International Journal of Neuropsychopharmacology (2012), 15, 1473–1487. f CINP 2011
doi:10.1017/S1461145711001635
ARTICLE
Impulsive action and impulsive choice are
mediated by distinct neuropharmacological
substrates in rat
Neil E. Paterson, Caitlin Wetzler, Adrian Hackett and Taleen Hanania
PsychoGenics Inc., 765 Old Saw Mill River Road, Tarrytown, USA
Abstract
Impulsivity is a heterogeneous construct according to clinical and preclinical behavioural measures
and there is some preliminary evidence indicating distinct neurobiological substrates underlying the
sub-components of impulsivity. Two preclinical assays, the five-choice serial reaction time task (5-CSRTT)
and the delayed discounting task (DDT), are hypothesized to provide measures of impulsive action
(premature responding) and impulsive choice (percent choice for delayed reward), respectively. In the
present studies, we show that the norepinephrine reuptake inhibitor atomoxetine attenuated premature
responding in the 5-CSRTT, but was ineffective in the DDT. The mixed dopamine/norepinephrine
reuptake inhibitor methylphenidate exhibited an opposite profile of effects. In addition, blockade of
5-HT2A/C receptors via ketanserin decreased premature responding but had no effects on percent choice
for delayed reward ; blockade of 5-HT2C receptors via SB 242084 had opposite effects. Follow-up studies
provided some limited evidence of additive effects of 5-HT2A/C receptor blockade on the effects of
atomoxetine on impulsive action. These studies demonstrate dissociable profiles of stimulant vs. nonstimulant attention deficit hyperactivity disorder medications and 5-HT subtype-selective ligands, in the
5-CSRTT and DDT assays. Thus, the present findings support the sub-categorization of impulsivity and
suggest that 5-HT receptor subtype-selective antagonists may provide therapeutic targets for disorders
characterized by different forms of impulsivity.
Received 13 June 2011 ; Reviewed 21 July 2011 ; Revised 12 September 2011 ; Accepted 30 September 2011 ;
First published online 18 November 2011
Key words : Atomoxetine, 5-CSRTT, delayed discounting task, methylphenidate, serotonin.
Introduction
Impulsivity is currently thought to be a heterogeneous
construct that comprises two subcomponents : impulsive action and impulsive choice (see Winstanley et al.
2006 for a review of animal models of impulsivity).
There is increasing evidence for dissociable neurobiological substrates underlying these constructs. The
current studies explore the neuropharmacological
substrates of impulsive action and impulsive choice,
each of which can be assessed with distinct translational assays. Premature responding in the five-choice
serial reaction time task (5-CSRTT), which has been
described in detail elsewhere (e.g. Bari et al. 2008), is
Address for correspondence : Dr N. E. Paterson, Behavioral
Pharmacology, PsychoGenics, Inc., 765 Old Saw Mill River Road,
Tarrytown, NY 10591, USA.
Tel. : 914-406-8058 Fax : 914-406-8090
Email : [email protected]
hypothesized to provide a measure of impulsive
action, i.e. a failure to withhold an inappropriate
response (e.g. Bizarro et al. 2004 ; Day et al. 2007). The
5-CSRTT is a preclinical analogue of the continuous
performance test in humans, which requires subjects
to monitor a succession of stimuli and respond only
when a specific stimulus is presented and to withhold
responding in the absence of a stimulus (see Young
et al. 2009 for a recent adaptation of the rodent
5-CSRTT). The delay discounting task (DDT) is a
cross-species task used to assess impulsive choice,
defined as intolerance of reward delay (Winstanley
et al. 2006), where high impulsivity detected in the task
is thought to be a trait of attention deficit hyperactivity
disorder (ADHD) patients and other patient populations. In animals, the DDT has been used to assess
preclinical efficacy of ADHD medications and additional explorations of neuropharmacological substrates of impulsivity.
1474
N. E. Paterson et al.
Atomoxetine (Strattera1 ; Lilly, USA), a norepinephrine transporter (NET) inhibitor (Wong et al.
1982), attenuates impulsivity in preclinical studies
(for review, see Eagle & Baunez, 2010) and is an efficacious ADHD medication (Chamberlain et al. 2007 ;
Michelson et al. 2003). Specifically, atomoxetine was
effective in 5-CSRTT (Blondeau & Dellu-Hagedorn,
2007 ; Navarra et al. 2008 ; Paterson et al. 2011 ;
Robinson et al. 2008), DDT (Robinson et al. 2008) and
the stop-signal reaction time task (Bari et al. 2009) in
rats. The anti-impulsive effects of atomoxetine may be
due to enhanced cortical noradrenergic and/or dopaminergic neurotransmission (Bymaster et al. 2002 ;
Swanson et al. 2006). The mixed dopamine (DA)/
norepinephrine (NE) reuptake inhibitor methylphenidate (Ritalin1 ; Novartis, Switzerland ; Han & Gu,
2006) increased NE and DA overflow in the prefrontal
cortex and striatum, unlike atomoxetine, which has no
striatal effects (Bymaster et al. 2002) and is an effective
ADHD medication (Shiels et al. 2009 ; Simpson &
Plosker, 2004). It is the distinct effects of atomoxetine
and methylphenidate on striatal DA (and perhaps NE)
that results in non-stimulant vs. stimulant effects, respectively. Based on selective neurochemical effects
in striatum and contrasting behavioural properties
in the 5-CSRTT (Navarra et al. 2008 ; Paterson et al.
2011), methylphenidate and atomoxetine can be distinguished as stimulant and non-stimulant ADHD
medications. Although Robinson et al. (2008) reported
efficacy of atomoxetine in the 5-CSRTT and the DDT,
we hypothesized that stimulant and non-stimulant
medications may exhibit different efficacy profiles in
5-CSRTT vs. DDT.
In addition to NE and DA, serotonin has also been
implicated in aspects of impulsivity. Specifically, administration of the 5-HT2C/B antagonist SER-082
was effective in decreasing impulsivity in the 5-CSRTT
but not the DDT ; the 5-HT2A/C antagonist ketanserin
exhibited opposite effects (Passetti et al. 2003 ;
Ruotsalainen et al. 1997 ; Talpos et al. 2006). Earlier
work exploring the effects of 5-HT2 subtype-selective
ligands on impulsive behaviours indicated that
5-HT2A blockade attenuated impulsivity (Fletcher
et al. 2007 ; Higgins et al. 2003 ; Winstanley et al. 2004),
whereas 5-HT2C blockade increased premature responding in the 5-CSRTT (Fletcher et al. 2007 ;
Winstanley et al. 2004). Most relevant to the 5-CSRTT
task variant used in the present study is the recent
finding that 5-HT2A blockade attenuated high levels
of premature responding observed under prolonged
inter-trial interval (ITI) (9 s) conditions (Fletcher et al.
2011). The current studies sought to replicate the
effects of ketanserin in a variable ITI 5-CSRTT and
DDT and extend the effects of the selective 5-HT2C
antagonist SB 242084 from the 5-CSRTT to the DDT in
rats.
In summary, impulsivity can be subdivided into
several distinct sub-components that, based on current
preclinical evidence, likely exhibit some degree of
distinct and shared neurobiological substrates. Based
on previous studies with ADHD medications and
selective 5-HT receptor subtype ligands, the present
studies assessed the effects of atomoxetine, methylphenidate, ketanserin and SB 242084 in the 5-CSRTT,
hypothesized to provide a measure of impulsive action, and the DDT, hypothesized to provide a measure
of impulsive choice. Based on the observed dissociable
effects of these ligands, follow-up studies determined
whether co-administration of 5-HT receptor subtype
ligands and NE/DA reuptake inhibitors yielded
additive effects on therapeutic efficacy.
Method
Subjects
Male Long–Evans rats (275–300 g) were obtained
from Harlan Laboratories (USA). Upon arrival, the
rats were assigned unique identification numbers (tail
marked). Rats were single-housed in standard or
OptiRAT cages (Animal Care Systems Inc., USA)
and acclimated for 7 d prior to commencing a food
restriction regimen : rats were held at 85 % of agematched free-feeding control body weights, receiving
approximately 10–20 g of rat chow daily. Water was
provided ad libitum, except during testing. Animals
were maintained on a 12-h light/dark cycle (lights on
07 : 00 hours EST) with room temperature maintained
at 22¡2 xC and the relative humidity maintained
at approximately 50 %. All animals were examined,
handled and weighed prior to initiation of the study
(during the week of habitation) to ensure adequate
health and suitability and to minimize non-specific
stress associated with testing. All efforts were made to
minimize discomfort of any sort at all times during the
conduct of the studies. Behavioural test sessions were
performed during the animal’s light cycle phase. All
experiments were approved by the Institutional
Animal Care and Use Committee of PsychoGenics,
Inc. (USA) in AAALAC-accredited facilities and in
accordance with the Guide to the Care and Use of
Laboratory Animals (NIH, 2010).
