The Many Facets of the Locomotor Response to a Novel Environment
Test: Theoretical Comment on Mitchell, Cunningham, and Mark (2005)
Rosalind Franklin University of Medicine and Science/The Chicago Medical School
Several animal studies have shown that there is a positive correlation between locomotor activity in
response to a novel environment and acquisition of drug self-administration behavior. This finding led
to the assumption that animals with heightened reactivity to novel environments are more sensitive to the
rewarding effects of drugs compared with animals with reduced reactivity. But are these individuals really
more responsive to drugs, or could they have enhanced sensitivity to rewards in general or even simply be
investigated these important matters. They reported that the locomotor response to a novel environment does
not predict responding for cocaine but reflects overall differences in the ability to learn operant tasks.
Keywords: addiction, locomotor response, learning, operant responding, cocaine
Large interindividual differences are found in drug responding,
both in humans and in animals (de Wit, Uhlenhuth, & Johanson,
1986; Piazza, Deroche, Rouge-Pont, & Le Moal, 1998). Some
subjects more easily acquire drug self-administration and develop
drug addiction, whereas others are more resistant. The presence of
such interindividual differences suggests that certain individuals
could be especially susceptible to the reinforcing effects of co-
caine; these individuals could develop self-administration behavior
more readily than subjects who are less sensitive to the drug
(Piazza et al., 1998, Piazza, Deroche-Gamonet, Rouge-Pont, & Le
Moal, 2000). Elucidating the neurobiological correlates of en-
hanced susceptibility to develop addiction is important, because it
may provide important information for understanding determi-
nants of drug addiction.
In 1989, Piazza, Deminiere, Le Moal, and Simon reported that
the locomotor response to a novel environment is positively cor-
related with the ability to acquire amphetamine self-administration
behavior. Rats with a high locomotor reactivity to a novel envi-
ronment (so-called “high responders” or HRs) show enhanced
acquisition of amphetamine self-administration behavior com-
pared with animals with a low reactivity to the same environment
(“low responders” or LRs). This was a critical finding, because it
allowed researchers to “predict,” for the first time, self-
administration behavior in the absence of any drug exposure in an
outbred population of animals. By simply screening rats for their
response to a novel environment, one could identify “drug-
vulnerable” or “drug-resistant” individuals.
This model of enhanced vulnerability to acquire self-
administration behavior led to many studies on the neurobiological
bases of such vulnerability. To give just some examples, drug-
vulnerable animals (HRs) were found to exhibit higher basal and
stimulated dopamine levels in the nucleus accumbens and striatum
compared with drug-resistant animals (LRs) (Bradberry, Gruen,
Berridge, & Roth., 1991; Hooks, Jones, Smith, Neill, & Justice
1991b; Piazza et al., 1991; Rouge-Pont, Piazza, Kharouby, Le
Moal, & Simon, 1993). HRs also show heightened firing activity
of dopamine neurons compared with LRs and decreased ability to
auto-regulate the impulse activity of these neurons (Marinelli &
White, 2000). This suggests that enhanced vulnerability to drugs
could be the result of a hyperactive dopaminergic system
(Marinelli & Piazza, 2005; Piazza et al., 1998).
Clearly, the validity of these results and their relevance to drug
vulnerability depend greatly on the nature of the relation between
the locomotor response to a novel environment and subsequent
drug self-administration behavior. Thus, the validity of the rela-
tion, together with the interesting findings reported by Mitchell,
Cunningham, and Mark. (2005) will be discussed below. Several
features will be examined, such as the specificity of the model with
respect to different addictive drugs, the importance of the experi-
mental conditions during testing, and the potential nature of the
differences in drug responding. In addition, a description of the
different methods of determining the locomotor response to a
novel environment will be provided.
Methods to Test Differences in the Locomotor Response
to a Novel Environment
Locomotor Activity to a Novel Environment Is a Relative
When placed in a novel environment such as an activity cham-
ber or a new home cage, not all animals respond with the same
I thank Terry E. Robinson and Ve ´ronique Deroche-Gamonet for very
helpful comments on this article and on the article by Mitchell et al.;
Jeffrey M. Grimm for valuable discussion concerning the methodological
details of his previously published studies; Paul Vezina and Pier Vincenzo
Piazza for giving helpful input and for providing data on the locomotor
response to the novel environment reported in this article; and, most of all,
Jennifer M. Mitchell for extremely helpful clarifications regarding the
details of her study.
Correspondence concerning this article should be addressed to Michela
Marinelli, Department of Cellular & Molecular Pharmacology, Rosalind
Franklin University of Medicine and Science/The Chicago Medical School,
2005, Vol. 119, No. 4, 1144–1151
Copyright 2005 by the American Psychological Association
0735-7044/05/$12.00 DOI: 10.1037/0735-7044.119.4.1144
level of activation; some animals exhibit high locomotor activity,
whereas others show lower counts. Two approaches are usually
used: The first involves correlation studies. In this case, individual
activity scores are plotted against individual responses for the
defined behavior (e.g., self-administration), which allows one to
determine whether a direct linear relation exists between locomo-
tor activity and the response of interest. The second approach
consists in dividing animals into HRs and LRs to this environment.
This separation is based either on a median split or on upper or
lower percentages of the population. Separating animals into HRs
and LRs is arbitrary (because locomotor scores are normally
distributed); however, it allows one to make better group compar-
isons, especially after treatments. For example, one can determine
whether a certain treatment affects vulnerable rats (HRs) more
than it affects the resistant ones (LRs).
