SEARCH

SEARCH BY CITATION

Keywords:

  • anxiety;
  • brain microdialysis;
  • chlorimipramine;
  • depression;
  • fluoxetine;
  • Roman high-avoidance and low-avoidance rats

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References

The selective breeding of Roman high- (RHA/Verh) and low-avoidance (RLA/Verh) rats for rapid versus poor acquisition of active avoidant behaviour has produced two behavioural phenotypes with different performances in a variety of animal models of anxiety, in which RLA/Verh rats are consistently more fearful than RHA/Verh rats. In addition, these two lines display different functional properties of brain neurotransmitters like serotonin (5-HT), known to be involved in the expression of anxiety- and depression-related behaviours. Therefore, we used brain microdialysis and [3H]-citalopram binding autoradiography to characterize further the neurochemical properties of 5-HTergic transmission in the two lines. No significant line-related differences were detected in the basal 5-HT output in the frontoparietal cortex (FPCx). In contrast, the increase in the cortical 5-HT output elicited by the systemic administration or the local application, via reverse dialysis, of chlorimipramine and fluoxetine was more robust in RHA/Verh than in RLA/Verh rats. Moreover, the binding signal of [3H]-citalopram to 5-HT re-uptake sites was more intense in the FPCx of RHA/Verh rats than in their RLA/Verh counterparts. These findings suggest that the functional tone of the 5-HTergic projection to the FPCx is stronger in the RHA/Verh line relative to the RLA/Verh line. It is proposed that RLA/Verh rats may be used as a model with heuristic value for studying the role of 5-HTergic transmission in anxiety and in the anxiolytic effects of monoamine re-uptake inhibitors.

Abbreviations used
DR

dorsal raphe nucleus

FPCx

frontoparietal cortex

HPLC

high-performance liquid chromatography

HSD

honest signifcant difference

5-HT

serotonin

MnR

median raphe nucleus

NA

norepinephrine

RHA/Verh

Roman high-avoidance rats

RLA/Verh

Roman low-avoidance rats

SSRI

selective serotonin re-uptake inhibitors

Alterations in central neurotransmitter function are known to play a pivotal role in the pathogenesis of mood disorders and anxiety. Thus, a reduced activity of a variety of central neurotransmitters such as the biogenic amines norepinephrine (NA) and serotonin (5-HT) is frequently observed in depressed patients (Kaplan et al. 1994). In keeping with this finding, antidepressant agents belonging to different chemical categories share the common property of increasing the synaptic concentrations of NA and 5-HT (Blier and de Montigny 1994; Kaplan et al. 1994; Artigas et al. 1996; Invernizzi et al. 1997). Furthermore, there is growing evidence that 5-HT transmission is involved in the control of the responses to stressful stimuli which, in turn, appear to play a causative role in the occurrence of mood disorders (Willner 1991). Accordingly, agents that are able to enhance the extracellular concentration of 5-HT in the brain, like the selective serotonin re-uptake inhibitors (SSRIs), can also decrease anxiety-related behaviours upon repeated administration to rats (Griebel et al. 1994; Zhang et al. 2000) and are clinically effective in the treatment of many of the psychiatric conditions classified as anxiety disorders in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association 1994; see also den Boer et al. 1987; Mancini and Ameringen 1996; Kindler et al. 1997; reviewed by Goddard et al. 1999).

Epidemiologic and genetic data derived from family, adoption and linkage studies strongly indicate that the genetic make-up is a significant factor in the development of mood and anxiety disorders (Crowe 1999; Sanders et al. 1999). The pattern of genetic inheritance, however, is clearly through complex mechanisms and, at this time, no specific associations between genes or gene markers and any of the mood or anxiety disorders have been unambiguously established in humans (Crowe 1999; Sanders et al. 1999). On the other hand, a recent study has characterized the genetic architecture of emotion-related behaviour in rodents by mapping quantitative trait loci that influence in a specific manner the responsiveness to fearful stimuli (Fernández-Teruel et al. 2002). This study, conducted in an F2 cross of inbred Roman high- (RHA/Verh) and low-avoidance rats (RLA/Verh), which differ in fear-related responses (Driscoll et al. 1998), has detected a locus on chromosome 5 that influences behaviour in a battery of experimental models of anxiety (Fernández-Teruel et al. 2002). The Swiss sublines of RHA/Verh and RLA/Verh are psychogenetically selected for, respectively, rapid versus poor acquisition of two-way active avoidance (Driscoll et al. 1998). This selection process has resulted in several behavioural and neurochemical line-related differences, with RLA/Verh rats being emotionally more reactive and displaying passive coping strategies in response to mild stressors as compared to RHA/Verh rats, the latter being normoemotional, or even hypoemotional, and behaving as active copers in response to stress (Corda et al. 1997a; Steimer et al. 1997; Driscoll et al. 1998). Several other differences in 5-HTergic, GABAergic and DAergic function in the CNS have been reported in these two lines (Driscoll et al. 1983; Corda et al. 1997a, 1997b; Giorgi et al. 1997). As regards the 5-HTergic system, RLA/Verh rats appear to have a decreased responsiveness of the cortical 5-HT projection systems to aversive stimuli (Corda et al. 1997b) as well as a reduced density of cerebral 5-HT uptake carrier sites relative to their RHA/Verh counterparts (Charnay et al. 1995). Together, these findings suggest that the RLA/Verh line could be used as a model for studying the neurochemical correlates of anxiety and the mechanisms underlying the anxiolytic effects of 5-HT re-uptake inhibitors. The present study was undertaken to further examine line-related differences in 5-HTergic neurotransmission. To this aim, we used brain microdialysis to investigate the in vivo effects of 5-HT re-uptake inhibitors administered systemically or intracerebrally on 5-HT output in the frontoparietal cortex (FPCx) of RHA/Verh and RLA/Verh rats. In addition, the regional distribution of the 5-HT transporter was examined in the two lines by measuring the binding of the selective ligand [3H]-citalopram to coronal brain slices.

