receptorswere downregulated upon chronic treatment with SSRIs might indicate relevance of thesereceptors in depression and anxiety (Barnes and Sharp. 1999). Interestingly, it has beenspeculated that the 5-HT7 receptor may play a role in the control of circadian rhythm.

Chapter 1

14

2.3 Autoreceptors / Heteroceptors

The activity of the serotonergic neuron is modulated by itself and several other systems bymeans of auto- and heteroceptors. Until now, the existence of at least 3 self-inhibitoryreceptors have been well established.

Autoreceptors

5-HT1A autoreceptors5-HT1A autoreceptors are present on the cell bodies of the dorsal and median raphe nuclei.Serotonergic nerve terminals, however, are devoid of 5-HT1A autoreceptors. Usingelectrophysiological recording to measure firing rates of serotonergic neurons in the DRN aswell as MRN, several authors have shown that the firing rate of serotonergic neurons isdecreased upon activation of these autoreceptors with selective 5-HT1A agonists. These effectswere shown to be reversed by putative 5-HT1A antagonists. This decreased firing rate iscaused by a 5-HT1A mediated opening of potassium channels, which in turn decreases theability of serotonergic neurons to depolarise.In addition to the effects of 5-HT1A autoreceptor activation on neuronal firing, it has beenshown, using microdialysis methods in freely moving animals, that activation of thesereceptors in the raphe decreases extracellular levels of serotonin in terminal areas. The same istrue in the raphe area itself in response to the decreased firing rate. This effect, similar to thedecrease in neuronal firing rate, could also be reversed by selective 5-HT1A antagonists(Bosker et al.1996).In addition to the inhibitory responses mediated by somatodendritically located 5-HT1A

receptors, the serotonergic system is also restrained by a long feedback loop. Although theprecise mechanisms underlying this mechanism are not exactly known, it has been shown to bemediated by post-synaptic located 5-HT1A receptors in several parts of the brain (Ceci et al.1994; Bosker et al. 1997; Casanovas et al.1999; Bosker et al. 2001).

5-HT1B autoreceptors5-HT1B autoreceptors are present on the terminals of serotonergic neurons. Upon activation ofthese receptors, release, as well as synthesis, of serotonin is directly inhibited. Infusion of 5-HT1B agonists, while measuring extracellular levels of serotonin, has been shown to decreaselevels of serotonin in several terminal areas in the brain. This effect could be blocked byadministration of selective 5-HT1B antagonists, indicative of its functionality as a terminalautoreceptor (Sharp et al. 1989; Hjorth and Tao 1991; Bosker et al. 1995).

5-HT1D autoreceptorsAlthough this receptor has been shown to be present in terminal areas, as well as in rapheregions, its function is not yet fully clear. In terminal regions it has been shown that thisreceptor does not predominantly control serotonin release, and might therefore be regarded asbeing of minor importance as compared to terminal 5-HT1B autoreceptors (Moret and Briley1997). In raphe regions, it was hypothesised that the 5-HT1D receptor might have an


15

autoreceptor function in controlling serotonin release. Although several studies have shown itsrelevance in controlling serotonin levels in the raphe, its relevance with regard to overallserotonin cell firing remains to be elucidated (Sprouse et al. 1997; el Mansari and Blier 1996;Starkey and Skingle 1994).

Other 5-HT receptors as autoreceptors:

5-HT2

Although local infusion of the 5-HT2 agonist DOI is devoid of any effects on serotonin releasein rat prefrontal cortex, it has been shown that systemic administration of this compounddecreases serotonin levels in the cortex. In addition, it also induced decreased neuronal firingin DRN. This effect however, could not be antagonised by co-administration of the 5-HT2

antagonist ketanserin (Wright et al.1990; Moret and Briley 1997). As the decrease establishedby the agonist could not be blocked by its concurrent antagonist, it is unlikely that 5-HT2

receptors are directly involved in serotonin release.

5-HT3

Local administration of the putative 5-HT3 agonist 2-methyl-5-HT has been shown to dose-dependently increase hippocampal 5-HT levels (Martin et al.1992). Administration of variousputative 5-HT3 antagonists, however, has yielded controversial results on extracellular levelsof 5-HT as measured with microdialysis. Ondansetron and (s)-zacopride, do not modifycortical 5-HT release, whereas (r)-zacopride decreased 5-HT in a dose dependent manner inrat prefrontal cortex, though all compounds are putative 5-HT3 antagonists (Barnes et al.1992). In ventral hippocampus, local infusion of 5-HT3 antagonist MDL 72222 did not haveany effect, but completely blocked the increase in 5-HT levels established by a local infusion of5-HT3 agonist 2-methyl-5-HT (Martin et al.1992). Thus, 5-HT3 receptors might act asreceptors which facilitate 5-HT release. Whether these receptors are located on theserotonergic neurons is unknown.

5-HT4

Systemic administration as well as local infusion of 5-HT4 agonists renzapride and 5-methoxytryptamine has been shown to increase extracellular levels of 5-HT in ventralhippocampus. Conversely, systemic and local application of 5-HT4 antagonist GR 125487Ddecreases extracellular levels of 5-HT. In addition, it also prevents the increase in 5-HT levelsestablished by local and systemic administration of the agonist renzapride (Ge and Barnes1996). Taken together, these data indicate that 5-HT4 receptors, similar to 5-HT3 receptors,might act as facilitatory receptors on serotonin release, though their presence on serotonergicterminals has not been proven yet.

Chapter 1

16

Heteroceptors:

The serotonergic system receives several innervations and conversely innervates numeroussystems. The present section merely illustrates a number of connections, as an overview of thecomplete interaction patterns would be beyond the scope of this thesis. For a more conciseoverview the reader is referred to several reviews on the interactions of the serotonergicsystem within the brain (Kennett 1998; Barnes and Sharp 1999).

Adrenergic:Alpha 1 heteroceptors are located on the cell bodies within the serotonergic nuclei. Severalauthors have shown that these receptors are tonically activated by norepinephrine and exert adriving force on serotonergic cell firing, especially within the DRN (Hjorth et al.1995). Theorigin of the noradrenergic innervation, although initially speculated to be derived from thelocus coeruleus, remains somewhat unclear. The innervation might be derived from otherbrainstem nuclei.

Alpha 2 heteroceptors are also present on serotonergic neurons, though their presence isrestricted to terminal areas were they, similar to 5-HT1B receptors, regulate serotonin releasedirectly (Tao and Hjorth 1992, de Boer et al.1996).

GABA-ergic receptors:GABAA receptors are present in the MRN and DRN nuclei. Their activation causes a decreasein firing rate of the serotonergic neurons in those nuclei. Whereas the serotonergic cells in theDRN are believed to be tonically inhibited by GABA, this seems not to be the case for theMRN nucleus (Tao and Auerbach. 1996).

GABAB receptors are located in the DRN and terminal areas like the nucleus accumbens.Although not tonically inhibited, stimulation of these receptors decreases 5-HT release interminal areas, and decrease 5-HT neuronal firing in DRN (Tao and Auerbach. 1996).

Dopaminergic receptors:Dopamine 1 (D1) receptors have been studied with respect to their possible contribution inregulating serotonin release. In the striatum, no effect on extracellular 5-HT levels wasobserved when D1 agonists were infused locally. Upon local infusion of mixed DA D1/D2

agonist apomorphine, an increase of serotonin release was observed in the DRN. This effecthowever, could not be blocked simultaneous application of the selective D1 antagonist SCH23390, indicative of absence of D1 receptors in DRN modulating serotonin release (Ferre andArtigas. 1993).


17

In contrast to D1 receptors, D2 receptors have been shown to be involved in serotonin release.As mentioned above, local infusion, as well as systemic administration of the mixed D1/D2

agonist apomorphine, or the D2 selective agonist LY 171,555 increased 5-HT levels in DRN.This could be blocked by local administration of the selective D2 antagonist raclopride. Thesedata indicate that D2 receptors exert a facilitating role on serotonin release in DRN (Ferre etal.1993) Local infusion of the D2 agonist sulpiride was shown to have little or no effect on 5-HT release in VTA. This observation, in addition to the absence of effects of localadministration of D2 agonists in striatum in serotonin extracellular levels, do not indicate thepresence of functional D2 receptors on serotonergic nerve terminals (Chen and Reith, 1995).In vitro experiments have shown that dopamine enhances excitability of dorsal raphe neurons,

this effect was found to be D2 mediated (Haj-Dahmane 2001).

Chapter 1

18

3. Neurobiology

3.1 Neurobiology of depression with respect to the serotonergic system

Several decades ago it was discovered that monoamine oxidase inhibitors were effectiveantidepressants. Soon tricyclic antidepressants were also found to be effective in this disorderand since their pharmacological effects comprised monoamine re-uptake inhibition, it washypothesised that a deficiency in monoamines might be responsible for the occurence ofdepression. This hypothesis as proposed by Bunney and Davis, as well as by Schildkraut in1965 is known as the monoamine hypothesis of depression.Although a large amount of research has been devoted to the evaluation of this hypothesis,several discrepancies have been encountered. As antidepressants have been shown to exerttheir monoamine enhancing effects immediately upon administration (Fuller 1994), the clinicalonset of action of these compounds is typically delayed for several weeks. In addition, severalantidepressants (like tianeptine and iprindole) have no inhibitory effects on re-uptake inhibitionalthough they do exert antidepressant activity. Thirdly, several monoamine re-uptake inhibitorslike cocaine do not posses antidepressant activity. Although apparently several features stillhave to be addressed with regard to the precise relevance of monoamine deficiency indepression, the involvement of monoamines in the pharmacology of depression has been wellestablished.

Relevance of deficits in serotonergic transmission in major depression

Although, as mentioned above, the monoamine hypothesis is not conclusive, several featuresindicate that deficits in serotonergic neurotransmission are involved in the pathophysiology ofdepression. These indications can be divided into peripheral and central parameters, dependingon the technique of evaluation.


19

Peripheral parameters:

Plasma parameters:

As methods for evaluation of in vivo brain pharmacology like PET have only become availablevery recently, brain dysfunction has been evaluated mainly by measuring peripheralparameters.Blood platelet dysfunction in the storage and or re-uptake of serotonin has been used forevaluation of dysfunctional serotonin transmission in several studies. In the majority of studies,lower uptake of serotonin was observed (Meltzer and Arora. 1991).The content of plasma or platelet serotonin has also been studied by several groups. Severalstudies have shown that 5-HT was lowered in plasma as well as platelets in depressed patients(Sarrias et al. 1987).

Precursor availability:

The amount of the precursor of serotonin, the aminoacid L-tryptophan, has been measured inplasma in order to look for a correlation with the presence of depression. In several studieslowered plasma concentrations were observed in depressed patients when compared tocontrols (Maes et al. 1990). Furthermore, as observed in a different study, a reduction in L-TRP availability by dietary exclusion can induce lowering of mood (Young et al. 1985).

Neuro-endocrine parameters:

The neuro-endocrine relationship has been used for studying serotonergic neurotransmission(Charney et al. 1982). In these experiments, neuro-hormones (cortisol, prolactin and growthhormone) responses are measured. The release of these hormones is hypothesised to beinfluenced by central serotonin neurotransmission. Differences in changes in these neuro-endocrine parameters upon administration of serotonergic compounds is thought to reflectdifferences in CNS serotonergic transmission.

Precursors:

When the serotonin precursor L-tryptophan was administered in high doses to untreateddepressed patients, an attenuated increase in prolactin and growth hormone levels wasobserved in depressed patients, indicative of subsensitive post-synaptic receptors in depression(Deakin et al. 1990; Charney et al.1984). Conversely, administration of L-tryptophan topatients treated with various antidepressant agents was shown to result in enhanced levels ofprolactin. This observation would indicate that responsiveness of postsynaptic receptors isenhanced after antidepressant treatment (Charney and Delgado. 1992).Administration of the serotonin precursor 5-hydroxytryptophan was found to induce enhancedlevels of cortisol in depressed patients. This observation was believed to be related to post-synaptic hypersensitivity (Meltzer et al. 1984). As simultaneously an inverse relation was

Chapter 1

20

observed between CSF levels of 5-HIAA, it was suggested that the cortisol effects wererelated to decreased serotonergic 5-HT metabolism (Koyama et al.1987).

Fenfluramine:

In untreated patients Siever et al. (1984) observed blunted effects of the serotonin releaserfenfluramine on serum prolactin levels. In addition, reduced prolactin responses tofenfluramine challenges were also observed in depressed patients with a history of suicideattempts (Coccaro et al. 1989). Enhanced levels of prolactin after administration offenfluramine have also been shown to be augmented after chronic treatment of patients withantidepressants (O'Keane et al. 1992). These observations would be indicative of decreasedpostsynaptic receptor sensitivity during depression, a feature which is reversed by chronictreatment with antidepressants.

5-HT1A agonists:

5-HT1A agonists have been shown to enhance plasma levels of cortisol and adrenocorticotropichormone (ACTH) upon systemic administration. The 5-HT1A agonist ipsapirone induced adecreased elevation of cortisol and ACTH in unmedicated depressed patients. Thisobservation was explained by a disrupted interaction of the serotonergic ligand andhypothalamic-pituitary-adrenal axis (HPA-axis) (Lesch et al. 1990). In depressed patients ithas been hypothesised that the increased levels of cortisol during baseline might be responsiblefor the downregulated sensitivity and responsivity of the post-synaptic 5-HT1A receptor(Deakin et al. 1990).

Central parameters:

As studies on brain function in psychiatric illnesses have been restricted to the availablemethods and minimal invasiveness, most work on brain dysfunction has been performed inCSF fluid and post-mortem tissue. Only since the development of PET it has become possibleto evaluate brain function in patients.

5-hydroxy-indole-acetic acid in cerebrospinal fluid (CSF):

Most investigators have observed decreased CSF levels in depressed patients when comparedto normal controls. This observation was thought to indicate decreased serotonin metabolismin the brain of depressed patients (Van Praag and Korf. 1971; Van Praag et al. 1970; Sjostrom1973; Goodwin et al.1977; Bowers 1976). More recent studies, however, have been unable toobserve similar relations, and found more evidence for low CSF 5-HIAA levels to be relatedto violence, and suicide, rather than depression (Faustman et al. 1991).


21

Post-mortem brain tissue:

Monoamine contentReports on 5-HT levels measured in post-mortem brain tissue has been somewhat conflicting.Whereas most studies indicate that 5-HIAA levels (and hence metabolic breakdown) are lowerin raphe nuclei of depressed patients (Meltzer et al. 1984; Beskow et al. 1976), conflictingresults exist on 5-HIAA and 5-HT content in other parts of the brain (Cheetham et al.1989).However, post-mortem artefacts and therapy induced changes in content may havecomplicated the results.

5-HT1A receptorsResults on 5-HT1A receptors in post-mortem tissue has been observed to be inconsistent.Several studies finding no change in receptors of people who committed suicide (Stockmeieret al. 1997; Lowther et al. 1997; Arranz et al. 1994; Matsubara et al. 1991; Dillon et al. 1991;Yates and Ferrier 1990). However, others have reported increased 5-HT1A receptors incortical areas.(Arrango et al. 1995; Joyce et al.1993). In addition, elevated numbers of 5-HT1A

receptors were also observed in raphe regions of suicides with major depression (Stockmeieret al. 1998).

5-HT2 receptorsSeveral studies have shown that 5-HT2 receptor binding sites are increased in frontal cortex ofdepressed patients (Arora and Meltzer 1989; Arango et al. 1992). The decreased receptordensity, which was observed in a different study, might be related to the use of antidepressantswhich reduce 5-HT2 receptor binding (Cheetham et al. 1988).

Serotonin-Reuptake[3H]Imipramine binding as an index of serotonin re-uptake-site density was observed to bereduced in brain tissue of depressed patients unrelated to the cause of death (Stanley et al.1982; Langer et al. 1981; Langer et al. 1981). However, regional dependent changes inimipramine binding were also observed (Gross-isseroff et al. 1989).

PET studies:

Analysis of the binding potential of the 5-HT1A antagonist WAY 100635 in depressed patientsusing PET suprisingly showed decreased 5-HT1A receptors in several brain areas (amongwhich frontal cortex and raphe regions). Treatment with SSRIs did not change bindingpotential of central 5-HT1A receptors (Drevets et al. 1999; Sargent et al. 2000).

Chapter 1

22

3.2 The mechanism of action of selective serotonin re-uptake inhibitors

Serotonin re-uptake inhibitors enhance serotonin levels by inhibiting re-uptake of serotoninwhich is released by serotonergic neurons. Brain extracellular serotonin levels has beenfrequently shown to be enhanced upon acute administration of SSRIs in animals (Fuller.1994). In addition, it was found that this effect was dependent on the dose of the SSRIadministered (Hjorth et al. 1997). However, the clinical onset of action of these substances indepressed patients is typically delayed for 2 to 4 weeks. In addition to the absence of anydirect correlation between the apparent serotonin enhancing capacities of SSRIs, noconclusive relationship between plasma levels and therapeutic response has ever beenobserved. However, clinical effective plasma levels have been identified (Devoto et al. 1992).This apparent differentiation between the acute and chronic effects of SSRIs has beenhypothesised to be due to the induction of several pharmacological changes in the brain uponchronic treatment with SSRIs. Several changes in brain function have been observed in animalsand humans upon chronic treatment (Baker and Greenshaw 1989). Although this thesisprimarily focuses on the function of serotonergic autoreceptors like 5-HT1A, 1B and 1D) severalother changes in brain pharmacology have also been observed and could, of course, influencethe therapy of depression and anxiety. For instance the sensitivity of several serotonergicreceptors (5-HT4, 5-HT7) has been shown to be decreased upon chronic treatment withantidepressants (Sleight et al. 1995; Bijak et al.1997). Additionally, chronic treatment ofanimals with SSRIs has also been shown to induce subsensitivity of β-adrenoceptors, a featurewhich was formerly believed to be restricted to the response of norepinephrinergicantidepressants (Petersen and Mork. 1996). Also changes in glutamatergic, as well as GABA-ergic neurotransmission have been observed after treatment with SSRIs. Whereas changes inglycine functionality on the NMDA receptor complex were observed for glutamatergicneurotransmission, decreased binding on GABAB receptors were observed when the GABA-ergic system was studied (Creese and Sibley 1981; Lloyd et al. 1985; Pilc and Lloyd 1984).Changes in serotonergic autoreceptor function in time has been hypothesised to be related tothe onset of action of antidepressant treatment with SSRIs. As mentioned before, serotonergicneurotransmission is controlled by auto-inhibitory mechanisms, which, to some extent, all aresubject to gradual adaptation upon chronic treatment with SSRIs.5-HT1A receptors have been observed to be desensitised upon chronic treatment with SSRIs inseveral studies, though other researchers have failed to show these effects, thus questioningthe robustness of this effect (Le Poul et al. 1995; Hjorth and Auerbach 1994; Bosker et al.1995; Invernizzi et al. 1994). The desensitisation was neither reflected in decreased numbersof receptors nor was it due to decreased affinity of the receptors for 5-HT1A ligands (Goodwin1996; Maura and Raiteri 1984), indicating some dysfunction of the signal-transmission of thereceptor.Desensitisation of 5-HT1B and 1D receptor has also been associated with the onset of action ofSSRIs. However, only few studies have shown these effects, though most researchers have notbeen able to observe any changes in function of these receptors when studied after chronictreatment (Auerbach and Hjorth 1995; Bosker et al. 1995; Bosker et al.1995; Gobbi et al.1997; Davidson and Stamford 1997; Chaput et al. 1986). Similar to the absence of changes in


23

ex vivo receptor parameters in 5-HT1A receptors, no changes were observed in 5-HT1B affinityand amount of receptors after treatment (Gobbi et al. 1997). As most studies have beenperformed after cessation of SSRI treatment, observation of desensitisation of serotonergic 5-HT1B/D receptors might have been hindered by rapid resensitisation, as was hypothesised in arecent study (Anthony et al. 2000).In addition to desensitisation of serotonergic autoreceptors during chronic treatment asdescribed above, several authors have observed desensitisation of the re-uptake site ofserotonin during chronic treatment (Pineyro et al. 1994). This observation, in combinationwith the changes in autoreceptor function as described above, has prompted severalresearchers to correlate the onset of action of SSRIs to decreased autorestraining function ofthe serotonergic system as a consequence of chronic SSRI treatment. Along this line,enhancing serotonergic neurotransmission above SSRI induced serotonergic elevation wouldincrease therapeutic efficacy as well as accelerate to onset of action.

Chapter 1

24

References

Abrahamowski (1995). Localisation of the 5-hydroxytryptophane2C receptor protein inhuman and rat brain using specific anitsera. Neuropharmacology, 34, 1635Anthony, J.P., Sexton, T.J. and Neumaier, J.F. (2000). Antidepressant-induced regulation of5-HT1B mRNA in rat dorsal raphe nucleus reverses rapidly after drug discontinuation. Journalof neuroscience research, 61, 82-87.Arango, V., Underwood, M.D. and Mann, J.J. (1992). Alterations in monoamine receptors inthe brain of suicide victims. J Clin Psychopharmacol, 12, 8-12.Arora, R.C. and Meltzer, H.Y. (1989). Increased serotonin 2 (5-HT2) receptor binding asmeasured by 3H-lysergic acid diethylamide (3H-LSD) in the blood platelets of depressedpatients. Life Sci, 44, 725-734.Arrango, V., Underwood, M.D., Gubbi, A.V. and Mann, J.J. (1995). Localised alterations inpre- and postsynaptic serotonin receptor sites in the ventrolateral prefrontal cortex of suicidevictims. Brain Research, 688, 121-133.Arranz, B., Eriksson, A., Mellerup, E., Plenge, E. and Maracusson, P. (1994). Brain 5-HT1A, 5-HT1D and 5-HT2 receptors in suicide victims. Biological Psychiatry, 35, 457-463.Auerbach, S.B. and Hjorth, S. (1995). Effect of chronic administration of the selectiveserotonin (5-HT) uptake inhibitor citalopram on extracellular 5-HT and apparent autoreceptorsensitivity in rat forebrain in vivo. Naunyn Schmiedebergs Arch Pharmacol, 352, 597-606.Baker, G.B.and Greenshaw, A.J. (1989). Effects of long-term administration ofantidepressants and neuroleptics on receptors in the central nervous system. Cell MolNeurobiol, 9, 1-44.Barnes, N.M., Cheng, C.H., Costall, B., Ge, J. and Naylor, R.J. (1992). Differentialmodulation of extracellular levels of 5-hydroxytryptamine in the rat frontal cortex by (R)- and(S)-zacopride. Br J Pharmacol, 107, 233-239.Barnes, N.M. and Sharp, T. (1999). A review of central 5-HT receptors and their function.Neuropharmacology, 38, 1083-1152.Beskow, J., Gottfries, C.G., Roos, B.E. and Winblad, D.B. (1976). Dtermination of mono-amine and mono-amine-metabolites in the human brain: Postmortem studies in a group ofsuicides and a control group. Acta Psychiatr Scand Suppl, 53, 7Bijak, M., Tokarski, K. and Maj, J. (1997). Repeated treatment with antidepressant drugsinduces subsensitivity to the excitatory effect of 5-HT4 receptor activation in the rathippocampus. Naunyn-Schmiedenberg's Arch Pharmacology, 355, 14-19.Bosker, F.J., Cremers, T.I.F.H., Jongsma, M.E., Westerink, B.H., Wikstrom, H.V., and denBoer, J.A. (2001). Acute and chronic effects of citalopram on postsynaptic 5-hydroxytryptamine (1A) receptor mediated feedback: a microdialysis study in the amygdala.Journal of Neurochem, 76, 1645-1653.Bosker, F.J., de Winter, T.Y., Klompmakers, A.A. and Westenberg, H.G. (1996). Flesinoxandose-dependently reduces extracellular 5-hydroxytryptamine (5-HT) in rat median raphe anddorsal hippocampus through activation of 5-HT1A receptors. Journal of Neurochem, 66,2546-2555.


25

Bosker, F.J., Klompmakers, A.A. and Westenberg, H.G. (1995). Effects of single and repeatedoral administration of fluvoxamine on extracellular serotonin in the median raphe nucleus anddorsal hippocampus of the rat. Neuropharmacology, 34, 501-508.Bosker, F.J., Klompmakers, A.A. and Westenberg, H.G. (1997). Post-synaptic 5-HT1Areceptors mediate 5-hydroxytryptamine release in the amygdala through a feedback to thecaudal linear raphe. Eur J Pharmacol, 333, 147-157.Bosker, F.J., van Esseveldt, K.E., Klompmakers, A.A. and Westenberg, H.G. (1995). Chronictreatment with fluvoxamine by osmotic minipumps fails to induce persistent functional changesin central 5-HT1A and 5-HT1B receptors, as measured by in vivo microdialysis in dorsalhippocampus of conscious rats. Psychopharmacology Berl, 117, 358-363.Bosker, F.J., Vrinten, D., Klompmakers, A.A. and Westenberg, H.G. (1997). The effects of a5-HT1A receptor agonist and antagonist on the 5-hydroxytryptamine release in the centralnucleus of the amygdala: a microdialysis study with flesinoxan and Way 100635. NaunynSchmiedebergs Arch Pharmacol, 355, 347-353.Bowers, M.B. (1976). Lumbar CSF 5-hydroxyindoleacetic acid and homovanillic acid inaffective syndromes. J Nerv Ment Dis, 158, 325Burns (1997). Regulation of serotonin-2C receptor G-protein coupling by RNA editing.Nature, 387, 303Casanovas, J.M., Hervas, I. and Artigas, F. (1999). Post-synaptic 5-HT1A receptors control5-HT release in the rat medial prefrontal cortex. Neuroreport, 10, 1441-1445.Ceci, A., Baschirotto, A. and Borsini, F. (1994). The inhibitory effect of 8-OH-DPAT on thefiring activity of dorsal raphe serotonergic neurons in rats is attenuated by lesion of theprefrontal cortex. Neuropharmacology, 33, 709-713.Chaput, Y., de Montigny, C. and Blier, P. (1986). Effects of a selective 5-HT reuptakeblocker, citalopram, on the sensitivity of 5-HT autoreceptors: electrophysiological studies inthe rat brain. Naunyn Schmiedebergs Arch Pharmacol, 333, 342-348.Charney, D.S. and Delgado, P.L. (1992). Current concepts of the role of serotonin function indepression and anxiety. In S.Z. Langer, N. Brunello, G. Racagni, & J. Mendlewicz (Eds.),Serotonin receptor subtypes: pharmacological significance and clinical implication. (pp. 89-104). Basel: Karger.Charney, D.S., Heninger, G.R. and Reinhard, J.F. (1982). The effect of i.v. L-tryptophane onprolactin, growth hormone and mood in healthy subjects. Psychopharmacology, 78, 38Charney, D.S., Heninger, G.R. and Sternberg, D.E. (1984). Serotonin function andmechanism of action of antidepressant treatment. Archives of General Psychiatry, 41, 359Cheetham, S.C., Crompton, M.R., Czudek, C., Horton, R.W., Katona, C.L. and Reynolds,G.P. (1988). Brain 5-HT2 receptor binding sites in depressed suicide victims. Brain Res, 443,272-280.Cheetham, S.C., Crompton, M.R., Czudek, C., Horton, R.W., Katona, C.L. and Reynolds,G.P. (1989). Serotonin concentrations and turnover in brains of depressed suicides. BrainResearch, 502, 332-340.

Chapter 1

26

Chen, N.H and Reith, M.E. (1995). Monoamine interactions measured by microdialysis in theventral tegmental area of rats treated systemically with (+/-)-8-hydroxy-2-(di-n-propylamino)tetralin. J Neurochem, 64, 1585-1597.Coccaro, E.F., Siever, L.J., Klar, H.M., Maurer, G., Cochrane, K., Cooper, T.B., Mohs, R.C.and Davis, K.L. (1989). Serotonergic studies in patients with affective and personalitydisorders: correlates with suicidal and impulsive aggressive behavior. Arch Gen Psychiatry, 46,587Creese, I., and Sibley, D.R. (1981). Receptor adaptations to centrally acting drugs. Annu RevPharmacol Toxicol, 21, 357-391.Daniel P.M., Moorhouse S.R. and Pratt O.E. (1976). Amino acid precursors of monoamineneurotransmitters and some factors influencing their supply to the brain. Psychol. Med., 6(2),277-286.Davidson, C. and Stamford, J.A. (1997). Chronic paroxetine desensitises 5-HT1D but not 5-HT1B autoreceptors in rat lateral geniculate nucleus. Brain Res, 760, 238-242.de Boer, T.H., Nefkens, F., van Helvoirt, A. and van Delft, A.M. (1996). Differences inmodulation of noradrenergic and serotonergic transmission by the alpha-2 adrenoceptorantagonists, mirtazapine, mianserin and idazoxan. J Pharmacol Exp Ther, 277, 852-860.Deakin, J.F.W., Pennell, I., Upadhyaya, A.J. and Lofthouse, R. (1990). A neuro-endocrinestudy of 5-HT function in depression: evidence for biological mechanisms of endogenous andpsychosocial causation. Psychopharmacology, 101, 85Devoto, P., De Montis, G. and Tagliamonte, A. (1992). Failure by tricyclic antidepressants toaffect the increase of dopamine extracellular concentrations produced by haloperidol in thecaudate and accumbens nuclei of rats. Biochem Pharmacol, 43, 507-511.Dillon, K.A., Gross-isseroff, R., Israeli, M. and Biegon, A. (1991). Autoradiographic analysisof serotonin 5-HT1A receptor binding in the human brain postmortem: effects of age andalcohol. Brain Res, 554, 56-64.Drevets, W.C., Frank, J., Price, J.C., Kupfer, D.J., Holt, D., Greer, P.J., Huang, Y., Gautier,C. and Mathis, C. (1999). PET imaging of serotonin 1A receptor binding in depression. BiolPsychiatry, 46, 1375-1387.Duxon (1997). Evidence for the expression of 5-HT2B receptor protein in the rat centralnervous system. Neuroscience, 76, 323el Mansari, M. and Blier, P. (1996). Functional characterization of 5-HT1D autoreceptors onthe modulation of 5-HT release in guinea-pig mesencephalic raphe, hippocampus and frontalcortex. Br J Pharmacol, 118, 681-689.Faustman, W.O., King, R.J., Faul, K.F., Moses, J.A., Brenson, K.L., Zarcone, V.P., andCsernansky, J.G. (1991). MMPI measures of impulsivity and depression correlate with CSF 5-HIAA and HVA in depression but not schizophrenia. Journal of Affective Disorders, 22,235-239.Ferre, S. and Artigas, F. (1993). Dopamine D2 receptor-mediated regulation of serotoninextracellular concentration in the dorsal raphe nucleus of freely moving rats. J Neurochem, 61,772-775.


27

Ferre, S., Cortes, R. and Artigas, F. (1994). Dopaminergic regulation of the serotonergicraphe-striatal pathway: microdialysis studies in freely moving rats. Journal of Neuroscience,14, 4839-4846.Frances, A., Pincus, H.A. and First, M.B. (1994). Diagnostic and Statistical Mqanual ofMental Disorders. Washington: American Psychiatric Association.Fuller, R.W. (1994). Uptake inhibitors increase extracellular serotonin concentration measuredby brain microdialysis. Life Sci, 55, 163-167.Ge, J. and Barnes, N.M. (1996). 5-HT4 receptor-mediated modulation of 5-HT release in therat hippocampus in vivo. Br J Pharmacol, 117, 1475-1480.Gobbi, M., Crespi, D., Foddi, M.C., Fracasso, C., Mancini, L., Parotti, L. and Mennini, T.(1997). Effects of chronic treatment with fluoxetine and citalopram on 5-HT uptake, 5-HT1Bautoreceptors, 5-HT3 and 5-HT4 receptors in rats. Naunyn Schmiedebergs Arch Pharmacol,356, 22-28.Goodwin, G.M. (1996). How do antidepressants affect serotonin receptors? The role ofserotonin receptors in the therapeutic and side effect profile of the SSRIs. J Clin Psychiatry, 57Suppl 4, 9-13.Goodwin, G.M., Rubovits, R. and Wehr, T.A. (1977). Serotonin and norepinephrine subgroupsin depression. Sci Proc Am Psychiatric Assoc, 130, 108Gross-isseroff, R., Israeli, M. and Biegon, A. (1989). Autoradiographic analysis of tritiatedimipramine binding in the human brain post-mortem: effect of suicide. Arch Gen Psychiatry,46, 237Haj-Dahmane (2001). D2-like dopamine receptor activation excites rat dorsal raphe 5-HTneurons in vitro. Eur. J. Neurosci. 14(1): 125-34Hjorth, S. and Auerbach, S.B. (1994). Lack of 5-HT1A autoreceptor desensitization followingchronic citalopram treatment, as determined by in vivo microdialysis. Neuropharmacology, 33,331-334.Hjorth, S., Bengtsson, H.J., Milano, S., Lundberg, J.F. and Sharp, T. (1995). Studies on therole of 5-HT1A autoreceptors and alpha 1-adrenoceptors in the inhibition of 5-HT release--I.BMY7378 and prazosin. Neuropharmacology, 34, 615-620.Hjorth, S. and Tao, R. (1991). The putative 5-HT1B receptor agonist CP-93,129 suppressesrat hippocampal 5-HT release in vivo: comparison with RU 24969. Eur J Pharmacol, 209,249-252.Hjorth, S., Westlin, D. and Bengtsson, H.J. (1997). WAY100635-induced augmentation ofthe 5-HT-elevating action of citalopram: relative importance of the dose of the 5-HT1A(auto)receptor blocker versus that of the 5-HT reuptake inhibitor. Neuropharmacology, 36,461-465.Hoyer (1994). International union of pharmacology classification of receptors for 5-hydroxytryptamine (serotonin). Pharmacological Reviews, 46, 157Invernizzi, R., Bramante, M. and Samanin, R. (1994). Chronic treatment with citalopramfacilitates the effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT1A receptors. Eur J Pharmacol, 260, 243-246.

Chapter 1

28

Joyce, J.N., Shane, A., Lexow, N., Winokur, A., Casanova, M.F. and Kleinman, J.E. (1993).Serotonin uptake sites and serotonin receptors are altered in the limbic system ofschizophrenics. Neuropsychopharmacology, 8, 315-336.Kennett, G.A. (1993). 5-HT1C receptors and their therapeutic relevance. Curr Opin InvestDrugs, 2, 317Kennett, G.A. (1994). In vivo properties of SB 200646A, a 5-HT2C/2B receptor antagonist.Br J Pharmacol, 111, 797Kennett, G.A. (1996). The effect of the 5-HT2B receptor agonist, BW 723C86, on threemodels of anxiety. Br J Pharmacol, 117,Kennet, G.A. (1998). Serotonin Receptors and their function.www.tocris.com ed.Koyama, T., Lowy, M.T. and Meltzer, H.Y. (1987). 5-hydroxytryptophane induced cortisolresponse and CSF 5-HIAA in depressed patients. American Journal of Psychiatry, 144, 334Langer, S.Z., Javoy-Agid, F., Raisman, R., Briley, M. and Aid Y. (1981). Distribution ofspecific high affinity binding sites for (3H) imipramine in human brain. Journal of Neurochem,37, 267Langer, S.Z., Zarifian, E., Briley, M. Raisman R. and Sechter D. (1981). High affinitybinding of (3H) imipramine in brain and platelets and its relevance to the biochemistry ofaffective disorders. Life Science, 29(3), 211Le Poul, E., Laaris, N., Doucet, E., Laporte, A.M., Hamon, M. and Lanfumey, L. (1995).Early desensitization of somato-dendritic 5-HT1A autoreceptors in rats treated with fluoxetineor paroxetine. Naunyn Schmiedebergs Arch Pharmacol, 352, 141-148.Lesch, K.P., Mayer, S., Disselkamp-Tietze, Hoh, A., Weismann, M., Osterheide, M. andSchulte, M. (1990). 5-HT1A receptor responsivity in unipolar depression: evaluation ofipsapirone-induced ACTH and cortisol secretion in patients and controls. Biol Psychiatry, 28,620Lloyd, K.G., Thuret, F. and Pilc, A. (1985). Upregulation of gamma-aminobutyric acid(GABA) B binding sites in rat frontal cortex: a common action of repeated administration ofdifferent classes of antidepressants and electroshock. Journal of Pharmacol Exp Ther, 235,191-199.Lowther, S., De Paermentier, F., Cheetham, S.C., Crompton, M.R., Katona, C.L.E. andHorton, R.W. (1997). 5-HT1A receptor binding sites in post-mortem brain samples fromdepressed suicides and controls. J Affect Disorders, 43, 199-207.Maes, M., Jacobs, M., Suy, E., MInner, B., Leclercq, C., Christiaens, F. and Raus, J. (1990).Suppressant effect of dexamethasone on the availability of plasma L-tryptophane and tyrosinein healthy controls and in depressed patients. Acta Psychiatria Scandinavia , 81, 19-23.Martin, K.F., Hannon, S., Phillips, I. And Heal, D.J. (1992). Opposing roles for 5-HT1B and5-HT3 receptors in the control of 5-HT release in rat hippocampus in vivo. Br J Pharmacol,106, 139-142.Matsubara, S., Arora, R.C.and Meltzer H.Y. (1991). Serotonergic measures in suicide brain :5-Ht1A binding sites in frontal cortex of suicide victims. Journal of Neural Transm, 85, 181-194.


29

Maura, G. and Raiteri, M. (1984). Functional evidence that chronic drugs induce adaptivechanges of central autoreceptors regulating serotonin release. European Journal ofPharmacology, 97, 309-313.McQuade, R. and Sharp, T. (1997). Functional mapping of dorsal and median raphe 5-hydroxytryptamine pathways in forebrain of the rat using microdialysis. J Neurochem, 69,791-796.Meltzer, H.Y. and Arora, R.C. (1991). Platelet serotonin studies in affective disorders:evidence for a serotonergic abnormality. In M. Sandler, A. Coppen, & S. Harnett (Eds.),5hydroxytryptamine in psychiatry: a spectrum of ideas. (pp. 50-89).Meltzer, H.Y., Lowy, M.T., Robertson, A., Goodnick, P., Perline, R. (1984). Effect of 5-hydroxytryptophane on serum cortisol levels in major affective disorders: III. Effect ofantidepressants and lithium carbonate. Archives of Gen. Psychiatry, 41, 391Mockus S.M. and Vrana K.E.( 1998). Advances in the molecular characterization oftryptophane hydroxylase. J. Mol. Neurosci., 10 (3), 163-179.Moret, C. and Briley, M. (1997). 5-HT autoreceptors in the regulation of 5-HT release fromguinea pig raphe nucleus and hypothalamus. Neuropharmacology, 36, 1713-1723.O'Keane, V., McLoughlin, D. and Dinan, T.G. (1992). D-fenfluramine-induced prolactin andcortisol release in major depression: response to treatment. Journal of Affect Disorders, 26,143-150.Petersen, B. and Mork, A. (1996). Chronic treatment with citalopram induces noradrenalinereceptor hypoactivity. A microdialysis study. Eur J Pharmacol, 300, 67-70.Phebus, et al. (1996). Characterisation of LY 334370, a potent and selective 5-HT1Freceptor agonist, in the neurogenic dural inflammation model of migraine pain. AM SocNeurosc, 22, 528.11(abstract)Pilc, A. and Lloyd, K.G. (1984). Chronic antidepressants and GABA "B" receptors: a GABAhypothesis of antidepressant drug action. Life Sci, 35, 2149-2154.Pineyro, G., Blier, P., Dennis, T. and de Montigny, C. (1994). Desensitization of the neuronal5-HT carrier following its long-term blockade. J Neurosci, 14, 3036-3047.Sargent, P.A., Kjaer, K.H., Bench, C.J., Rabiner, E.A., Messa, C., Meyer, J., Gunn, R.N.,Grasby, P.M. and Cowen, P.J. (2000). Brain serotonin1A receptor binding measured bypositron emission tomography with [11C] WAY 100635. Archives of Gen Psychiatry, 57,174-180.Sarrias, M.J., Artigas, F., Marinez, E., Gelpi, E., Alvarez, E., Udina, C. and Casa, M.(1987). Decreased plasma-serotonin in melancholic patients: a study with clomipramine. BiolPsychiatry, 22, 1429-1438.Sharma, H.S. and Dey, P.K. (1986). Probable involvement of 5-hydroxytryptamine inincreased permeability of blood-brain barrier under heat stress in young rats.Neuropharmacology, 25, 161-167.Sharma, H.S. and Dey, P.K. (1987). Influence of long-term acute heat exposure on regionalblood-brain barrier permeability, cerebral blood flow and 5-HT level in consciousnormotensive young rats. Brain Res, 424, 153-162.

Chapter 1

30

Sharma, H.S., Olsson, Y. and Dey, P.K. (1990). Changes in blood-brain barrier and cerebralblood flow following elevation of circulating serotonin level in anesthetized rats. Brain Res,517, 215-223.Sharp, T., Bramwell, S.R. and Grahame Smith, D.G. (1989). 5-HT1 agonists reduce 5-hydroxytryptamine release in rat hippocampus in vivo as determined by brain microdialysis. BrJ Pharmacol, 96, 283-290.Siever, L.J., Murphy, D.L., Slater, S., de la Vega, E., Lipper S. (1984). Plasma prolactinchanges following fenfluramine in depressed patients compared to controls: an evaluation ofcentral serotonergic responsivity in depression. Life Sci, 34, 1029Sjostrom, R. (1973). 5-hydroxy-indole acetic acid and homovanillic acid in cerebrospinal fluidin manic-depressive psychosis and the effect of probenecid treatment. Eur J Clin Pharmacol,6, 75Sleight, A.J., Carolo, C., Petit, N., Zwingelstein, C. and Bourson, A. (1995). Identification of5-hydroxytryptamine7 receptor binding sites in rat hypothalamus: sensitivity to chronicantidepressant treatment. Mol Pharmacol, 47, 99-103.Sprouse, J., Reynolds, L. and Rollema, H. (1997). Do 5-HT1B/1D autoreceptors modulatedorsal raphe cell firing? In vivo electrophysiological studies in guinea pigs with GR127935.Neuropharmacology, 36, 559-567.Stanley, M., Virgilio, J. and Gershon, S. (1982). Tritiated imipramine binding sites aredecreased in the frontal cortex of suicides. Science, 216, 1337Starkey, S.J. and Skingle, M. (1994). 5-HT1D as well as 5-HT1A autoreceptors modulate HTrelease in the guinea-pig dorsal raphe nucleus. Neuropharmacology, 33, 393-402.Stockmeier, C.A., Dilley, G.E., Shapiro, S.A., Overholser, J.C., Thompson, P.A. and Meltzer,H.Y. (1997). Serotonin receptors in suicide victims with major depression.Neuropsychopharmacology, 16, 162-173.Stockmeier, C.A., Shapiro, S.A., Dilley, G.E., Kolly, T.N., Friedman, L. and Rajkowska, G.(1998). Increase in serotonin-1A autoreceptors in midbrain of suicide victims with majordepression-postmortem evidence for decreased serotonergic activity. J Neurosci, 18, 7394-7401.Tao and Hjorth (1992). Alpha-2 adrenoceptor modulation of rat ventral hippocampal 5-hydroxytryptamine release in vivo. Naunyn Schmiedebergs Arch Pharmacol. 1992,345(2):137-43.Tao, R., Ma, Z. and Auerbach, S.B. (1996). Differential regulation of 5-hydroxytryptaminerelease by GABAA and GABAB receptors in midbrain raphe nuclei and forebrain of rats. Br JPharmacol, 119, 1375-1384.Van Praag, H.M. and Korf, J. (1971). Endogenous depression with and without disturbancesin the 5-hydroxytryptamine metabolism: a biochemical classification? psychopharmacology,19, 148Van Praag, H.M., Korf, J. and Puite, J. (1970). 5-hydroxy-indole-acetic acid levels incerebropsinal fluid of depressive patients treated with probenecid. Nature, 225, 1259Wright, I.K., Garratt, J.C. and Marsden, C.A. (1990). Effects of a selective 5-HT2 agonist,DOI, on 5-HT neuronal firing in the dorsal raphe nucleus and 5-HT release and metabolism inthe frontal cortex. Br J Pharmacol, 99, 221-222.


31

Wurch, T., Palmier, C., Colpaert, F.C. and Pauwels, P.J. (1997). Sequence and functionalanalysis of cloned guinea pig and rat serotonin 5-HT1D receptors: common pharmacologicalfeatures within the 5-HT1D receptor subfamily. Journal of Neurochemistry, 68, 410-418.Yates, M. and Ferrier, I.M. (1990). 5-HT1A receptors in major depression. Journal ofPsychopharmacology, 4, 69-74.Young, S.N., Smith, S.E., Pihl, R. and Ervin, F.R. (1985). Tryptophane depletion causes arapid lowering of mood in normal males. Psychopharmacology, 87, 173-177.

Chapter 1

32

Clinical efficacy of antidepressants in depression

33

Chapter 2


Chapter 2

34

Introduction

HistoryOnce the therapeutic effects of iproniazid and imipramine in major depression wererecognized, pharmaceutical companies intensified research programs to develop superiorantidepressants. Progress has undeniably been made with respect to side effects and safety. Weare only on the verge, however, of dealing with issues such as efficacy and onset of action.The first class of antidepressants, to which iproniazid and pargyline belong, increasenorepinephrine, dopamine, and serotonin levels in the brain by interfering with their metabolicpathways through inhibition of the mitochondrial enzyme monoamine oxidase (MAO). MAOinhibitors (MAOIs) are effective in major depression, although they do have pronounced sideeffects. Moreover, their mechanism of action is prone to hazardous interactions thatoccasionally lead to hypertensive crises, serotonin syndrome, and severe intoxications.The second class of antidepressants, tricyclic antidepressants (TCAs) such as imipramine,block the reuptake of monoamines. TCAs are effective antidepressants and interactions withother exogenous compounds are less severe than with the MAOIs. Nevertheless, manyunwanted side effects have been reported, most of which are related to the complexpharmacological profile of TCAs. The most serious unwanted side effect is that TCAs are notsafe in overdose, which may lead to life-threatening situations.Several classes of antidepressants have evolved from the classic tricyclic antidepressants,which may be superior in terms of safety profile and lack of side effects. With regard toefficacy, however, little progress has been made. All antidepressants have a delayed onset ofaction (2-6 weeks) and the percentage of nonresponders is relatively large (30-40%). Researchis therefore increasingly focused on improving efficacy and onset of action. In order to reachthis goal, several pharmacological concepts have been investigated over the last decades. Thepresent chapter discusses these concepts, which range from single- to multiple-actionstrategies.

Confounding factors in antidepressant research

Subclassification and comorbidityThe clinical evaluation of antidepressant drugs is hampered by the multiplicity of the diseaseitself. Depression is an ‘umbrella’ term for several clusters of clinical symptoms. Great efforthas been made to develop a structural subclassification system, e.g. DSM-IV. Diagnosticcriteria may be too coarse, however, to guarantee that a coherent group of patients is enteringthe clinical studies.An extra confounding factor is comorbidity with anxiety and other psychiatric disorders, whichblurs the boundaries of the illness and further complicates rational treatment. Better insightinto the biological mechanisms underlying the different subforms of depression and comorbiddisease will hopefully help to improve the diagnostic criteria.


35

Response and onset of actionEvaluation of antidepressant activity is also complicated by differences in design and scoringmethods between clinical studies. Rating scales, such as the Hamilton Rating Scale forDepression (HAM-D), may improve objectivity but substantial differences between open andcontrolled trials have often been reported. In addition, although objective scoring is improvedby rating scales, differences in scoring between investigators enhances noise, which in turnincreases the statistical power needed to show differences in efficacy when antidepressantdrugs are compared.

Co-medicationIdeally, a drug’s therapeutic effect is measured in a biological matrix that is free of interferingexogenous substances. In practice, depressed patients are seldom drug-naive. An extensivewashout period certainly helps, but pharmacological interference cannot be excludedcompletely; this phenomenon is sometimes neglected in the overall outcome analysis.Moreover, participants in clinical trials often receive additional medication to relieve thesymptoms of comorbid disease. Benzodiazepines are likely to interfere with the pharmacologyof an antidepressant, which complicates its clinical evaluation considerably. Althoughundesirable, this situation is sometimes inevitable.

Etiology, gender, genetics, and pharmacokineticsSeveral other factors may influence the pharmacology of antidepressants. Differences inetiology, gender, genetics and pharmacokinetics potentially complicate the evaluation process,but are often neglected or not examined due to low statistical power. Confounding factors likethese will certainly gain increased attention in future studies.

Chapter 2

36

Antidepressant drugs

Classical antidepressantsMAOIs and TCAs belong to the first generation of antidepressant drugs. Severalaugmentation strategies to improve the therapeutic efficacy of these drugs have beenattempted in the past, but these strategies are beyond the scope of this chapter and willtherefore not be discussed here.

1. Selective compounds: SSRIs

Several lines of evidence indicate an involvement of serotonin in depression. The notion thatLSD, a potent serotonin receptor antagonist, could induce mood changes in humans gave riseto the idea that serotonin could be involved in the etiology of mood disorders (Schildkraut1965, Woolley and Shaw 1963, Gaddum 1963). This idea was strengthened by the observationthat reserpine, a monoamine depleter, was able to induce depressive symptoms in humans(Carlsson et al. 1957). The effectiveness of MAOIs and TCAs in the treatment of depressiongave further support for a role of serotonin in depression. Finally, this insight has resulted inthe development of a new generation of antidepressants, the selective serotonin reuptakeinhibitors (SSRIs, structures 1).