I should note that there is no absolute value that can define high
versus low activity counts. Because of this, comparisons of loco-
motor activity across individuals can only be made within a group
of rats that are screened at the same time, under exactly the same
conditions (generally, ?10 rats per screening are required to have
a statistically meaningful sample of the population; initial studies
by Piazza et al. (1989) screened 16 rats at a time). Pooling
locomotor counts across groups that are tested at different times is
inappropriate (both for correlation analyses and for group compar-
isons), because it confounds differences between individuals with
those between experimental conditions (day of testing, time of day,
and so on). Therefore, if results are to be combined across exper-
iments, individual activity scores should not be expressed as ab-
solute values; they should be normalized according to the activity
levels of each screened batch (e.g., see W. Lu, Marinelli, Xu,
Worley, & Wolf, 2002). Unfortunately, this is not done in many
studies, which reduces the ability to interpret the results.
Activity Scores Differ According to the Type of
Environment Used and the Duration of the Test
The types of novel environments that have been used in this
model vary across laboratories. Different factors, such as shape,
size, lighting, and so forth, can determine the amount and duration
of the novel environment–induced locomotor activation; the
smaller the environment, the quicker that the novelty-induced
locomotion will subside (Eilam, 2003). The initial apparatus used
by Piazza et al. (1989) was a circular corridor created between two
concentric cylinders (Figure 1a). Animals move within this corri-
dor, which is identical throughout the apparatus. This physical
arrangement is an important consideration because the absence of
corners or a center eliminates potential disparities that may be due
to a different locomotion used in open versus cornered spaces.
When placed in a large open field instead (Figure 1b), animals with
high anxiety levels tend to avoid the central portion of the appa-
ratus and locomote more in the periphery, sometimes showing
thigmotaxis. This reaction could potentially result in a greater
distance traveled, despite lower exploration of the entire arena
(Eilam, 2003). Often researchers measure response to novel envi-
ronments by simply placing an animal in a new cage (Figure 1c)
that is physically similar to the home cage. Regardless of the
apparatus, an animal’s response to the novel environment is usu-
ally characterized by initial strong exploratory behavior (locomo-
tion, sniffing, and rearing) that gradually diminishes and eventu-
ally ceases once the novelty of the environment has subsided.
The level of activity in response to a novel environment is
usually measured over periods of 1–2 hr of exposure to the
environment. Such long periods are required to detect differences
in the way the animals react to the environment as well as
differences in the way animals habituate to the environment. In
fact, many studies have shown that HRs take longer to habituate to
the novel environment (i.e., to reach trough activity scores) com-
pared with LRs. Figure 2 shows locomotor activity between HRs
and LRs obtained by three different laboratories that have success-
fully used this approach to determine markers of drug vulnerability
across subjects. HR–LR differences are observed after 10 min of
exposure to the environment and continue until the animals have
become habituated to the environment. Mitchell and colleagues
reported that animals in their study were screened for 15 min,
which probably allowed detection only of differences in reactivity
rather than in habituation to the environment.
corridor, (b) open field, and (c) novel cage that is physically similar to the home cage. The sizes of these
apparatuses vary across laboratories; approximate measurements used in the literature appear on the images.
Image (a) was obtained by courtesy of Ime ´tronic (www.imetronic.com); image (b), model ENV-515, was
obtained by courtesy of Med Associates (www.med-associates.com).
Different test chambers used to test rats’ locomotor response to a novel environment: (a) circular
The Locomotor Response to a Novel Environment
Predicts Subsequent Self-Administration Behavior
Generalization to Different Species and Drugs of Abuse
The initial findings by Piazza et al. (1989) showing that rats with
higher locomotor scores exhibit greater acquisition of a low dose
of amphetamine were replicated by several other researchers.
These studies confirmed that locomotor response to novelty pre-
dicts the ability to acquire and maintain self-administration of
amphetamine (Cain, Smith, & Bardo, 2004; Klebaur, Bevins,
Segar, & Bardo, 2001) but also extended the findings to other
abused drugs such as cocaine (Grimm & See, 1997; Mantsch, Ho,
Schlussman, & Kreek, 2001; Marinelli & White, 2000; Piazza et
al., 2000), nicotine (Suto, Austin, & Vezina, 2001), morphine
(Ambrosio, Goldberg, & Elmer, 1995), and ethanol (Nadal, Ar-
mario, & Janak, 2002). In addition, this trait was observed both in
rats and in mice, suggesting that the finding can be generalized to
different species (Marinelli & Piazza, 2005).