Animals

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References

Outbred male RHA/Verh and RLA/Verh rats were shipped from the Institute of Animal Science (Zürich, Switzerland) to the University of Cagliari, and were kept in the animal house of the Department of Toxicology under controlled temperature and lighting conditions (23 ± 1°C; lights on: 08:00–20:00 h) for at least 2 weeks before beginning the experiments, in order to ensure habituation to the new environment. Food and water were available ad libitum and contact with the animal house maintenance personnel was limited to a single attendant. At the time of the experiments, animals were four to 6 months old and their body weight ranged from 400 to 500 g. All procedures met guidelines and protocols approved by the European Economic Community (EEC Council 86609; D.L. 27.01. 1992, no. 116) and by the Ethical Commission for Animal Care and Use of the University of Cagliari.

Rats were anaesthetized with chloral hydrate [400 mg/kg, intraperitoneally (i.p.)] and placed in a stereotaxic frame with the incisor bar positioned at + 5.0 mm relative to the interaural line. Transversal dialysis probes were constructed using AN69 fibres (Hospal Dasco, Bologna, Italy) and implanted in the FPCx, using the following stereotaxic co-ordinates relative to bregma: AP, + 0.40 mm and V, − 2.40 mm. As shown in Fig. 1(a), the active region of the dialysis membrane was 7 mm in length (3.5 mm in each cerebral hemisphere).

image

Figure 1. (a) Schematic representation of the horizontal dialysis probe implanted in the frontoparietal cortex (FPCx). The dialysing portion of the probes, indicated in white, was located in the FPCx, across the cingulate cortex, area 1 (Cg1), the secondary motor cortex (M2), the primary motor cortex (M1), and the primary somatosensory cortex [forelimb region (S1FL), dysgranular region (S1DZ) and jaw region (S1J)]. AP coordinate: + 0.48 mm relative to bregma (Paxinos and Watson 1998). (b) Basal 5-HT output: Results are expressed as fmol/20 µL sample and are the mean ± SEM of 20 RHA/Verh (▪) and 23 RLA/Verh rats (□).

Download figure to PowerPoint

One day after surgery, brain dialysis experiments were conducted on freely moving rats between 09:00 and 16:00 h, as follows. Ringer's solution (147 mm NaCl, 3.4 mm CaCl2 and 4 mm KCl) was pumped through the dialysis probe at a constant rate of 1 µL/min. Dialysate samples (20 µL) were collected at 20-min intervals and injected without purification into an high-performance liquid chromatography (HPLC) system equipped with a Supelco reverse phase column (LC-18 DB; 150 × 4.5 mm; particle size: 5 µm; Supelco, Bellefonte, PA, USA) and an electrochemical detector (CU 04A2; Antec, Leiden, the Netherlands). The oxidation potential of the cell was set at + 600 mV. The mobile phase [100 mm NaH2PO4/Na2HPO4, 0.1 mm Na2-EDTA, 0.48 mmn-octyl sodium sulfate and 22% methanol (v/v), pH 5.70] was pumped at a flow rate of 1.2 mL/min with an HPLC pump (Model 422, Kontron Instruments, Everett, MA, USA). The sensitivity limit of the assay for 5-HT was 3 fmol/sample (signal to noise ratio ≅ 3).

In all the brain microdialysis experiments, four animals (i.e. two from each line) were studied in parallel. Treatments were started when the basal 5-HT output had become stable (i.e. at least four consecutive samples differed from each other by less than 10%). Chlorimipramine and fluoxetine were dissolved in saline (0.9% NaCI in distilled water) and injected at the dose of 10 mg/kg. Chlorimipramine was given i.p. and fluoxetine was injected subcutaneously (s.c.). In another set of experiments, chlorimipramine and fluoxetine were added to the Ringer's solution and applied locally through the probe by reverse dialysis.

At the end of the experiment, rats were anaesthetized with chloral hydrate (400 mg/kg, i.p.) and perfused transcardially with 50 mL of physiological saline, followed by 100 mL of 10% formalin. Brains were immediately removed and stored at room temperature in 10% formalin for at least 2 days. Subsequently, brains were cut into serial, 100 µm thick, coronal sections with a vibratome in order to determine the location of the probes, using the atlas of Paxinos and Watson (1998) as a reference.

Six rats from each line were killed by decapitation and their brains were rapidly removed and frozen in powdered dry ice for 5 min. The brains were then wrapped in parafilm and kept frozen at −70°C. Serial coronal sections (14 µm thick) were cut from the frozen brains on a Leitz cryostat (working temperature: − 22°C). The sections were thaw-mounted on 3-aminopropyl-triethoxysilane-coated (Sigma, St Louis, MO, USA) glass slides and kept in plastic boxes at −70°C until used. Three sections for specific and one for non-specific signal from different brain levels of each animal (AP, relative to bregma: 1.70, −0.26, − 5.80 and − 7.30 mm) were processed in parallel for receptor autoradiography.

Tissue sections were preincubated for 15 min at room temperature in 50 mm Tris–HCl buffer containing 120 mm NaCl and 5 mm KCl, pH 7.6, to remove any endogenous 5-HT from the tissue. The tissue sections were then incubated for 60 min at room temperature in the same buffer solution, containing 0.7 nm[3H]-citalopram (specific activity: 85 Ci/mmol). The reaction was stopped by dipping the slides in ice cold assay buffer followed by two washes (10 min each) in fresh ice cold assay buffer. Non-specific binding was determined in the presence of 1 µm paroxetine. The slides were then dried under a stream of cold dry air and apposed to autoradiographic film (3H-Hyperfilm, Amersham, Piscataway, NJ, USA) along with radioactivity standard slides (3H-Microscales, Amersham). After exposure at 4°C for 30 days, the films were developed by hand using Kodak D-19 developer and Kodak (Rochester, NY, USA) fixer.