SSRIs and the serotonergic systemThe serotonergic innervation of the brain mainly originates in the dorsal raphe nucleus (DRN)and the median raphe nucleus (MRN). These nuclei innervate a variety of structures within thebrain, with a topical organization with respect to several brain areas. Whereas the prefrontalcortex is mainly innervated by the DRN, the dorsal hippocampus is mostly innervated by theMRN.SSRIs increase extracellular levels of serotonin immediately upon administration (Fuller 1994).Yet their therapeutic effect is typically delayed for several weeks. This apparent discrepancymay be explained as follows. At least two types of serotonin (5-HT) autoreceptor are foundon the serotonergic neuron. 5-HT1A receptors are present in the somatodendritic area;activation of these receptors decreases neuronal firing, which results in less serotonin beingreleased from the axon terminals. 5-HT1B receptors are located on the terminals of serotoninneurons; when they are activated, serotonin release is directly inhibited. There is growingevidence that postsynaptic 5-HT1A receptors are also involved in the control of serotoninrelease, through a large feedback loop from terminal to cell body region (Bosker et al. 1997).It is very likely that these autorestraining processes counteract the initial effect of SSRIs.Following chronic administration of SSRIs it has been shown that at least the 5-HT1A

receptors (presynaptically as well as postsynaptically) desensitize (Cremers et al. 2000).Arguably, this adaptive process enhances the effect of the SSRI on serotonergicneurotransmission. This cascade of events may partly explain the delayed onset of action ofSSRIs.


37

SSRIs and pharmacokineticsThe pharmacokinetics of SSRIs are generally well documented, although no correlation hasbeen found between plasma levels and the antidepressant effect. Clinically effective plasmalevels have been described, however (Baumann and Rochat 1995).The SSRIs form a heterogeneous group in terms of affinity for liver enzymes. Citalopram doesnot influence the metabolism of other drugs, while other SSRIs have prominent interactions(Sproule et al. 1997).

Clinical trials: SSRI versus TCA and SSRI versus SSRIMany studies have been performed to evaluate the efficacy of SSRIs in the treatment ofvarious subtypes of depression. An elaborate discussion of individual studies is beyond thescope of this chapter, however.Several meta-analyses of short-term comparative studies have investigated the efficacy ofSSRIs compared with TCAs in major depression (Anderson 1998, Anderson and Tomenson1995, Bech and Cialdella 1992, Bech 1989, Mendlewicz 1992, Montgomery et al. 1994, Songet al. 1993, Steffens et al. 1997, Tignol et al. 1992). The majority of these papers did notobserve differences in efficacy between the two classes of antidepressant drugs. In one meta-analysis (Anderson 1998) it was found that some TCAs, in particular amitriptyline, may bemore effective than SSRIs in depressed patients. A study of combined fluvoxamine data

N

O

O

O

H

OF3C

NCH3

Structure 1.1 Paroxetine Structure 1.2 Fluoxetine

O

F

N

N

NH

H

CH3

Cl

ClStructure 1.3 Citalopram Structure 1.4 Sertraline

Chapter 2

38

showed that the response rate with this antidepressant was comparable to that seen withtricyclics and tetracyclics (Mendlewicz 1992).Most short-term, controlled, comparative studies on efficacy did not reveal differences amongthe various SSRIs in the treatment of major depression (Aguglia et al.1993, Bennie et al.1995,De Wilde et al.1993, Eskelius et al.1997, Franchini et al. 1997,Geretsegger et al.1994,Haffmans et al.1996, Kiev and Feiger 1997, Patris et al.1996, Rapaport et al.1996, Tignol1993). Some studies, however, did find indications for an earlier onset of action of citalopramand paroxetine compared with fluoxetine (De Wilde et al.1993, Geretsegger et al 1994, Patriset al. 1996, Tignol 1993).

Tolerability and long-term efficacySeveral studies have analyzed the adverse-effect profile of SSRIs. In addition togastrointestinal effects, headaches and ‘stimulant adverse effects’ like agitation, anxiety, andinsomnia were frequently reported (Baldwin and Johnson 1995, Boyer and Blumhardt 1992,Cooper 1988, Doogan 1991, Wagner et al.1994).Since depression is a chronic recurrent condition, long-term treatment is often required inorder to minimize the chance of relapse. Several studies have been performed on relapseduring placebo treatment, after initial successful treatment with SSRIs. All studies show thatchronic treatment with SSRIs was superior with respect to reappearance of depression and thetime to relapse. No evidence was found for differences between SSRIs or between SSRIs andTCAs in terms of relapse features (Anderson and Tomenson 1995, Doogan and Caillard 1992,Franchini et al.1997, Keller et al.1998a, Keller et al.1998b, Montgomery and Dunbar 1993a,Montgomery et al. 1993b, Montgomery and Kasper 1995a, Robert and Montgomery 1995,Rush et al.1998).Although the long-term efficacies of SSRIs and TCAs are similar, the tolerance of SSRIs isclearly superior to TCAs (Song et al.1993, Montgomery et al. 1994, Anderson and Tomenson1995, Anderson 1998, Meanna et al. 1992).

Augmentation with 5-HT1A receptor antagonistsSSRIs are believed to exert their antidepressant effect partly through desensitization of 5-HT1A

receptors (Blier et al.1987). Following this line of thought, it was hypothesized that co-administration with a 5-HT1A receptor antagonist would instantaneously mimic thisdesensitization, thereby enhancing efficacy and reducing the lag time of the antidepressanttreatment (Artigas 1993). Since selective 5-HT1A receptor antagonists were not available forclinical use, investigators have used the mixed ß-adrenoceptor/5-HT1A receptor antagonistpindolol in order to test this hypothesis (stucture 1.5).


39

Evaluation of this hypothesis has yielded mixed results depending on the setup of the clinicaltrial.The first trials on the effect of co-administration of pindolol with SSRIs in patients sufferingfrom major depression have all been performed in an open and not placebo-controlled manner.All open trials on co-administration of pindolol with several different SSRIs have reportedpositive effects on latency to improvement and effectiveness in refractory depression (table 1).In line with the hypothesis, no augmenting effects were observed when pindolol was co-administered with tricyclic antidepressant drugs devoid of serotonin reuptake inhibitoryproperties.In contrast with these preliminary studies, double-blind placebo controlled trials have yieldedcontroversial results (table 2). When increased efficacy was evaluated versus placebo based onreductions in HAM-D scores or respons rate at end-point, only 3 out of 9 studies were insupport of enhanced efficacy. A decreased latency in therapeutic respons was only observed in5 out of 9 cases, whereas no superiority was observed in treatment resistance.

N

O

OHN

Structure 1.5 Pindolol

Chapter 2

40

SSRI Dose SSRI Dose pindolol Number

of

subjects

Treatment

resistance (TR)

or treatment (T)

Outcome

(response)

Reference

Paroxetine 20 mg/day 2.5 mg t.i.d. 7 T 5/7 Artigas et al.

1994

Paroxetine 20 mg/day 2.5 mg t.i.d. 9 T 7/9 Blier and

Bergeron 1995

Paroxetine 20 mg/day 2.5 mg t.i.d. 3 T 1/3 Blier et al. 1997

Paroxetine 20-40

mg/day

2.5 mg t.i.d. 3 TR 3/3 Artigas et al.

1994

Paroxetine 20-40

Mg/day

2.5 mg t.i.d. 8 TR ∆-HAMD

32-10

Blier and

Bergeron 1995

Paroxetine 20-40

mg/day

2.5 mg t.i.d. 3 TR 1/3 Dinan and Scott

1996

Fluoxetine 20-40

mg/day

2.5 mg t.i.d. 3 TR ∆-HAMD

29-11

Blier and

Bergeron 1995

Fluoxetine 20-60

mg/day

2.5 mg t.i.d. 6 TR 1/6 Dinan and Scott

1996

Fluvoxamine 50-100

mg/day

2.5 mg t.i.d. 8 T 0/8 Blier et al. 1997

Fluvoxamine 200 mg/day 2.5 mg t.i.d. 1 TR 1/1 Artigas et al.

1994

Sertraline 50-100

mg/day

2.5 mg t.i.d. 5 TR 0/5 Blier and

Bergeron 1995

Sertraline 50-100

mg/day

2.5 mg t.i.d. 4 TR ¼ Dinan and Scott

1996

Trimipamine 75-100

mg/day

2.5 mg t.i.d. 5 T 0/5 Blier et al. 1997

Moclobemide 900 mg/day 2.5 mg t.i.d. 2 TR 0/2 Blier and

Bergeron 1995

Nefazodone 100 mg/day 2.5 mg t.i.d. 20 T 15/20 Barkish et al.

1997

Phenelzine 60 mg/day 2.5 mg t.i.d. 1 TR 1/1 Artigas et al.

1994

Imipramine 200 mg/day 2.5 mg t.i.d. 1 TR 1/1 Artigas et al.

1994

Table 1. Efficacy of pindolol co-administration in open trials


41

SSRI Dose SSRI Dose pindolol Duration

of study

(weeks)

Number

of

subjects

Treatment

resistance

(TR) or

treatment (T)

Increased

efficacy

Decrease

d latency

Reference

Fluoxetine 40 mg/day 2.5 mg t.i.d. 1.5 28 TR - - Perez et

al. 1999

Fluoxetine 20-40

mg/day

2.5 mg t.i.d. 2 8 T - - Moreno et

al. 1997

Fluoxetine 20 mg/day 2.5 mg t.i.d. 5 21 30 % TR + - Maes et

al. 1999

Fluoxetine 20 mg/day 2.5 mg t.i.d. 6 86 50 % TR - - Berman et

al. 1997,

1999

Fluoxetine 20 mg/day 2.5 mg t.i.d. 6 111 T + + Perez et

al. 1997

Paroxetine 20 mg/day 2.5 mg t.i.d. 6 80 T - + Tome et

al. 1997

Paroxetine 20 mg/day 2.5 mg t.i.d. 4 63 T + + Zanardi et

al. 1997

Paroxetine 20 mg/day 5 mg t.i.d. 4 100 T - + Bordet et

al. 1998

Paroxetine 40 mg/day 2.5 mg t.i.d. 1.5 20 TR - - Perez et

al. 1999

Fluvoxam

ine

200 mg/day 2.5 mg t.i.d. 1.5 6 TR - - Perez et

al. 1999

Clomipra

mine

150 mg/day 2.5 mg t.i.d. 1.5 26 TR - - Perez et

al. 1999

Table 2. Efficacy of pindolol co-administration in double-blind placebo-controlled clinicaltrials

Although several trials have observed beneficial effects of co-administration of pindolol withSSRIs, there is increasing evidence that these effects may not be due to antagonism ofsomatodendritic 5-HT1A receptors by pindolol. Several preclinical studies indicate thatpindolol has agonistic properties at 5-HT1A receptors in the raphe nuclei in vivo. Moreover,plasma levels of pindolol are low and the drug does not penetrate the brain easily due to itspolar character. It is therefore questionable whether the dose of pindolol used in clinicalstudies is sufficient to mimic desensitization of 5-HT1A autoreceptors. More studies withspecific and potent 5-HT1A receptor antagonists are definitely needed to support thisinteresting concept.

Chapter 2

42

Augmentation with 5-HT1A receptor agonists5-HT1A receptor partial agonists have weak antidepressant properties. Preclinical studies haveshown that 5-HT1A receptor agonists induce rapid desensitization of 5-HT1A receptors (Kreissand Lucki 1997). Theoretically, co-administration with a 5-HT1A receptor partial agonist mayimprove the efficacy and lag time of an SSRI, depending on the size and rate of desensitizationinduced by the 5-HT1A receptor partial agonist.Clinical evaluation of this concept has shown beneficial effects, although the evidence is notconclusive (Bouwer and Stein 1997, Jacobsen 1991, Joffe and Schuller 1993, Harvey andBallon 1995, Landen et al. 1998). More studies are needed to explore this interesting butsomewhat paradoxical idea.

2. Selective norepinephrine reuptake inhibitors (reboxetine)

As with serotonin, a role for norepinephrine in depression was suggested by the depressogenicfeatures of reserpine. Abundant evidence is present linking dysfunction of thenorepinephrinergic system to depression (Leonard 1997b). In addition, several TCAs are alsovery potent norepinephrine reuptake inhibitors.The idea that inhibition of norepinephrine uptake, alongside serotonin reuptake inhibition,could be beneficial in treating depression prompted the development of selectivenorepinephrine reuptake inhibitors, of which reboxetine is currently the only marketed drug(structure 2.1).

Reboxetine and the noradrenergic systemThe noradrenergic system originates mainly in the locus coeruleus. α2-autoreceptors arelocated on both axon terminals and cell bodies, thus establishing an effective self-regulationsystem similar to that seen in the serotonergic neuron.Post-mortem studies of the frontal cortex of suicide victims revealed that both the density andaffinity of these receptors were increased (Callado et al. 1998, Meanna et al. 1992). Inaddition, α2- adrenoceptors may become supersensitive during depression (Charney et al.1981, Spyraki and Fibiger 1980).Although chronic administration of desimipramine has been shown to effectively reduce thesupersensitivity of α2 -adrenoceptors, a recent preclinical study with reboxetine failed todemonstrate changes in receptor function following chronic treatment (Charney et al.1981,Sacchetti et al. 1999).

β-adrenoceptors are located postsynaptically. Upregulation of these receptors has consistentlybeen observed in patients with depression, whereas downregulation of these receptors isregarded as a marker for antidepressant activity (Leonard 1997a).

The relevance of α2- and β-adrenoceptor downregulation/desensitization for reboxetine’santidepressant effect has yet to be established.


43

Pharmacokinetics of reboxetineReboxetine is administered on a daily basis at a dose ranging from 8 to 20 mg. Due to itsrelative short half-life, the daily dose must be divided over the day. The bioavailability ofreboxetine is more than 60%, and it has no active metabolites (Table 3)

Dose

(mg/day)

Bioavailabil

ity

Protein

binding

Vd (l/kg) Half-life

range (h)

Steady state

(ng/ml)

Active

metabolites

Reboxetine 8-20 >60 97 0.5 12-16 50-160 No

Venlafaxine 75-225 92 27 2-23 2-11 50-150 Yes

Milnacipran 50-200 85 13 5.3 8 No

Trazodone 200-600 60-80 89-95 5-9 700-1600 Yes

Nefazodone 50-600 >20 99 0.2-1 2-8 150-1000 Yes

Mirtazapine 15-45 50 85 4.5 13-34 20-40 No

Table3: Pharmacokinetic summary of newer antidepressants. Data from (Aguglia et al. 1993,Davis et al. 1997, Landsay de Vane 1998, Reboxetine package insert 1997, Spencer and Wilde1998, Venlafaxine package insert 1999).

Reboxetine versus TCAsSeveral trials have investigated the efficacy of reboxetine compared with TCAs. In one short-term study, reboxetine was found to be at least as effective as imipramine (Montgomery1997). Analysis of pooled data from four double-blind outpatient studies also showed nodifferences between reboxetine and imipramine (Massana and Moller 1998a). Another study,comparing imipramine and reboxetine in depressed and dysthymic elderly patients, reportedbetter efficacy of reboxetine in dysthymic patients but not in depressed patients (Dubini et al.1997, Katona et al. 1999).Only one short-term study has compared reboxetine with desimipramine. Equal or superioractivity of reboxetine was observed (Montgomery 1997).

N

O

H

H

OO

CH3

Structure 2.1 Reboxetine

Chapter 2

44

Reboxetine versus SSRIsFluoxetine is the only SSRI that has been compared with reboxetine. Short-term evaluationhas shown that the efficacy of reboxetine is similar to fluoxetine (Massana 1998b,Montgomery 1997). Pooling of four double-blind comparison studies, however, revealedincreased efficacy of reboxetine compared with fluoxetine in depressed outpatients imipramine(Massana and Moller 1998a).Subset analysis on severe depression in several trials showed that reboxetine was superior tofluoxetine (Massana 1998, Montgomery 1999a).

Tolerability and long-term efficacyReboxetine has been shown to be well tolerated in short-term studies. Adverse events, whichhave been more frequently observed in reboxetine- versus placebo-treated patients, were drymouth, constipation, insomnia, increased sweating, tachycardia, vertigo, urinary hesitancyand/or retention, and impotence (Mucci 1997).Comparison of reboxetine with imipramine and desimipramine revealed a beneficial profilewith respect to a number of common side effects like hypotension, dry mouth, and tremor(Ban et al. 1998, Berzewski et al. 1997). When reboxetine was compared with fluoxetine,patients were observed to be less likely to experience ‘stimulant adverse effects’ as well asgastrointestinal effects (Mucci 1999).Reboxetine was shown to be effective in the long-term treatment of depression, given itssuperior efficacy in a 1-year placebo-controlled study (Montgomery 1997).


45

Multiple action compounds

RationaleAnderson and colleagues have compared the efficacy of SSRIs and TCAs. No overalldifferences were found between the groups, although amitriptyline was superior in efficacycompared with SSRIs (Anderson 1998, Anderson and Tomenson 1994). Similar results werefound when the SSRI fluoxetine was compared with nortriptyline (Roose et al.1994). Basedon the high affinity of these compounds for both the serotonin and norepinephrine reuptakesite, it was speculated that compounds acting on both serotonin and norepinephrine would besuperior to single-action antidepressants such as the SSRIs and reboxetine (see Chapter 5).This idea has prompted the development of a group of dual-action antidepressants, which canbe subdivided as follows:

I) dual-action reuptake inhibitors (venlafaxine, milnacipran, nomifensin*, and Bupropion*)II) receptor antagonist/reuptake inhibitors (trazodone and nefazodone)III) heterocyclics with no affinity for the re-uptake site (mirtazapine).

* This chapter is focusing on serotonin and norepinephrine. The dopamine reuptake inhibitorsnomifensine and bupropion will therefore not be discussed here.

3.1 Serotonin-norepinephrine reuptake inhibitors: venlafaxine

Pharmacology of venlafaxineVenlafaxine has considerable affinity for both the norepinephrine and serotonin reuptake sites,and is therefore called a serotonin-norepinephrine reuptake inhibitor (SNRI)(structure 3.1).The mechanism of action may be deduced from the SSRIs and reboxetine, insofar asvenlafaxine can desensitize serotonin autoreceptors and may induce rapid β- adrenoceptordesensitization. In contrast to the classical TCAs, venlafaxine does not bind to histaminergic,cholinergic, or adrenergic receptors, thereby avoiding side effects such as hypotension andsedation.

Pharmacokinetics of venlafaxineVenlafaxine is marketed as regular venlafaxine (venlafaxine-IR). Due to the fast elimination ofvenlafaxine (elimination half-life of four hours), it is also available in an extended-releaseformulation (venlafaxine-XR) that enables venlafaxine to be administered once daily.Venlafaxine does not substantially inhibit CYP2C9, 2D6, 1A2, 3A3, and 3A4, and is thereforenot likely to interfere with the metabolism of other drugs (Ereshefsky 1996).Venlafaxine has an ascending dose-response curve (Preskorn 1994). This observation isconsistent with preclinical data demonstrating that norepinephrine uptake inhibition is seen atdoses exceeding those for serotonin reuptake inhibition. Clinical observations also strengthenthis idea. Norepinephrine uptake inhibition becomes relevant at doses above 150 mg/day, andclinical superiority with respect to SSRIs is apparent at doses from 225 mg/day (Muth et al.1986, Clerc et al. 1994).

Chapter 2

46

Venlafaxine versus TCAsVenlafaxine (25-75 mg twice daily) has been compared with imipramine in a double-blind 13-week study in outpatients with mild-to-moderate intensity (Lecrubrier et al. 1997). Comparedwith imipramine, venlafaxine-treated patients showed greater reductions in Montgomery-Asberg Depression Rating Scale (MADRS) total scores from week 4 until the end of thestudy, showing a significant difference in favor of venlafaxine at week 4. Additionally,response rates and clinical global impression (CGI) indices over time were superior forvenlafaxine recipients compared with imipramine recipients.Venlafaxine-XR has been compared with TCAs in a meta-analysis (Einarson et al. 1999).Therapeutic success was defined as a 50% decrease in HAM-D or MADRS scores. Using thisapproach, the authors showed that venlafaxine-XR was significantly superior to TCAs.

Venlafaxine versus SSRIsSeveral trials have been conducted comparing venlafaxine and venlafaxine-XR with SSRIs. Anumber of these studies indicate that venlafaxine could be superior to SSRIs.In a 6-week study comparing venlafaxine with fluoxetine in severely depressed inpatients,superior efficacy was observed for venlafaxine (Clerc et al 1994), while an 8-week study alsorevealed an enhanced response in patients on venlafaxine compared with fluoxetine (Dierick etal. 1995). Another study, however, did not find that the efficacy of venlafaxine was superior tofluoxetine in depressive outpatients (Costa e Silva 1998).Venlafaxine-XR enhances tolerability and ease of administration. A study investigating theefficacy of venlafaxine-XR versus fluoxetine suggests that venlafaxine-XR brings aboutgreater remission than fluoxetine. In addition, CGI indices were higher in venlafaxine- versusfluoxetine-treated patients (Rudolph and Derivan 1997). Additionally, a meta-analysis hasshown that venlafaxine-XR was superior to SSRIs as a group when success rates based oncriteria of a 50% decrease in HAM-D and MADRS scores were evaluated (Einarson et al.1999).

Tolerability and long-term efficacyAs venlafaxine is a very potent inhibitor of serotonin reuptake, it shares several adverse eventsin common with the SSRIs. The most common adverse events are nausea, dizziness, andsomnolence. Just like SSRIs, venlafaxine can induce sexual dysfunction after several weeks of

OH

N

O

Structure 3.1 Venlafaxine


47

treatment. When the daily dose of venlafaxine exceeds 225 mg, adverse events may be inducedthat differ from SSRIs (e.g. hypertension, sweating, and tremors).

Chapter 2

48

3.2 Noradrenaline/serotonin reuptake inhibitors: milnacipran

Pharmacology of milnacipranLike venlafaxine, milnacipran blocks norepinephrine reuptake sites in addition to serotoninreuptake sites (structure 3.2). Arguably, its mechanism of action is comparable to that ofvenlafaxine. Differences in efficacy between milnacipran and venlafaxine may be attributed todifferent pharmacokinetics.

Pharmacokinetics of milnacipranThe bioavailability of milnacipran after oral administration is around 85%. The variation inplasma levels between patients is low, and steady-state conditions are realized with a singledaily dose. No active metabolites have been found in humans (Table 3).

Milnacipran versus TCAs

Several clinical trials have compared milnacipran with amitriptyline, imipramine, andclomipramine.Milnacipran was found to be inferior or equal to amitriptyline, depending on the milnaciprandose used in the trials. At a dose of 50 mg/day, milnacipran was clearly inferior toamitriptyline in terms of onset of action and efficacy after 4 weeks. At a dose of 100 mg/day,onset of action was still inferior but the efficacy after four weeks was comparable toamitriptyline (Ansseau et al. 1989a). Evaluation was based on reductions in MADRS andHDRS (Hamilton Depression Rating Scale) scores, as well as CGI efficacy scores. In a latertrial that compared milnacipran 200 mg/day with amitriptyline, CGI efficacy scores weresignificantly greater in milnacipran-treated patients than in amitriptyline-treated patients(Ansseau et al. 1989b).Several studies have compared the efficacy of imipramine, clomipramine, and milnacipran inmajor depression (Clerc et al.1990, Kasper et al. 1996, Matsubara et al. 1996, Puech et al.1997, Steen and den Boer 1997, Yamashita et al. 1995). Despite the suboptimal dose ofmilnacipran (50 mg/day), pooled data from these studies did not reveal significant differencesbetween milnacipran and the two tricyclic antidepressants. HDRS and/or MADRS scores wereused to assess antidepressant efficacy (Kasper et al. 1996).

NH2

NO

Figure 3.2 Milnacipran


49

In a Japanese trial, treatment with milnacipran (50-150 mg/day) was superior to imipramineafter one week of treatment, although no differences were observed at the end of the study(week 4) (Yamashita et al. 1995, Matsubara et al. 1996). Analysis was based on the followingscales: HDRS, CGI, and CPRGDRSDUC (Clinical Psychopharmacology Research GroupDepression Rating Scale for Doctor’s Use). The efficacy of milnacipran (50 mg/day) versusimipramine in elderly people was not significantly different in terms of HDRS, MADRSscores, and other indices in an 8-week trial (Tignol et al 1998).The efficacies of milnacipran and clomipramine were comparable in several trials, albeit with aslight advantage demonstrated for clomipramine. In one study, where milnacipran (100mg/day) was compared with clomipramine in major depressive patients, no differences wereobserved at any time during 3 weeks of treatment (Clerc et al. 1990). Two trials, each lasting26 weeks, have yielded controversial results. In the first study, no clinical efficacy of eithermilnacipran or clomipramine could be demonstrated, probably due to the high withdrawalrates and the inclusion of therapy-resistant patients (Steen and den Boer 1997). The secondtrial, however, clearly favored clomipramine in terms of end-point efficacy, onset of action,and response in severely depressed patients. Interestingly, both studies used an optimal dose ofmilnacipran 200 mg/day. Significant differences were found in the latter study in terms of theHDRS scores. MADRS and CGI scores were slightly in favor of clomipramine but thedifferences were not significant (Leinonen 1997).

Milnacipran versus SSRIsMilnacipran has been compared with fluoxetine and fluvoxamine in several clinical trials(Ansseau et al. 1994, Ansseau et al. 1991,Guelfi et al. 1997, Lopez-Ibor et al. 1996No differences were observed at the end-point (12 weeks) when milnacipran (50 and 100 mgtwice daily) was compared with fluoxetine, in terms of HDRS, MADRS, and CGI scores.Milnacipran recipients responded more rapidly than fluoxetine recipients, however, asmeasured by reductions in HDRS and MADRS scores at week 4 (Guelfi et al. 1997, Lopez-Ibor 1996). In another study, fluoxetine was superior to milnacipran (100mg once daily) after6 weeks in outpatients suffering from major depression. Reductions in MADRS, HDRS, andCGI scores were significantly different between fluoxetine and milnacipran recipients. Sincemilnacipran was dosed once daily in this study, and the usual dosage is twice daily,pharmacokinetics may explain the unfavorable results seen with milnacipran (Ansseau et al.1994).A comparison of fluvoxamine with milnacipran (75-100 twice daily), either with or without aloading dose, did not produce any significant differences in reductions of MADRS, HDRS, orCGI scores after 4 weeks in major depressive inpatients (Ansseau et al. 1991). Milnacipran (50mg twice daily) produced larger reductions in MADRS but not in HDRS scores whencompared with fluvoxamine (Lopez-Ibor 1996).When data from two comparative studies of milnacipran (50 mg twice daily) versus eitherfluoxetine or fluvoxamine were pooled, they indicated significantly better results formilnacipran than for the SSRIs. Reductions in HDRS and MADRS scores were significantlygreater for milnacipran (Lopez-Ibor 1996).

Chapter 2

50

Tolerability and long-term efficacyMilnacipran was associated with a higher incidence of vertigo, sweating, anxiety, hot flushes,and dysuria when compared with placebo. When the adverse effects of TCAs were comparedwith those of milnacipran, patients on TCAs reported a significantly greater number of eventsof dry mouth, constipation, tremor, sweating, somnolence, tiredness, and vertigo. Only nauseawas observed more frequently in SSRI-treated patients than in milnacipran-treated patients(Puech et al. 1997).


51

4.1 Mixed serotonin antagonist/reuptake inhibitors: trazodone

Pharmacology of trazodoneTrazodone is a potent 5-HT2A and 5-HT2C receptor antagonist (structure 4.1). In addition, it isa weak inhibitor of serotonin reuptake (Table 2). Its metabolite, mCPP, is an agonist at 5-HT2C receptors but an antagonist at 5-HT2A receptors, thus it partly counteracts the effect oftrazodone (Fiorella et al. 1995). Taking into account its pharmacokinetic profile, theantidepressant effect of trazodone is most likely related to serotonin reuptake inhibition.Nevertheless, antagonism of 5-HT2 receptors may be beneficial, since these receptors areupregulated in suicide patients with a history of depression (Mann et al. 1986).

5-HT NE DA α1 α2 H HT2a HT1a Musc

Reboxetine 10702 82 >10,0003 10,0002 43,0002 14002 62502 -- 39002

Venlafaxine 392 2132 28002 39,9212 >100,0002 12,9092 >100,0002 >100,0002 29,9662

Milnacipran 111 441 >100001 >10,0001 >10,0001 >10,0001 >10,0001 >10,0001 >10,0001

Trazodone 1583 83003 71003 122 1062 292 202 422 12,1882

Nefazodone 2003 3573 3573 62 842 302 72 522 4,5692

Mirtazapine 4,7623 >10,0003 >10,0003 5003 1403 0.13 163 7143 6663

Table 2. In-vitro affinities for receptors and synaptisome reuptake inhibition. Data are IC501,

Ki2, and Kd

3 values (nM). Data from (Davis et al.1997, Vanlafaxine package insert 1999,Montgomery 1999b, Owens et al. 1997, Spencer and Wilde 1998, Moret et al. 1985, Taylor etal. 1995).5-HT; potency on serotonergic re-uptake inhibition. NE ; potency on norepinephrine re-uptake inhibition. DA ;

potency on dopaminergic re-uptake inhibition. α1&2, H, HT2a, HT1A and Musc.; potencies on alpha adrenergic

1 and 2, histaminergic, serotonin 2A and 1A and muscarinergic receptors.

Pharmacokinetics of trazodoneTrazodone has a short elimination half-life. The daily dose of 200-600 mg must therefore bedivided over the day (Table 1).

N N

NO

N N

Cl

Figure 4.1 Trazodone

Chapter 2

52

Trazodone versus TCAsA meta-analysis has compared the efficacy of trazodone with imipramine. Using HDRS ratingscales, no differences were found between the two drugs (Workmann and Short 1993).Moreover, a review of 25 clinical studies on the efficacy of trazodone in depression hasreported equal efficacy for imipramine and trazodone Patten 1992). Additionally, other TCAsdid not differ significantly in terms of efficacy when compared with trazodone (Marek et al.1992).

Trazodone versus SSRIsTrazodone has good antidepressant properties compared with SSRIs, but it has no effect inpanic disorder and OCD (Obsessive Compulsive Disorder) (Marek et al. 1992, Pigott et al.1992).

Tolerability and long-term efficacyDizziness, drowsiness, dry mouth, and nausea are the most frequently reported adverse effectswhen trazodone was compared with placebo.


53

4.2 Mixed serotonin antagonist/reuptake inhibitors: nefazodone

Pharmacology of nefazodoneNefazodone is a potent 5-HT2A receptor antagonist with moderate antagonism at the α1 –adrenoceptors (structure 4.2). It has a considerable affinity for serotonin and norepinephrinereuptake sites too, but these properties were recognized much later (Table 4). Nefazodonemay be superior to TCAs in combining serotonin and norepinephrine reuptake inhibition with5-HT2A receptor antagonism. Upregulation has been reported for 5-HT2A receptors in suicidevictims with a history of depression (Mann et al. 1986).

Pharmacokinetics of nefazodoneNefazodone is absorbed rapidly and completely after oral administration. Due to extensivefirst-pass metabolism, the bioavailability of nefazodone is approximately 20%. Nefazodone iswidely distributed throughout the body, including the central nervous system. The volume ofdistribution ranges from 0.22 to 0.87 l/kg. Nefazodone is metabolized into severalpharmacologically active metabolites such as hydroxynefazodone, mCPP, p-hydroxynefazodone, and a triazoledione metabolite (Davis et al. 1997). The daily dose ofnefazodone is 50-600 mg, given once or in two divided doses. As with milnacipran, it has anascending dose-response curve. Nefazodone’s efficacy at lower doses (300-400 mg/day) maybe comparable to that of the SSRIs, but it could exceed the latter’s efficacy at higher doses(500mg/day). Analysis of data in the literature has shown that higher doses of nefazodone (>500 mg/day) do not improve antidepressant efficacy (Dale Horst and Preskorn 1998).

Nefazodone versus TCAsNefazodone has been compared with imipramine and amitriptyline (Cohn et al. 1996, D’Amicoet al. 1990, Fawcett et al. 1995, Feighner et al. 1996, Fountaine et al. 1994, Marcus andMendels 1996, Mendels et al. 1995, van Moffaert et al. 1994).Several trials have reported a similar overall clinical efficacy for imipramine and nefazodone(Cohn et al. 1996, Fountaine et al. 1994, Rickels et al. 1994, Rickels et al. 1995). The efficacyof nefazodone appeared to be critically related to the dose. At high dosages (100-600

CH3

N

N N

O

N N

Cl

O

Figure 4.2 Nefazodone

Chapter 2

54

mg/day), nefazodone was as effective as imipramine, whereas low dosages (50-300 mg/day)were less efficacious (Fountaine et al. 1994, Mendels et al 1995, van Moffaert et al. 1994).Amitriptyline has shown superior efficacy compared with nefazodone in patients withmoderate-to-severe depression and in bipolar affective disorder (Ansseau et al. 1994b).

Nefazodone versus SSRIsNefazodone has been compared with paroxetine, sertraline, and fluoxetine in several clinicaltrials (Armitage et al 1996, Baldwin et al 1996, Cassano et al. 1993, Feiger et al. 1996, Riouxet al. 1996). No indications for differences in efficacy have been observed between nefazodoneand any of the SSRIs.

Tolerability and long-term efficacy The most frequently reported adverse events observed with nefazodone are nausea,somnolence, dry mouth, dizziness, constipation, asthenia, light-headedness, and blurred vision.Compared with TCAs, nefazodone was observed to induce less dry mouth, constipation, andblurred vision. No abnormal weight gain has been observed with nefazodone, although weightgain was reported for TCAs and trazodone.When nefazodone was compared with SSRIs, more activating symptoms (agitation, anxiety,tremor, insomnia, and nervousness), diarrhea, sweating, anorexia, nausea, and male sexualdysfunction were observed in SSRI versus nefazodone recipients. Dry mouth, dizziness,constipation, visual disturbances, and confusion were more common in nefazodone versusSSRI recipients (Davis et al. 1997).


55

5.1 Mixed serotonin/noradrenaline antagonist: mirtazapine

Pharmacology of mirtazapineIt is thought that the dual mode of action of mirtazapine operates via the blockade of α2-adrenoceptors in combination with 5-HT2 and 5-HT3 receptor antagonism (structure 5.1). Thismechanism of action refers to the interplay between the norepinephrine system and theserotonin system in the brain. The norepinephrine neuron carries α2-autoreceptors in theterminal as well as in the cell body region.

Antagonism of these receptors by mirtazapine has been shown to enhance norepinephrineneurotransmission. Noradrenaline interacts with the serotonergic system in two ways. First,α2-heteroceptors are present on the terminals of serotonin neurons, activation of whichattenuates serotonin release. Second, α1-heteroceptors are present in the cell body region ofthe serotonergic system, activation of which increases the firing rate of the serotonergicneurons, thus enhancing serotonin neurotransmission. Mirtazapine blocks presynaptic α2-adrenoceptors, which enhances noradrenaline release in the raphe nuclei. Arguably, α1-adrenoceptors on serotonin cell bodies in the raphe nuclei become increasingly activated,thereby enhancing the firing rate of the serotonergic neurons. Antagonism of 5-HT2 and 5-HT3

receptors has been hypothesized to further concentrate the enhanced serotoninneurotransmission towards 5-HT1A receptors.

Pharmacokinetics of mirtazapineThe pharmacokinetic parameters of patients have been summarized in Table 1. Based on therecommended daily dose of 15-45 mg, average steady-state plasma levels of mirtazapine rangebetween 30-300 nM. Since protein binding of mirtazapine is approximately 85%, freemirtazapine levels will be between 4.5 and 45 nM. The free brain concentration of mirtazapinewill maximally range between 4.5 and 45 nM. Since mirtazapine is not a very potent α2-adrenoceptor antagonist (Ki ~ 100nM, Kd ~ 140 nM), binding of mirtazapine to thesereceptors will not be substantial at clinical doses. The latter argument may urge for areconsideration of the proposed mechanism.

NN

NCH3

Structure 5.1 Mirtazapine

Chapter 2

56

Mirtazapine does not inhibit CYP450 enzymes and is therefore not expected to interact withthe metabolism of other drugs (Table 3).

1A2 2C 2D6 3A4

Reboxetine 0 0 0 0

Venlafaxine 0 0 0 (+) 0

Milnacipran 0 0 0 0

Trazodone 0 0 0 ++++

Nefazodone 0 0 0 ++++

Mirtazapine 0 0 0 0

Table 5: Relative inhibition of CYP450. Data from (Spencer and Wilde 1998, Lindsay deVane 1998).

Clinical trials: mirtazapine versus TCAsComparisons of mirtazapine with tricyclic antidepressants have mainly focused onamitriptyline, imipramine, clomipramine, and doxepine. In a review by Davis and Wilde (Davisand Wilde 1996), it was reported that mirtazapine has a similar efficacy to TCAs in bothoutpatients and inpatients with major depression.Several trials comparing mirtazapine with amitriptyline have observed similar HDRS responserates between the two compounds (Bremner 1995, Mullin et al. 1996, Smith et al 1990,Zivkov and de Jongh 1995). Concurrent meta-analyses have confirmed the equal efficacy ofboth drugs (Stahl et al. 1997, Fawcett and Barkin 1998, Kasper et al 1997, Kasper 1995).Although efficacy was observed to be similar, relapse rates were lower and sustainedremission rates were higher for mirtazapine in a placebo-controlled continuation trial(Montgomery et al 1998).Mirtazapine has been compared with imipramine in inpatients in a study that claimed asuperior response rate for imipramine (Bruijn et al . 1996). When the efficacy on severalsymptom clusters was investigated, mirtazapine appeared to be effective in clusters related toanxiety and sleep, whereas imipramine was effective on core symptoms of depression (Bruijnet al 1999)No indications for a different efficacy for mirtazapine and clomipramine were found in twodouble-blind randomized trials (Richou et al. 1995, Peyron 1996).Response rates at study end-point were similar in mirtazapine and doxepine recipients in adouble-blind multicenter trial (Martilla et al. 1995).

Clinical trials: mirtazapine versus SSRIsMirtazapine had a superior onset of action compared with fluoxetine. The reductions in HAM-D score were significantly greater at weeks 3 and 4. Response rates (>50% reduction inHAM-D score) were also significantly higher at week 4 (Wheatley et al. 1998).When compared with paroxetine, mirtazapine was significantly superior in terms of reductionin HDRS scores at various time points (Benkert et al. 1998). The decrease in HAM-D score


57

was higher for mirtazapine at week 1. Mirtazapine was also significantly superior to fluoxetineat weeks 1 and 4 when responder dates were evaluated.A multicenter, double-blind, randomized study in inpatients and outpatients with majordepressive disorder has compared mirtazapine with citalopram. The decrease in baselineMADRS scores was greater for mirtazapine than for citalopram at week 2. Additional scoringssuch as the Hamilton Rating Scale for Anxiety (HAM-A) and CGI also showed superiorreductions for mirtazapine versus citalopram at week 2. Response rates, as defined by CGI,were better for mirtazapine at weeks 1 and 2 (Agren et al. 1998).

Tolerability and long-term efficacyThe results of clinical trials suggest that patients are more likely to comply with mirtazapinethan with TCA treatment. An overview of the available data showed that amitriptyline wasinferior to mirtazapine in terms of long-term dropouts due to adverse events (Montgomery1995b).

Chapter 2

58

ConclusionsThe latest generation of antidepressants may have several advantages in comparison toMAOIs, TCAs, and SSRIs.

• Safety profile and interaction with the metabolism of other drugs are much improvedcompared with MAOIs and TCAs.

• Unfortunately, elimination rates (except for mirtazapine) have increased, making itnecessary to use multiple daily doses. By using extended-release formulations such asvenlafaxine-XR, however, the inconvenient multiple dosages can be avoided. On theother hand, rapid elimination is beneficial when patients change antidepressants.

• Selective norepinephrine reuptake inhibition (Reboxetine) may be superior to selectiveinhibition of the serotonin reuptake site, although enhanced efficacy with respect toTCAs has not been observed in the majority of comparative studies.

• Clinical studies have shown that some, but not all, of the multiple-actionantidepressants tend to have better short-term efficacy than TCAs and SSRIs (Table4).

• The mixed serotonin/norepinephrine reuptake inhibitor venlafaxine has shown superiorefficacy compared with TCAs and SSRIs in the majority of short-term trials.Unexpectedly, milnacipran, with almost the same pharmacological profile asvenlafaxine, was not found to be superior to the SSRIs and TCAs. The less convincingresults for milnacipran may be related to the pharmacokinetic properties of thecompound.

• Although not reviewed in this chapter, combined dopaminergic/noradrenergiccompounds (nomifensine and bupropion) do not seem to be superior to SSRIs andTCAs in depression. These drugs, although originally marketed for depression, haveboth been withdrawn as antidepressants and were therefore not evaluated extensively.

• Antidepressants with a mixed profile of Selective Serotonin Reuptake inhibition and 5-HT1A antagonism and agonism could have beneficial effects over SSRIs, though theirexact efficacy has to be evaluated with selective and potent ligands. Comparison versusTCA’s has not been performed yet.

• The mixed serotonin antagonist/reuptake inhibitors nefazodone and trazodone do nothave an advantage over regular treatment with TCAs and SSRIs. Trazodone may be aseffective, or slightly superior to, TCAs whereas nefazodone has equal or inferiorefficacy compared with TCAs. Compared with SSRIs, nefazodone was equallyeffective.

• Mirtazapine is comparable to the TCAs on short-term efficacy, but a majority ofstudies indicate that mirtazapine is superior to SSRIs.


59

TCA SSRI

TCA TCA = SSRI

SSRI TCA = SSRI

SSRI + 5-HT1A

antagonist

Combination > SSRI

SSRI + 5-HT1A

agonist

Combination > SSRI

Reboxetine TCA = Reboxetine Reboxetine > SSRI

Venlafaxine Venlafaxine > TCA Venlafaxine > SSRI

Milnacipran Milnacipran < TCA Milnacipran > SSRI

Trazodone Trazodone = TCA Trazodone = SSRI

Nefazodone Nefazodone < TCA Nefazodone = SSRI

Mirtazapine Mirtazapine = TCA Mirtazapine > SSRI

Table 4. Short-term comparative efficacy of new antidepressants versus tricyclicantidepressants and selective serotonin reuptake inhibitors based on reduction in HDRS andCGI.

In conclusion, it can be stated that the majority of new antidepressants tend to be superior toSSRIs in short-term trials. Compared with the TCAs, venlafaxine may be the only multiple-action compound with somewhat better antidepressant properties.Depression is a recurrent condition. Future trials should therefore evaluate the efficacy ofnewer antidepressants over a longer period of time (years), thereby evaluating patientcompliance as well as sustained antidepressant efficacy. In addition, patient sub-group analysisof trials would indicate the long- and short-term efficacies of antidepressants over thedepression sub-population. Only in this way can the clinical efficacy of the new antidepressantsbe evaluated thoroughly.

Chapter 2

60

References

Agren H, Leinonen E, Skarstein, J Helsdingen. Efficacy and tolerability of mirt azapine vs.citalopram: a double -blind, randomized study in patients with major depressive disorder[Abstract]. In: 37th Annual Meeting of the American College of Neuropsychopharmacology1998;311.Aguglia E, Casacchia M, Cassano GB, Faravelli C., Ferrari G., Giordano P., Pancheri P.,Ravizza L., Trabucchi M., Bolino F. .Double-blind study of the efficacy and safety ofsertraline versus fluoxetine in major depression. Int Clin Psychopharmacol 1993;8:197-202.Anderson IM, Tomenson BM. The efficacy of selective serotonin reuptake inhibitors indepression: a meta-analysis of studies against tricyclic antidepressants. J Psychopharmacology1994;8:238-49.Anderson IM, Tomenson BM. Treatment discontinuation with selective serotonin reuptakeinhibitors compared with tricyclic antidepressants: a meta-analysis. Br Med J 1995;310:1433-8.Anderson IM. SSRIs versus tricyclic antidepressants in depressed inpatients: a meta-analysisof efficacy and tolerability. Depress Anxiety 1998;7:11-7.Ansseau M, von Frenckell R, Mertens C, de Wilde J., Botte L., Devoitille J.M., Evrard J.L.,De Nayer A., Darimont P., Dejaiffe G. Controlled comparison of two doses of milnacipran(F2207) and amitriptyline in major depressive inpatients. Psychopharmacology Berl1989a;98:163-8.Ansseau M, von Frenckell R, Papart P et al. Controlled comparison of milnacipran (F2207)200 mg and amitriptyline in endogenous depressive inpatients. Hum Psychopharmacology1989b;4(3):221-7.Ansseau M, von Frenckell R, Gerard M, de Wilde J., Botte L., Devoitille J.M., Evrard J.L.,De Nayer A., Darimont P., Dejaiffe G., et al. Interest of a loading dose of milnacipran inendogenous depressive inpatients. Comparison with the standard regimen and withfluvoxamine. Eur Neuropsychopharmacol 1991;1:113-21.Ansseau M, Papart P, Troisfontaines B et al. Controlled comparison of milnacipran andfluoxetine in major depression. Psychopharmacology 1994a;114:131-7.Ansseau M, Darimont P, Lecoq A., De Nayer A., Evrard J.L., Kremer P., Devoitille J.M.,Dierick M., Mertens C., Messotten. Controlled comparison of nefazodone and amitriptyline inmajor depressive inpatients. Psychopharmacology 1994b;115:254-60.Armitage R, Rush AJ, Gillin JC, Yonkers K.A., Winokur A., Moldofsky H., Vogel G.W.,Kaplita S.B., Fleming J.B., Montplaisir J., Erman M.K., Albala B.J., Mcquade R.D.Comparative effects of nefazodone & fluoxetine on sleep in outpatients with major depressivedisorder. Biol Psychiatry 1996;39:623.Artigas F. 5-HT and antidepressants: new views from microdialysis studies. Trends PharmacolSci 1993;14(7):262Artigas F, Perez V, Alvarez E. Pindolol induces a rapid improvement of depressed patientstreated with serotonin reuptake inhibitors [letter] [see comments]. Arch Gen Psychiatry1994;51:248-51.


61

Baldwin DS, Johnson RN. Tolerability and safety of citalopram. Rev Contemp Pharmacother1995;6:315-25.Baldwin DS, Hawley CJ, Abed RT et al. A multicentre double-blind comparison ofnefazodone and paroxetine in the treatment of outpatients with moderate-to-severedepression. J Clin Psychiatry 1996;57(Suppl. 2):46-52.Ban TA, Gaszner P, Aguglia E et al. Clinical efficacy of reboxetine: a comparative study withdesimipramine- with methodological considerations. Hum Psychopharmacology 1998;13:529-39.Barkish D, Hooper CL, Thornton MD, Weins A, Miller CA, Thibaudeau CA. Fast onset: anopen study of the treatment of major depressive disorder with nefazodone and pindololcombination therapy. Int Clin Psychopharmacol 1997;12:91-7.Baumann P, Rochat B. Comparative pharmacokinetics of selective serotonin re-uptakeinhibitors: a look behind the mirror. Int Clin Psychopharmacol 1995;10(Suppl. 1):15-21.Bech P. Clinical effects of selective serotonin reuptake inhibitors. In: Dahl SG, Gram LF, eds.Clinical pharmacology in psychiatry. Berlin: Springer, 1989;81-93.Bech P, Cialdella P. Citalopram in depression: meta-analysis of intended and unintendedeffects. Int Clin Psychopharmacol 1992;6(suppl. 5):45-54.Benkert O, Szegedi A, Kohnen R. Rapid onset of action in major depression: a comparativetrial of mirtazapine and paroxetine. [Poster]. In: 37th Annual Meeting of the American Collegeof Neuropsychopharmacology 1998;Puerto Rico.Bennie EH, Mullin JM, Martindale JJ. A double-blind multicenter trial comparing sertralineand fluoxetine in outpatients with major depression. J Clin Psychiatry 1995;56:229-37.Berman RM, Darnell AM, Miller HL, Anand A, Charney DS. Effect of pindolol in hasteningresponse to fluoxetine in the treatment of major depression: a double-blind, placebo-controlledtrial. Am J Psychiatry 1997;154(1):37-43.Berman RM, Anand A, Cappiello A, Miller HL, Hu XS, Oren DA, Charney DS. The use ofpindolol with fluoxetine in the treatment of major depression: final results form a double-blind,placebo-controlled trial. Biol Psychiatry 1999;45:1170-17.Berzewski H, van Moffaert M, Gagiano CA. Efficacy and tolerability of reboxetine comparedwith imipramine in a double-blind study in patients suffering from major depressive episodes.Eur Neuropsychopharmacol 1997;7(Suppl. 1):S37-47.Blier P, de Montigny C, Chaput Y. Modifications of the serotonin system by antidepressanttreatments: implications for the therapeutic response in major depression. J ClinPsychopharmacol 1987;7(6 Suppl):24S-35S.Blier P, Bergeron R. Effectiveness of pindolol with selected antidepressant drugs in thetreatment of major depression. J Clin Psychopharmacol 1995;15:217-22.Blier P, Bergeron R, de Montigny C. Selective activation of postsynaptic 5-HT1A receptorsinduces rapid antidepressant response. Neuropsychopharmacology 1997;16(5):333-8.Bordet R, Thomas P, Dupuis B. Effect of pindolol on onset of action of paroxetine in thetreatment of major depression: intermediate analysis of a double-blind, placebo-controlledtrial. Reseau de Recherche et d'Experimentation Psychopharmacologique. Am J Psychiatry1998;155(10):1346-51.