Importance of the Experimental Conditions in Acquisition
Self-administration studies are sometimes performed with food-
restricted animals or previously food-restricted animals that have
been trained to respond for food. Although this practice facilitates
subsequent drug intake (K. D. Carr, 2002; L. Lu, Shepard, Scott
Hall, & Shaham, 2003), it also introduces the potential confound of
stress. Studies on HR–LR differences in drug intake reviewed
below, as well as the study described by Mitchell et al., were
performed with ad libitum–fed animals and are thus free of this
HR–LR differences in the acquisition of self-administration
behavior are only observed for low drug doses. It has been pro-
posed that these doses are too low to be reinforcing in LRs but not
in HRs (Piazza et al., 1998). In other words, HRs could be more
sensitive than LRs to low, typically nonreinforcing doses of psy-
chostimulant drugs. In fact, when high drug doses are used, both
HRs and LRs learn to self-administer psychostimulant drugs
equally well (Piazza et al., 2000). Although HR–LR differences in
acquisition of self-administration behavior are only detectable at
low drug doses, these doses should be high enough to maintain
self-administration behavior across sessions. If doses are very low,
HR–LR differences can wane over time because both groups of
animals decrease responding across sessions (Pierre & Vezina,
In the study by Mitchell et al. (2005), acquisition of cocaine
self-administration was performed at a moderate dose (250 ?g/kg
per infusion), which should produce reliable acquisition behavior
in most animals. However, animals were trained to acquire cocaine
self-administration on a fixed ratio 5:1 schedule of reinforcement
(i.e., they were required to lever-press five times to obtain each
cocaine infusion). Acquisition in these conditions produces low
rates of self-administration behavior (see figure 3a in Mitchell et
al., 2005), which could explain, at least in part, why HR–LR
differences were present on Days 2–4 of testing but were not
maintained over time. Another reason why HR–LR differences
could be reduced over time is not because self-administration
decreases in all animals (Pierre & Vezina, 1997) but because LRs
increase responding across sessions. They “catch up” and eventu-
ally do not differ from HRs. This increased response could be due
to improved learning across sessions, which is what Mitchell and
colleagues proposed happened with their LR rats. A similar idea
responders (HRs) and low responders (LRs) show differential reactivity and adaptation to the novel environment.
HRs take longer to habituate to the test environment compared with LRs. These profiles are examples of data
obtained in different laboratories: (a) P.V. Piazza, n ? 8 per group, division based on median split, animals
screened toward the end of the light phase (approximately 2-4 hr before lights out); (b) M. Marinelli, n ? 6 per
group, division based on median split, animals screened toward the end of the light phase; (c) P. Vezina, n ?
9–10 per group, division based on upper and lower quartiles, animals screened during the dark phase; modified
from data appearing in Suto, Austin, & Vezina, 2001. Vertical lines above and below the points are error bars.
Figures are obtained by courtesy of these authors.
Profiles of the rats’ locomotor response to a novel environment. Animals identified as high
was proposed in a study on ethanol self-administration (Nadal et
al., 2002) that illustrated that responding in LRs can increase
Although some studies show that HR–LR differences in self-
administration behavior can decrease over time, numerous other
studies have shown the opposite. Using cocaine or amphetamine as
the reinforcer, researchers have shown that HRs do not differ from
LRs on the first few days of testing (Days 1–5, according to the
study); differences in drug intake develop only over subsequent
days of drug exposure (Marinelli & White, 2000; Piazza et al.,
1989; Piazza et al, 1990), even when animals are exposed to
extended (10 hr/day) daily cocaine access (Mantsch, Ho,
Schlussman, & Kreek, 2001). Figure 3 shows an example of how
HR–LR differences in self-administration behavior develop after 5
days and are maintained for the remainder (the following 5 days)
of the study. In this experiment, animals start off with similar
responding abilities; however, after having experienced the drug,
HRs perceive the drug as being more reinforcing than LRs.
However, in the studies cited above, drug responding was mea-
sured for only 5–10 days, so it was questionable whether HR–LR
differences would persist over protracted periods. In experiments
carried out over longer periods of time (approximately 1 month),
HR rats were found to respond more for psychostimulant drugs
across days, doses, and ratios than did LR rats (Klebaur et al.,
2001; Piazza et al., 2000; Suto et al., 2001). Similarly, subordinate
monkeys, which show higher locomotor activity levels, exhibit
greater cocaine intake as measured over at least 1 month (Morgan
et al., 2000, 2002). These studies indicate the stability of the
HR–LR trait over time, at least in conditions of sustained self-
administration behavior across days (Pierre & Vezina, 1997).
Importance of the Duration of the Locomotor Response to
Most studies showing strong relationships between locomotor
response to a novel environment and self-administration behavior
used long (? 60 min) as opposed to short (10–15 min) locomotor
screening periods. As mentioned above, it is possible that screen-
ing animals for long periods of time could “capture” a behavioral
trait that predicts different behavioral aspects than those revealed
by short screening times. Figure 4 examines the relationship be-
tween cocaine self-administration and the locomotor response to a
novel environment test, measured over 10, 30, or 120 min. Cor-
relations are strongest when the duration of the locomotor activity
test is the longest (120 min). Almost identical results were reported
by Suto et al. (2001) for nicotine self-administration. Thus, the
correlation coefficients obtained between locomotor counts ob-
served at different times during the response to novelty screen and
nicotine self-administration were strong when long durations of
the locomotor test were taken into account but were absent when
short (?60 min) periods were considered. Thus, screening animals
for short periods produces correlations that are weak (see Figure
4), transiently present (Mitchell et al., 2005), or significant when
responding is cumulated across days (Grimm & See, 1997). In the
latter study, correlations were best observed when taking into
account 10 min rather than 5 min of screening time. Together,
these data indicate that testing animals for their locomotor re-
sponse to a novel environment for very short periods does not
produce a robust model that can be used to predict cocaine self-
administration behavior; longer periods of screening are preferable
to obtain the strongest correlations. This finding suggests that the
animal’s adaptation to the environment is an important factor that
needs to be considered when evaluating the locomotor activity in
response to novelty. However, although this factor applies to
psychostimulants, it might not apply to different drugs such as
ethanol; for ethanol self-administration, correlations are obtained
when animals are screened for short (10 min) but not long (80 min)
periods (Nadal et al., 2002).
