Neuroscience Letters 322 (2002) 169–172
www.elsevier.com/locate/neulet
Dissociation of the associative and visceral sensory components of
taste aversion learning by tetrodotoxin inactivation of the
parabrachial nucleus in rats
M.A. Ballesteros a, F. González a, I. Morón a, I. DeBrugada a, A. Cándido a, M. Gallo a,b,*
a
Faculty of Psychology, Department of Experimental Psychology and Physiology of Behavior, University of Granada, Campus Cartuja,
Granada -18071, Spain
b
Institute of Neurosciences Federico Oloriz, University of Granada, Granada, Spain
Received 4 October 2001; received in revised form 17 January 2002; accepted 17 January 2002
Abstract
The parabrachial nucleus (PBN) has been proposed as the associative site for conditioned taste aversion. Previous
evidence has shown that functional blockade of the PBN by tetrodotoxin (TTX) produces retrograde disruption of
lithium-induced taste aversions in rats. However, given the PBN role in processing visceral cues and the long duration
of the lithium-induced aversive effects, an interpretation based on lithium chloride processing deficits can not be ruled
out. The aim of the present study was to use the unconditioned stimulus (US) pre-exposure phenomenon to explore the
effect of PBN inactivation by intracerebral TTX microinjections on visceral processing. Three intraperitoneal (i.p.) lithium
chloride injections (0.15 M; 2% b.w.) applied before the conditioning session, but not isotonic saline i.p. injections,
interfered with the acquisition of a learned aversion to a cider vinegar solution (3%) in cannulated control rats. Bilateral
PBN inactivation by TTX (10 ng) applied immediately after each LiCl injections disrupted the US pre-exposure effect, thus
confirming its sensory role. However, PBN inactivation 30 min after LiCl injections did not interfere with the US preexposure effect, in spite of the fact that an identically timed PBN blockade after the acquisition trial disrupted the
acquisition of taste aversions. These results stand for the associative role of PBN in taste aversion learning induced
by lithium chloride, independent of its sensory role. It is concluded that PBN activity is required after the conditioning
trial for the taste-visceral association to take place. q 2002 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Conditioned taste aversion; Unconditioned stimulus pre-exposure; Parabrachial; Rat; Reversible blockade; Tetrodotoxin
Conditioned taste aversion (CTA) is a robust type of
learning that involves the presentation of a taste solution
as the conditioned stimulus and an illness-inducing aversive
agent as the unconditioned stimulus (US). CTA is readily
acquired in one-trial by rats and they avoid the taste in later
presentations [4]. Although a variety of aversive agents
induce CTA, intraperitoneal (i.p.) injections of lithium
chloride (LiCl) have been the US more widely used [4].
Evidence from permanent and reversible lesion studies has
pointed to the parabrachial pontine area (PBN), the second
relay station both in taste and visceral sensory pathways, as
the primary associative locus (for review see Refs.
[4,11,16,17]). Data obtained applying reversible brain inactivation induced by tetrodotoxin (TTX), a blocker of the
* Corresponding author. Tel.: 134-958-243771; fax: 134-958246239.
E-mail address: mgallo@ugr.es (M. Gallo).
voltage-dependent sodium channels, showed that inactivating the entire PBN during conditioning, either before [3,7]
or after the LiCl injection [8], disrupted CTA acquisition.
The PBN maximal blockade induced by TTX lasted for 30–
120 min decaying exponentially and completely disappearing 24 h after its administration [18]. As the brain was intact
during taste processing either during acquisition or testing,
the authors excluded a gustatory deficit. However, a deficit
of visceral processing, although it did not seem feasible,
could not be excluded. In spite of the fact that PBN blockade
took place after the LiCl injection [8], the long duration and
the unknown temporal parameters of the aversive effects
induced by LiCl precluded a definitive interpretation of
the results in terms of associative deficits.
An associative role of PBN in CTA, independent of its
visceral processing function, has also been supported by
reversible lesion studies using USs, other than LiCl, such
as amphetamine [3] and body-rotation [6]. However, LiCl is
0304-3940/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved.