Chapter 2

62

Bosker F, Vrinten D, Klompmakers A, Westenberg H. The effects of a 5-HT1A receptoragonist and antagonist on 5-hydroxytryptamine release in the central nucleus of the amygdala:a microdialysis study with flesinoxan and WAY 100635. Naunyn Schmiedebergs ArchPharmacol 1997a;355:347-53.Bouwer C, Stein DJ. Buspirone is an effective augmenting agent of serotonin selective re-uptake inhibitors in severe treatment refractory depression. S Afr Med J 1997;87(Suppl.4):534-7.Boyer WF, Blumhardt CL. The safety profile of paroxetine. J Clin Psychiatry1992;53(Suppl.):61-6.Bremner JD. A double blind placebo comparison of ORG 3770, amitryptiline, and placebo inmajor depression. J Clin Psychiatry 1995;(56):519-25.Bruijn JA, Moleman P, Mulder PGH. A double-blind fixed blood-level study comparingmirtazapine with imipramine in depressed in-patients. Psychopharmacology 1996;127:231-7.Bruijn JA, Moleman P, Mulder PGH, van den Broek WW. Depressed in-patients responddifferentially to imipramine and mirtazapine. Pharmacopsychiatry 1999;32(3):87-92.Callado LF, Meana JJ, Grijalba B et al. Selective increase of alpha 2A-adrenoceptor agonistbinding sites in brains of depressed suicide victims. J Neurochem 1998;70:1114-23.Carlsson A, Lindqvist M, Magnusson T. 3,4-dihydroxyphenylalanine and 5-hydroxytryptophane as reserpine antagonists. Nature 1957;180:1200.Cassano GB, Gammans RE, Zibellini M et al. A multicentre double-blind comparison ofnefazodone and fluoxetine in the treatment of patients with mood disorder, moderately toseverely depressed. Data on file 1993;Bristol-Myers Squibb(Protocol CN 104-052).Charney DS, Heninger GR, Sternberg DE et al. Presynaptic adrenergic sensitivity indepression: the effect of long-term desimipramine treatment. Arch Gen Psychiatry1981;38:1334-40.Clerc GE, Pagot R, Bouchard JM et al. Therapeutic contribution of milnacipran andclomipramine during a 3-month treatment: results of a comparative study. PsychiatrPsychobiol 1990;5(2):137-43.Clerc GE, Ruimy P, Verdeau-Pailles J. A double-blind comparison of venlafaxine andfluoxetine in patients hospitalized for major depression and melancholia. Int ClinPsychopharmacol 1994;9:139-43.Cohn CK, Robinson DS, Roberts DL et al. Responders to antidepressant drug treatment: astudy comparing nefazodone, imipramine, and placebo in patients with major depression. JClin Psychiatry 1996;57(Suppl. 2):15-8.Cooper GL. The safety of fluoxetine - an update. Br J Psychiatry 1988;15(Suppl. 3):77-86.Costa e Silva J. Randomized, double-blind comparison of venlafaxine and fluoxetine inoutpatients with major depression. J Clin Psychiatry 1998;59:352-7.Cremers TIFH, Spoelstra EN, de Boer P., Bosker F.J., Mork A., den Boer J.A., WesterinkB.H.C., Wikstrom H.V. Desensitisation of serotonergic autoreceptors uponpharmacokinetically monitored chronic treatment with citalopram. Eur J Pharmacol2000;397(2-3):351-7.D'Amico MF, Roberts DL, Robinson DS et al. Placebo-controlled dose-ranging trial designs inPhase II development of nefazodone. Psychopharm Bull 1990;26(1):147-50.


63

Dale Horst W, Preskorn SH. Mechanisms of action and clinical characteristics of three atypicalantidepressants: venlafaxine, nefazodone, bupropion. J Affect Disord 1998;51:237-54.Davis R, Wilde MI. Mirtazapine: a review of its pharmacology and therapeutic potential in themanagement of major depression. CNS Drugs 1996;5:389-402.Davis R, Whittington R, Bryson HM. Nafazodone: A review of its pharmacology and clinicalefficacy in the management of major depression. Drugs 1997;53(4):608-36.Dinan TG, Scott LV. Does pindolol induce a rapid improvement in depressed patients resistantto serotonin reuptake inhibitors? Journal of serotonin research 1996;3:119-21.De Wilde J, Spiers R, Mertens C et al. A double-blind, comparative, multicentre studycomparing paroxetine with fluoxetine in depressed patients. Acta Psychiatr Scand1993;87:141-5.Dierick M, Ravizza L, Realini R et al. A double-blind comparison of venlafaxine andfluoxetine for the treatment of major depression in outpatients. Prog NeuropsychopharmacolBiol Psychiatry 1995;30:57-71.Doogan DP. Toleration and safety of sertraline: experience worldwide. Int ClinPsychopharmacol 1991;6(Suppl.2):47-56.Doogan DP, Caillard V. Sertraline in the prevention of depression. Br J Psychiatry1992;160:217-22.Dubini A, Bosc M, Polin V. Noradrenaline-selective antidepressant therapy: differential effectson social functioning. J Psychopharmacology 1997;11(Suppl.4):S17-23.Einarson TR, Arikian SR, Casciano J, Doyle JJ. Comparison of extended-release venlafaxine,selective serotonin reuptake inhibitors, and tricyclic antidepressants in the treatment ofdepression: a meta-analysis of randomized controlled trials. Clin Ther 1999;2:296-308.Ereshefsky L. Drug-drug interactions involving antidepressants: focus on venlafaxine. J ClinPsychopharmacol 1996;16(Suppl. 2):37S-53S.Eskelius L, Von Knorring L, Eberhard G. A double-blind multicentre trial comparingsertraline and citalopram in patients with major depression treated in general practice. Int ClinPsychopharmacol 1997;12:323-31.Fawcett J, Marcus RN, Anton SF. Response of anxiety and agitation symptoms duringnefazodone treatment of major depression. J Clin Psychiatry 1995;56(Suppl. 6):37-42.Fawcett J, Barkin RL. A meta-analysis of eight randomized, double-blind controlled clinicaltrials of mirtazapine for the treatment of patients with major depression and symptoms ofanxiety. J Clin Psychiatry 1998;59:123-7.Feiger A, Kiev A, Shrivastava RK et al. Nefazodone versus sertraline in outpatients withmajor depression: focus on efficacy, tolerability and effects on sexual function and satisfaction.J Clin Psychiatry 1996;57(Suppl. 2): 53-62.Feighner JP, Bennett ME, Roberts DL et al. A double-blind trial of nefazodone versusplacebo in depressed inpatients [Abstract]. In: 149th Meeting of the American PsychiatricAssociation 1996;168-No NR370.Fiorella D, Rabin RA, Winter JC. The role of the 5-HT2A and 5-HT2C receptors in thestimulus effects of m-chlorophenylpiperazine. Psychopharmacology 1995;119:222-30.Fountaine R, Ontiveros A, Elie R et al. A double-blind comparison of nefazodone, imipramineand placebo in major depression. J Clin Psychiatry 1994;54:234-41.

Chapter 2

64

Franchini L, Gasperinin M, Perez J et al. A double-blind study of long-term treatment withsertraline or fluvoxamine for prevention of highly recurrent unipolar depression. J ClinPsychiatry 1997;58:104-7.Fuller RW. Uptake inhibitors increase extracellular serotonin concentration measured by brainmicrodialysis. Life Sci 1994;55:163-7.Gaddum JH. Chemical transmission in the central nervous system. Nature 1963;197:741-3.Geretsegger C, Bohmer F, Ludwig M. Paroxetine in the elderly patient: randomizedcomparison with fluoxetine of efficacy, cognitive and behavioral effects. Int ClinPsychopharmacol 1994;9:25-9.Guelfi J, Ansseau M, Corruble E et al. A double blind comparison of the efficacy and safetyof milnacipran and fluoxetine in severly depressed inpatients. [data on file] 1997;Instituede(recherche):Pierre Fabre.Haffmans PM, Timmerman L, Hoogduin CA. Efficacy and tolerability of citalopram incomparison with fluvoxamine in depressed outpatients: a double-blind multicenter study. IntClin Psychopharmacol 1996;11:157-64.Harvey KV, Balon R. Augmentation with buspirone: a review. Ann Clin Psychiatry1995;7:143-7.Jacobsen FM. Possible augmentation of antidepressant response by buspirone. J ClinPsychiatry 1991;52:217-20.Joffe RT, Schuller DR. An open study of buspirone augmentation of serotonin reuptakeinhibitors in refractory depression. J Clin Psychiatry 1993;54:269-71.Kasper S. Clinical efficacy of mirtazapine: a review of meta-analyses of pooled data. Int ClinPsychopharmacol 1995;10(4):25-35.Kasper S, Pletan Y, Solles A et al. Comparative studies with milnacipran and tricyclicantidepressants in the treatment of patients with major depression: a summary of clinical trialresults. Int Clin Psychopharmacol 1996;11(Suppl. 4):35-9.Kasper S, Zivkov M, Roes KCB et al. Pharmacological treatment of severely depressedpatients: a meta-analysis comparing efficacy of mirtazapine and amitriptyline.Neuropsychopharmacol 1997;7:115-24.Katona C, Bercoff E, Chiu E et al. Reboxetine versus imipramine in the treatment of elderlypatients with depressive disorders: a double-blind randomised trial. J Affect Disord1999;55:203-13.Keller MB, Kocsis JH, Thase ME et al. Maintenance phase efficacy of sertraline for chronicdepression: a randomized controlled study. JAMA 1998b;280:1665-72Keller MB, Gelenberg AJ, Hirschfeld RM et al. The treatment of chronic depression, part 2: adouble-blind, randomized trial of sertraline and imipramine. J Clin Psychiatry 1998;59:598-607.Kiev A, Feiger A. A double-blind comparison of fluvoxamine and paroxetine in the treatmentof depressed outpatients. J Clin Psychiatry 1997;58:146-52.Kreiss D., Lucki. Chronic administration of the 5-HT1A receptor agonist 8-OH-DPATdifferentially desensitizes 5-HT1A autoreceptors of the dorsal and median raphe nuclei.Synapse 1997; 25: 107-116.


65

Landen M, Bjorling G, Agren H, Fahlen T. A randomized, double-blind, placebo-controlledtrial of buspirone in combination with an SSRI in patients with treatment-refractorydepression. J Clin Psychiatry 1998;59(12):664-8.Lecrubier Y, Bourin M, Moon CAL et al. Efficacy of venlafaxine in depressive ilness in generalpractice. Acta Psychiatr Scand 1997;95:485-93.Leinonen E, Lepola U, Koponen H et al. Long-term efficacy and safety of milnaciprancompared to clomipramine in patients with major depression. Acta Psychiatr Scand1997;96:497-504.Leonard BE. Noradrenaline in basic models of depression. Eur Neuropsychopharmacol1997a;7(Suppl. 1):S11-6.Leonard B.E. The role of noradrenaline in depression: a review. J. Psychopharmacology.1997b; 11 (suppl. 4): S39-47Lindsay DeVane C. Differential pharmacology of newer antidepressants. J Clin Psychiatry1998;59(Suppl. 20):85-93.Lopez-Ibor J, Guelfi J, Pletan Y. Milnacipran and selective serotonin reuptake inhibitors inmajor depression. Int Clin Psychopharmacol 1996;11(Suppl. 4):41-6.Maes M, Libbrecht I, van Hunsel F, Campens D, Meltzer HY. Pindolol and mianserinaugment the antidepressant activity of fluoxetine in hospitalized major depressed patients,including those with therapy resistance. J Clin Psychopharmacol 1999;19:177-82.Mann JJ, Stanley M, McBride PA et al. Increased serotonin2 and beta-adrenergic receptorbinding in the frontal cortices of suicide victims. Arch Gen Psychiatry 1986;43:954-9.Marek GJ, McDougle CJ, Price LH, Seiden LS. A comparison of trazodone and fluoxetine:implications for a serotonergic mechanism of antidepressant action. Psychopharmacology1992;109:2-11.Marcus RN, Mendels J. Nefazodone in treatment of severe, melancholic, and recurrentdepression. J Clin Psychiatry 1996;57(Suppl. 2):19-23.Marttila M, Jaaskelainen J, Jarvi R et al. A double blind study comparing the efficacy andtolerability of mirtazapine and doxepin in patients with major depression. EurNeuropsychopharmacol 1995;5(4):441-6.Massana J, Moller H. In severe depression, reboxetine is as effective as imipramine and moreeffective than fluoxetine [Abstract]. Proc Ann Meeting Am Psychiatric Assoc, Toronto,Canada 1998a.Massana J. Reboxetine versus fluoxetine: an overview of efficacy and tolerability. J ClinPsychiatry 1998b;59(suppl. 14):8-10.Matsubara R, Onodera I, Ito K et al. A double-blind comparison of milnacipran andimipramine in depressive patients. Eur Neuropsychopharmacol 1996;6(Suppl. 3):48.McAskill R, Mir S, Taylor D. Pindolol augmentation of antidepressant therapy. Br J Psychiatry1998;173:203-8.Meana JJ, Barturen F, Garcia-Sevilla JA. Alpha 2-adrenoceptors in the brain of suicidevictims: increased receptor density associated with major depression. Biol Psychiatry1992;1:471-90.

Chapter 2

66

Mendels J, Reimherr F, Marcus RN et al. A double-blind, placebo-controlled trial of two doseranges of nefazodone in the treatment of depressed outpatients. J Clin Psychiatry1995;56(Suppl.6):30-6.Mendlewicz J. Efficacy of fluvoxamine in severe depression. Drugs 1992;43(suppl. 2):32-7.Montgomery SA, Dunbar GC. Paroxetine is better than placebo in relapse prevention and theprophylaxis of recurrent depression. Int Clin Psychopharmacol 1993a;8:189-95.Montgomery SA, Rasmussen JG, Tanghoj P. A 24-week study of 20 mg citalopram, 40 mgcitalopram and placebo in the prevention of relapse of major depression. Int ClinPsychopharmacol 1993b;8:181-8.Montgomery SA, Henry J, McDonald G et al. Selective serotonin reuptake inhibitors: meta-analysis of discontinuation rates. Int Clin Psychopharmacol 1994;9:47-53.Montgomery SA, Kasper S. Comparison of compliance between serotonin reuptake inhibitorsand tricyclic antidepressants: a meta-analysis. Int Clin Psychopharmacol 1995a;9(Suppl. 4):33-40.Montgomery SA. Safety of mirtazapine: a review. Int Clin Psychopharmacol 1995b;10(Suppl.4):37-45.Montgomery SA. Reboxetine: additional benefits to the depressed patient. JPsychopharmacology 1997;11(suppl. 4):S9-15.Montgomery SA, Reimitz PE, Zivkov M. Mirtazapine versus amitryptyline in the long-termtreatment of depression: a double blind placebo controlled study. Int Clin Psychopharmacol1998;13:63-73.Montgomery SA. Managing the severely ill and long-term depressed. Int J Psych Clin Pract1999a;3(suppl. 1):S13-7.Montgomery SA. Predicting response: noradrenaline reuptake inhibition. Int ClinPsychopharmacol 1999b;14(Suppl. 1):S21-6.Moreno FA, Gelenberg AJ, Bachar K, Delgado PL. Pindolol augmentation of treatment-resistant depressed patients. J Clin Psychiatry 1997;58:437-9.Moret C, Charveron M, Finberg JP et al. Biochemical profile of midalcipran (F2207),1-phenyl-1-diethyl-aminocarbonyl-2-aminomethyl-cyclopropane (Z) hydrochloride, a potentialfourth generation antidepressant drug. Neuropharmacology 1985;24:1211-9.Mucci M. Reboxetine: a review of antidepressant tolerability. J Psychopharmacology1997;11:S33-7.Mullin J, Lodge A, Bennie E et al. A multicentre, double-blind, amitriptyline controlled studyof mirtazapine in patients with major depression. J Psychopharmacology 1996;10(3):235-40.Muth EA, Haskins JT, Moyer JA et al. Antidepressant biochemical profile of the novel bicycliccompound Wy-45,030 an ethylcyclohexanol derivative. Biochemical Pharmacology1986;35:4493-7.Nelson JC. Augmentation strategies in depression 2000. J Clin Psychiatry 2000;61(Suppl.2):13-9.Owens MJ, Morgan WN, Plott Sj, Nemeroff CB. Neurotransmitter receptor and transporterbinding profile of antidepressants and their metabolites. J Pharmacol Exp Ther1997;283:1305-22.


67

Patten SB. The comparative efficacy of trazodone and imipramine in the treatment ofdepression. CMAJ 1992;146:1177-82.Patris M, Bougerol T, Charbonnier JF et al. Citalopram versus fluoxetine: a double-blind,controlled, multicentre, phase III trail in patients with unipolar major depression treated ingeneral practice. Int Clin Psychopharmacol 1996;11:129-36.Perez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Randomised, double-blind, placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment [seecomments]. Lancet 1997;349:1594-7.Perez V, Soler J, Puigdemont D et al. A double-blind, randomized, placebo-controlled trial ofpindolol augmentation in depressive patients resistant to serotonin re-uptake inhibitors. ArchGen Psychiatry 1999;56:375-9.Peyron E. Efficacy of mirtazapine vs clomipramine in severely depressed, hospitalized patients[Abstract]. Eur Neuropsychopharmacol 1996;6(3):46-7.Pigott TA, L'Heueux F, Rubenstein CS et al. A double-blind placebo-controlled study oftrazodone in patients with obsessive compulsive disorder. J Clin Psychopharmacol1992;12:156-62.Puech A, Montgomery SA, Prost JF et al. Milnacipran, a new serotonin and noradrenalinereuptake inhibitor: an overview of its antidepressant activity and clinical tolerability. Int ClinPsychopharmacol 1997;12:99-108.Preskorn SH. Antidepressant drug selection: criteria and options. J Clin Psychiatry 1994;55:6-22.Rapaport M, Coccaro E, Sheline Y et al. A comparison of fluvoxamine and fluoxetine in thetreatment of major depression. J Clin Psychopharmacol 1996;16:373-8.Reboxetine European Package insert. In: Pharmacia and Upjohn 1997.Richou H, Ruimy P, Charbaut J et al. A multicentre, double-blind, clomipramine-controlledefficacy and safety study of Org 3770. Hum Psychopharmacology 1995;10:263-71.Rioux P, Kibleur Y, Frachon O et al. A double-blind comparison of nefazodone and fluoxetinein depressed patients [Abstract]. In: 149th Meeting of the American Psychiatric Association1996;164-No NR 359.Rickels K, Schweizer E, Clary C, et al. Nefazodone and imipramine in major depression: aplacebo-controlled trial. Br J Psychiatry 1994;164:802-5.Rickels K, Robinson DS, Schweizer E et al. Nefazodone: aspects of efficacy. J Clin Psychiatry1995;56(Suppl. 6):43-6.Robert P, Montgomery SA. Citalopram in doses of 20-60 mg is effective in depression relapseprevention: a placebo-controlled 6 month study. Int Clin Psychopharmacol 1995;10(Suppl.1):29-35.Roose SP, Glassman AH, Attia E. Comparative efficacy of selective serotonin reuptakeinhibitors and tricyclics in the treatment of melancholia. Am J Psychiatry 1994;151:1735-9.Rudolph RL, Derivan AT. Double-blind, placebo-controlled study of once daily venlafaxineXR and fluoxetine in depressed outpatients. Presented at: 36th Annual Meeting of theAmerican College of Neuropsychopharmacology, Kamuela, Hawaii, 8-12 December, 1997.Rush AJ, Koran LM, Keller MB et al. The treatment of chronic depression, part 1: studydesiign and rationale for evaluating the comparative efficacy of sertraline and imipramine as

Chapter 2

68

acute, crossover, continuation, and maintenance phase therapies. J.Clin Psychiatry1998;59:589-97.Sacchetti G, Bernini M, Bianchetti A et al. Studies on the acute and chronic effect ofreboxetine on extracellular norepinephrine and other monoamines in the rat brain. Br JPharmacol 1999;128(6):1332-8.Schildkraut JJ. The catecholamine hypothesis of affective disorders: a review of thesupporting evidence. Am J Psychiatry 1965;122:509-21.Smith WT, Glaudin V, Panagides J et al. Mirtazapine vs. amitriptyline vs. placebo in thetreatment of major depressive disorder. Psychopharm Bull 1990;26(2):191-6.Song F, Freemantle N, Sheldon TA et al. Selective serotonin reuptake inhibitors: meta-analysis of efficacy and acceptability. Br Med J 1993;306:683-7.Spencer CM, Wilde MI. Milnacipran: A review of its use in depression. Drugs1998;56(3):405-27.Sproule BA, Naranjo CA, Brenmer KE, Hassan PC. Selective serotonin re-uptake inhibitorsand CNS drug interactions: A critical review of the evidence. Clin Pharmacokinetics1997;33(6):454-71.Spyraki C, Fibiger HC. Functional evidence for subsensitivity of noradrenergic alpha 2receptors after chronic desimipramine treatment. Life Sci 1980;27:1863-7.Stahl S, Zivkov M, Reimitz PE et al. Meta-analysis of randomized, double-blind, placebo-controlled, efficacy and safety studies of mirtazapine versus amitriptyline in major depression.Acta Psychiatr Scand 1997;96(suppl. 391):22-30.Steen A, den Boer JA. A double-blind six month comparative study of milnacipran andclomipramine in major depressive disorder. Int Clin Psychopharmacol 1997;12:269-81.Steffens DC, Krishnan KR, Helms MJ. Are SSRI's better than TCA's: a meta-analysis.Depress Anxiety 1997;6:10-8.Taylor DP, Carter RB, Eison AS et al. Pharmacology and neurochemistry of nefazodone, anovel antidepressant drug. J Clin Psychiatry 1995;56(Suppl.6):3-11.Tignol J, Stoker MJ, Dunbar GC. Paroxetine in the treatment of melancholia and severedepression. Int Clin Psychopharmacol 1992;7:91-4.Tignol J. A double-blind, randomized, fluoxetine-controlled, multicenter study of paroxetine inthe treatment of depression. J Clin Psychopharmacol 1993;13(Suppl. 2):18-22.Tignol J, Pujol-Domenech J, Chartres JP et al. Double-blind study of the efficacy and safetyof milnacipran (100 mg/day) and imipramine (100 mg/day) in elderly patients with majordepressive episode. Acta Psychiatr Scand 1998;97:157-65.Tome MB, Isaac MT, Harte R, Holand C. Paroxetine and pindolol : a randomised trial ofserotonergic autoreceptor blockade in the reduction of antidepressant latency. Int ClinPsychopharmacol 1997;12:81-9.Van Moffaert M, Pregaldien JL, von Frenckell R et al. A double-blind comparison ofnefazodone and imipramine in the treatment of depressed patients. New Trends Exp ClinPsychiatry 1994;10(2):85-7.Venlafaxine USA package insert. In: Wyeth Ayerst 1999.Wagner W, Zaborny BA, Gray TE. Fluvoxamine: a review of its safety profile in world-widestudies. Int Clin Psychopharmacol 1994;9:223-7.


69

Wheatley DP, van Moffaert M, Timmerman L et al. Mirtazapine: efficacy and tolerability incomparison with fluoxetine in patients with moderate to severe major depressive disorder. JClin Psychiatry 1998;59:306-12.Woolley DW, Shaw E. A biochemical and pharmacological suggestion about certain mentaldisorders. Science 1954;119:587-8.Workman EA, Short DD. Atypical antidepressants versus imipramine in the treatment ofdepression: a meta-analysis. J Clin Psychiatry 1993;54(1):5-12.Yamashita I, Matsubara R, Onodera I et al. Clinical evaluation of milnacipran hydrochloride(TN-912) on depression and depressive states. Rinsho Iyaku 1995;11(4):819-42.Zanardi R, Artigas F, Franchini L, Sforzini L, Gasperini M, Smeraldi E, Perez V. How longshould pindolol be associated with paroxetine to improve the antidepressant response? J ClinPsychopharmacol 1997;17:446-50.Zivkov M, De Jongh GD. Org 3770 versus amitriptyline: a 6-week randomized double-blindmulticentre trial in hospitalized depressed patients. Hum Psychopharmacology 1995;10:173-80.

Chapter 2

70

Augmentation with a 5-HT1A, but not a 5-HT1B receptor antagonist critically depends on the dose of citalopram

71

Chapter 3

Augmentation with a 5-HT1A, but not a 5-HT1B

receptor antagonist critically depends on the dose of

citalopram.

Chapter 3

72

Abstract

Pharmacokinetic and pharmacodynamic parameters of the selective serotonin re-uptakeinhibitor (SSRI) citalopram [1-(3-dimethylaminopropyl) -1-(4-fluorophenyl)-5-phtalancarbonitril] were determined in order to find optimal conditions for augmentation ofits effect on extracellular serotonin (5-HT) through blockade of 5-HT1A and 5-HT1B

autoreceptors. Citalopram dose-dependently (0.3 to 10 µmol/kg s.c.) increased serotoninlevels in ventral hippocampus of conscious rats. At plasma levels above approximately 0.15µM, the effect of citalopram on extracellular 5-HT was augmented by both a 5-HT1A (Way100635, 1 micromol/kg s.c. (N-[2-[4-(2-mehoxyphenyl)-1-piperazinyl]ethyl]-N-(2-pyridil)cyclohexanecarboxamide trihydrochloride)) and a 5-HT1B receptor antagonist (GR127935, 1 µmol/kg s.c., (2’-methyl-4’-(5-methyl-[1,2,4]oxadiazol-3-yl)biphenyl-4-carboxylicacid [4-methoxy]-3-(4-methylpiperazin-1-yl)phenyl]amide)). However, at plasma levels ofthe SSRI below 0.15 µM, the effects of the antagonists diverged viz. the 5-HT1B receptorantagonist was still able to potentiate citalopram’s effect on extracellular 5-HT, while the 5-HT1A receptor antagonist was no longer effective. These results suggest that in contrast to 5-HT1B autoreceptors, indirect activation of 5-HT1A autoreceptors by citalopram is criticallyrelated to the dose of SSRI administered. The latter may have consequences for SSRIaugmentation strategies with 5-HT1A receptor antagonists in the therapy of depression andanxiety disorders.


73

1. Introduction

Selective serotonin re-uptake inhibitors (SSRIs) increase brain extracellular serotonin (5-HT)concentrations in animals acutely following administration (Fuller, 1994). In contrast, theclinical effects of these substances are typically delayed for several weeks (Baumann, 1992;Fuller, 1994). This notion suggests the occurrence of adaptive changes triggered by thesustained elevated 5-HT levels. Several lines of evidence suggest that these changes involvedesensitization of 5-HT autoreceptors. Attenuation of auto-inhibitory processes increases theoverall serotonergic neurotransmission, and this process has been hypothesized to underliethe therapeutic effects of SSRIs (Blier et al., 1987).

At least three types of presynaptic 5-HT receptor regulatory mechanisms are known.Somatodendritic 5-HT1A autoreceptor activation decreases 5-HT release in terminal areas viainhibition of serotonergic cell firing (Arborelius et al., 1994; Bosker et al., 1994). Severalauthors have reported a reduced inhibitory effect of the 5-HT1A receptor agonist 8-hydroxy-2-(di-n-propylamino)tetraline (8-OH-DPAT) on terminal 5-HT release or raphe nuclei cellfiring following chronic treatment with SSRIs (Invernizzi et al., 1994; Kreiss and Lucki,1995; Le Poul et al., 1995). Others, however, failed to observe any such effects (Hjorth andAuerbach, 1994; Bosker et al., 1995).Activation of terminal 5-HT1B receptors (formerly 5-HT1Dβ) also decreases 5-HT release(Bosker et al., 1995). Similarly, data concerning 5-HT1B autoreceptor desensitizationfollowing chronic treatment with SSRIs are controversial (Chaput et al., 1986; Auerbach andHjorth, 1995; Bosker et al., 1995; Moret and Briley, 1996; Davidson and Stamford, 1997).The third type of regulatory mechanism may be through somatodendritic 5-HT1D (formerly 5-HT1Dα) receptors (Starkey and Skingle, 1994; Sprouse et al., 1997). Activation of thesereceptors decreases 5-HT release in the cell body region, which may indirectly influenceterminal 5-HT release through interplay with somatodendritic 5-HT1A autoreceptors.

In contrast to receptor density and affinity, which can be adequately assessed by means of in-vitro binding studies (Bmax and Kd), the desensitization of auto-inhibitory processes can onlybe measured in functional models, for example by using microdialysis or electrophysiologicalmethods. In addition to using selective 5-HT1A and 5-HT1B receptor agonists, the SSRIsthemselves can be used as probes for measuring desensitization processes. Eventually, theattenuation of release-restraining processes will lead to an additional increase of extracellular5-HT. Using this latter approach, microdialysis studies have indeed reported additionalincreases in extracellular 5-HT following chronic treatment with SSRIs (Invernizzi et al.,1994; Auerbach and Hjorth, 1995), although another study failed to observe these effects(Bosker, et al. 1995). In addition, an electrophysiological study has reported that inhibition ofdorsal raphe nucleus 5-HT cell firing by citalopram was time-dependently abolished duringchronic treatment (Chaput et al., 1986).

Apart from differences in animal strains (Schoups and De Potter, 1988), the discrepantfindings may have a variety of causes. For instance, McQuade et al. (McQuade and Sharp,

Chapter 3

74

1997) have demonstrated that forebrain regions are not evenly innervated by dorsal raphenucleus and median raphe nucleus. It can be speculated that 5-HT1A autoreceptors on 5-HTcell bodies within these nuclei differ in their susceptibility to desensitization, which mayexplain some of the discrepancies (Kreiss and Lucki, 1994; Kreiss and Lucki., 1997).An obvious source of variance are differences in dose regimen of the SSRIs used in thestudies. Since plasma half-lives of drugs in rodents are generally much shorter than inhumans, multiple or very high doses are not uncommon in order to mimic human plasmalevels. Pharmacologically inactive, strongly fluctuating or even toxic plasma levels thusestablished, may have contributed to the varying results.

An alternative approach to test Blier’s desensitization hypothesis (Blier et al., 1987) is theadministration of SSRIs with concomitant autoreceptor blockade. Arguably, blocking theautoreceptors instantaneously mimics the desensitization process. Animal studies havedemonstrated that co-administration of a 5-HT1A receptor antagonist causes an additionalincrease of extracellular serotonin levels (Hjorth, 1993; Hjorth et al., 1996; Gobert et al.,1997), which is afterall the basis of Blier’s hypothesis. Consequently, co-administration of anSSRI and 5-HT1A and/or 5-HT1B antagonists should accelerate the onset of therapeutic effect.Following this line of thought, clinical studies have reported on hastening and/orimprovement of response to antidepressant treatment by co-administration of the β-adrenoceptor and 5-HT1A/B receptor antagonist pindolol.Several groups have indeed claimed a more rapid onset of action, larger reductions inHamilton depression scores in time, and sometimes beneficial effects in therapy resistantdepression (Artigas et al., 1994; Perez et al., 1997; McAskill et al., 1998). On the other handseveral equally well executed studies were not able to support these findings (Berman et al.,1997; McAskill et al., 1998).To explain this discrepancy several explanations have been advanced such as differences inpatient groups, different properties of the SSRIs, etc.An alternative explanation may be found in varying SSRI plasma levels between patients,which may lead to different degrees of autoreceptor activation. Arguably, this would lead to asubstantial variation in antagonist effect between patients.

In the present study, the contribution of 5-HT1A and 5-HT1B autoreceptors in restraining SSRIevoked 5-HT release in rat ventral hippocampus was investigated by co-administration ofselective antagonists and different doses of citalopram, currently the most selective serotoninre-uptake inhibitor (Hyttel, 1994). The selective 5-HT1A and 5-HT1B receptor antagonistsused in the study were WAY 100635 and GR 127935, respectively. In addition, citalopramplasma concentrations were measured to relate brain pharmacology with plasma levels of theSSRI.


75

2. Materials and Methods

2.1 Animals

Male albino rats of a Wistar-derived strain (285-320 g; Harlan, Zeist, The Netherlands) wereused for the experiments. Upon surgery, rats were housed individually in plastic cages (35 x35 x 40 cm), and had free access to food and water. Animals were kept on a 12 h lightschedule (light on 7:00 a.m.). The experiments are concordant with the declarations ofHelsinki and were approved by the animal care committee of the faculty of mathematics andnatural science of the University of Groningen .

2.2 Drugs

The following drugs were used: Citalopram hydrobromide [1-(3-dimethylaminopropyl) -1-(4-fluorophenyl)-5-phtalancarbonitril] (kindly donated by Lundbeck (Denmark), courtesy Dr.Sanchez), Way 100635 (N-[2-[4-(2-mehoxyphenyl)-1-piperazinyl]ethyl]-N-(2-pyridil)cyclohexanecarboxamide trihydrochloride) and (GR 127935 2’-methyl-4’-(5-methyl-[1,2,4]oxadiazol-3-yl)biphenyl-4-carboxylic acid [4-methoxy]-3-(4-methylpiperazin-1-yl)phenyl]amide) (both compound were synthesized in our laboratory, courtesy Dr. Y. Liaoand Mrs. M. Mensonides). Drugs were dissolved in saline except for GR 127935, which wasdissolved in saline with a drop of acetic acid. Substances were injected subcutaneously in avolume of 1 ml per kg.

2.3 Surgery

Microdialysis experiments were performed using home made I-shaped probes, made ofpolyacrylonitrile / sodium methyl sulfonate copolymer dialysis fiber (i.d. 220 µm, o.d. 0.31µm, AN 69, Hospal, Italy).The exposed length of the membranes was 4 mm.Preceding surgery rats were anesthetized by means of an intraperitoneal injection of 400mg/kg chloral hydrate. Lidocaine-HCl, 10 % (m/v) was used for local anesthesia. Rats wereplaced in a stereotaxic frame (Kopf, USA), and probes were inserted into the ventralhippocampus (coordinates: IA: +3.7 mm, lateral : +4.8 mm, ventral: -8.0 mm from the duramater, Paxinos and Watson, 1982) and secured with dental cement.

Pharmacokinetic experiments were performed in a separate group of rats withoutmicrodialysis probes. Apart from this, conditions were comparable with the microdialysisanimals, including anaesthesia. Blood was drawn through a canula made of silicon tubing,which was inserted into the right jugular vein for 3.8 mm, during chloralhydrate anaesthesia.The tubing was transferred subcutaneously to the scull of the rat, and a stainless steel inletwas connected to the tubing. The inlet was mounted on the skull with dental cement andsurgical screws. After insertion, canulas were filled with a PVP solution (55 %polyvinylpyrrolidon in 500 IE/ml heparin) in saline to prevent blood clotting.

Chapter 3

76

2.4 Microdialysis experiments

Rats were allowed to recover for at least 24 h. Probes were perfused with artificialcerebrospinal fluid containing 147 mM NaCl, 3.0 mM KCl, 1.2 mM CaCl2, and 1.2 mMMgCl2, at a flow-rate of 1.5 µl / min (Harvard apparatus, South Natick, Ma., USA). Sampleswere collected on-line in a 20 µl loop and injected automatically onto the column every 15min.

2.5 Pharmacokinetic experiments

Preceding the experiments, rats were allowed to recover from surgery for 24 h. Bloodsamples (0.35 ml) were drawn at 0, 15, 30, 60, 120, 240 and 360 min after injection ofcitalopram (3 and 10 µmol/kg s.c.). Samples were transferred to 1.5 ml eppendorf vials,containing 5 µl heparin (500 IE/ml saline), mixed and immediately transferred to a chilledcentrifuge (MSE, England), and centrifuged for 15 min at 3,000 rpm.To measure citalopram plasma concentrations at the dose of 0.1 µmol/kg, it appearednecessary to introduce a concentration step. To this end animals were anesthetized by etheranesthesia 30 min after receiving the citalopram injection. Vacutainers (Becton Dickinson,England) were used in order to obtain at least 5 ml of plasma by means of cardiac puncture.

2.6 Analysis

2.6.1. Serotonin analysis:5-HT was analyzed using high pressure liquid chromatography (HPLC) with electrochemicaldetection. The HPLC pump (Shimadzu LC-10 AD liquid chromatograph) was connected to areversed phase column (phenomenex hypersil 3 : 3 µm, 100 x 2.0 mm, C18, Bester,Amstelveen, the Netherlands) followed by an electrochemical detector (Antec Leyden,Leiden, the Netherlands), working at a potential setting of 500 mV vs. Ag/AgCl reference.The mobile phase consisted of 5 g/l di-ammoniumsulfate, 500 mg/l ethylene diamino tetraacetic acid (EDTA), 50 mg/l heptane sulphonic acid, 30 µl/l of triethylamine and 4.5 % v/vMeOH at a pH of 4.65. Flow-rate of the mobile phase was 0.4 ml/min. The detection limitwas 1-2 fmol 5-HT per 20 µl sample (signal to noise ratio 2).

2.6.2. Citalopram analysis:Citalopram was measured according to Øyehaug et al. (1982), with minor modifications.Briefly, to 150 µl plasma samples 75 µl of the internal standard LU 10-202 (2 µM) and 30 µlof 1N NaOH were added. Samples were extracted twice by mechanically shaking for 3 min.with 3 ml of diethyl ether. The ether layers were then transferred to 10 ml evaporating tubes,and 150 µl of 0.1 N HCl was added. The ether was evaporated in a water-bath at 40 °C undera stream of nitrogen. The HCl layer was washed once with 0.5 ml ether. 50 µl samples wereinjected onto the column. Extraction recovery of citalopram, metabolites and internal


77

standard were approximately 95 %. Citalopram plasma levels were corrected for recovery ofthe internal standard.With the experiments in which 5 ml plasma was obtained via the vacutainer, a similarprotocol was used, but the plasma sample was now extracted three times with 10 ml of ether.After extraction, the combined ether layers were evaporated into 150 µl of 0.1 N HCl,thereby concentrating the samples 33.3 times. Under these conditions, relative recovery of theextraction for citalopram, metabolites and internal standard was lower and amounted toapproximately 70 %. Citalopram plasma levels were corrected for recovery of the internalstandard.An HPLC/auto-injector (1084B Liquid Chromatograph, Hewlett-Packard) was used, incombination with a fluorescence detector (470 Scanning Fluorescence detector, Waters,England) operating at an absorption wavelength of 240 nm, an emission wavelength of 296nm, and a slit-width of 12 nm. Separation was performed using a Supelcosil HPLC column (5µm, C18, 250 x 46 mm, Supelco, the Netherlands), at ambient temperature. The mobilephase consisted of 46 % v/v acetonitrile, 54 % v/v potassium dihydrogen phosphate buffer(4.3 g/l), and 30 µl/l triethylamine, at a pH value of 3.0. Flow-rate in this system was 0.75ml/min. The detection limits for “on column injections” of citalopram, desmethyl-citalopramand didesmethyl-citalopram were 8, 7 and 5 nM, respectively (signal to noise = 2).

2.7 Data presentation and statistics

Four consecutive microdialysis samples with less then 20 % variation were taken as controland set at 100 %. Data are presented as percentages of control level (mean + S.E.M.).Statistical analysis was performed using Sigmastat for windows (Jandel Corporation).Treatment effects were compared using two way analysis of variance (ANOVA) for repeatedmeasurements, followed by Dunnett’s test. Level of significance level was set at p<0.05.Dose response curves and ED50s were calculated using origin software.Pharmacokinetic data were fitted using Multifit (copyright Dr. J.H. Proost, Dept. ofpharmacokinetics and drug delivery, University of Groningen, the Netherlands).

Chapter 3

78

3. Results


Plasma levels were determined of 3 doses of citalopram (0.1, 3 and 10 µmol/kg).Subcutaneous administration of 10 µmol/kg citalopram rapidly established plasma levels ofabout 0.5 µM. During the time span of the pharmacological experiment (240 min.), plasmalevels gradually decreases to about 0.15 µM. Following administration of 3 µmol/kgcitalopram, plasma levels of about 0.13 µM were found to occur within 15 min. After 240min. plasma levels decreased to approximately 0.01 µM (Fig. 1). Desmethyl- anddidesmethyl metabolites of citalopram could not be detected in the plasma afteradministration of either dose.The three doses of citalopram showed a linear relation with the respective plasmaconcentrations (Fig. 2). From the plasma time-concentration profile after 10 µmol/kg s.c.citalopram, several pharmacokinetic parameters were calculated. T1/2 was 110 + 13 min,volume of distribution 20 + 4 l/kg, and clearance 0.13 + 0.02 l/kg/min.

Figure 1. Plasmaconcentrations in time following subcutaneous injection of 3.0 (n=4,�) and10 µmol/kg (n=4,�) citalopram. Min, mean and max denote the respective minimal, meanand maximal borders of the human clinically effective therapeutic window.

0 100 200 300 400

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

min

mean

max

Cita

lop

ram

(µ M

)

Time (min)


79

3.2 Pharmacodynamic experiments

3.2.1. Basal extracellular 5-HT levels

Basal levels of 5-HT in ventral hippocampus were 7.8 + 0.7 fmol/15 min fraction (mean +S.E.M., n = 72). No significant differences were observed between the different treatmentgroups.

Figure 2. Linear relationship between the administered dose of citalopram and the plasmaconcentration measured at t=30 min. (n=4) (r =1.00).

3.2.2 Effect of subcutaneous administration of citalopram on extracellular 5-HT levels

Subcutaneous administration of 0.1, 0.3, 1.0, 3.0 and 10.0 µmol/kg citalopram dose-dependently increased extracellular 5-HT in ventral hippocampus (Fig 3).A dose of 0.1 µmol / kg citalopram had no significant effect on extracellular 5-HT in ventralhippocampus (F(1,116)=0.729, P > 0.05)). The next dose induced a short lasting increase ofextracellular 5-HT to about 225 % of control values (F(1,107)=2.66, P < 0.05)). Post-hocanalysis revealed that the effect was significant at t=60 min. Increasing the dose to 1 µmol/kgelevated 5-HT levels significantly to 275 % for the remainder of the experiment(F(1,121)=5.26, P < 0.05)). Post-hoc analysis revealed significant increases from t=90 to 120min. A dose of 3 µmol/kg elicited an increase in extracellular 5-HT of 350 %(F(1,121)=22.3, P < 0.05)). Post-hoc analysis revealed significant differences from t=30 tot=150 min. A dose of 10 µmol/kg increased 5-HT to a comparable extent (F(1,131)=18.8, P

0.01 0.1 1 101E-3

0.01

0.1

1

Cita

lopr

am (

µ M)

Dose (µmol/kg s.c.)

Chapter 3

80

< 0.05)). Post-hoc analysis revealed significant effects from t=45 to t=150. From the doseresponse curve, using the area under the curve (AUC) from t=0 to t=150 min, the half-maximal dose was estimated to be 0.62 µmol/kg (Fig 4).

Figure 3. Effect of subcutaneous administration of citalopram on extracellular levels of 5-HTin ventral hippocampus of conscious and freely moving rats. Citalopram or vehicle wasadministered at t=0 min: Saline ( n = 7 ; �), Citalopram 0.1 µmol/kg (n = 5; �), Citalopram0.3 µmol/kg (n = 4 ; �), Citalopram 1.0 µmol/kg (n = 5 ; �), Citalopram 3.0 µmol/kg (n = 5; �), Citalopram 10 µmol/kg (n = 6 ; ++++ ).

3.2.3. Effect of co-administration of the 5-HT1A receptor antagonist WAY 100635 andcitalopram on extracellular 5-HT levels

A low dose of citalopram (1 µmol/kg) increased 5-HT levels, but co-administration of the 5-HT1A antagonist did not potentiate this effect (F(1,117)=1.129, P > 0.05))(Fig. 5). Co-administration of an intermediate dose of citalopram (3 µmol/kg) and the 5-HT1A receptorantagonist showed a tendency towards an additional increase. However, this effect was notstatistically significant (F(1,108)=0.648, P > 0.05)) (Fig. 6). At a dose of 10 µmol/kgcitalopram, the increase in extracellular 5-HT was significantly augmented by the 5-HT1A

receptor antagonist (F(1,118)=2.71, P < 0.05))(Fig.7).From the AUC (0 to 150 min ) it can be seen that augmentation by WAY 100635 occurs atdoses of the SSRI exceeding 3 µmol/kg (Fig. 8).

-100 -50 0 50 100 1500

200

400

5-H

T %

Bas

al L

eve

l

Time (min)


81

Figure 4. Dose-response curve of citalopram induced 5-HT increases in ventralhippocampus. release. Data were calculated as area under the curve (AUC), from t=0 tot=150 min. Dose 1.0, 3.0 and 10 µmol/kg elicit an statistical difference in AUC vs. SalineAUC (P < 0.05). The calculated ED50 of this effect was 0.62 µmol/kg s.c..

3.2.4. Effect of co-administration of the 5-HT1B receptor antagonist GR 127935 andcitalopram on extracellular 5-HT levels

Co-administration of the selective 5-HT1B receptor antagonist GR 127935 clearly potentiatedthe citalopram induced increase in extracellular 5-HT. The effect was already significant at adose of 1 µmol/kg of citalopram (F(1,79) = 3.09, P < 0.05)) (Fig. 5). Augmentation wasobserved to become progressively stronger when the dose of citalopram was increased(citalopram 3 µmol/kg s.c. (F(1,107) = 2.73, P < 0.05), citalopram 10 µmol/kg s.c. (F(1,118)= 14.6, P < 0.05) (Fig. 5 and 6, respectively). Conversion of these data into AUCs (0 to 150min) clearly shows that augmentation by the 5-HT1B receptor antagonist is already evident atlow doses of the SSRI, while under the same conditions the 5-HT1A receptor antagonist isdevoid of effect (Fig. 8).

1E-3 0.01 0.1 1 10

15000

20000

25000

30000

35000

40000

45000

50000

55000

**

*

Cit 10Cit 1Cit 0.1Saline

5-H

T (

AU

C %

X m

in )

Dose (µmol/kg s.c.)

Chapter 3

82

Figure 5. Effects of citalopram 1.0 µmol/kg and WAY 100635 1.0 µmol/kg or GR 1279351.0 µmol/kg, alone, or co-administered, on extracellular 5-HT levels in ventral hippocampus.Citalopram 1.0 µmol/kg (�, n = 5), WAY 100635 (�, n = 3), citalopram 1.0 µmol/kg andWAY 100635 1.0 µmol/kg (�, n = 6 ), GR 127935 (�, n = 4), citalopram 1.0 µmol/kg andGR 127935 1.0 µmol/kg (� , n = 5) (* P < 0.05).

-100 -50 0 50 100 1500

200

400

600

800

1000

1200

*

5-H

T %

Ba

sal L

evel

Time (min)


83

4. Discussion

The present study clearly demonstrates that in contrast to a 5-HT1B receptor antagonist,augmentation by a 5-HT1A antagonist critically depends on the dose of citalopram. Severalmicrodialysis studies have investigated the acute effects of SSRIs on extracellular 5-HTlevels, but none of them has related these data to plasma concentrations of the SSRI. To ourknowledge, this is the first microdialysis study is the first in which measurement ofpharmacodynamic effects of a selective serotonin re-uptake inhibitor is combined with anestimation of some of its pharmacokinetic parameters.

Figure 6. Effects of citalopram 3.0 µmol/kg and WAY 100635 1.0 µmol/kg or GR 1279351.0 µmol/kg, on extracellular 5-HT levels in ventral hippocampus. Citalopram 3.0 µmol/kg(�, n = 5), citalopram 3.0 µmol/kg and Way 100635 1.0 µmol/kg (�, n = 5), citalopram 3.0µmol/kg and GR 127935 1.0 µmol/kg (�, n = 5) (* P < 0.05).

4.1 Pharmacokinetic experimentsRestrictions imposed by the detection limit of the citalopram assay and the maximal amountof plasma that could reasonably be taken from the canula in the jugular vein made itnecessary to use cardiac puncture (5 ml) for the lowest dose of citalopram. Concentration ofthese samples enabled us to estimate plasma levels even at the lowest dose of citalopram (0.1µ mol/kg). Due to large variation of citalopram plasma levels 15 min. after injection, bloodsamples were taken 30 min after administration. Citalopram plasma levels measured at thistimepoint displayed a linear relationship with the corresponding doses of 0.1, 3.0 and 10.0

-100 -50 0 50 100 1500

200

400

600

800

1000

1200

** *

*

5-H

T %

Bas

al L

evel

Time (min)

Chapter 3

84

µmol/kg. In principle the plasma levels of all other doses used in the present study can beestimated from this linear relation.Evaluation of treatment studies with citalopram in major depression indicates that clinicallyeffective plasma levels range between 0.12 µM and 0.88 µM (Baumann, 1992). In the presentstudy in rats similar plasma levels of citalopram were obtained only at doses of 3 and 10µmol/kg. At the highest dose of citalopram, plasma levels were maintained within this rangefor at least 4 h. Due to rapid elimination in rats, plasma levels of the lower dose alreadydropped below 0.12 µM after 45 min. The pharmacokinetic parameters after 10 µmol/kgcitalopram, are somewhat different from the data initially published by Overø et al, andindicate a faster elimination in the present study (Fredricson Overø, 1982). Different ratstrains, route of administration and protocols may be responsible for this discrepancy(Fredricson Overø, 1982).Although it is realized that pharmacokinetics and pharmacodynamics differ between rats andhumans, we used the reported clinically effective plasma levels as a starting point for ourexperiments.In vivo citalopram is metabolized into several metabolites, of which the desmethyl and thedidesmethyl analogues are the most prominent (Oyehaug et al., 1982).Plasma levels of the metabolites, in depressive patients treated with citalopram under steadystate conditions, have been reported to be at about 2-3 times lower as with the parentcompound (Baumann, 1992). Compared to citalopram, the metabolites have a 10 times loweraffinity for the 5-HT re-uptake site. Arguably, their pharmacological effects, if any, will notbe predominant (Hyttel, 1994). In the present study we could not detect any of the citaloprammetabolites.

4.2 Pharmacological experimentsIn the present study, systemic administration of citalopram induced a dose dependent increaseof extracellular 5-HT levels in the ventral hippocampus of the rat. This observation is inagreement with several other studies (Bel and Artigas, 1992; Hjorth et al., 1997; Invernizzi etal., 1997; Romero and Artigas, 1997; Gundlah et al., 1998), (review by Fuller (1994)). Athigher doses, the effect of citalopram reached a plateau value. Suprisingly, citalopram plasmaconcentrations at this plateau value were in the same range as reported for clinically effectiveplasma levels (Baumann, 1992).The relation between dose of the SSRI and its effect on extracellular 5-HT levels has not verywell been investigated. Hjorth et al., using doses of citalopram comparable to those in thepresent study, reported that increasing the citalopram dose from 0.5 to 5 mg/kg (1.25 and12.5 µmol/kg resp.) did not proportionally increase its effect on 5-HT levels in the ventralhippocampus (Hjorth et al., 1997). The latter indicates that plateau values had been reachedsomewhere between these doses, which is in agreement with our data.In the present study, administration of 0.1 µmol/kg citalopram had no effect on 5-HT levelsin the ventral hippocampus. The next dose of 0.3 µmol/kg transiently increased extracellular5-HT. Plasma level of citalopram (t=30 min) at the lowest dose was approximately 10 nM.This plasma value should be kept in mind when performing chronic experiments withcitalopram. If experiments after cessation of chronic treatment are performed when plasma


85

levels are still above this value, effects of pharmacological probes may be interfered byresidual effects of the SSRI.