Beyond Acquisition of Drug Self-Administration
Further studies have been performed to determine whether
HR–LR differences in drug responding are restricted to acquisition
of low drug doses. Using dose–response paradigms, researchers
have shown that HRs exhibit a vertical upward shift of the dose–
response curve for cocaine (Klebaur et al., 2001; Piazza et al.,
2000), suggesting that the reinforcing efficacy of the drug is
greater in these animals. Furthermore, HRs show greater respond-
ing for psychostimulants in progressive ratio schedules of rein-
forcement (Grimm & See, 1997; Suto et al., 2001), suggesting that
HRs work harder to obtain drug rewards. Although the locomotor
response to a novel environment predicts acquisition and mainte-
nance of self-administration behavior, it does not appear to predict
drug-seeking behavior or the loss of control that is usually typical
of addicted humans. Thus, HRs and LRs show similar responding
during extinction paradigms, and they reinstate to a similar extent
when presented with drug cues (Deroche-Gamonet, Belin, & Pi-
azza, 2004; Sutton, Karanian, & Self, 2000). After protracted
self-administration (?2 months) HRs and LRs also seem to show
similar drug intake in the presence of adverse consequences and in
other paradigms that are indicative of compulsive drug seeking
responder rats (HRs represented by dark circles) and low responder rats
(LRs, represented by white circles, n ? 5 per group). Responder status was
based on the median split over 120 min of locomotor activity testing). HRs
and LRs do not differ for cocaine intake over the first five self-
administration sessions; differences in drug intake develop from Day 6 and
remain until the end of the experiment. In this experiment, nose pokes in
the active hole resulted in an infusion of cocaine; a time-out of 5 s was used
after each infusion. Vertical lines above the points are error bars. Unlike
previous studies (Marinelli & White, 2000), in this experiment, no discrete
cues (light) were associated with responding in the active hole (Marinelli,
Self-administration of 165 ?g/kg of cocaine tested in high
(Deroche-Gamonet et al., 2004). These findings indicate that lo-
comotor response to a novel environment can be to used to detect
differences in sensitivity to psychostimulant drugs and possibly
also in the motivation to obtain drugs, but they do not predict drug
seeking or loss of control.
What Else Could the Locomotor Response to a Novel
Environment Be Used to Predict?
The above data indicate that locomotor response to a novel
environment can be used to predict subsequent behavior in some
kinds of self-administration tests (acquisition and maintenance), at
least if the animals are screened for the locomotor response to
novelty test for long periods of time and if the training conditions
are appropriate to produce sustained self-administration behavior
across days. The question could arise as to whether these findings
reflect differences in cocaine sensitivity or simply reflect behav-
ioral characteristics that could be unrelated to drugs.
Differences in Sampling Rates?
Animals that locomote more in a novel environment could also
show nonspecific increases in responding (lever-pressing or nose-
poking behavior) that could be erroneously interpreted as greater
self-administration behavior. To test potential differences in sam-
pling rates, investigators have tested HR and LR animals for
responding in the inactive device or for responding in the absence
of the drug. Results indicate that HRs and LRs do not show
generalized differences in nonreinforced responding and that there
is no relationship between the locomotor response to a novel
environment and nose-poking behavior (Cain et al., 2004;
Marinelli & White, 2000; Piazza et al., 1990). Thus, it is unlikely
that variations in sampling rates can account for individual differ-
ences in drug responding and intake.
Differences in Responding for Rewards in General?
If locomotor response to a novel environment is predictive of
responding for different drugs of abuse, it is possible that it is
predictive of responding for rewards in general. This theory can be
tested by examining operant responding for nondrug rewards. In
two studies, HRs and LRs responded equally for light cues
(Marinelli & White, 2000; Piazza et al., 1990); however, it should
be noted that although responding was maintained over 7 days,
levels of responding were very low. As mentioned above, when
subsequent cocaine self-administration behavior (reported on Days 1, 3, 5 and 7 of testing). The strongest and
more consistent correlations are obtained by considering longer screening times. For all conditions, correlations
are absent on the first day of testing for cocaine self-administration behavior. When the locomotor response to
a novel environment is tested over 120 min, correlations develop on Day 3 of self-administration testing and are
maintained until the end of the experiment (Day 7). In addition, the strength of the correlation increases over
time. These correlations are obtained by analyzing data that were reported partially in a previous publication
(Marinelli & White, 2000). The r values represent the correlation coefficient obtained with Pearson’s correlation
test, p values reported in the figures indicate the level of significance; values are underlined when p ? .05.
Relation between the locomotor response to a novel environment tested over 10, 30, or 120 min and
responding is low, HR–LR differences are rarely robust. In fact,
responding for stimuli with greater reinforcing properties yields
different results. Relative to LRs, HRs show faster acquisition of
sucrose self-administration (Klebaur et al., 2001). On a similar
line, animals that show greater amphetamine-induced locomotion
and dopamine release also show greater preference for a sucrose
solution (Sills & Crawley, 1996; Sills & Vaccarino, 1994), sug-
gesting a positive relation between the ability to respond to natural
and drug rewards. Additionally, when tested for operant respond-
ing for food, HRs display higher sensitivity to the reinforcing
properties of food (Dellu, Piazza, Mayo, Le Moal, & Simon,
1996). These data were confirmed in the study by Mitchell et al.
