Cerebral Cortex, Vol. 13, No. 8, 870-882,
August 2003
© 2003 Oxford University Press
In Vivo Modulation of the Activity of Pyramidal Neurons in the Rat Medial Prefrontal Cortex by 5-HT2A Receptors: Relationship to Thalamocortical Afferents
Department of Neurochemistry, Institut d’Investigacions Biomèdiques de Barcelona (CSIC), IDIBAPS, 08036 Barcelona, Spain
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionNotesReferences |
|---|
The activation of 5-HT2A receptors in medial prefrontal cortex (mPFC) by the hallucinogen DOI increases the firing activity of dorsal raphe (DR) 5-HT neurons and prefrontal 5-HT release. Here we show that the i.v. administration of DOI markedly affected the firing rate of identified pyramidal neurons recorded extracellularly. DOI excited (481%) 21/56 neurons, inhibited (11%) 17/56 neurons and left the rest unaffected (overall 2.4-fold increase in firing rate). Both effects were antagonized by 5-HT2A receptor blockade. 5-HT2A-mediated orthodromic excitations were recorded in pyramidal neurons projecting to DR after electrical stimulation of this nucleus. We also examined whether the effects of DOI in mPFC involve thalamic excitatory inputs. The disinhibition of the mediodorsal and centromedial nuclei of the thalamus by local bicuculline resembled the effects of DOI as it increased pyramidal cell firing and 5-HT release in mPFC. However, the selective activation of prefrontal µ-opioid and mGlu II receptors counteracted the effects of the thalamic disinhibition but not those of DOI. Moreover, extensive thalamic lesions did not alter the effect of DOI on pyramidal cell firing and 5-HT release. We conclude that DOI increases the activity of the mPFC-DR circuit by an action on postsynaptic 5-HT2A receptors unrelated to thalamocortical afferents.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionNotesReferences |
|---|
The prefrontal cortex is involved in a large number of higher brain functions (working memory, cognition, affect, behavioral inhibition, etc.) and controls hierarchically the activity of subcortical structures (Fuster, 1997
; Miller and Cohen, 2001). Pyramidal neurons in cortical layer V integrate excitatory inputs from cortical and subcortical areas. The latter include, among others, several nuclei of the thalamus, such as the medio-dorsal (MD) and centromedial (CM) nuclei (Berendse and Groenewegen, 1991; Fuster, 1997; Kuroda et al., 1998; Van der Werf et al., 2002). Modulatory inputs arise from the brainstem aminergic nuclei (raphe nuclei, ventral tegmental area and locus coeruleus) (Azmitia and Segal, 1978; Phillis, 1984; Kosofsky and Molliver, 1987; Durstewitz et al., 2000; Lewis and O’Donnell, 2000). Signal integration in pyramidal neurons is exerted at various cellular levels, with a key role played by the large apical dendrites. These are highly enriched in serotonergic 5-HT2A receptors which are also present in large and medium-size GABAergic interneurons that control the activity of pyramidal neurons in local microcircuits (Willins et al., 1997; Jakab and Goldman-Rakic, 1998, 2000; Jansson et al., 2001; Martín-Ruiz et al., 2001). Hallucinogens like LSD or DOI are agonists and atypical antipsychotics are antagonists at 5-HT2A receptors (Kroeze and Roth, 1998; Meltzer, 1999). Also, atypical antipsychotics can augment the effects of antidepressant drugs (Ostroff and Nelson, 1999; Shelton et al., 2001; Marangell et al., 2002). Moreover, although the physiological role of cortical 5-HT2A receptors remains largely unknown, recent data suggest their involvement in working memory (Williams et al., 2002). Together, these observations suggest that 5-HT2A receptor blockade may be beneficial for the treatment of severe psychiatric disorders.
Recently, we showed that the selective activation of prefrontal 5-HT2A receptors by DOI increased the firing activity of 5-HT neurons in the dorsal raphe nucleus (DR) and the 5-HT release in mPFC (Martín-Ruiz et al., 2001). These effects are presumably mediated by an increased excitatory transmission through pyramidal axons projecting to DR 5-HT neurons (Aghajanian and Wang, 1977; Sesack et al., 1989; Takagishi and Chiba, 1991; Hajós et al., 1998; Peyron et al., 1998; Celada et al., 2001). However, conflicting results have been reported in regards to the effects of 5-HT2A receptor activation on cortical pyramidal cells. Thus, the microiontophoretic application of DOI to anesthetized rats predominantly inhibited prefrontal neurons, although low ejection currents potentiated the excitatory effect of glutamate (Ashby et al., 1989, 1990). Using in vitro intra-cellular recordings of pyramidal neurons, depolarizing and hyperpolarizing effects of 5-HT through the activation of 5-HT2A receptors have been reported (Araneda and Andrade, 1991; Tanaka and North, 1993; Aghajanian and Marek, 1997, 1999, 2000; Arvanov et al., 1999; Zhou and Hablitz, 1999). DOI has been shown to prolong late EPSCs evoked in layer V pyramidal neurons by electrical stimulation of afferent fibers (Aghajanian and Marek, 1999). However, an examination of the in vivo systemic actions of DOI on the firing activity of pyramidal cells is lacking. Moreover, the cellular site(s) responsible for the above effects are not fully elucidated. Hence, despite the abundance of 5-HT2A receptors in pyramidal neurons compared to other cellular locations (Willins et al., 1997; Jakab and Goldman-Rakic, 1998, 2000), the 5-HT-induced EPSCs in layer V pyramidal neurons have been attributed to the stimulation of terminal 5-HT2A receptors on thalamocortical afferents on the basis of lesion and pharmacological studies (Marek and Aghajanian, 1998; Marek et al., 2001). Also, thalamic lesions partly prevented the increase in c-fos expression produced mainly in non-pyramidal neurons of somatosensory cortex by a large systemic dose of DOI (Scruggs et al., 2000).
