Download Mapping Pavlovian Conditioning Effects on the Brain: Blocking

Mapping Pavlovian Conditioning Effects on the Brain:
Blocking, Contiguity, and Excitatory Effects
DIRK JONES AND F. GONZALEZ-LIMA
Department of Psychology and Institute for Neuroscience, University of Texas at Austin, Austin, Texas 78712
Received 26 December 2000; accepted in final form 17 April 2001
In modern learning theory, Pavlovian conditioning has come
to be understood as more than simple associations; evidence
shows that organisms learn new responses based on their
current knowledge (Domjan et al. 2000). The blocking effect
on Pavlovian conditioning (Kamin 1969) provided an early
demonstration that learning about a stimulus depended not
simply on contiguity but on previous learning. In the Kamin
blocking effect, the presence of a previously conditioned stimulus blocks the conditioned response to a newly introduced
stimulus. For example, subjects exposed to a compound stimulus (tone plus light) paired with shock (Fig. 1A) exhibit
conditioned responses to both the tone and the light. However,
subjects first exposed to a number of light and shock pairings
(phase I, Fig. 1B) followed later by the tone/light compound
paired with the shock (phase II, Fig. 1B), show conditioned
responses to the light but not to the tone. Of importance is that
all subjects received the same tone-shock contiguity. In other
words, in both groups the tone occurred in close temporal
proximity with the shock. However, conditioned responses to
the tone are “blocked” in subjects with a history of light-shock
pairings. This blocking phenomenon has been regarded as one
of the most significant observations in Pavlovian conditioning
(Fanselow 1998). In addition, recent findings (Jones et al.
1997) showing that schizophrenics lack the development of
blocking render this phenomenon of interest to a wider audience.
Neuroimaging techniques using glucose analogues (Sokoloff 1992) such as fluorodeoxyglucose (FDG) autoradiography
have increased the ability to study the functioning nervous
system, including behavioral and learning effects (GonzalezLima 1992). This approach has made possible the simultaneous
examination of functional maps of learning-related activity
throughout the rat brain. For example, researchers using this
approach were the first to report functional brain maps produced by habituation and arousal (Gonzalez-Lima and Scheich
1984, 1985; Gonzalez-Lima et al. 1989; Toga and Collins
1981), conditioned excitation (Gonzalez-Lima and Scheich
1984, 1986), conditioned inhibition (McIntosh and GonzalezLima 1994, 1998), and extinction (Nair and Gonzalez-Lima
1999). In addition, differential conditioning of tones has been
shown to evoke specific metabolic differences in a thalamostriate circuit as well as in the auditory system (Gonzalez-Lima
1992; Gonzalez-Lima and Agudo 1990; Scheich et al. 1997).
McIntosh and Gonzalez-Lima (1993–1995) have shown FDG
increases in the auditory system, basal forebrain, and cerebellum evoked by a tone conditioned as an excitor relative to the
same tone conditioned as inhibitor. However, little work has
been done on the neural mechanisms of the Kamin blocking
effect.
Most of the available neurobiological studies of blocking
have assessed the effects of lesions in the hippocampus or the
amygdala (Gallo and Candido 1995; Holland and Gallagher
1993; Rickert et al. 1978; Solomon 1977). Lesions of the
amygdala did not prevent blocking (Holland and Gallagher
Address for reprint requests: F. Gonzalez-Lima, Behavioral Neuroscience,
Mezes Hall 330, University of Texas, Austin, TX 78712 (E-mail: [email protected]).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked ‘‘advertisement’’
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
INTRODUCTION
www.jn.org
0022-3077/01 $5.00 Copyright © 2001 The American Physiological Society
809
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
Jones, Dirk and F. Gonzalez-Lima. Mapping Pavlovian conditioning effects on the brain: blocking, contiguity, and excitatory effects. J
Neurophysiol 86: 809 – 823, 2001. Pavlovian conditioning effects on
the brain were investigated by mapping rat brain activity with fluorodeoxyglucose (FDG) autoradiography. The goal was to map the
effects of the same tone after blocking or eliciting a conditioned
emotional response (CER). In the tone-blocked group, previous learning about a light blocked a CER to the tone. In the tone-excitor group,
the same pairings of tone with shock US resulted in a CER to the tone
in the absence of previous learning about the light. A third group
showed no CER after pseudorandom presentations of these stimuli.