(2005), who reported greater responding for food in animals with
greater locomotor response to a novel environment (Figure 4b in
Mitchell et al.). Greater responding was most evident on Days 3–8
(see Results section in Mitchell et al., 2005) and decreased on
Days 9–12. This indicates that high locomotor response to a novel
environment is coupled to high reactivity to rewards in general and
not just to drug rewards.
Differences in Learning?
If animals showing greater locomotor response to a novel envi-
ronment over 15 min show greater responding for many rewards,
they could simply be better learners. To test this possibility,
Mitchell and colleagues (2005) studied responding for cocaine
after having submitted animals to extensive food training on the
operant task. After having learned operant responding (with food
as the reinforcer), HRs and LRs no longer differed for cocaine
self-administration, suggesting that previously observed differ-
ences in drug responding were simply the consequence of differ-
ences in learning capacities. Unfortunately, cocaine responding
was only measured over 3 days, so it is difficult to determine
whether differences in drug intake would have developed during
subsequent days of testing (as shown in Figure 3, correlations
between locomotor response to a novel environment and self-
administration often develop after 3 days of testing). However, it
is interesting that HRs and LRs did not differ in responding for
cocaine over 3 days if they had previously been trained to respond
for food. In addition, the absence of a correlation between loco-
motor response to a novel environment and self-administration
apparently was maintained over subsequent days as well (J. M.
Mitchell, personal communication, March 2005). Finally, Mitchell
et al. found that results were similar when two rats that appeared
to be outliers (see Figure 1a) were removed (J. M. Mitchell,
personal communication, March 2005). Possibly, in these condi-
tions, differences in cocaine responding were indeed a conse-
quence of differences in the ability to learn an operant task.
The idea that response to a novel environment is associated with
better instrumental learning is corroborated by the positive corre-
lation between locomotor response to novelty and the day on
which the learning of the food self-administration task was
achieved (Figure 4a in Mitchell et al., 2005). The “learning thresh-
old” for self-administration behavior is determined with a novel
and thoughtful method that uses the trough of the bimodal distri-
bution of lever-pressing over self-administration sessions. This
method allows objective measurements of learning, without con-
founds of arbitrary definitions of learning.
Mitchell et al. (2005) proposed further evidence for learning dif-
ferences between their HRs and LRs. In animals that received no
previous training on the operant task, a positive correlation was found
between locomotor response to a novel environment and cocaine
self-administration on Days 2–4 of training but not on subsequent
days (Days 5–7). Similarly, for food self-administration, correlations
were present during Days 3–8 of testing but not on Days 9–12. As
time to “practice” and learn self-administration behavior, differences
between HRs and LRs disappeared. This result implies that HRs, at
least those in the study by Mitchell et al., are better learners but not
necessarily greater “addicts” or greater responders for rewards in
general. Given that animals start off with a moderately difficult
schedule of reinforcement (they have to perform five lever presses to
obtain each cocaine infusion), it is indeed possible that differences in
learning could account for differences in drug intake under this
Although differences in learning probably did contribute to
differences in drug intake in the model by Mitchell et al. (2005),
differences in learning are unlikely to explain HR–LR differences
in drug responding that have been observed over protracted peri-
ods of time in other studies. For example, in dose–response stud-
ies, animals are first trained to respond for a high cocaine dose
(1000 ?g/kg per infusion) over 10 days on a fixed ration 1:1 of
reinforcement (1 response: 1 infusion); once all animals have
learned the task equally well, the drug dose is decreased progres-
sively across days. In these conditions, HRs still show greater
responding for cocaine relative to LRs, as evidenced by a vertical
upward shift of the dose–response curve for cocaine (Piazza et al.,
2000). Similar results have been reported in monkeys (Morgan et
al., 2000, 2002). This indicates that, once the behavior has been
learned, HR–LR differences in drug responding are still main-
tained and suggests that differences in learning cannot always
account for differences in drug responding.
Assuming that animals that show greater drug intake are better
learners, it could be interesting to determine whether they are
better learners in general or whether this is only restricted to
responding for rewards on operant tasks. Such studies are limited;
one study reported no difference in HR–LR performance in a
recognition memory task for cognitive abilities tested during adult-
hood; memory impairments in HRs, however, did develop during
aging (Dellu, Mayo, Vallee, Le Moal, & Simon, 1994). Another
study showed the absence of a relationship between locomotor
response to a novel environment and performance on an eight-arm
radial maze (Dellu-Hagedorn, 2005). In addition, in that study,
animals with poor learning capacities showed the greatest
amphetamine-induced locomotion, and a negative correlation was
found between learning and drug reactivity. These results indicate
that animals with greater reactivity to drugs do not differ from HRs
and LRs in learning capacities in general. If they do, it is only for
the learning of operant responding.
What About Other Drug-Related Responses?