Apparatus
The test apparatus consisted of 10 aluminium and
Plexiglas chambers with grid floors (width 31.5 cm,
Pharmacologic dissociation of impulsivity subcategories
depth 25.0 cm, height 33.0 cm), housed in soundattenuating cabinets (Med Associates, USA). Each
cabinet was fitted with a low-level noise extractor fan.
The left wall of each chamber was concavely curved,
with five apertures evenly spaced, located approximately 2.5 cm from the floor. Each aperture contained
a standard 3 W LED to serve as stimulus lights. The
opposite wall contained a food magazine, located approximately 3.0 cm from the floor. A retractable lever
(approximately 2 cm long) was positioned on either
side of the food magazine (approximately 2 cm from
the grid floor). Each chamber was illuminated with a
1.5 W house-light located above the food magazine,
1 cm from the chamber ceiling. After each test session,
the apparatus was cleaned with 70 % ethanol ; at least
20 min elapsed before the next test session. The test
session functions were controlled and all data were
captured, by a computer using K-Limbic software
(Conclusive Software Ltd, distributed by Med
Associates, USA).
Experimental procedures
5-CSRTT
Animals were trained to monitor the five apertures for
stimulus light illumination. Each session was initiated
by the illumination of the house-light and the delivery
of a food reward into the magazine. The first trial
began when the rat nose-poked into the magazine to
obtain the food pellet. After the ITI (5 s duration), one
of the stimulus lights was illuminated for 500 ms. The
rat should nose-poke in the illuminated aperture
either during or within 5 s of the stimulus light
illumination. Such a response was defined as a correct
response and was rewarded with delivery of a food
pellet. Collection of the pellet initiated the next trial.
A nose-poke response in a non-illuminated aperture
(incorrect response) or a nose-poke after the 5 s limited
hold (missed trial) resulted in termination of the trial
with extinction of the house-light and imposition of
a time-out period (5 s duration). Nose-pokes made
during the ITI (premature responses) and nose-pokes
made after the first response to the stimulus presentation (perseverative response) were punished by imposition of the time-out ; premature responses resulted
in a reset of the ITI. Rats took 40–60 sessions to acquire
the task.
After acquisition of the 5-CSRTT [>70 % correct,
<30 % omissions, minimum 50 trials completed
(i.e. excluding omissions) per session for five consecutive sessions], drug testing began. During variable
ITI test sessions, the ITI was 10, 7, 5 or 4 s in duration
(presented in banks of four trials ; each ITI duration
1475
was presented once in a randomized order within each
bank) for 60 min or 200 trials. The criterion of 70 %
correct, rather than the more commonly used 80 %,
was utilized in order to increase the dynamic range of
the test and thereby allow the detection of attentionenhancing effects of pharmacological manipulations.
Measures obtained during the test sessions were :
(1) percent correct, defined as the number of correct
trials r100, divided by the total number of correct and
incorrect trials ; (2) percent omissions, defined as
responding beyond the 5 s limited hold or failing
to respond, expressed as the number of missed
trials r100 divided by the total number of trials ;
(3) premature responding, defined as the total number
of nose-poke responses made during the ITI ; (4) perseverative responding, defined as the total number of
additional responses emitted after the initial nosepoke within a single trial ; (5) correct latency, defined
as the time taken to make a correct response after the
illumination of the stimulus ; (6) incorrect latency, defined as the time taken to make an incorrect response
after the illumination of the stimulus ; (7) magazine
latency, defined as the time taken to enter the magazine to collect the food pellet after making a correct
response.
DDT
Animals were trained to respond on one lever for
a large (three pellets) reward delivered at variable
delays (0, 10, 20 and 40 s) after pressing the lever and
to respond on a different lever for a small (one pellet)
reward delivered immediately after the lever press.
Specifically, sessions began with illumination of the
house and magazine light. Subjects initiated all trials
by making a nose-poke into the magazine. A maximum of 10 s was allowed for nose-poking, after which
time the house and magazine lights were extinguished
and a missed trial was recorded. Trials were presented
in four distinct blocks of 12 trials : the first four trials of
each block were forced trials (i.e. only one lever was
presented ; each lever was presented twice) and the
final eight trials were free choice trials (i.e. both levers
were presented). A maximum of 10 s was allowed for
lever pressing, after which time the lever(s) were
retracted and a missed trial was recorded. In free choice
trials, the rats received either a single or multiple
(three) food pellets after making a lever press, according to the designation of the lever as immediate
(single) or delayed (multiple) reward. Delays for the
delayed/large reinforcer progressively increased
across the four trial blocks as follows : 0, 10, 20 and
40 s delay. The designation of left/right lever as
1476
N. E. Paterson et al.
delayed/immediate reward was counter-balanced
across subjects. Each trial was separated by a 60 s ITI,
during which the house-light was extinguished.
Sessions were terminated after 48 trials (approximately 60 min).
After the rats exhibited a stable baseline performance, drug testing was initiated. Stable baseline
performance was determined by analysis of variance
(ANOVA) of two contiguous three-session blocks indicating no main or interaction effect of the threesession block and a significant main effect of delay
with significant post-hoc comparisons between long
and short delays (Robinson et al. 2008). The measures
obtained during DDT sessions were : (1) percent preference for the large reinforcer as a function of delay :
calculated as the number of choices for the large
reinforcer/(total number of choices for large+small
reinforcers)r100 ; calculated for each specific delay
duration ; (2) percent omissions : failures to respond
when the magazine was illuminated or the lever was
presented : calculated as the number of omissions/
(total number of correct responses+omissions)r100
(forced and choice trials) ; (3) response latency : the
time to lever press from the extension of the levers
(for choice trials only) ; (4) magazine latency : the
time from food pellet delivery to collection of the food
pellet from the magazine (for all trials).
Data analyses
Data were analysed via one- (study defined as
a within-subjects factor) or two-way ANOVAs (ITI/
delay duration and dose defined as within-subjects
factors). Where the assumption of equal variance was
violated, data were transformed to the square root
prior to ANOVA. ANOVAs were followed, when appropriate, by Student Newman–Keuls post-hoc tests.
Results
Effects of repeated testing on baseline performance
levels
For animals that were used in more than one study,
baseline performance measures prior to the initiation
of each study were compared via one- or two-way
ANOVA.
5-CSRTT
ANOVA revealed no effects of repeat testing, except
for a significant increase in perseverative responding
in the baseline prior to SB 242084 (13.2¡2.7 responses)
vs. atomoxetine (7.9¡1.3 responses : F1,9=6.20,
p<0.05) for the 10 rats that were used in both studies.
For the seven rats used in three studies, there were no
significant differences in any performance measures
during baseline pre-study testing.
Drugs
Methylphenidate hydrochloride was obtained from
Sigma-Aldrich, USA. Tomoxetine hydrochloride,
ketanserin and SB 242084 were obtained from Tocris
Biosciences, USA. Test compounds were dissolved in
sterile saline and administered i.p. in a volume of
1 ml/kg (doses expressed as salt form). Atomoxetine,
methylphenidate, ketanserin and SB 242084 were administered with pre-treatment times of 30 min. In the
co-administration studies (expt 5), test compounds
were administered immediately after each other in
two discrete injections. In all studies, all rats received
all drug treatments, according to a randomized order,
counter-balanced and within-subjects design (‘ Latin
square’ design). The 5-CSRTT studies were performed
in nine to 11 rats and the DDT studies were performed
in 10–11 rats. In all cases, animals entered a study only
if they exhibited stable baseline performance that met
the acquisition criteria specified above. For the
5-CSRTT studies, rats were used for one to three experiments. For the DDT studies, rats were used in two
to five experiments. Drug tests were performed on
Wednesdays and Fridays of each week, when subjects
met pre-determined performance criteria.
DDT
ANOVA indicated no differences in any performance
measures for the rats that were used in repeated
studies (six rats – two DDT studies ; five rats – three
DDT studies ; three rats – four studies ; three rats – five
studies). For all groups, there was a significant effect of
delay (F3,15=27.79, p<0.001 ; F3,12=128.58, p<0.001 ;
F3,6=38.46, p<0.001 ; F3,6=50.44, p<0.001 ; respectively).
