5-HT6 receptor antagonists improve performance in an attentional set shifting task in rats
- First Online:
- Cite this article as:
- Hatcher, P.D., Brown, V.J., Tait, D.S. et al. Psychopharmacology (2005) 181: 253. doi:10.1007/s00213-005-2261-z
- 256 Views
Rationale and objective
Performance on the Wisconsin Card Sorting Test (WCST), which requires patients to ‘shift attention’ between stimulus dimensions (sorting categories), is impaired in diseases such as schizophrenia. The rat attentional set shifting task is an analogue of the WCST. Given that 5-HT6 receptor antagonists improve cognitive performance and influence cortical neurochemistry in rats, the present study investigated the effects of 5-HT6 receptor antagonists upon attentional set shifting in rats.
Rats were tested in this paradigm following sub-chronic SB-399885-T or SB-271046-A (both 10 mg kg−1 bid, p.o. for 8 days prior to testing and either 4 or 2 h prior to testing on day 9, respectively). Rats were trained to dig in baited bowls for a food reward and to discriminate based on odour or digging media (Habituation, day 8). In a single session (day 9), rats performed a series of discriminations, including reversals (REV), intra-dimensional (ID) and extra-dimensional (ED) shifts.
Neither compound altered performance during Habituation. On the test day, both SB-399885-T and SB-271046-A reduced the total trials to reach criterion and the total errors made when data were collapsed across all discriminations (P<0.05–0.01). Further, both compounds significantly reduced the trials to criterion for REV-1 (P<0.05–0.01) and abolished the ID/ED shift. SB-399885-T, but not SB-271046-A, reduced trials required to complete the ED shift (P<0.05) and the number of errors made during completion of the ID (P<0.05) and ED shifts (P<0.01).
5-HT6 receptor antagonists improved performance in the attentional set shifting task and may have therapeutic potential in the treatment of disorders where cognitive deficits are a feature, including schizophrenia.
KeywordsSB-271046-ASB-399885-TIntra-dimensional shift (ID)Extra-dimensional shift (ED)Wisconsin Card Sorting Task (WCST)Cognition
Schizophrenic patients perform poorly on neuropsychological tests sensitive to frontal lobe function (Kolb and Wilshaw 1983; Pantelis et al. 1999). One such test used is the attentional set-shifting paradigm, the Wisconsin Card Sorting Test (WCST) (Berg 1948). In the WCST subjects are required to ‘shift attention’ between different stimulus dimensions (sorting categories). Schizophrenic patients achieve fewer sorting categories compared to controls and make significantly more perseverative errors (Kolb and Wilshaw 1983; Pantelis et al. 1999). However, the underlying cognitive deficit remains unclear, as successful performance of the WCST requires motivation, attentional memory and learning processes and/or intact frontal lobe functioning. It is unknown whether schizophrenic patients have the same cognitive deficits as, for example, patients with specific brain lesions. In an attempt to separate the different cognitive components of the task, two types of set-shift have been described. The intra-dimensional (ID) shift involves transfer of a rule within the same sorting category, whereas the extra-dimensional (ED) shift involves the transfer of attention across different sorting categories. Thus, the ID shift is related to the ability of the subject to be aware of the conceptual category in which they are responding, whereas the ED shift is the basis for the achievement of novel sorting categories (Pantelis et al. 1999).
The procedure (perceptual attentional set shifting) used here is analogous to the WCST used in man (Birrell and Brown 2000) and followed the development of a primate version (Roberts et al. 1988). The test has been adapted such that rats are trained to dig in bowls for a hidden food reinforcement. The bowls are presented in pairs, only one of which is baited. In this task, the rat is required to select the bowl in which to dig by either its odour or the digging medium that fills it. In a single test session, rats perform a series of discriminations, including acquisition of the discrimination, reversals, ID and ED shifts, in a manner that mirrors the WCST.
Birrell and Brown (2000) showed that lesions of rat medial prefrontal cortex resulted in a selective impairment of the ED shift, with no impairment in acquisition of the task or reversal learning. This provides evidence that the rat medial frontal cortex is not just concerned with spatial tasks or working memory, but that the deficits reflect impairment of an attentional system. In contrast, lesion of the orbital prefrontal cortex leaves attentional set shifting intact, but impairs reversal learning (McAlonan and Brown 2003). Other reports, using a variant of this paradigm, describe specific impairment of the ED shift in aged rats and following lesion of the posterior parietal cortex (Barense et al. 2002; Fox et al. 2003).