A different approach to assessing whether locomotor response
to a novel environment is potentially related to drug sensitivity is
to examine its relation with other behavioral responses to drugs
that do not involve operant responding. Conditioned place prefer-
ence has often been considered an index of the rewarding effects
of drugs of abuse (G. D. Carr, Fibiger, & Phillips, 1989; Hoffman,
1989; Tzschentke, 1998). In mice, there appears to be a positive
relationship between locomotor response to a novel environment
and the amount of time spent in the drug-paired environment of the
conditioned place preference test (Orsini, Buchini, Piazza, Puglisi-
Allegra, & Cabib, 2004). In rats, however, data consistently show
that locomotor response to novelty does not predict place prefer-
ence behavior (Erb & Parker, 1994; Gong, Neill, & Justice, 1996;
Klebaur & Bardo, 1999; Kosten & Miserendino, 1998). Although
a useful test, place conditioning measures changes in the threshold
dose of psychostimulant required to produce conditioning; how-
ever, once the response is induced, the intensity of its effects does
not change as a function of drug dose (Costello, Carlson, Glick, &
Bryda, 1989). This test, therefore, is well adapted to measure
horizontal shifts in dose–response functions, but it is not well
suited to assess changes in the intensity of the rewarding effects of
drugs (Bardo & Bevins, 2000), such as those seen between HRs
and LRs (Piazza et al., 2000).
Other behavioral responses are therefore useful to examine. HRs
show stronger contextual conditioning to drugs (Jodogne,
Marinelli, Le Moal, & Piazza, 1994), which is in line with Mitchell
et al.’s (2005) idea that these animals could show increased learn-
ing capacities. However, HRs also show greater locomotor reac-
tivity to an acute injection of a psychostimulant drug, which is
clearly independent of learning. These HR–LR differences are
present over a wide range of drug doses. In addition, they are not
confounded by differences in locomotor response to the environ-
ment per se because they are present after the animal has habitu-
ated to the test chamber (Exner & Clark, 1993; Hooks et al.,
1991b; Hooks, Jones, Neill, & Justice, 1992; Piazza et al., 1989;
1990; 1998). Finally, HRs develop behavioral sensitization more
readily than do LRs (Hooks, Jones, Liem, & Justice, 1992; Hooks,
Jones, Neill, & Justice, 1992; Hooks, Jones, Smith, Neill, & Justice
1991a; Pierre & Vezina, 1997). Behavioral sensitization is impor-
tant in that it reflects drug-induced neuroadaptations that could
have an important role in the development of addiction (Robinson
& Berridge, 1993; 2001). Overall, these findings show that loco-
motor response to a novel environment is predictive of several
drug-associated behaviors. Some of these, but not all, can be
explained by potential differences in learning across animals.
In summary, the locomotor response to a novel environment test
has proven useful for predicting differences in drug sensitivity. How-
ever, several considerations should be taken into account. These
include screening animals for long enough periods of time to capture
individual differences in habituation to the environment and recording
relative locomotor scores rather than absolute counts. In addition,
acquisition of drug self-administration should be tested over several
days to allow differences in responding to emerge; drug doses should
also be low enough to allow detection of differences in drug sensi-
sessions; and finally, the schedule of reinforcement should be easy
enough to reduce potential differences in learning across animals.
criteria, it is still extremely valuable because it showed that the ability
to learn instrumental responding and drug self-administration behav-
enhanced locomotor reactivity to a novel environment exhibited
higher cocaine self-administration behavior during their first days of
exposure to the drug; however, these animals showed higher drug
responding simply because they were better instrumental learners. In
fact, once the behavior was learned, differences in drug responding
were no longer present.
Regardless of whether individual differences in drug taking can be
predicted by the locomotor response to a novel environment or not,
this article clearly shows that initial differences in cocaine intake
(under moderately demanding schedules of reinforcement) can be
explained by differences in the animal’s ability to learn operant
responding. Thus, when trying to determine the neurobiological bases
control studies to determine the nature of these differences. Without
such controls, one risks studying the neurobiological mechanisms
underlying differences in learning abilities, rather than those under-
lying differences in the reinforcing properties of drugs.
Ambrosio, E., Goldberg, S. R., & Elmer, G. I. (1995). Behavior genetic
investigation of the relationship between spontaneous locomotor activity
and the acquisition of morphine self-administration behavior. Behav-
ioural Pharmacology, 6, 229–237.
Bardo, M. T., & Bevins, R. A. (2000). Conditioned place preference: What
does it add to our preclinical understanding of drug reward? Psycho-
pharmacology (Berl), 153, 31–43.
Bradberry, C. W., Gruen, R. J., Berridge, C. W., & Roth, R. H. (1991).
Individual differences in behavioral measures: Correlations with nucleus
accumbens dopamine measured by microdialysis. Pharmacology, Bio-
chemistry and Behavior, 39, 877–882.
Cain, M. E., Smith, C. M., & Bardo, M. T. (2004). The effect of novelty
on amphetamine self-administration in rats classified as high and low
responders. Psychopharmacology (Berl), 176, 129–138.
Carr, G. D., Fibiger, H. C., & Phillips, A. G. (1989). Conditioned place
preference as a measure of drug reward. In J. M. Liebman & S. J. Cooper
(Eds.), The neuropharmacological basis of reward (pp. 264–319). Ox-
ford, United Kingdom: Oxford University Press.
Carr, K. D. (2002). Augmentation of drug reward by chronic food restric-
tion: Behavioral evidence and underlying mechanisms. Physiology and
Behavior, 76, 353–364.