PII: S03 04 - 394 0( 0 2) 00 09 4- 0
170
M.A. Ballesteros et al. / Neuroscience Letters 322 (2002) 169–172
the standard US for inducing CTA in the majority of lesion
studies addressing the specific location of the associative
locus inside the PBN.
Hence, in the present study we have tested if post-training
TTX inactivation of the PBN interferes with the ability of
LiCl to act as a US using a CTA procedure. We have taken
advantage of the US pre-exposure learning effect [10]. It is
well known that the US experience prior to the conditioning
trial results in taste aversions weaker than those exhibited by
non-pre-exposed control animals [1]. The advantage of this
learning phenomenon is that it allows us to temporally
dissociate the presentation of lithium chloride as a US in
the pre-exposure phase and the taste-lithium association in
the conditioning phase. Thus, TTX-induced reversible
blockade of the PBN can be applied after LiCl injection
during the pre-exposure phase, leaving intact the brain for
the conditioning and testing phases.
The first experiment was aimed to replicate the basic
Ivanova and Bures [8] findings using shorter delays and a
one-bottle test instead of the choice-test. The entire PBN
was inactivated 30 or 60 min after CTA acquisition by TTX
microinjections. Twenty seven naı̈ve male Wistar rats,
weighing 280–320 g, were assigned to three different
groups: Cannula (n ¼ 8); PBN30 (n ¼ 9); and PBN60
(n ¼ 10). They were provided by the breeding colony of
the University of Granada and individually housed in a
room with constant temperature and a 12:12 h light cycle.
Food was available ad libitum.
All the animals were subjected to the surgical procedure
under pentobarbital anesthesia (50 mg/kg), in order to
implant chronic guiding cannulae for intracerebral (i.c)
microinjections of TTX in the PBN. After having been
fixed in the stereotaxic apparatus with bregma and lambda
at the same height, small trephine openings were drilled in
the exposed skull, anchoring screws were positioned, the
guiding cannulae were inserted bilaterally and fixed to the
skull with acrylate. The coordinates from bregma, taken
from Paxinos and Watson [9] atlas, were the following:
AP 29.2; ML ^1.8. The guiding cannulae were 10 mm
long stainless steel tubings (0.71 mm o.d.; 0.41 mm i.d.).
They were inserted 4 mm below the skull and closed with
loosely fitting mandrels.
After a week recovery period, the animals were adapted
to the water deprivation schedule, with water available daily
for 30 min during 1 week. The volume consumed was
recorded by weighing the bottles before and after the drinking period and estimating the difference. The learning
procedure lasted 4 days and took place in the home cages,
except for LiCl and TTX injections. On day 1, during the
daily 30 min drinking period, a bottle containing 10 ml of a
sodium saccharin solution (0.1%) instead of water was
available and the amount ingested was recorded. Immediately after, the animals received an i.p. injection of LiCl
(0.15 M; 2% b.w.). Thirty (PBN30 group) or sixty minutes
(PBN60 group) later, the rats were hand restrained and
received i.c. TTX microinjections. Injection needles (0.3
mm o.d.; 0.15 mm. i.d.) connected to 10 ml Hamilton
syringes were inserted 13 mm deep into the 10 mm long
guiding cannulae to a point 7 mm below the skull surface to
inject bilaterally 10 ng of TTX in 1 ml of saline into the
PBN. The cannulated animals (Cannula group) received the
same behavioral training but no TTX injection. On day 2,
water was available. On day 3, the rats were allowed to
drink saccharin during the 30 min drinking period and the
amount ingested was recorded.