Expt 1 : effects of atomoxetine, methylphenidate,
ketanserin or SB 242084 in variable ITI 5-CSRTT
Atomoxetine
Premature responding decreased as the ITI duration
decreased (ITI : F3,24=102.61, p<0.001) and atomoxetine attenuated the high levels of premature responding observed at longer ITI trials (doserITI :
F9,72=2.53, p<0.05 ; dose : F3,24=10.73, p<0.001). Posthoc tests indicated that 0.5, 1.0 and 2.0 mg/kg atomoxetine significantly decreased premature responding compared to vehicle at 10, 7 and 5 s ITI trials
(Fig. 1 a). Percent accuracy (F3,24=5.18, p<0.01), correct
Pharmacologic dissociation of impulsivity subcategories
(b)
Premature responses
120
100
Saline
ATX 0.5 mg/kg
ATX 1.0 mg/kg
ATX 2.0 mg/kg
80
n=9
60
*
*
*
40
20
0
120
Premature responses
(a)
* * *
7
100
80
n = 10
60
40
20
* *
*
10
Vehicle
MP 0.5 mg/kg
MP 1.0 mg/kg
MP 2.0 mg/kg
MP 4.0 mg/kg
5
0
4
10
7
ITI duration (s)
(d)
120
Saline
Ket 0.1 mg/kg
Ket 0.3 mg/kg
Ket 1.0 mg/kg
100
80
60
n = 11
*
*
40
20
0
*
10
7
5
4
ITI duration (s)
Premature responses
Premature responses
(c)
1477
120
Vehicle
SB 0.3 mg/kg
SB 1.0 mg/kg
SB 3.0 mg/kg
100
80
*
*
n = 10
*
60
40
* *
20
*
5
4
0
10
7
5
4
ITI duration (s)
ITI duration (s)
Fig. 1. Effects of atomoxetine (ATX), methylphenidate (MP), ketanserin (Ket) or SB 242084 (SB) on premature responding in
the variable inter-trial interval (ITI) five-choice serial reaction time task. The graphs depict the effects of ATX (a), MP (b), Ket
(c) or SB (d ) on premature responding. Data are expressed as mean¡S.E.M. * p<0.05 indicates significant differences compared
to vehicle at specific ITI durations.
(F3,24=17.75, p<0.001) and incorrect (F3,24=24.06,
p<0.001) response latencies increased as the ITI
duration decreased. Percent omissions decreased as
the ITI decreased (F3,24=8.68, p<0.001). Atomoxetine
had no effects on any measures other than premature
responding (Table 1).
p<0.05) and methylphenidate dose (F4,36=4.17,
p<0.01) in a nonlinear fashion. There were no effects
of ITI on magazine latency or perseverative responding. There were no effects of methylphenidate dose on
any measure other than correct response latency
(Table 2).
Methylphenidate
Ketanserin
Premature responding decreased as the ITI duration
decreased (F3,27=42.67, p<0.001), but there was no
effect of methylphenidate on premature responding.
Post-hoc tests indicated that premature responding
was significantly higher at 10 and 7 s ITI trials compared to data collapsed across 5 and 4 s trials (Fig. 1 b).
As ITI duration decreased, percent correct (F3,27=9.25,
p<0.001) and incorrect (F3,27=9.07, p<0.001) response
latency increased ; percent omissions decreased
(F3,27=20.71, p<0.001). Correct response latency
varied as a function of ITI duration (F3,27=3.25,
Ketanserin (dose : F3,30=5.63, p<0.01) attenuated
higher levels of premature responding observed at
the longer ITI durations (ITI : F3,30=109.75, p<0.001 ;
ITIrdose : F9, 90=2.65, p<0.01). Post-hoc tests indicated
that premature responding at 10 and 7 s ITI was significantly attenuated by administration of ketanserin
at 0.3 and 1.0 mg/kg compared to vehicle (Fig. 1 c).
Percent correct was decreased at 10 and 7 s ITI durations compared with 4 s ITI durations (F3,30=4.86,
p<0.01). Ketanserin (F3,30=2.69, p<0.05) increased
accuracy across all ITI trials ; post-hoc tests indicate that
1478
N. E. Paterson et al.
Table 1. The effects of atomoxetine on performance in the variable ITI 5-CSRTT
ITI
Measure
0 mg/kg
0.5 mg/kg
1.0 mg/kg
2.0 mg/kg
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
Pct Corr
Pct Corr
Pct Corr
Pct Corr
Pct Omis
Pct Omis
Pct Omis
Pct Omis
Pers Resp
Pers Resp
Pers Resp
Pers Resp
CRL
CRL
CRL
CRL
IRL
IRL
IRL
IRL
ML
ML
ML
ML
72.70¡5.19#
70.09 ¡3.10#
83.38¡2.87
80.05¡4.06
29.97¡6.27#
18.04¡5.22
11.29¡3.78
13.34¡4.34
3.56¡1.14
4.89¡0.77
2.89¡0.63
3.89¡0.90
0.69¡0.06
0.63¡0.07#
0.75¡0.06
0.88¡0.07#
1.54¡0.22#
1.26¡0.12#
1.57¡0.25
2.42¡0.24
1.37¡0.07
1.98¡0.28
1.51¡0.10
1.71¡0.28
72.81¡4.23#
72.93¡4.34#
80.76¡2.62
78.68¡2.38
30.53¡7.27#
17.58¡4.63
12.16¡2.40
17.77¡2.90
7.11¡4.18
3.89¡0.81
4.00¡0.71
4.44¡1.43
0.75¡0.07
0.60¡0.05#
0.70¡0.04
0.86¡0.05#
1.03¡0.16#
1.44¡0.15#
2.16¡0.28
2.59¡0.21
1.61¡0.15
2.41¡0.60
2.40¡0.45
1.70¡0.17
75.38¡4.99#
78.67¡3.09#
83.38¡3.00
79.92¡2.08
37.45¡7.19#
15.53¡4.27
16.14¡3.66
21.97¡11.07
6.33¡1.26
2.89¡0.72
3.44¡0.84
4.56¡1.07
0.71¡0.07
0.64¡0.05#
0.82¡0.06
0.90¡0.12#
1.23¡0.25#
1.31¡0.16#
2.42¡0.19
2.33¡0.22
2.38¡0.61
1.81¡0.18
1.89¡0.20
3.18¡1.47
81.20¡3.53#
80.98¡3.90#
81.32¡3.03
81.91¡2.70
34.21¡7.27#
17.91¡3.83
16.24¡3.93
28.48¡7.00
3.67¡1.08
2.44¡0.63
2.22¡0.49
3.89¡0.86
0.66¡0.06
0.69¡0.05#
0.83¡0.04
0.91¡0.08#
1.11¡0.21#
1.59¡0.20#
2.16¡0.24
2.18¡0.17
2.02¡0.35
1.63¡0.08
2.19¡0.33
1.83¡0.10
ITI, Inter-trial interval ; 5-CSRTT, five-choice serial reaction time task ; Pct Corr, Percent correct ; Pct Omis, percent omissions ;
Pers Resp, perserverative responding ; CRL, correct latency ; IRL, incorrect latency ; ML, magazine latency.
Data are presented as mean¡S.E.M.
#
p<0.05 indicates significant differences compared to 5-s ITI trials, collapsed across treatment conditions.
1.0 mg/kg ketanserin increased accuracy compared to
vehicle (Table 3). Percent omissions increased as
the ITI duration decreased (F3,30=6.74, p<0.01), but
ketanserin had no effect on omissions. Mean correct
response latency varied as a function of ITI (F3,30=8.88,
p<0.001) and dose (ITIrdose interaction : F9, 90=2.40,
p<0.05). Post-hoc tests indicated that at 10 s ITI,
0.3 mg/kg ketanserin significantly increased correct
response latency compared to vehicle (Table 3). Mean
incorrect response latency increased as the ITI duration decreased (F3,30=19.51, p<0.001) ; post-hoc tests
indicated that 5 s and 4 s trials were associated with
significantly longer mean incorrect response latencies
compared with 10 s trials, irrespective of ketanserin
dose. Administration of ketanserin (F3,30=2.98, p<
0.05) prolonged mean incorrect latency ; post-hoc tests
indicated that 0.3 mg/kg ketanserin increased mean
incorrect response latency compared to vehicle, across
all ITI durations (Table 3). Neither perseverative responding nor magazine latency was affected by ITI or
ketanserin.
SB 242084
Premature responding was increased at longer
ITI durations (F3,27=34.12, p<0.001) and SB 242084
(F3,27=4.74, p<0.01) further increased levels of
premature responding at longer ITI values (F9,81=2.17,
p<0.05). Post-hoc tests indicated that premature responding was significantly increased after administration of 0.3, 1.0 and 3.0 mg/kg SB 242084 compared
to vehicle at 10 s ITI and after administration of 1.0 and
3.0 mg/kg compared to vehicle at 7 s ITI (Fig. 1 d ).