Using extracellular recordings of projection neurons and intracerebral microdialysis, in the present study we sought to examine whether DOI enhances the firing activity of projection pyramidal neurons in mPFC, and whether these effects are dependent on 5-HT2A receptors putatively located in thalamocortical afferents.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionNotesReferences |
|---|
Experimental Design
To experimentally assess the above points, we examined the effects of the administration of DOI on the firing rate of projection pyramidal neurons and 5-HT release in mPFC. We also examined the effects of increasing the excitatory thalamic input onto mPFC neurons on the pyramidal firing rate and 5-HT release. Thalamic disinhibition was carried out by locally applying bicuculline in the mediodorsal (MD) and centromedial (CM) nuclei, which project to the cingulated and limbic parts of the mPFC (Berendse and Groenewegen, 1991; Fuster, 1997; Van der Werf et al., 2002). Finally, we assessed the effects of drugs modulating glutamate release from thalamocortical afferents and of extensive thalamic lesions on the DOI-induced effects. Specifically, we tested the hypotheses that (i) DOI may enhance the firing rate of pyramidal neurons in mPFC in vivo, thus resulting in an increased excitatory output to subcortical structures, and (ii) this effect may be independent of thalamocortical inputs. The effects of the excitatory thalamic input on pyramidal cell firing and 5-HT release in mPFC have been studied using single unit recordings and dual probe microdialysis (mPFC and thalamus), respectively. To increase the excitatory input onto mPFC, we applied bicuculline by reverse dialysis through the thalamic probe in order to disinhibit thalamocortical afferents. Also, by performing lesions of the thalamic nuclei projecting to mPFC, we studied whether the effects of DOI on 5-HT release and on pyramidal cell firing depend on the integrity of thalamic inputs onto this cortical area.
Animals
Male albino Wistar rats (Iffa Credo, Lyon, France) weighing 250–320 g and kept in a controlled environment (12 h light–dark cycle and 22 ± 2°C room temperature) with food and water provided ad libitum, were used in in vivo experiments. Animal care followed the European Union regulations (OJ of EC L358/1 18/12/1986).
Drugs and Reagents
5-HT oxalate, (-)-bicuculline methiodide, DOI (1-[2,5-dimethoxy-4-iodo-phenyl-2-aminopropane]), (+)-MK-801 (dizocilpine) and picrotoxinin were from Sigma/RBI (Natick, MA). 1S,3S-Aminocyclopentane dicarboxylic acid (1S,3S-ACPD), DAMGO and NBQX (2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo[f]quinoxaline-7-sulfonamide) were from Tocris (Bristol, UK). Citalopram
HBr, LY-379268 {(-)-2-oxa-4-aminobicyclo-[3.1.0]hexane-4,6-dicarboxylate]} and M100907 [R-(+)-alpha-(2,3-dimethoxyphenil)-1-[4-fluorophenylethyl]-4-piperidinemethanol] (Lilly code LY 368675) were from Lundbeck A/S and Eli Lilly & Co., respectively. Other materials and reagents were from local commercial sources. For the assessment of local effects, drugs were dissolved in the perfusion fluid and applied by reverse dialysis at the stated concentrations. Concentrated solutions (1 mM; pH adjusted to 6.5–7 with NaHCO3 when necessary) were stored frozen (-80°C) and working solutions were prepared daily by dilution. Concentrations are expressed as free bases. In microdialysis experiments, control rats were perfused for the entire experiment with artificial cerebrospinal fluid. The concentrations of DOI and other agents used in the present microdialysis experiments were determined in previous studies (Martín-Ruiz et al., 2001) or in pilot studies. Nominal concentrations are much greater than the in vitro drug affinities for receptors in order to allow for sufficient stimulation of the desired receptors in the in vivo conditions. Indeed, only a small percentage of the drugs applied by reverse dialysis at the flow rate used enter the brain compartment and then, they are continuously cleared by the CSF and the systemic circulation. The bars in the figures show the period of drug application (corrected for the void volume of the system). In experiments involving systemic administration, the drugs were administered i.v. at the doses stated.
Electrophysiological Experiments
We carried out three types of electrophysiological experiments involving the extracellular recording of pyramidal neurons in mPFC. We examined effects of the i.v. administration of DOI (0.05–0.6 mg/kg in cumulative doses) on the firing rate of pyramidal cells. In some instances, the 5-HT2A receptor antagonist M100907 (0.12–1 mg/kg i.v. in cumulative doses) was administered after DOI to examine the involvement of 5-HT2A receptors in the action of DOI. Secondly, we assessed the effects of endogenous 5-HT on the activity of pyramidal neurons in mPFC. To this end, the DR was stimulated (0.15–2 mA, 0.2 ms square pulses, 0.9 Hz) and peristimulus time histograms (PSTH) were constructed in baseline conditions and after the administration of M100907. Finally, we assessed the effects of the application of bicuculline in the CM + MD nuclei of the thalamus. Bicuculline (0.1–0.3 mM; dissolved in aCSF) was infused through a stainless steel (32 g) cannula (n = 3) implanted in the thalamus (tip coordinates: AP -3.6, L -0.7, DV -6.5). The cannula was attached to a 10 µl Hamilton syringe by a Teflon tubing. A microinfusion pump (Bioanalytical Systems Inc., West Lafayette, IN) was used. After recording baseline spontaneous activity for at least 5 min, 0.2 µl of bicuculline (0.1–0.3 mM) were infused through the cannula over the course of 1 min. This volume has been reported to diffuse to a maximum effective diameter of 0.4–0.6 mm (Myers, 1971). In another group of rats, we applied bicuculline (1 mM, dissolved in the aCSF) in the CM + MD thalamic nuclei by reverse dialysis using a modified dialysis probe designed to minimize dead volume. Flow rate was adjusted at 3 µl/min to reduce the onset of delivery of bicuculline to the brain tissue. Basal firing activity of pyramidal cells was recorded and then the aCSF used to perfuse the probes was changed to one containing bicuculline.