The 5-HT6 receptor is one of the 14 known receptors for 5-HT (see Barnes and Sharp 1999). 5-HT6 receptors are localised in the CNS to cerebral cortex, striatum, olfactory tubercle, nucleus accumbens, hypothalamus and hippocampus, with more moderate expression in such areas as the thalamus and substantia nigra (Hamon et al. 1999; Gerard et al. 1996, 1997; Ruat et al. 1993; Ward et al. 1995). The physiological role of 5-HT6 receptors is still not fully understood, although these receptors have been implicated in the modulation of CNS acetylcholine function. For example, antisense oligonucleotides directed against 5-HT6 receptors induce yawning, chewing and stretching behaviours, which are blocked by the muscarinic antagonist atropine (Bourson et al. 1995; Bentley et al. 1999). Further, the selective 5-HT6 receptor antagonist, RO 04-6790, increased yawning and stretching behaviours, an effect that was reversed by atropine and scopolamine (Bentley et al. 1999), and reversed scopolamine-induced rotation in 6-OHDA lesioned rats (Bourson et al. 1998). A second selective 5-HT6 receptor antagonist, SB-271046A, did not induce yawning when administered alone, but enhanced physostigmine-induced yawning (Routledge et al. 2000). Further, a representative 5-HT6 receptor antagonist from a novel chemical series of aryl pyridyl sulfones reversed a scopolamine-induced deficit in passive avoidance (after acute and subchronic dosing) and caused a twofold increase in frontal cortex acetylcholine (Ach) (Riemer et al. 2003).
Together with the high level of expression of 5-HT6 receptors in areas such as the hippocampus and cortex, these data provide evidence for a role of the 5-HT6 receptor in memory processes (see review by Woolley et al. 2004). Indeed, administration of antisense oligonucleotide directed against the 5-HT6 receptor (Bentley et al. 1997) and several selective 5-HT6 receptor antagonists, RO 04-6790 (Woolley et al. 2000), and SB-271046-A (Rogers and Hagan 2001) have been shown to enhance retention, but not acquisition, of a spatial learning task in the rat (Morris Water Maze). Further, 5-HT6 receptor antagonists also decrease scopolamine-induced deficits in an object recognition task in rats (Woolley et al. 2000) and passive avoidance in mice (Riemer et al. 2003).
Some atypical antipsychotic drugs also have affinity for the 5-HT6 receptor (e.g. clozapine, pKi=7.96; olanzapine, pKi=7.95; sertindole, pKi=9.05) (Roth et al. 1994), leading to the suggestion that this receptor may be important in the response to these drugs (Roth et al. 2004). Clozapine down-regulates 5-HT6 receptor density in HeLa cells in vitro (Zhukovshaya and Neumaier 2000). Furthermore, 14 days of treatment with clozapine, but not haloperidol, causes down-regulation of 5-HT6 receptor mRNA in the hippocampus, but not in the striatum, nucleus accumbens, olfactory tubercle or limbic cortex of rats (Frederick and Meador-Woodruff 1999). The difference in the effect of the two antipsychotics upon 5-HT6 receptor mRNA and the high levels of 5-HT6 receptors in areas where antipsychotics are thought to exert their effect, has led to the investigation of 5-HT6 receptors and their contribution to the therapeutic properties of these drugs. Indeed, the relatively higher affinity for 5-HT6 receptors may explain some of the beneficial effects of clozapine upon cognitive impairments in schizophrenia (e.g. see Meltzer and McGurk 1999; Roth et al. 2004), which include improved attention and verbal fluency and moderate evidence for improved executive function.
SB-271046-A (Bromidge et al. 1999; Routledge et al. 2000) and SB-399885-T are brain penetrant and highly selective 5-HT6 receptor antagonists (Hirst et al., personal communication). (N-[3,5-Dichloro-2-(methoxy)phenyl]-4-(methoxy)-3-(1-piperazinyl)benzenesulfonamide (SB-399885) has high affinity for human 5-HT6 receptors (pKi 9.1) and shows over 200-fold selectivity over all other receptors, ion channels and enzymes tested to date. Like SB-271046-A, subchronic administration of SB-399885 improved retention, but not acquisition, of the rat water maze task. Further, SB-399885 reversed a scopolamine-induced deficit in a novel object recognition paradigm and increased extracellular dopamine and acetylcholine levels in the medial prefrontal cortex (Hirst et al., personal communication).
Therefore, the aim of the present study was to investigate the effects of subchronic administration of these two selective 5-HT6 receptor antagonists on the performance of rats in the attentional set shifting task. A preliminary account of this work has been published in abstract form (Hatcher et al. 2002).