There were no effects of variable ITI or SB 242084 on
percent correct. Percent omissions increased as the ITI
duration decreased (F3,27=19.49, p<0.001) and there
was a significant ITIrSB 242084 dose interaction
effect (F9,81=2.89, p<0.01). Post-hoc tests indicated that
administration of 0.3 mg/kg SB 242084 significantly
decreased percent omissions compared to vehicle
at 10 s ITI (Table 4). Mean correct response latency
varied across ITI durations (F3,27=9.77, p<0.001) and
SB 242084 doses (F3,27=3.26, p<0.05). Post-hoc tests
Pharmacologic dissociation of impulsivity subcategories
1479
Table 2. The effects of methylphenidate on performance in the variable ITI 5-CSRTT
ITI
Measure
0 mg/kg
0.5 mg/kg
1.0 mg/kg
2.0 mg/kg
4.0 mg/kg
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
Pct Corr
Pct Corr
Pct Corr
Pct Corr
Pct Omis
Pct Omis
Pct Omis
Pct Omis
Pers Resp
Pers Resp
Pers Resp
Pers Resp
CRL
CRL
CRL
CRL
IRL
IRL
IRL
IRL
ML
ML
ML
ML
73.61¡3.53#
71.22¡2.70#
80.81¡3.31
79.07¡5.09
32.80¡4.36#
20.93¡3.64
11.44¡2.50
12.25¡1.88
4.80¡1.42
5.90¡1.56
5.80¡1.40
4.50¡1.15
0.67¡0.07
0.60¡0.04
0.62¡0.05
0.74¡0.05
1.16¡0.27
1.13¡0.17
1.57¡0.12
2.39¡0.21#
1.50¡0.14
1.85¡0.30
1.98¡0.32
1.85¡0.29
74.51¡3.65#
75.29¡3.43#
81.68¡3.25
80.99¡3.43
37.53¡5.57#
20.76¡3.58
9.33¡2.15
11.16¡2.30
5.90¡1.45
4.20¡0.89
4.40¡1.49
4.00¡1.14
0.69¡0.07
0.67¡0.04
0.66¡0.04
0.83¡0.07
1.74¡0.34
1.27¡0.10
1.70¡0.22
2.15¡0.21#
1.83¡0.21
1.55¡0.17
2.24¡0.58
2.12¡0.48
72.93¡3.64#
77.62¡3.16#
79.25¡2.96
81.68¡3.06
35.77¡5.60#
19.82¡4.09
11.81¡2.99
13.08¡2.04
4.10¡0.59
3.80¡0.73
3.90¡0.85
2.50¡0.85
0.72¡0.07
0.66¡0.04
0.71¡0.05
0.79¡0.07
1.31¡0.20
1.23¡0.16
1.86¡0.16
2.20¡0.13#
1.76¡0.41
1.41¡0.16
1.72¡0.33
1.69¡0.31
67.83¡3.89#
71.70¡2.87#
80.11¡4.49
84.83¡4.39
31.58¡7.18#
19.52¡4.28
10.56¡2.75
11.97¡4.01
4.90¡0.89
5.20¡0.98
3.40¡0.54
3.60¡1.05
0.73¡0.06
0.63¡0.03
0.67¡0.03
0.75¡0.05
1.59¡0.24
1.48¡0.15
1.39¡0.26
2.07¡0.28#
1.51¡0.21
1.97¡0.40
2.03¡0.38
1.91¡0.44
72.72¡3.46#
74.21¡2.60#
80.90¡2.55
84.22¡3.02
37.21¡5.49#
20.67¡4.40
11.14¡2.02
12.89¡2.12
6.30¡1.34
7.40¡2.13
6.20¡2.23
5.80¡2.29
0.65¡0.04
0.57¡0.05
0.63¡0.04
0.63¡0.04
1.86¡0.22
1.32¡0.19
1.58¡0.16
1.82¡0.28#
2.35¡0.68
2.11¡0.60
2.70¡0.91
2.13¡0.69
ITI, Inter-trial interval ; 5-CSRTT, five-choice serial reaction time task ; Pct Corr, Percent correct ; Pct Omis, percent omissions ;
Pers Resp, perserverative responding ; CRL, correct latency ; IRL, incorrect latency ; ML, magazine latency.
Data are presented as mean¡S.E.M.
#
p<0.05 indicates significant differences compared to 5-s ITI trials, collapsed across treatment conditions.
indicated that mean correct response latency was decreased at the 7 s ITI duration compared to all other
ITI values ; administration of 0.3 mg/kg SB 242084
decreased mean correct response latency compared
to vehicle, across all trials (Table 4). Mean incorrect
latency increased as the ITI duration decreased (F3,27=
19.24, p<0.001) ; there was no effect of SB 242084
administration. Post-hoc tests indicated that incorrect
latency at 4 s ITI trials was significantly increased
compared with all other trials. Neither ITI duration
nor SB 242084 administration altered perseverative
responding or magazine latency (Table 4).
Expt 2 : effects of atomoxetine, methylphenidate,
ketanserin or SB 240284 in DDT
Atomoxetine
Delay resulted in decreased preference for the large
reward (F3,30=131.35, p<0.0001). Atomoxetine had no
effects on percent preference for large reward as a
function of delay but there was a significant main
effect of dose (F3,30=4.33, p<0.05). Post-hoc tests
indicated that 1.0 mg/kg atomoxetine decreased tolerance for reward delay for data collapsed across all
delays (Fig. 2a). In addition, there were no effects of
atomoxetine on percent omissions, choice or magazine
latencies (Table 5).
Methylphenidate
Methylphenidate increased percent preference for
large reward and varied as a function of dose and
delay (dose : F3,30=6.62, p<0.01 ; delay : F3,30=156.23,
p<0.0001 ; doserdelay : F9, 90=3.13, p<0.01). Post-hoc
tests indicated that methylphenidate significantly
increased percent preference for large reward compared to vehicle at two intermediate delays, 10 s (1.0,
2.0 and 4.0 mg/kg methylphenidate) and 20 s (2.0 and
4.0 mg/kg methylphenidate). In addition, administration of 2.0 mg/kg methylphenidate significantly
increased large reward preference compared with
1.0 mg/kg methylphenidate (Fig. 2 b). Methylphenidate had no effects on percent omissions, choice or
magazine latencies (Table 5).
1480
N. E. Paterson et al.
Table 3. The effects of ketanserin on performance in the variable ITI 5-CSRTT
ITI (s)
Measure
0 mg/kg
0.1 mg/kg
0.3 mg/kg
1.0 mg/kg
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
Pct Corr
Pct Corr
Pct Corr
Pct Corr
Pct Omis
Pct Omis
Pct Omis
Pct Omis
Pers Resp
Pers Resp
Pers Resp
Pers Resp
CRL
CRL
CRL
CRL
IRL
IRL
IRL
IRL
ML
ML
ML
ML
67.75¡3.80
71.41¡3.01
77.17¡2.49
76.09¡2.74
33.12¡6.62
18.30¡2.88
12.13¡1.08
18.43¡2.2
3.46¡0.91
4.27¡1.13
3.55¡0.90
4.00¡0.81
0.59¡0.04
0.80¡0.10
0.70¡0.04
0.89¡0.08
1.37¡0.10
1.27¡0.07
1.93¡0.19
2.24¡0.21
1.54¡0.28
1.98¡0.45
1.59¡0.18
1.56¡0.20
72.40¡2.98
73.83¡3.67
80.20¡3.05
76.48¡2.94
34.30¡6.88
17.39¡3.54
12.24¡3.15
18.04¡3.15
3.00¡0.86
3.18¡0.55
1.91¡0.53
4.09¡1.38
0.69¡0.06
0.67¡0.06
0.78¡0.06
0.86¡0.05
1.37¡0.12
1.52¡0.18
2.11¡0.10
2.27¡0.21
2.48¡0.76
1.48¡0.14
1.73¡0.36
1.70¡0.27
73.74¡4.50
71.34¡2.63
79.52¡2.29
75.28¡3.24
28.59¡4.94
15.33¡3.18
16.77¡1.68
20.39¡3.17
3.36¡0.65
4.00¡0.73
4.09¡0.85
3.91¡1.15
0.79¡0.10**
0.69¡0.05
0.77¡0.07
0.97¡0.09
1.53¡0.18
1.66¡0.16
2.31¡0.29
2.27¡0.22
2.37¡0.60
2.82¡0.60
1.99¡0.59
1.82¡0.35
72.60¡3.67
74.98¡4.08
79.83¡2.27
83.18¡3.04
31.72¡5.74
19.38¡2.24
15.25¡2.11
25.49¡3.58
3.55¡1.08
2.91¡0.56
3.64¡0.79
4.00¡0.74
0.71¡0.05
0.64¡0.03
0.87¡0.05
0.91¡0.06
1.13¡0.14
1.58¡0.20
2.08¡0.19
2.57¡0.15
2.70¡0.73
2.08¡0.31
2.67¡0.90
2.59¡0.84
ITI, Inter-trial interval ; 5-CSRTT, five-choice serial reaction time task ; Pct Corr, Percent correct ; Pct Omis, percent omissions ;
Pers Resp, perserverative responding ; CRL, correct latency ; IRL, incorrect latency ; ML, magazine latency.
Data are presented as mean¡S.E.M.
**p<0.01 indicates a significant difference compared to vehicle-treated controls at a specific ITI value.
Ketanserin
Administration of ketanserin had no effects on percent
preference for delayed reward. Increased delay was
associated with decreased preference for large reward
(F3,27=147.66, p<0.0001 ; Fig. 2 c). Administration of
ketanserin had no effects on percent omissions or
magazine latency. Nonetheless, there was a significant
effect of ketanserin dose on choice latency (F3,27=2.97,
p<0.05) ; post-hoc tests indicated that 1.0 mg/kg
significantly increased choice latency compared to
vehicle (Table 5).