Single-unit extracellular recordings of pyramidal neurons were performed as follows. Rats were anesthetized (chloral hydrate 400 mg/kg i.p.) and positioned in a stereotaxic apparatus. Additional doses of chloral hydrate (80 mg/kg) were administered i.v. through the femoral vein. Body temperature was maintained at 37°C throughout the experiment with a heating pad. All wound margins and points of contact between the animal and the stereotaxic apparatus were infiltrated with lidocaine solution (5%). In order to minimize pulsation, the atlanto-occipital membrane was punctured to release some CSF. Stimulating electrodes were stereotaxically positioned in the DR or ventral tegmental area (VTA) (coordinates: DR, AP -7.8, L -3.1, DV -6.8 with an angle of 30° to avoid the sinus; VTA, AP -6.0, L -0.5, DV -8.2) and were stimulated at 0.15–2 mA, 0.2 ms square pulses, 0.9 Hz. Pyramidal neurons were recorded extracellularly with glass micropipettes pulled from 2.0 mm capillary glass (WPI, Saratosa, FL) on a Narishige PE-2 pipette puller (Narishige Sci. Inst., Tokyo, Japan). Microelectrodes were filled with 2 M NaCl. Typically, impedance was between 4 and 10 M
. Constant current electrical stimuli were generated with a Grass stimulation unit S-48 connected to a Grass SIU 5 stimulus isolation unit. Single unit extracellular recordings were amplified with a Neurodata IR283 (Cygnus Technology Inc., Delaware Water Gap, PA), postamplified and filtered with a Cibertec amplifier (Madrid, Spain) and computed on-line using a DAT 1401plus interface system Spike2 software (Cambridge Electronic Design, Cambridge, UK).
Descents were carried out at AP +3.2 to +3.4, L -0.5 to -1.0, DV -1.1 to -4.8 below the brain surface. All recorded pyramidal neurons were identified by antidromic activation from DR and/or VTA and collision extinction with spontaneously occurring spikes (Fuller and Schlag, 1976). Neurons without antidromic activation or without spontaneous firing activity were discarded.
Surgery and Microdialysis Procedures
An updated description of the microdialysis procedures used can be found in Adell and Artigas (Adell and Artigas, 1998). Briefly, anesthetized rats (pentobarbital, 60 mg/kg i.p.) were stereotaxically implanted with concentric microdialysis probes equipped with a Cuprophan membrane. In experiments involving the local effect of DOI in mPFC, rats were implanted with one probe in mPFC (AP +3.2, L -0.8, DV -6.0; probe tip: 4 mm of active membrane, between DV -1.5 to -5.5) [coordinates in mm (Paxinos and Watson, 1986)]. In experiments involving thalamic disinhibition rats were implanted with two probes (see below). On the day following implants, microdialysis experiments were performed in freely moving (DOI effect) or chloral hydrate anesthetized rats (thalamic disinhibition). The probes in mPFC were perfused at 1.5 µl/min with artificial cerebrospinal fluid (aCSF; 125 mM NaCl, 2.5 mM KCl, 1.26 mM CaCl2 and 1.18 mM MgCl2) containing 1 µM citalopram. After a 1 h stabilization period, four fractions were collected to obtain basal values before local administration of drugs by reverse dialysis. Successive 20 min (30 µl) dialysate samples were collected. In most experiments, the partial 5-HT2A/2C receptor agonist DOI was applied alone for 2 h (six fractions), followed by its application in combination with other drugs for another 2 h period.
For experiments in chloral hydrate anesthetized rats, these were implanted with two microdialysis probes, one in mPFC (as above) and a second one in a thalamic area sampling parts of the CM an MD nuclei of the thalamus (AP -3.6, L -0.7, DV -6.5; probe tip 1.5 mm). Bicuculline (1 mM, dissolved in aCSF) was perfused through the probe implanted in the thalamus at 1.5 µl/min and the effects on the 5-HT release in mPFC were analyzed. These experiments were aimed at disinhibiting thalamic excitatory afferents to mPFC and examining the effects on distal 5-HT release. We also examined the modulation of these effects on prefrontal 5-HT release by drugs perfused through the probe in mPFC (NBQX, 1S,3S-ACPD, LY-379268 and DAMGO).
The concentration of 5-HT in dialysate samples was determined by HPLC, as described (Adell and Artigas, 1998). 5-HT was separated using a Beckman (San Ramon, CA) 3 µm particle size column and detected with a Hewlett Packard 1049 electrochemical detector at +0.6 V. Retention time was between 3.5 and 4 min and the limit of detection was typically 1 fmol/sample.
Thalamic Lesions
Electrolytic lesions of several thalamic nuclei projecting to the mPFC (Berendse and Groenewegen, 1991; Van der Werf et al., 2002) were performed by passing 1.5 mA (two pulses of 10 s each; Grass stimulation unit S-48 connected to a Grass SIU 5 stimulus isolation unit) at three different localizations: (a) AP -1.8, L -0.7, DV -5.7, (b) AP -2.7, L -0.7, DV -6.3, and (c) AP -3.6, L -0.7, DV -6.5 (Fig. 1). The tip of electrodes was peeled off (~0.5 mm) and poles were slightly separated (~0.5 mm) to affect a larger tissue area. The coordinates used were those employed by Marek et al. (Marek et al., 2001) to perform chemical (NMDA) thalamic lesions and were intended to lesion the two main nuclei projecting densely to the mPFC, CM and MD as well as several neighboring nuclei which also project to the mPFC area sampled by microdialysis probes and recording electrodes in the present study. These included, among others, the paraventricular (PV), paracentral (PC) and paratenial (PT) nuclei. Sham controls were stereotaxically implanted with the electrodes but no current was passed. Single unit recordings and microdialysis experiments were conducted 8 days after the lesion. One microdialysis experiment was carried out in rats lesioned 3 days before.
|
Histological Examinations
At the end of the experiments, rats were killed by an overdose of anesthetic. The placement of the dialysis probes was examined by perfusion of Fast green dye and visual inspection of the probe track after cutting the brain at the appropriate level. In experiments involving disinhibition of thalamic afferents to mPFC the placement of the thalamic probes or cannulae was verified histologically with Neutral Red staining after transcardial perfusion with saline followed by 10% formalin solution (Sigma). Lesioned animals were examined in the same way. Brains were post-fixed, sagitally sectioned (80 µm) and stained with Neutral Red.