Figure 7. Effects of citalopram 10.0 µmol/kg and WAY 100635 1.0 µmol/kg or GR 1279351.0 µmol/kg, on extracellular 5-HT levels in ventral hippocampus. Citalopram 10.0 µmol/kg(�, n = 6), citalopram 10.0 µmol/kg and WAY 100635 1.0 µmol/kg (�, n = 5), citalopram10.0 µmol/kg and GR 127935 1.0 µmol/kg (�, n = 5) (* P < 0.05).

Activation of somatodendritic 5-HT1A autoreceptors reduces neuronal firing of 5-HT raphenuclei cells (Arborelius et al., 1994). SSRIs have been shown to increase extracellular 5-HTin both terminal and cell body areas. The increase in extracellular 5-HT in the cell bodyregion activates 5-HT1A autoreceptors, thereby decreasing 5-HT neuronal firing (Chaput etal., 1986).The latter process counteracts the primary effect of the SSRI. In theory co-administration of a 5-HT1A receptor antagonist should prevent this secondary effect of theSSRI. Indeed, several studies have shown that co-administration of a 5-HT1A receptorantagonist potentiates the effect of an SSRI on extracellular 5-HT (Hjorth et al., 1996; Gobertet al., 1997; Gundlah et al., 1997). The present study corroborates and extends these findings.The 5-HT1A receptor antagonist WAY 100635 potentiated the effect of citalopram onextracellular 5-HT, but the effect appeared critically dependent on the plasma level of theSSRI. Apparently a certain level of feedback regulation is necessary to show the effect of a 5-HT1A receptor antagonist. This finding supports the study by Hjorth et al. (1997), whereinWAY 100635 augmented the effect of a high, but not a low dose of citalopram.In addition to the dependence of augmentation on the SSRI plasma level, Hjorth et al alsoshowed that this effect is critically dependent on the dose of 5-HT1A antagonist (Hjorth et al.,1997). In order to eliminate the possibility that augmentation remained absent due to

-100 -50 0 50 100 1500

200

400

600

800

1000

1200

1400

****

*

*

* * * ***

5-H

T %

Ba

sal L

eve

l

Time (min)

Chapter 3

86

insufficient presence of 5-HT1A antagonist, a supramaximal dose was administered. Co-administration of an even higher amount of Way 100635 (10 µmol/kg) revealed that theaugmentation was maximal with regard to the dose of 5-HT1A antagonist administered (datanot shown).

Figure 8. Effects of citalopram 10 µmol/kg, alone, or combined with WAY 100635 1.0µmol/kg or GR 127935 1.0 µmol/kg. Data are presented as AUC from t=0 to 150 min.Citalopram (�), citalopram combined with WAY 100635 (�), citalopram combined withGR127935 (�). Min denotes entrance of plasma citalopram levels in the therapeutic window.

In addition to the somatodendritic 5-HT1A autoreceptor mediated feedback, 5-HT release isalso controlled by 5-HT1B terminal autoreceptors. Accordingly, simultaneous systemicadministration of the putative 5-HT1B receptor antagonist GR127935 and an SSRI has beenshown to augment the effect of the latter on extracellular 5-HT ( Rollema et al., 1996; Gobertet al., 1997; Sharp et al., 1997).In contrast with WAY 100635, augmentation by GR 127935 appeared independent of thedose of citalopram. Apparently the level of feedback regulation is less critical for 5-HT1B

receptor antagonists.Notably, GR 127935 also has moderate to high affinity for 5-HT1D receptors, suggesting thatthere is room for at least partial involvement of somatodendritic 5-HT1D autoreceptors in theobserved effects (Sprouse et al., 1997). Arguably, blockade of the somatodendritic 5-HT1D

autoreceptors further increases 5-HT levels in the cell body region. Theoretically, this willlead to an increased activation of 5-HT1A autoreceptors in the same area, thereby decreasingfiring rate and terminal 5-HT release. The latter effect would counteract the effect of GR127935 through terminal 5-HT1B autoreceptors. Anyhow, the strong potentiation of the effect

1E-3 0.01 0.1 1 100

20000

40000

60000

80000

100000

120000

140000min

5-H

T A

UC

( %

X m

in)

Dose citalopram µmol/kg s.c.


87

of citalopram by GR 127935 suggests that an effect through 5-HT1D receptors, if any, is ofminor importance at this dose of antagonist.

Arguably, pharmacokinetic interactions between citalopram and the 5-HT receptorantagonists could also play a role. However, citalopram concentrations, measured locally inthe brain by means of microdialysis, were not affected by co-administration of either WAY100635 or GR 127935, which contradicts any putative influence of the antagonists on the bio-availability of citalopram (data not shown). In theory, citalopram could also influence theavailability of the 5-HT antagonists in the brain. For this reason a supra maximal antagonistdose of WAY 100635 was chosen to minimize the risk that an increased availability of theantagonist would further enhance the effect on 5-HT levels. Although no effort was made torelate the effects of GR 127935 to its dose, it can be stated that the observed effects of thecompound would be even stronger at higher doses, thereby decreasing the necessity to use ofa higher dose in the present study. Moreover, citalopram has been shown to lack importantpharmacokinetic interactions through inhibition of cytochrome p450 iso-enzymes, whichrenders increase of availability of the antagonist rather unlikely (Sproule et al., 1997).

4.3 Pharmacodynamic and pharmacokinetic considerationsIn the present study potentiation of citalopram’s effect on extracellular 5-HT by the 5-HT1A

receptor antagonist WAY 100635 appeared critically dependent on the plasma levels of theSSRI. It is of note that augmentation could only be observed when the effect of citalopramhad reached its plateau value. Much to our surprise, the corresponding citalopram plasmalevels were similar to clinically effective plasma levels of the drug. Despite the fact that it isdifficult or even impossible to extrapolate data obtained from rats to humans, this still mayhave implications for SSRI augmentation studies with 5-HT1A receptor antagonists indepressed patients.For example, pindolol augmentation studies in major depression have yielded mixed results(McAskill, et al. 1998), which may be partly due to differences in SSRI plasma levelsbetween patients. It can be speculated that patients with high SSRI plasma concentrationswould benefit more from the addition of pindolol than patients with lower plasma values.

Selective 5-HT1B receptor antagonists may have great potential in SSRI augmentationstrategies. The present study indicates that this type of augmentation is more prominent thanthrough 5-HT1A receptors and is not restricted to clinically effective plasma levels of theSSRI. It can be argued that they may be superior to 5-HT1A receptor antagonists in hasteningresponse to antidepressant treatment. Moreover, they may be beneficial for patients with lowSSRI plasma levels, viz. rapid metabolizers. An additional argument for this combinationcould be the use of a lower dose of the SSRI, which may diminish side effects.In view of the latter considerations we expect that pharmacokinetics will become increasinglyimportant in future clinical studies.

Assessment of pharmacokinetic parameters is also indispensable in pre-clinical studies. Forexample, desensitization of somatodendritic 5-HT1A autoreceptors is believed to be the result

Chapter 3

88

of chronic activation of this receptor type. It can be argued that SSRI plasma concentrationsshould be kept within the therapeutic window to insure permanent activation. Evaluation ofchronic treatment studies suggests a relation between the latter criterion and desensitizationof 5-HT1A autoreceptors (Chaput et al., 1986; Moret and Briley, 1990; Invernizzi et al., 1994;Moret and Briley, 1996). Authors who failed to maintain SSRI levels within the therapeuticwindow, consequently failed to observe effects upon chronic treatment (Hjorth andAuerbach, 1994; Auerbach and Hjorth, 1995; Bosker et al., 1995; Gundlah et al., 1997).One might suppose that the relation found in the present study between 5-HT1A and 5-HT1B

receptor antagonist’s effects and SSRI plasma levels may be indicative for the occurrence ofdesensitization processes. Following this reasoning, 5-HT1B receptors would desensitizeirrespective of SSRI plasma levels. This happens not to be the case. Desensitization ofterminal 5-HT1B receptors is even more controversial than is the case with 5-HT1A

autoreceptors (Chaput et al., 1986; Auerbach and Hjorth, 1995; Bosker et al., 1995; Bosker etal., 1995). Apparently, differences in animal strain, SSRI and brain area also play animportant role in explaining discrepant findings.

4.4 ConclusionThe effect of citalopram on extracellular 5-HT is potentiated by a 5-HT1A and a 5-HT1B

receptor antagonist, albeit in a different dose range of the SSRI. Moreover, the studyindicates that assessment of SSRI pharmacokinetics could be important for evaluating theirpharmacological effects in relation to the clinical effects. It should be realized thatextrapolation to human studies should be done carefully, taking into account differences inbrain SSRI concentrations, receptor phamacology and neuroanatomy.

Acknowledgments:The skillful assistance of M. Mensonides-Harsema and J.B. de Vries is gratefullyacknowledged. We wish to thank “Epilepsie centrum Kempenhagen” and H.M.H.G. Cremersfor their kind donation of HPLC equipment.


89

References

Arborelius, L., Hook, B.B., Hacksell, U., Svensson, T.H., 1994. The 5-HT1A receptorantagonist (S)-UH-301 blocks the (R)-8-OH-DPAT-induced inhibition of serotonergic dorsalraphe cell firing in the rat. J. Neural Transm. Gen. Sect. 96, 179-186.Artigas, F., Perez, V., Alvarez, E., 1994. Pindolol induces a rapid improvement of depressedpatients treated with serotonin reuptake inhibitors. Arch. Gen. Psychiatry 51, 248-251.Auerbach, S.B., Hjorth, S., 1995. Effect of chronic administration of the selective serotonin(5-HT) uptake inhibitor citalopram on extracellular 5-HT and apparent autoreceptorsensitivity in rat forebrain in vivo. Naunyn Schmiedebergs Arch. Pharmacol. 352, 597-606.Baumann, P., 1992. Clinical pharmacokinetics of citalopram and other selective serotonergicreuptake inhibitors (SSRI). Int. Clin. Psychopharmacol. 6 Suppl 5, 13-20.Bel, N., Artigas, F., 1992. Fluvoxamine preferentially increases extracellular 5-hydroxytryptamine in the raphe nuclei: an in vivo microdialysis study. Eur. J. Pharmacol 229,101-103.Berman, R.M., Darnell, A.M., Miller, H.L., Anand, A., Charney, D.S., 1997. Effect ofpindolol in hastening response to fluoxetine in the treatment of major depression: a double-blind, placebo-controlled trial. Am. J. Psychiatry 154, 37-43.Blier, P., de Montigny, C., Chaput, Y., 1987. Modifications of the serotonin system byantidepressant treatments: implications for the therapeutic response in major depression. J.Clin. Psychopharmacol. 7, 24S-35S.Bosker, F.J., Donker, M.G., Klompmakers, A.A., Kurata, K., Westenberg, H.G., 1994. 5-Hydroxytryptamine release in dorsal hippocampus of freely moving rats: modulation bypindolol. Prog. Neuropsychopharmacol. Biol. Psychiatry 18, 765-778.Bosker, F.J., Klompmakers, A.A., Westenberg, H.G., 1995. Effects of single and repeated oraladministration of fluvoxamine on extracellular serotonin in the median raphe nucleus anddorsal hippocampus of the rat. Neuropharmacology 34, 501-508.Bosker, F.J., van Esseveldt, K.E., Klompmakers, A.A., Westenberg, H.G., 1995. Chronictreatment with fluvoxamine by osmotic minipumps fails to induce persistent functionalchanges in central 5-HT1A and 5-HT1B receptors, as measured by in vivo microdialysis indorsal hippocampus of conscious rats. Psychopharmacology Berl. 117, 358-363.Chaput, Y., de Montigny, C., Blier, P., 1986. Effects of a selective 5-HT reuptake blocker,citalopram, on the sensitivity of 5-HT autoreceptors: electrophysiological studies in the ratbrain. Naunyn Schmiedebergs Arch. Pharmacol. 333, 342-348.Davidson, C., Stamford, J.A., 1997. Chronic paroxetine desensitises 5-HT1D but not 5-HT1B

autoreceptors in rat lateral geniculate nucleus. Brain. Res. 760, 238-242.Fredricson Overø, K., 1982. Kinetics of citalopram in test animals; drug exposure in safetystudies. Prog. Neuropsychopharmacol. Biol. Psychiatry. 6, 297-309.Fuller, R.W., 1994. Uptake inhibitors increase extracellular serotonin concentration measuredby brain microdialysis. Life Sci. 55, 163-167.Gobert, A., Rivet, J.M., Cistarelli, L., Millan, M.J., 1997. Potentiation of the fluoxetine-induced increase in dialysate levels of serotonin (5-HT) in the frontal cortex of freely moving

Chapter 3

90

rats by combined blockade of 5-HT1A and 5-HT1B receptors with WAY 100,635 and GR127,935. J. Neurochem. 68, 1159-1163.Gundlah, C., Hjorth, S., Auerbach, S.B., 1997. Autoreceptor antagonists enhance the effect ofthe reuptake inhibitor citalopram on extracellular 5-HT: this effect persists after repeatedcitalopram treatment. Neuropharmacology 36, 475-482.Gundlah, C., Simon, L.D., Auerbach, S.B., 1998. Differences in hypothalamic serotoninbetween estrous phases and gender: an in vivo microdialysis study. Brain Research 785, 91-96.Hjorth, S., 1993. Serotonin 5-HT1A autoreceptor blockade potentiates the ability of the 5-HTreuptake inhibitor citalopram to increase nerve terminal output of 5-HT in vivo: amicrodialysis study. J. Neurochem. 60, 776-779.Hjorth, S., Auerbach, S.B., 1994. Lack of 5-HT1A autoreceptor desensitization followingchronic citalopram treatment, as determined by in vivo microdialysis. Neuropharmacology33, 331-334.Hjorth, S., Bengtsson, H.J., Milano, S., 1996. Raphe 5-HT1A autoreceptors, but notpostsynaptic 5-HT1A receptors or beta-adrenoceptors, restrain the citalopram-induced increasein extracellular 5-hydroxytryptamine in vivo. Eur. J. Pharmacol. 316, 43-47.Hjorth, S., Westlin, D., Bengtsson, H.J., 1997. WAY100635-induced augmentation of the 5-HT-elevating action of citalopram: relative importance of the dose of the 5-HT1A

(auto)receptor blocker versus that of the 5-HT reuptake inhibitor. Neuropharmacology 36,461-465.Hyttel, J., 1994. Pharmacological characterization of selective serotonin reuptake inhibitors(SSRIs). Int. Clin. Psychopharmacol. 9 Suppl 1, 19-26.Invernizzi, R., Bramante, M., Samanin, R., 1994. Chronic treatment with citalopramfacilitates the effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT1A receptors. Eur. J. Pharmacol. 260, 243-246.Invernizzi, R., Velasco, C., Bramante, M., Longo, A., Samanin, R., 1997. Effect of 5-HT1A

receptor antagonists on citalopram-induced increase in extracellular serotonin in the frontalcortex, striatum and dorsal hippocampus. Neuropharmacology 36, 467-473.Kreiss, D.S., Lucki, I., 1994. Differential regulation of serotonin (5-HT) release in thestriatum and hippocampus by 5-HT1A autoreceptors of the dorsal and median raphe nuclei. J.Pharmacol. Exp. Ther. 269, 1268-1279.Kreiss, D.S., Lucki, I., 1995. Effects of acute and repeated administration of antidepressantdrugs on extracellular levels of 5-hydroxytryptamine measured in vivo. J. Pharmacol. Exp.Ther. 274, 866-876.Kreiss, D.S., Lucki, I., 1997. Chronic administration of the 5-HT1A receptor agonist 8-HODPAT differentially desensitizes 5-HT1A autoreceptors of the dorsal and median raphenuclei. Synapse 25, 107-116.Le Poul, E., Laaris, N., Doucet, E., Laporte, A.M., Hamon, M., Lanfumey, L., 1995. Earlydesensitization of somato-dendritic 5-HT1A autoreceptors in rats treated with fluoxetine orparoxetine. Naunyn Schmiedebergs Arch. Pharmacol. 352, 141-148.McAskill, R., Mir, S., Taylor, D., 1998. Pindolol augmentation of antidepressant therapy. Br.J. Psychiatry 173, 203-208.


91

McQuade, R., Sharp, T., 1997. Functional mapping of dorsal and median raphe 5-hydroxytryptamine pathways in forebrain of the rat using microdialysis. J. Neurochem. 69,791-796.Moret, C., Briley, M., 1990. Serotonin autoreceptor subsensitivity and antidepressant activity.Eur. J. Pharmacol. 180, 351-356.Moret, C., Briley, M., 1996. Effects of acute and repeated administration of citalopram onextracellular levels of serotonin in rat brain. Eur. J. Pharmacol. 295, 189-197.Øyehaug, E., Ostensen, E.T., Salvesen, B., 1982. Determination of the antidepressant agentcitalopram and metabolites in plasma by liquid chromatography with fluorescence detection.J. Chromatogr. 227, 129-135.Pauwels, P.J., Palmier, C., 1995. Functional effects of the 5-HT1D receptor antagonist GR127,935 at human 5-HT1D alpha, 5-HT1D beta, 5-HT1A and opossum 5-HT1B receptors. Eur. J.Pharmacol. 290, 95-103.Perez, V., Gilaberte, I., Faries, D., Alvarez, E., Artigas, F., 1997. Randomised, double-blind,placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment.Lancet 349, 1594-1597.Rollema, H., Clarke, T., Sprouse, J.S., Schulz, D.W., 1996. Combined administration of a 5-hydroxytryptamine (5-HT)1D antagonist and 5-HT reuptake inhibitor synergistically increases5-HT release in guinea pig hypothalamus in vivo. Journal of neurochemistry 67, 2204-1996.Romero, L., Artigas, F., 1997. Preferential potentiation of the effects of serotonin uptakeinhibitors by 5-HT1A receptor antagonists in the dorsal raphe pathway: role ofsomatodendritic autoreceptors. J. Neurochem. 68, 2593-2603.Schoups, A.A., De Potter, W.P., 1988. Species dependence of adaptations at the pre- andpostsynaptic serotonergic receptors following long-term antidepressant drug treatment.Biochem. Pharmacol. 37, 4451-4460.Sharp, T., Umbers, V., Gartside, S.E., 1997. Effect of a selective 5-HT reuptake inhibitor incombination with 5-HT1A and 5-HT1B receptor antagonists on extracellular 5-HT in rat frontalcortex in vivo. Br. J. Pharmacol. 121, 941-946.Sproule, B.A., Naranjo, C.A., Brenmer, K.E., Hassan, P.C., 1997. Selective serotoninreuptake inhibitors and CNS drug interactions. A critical review of the evidence. ClinPharmacokinet. 33, 454-471.Sprouse, J., Reynolds, L., Rollema, H., 1997. Do 5-HT1B/1D autoreceptors modulate dorsalraphe cell firing? In vivo electrophysiological studies in guinea pigs with GR127935.Neuropharmacology 36, 559-567.Starkey, S.J., Skingle, M., 1994. 5-HT1D as well as 5-HT1A autoreceptors modulate 5-HTrelease in the guinea-pig dorsal raphe nucleus. Neuropharmacology 33, 393-402.

Chapter 3

92

Desensitisation of serotonergic autoreceptors upon pharmacokinetically monitored chronic treatment withcitalopram

93

Chapter 4

Desensitisation of serotonergic autoreceptors upon

pharmacokinetically monitored chronic treatment

with citalopram.

Chapter 4

94

Abstract

Rats were chronically treated with the selective serotonin re-uptake inhibitor citalopram [1-(3-dimethylaminopropyl)-1-(4-fluorophenyl)-5-phtalancarbonitril], by means of osmoticminipumps. Using an infusion concentration of 50 mg/ml citalopram, steady state plasmaconcentrations of approximately 0.3 µM citalopram were maintained for 15 days. Citalopramplasma levels dropped below pharmacologically active concentrations 48 h after removal ofthe minipumps. Although chronic treatment with citalopram did induce an attenuatedresponse by extracellular levels of 5-HT (5-hydroxytryptamine) after systemic administrationof the 5-HT1A receptor agonist 8-OH-DPAT (8-hydroxy-2-(di-n-propylamino)tetralin), noeffect of chronic citalopram treatment was observed when 5-HT1B receptor function wasevaluated with a local infusion of 5-HT1B/D receptor agonist, sumatriptan (3-[2-dimethylamino]ethyl-N-methyl-1H-indole-5methane sulphonamide). Controversially, noaugmentation of the increase of 5-HT levels was observed upon systemic administration ofcitalopram. It is concluded that, although chronic treatment with citalopram does inducedesensitisation of 5-HT1A receptors, the absence of augmented effects of citalopram on 5-HTlevels indicates that other mechanisms compensate for the loss of autoreceptor control.


95

1. Introduction

Selective serotonin re-uptake inhibitors increase brain extracellular 5-HT (5-hydroxytryptamine) levels immediately upon administration (Fuller, 1994). The concurrentantidepressant effect in patients is, however, delayed for several weeks (Baumann, 1992).

Changes in brain pharmacology during chronic treatment of animals with selective serotoninre-uptake inhibitors has been a topic of research for many years. Most of the studiesinvestigating adaptation of the serotonergic system upon chronic treatment of animals withselective serotonin re-uptake inhibitors have focused on the evaluation of desensitisation of 5-HT autoreceptors.

Somatodendritic 5-HT1A autoreceptors decrease neuronal firing upon activation (Arboreliuset al., 1994). Several authors have shown that selective serotonin re-uptake inhibitors activatethese receptors indirectly by increasing extracellular levels of 5-HT in raphe nuclei. Theextent of this activation, however, is related to the dose of selective serotonin re-uptakeinhibitor used (Hjorth et al., 1997, Cremers et al., 2000). Although some authors reported thatthe decrease of extracellular 5-HT after systemic administration of the 5-HT1A receptoragonist, 8-OH-DPAT (8-hydroxy-2-(di-n-propylamino)tetralin), was attenuated upon chronictreatment of animals with selective serotonin re-uptake inhibitors (Invernizzi et al., 1994;Kreiss and Lucki, 1995; Le Poul et al., 1995), others have failed to observe similar effects(Hjorth and Auerbach, 1994; Bosker et al., 1995a,b).

Like the discrepancies regarding possible desensitisation of 5-HT1A autoreceptors,desensitisation of 5-HT1B receptors located on the serotonergic nerve terminals, is alsosomewhat controversial. Although an initial electrophysiological study provided evidence fordesensitisation of 5-HT1B receptors upon chronic treatment with selective serotonin re-uptakeinhibitors (Chaput et al., 1986), various studies using microdialysis could not confirm thisconclusion (Auerbach and Hjorth, 1995; Bosker et al., 1995a,b; Moret and Briley, 1996;Davidson and Stamford, 1997).

There might be several reasons for these controversial results. First, species and animalstrains may differ in their susceptibility to desensitisation (Schoups and De Potter, 1988).Secondly, as elimination of substances in animals, especially rodents, is much faster than inhumans, chronic treatment regimens are characterised by high and multiple dosing of thedrugs. Although a relation between the antidepressant effect and plasma levels of theselective serotonin re-uptake inhibitors is not well established, clinically effective plasmalevels have been reported (Baumann, 1992). In this respect we have recently demonstratedthat selective serotonin re-uptake inhibitor induced 5-HT1A activation is critically dependenton the plasma levels of the selective serotonin re-uptake inhibitor (Cremers et al., 2000).However, in animal studies on chronic treatment with selective serotonin re-uptake inhibitors,little attention was given to the role of plasma levels. Therefore the controversial data

Chapter 4

96

regarding desensitisation of autoreceptors might be explained by inadequacy of plasma levelsof the selective serotonin re-uptake inhibitor during the chronic treatment.

In the present microdialysis study, desensitisation of autoreceptors was studied after chronictreatment of rats with citalopram [1-(3-dimethylaminopropyl)-1-(4-fluorophenyl)-5-phtalancarbonitril]. To achieve stable plasma levels, citalopram was administered by meansof osmotic minipumps. Plasma concentrations of citalopram and its less potent des- anddidesmethyl metabolites were followed during the 15-day period. After a wash-out period of2 days was determined, possible changes in 5-HT receptor sensitivity were challenged byrecording the effects of administration of the 5-HT1A receptor agonist, 8-OH-DPAT, the 5-HT1B/D receptor agonist, sumatriptan (3-[2-dimethylamino]ethyl-N-methyl-1H-indole-5methane sulphonamide), or citalopram itself, on extracellular 5-HT in the ventralhippocampus.


97


2.1. Animals and drug administration

Male albino rats of a Wistar-derived strain (285-320 g; Harlan, Zeist, The Netherlands) wereused for the experiments. The rats were housed in plastic cages (35 x 35 x 40 cm), and hadfree access to food and water. The experiments conformed with the declarations of Helsinkiand were approved by the animal care committee of the Faculty of Mathematics and NaturalScience of the University of Groningen .

The following drugs were used: citalopram hydrobromide [1-(3-dimethylaminopropyl)-1-(4-fluorophenyl)-5-phtalancarbonitril] , des-, didesmethyl metabolites and internal standard [1-(3-dimethylaminopropyl)-1-(4-bromophenyl)-5-phtalancarbonitril] (kindly donated byLundbeck (Denmark), courtesy of Dr. Sanchez), (+)-8-OH-DPAT (8-hydroxy-2-(di-n-propylamino)tetralin) (RBI, Natick,USA), sumatriptan (3-[2-dimethylamino]ethyl-N-methyl-1H-indole-5methane sulphonamide). (Glaxo Wellcome,UK.).

2.2. Surgery

Osmotic minipumps (2ML2 Alzet, USA, 5 µl/h, 2 weeks) were filled with 50 mg/mlcitalopram hydrobromide under aseptic conditions. During brief isoflurane anaesthesia (2.5%, 400ml/min N2O, 600 ml/min O2), minipumps were implanted subcutaneously on the leftside of the back of the rat. In a subgroup, 4 rats were provided with an additional jugular veincanula in order to determine plasma citalopram levels during and after 15 days of treatment.

After 15 days, the rats were anaesthetised with chloral hydrate (400 mg/kg), and the osmoticminipumps were removed. The remaining subcutaneous cavity was flushed twice with 5 mlof sterile saline. Home-made I-shaped microdialysis probes, made of polyacrylonitrile /sodium methyl sulfonate copolymer dialysis fiber (i.d. 220 µm, o.d. 0.31 µm, AN 69, Hospal,Italy) were implanted bilaterally in the ventral hippocampus. The exposed length of themembranes was 4 mm. Lidocaine-HCl, 10 % (m/v) was used for local anaesthesia. The ratswere placed in a stereotaxic frame (Kopf, USA), and probes were inserted into the ventralhippocampus (co-ordinates: IA: +3.7 mm, lateral : +4.8 mm, ventral: -8.0 mm from the duramater, Paxinos and Watson, 1982) and secured with dental cement.

2.3. Pharmacokinetic and microdialysis experiments

Blood samples (0.35 ml) were drawn on day 3, 7, 12, 15,16,17 after implantation of theosmotic minipumps. Samples were transferred to 1.5-ml Eppendorf vials, containing 5 µlheparin (500 IE/ml saline), mixed and immediately centrifuged for 15 min at 3,000 rpm at 4ºC (Chillspin, MSE, England). Plasma samples were stored at -80 ºC until analysis.

Chapter 4

98

The rats were allowed to recover for at least 48 h after insertion of the microdialysis probes.The probes were perfused with Ringer solution containing 147 mM NaCl, 3.0 mM KCl, 1.2mM CaCl2, and 1.2 mM MgCl2, at a flowrate of 1.5 µl / min (Harvard apparatus, SouthNatick, Ma., USA). Samples were collected on-line in a 20-µl loop and injected automaticallyonto the column every 15 min.

2.4. Analysis

2.4.1. Serotonin5-HT was analysed using high pressure liquid chromatography (HPLC) with electrochemicaldetection. The HPLC pump (Shimadzu LC-10 AD liquid chromatograph) was connected to areversed phase column (Phenomenex hypersil 3 : C18 , 3 µm, 100 x 2.0 mm, C18, Bester,Amstelveen, the Netherlands) followed by an electrochemical detector (Antec Leyden,Leiden, the Netherlands), working at a potential setting of 500 mV vs. Ag/AgCl reference.The mobile phase consisted of 5 g/l di-ammoniumsulfate, 500 mg/l ethylene diamino tetraacetic acid (EDTA), 50 mg/l heptane sulphonic acid, 30 µl/l of triethylamine, and 4.5 % v/vmethanol, at a pH of 4.65. The flow-rate of the mobile phase was 0.4 ml/min. The detectionlimit was 0.2 fmol 5-HT per 20-µl sample (signal to noise ratio 2).

2.4.2. CitalopramCitalopram was measured according to Øyehaug et al. (1982), with minor modifications.Briefly, 75 µl of the internal standard LU 10-202 (2 µM) and 30 µl of 1N NaOH were addedto 150 µl plasma samples. Samples were extracted twice by mechanically shaking for 3 minwith 3 ml of diethyl ether. The ether layers were then transferred to 10-ml evaporating tubes,and 150 µl of 0.1 N HCl was added. The ether was evaporated in a water-bath at 40 °C undera stream of nitrogen. The HCl layer was washed once with 0.5 ml ether. Samples, 50 µl, wereinjected onto the column. Extraction recovery of citalopram, metabolites and internalstandard was approximately 95 %. Plasma levels were corrected for recovery of the internalstandard.An HPLC/auto-injector (1084B Liquid Chromatograph, Hewlett-Packard) was used, incombination with a fluorescence detector (470 Scanning Fluorescence detector, Waters,England) operating at an absorption wavelength of 240 nm, an emission wavelength of 296nm, and a slit-width of 12 nm. Separation was performed using a Supelcosil HPLC column (5µm, C18, 250 x 46 mm, Supelco, the Netherlands), at ambient temperature. The mobilephase consisted of 46 % v/v acetonitrile, 54 % v/v potassium dihydrogen phosphate buffer(4.3 g/l), and 1mM tetramethylammonium, at pH 3.0. The flowrate in this system was 0.75ml/min. The detection limits for “on column injections” of citalopram, desmethyl-citalopramand didesmethyl-citalopram were 5, 5 and 3 nM, respectively (signal to noise = 2).


99

2.5. Data presentation and statistics

Four consecutive microdialysis samples with less than 20 % variation were taken as control,and set at 100 %. The data are presented as percentages of the control level (mean + S.E.M.).Statistical analysis was performed using Sigmastat for Windows (Jandel Corporation).Treatment effects were compared using two-way analysis of variance (ANOVA) for repeatedmeasurements, followed by the Mann-Whitney Rank Sum Test. Significance was set atP<0.05.

Chapter 4

100

3. Results

3.1. Citalopram plasma levelsInfusion rate calculations, based on acute subcutaneous administration of citalopram(Clcalculated= 0.142 l/min), predicted that plasma levels of 500 nM would be obtained when theosmotic minipumps were filled with 100 mg/ml. Empirical evaluation showed, however, thatthese plasma levels were around 1 µM (data not shown).Figure 1 shows that treatment of rats with osmotic minipumps (50 mg/ml citalopram),resulted in stable citalopram plasma levels of around 0.3 µM (0.30 + 0.03 µM, Fig. 1).Desmethyl- and di-desmethyl metabolites were found to be around 0.065 and 0.2 µMrespectively).Upon removal of the minipumps, a gradual decline in plasma levels of citalopram andmetabolites was observed. After 48 h, the plasma levels of citalopram had dropped to below 9nM.

Figure 1. Plasma levels of citalopram (�), desmethylmetabolite (�), and didesmethylmetabolite (�) (n=4), during subcutaneous infusion of 0.25 mg/h citalopram by means ofosmotic minipumps. The arrow denotes removal of the minipumps.

3.2. Basal 5-HT LevelsThe basal levels of 5-HT in dialysates did not differ between citalopram- and saline-treatedrats 48 h after removal of the minipumps: saline-treated rats 6.66 + 0.82 fmol/sample (n=26);citalopram-treated rats 7.26 + 0.56 fmol/sample (n=27); P=0.219, n.s.).

3.3. 8-OH-DPAT challengeThe effect of 8-OH-DPAT (0.1 mg/kg s.c.), that clearly decreased 5-HT levels in ventralhippocampus, was completely abolished after chronic treatment of the animals with

2 4 6 8 10 12 14 16 181E-3

0.01

0.1

1 Max. clinical plasmalevel

pharmacological inactive plasma level

Min. clinical plasmalevel

Con

cent

ratio

n µ M

Time (days)


101

citalopram (F(1,108)=2.65 P<0.05). Post hoc analysis revealed significant differences fromt=30 to t=90 min (Fig.2).After subcutaneous administration of a higher dose of 8-OH-DPAT (0.2 mg/kg), 5-HT levelsdecreased to about 40 % of the controls. Chronic pretreatment with citalopram did notprevent the maximum effect, but attenuated the effect on 5-HT levels during the reboundphase (F(1,83)=2.64 P<0.05. The effects were significantly different from t=90 to t=135 (Fig.3).

Figure 2. Effect of chronic treatment of rats for 15 days with citalopram on 5-HT1Areceptoragonist (+)-8-HO-DPAT 0.1 mg/kg-induced decrease in 5-HT levels. Citalopram-treated (n =5; �), saline-treated (n = 8 ; �). * P<0.05.

3.4. Sumatriptan challengeLocal infusion of 500 nM of sumatriptan in the ventral hippocampus for 60 min decreased thelevels of 5-HT to about 50 % of the controls. No significant differences were observedbetween citalopram- and saline-treated animals (F(1,128)=0.389, n.s.) (Fig. 4).Local infusion of 1000 nM of sumatriptan for 60 min decreased the levels of 5-HT to about30 % of the controls. Again, no significant differences were observed between citalopram-and saline-treated animals (F(1,175)=1.549, n.s.) (Fig. 5).

3.5. Citalopram challengeNo differences were found between treated and untreated animals after systemicadministration of 10 µmol/kg citalopram (F(1,98)=0.959, n.s.). The maximum increase inboth groups was about 450 %. No differences in time-effect were observed (Fig. 6).

-100 -50 0 50 100 1500

20

40

60

80

100

120

140

160

180

200

*

**

*

*

5-H

T %

Bas

al L

eve

l

Time (min)

Chapter 4

102

Figure 3. Effect of chronic treatment of rats for 15 days with citalopram on 5-HT1A receptoragonist (+)-8-HODPAT 0.2-mg/kg induced decrease in 5-HT levels. Citalopram-treated (n =6 ; �), saline-treated (n = 6 ; �). * P<0.05.

-100 -50 0 50 100 150 2000

20

40

60

80

100

120

140

160

180

200

*

*

**

% b

asa

l Le

vel

Time (min)


103

4. Discussion

Desensitisation of 5-HT autoreceptors upon chronic treatment of animals with selectiveserotonin re-uptake inhibitors has been a topic of interest for several years. Although variousstudies have found evidence for the occurrence of these processes, others have consistentlyfailed to show any of these effects. As pharmacokinetic measurements are not common inpharmacological research, dose regimens used were based on estimates rather than onpharmacokinetic data. The present study showed that, if a properly pharmacokineticallyvalidated chronic treatment regimen is used, 5-HT1A autoreceptor desensitisation can beobserved.

Figure 4. Effect of chronic treatment of rats for 15 days with citalopram on the decrease in 5-HT levels produced by local infusion of 500 nM of 5-HT1B/D receptor agonist, sumatriptan.Citalopram-treated (n = 4; �), saline-treated (n = 8; �).

Recent work has shown that selective serotonin re-uptake inhibitor induced activation of 5-HT autoreceptors is critically related to the dose and concurrent plasma- and brain levels ofthe selective serotonin re-uptake inhibitor as well as the autoreceptor antagonist (Hjorth et al.,1997). In a previous study, we have shown that activation of 5-HT1A autoreceptors is onlyobserved on acute administration of citalopram when plasma levels of citalopram are above150 nM. Selective serotonin re-uptake inhibitor-induced activation of 5-HT1B receptors wasalready observed at much lower plasma levels of the selective serotonin re-uptake inhibitor(Cremers et al., 2000).As elimination of exogenous substances is in general much faster in rodents than in humans,chronic treatment of animals with selective serotonin re-uptake inhibitors is characterised bydaily multiple high dosing. Consequently, high plasma levels followed by a rapid decline tolow and inactive concentrations might explain why several authors have failed to observe anydesensitisation of 5-HT autoreceptors. As plasma levels of the selective serotonin re-uptake

-100 -50 0 50 100 150 200 2500

20

40

60

80

100

120

140

160

180

200

Sum 500nM

% B

asal

Lev

el

Time (min)

Chapter 4

104

inhibitor are critically important and fluctuations in plasma levels are inherent to systemicadministration, the present study made use of osmotic minipumps in an attempt to achievestable plasma levels, which were shown to produce autoreceptor activation .

Figure 5. Effect of chronic treatment of rats for 15 days with citalopram on the decrease in 5-HT levels produced by local infusion of 1000 nM of 5-HT1B/D receptor agonist, sumatriptan.Citalopram-treated (n = 8; �), saline-treated (n = 8; �).

From pharmacokinetic data obtained after acute subcutaneous administration of citalopram,we calculated which infusion concentration was needed to establish citalopram plasma levelsknown to activate both autoreceptors. From the time-concentration profile of acutelyadministratered 10 µmol/kg citalopram in rats who received surgery under isofluraneanaesthesia, a clearance (Cl) of 0.142 l/min was calculated (data not shown). Calculating theinfusion rate (Ro) with the equation Ro=Css X Cl, we first evaluated minipumps filled with100 mg/ml saline (aiming at a steady state (Css) of 500 nM in 300-gram rats). The plasmalevels, however, were found to be around 1 µM, which was above clinically effectiveconcentrations. Lowering the infusion concentration to 50 mg/ml was found to produce stableplasma levels for 2 weeks, ensuring chronic activation of autoreceptors. In addition, theselevels were also within the limits of clinically effective plasma levels of citalopram inhumans (0.12 µM-0.84 µM (Baumann, 1992)). Desmethyl and didesmethyl metabolites ofcitalopram were quantified during chronic treatment. The drug-to-metabolites ratios observedin rats were somewhat different from those observed in humans. Whereas, in rats, the ratiobetween parent, desmethyl and didesmethyl was about 10-1-5 respectively, thecorresponding ratios in humans are about 10-5-1, respectively. However, as the affinity of themetabolites for the re-uptake site is at least 10 times lower than that of the parent compound,their effects, if any, will not be predominant (Hyttel, 1994).

-100 -50 0 50 100 150 200 2500

20

40

60

80

100

120

140

160

180

200

Sum 1000 nM

% B

asal

Lev

el

Time (min)


105

Figure 6. Effect of chronic treatment of rats for 15 days with citalopram on citalopraminduced increase in 5-HT levels. Citalopram-treated (n = 4; �), saline-treated (n = 5; �).

Upon removal of the minipumps the T1/2 value for elimination was dramatically increased (2days, vs 2 h (Cremers et al., 2000). Apparently, the pharmacokinetic parameters of citalopramchange after chronic treatment because of redistribution in the body. This dramatic decreasein elimination has consequences for chronic experiments, as pharmacological challenges ontermination of the treatment should be postponed until selective serotonin re-uptake inhibitorconcentrations are below pharmacologically active levels. In a previous study (Cremers et al.,2000) we determined the threshold of the pharmacologically active plasma level to be around9 nM. Forty eight hours after removal of the minipumps, the plasma levels of citalopramwere indeed below this border. Although the didesmethyl metabolite did not reach this limitwithin 48 h, due to lower affinity for the 5-HT re-uptake side (Hyttel, 1994), it is assumed tobe without interfering effects. It is of interest to note that the basal 5-HT levels 48 h afterremoval of the minipumps did not different between saline-and citalopram-treated rats, whichalso indicates the absence of interfering amounts of residual citalopram.Systemic administration of the putative 5-HT1A receptor agonist, 8-OH-DPAT, had markeddose-dependent effects after chronic treatment with citalopram. Whereas the effect of a 0.1mg/kg dose was completely abolished after chronic treatment, administration of 0.2 mg/kgproduced near similar maximal responses, with the citalopram-treated rats showing a fasterrecovery of 5-HT levels. Apparently, chronic treatment of animals with citalopram induces ashift in the response curve of 5-HT1A receptor agonists, indicative of somatodendritic 5-HT1A

receptor desensitisation. Although as mentioned above, not all authors have observed thisdesensitisation upon chronic treatment. Several authors also found attenuated effects of 8-OH-DPAT after chronic treatment with selective serotonin re-uptake inhibitors (Invernizzi etal., 1994; Kreiss and Lucki, 1995; Le Poul et al., 1995). Extrapolating the data on citaloprampharmacokinetics to the chronic treatment regimens used in these latter experiments, showsthat researchers using higher and frequent doses were more likely to find these effects than

-100 -50 0 50 100 1500

100

200

300

400

500

600

700

800

% B

asal

Lev

el

Time (min)

Chapter 4

106

when lower doses were used (Chaput et al., 1986; Moret and Briley, 1990; Invernizzi et al.,1994; Moret and Briley, 1996). Although in the present study the effect of 8-OH-DPAT wasindeed abolished, or at least attenuated, it cannot be neglected that this attenuation could havebeen due to decreased availability of the agonist. Differences in elimination due to inductionof kinetics, however, should not be very relevant, as citalopram is described to be withoutmajor interaction with other drugs (Sproule et al., 1997). The possibility of bluntedabsorption due to tissue necrosis produced by the osmotic minipump was eliminated byinjection of the agonist on the side contralateral to the area were the osmotic minipump waspositioned. Although there are no reasons to suspect lower bioavailability of the agonist incitalopram- vs. saline-treated rats, local application in the raphe, or pharmacokineticexperiments should elucidate this point.Local administration of two concentrations of the 5-HT1B/D receptor agonist, sumatriptan,decreased 5-HT levels equally in citalopram- and saline-treated rats. Chronic treatment of ratswith plasma levels which were shown to activate 5-HT1B receptors apparently does notdesensititize these autoreceptors (Cremers et al., 2000). Chaput et al. (1986), using an indirectelectrophysiological approach, found evidence for desensitisation of this receptor, but nostudy using the microdialysis method has ever shown desensitisation of this receptor(Auerbach and Hjorth, 1995; Bosker et al., 1995a,b; Moret and Briley, 1996; Davidson andStamford, 1997). As 5-HT1B receptors have been shown to be activated by low doses ofselective serotonin re-uptake inhibitor (Rollema et al., 1996, Cremers et al., 2000), fasterdesensitisation than of 5-HT1A receptors would be expected. Rapid resensitisation of the 5-HT1B receptor is a possible explanation for the observed absence of effects. As the currentprotocol was based on challenge of the system after elimination of the selective serotonin re-uptake inhibitor, such effects cannot be excluded.Although terminal 5-HT1B receptors apparently were not desensitised in our setup, the 5-HT1A receptors were. This would mean that, upon challenge with a dose of selectiveserotonin re-uptake inhibitor which activates 5-HT1A as well as 5-HT1B receptors, selectiveserotonin re-uptake inhibitor-evoked increases in 5-HT levels would be augmented incitalopram-treated animals. However, in our experiments this effect was not observed. Theremight be several reasons for this observation. First, since citalopram treatment induced a shiftin 5-HT1A receptor agonist-induced response, rather than abolished the effect, any remaining5-HT1A autoreceptor function might have sufficed to preserve proper autoreceptor function.Treatment of animals for longer periods should elucidate whether this would lead toenhancement of the shift in receptor function, with consecutive augmentation of selectiveserotonin re-uptake inhibitor-induced effects. Second, as desensitisation of the serotonintransporter has been described after chronic administration of selective serotonin re-uptakeinhibitor, this effect might counteract the desensitisation of autoreceptors (Blier andBouchard, 1994). Chronic treatment with selective serotonin re-uptake inhibitor would thenbe a “resetting of the system”, rather than an enhancement of serotonergic neurotransmission.Based on this possibility, chronic treatment of animals with selective serotonin re-uptakeinhibitors followed by physiological activation of the serotonergic system should be tried inorder to unravel the function of desensitisation of the 5-HT1A autoreceptor.


107

In conclusion, the present study demonstrated that rational chronic treatment is only possibleif it is accompanied by proper pharmacokinetic validation. Several discrepancies in theliterature regarding effects after chronic treatment with drugs might be strongly related tofailure to produce relevant and stable plasma levels of drugs in animals. In addition, if stableand relevant plasma levels are ensured, pharmacological changes would be easier toextrapolate to a clinical setting.Although, in the present study, 5-HT1A desensitisation was observed, no augmented 5-HTlevels were observed after a single selective serotonin re-uptake inhibitor challenge. If thisaugmentation were not relevant for the antidepressive action, but instead, the desensitisationof 5-HT1A receptors which control the firing rate of the serotonergic system, co-administration of a 5-HT1A receptor antagonist with selective serotonin re-uptake inhibitorswould be more useful than co-administration of a 5-HT1B receptor antagonist.

Chapter 4

108

References

Arborelius, L., Hook, B.B., Hacksell, U., Svensson, T.H., 1994. The 5-HT1A receptorantagonist (S)-UH-301 blocks the qR)-8-OH-DPAT-induced inhibition of serotonergic dorsalraphe cell firing in the rat. J. Neural. Transm. Gen. Sect. 96, 179-186.Auerbach, S.B., Hjorth, S., 1995. Effect of chronic administration of the selective serotonin(5-HT) uptake inhibitor citalopram on extracellular 5-HT and apparent autoreceptorsensitivity in rat forebrain in vivo. Naunyn Schmiedebergs Arch. Pharmacol. 352, 597-606.Baumann, P., 1992. Clinical pharmacokinetics of citalopram and other selective serotonergicreuptake inhibitors (SSRI). Int. Clin. Psychopharmacol. 6 Suppl. 5, 13-20.Blier, P., Bouchard, C., 1994. Modulation of 5-HT release in the guinea-pig brain followinglong-term administration of antidepressant drugs. Br. J. Pharmacol. 113, 485-495.Bosker, F.J., Klompmakers, A.A., Westenberg, H.G., 1995a. Effects of single and repeatedoral administration of fluvoxamine on extracellular serotonin in the median raphe nucleus anddorsal hippocampus of the rat. Neuropharmacology 34, 501-508.Bosker, F.J., van Esseveldt, K.E., Klompmakers, A.A., Westenberg, H.G., 1995b. Chronictreatment with fluvoxamine by osmotic minipumps fails to induce persistent functionalchanges in central 5-HT1A and 5-HT1B receptors, as measured by in vivo microdialysis indorsal hippocampus of conscious rats. Psychopharmacology Berl. 117, 358-363.Chaput, Y., de Montigny, C., Blier, P., 1986. Effects of a selective 5-HT reuptake blocker,citalopram, on the sensitivity of 5-HT autoreceptors: electrophysiological studies in the ratbrain. Naunyn Schmiedebergs Arch. Pharmacol. 333, 342-348.Cremers, T.I.F.H., de Boer, P., Liao, Y., Bosker F.J., den Boer, J.A., Westerink, B.H.C.,Wikström, H.V., 2000. Augmentation with a 5-HT1A, but not a 5-HT1B receptor antagonistcritically depends on the dose of citalopram. A pharmacodynamic and pharmacokineticstudy. Eur. J. Pharmacol. In press.Davidson, C., Stamford, J.A., 1997. Chronic paroxetine desensitises 5-HT1D but not 5-HT1B

autoreceptors in rat lateral geniculate nucleus. Brain Res. 760, 238-242.Fuller, R.W., 1994. Uptake inhibitors increase extracellular serotonin concentration measuredby brain microdialysis. Life Sci. 55, 163-167.Hjorth, S., Auerbach, S.B., 1994. Lack of 5-HT1A autoreceptor desensitization followingchronic citalopram treatment, as determined by in vivo microdialysis. Neuropharmacology33, 331-334.Hjorth, S., Westlin, D., Bengtsson, H.J., 1997. WAY100635-induced augmentation of the 5-HT-elevating action of citalopram: relative importance of the dose of the 5-HT1A

(auto)receptor blocker versus that of the 5-HT reuptake inhibitor. Neuropharmacology 36,461-465.Hyttel, J., 1994. Pharmacological characterization of selective serotonin reuptake inhibitors(SSRIs). Int. Clin. Psychopharmacol. 9 Suppl 1, 19-26.Invernizzi, R., Bramante, M., Samanin, R., 1994. Chronic treatment with citalopramfacilitates the effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT1A receptors. Eur. J. Pharmacol. 260, 243-246.


109

Kreiss, D.S., Lucki, I., 1995. Effects of acute and repeated administration of antidepressantdrugs on extracellular levels of 5-hydroxytryptamine measured in vivo. J. Pharmacol. Exp.Ther. 274, 866-876.Le Poul, E., Laaris, N., Doucet, E., Laporte, A.M., Hamon, M., Lanfumey, L. 1995. Earlydesensitization of somato-dendritic 5-HT1A autoreceptors in rats treated with fluoxetine orparoxetine. Naunyn Schmiedebergs Arch. Pharmacol. 352, 141-148.Moret, C., Briley, M., 1990. Serotonin autoreceptor subsensitivity and antidepressant activity.Eur. J. Pharmacol. 180, 351-356.Moret, C., Briley, M., 1996. Effects of acute and repeated administration of citalopram onextracellular levels of serotonin in rat brain. Eur. J. Pharmacol. 295, 189-197.Oyehaug, E., Ostensen, E.T., Salvesen, B., 1982. Determination of the antidepressant agentcitalopram and metabolites in plasma by liquid chromatography with fluorescence detection.J. Chromatogr. 227, 129-135.Rollema, H., Clarke, T., Sprouse, J.S., Schulz, D.W., 1996. Combined administration of a 5-hydroxytryptamine (5-HT)1D antagonist and 5-HT reuptake inhibitor synergistically increases5-HT release in guinea pig hypothalamus in vivo. Journal of neurochemistry 67, 2204-1996.Schoups, A.A., De Potter, W.P., 1988. Species dependence of adaptations at the pre- andpostsynaptic serotonergic receptors following long-term antidepressant drug treatment.Biochem. Pharmacol. 37, 4451-4460.Sproule, B.A., Naranjo, C.A., Brenmer, K.E., Hassan, P.C., 1997. Selective serotoninreuptake inhibitors and CNS drug interactions. A critical review of the evidence. Clin.Pharmacokinet. 33, 454-471.