Brain systems involved in the various associative effects of Pavlovian
conditioning were identified, and their functional significance was
interpreted in light of previous FDG studies. Three conditioning
effects were mapped: 1) blocking effects: FDG uptake was lower in
medial prefrontal cortex and higher in spinal trigeminal and cuneate
nuclei in the tone-blocked group relative to the tone-excitor group. 2)
Contiguity effects: relative to pseudorandom controls, similar FDG
uptake increases in the tone-blocked and -excitor groups were found
in auditory regions (inferior colliculus and cortex), hippocampus
(CA1), cerebellum, caudate putamen, and solitary nucleus. Contiguity
effects may be due to tone-shock pairings common to the toneblocked and -excitor groups rather than their different CER. And 3)
excitatory effects: FDG uptake increases limited to the tone-excitor
group occurred in a circuit linked to the CER, including insular and
anterior cingulate cortex, vertical diagonal band nucleus, anterior
hypothalamus, and caudoventral caudate putamen. This study provided the first large-scale map of brain regions underlying the Kamin
blocking effect on conditioning. In particular, the results suggest that
suppression of prefrontal activity and activation of unconditioned
stimulus pathways are important neural substrates of the Kamin
blocking effect.
810
D. JONES AND F. GONZALEZ-LIMA
1993), and hippocampal lesions have had mixed results (Garrud et al. 1984; Rickert et al. 1978; Solomon 1977). Hippocampal unit responses to a blocked light appear at the first compound trial but do not occur at the end of conditioning
(Neuenschwander-El Massioui et al. 1991). Identification of
the neural mechanisms involved in the blocking effect may
help to differentiate among various theories of blocking of
Pavlovian conditioning based on behavioral studies because
there is conflicting behavioral evidence for each theory.
For example, some theories limit attention to a redundant
conditioned stimulus (CS) as an explanation of why a blocked
CS fails to elicit a conditioned response (Kamin 1969; Mackintosh 1983; Pearce and Hall 1980). These theories emphasize
changes in processing of the blocked CS. In contrast, the
influential Rescorla-Wagner model (1972) limits the amount of
learning that an unconditioned stimulus (US) can support. This
theory emphasizes changes in US processing as an explanation
of the blocking effect. A third group of theories views blocking
as an instance of conditioned response (CR) inhibition independent of changes in CS or US processing (Balaz et al. 1982;
J Neurophysiol • VOL
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
FIG. 1. Schematic depiction of the ideas of contiguity and the Kamin
blocking effect. The lines indicate time of presentation of various stimuli. In
Pavlovian conditioning to a single stimulus, the conditioned and unconditioned
stimuli (CS) and (US) are presented in close temporal proximity (A) and a
conditioned emotional response is elicited. B: the 2 phases of training in the
blocking design. CSE is the light paired with shock US that is conditioned as
excitor (elicits a conditioned response) in phase I. CSB is the tone CS that is
paired with the shock US in compound with the light during phase II training.
In the tone-blocked group, this tone CS is blocked (fails to elicit a conditioned
response) due to phase I training.
Miller et al. 1993; Schachtman et al. 1983). Although these
three types of theories emphasize either CS, US or CR behavioral mechanisms, their implications can be considered in light
of functional neural data.
For example, differences in neural modification of CS pathways by excitor and blocked CSs would provide support for CS
inattention interpretations of blocking (e.g., Mackintosh 1983;
Pearce and Hall 1980). Neuronal recordings in the visual
system have been carried out during administration of a modified blocking procedure (Kinkaide and Walley 1974), but a
clear interpretation of the resulting data regarding the blocking
effect is confounded by the presence of inhibitory training.
Recent work using the presence of fos protein has implicated
the auditory and visual parts of the thalamic reticular nucleus
as mediating attentional properties involved in blocking (McAlonan et al. 2000). Although interesting, this fos study
showed no convincing evidence of blocking because the compound stimuli were as effective as the blocked stimuli. Hence
the results can be explained not by a blocking effect but by
differential excitatory effects of CSs as we have shown previously in the reticular thalamic nucleus using FDG uptake after
differential conditioning (Gonzalez-Lima 1992).