Data and Statistical Analysis
Microdialysis results are expressed as fmol/fraction (uncorrected for recovery) and shown in figures as percentages of basal values (individual means of four pre-drug fractions). The statistical analysis was carried out using one- or two-way ANOVA for repeated measures followed by Duncan test. The effects of DOI or thalamic disinhibition with bicuculline on the 5-HT release in mPFC were assessed using one-way ANOVA of the 5-HT values (four stable basal fractions plus all post-drug fractions). In experiments devised to reverse the effect of the thalamic disinhibition or DOI application on 5-HT release, a second drug was applied through the microdialysis probe in mPFC. The assessment of the effects of this second drug was done with fractions 7–16. These correspond to four fractions with an stabilized effect of DOI or thalamic disinhibition on 5-HT release in mPFC followed by the six fractions (2 h) in which additional drugs were applied in mPFC. The effect of DOI on pyramidal neuron firing was assessed using paired Student’s t-test. Changes in firing rate were quantified by averaging the values after i.v. drug injection (omitting the first minute). Data are expressed as the mean ± SEM. Statistical significance has been set at the 95% confidence level (two-tailed).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionNotesReferences |
|---|
Firing Characteristics of Projection Neurons in mPFC
A total of 144 projection pyramidal neurons recorded in untreated rats have been included in this study (15 additional neurons were recorded in rats with thalamic lesions, whose characteristics are given below). All them were identified by antidromic activation from the DR (n = 104) or the VTA (n = 48) as described in Materials and Methods (Fig. 1A,E
). In some instances, pyramidal neurons projecting to both midbrain structures were recorded (n = 10). However, since no systematic stimulation from both midbrain nuclei were performed, these figures are not representative of the actual proportion of mPFC neurons innervating the DR and VTA concurrently. Two additional neurons were identified by antidromic activation from MD nucleus. The mean firing rate was 1.52 ± 0.14 spikes/s (n = 144). According to previously reported data (Dégenètais et al., 2002), we identified neurons with regular spiking and inactivating burst firing modes. Figure 2 shows typical examples of the waveform and firing characteristics of the recorded neurons as well as the identification by antidromic activation and collision test of a prefrontal neuron projecting to the DR. The calculated average conduction velocity (linear distance/latency of antidromic spikes) was 0.93 ± 0.03 m/s for the neurons projecting to the DR (n = 104) and 1.89 ± 0.15 m/s for those projecting to the VTA (n = 48). These values fall within the range of previously reported values for cortical pyramidal neurons (Thierry et al., 1983; Peterson et al., 1990).
|
Effects of DOI on the Firing Rate of Projection Pyramidal Neurons
The i.v. administration of DOI resulted in three types of responses in pyramidal neurons of the mPFC. Typically, pyramidal neurons responded to doses of DOI between 0.05 and 0.3 mg/kg i.v. One additional neuron responded to 0.6 mg/kg i.v. DOI enhanced the firing rate of 38% of the neurons examined (21/56). The mean firing rate rose from 1.34 ± 0.36 spikes/s (baseline) to 6.45 ± 0.97 spikes/s (481%; n = 21). When considering the individual change from baseline, DOI increased the firing rate to 1525 ± 549%. This difference (1525 versus 481%) is due to the low basal firing rate of some of the recorded units, which made the change (expressed as percentage of basal) much greater than the ratio between mean baseline and post-DOI firing rates in the 21 neurons examined. Another 32% of projection neurons (18/56) were unaffected by DOI administration (up to 0.3–0.8 mg/kg i.v.) whereas the rest (17/56, a 30%) experienced reductions in cell firing (to 11 ± 3% on average). Baseline firing rate of the neurons excited or inhibited by DOI did not significantly differ (1.34 ± 0.36 spikes/s for excitations versus 1.20 ± 0.33 spikes/s for inhibitions). Summing over all neurons examined, the systemic administration of DOI significantly increased the output of mPFC projection neurons a 238%, from 1.23 ± 0.18 to 2.93 ± 0.53 spikes/s (n = 56; P < 0.002, paired t-test). Figure 3 shows representative examples of projection pyramidal neurons in mPFC excited and inhibited by the systemic administration of DOI.
|
The reduction in firing rate elicited by DOI was antagonized by the selective 5-HT2A receptor antagonist M100907 (0.1–0.5 mg/kg; n = 6; Fig. 3C
) in all cases examined. The inhibitory effect of DOI was also reversed by picrotoxinin administration (1–3 mg/kg i.v.; n = 5; Fig. 3D). The DOI-induced excitations were reversed by M100907 (0.1–0.5 mg/kg i.v.) in six out of nine cases examined. Because DOI is also a partial agonist at 5-HT2C receptors, we cannot discard that its effects on some pyramidal neurons may be mediated by this receptor subtype. Alternatively, it may occur that M100907 cannot reverse the pyramidal excitation once these neurons become extensively depolarized, likely by activation of ionotropic glutamate receptors. Indeed, in one of these M100907-insensitive units, the effect of DOI was fully reversed by the mGluR II agonist LY-379268 (Fig. 3B). This agent could reverse the DOI-induced excitations of pyramidal neurons in four out of five neurons (up to 4 mg/kg i.v.). Effects of DR Electrical Stimulation on mPFC Pyramidal Neurons
To examine whether the excitatory effect of DOI could be also elicited by endogenous 5-HT, we recorded pyramidal neurons in mPFC during the electrical stimulation of the DR (see Methods). Because we were interested on putative 5-HT2A-mediated excitations, neurons that were only inhibited by DR stimulations were not considered. Peristimulus time histograms showed the presence of excitations (Fig. 4), which had a mean duration of 80 ± 8 ms and a mean latency of 82 ± 8 ms (n = 19). Six out of the 19 excitations recorded were preceded by marked inhibitions, as in the neuron shown in Figure 4. The success rate of orthodromic activation varied between neurons and was 51 ± 8% on average. In 18 out of the 19 pyramidal neurons examined, orthodromic and antidromic excitations were recorded, showing the existence of a strong reciprocal DR-mPFC connectivity (Fig. 4). The DR-induced excitations were reversed by the i.v. administration of M100907 in seven out of eight neurons tested (0.2–0.8 mg/kg i.v.) (P < 0.01 t-test, n = 7; Fig. 3B). The success rate before and after M100907 was significantly different (from 62 ± 12% to 15 ± 4%; P < 0.01). The duration and latency of the seven excitations inhibited by M100907 did not differ from those of the other 11 neurons in which reversal by M100907 was not attempted.