SB 242084
Delay resulted in decreased preference for the large
reward (F3,30=147.09, p<0.001). SB 242084 had no
effects on percent preference for large reward as a
function of delay (doserdelay interaction effect :
F9, 90=1.59, n.s.), but there was a significant main effect
of dose (F3,30=3.77, p<0.05). For data collapsed across
all delays, administration of 3.0 mg/kg SB 242084
significantly increased percent choice for large
reward compared to saline (Fig. 2 d). Administration of
SB 242084 had no effects on percent omissions or
magazine latencies, but significantly decreased choice
latencies (F3,30=5.57, p<0.01). Post-hoc tests indicated
that all doses of SB 242084 (0.3, 1.0 and 3.0 mg/kg)
significantly decreased choice latency compared to
saline (Table 5).
Expt 3 : the effects of 5-HT2A/C or 5-HT2C receptor
blockade on the anti-impulsive effects of atomoxetine
or methylphenidate in the 5-CSRTT and DDT
Co-administration of ketanserin and atomoxetine in the
5-CSRTT
Premature responding was increased at prolonged
ITI intervals (F3,27=132.31, p<0.001) and coadministration of ketanserin and atomoxetine attenuated the increased premature responding
Pharmacologic dissociation of impulsivity subcategories
1481
Table 4. The effects of SB 240284 on performance in the variable ITI 5-CSRTT
ITI (s)
Measure
0 mg/kg
0.3 mg/kg
1.0 mg/kg
3.0 mg/kg
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
Pct Corr
Pct Corr
Pct Corr
Pct Corr
Pct Omis
Pct Omis
Pct Omis
Pct Omis
Pers Resp
Pers Resp
Pers Resp
Pers Resp
CRL
CRL
CRL
CRL
IRL
IRL
IRL
IRL
ML
ML
ML
ML
75.44¡4.51
74.45¡3.51
81.63¡3.44
78.32¡3.21
45.64¡7.14
20.91¡3.21
15.97¡3.68
22.50¡4.11
5.60¡1.31
4.10¡0.57
4.90¡0.78
5.60¡1.42
0.85¡0.08
0.69¡0.04
0.75¡0.04
0.81¡0.06
1.14¡0.27
1.40¡0.16
1.87¡0.20
2.35¡0.17
1.89¡0.29
1.67¡0.18
1.84¡0.20
2.29¡0.42
72.10¡3.51
78.60¡4.57
80.70¡3.29
78.35¡2.79
34.22¡5.41**
27.33¡4.63
16.07¡3.30
20.75¡4.27
3.90¡0.85
5.70¡1.04
3.20¡0.59
7.80¡2.79
0.69¡0.05
0.60¡0.04
0.70¡0.05
0.65¡0.03
1.53¡0.20
1.20¡0.16
1.45¡0.12
2.28¡0.18
1.68¡0.30
1.76¡0.22
1.98¡0.29
1.76¡0.24
75.35¡4.10
74.84¡3.31
78.02¡3.90
78.99¡2.35
42.29¡5.22
24.41¡4.20
35.76¡2.41
18.03¡3.93
5.20¡1.18
4.50¡0.97
5.90¡0.46
5.60¡0.86
0.72¡0.06
0.62¡0.04
0.64¡0.04
0.80¡0.09
1.41¡0.18
1.39¡0.17
1.40¡0.15
1.89¡0.08
1.51¡0.15
2.03¡0.33
1.85¡0.28
1.55¡0.15
69.51¡5.78
71.91¡3.80
78.15¡3.58
76.17¡3.40
47.60¡6.72
25.40¡4.15
15.46¡3.39
16.76¡2.81
4.30¡0.76
4.20¡1.00
4.40¡0.75
5.20¡0.89
0.71¡0.07
0.59¡0.03
0.75¡0.09
0.79¡0.06
1.43¡0.17
1.20¡0.14
1.90¡0.17
2.02¡0.18
1.53¡0.14
1.29¡0.09
1.80¡0.24
1.45¡0.16
ITI, Inter-trial interval ; 5-CSRTT, five-choice serial reaction time task ; Pct Corr, Percent correct ; Pct Omis, percent omissions ;
Pers Resp, perserverative responding ; CRL, correct latency ; IRL, incorrect latency ; ML, magazine latency.
Data are presented as mean¡S.E.M.
**p<0.01 indicates a significant difference compared to vehicle-treated controls at a specific ITI value.
(F12,108=12.32, p<0.001 ; dose F4,36=29.59, p<0.001).
Post-hoc tests indicated that co-administration of
0.1 mg/kg ketanserin and all doses of atomoxetine
(including vehicle) decreased premature responding
compared to vehicle–vehicle controls at 10 s ITI.
Also, co-administration of 0.1 mg/kg ketanserin
plus 0.5–2.0 mg/kg atomoxetine was significantly
more effective than 0.1 mg/kg ketanserin plus vehicle.
Co-administration of 0.1 mg/kg ketanserin plus 1.0
or 2.0 mg/kg atomoxetine was significantly more
effective than 0.1 mg/kg ketanserin plus 0.5 mg/kg
atomoxetine, and 0.1 mg/kg ketanserin plus 2.0 mg/
kg atomoxetine was more effective than 0.1 mg/kg
ketanserin plus 1.0 mg/kg atomoxetine (Fig. 3 a).
At 7 s ITI, co-administration of 0.1 mg/kg ketanserin
plus 0.5–2.0 mg/kg atomoxetine significantly decreased premature responding compared with vehicle–
vehicle controls and compared with 0.1 mg/kg
ketanserin plus vehicle. Finally, co-administration of
0.1 mg/kg ketanserin plus 2.0 mg/kg atomoxetine
was significantly more effective than 0.1 mg/kg
ketanserin plus 0.5 mg/kg atomoxetine (Fig. 3 a).
Percent correct and percent omissions also changed as
a function of ITI and treatment (significant interaction
effects). Post-hoc tests indicated that at 7 s ITI, coadministration of 0.1 mg/kg ketanserin plus 1.0 mg/
kg atomoxetine significantly increased accuracy
compared with vehicle–vehicle controls (Table 6). At
5 s ITI, ketanserin plus 1.0 or 2.0 mg/kg atomoxetine
significantly decreased percent omissions compared
with vehicle–vehicle controls. At 4 s ITI, ketanserin
plus 0.5–2.0 mg/kg atomoxetine significantly decreased percent omissions compared with vehicle–
vehicle controls. Perseverative responding and correct
response latency changed as a function of ITI, with
significantly different correct response latencies at
10 s (shorter), 7 s (shorter) or 4 s (longer) ITI trials.
Significant effects of ITI and treatment were observed
for incorrect response latency. Post-hoc tests indicated
that incorrect response latencies were significantly
N. E. Paterson et al.
1482
(b)
Saline
ATX 0.5 mg/kg
ATX 1.0 mg/kg
ATX 2.0 mg/kg
Percent choice (large reward)
100
80
n = 11
60
40
20
Saline
MP 1.0 mg/kg
MP 2.0 mg/kg
MP 4.0 mg/kg
100
Percent choice (large reward)
(a)
**
80
**
n = 11
**
*
60
**#
40
20
**
0
0
0
20
10
Reinforcer delay (s)
40
Percent choice (large reward)
10
20
Reinforcer delay (s)
40
(d)
Saline
Ket 0.1 mg/kg
Ket 0.3 mg/kg
Ket 1.0 mg/kg
100
80
n = 10
60
40
20
100
Percent choice (large reward)
(c)
0
*
80
Saline
SB 0.3 mg/kg
SB 1.0 mg/kg
SB 3.0 mg/kg
n = 11
60
40
20
0
0
0
10
20
Reinforcer delay (s)
40
0
10
20
Reinforcer delay (s)
40
Fig. 2. Effects of atomoxetine (ATX), methylphenidate (MP), ketanserin (Ket) or SB 242084 (SB) on reward delay tolerance
in the delay discounting task. The graphs depict the effects of ATX (a), MP (b), Ket (c) or SB (d) on percent choice of large
reward across variable delays. Data are expressed as mean¡S.E.M. (a, d). * p<0.05, ** p<0.01 indicate significant differences
compared to vehicle for data collapsed across all delay intervals. (b) * p<0.05, ** p<0.01 indicate significant differences
compared to vehicle and # p<0.05 indicates significant differences compared to 1.0 mg/kg methylphenidate at specific
reward delays.
different at 4, 7 or 10 s compared with 5 s ITI trials ; coadministration of ketanserin and 2.0 mg/kg atomoxetine significantly prolonged latency compared
with vehicle–vehicle controls. Magazine latencies
were stable across different ITI trials and treatments
(Table 6).
Co-administration of SB 242084 and methylphenidate
in the DDT
Co-administration of SB 242084 and methylphenidate
significantly increased percent preference at specific
reward delay intervals (delayrtreatment interaction
effect : F12,120=1.85, p<0.05 ; treatment : F4,40=16.14,
p<0.001 ; delay : F3,30=123.42, p<0.001). Post-hoc tests
indicated that, compared with vehicle–vehicle, coadministration of SB 242084 (1 mg/kg) and methylphenidate (0, 1, 2 and 4 mg/kg) significantly increased
percent choice for delayed reward at 10 and 20 s
intervals. In addition, 1 mg/kg SB 242084 plus 1, 2 and
4 mg/kg methylphenidate significantly increased
percent choice for delayed reward compared with
vehicle–vehicle controls at the 40 s delay (Fig. 3 b). Coadministration of SB 242084 and methylphenidate had
no effects on percent omissions or magazine latency,
but significantly shortened choice latency (F4,40=9.57,
p<0.0001). Post-hoc tests indicated that all treatments
containing 1 mg/kg SB 242084 significantly decreased
choice latency compared with vehicle–vehicle controls
(Table 7).