Costello, N. L., Carlson, J. N., Glick, S. D., & Bryda, M. (1989). Dose-
dependent and baseline-dependent conditioning with d-amphetamine in
the place conditioning paradigm. Psychopharmacology (Berl), 99, 244–
Dellu, F., Mayo, W., Vallee, M., Le Moal, M., & Simon, H. (1994).
Reactivity to novelty during youth as a predictive factor of cognitive
impairment in the elderly: A longitudinal study in rats. Brain Research,
Dellu, F., Piazza, P. V., Mayo, W., Le Moal, M., & Simon, H. (1996).
Novelty-seeking in rats: biobehavioral characteristics and possible rela-
tionship with the sensation-seeking trait in man. Neuropsychobiology,
Dellu-Hagedorn, F. (2005). Spontaneous individual differences in cogni-
tive performances of young adult rats predict locomotor response to
amphetamine. Neurobiology of Learning and Memory, 83, 43–47.
Deroche-Gamonet, V., Belin, D., & Piazza, P. V. (2004, August 13).
Evidence for addiction-like behavior in the rat. Science, 305, 1014–
de Wit, H., Uhlenhuth, E. H., & Johanson, C. E. (1986). Individual
differences in the reinforcing and subjective effects of amphetamine and
diazepam. Drug and Alcohol Dependence, 16, 341–360.
Eilam, D. (2003). Open-field behavior withstands drastic changes in arena
size. Behavioural Brain Research, 142, 53–62.
Erb, S. M., & Parker, L. A. (1994). Individual differences in novelty-
induced activity do not predict strength of amphetamine-induced place
conditioning. Pharmacology, Biochemistry and Behavior, 48, 581–586.
Exner, M., & Clark, D. (1993). Behaviour in the novel environment
predicts responsiveness to d-amphetamine in the rat: A multivariate
approach. Behavioural Pharmacology, 4, 47–56.
Gong, W., Neill, D. B., & Justice, J. B., Jr. (1996). Locomotor response to
novelty does not predict cocaine place preference conditioning in rats.
Pharmacology, Biochemistry and Behavior, 53, 191–196.
Grimm, J. W., & See, R. E. (1997). Cocaine self-administration in ovari-
ectomized rats is predicted by response to novelty, attenuated by 17-beta
estradiol, and associated with abnormal vaginal cytology. Physiology
and Behavior, 61, 755–761.
Hoffman, D. C. (1989). The use of place conditioning in studying the
neuropharmacology of drug reinforcement. Brain Research Bulletin, 23,
Hooks, M. S., Jones, G. H., Liem, B. J., & Justice, J. B., Jr. (1992).
Sensitization and individual differences to IP amphetamine, cocaine, or
caffeine following repeated intracranial amphetamine infusions. Annals
of the New York Academy of Sciences, 654, 444–447.
Hooks, M. S., Jones, G. H., Neill, D. B., & Justice, J. B., Jr. (1992).
Individual differences in amphetamine sensitization: Dose-dependent
effects. Pharmacology, Biochemistry and Behavior, 41, 203–210.
Hooks, M. S., Jones, G. H., Smith, A. D., Neill, D. B., & Justice, J. B., Jr.
(1991a). Individual differences in locomotor activity and sensitization.
Pharmacology, Biochemistry and Behavior, 38, 467–470.
Hooks, M. S., Jones, G. H., Smith, A. D., Neill, D. B., & Justice, J. B., Jr.
(1991b). Response to novelty predicts the locomotor and nucleus ac-
cumbens dopamine response to cocaine. Synapse, 9, 121–128.
Jodogne, C., Marinelli, M., Le Moal, M., & Piazza, P. V. (1994). Animals
predisposed to develop amphetamine self-administration show higher
susceptibility to develop contextual conditioning of both amphetamine-
induced hyperlocomotion and sensitization. Brain Research, 657, 236–244.
Klebaur, J. E., & Bardo, M. T. (1999). Individual differences in novelty
seeking on the playground maze predict amphetamine conditioned place
preference. Pharmacology, Biochemistry and Behavior, 63, 131–136.
Klebaur, J. E., Bevins, R. A., Segar, T. M., & Bardo, M. T. (2001).
Individual differences in behavioral responses to novelty and amphet-
amine self-administration in male and female rats. Behavioural Phar-
macology, 12, 267–275.
Kosten, T. A., & Miserendino, M. J. (1998). Dissociation of novelty- and
cocaine-conditioned locomotor activity from cocaine place conditioning.
Pharmacology, Biochemistry and Behavior, 60, 785–791.
Lu, L., Shepard, J. D., Scott Hall, F., & Shaham, Y. (2003). Effect of
environmental stressors on opiate and psychostimulant reinforcement,
reinstatement and discrimination in rats: A review. Neuroscience and
Biobehavioral Reviews, 27, 457–491.
Lu, W., Marinelli, M., Xu, D., Worley, P. F., & Wolf, M. E. (2002).
Amphetamine and cocaine do not increase Narp expression in rat ventral
tegmental area, nucleus accumbens or prefrontal cortex, but Narp may
contribute to individual differences in responding to a novel environ-
ment. European Journal of Neuroscience, 15, 2027–2036.
Mantsch, J. R., Ho, A., Schlussman, S. D., & Kreek, M. J. (2001).
Predictable individual differences in the initiation of cocaine self-
administration by rats under extended-access conditions are dose-
dependent. Psychopharmacology (Berl), 157, 31–39.