|
Effect of the Thalamic Disinhibition on the Activity of Pyramidal Neurons
The single-point application of bicuculline in the CM + MD nuclei (0.2 µl, 0.1–0.3 mM, over the course of 1 min) elicited a dramatic increase in the activity of projection neurons in mPFC. On average, the increase in firing rate was 2423 ± 500% (n = 3). Figure 5B shows a representative example of a neuron excited by the application of bicuculline in the thalamus. In several other neurons depolarization block occurred soon after a rapid and dramatic increase in firing rate (data not shown). The data of these neurons was not included in the calculations. We then applied bicuculline 2 mM in the CM + MD nuclei by reverse dialysis, thus enabling a more controlled and stable delivery of the drug. The application of bicuculline by this method also resulted in a very marked increase of the firing rate of projection neurons in mPFC, which attained 1881 ± 564% of baseline on average (n = 9). Figure 5A shows a representative example of a projection neuron in mPFC whose firing rate increased from 0.34 to 5.21 spikes/s after the application of bicuculline by reverse dialysis in the CM and MD nuclei. When examined, the increase in pyramidal firing rate was reversed by the i.v. administration of the mGluR II agonist LY-379268 (0.5–3.5 mg/kg) (n = 2; Fig. 5A).
|
The application of 2 mM bicuculline by reverse dialysis in the CM and MD nuclei of chloral hydrate anesthetized rats, to exactly mimic the experimental conditions in single unit recording experiments, resulted in a 3-fold increase of the 5-HT release in the mPFC [F(10,30) = 7.05, P < 0.00002, one-way repeated measures ANOVA] (Fig. 5C
). Effect of the Thalamic Disinhibition on 5-HT Release in mPFC in Anesthetized Rats
The application by reverse dialysis of 1 mM bicuculline in the CM + MD nuclei of the thalamus increased significantly the 5-HT output in mPFC of chloral hydrate anesthetized rats to a maximal increase of 175 ± 11% of baseline [F(15,45) = 5.86, P < 0.000002, one-way repeated measures ANOVA; Fig. 6A]. This effect was reversed by the concurrent application of the µ-opioid agonist DAMGO (100 µM) in mPFC, which fully attenuated the increase in 5-HT elicited by the thalamic disinhibition [F(9,27) = 6.62, P < 0.00006 one-way ANOVA of fractions 7–16; Fig. 6A]. DAMGO alone (10, 30 and 100 µM) did not affect the basal 5-HT release (data not shown). Likewise, the application in mPFC of the AMPA/KA antagonist NBQX (300 µM) or the mGluR II/III agonist 1S,3S-ACPD (300 µM) also antagonized the increase in mPFC 5-HT release produced by thalamic disinhibition [F(9,27) = 4.64, P < 0.00009 and F(9,27) = 7.16, P < 0.00003, respectively, one-way repeated measures ANOVA; Fig. 6B,C]. The selective mGluR II agonist LY-379268 appeared to be less effective and elicited only a partial reversal at 0.3–1 mM (three fractions each) [F(9,36) = 7.66, P < 0.00003; Fig. 6D]. A higher concentration (3 mM) failed to antagonize the effect of thalamic disinhibition and increased 5-HT further (data not shown).
|
Effect of DOI on 5-HT Release in mPFC
As previously observed (Martín-Ruiz et al., 2001) the application of DOI 100 µM in mPFC significantly increased the local 5-HT release [F(9,63) = 11.5, P < 0.000001, one-way repeated measures ANOVA] (Fig. 7). This effect is due to the selective activation of 5-HT2A receptors, as it is fully blocked by co-perfusion of the selective 5-HT2A antagonist M100907 (Martín-Ruiz et al., 2001). Moreover, it is also reversed by NBQX and 1S,3S-ACPD (Martín-Ruiz et al., 2001), as observed for the effects of thalamic disinhibition (Fig. 6). Since 100 µM DAMGO application in mPFC reversed the effect of thalamic disinhibition (Fig. 6A), we examined whether it could antagonize or reverse the effect of DOI on 5-HT release in mPFC. The concurrent application of DAMGO 100 µM did not prevent the increase in 5-HT release induced by DOI (Fig. 7A). Two-way ANOVA revealed a significant effect of time [F(9,90) = 19.43, P < 0.00001] but not of the group or time x group interaction. Also, DAMGO was equally ineffective to reverse the effect of DOI [F(9,36) = 0.74, P = 0.66; Fig. 7B]. The selective mGluR II agonist LY-379268 also failed to reverse the effect of DOI in a wide range of concentrations (100 µM–1 mM) (Fig. 7C). In fact, at 100 µM, LY-379268 further enhanced the 5-HT release induced by DOI [F(9,27) = 4.91, P < 0.0006]. This effect is possibly related to the reported enhancement of 5-HT release induced by this agent after systemic administration, although lower local doses (3–10 µM) did not elicit such effect (Cartmell et al., 2001).