Chapter 4

110

Topical differences in serotonergic autoreceptor control; effects of chronic treatment with citalopram

111

Chapter 5

Topical differences in serotonergic autoreceptor

control; effects of chronic treatment with citalopram

Chapter 5

112

Abstract

The mechanism of action of selective serotonin reuptake inhibitors (SSRIs) has beenhypothesised to relate to the induction of enhanced serotonergic neurotransmission due todesensitisation of autoreceptors. However, little is known of regional differences inautoreceptor control.Using dual probe microdialysis the acute and chronic effects of the SSRI citalopram onserotonin release in four brain structures were studied. Increasing doses of citalopram incombination with the 5-HT1A receptor antagonist Way 100635 or the 5-HT1B/D receptorantagonist GR 127935 were used to investigate the participation of the autoreceptors in theeffects of the SSRI on serotonin release.In all four areas we found a dose dependent increase of extracellular serotonin, reaching aplateau value at citalopram doses exceeding 3 µmol/kg. Depending on the dose of citalopramthe increase in extracellular serotonin in the dorsal raphe-prefrontal cortex system appeared tobe restrained through 5-HT1A autoreceptors, however, the contribution of 5-HT1B/D receptorswas negligible. In contrast, the increase in extracellular serotonin in the median raphe-dorsalhippocampus system was counteracted by both types of autoreceptors. At low doses ofcitalopram the 5-HT1A receptor played the most prominent role, however at higher doses itwas surpassed by the 5-HT1B/D receptor.Chronic treatment produced a rather different picture; whereas the dorsal hippocampus 5-HTlevels were relatively insensitive to chronic treatment, while major changes were observed inboth the median and dorsal raphe nucleus and the median prefrontal cortex.The present study shows that the participation of autoreceptors in the effect of citalopram isdependent on the brain area and the dose of SSRI. After chronic treatment, the observedeffects can only partly be explained by functional changes of 5-HT1A and 5-HT1B receptors.


113

1. Introduction

The clinical efficacy of selective serotonin re-uptake inhibitors (SSRIs) in several psychiatricsyndromes such as depression and anxiety disorders has been established in several large-scaleclinical trials. Their therapeutic effect is typically delayed for 2 to 4 weeks and can thereforenot readily be explained by the immediate effect on serotonin re-uptake. It has beenhypothesised that the delay in clinical efficacy relates to the time needed to desensitise severalserotonin receptors throughout the brain (Blier et al. 1987).Serotonergic neurotransmission is subject to several auto-inhibitory processes. Activation of5-HT1A receptors located within the raphe nuclei inhibits serotonergic neuronal firing(Arborelius et al. 1994). Administration of SSRIs apparently increases serotonin levels withinthese nuclei to such extent that serotonergic neuronal firing is decreased through activation of5-HT1A receptors (Arborelius et al. 1995). This secondary effect counteracts the effect ofreuptake inhibition on serotonin levels. However, the persistent activation of these receptorseither directly by 5-HT1A agonists or indirectly by SSRIs, has been reported to desensitisethese receptors resulting in an increased levels of 5-HT (Invernizzi et al. 1994; Kreiss andLucki 1997). As the time period of chronic treatment with antidepressants needed fordesensitisation of 5-HT1A autoreceptors was similar to the latency time observed in patients(2-4 weeks), it was hypothesised that these adaptive changes might be partly responsible forthe delayed onset of the therapeutic effect.5-HT1B receptors are located on the nerve terminals of serotonergic neurons. Activation ofthese receptors directly inhibits the release of serotonin. Desensitisation of 5-HT1B

autoreceptors following chronic antidepressant treatment has also been reported, but thisinitial observation (Blier et al. 1988) appeared difficult to reproduce by other groups (Gobbi etal. 1997; Auerbach and Hjorth 1995; Bosker et al. 1995).Next to somatodendritic 5-HT1A receptors, the raphe nuclei have also been shown to posses 5-HT1D receptors (Pineyro et al. 1995). The precise function of these receptors has not beenconclusively shown, but part of their function may be regulation of serotonin release within theraphe nuclei (Sprouse et al. 1997; Pineyro and Blier 1996; Starkey and Skingle 1994; Pineyroet al. 1995). Research is seriously hindered by the lack of selective compounds for thisreceptor subtype. As a consequence little is known how they respond to chronic treatmentwith SSRIs.Most research on serotonin release of the serotonergic system during acute and chronictreatment with SSRIs has been focussed on the function of the autoreceptors. Nevertheless,several studies have suggested that serotonin release in terminal areas is also under control ofpostsynaptic 5-HT1A receptors through a neuronal feedback loop projecting back to the raphenuclei (Casanovas et al. 1999; Bosker et al. 1997; Ceci et al. 1994).The concept of serotonin autoreceptor desensitisation has lead several research groups topropose that co-administration of SSRIs with a 5-HT1A receptor antagonist might decrease thelatency time of SSRIs, and improve their clinical efficacy (Artigas 1993, Hjorth 1993).However, clinical evaluation of this augmentation strategy has been hindered by the absence ofclinically available selective and potent 5-HT1A antagonists. This may be one of the reasons

Chapter 5

114

that clinical studies with the mixed beta/5-HT1A receptor antagonist pindolol have yieldedcontradictory results (review Nelson 2000).Augmentation with 5-HT1A and 5-HT1B/D receptor antagonists in the treatment of depressionshow great promise. However, the exact effects of these approaches on extracellular serotoninin various brain regions, especially in relation to the dose of SSRI, are not unequivocallyestablished (Cremers et al. 2000; Romero and Artigas 1997).

Using dual probe microdialysis we have measured the effect of increasing doses of the SSRIcitalopram on extracellular serotonin in the dorsal raphe nucleus and prefrontal cortex, an areawhich is believed to be preferentialy innervated by the dorsal raphe nucleus. In addition wealso measured extracellular serotonin in the median raphe nucleus and dorsal hippocampus.Furthermore, the role of autoreceptors was studied by co-administration of citalopram with a5-HT1A or a 5-HT1B receptor antagonist.In a second part of the study, animals were chronically treated with citalopram vs. saline(osmotic minipumps, (Cremers et al. 2000)). The effect of chronic treatment with citalopramon basal and SSRI-evoked levels of 5-HT was evaluated.

The following questions were evaluated;

---First, are the citalopram dose-response curves comparable for all measured areas?---Second, are there differences in autoreceptor restrainment?---Third, are there differences in chronic treatment effects between the four brain areasmentioned and if so can they be explained by changes in autoreceptor function?


115


2.1 Animals


2.2 Drugs

The following drugs were used: Citalopram hydrobromide (kindly donated by Lundbeck(Denmark), courtesy Dr. Sanchez), Way 100635 and GR 127935 (both compound weresynthesised in our laboratory, courtesy Dr. Y. Liao and Mrs M. Mensonides). Drugs weredissolved in saline except for GR 127935, which was dissolved in saline with a drop of aceticacid. Substances were injected subcutaneously in a volume of 1 ml per kg.

2.3 Surgery

acute;Microdialysis of brain serotonin levels was performed using home made I-shaped probes,made of polyacrylonitrile / sodium methyl sulfonate copolymer dialysis fiber (i.d. 220 µm, o.d.0.31 µm, AN 69, Hospal, Italy). The exposed length of the membranes was 2 mm for dorsalhippocampus, 4 mm for prefrontal cortex, 2 mm for median raphe nucleus and 1.5 mm fordorsal raphe nucleus. Preceding surgery rats were anaesthetised by means of an intraperitonealinjection of Ketamine/xylazine 50/8 mg/kg i.p.), after premedication with 5mg/kg s.c. ofmidazolam. Lidocaine-HCl, 10 % (m/v) was used for local anaesthesia. Rats were placed in astereotaxic frame (Kopf, USA), and probes were inserted into Dorsal Raphe Nucleus (DRN, L-1.4 mm; IA: +1.2 mm, V: -7.0 mm angle 10o), Median Raphe Nucleus (MRN, L –1.4 mm;IA: +1.2 mm, V: -9.0 mm angle 10o), medial Prefrontal Cortex (PFC, L –0.9 mm; AP: +3.5mm relative to bregma; V: -6.0 mm from Dura Mater) and Dorsal Hippocampus (D Hippo, L:-2.9 mm; AP: -4.0 relative to bregma; V: -5.5 from the skull and the toothbar set at +5.0 mm(Paxinos and Watson, 1982).After insertion, probes were secured with dental cement.

Chapter 5

116

Chronic treatment;Osmotic minipumps (2ML2 Alzet, USA, 5 µl/h, 2 weeks) were filled with 50 mg/mlcitalopram hydrobromide under aseptic conditions. During brief isoflurane anaesthesia (2.5 %,400ml/min N2O, 600 ml/min O2), minipumps were implanted subcutaneously on the left sideof the back of the rat.After 15 days, the rats were anaesthetised and intracerebral probes were implanted asdescribed above. During this surgery also the osmotic minipumps were removed and theremaining subcutaneous cavity was flushed twice with 5 ml of sterile saline.


Rats were allowed to recover for at least 24 h. Probes were perfused with artificialcerebrospinal fluid containing 147 mM NaCl, 3.0 mM KCl, 1.2 mM CaCl2, and 1.2 mMMgCl2, at a flow-rate of 1.5 µl / min (Harvard apparatus, South Natick, Ma., USA). 15 minutemicrodialysis samples were collected in HPLC vials containing 7.5 µl 0.02 M acetic acid forserotonin analysis.

2.5 Serotonin analysisTwenty-µl microdialysate samples were injected via an autoinjector (CMA/200 refrigeratedmicrosampler, CMA, Sweden) into a 100 x 2.0 mm C18 Hypersil 3 µm column (Bester,Amstelveen, the Netherlands) and separated with a mobile phase consisting of 5 g/L di-ammoniumsulfate, 500 mg/L EDTA, 50 mg/L heptane sulphonic acid, 4 % methanol v/v, and30 µl/L of triethylamine, pH 4.65 at 0.4 ml/min (Shimadzu LC-10 AD) 5-HT was detectedamperometrically at a glassy carbon electrode at 500 mV vs Ag/AgCl (Antec Leyden, Leiden,The Netherlands). The detection limit was 0.5 fmol 5-HT per 20 µl sample (signal to noiseratio 3).


Four consecutive microdialysis samples with less then 20 % variation were taken as controland set at 100 %. Data are presented as areas under the curve (AUC) calculated as percentagetimes minutes over a period of 150 minutes after injection, or as percentages of control level(mean + S.E.M.) in time. Statistical analysis was performed using Sigmastat for windows(SPSS, Jandel Corporation). AUC data were compared using Mann Whitney Rank Sum test.Effects observed after chronic treatment were compared using two way analysis of variance(ANOVA) for repeated measurements, followed by Mann Whitney rank sum test. Level ofsignificance level was set at p<0.05.


117

3. Results

3.1 Medial Prefrontal CortexAcute: Basal levels of 5-HT in prefrontal cortex were 4.66 +0.30 fmol/sample (n=71). Acuteadministration of increasing doses of citalopram enhanced 5-HT levels in prefrontal cortexdose dependently. A plateau was reached at doses of citalopram exceeding 3 µmol/kg. Co-administration of citalopram with the 5-HT1A antagonist Way 100635 augmented the SSRIinduced increase in 5-HT levels. This enhancement occurred irrespective of the dose ofcitalopram administered (1 µmol/kg (P=0.0016) and 10 µmol/kg (P=0.0104) µmol/kg). Theputative 5-HT1B/D antagonist GR 127935 failed to enhance 5-HT levels when co-administeredwith citalopram (fig1).

Fig 1. Effect of citalopram (µmol/kg s.c.), alone or combined with 5-HT1A antagonist Way100635 1 µmol/kg s.c., or 5-HT1B/D antagonist GR 127935 1 µmol/kg s.c. on 5-HT levels inprefrontal cortex (data are area under the curve; % X min, * P < 0.05 ). � = citalopram andsaline (saline n =7, cit 0.1 µmol/kg n = 4, cit 0.3 µmol/kg n=3, cit 1 µmol/kg n= 5, cit 3µmol/kg n=6, cit 10 µmol/kg n=7), � = citalopram and Way 100635 (cit 0.1 µmol/kg andway n= 4, cit 1 µmol/kg and way n=8, cit 10 µmol/kg and way n=8), � = citalopram and GR127935 (cit 0.1 µmol/kg and GR n= 4, cit 1 µmol/kg and GR n=5, cit 10 µmol/kg and GRn=4)

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

**

1010.1

saline

mPFC

AU

C (

% X

min

)


Chapter 5

118

Fig 2. Effect of chronic treatment for 15 days with citalopram (�, b, n= 5) or saline (�, a,n=4) on effect of subcutaneous challenge of 10 µmol/kg citalopram on 5-HT levels inprefrontal cortex (inlay absolute outputs (fmol/sample), * P < 0.05).

Chronic: After chronic treatment for 14 days, basal levels of 5-HT were significantlyincreased (P = 0.049) for the citalopram group compared to saline treated animals, amountingto 6.46 + 1.28 fmol/sample (n=5) and 3.43 + 0.87 fmol/sample ( n=4), respectively. Achallenge with 10 µmol/kg s.c. of citalopram showed a similar relative increase of 5-HT levelsin saline and citalopram treated animals(F (1,95) = 1.08 P= 0.38)(fig2.)

3.2 Dorsal Raphe NucleusAcute: Basal levels of serotonin in dorsal raphe nucleus were 12.32 +2.05 fmol/sample (n=48).The dose effect curve of citalopram in the dorsal raphe was comparable with the one observedin the median prefrontal cortex, also levelling off at doses exceeding 3 µmol/kg s.c.. Whereasthe 5-HT1A antagonist Way 100635 enhanced 5-HT levels in prefrontal cortex at low- as wellas at high doses, augmentation of 5-HT levels in DRN was only observed at the highest doseof the re-uptake inhibitor (P=0.0025). Similar to median prefrontal cortex, co-administrationof the 5-HT1B/D receptor antagonist GR 127935 did not significantly augment the effect ofcitalopram in the dorsal raphe at any dose tested (fig 3.).

salin

e

cita

lop

ram

0

1

2

3

4

5

6

7

8

9

10

*

fmo

l/sa

mp

le

-100 -50 0 50 100 1500

200

400

600

ba

Citalopram 10 µmol/kg s.c.

14 days saline 14 days citalopram

% B

asa

l Le

vel

Time (min)


119

Fig 3. Effect of citalopram (µmol/kg s.c.), alone or combined with 5-HT1A antagonist Way100635 1 µmol/kg s.c., or 5-HT1B/D antagonist GR 127935 1 µmol/kg s.c. on 5-HT levels inDorsal Raphe Nucleus (data are area under the curve; % X min, * P < 0.05 ). � = citalopramand saline (saline n = 5, cit 0.1 µmol/kg n = 3, cit 0.3 µmol/kg n = 3, cit 1 µmol/kg n = 5, cit 3µmol/kg n = 3, cit 10 µmol/kg n = 7), � = citalopram and Way 100635 (cit 0.1 µmol/kg andway n = 5, cit 1 µmol/kg and way n = 6, cit 10 µmol/kg and way n = 5), � = citalopram andGR 127935 (cit 0.1 µmol/kg and GR n = 6, cit 1 µmol/kg and GR n = 6, cit 10 µmol/kg andGR n = 4)

Chronic: No significant differences in basal output levels in DRN were observedbetween animals chronically treated with saline (22.35 + 13.7 fmol/sample, n=4) andcitalopram (25.30 + 4.91 fmol/sample, n=6). A challenge with 10 µmol/kg s.c. of citalopramshowed a similar relative increase of 5-HT levels in saline and citalopram treated animals(F(1,96) = 0.63) (fig 4).

3.3 Dorsal HippocampusAcute: Basal output levels from dorsal hippocampal were 2.94 + 0.4 fmol/sample (n=44).Systemic administration of citalopram elevated 5-HT levels in a dose dependent fashion.Again, a plateau was attained with doses of citalopram exceeding 3 µmol/kg. Simultaneousadministration of citalopram and the 5-HT1A receptor antagonist yielded augmented levels of5-HT, but this depended on the dose of SSRI used. Whereas at a dose of 1µmol/kg ofcitalopram Way 100635 enhanced the increased levels of 5-HT (P = 0.0159) no augmentationwas observed when Way was combined with 10 µmol/kg of the SSRI. When citalopram wascombined with the 5-HT1B/D receptor antagonist GR 127935 augmentation was seen at anydose of citalopram (1 µmol/kg s.c., P = 0.0095; 10 µmol/kg s.c., P = 0.0286) (fig 5.).

0

20000

40000

60000

80000

100000

120000

140000

DRN

*

101.00.1saline

AU

C (

% X

min

)


Chapter 5

120

Fig 4. Effect of chronic treatment for 15 days with citalopram (�, b, n = 6) or saline (�, a, n= 4) on effect of subcutaneous challenge of 10 µmol/kg citalopram on 5-HT levels in DorsalRaphe Nucleus (inlay absolute outputs (fmol/sample), * P < 0.05).

Chronic: No significant differences in basal output levels in dorsal hippocampus wereobserved between animals chronically treated with saline (4.32 + 1.61 fmol/sample (n=4) andcitalopram (4.98 + 0.99 fmol/sample (n=5). A challenge with 10 µmol/kg s.c. of citalopramshowed a similar relative increase of 5-HT levels in saline and citalopram treated animals(F(1,73) = 1.43)(fig 6.).

3.3 Median Raphe NucleusAcute: Output levels of median raphe nucleus were 10.32 + 1.73 fmol/sample (n = 48). Inconsonance with the other brain areas tested citalopram increased extracelular serotonin in adose dependent fashion, with the effect levelling off at doses exceeding 3 µmol/kg. Thecombination of citalopram and 5-HT1A receptor antagonist Way 100635 showed a similarpicture as in the dorsal hippocampus. The increase of extracellular serotonin by 1 µmol/kgcitalopram was augmented by Way 100635 (P = 0.0286), but no augmentation was seen whenthe antagonist was combined with 10 µmol/kg citalopram. The combined administration ofcitalopram with 5-HT1B/D antagonist GR 127935 gave opposite results. Augmentation couldonly be observed at 10 µmol/kg of citalopram (P = 0.0303) (fig 7).

salin

e

cita

lopr

am

0

5

10

15

20

25

30

35

40

45

50

fmol

/sam

ple

-100 -50 0 50 100 1500

100

200

300

400

500

ba



% B

asa

l Lev

el

Time (min)


121

Fig 5. Effect of citalopram (µmol/kg s.c.), alone or combined with 5-HT1A antagonist Way100635 1 µmol/kg s.c., or 5-HT1B/D antagonist GR 127935 1 µmol/kg s.c. on 5-HT levels inDorsal Hippocampus (data are area under the curve; % X min, * P < 0.05 ). � = citalopramand saline (saline n = 3, cit 0.1 µmol/kg n = 4, cit 0.3 µmol/kg n = 3, cit 1 µmol/kg n = 4, cit10 µmol/kg n = 4), � = citalopram and Way 100635 (cit 0.1 µmol/kg and way n = 4, cit 1µmol/kg and way n = 5, cit 10 µmol/kg and way n = 4), � = citalopram and GR 127935 (cit0.1 µmol/kg and GR n = 6, cit 1 µmol/kg and GR n = 6, cit 10 µmol/kg and GR n = 4)

Chronic: Median raphe output levels were significantly enhanced when animals were treatedfor 15 days with citalopram (saline treated 14.28 + 4.11 fmol/sample (n=6), citalopram treated56.06 + 6.44 fmol/sample (n=5)). On the other hand, a challenge with 10 µmol/kg ofcitalopram had a marginal effect on extracellular serotonin levels, which was significantlydifferent from the effect measured in saline treated animals (F(1,113)=5.80) (fig 8.).

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

*

*

*

101.00.1saline

D Hippo

AU

C (

% X

min

)


Chapter 5

122

Fig 6. Effect of chronic treatment for 15 days with citalopram (�, b, n = 6) or saline (�, a, n= 3) on effect of subcutaneous challenge of 10 µmol/kg citalopram on 5-HT levels in DorsalHippocampus (inlay absolute outputs (fmol/sample), * P < 0.05).

salin

e

cita

lopr

am

0

1

2

3

4

5

6

7

8

9

10

fmo

l/sam

ple

-100 -50 0 50 100 1500

100

200

300

400

500

600

700

800

900

1000

ba



% B

asa

l Lev

el

Time (min)


123

4. Discussion and conclusion

Several studies indicate that regional differences of the effect of SSRIs on extracellularserotonin exist within the brain (Fuller 1994; Romero and Artigas 1997). An explanation maybe found in regional differences in autoreceptor control. To investigate this issue we haveinvestigated the acute and chronic effects of citalopram on extracellular serotonin in four areasin the rat brain. Since autoreceptor function may critically depend on the citalopram dose(Cremers et al. 2000) and may be regionally specific, we have included a 5-HT1A and 5-HT1B

receptor antagonist in the acute experiments.

4.1 Citalopram dose response study in four brain areasIn the present study, citalopram was found to induce a dose dependent increase in 5-HT levelsin all brain areas tested. The effect of citalopram on extracellular serotonin reached a plateauin all areas tested at doses exceeding 3 µmol/kg. In an earlier study we observed that theconcurrent plasma levels of citalopram at this dose were in the same range as those reportedfor patients treated with Cipramil, which might have its relevance in treatment (0.12-0.88µM, (Baumann 1996; Cremers et al. 2000)).

4.2a Autoreceptor control in prefrontal cortex and dorsal raphe nucleus.Restrainment of citalopram’s effect through the autoreceptors was very different from ourprevious observations in ventral hippocampus(Cremers et al. 2000). In median prefrontalcortex suppression of SSRI enhanced serotonin levels by 5-HT1A autoreceptors was observedat any dose of the SSRI, but no significant effect of the 5-HT1B/D receptor antagonist wasfound at any dose of the SSRI. This observation confirms earlier findings in this brain areafrom other investigators(Romero and Artigas 1997; Sharp et al. 1997).

As a vast majority of serotonergic innervation of the prefrontal cortex originates from thedorsal raphe nucleus, we thought it was of interest to compare their responses (McQuade andSharp 1997). Interestingly, the observations in the DRN did not completely match those madein the median prefrontal cortex. Although in prefrontal cortex, 5-HT1A receptor restrainingproperties were observed at any dose of the SSRI, this event was only observed in the DRN atthe highest dose of the SSRI. Apparently, additional regulatory processes in the cell body andterminal area play a role here.

Chapter 5

124

Fig 7. Effect of citalopram (µmol/kg s.c.), alone or combined with 5-HT1A antagonist Way100635 1 µmol/kg s.c., or 5-HT1B/D antagonist GR 127935 1 µmol/kg s.c. on 5-HT levels inMedian Raphe Nucleus (data are area under the curve; % X min, * P < 0.05 ). � = citalopramand saline (saline n = 6, cit 0.1 µmol/kg n = 4, cit 0.3 µmol/kg n = 4, cit 1 µmol/kg n = 4, cit 3µmol/kg n = 3, cit 10 µmol/kg n = 6), � = citalopram and Way 100635 (cit 0.1 µmol/kg andway n = 6, cit 1 µmol/kg and way n = 4, cit 10 µmol/kg and way n = 4), � = citalopram andGR 127935 (cit 0.1 µmol/kg and GR n = 4, cit 1 µmol/kg and GR n = 4, cit 10 µmol/kg andGR n = 5).

Fig 8. Effect of chronic treatment for 15 days with citalopram (�, b, n = 5) or saline (�, a, n= 6) on effect of subcutaneous challenge of 10 µmol/kg citalopram on 5-HT levels in MedianRaphe Nucleus (inlay absolute outputs (fmol/sample), * P < 0.05).

salin

e

cita

lopr

am

0

10

20

30

40

50

60

70

80

90

100

*

fmol

/sam

ple

-100 -50 0 50 100 1500

100

200

300

400

500

ba

citalopram 10 µmol/kg s.c.


% B

asa

l Lev

el

Time (min)

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

MRN

*

*

101.00.1saline

AU

C (

% X

min

)



125

Presuming that serotonin levels in the raphe nuclei reflect neuronal firing, activation ofsomatodendritic 5-HT1A receptors or postsynaptic 5-HT1A receptors attenuates the increaseinduced by the SSRI in raphe and cortex, thus explaining the observed effects. For theobserved difference between dorsal raphe and median prefrontal cortex at 1 µmol/kgcitalopram, several explanations can be found. A postsynaptic 5-HT1A receptor mediatedshort-loop type that acts locally in the PFC could explain the present finding. Second, ourassumption of an exclusive innervation of the cortex by the dorsal raphe may not be entirelycorrect.In the present study we could not demonstrate activation of 5-HT1B/D receptors in mPFC andDRN at any dose of citalopram. Several authors have investigated the relevance of 5-HT1B/D

receptors in relation to 5-HT levels and neuronal firing in DRN (Sprouse et al. 1997; Pineyroand Blier 1996; Pineyro et al. 1996; Pineyro et al. 1995; Starkey and Skingle 1994). Thoughsome indications were observed that 5-HT1B/D receptors might act as release regulatingautoreceptors in the DRN, the present data do not corroborate these findings.

4.2b Autoreceptor control in dorsal hippocampus and median raphe nucleus

Autoreceptor control was generally comparable between dorsal hippocampus and medianraphe nucleus. Extracellular serotonin in dorsal hippocampus as well as median raphe wasclearly under influence of 5-HT1A autoreceptors, which is in agreement with previous dualprobe microdialysis studies (Bosker et al. 1996; Bosker et al. 1994). However, 5-HT1A

receptor mediated restrainment was only observed at low doses of the SSRI (1 µmol/kg). Atthe highest dose of citalopram, augmentation was no longer detectable, which could mean thatthe 5-HT1B receptor had taken over control. Although the data for dorsal hippocampus andmedian raphe are broadly compatible with each other they do not match those from dorsalraphe nucleus and prefrontal cortex and previous data from ventral hippocampus (Cremers etal. 2000).The functionality and predominance of 5-HT1B autoreceptors in dorsal hippocampus has beenwell established (Bosker et al. 1995). This likely explains the augmentation of extracellularserotonin levels irrespective of the dose of citalopram (Bosker et al. 1995).Augmentation by the 5-HT1B receptor antagonist in the median raphe nucleus is not so easy toexplain, as it has not been thoroughly investigated. Possibly 5-HT1B and/or 5-HT1D receptorsare involved the control of somatodendritic release or alternatively in that from recurrentcolaterals as has been suggested for the DRN (Sprouse et al. 1997).Another explanation might also possible be relevant for the present observations. The effect of5-HT1A antagonist induced augmentation was not present at higher doses of citalopram (10µmol/kg). However, at these doses, augmentation could be observed during co-administrationwith 5-HT1B/D antagonists. Possibly, at these doses, a non-serotonergic feedback mechanism isactivated outsite the raphe through activation of 5-HT1B/D receptors. This would explain that5-HT1A antagonist induced augmentation is absent at higher doses of citalopram.

Chapter 5

126

4.3 Implications for the therapy of depression and anxiety disordersAn important finding of the present study is that the effects of citalopram on extracellularserotonin in the dorsal raphe-cortex and the median raphe-hippocampus are modulated in adifferent way by 5-HT1A and 5-HT1B/D autoreceptors. This may offer the opportunity toselectively modulate the effect of citalopram on serotonin levels by a 5-HT1A receptorantagonist for the dorsal raphe-cortex and by a 5-HT1B receptor antagonist for the medianraphe-hippocampus. It has been suggested that dorsal raphe-cortex is implicated in depression,while median raphe-hippocampus is more involved in anxiety disorders. Since SSRIs are usedfor both types of psychiatric disorders, combining them with a 5-HT1A receptor antagonistmay improve selectivity for depression, while the combination with a 5-HT1B receptorantagonist may be more beneficial for treatment of anxiety disorders.

4.4 Chronic treatmentIn the present study the osmotic minipumps were removed after two weeks of treatment.Microdialysis experiments were performed 48 hours after removal. As was observed in anearlier study, this time-period ensures that plasma levels of citalopram have dropped belowpharmacologically active levels, thus enabling the use of saline treatment as control (Cremerset al. 2000).

4.4a Basal Levels.Several changes in basal serotonin levels were observed using the present treatment protocol.In the prefrontal cortex, serotonin levels were significantly enhanced after chronic treatmentwith citalopram. This observation was also made by several other authors (Bel and Artigas.1993). These results might indicate that the level of auto-inhibition is decreased due to thechronic treatment with citalopram. As from the acute data it is apparent that no 5-HT1B

autoreceptor functionality is present in the prefrontal cortex, the current observations might bea reflection of 5-HT1A autoreceptor desensitisation. In addition, other features, likedesensitisation of feedback loops and the re-uptake site should not be discarded.In Dorsal Raphe, no significant changes in basal serotonin output levels were observed whencitalopram treatment was compared to saline treatment. However, when citalopram treatedanimals were compared to acute experiments, enhanced serotonin levels were observed due tothe treatment with citalopram. The basal levels of saline treated animals might show equallyenhanced levels as the citalopram treated animals, though the variation in the saline data wereof such extent that no conclusions can be drawn. Given the apparent connection between theDRN and mPFC, enhanced levels would indeed have been expected in DRN if 5-HT1A

desensitisation would have occurred. Indeed, several electrophysiological studies have shownthat ssri-induced inhibition of neuronal firing is restored after 2 weeks of chronic treatment.This effect was shown to be due to a desensitisation of 5-HT1A autoreceptors. The relevanceof the interplay with the apparent presence of the 5-HT1B receptor, however, should be furtherelucidated.

No effects of chronic treatment with citalopram was observed on basal levels of serotonin indorsal hippocampus. This observation would indicate that no apparent change in autoreceptor


127

function has occurred. In line with this observation, several studies have investigated the effectof chronic treatment with SSRIs on hippocampal serotonin levels, and none has ever observedany differences.Significant enhanced levels of serotonin were observed in MRN after chronic treatment withcitalopram. This observation could reflect a decreased functionality of 5-HT1A autoreceptors.Oddly, as described above, no concurrent enhanced level of serotonin was observed in dorsalhippocampus, as was observed for the mPFC-DRN innervation. Given the data from the acuteexperiments, it is interesting that the dorsal hippocampus in opposite to mPFC is under controlof 5-HT1B autoreceptors in addition to the somatodendritic 5-HT1A receptors. As nodesensitisation of these receptors has been profoundly observed, any desensitation of 5-HT1A

receptors might be effectively counteracted by the 5-HT1B receptor.

4.4b Citalopram challenge after chronic treatment.No changes in citalopram induced percentual increase of 5-HT levels was observed in mPFCor DRN. However, since the basal levels of serotonin in these areas were already elevated, thecurrent observation shows that the total serotonergic neurotransmission is elevated afterchronic treatment. Again, these events might be related to the desensitisation of 5-HT1A

autoreceptors.No changes in the percentual change of serotonin was observed in dorsal hippocampus whenthe animal was challenged with citalopram upon chronic treatment with the SSRI. Apparentlyno elevation of either form of serotonergic neurotransmission (basal or SSRI evoked) isobserved after chronic treatment. As the hippocampus is under control of 5-HT1B

autoreceptors in addition to 5-HT1A, the integrity of these receptors might remain intact,thereby effectively controlling serotonin neurotransmission. The concurrent absence of anyeffect of citalopram on serotonin levels in MRN, however, was somewhat suprising, but mightbe related to its already elevated basal levels. Possible explenations of this event mightoriginate in a desensitised re-uptake site or 5-HT1A receptor. Absence of any reflection of thisobservation in it projection area, the dorsal hippocampus, again hints towards an effective 5-HT1B receptor control in terminal areas.

Chapter 5

128

ConclusionsAcute-5-HT1A as well as 5-HT1B Autoreceptor control is related to the dose of SSRI and the brainarea under investigation.-mPFC and is innervating cell body area, DRN, is firmly under control of 5-HT1Aautoreceptors. Whereas the auto-inhibition of DRN serotonin levels is also under influence of5-HT1B receptors at higher doses of the SSRI, no relevance of these autoreceptors is presentin mPFC.-The dorsal hippocampus and its innervating nucleus (MRN) are both under influence of 5-HT1A autoreceptors at low doses of the SSRI, whereas, this effect is completelyovershadowed by 5-HT1B receptors at higher doses of the SSRI in both areas. Only the dorsalhippocampus is also restricted by 5-HT1B receptors at low doses of the citalopram.

Chronic-While chronic treatment with SSRIs apparently induces enhanced serotonergicneurotranmission in the mPFC-DRN pathway, this effect was not observed in the d. hippo-MRN connection.


129

References

Arborelius, L., Hook, B.B., Hacksell, U. and Svensson, T.H. (1994). The 5-HT1A receptorantagonist (S)-UH-301 blocks the qR)-8-OH-DPAT-induced inhibition of serotonergic dorsalraphe cell firing in the rat. J Neural Transm Gen Sect, 96, 179-186.Arborelius, L., Nomikos, G.G., Grillner, P., Hertel, P., Hook, B.B., Hacksell, U. andSvensson, T.H. (1995). 5-HT1A receptor antagonists increase the activity of serotonergic cellsin the dorsal raphe nucleus in rats treated acutely or chronically with citalopram. NaunynSchmiedebergs Arch Pharmacol, 352, 157-165.Artigas, F. (1993). 5-HT and antidepressants: new views from microdialysis studies. TrendsPharmacol Sci, 14, 262Auerbach, S.B. and Hjorth, S. (1995). Effect of chronic administration of the selectiveserotonin (5-HT) uptake inhibitor citalopram on extracellular 5-HT and apparent autoreceptorsensitivity in rat forebrain in vivo. Naunyn Schmiedebergs Arch Pharmacol, 352, 597-606.Baumann, P. (1996). Pharmacokinetic-pharmacodynamic relationship of the selectiveserotonin reuptake inhibitors. Clin Pharmacokinet, 31, 444-469.Bel, N. and Artigas, F. (1993). Chronic treatment with fluvoxamine increases extracellularserotonin in frontal cortex but not in raphe nuclei. Synapse, 15, 243-245.Blier, P., Chaput, Y. and de Montigny, C. (1988). Long-term 5-HT reuptake blockade, butnot monoamine oxidase inhibition, decreases the function of terminal 5-HT autoreceptors: anelectrophysiological study in the rat brain. Naunyn Schmiedebergs Arch Pharmacol, 337, 246-254.Blier, P., de Montigny, C. and Chaput, Y. (1987). Modifications of the serotonin system byantidepressant treatments: implications for the therapeutic response in major depression. J ClinPsychopharmacol, 7, 24S-35S.Bosker, F., Klompmakers, A.N. and Westenberg, H. (1994). Extracellular 5-hydroxytryptamine in median raphe nucleus of the conscious rat is decreased by nanomolarconcentrations of 8-hydroxy-2-(di-n-propylamino) tetralin and is sensitive to tetrodotoxin. JNeurochem, 63, 2165-2171.Bosker, F., Vrinten, D., Klompmakers, A. and Westenberg, H. (1997). The effects of a 5-HT1A receptor agonist and antagonist on the 5-hydroxytryptamine release in the centralnucleus of the amygdala: a microdialysis study with flesinoxan and WAY 100635. NaunynSchmiedebergs Arch Pharmacol, 355, 347-353.Bosker, F.J., de Winter, T.Y., Klompmakers, A.A. and Westenberg, H.G. (1996). Flesinoxandose-dependently reduces extracellular 5-hydroxytryptamine (5-HT) in rat median raphe anddorsal hippocampus through activation of 5-HT1A receptors. J Neurochem, 66, 2546-2555.Bosker, F.J., Klompmakers, A.A. and Westenberg, H.G. (1995). Effects of single and repeatedoral administration of fluvoxamine on extracellular serotonin in the median raphe nucleus anddorsal hippocampus of the rat. Neuropharmacology, 34, 501-508.Bosker, F.J., van Esseveldt, K.E., Klompmakers, A.A. and Westenberg, H.G. (1995). Chronictreatment with fluvoxamine by osmotic minipumps fails to induce persistent functional changesin central 5-HT1A and 5-HT1B receptors, as measured by in vivo microdialysis in dorsalhippocampus of conscious rats. Psychopharmacology Berl, 117, 358-363.

Chapter 5

130

Casanovas, J.M., Hervas, I. and Artigas, F. (1999). Post-synaptic 5-HT1A receptors control5-HT release in the rat medial prefrontal cortex. Neuroreport, 10, 1441-1445.Ceci, A., Baschirotto, A. and Borsini, F. (1994). The inhibitory effect of 8-OH-DPAT on thefiring activity of dorsal raphe serotonergic neurons in rats is attenuated by lesion of theprefrontal cortex. Neuropharmacology, 33, 709-713.Cremers, T.I.F.H., de Boer, P., Liao, Y., Bosker, F.J., den Boer, J.A., Westerink, B.H. andWikstrom, H.V. (2000). Augmentation with a 5-HT1A, but not a 5-HT1B receptor antagonistcritically depends on the dose of citalopram. Eur J Pharmacol, 397, 63-74.Cremers, T.I.F.H., Spoelstra, E.N., de Boer, P., Bosker, F.J., Mork, A., den Boer, J.A.,Westerink, B.H., & Wikstrom, H.V. (2000). Desensitisation of 5-HT autoreceptors uponpharmacokinetically monitored chronic treatment with citalopram. Eur J Pharmacol, 397, 351-357.Fuller, R.W. (1994). Uptake inhibitors increase extracellular serotonin concentration measuredby brain microdialysis. Life Sci, 55, 163-167.Gobbi, M., Crespi, D., Foddi, M.C., Fracasso, C., Mancini, L., Parotti, L. and Mennini, T.(1997). Effects of chronic treatment with fluoxetine and citalopram on 5-HT uptake, 5-HT1Bautoreceptors, 5-HT3 and 5-HT4 receptors in rats. Naunyn Schmiedebergs Arch Pharmacol,356, 22-28.Hjorth, S. (1993). Serotonin 5-HT1A autoreceptor blockade potentiates the ability of the 5-HT reuptake inhibitor citalopram to increase nerve terminal output of 5-HT in vivo: amicrodialysis study. J. Neurochem. 60(2); 776-779.Invernizzi, R., Bramante, M. and Samanin, R. (1994). Chronic treatment with citalopramfacilitates the effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT1A receptors. Eur J Pharmacol, 260, 243-246.Kreiss, D.S. and Lucki, I. (1997). Chronic administration of the 5-HT1a receptor agonist 8-HODPAT differentially desensitizes 5-HT1a autoreceptors of the dorsal and median raphenuclei. Synapse, 25, 107-116.McQuade, R. and Sharp, T. (1997). Functional mapping of dorsal and median raphe 5-hydroxytryptamine pathways in forebrain of the rat using microdialysis. J Neurochem, 69,791-796.Nelson, J.C. (2000). Augmentation strategies in depression 2000. J Clin Psychiatry, 61, 13-19.Pineyro, G. and Blier, P. (1996). Regulation of 5-hydroxytryptamine release from ratmidbrain raphe nuclei by 5-hydroxytryptamine1D receptors: effect of tetrodotoxin, G proteininactivation and long-term antidepressant administration. J Pharmacol Exp Ther, 276, 697-707.Pineyro, G., de Montigny, C. and Blier, P. (1995). 5-HT1D receptors regulate 5-HT releasein the rat raphe nuclei. In vivo voltammetry and in vitro superfusion studies.Neuropsychopharmacology, 13, 249-260.Pineyro, G., de Montigny, C., Weiss, M. and Blier, P. (1996). Autoregulatory properties ofdorsal raphe 5-HT neurons: possible role of electrotonic coupling and 5-HT1D receptors inthe rat brain. Synapse, 22, 54-62.


131

Romero, L. and Artigas, F. (1997). Preferential potentiation of the effects of serotonin uptakeinhibitors by 5-HT1A receptor antagonists in the dorsal raphe pathway: role ofsomatodendritic autoreceptors. J Neurochem, 68, 2593-2603.Sharp, T., Umbers, V. and Gartside, S.E. (1997). Effect of a selective 5-HT reuptake inhibitorin combination with 5-HT1A and 5-HT1B receptor antagonists on extracellular 5-HT in ratfrontal cortex in vivo. Br J Pharmacol, 121, 941-946.Sprouse, J., Reynolds, L. and Rollema, H. (1997). Do 5-HT1B/1D autoreceptors modulatedorsal raphe cell firing? In vivo electrophysiological studies in guinea pigs with GR127935.Neuropharmacology, 36, 559-567.Starkey, S.J. and Skingle, M. (1994). 5-HT1D as well as 5-HT1A autoreceptors modulate 5-HT release in the guinea-pig dorsal raphe nucleus. Neuropharmacology, 33, 393-402.

Chapter 5

132

Acute and Chronic Effects of Citalopram on Postsynaptic 5-Hydroxytryptamine1A (5-HT1A) Receptor MediatedFeedback: A Microdialysis Study in the Amygdala

133

Chapter 6

Acute and Chronic Effects of Citalopram on

Postsynaptic 5-Hydroxytryptamine1A (5-HT1A)

Receptor Mediated Feedback: A Microdialysis

Study in the Amygdala

Chapter 6

134

Abstract

Microdialysis was used to assess the involvement of postsynaptic 5-HT1A receptors in theregulation of extracellular 5-HT in the amygdala. Local infusion of the 5-HT1A receptoragonist flesinoxan (0.3, 1, 3 µM) for 30 min into the amygdala maximally decreased 5-HT to50% of basal level. Systemic administration of citalopram (10 µmol/kg) increased 5-HT to175% of basal level. Local infusion of 1 µM of the 5-HT1A receptor antagonist WAY 100.635into the amygdala augmented the effect of citalopram to more than 500% of basal 5-HT level.5-HT1A receptor responsiveness after chronic citalopram treatment was determined in twoways. First, by local infusion of 1 µM flesinoxan for 30 min into the amygdala, which showeda significant 63 % reduction in response (AUC) for the citalopram group compared to thesaline group. Second, by systemic administration of citalopram (10 µmol/kg), which increased5-HT to 350 % of basal level. The effect was larger than in untreated animals, but moreimportant, local infusion of 1 µM WAY 100.635 into the amygdala now failed to augment theeffect of citalopram. Both the flesinoxan and WAY 100.635 data suggest an involvement ofpostsynaptic 5-HT1A receptor mediated feedback in the amygdala, which diminishes followingchronic citalopram treatment.


135

1. Introduction

Selective serotonin reuptake inhibitors (SSRIs) have proven effective in the treatment of majordepression (Gravem et al., 1987; Montgomery et al., 1992) and various anxiety disorders suchas panic disorder (Den Boer and Westenberg, 1988), obsessive-compulsive disorder(Montgomery et al., 1993) and social phobia (Van Vliet et al., 1994). Characteristic for SSRItreatment is a 3-6 week delay in therapeutic response. This delay, which may partly depend onthe nature and severity of the illness, suggests adaptive changes taking place, possibly at thelevel of serotonin receptors.SSRIs increase serotonin (5-HT) in the vicinity of serotonergic neurons through blockade ofthe 5-HT carrier (for review see Fuller, 1994; Hyttel, 1994). Arguably 5-HT1A autoreceptorson 5-HT cellbodies and 5-HT1B receptors on axon terminals become increasingly activated.Stimulation of 5-HT1A receptors reduces the firing rate of 5-HT neurons (Sprouse andAghajanian, 1987), release of 5-HT (Bosker et al., 1994, 1996) and 5-HT synthesis (Hutson etal., 1989), while stimulation of 5-HT1B presynaptic receptors decreases 5-HT release (Sharpet al., 1989; Limberger et al., 1991). These secondary phenomena are believed to limit theeffect of reuptake inhibition on the 5-HT release in the terminal regions.The delayed response to antidepressant treatment may be connected to a gradual loss ofresponsiveness of 5-HT autoreceptors following chronic treatment (Blier et al., 1987a).Theoretically, the latter condition can be mimicked instantaneously by using a combination ofSSRI and 5-HT autoreceptor antagonist. Intracerebral microdialysis studies have reported anaugmented increase in extracellular 5-HT using this approach (e.g. Invernizzi et al., 1992;Hjorth, 1993; Gobert et al., 1997; Sharp et al., 1997). The clinical relevance of this concept(Artigas, 1993, Hjorth 1993) has been investigated with the 5-HT1A and β- adrenergicreceptor antagonist pindolol (e.g. Artigas et al., 1994; Blier and Bergeron, 1995; Berman etal., 1997).Feedback control of firing rate and 5-HT release through somatodendritic 5-HT1A andpresynaptic 5-HT1B/D autoreceptors is now well established. In addition to the autoreceptors,postsynaptic 5-HT1A receptors may also play a role in the regulation of firing rate and releaseof serotonergic neurons. The existence of a neuronal feedback loop between projection areasand raphe nuclei had already been hypothesized by Blier and De Montigny (1987b,c). The ideais supported by several studies (e.g. Ceci et al.,1994; Romero et al.,1994; Bosker et al.,1997b; Casanovas et al., 1999). In the dual probe microdialysis study by Bosker et al. (1997b)flesinoxan, a selective 5-HT1A receptor agonist, was locally infused into the central nucleus ofthe amygdala (CeA). This decreased extracellular 5-HT both locally and in the caudal linearraphe nucleus. Simultaneous infusion of the 5-HT1A receptor antagonist WAY 100.635 intothe CeA, but not into the caudal linear raphe nucleus, blocked the effects of flesinoxan.The effect of chronic treatment on receptor function is usually assessed through systemicadministration of pharmacological probes. This makes it difficult to discriminate between preand postsynaptic receptor mediated feedback. The present microdialysis study has investigatedthe involvement of the latter type of feedback in the acute and chronic effects of the SSRIcitalopram on extracellular 5-HT in the CeA.

Chapter 6

136

The function of the postsynaptic 5-HT1A receptors was assessed by local infusions of the 5-HT1A receptor agonist flesinoxan (Bosch et al., 1994; Bosker et al., 1996, 1997a,b) and the 5-HT1A receptor antagonist WAY 100.635 (Fletcher et al., 1994; Munday et al., 1994).


137


2.1 Animals

Male Wistar rats (Harlan, Zeist, The Netherlands) weighing 285-320 g were housed three percage under standard conditions (22-24 °C, 12/12 h light/dark cycle, food and water ad libitum)for at least 3 days until insertion of the microdialysis probes. After surgery the animals werehoused separately. All animal experiments were in accordance with the declarations ofHelsinki and were approved by the Animal Care Committee of the Faculty of Mathematics andNatural Science of the University of Groningen.

2.2 Surgery

Acute experiments: Rats were anesthetized with chloral hydrate (400 mg/kg, i.p.). Chloralhydrate was used to enable comparisons with a previous study (Bosker et al., 1997b). Homemade I-shaped probes (i.d. 220 µm, o.d. 310 µm, AN 69, Hospal, Italy), made ofpolyacrylonitrile / sodium methyl sulphonate copolymer dialysis fiber (Santiago and Westerink,1990) were implanted bilaterally in the vicinity of the central nucleus of the amygdala. Theexposed length of the membranes was 1 mm. Lidocaine-HCl, 10 % (m/v) was used for localanaesthesia. Rats were placed in a stereotaxic frame (Kopf, USA), and probes were inserted atthe following coordinates: incisorbar at −3.5 mm, A: −2.6 mm, L: +/− 4.2 mm, V: 9.1 mmfrom bregma and skull surface (Paxinos and Watson, 1982) and secured with dental cement.The animals were allowed to recover for 2 days.Chronic experiments: Osmotic minipumps (2ML2 Alzet, USA, 5 µl/h, 2 weeks) were filledwith a solution of 50 mg/ml citalopram hydrobromide under aseptic conditions. During briefisoflurane anesthesia (2.5 %, 400ml/min N2O, 600 ml/min O2), minipumps were implantedsubcutaneously on the left side of the back of the rat. In four animals a cannula was insertedinto the jugular vein to monitor citalopram levels during and after the 14 days treatmentperiod. After 14 days, rats were anesthetized with chloral hydrate (400 mg/kg, i.p.), and theosmotic minipumps were removed. The exposed subcutaneous cavity was flushed twice with 5ml of sterile saline. Hereafter microdialysis probes were implanted bilaterally in the vicinity ofthe CeA (see acute experiments).


Blood samples (0.35 ml) were drawn on day 3, 7, 12, 15, 16 and 17 after implantation of theosmotic minipumps. Samples were transferred to 1.5 ml Eppendorf vials, containing 5 µlheparin (500 IE/ml saline), mixed and immediately centrifuged for 15 min at 4,000 g(Chillspin, MSE, England).

Chapter 6

138


Microdialysis experiments were performed on both the second and third day after surgery tolimit the number of animals. Administration of drugs was randomly allocated to the secondand third day using a balanced design. Experiments with and without reuptake inhibitor in theperfusion fluid were randomly performed in the left and right amygdala. None of the animalsreceived the same treatment twice. The probes were perfused with Ringer (147 mM NaCl, 3.0mM KCl, 1.2 mM CaCl2, and 1.2 mM MgCl2, pH 6-7) using a Harvard micro infusion pump(Harvard, South Natick, Ma., USA) at a constant flow rate of 1.5 µl/min. Flesinoxan wasinfused for 30 min. WAY 100.635 and citalopram were infused for 210 min. Samples werecollected in mini vials containing 7.5 µl of 0.02 M acetic acid, mixed and placed in a chilledautomatic injection apparatus (Carnegie, Sweden). For systemic administration (s.c.)citalopram was dissolved in saline. For local administration citalopram, flesinoxan and WAY100.635 were dissolved in the perfusion fluid and infused via retrograde microdialysis into theCeA. Doses were based on earlier studies on related experiments (Bosker et al., 1997b;Cremers et al., 2000a,b). All experiments were performed in conscious and freely movinganimals.