Marinelli M., Koeltzow T. E., & White F. (1998, November 7–12) Indi-
vidual response to novelty predicts activity of dopamine cells in the
ventral tegmental area. Society for Neuroscience, 28thAnnual Meeting,
Marinelli, M., & Piazza, P. V. (2005). Glucocorticoid hormones, individual
differences, and behavioral and dopaminergic responses to psychostimu-
lant drugs. In T. Steckler, N. H. Kalin, & J. M. H. M. Reul (Eds.),
Handbook on stress and the brain: Pt. 2. Integrative and clinical aspects
(pp. 89–111). Amsterdam: Elsevier Science.
Marinelli, M., & White, F. J. (2000). Enhanced vulnerability to cocaine
self-administration is associated with elevated impulse activity of mid-
brain dopamine neurons. Journal of Neuroscience, 20, 8876–8885.
Mitchell, J. M., Cunningham, C. L., & Mark, G. P. (2005). Locomotor
activity predicts acquisition of self-administration behavior but not co-
caine intake. Behavioral Neuroscience, 119, 464–472.
Morgan, D., Grant, K. A., Gage, H. D., Mach, R. H., Kaplan, J. R., Prioleau,
cocaine self-administration. Nature Neuroscience, 5, 169–174.
Morgan, D., Grant, K. A., Prioleau, O. A., Nader, S. H., Kaplan, J. R., &
Nader, M. A. (2000). Predictors of social status in cynomolgus monkeys
(Macaca fascicularis) after group formation. American Journal of Pri-
matology, 52, 115–131.
Nadal, R., Armario, A., & Janak, P. H. (2002). Positive relationship
between activity in a novel environment and operant ethanol self-
administration in rats. Psychopharmacology (Berl), 162, 333–338.
Orsini, C., Buchini, F., Piazza, P. V., Puglisi-Allegra, S., & Cabib, S.
(2004). Susceptibility to amphetamine-induced place preference is pre-
dicted by locomotor response to novelty and amphetamine in the mouse.
Psychopharmacology (Berl), 172, 264–270.
Piazza, P. V., Deminiere, J. M., Le Moal, M., & Simon, H. (1989,
September 29). Factors that predict individual vulnerability to amphet-
amine self-administration. Science, 245, 1511–1513.
Piazza, P. V., Deminiere, J. M., Maccari, S., Mormede, P., Le Moal, M., &
Simon, H. (1990). Individual reactivity to novelty predicts probability of
amphetamine self-administration. Behavioural Pharmacology, 1, 339–345.
Piazza, P. V., Deroche, V., Rouge-Pont, F., & Le Moal, M. (1998).
Behavioral and biological factors associated with individual vulnerabil-
ity to psychostimulant abuse. NIDA Research Monograph, 169, 105–133.
Piazza, P. V., Deroche-Gamonet, V., Rouge-Pont, F., & Le Moal, M.
(2000). Vertical shifts in self-administration dose-response functions
predict a drug-vulnerable phenotype predisposed to addiction. Journal of
Neuroscience, 20, 4226–4232.
Piazza, P. V., Rouge-Pont, F., Deminiere, J. M., Kharouby, M., Le Moal,
M., & Simon, H. (1991). Dopaminergic activity is reduced in the
prefrontal cortex and increased in the nucleus accumbens of rats predis-
posed to develop amphetamine self-administration. Brain Research,
Pierre, P. J., & Vezina, P. (1997). Predisposition to self-administer am-
phetamine: The contribution of response to novelty and prior exposure to
the drug. Psychopharmacology (Berl), 129, 277–284.
Robinson, T. E., & Berridge, K. C. (1993). The neural basis of drug
craving: An incentive-sensitization theory of addiction. Brain Research
Reviews, 18, 247–291.
Robinson, T. E., & Berridge, K. C. (2001). Incentive-sensitization and
addiction. Addiction, 96, 103–114.
Rouge-Pont, F., Piazza, P. V., Kharouby, M., Le Moal, M., & Simon, H.
(1993). Higher and longer stress-induced increase in dopamine concen-
trations in the nucleus accumbens of animals predisposed to amphetamine
self-administration. A microdialysis study. Brain Research, 602, 169–174.
Sills, T. L., & Crawley, J. N. (1996). Individual differences in sugar
consumption predict amphetamine-induced dopamine overflow in nu-
cleus accumbens. European Journal of Pharmacology, 303, 177–181.
Sills, T. L., & Vaccarino, F. J. (1994). Individual differences in sugar
intake predict the locomotor response to acute and repeated amphet-
amine administration. Psychopharmacology (Berl), 116, 1–8.
Suto, N., Austin, J. D., & Vezina, P. (2001). Locomotor response to
novelty predicts a rat’s propensity to self-administer nicotine. Psycho-
pharmacology (Berl), 158, 175–180.
Sutton, M. A., Karanian, D. A., & Self, D. W. (2000). Factors that
determine a propensity for cocaine-seeking behavior during abstinence
in rats. Neuropsychopharmacology, 22, 626–641.
Tzschentke, T. M. (1998). Measuring reward with the conditioned place
preference paradigm: A comprehensive review of drug effects, recent
progress and new issues. Progress in Neurobiology, 56, 613–672.
Received May 9, 2005
Revision received May 10, 2005
Accepted May 11, 2005 ?