|
Influence of Thalamic Lesions on the Effects of DOI in mPFC
Figure 8 shows examples of the thalamus of control (sham-operated) and lesioned rats. These were subjected to electrolytic lesions 3 and 8 days before experiments, as described in Methods. All brain sections showed extensive destruction of the centromedial and mediodorsal nuclei and also of the paracentral and paratenial nuclei. Additionally, most animals (~70%) exhibited moderate to severe lesions of the anteromedial, centrolateral, parafascicular, paraventricular and reuniens nuclei, which also project to the cingulated and limbic portions of the mPFC.
|
The ability of DOI to modulate the firing rate of pyramidal neurons in mPFC was unaffected by the prior lesion (8 days) of the CM and MD nuclei of the thalamus (n = 9) (Fig. 9
). All neurons recorded were also identified as projection neurons by antidromic stimulation from the DR or VTA, as in control rats. Baseline firing rate was significantly lower in pyramidal neurons from lesioned rats compared to the rest of pyramidal neurons (0.51 ± 0.12 spikes/s versus 1.52 ± 0.14 spikes/s, n = 15 and 144, respectively; P < 0.00001). Four projection neurons were excited by DOI administration (0.05–0.3 mg/kg i.v.), an effect antagonized by M100907 in all cases examined (3/3; 0.2–0.5 mg/kg i.v.) (Fig. 9A). Likewise, four neurons were inhibited by DOI (0.05–0.2 mg/kg i.v.), an effect also antagonized by M100907 (0.2–0.5 mg/kg i.v.) (Fig. 9B). One neuron was unaffected by DOI administration (up to 0.3 mg/kg i.v.).
|
The application of DOI 100 µM in the mPFC of rats subjected to sham lesions of the thalamus increased the 5-HT release to an extent similar to that seen in naïve animals (~200% of baseline; non-significant difference with respect to naïve rats, two-way repeated measures ANOVA; Fig. 10
). Likewise, DOI application elicited a comparable increase of 5-HT release in rats subjected to electrolytic lesions of the thalamus 3 and 8 days before (Fig. 10). Two-way repeated measures ANOVA revealed a significant effect of the time [F(9,207) = 27.4 P < 0.000001] but not of the group or time x group interaction at any of the lesion times versus sham rats. In two rats lesioned 8 days before, the subsequent co-perfusion of M100907 (300 µM) fully reversed the DOI-stimulated 5-HT release (to ~70% of baseline), indicating that the effect of DOI was due to the activation of 5-HT2A receptors (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionNotesReferences |
|---|
5-HT2A receptors are widely expressed in prefrontal cortex of rodent and primate brain (see Introduction). Their involvement in cognition and mood is supported by the fact that hallucinogens (LSD, DOI or chemically related compounds) act as partial agonists whereas atypical antipsychotics and some antidepressant drugs, like nefazodone or trazodone, are 5-HT2A receptor antagonists (Kroeze and Roth, 1998
; Meltzer, 1999). More recently, direct evidence for the involvement of 5-HT2A in cognition has been provided (Williams et al., 2002). 5-HT2A receptor activation has been shown to exert a variety of actions in prefrontal neurons in vitro (Araneda and Andrade, 1991; Tanaka and North, 1993; Aghajanian and Marek, 1997, 1999, 2000; Arvanov et al., 1999; Zhou and Hablitz, 1999). The present study shows that DOI markedly affects the firing activity of prefrontal pyramidal neurons in vivo, and in particular, of neurons projecting back to the DR. Also, the physiological stimulation of the DR elicited 5-HT2A-mediated excitations in such DR-projecting pyramidal neurons.
DOI markedly excited nearly 40% of the neurons examined. A lower proportion of neurons (30%) were inhibited and the rest were unaffected by the DOI doses used (in most instances, up to 300 µg/kg). This suggests that the prefrontal output to subcortical structures is affected by DOI in an uneven manner. However, the average effect of DOI is excitatory (2.4-fold increase in firing rate), which is consistent with the increase in 5-HT cell firing and 5-HT release produced by the local application of DOI in mPFC (Martín-Ruiz et al., 2001). Given the existence of anatomical projections from the mPFC to the DR and the control exerted by the mPFC on DR 5-HT neurons (Hajós et al., 1998; Celada et al., 2001), the local effect of DOI on mPFC 5-HT release was interpreted as resulting from an enhanced excitatory output from the mPFC, but definite evidence that DOI might exert such an effect on projection neurons was lacking. However, in previous studies DOI was applied locally whereas here we examined its systemic effect on cell firing, which raises the possibility that, in addition to local (mPFC) 5-HT2A receptors, other 5-HT2A-rich cortical areas (Pazos et al., 1985) which project to mPFC may also be involved.
The DOI-mediated increase of pyramidal activity involved 5-HT2A receptors since they were reversed by M100907. However, M100907 failed to antagonize the excitatory effect of DOI in some units, which may indicate the additional involvement of pyramidal 5-HT2C receptors (Clemett et al., 2000) or the inability of M100907 to reverse a full-blown activation of pyramidal neurons once initiated.
Interestingly, the electrical stimulation of the DR also elicited orthodromic excitations in pyramidal neurons which were blocked by M100907, showing that endogenous 5-HT can activate pyramidal neurons through 5-HT2A receptors in vivo. In most cases examined (18/19) the excited neurons projected to the DR, which indicates the existence of a strong reciprocal mPFC-DR interplay involving 5-HT2A receptors. The DR-induced excitations of pyramidal neurons were evoked at a stimulation rate (0.9 Hz), close to the firing rate of 5-HT neurons (typically 1 Hz in the anesthetized rat), which suggests that prefrontal excitations may occur following the physiological activation of DR neurons by incoming excitatory stimuli. Interestingly, the activation of prefrontal neurons elicited by a delayed response task in monkey prefrontal cortex was reversed by M100907, indicating the involvement of this receptor in pyramidal excitations (Williams et al., 2002).