No differences were found among the groups in the
water intake, either before training or before testing. All
the animals in each of the three groups drank the maximum
amount of available saccharin (10 ml) during the conditioning session. Fig. 1a summarizes the results of the
saccharin one-bottle test. A one way ANOVA showed a
significant group effect (F½2; 24 ¼ 13:15; P , 0:01). Post-
Fig. 1. (a) Mean (^SEM) saccharin intake during the one-bottle
test of the different groups in Exp 1. (Cannula, non-injected
cannulated control group; PBN30, group subjected to bilateral
TTX blockade of the PBN 30 min after LiCl injection in the conditioning day; PBN60, group subjected to bilateral TTX blockade of
the PBN 60 min after LiCl injection in the conditioning day). (b)
Mean (^SEM) vinegar intake during the one-bottle test of the
different groups in Exp 2, showing disruption of the US preexposure phenomenon by TTX blockade of the PBN immediately after lithium chloride pre-exposures (PBN0) and a
preserved US pre-exposure effect in those groups receiving
TTX blockade of the PBN 30 min after lithium chloride pre-exposure (PBN30). (Pre, Pre-exposed group; Ctrl, Non-pre-exposed
group; *: P , 0:05; **:P , 0:01)
M.A. Ballesteros et al. / Neuroscience Letters 322 (2002) 169–172
hoc Newman–Keuls comparisons revealed that both the
PBN30 (P , 0:01) and the PBN60 (P , 0:05) groups
showed higher saccharin intake, i.e. impaired CTA,
compared with the non-injected Cannula group. There
were also significant differences between the two injected
groups (P , 0:01), the group PBN30 drinking more
saccharin than the group PBN60. The impairment of
LiCl-induced CTA acquisition by post-training TTX inactivation of the PBN seen in the first experiment is in agreement with previous reports [8].
The second experiment was intended to explore an interpretation of this retrograde impairment of CTA acquisition
based on sensory deficits. A standard behavioral procedure
to induce the US pre-exposure phenomenon that included
three LiCl pre-exposures was applied [1]. PBN functional
inactivation by TTX was applied either immediately (Exp
2a) or 30 min after each LiCl pre-exposure (Exp 2b), leaving
intact the brain during conditioning and testing. If the US
pre-exposure phenomenon does not appear and the preexposed and the non-pre-exposed groups show similar
taste aversions, it can be assumed that PBN blockade has
disrupted the processing of LiCl. However, the appearance
of the US pre-exposure effect would demonstrate that LiCl
processing before the PBN blockade is sufficient to act as a
US, discarding visceral processing deficits as the explanation for the results of Exp 1.
A total number of seventy two naı̈ve male Wistar rats,
weighing 280–320 g, were used (38 in Exp 2a and 38 in Exp
2b). According to the reversible lesion treatment they were
divided in two groups (TTX vs. Cannula) and depending on
the behavioral treatment the animals in each group were
further assigned to one of two behavioral groups (Pre vs.
Ctrl). Thus, the animals were distributed in four groups:
TTX-Pre (Exp 2a, n ¼ 11; Exp 2b, n ¼ 8); TTX-Ctrl (Exp
2a, n ¼ 11; Exp 2b, n ¼ 8); Cannula-Pre (Exp 2a, n ¼ 8;
Exp 2b, n ¼ 9); and Cannula-Ctrl (Exp 2a, n ¼ 8; Exp 2b,
n ¼ 9). Housing, deprivation conditions, surgery and microinjection procedures were identical to those described in
Exp 1.
The learning procedure included a first lithium pre-exposure phase, lasting 6 days, and a second CTA training phase,
lasting 4 days. On days 1, 3 and 5 all the animals drank
water during the daily drinking period. Immediately after,
the animals in the pre-exposed groups (Pre) received an i.p.
injection of lithium chloride (0.15 M; 2% b.w.), while
animals in the non-pre-exposed control groups (Ctrl)
received an i.p. injection of a similar amount of physiological saline. Immediately after (Exp 2a; PBN0) or 30 min
later (Exp 2b; PBN30) the rats belonging to the PBN groups
were hand restrained and received i.c. TTX microinjections
in the PBN. Cannulated groups (Cannula) did not receive
i.c. microinjection. Days 2, 4 and 6 were recovery days with
water available.
Eleven days after the pre-exposure phase a CTA procedure similar to that described in Exp 1, except for the taste
solution used, was applied. The reason for changing the
171
taste solution in this experiment was that an attempt
performed during the inter-phase interval to obtain the US
pre-exposure effect using a saccharin solution failed both in
cannulated and TTX-injected groups. As there were no
differences among the groups the data are not relevant for
the present purpose. On day 1, during the daily 30 min
drinking period, a bottle containing 10 ml of cider vinegar
in tap water (3%) instead of water was available and the
amount ingested was recorded. Immediately after, the
animals received an i.p. injection of LiCl (0.15 M; 2%
b.w.). On days 2 and 3, water was available. One bottle
test took place on day 4, with the vinegar solution available
during the 30 min drinking period.