2.5 Analytical procedure

Citalopram: Citalopram was measured according to Øyehaug et al. (1982), with minormodifications (Cremers et al., 2000b). The detection limit for citalopram was 5 pmol/mlplasma (signal / noise ratio = 2).Serotonin: Analysis of 5-HT was performed by HPLC with electrochemical detection. Briefly,20 µl samples were injected into a Shimadzu LC-10 AD high performance liquidchromatograph equipped with a 10 cm reversed phase column (phenomenex hypersil 3 : 3 µm,C18, 100 x 2.0 mm, Bester, Amstelveen, the Netherlands) and an electrochemical detector(Antec Leyden B.V., Leiden, the Netherlands), at a potential setting of 500 mV vs. Ag/AgClreference electrode. Chromatography was performed at 30 °C using the integrated columnoven of the Antec potentiostate. The mobile phase consisted of 5 g/l di-ammoniumsulfate, 500mg/l ethylene diamino tetra acetic acid (EDTA), 50 mg/l heptane sulphonic acid, 30 µl/ltriethylamine, 4% methanol, adjusted to pH 4.65. The flow rate was 0.4 ml/min. The detectionlimit for 5-HT was 0.2 fmol/20 µl sample (signal / noise ratio =2).

2.6 Histology

Following the termination of each experiment, animals were anesthetized with chloral hydrate,the probes were flushed successively with a 5% aqueous FeCl3 solution, H2O and a 5%aqueous K4Fe(CN)6 solution for 15 minutes (Kendrick, 1990). The iron deposited in the tissuesurrounding the probe is stained green-blue. The diffusion area is normally no more than 1 mmfrom the membrane. The animals were killed by decapitation, the brains removed and fixed in a5% formaldehyde solution. After at least 2 days the brains were cut in 150 µm slices using avibratome. The position of the probe was verified microscopically by the track of the probe


139

through the brain and the green-blue staining of the tissue surrounding the probe tip. Datafrom few animals with improper probe placement were excluded (~ 5%). In this respect, adistance > 150 µm from the CeA was used as exclusion criterion. Due to the occasionalobstruction of a microdialysis probe or analytical failure the overall success rates were lowerand amounted to approximately 80 %.

2.7 Statistics

The data are presented as percentage of basal values calculated as individual means of the firstfour consecutive microdialysis samples. Statistical analysis was performed using Sigmastat forwindows (Jandel Corporation). Treatment effects were evaluated using one way ANOVA forrepeated measurements, followed by Dunnet’s test or two way ANOVA for repeatedmeasurements, followed by Mann-Whitney rank sum test. The level of significance level wasset at p<0.05.

2.8 Materials

The following drugs were used: Citalopram hydrobromide, metabolites and internal standard(kindly donated by Lundbeck, Denmark), flesinoxan (kindly donated by SolvayPharmaceuticals, the Netherlands) and WAY 100.635 (synthesized at our medicinal chemistrylaboratory).

Chapter 6

140

3. Results

3.1 Basal conditionsWithout citalopram in the perfusion fluid the basal levels of extracellular 5-HT in the CeAwere 2.6 + 0.3 fmol/15 min (SEM, n=37). As expected with 1 µM of citalopram in theperfusion fluid 5-HT levels were higher, amounting to 12.1 + 1.8 fmol/15 min (SEM, n=22).This increase in 5-HT levels (~ 470%) is concordant with the increase in 5-HT observed in adedicated experiment, where 1 µM of citalopram maximally increased 5-HT levels to 480% ofbasal level (Fig. 1). Using one way ANOVA the effect was significant (χ2

10 = 29.3; p=0.0056).Infusion of flesinoxan by retrograde microdialysis into the CeA for 30 min decreased 5-HTlevels significantly (Table 1). At the highest dose of flesinoxan extracellular 5-HT maximallydecreased to 50% of basal level (Fig. 2). The same experiment with 1 µM of citalopram in theperfusion fluid also showed significant decreases of 5-HT levels, but the effects were lessmarked (Table 1). Maximal decrease was to 69% of basal level (Fig. 3).

Figure 1. Effect of local infusion of 1 µM of citalopram on extracellular 5-HT in the centralnucleus of the amygdala. Data are given as percentage of controls + SEM and are the averageof four experiments. Horizontal bar indicates the period of citalopram infusion. * indicates p<0.05.

3.2 Acute effect of citalopramSystemic administration of 10 µmol/kg of citalopram maximally increased extracellular 5-HTin the amygdala to 175% of basal level (Table 2)(Fig. 4). Blockade of 5-HT1A receptors bylocal infusion of 1 µM of WAY 100.635 augmented the systemic effect of citalopram to >500% of controls (Fig. 4). The latter increase was significantly different from the effect in

-100 -50 0 50 100 150 200 2500

100

200

300

400

500

600

700

*

** * *

* **

*

Ext

race

llula

r 5

-HT

% o

f con

tro

l

min


141

absence of WAY 100.635 (Table 2). The antagonist was infused from t=0 and had no effecton basal 5-HT levels.

Figure 2A. Effect of local infusion of flesinoxan on extracellular 5-HT in the central nucleusof the amygdala. � 300 nM; � 1000 nM; � 3000 nM.Data are given as percentage of controls + SEM and are the average of 7, 6 and 8experiments, respectively. Horizontal bar indicates the period of flesinoxan infusion. Asterisks

not shown. See section results for the statistical evaluation.Figure 2B. Effect of local infusion of flesinoxan. Effects on extracellular 5-HT arerepresented as the area under curve (AUC). * indicates p< 0.05

Chapter 6

142

Figure 3A. Effect of local infusion of flesinoxan on extracellular 5-HT in the central nucleusof the amygdala in the presence of 1 µM of citalopram. � 300 nM; � 1000 nM; � 3000 nM.Data are given as percentage of controls + SEM and are the average of 4, 7 and 4experiments, respectively. Horizontal bar indicates the period of flesinoxan infusion. Asterisksnot shown. See section results for the statistical evaluation. Figure 3B. Effect of local infusionof flesinoxan in the presence of 1 µM of citalopram. Effects on extracellular 5-HT arerepresented as the area under curve (AUC). * indicates p< 0.05.

3.3 Chronic effect of citalopram Citalopram vs. saline was chronically administered via osmotic minipumps for 14 days. Basal5-HT levels were 2.5 + 0.3 and 2.6 + 0.3 fmol/15 min for the saline (n=6) and citalopram(n=6) group, respectively. Assessment of postsynaptic 5-HT1A receptor response in the CeAwas carried out in two ways.First, by local infusion of 1 µM of flesinoxan for 30 min, which showed a 63 % reduction ineffect for the citalopram group compared to the saline group (Table 1)(Fig. 5a).Second, by systemic injection of citalopram (10 µmol/kg) with and without 1 µM of WAY100.635 locally perfused into the CeA. The latter experiments were performed in a differentgroup of animals, which had been chronically treated with citalopram. In the presence andabsence of WAY 100.635, citalopram increased 5-HT in the CeA to approximately 350 % ofbasal value (Table 2)(Fig. 6).With citalopram in the perfusion fluid, basal 5-HT levels for saline (n=7) and citalopram (n=7)treated animals were 11.4 + 1.9 and 9.4 + 1.4 fmol/15 min, respectively. Perfusion withflesinoxan in the presence of 1 µM of citalopram in the perfusion fluid failed to showsignificant differences between the groups (Table 1)(Fig. 5b).


143

Figure 4. Effect of local infusion of 1µM of WAY 100.635 on the systemic effect of 10µmol/kg s.c. of citalopram on extracellular 5-HT in the central nucleus of the amygdala. �

without WAY 100.635; � with WAY 100.635. Data are given as percentage of controls +SEM and are the average of 7 experiments. Horizontal bar indicates period of WAY 100.635infusion. The arrow indicates the systemic administration of citalopram. * indicates p< 0.05.

3.4 Citalopram plasma levels Chronic treatment of rats with osmotic minipumps (50 mg/ml citalopram), resulted in stablecitalopram plasma levels of 0.30 + 0.03 µM (n=4). Desmethyl- and didesmethyl metabolitescould also be demonstrated during chronic treatment (0.06 + 0.01 µM and 0.18 + 0.01 µM,respectively). Removal of the minipumps gradually declined plasma levels of citalopram and itsmetabolites. After 48 hours, plasma levels of citalopram were 8.2 + 0.3 nM (n=4). Desmethyland didesmethyl metabolites were 0.48 + 0.30 nM (n=4) and 14.4 + 1.0 nM (n=4),respectively. The very low concentrations of citalopram and metabolites were estimated fromblood taken by cardiac puncture, which was concentrated 33 times in the extraction procedure(Cremers 2000b).

-100 -50 0 50 100 150 200 2500

100

200

300

400

500

600

700

800

enter text here

* * **

*

Ext

race

llula

r 5-

HT

% o

f con

tro

l

min

Chapter 6

144

Figure 5A. Chronic treatment with citalopram vs. saline. Effect of local infusion of 1000 nMof flesinoxan into the central nucleus of the amygdala. � chronic saline treatment; � chroniccitalopram treatment. Data are given as percentage of controls + SEM and are the average ofsix experiments. Horizontal bar indicates the period of flesinoxan infusion. * indicates p< 0.05.Figure 5B. Chronic treatment with citalopram vs. saline. Effect of local infusion of 1000 nMof flesinoxan into the central nucleus of the amygdala in the presence of 1µM of citalopram inthe perfusion fluid. � chronic saline treatment; � chronic citalopram treatment. Data are givenas percentage of controls + SEM and are the average of six experiments. Horizontal barindicates the period of flesinoxan infusion. * indicates p< 0.05.


145

4. Discussion

4.1 Basal conditions

Basal 5-HT levels with 1 µM of citalopram in the perfusion fluid were lower than thosepreviously reported (Bosker et al., 1997b) in the presence of 10 µM of fluvoxamine (12.1 vs.20.3 fmol/15 min), which may partly relate to the sub-optimal citalopram concentration. Basal5-HT levels were much lower without re-uptake inhibitor in the perfusion fluid. However, therelative decrease of extracellular 5-HT by locally infused flesinoxan was now morepronounced. One could speculate that the increased 5-HT levels induced by re-uptakeinhibition have caused a shift to the upper less steep part of the flesinoxan dose responsecurve.The primary reason for the flesinoxan concentration-response study was to find the optimalcondition as pharmacological probe in the chronic part of the study.

4.2 Acute treatment with citalopram

Systemic administration of 10 µmol/kg of citalopram increased extracellular 5-HT in the CeAto 175% of basal level. This increase is modest compared to the ventral hippocampus, wherethe same dose of citalopram increased 5-HT to 325% of basal level (Cremers et al., 2000a).Previous experiments in the ventral hippocampus have shown that a dose of 10 µmol/kg ofcitalopram maximally increased extracellular 5-HT (Cremers et al., 2000a). Assuming a similardose-effect relation for reuptake inhibition in ventral hippocampus and CeA, one couldspeculate that extracellular 5-HT is more tightly controlled in the latter area. It is alsonoteworthy that the effect of systemically administered flesinoxan on extracellular 5-HT wasmore pronounced in CeA (Bosker et al., 1997a) compared to dorsal hippocampus (Bosker etal., 1996).Several studies have reported different effect sizes in terminal areas by systemicallyadministered SSRIs (see Fuller, 1994). This is generally attributed to a divergent involvementof release controlling somatodendritic 5-HT1A autoreceptors in the raphe nuclei (Kreiss andLucki, 1994). The more than 400% increase of extracellular 5-HT by local infusion of 1 µM ofcitalopram compared to the 175% following systemic administration suggests a stronginvolvement of somatodendritic 5-HT1A receptor mediated feedback. Therefore, the markedaugmentation of citalopram’s effect by local infusion of the 5-HT1A receptor antagonist WAY100.635 into the CeA came rather unexpected. Apparently, postsynaptic 5-HT1A receptormediated feedback is also strongly activated by citalopram. It is feasible that, next to adifferential involvement of raphe nuclei, differences in postsynaptic 5-HT1A receptor mediatedfeedback between terminal areas also play a role in the varying effect sizes of SSRIs.Interestingly, Sharp et al. (1989) did not observe an effect on extracellular 5-HT followinglocal application of 10 µM of 8-OH-DPAT in ventral hippocampus. In addition, Hjorth et al.(1996) could not demonstrate restraint of the citalopram induced increase in ventralhippocampal extracellular 5-HT by postsynaptic 5-HT1A receptors.

Chapter 6

146

Generalizations regarding postsynaptic 5-HT1A receptor mediated feedback cannot be madeeasily. Clearly more studies in different terminal areas are needed. However, taking intoaccount the earlier findings in cortex (Ceci et al., 1994; Casanovas et al., 1999) and striatum(Romero et al., 1994) it is rather unlikely that this type of feedback is restricted to a few areasin the brain.

Figure 6. Chronic treatment with citalopram. Effect of local infusion of 1µM of WAY100.635 on the systemic effect of 10 µmol/kg s.c. of citalopram on extracellular 5-HT in thecentral nucleus of the amygdala. � without WAY 100.635; � with WAY 100.635. Horizontalbar indicates period of WAY 100.635 infusion. The arrow indicates the systemicadministration of citalopram. * indicates p< 0.05.

4.3 Chronic treatment with citalopram

The mechanism underlying the therapeutic effects of SSRIs is far from clear. Several studiessuggest that desensitization of 5-HT1A receptors may at least partly be involved. The presentstudy supports and extends this idea for the postsynaptic 5-HT1A receptors in the amygdala.Local infusion of 1 µM of the 5-HT1A receptor agonist flesinoxan into the CeA induced asignificantly less marked decrease of extracellular 5-HT in the chronic citalopram groupcompared to the chronic saline group, which suggests desensitization of postsynaptic 5-HT1A

receptors in the amygdala. The decrease induced by flesinoxan in the chronic saline group wassimilar to the effect size in untreated animals, indicating that the osmotic minipump itself hadlittle effect on the outcome. Using the AUC-method postsynaptic 5-HT1A receptorresponsiveness at a concentration of 1000 nM of flesinoxan was estimated to be 80% for thesaline group and 35% for the citalopram treated animals.Similar experiments in the presence of 1 µM of citalopram showed no significant differencesbetween the chronic citalopram group and the chronic saline group. In view of the smallereffect size for flesinoxan under this condition this outcome may not be surprising. Of more

-100 -50 0 50 100 150 200 2500

50

100

150

200

250

300

350

400

450

500

550

600

Ext

race

llula

r 5

-HT

% o

f co

ntro

l

min


147

concern are the similar basal 5-HT levels in the presence of citalopram. In view of thedecreased responsiveness of postsynaptic 5-HT1A receptors following chronic treatment withcitalopram we would at least have expected higher basal 5-HT levels for the chroniccitalopram group. A somewhat paradoxical argument would be that the lack of effect on basal5-HT levels agrees with the negative flesinoxan data obtained when citalopram was present inthe perfusion fluid (Fig. 6). Anyhow, local reuptake inhibition compromises the assessment ofpostsynaptic 5-HT1A receptor sensitivity, and should therefore be avoided in future studies.A systemic challenge with citalopram had a more profound effect on 5-HT levels in animalschronically treated with citalopram than in drug-naive animals. On the other hand we did notobserve an increase in basal levels following the washout period. Both observations areconsistent with the notion that patients improved by their SSRI do so while on the drug.Furthermore, when withdrawn prematurely from their SSRI regimen, patients relapse rapidlywhich is also consistent with the present data.Simultaneous infusion of WAY 100.635 into the CeA now failed to augment the effect ofcitalopram on extracellular 5-HT. This observation indicates that postsynaptic 5-HT1A mayhave become less responsive following chronic treatment with the SSRI. We would haveexpected a comparable increase of 5-HT by citalopram in the presence of WAY 100.635 incitalopram treated (Fig. 6) and drug-naive animals (Fig. 4), which appeared not to be the case.On retrospection, if desensitization of the postsynaptic 5-HT1A receptors involves a decreasein affinity (increase in Kd) this may also influence the binding of WAY 100.635. In principle,this (putative) change in affinity can be estimated from the flesinoxan concentration responsecurve. Using the equation for a single binding site (NH=1) a shift in responsiveness from 80%to 35% corresponds with a 5-6 fold decrease in affinity. Arguably, a higher concentration ofthe antagonist might have augmented the effect of citalopram following chronic treatment. Anadditional confounding factor could be that other neuronal components e.g. the 5-HT carrierand 5-HT1B receptors also underwent adaptive changes, which complicates the interpretationof the results considerably. Definitely more work is needed to investigate the interplaybetween the various neuronal components in the acute and chronic effects of SSRIs.Nevertheless, the absence of augmentation by WAY 100.635 and the significantly lesser effectof flesinoxan in animals chronically treated with citalopram both support the idea thatpostsynaptic 5-HT1A receptors in the amygdala have become less responsive following chronictreatment with citalopram. Future work in other brain areas, while using different SSRIs, isneeded to determine how important this type of feedback is as target for antidepressanttreatment.

4.4 Citalopram pharmacokinetics

The focus of the present study is on the pharmacodynamic aspects of citalopramadministration, but some pharmacokinetic aspects were also taken into consideration.First, we wanted to achieve stable and “sufficiently” high citalopram plasma levels in thechronic experiments. Effective citalopram plasma concentrations reported for depressedpatients treated with Cipramil range between 0.13 and 0.88 µM (Baumann, 1992). We haveused this range as the starting point for chronic treatment.

Chapter 6

148

SSRI plasma half lives are much shorter in rodents compared to humans. To obtain steady-state conditions for citalopram in rats multiple injections would have been necessary, which isstressful for the animals. Citalopram’s excellent solubility in saline enabled us to use osmoticminipumps and yet to achieve citalopram plasma levels within the clinically effective range.The effect of the metabolites on 5-HT uptake is considerably less than for citalopram.Therefore, their contribution to the acute and chronic effects of citalopram may be minor(Hytell, 1994).Second, we wanted to be sure that after the removal of the minipumps residual citalopram wasso low that it would not interfere with the microdialysis experiments. Two measures weretaken to achieve this goal. The cavity wherein the minipump was housed was washed twicewith saline and the microdialysis experiments were performed two days after the removal ofthe minipumps. After two days citalopram plasma levels had dropped to 8.2 nM (Cremers etal., 2000b). More important, citalopram was no longer pharmacologically active at theselevels, as witnessed by the comparable basal 5-HT levels for the citalopram and saline group.

4.5 Conclusions

The present study supports previous findings on postsynaptic 5-HT1A receptor mediatedfeedback and demonstrates that this type of feedback is strongly involved in the acute effect ofcitalopram on extracellular 5-HT in the amygdala. The study also demonstrates that this 5-HT1A receptor mediated feedback is reduced following chronic citalopram treatment.


149

- citalopram + citalopramAcute (figs 2 & 3)

Flesinoxan 300nM S χ215 = 29.2; p=0.0152 NS

Flesinoxan 1000nM S χ215 = 27.0; p=0.0286 S χ2

15 = 40.0; p= 0.0005Flesinoxan 3000nM S χ2

15 = 51.0; p< 0.0001 S χ215

= 32.5; p= 0.0055Chronic (fig 5)

Flesinoxan 1000nMSaline vs.citalopram

S F(1,112)= 4.380 NS F(1,207)=1.759

Table 1 Flesinoxan: statistical data

Citalopram Citalopram + WAY 100635Acute (fig 4) S χ2

12 = 29.2, p= 0.0037Cit vs Cit + WAY 100635 S F(1,183)= 5,33Chronic (fig 6) S χ2

10= 19,7; p=0.0311 S χ210=27,9; p=0.0018

Cit vs Cit + WAY 100635 NS F(1,130)=0.261Table 2 Augmentation with WAY 100635: statistical data

Chapter 6

150

References

Artigas F (1993) 5-HT and antidepressants: new views from microdialysis studies. TrendsPharmacol Sci 14, 262-263.Artigas F, Perez V, Alvarez E (1994) Pindolol induces a rapid improvement of depressedpatients treated with serotonin reuptake inhibitors. Arch Gen Psychiatry 51, 248-251.Baumann P (1992) Clinical pharmacokinetics of citalopram and other selective serotonergicreuptake inhibitors (SSRI). Int Clin Psychopharmacol 6 Suppl 5, 13-20.Berman R M, Darnell A M, Miller HL, Anand A, Charney DS (1997) Effect of pindolol inhastening response to fluoxetine in the treatment of major depression: a double-blind, placebo-controlled trial. Am J Psychiatry 154, 37-43.Blier P, De Montigny C, Chaput Y (1987a) Modifications of the serotonin system byantidepressant treatments: implications for the therapeutic response in major depression. J ClinPsychopharmacol 7, 24S-35S.Blier P, de Montigny C, Tardif D (1987b) Short-term lithium treatment enhancesresponsiveness of postsynaptic 5-HT1A receptors without altering 5-HT autoreceptorsensitivity: an electrophysiological study in the rat brain. Synapse 1, 225-232. Blier P, De Montigny C (1987c) Modification of 5-HT neuron properties by sustainedadministration of the 5-HT1A agonist gepirone: electrophysiological studies in the rat brain. Synapse1, 470-480.Blier P, Bergeron R (1995) Effectiveness of pindolol with selected antidepressant drugs in thetreatment of major depression. J Clin Psychopharmacol 15, 217-222.Bosch AI, van Hove N, Long SK, de Rond EJ (1994) Systemic flesinoxan depresses hippocampal5-HT release in the freely-moving guinea-pig. (Abstr) Br J Pharmacol 112, 94P.Bosker FJ, Klompmakers AA, Westenberg HGM (1994) Extracellular5-hydroxytryptamine in median raphe nucleus of the conscious rat is decreased by nanomolarconcentrations of 8-hydroxy-2-(di-n-propylamino)tetralin and is sensitive to tetrodotoxin. JNeurochem 63, 2165-2171.Bosker FJ, de Winter TYCE, Klompmakers AA, Westenberg HGM (1996) Flesinoxan dose-dependently reduces extracellular 5-hydroxytryptamine (5-HT) in rat median raphe and dorsalhippocampus through activation of 5-HT1A receptors. J Neurochem 66, 2546-2555.Bosker FJ, Vrinten D, Klompmakers AA, Westenberg HGM (1997a) The effects of a 5-HT1A

receptor agonist and antagonist on the 5-hydroxytryptamine release in the central nucleus of theamygdala: a microdialysis study with flesinoxan and Way 100635. Naunyn Schmiedeberg's ArchPharmacol 355, 247-253.Bosker FJ, Klompmakers AA, Westenberg HGM (1997b) Postsynaptic 5-HT1A receptors mediate5-hydroxytryptamine release in the amygdala through a feedback to the caudal linear raphe. Eur JPharmacol 333, 147-157.Casanovas JM, Hervas I, Artigas F (1999) Postsynaptic 5-HT1A receptors control 5-HT release inthe rat medial prefrontal cortex. Neuroreport 10, 1441-1445.Ceci A, Baschirotto A, Borsini F (1994) The inhibitory effect of 8-OH-DPAT on the firing activityof dorsal raphe serotoninergic neurons in rats is attenuated by lesion of the frontal cortex.Neuropharmacology 33, 709-713.


151

Cremers TIFH, de Boer P, Liao Y, Bosker FJ, den Boer JA, Westerink BHC, Wikstrom HV(2000a) Augmentation with a 5-HT1A, but not a 5-HT1B receptor antagonist depends on thedose of citalopram. Eur J Pharmacol, 397, 63-74.Cremers TIFH, Spoelstra EN, de Boer P, Bosker FJ, Mork A, den Boer JA, Westerink BHC,Wikstrom HV (2000b) Desensitisation of 5-HT autoreceptors upon pharmacokineticallymonitored chronic treatment with citalopram. Eur J Pharmacol, 397, 351-357.Den Boer JA, Westenberg HGM (1988) Effect of a serotonin and noradrenaline uptakeinhibitor in panic disorder: a double-blind comparative study with fluvoxamine andmaprotiline. Int Clin Psychopharmacol 3, 59-74.Fletcher A, Bill DJ, Cliffe I.A, Forster EA, Jones D, Reilly Y (1994) A pharmacological profile ofWAY-100635, a potent and selective 5-HT1A receptor antagonist. (Abstr) Br J Pharmacol 112,P91.Fuller RW (1994) Uptake inhibitors increase extracellular serotonin concentration measuredby brain microdialysis. Life Sci 55, 163-167.Gobert A., Rivet JM, Cistarelli L, Millan MJ (1997) Potentiation of the fluoxetine-inducedincrease in dialysate levels of serotonin (5-HT) in the frontal cortex of freely moving rats bycombined blockade of 5-HT1A and 5-HT1B receptors with WAY 100,635 and GR 127,935. JNeurochem 68, 1159-1163.Gravem A, Amthor KF, Astrup C, Elgen K, Gjessing LR, Gunby B, Pettersen RD, KyrdalenL, Vaadal J, Ofsti E, et al. (1987) A double-blind comparison of citalopram (Lu 10-171) andamitriptyline in depressed patients. Acta Psychiatr Scand 75, 478-486.Hjorth S, Auerbach SB (1994) Further evidence for the importance of 5-HT1A autoreceptorsin the action of selective serotonin reuptake inhibitors. Eur J Pharmacol 260, 251-255.Hjorth S (1993) Serotonin 5-HT1A autoreceptor blockade potentiates the ability of the 5-HTreuptake inhibitor citalopram to increase nerve terminal output of 5-HT in vivo: amicrodialysis study. J Neurochem 60, 776-779.Hjorth S, Bengtsson HJ, Milano S (1996) Raphe 5-HT1A autoreceptors, but not postsynaptic5-HT1A receptors or beta-adrenoceptors, restrain the citalopram-induced increase inextracellular 5-hydroxytryptamine in vivo. Eur J Pharmacol 316, 43-47.Hutson PH, Sarna GS, O’Connell MT, Curzon G (1989) Hippocampal 5-HT synthesis andrelease in vivo is decreased by infusion of 8-OH-DPAT into the nucleus raphe dorsalis.Neurosci Lett 100, 276-280.Hyttel J (1994) Pharmacological characterization of selective serotonin reuptake inhibitors(SSRIs). Int Clin Psychopharmacol 9 Suppl 1, 19-26.Invernizzi R, Belli S, Samanin R (1992) Citalopram’s ability to increase the extracellularconcentrations of serotonin in the dorsal raphe prevents the drug’s effect in the frontal cortex.Brain Res 584, 322-324.Kendrick KM (1990) Microdialysis measurement of in vivo neuropeptide release. J NeurochemMethods 34, 35-46.Kreiss DS, Lucki I (1994) Differential regulation of serotonin (5-HT) release in the striatumand hippocampus by 5-HT1A autoreceptors of the dorsal and median raphe nuclei. J PharmacolExp Ther 269, 1268-1279.

Chapter 6

152

Limberger N, Deicher R, Starke K (1991) Species differences in presynaptic serotoninautoreceptors: mainly 5-HT1B but possibly in addition 5-HT1D in the rabbit and guinea-pigbrain cortex. Naunyn Schmiedeberg’s Arch Pharmacol 343, 353-364.Montgomery SA, Rasmussen JGC, Lyby K, Connor P, Tanghoj P (1992) Dose responserelationship of citalopram 20 mg, citalopram 40 mg, and placebo in the treatment of moderateand severe depression. Int Clin Psychopharmacol 6 S5, 65-70.Montgomery SA, McIntyre A, Osterheider M, Sarteschi P, Zitterl W, Zohar J, Birkett M,Wood AJ (1993) A double-blind, placebo-controlled study of fluoxetine in patients with DSM-III-R obsessive-compulsive disorder. The Lilly European OCD Study Group. EurNeuropsychopharmacol 3, 143-152.Munday MK, Fletcher A, Marsden CA (1994) Effect of the putative 5-HT1A antagonist WAY100635 on 5-HT neuronal firing in the guinea pig dorsal raphé nucleus. (Abstr) Br J Pharmacol112, 93P.Oyehaug E, Ostensen ET, Salvesen B (1982) Determination of the antidepressant agentcitalopram and metabolites in plasma by liquid chromatography with fluorescence detection. JChromatography 227, 129-135.Paxinos G, Watson C (1982) The Rat Brain in Stereotaxic Coordinates. (Sydney: Academic Press).Romero L, Celada P, Artigas F (1994) Reduction of in vivo striatal 5-hydroxytryptamine releaseby 8-OH-DPAT after inactivation of Gi/Go proteins in dorsal raphe nucleus. Eur J Pharmacol 265,103-106.Santiago M, Westerink BHC (1990) Characterization of the in vivo release of dopamine asrecorded by different types of intracerebral microdialysis probes. Naunyn-Schniedeberg’s ArchPharmacol 342, 407-414.Sharp T, Bramwell SR, Hjorth S, Grahame-Smith DG (1989) Pharmacological characterization of8-OH-DPAT-induced inhibition of rat hippocampal 5-HT release in vivo as measured bymicrodialysis. Br J Pharmacol 98, 989-997.Sharp T, Umbers V, Gartside SE (1997) Effect of a selective 5-HT reuptake inhibitor incombination with 5-HT1A and 5-HT1B receptor antagonists on extracellular 5-HT in ratfrontal cortex in vivo. Br J Pharmacol 121, 941-946.Sprouse JS, Aghajanian GK (1987) Electrophysiological responses of serotonergic dorsal rapheneurons to 5HT1A and 5-HT1B agonists. Synapse 1, 3-9.Van Vliet IM, Den Boer JA, Westenberg HGM (1994) Psychopharmacological treatment ofsocial phobia; a double-blind placebo-controlled study with fluvoxamine. Psychopharmacology115, 128-134.

Is the beneficial antidepressant effect of co-administration of pindolol really due to somatodendriticautoreceptor antagonism?

153

Chapter 7

Is the beneficial antidepressant effect of co-

administration of pindolol really due to

somatodendritic autoreceptor antagonism ?

Chapter 7

154

Abstract

The combination of SSRIs with the beta adrenoceptor / 5HT1A receptor antagonist pindolol iscurrently investigated, based on the concept that 5-HT1A receptor blockade would eliminate theneed for desensitization of presynaptic 5-HT1A receptors and therefore hasten the onset of actionand improve the efficacy of SSRIs. However, since pindolol plasma levels after 2.5 mg t.i.d. areabout 60 nM, and the Ki for the 5-HT1A receptor is 30 nM, it is questionable whether pindolollevels in the brain would be sufficient to antagonise 5-HT1A receptors.Using microdialysis and jugular vein catherisation, the ability of systemically administeredpindolol to antagonise central 5-HT1A and beta adrenoceptors was studied, while simultaneouslymonitoring pindolol plasma and brain concentration.Augmentation of paroxetine-induced increases in extracellular 5-HT levels in ventralhippocampus was only observed at steady state plasma levels exceeding 7,000 nM (concurrentbrain levels 600 nM). In contrast, antagonism of beta agonist induced increases of brain c-AMPlevels was already observed at pindolol plasma levels of 70 nM (concurrent brain levels < 3nM)At plasma levels that are observed in patients after t.i.d. 2.5 mg (~60 nM), pindolol producesonly a partial blockade of presynaptic 5-HT1A autoreceptors and does not augment the SSRI-induced 5-HT increase in guinea pig brain. It is therefore very unlikely that the favourableeffects of combining pindolol with SSRIs, as reported in a number of clinical studies, are due to5-HT1A antagonism. Since pindolol completely blocks central beta-adrenorecepors at clinicallyrelevant plasma levels, it is possible that beta adrenoceptor antagonism is involved in mediatingpindolol’s beneficial effects.


155

1. Introduction

Selective serotonin re-uptake inhibitors (SSRIs) are believed to exert their antidepressiveeffect by increasing the availability of extracellular 5-HT for interaction with post-synaptic 5-HT receptors. However, increased 5-HT levels will also activate cell body 5-HT1A

autoreceptors, resulting in reduced cell firing rate and ultimately in decreased terminal 5-HTrelease. Chronic administration of SSRIs has been shown to desensitise 5-HT1A

autoreceptors, thereby relieving the serotonergic neuron of its auto-inhibitory regulation. Thissequence of events is thought to be responsible for the slow onset of action of SSRI’s, whichoften requires 3-4 weeks before clinical effects are observed (Blier et al. 1987). Therefore, itwas hypothesised that co-administration of an SSRI with a 5-HT1A autoreceptor antagonistwould lead to a more rapid onset of antidepressive action (Artigas 1993, Blier et al. 1994).Several clinical trials have indeed shown that combining an SSRI with the mixed beta-adrenoceptor/5-HT1A antagonist, pindolol, might be beneficial in the treatment of depression,although others failed to observe a more rapid onset of action or a superior effect of thecombination (McAskill et al.1998, Nelson 2000). The use of various SSRIs and diverseinclusion criteria for the trials might at least partially explain these discrepancies. However,it is obvious that pharmacokinetic considerations should be taken into account as well, sincethe augmentation of the SSRI evoked increase in extracellular 5-HT levels in rat or guineapig brain is known to be dependent on the dose, and consequently brain levels, of both theSSRI and the 5-HT1A receptor antagonist (Hjorth et al.1997, Cremers et al. 2000).In the majority of the animal studies the minimal pindolol dose required to augment the SSRIeffect on 5-HT levels ranges from 8 to 15 mg/kg i.p. or s.c. (Dreshfield et al.1996, Hjorth1996, Romero et al. 1996). Although the elimination half-life of pindolol in rodents is muchfaster than in humans (20 min vs. 3.5 h), these doses will still induce very high plasma levels.Based on a volume of distribution of 7.6 L/kg for the active enantiomer in rats and assumingthat subcutaneous absorption is not rate limiting, 8-15 mg/kg (-) pindolol will give initialplasma concentrations of approximately 6 µM (Ci=Dose/Vd; Hasegawa et al. 1989). Theseplasma levels are two orders of magnitude higher than those found in patients undergoingcombination therapy with SSRIs and 2.5–5 mg t.i.d. pindolol, when plasma levels are foundto be 30-60 nM (Moffat et al.1986, Hasegawa et al.1989, Perez et al. 1999). Because pindololis a rather polar compound (log P=–0.9; Moffat et al.1986) and 50% protein bound (Belpaireet al. 1982), it will not easily penetrate into the brain and free pindolol concentrations in thebrain are expected to be much lower than plasma concentrations. With a Ki for the 5-HT1A

receptor in the same range as the clinical plasma concentrations (30 nM (Boddeke 1992)), itis therefore very unlikely that the brain levels of pindolol after 2.5-5 mg t.i.d. are high enoughto block central presynaptic 5-HT1A receptors in the raphe nuclei.The present study addresses these inconsistencies by examining the pharmacokinetics andpharmacodynamics of pindolol in guinea pigs, during co-administration with the SSRIparoxetine. We determined to what extent different pindolol plasma levels affected theparoxetine-induced increase in extracellular 5-HT in rat ventral hippocampus, as an estimateof the degree of 5-HT1A receptor blockade. In addition, since pindolol is a more potent beta-adrenoceptor antagonist (Ki= 2 nM) than 5-HT1A antagonist, we also determined the effect of

Chapter 7

156

pindolol administration on the beta-receptor agonist-induced stimulation of c-AMP formation(Beer et al. 1988). During these experiments plasma levels of pindolol and paroxetine, as wellas pindolol brain levels in the raphe nuclei, were measured to correlate actual brain andplasma levels with the pharmacological effects. Since pindolol has considerable affinity forrat 5-HT1B, but not for human and guinea pig 5-HT1B receptors, the present study wasperformed in freely moving guinea pigs in order to prevent interference by 5-HT1B

antagonism of pindolol during SSRI administration (Moret et al. 1997, Zgombick et al.1997). In addition, the affinity of (-)-pindolol for guinea pig and human 5-HT1A receptors isnear similar (Raurich et al. 1999)


157


2.1 Animals and drugs

Male albino guinea pigs of a Dünkin Hartley strain (300-400 g; Harlan, Zeist, TheNetherlands) were housed in cages (32 x 40 x 40 cm) with free access to food and water. Allexperiments were concordant with the declarations of Helsinki and were approved by theAnimal Care Committee of the Faculty of Mathematics and Natural Science of the Universityof Groningen.The following drugs were used: paroxetine (SKB, West Sussex, UK), (+) pindolol, (+)-isoprenaline (RBI, Natick, USA) and dibucaine (Sigma, St. Louis, USA). Paroxetine wasdissolved in water and injected subcutaneously. Pindolol was dissolved in glacial acetic acid(10,100 and 1000 mg/ml), diluted with saline to 0.1,1 and 10 mg/ml and adjusted to pH 5.5.Isoprenaline solutions for local administration were freshly prepared daily in a concentrationof 10 µM in Ringer’s solution containing 5 µM of ascorbic acid to prevent oxidation ofisoprenaline.

2.2 Surgery and microdialysis

Guinea pigs were anaesthetised with ketamine/xylazine (50/8 mg/kg), after premedicationwith midazolam (5 mg/kg s.c.) and using lidocaine-HCl 10 % (m/v) for local anaesthesia.The animals were placed in a stereotaxic frame (Kopf, USA) and home-made I-shaped probeswere implanted and secured with dental cement, using the following co-ordinates Luparello(1965) for 5-HT and cAMP measurements in ventral hippocampus: anterior from intra-aural+ 4.9 mm, lateral from midline +/- 6.5 mm, ventral from dura - 9.0 mm; for pindololmeasurements in the raphe nuclei: anterior + 2.8 mm, lateral + 2.6 mm ventral - 11.0 mm atan angle of 12°. Probes for serotonin and cAMP measurements were constructed with apolyacrylonitrile/sodium methyl sulphonate copolymer dialysis fibre ( 4 mm open surface,i.d. 220 µm, o.d. 310 µm, AN 69, Hospal, Italy), for pindolol measurements with a cellulosemembrane (5 mm open surface, i.d.200 µm, Spectra/Por®, Hollow fiber bundles) (Santiago &Westerink 1990). A polyethylene canula (8 cm, i.d. 0.5 mm, o.d. 1.0 mm) was positionedsubcutaneously in the neck region for pindolol infusion and a silicon canula was inserted 4.2mm into the right jugular vein for blood sampling. The tubing was guided subcutaneously tothe skull and connected with a stainless steel inlet, which was mounted on the skull withdental cement and surgical screws. After insertion, the canula was filled with a PVP solution(55 % polyvinylpyrrolidon in 500 IE/ml heparin) in saline to prevent blood clotting.Postoperative analgesia was accomplished by an intramuscular injection of 0.1 mg/kgbuprenorphine and the animals were allowed to recover for at least 24 hours.At the day of the experiment, microdialysis probes were connected with flexible PEEKtubing to a microperfusion pump (Harvard apparatus, South Natick, MA, USA) and perfusedwith artificial cerebrospinal fluid (aCSF), containing 147 mM NaCl, 3.0 mM KCl, 1.2 mMCaCl2, and 1.2 mM MgCl2. Flow rates were 1.5 µl/min for the measurement of extracellular5-HT and c-AMP and 0.1 µl/min for pindolol.

Chapter 7

158

2.3 Effect of pindolol on paroxetine-induced 5-HT increase

The effects of two doses of paroxetine (0.5 and 5 mg/kg s.c.) on extracellular 5-HT levelswere determined during continuous s.c. infusions with saline (controls) or with increasingconcentrations of pindolol. The infusions started 3 hours prior to drug administration. Thepindolol infusions were chosen to give steady state plasma levels of approximately 70, 700and 7,000 nM. Fifteen-minute dialysate samples were collected in vials containing 7.5 µl of0.02 mM acetic acid.

2.4 Effect of pindolol on isoprenaline-induced c-AMP increase

A 10 µM isoprenaline solution in aCSF was perfused in the ventral hippocampus via themicrodialysis probe, during continuous s.c. infusion with saline or with increasingconcentrations of pindolol as described above. Microdialysate samples (45 µl) were collectedin glass vials.

2.5 Pindolol brain concentratons

In a separate experiment, pindolol concentrations were measured in microdialysates from theraphe nuclei for each of the infusion conditions described above. Although extracellular brainconcentrations of exogenous substances are conveniently determined by microdialysis, acorrection for the in vivo recovery of the probe is required (Menacherry 1992). By using anultra-slow flow rate of 0.l µl/min, which gives approximately 100% in vivo recovery, noexogenous markers are needed to correct for in vivo recovery. We found that the cellulosedialysis membrane rapidly followed pindolol concentration changes in the external mediumand that the in vitro recovery approached 100 % at this flow rate. Microdialysateconcentrations are therefore not corrected for recovery. Probes were perfused at 0.1 µl /minwith aCSF, while pindolol was continuously infused via the s.c. canula. 60 Min dialysatesamples were collected on-line in a 20 µl HPLC loop and assayed by HPLC andelectrochemical detection.

2.6 Plasma concentrations of pindolol and paroxetine

Blood samples (0.3 ml) were drawn via the jugular vein canula at different time points todetermine steady state conditions for pindolol or the time-concentration profile of paroxetine.Samples were collected in 1.5 ml eppendorf vials, containing 5 µl heparin (500 IE/mLsaline), mixed and centrifuged for 15 min at 3,000 rpm and 4oC (MSE, England). Aliquots ofthe supernatants were assayed for pindolol and paroxetine.

2.7 Assays

5-HT: Twenty-µl microdialysate samples were injected via an autoinjector (CMA/200refrigerated microsampler, CMA, Sweden) on a 100 x 2.0 mm C18 Hypersil 3 µm column(Bester, Amstelveen, the Netherlands) and separated with a mobile phase consisting of 5 g/Ldi-ammoniumsulfate, 500 mg/L EDTA, 50 mg/L heptane sulphonic acid, 4 % methanol v/v,


159

and 30 µl/L of triethylamine, pH 4.65 at 0.4 ml/min (Shimadzu LC-10 AD) 5-HT wasdetected amperometrically at a glassy carbon electrode at 500 mV vs Ag/AgCl (AntecLeyden, Leiden, The Netherlands). The detection limit was 0.5 fmol 5-HT per 20 µl sample(signal to noise ratio 3).Paroxetine: Plasma paroxetine levels were measured after liquid-liquid extraction withdiethylether. To 250 µl plasma samples 75 µl of a 5 µg/ml dibucaine solution in 1 % m/vNaHCO3 was added. Samples were extracted thrice by mechanically shaking for 3 minuteswith 4 ml of diethyl ether. The ether layers were then transferred to 10 ml evaporating tubes,and 150 µl of 0.1 N HCl was added. The ether was evaporated in a water-bath at 40 °C undera stream of nitrogen. The HCl layer was washed once with 0.5 ml ether and 50 µl sampleswere injected onto the column. An HPLC/auto-injector (1084B Liquid Chromatograph,Hewlett-Packard) was used, in combination with a fluorescence detector (470 ScanningFluorescence detector, Waters, England) operating at an absorption wavelength of 295 nm, anemission wavelength of 365 nm, and a slit-width of 12 nm. Paroxetine was separated over aSupelcosil HPLC column (5 µm, C18, 250 x 46 mm, Supelco, the Netherlands) using amobile phase consisting of 46 % v/v acetonitrile, 54 % v/v potassium dihydrogen phosphatebuffer (4.3 g/l), and 30 µl/l triethylamine, pH 3.0, at 1 ml/min. Recovery of paroxetine andthe internal standard was approximately 100 %, and concentrations were calculated using theinternal standard. The detection limit of the assay was 8 nM (signal to noise = 3).Pindolol in plasma: Pindolol plasma concentrations were also determined after liquid-liquidextraction with diethylether. To 100 µl of plasma, 50 µl of 1 N NaOH was added and sampleswere extracted thrice by mechanically shaking for 3 minutes with 4 ml of diethyl ether. Thecombined organic layers were evaporated into 100 µl of 1 mM HCl and 20 µl samples wereinjected onto the column. The detection limit of the assay was 1 nM (signal to noise = 3).Recovery of pindolol, determined by extracting plasma samples spiked with knownconcentrations, was about 95 % and all samples were corrected for recovery. Analysis wasperformed as described below.Pindolol in microdialysates: Pindolol concentrations in raphe nuclei dialysate were measuredby HPLC-ECD. Samples were collected on-line and injected automatically every 60 min andseparated over a Supelcosil column, with a mobile phase consisting of 4.27 g/l sodiumacetate, 221 mg/l tetramethyl ammonium, 100 mg/l ethylenediaminotetra-acetic acid, and 10% v/v acetonitrile, pH value 4.25, at 1 ml/min (LKB 2150, Pharmacia LKB, Sweden).Pindolol was detected amperometrically at a glassy carbon electrode set at 850 mV vsAg/AgCl (Antec Leyden, Leiden, the Netherlands). The detection limit for pindolol was 3nM.c-AMP: Concentrations of cAMP in microdialysates from the ventral hippocampal weredetermined as described by Svensson et al. (Svensson et al. 1990). Briefly, 10 µl of 0.5 Msodium acetate and 10 µl 55% chloroacetaldehyde (Fluka) were mixed with 45 µlmicrodialysate in vials which were capped, heated in a boiling water bath for 20 min, injected(1084B Liquid Chromatograph, Hewlett-Packard) onto a Supelcosil guard plus analyticalcolumn (3C18 3 µm, 15 cm x 2.1 mm; Security Guard, Phenomenex, Bester, Amstelveen, theNetherlands). c-AMP was eluted with 1mM tetramethylammonium, 9 % tetrahydrofuran, pH4.5, at a flow rate of 0.4 ml/min and detected fluorimetrically (at absorption and emission

Chapter 7

160

wavelengths of 275 nm and 418 nm, respectively (slit-width 12 nm)). The limit of detectionwas 10.5 fmol/sample (signal to noise =2).2.8 Protein bindingProtein binding of pindolol was determined by spiking guinea pig plasma with knownconcentrations of pindolol dissolved in phosphate buffer pH 7.6. After a 1 hour incubation atroom temperature, samples were transferred to ultrafiltration tubes (Amicon, cut-off 3 kD),and spun at 1000 rpm for 30 minutes. The free fraction was determined by HPLC-ECD asdescribed above.


Four consecutive microdialysis samples with less then 20 % variation in levels of 5-HT andc-AMP were taken as baseline levels and set at 100 %. Drug effects were expressed aspercentages of basal level (mean + SEM). Statistical analysis was performed using Sigmastatfor windows (Jandel Corporation). Treatment effects were compared versus saline treatmentusing two way ANOVA for repeated measurements, followed by the Student Newman Keulspost-hoc test . Treatment effects were compared versus control values using one wayANOVA for repeated measurements on ranks. The level of significance level was set atp<0.05.


161

3. Results

3.1 Pharmacokinetics of pindolol and paroxetineSubcutaneous infusion of pindolol resulted in steady state plasma levels at approximately 3hours after initiation of the infusions. Infusions with 0.1 mg/kg/ml/h produced stable pindololplasma levels of about 70 nM. Ten-fold increases in the infusion concentration of pindolol (1mg/kg/ml/h and 10 mg/kg/ml/h) resulted in ten-fold higher plasma levels to approximately700 and 7,000 nM, respectively (figure 1).

Figure 1. Plasma levels of pindolol during subcutaneous infusion. � pindolol 0.1mg/kg/ml/h (n=3), � pindolol 1 mg/kg/ml/h (n=4), and � pindolol 10 mg/kg/ml/h (n=5).

The plasma protein binding of pindolol was about 50% over a range of concentrations (n = 2-4, fig 2a) and did not change in the presence of 1 µM paroxetine (n = 2, fig 2b).

0 2 4 6 81

10

100

1000

10000

Pin

dolo

l pla

sma

conc

. (nM

)

Time (h)

Chapter 7

162

Figure 2. (a) percentage protein binding of pindolol at several concentrations (n=2-4), (b)percentage protein binding of pindolol at 1 µM concentration, with and without the presenceof paroxetine 1 µM (n=4).

During pindolol steady state plasma concentrations of 70 nM, pindolol was not detectable indialysates from the raphe nuclei (< 3 nM). However, at higher pindolol plasmaconcentrations, the extracellular brain levels of pindolol were approximately 30 nM during 1mg/kg/ml/h infusion, and 600 nM during 10 mg/kg/ml/h infusion (figure 3). Administrationof 5 mg/kg s.c. paroxetine during 10 mg/kg/ml/h pindolol infusions, had no effect on theplasma or brain levels of pindolol (figure 4a-b).Subcutaneous administration of paroxetine rapidly induced measurable plasma concentrationsof the drug. Peak plasma levels after 0.5 mg/kg s.c. paroxetine were 0.5 µM, which declinedwith an elimination half live of about 75 min. After 5 mg/kg paroxetine s.c. peak plasmalevels were 2 µM, after which elimination was bi-phasic, with apparent elimination half livesof 220 min and 800 min, respectively (fig 5).

3.2 Effect of pindolol on the paroxetine-induced 5-HT increaseAbsolute output of 5-HT from guinea pig ventral hippocampus was 8.79 +0.67 fmol per 15minute sample (n=38).Subcutaneous administration of 0.5 mg/kg paroxetine during saline infusion increasedhippocampal 5-HT levels about 2-fold. When 0.5 mg/kg paroxetine was administered duringpindolol infusions to give steady state plasma concentrations of 70 or 700 nM pindolol, noaugmentation of the SSRI effect was observed (F1,140=0.68 and F1,129=1.65, respectively).However, during steady state levels of 7,000 nM pindolol (10 mg/kg/ml/h s.c.), the effect ofparoxetine on extracellular 5-HT levels was significantly potentiated to a 3–fold increase(F1,128= 2.86; figure 6).


163

Essentially the same results were obtained with a 10 times higher dose of paroxetine. The 5-HT increase induced by 5 mg/kg s.c. paroxetine was about 3.5-fold and was unchanged atsteady state pindolol levels of 70 nM (F1,101=0.57) and 700 nM (F1,116=0.58), but wassignificantly greater when pindolol plasma levels were 7,000 nM (F1,117=4.37; figure 7).

Figure 3. Brain levels of pindolol during subcutaneous infusion. � pindolol 1 mg/kg/ml/h(n=3), and � pindolol 10 mg/kg/ml/h (n=3). Brain levels of 0.1 mg/kg/ml/h infusion ofpindolol could not be detected.