Images corresponding to coronal brain sections were captured using a microcomputer-based system interfaced with a Hewlett-Packard scanner (optical resolution: 600 × 600 dpi). The densitometric analysis was performed using program Scion-Image for Windows. Optical density readings were performed using a grayscale ranging from 0 (white) to 255 (black). Binding densities were calculated from a calibration curve obtained for the radioactivity standard slides according to Hrdina et al. (1990) and were expressed as femtomoles per mg of protein. The density of non-specific binding was measured in corresponding areas of adjacent sections and subtracted from the total binding to obtain the specific binding.

Statistics

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References

The basal 5-HT output was calculated as the mean of at least four consecutive samples differing from each other by no more than 10%, and the mean baseline 5-HT output values for each line were compared with the Student's t-test (two-tailed). The brain microdialysis data from the chlorimipramine and fluoxetine treatment experiments are expressed as a percentage of the respective pre-treatment mean value. The raw data, that is, untransformed values expressed as fmoles/20 µL, from the brain microdialysis experiments were analysed by two-way anova for repeated measures over time. When appropriate (i.e. p for the main factors and their interaction < 0.05), anova was followed by pair-wise comparisons versus within-group baseline values and versus the time-matched value of the other line using the honest significant difference (HSD) Tukey test for equal or unequal sample N, as required. Neurochemical results from rats bearing incorrectly implanted probes were not used for statistical evaluations.

Binding autoradiography data were analysed with the Student's t-test for unpaired groups (two-tailed). Statistical significance was set at p < 0.05.

Drugs and reagents

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References

[3H]-citalopram was obtained from Amersham Biosciences (Little Chalfont, Buckinghamshire, UK). Chlorimipramine was kindly provided by Ciba Geigy (Saronno, Italy); paroxetine was a gift from SmithKline and Beecham Pharmaceutical (Harlow, UK) and fluoxetine HCl was obtained from Tocris (London, UK). All other drugs and reagents were purchased from local commercial suppliers.

Brain microdialysis

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References

As shown in Fig. 1(a), the dialysing portion of the horizontal probes was located in the FPCx, across the cingulate cortex (area 2), the secondary motor cortex, the primary motor cortex and the primary somatosensory cortex (forelimb, dysgranular and jaw regions; Paxinos and Watson 1998). No line-related differences were detected in the basal 5-HT output in this brain area (13.30 ± 0.63 fmol/20 µL and 13.09 ± 0.65 fmol/20 µL in RHA/Verh and RLA/Verh rats, respectively, see Fig. 1b).

Chlorimipramine and fluoxetine were administered systemically at doses that induced maximal increments in 5-HT output in both lines (results of pilot studies using 1, 5, 10 and 20 mg/kg of each compound, not shown). Chlorimipramine (10 mg/kg, i.p.) produced a more robust and longer lasting increase in 5-HT output in the FPCx of RHA/Verh rats than in RLA/Verh rats. In RHA/Verh rats, the extracellular concentrations of 5-HT were significantly larger than the respective basal value between 60 and 180 min, and the maximal increment (310%) was observed at 120 min after drug injection, whereas in RLA/Verh rats the 5-HT output was significantly larger than the respective basal value between 60 and 180 min with a smaller peak effect (203%) at 120 min after treatment (Fig. 2a). A two-way anova revealed a significant difference for both main factors and for the interaction line × time (F1,9 : 6.66, p < 0.001). In addition, pair-wise comparisons with the HSD Tukey test revealed significant differences between the lines from 80 to 180 min after drug injection (see legend to Fig. 2a).

image

Figure 2. (a) Effect of chlorimipramine (10 mg/kg, i.p.) and (b) of fluoxetine (10 mg/kg, s.c.) on 5-HT output in the FPCx of RHA/Verh (□) and RLA/Verh rats (○). Results are expressed as a percentage of the respective pretreatment baseline value and are the mean ± SEM of the number of animals shown in parentheses. Chlorimipramine and fluoxetine were injected at the arrow. The experimental groups were compared by two-way anova for repeated measures over time, followed by multiple pair-wise comparisons with the HSD Tukey test. Filled symbols (▪, ●): p < 0.05 versus the within-line baseline value. *p < 0.05 versus time-matched value for RLA/Verh rats.

Download figure to PowerPoint

Similar results were obtained with fluoxetine (10 mg/kg, s.c.). The extracellular concentrations of 5-HT in the FPCx of RHA/Verh rats were significantly larger than the respective basal value between 40 and 180 min, and the maximal increment (255%) was observed at 160 min after drug injection. By contrast, in RLA/Verh rats, the increment in 5-HT output relative to the respective basal value was significant between 60 and 180 min with a maximal increment (196%) at 160 min after treatment (Fig. 2b). anova showed a significant difference for both main factors and their interaction (F1,9 : 4.91, p < 0.001). Moreover, post-hoc pair-wise comparisons showed significant line-related differences between 80 and 180 min after fluoxetine injection (see legend to Fig. 2b).

It is noteworthy that, in RHA/Verh rats, the maximal effect of chlorimipramine (310%) tended to be larger than that of fluoxetine (255%, compare Figs 2a and b). This difference may be due to the metabolic transformation of chlorimipramine in a demethylated metabolite that inhibits NA uptake, whereas fluoxetine and its active metabolites are selective blockers of the 5-HT transporter (see Discussion).