Auditory CS learning-dependent modification of neuronal
activity in the primary auditory system and its secondary
projection fields is well documented (Gonzalez-Lima and
Agudo 1990; Gonzalez-Lima and Scheich 1984, 1986; Scheich
et al. 1997; Weinberger 1995; Weinberger and Diamond 1987).
Some have interpreted these increases in neuronal activity as
underlying increases in attention to a particular CS (Mesulam
1990; Posner et al. 1987). This laboratory published a report of
changes in the auditory system during blocking of a tone CS
using a marker of metabolic capacity, cytochrome oxidase
(Poremba et al. 1997). The regions with cytochrome oxidase
differences were limited to areas of the auditory system receiving US somatosensory inputs, such as the dorsal cochlear
nucleus, the inferior colliculus, and the secondary auditory
cortex. These regions receive direct anatomical convergence of
the CS (auditory) and US (somatosensory) pathways, suggesting that the influence of US pathways may be relevant for the
blocking effect. This provided support for a US-based explanation of blocking. Additional evidence suggests that blocking
of eyeblink conditioning depends on a negative feedback circuitry related to the US (Kim et al. 1998); this is also consistent
with the US-based interpretation of blocking afforded by the
Rescorla-Wagner model (Rescorla and Wagner 1972).
The current paper investigates the Kamin blocking effect
and contiguity effects in a FDG mapping of rat auditory and
extra-auditory regions implicated in learning-dependent,
evoked activity differences in our previous FDG studies of
conditioning. The results suggest that the blocking of conditioned emotional responses (CER) to a tone is not simply
dependent on reductions in CS auditory processing pathways
as would be anticipated by CS inattention theories or reductions in extra-auditory circuits implicated in CS-US contiguity
as would be anticipated by an attention or CS-US associative
deficit. Rather, blocking of an excitatory tone CER may result
from metabolic suppression of the prefrontal cortex as would
be anticipated by theoretical accounts of blocking based on CR
inhibition (Balaz et al. 1982; Miller et al. 1993; Schachtman et
al. 1983) and activation of lower US somatosensory pathways
NEUROIMAGING OF CONDITIONING AND BLOCKING EFFECTS
as would be anticipated by theoretical accounts of blocking
based on US saturation (Rescorla and Wagner 1972).
D. Jones submitted this work in partial fulfillment of the
requirements for a Ph.D. degree at the University of Texas at
Austin.
METHODS
Subjects and apparatus
Experimental stimuli
Two Wavetek sweep/modulation generators produced the acoustic
stimuli. Tones were presented through two speakers, one mounted on
top and one mounted in the front of the operant chamber, to produce
whole-field stimulation. Two frequency-modulated (FM) acoustic
stimuli were used: a low-frequency FM tone (1–2 kHz, 65 dB SPL)
and a high-frequency FM tone (10 –20 kHz, 65 dB SPL). The lowfrequency tone was used as the CS, while the high-frequency tone was
used as a comparator stimulus. A visual stimulus was provided by two
flashing white lights (intensity of 200 foot-candles measured 2 cm
from the source). The light source was outside the clear Plexiglas side
of the chamber approximately 5 cm from the front. Background
illumination was provided by a red light mounted outside the rear of
the chamber (intensity of 100 ft.-cd measured 2 cm from the source).
The grid floor of the chamber was wired for US delivery to a Lafayette
Instruments master shocker (Lafayette IN). The US was a 0.75-s,
0.5-mA footshock. In CS-US paired trials, the US co-terminated with
the 15-s CS. Intertrial intervals ranged randomly between 60 and
180 s. Presentation of stimuli and collection of drink data were
controlled by computer programs created using MED-PC behavioral
programming language (MED Associates).
This average level of baseline drinking was used to order subjects and
alternatively assign them to the three groups. This was done to obviate
any group differences in individual water consumption prior to formal
training (Table 1). During the course of the experiments, several
subjects were lost to difficulties in tissue processing and data acquisition. This resulted in n ⫽ 6 for the tone-blocked group, n ⫽ 7 for the
tone-excitor group, and n ⫽ 7 for the pseudorandom group. The
previous groups were equated on stimulus history, therefore the use of
additional CS alone or US alone groups were avoided as they would
be confounded by different stimulus histories.