Materials and methods
Male Hooded Lister rats (Charles River UK Ltd.), weighing 250–300 g at the start of the experiment, were used. Animals were housed in groups of four, in a temperature-controlled environment (21±1°C) and maintained on a 12-h light/dark cycle (lights on 07:00-19:00 hours). Rats were fed a restricted diet (SDS Witham UK) for at least 7 days before testing began and throughout the experiment (15 g rat−1 day−1), and given ad libitum access to water.
Experiments were carried out in accordance with the United Kingdom Animals (Scientific Procedures) Act, 1986 and conformed to GlaxoSmithKline ethical guidelines.
The apparatus and methods are a modified version of those described by Birrell and Brown (2000) and consisted of a modified opaque plastic rat housing cage (40×70×18 cm) with Plexiglas panels used to divide one third of the length of the cage into two sections (choice area). The digging bowls were placed in these sections with a solid divider between them. Inserted into the Plexiglas panel were two independent removable dividers which separated the rat (start area) from the two sections where the bowls were placed (left and right side, respectively). The rat was given access to these bowls by lifting the divider(s) which could also be used to block access to the bowls, such as in training or when an incorrect choice was made.
Habituation and simple discrimination training
On the day before testing, rats were habituated to the testing area and trained to dig in the bowls filled with sawdust to retrieve the food reward (quarter of a honey nut loop, Kelloggs, UK). Rats were given free access to two sawdust-filled bowls, containing approximately ten 1/4 loops, in the start box of the testing cage for approximately 30 min. After this time bowls were placed in the choice area and baited with a food reward. The left or right divider was raised (chosen randomly) and the rat allowed to retrieve the food reward. Once the rat had eaten the reinforcement, the divider was lowered to force the rat to leave the choice area. Alternation between shutters continued until the rat was reliably digging in both bowls to retrieve the rewards (minimum of six trials with successful retrieval on each side).
Rats were then trained on a series of two simple discriminations (SDs) (odour of lemon or rose shavings; digging medium of pebbles or cardboard pieces) to a criterion of six consecutive correct trials. The first four trials at the beginning of each discrimination were discovery trials (not included in six trials to criteria), where the rat was allowed to dig in both bowls to discover which was the rewarded odour/media. As soon as the rat had explored one bowl, the divider was lowered so that the rat was forced to investigate the other bowl. This occurred irrespective of which bowl was investigated first. In subsequent trials, an incorrect choice terminated the trial. The test continued until a criterion of six consecutively correct trials had been met. Rats were allowed to approach, sniff bowls and climb over bowls, but only an attempt to dig was scored as a response. All rats were trained to the same discriminations in the same order. These exemplars were not used again on the test day.
Order of discriminations
Intradimensional (ID) Shift
Extradimensional (ED) Shift
Exemplars used (presented in pairs within each dimension)
Smooth Twinnings tea (M1)
Rough Twinnings tea (M2)
Wood shavings (M4)
Shredded white paper (M5)
Shredded brown paper (M6)
N-(3,5-Dichloro-2-methoxyphenyl)-4-methoxy-3-piperazin-1-yl-benzenesulfonamide tosylate (SB-399885-T) and 5-chloro-3-methylbenzo[b]thiophene-2-sulfonic acid (4-methoxy-3-piperazin-1-ylphenyl) amide hydrochloride (SB-271046-A) (both Medicinal Chemistry, GlaxoSmithKline) were tested in two separate experiments. Preliminary studies showed that acute dosing with SB-271046 did not improve performance in this task (D. Tait, unpublished data), therefore SB-399885-T, SB-271046-A (both 10.0 mg kg−1 p.o.; 2 ml kg−1 b.w.) or vehicle (1% methyl cellulose water) were dosed bi-daily for 8 days, with habituation to the test on day 8. This dosing regime was similar to that used in previous water maze studies (e.g. Rogers and Hagan 2001). On day 9 rats were dosed with SB-399885-T (pretreatment time 4 h) or SB-271046-A (pretreatment time 2 h) and tested in the set shifting task. All doses are expressed as mg kg−1 free base. Doses/pre-treatment times were chosen from previous in vivo behavioural studies and pharmacokinetic data (Bromidge et al. 1999; Routledge et al. 2000; Hirst et al., personal communication).