There were no significant differences among the groups in
water intake, nor in vinegar intake during the acquisition
day in the conditioning phase. Fig. 1b summarizes the
results of the vinegar one-bottle test in Exp 2. In order to
allow comparisons between Exp 2a and 2b a 3 £ 2 ANOVA
Lesion (PBN0, PBN30, Cannula) £ Behavioral Procedure
(Pre, Ctrl) was accomplished. Cannula groups were
obtained by joining the data of Exp 2a and 2b, as there
were no differences between them. ANOVA analysis
revealed a significant effect of lesion (F½2; 66 ¼ 8:99;
P , 0:01), behavioral procedure (F½1; 66 ¼ 32:82;
P , 0:01) and the interaction Lesion £ Behavioral Procedure (F½2; 66 ¼ 4:36; P , 0:02). Post-hoc Newman–Keuls
comparisons showed that Cannula groups drank more vinegar than both PBN0 (P , 0:01) and PBN30 (P , 0:01)
groups. An analysis of the interaction showed the US preexposure effect both in Cannula (P , 0:01) and PBN30
groups (P , 0:01), with pre-exposed groups drinking a
greater amount of the vinegar solution than the non-preexposed control groups. However, the US pre-exposure
phenomenon did not appear in the PBN0 groups
(P . 0:07), with similar intakes in the Pre and Ctrl groups.
A one-way ANOVA showed significant differences between
the pre-exposed groups (F½2; 33 ¼ 9:58; P , 0:01), with
the cannulated group drinking more vinegar than both
PBN0 (P , 0:01) and PBN30 (P , 0:01). There were no
differences between the Ctrl groups (F½2; 33 ¼ 0:97;
P . 0:38).
In both experiments, after completion of the behavioral
procedure, the rats were deeply anesthetized with pentobarbital and intracardially perfused with saline followed by
10% formalin. The brains were dissected and stored in
formalin. Several days after, they were cut with a freezing
microtome. Coronal sections stained with cresyl violet were
examined for injection needle tracks. The histological
analysis confirmed the location of the injection needles
above the PBN, while showing negligible tissue damage
in a small area circumscribed to the needle track. Given
that the blockade induced by 10 ng of TTX affects a spherical volume of tissue about 3 mm in diameter [18], the
entire PBN was inactivated.
The main finding of Exp 2 was that the retrograde disruption of CTA induced by PBN inactivation applied 30 min
172
M.A. Ballesteros et al. / Neuroscience Letters 322 (2002) 169–172
after training in Exp 1 can not be exclusively attributed to
deficits of LiCl aversive properties processing. The presence
of the US pre-exposure effect in those groups with PBN
blockade 30 min after each LiCl pre-exposure demonstrates
that this time is enough for some of the aversive effects of
i.p. LiCl injections to be processed and to act as a US in
CTA. This is consistent with previous data showing that the
behavioral effects of an i.p. injection of lithium chloride
may appear 10 min later [4] and those indicating that 20
min are enough to associate LiCl aversive effects with a
taste [15]. Moreover, the absence of the US pre-exposure
phenomenon in those groups receiving TTX microinjections
immediately after lithium pre-exposures demonstrate that
PBN was effectively inactivated and that the activity of
the area is critical for processing the US properties of
LiCl in this specific CTA task. This is consistent with the
well known role of this area in processing the aversive
properties of LiCl, that are relevant for CTA [2,12–14].
The acquisition of learned taste aversions by the non-preexposed groups subjected to previous PBN blockade, that
were of similar magnitude to those aversions seen in the
control cannulated groups, confirmed the histological findings and previous data demonstrating that three TTX injections did not induce permanent PBN lesions [5].
In all, the present results definitively support the proposed
PBN role in the taste-lithium association and memory formation using a standard CTA procedure [4,7,8,11,15–17].