Training followed three basic phases; phase I consisted of light
training, phase II consisted of compound training, and phase III was
FDG testing. In the tone-blocked group, phase I excitatory training
consisted of conditioning the flashing light as an excitatory CS (i.e.,
light will evoke CER) through four reinforced trials a day on days
5– 8. In the tone-excitor group, the light and shock were presented
four trials a day for days 5– 8. Unlike the tone-blocked group, all but
one of the light and US presentations were unpaired in the toneexcitor group. The light presentations were given in one session, and
US presentations were in a different session on the same day. These
sessions were separated by at least 1 h. The remaining light and shock
stimulus in the tone-excitor group was given as a paired presentation
that randomly occurred on training day 6 or 7. This resulted in a 100%
probability of light-shock pairing in the tone-blocked group and a
probability of 0.0625 for light-shock pairing in the tone-excitor group.
This was done to minimize any potential for the light in the toneexcitor group to acquire properties consistent with conditioned excitation or inhibition. Phase II proceeded identically for both groups.
The tone/light compound was paired with footshock four times a day
on days 9, 10, and 14 for a total number of 12 reinforced presentations.
The pseudorandom group received pseudorandom presentations of
the same tones, lights, and foot shocks. These were not explicitly
unpaired stimuli presentations because there was less than a 0.04
probability of tone or light being paired with footshock (1 light/
footshock and 1 tone/footshock pairing over 7 days). This served to
prevent the stimuli from acquiring a reliable predictive value of
footshock delivery (Poremba et al. 1998). Therefore the number of
days and stimuli presentations was the same in all groups. For a
summary of the experimental design, see Table 2.
In all subjects, 3 days of probe trials measured the degree of
conditioning to the context, light, tone, and a comparator stimulus on
days 11–13 and 15. Day 11 consisted of exposure to the training
context and observation of any contextual fear conditioning. The
probe sessions on days 12 and 13 consisted of tests of responding to
three presentations of the discrete stimuli as well as summation tests
TABLE
1. Summary of behavioral training
Description
Day
Behavioral protocol
Drinking baseline in operant chamber
Phase I trials with four paired or unpaired light & US
presentations
Phase II excitatory training with 4 reinforced compound
trials
Post train probe of behavior to context, no discrete stimuli
presented
Post train probe of conditioning, 3 presentations of each
stimulus
Reinstatement training with 4 reinforced compound trials
Probe on day of FDG test, 3 presentations of tone alone
FDG post injection session with target tone
Training involved the CER or conditioned fear paradigm. In this
protocol, training establishes a stimulus as predictive of an aversive
event and the degree of conditioning is assessed by measuring disruption of ongoing behavior such as drinking. Total drink time during
a 15-min period in the training context was collected on days 1– 4.
Total number of stimulus presentations is equal in the three groups. Only
reinforcement history with the light stimulus in phase I is different between
blocked and excitor groups. The pseudorandom group had only one paired trial
with the light and one with the tone. US, unconditioned stimulus; FDG,
fluorodeoxyglucose.
J Neurophysiol • VOL
86 • AUGUST 2001 •
www.jn.org
1–4
5–8
9–10
11
12–13
14
15
15
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
Subjects were 32 male Long-Evans black-hooded rats weighing an
average of 100 g when received from the supplier (Harlan, Indianapolis, IN). Subjects were housed in Association for Assessment and
Accreditation of Laboratory Animal Care-approved facilities under
standard laboratory conditions, two to three to a cage with free access
to food and water. Following 1 wk of acclimation and handling, all
subjects were placed on restricted water for the duration of training.
Training was conducted in a standard operant conditioning chamber
as described elsewhere (Poremba et al. 1997, 1998). Two operant
chambers (28 ⫻ 22 ⫻ 22 cm, model No. 80000, MED Associates, St.