The 5-HT carrier is located in the axon terminals as well as in the somatodendritic portion of 5-HT neurones (Hrdina et al. 1990; Qian et al. 1995; Sur et al. 1996; Bengel et al. 1997). The effect of the systemic administration of compounds like chlorimipramine and fluoxetine on the extracellular 5-HT concentration is therefore determined by the inhibition of neurotransmitter re-uptake at both sites, namely, axon terminals and cell bodies. To assess the effects of these two compounds on 5-HT output that are mediated exclusively by the inhibition of uptake sites located in the nerve endings, we applied local intracortical perfusions of chlorimipramine or fluoxetine through the dialysis probe. Pilot experiments using different concentrations of chlorimipramine and fluoxetine (1, 10, 30 and 100 µm) showed that maximal increases of cortical 5-HT output were induced by both compounds at 30 µm and 100 µm (not shown). Therefore, in all subsequent experiments, local perfusions with chlorimipramine or fluoxetine were performed at a concentration of 100 µm, which caused a maximal effect in both lines (Figs 3a and b).

image

Figure 3. (a) Effect of chlorimipramine (100 µm) and (b) of fluoxetine (100 µm) given through the dialysis probe on 5-HT output in the FPCx of RHA/Verh (□) and RLA/Verh rats (○). Results are expressed as a percentage of the respective pre-treatment baseline value and are the mean ± SEM of the number of animals shown in parentheses. The horizontal bars indicate the infusion of chlorimipramine and fluoxetine. The experimental groups were compared by two-way anova for repeated measures over time, followed by multiple pair-wise comparisons with the HSD Tukey test. Filled symbols (▪, ●): p < 0.05 versus the within-line baseline value. *p < 0.05 versus time-matched value for RLA/Verh rats.

Download figure to PowerPoint

The increment in 5-HT output induced by the local perfusions of chlorimipramine or fluoxetine was much larger than that obtained after the systemic administration of supramaximal doses of either compound (compare Figs 2a and b with Figs 3a and b). Moreover, also after intracortical infusion, the effects of chlorimipramine and fluoxetine were more robust in RHA/Verh than in RLA/Verh rats. The extracellular concentrations of 5-HT in the FPCx of RHA/Verh rats were significantly larger than the respective basal value between 80 and 180 min after the start of the intracortical perfusion with 100 µm chlorimipramine, and the maximal increment (870%) was observed at 160 min, whereas in RLA/Verh rats the cortical 5-HT output was significantly larger than the respective basal value between 80 and 180 min with a smaller peak effect (582%) after 160 min of perfusion (Fig. 3a). A two-way anova revealed a significant difference for the main factors line and time, and for their interaction (F1,9 : 10.44, p < 0.001), and pair-wise comparisons indicated significant line-dependent differences between 80 and 180 min after starting the infusion (see legend to Fig. 3a). Likewise, the 5-HT output in the FPCx of RHA/Verh rats was significantly larger than the respective basal value between 60 and 180 min of perfusion with 100 µm fluoxetine, with a maximal effect (1005%) at 160 min, whereas in RLA/Verh rats the cortical 5-HT output was significantly larger than the respective basal value between 100 and 180 min with a smaller peak effect (673% at 180 min, see Fig. 3b). Also in this case, anova indicated significant difference for the main factors line and time, and for their interaction (F1,9 : 6.26, p < 0.001), and pair-wise comparisons revealed significant line-dependent differences at several time points after starting the infusion (see legend to Fig. 3b).

5-HT uptake sites

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References

The regional distribution of 5-HT uptake sites was studied using the specific ligand [3H]-citalopram. Autoradiographic images from coronal brain sections at four different AP levels relative to bregma are shown in Fig. 4 and quantitative data are summarized in Table 1. The intensity of the binding signal ranged more than 10-fold between various brain regions from a high of > 350 fmol/mg protein in the superior colliculus, substantia nigra, CA1 field of hippocampus, dorsal raphe nucleus (DR) and median raphe nucleus (MnR) to a low of < 20 fmol/mg protein in the pontine nuclei (Fig. 4d) and a binding signal indistinguishable from background in the cerebellum (not shown).

image

Figure 4. Autoradiographic visualization of [3H] citalopram binding to coronal brain sections of RHA/Verh (left) and RLA/Verh rats (right). Images were captured directly from the autoradiographic films as described under ‘Methods’. Increased whiteness corresponds to higher levels of grains associated with binding of [3H] citalopram. Shown are images obtained from the following AP levels relative to bregma: (a) 1.70 mm, (b) – 0.26 mm, (c) − 5.80 mm and (d) − 7.30 mm. The figures adjacent to each autoradiographic section level are the respective schematic representations obtained from Paxinos and Watson (1998). PPTg, pedunculopontine tegmental nucleus. All the other abbreviations are listed in Table 1.

Download figure to PowerPoint

Table 1.  Quantitative autoradiographic analysis of specific [3H] Citalopram binding to brain sections of RHA/Verh and RLA/Verh rats
AbbreviationStructureSpecific [3H] citalopram binding (fmol/mg protein)
RHA/VerhRLA/Verh
  1. Binding sites were labelled by incubating the sections with 0.7 nm[3H] Citalopram and quantitative densitometry of autoradiograms was performed as described under ‘Methods’. Abbreviations and anatomical terminology are derived from Paxinos and Watson (1998). Shown are the means ±SEM of measurements from six rats for each line. In each brain, three determinations were obtained from each of three sections for each area at four different AP levels relative to bregma: 1.70 mm, – 0.26 mm, − 5.80 mm and − 7.30 mm. *p < 0.05 versus the RHA/Verh group (two-tailed Student's t-test).