3.3 Effect of pindolol on the isoprenaline-induced c-AMP increaseBasal extracellular cAMP levels in microdialysates from ventral hippocampus were 739.0 +157.6 fmol/sample (n= 8). These basal levels were somewhat higher than reported in previousstudies (Stone 1988), probably because of experimental differences, such as dialysis surface,membranes, brain area and species. Continuous local administration of 10 µM of the betaadrenoceptor agonist isoprenaline via the hippocampal probe produced significant increasesin extracellular cAMP levels. A maximal 2–fold increase over basal levels was reached 3hours after drug administration (Χ2

6=13.3, p<0.05). When isoprenaline (10 µM) wasadministered in the presence of pindolol, at steady state levels of 70 nM, the increase in c-AMP formation was completely blocked (F1,61=11.13; figure 8).

0 2 4 6 81

10

100

1000

Pin

dolo

l bra

in c

onc

. (nM

)

Time (h)

Chapter 7

164

Figure 4. (a) Effect of paroxetine 5 mg/kg s.c. on plasma levels of pindolol duringsubcutaneous infusion of 10 mg/kg/ml/h pindolol (� saline injection (n=3), � paroxetine 5mg/kg s.c. (n=3)) (b) Effect of paroxetine 5 mg/kg s.c. on brain levels of pindolol duringsubcutaneous infusion of 10 mg/kg/ml/h pindolol (� saline injection (n=3), � paroxetine 5mg/kg s.c. (n=3)


165

4. Discussion

Several clinical studies have investigated the putative beneficial effects of co-administrationof a 5-HT1A antagonist with antidepressive SSRI treatment, based on the hypothesis that 5-HT1A receptor blockade will hasten the onset of the effects of SSRIs (see Introduction).Although some studies with the beta-adrenergic/5-HT1A receptor antagonist pindolol foundevidence for enhanced efficacy of pindolol-SSRI combinations (McAskill et al. 1998, Nelson2000), pindolol plasma levels after 2.5-5 mg t.i.d. (30-60 nM), seem insufficient toeffectively block central 5-HT1A autoreceptors, and consequently to augment the SSRI-induced 5-HT increase. To investigate at what plasma or brain levels pindolol affects theserotonergic effects of SSRIs in more detail, the present study determined the effects ofincreasing plasma pindolol concentrations on the paroxetine-induced 5-HT increase in guineapig hippocampus.

Figure 5. Plasma levels of paroxetine upon subcutaneous administration of 0.5 mg/kg (�,n=3) and 5 mg/kg (�, n=4)

Pindolol augmentation of the effect of paroxetineContinuous subcutaneous infusions with 0.1, 1.0 and 10 mg/kg/ml/h pindolol, resulted instable pindolol plasma levels of 70, 700 and 7,000 nM, respectively. Pindolol metabolites,which have negligible affinity for 5-HT receptors and are thus not expected to have anypharmacological effect in patients (Ohnhaus et al. 1982), were not detected under theseconditions. During each of the three pindolol steady state conditions, guinea pigs werechallenged with two subcutaneous paroxetine doses, 0.5 and 5.0 mg/kg, and the increase inextracellular 5-HT levels was compared with the increase produced by paroxetine during s.c.

0 200 400 600 8001E-3

0.01

0.1

1

10

Par

oxet

ine

pla

sma

conc

. ( µ

M)

Time (min)

Chapter 7

166

infusion with saline. The two doses of paroxetine were chosen to cover the entire range ofplasma concentrations of paroxetine observed in depressed patients in the clinic (Dechant etal. 1991; Baumann 1992). Paroxetine metabolites, which have low affinity for the 5-HTtransporter, were not detected after administration of either dose (Baumann 1992).

Figure 6. Effect of pindolol pretreatment on paroxetine 0.5 mg/kg s.c. induced increase inextracellular levels of 5-HT. � saline 1 ml/kg/h (n=7), � pindolol 0.1 mg/kg/ml/h (n=6), �pindolol 1 mg/kg/ml/h (n=6), and � pindolol 10 mg/kg/ml/h (n=5), * P<0.05.

At steady state plasma levels of 70 and 700 nM, pindolol failed to change the effect of eitherdose of paroxetine on extracellular 5-HT levels. Only when pindolol plasma levels wereincreased to 7,000 nM, the effects of both doses of paroxetine were significantly greater.These findings are in good agreement with several reports on the augmentation of the SSRI-induced 5-HT increase by pindolol co-administration in rats, which is only observed atpindolol doses exceeding 8 mg/kg s.c. (Dreshfield et al.1996; Hjorth 1996; Romero et al.1996). Doses of 8 mg/kg s.c. will result in pindolol plasma concentrations of approximately6,000 nM (see Introduction), which is in the same range as we found to be required forpotentiation of the SSRI effect by pindolol. It might be questioned whether similarpharmacokinetic-pharmacodynamic relations would be observed when serotonin levels wouldbe measured in terminal areas other than ventral hippocampus. As no data exist on regionaldifferences in dose-dependency of 5-HT1A antagonist induced augmentation, additionalresearch should elucidate this topic. However, since antagonism of 5-HT1A receptors wouldin theory be a concentration dependent process, and not like the effects of SSRI’s acomposition of re-uptake inhibition, and autoreceptor functionality, no differences would beexpected. Microdialysis measurements of the actual free brain concentrations of pindolol atits proposed site of action, the raphe nuclei where presynaptic 5-HT1A autoreceptors arelocated, confirmed that pindolol, which is hydrophilic (logP=-0.90) and 50% protein-bound,does not readily cross the blood-brain-barrier. Brain levels at steady state plasma levels of 70

-100 0 100 200 300 4000

50

100

150

200

250

300

350

400

***

Paroxetine 0.5 mg/kg s.c.Start s.c. pindolol infusion

5-H

T (

% B

asal

Le

vel)

Time (min)


167

nM were not detectable (<3 nM), whereas at plasma concentrations of 700 and 7,000 nM,pindolol extracellular levels in the raphe were about 30 nM and 600 nM, respectively.

Figure 7. Effect of pindolol pretreatment on paroxetine 5 mg/kg s.c. induced increase inextracellular levels of 5-HT. � saline 1 ml/kg/h (n=6), � pindolol 0.1 mg/kg/ml/h (n=4), �pindolol 1 mg/kg/ml/h (n= 5), and � pindolol 10 mg/kg/ml/h (n=5), * P<0.05.

Since it has been suggested that pharmacokinetic interactions between SSRIs and pindololcould possibly lead to changes in drug levels, we also determined the effect of paroxetine onpindolol plasma and brain levels. Since we found no evidence for interaction between the twodrugs at the plasma levels studied, and no effect on the protein binding of pindolol, it isunlikely that the beneficial effects of pindolol addition is correlated with increased pindolollevels.

Implications for 5-HT1A receptor occupancyThe finding that pindolol only enhances the paroxetine-induced 5-HT increase at plasmalevels that are 100-fold higher than clinical pindolol plasma levels, suggests that it is highlyunlikely that SSRI augmentation is due to blockade of presynaptic somatodendritic 5-HT1A

autoreceptors. At clinical plasma levels of 30-60 nM (Moffat et al.1986; Hasgawa et al.1989,Perez 1999), pindolol concentrations in the raphe are ≤ 1 nM and will occupy less than 10%of the 5-HT1A receptors, insufficient to block their activation by the endogenous agonist, 5-HT (see below). This estimate is in reasonable agreement with a recent PET study involunteers who were not receiving SSRI treatment (Andree et al.1999). After a single dose of10 mg (-) pindolol and at plasma levels of 113-153 nM, the 5-HT1A receptor occupancy in thedorsal raphe ranged from 7-25%. It is however likely that receptor occupancy duringcombination therapy with 2.5 mg t.i.d. pindolol will be markedly less, since plasma levelswill be 2-4 fold lower, while at the same time endogenous 5-HT levels will be elevated in thepresence of the SSRI. This is easily demonstrated by extrapolating IC50 values for pindolol in

-100 0 100 200 300 4000

50

100

150

200

250

300

350

400

450

500

550

600

***

Paroxetine 5 mg/kg s.cStart infusion pindolol

**

5-H

T (

% B

asa

l Lev

el)

Time (min)

Chapter 7

168

the absence and presence of an SSRI with the equation, IC50=Ki(1+Cag/KDag) (Cheng &Prusoff 1973). Assuming that brain concentrations of the agonist, 5-HT, are about 1.5 nM(Adell et al. 1991), the IC50 of pindolol for the 5-HT1A receptor is calculated to be 42 nMduring basal conditions (KD 5-HT=3.8 nM (Peroutka 1986), Ki pindolol =30 nM (Boddeke 1992)).After SSRI administration, endogenous 5-HT concentrations are 2-3.5 times higher and underthese conditions the equation predicts that the IC50 is about 54-71 nM and that the receptoroccupancy will be lower than found under basal conditions.

Our finding that free pindolol raphe concentrations should be about 600 nM, or 20 timeshigher than the Ki, to augment the SSRI effect, is consistent with the fact that a strongreceptor occupancy is required for functional 5-HT1A receptor antagonism. This is supportedby the finding that maximal inhibitory concentrations of pindolol at the human 5-HT1A

receptor are also approximately 600 nM (Raurich et al. 1999). In addition, although pindololwas originally assumed to be a full 5-HT1A receptor antagonist, accumulating evidencesuggests that pindolol has actually the properties of a partial agonist (De Vivo &Maayani S.1990, Newman-Tancredi et al. 1998, Fornal et al. 1999, Sprouse et al. 1998) which mayfurther explain the absence of augmentation at low pindolol concentrations. Taken together,these observations are in line with the proposed existence of a substantial 5-HT1A receptorreserve in the raphe nuclei (Cox et al.1993), which demands strong receptor occupation by anantagonist or partial agonist to reduce the effects of an agonist, in particular a full agonistsuch as 5-HT, on 5-HT1A autoreceptors.

Other receptor mechanisms: beta-adrenoceptor blockadeSince at clinical plasma levels the pindolol brain concentrations are insufficient to completelyblock 5-HT1A autoreceptors, other receptor mechanisms that are sensitive to low plasma andbrain levels of pindolol may be involved in the augmentation of the antidepressive effects ofSSRIs. The very high affinity of pindolol for the beta-adrenoceptor (Ki= 2 nM), prompted usto investigate one possible other mechanism, central beta-adrenoceptor blockade. Weassessed the functional beta-antagonist potency of pindolol in vivo via beta-adrenoreceptormediated effects on c-AMP formation. Beta-adrenoceptor stimulation has been shown toproduce increases in extracellular brain levels of cAMP, an effect that could be blocked byselective beta antagonists, demonstrating the beta adrenergic origin of the c-AMP sampled bymicrodialysis (Petersen et al. 1996; Egawa et al. 1988). We also found that local infusionswith 10 µM of the beta-adrenoceptor agonist isoprenaline in the ventral hippocampusproduced 2-fold increases in extracellular c-AMP levels. When the same concentrationisoprenaline was administered during continuous s.c. pindolol infusion (0.1 mg/kg/ml/h) thatgave steady state plasma concentrations of 70 nM, the c-AMP increase was completelyabolished. Whereas plasma levels of 70 nM pindolol thus failed to augment the serotonergiceffect of an SSRI, the same pindolol concentrations were able to completely inhibit theincrease in c-AMP levels produced by local infusion with a beta-adrenoceptor agonist. Thisdemonstrates that at clinically relevant plasma levels of pindolol, beta adrenoceptors arefunctionally blocked. Antagonism of beta-adrenoceptors could be clinically relevant as


169

several authors have shown that beta-adrenoceptors desensitise during treatment withantidepressants (Petersen et al. 1996). Co-administration of SSRIs with beta antagonistswould then mimic this desensitisation and therefore might be beneficial for the treatment ofdepression.

Figure 8. Effect of pindolol 0.1 mg/kg/ml/h pretreatment on local infusion of isoprenaline(10µM) induced increase in extracellular levels of cAMP. � saline 1 ml/kg/h (n=4), �pindolol 0.1 mg/kg/ml/h (n=4), * P<0.05.

In conclusion, the present study provides pharmacokinetic and pharmacodynamic evidencethat pindolol, at the plasma levels observed in clinical studies with 2.5-5 mg t.i.d. pindolol,does not augment the 5-HT increasing effect of the SSRI paroxetine in the guinea pig and thatat these levels 5-HT1A autoreceptors in the raphe nuclei are only partially occupied. Inaddition, pindolol plasma and brain levels are not affected by the simultaneous administrationof paroxetine. Therefore, the putative beneficial effects of co-administration of pindolol withan SSRI are likely mediated via other receptor mechanisms, such as for instance central beta-adrenoreceptor blockade, since beta-adrenergic receptors are completely blocked by pindololat clinically relevant plasma levels.

-100 0 100 200 300 400 5000

50

100

150

200

250

300

isopren 10 microMStart s.c. infusion

*

cAM

P (

% B

asal

Lev

el)

Time (min)

Chapter 7

170

References

Adell A., Carceller A. and Artigas F. (1991). Regional distribution of extracellular 5-hydroxytryptamine and 5-hydroxyindoleacetic acid in the brain of freely moving rats. Journalof Neurochemistry 56 (2): 709-712Andree B., Thorberg S., Halldin C. and Farde L. (1999). Pindolol binding to 5-HT1Areceptors in the human brain confirmed with positron emission tomography.psychopharmacology 144 : 303-305Artigas F. (1993). 5-HT and antidepressants: new views from microdialysis studies. Trendspharmacol. Sci. 14 (7) : 262Baumann P. (1992). Clinical pharmacokinetics of citalopram and other selective serotonergicreuptake inhibitors (SSRI). Int Clin Psychopharmacol 6 Suppl 5 : 13-20Beer M., Richardson A., Poat J., Iverson L.L. and Stahl S.M. (1988). In vitro selectivity ofagonists and antagonists for beta1- and beta 2-adrenoceptor subtypes in rat brain. Biochem.Pharmacol. 15; 37 (6):1145-51Belpaire F.M., Bogaert M.G. and Rossenue M. (1982). Binding of β-adrenoceptor blockingdrugs to human serum albumin, to α1-acid glycoprotein and to human serum. Eur J ClinPharmacol 22 : 253-256Blier P. and de Montigny C. (1994). Current advances and trends in the treatment ofdepression [see comments]. Trends Pharmacol Sci 15 : 220-226Blier P., de Montigny C. and Chaput Y. (1987). Modifications of the serotonin system byantidepressant treatments: implications for the therapeutic response in major depression. JClin Psychopharmacol 7 : 24S-35SBoddeke H.W.G., Fargin A., Raymond J.R., Schoeffter P. and Hoyer D. (1992).Agonist/antagonist interactions with cloned human 5-HT1A receptors: variations in intrinsicactivity studied in transfected HeLa cells Naunyn-Schmiedeberg’s Arch Pharmacol 345 :257-263Cheng Y-C and Prusoff W.H. (1973). Relationship between the inhibition constant (Ki) andthe conentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymaticreaction. Biochemical Pharmacology 22 : 3099-3108Cox R.F., Meller E. and Wazczak B.L. (1993). Electrophysiological evidence for a largereceptor reserve for inhibition of dorsal raphe neuronal firing by 5-HT1A antagonists.Synapse 14 : 297-304Cremers T.I.F.H., de Boer P., Liao Y., Bosker F.J., den Boer J.A., Westerink B.H.C. andWikström H.V. (2000). Augmentation with a 5-HT1A, but not a 5-HT1B receptor antagonistcritically depends on the dose of citalopram. Eur J Pharmacol. 26 ; 397(1):63-74.Dechant K.L. and Clissold S.P. (1991). Paroxetine. A review of its pharmacodynamic andpharmacokinetic properties, and therapeutic potential in depressive illness. Drugs 41 : 225-253De Vivo M. and Maayani S. (1990). Stimulation and inhibition of adenylyl cyclase by distinct5-hydroxytryptamine receptors. Biochem Pharmacol 1 ; 40(7):1551-8


171

Dreshfield L.J., Wong D.T., Perry K.W. and Engleman E.A. (1996). Enhancement offluoxetine-dependent increase of extracellular serotonin (5-HT) levels by (-)-pindolol, anantagonist at 5-HT1A receptors. Neurochem Res 21 : 557-562Egawa M., Hoebel B.G. and Stone E.A. (1988). Use of microdialysis to measure brainnoradrenergic receptor function in vivo. Brain Research 458 : 303-308Fornal C.A., Martin F.J., Metzler C.W. and Jacobs B.L. (1999). Pindolol suppressesserotonergic neuronal activity and does not block the inhibition of serotonergic neuronsproduced by 8-hydroxy-2-(di-n-propylamino)tetralin in awake cats. The journal ofpharmacology and experimental therapeutics 291 : 229-238Hasegawa R., Murai Kushiya M., Komuro T. and Kimura T. (1989). Stereoselectivedetermination of plasma pindolol in endotoxin-pretreated rats by high-performance liquidchromatography. J Chromatogr 494 : 381-388Hjorth S. (1996). (-)-Pindolol, but not buspirone, potentiates the citalopram-induced rise inextracellular 5-hydroxytryptamine. Eur J Pharmacol 303 : 183-186Hjorth S., Westlin D. and Bengtsson H.J. (1997). WAY100635-induced augmentation of the5-HT-elevating action of citalopram: relative importance of the dose of the 5-HT1A(auto)receptor blocker versus that of the 5-HT reuptake inhibitor. Neuropharmacology 36 :461-465McAskill R., Mir S. and Taylor D. (1998). Pindolol augmentation of antidepressant therapy.Br J Psychiatry 173 : 203-208Menacherry S., Hubert W. and Justice J.B. (1992). In vivo calibration of microdialysisprobes for exogenous compounds. Anal Chem 64 : 577-583Moffat A.C., Jackson J.V., Moss M.S. and Widdop B. (1986). Clarke's isolation andidentification of drugs. London: Pharmaceutical press.Moret C. and Briley M. (1997). Effects of milnacipran and pindolol on extracellularnoradrenaline and serotonin levels in guinea pig hypothalamus. Journal of neurochemistry.69(2) : 815-822Nelson J.C. (2000). Augmentation strategies in depression 2000, Journal of clinicalpsychiatry 61 [suppl 2]: 13-19Newman-Tancredi A., Chaput C., Gavaudan S., Verriele L. and Millan M.J. (1998). Agonistand antagonist actions of (-)pindolol at recombinant, human serotonin1A (5-HT1A) receptors.Neuropsychopharmacology 18 : 395-397Ohnhaus E.E., Heidemann H., Meier J. and Maurer G. (1982). Metabolism of pindolol inpatients with renal failure, Eur. J. Clin. Pharmacol 22 : 423-428Petersen B. and Mork A. (1996). Chronic treatment with citalopram induces noradrenalinereceptor hypoactivity. A microdialysis study. Eur J Pharmacol 300 : 67-70Perez V., Soler J. Puigdemont D., Alvarez E. and Artigas F. (1999). A double-blind,randomized, placebo-controlled trial of pindolol augmentation in depressive patients resistantto serotonin reuptake inhibitors. Arch Gen Psychiatry 56 (4) : 375-379Peroutka S.J. (1986) Pharmacological differentiation and characterisation of 5-HT1A, 5-HT1B and 5-HT1C binding sites in rat frontal cortex. J. Neurochem 4 (2) : 529-540.

Chapter 7

172

Raurich A., Mengod G., Artigas F. and Cortes R. (1999). Displacement of the binding of 5-HT1A receptor ligands to pre- and postsynaptic receptors by (-)pindolol. A comparativestudy in rodent, primate and human brain. Synapse 34 : 68-76Romero L., Bel N., Artigas F., de Montigny C. and Blier P. (1996). Effect of pindolol on thefunction of pre- and postsynaptic 5-HT1A receptors: in vivo microdialysis andelectrophysiological studies in the rat brain [published erratum appears inNeuropsychopharmacology 1997 Jan;16(1):91]. Neuropsychopharmacology 15 : 349-360Santiago M. and Westerink B.H. (1990). Characterization of the in vivo release of dopamineas recorded by different types of intracerebral microdialysis probes. Naunyn SchmiedebergsArch Pharmacol 1990 Oct; 342(4):407-14Sprouse J., Braselton J. and Reynolds L. (1998). 5-HT1A agonist activity of pindolol: reversalof the inhibitory effects on cell firing in the dorsal raphe nucleus but not in the hippocampusby WAY-100,635. Ann N Y Acad Sci 861 : 274-275Svensson J.O. and Jonzon B. (1990). Determination of adenosine and cyclic adenosinemonophosphate in urine using solid-phase extraction and high-perfomance liquidchromatography with fluorimetric detection. J Chromatogr 529 : 437-441Zgombick J.M. and Bard J.A., Kucharewicz S.A., Urquhart D.A., Weinshank R.L., BranchekT.A. (1997). Molecular cloning and pharmacological characterization of guinea pig 5-HT1B

and 5-HT1D receptors. Neuropharmacology 36 : 513-524

Attenuation of SSRI-induced increases in extracellular brain 5-HT by benzodiazepines

173

Chapter 8

Attenuation of SSRI-induced increases in

extracellular brain 5-HT by benzodiazepines.

Chapter 8

174

Abstract

Enhanced serotonergic neurotransmission is generally thought to be the neurochemical basisof the antidepressant effects of Selective Serotonin Reuptake Inhibitors (SSRIs).The anxiolytic benzodiazepines, on the other hand, have been shown to decrease serotonergicneurotransmission . Since depressed patients are frequently treated with a combination ofSSRI’s and benzodiazepines, we investigated the effects of co-administration of these drugson extracellular levels of serotonin (5-HT) in guinea pig brain.Using microdialysis of 5-HT in ventral hippocampus of freely moving guinea pigs, weinvestigated the effects of the typical benzodiazepines oxazepam and temazepam, alone andin combination with the SSRI paroxetine on extracellular levels of 5-HT.Paroxetine alone increased extracellular 5-HT levels in hippocampus to about 350 % ofcontrol values,whereas oxazepam and temazepam each produced small decreases to 90% and70% of control levels, respectively when administered alone. The combined administration ofparoxetine with oxazepam as well as temazepam significantly attenuated the increase of 5-HT levels induced by paroxetine. Since the co-administration of SSRIs with benzodiazepines attenuates the serotonergictransmission, but enhances the clinical effects of SSRI’s in depressed patients, it might bequestioned whether the mechanism of the antidepressant action of SSRI’s is completely dueto enhanced serotonergic neurotransmission.


175

1. Introduction

Selective Serotonin Re-uptake Inhibitors (SSRIs) are generally believed to exert theirantidepressant effects by enhancing serotonergic neurotransmission (Blier et al. 1987). Uponacute administration of SSRIs, extracellular brain 5-HT levels are increased (Fuller 1994), butthese 5-HT enhancing effects are restricted by the counteraction of release modulatingserotonergic autoreceptors . It has been shown that during chronic treatment theseautoreceptors are desensitized, thereby potentiating the effect of SSRIs on brain 5-HT levels(Blier et al.1987; Invernizzi et al. 1994). Since the gradual desensitization of autoreceptors isthought to produce, at least partly, the antidepressant effects of SSRIs, it was hypothesizedthat co-administration of an SSRI with an autoreceptor antagonist would instantaneouslymimic this desensitization, and therefore reduce the time until onset of antidepressant action(Artigas 1993, Hjorth 1993). Indeed, clinical trials investigating the effects of co-administration of an SSRI with the putative 5-HT1A antagonist pindolol have shownpromising results, indicative of the beneficial effects of augmented 5-HT levels in thetreatment of depression (McAskill et al.1998).Depressed patients are often simultaneously treated with benzodiazepines because ofcomorbidity of depression and anxiety. Although patients who were being treated with otherdrugs are excluded from clinical trials that investigate the effect of SSRIs or combinations ofSSRIs with pindolol , patients on benzodiapines are usually included in these studies (Bordetet al.1998; Perez et al. 1997). Previous preclinical studies had shown that administration ofbenzodiazepines decreases the extracellular levels of 5-HT in a wide variety of brainstructures in rats and guinea pigs (Rex et al.1993; Pei et al.1989,Gibson et al.1996). Sincethis 5-HT decreasing effect could possibly counteract the 5-HT increase produced by SSRI’swe investigated the effect of co-administration of the SSRI paroxetine with two commonlyused benzodiazepines on extracellular 5-HT levels in guinea pig brain.

Chapter 8

176


2.1 Animals and drug administration

Male albino guinea pigs of a Dünkin Hartley strain (300-400 g; Harlan, Zeist, TheNetherlands were housed in cages (32 x 40 x 40 cm), and had free access to food and water.The experiments are concordant with the declarations of Helsinki and were approved by theanimal care committee of the fbaculty of mathematics and natural science of the University ofGroningen .

The following drugs were used: paroxetine (SKB, West Sussex, UK) , Oxazepam andTemazepam (Bufa, the Netherlands). Paroxetine was dissolved in ultrapure water. Oxazepamand temazepam were dissolved in 10 % v/v solutol and ultrapure water. In experiments witha drug alone, the appropriate vehicle for the combination was injected before or after thedrug.

2.2 Surgery

Preceding surgery, the animals were anaesthetised by means of an intraperitoneal injection ofketamine/xylazine (50/8 mg/kg), after premedication with midazolam (5 mg/kg s.c.).Lidocaine-HCl, 10 % (m/v) was used for local anaesthesia. The animals were placed in astereotaxic frame (Kopf, USA), and home made I-shaped probes (polyacrylonitrile / sodiummethyl sulphonate copolymer dialysis fibre; 4 mm open surface, i.d. 220 µm, o.d. 310 µm,AN 69, Hospal, Italy) were inserted into the ventral hippocampus (co-ordinates: IA: + 4.9mm, lateral : +/- 6.5 mm, ventral: - 9.0 mm from the dura mater (Luparello, stereotaxic atlas)and secured with dental cement. Postoperative analgesia was accomplished by anintramuscular injection of 0.1 mg/kg buprenorphine.


Guinea pigs were allowed to recover for at least 24 h, after which theprobes were perfusedwith artificial CSF (147 mM NaCl, 3.0 mM KCl, 1.2 mM CaCl2, and 1.2 mM MgCl2.) at aflow-rate of 1.5 µl / min (Harvard apparatus, South Natick, Ma., USA). 15 minute sampleswere collected in vials containing 7.5 µl of 0.02 mM acetic acid.

2.4 5-HT assay

Concentrations of 5-HT in microdialysates were measured by HPLC with electrochemicaldetection. Twenty µl samples were injected onto a reversed phase column (PhenomenexHypersil 3 : 3 µm, 100 x 2.0 mm, C18, Bester, Amstelveen, the Netherlands) by anautoinjector (CMA/200 refrigerated microsampler, Carnegie Medicine, Sweden). The mobile


177

phase consisted of 5 g/l di-ammoniumsulfate, 500 mg/l ethylene diamino tetra acetic acid(EDTA), 50 mg/l heptane sulphonic acid, 4 % methanol v/v, and 30 µl/l of triethylamine, at apH of 4.65, and was delivered at aflow-rate of 0.4 ml/min (Shimadzu LC-10 AD liquidchromatograph). 5-HT was detected electrochemically at a glassy carbon electrode set at aworking potential of 500 mV vs. Ag/AgCl (Antec Leyden, Leiden, the Netherlands). Thedetection limit was 0.5 fmol 5-HT per 20 µl sample (signal to noise ratio 3).


Four consecutive microdialysis samples with less then 20 % variation were taken as controland set at 100 %. Data are presented as percentages of control level (mean + SEM). Statisticalanalysis was performed using Sigmastat for Windows (Jandel Corporation). Treatmenteffects were compared versus saline treatment using two way ANOVA for repeatedmeasurements, followed by Student’s Newman Keuls post-hoc test. Treatment effects werecompared versus basal values using one way ANOVA for repeated measurements on ranks.Level of significance level was set at p<0.05.

Chapter 8

178

3. Results

3.1 Serotonin basal levels:

Basal levels of 5-HT in dialysates from guinea pig ventral hippocampus were 8.87 + 0.84fmol/sample (mean + S.E.M.) (n= 27). No significant differences were observed between thedifferent treatment groups.

3.2 Oxazepam administration and paroxetine + oxazepam co-administration:

Figure 1 shows the effect of administration of oxezepam alone and in combination withparoxetine. Oxazepam alone significantly decreased extracellular levels of 5-HT in guineapig ventral hippocampus to about 90 % of basal values (χ2

10 = 18.5, p<0.05). Paroxetinealone increased extracellular 5-HT levels to about 350 % of control values (χ2

10 = 28.9;p<0.05). Injection of oxazepam 30 minutes prior to administration of paroxetine significantlyattenuated the 5-HT produced by paroxetine alone to 250% of basal levels (F(1,151) = 2,73;p<0.05).

Figure 1. Effect of administration of 5 mg/kg paroxetine (�, n = 10, vehicle t=0, paroxetinet=30), oxazepam 1µmol/kg (�, n = 5 , oxazepam t=0, vehicle t=30), and paroxetine 5 mg/kgtogether with oxazepam (�, n = 4 ,oxazepam t=0, paroxetine t=30). * denote significant vs.paroxetine alone.

-100 -50 0 50 100 150 2000

100

200

300

400

500

600

* **

*

*

oxazepam

paroxetine

5-H

T %

bas

al le

vel

Time (min)


179

Figure 2. Effect of administration of 5 mg/kg paroxetine (�, n = 10, vehicle t=0, paroxetinet=30), temazepam 1µmol/kg (�, n = 4 , temazepam t=0, vehicle t=30), and paroxetine 5mg/kg together with temazepam oxazepam (�, n = 4 , temazepam t=0, paroxetine t=30). *denote significant vs. paroxetine alone.

3.3 Temazepam administration and paroxetine + temazepam co-administration:

Administration of temazepam alone induced a significant decrease of extracellular levels of5-HT levels to about 70 % of basal values (χ2

10 = 20, p<0.05). Similar to oxazepam, co-administration of temazepam with paroxetine induced a blunted response to paroxetine andextracellular 5-HT levels increased to only 225% (F(1, 150) = 2.10; p<0.05).

-100 -50 0 50 100 150 2000

100

200

300

400

500

600

*** **

*

Temazepam

paroxetine

5-H

T %

bas

al l

evel

Time (min)

Chapter 8

180

4. Discussion

The present data show that co-administration of benzodiazepines and the SSRI paroxetineattenuates the effect of the SSRI on brain extracellular 5-HT levels.The few studies that investigated the effect of benzodiazepine administration on theserotonergic system showed that these drugs have an inhibitory influence on the firing rate of5-HT neurons and decrease terminal 5-HT release in several brain structures of rats (Pei etal. 1989; Gibson et al., 1996) and guinea pigs (Rex et al.1993). It has been suggested thatthis inhibitory effect of benzodiazepines on the serotonergic system is a consequence ofenhanced GABA-ergic transmission, resulting from the binding of benzodiazepines to thebenzodiazepine/GABAA receptor complex. (Pei et al.1989).As mentioned in the introduction, SSRIs are believed to exert their clinical effect byenhancing serotonergic neurotransmission. In support of this, clinical and preclinical studieshave indeed shown evidence for the positive effects of enhanced serotonergic levels intreatment of depression. Analysis of literature revealed that patients treated for depression arefrequently also treated with benzodiazepines and that in trials designed to evaluate the effectsof enhanced serotonergic levels by co-administration of SSRIs with the autoreceptorantagonist pindolol, patients on benzodiazepines were included (Bordet et al.1998; Perez etal.1997).It is clear from the present study that co-administration of benzodiazepines such as oxazepamand temazepam, attenuates the effect of SSRIs on extracellular 5-HT levels and thus the SSRI–induced effect on serotonergic transmission. Therefore if enhanced serotonergicneurotransmission is indeed responsible for the antidepressant action of SSRI’s, one wouldexpect that co-administration of benzodiazepines to patients on SSRIs would also diminishthe antidepressant effects,. This effect however, is not observed in the clinic, in contrast,some authors have suggested that co-administration of an SSRI with a benzodiazepine issuperior to SSRI treatment alone (Smith et al.1998). Co-morbidity of anxiety with depressionin addition to the well-known anxiolytic effect of benzodiazepine could be a feasibleexplanation for this apparent discrepancy (Keller et al.1995). Furthermore, since some potentbenzodiazepines have antidepressant activity of their own, this might further complicate theexplanation of the overall clinical outcome of combined administration of both classes ofcompounds (Petty et al. 1995, Sussman 1998).Theoretically, the observed pharmacological effects of co-administration of oxazepam andtemazepam with paroxetine might be explained by pharmacokinetic interference (druginteractions) between the compounds. However, since oxazepam and temazepam are directlyconjugated, interference does not seem likely between these specific benzodiazepines andparoxetine (Sproule et al.1997).Although the present findings seem to be at odds with the idea that enhanced serotonin levelsare associated with an antidepressant response, the situation might be different after chronictreatment. Since chronic benzodiazepine treatment has been shown to result in subsensitivityof GABA-ergic receptors in the dorsal raphe nucleus (Wilson and Gallager 1988), theseeffects might counteract the inhibitory effects of GABA in the DRN and augment the SSRIinduced 5-HT increases (Tao and Auerbach 1996).


181

It would require further preclinical studies to evaluate whether chronic benzodiazepine andSSRI co-administration attenuates the serotonergic effects of SSRI’s (thus questioning therelevance of enhanced 5-HT levels for the antidepressant response), or that this combinationrapidly desensitizes GABA-ergic mechanisms, leading to an augmented response to SSRIs.

Chapter 8

182

References

Artigas F. (1993). 5-HT and antidepressants: new views from microdialysis studies. TrensPharmacol. Sci. 14(7) :262Blier P., de Montigny C., Chaput Y. (1987). Modifications of the serotonin system byantidepressant treatments: implications for the therapeutic response in major depression. JClin Psychopharmacol 7 : 24S-35SBordet R., Thomas P., Dupuis B. (1998). Effect of pindolol on onset of action of paroxetine inthe treatment of major depression: intermediate analysis of a double-blind, placebo-controlledtrial. Reseau de Recherche et d'Experimentation Psychopharmacologique. Am J Psychiatry155 : 1346-1351Clifford E.M., Gartside S.E., Umbers V., Cowen P.J., Hajos M.,Sharp T. (1998).Electrophysiological and neurochemical evidence that pindolol has agonist properties at the5-HT1aautoreceptor in vivo. Br J Pharmacol 124 : 206-212Fuller R.W. (1994). Uptake inhibitors increase extracellular serotonin concentration measuredby brain microdialysis. Life Sci 55 : 163-167Gibson E.L., Barnfield A.M.C., Curzon G., (1996) Dissociation of effects of chronicdiazepam treatment and withdrawal on hippocampal dialysate 5-HT and MCPP-inducedanxiety in rats. Behavioral Pharmacology 7: 185-193Hjorth, S. (1993), Serotonin 5-HT1A autoreceptor blockade potentiates the ability of the 5-HT reuptake inhibitor citalopram to increase nerve terminal output of 5-HT in vivo: amicrodialysis study. J. Neurochem. 60(2); 776-779.Invernizzi R., Bramante M.,Samanin R. (1994). Chronic treatment with citalopram facilitatesthe effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT1Areceptors. Eur J Pharmacol 260 : 243-246Keller M.B., Hanks D.L. (1995). Anxiety symptom relief in depression treatment outcome. J.Clin. Psychiatry 57 (suppl. 6) : 22-29McAskill R., Mir S.,Taylor D. (1998). Pindolol augmentation of antidepressant therapy. Br JPsychiatry 173 : 203-208Pei Q., Zetterstrom T., Fillenz M. (1989). Both systemic and local administration ofbenzodiazepine agonists inhibit the in vivo release of 5-HT from ventral hippocampus.Neuropharmacology 28 : 1061-1066Perez V., Gilaberte I., Faries D., Alvarez E., Artigas F. (1997). Randomised, double-blind,placebo-controlled trial of pindolol in combination with fluoxetine antidepressant treatment[see comments]. Lancet 349 : 1594-1597Rex A., Marsden C.A., Fink H. (1993). Effect of diazepam on cortical 5-HT release andbehaviour in the guinea-pig on exposure to the elevated plus maze. Psychopharmacology 110: 490-496Petty F., Trivedi M.H., Fulton M., Rush A.J. (1995). Benzodiazepines as antidepressants :Does GABA play a role in depression? Biol. Psychiatry 38 : 578-591Smith W.T., Londborg P.D., Glaudin V., Painter J.R. (1998). Short-term augmentation offluoxetine with clonazepam in the treatment of depression: a double blind study. Am JPsychiatry 155 : 1339-1351


183

Sproule B.A., Naranjo C.A., Bremner K.E., Hassan P.C. (1997). Selective serotonin reuptakeinhibitors and CNS drug interactions. Clin. Pharmacokinet. 33 (6) : 454-471Sussman N. (1998). Anxiolytic antidepressant augmentation, J. Clinical Psychiatry 59 (Suppl5) : 42-48Tao R., Ma Z., Auerbach S.B. (1996). Differential regulation of 5-hydroxytryptamine releaseby GABAA and GABAB receptors in midbrain raphe nuclei and forebrain of rats. BritishJournal of Pharmacology 119 : 1375-1384Wilson M.A., Gallager, D.W. (1988). GABAergic subsensitivity of dorsal raphe neuronsinvitro after chronic benzodiazepine treatment in vivo. Brain Research 473 : 198-202.

Chapter 8

184

Augmentation of antidepressive effects of SSRIs by co-administration of 5-HT2C antagonists.

185

Chapter 9

Augmentation of antidepressive effects of SSRIs

by co-administration of 5-HT2C antagonists.

Chapter 9

186

Abstract

The current study describes a new pharmacological concept for augmentation ofantidepressant treatment by co-administration of an SSRI with a 5-HT2C antagonist. Initiallywe observed an augmentation of the effect of the SSRI citalopram on central 5-HT levels bynon-selective 5-HT2 antagonists like ketanserin and irindalone. Co-administration ofcitalopram with selective 5-HT2C antagonists RS 102221 or SB 242084 enhanced the effectsof citalopram on extracellular 5-HT levels, whereas no additional effect was observed whencitalopram was co-administered with the selective 5-HT2A antagonist MDL 100907.Local infusion of ketanserin in terminal areas of serotonergic neurons as well as in the raphenuclei, while systemically administering citalopram, showed enhanced levels of 5-HT in bothcombinations, indicating an interaction at the level of nerve terminals as well as cell bodies.Local administration of citalopram in terminal areas also showed augmented 5-HT levels whenketanserin was administered systemically, though only at high doses of ketanserin. As localadministration of SSRI plus ketanserin did not increase the effects of the SSRI, the interactionof the SSRI with the re-uptake site is not expected to take place in the vicinity of the probe.Given the present data, the interaction is hypothesised to occur via modulation by 5-HT2C

receptors of a feedback-loop.


187

1. Introduction

All antidepressants which have been marketed thusfar are characterised by their capacity toenhance the extracellular levels of central monoamines (Hyttel 1994). Although the earlierantidepressants like monoamine oxidase inhibitors and the so-called tricyclic antidepressantsare effective in the treatment of depression, these compounds have disadvantages as theyinduce numerous side-effects in addition to their relative unsafety in overdose. Therefore morespecific compounds have been developed which, due to their selective profile (limited to re-uptake inhibition of monoamines), have less pronounced side-effects, and are more safe inoverdose. These Selective Serotonin Re-uptake Inhibitors (SSRIs) as well as SelectiveNorepinephrine re-uptake inhibitors (SNRIs) however, although safe in use, still have a limitedclinical efficacy. Upon initiation of treatment, onset of action of antidepressants takes typically2 to 4 weeks, while side-effects are fully expressed during this period. In addition, 30 % ofpatients starting antidepressant treatment do not respond after several weeks. Therefore, thereis still considerable room for enhanced therapeutic efficacy of antidepressants.The mechanism of action of the current antidepressants is thought to be related to dynamicprocesses which are brought upon by a chronic elevation of monoaminergic neurotransmission(Blier et al. 1987). Several researchers have shown that serotonergic autoreceptors aredesensitised upon chronic treatment with SSRIs. This desensitisation occurs in a similar time-frame as the onset of antidepressant activity in the clinic (2-4 weeks), and has therefore beenthought to be related to antidepressant activity rather than the mere elevation of monoaminelevels, which occurs within hours after initiation of therapy (Invernizzi et al. 1994; Fuller1994; Cremers et al. 2000). Based on these experiments it was hypothesised that co-administration of an SSRI with its relevant autoreceptor antagonists (5-HT1A or 5-HT1B)would instantaneously mimic the dynamic changes established by chronic administration ofSSRIs and therefore enhance antidepressant activity and accelerate the onset of action (Blieret al. 1987; Artigas 1993, Hjorth 1993).Clinical evaluation of this hypothesis has been hampered by the absence of the clinicalavailability of selective and potent antagonists. Although initial trials indicated an enhancedeffectiveness during co-administration of SSRIs with the mixed ß/5-HT1A antagonist pindolol,it has been questioned whether this compound is an appropriate selective ligand (McAskill etal. 1998; Cremers et al. 2000).Enhancement of the effect of SSRIs after co-administration with 5-HT1A and 5-HT1B

antagonists could provide an elegant and powerful tool for the treatment of depression.However, the augmentation of central serotonin levels itself could facilitate the developmentof several serotonin related side-effects like the serotonergic syndrome.In the present study we investigated the possibility of using 5-HT2 antagonism as a tool foraugmentation of the effects of SSRIs. As 5-HT2 antagonists possess prominent anxiolyticactivity, as well as anti-aggressive properties, augmentation using these compounds wouldproduce an effective, but also a safe way to enhance antidepressant activity.

Chapter 9

188


2.1 Animals


2.2 DrugsCitalopram hydrobromide, irindalone, RS 102221 and SB 242084 were synthesised at andobtained from Lundbeck A/S (Copenhagen, Denmark). MDL 100907, mepyramine, prazosine,ketanserin, ritanserin and (+)- DOI were obtained at RBI (Natick, USA).

2.3 Surgery

Microdialysis of brain serotonin levels was performed using home made I-shaped probes,made of polyacrylonitrile / sodium methyl sulfonate copolymer dialysis fiber (i.d. 220 µm, o.d.0.31 µm, AN 69, Hospal, Italy). The exposed length of the membranes was 4 mm for ventralhippocampus, 4 mm for raphe nuclei, 2 mm for median raphe and 1.5 mm for dorsal raphenucleus.Preceding surgery rats were anaesthetised by means of an intraperitoneal injection ofketamine/xylazine 50/8 mg/kg i.p.), after premedication with 5mg/kg s.c. of midazolam.Lidocaine-HCl (10 % (m/v)) was used for local anaesthesia. Rats were placed in a stereotaxicframe (Kopf, USA), and probes were inserted into Ventral Hippcampus (V. Hippo, L +4.8mm, IA: +3.7 mm, V: -8.0 mm), Dorsal Raphe Nucleus (DRN, L -1.4 mm; IA: +1.2 mm, V: -7.0 mm angle 10o), Median Raphe Nucleus (MRN, L –1.4 mm; IA: +1.2 mm, V: -9.0 mmangle 10o), Raphe Nuclei (RN, L –1.4 mm; IA: +1.2 mm, V: -9.0 mm angle 10o; Paxinos andWatson, 1982). After insertion, probes were secured with dental cement.


Rats were allowed to recover for at least 24 h. Probes were perfused with artificialcerebrospinal fluid containing 147 mM NaCl, 3.0 mM KCl, 1.2 mM CaCl2, and 1.2 mMMgCl2, at a flow-rate of 1.5 µl / min (Harvard apparatus, South Natick, Ma., USA). 15 minutemicrodialysis samples were collected in HPLC vials containing 7.5 µl 0.02 M acetic acid forserotonin analysis.


189

2.5 Serotonin analysis

Twenty-µl microdialysate samples were injected via an autoinjector (CMA/200 refrigeratedmicrosampler, CMA, Sweden) onto a 100 x 2.0 mm C18 Hypersil 3 µm column (Bester,Amstelveen, the Netherlands) and separated with a mobile phase consisting of 5 g/L di-ammoniumsulfate, 500 mg/L EDTA, 50 mg/L heptane sulphonic acid, 4 % methanol v/v, and30 µl/L of triethylamine, pH 4.75 at a flow of 0.4 ml/min (Shimadzu LC-10 AD). 5-HT wasdetected amperometrically at a glassy carbon electrode at 500 mV vs Ag/AgCl (Antec Leyden,Leiden, The Netherlands). The detection limit was 0.5 fmol 5-HT per 20 µl sample (signal tonoise ratio 3).


Four consecutive microdialysis samples with less then 20 % variation were taken as controland set at 100 %. Data are presented as areas under the curve (AUC) calculated as percentagetimes minutes over a period of 150 minutes after injection. Statistical analysis was performedusing Sigmastat for Windows (Jandel Corporation). AUC data were compared using MannWhitney Rank Sum test. Level of significance level was set at p<0.05.

Chapter 9

190

3. Results

Basal outputs:Basal outputs of 5-HT levels in rat ventral hippocampus were 4.39 + 0.24 fmol/sample (n =92). When citalopram 10 µM was applied locally in the in the hippocampus basal output roseto 30.34 + 4.56 fmol/sample (n = 14). In raphe regions, absolute output levels during localadministration of 10 µM of citalopram were 69 + 19.78 fmol/sample (n=18) for DRN and119.12 + 19.41 fmol/sample (n = 21) for MRN.

Effect of combined administration of citalopram with 5-HT2 related ligands on hippocampal5-HT levels.

Fig. 1 shows the effect of citalopram (10 µmol/kg s.c.) on 5-HT levels of rat ventralhippocampus recalculated to areas under the curve (n=7). Putative 5-HT2A/C antagonistketanserin augmented 5-HT levels when given simultaneously with citalopram, however, noeffect was observed when ketanserin was injected alone (n=5, n.s., data not shown). Thepotentiating effect was observed to be dose-dependent, and was already maximal at a dose of10 nmol/kg s.c. (n=4, n=4 and n=4 for ketanserin 1, 10 and 100 nmol / kg s.c., P<0.05 for 10and 100 nmol/kg s.c.). 5-HT2A/C antagonist irindalone also augmented the effect of citalopramin a dose-related fashion (n=3, n=9, n=5, P<0.05 for dose 1000 nmol/kg s.c.).

Fig. 1 Effect of co-administration of citalopram (10 µmol/kg s.c., dashed line representssaline administration) with several 5-HT2 ligands on 5-HT levels in ventral hippocampus (*P<0.05).

cita

lopr

am 1

0 m

icro

mol

+MD

L 10

0907

1m

icro

mol

+ket

anse

rin 1

nm

ol

+ket

anse

rin 1

0 nm

ol

+ket

anse

rin 1

00 n

mol

+irin

dalo

n 10

nm

ol

+irin

dalo

ne 1

00 n

mol

+irin

dalo

ne 1

000

nmol

+RS

1022

21 1

mic

rom

ol

+SB

2420

84 1

mic

rom

ol

+DO

I 1 m

icro

mol

+DO

I 10

mic

rom

ol

0

50000

100000

150000

200000

**

***

AU

C (

% X

min

)


191

Combined administration of citalopram with 1 µmol/kg of the selective 5-HT2A antagonistMDL 100907 did not augment the increase of 5-HT induced by citalopram (n=5, P>0.05).Administration of two selective 5-HT2C antagonists RS 102221 and SB 242084 both in a doseof 1 µmol/kg s.c. significantly enhanced the increase of 5-HT levels established bycitalopram.(n=4, n=4, P < 0.05). Co-administration of citalopram with 5-HT2A/C agonist (+DOI) did not affect 5-HT release (1 µmol/kg s.c. n=7, 10 µmol/kg s.c. n=4).

Infusion of ketanserin with subcutaneous administration of citalopram

Administration of ketanserin 1 µM locally in ventral hippocampus, as well as in raphe nuclei,starting 90 minutes prior to citalopram 10 µmol/kg s.c. injection elicited augmented levels of5-HT (Fig 2, control n=10, n=4 infusion in ventral hippocampus P<0.05, n=4 infusion inraphe nuclei P<0.05). Administration of ketanserin in hippocampus and raphe did not modify5-HT levels (data not shown).

Fig 2. Effect of systemic administration of citalopram (10 µmol/kg s.c.) on hippocampal 5-HT levels during local administration of ketanserin (1 µM) in v.hippocampus or in raphenuclei(* P<0.05).

Infusion of citalopram with subcutaneous administration of ketanserin

Subcutaneous administration of ketanserin (100 and 1000 nmol/kg s.c.) induced a dosedependent increase of 5-HT levels in ventral hippocampus, when re-uptake was locallyblocked by retrograde infusion of 10 µM citalopram in ventral hippocampus (Fig. 3; controln=5, ketanserin 100 nmol/kg s.c. n=5, P>0.05, ketanserin 1000 nmol/kg s.c. n=5, P<0.05).However, the dose of ketanserin needed was much higher than the dose needed in otherparadigms (fig 1), which may indicate the relevance of other receptors at this dose. Local

0

50000

100000

150000

200000

Ketanserin inRaphe

Ketanserin in Hippocampus

Ringer

**

AU

C (

% X

min

)

Chapter 9

192

application of ketanserin (1 µM) in the presence of 10 µM citalopram was devoid of any effecton hippocampal 5-HT levels (n=5, P>0.05).

Fig 3. Effect of systemic (s.c.; nmol/kg) or local (µM) administration of ketanserin on 5-HTlevels in ventral hippocampus during local perfusion of 10 µM of citalopram (* P<0.05).

Effect of ketanserin during activation of long feedback loops

Systemic administration of citalopram 10 µmol/kg s.c. elicited a significant decrease in 5-HTlevels in dorsal as well as median raphe nucleus when simultaneously re-uptake was locallyblocked by infusion of 10 µM citalopram. This decrease unmasks the long-feedback eventsoccurring during systemic administration of an SSRI. Co-administration of 100 nmol/kg s.c.ketanserin failed to significantly alter the decrease established by the systemic administration ofcitalopram in both raphes (Fig 4. saline DRN n=5; Ketanserin 100 nmol/kg s.c. DRN n=4,P>0.05 vs saline; Citalopram 10 µmol/kg s.c. DRN n=5, P<0.05 vs. saline; Citalopram 10µmol/kg and ketanserin 100 nmol/kg s.c. DRN n=4, P>0.05 vs. citalopram and P<0.05 vs.ketanserin, Fig 5. saline MRN n=6; Ketanserin 100 nmol/kg s.c. MRN n=4, P>0.05 vs saline;Citalopram 10 µmol/kg s.c. MRN n=7, P<0.05 vs. saline; Citalopram 10 µmol/kg andketanserin 100 nmol/kg s.c. MRN n=4, P>0.05 vs. citalopram and P<0.05 vs. ketanserin).