The data for total trials to criteria, number of errors and time per test trial were recorded for each rat for each discrimination. To correct for heterogeneity of variance in the treatment groups, data was either log transformed, logit transformed or arcsine transformed. Number of trials to criteria was logit transformed and number of errors were arcsin transformed as proportions of, respectively, (1) the maximum number of trials to criteria for the study, or (2) the actual number of trials to criteria made in each discrimination period. Data were analysed using a repeated measures ANOVA with two factors, one within subjects (test: SD, CD, REV-1, ID, REV-2, ED, REV-3) and one between subjects (treatment: drug or vehicle). Post-hoc planned comparisons were made using LSD tests. One-sided LSD tests were used where only a change in one direction was of scientific interest (tests of increase for reversals, and test of decrease in drug treatment from vehicle). Two-sided LSD tests were used for all other comparisons. In addition, total trials to criteria, total errors and average trial times (collapsed across all discriminations) were analysed by Student’s t test.
SB-399885-T did not affect any of the performance measures during habituation and simple discrimination training (data not shown).
Trials to criteria
Data from the separate stages of the task are shown in Fig. 1. SB-399885-T did not influence performance at SD, CD and ID shift stages compared with vehicle-treated rats. All rats required more trials to criterion for REV-1 compared to CD (P<0.01), although this was reduced by SB-399885-T treatment (P<0.01 compared with vehicle-treated controls).
Vehicle-treated rats required more trials to reach criteria during the ED shift compared to the ID shift (P<0.01), but this effect was absent in SB-399885-T-treated rats. Further, SB-399885-T significantly decreased the number of trials to criterion during the ED shift compared to vehicle-treated animals (P<0.05).
Number of errors
SB-399885-T decreased the total number of errors collapsed across all discriminations when compared to vehicle-treated rats (P<0.01; Fig. 1, inset).
Analysis of the stages SD, CD and REV-1 showed that performance was not significantly different between vehicle and SB-399885-T-treated rats. However, SB-399885-T significantly decreased the number of errors made during both the ID [from 1.75±0.5 (Veh) to 0.63±0.3, P<0.05] and ED [from 3.5±0.4 (Veh) to 1.0±0.3, P<0.01] shifts compared to vehicle-treated rats.
Mean time per test trial (s)
Mean trial time (±SEM) in test trials (s)
SB-271046 did not affect any of the performance measures during habituation and simple discrimination training (data not shown).
Trials to criteria
Analysis of the stages SD, CD, ID or ED shift show that performance was not significantly different between vehicle- and SB-271046-A-treated animals. All rats required more trials to reach criteria for REV-1 when compared to CD (P<0.01), although this was reduced by SB-271046-A (P<0.05 compared with vehicle-treated controls). Vehicle-treated rats required more trials to reach criteria during the ED shift compared with the ID shift (P<0.01), but this effect was not observed in rats treated with SB-271046-A.
Number of errors
SB-271046-A decreased the total number of errors collapsed across all discriminations when compared to vehicle-treated animals (P<0.05, Fig. 2 inset).
Analysis of all the individual test stages showed that performance was not significantly different between vehicle and SB-271046-A-treated animals. All rats made more errors during the ED shift than the ID shift: the increase in the number of errors made was significant in vehicle-treated rats [from 0.4±0.2 (ID) to 5±0.7 P<0.01], but missed significance in SB-271046-A-treated rats [from 0.5±0.3 (ID) to 2.1±0.7, P=0.06].
Table 3 shows the mean time spent on test trials collapsed across all discriminations. SB-271046-A did not influence the mean time per test trial.
The aim of the present study was to investigate the effects of subchronic dosing with the selective 5-HT6 receptor antagonists, SB-399885-T and SB-271046-A, on perceptual attentional set shifting in the rat. This test, an analogue of the WCST, requires rats to shift attention between different sorting categories to test conceptual awareness (i.e. awareness of the sorting category in which they are responding) and the achievement of novel sorting categories (i.e. ability to switch attention to a new stimulus dimension or cognitive flexibility). Both selective 5-HT6 receptor antagonists improved the overall performance of this task, and specifically improved reversal learning and attentional set-shifting.
Rats readily learnt to discriminate food bowls on the basis of odour and the digging medium with which the bowls were filled and also learnt to inhibit digging in unbaited bowls. In these studies, the first reversal (REV-1) required the most trials to reach criteria and consequently, also made the largest numbers of errors.