Plate 11 (AP: 1.70 mm)
TuOlfactory tubercle344 ± 6282 ± 20*
M1Primary motor cortex140 ± 9102 ± 3*
M2Secondary motor cortex71 ± 1055 ± 4
LSILateral septal nucleus, intermediate382 ± 11286 ± 16*
AcbCAccumbens nucleus, core91 ± 777 ± 9
AcbShAccumbens nucleus, shell157 ± 11124 ± 17
Plate 18 (AP: – 0.26 mm)
M1Primary motor cortex60 ± 737 ± 4*
M2Secondary motor cortex57 ± 1053 ± 13
CpulCaudate putamen (lateral compartment)134 ± 680 ± 4*
MfbMedial forebrain bundle328 ± 12334 ± 15
IcjIslands of Calleja444 ± 14408 ± 13
Plate 43 (AP: – 5.80 mm)
SuGSuperficial gray layer of superior colliculus392 ± 12356 ± 10*
CA1Field CA1 of Ammon's horn396 ± 13352 ± 4*
DLPAGDorsolateral periaqueductal gray284 ± 12238 ± 8*
IPRInterpeduncular nucleus, rostral464 ± 10442 ± 16
SNCDSubstantia nigra, compact part, dorsal tier421 ± 19437 ± 13
SNRSubstantia nigra, reticular part455 ± 12463 ± 21
Plate 49 (AP: – 7.30 mm)
DRDDorsal raphe nucleus, dorsal part464 ± 28419 ± 17
DRVDorsal raphe nucleus, ventral part466 ± 12428 ± 16
MnRMedian raphe nucleus340 ± 14350 ± 8

Most forebrain structures receive parallel innervation from B7 (DR) and B8 cells (MnR; Halliday et al. 1995). The ascending 5-HT projections diverge into different fibre pathways to innervate their main target structures: the olfactory bulb, hypothalamus, thalamus, septal area, striatum, hippocampus and cerebral cortex (Halliday et al. 1995). As can be seen from Figs 4(a–d), the [3H]-citalopram binding signal was particularly intense in all the forebrain areas containing a high density of 5-HT nerve endings.

A more intense [3H]-citalopram binding signal was detected in several forebrain areas of RHA/Verh rats as compared to their RLA/Verh counterparts, including the intermediate part of the lateral septal nucleus, the olfactory tubercle, and the lateral part of the caudate putamen nucleus (Figs 4a and b). Importantly, the density of 5-HT uptake sites was larger in the FPCx of RHA/Verh versus RLA/Verh rats (specifically, the primary motor cortex, M1), corresponding to the brain area where the dialysis probes were implanted (Figs 1a and 4a and b). Significant differences between the two lines were also observed in the field CA1 of the ventral hippocampus (Fig. 4c).

The 5-HT projections in the mesencephalon are found in central gray structures, the superficial layers of the superior colliculus, the cranial nerve nuclei (except those supplying the extraocular muscles), the interpeduncular complex, the substantia nigra and other dopaminergic cell groups. As shown in Fig. 4(c and d), all these areas were intensely labelled with [3H]-citalopram and the density of 5-HT uptake sites was larger in the superficial gray layer of the superior colliculus of RHA/Verh than RLA/Verh rats. The majority of midbrain structures receive parallel innervation from B7 5-HT cells (i.e. DR), with a minor contribution from MnR (i.e. B8; Halliday et al. 1995). Accordingly, the [3H]-citalopram labelling was very intense in the DR and MnR, with no significant differences across the two lines (Fig. 4d).

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References

Two major findings emerged from the present study. First, the increase in the cortical 5-HT output elicited by the systemic administration or the local perfusion of chlorimipramine and fluoxetine was greater in RHA/Verh than in RLA/Verh rats. Second, in many forebrain areas that receive 5-HT projections originating in the DR, such as the FPCx, or in the MnR, like the hippocampus, the density of 5-HT re-uptake sites was larger in RHA/Verh rats than in their RLA/Verh counterparts. These findings confirm and extend previous reports indicating that these two lines differ in the neurochemical properties of their central 5-HT pathways (Driscoll et al. 1983; Charnay et al. 1995; Corda et al. 1997b) and are consistent with the view that the 5-HTergic tone of the DR projection to the FPCx is stronger in the RHA/Verh line relative to the RLA/Verh line.

It is noteworthy that, in RHA/Verh rats, the maximal effect of systemic chlorimipramine tended to be larger than that of fluoxetine, a selective inhibitor of the 5-HT transporter in vitro and in vivo (compare Figs 2a and b), whereas in RLA/Verh rats the effects of these two compounds were completely superimposable. It may be speculated that the pharmacokinetic properties of chlorimipramine contribute to this line-related differential effects: although this compound is a selective inhibitor of the 5-HT transporter in vitro, a significant fraction of a systemic chlorimipramine dose is transformed in a demethylated metabolite that is able to inhibit the NA transporter (Balant-Gorgia et al. 1991). The increase in the concentration of NA induced by demethyl-chlorimipramine at certain synaptic sites may contribute to the effect of systemic chlorimipramine on cortical 5-HT output. Thus, is has been shown that, in the DR, NA exerts a facilitatory influence on 5-HT release (Baraban and Aghajanian 1980; Bel and Artigas 1996; Adell and Artigas 1999); in contrast, NA decreases the firing rate of noradrenergic cells in the locus coeruleus, thereby attenuating NA release in projection areas such as the DR (Mateo et al. 1998). Therefore, in addition to the inhibition of the 5-HT uptake, these indirect NA-mediated mechanisms in the DR and locus coeruleus could account for the line-related difference in the cortical 5-HT response to systemic chlorimipramine as compared to systemic fluoxetine. The neural bases of this dissociation remain unclear but may reflect important line-related differences in noradrenergic mechanisms that warrant further investigation.