The activation of 5-HT2A receptors may increase the efficacy of AMPA-mediated inputs onto pyramidal neurons (see below). Among other possible sources of glutamate, it could arise from both serotonergic and non-serotonergic neurons of the DR projecting to mPFC which express the vesicular glutamate transporter 3 (VGLUT3) (Gras et al., 2002). Indeed, glutamate was identified as a co-transmitter with 5-HT in cell culture (Johnson, 1994). However, the latency of the DR-evoked excitations is greater than that reported for glutamatergic axons (Maurice et al., 1999; Celada et al., 2001) and in most instances, the excitations were reversed by the 5-HT2A antagonist M100907. However, we cannot fully exclude that excitations were produced by co-release of 5-HT and glutamate from serotonergic axons in mPFC.
There is a substantial overlap in rat cortex between 5-HT axon terminals and 5-HT2 receptors (Blue et al., 1988). However, the cellular localization of 5-HT2A receptors responsible for the excitatory effects of 5-HT and DOI is unclear. 5-HT2A receptors are mainly localized to pyramidal neurons and GABAergic interneurons in mPFC (Willins et al., 1997; Jakab and Goldman-Rakic, 1998, 2000; Martín-Ruiz et al., 2001). The latter are large and medium-size parvalbumin- and calbindin-containing interneurons involved in the feed-forward inhibition of pyramidal neurons (Jakab and Goldman-Rakic, 2000). In the rat frontoparietal cortex, 5-HT axons have been observed that run in parallel with the apical dendrites of pyramidal neurons expressing 5-HT2A receptors (Jansson et al., 2001). Additionally, a lower proportion of 5-HT2A receptors was found pre-synaptically (Jakab and Goldman-Rakic, 1998; Miner et al., 2003). The activation of 5-HT2A receptors by 5-HT or 5-HT2A/2C agonists (DOI, DOB) exerts complex effects on prefrontal neurons. Thus, the microiontophoretic application of DOI suppressed the firing activity of presumed pyramidal neurons in anesthetized rats but enhanced the excitatory effect of glutamate at low ejection currents (Ashby et al., 1989, 1990). In vitro recordings of identified pyramidal neurons in prefrontal slices has revealed that 5-HT2A receptor activation increases spontaneous EPSCs and depolarizes the recorded cells (Araneda and Andrade, 1991; Tanaka and North, 1993; Aghajanian and Marek, 1997, 1999, 2000; Zhou and Hablitz, 1999). In addition, 5-HT can elicit 5-HT2A-mediated IPSCs through the activation of GABA synaptic inputs (Zhou and Hablitz, 1999). Hence, the inhibitory responses elicited by systemic DOI in the present study may be attributable to the activation of 5-HT2A receptors in GABA interneurons that locally control pyramidal cell activity.
The excitatory effects of DOI appear to involve interaction with glutamatergic transmission. As mentioned above, DOI could increase the excitatory effects of glutamate on prefrontal neurons (Ashby et al., 1989, 1990). Likewise, the 5-HT2A receptor-mediated EPSCs evoked by 5-HT in prefrontal slices are canceled by blockade of AMPA receptors and mGluR II receptor activation (Aghajanian and Marek, 1997, 1999). Moreover, the modulation of prefrontal NMDA transmission by 5-HT and DOB appears to involve pre- and postsynaptic 5-HT2A receptors (Arvanov et al., 1999). Our observation that the selective mGluR II agonist LY-379268 reversed the excitatory effect of DOI on pyramidal neurons in vivo is consistent with these in vitro observations. Also, the local increase of prefrontal 5-HT release elicited by DOI was blocked by NBQX (AMPA/KA antagonist) and 1S,3S-ACPD (mGluR II/III agonist) and was mimicked by AMPA application and blockade of glutamate reuptake (Martín-Ruiz et al., 2001). However, the NMDA receptor antagonist MK-801 failed to reverse the effect of DOI on 5-HT release.
In the present study, we used the disinhibition of the CM and MD nuclei of the thalamus which project densely to the mPFC (Berendse and Groenewegen, 1991; Fuster, 1997; Van der Werf et al., 2002) to examine whether an increase in thalamocortical inputs to the mPFC would mimic the excitatory effects of DOI on cell firing and 5-HT release. We also sought to examine whether the action of DOI implied an increase in thalamocortical excitatory inputs through the activation of terminal 5-HT2A receptors. Thalamic disinhibition increases c-fos expression in mPFC (Erdtsieck-Ernste et al., 1995; Bubser et al., 1998) by enhancing excitatory inputs from mediodorsal afferents that terminate on dendritic spines of pyramidal dendrites (Kuroda and Price, 1991; Kuroda et al., 1998). Previous studies had determined that stimulation of the MD nucleus increases AMPA inputs onto prefrontal pyramidal neurons (Pirot et al., 1994). Consistent with this synaptic connectivity, the thalamic application of bicuculline dramatically enhanced the firing rate of pyramidal neurons. This effect was reversed by the mGluR II agonist LY-379268, which reduces glutamate release (Cartmell and Schoepp, 2000).