Control rats also require significantly more trials to reach criteria for the ED shift than for the ID shift, suggesting the requirement of an attentional set shift (“ID-ED shift”). Reversal learning does not require attention to be shifted, because the dimensions and exemplars remain the same. The previously correct exemplar is now incorrect and the rat must respond to the previously non-reinforced exemplar. Continuing to respond to the previously correct exemplar (perseveration) impairs reversal learning. As such, reversal learning assesses the animal’s awareness of the sorting category in which it is responding and its ability to switch responses to the relevant stimulus. All rats required more trials to complete the reversals than for either initial acquisition (SD and CD stages) or the ID shift. This effect was most marked during REV-1, where both SB-399885-T and SB-271046-A significantly decreased the number of trials to criteria when compared to vehicle-treated rats. Further, SB-399885-T decreased the number of errors made during REV-1. It may be that 5-HT6 receptor antagonists specifically influence reversal learning. However, as rats require the largest number of trials to reach criterion in this part of the paradigm, REV-1 may simply be the most sensitive to drug effects. The magnitude of the response to REV-2 and REV-3 was not as large as that observed for REV-1. This may be due to the novelty of REV-1 and reveals a “practice” component in this paradigm.
All vehicle-treated rats required significantly more trials to reach criteria for the ED shift compared to the ID shift. This effect was absent in drug-treated rats, i.e. rats responded in a similar way during both the ID and ED components of the procedure, suggesting that 5-HT6 receptor antagonists improve attentional set shifting in rats. Both SB-399885-T and SB-271046-A decreased the number of trials to criteria during the ED shift when compared to vehicle-treated rats, although this effect was only significant in rats treated with SB-399885-T. A similar effect was observed with the number of errors made during the ED shift.
Several studies have shown that the shifting of rules and attentional set is impaired by lesions of the medial prefrontal cortex (Ragozzino et al. 1999; Birrell and Brown 2000). The performance enhancing effects of the 5-HT6 receptor antagonists in this task (and others) may be attributed to their influence upon ACh and dopamine (DA) in the medial prefrontal cortex (Riemer et al. 2003; Hirst et al., personal communication). In the study by Hirst et al. an acute dose of SB-399885-T (10 mg kg−1) which increased cortical Ach was ineffective at increasing DA levels, contrasting with the effects of lower doses. In this study, SB-399885-T caused a clear trend towards elevated noradrenaline (NA) at all doses. Recent studies by Lacroix et al. (2004) demonstrate significant increases in extracellular NA and DA in the rat medial prefrontal cortex, complementing the findings of Hirst et al. (personal communication). There is considerable evidence supporting the role for both of these neurotransmitter systems, within the medial prefrontal cortex and other regions, in the modulation of cognitive function (reviewed by Jay 2003; Pare 2003). However, the effects of subchronic twice daily dosing with a 5-HT6 receptor antagonist upon cortical ACh, DA and NA are currently unknown.
Repeated dosing with SB-271046-A and SB-399885-T, albeit for 40 days, increased the number of cells positive for the polysialylated form of the neural cell adhesion molecule (PSA-NCAM) in the dentate gyrus of rats (Regan et al. 2003). Similar effects are evident after only 8 days of treatment (Regan, unpublished data). PSA-NCAM has been implicated in the support of structural reorganizations and synaptic plasticity, which may accompany memory consolidation. These authors suggest that the cognition enhancing effects of 5-HT6 receptor antagonists may be mediated in part by facilitation of activity-dependent synapse remodelling in the hippocampus.
The present report provides further evidence for a role for selective 5-HT6 receptor antagonists to improve cognitive function. In contrast, Lindner et al. (2003) failed to detect any effects of Ro 04-6790 or SB-271046 upon acquisition of an autoshaping task, scopolamine-induced deficits in contextual fear-conditioning, or retention of a water maze task. However, there is an increasing body of evidence generated in a number of different laboratories with structurally different compounds supporting the role of 5-HT6 receptors in cognitive function (see review by Woolley et al. 2004). For example, Ro 04-6790 reverses a scopolamine-induced deficit in learning consolidation in an autoshaping task (Meneses 2001), improves retention of the Morris water maze task (Woolley et al. 2001) and reverses a scopolamine-, but not a raclopride-, induced deficit in a rodent test of recognition memory (Woolley et al. 2003). Another 5-HT6 receptor antagonist, SB-357134 improved water maze retention (Rogers and Hagan 2001) and more recently, SB-271046 has been shown to reverse scopolamine-impaired consolidation of a passive avoidance task in mature rats and subchronic dosing improves spatial task deficits in the Morris water maze in aged rats (Foley et al. 2004).
In summary, subchronic administration of two separate 5-HT6 receptor antagonists, SB-271046-A and SB-399885-T, improved performance of an attentional set shifting task, which suggests increased cognitive flexibility. These findings suggest a potential role for 5-HT6 receptor antagonist to treat cognitive impairments seen in disorders such as schizophrenia. Future studies will evaluate the effects of antipsychotics with differing affinities for 5-HT6 receptors upon performance of this task.