Under steady-state conditions, the concentration of 5-HT in the extracellular fluid is determined by the balance between two main processes, namely, neurotransmitter release and re-uptake. In addition, neurotransmitter release is strongly dependent on the frequency and pattern (i.e. tonic vs. phasic) of the firing activity of 5-HT neurones (Carboni and Di Chiara 1989; Bel and Artigas 1996). The activation of somatodendritic 5-HT1A autoreceptors of DR and MnR cells gives rise to a dramatic decrease in the firing rate of these neurones, thereby attenuating the amount of neurotransmitter released by the 5-HTergic axon terminals in the projection areas (Blier and de Montigny 1994; Artigas et al. 1996; Casanovas and Artigas 1996; Invernizzi et al. 1997). It is still unclear whether the extracellular 5-HT in the midbrain nuclei is released from cell bodies and dendrites or from axon terminals of afferent and/or efferent fibres (i.e. axon collaterals) located within these nuclei (Halliday et al. 1995). Whatever the site of 5-HT release, antidepressant compounds that inhibit 5-HT re-uptake behave as indirect 5-HT1A autoreceptor agonists by increasing the extracellular concentration of the neurotransmitter in the midbrain nuclei. This effect attenuates the increment in the extracellular 5-HT concentration in the projection areas provoked by the systemic administration of neurotransmitter re-uptake blockers (Blier and de Montigny 1994; Artigas et al. 1996; Casanovas and Artigas 1996; Invernizzi et al. 1997). Accordingly, in both lines, the maximal increase in cortical 5-HT output induced by the systemic administration of chlorimipramine and fluoxetine was two- to threefold smaller than the effect observed after the local perfusion of fully effective concentrations of either compound (compare Figs 2a and b with Figs 3a and b). On the other hand, in keeping with their systemic effects, the local administration of chlorimipramine and fluoxetine determined a much greater increment of 5-HT output in RHA/Verh than in RLA/Verh rats, suggesting that the rate of cortical 5-HT release is faster in the former line. Moreover, because the basal 5-HT concentrations in the extracellular space of the FPCx, which are determined by the ratio between the rates of neurotransmitter release and uptake, were very similar in both lines, it may be speculated that the rate of cortical 5-HT uptake is also faster in the RHA/Verh line, relative to their RLA/Verh counterpart.

To test the above hypothesis, we characterized the regional distribution of 5-HT uptake sites using [3H]-citalopram autoradiography. Most 5-HT fibres projecting to forebrain and midbrain areas originate in the DR and MnR cell groups of the midbrain raphe complex (Halliday et al. 1995). Corroborating previous reports in other rat stocks and animal species, the [3H]-citalopram binding signal was very high in the DR and MnR (Hrdina et al. 1990; Bengel et al. 1997), although no line-related differences were detected in either cell group. The latter finding is in good agreement with previous reports using binding methods in brain homogenates (Charnay et al. 1995; Kulikov et al. 1995).

Studies on the cellular distribution of both the mRNA encoding the 5-HT transporter (assessed by in situ hybridization autoradiography) and the 5-HT carrier protein (monitored by means of binding assays using selective ligands and site-specific antibodies), strongly support the view that the neurotransporter is predominantly located in 5-HT neurones (Hrdina et al. 1990; Qian et al. 1995; Sur et al. 1996; Bengel et al. 1997). Consistent with these findings, the [3H]-citalopram binding signal was also very intense in several forebrain and midbrain areas that receive projections from the 5-HT cells of the DR and MnR. In addition, the binding signal in many of these terminal fields was significantly larger in RHA/Verh than in RLA/Verh rats. Some of these areas, like the FPCx (specifically, the primary motor cortex, M1), the caudate putamen nucleus and the superior colliculus, are predominantly innervated by the DR, whereas most 5-HT fibers innervating other forebrain regions, such as the CA1 field of the ventral hippocampus, originate from MnR cells (Halliday et al. 1995). These findings corroborate and extend the results of a previous study showing that [3H]-paroxetine binding to brain homogenates is greater in the hippocampus and frontal cortex, but not in the brain stem, of RHA/Verh as compared with RLA/Verh rats (Charnay et al. 1995). The present autoradiographic study, in which a single concentration of [3H]-citalopram was used, does not allow us to distinguish between changes in binding affinity and modifications in the density of 5-HT uptake sites. However, as the differences in [3H]-paroxetine binding were due to modifications in the Bmax, without significant changes in the Kd values between the lines, Charnay et al. (1995) have proposed that the number of uptake sites in the projection areas is larger in RHA/Verh than RLA/Verh rats, whereas the affinity of the transporter for 5-HT is similar in the two lines. Whatever the mechanism(s) involved, the results of our binding studies provide additional experimental support to the hypothesis that differences in the functional properties of the 5-HT pathways may influence the distinct behavioural profiles of the Roman lines, which are the product of bi-directional selection for extreme performance in the two-way active avoidance paradigm and represent one of the most thoroughly investigated animal models of anxiety. These lines differ in many other respects, at both the behavioural and the neurochemical levels: when exposed to a variety of environmental and pharmacological stressors, RLA/Verh rats show more pronounced emotional responses such as defecation, freezing, or self-grooming and a comparatively higher activation of the hypothalamus–pituitary–adrenal axis, i.e. increased corticosterone and corticotropin secretion, compared to RHA/Verh rats (Corda et al. 1997a; Steimer et al. 1997; Driscoll et al. 1998; Fernández-Teruel et al. 2002).

Collectively, the results of the brain microdialysis and [3H]-citalopram binding autoradiography studies are consistent with the view that the basal functional tone of the cortical 5-HT projections is more robust in normoemotional RHA/Verh rats than in their more fearful RLA/Verh counterparts. In this context, it is noteworthy that the relationship between serotonergic function and emotionality is a controversial issue, particularly as regards the different effects on anxiety levels in experimental animals and humans observed upon acute or chronic treatments with agents that modify central 5-HT transmission. Thus, pre-clinical studies show that drugs that reduce serotonergic transmission have anxiolytic effects (Handley 1995) and, accordingly, it has been suggested that the anxiolytic properties of 5-HT1A agonists in experimental animals are due to the decrease in the firing rate of serotonergic neurones caused by the activation of somatodendritic autoreceptors (De Vry 1995). Moreover, in vivo microdialysis studies in rats show that stressful stimuli increase 5-HT release in the amygdala, mediaI pre-frontal cortex and raphe nuclei (Adell et al. 1997; Corda et al. 1997b). On the other hand, the acute administration of 5-HT re-uptake inhibitors has been reported to produce anxiogenic effects, whereas chronic treatment with these agents decreases anxiety-related behaviour in rats (Griebel et al. 1994; Zhang et al. 2000). Similarly, clinical studies indicate that SSRIs have anxiolytic effects, particularly in patients suffering from generalized anxiety and panic disorders (Goddard et al. 1999). However, the anxiolytic effect of SSRIs is observed after a 2- to 3-week delay from the onset of medication (Goddard et al. 1999), suggesting that these compounds produce neuroadaptive changes in serotonergic transmission, which may be responsible for their therapeutic effects. Thus, after chronic SSRI administration, the increase in synaptic levels of 5-HT leads to desensitization of both, somatodendritic (Blier and de Montigny 1994; Artigas et al. 1996) and post-synaptic (Li et al. 1996) 5-HT1A receptors. Considered together, these findings suggest that the increment in 5-HT availability upon acute re-uptake inhibition results in an initial increase in emotionality, which turns into a reduction in anxiety levels during chronic treatment. Likewise, the stronger basal serotonergic tone of RHA/Verh rats as compared with RLA/Verh rats may result in a decreased sensitivity of post-synaptic 5-HT1A receptors in the former line relative to the latter. This line-related difference in 5-HT-mediated transmission may be involved in the less intense emotional reactivity of RHA/Verh rats relative to their RLA/Verh counterparts.