Discussion
The present studies indicated dissociation between
the effects of stimulant and non-stimulant ADHD
Pharmacologic dissociation of impulsivity subcategories
1483
Table 5. The effects of atomoxetine (ATX), methylphenidate (MP), ketanserin and SB 242084 on performance in the delay
discounting task
Measure(s)
Drug
0 mg/kg
0.5 mg/kg
1.0 mg/kg
2.0 mg/kg
Percent omissions
Choice latency
Magazine latency
Percent omissions
Choice latency
Magazine latency
Percent omissions
Choice latency
Magazine latency
Percent omissions
Choice latency
Magazine latency
ATX
ATX
ATX
MP
MP
MP
Ketanserin
Ketanserin
Ketanserin
SB 242084
SB 242084
SB 242084
0.38¡0.38
0.75¡0.12
0.79¡0.15
0.0¡0.0
0.71¡0.09
0.64¡0.05
1.88¡0.85
0.80¡0.08
0.66¡0.11
1.89¡0.86
0.90¡0.10
1.19¡0.29
0.0¡0.0
0.76¡0.11
0.74¡0.10
0.0¡0.0
0.72¡0.09
1.20¡0.43
1.25¡0.71
0.88¡0.08
0.75¡0.17
0.57¡0.41
0.77¡0.07*
0.77¡0.18
0.76¡0.76
0.77¡0.08
0.65¡0.09
0.95¡0.76
0.72¡0.11
1.01¡0.28
2.71¡1.36
0.90¡0.08
0.81¡0.19
0.19¡0.19
0.76¡0.05*
0.88¡0.17
0.0¡0.0
0.85¡0.12
0.61¡0.08
0.0¡0.0
0.76¡0.11
0.99¡0.24
5.00¡3.12
0.95¡0.10*
0.76¡0.16
0.76¡0.58
0.69¡0.07**
0.99¡0.20
Data are presented as mean¡S.E.M.
* p<0.05, **p<0.01 indicate a significant difference from vehicle.
medications in two preclinical assays hypothesized to
measure two forms of impulsivity and also demonstrated a similar dissociation for 5-HT2 receptor subtype-selective ligands. Specifically, the selective
NET inhibitor atomoxetine attenuated high levels
of impulsive action observed in the ITI variant of the
5-CSRTT in rats, but decreased tolerance for reward
delay measured in the DDT. Blockade of 5-HT2A/C
receptors, via ketanserin, mimicked the effects of atomoxetine ; co-administration of both compounds may
have exerted additive effects in the 5-CSRTT. In contrast, the mixed DA/NE reuptake inhibitor methylphenidate had no effects on premature responding in
the 5-CSRTT, but increased tolerance for reward delay
in the DDT. The 5-HT2C antagonist SB 242084 mimicked the effects of methylphenidate. There was minimal evidence, in the current study, for additive effects
of the compounds in the DDT. These data point
toward distinct neuropharmacological substrates for
impulsive action (premature responding in 5-CSRTT)
and impulsive choice (reward delay tolerance in the
DDT) and suggest 5-HT2 receptor subtype selective
ligands may be potential medications for ADHD and
other disorders characterized by high impulsivity.
Previously, atomoxetine exhibited anti-impulsivity
properties in both 5-CSRTT and DDT in rats (Robinson
et al. 2008), in contrast with the present studies.
Comparison of the delay discounting data in the
vehicle-treated groups across these studies indicates
a more limited extent of delay discounting in the
Robinson study, with percent choice of delayed reward at approximately 50 % at 10 s delay and 30–35 %
from 20 to 60 s delays. By contrast, the present study
showed percent choice values of approximately 40, 20
and 10 % at 10, 20 and 40 s delays, respectively. Other
possible reasons for the discrepancy between the
DDT data in the two studies may be related to rat
strain, light cycle, housing or experimental procedural
differences (e.g. number of free/forced trials or the
delays used). It should be noted that the effects
of atomoxetine in Robinson et al. (2008) were limited to
a single intermediate dose (1.0 mg/kg) following a
simple main effect of dose. The effects of atomoxetine
in the 5-CSRTT were consistent, however, with published reports (Robinson et al. 2008), including studies
utilizing a variable ITI task as used here (Navarra et al.
2008). The observed effects of the 5-HT2A/C antagonist
ketanserin were also consistent with previous studies
(Passetti et al. 2003 ; Ruotsalainen et al. 1997 ; Talpos
et al. 2006). Comparison of the behavioural properties
of ketanserin against those of 5-HT2C and 5-HT2A
selective ligands (Fletcher et al. 2007 ; Higgins et al.
2003 ; Winstanley et al. 2004) indicates that it is likely
that the effects of ketanserin are mediated via 5-HT2A
receptors. Co-administration of atomoxetine with
ketanserin may have exerted additive effects in the
5-CSRTT. Specifically, the addition of ketanserin
(0.1 mg/kg) resulted in dose-dependent effects of atomoxetine, which exhibited a relatively flat dose–
response function when administered alone, at 10 s ITI
trials. Further, at 7 s ITI trials, the combination of atomoxetine and ketanserin was effective at the three
highest atomoxetine doses tested ; when tested in the
absence of ketanserin, only 2.0 mg/kg atomoxetine
N. E. Paterson et al.
1484
(a)
120
Saline-saline
Ket 0.1 mg/kg-ATX 0.0 mg/kg
Ket 0.1 mg/kg-ATX 0.5 mg/kg
Ket 0.1 mg/kg-ATX 1.0 mg/kg
Ket 0.1 mg/kg-ATX 2.0 mg/kg
Premature responses
100
80
n = 10
*
60
#
*
*
40
20
+
# ^
+
* #
*
*
*
# # +
* #
* ** *
*
0
10
7
5
ITI duration (s)
4
(b)
Percent choice (large reward)
100
**
80
Vehicle-vehicle
SB 1.0 mg/kg-MP 0.0 mg/kg
SB 1.0 mg/kg-MP 1.0 mg/kg
SB 1.0 mg/kg-MP 2.0 mg/kg
SB 1.0 mg/kg-MP 4.0 mg/kg
**
60
n = 11
**
**
40
**
*
* *
20
*
**
*
0
0
10
20
Reinforcer delay (s)
40
Fig. 3. The effects of 5-HT2A/C or 5-HT2C receptor
blockade on the anti-impulsive effects of atomoxetine or
methylphenidate. The graphs depict the effects of ketanserin
(Ket) or SB 242084 (SB) on the effects of atomoxetine
(ATX) in the five-choice serial reaction time task (a) or
methylphenidate in the delay discounting task (b). Premature
responding data (a) and percent choice of large reward data
(b) are expressed as mean¡S.E.M. * p<0.05, ** p<0.01 indicate
significant differences compared to vehicle–vehicle at specific
inter-trial interval (ITI) or reward delay values. There were
significant differences compared to co-administration of
Ket and 0.0 (# p<0.05), 0.5 (+ p<0.05) and 1.0 (^ p<0.05)
mg/kg atomoxetine at specific ITI values.
was effective and 0.1 mg/kg ketanserin was ineffective. 5-HT2A blockade selectively increased cortical
(Mørk et al. 2009) and not striatal (Uslaner et al. 2009)
DA. Thus, 5-HT2A blockade could enhance the antiimpulsivity effects of atomoxetine that are likely
mediated by enhanced NE or DA overflow in the
prefrontal cortex. Nonetheless, it should be noted that
ketanserin was effective at 10 s ITI in the second but
not the first study ; further, the second study lacks an
atomoxetine-only condition.
The effects of methylphenidate in the present study
are consistent with previous studies utilizing 5-CSRTT
and DDT. Methylphenidate has been reported previously to exhibit stimulant-like effects in 5-CSRTT
and other assays (Bizarro et al. 2004 ; Milstein et al.
2010 ; Navarra et al. 2008 ; Puumala et al. 1996). In the
present study, higher doses of methylphenidate
were associated with significantly increased tolerance
for delayed reward and non-significantly increased
premature responding in the 5-CSRTT. Similarly,
SB 242084 significantly increased tolerance for delayed
reward (significant main effect in expt 2 and significant effects of 1.0 mg/kg plus vehicle at 10 and 20 s
delays in expt 3) and significantly increased premature
responding in the 5-CSRTT (1.0 and 3.0 mg/kg at
10 and 7 s ITI trials). It should be pointed out that,
although 1.0 mg/kg SB 242084 is reliably selective for
5-HT2C (e.g. Kennett et al. 1997 ; Quérée et al. 2009), it is
possible that 3.0 mg/kg SB 242084 may have some effects on additional targets, such as 5-HT2A receptors.