Albans, VT) were used for training and testing. These chambers were
enclosed in sound-attenuating boxes to eliminate extraneous acoustic
stimuli. The front and back walls of the operant chambers were
aluminum and the sides were clear Plexiglass. A 5-cm hole in the front
wall of each chamber allowed access to a drinking tube. An infrared
beam from a photo source module was focused directly in front of the
drinking tube through clear glass rods (MED Associates). This was
used to monitor drink time and ensure that the beam was broken only
by drinking rather than some other activity (i.e., exploration of recess,
freezing with head in the recess). Time spent drinking was recorded
on a AT&T PC 6300. Water was provided for 15 min in the training
context during baseline training trials and for a 45-min period in the
home cage daily. Protocols for these experiments were approved by
the University of Texas at Austin Institutional Animal Care and Use
Committee, and conform to all applicable federal and National Institutes of Health guidelines for the care and use of animals in research.
811
812
D. JONES AND F. GONZALEZ-LIMA
TABLE
2. Experimental design
Groups
Phase I
Phase II
Tone CER
FDG Test
Tone blocked
Tone excitor
Pseudorandom
L3S
L, S
L, S
TL 3 S
TL 3 S
T, L, S
No
Yes
No
T
T
T
T, 1 to 2-kHz tone FM sweep; L, flashing light; TL, tone/light compound;
S, footshock. Both the blocked and excitor groups have tone conditioned
stimulus (CS)-US contiguity in phase II, but the blocked group does not show
a conditioned emotional response (CER) to the tone while the excitor group
does. The pseudorandom group has no reliable contiguity and does not show
a CER.
Tissue processing and autoradiography
On the day of FDG administration, all animals were allowed
15-min free access to water prior to the FDG uptake period. This was
done to equate groups on water access and avoid dehydration during
the FDG uptake period. FDG administration [an intraperitoneal injection of 18 ␮Ci/100 g body wt of universally labeled [14C]2-deoxy-2fluoro-glucose (2-DG; 360 mCi/mM specific activity; American Radiolabeled Chemicals, St. Louis, MO)] in 0.36 ml sterile saline
followed. FDG was used, as opposed to 2-DG, because rates of
J Neurophysiol • VOL
Imaging analysis
Using the same slides previously exposed to X-ray film, sections
were stained for Nissl substance with cresyl violet and used to
corroborate selection of regions of interest for analysis. A stereotaxic
atlas (Paxinos and Watson 1986) and our Nissl/cytochrome oxidase
atlas (Gonzalez-Lima and Cada 1998) of the rat brain were used to
identify brain sections that approximately matched the levels of sections between Bregma ⫹3.7 and ⫺13.3 mm. At each Bregma level,
three adjacent sections up to 120 ␮m of anteroposterior distance were
sampled (Fig. 4). Both gray and white matter landmarks were used to
locate each region analyzed at the same level in all subjects. The
regions of interest (ROI) were selected based on our previous FDG
studies of conditioning that showed tone conditioning effects on
auditory and extra-auditory integrative regions (Gonzalez-Lima 1989,
1992; Gonzalez-Lima and Agudo 1990; Gonzalez-Lima and Scheich
1984 –1986; Gonzalez-Lima et al. 1989). In addition, control regions
were analyzed including white matter (optic tract and spinal trigeminal tract) and regions to control for nonspecific motor and arousal
effects such as primary motor cortex (lateral frontal cortex) and
reticular formation (lateral and magnocellular reticular nuclei).
The autoradiographs were analyzed using an image-processing
system consisting of a high-gain video camera, a Targa-M8 image
capture board, a 486 computer, Sony color monitor, DC-powered
illuminator, and JAVA software (Jandel Scientific, San Rafael, CA).
The system was calibrated using the plastic 14C standards from
Amersham.
Multiple readings were taken from left and right sides of each
section to avoid artifacts. For each region measured, the size of the
densitometer window was set to approximate the size of the whole
region. In cases of odd-shaped ROIs, a smaller size window was used,
and adjacent measurements were taken and averaged to sample the
entire region. The averaged measures resulted in one mean value per
region for each subject. The size of the window was held constant
across subjects as was the number of readings for each ROI. Thus the
area measured was constant across all animals and was limited to the
anatomical region of interest.
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
of compound stimuli (Table 1). On day 15, 1 h prior to the FDG
session, a final probe of three presentations of the CS tone alone was
used to confirm behavioral effects before the FDG administration.
Measures of drink time and behavioral counts of freezing were taken
during the probe sessions.