In summary, this study is the first to report in vivo experimental evidence, obtained using brain microdialysis, suggesting that the functional tone of the cortical 5-HT projections of RLA/Verh rats is weaker than that of their RHA/Verh counterparts. Together, these findings support the views that the RLA/Verh line may represent a useful genetic model to investigate the role of different aspects of 5-HT-mediated transmission in the expression of anxiety-like behaviours and in the pharmacodynamics of the anxiolytic effects of 5-HT re-uptake inhibitors.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Animals
  5. Brain microdialysis
  6. Surgery
  7. Sample collection and chromatographic assays
  8. Treatments
  9. Histology
  10. [3H]-citalopram binding
  11. Brain sections
  12. Binding assay
  13. Quantitative analysis of autoradiograms
  14. Statistics
  15. Drugs and reagents
  16. Results
  17. Brain microdialysis
  18. 5-HT uptake sites
  19. Discussion
  20. Acknowledgements
  21. References
  • Adell A. and Artigas F. (1999) Regulation of the release of 5-hydroxytryptamine in the median raphe nucleus of the rat by catecholaminergic afferents. Eur. J. Neurosci. 11, 23052311.
  • Adell A., Casanovas J. M. and Artigas F. (1997) Comparative study in the rat of the actions of different types of stress on the release of 5-HT in raphe nuclei and forebrain areas. Neuropharmacology 36, 735741.
  • American Psychiatric Association. (1994) Diagnostic and Statistical Manual of Mental Disorders, 4th edn, pp. 393444. American Psychiatric Association, Washington, DC.
  • Artigas F., Romero L., De Montigny C. and Blier P. (1996) Acceleration of the effect of selected antidepressant drugs in major depression by 5-HT1A antagonists. Trends Neurosci. 19, 378383.
  • Balant-Gorgia A. E., Gex-Fabry M. and Balant L. P. (1991) Clinical pharmacokinetics of clomipramine. Clin. Pharmacokinet. 20, 447462.
  • Baraban J. M. and Aghajanian G. K. (1980) Suppression of firing activity of 5-HT neurons in the dorsal raphe by α-adrenoceptor antagonists. Neuropharmacology 19, 355363.
  • Bel N. and Artigas F. (1996) In vivo effects of the simultaneous blockade of serotonin and norepinephrine transporters on serotonergic function: microdialysis studies. J. Pharmacol. Exp. Ther. 278, 10641072.
  • Bengel D., Johren O., Andrews A. M., Heils A., Mossner R., Sanvitto G. L., Saavedra J. M., Lesch K. P. and Murphy D. L. (1997) Cellular localization and expression of the serotonin transporter in mouse brain. Brain Res. 778, 338345.
  • Blier P. and De Montigny C. (1994) Current advances and trends in the treatment of depression. Trends Pharmacol. Sci. 15, 220226.
  • Den Boer J. A., Westenberg H. G., Kamerbeek W. D., Verhoeven W. M. and Kahn R. S. (1987) Effect of serotonin uptake inhibitors in anxiety disorders; a double-blind comparison of clomipramine and fluvoxamine. Int. Clin. Psychopharmacol. 2, 2132.
  • Carboni E. and Di Chiara G. (1989) Serotonin release estimated by transcortical dialysis in freely-moving rats. Neuroscience 32, 637645.
  • Casanovas J. M. and Artigas F. (1996) Differential effects of ipsapirone on 5-hydroxytryptamine release in the dorsal and median raphe neuronal pathways. J. Neurochem. 67, 19451952.
  • Charnay Y., Steimer T., Huguenin C. and Driscoll P. (1995) [3H] Paroxetine binding sites: brain regional differences between two psychogenetically selected lines of rats. Neurosci. Res. Comm. 16, 2935.
  • Corda M. G., Lecca D., Piras G., Di Chiara G. and Giorgi O. (1997a) Biochemical parameters of dopaminergic and GABAergic neurotransmission in the CNS of Roman high-avoidance and Roman low-avoidance rats. Behav. Genet. 27, 527536.
  • Corda M. G., Lecca D., Piras G., Di Chiara G. and Giorgi O. (1997b) Tail-pinch and pentylenetetrazol increase the release of serotonin in the cerebral cortex of Roman high-avoidance, but not low avoidance rats. Neurosci. Meeting Abstract 23, 1850.
  • Crowe R. R. (1999) Molecular genetics of anxiety disorders. In: Neurobiology of Mental Illness (CharneyD. S., NestlerE. J. and BunneyB. S., eds), pp. 451462. Oxford University Press, New York.
  • De Vry J. (1995) 5-HT1A receptor agonists: recent developments and controversial issues. Psychopharmacology 121, 126.
  • Driscoll P., Dedek J., Martin J. R. and Zivkovic B. (1983) Two-way avoidance and acute shock stress induced alterations of regional noradrenergic, dopaminergic and serotonergic activity in Roman high- and low-avoidance rats. Life Sci. 24, 17191725.