Salin

e

Keta

nser

in 1

00nm

ol

Keta

nser

in 1

000

nmol

Keta

nser

in 1

Mic

roM

0

5000

10000

15000

20000

25000

*

AU

C (

% X

min

)


193

Fig 4. Absence of effect of combined administration of ketanserin 100 nmol/kg s.c.and citalopram 10 µmol/kg on 5-HT levels in DRN compared to citalopram 10 µmol/kgalone during local perfusion of 10 µM of citalopram in DRN (* P<0.05).

Fig 5. Absence of effect of combined administration of ketanserin 100 nmol/kg s.c. andcitalopram 10 µmol/kg on 5-HT levels in MRN compared to citalopram 10 µmol/kg aloneduring local perfusion of 10 µM of citalopram in MRN (* P<0.05).

saline ket cit cit/ket0

5000

10000

15000

20000

**

DRN

AU

C (

% X

min

)

saline ketans cit cit/ket0

5000

10000

15000

20000

**

MRN

AU

C (

% X

min

)

Chapter 9

194

3. Discussion and conclusion

Several studies have been performed to evaluate the effects of 5-HT2 ligands on serotonergicneurotransmission. Most 5-HT2 antagonists have not been observed to induce changes inextracellular levels of 5-HT (Lakoski and Aghajanian 1985; Sharp et al. 1989; Devaud andHollingsworth 1991). The 5-HT2 agonist, DOI, was reported to decrease extracellularserotonin levels in terminal areas (Wright et al. 1990). However, as none of the effectsestablished by DOI could be blocked by 5-HT2 antagonists, the involvement of the 5-HT2

receptor in the action of DOI was questioned (Kidd et al. 1991).Apparently, 5-HT2A/C receptors are not relevant for the control of 5-HT levels during basalconditions. However, when the serotonergic system is challenged by administration of anSSRI, it was observed in the present study that ketanserin and irindalone augmented the SSRIinduced increase of serotonin levels.Subsequently we investigated which of the 5-HT2 receptors are involved in controlling 5-HTlevels. Co-administration of citalopram with the putative 5-HT2A antagonist MDL 100907, atdoses which have been shown to induce CNS mediated effects, did not induce anyaugmentation as was also observed by several other researchers (Hatanaka et al. 2000; Scottand Heath 1998; Zhang et al. 2000).However, when citalopram was combined with the selective 5-HT2C antagonists SB 242084 orRS 102221, a clear augmentation of extracellular serotonin levels was observed. Injection ofthe selective compounds alone did not modify 5-HT levels to any extent, indicative that thisinteraction is not present in absence of the SSRI. From these data we conclude that 5-HT2C

receptors are responsible for the augmentation observed with ketanserin and irindalone.5-HT2C receptors are abundantly found in the central nervous system. High levels have beenobserved in chorioid plexus, cerebral cortex, hippocampus, striatum and substantia nigra(Barnes and Sharp 1999; Hoyer et al. 1994). The localisation of these receptors has beenpostulated to be postsynaptic, but some indications are found that they might also be locatedon serotonergic terminals (Mengod et al. 1990, Abramowski et al. 1995). However, sincestimulated release of serotonin from isolated brain tissue in vitro was not modified byketanserin, this possibility should be further elucidated (Moret and Briley 1997; Fink et al.1995).Next we investigated the mechanism of action of the augmentation by 5-HT2C antagonists.When ketanserin was locally applied to the ventral hippocampus preceding to systemicadministration of citalopram, an augmented effect of the SSRI was observed. Also whenketanserin was applied to the raphe nuclei, systemic administration of citalopram elicitedenhanced levels of 5-HT in hippocampus. These experiments show that the site of action ofthe 5-HT2C antagonist is locally in terminal areas, but also in the cell-body region.Local administration of citalopram in the ventral hippocampus followed by systemicadministration of ketanserin induced enhanced levels of 5-HT, possibly indicative of a localinteraction between the re-uptake inhibition and 5-HT2C antagonism. However, the dose ofketanserin needed to establish the effect was considerably higher than the dose needed foraugmentation in other experiments (e.g. augmentation established in fig 1). In addition to theabsence of any effect on 5-HT levels when ketanserin was applied locally during local


195

application of citalopram it should be concluded that the interaction of serotonin with 5-HT2C

receptors is local, while the interaction of the SSRI with the re-uptake site also has to takeplace elsewhere in the brain.Using a paradigm in which long-feedback loops from terminal to raphe regions can beunmasked, it was observed that ketanserin could not antagonise the activation of long-feedback loops to dorsal as well as median raphe. In addition, co-administration of ketanserinwith citalopram tended to induce a larger decrease of 5-HT levels, indicative for enhancedserotonin levels in terminal regions.A short-feedback loop would be the most logic explanation for the observed effects. Apossible mechanism might be related to results of a study by Stanford and Lacey (Stanford etal. 1996), who observed an interaction of 5-HT2C receptors and GABA release. Activation of5-HT2C receptors, located as heteroceptors on GABA-ergic terminals, facilitated GABA-ergicrelease. This enhanced GABA release might decrease serotonergic neurotransmission, as hasbeen shown by Tao and Auerbach (1996). As 5-HT2C receptor antagonists were devoid of anyeffect on basal 5-HT levels, this would indicate that no GABA-ergic tone exists under basalconditions. However, when citalopram was administered, 5-HT2C antagonists were able toenhance 5-HT levels. This means that the presence of an SSRI would enhance GABA-ergicrelease, which in turn would attenuate the effect of the SSRI on terminal 5-HT levels.However, when a 5-HT2C antagonist is administered, the facilitation of GABA release byserotonin is antagonised. By attenuating GABA release, serotonin release would be facilitated.The clinical beneficial effects of the current combination might be more favourable aboveaugmentation of the effect of a SSRI with a 5-HT1A or 5-HT1B antagonist. First, 5-HT2antagonists have been shown to be effective as antidepressants. Trazodone, nefazodone,mianserin and mirtazepine, compounds which are all 5-HT2C antagonists, have been shown tobe effective in depression. Clinical evaluation of combined treatment of depressive patientswith SSRIs and mianserin was shown to have superior clinical efficacy in depressive patients(Maes et al. 1999), indicative for the effectiveness of the present concept in the clinic. Inaddition, 5-HT2C antagonists have been observed to behave as anxiolytics in animals (Bagdy2001), whereas their agonists conversely behave as anxiogenic in animals (Kennett et al. 1997;Kennett. 1994). Furthermore, 5-HT2C antagonists have been observed to increase sleep quality(Kennett 1993).5-HT1A and 5-HT1B antagonists, when given alone, have not been shown to be effective asantidepressants in animal models of depression (Moser 1999, Papp 2002). Combinedadministration of these compounds with SSRIs, however, showed some evidence of enhancedefficacy (Da Rocha 1997), although in several studies no augmentation or even reversal ofbehavioural effects could be observed (O’Neill 1996, Moser 1999, Gardier 2001, Papp 2002).Whereas 5-HT2C antagonists posses prominent anxiolytic activity, limited preclinical data showthat 5-HT1A and 5-HT1B antagonists exert anxiolytic activity (Olivier 1998, Mendoza 1999).Several studies did not observed any change or even anxiogenic effects after administration of5-HT1A and 5-HT1B antagonists (Benjamin 1990, Remy 1996). Although SSRIs are anxiolyticdrugs, they exert some anxiogenic activity in the first period of treatment. Whereas the 5-HT2C

antagonist SB 242084 was shown to reverse SSRI induced anxiety in rats, the 5-HT1A

antagonist WAY 100635 failed to do so (Bagdy 2001).

Chapter 9

196

Taken together, the present pharmacological concept of co-administration of a 5-HT2C

antagonist with SSRIs in depression and anxiety, is expected to behave superior to treatmentwith SSRIs alone. In addition, based on preclinical as well as early clinical evaluation, it isexpected that this combination will act superior to 5-HT1A as well as 5-HT1B antagonistinduced augmentation of antidepressant effects of SSRIs.


197

References

Abramowski D. Rigo M., Duc D., et al. (1995), Localisation of the 5-HT2C receptor protein inhuman and rat brain using specific antisera. Neuropharmacology 34, 1635-1645Artigas, F. (1993). 5-HT and antidepressants: new views from microdialysis studies. TrendsPharmacol Sci, 14, 262Bagdy G., Graf M., Anheuer Z.E., Modos E.A. and Kantor S. (2001). Anxiety-like effectsinduced by acute fluoxetine, sertraline or mCPP treatment are reversed by pretreatment withthe 5-HT2C receptor antagonist SB 242084 but not the 5-HT1A receptor antagonist WAY100635. The Int. J. of Neuropsychopharmacology 4; 399-408.Barnes, N.M. and Sharp, T. (1999). A review of central 5-HT receptors and their function.Neuropharmacology, 38, 1083-1152.Benjamin D., Lal H. and Meyerson L.R. 1990. The effects of 5-HT1B characterizing agents inthe mouse elevated plus-maze. Life Sci. 47(3); 195-203.Blier, P., de Montigny, C. and Chaput, Y. (1987). Modifications of the serotonin system byantidepressant treatments: implications for the therapeutic response in major depression. J ClinPsychopharmacol, 7, 24S-35S.Cremers, T.I.F.H., Spoelstra, E.N., de Boer, P., Bosker, F.J., Mork, A., den Boer, J.A.,Westerink, B.H. and Wikstrom, H.V. (2000). Desensitisation of 5-HT autoreceptors uponpharmacokinetically monitored chronic treatment with citalopram. Eur J Pharmacol, 397, 351-357.Da-Rocha M.A., Puech A.J. and Thiebot M.H. (1997). Influence of anxiolytic drugs on theeffects of specific serotonin reuptake inhibitors in the forced swimmingtest in mice. J.Psychopharmacology, 11(3):211-8.Devaud, L.L. and Hollingsworth, E.B. (1991). Effects of the 5-HT2 receptor antagonist,ritanserin, on biogenic amines in the rat nucleus accumbens. Eur J Pharmacol, 192, 427-429.Fink, K., Zentner, J. and Gothert, M. (1995). Subclassification of presynaptic 5-HTautoreceptors in the human cerebral cortex as 5-HT1D receptors. Naunyn SchmiedebergsArch Pharmacol, 352, 451-454.Fletcher, A., Cliffe, I.A. and Dourish, C.T. (1993). Silent 5-HT1A receptor antagonists: utilityas research tools and therapeutic agents. Trends Pharmacol Sci, 14, 41-48.Fuller, R.W. (1994). Uptake inhibitors increase extracellular serotonin concentration measuredby brain microdialysis. Life Sci, 55, 163-167.Gardier A.M., Trillat A.C., Malagie I., David D., Hascoet M., Colombel M.C., Jolliet P.,Jacquot C., Hen R. and Bourin M. (2001). 5-HT1B serotonin receptors and antidepressanteffect of selective serotonin reuptake inhibitors. C.R. Acad Sci. III 324(5): 433-41.Hatanaka, K., Yatsugi, S. and Yamaguchi, T. (2000). Effect of acute treatment with YM922on extracellular serotonin levels in the rat frontal cortex. Eur J Pharmacol, 395, 23-29.Hjorth, S. (1993), Serotonin 5-HT1A autoreceptor blockade potentiates the ability of the 5-HT reuptake inhibitor citalopram to increase nerve terminal output of 5-HT in vivo: amicrodialysis study. J. Neurochem. 60(2); 776-779.Hoyer, et al. (1994). International union of pharmacology classification of receptors for 5-hydroxytryptamine (serotonin). Pharmacological Reviews, 46, 157

Chapter 9

198

Hyttel, J. (1994). Pharmacological characterization of selective serotonin reuptake inhibitors(SSRIs). Int Clin Psychopharmacol, 9 Suppl 1, 19-26.Invernizzi, R., Bramante, M. and Samanin, R. (1994). Chronic treatment with citalopramfacilitates the effect of a challenge dose on cortical serotonin output: role of presynaptic 5-HT1A receptors. Eur J Pharmacol, 260, 243-246.Kennett. (1994). In vivo properties of SB 200646A, a 5-HT2C/2B receptor antagonist. Br JPharmacol, 111, 797Kennett, G.A., Wood, M.D., Bright, F., Trail, B., Riley, G., Holland, V., Avenell, K.Y., Stean,T., Upton, N., Bromidge, S., Forbes, I.T., Brown, A.M., Middlemiss, D.N. and Blackburn,T.P. (1997). SB 242084, a selective and brain penetrant 5-HT2C receptor antagonist.Neuropharmacology, 36, 609-620.Kidd, E.J., Garratt, J.C. and Marsden, C.A. (1991). Effects of repeated treatment with 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane (DOI) on the autoregulatory control of dorsalraphe 5-HT neuronal firing and cortical 5-HT release. Eur J Pharmacol, 200, 131-139.Lakoski, J.M. and Aghajanian, G.K. (1985). Effects of ketanserin on neuronal responses toserotonin in the prefrontal cortex, lateral geniculate and dorsal raphe nucleus.Neuropharmacology, 24, 265-273.Maes, M., Libbrecht, I., van Hunsel, F., Campens, D. and Meltzer, H.Y. (1999). Pindolol andmianserin augment the antidepressant activity of fluoxetine in hospitalized major depressedpatients, including those with therapy resistance. J Clin Psychopharmacol, 19, 177-182.McAskill, R., Mir, S. and Taylor, D. (1998). Pindolol augmentation of antidepressant therapy.Br J Psychiatry, 173, 203-208.Mendoza D.L., Bravo H.A. and Swanson H.H. (1999). Antiaggressive and anxiolytic effects ofgepirone in mice, and their attenuation by WAY 100635. Pharmacol., Biochem. AndBehavior, 62(3); 499-509.Mengod G., Nguyen H., Le H., et al. (1990). The distribution and cellular localisation of theserotonin 1C receptor mRNA in the rodent brain examined by in situ hybridizationhistochemistry. Comparison with receptor binding distribution. Neuroscience 35,577-591.Moret, C. and Briley, M. (1997). 5-HT autoreceptors in the regulation of 5-HT release fromguinea pig raphe nucleus and hypothalamus. Neuropharmacology, 36, 1713-1723.Moser P.C., Sanger D.J., (1999). 5-HT1A receptor antagonists neither potentiate nor inhibitthe effects of fluoxetine and befloxatone in the forced swom test in rats. Eur. J. Pharm. 14;372(2): 127-34.Olivier B., Molewijk H.E., van der Heyden J.A.M., van Oorschot R., Ronken E., Mos J. andMiczek K.A. 1998. Ultrasonic vocalisation in rat pups: effects of serotonergic ligands.Neuroscience and biobehavioral reviews 23; 215-227.O’Neill M.F., Fernandez A.G. and Palacios J.M. (1995). GR 127935 blocks locomotor andantidepressant-like effects of RU 24969 and the action of antidepressants in the mouse tailsuspension test . Parmacol., Biochem. And behavior 53; 535-539.Papp M., Nalepa I., Antkiewicz-Michaluk L. and Sanchez C. (2002). Beavioural andbiochemical studies of citalopram and WAY 100635 in rat chronic mild stress model.Pharmacol., Biochem. And behavior 72; 465-474.


199

Ramboz, S., Saudou, F., Amara, D.A., Belzung, C., Segu, L., Misslin, R., Buhot, M.C. andHen, R. (1996). 5-HT1B receptor knock-out-behavioural consequences. Behav Brain Res,73, 305-312.Schlicker, E., Werner, U., Hamon, M., Gozlan, H., Nickel, B., Szelenyi, I. and Gothert, M.(1992). Anpirtoline, a novel, highly potent 5-HT1B receptor agonist withantinociceptive/antidepressant-like actions in rodents. Br J Pharmacol, 105, 732-738.Schoeffter, P. and Hoyer, D. (1989). How selective is GR 43175? Interactions with functional5-HT1A, 5-HT1B, 5-HT1C and 5-HT1C receptors. Naunyn Schmiedebergs Arch Pharmacol,340, 135-138.Scott, D.O. and Heath, T.G. (1998). Investigation of the CNS penetration of a potent 5-HT2Areceptor antagonist (MDL 100,907) and an active metabolite (MDL 105,725) using in vivomicrodialysis sampling in the rat. Journal of Pharmaceutical and biomedical analysis, 17, 17-25.Sharp, T., Bramwell, S.R., Hjorth, S. and Grahame Smith, D.G. (1989). Pharmacologicalcharacterization of 8-OH-DPAT-induced inhibition of rat hippocampal 5-HT release in vivo asmeasured by microdialysis. Br J Pharmacol, 98, 989-997.Stanford, I.M. and Lacey, M.G. (1996). Differential actions of serotonin, mediated by 5-HT2C receptors, on GABA-mediated synaptic input to rat substantia nigra pars reticulataneurons in vitro. J Neurosci, 16, 7566-7573.Tao, R., Ma, Z. and Auerbach, S.B. (1996). Differential regulation of 5-hydroxytryptaminerelease by GABAA and GABAB receptors in midbrain raphe nuclei and forebrain of rats. Br JPharmacol, 119, 1375-1384.Wilkinson, L.O. and Dourish, C.T. (1991). Serotonin and animal behaviour. In Anonymous,Serotonin receptor subtypes: basic and clinical aspects.Wright, I.K., Garratt, J.C. and Marsden, C.A. (1990). Effects of a selective 5-HT2 agonist,DOI, on 5-HT neuronal firing in the dorsal raphe nucleus and 5-HT release and metabolism inthe frontal cortex. Br J Pharmacol, 99, 221-222.Zhang, W., Perry, K.W., Wong, D.T., Potts, B.D., Bao, J., Tollefson, G.D. and Bymaster,F.P. (2000). Synergistic effects of olanzepine and other antipsychotic agents in combinationwith fluoxetine on norepinephrine and dopamine release in rat prefrontal cortex.Neuropsychopharmacology, 23, 250-262.

Chapter 9

200

Summary and conclusions

201

Chapter 10

Summary and Conclusions

Monoamines have been the primary target in depression research for almost half a century.The role these compounds play in the pathofysiology of depression is far from clear, butseveral forms of pharmacological treatments have evolved that modulate the function ofmonoaminergic systems in the brain. The first antidepressants, monoamine oxidase inhibitors(e.g. iproniazid) and tricyclic antidepressants (e.g. imipramine) had many unwanted sideeffects and were not safe in overdose. In the last decades the latter have markedly improved,but issues such as treatment resistance and onset of action still need attention.Most modern antidepressants are based on the tricyclics. These compounds enhancemonoamine levels in the brain by inhibiting the re-uptake of released monoamines into theneurones. Next to re-uptake inhibition tricyclic antidepressants have considerable affinitiesfor several neurotransmitter receptors. This may explain why tricyclics, although veryeffective antidepressants, induce so many unwanted side-effects. Since long term treatment indepression is important to prevent relapse, it is obvious that these compounds, thougheffective in early treatment, become less efficient due to treatment discontinuation of thepatient. In addition, they are toxic in overdose. Accordingly, the strategy of thepharmaceutical industry has been to isolate their monoamine reuptake inhibitory properties,thereby reducing cholinergic, adrenergic and histaminergic side effects. This has resulted inmore or less selective reuptake inhibitors for serotonin (e.g. paroxetine), noradrenaline (e.g.reboxetine) and dopamine (e.g. buproprion). These modern antidepressants have significantlyless side effects, but antidepressant activity may be somewhat less compared with tricyclics,such as imipramine. To improve antidepressant activity several strategies have been usedsuch as combining serotonin and noradrenaline reuptake inhibition in one molecule (e.g.

Chapter 10

202

venlafaxine) or serotonin receptor antagonists that also block the reuptake of serotonin (e.g.nefazodone). A different approach was used for the heterocyclic compound mirtazepine.This very successful antidepressant has been proposed to enhance monoamineneurotransmission through blockade of adrenergic auto- and heteroceptors on thenoradrenergic and serotonergic systems in the brain. Mirtazepine is a potent antidepressantwith positive short- and long-term treatment prognoses as well as a good safety profile.Several other putative antidepressant drugs are currently being investigated for their use inthe clinic (e.g. substance P antagonists, corticosteroid (releasing factor) antagonists and theherbal preparation St. John Worth), but these fall beyond the scope of this thesis.By far the most widely used antidepressants are the selective serotonin inhibitors (SSRIs).They primarily act on the serotonergic system and are relatively safe in their therapeutic use.Much research has been dedicated to their mechanism of action, which may eventually leadto new stategies that are effective in treatment resistant patients. A confounding factor ofantidepressant treatment is the delay in therapeutic response, which depends on the natureand severity of the illness, and may last from 2 to 6 weeks. During this time patientsexperience side effects while symptoms of depression do not improve, which has a rathernegative effect on the patient’s motivation to continue antidepressant therapy.

The present thesis can be divided into three parts. In the first part (chapter 3 to 6), themechanism of action of the SSRI citalopram is investigated for the acute administration aswell as after 2 weeks of treatment with the SSRI. In the second part (chapter 7 and 8), the“pindolol augmentation strategy” is critically evaluated with respect to pharmacokinetics andpharmacodynamics, and the pharmacological interaction between benzodiazepines and SSRIsis discussed. In the third part (chapter 9), a new concept to augment SSRI induced increasesin extracellular 5-HT levels is described. In this concept, the inhibition of the 5-HT2C receptorsubtype plays an important role.Chapter 3 describes the effects of the SSRI citalopram on 5-HT levels in the ventralhippocampus of freely rats after acute administration. In the serotonergic system there are atleast two forms of auto-regulation. Presynaptic 5-HT1A receptors are involved in a feedbackmechanism that is localized on cellbodies, whereas a post-synaptic mechanism is acting vianeuronal feedback loops (chapter 6). Upon activation, both types of receptors reduce thefiring rate of the serotonergic neurones, thereby decreasing serotonergic neurotransmission inthe brain. In certain brain areas 5-HT1B receptors are located on the terminals of serotonergicneurons. These receptors inhibit 5-HT release when activated.The mechanism of action of SSRIs has been hypothesised to be related to a gradualdesensitisation of serotonergic autoreceptors. Therefore we have investigated the influence ofautoreceptor blockade on citalopram induced increases in extracellular 5-HT levels. Severaldoses of citalopram were used incombination with a selective 5-HT1A receptor antagonist or aselective 5-HT1B receptor antagonist. Citalopram dose-dependently increased 5-HT levels inventral hippocampus. The effect reached a plateau value at plasma levels in the rangeobserved in depressed patients treated with Cipramil. Autoreceptor control varieddepending on the type of autoreceptor as well as the dose of citalopram used. Whereas the 5-


203

HT1A autoreceptor control was only observed at plasma levels in the clinical range, 5-HT1B

autoreceptor control was present at any dose of the SSRI.In Chapter 4 the function of the autoreceptors was measured following chronic treatment. Asrodents eliminate exogenous substances very rapidly, chronic treatment would demandmultiple daily injections to establish steady state conditions. This is very uncomfortable forthe animals. Instead we have used osmotic minipumps, which resulted in stable plasma levelswithin the clinical range. After 2 weeks of chronic treatment with citalopram the effect ofsystemic administration of the 5-HT1A receptor agonist 8-OH-DPAT was attenuated, whichsuggests desensitisation of these receptors. In contrast a challenge with a 5-HT1B receptoragonist failed to show significant changes. A challenge with citalopram had a similar effecton extracellular 5-HT in chronically citalopram treated compared to saline treated animals,which indicates that 5-HT1B autoreceptors posses sufficient restraining properties tocompensate for the loss of 5-HT1A receptor function.There is increasing evidence that control by autoreceptors varies between brain areas. InChapter 5 we have investigated the function of the autoreceptors in different brain areas,using protocols similar to those employed in chapter 3 and 4. In the cortex-dorsal raphepathway (mPFC-DRN), it was observed that in the cortical serotonin release was under theinfluence of 5-HT1A receptors at subclinical as well as clinical plasma levels of citalopram.However, since 5-HT1A receptor induced feedback in the DRN was only observed at clinicalplasma levels of the SSRI, restrainment of cortical 5-HT at lower plasma levels may bederived from long-feedback loop mechanisms or median raphe projections. No 5-HT1B

autoreceptor functionality was observed in this brain area. The dorsal hippocampal-medianraphe (DH-MRN) pathway was observed to be under influence of 5-HT1A receptorsdepending on the dose of the SSRI. Whereas at intermediate doses (subclinical plasma levels)DH as well as MRN were under the influence of 5-HT1A receptors, this effect was surpassedby a 5-HT1B receptor mediated event at higher doses (clinical plasma levels). This may hint atthe presence of 5-HT1B receptor mediated long loop type of feedback, which is activated athigher citalopram doses.Similar experiments were performed after chronic treatment with citalopram to investigatewhether this had an effect on autoreceptor control in the two neuronal systems. Animals werechallenged with a high dose of citalopram after being treated for 2 weeks with the SSRI.Basal 5-HT levels in the PFC, as well as in MRN were significantly enhanced after chronictreatment, indicating enhanced tonic activation of these areas after treatment. With respect tothe enhanced levels in the PFC it can be argued that these changes in basal levels only takeplace in areas devoid of 5-HT1B autoreceptors, an observation which may be relevant for thetreatment of depression. Although the relative effect of a citalopram challenge after chronictreatement with citalopram remained similar in PFC, DH and DRN, an attenuated effect wasobserved in the MRN, further indicating the changed tonicity of this system. It is concludedfrom the present chapter that a delicate balance exists with respect to autoreceptor functionswithin the brain. It is likely that the relative autoreceptor distribution plays a pivotal role inthe chronic effects of antidepressants in various brain areas.In Chapter 6, the function of a 5-HT1A receptor mediated long loop type of feedback fromthe central nucleus of the amygdala back to the caudal linear raphe was investigated. Special

Chapter 10

204

attention was paid to functional changes following chronic treatment with citalopram. Afteractivation of the feedback loop by local application of 5-HT1A agonist flesinoxan in theamygdala and by observing 5-HT levels in the amygdala, the activity of the long-feedbackloop was evaluated after 2 weeks of treatment with citalopram. In line with thedesensitisation of 5-HT1A receptors in raphe nuclei, it was observed that the post-synaptic 5-HT1A receptors desensitise upon chronic treatment, thereby disrupting inhibitory long loopfeedback mechanisms, enabling enhanced serotonergic neurotransmission in the amygdala.From these data it was concluded that enhanced serotonergic neurotransmission in theamygdala could have its contribution to the antidepressive effect of SSRIs.As the mechanism of action of SSRIs is thought to invlove the gradual desensitisation ofserotonergic autoreceptors, theoretically the combined administration of an SSRI with a 5-HT1A antagonist would enhance the therapeutic efficacy as well as accalerate the onset ofaction of the antidepressant effect. As no selective and potent 5-HT1A antagonist was presentfor clinical evaluation, investigators have used the mixed beta/5-HT1A antagonist pindolol forthe proof of this principle. Although pindolol is a moderately potent antagonist of 5-HT1A

receptors, it was calculated that at the plasma levels used in the clinic, this compound wouldnever induce sufficient antagonism to exert its desired effects. In the Chapter 7 weinvestigated the capability of different plasma levels of pindolol to augment paroxetineinduced increase in 5-HT extracellular levels in the brain of guinea pigs. Clinical plasmalevels of pindolol were incapable of augmenting paroxetine induced increases in 5-HT levels.Only at plasma levels 100 fold higher than those observed in the clinic, augmentation of 5-HT levels was observed. These data indicate that the beneficial effects observed during co-administration of pindolol with an SSRI in depressed patients, might be derived from betareceptor antagonism rather than 5-HT1A antagonism.Concerning antidepressant treatment, few attention is paid on the abundant use poly-pharmacy in psychiatric patients. Although antagonism of 5-HT1A receptors duringadministration of SSRIs might induce augmented levels of 5-HT in the brain, this effectmight be influenced by co-administration of other compounds. As benzodiazepines arefrequently used during antidepressant treatment we investigated the effect of the co-administration of SSRIs with benzodiazepines in Chapter 8. The increase of serotonin levelsby paroxetine injection was observed to be attenuated by pre-treatment of the animal withoxazepam or temazepam. As clinical data from co-administration of SSRIs withbenzodiazepine indicate enhanced rather than decreased clinical effects, the present studyquestions the relevance of enhanced 5-HT levels in the treatment of depression.In Chapter 9 we describe the augmention which is observed when non-selective 5-HT2

antagonstic drugs are co-administered with SSRIs. It was found that 5-HT2C antagonism wasresponsible for the enhancement of central 5-HT release. By local administration ofketanserin and/or citalopram in terminal as well as cell body regions, the topical organisationof the mechanism of augmentation was evaluated. Antagonism of 5-HT2C receptors in bothterminal and cell bodies was found to augment the effects of citalopram on 5-HT levels.Given the capability for augmentation as well as the anxiolytic, sleep-enhancing and side-effect profile of the combination of an SSRI with 5-HT2C antagonism, it is expected that thiscombination will act superior to SSRI treatment alone. In addition, since preclinical results of


205

beneficial effects 5-HT1A and 5-HT1B receptor antagonists on sleep and anxiety are not verypromising, the current concept is expected to be beneficial over augmentation by 5-HT1

antagonists.

In conclusion, it is shown in this thesis that 5-HT1A and 5-HT1B autoreceptor control duringchronic treatment with SSRIs is different in various brain structures. Delicate changes inautoreceptor functionality which are brought upon by chronic treatment with SSRIs might bemimicked by co-administration of different serotonin receptor antagonists (5-HT1A, 5-HT1B)with different doses of SSRIs.Augmentation induced by co-administration of a 5-HT2C antagonist with a SSRI is expectedto become a powerfull approach in the treatment of depression as well as in anxiety.The concepts that are described and evaluated in the current thesis need clinical evaluationusing proper compounds. To prevent misinterpretation of the generated results, these trialsshould be performed while pharmacokinetics are closely monitored. Additionally, clinicalprotocols should exclude possible pharmacological interfering compounds likebenzodiazepines.

Chapter 10

206

Samenvatting voor de leek

207


Depressie

Iedereen kent wel iemand die depressief is dan wel antidepressiva gebruikt. Depressiviteit enantidepressiva zijn termen die steeds vaker opduiken in de samenleving en dat zal welcorreleren met de hectiek en stress die onze huidige samenleving met zich meebrengt. Hetwordt zelfs verwacht dat in het komende decennium depressie zal uitgroeien tot de ziekte diede belangrijkste oorzaak is dat mensen niet meer een normaal arbeidzaam leven kunnenvolgen.

Alhoewel depressie al in de Griekse oudheid beschreven werd, is in de laatste decennia veelwerk gedaan aan het classificeren van deze ziekte. Het blijkt namelijk dat deze ziekte niet eenduidelijk aanwijsbaar ziektebeeld is zoals bijvoorbeeld de ziekte van Parkinson. Depressie isnamelijk een samenraapsel van allerlei vormen van gevoelsafwijkingen. Waar vroeger al dezeafwijkingen eenvoudig benoemd werden met dezelfde naam, is er in de laatste 30 jaar eenonderverdeling gemaakt die hedentendage door veel psychiaters, artsen en psychologengebruikt wordt. Deze classificering is onder meer beschreven in het “Diagnostic andStatistical Manual” (DSM). Aan de hand van dit handboek kan de behandelaar de patientdiagnostiseren. Dit gebeurd met behulp van een lijst met symptomen die betrekking hebbenop bijvoorbeeld somberheid, slaapkwaliteit en zelfmoordneigingen. Eén en ander heeft ertoegeleidt dat tegenwoordig depressie uiteenvalt in subvormen op basis van bijvoorbeeld het aldan niet aanwezig zijn van schizofrene eigenschappen, zuiver depressieve eigenschappendanwel afwisselend aanwezig zijn van deze eigenschappen. Tevens wordt er rekening meegehouden dat de depressie bijvoorbeeld door het gebruik van drugs en of medicijnenveroorzaakt kan zijn. Verder wordt nu ook gelet op het aanwezig zijn van andere symptomen,waarvan men weet dat ze dikwijls aanwezig zijn bij depressie, zoals bijvoorbeeld angst.

Serotonine en antidepressiva

Sinds ongeveer een halve eeuw wordt er veel onderzoek gedaan om de onderliggendeafwijkingen in kaart te brengen die verantwoordelijk zijn voor depressie en daaraangerelateerde symptomen. Al lang geleden werd ontdekt dat een stof die ontwikkeld was tegentuberculose, goed werkte tegen depressie (isoniazide). Men ontdekte dat deze stof de afbraakvan boodschappers van de hersenen (de zogenaamde monoaminerge neurotransmitters alsnoradrenaline en serotonine) remde, waardoor deze neurotransmitters verhoogd aanwezigwaren in de hersenen. Deze bevinding heeft ertoe geleid dat men concludeerde dat kennelijkeen afwijking in deze neurotransmitters de basis was van depressie. Spoedig werden anderestoffen tegen depressie ontdekt. Deze stoffen remde niet de afbraak, maar gingen de opnamevan deze stoffen in neuronen tegen, waardoor ook de aanwezigheid van neurotransmittersverhoogd kon worden. Alhoewel deze zogenaamde “tricyclische antidepressiva” effectiefwaren tegen depressie hadden deze middelen sterke bijwerkingen. Daarom werden er in de

208

daaropvolgende jaren “schonere” middelen ontwikkeld die alleen nog maar de heropnameremden. Deze selectieve serotonine heropname remmers (SSRIs) waarvan Prozac® zeerbekend is, zijn tegenwoordig de eerste keus in de behandeling van depressie.

Hoewel, zoals eerder beschreven, de behandeling van depressie vooruitgang heeft geboekt inde vorm van de ontwikkeling van middelen die minder bijwerkingen hebben en daardoorveiliger zijn, kleven er nog steeds nadelen aan deze stoffen. Het blijkt namelijk dat dezestoffen in slechts 60 to 80 % van de mensen die ermee behandeld worden daadwerkelijk eengoed antidepressief effect hebben. Tevens duurt het ook nog een aantal weken voordat dezemiddelen beginnen te werken.Om deze onvolkomenheden weg te werken wordt er tegenwoordig veel onderzoek gedaannaar het verbeteren van de effectiviteit van SSRIs. In dit proefschrift worden een aantalstudies beschreven die een aantal mogelijkheden evalueren om het antidepressief effect vanSSRIs te versterken.

De werking van antidepressiva

Om de werking van antidepressiva te kunnen begrijpen is een stukje onderliggende kennisvan de hersenen nodig. Hersenen bestaan uit een groot aantal neuronen. Deze neuronen zijncellen die bestaan uit een cellichaam (soma) en uitlopers (axonen). Boodschappen wordendoorgegeven door een electrisch stroompje dat in het cellichaam begint, waarna deelectriciteit via de uitlopers reist naar het gebied waar de boodschap naar toe moet. Daaraangekomen geeft het neuron zijn booschap door aan een volgend neuron door middel vanafgifte van neurotransmitters zoals de bovenbeschreven serotonine (5-HT) en noradrenaline(NE). Het neuron dat de boodschap ontvangt heeft “voelers” aan de buitenkant (receptoren)die de uitgestoten serotonine en noradrenaline opvangen, waardoor de boodschap isdoorgegeven.Het functioneren van de afgifte van serotonine en andere neurotransmitters is niet alleenafhankelijk van het doorgeven van activiteit. Op het neuron zelf zijn namelijk een aantalvoel-systemen aanwezig die de totale afgifte van neurotransmitter binnen de perken houden.In het cellichaam gebied is het de 5-HT1A receptor die, wanneer hij voelt dat er veelserotonine is, terugkoppelt naar het neuron en zo de activiteit beheerst. Tevens bevindt zichaan het einde van het axon een receptor met eenzelfde functie nl. de 5-HT1B receptor. Alsdeze receptor teveel serotonine voelt dan houdt deze rechtstreeks de afgifte ervan tegen.Zoals boven al aangegeven is de geldende hypothese dat ten tijde van depressie de functievan het serotonine systeem verminderd is. Alhoewel SSRIs serotonine verhogen, blijkt dat deterugregulerende mechanismen (5-HT1A en 5-HT1B receptoren) in het begin zo sterkterugkoppelen, dat er weinig van het serotonine verhogend effect merkbaar is. Na verloopvan tijd vermindert deze terugkoppeling echter, waardoor de SSRIs hun volledigewerkzaamheid kunnen uitoefenen. Deze achtereenvolgende gebeurtenissen zoudenverantwoordelijke zijn voor het feit dat het enige weken duurt voor SSRIs beginnen tewerken.


209

Studies in dit proefschrift

Als deze gebeurtenissen inderdaad verantwoordelijk zijn voor de werkzaamheid van SSRIs,zou men door het remmen van de werkzaamheid van de terugregulerende mechanismen eenantidepressivief effect kunnen krijgen.In hoofdstuk 3 tot en met 7 werd gekeken hoe de bovenstaande theorie werkt in proefdieren.In hoofdstuk 3 werd onderzocht hoe het geven van steeds hogere doseringen van de SSRIcitalopram (Cipramil®) de serotonine niveaus uit de axonen verhoogd in een gedeelte van dehersenen dat belangrijke veranderingen laat zien tijdens depressie (de ventrale hippocampus).Er kon een duidelijk effect worden gezien van de zelfregulerende mechanismen van hetserotonine systeem. Bij hogere doseringen van citalopram werden geen verdere stijgingenvan serotonine waargenomen. Opmerkelijk is dat op het punt waar deze stijgingen maximaalwerden, de bloedconcentraties van citalopram in de rat overeenkwamen met de bloedwaardendie in de mens gezien worden. Als de zelfregulerende mechanismen geremd werden, dan konduidelijk worden waargenomen dat 5-HT1B receptoren actief werden bij elke dosering van deSSRI. De 5-HT1A receptoren werden alleen actief bij de hogere doseringen.Nadat we in hoofdstuk 3 lieten zien dat de activatie van de zelfregulerende mechanismensterk afhankelijk was van de dosering van citalopram, werden in hoofdstuk 4 dieren tweeweken behandeld met citalopram in dezelfde bloedconcentraties die bij patienten in de kliniekgebruikt worden. Er kon na twee weken duidelijk een afzwakking van de 5-HT1A

functionaliteit waargenomen worden, terwijl dit voor 5-HT1B receptoren niet werd gezien. Intotaal kon geen versterkt effect gezien worden wanneer de SSRI gegeven werd na dechronische behandeling. Aangezien door de langdurige toediening de zelfregulerendemechanismen deels afgezwakt waren zou dit wel te verwachten zijn.Nu er gezien was dat de zelfregulatie van de serotonine vrijmaking in de hersenen sterkafhankelijk kon zijn van de dosering van de SSRI, was het interessant om te kijken hoe dezeprocessen in andere hersengebieden plaatsvonden. In hoofdstuk 5 is gekeken hoe in deprefrontale cortex (PFC) en dorsale hippocampus (DH) deze mechanismen werken. In dezegebieden komen de axonen uit waarvan de cellichamen liggen in respectivelijk de dorsale enmediane raphe. Het bleek dat de PFC sterk onder controle staat van de 5-HT1A receptor. DeDH daarentegen staat onder controle van 5-HT1B receptor en alleen bij lage concentraties vande SSRI ook onder controle van de 5-HT1A receptor. Na chronische behandeling van dedieren met een SSRI was het duidelijk dat alleen de gebieden die hoofdzakelijk onder invloedvan de 5-HT1A receptoren staan, een totaal verhoogd effect van de SSRI lieten zien. Dezeobservaties benadrukken dus het belang van de regio-specifieke effecten van SSRIs in dehersenen. Tevens is het ook gezien de data van hoofdstuk 4 duidelijk dat overal waar een 5-HT1B receptor meespeelt, geen verandering van de effecten van een SSRI kan waargenomenworden.Alhoewel een groot deel van de zelfregulatie van neuronen afkomstig is van receptoren dieop het neuron zelf zitten, blijkt het ook mogelijk dat de afgifte van een neurotransmitter eenander neuron activeert, die dan op zijn beurt de boodschap via een axon naar cellichaam vanhet eerste neuron stuurt, om zo zijn activiteit te remmen. Deze manier van terugkoppelingwordt “long feedback loop” genoemd. In hoofdstuk 6 is gekeken wat het effect van

210

chronische behandeing met SSRIs is op dit mechanisme. We konden een duidelijkeafzwakking van deze terugkoppelig zien, die zich reflecteerde in een verhoogd effect van deSSRI.Om de hypothese te testen dat het remmen van de 5-HT1A receptor tijdens de behandelingmet een SSRI een versnelde en verbeterde activiteit van de antidepressiva zou geven, hebbenpsychiaters in de kliniek de stof pindolol gebruikt. Aangezien er geen selectieve en potenteblokkers beschikbaar waren moesten deze studies gedaan worden met deze stof, die naast 5-HT1A blokkade ook beta noradrenerge receptoren blokkeert. Alhoewel de meeste van dezestudies positief zijn over het klinisch effect van pindolol, lijkt de dosering die gebruikt werdin de kliniek veel te laag te zijn om enig effect te kunnen geven. In hoofdstuk 7 isgeëvalueerd of de beoogde effecten van pindolol (versterking van het effect van de SSRI opserotonine niveau’s) daadwerkelijk optreden bij de bloedconcentraties die in de kliniekoptreden. Er kon duidelijk gezien worden dat deze effecten pas optraden bij 100 voudighogere concentraties van pindolol. Het is daarom onwaarschijnlijk dat de gunstige effectenvan pindolol bij depressie door zijn 5-HT1A remming komen. De algenoemde remming vanbeta noradrenerge receptoren, zou een verklaring kunnen zijn voor de succesvolle resultaten.Patienten met depressie worden vaak behandeld met meerdere geneesmiddelen tegelijkertijd.De eerder genoemde verhoogde aanwezigheid van angst en slecht slapen bij depressie is dereden dat veel van deze patienten naast SSRIs ook benzodiazepines zoals bijvoorbeeldValium® krijgen. In hoofdstuk 8 is gekeken naar het effect van benzodiazepines op deverhoging van serotonine niveau’s door een SSRI. Het blijkt dat benzodiazepines het effectvan SSRIs op serotonine tegengaan. Aangezien de werkzaamheid van SSRIs in de kliniekwordt toegeschreven aan de het verhogen van serotonine niveau’s, zou dit moeten betekenendat deze gelijktijdige toediening de antidepressieve activiteit van SSRIs zou verlagen. Nublijkt uit klinische studies dat dit niet het geval is, maar dat deze stoffen zelfs de werking inde kliniek versterken. Mogelijk versterken benzodiazepines het antidepressief effect vanSSRIs via andere mechanismen, zoals verbetering van slaap en vermindering van angst.In hoofdstuk 9 wordt ten slotte een nieuwe manier beschreven om het effect van SSRIs op deserotonine niveau’s te verhogen. Door een SSRI gelijktijdig toe te dienen met een 5-HT2C

antagonist werd een versterkt effect gezien op serotonine niveau’s. Uit de literatuur blijkt dat5-HT2C receptoren, in tegenstelling tot 5-HT1A en 5-HT1B receptoren, geen serotonergeautoreceptoren zijn. Het mechanisme dat de versterking van het effect veroorzaakt, werktkennelijk anders dan bij 5-HT1A en 5-HT1B blokkers. Door een aantal experimenten uit tevoeren, kon geconcludeerd worden dat de versterkende werking van 5-HT2C blokkers berustop het afzwakken van remmende signalen, die elders vanuit de hersenen komt.Kennelijk verhoogd het gezamelijk geven van SSRIs met 5-HT2C remmers de effectiviteit opserotonine niveau’s. Tevens blijken 5-HT2C remmers zelf ook gunstige effecten tegen angst,slecht slapen en aggressie te hebben. In hoofdstuk 8 werd al gezien dat stoffen die, alhoewelze het effect van SSRIs op serotonine tegengaan, toch een verbeterd antidepressief effectgeven omdat ze mogelijk slaap verbeteren en tegen angst werken. 5-HT1A en 5-HT1B blokkershebben deze eigenschappen niet. Daarom is te verwachten dat het combineren van 5-HT2C

antagonisten met SSRIs niet alleen beter werkt dan SSRI behandeling alleen, maar ook beterdan SSRI behandeling tegelijkertijd met 5-HT1A en 5-HT1B blokkers.

Dankwoord

211

Dankwoord

Zoals gebruikelijk was ook deze promotie een project met hoogte- en dieptepunten en wintde aanhouder. Aan het einde van de rit wil ik natuurlijk een aantal mensen bedanken. Zonderde hulp en steun van deze mensen was het waarschijnlijk niet mogelijk geweest om ditproefschrift in deze vorm voor elkaar te krijgen.

In min of meer chronologische volgorde;Alleerst wil ik natuurlijk Hakan Wikström bedanken. Hakan was vanaf het begin mijnpromotor en heeft alle escapades rondom dit proefschrift meegemaakt. De stijl van Hakan’sbegeleiding deed af en toe uit een scene van “karakter” van Bordewijk denken (in positievezin). Achteraf ben ik daar zeer dankbaar voor, want zo leer je om eigen projecten te trekken.Tevens ben ik je zeer erkentelijk voor alle kansen en tevens alle vrijheid die je me gegevenhebt.Peter de Boer was mijn dagelijkse begeleider toen ik begon met mijn promotie. Dankzij Peterkon ik op een “reidende trein” springen. Helaas heeft Peter de Universiteit verlaten, maar hijis goed terecht gekomen. Ik mis echter de discussies en je wetenschappelijke diepzinnigheidnog steeds.Fokko Bosker begon in Groningen ongeveer het moment dat Peter de universiteit verliet. Hetklikte meteen. Avonden hebben we in restaurants en cafés over de wetenschap gepraat. Hetfeit dat hoofdstuk 6 met jou naam als eerste auteur gepubliceerd is, zegt veel over de hogemate van wetenschappelijke interactie. Ook hadden we grootse plannen voor de toekomstover hoe iedereen binnen de wetenschap moest samenwerken om een geoliede machine tekrijgen. We zijn er nog een aardig eind mee gekomen ook. Fokko, ik ben je enorm dankbaarvoor alle steun, kennis, enthousiasme en vriendschap die je me in de afgelopen jaren gegevenhebt.Het overgrote deel van het werk in dit proefschrift is uitgevoerd in het lab van de “godfather”van microdialyse Ben Westerink. Alhoewel in het begin van mijn promotie de samenwerkingnog niet zo aanwezig was, is dat de laatste jaren sterk gegroeid. Ik ben je zeer dankbaar voorde hoge mate van vrijheid en vertrouwen die je mij gegeven hebt. Tevens hebben weinmiddels samen “Brains-on-Line” opgezet. Ik ben je enorm dankbaar dat je mee bent gegaanin deze uitdaging.Natuurlijk was al het werk niet mogelijk zonder de mensen op het lab. In eerste instantie wilik natuurlijk Jan de Vries bedanken. Tevens ben ik alle andere analisten en biotechnici op hetlab veel dank verschuldigd (Christa, Karola en Alle). Ik heb ook veel genoten en geleerd vande samenwerking met de andere AIO’s en wetenschappers op het lab (Minke, Marieke, Yuki,Olga, Wia, Eitan, Weite). De samenwerking met alle doctoraal studenten was altijd zeervruchtbaar, alsmede zeer gezellig (bedankt: Martin, Tom, Milly, Robert, Loes, Gunnar,Hylke-Jan, Jasper en Roland). Natuurlijk kan ik ook Edje niet vergeten. Zeer therapeutischwas het om al vissend of Boot-schurend de gang van zaken omtrend het lab evalueren.

Dankwoord

212

Verder is natuurlijk de gezelligheid en steun van de “orginele vakgroep medicinal chemistry”niet te verwaarlozen. Erik, Pieter, Ulrik, Evert, Jonas, Durk, Yi, Marquerite, Nynke, Margrieten Janneke en natuurlijk Cor, bedankt.Ongetwijfeld vergeet ik nog een heel aantal mensen. In een poging om volledig te zijn, kan ikde volgende mensen niet vergeten aangezien. Hun deur stond altijd open, zodat ik weerbinnen kon vallen met praktische en theoretische vragen. De Vakgroep klinischefarmacologie (Egbert), farmacokinetiek (Jan Visser), Moleculaire farmacologie (prof.Zaagsma en Ad), biomonitoring and Sensoring (Bert, Irene, Klaas-Jan en Rene). Biologie(Jan en Tibor).Of course, I should not forget to thank my colleagues at Lundbeck. In the years that we havebeen collaborating I have always admired the high degree of scientific approach that ischaracteristic for this company. In addition, I have always enjoyed coming up to Denmarkboth scientifically as well as socially. Arne, Sandy, Jørn, Klaus-Peter, Peter, Jesper, thankyou.Nu ik toch de mensen in het buitenland aan het bedanken ben, kan ik natuurlijk HansRollema niet vergeten. Ik bewonder hoe je om 6 uur ‘s ochtends al op je werk kan zijn en danook nog de scherpheid van geest te hebben om repliek te geven op allerlei manuscripten enrare histamine proeven. Hans, enorm bedankt.At this point I would like to thank the reading committee for spending their valuable timestruggling through this manuscript (Professor Korf, Professor Zaagsma and Professor Hjorth).Especially Prof. Hjorth, who is an expert in the serotonin field, spent a lot of time checkingthis manuscript and providing relevant comments, Stephan, Many thanks.

Als laatste wil ik natuurlijk mijn familie bedanken voor al hun steun. Vooral mijn oudershebben altijd onvoorwaardelijk achter me gestaan en steun gegeven, en waar mogelijk goederaad gegeven. Pap, mam, bedankt.Dit proefschrift was niet tot stand gekomen zonder mijn kersverse vrouw Marta. Alhoewel zezelf een tijd onderzoek heeft gedaan, heeft ze onmiddellijk afstand genomen van dit rarewereldje (slimme meid). Ondanks alles heeft ze me altijd gesteund; nachtelijk woelen enmalen, hoogte- en dieptepunten, ze stond altijd voor me klaar. Kortom; Mart bedankt.

Documents

University of Groningen SSRI augmentation strategies ...(5 centuries B.C.), these terms do not match the psychiatric classifications which are used nowadays. In the Diagnostic and