As an index of excitatory conditioning, suppression of drinking was
measured and reported in the form of a suppression ratio. Time spent
drinking was collected during a 15-s period prior to the onset of the
CS and during the 15-s CS presentation period. Drink time prior to CS
presentation (pre-CSdrink-time) was used as an indication of baseline
drinking, while drink time during the CS presentation (CSdrink-time)
indicated any disruption caused by the stimulus presentation. A ratio
that expressed this behavioral suppression was calculated as follows:
CSdrink-time/(pre-CSdrink-time ⫹ CSdrink-time) (Kamin 1969). Suppression ratios range from 0 to 0.5. A value of zero indicates strong
conditioned suppression to the CS, and 0.5 indicates no change in
drinking behavior as a result of the CS presentation.
In addition to suppression measures, direct measures of freezing
behavior were taken. Researchers blind to experimental condition
monitored each subject’s behavior. Behavior was monitored in 3-s
intervals during the 15-s CS presentations. In each period, a freezing
response was recorded if the rat was standing motionless with short
rapid breathing as indicated by slight body movement. A total of five
counts were possible for each 15-s CS period. For example, if a
subject exhibited freezing behavior continuously during the CS period, a count of five for freezing was recorded.
At the time of the FDG uptake test, all subjects were presented with
the same low-frequency tone unreinforced in a context changed from
that of training. Adding a smooth metal plate on the floor, placing
horizontal black lines over the Plexiglas front and swabbing an iodine
surgical scrub solution (Betadine) on the walls and floor of the
chamber changed the context. Since rats show learned responses to the
training context (Miller et al. 1993), it was important to change the
context to control for context generated cues. While the effect of
associations involving contextual cues on brain metabolism is interesting, the present study was aimed at revealing the effects of a
discrete CS. Presentation of the CS outside of the training context
leads to robust CERs in this paradigm (McIntosh and Gonazalez-Lima
1993–1995, 1998). Furthermore, changes in context following phase
II training does not attenuate the blocking effect (Giftakis et al. 1998).
During the FDG uptake period, the brain effects were evoked by the
tone CS rather than other learned cues.
transport of FDG across the blood-brain barrier and phosphorylation
in the cell are significantly better than those of 2-DG and because
FDG universal labeling allows for a greater signal (Gonzalez-Lima
1992). Immediately following injection, the subject was placed in the
test chamber, and stimulus presentation was initiated. The test period
was 60 min, during which 5-s tone (1–2 kHz sweep) presentations
were separated by 1-s intervals as this stimulation has resulted in
optimal evoked FDG uptake and no evidence of CER extinction under
our testing conditions (McIntosh and Gonzalez-Lima 1993–1995,
1998).
Following completion of the testing period, animals were removed
from the testing room and immediately decapitated. Brains were
removed rapidly and frozen in isopentane at ⫺40°C. The brain was
then mounted and sectioned at 40 ␮m at ⫺20°C in a cryostat
(Reichert-Jung 2800 Frigocut E). Sections were picked up on slides at
room temperature and dried on a hot plate at 60°C for later processing
with X-ray film.
Dried slides were mounted on black paper, opposed to Kodak EB-1
X-ray film and placed in Kodak X-O-Matic cassettes. Acrylic microscale standards of known 14C concentrations (Amersham, Arlington
Heights, IL) were placed with each piece of film. Following a 2-wk
exposure period, film was developed in Kodak D-19 developer at
25°C for 2 min, rinsed in Kodak indicator stop, fixed for 10 min, and
rinsed in water for 10 min. Variability in exposure time and development was obviated by the use of standards, and further readings
were automatically expressed as 14C isotope incorporation (nCi/g
tissue).
NEUROIMAGING OF CONDITIONING AND BLOCKING EFFECTS
813
Zij(group 2))/公(1/ng1) ⫹ (1/ng2) was used to compare across group
differences evaluated at ␣ ⫽ 0.01 level. The use of this data set of
correlations for structural equation modeling revealed evidence of
multicollinearity: i.e., there were several correlations above 0.8 in
each group. This presents problems in drawing inferences about the
nature of the path coefficients in structural modeling (Asher 1983).