  • Driscoll P., Escorihuela R. M., Fernández-Teruel A., Giorgi O., Schwegler H., Steimer T., Wiersma A., Corda M. G., Flint J., Koolhaas J. M., Langhans W., Schulz P. E., Siegel J. and Tobeña A. (1998) Genetic selection and differential stress responses: The Roman lines/strains of rats. Ann. N.Y. Acad. Sci. 851, 501510.
  • Fernández-Teruel A., Escorihuela R. M., Gray J. A., Aguilar R., Gil L., Gimenez-Llort L., Tobeña A., Bhomra A., Nicod A., Mott R., Driscoll P., Dawson G. R. and Flint J. (2002) A quantitative trait locus influencing anxiety in the laboratory rat. Genome Res. 12, 618626.
  • Giorgi O., Corda M. G., Carboni G., Frau V. and Di Chiara G. (1997) Effects of cocaine and morphine in rats from two psychogenetically selected lines: a behavioral and brain dialysis study. Behav. Genet. 27, 537546.
  • Goddard A. W., Coplan J. D., Gorman J. M. and Charney D. S. (1999) Principles of the pharmacotherapy of anxiety disorders. In: Neurobiology of Mental Illness (CharneyD. S., NestlerE. J. and BunneyB. S., eds), pp. 548563. Oxford University Press, New York.
  • Griebel G., Moreau J. L., Jenck F., Misslin R. and Martin J. R. (1994) Acute and chronic treatment with 5-HT reuptake inhibitors differentially modulate emotional responses in anxiety models in rodents. Psychopharmacology 113, 463470.
  • Halliday G., Harding A. and Paxinos G. (1995) Serotonin and tachykinin systems. In: The Rat Nervous System, 2nd edn. (PaxinosG., ed.), pp. 929974. Academic Press, San Diego.
  • Handley S. L. (1995) 5-Hydroxytryptamine pathways in anxiety and its treatment. Pharmacol. Ther. 66, 103148.
  • Hrdina P. D., Foy B., Hepner A. and Summers R. J. (1990) Antidepressant binding sites in brain: autoradiographic comparison of [3H] paroxetine and [3H] imipramine localization and relationship to serotonin transporter. J. Pharmacol. Exp. Ther. 252, 410418.
  • Invernizzi R., Velasco C., Bramante M., Longo A. and Samanin R. (1997) Effect of 5-HT1A receptor antagonists on citalopram-induced increase in extracellular serotonin in the frontal cortex, striatum and dorsal hippocampus. Neuropharmacology 36, 467473.
  • Kaplan H. I., Sadock B. J. and Grebb J. A. (1994) Synopsis of Psychiatry, 7th edn, pp. 516572. Williams & Wilkins, Baltimore.
  • Kindler S., Dolberg O. T., Cohen H., Hirschmann S. and Kotler M. (1997) The treatment of co-morbid premature ejaculation and panic disorder with fluoxetine. Clin. Neuropharmacol. 20, 466471.
  • Kulikov A., Castanon N., Mormède P. and Chaouloff F. (1995) Cerebral tryptophan hydroxylase activity, and 5-HT1A receptor, 5-HT2A receptor, and 5-HT transporter binding in grouped and isolated Roman RHA and RLA rats: relationships with behaviours in two models of anxiety. Psychopharmacology 121, 385395.
  • Li Q., Muma N. A. and Van De Kar L. D. (1996) Chronic fluoxetine induces a gradual desensitization of 5-HT1A receptors: reductions in hypothalamic and midbrain Gi and Go proteins and in neuroendocrine responses to a 5-HT1A agonist. J. Pharmacol. Exp. Ther. 279, 10351042.
  • Mancini C. and Ameringen M. V. (1996) Paroxetine in social phobia. J. Clin. Psychiatry 57, 519522.
  • Mateo Y., Pineda J. and Meana J. J. (1998) Somatodendritic alpha2-adrenoceptors in the locus coeruleus are involved in the in vivo modulation of cortical noradrenaline release by the antidepressant desipramine. J. Neurochem. 71, 790798.
  • Paxinos G. and Watson C. (1998) The Rat Brain in Stereotaxic Coordinates, 4th edn. Academic Press, San Diego.
  • Qian Y., Melikian H. E., Rye D. B., Levey A. I. and Blakely R. D. (1995) Identification and characterization of antidepressant-sensitive serotonin transporter proteins using site-specific antibodies. J. Neurosci. 15, 12611274.
  • Sanders A. R., Sevilla D., Detera-Wadleigh S. D. and Gershon E. S. (1999) Molecular genetics of mood disorders. In: Neurobiology of Mental Illness (CharneyD. S., NestlerE. J. and BunneyB. S., eds), pp. 299316. Oxford University Press, New York.
  • Steimer T., La Fleur S. and Schulz P. (1997) Neuroendocrine correlates of emotional reactivity and coping in male rats from the Roman high (RHA/Verh)- and low (RLA/Verh)-avoidance lines. Behav. Genet. 27, 503512.
  • Sur C., Betz H. and Schloss P. (1996) Immunocytochemical detection of the serotonin transporter in rat brain. Neuroscience 73, 217231.
  • Willner P. (1991) Animal models of depression. In: Behavioural Models in Psychopharmacology: Theoretical, Industrial and Clinical Perspectives. (Willner, P., ed.), pp. 89174. Cambridge University Press, Cambridge.
  • Zhang Y., Raap D. K., Garcia F., Serres F., Ma Q., Battaglia G. and Van de Kar L. D. (2000) Long-term fluoxetine produces behavioral anxiolytic effects without inhibiting neuroendocrine responses to conditioned stress in rats. Brain Res. 855, 5866.