The 5-HT2C/B antagonist SER-082 was effective in decreasing impulsivity in the 5-CSRTT but not the DDT
(Talpos et al. 2006). There was some limited evidence
for additive effects of methylphenidate and SB 242084
in the DDT, although a shift in the decay curve for the
saline condition, coupled with the significant effects of
1.0 mg/kg SB 242084 in the follow-up vs. the initial
study, preclude any firm conclusions. Mechanistically,
it is possible that SB 242084 facilitated methylphenidate-induced enhanced DA signalling in the striatum
in a manner similar to that observed for the psychostimulant cocaine (Navailles et al. 2004). The antiimpulsivity profiles of methylphenidate and SB 242084
are matched by d-amphetamine (Wiskerke et al. 2011)
and it is hypothesized that the dopaminergic effects of
d-amphetamine are responsible (van Gaalen et al.
2006), most likely in the striatum. Thus, it appears
likely that the DDT is sensitive to altered DA overflow
in the striatum.
The present studies demonstrated differential
effects of blockade of DA or NE reuptake, 5-HT2C or
5-HT2A/C receptors, on premature responding in the
5-CSRTT and tolerance to reward delay in the DDT,
hypothesized to reflect distinct facets of impulsivity.
Previous studies identified distinct dopaminergic
pathways for impulsive action and impulsive choice
(Pattij & Vanderschuren, 2008 ; Winstanley et al. 2006).
More recently, naloxone was shown to attenuate
amphetamine-induced increased impulsive action, but
had no effects of amphetamine-induced attenuation
of impulsive choice (Wiskerke et al. 2011). In utero
nicotine exposure was also shown to disrupt 5-CSRTT
performance in adult rats, including increased
Pharmacologic dissociation of impulsivity subcategories
1485
Table 6. The effects of co-administration of ketanserin (ket) and atomoxetine (ATX) on performance in the variable ITI 5-CSRTT
ITI (s)
Measure
Ket
0.0+ATX
0+0
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
10
7
5
4
Pct Corr
Pct Corr
Pct Corr
Pct Corr
Pct Omis
Pct Omis
Pct Omis
Pct Omis
Pers Resp
Pers Resp
Pers Resp
Pers Resp
CRL
CRL
CRL
CRL
IRL
IRL
IRL
IRL
ML
ML
ML
ML
70.30¡4.44
71.03¡3.22
79.98¡4.33
76.73¡4.46
31.47¡5.71
18.30¡3.01
14.94¡2.49
18.32¡3.94
5.30¡1.33
3.90¡0.99
5.20¡0.89
5.70¡1.57
0.68¡0.08#
0.63¡0.05#
0.82¡0.08
0.90¡0.12#
1.35¡0.18#
1.36¡0.15#
1.71¡0.25
2.50¡0.22#
1.74¡0.27
1.58¡0.20
1.68¡0.26
2.00¡0.33
Ket
0.1+ATX
0.0 mg/kg
Ket
0.1+ATX
0.5 mg/kg
Ket
0.1+ATX
1.0 mg/kg
Ket
0.1+ATX
2.0 mg/kg
73.16¡4.23
75.66¡3.56
82.36¡3.00
76.53¡3.40
36.80¡6.03
20.47¡2.83
15.02¡3.67
22.52¡5.81
3.40¡0.52
3.20¡0.83
2.90¡0.78
4.20¡1.03
0.67¡0.05#
0.72¡0.04#
0.75¡0.05
0.96¡0.09#
1.09¡0.21#
1.50¡0.15#
2.23¡0.23
2.76¡0.19#
1.45¡0.07
1.96¡0.49
1.78¡0.29
1.92¡0.40
76.40¡2.33
79.49¡2.90
79.78¡3.10
73.94¡4.22
29.78¡6.20
17.01¡4.19
22.85¡4.97
31.17¡6.65**
3.00¡0.47
3.20¡0.61
2.80¡0.49
5.60¡1.04
0.64¡0.05#
0.74¡0.04#
0.85¡0.06
1.13¡0.11#
1.30¡0.13#
1.71¡0.14#
2.26¡0.15
2.58¡0.21#
2.40¡0.72
1.72¡0.23
1.77¡0.20
2.32¡0.51
74.58¡2.50
84.73¡2.49**
81.30¡2.24
69.38¡3.37
28.31¡5.72
22.63¡4.98
29.76¡5.30**
41.93¡6.29**
3.80¡0.85
3.20¡0.73
4.00¡1.26
4.70¡0.98
0.64¡0.03#
0.64¡0.03#
0.87¡0.07
0.97¡0.10#
1.62¡0.14#
1.58¡0.22#
2.08¡0.24
2.55¡0.23#
2.00¡0.31
2.57¡0.85
2.12¡0.50
1.60¡0.14
71.10¡1.72
80.33¡2.42
77.44¡4.35
80.83¡4.21
29.23¡4.55
23.02¡5.49
36.24¡6.43**
43.09¡6.07**
2.90¡0.55
2.0¡0.67
4.20¡0.73
4.10¡0.80
0.81¡0.10#
0.76¡0.08#
0.88¡0.10
0.91¡0.08#
1.59¡0.18##
2.21¡0.34##
2.59¡0.34#
2.44¡0.31##
2.70¡0.73
2.15¡0.64
2.30¡0.39
3.03¡1.18
ITI, Inter-trial interval ; 5-CSRTT, five-choice serial reaction time task ; Pct Corr, Percent correct ; Pct Omis, percent omissions ;
Pers Resp, perserverative responding ; CRL, correct latency ; IRL, incorrect latency ; ML, magazine latency.
Data are presented as mean¡S.E.M.
**p<0.01 indicates significant differences compared to vehicle–vehicle control group, within specific ITI values.
#
p<0.01 indicates a significant difference compared to 5-s ITI trials, collapsed across treatment conditions.
#
p<0.01 indicates a significant difference compared to vehicle–vehicle control, collapsed across all ITI values.
Table 7. The effects of co-administration of SB 242084 (SB) and methylphenidate (MP) performance in the delay discounting task
Measure(s)
0 mg/kg
SB+0.0 mg/kg
MP
1 mg/kg
SB+0.0 mg/kg
MP
1 mg/kg
SB+1.0 mg/kg
MP
1 mg/kg
SB+2.0 mg/kg
MP
1 mg/kg
SB+4.0 mg/kg
MP
Percentt omissions
Choice latency
Magazine latency
0.19¡0.19
0.81¡0.09
0.52¡0.05
0.57¡0.29
0.60¡0.06**
0.81¡0.22
0.38¡0.38
0.65¡0.09**
0.71¡0.16
0.19¡0.19
0.57¡0.06**
0.93¡0.28
0.0¡0.0
0.61¡0.07**
0.686¡0.15
Data are presented as mean¡S.E.M.
**p<0.01 indicates a significant difference compared to vehicle–vehicle controls.
premature responding, but had no effects in the DDT
(Schneider et al. 2011). Nonetheless, morphine increased both forms of impulsivity (Pattij et al. 2009).
Overall, the present studies add to the existing
evidence that impulsive action and impulsive choice
are mediated by distinct neuropharmacological substrates. Interestingly, 5-HT2C and 5-HT2A receptor
blockade was previously shown to worsen or enhance
1486
N. E. Paterson et al.
reversal learning, a measure of compulsivity that requires inhibition of a previously learned response
(Boulougouris et al. 2008). Thus, the dissociable effects
of 5-HT2A and 5-HT2C blockade demonstrated in
studies of impulsivity, including the current work, can
be extended to another type of behaviour.
In summary, the present studies demonstrate dissociable effects of stimulant and non-stimulant ADHD
medications, and 5-HT2 receptor subtype-selective
ligands, on premature responding in the 5-CSRTT and
tolerance to reward delay in the DDT, hypothesized to
reflect distinct sub-categories of impulsivity. These
results add to a body of work, specifically the novel
demonstration of the effects of SB 242084 in the DDT
and the differential effects of atomoxetine in the DDT
and 5-CSRTT, that supports the sub-categorization of
impulsivity based on distinct neurobiological substrates. Finally, the observed effects of 5-HT2 subtypeselective antagonists suggests that the 5-HT2A and
5-HT2C receptors may be valuable targets for the development of novel anti-impulsivity medications that
could potentially be used in conjunction with existing,
approved ADHD medications.
Acknowledgements
None.
Statement of Interest
None.
References
Bari A, Dalley JW, Robbins TW (2008). The application of
the 5-choice serial reaction time task for the assessment of
visual attentional processes and impulse control in rats.
Nature Protocols 3, 759–767.
Bari A, Eagle DM, Mar AC, Robinson ES, Robbins TW
(2009). Dissociable effects of noradrenaline, dopamine,
and serotonin uptake blockade on stop task performance
in rats. Psychopharmacology 205, 273–283.
Bizarro L, Patel S, Murtagh C, Stolerman IP (2004).