The total coefficient of determination (an index of the amount of
variation in the observed variables explained by the model) should
have a value between zero and one. This ratio of the determinants of
the residual to fitted covariance matrices (Jöreskog and Sörbom 1989)
was below 0.5 for all groups, suggesting the model was accounting for
little of the observed variance. In some cases, this value was negative
(pseudorandom ⫽ ⫺1.16; excitor ⫽ ⫺0.84), indicating that the observed variance was being accounted for by residual values rather than
the path coefficients. For these reasons, it was concluded that covariance structural equation modeling was not informative as applied to
this particular data set.
Statistical analysis
In addition to the individual regions of interest, a measure of global
brain activity was taken by averaging the isotope incorporation from
the entire set of sections in the series of a given subject. This global
measure of activity was used as a reference for group differences in
overall metabolic activity. These global measures did not differ in the
experimental groups (P ⬍ 0.98, 2-tailed t-test), but there was a
significantly lower level of global FDG labeling in the pseudorandom
group. Due to this overall difference, regional FDG incorporation was
adjusted based on white-matter readings, and the estimated marginal
means were compared. Following this correction, measures of whole
brain activity did not differ across groups. An ANOVA was performed following the statistical removal of nonspecific differences in
overall FDG incorporation. Data are presented in the form of mean
FDG uptake for each ROI adjusted by the covariate white-matter
readings. The statistical significance of the results at P ⬍ 0.01 was
tested with confidence intervals produced through the ANCOVA
analysis. Each ROI comparison is treated as independent of the others
(a premise of regional brain mapping) and one simply accepts that 1
of 100 comparisons (for P ⫽ 0.01) may be type 1 errors. This is a
standard procedure in neuroimaging studies that cannot apply any
correction for multiple comparisons due to the large number of ROIs
analyzed (Nobrega 1992). All structures were analyzed at P ⬍ 0.01
level, and systematic bias within group error variance were minimized
through the use of the ANCOVA procedure.
Behavioral data were analyzed by repeated measures ANOVA with
tests for simple effects where appropriate. The covariance structure of
the ROIs was also examined. Pairwise correlations were calculated
between pairs of normalized values of regional glucose metabolic
activity within each group using SPSS 8.0 software. To control for
inflation of family-wise error rates due to the large number of correlations relative to the sample size, a “jackknife” procedure was used
(Shao and Dongsheng 1995). In this procedure, the calculation of
pairwise correlations is repeated for each possible n ⫺ 1 subset of the
sample. Only those correlations that are significant across all subsets
are used for the final analysis. Thus spuriously high correlations that
are dependent on a single subject are screened out.
Areas that survived the jackknife procedure were tested to see if
they had a correlation coefficient that differed significantly between
groups. The Fisher Z transformation was used to convert each correlation to a Z score. A test statistic of the form Z ⫽ (Zij(group 1) ⫺
J Neurophysiol • VOL
RESULTS
Behavioral results
Conditioned suppression was confirmed in probe trials prior
to FDG administration. The presentation of the tone CS was
expected to produce distinct behaviors in the three groups. In
the tone-blocked group, little or no interruption or suppression
of drinking was predicted for the tone, whereas in the toneexcitor group, significant suppression of drinking was expected
for the tone. This constitutes the traditional comparison used to
demonstrate the Kamin blocking effect. The light was expected
to produce CERs in both groups, as it was paired with shock in
phase I of the tone-blocked group and phase II of both groups
(Table 2). The pseudorandom group was expected to show no
CERs to the tone or the light.
The conditioned suppression ratio for the tone CS in the
FIG. 3. Freezing data for the 3 groups to the different test stimuli in the
probe sessions. Behavioral counts represent the mean of observed behavior
during the 2nd 2 15-s CS presentations. Each 15-s CS period was broken down
into 5 bins of 3 s each. The significantly lower freezing count (*) in the
tone-blocked relative to the tone-excitor group is further support that the light
was effective in blocking the behavioral expression of conditioned emotional
response (CER) to the tone. As expected, both blocked and excitor groups
showed significantly greater freezing to the light alone or in compound with
the tone as compared with the pseudorandom group.