The thalamic disinhibition also increased the prefrontal 5-HT release in dual probe experiments performed in freely moving and anesthetized rats, paralleling the effects on pyramidal cell firing. This effect was reversed by the application in mPFC of agents that reduce/block glutamatergic transmission, such as 1S,3S-ACPD, LY-379268 and NBQX. This suggests that an increased excitatory input onto mPFC neurons may activate descending excitatory pathways to the DR, which eventually results in an enhanced 5-HT release in mPFC (see scheme in Fig. 11). The stimulatory effect of the thalamic disinhibition on prefrontal 5-HT release was also reversed by the application in mPFC of DAMGO, a µ-opioid agonist. The DAMGO-induced reduction is likely accounted for by a presynaptic suppression of excitatory inputs onto mPFC since µ-opioid receptors suppress inhibitory and excitatory transmission in brain (Jolas and Aghajanian, 1997; Lüscher et al., 1997). Also, cortical µ-opioid receptors are considered to be present on thalamocortical afferents due to the mismatch between the receptor protein (mainly in cortex) and its encoding mRNA (mainly in thalamus) (Delfs et al., 1994; Zastawny et al., 1994). In agreement, lesions of the thalamic nuclei projecting to mPFC reduced cortical [3H]DAMGO binding (Marek et al., 2001).
|
The parallel effects of DOI and thalamic disinhibition (e.g. both increased 5-HT release and pyramidal output) raised the possibility that DOI might act by activating the release of glutamate from thalamocortical afferents, as observed for 5-HT in the in vitro slice preparation (Aghajanian and Marek, 1997
, 1999). However, this view is discordant with several observations from the present study. First, the application of DAMGO and LY-379268 in mPFC fully and partly reversed, respectively, the 5-HT increase produced by thalamic disinhibition but failed to attenuate the effect of DOI on 5-HT release. The partial suppression of 5-HT release elicited by LY-379268 may seem at variance with the full reversal of the effect of the thalamic disinhibition on pyramidal cell firing, given the ability of this agent to reduce presynaptic glutamate release (Schoepp et al., 1999; Cartmell and Schoepp, 2000). However, LY-379268 increases 5-HT release by itself (Cartmell et al., 2001) (see also Fig. 7C), an action which may mask the 5-HT decrease elicited by reducing excitatory inputs. The dissimilar effects of 1S, 3S-ACPD and LY-379268 on the thalamic disinhibition paradigm and on DOI’s effects may indicate the involvement of mGluR receptors other than mGluR II in the effect of 1S,3S-ACPD (Martín-Ruiz et al., 2001). The lack of any effect of DAMGO on the DOI-stimulated 5-HT release (but reversal of the effects of thalamic disinhibition) also suggests that the effect of DOI does not involve an increase in excitatory inputs from thalamocortical afferents.
Second, the effects of DOI on 5-HT release and pyramidal cell firing were unaffected by lesions of the thalamic nuclei that project to mPFC (Berendse and Groenewegen, 1991; Van der Fuster, 1997; Werf et al., 2002). Chemical lesions (NMDA) applied at the coordinates used herein partly attenuated the EPSCs induced by 5-HT in layer V pyramidal neurons of the mPFC (Marek et al., 2001). Electrolytic lesions caused an extensive brain damage of the centromedial, mediodorsal, paracentral and paratenial nuclei, which project densely to subdivisions of the mPFC, and other thalamic nuclei in their vicinity (anteromedial, centrolateral, parafascicular, paraventricular and reuniens nuclei). The lesion probably affected ascending fibers from other thalamic nuclei as well. Eight days after the electrolytic lesions, the ability of DOI to modulate pyramidal cell firing and 5-HT release in mPFC was unaltered. This cannot be attributed to an insufficient tissue damage, since all animals included showed extensive loss of thalamic tissue. Also, 1 week appears to be a sufficient time to allow for degeneration of existing terminals in mPFC since electrolytic lesions performed at this time post-lesion altered the thalamic influence on a variety of brain areas, including the prefrontal cortex (Raos and Savaki, 1995; Raos et al., 1995). Moreover, electrolytic lesions of the MD nucleus performed 4 days before markedly affected spatial memory, which suggests an impaired thalamocortical input (Stokes and Best, 1988). Interestingly, in our experiments, the baseline firing of pyramidal neurons in lesioned rats was one-third of that in untreated animals (0.51 versus 1.51 spikes/s), which likely reflects the marked loss of excitatory thalamic inputs.
Thus, on account of the above evidence, we suggest that DOI does not affect pyramidal cell firing (excitations and inhibitions) and 5-HT release by acting on 5-HT2A receptors putatively located on thalamocortical afferents to mPFC. This view is consistent with a recent study on the ultrastructural localization of 5-HT2A receptors in the rat mPFC. 5-HT2A receptor immunoreactivity was mainly found postsynaptically in dendritic shafts and spines (73% of labeled profiles) whereas the rest was located presynaptically on axons that do not display the typical morphology of glutamatergic axons (Miner et al., 2003).
In summary, the present study shows that the in vivo activation of 5-HT2A receptors by DOI has profound effects on pyramidal cell activity and, by virtue of descending mPFC-DR excitatory afferents, on 5-HT neurons. In turn, the electrical stimulation of the DR can elicit 5-HT2A-mediated excitations on pyramidal neurons. Taken together, these observations indicate the existence of a close functional relationship between the mPFC and the serotonergic system in rat brain (mPFC-DR circuit; Fig. 11). Conceivably, pyramidal 5-HT2A receptors mediate the excitatory effects of DOI whereas those present in GABA interneurons mediate inhibitory effects. Thalamic disinhibition resembled the excitatory effects of DOI, yet pharmacological differences (activation of µ-opioid receptors and mGluR II) suggest that DOI does not act by enhancing glutamatergic inputs from thalamocortical afferents, a view also supported by the absence of effects of the thalamic lesions on DOI’s action. However, since the excitatory effect of DOI appears to depend on glutamatergic inputs (Martín-Ruiz et al., 2001) (this study) it may involve a 5-HT2A-mediated synergism with AMPA inputs [for instance, increasing Ca2+ entry (Porter et al., 1999)] onto pyramidal neurons. These glutamatergic inputs may arise from different cortical or subcortical projections to the mPFC (Fuster, 1997), an issue deserving further investigation. The preservation of such inputs in vivo, even in rats with thalamic lesions may perhaps explain the apparent discrepancy between the present in vivo results and the in vitro observations by Aghajanian and Marek (Aghajanian and Marek, 1997, 1999, 2000) in the slice preparation. The present observations may help to clarify the neurobiological substrate of the action of hallucinogens and atypical antipsychotics, acting as agonists and antagonists, respectively, at cortical 5-HT2A receptors.
|
Notes |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussion< |
|---|