Differential effects of psychomotor stimulants on
attentional performance in rats : nicotine, amphetamine,
caffeine and methylphenidate. Behavioral Pharmacology
15, 195–206.
Blondeau C, Dellu-Hagedorn F (2007). Dimensional analysis
of ADHD subtypes in rats. Biological Psychiatry 61,
1340–1350.
Boulougouris V, Glennon JC, Robbins TW (2008).
Dissociable effects of selective 5-HT2A and 5-HT2C
receptor antagonists on serial spatial reversal learning in
rats. Neuropsychopharmacology 33, 2007–2019.
Bymaster FP, Katner JS, Nelson DL, Hemrick-Luecke SK,
et al. (2002). Atomoxetine increases extracellular levels
of norepinephrine and dopamine in prefrontal cortex of
rat : a potential mechanism for efficacy in attention
deficit/hyperactivity disorder. Neuropsychopharmacology
27, 699–711.
Chamberlain SR, Del Campo N, Dowson J, Müller U,
et al. (2007). Atomoxetine improved response inhibition
in adults with attention deficit/hyperactivity disorder.
Biological Psychiatry 62, 977–984.
Day M, Pan JB, Buckley MJ, Cronin E, et al. (2007).
Differential effects of ciproxifan and nicotine on
impulsivity and attention measures in the 5-choice
serial reaction time test. Biochemical Pharmacology 73,
1123–1134.
Eagle DM, Baunez C (2010). Is there an
inhibitory-response-control system in the rat?
Evidence from anatomical and pharmacological
studies of behavioral inhibition. Neuroscience and
Biobehavioral Reviews 34, 50–72.
Fletcher PJ, Rizos Z, Noble K, Higgins GA (2011).
Impulsive action induced by amphetamine, cocaine and
MK801 is reduced by 5-HT(2C) receptor stimulation and
5-HT(2A) receptor blockade. Neuropharmacology 61,
468–477.
Fletcher PJ, Tampakeras M, Sinyard J, Higgins GA (2007).
Opposing effects of 5-HT(2A) and 5-HT(2C) receptor
antagonists in the rat and mouse on premature
responding in the five-choice serial reaction time test.
Psychopharmacology 195, 223–234.
Han DD, Gu HH (2006). Comparison of the monoamine
transporters from human and mouse in their sensitivities
to psychostimulant drugs. BMC Pharmacology 6, 6.
Higgins GA, Enderlin M, Haman M, Fletcher PJ (2003).
The 5-HT2A receptor antagonist M100,907 attenuates
motor and ‘impulsive-type’ behaviours produced by
NMDA receptor antagonism. Psychopharmacology 170,
309–319.
Kennett GA, Wood MD, Bright F, Trail B, et al. (1997).
SB 242084, a selective and brain penetrant 5-HT2C
receptor antagonist. Neuropharmacology 36, 609–620.
Michelson D, Adler L, Spencer T, Reimherr FW, et al.
(2003). Atomoxetine in adults with ADHD : two
randomized, placebo-controlled studies. Biological
Psychiatry 53, 112–120.
Milstein J, Dalley J, Robbins T (2010).
Methylphenidate-induced impulsivity : pharmacological
antagonism by beta-adrenoreceptor blockade. Journal of
Psychopharmacology 24, 309–321.
Mørk A, Witten LM, Arnt J (2009). Effect of sertindole
on extracellular dopamine, acetylcholine, and glutamate
in the medial prefrontal cortex of conscious rats :
a comparison with risperidone and exploration
of mechanisms involved. Psychopharmacology
206, 39–49.
National Institutes of Health (2010). Guide for the Care and
Use of Laboratory Animals, 8th edn. Washington, DC :
The National Academies Press.
Pharmacologic dissociation of impulsivity subcategories
Navailles S, De Deurwaerdère P, Porras G, Spampinato U
(2004). In vivo evidence that 5-HT2C receptor antagonist
but not agonist modulates cocaine-induced dopamine
outflow in the rat nucleus accumbens and striatum.
Neuropsychopharmacology 29, 319–326.
Navarra R, Graf R, Huang Y, Logue S, et al. (2008). Effects
of atomoxetine and methylphenidate on attention and
impulsivity in the 5-choice serial reaction time test. Progress
in Neuropsychopharmacology and Biological Psychiatry 32,
34–41.
Passetti F, Dalley JW, Robbins TW (2003). Double
dissociation of serotonergic and dopaminergic
mechanisms on attentional performance using a rodent
five-choice reaction time task. Psychopharmacology 165,
136–145.
Paterson NE, Ricciardi J, Wetzler C, Hanania T (2011).
Sub-optimal performance in the 5-choice serial reaction
time task in rats was sensitive to methylphenidate,
atomoxetine and d-amphetamine, but unaffected by
the COMT inhibitor tolcapone. Neuroscience Research 69,
41–50.
Pattij T, Schetters D, Janssen MC, Wiskerke J, et al. (2009).
Acute effects of morphine on distinct forms of impulsive
behavior in rats. Psychopharmacology 205, 489–502.
Pattij T, Vanderschuren LJ (2008). The neuropharmacology
of impulsive behaviour. Trends in Pharmacological Sciences
29, 192–199.
Puumala T, Ruotsalainen S, Jäkälä P, Koivisto E, et al.
(1996). Behavioral and pharmacological studies on the
validation of a new animal model for attention deficit
hyperactivity disorder. Neurobiology of Learning and
Memory 66, 198–211.
Quérée P, Peters S, Sharp T (2009). Further pharmacological
characterization of 5-HT(2C) receptor agonist-induced
inhibition of 5-HT neuronal activity in the dorsal raphe
nucleus in vivo. British Journal of Pharmacology 158,
1477–1485.
Robinson ES, Eagle DM, Mar AC, Bari A, et al. (2008).
Similar effects of the selective noradrenaline reuptake
inhibitor atomoxetine on three distinct forms of
impulsivity in the rat. Neuropsychopharmacology 33,
1028–1037.
Ruotsalainen S, Sirviö J, Jäkälä P, Puumala T, et al. (1997).
Differential effects of three 5-HT receptor antagonists on
the performance of rats in attentional and working
memory tasks. European Neuropsychopharmacology 7,
99–108.
Schneider T, Ilott N, Brolese G, Bizarro L, et al.
(2011). Prenatal exposure to nicotine impairs performance
1487
of the 5-choice serial reaction time task in adult rats.
Neuropsychopharmacology 36, 1114–1125.
Shiels K, Hawk LW, Reynolds B, Mazzullo RJ, et al. (2009).
Effects of methylphenidate on discounting of delayed
rewards in attention deficit/hyperactivity disorder.
Experimental and Clinical Psychopharmacology 17, 291–301.
Simpson D, Plosker GL (2004). Atomoxetine : a review of
its use in adults with attention deficit hyperactivity
disorder. Drugs 64, 205–222.
Swanson CJ, Perry KW, Koch-Krueger S, Katner J, et al.
(2006). Effect of the attention deficit/hyperactivity disorder
drug atomoxetine on extracellular concentrations of
norepinephrine and dopamine in several brain regions
of the rat. Neuropharmacology 50, 755–760.
Talpos JC, Wilkinson LS, Robbins TW (2006). A comparison
of multiple 5-HT receptors in two tasks measuring
impulsivity. Journal of Psychopharmacology 20, 47–58.
Uslaner JM, Smith SM, Huszar SL, Pachmerhiwala R, et al.
(2009). Combined administration of an mGlu2/3 receptor
agonist and a 5-HT 2A receptor antagonist markedly
attenuate the psychomotor-activating and neurochemical
effects of psychostimulants. Psychopharmacology 206,
641–651.
van Gaalen MM, Brueggeman RJ, Bronius PF, Schoffelmeer
AN, et al. (2006). Behavioral disinhibition requires
dopamine receptor activation. Psychopharmacology 187,
73–85.
Winstanley CA, Eagle DM, Robbins TW (2006). Behavioral
models of impulsivity in relation to ADHD : translation
between clinical and preclinical studies. Clinical
Psychology Review 26, 379–395.
Winstanley CA, Theobald DE, Dalley JW, Glennon JC,
Robbins TW (2004). 5-HT2A and 5-HT2C receptor
antagonists have opposing effects on a measure of
impulsivity : interactions with global 5-HT depletion.
Psychopharmacology 176, 376–385.
Wiskerke J, Schetters D, van Es IE, van Mourik Y, et al.
(2011). m-Opioid receptors in the nucleus accumbens
shell region mediate the effects of amphetamine on
inhibitory control but not impulsive choice. Journal for
Neuroscience 31, 262–272.
Wong DT, Threlkeld PG, Best KL, Bymaster FP (1982).
A new inhibitor of norepinephrine uptake devoid of
affinity for receptors in rat brain. Journal of Pharmacology
and Experimental Therapeutics 222, 61–65.
Young JW, Light GA, Marston HM, Sharp R, Geyer MA
(2009). The 5-choice continuous performance test :
evidence for a translational test of vigilance for mice.
Public Library of Science One 4, e4227.