This research was supported by the CICYT grants PB981309 and PB98-1362 (Spain). Morón was recipient of a
predoctoral grant of the Junta de Andalucı́a (Spain). The
authors are greatly indebted to Ms M. Burnett for her helpful
suggestions with the English. We also wish to thank J.C.
Rodriguez Garcia and A. Molina for their technical assistance.
[1] Aguado, L., Brugada, I. and Hall, G., Effects of retention
interval on the US preexposure phenomenon in flavor aversion learning, Learn. Motiv., 28 (1997) 311–322.
[2] Agüero, A., Arnedo, M., Gallo, M. and Puerto, A., The functional relevance of the lateral parabrachial nucleus in
lithium chloride-induced aversion learning, Pharmacol.
Biochem. Behav., 45 (1993) 973–978.
[3] Bielavska, E. and Bures, J., Universality of parabrachial
mediation of conditioned taste aversion, Behav. Brain
Res., 60 (1994) 35–42.
[4] Bures, J., Bermudez-Rattoni, F. and Yamamoto, T., Conditioned Taste Aversion: Memory of a Special Kind, Oxford
University Press, Oxford, 1998.
[5] Gallo, M., Bielavska, E., Roldán, G. and Bures, J., Tetrodotoxin inactivation of the gustatory cortex disrupts the effect
of the NMDA antagonist ketamine on latent inhibition of
conditioned taste aversion in rats, Neurosci. Lett., 240
(1998) 61–64.
[6] Gallo, M., Márquez, S.L., Ballesteros, M.A. and Maldonado,
A., Functional blockade of the parabrachial area by tetrodotoxin disrupts the acquisition of conditioned taste aversion
induced by motion-sickness in rats, Neurosci. Lett., 265
(1999) 57–60.
[7] Ivanova, S.F. and Bures, J., Acquisition of conditioned taste
aversion in rats is prevented by tetrodotoxin blockade of a
small midbrain region centered around the parabrachial
nuclei, Physiol. Behav., 48 (1990) 543–549.
[8] Ivanova, S.F. and Bures, J., Conditioned taste aversion is
disrupted by prolonged retrograde effects of intracerebral
injection of tetrodotoxin in rats, Behav. Neurosci., 104
(1990) 948–954.
[9] Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic
Coordinates, San Diego, Academic Press, 1986.
[10] Randich, A. and Lolordo, V.M., Associative and non-associative theories of the UCS preexposure phenomenon:
implications for Pavlovian conditioning, Psychol. Bull., 86
(1979) 523–548.
[11] Reilly, S., The parabrachial nucleus and conditioned taste
aversion, Brain Res. Bull., 48 (1999) 239–254.
[12] Reilly, S. and Trifunovic, R., Lateral parabrachial nucleus
lesions in the rat: aversive and appetitive gustatory conditioning, Brain Res. Bull., 52 (2000) 269–278.
[13] Reilly, S. and Trifunovic, R., Lateral parabrachial nucleus
lesions in the rat: neophobia and conditioned taste aversion, Brain Res. Bull., 55 (2001) 359–366.
[14] Sakai, N. and Yamamoto, T., Role of the medial and lateral
parabrachial nucleus in acquisition and retention of conditioned taste aversion in rats, Behav. Brain Res., 93 (1998)
63–70.
[15] Spector, A.C., Breslin, P. and Grill, H.J., Taste reactivity as a
dependent measure of the rapid formation of conditioned
taste aversion: a tool for the neural analysis of taste-visceral
associations, Behav. Neurosci., 102 (1988) 942–952.
[16] Spector, A.C., Gustatory function in the parabrachial nuclei:
implications from lesion studies in rats, Rev. Neurosci., 6
(1995) 143–175.
[17] Yamamoto, T., A neural model for taste aversion learning,
In K. Kurihara and N. Suzuki (Eds.), Olfaction and Taste XI,
Springer-Verlag, Tokyo, 1994, pp. 471–474.
[18] Zhuravin, I.A. and Bures, J., Extent of tetrodotoxin induced
blockade examined by pupillary paralysis elicited by intracerebral injection of drug, Exp. Brain Res., 93 (1991) 687–
690.