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
FIG. 2. Suppression ratios for the tone alone probe session on day 15. A
score near 0.0 indicates strong conditioned responding, while a score near 0.5
indicates no change in drinking with the presentation of the tone. Some
suppression seen on the 1st presentation of the tone was expected as animals
will suppress to any initial stimulus presentation when first in the training
boxes. The significantly lower suppression ratios (*) in the tone-excitor group
compared with the tone-blocked group are evidence of the blocking effect
[F(1, 12) ⫽ 101.47, P ⬍ 0.0001, repeated-measures ANOVA].
814
D. JONES AND F. GONZALEZ-LIMA
tone-blocked group was always significantly higher than that of
the tone-excitor group [F(1, 12) ⫽ 101.47, P ⬍ 0.0001, repeated-measures ANOVA], indicating the previous training
with the light effectively blocked conditioned suppression to
the tone in the tone-blocked group (Fig. 2).
However, the novel high-frequency comparator tone failed
to produce significantly greater suppression of drinking in the
tone-blocked group (0.28 ⫾ 0.05; mean ⫾ SE). A nonsignificant suppression ratio to the comparator tone in the toneexcitor group (0.18 ⫾ 0.06) is most probably indicative of
some generalization from the CS tone. As expected, both
conditioned groups showed a strong suppression to the light
alone (tone excitor, 0.14 ⫾ 0.04; tone blocked, 0.03 ⫾ 0.02)
and to the compound (tone excitor, 0.06 ⫾ 0.02; tone blocked,
0.03 ⫾ 0.02).
Furthermore there was significantly less freezing to the CS
tone in the tone-blocked group than in the tone-excitor group
[F(1, 11) ⫽ 9.18, P ⬍ 0.03; Fig. 3]. As anticipated, the
pseudorandom group showed no evidence of CERs, with mean
behavioral counts of freezing of 0.93 ⫾ 0.13 (Fs ⬍ 1.0).
The possibility that group differences in responding to the
tone result from tone-excitor group enhanced responding to
the tone due to putative inhibitory conditioning to the light
J Neurophysiol • VOL
is unlikely. The possibility of an inhibitory association with
the light was lessened by the lack of an explicitly unpaired
light-US relationship (probability of pairing ⫽ 0.0625).
Additionally, the similar level of freezing to the light in both
the tone-excitor and tone-blocked group is inconsistent with
inhibitory conditioning of the light in the tone-excitor group
(Fig. 3).
FDG uptake in the auditory system
Figure 4 illustrates the location of ROIs and highlights
significant differences across groups. There were no differences in mean FDG uptake evoked by the tone in the auditory
system between tone-blocked and -excitor groups (Table 3).
But both conditioned groups showed regional means significantly greater than the pseudorandom group at P ⬍ 0.01. These
conditioning effects were localized to the inferior colliculus
and the auditory cortex. The largest effect was in the central
nucleus of the inferior colliculus, as in the first FDG studies of
Pavlovian conditioning (Gonzalez-Lima and Scheich 1984,
1986). The central nucleus in tone-blocked and -excitor groups
showed mean increases of 36% as compared with the effect of
the same tone in the pseudorandom group.
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
FIG. 4. Images taken from the autoradiographs representing
the approximate levels at which readings were taken. Regions
of interest are indicated, and areas that reached statistical
significance (P ⬍ 0.01) following the ANCOVA analysis are
labeled. Bregma levels are indicated below each level.
NEUROIMAGING OF CONDITIONING AND BLOCKING EFFECTS
TABLE
815
3. FDG uptake in the auditory structures
Structure
Dorsal cochlear nucleus (DCN)
Ventral cochlear nucleus-posterior (VCP)
Ventral cochlear nucleus-anterior (VCA)
Lateral superior olivary nucleus (LSO)
Medial superior olivary nucleus (MSO)
Trapezoid body nucleus (TBN)
Lateral lemniscus nucleus-intermediate (ILL)
Lateral lemniscus nucleus-ventral (VLL)
Inferior colliculus nucleus-dorsal (ICD)
Inferior colliculus nucleus-central (ICC)
Inferior colliculus nucleus-external (ICE)
Medial geniculate nucleus-dorsal (MGD)
Medial geniculate nucleus-medial (MGM)
Medial geniculate nucleus-ventral (MGV)
Auditory cortex, area 1 (TE1)
Auditory cortex, area 2 (TE2)
Auditory cortex, area 3 (TE3)
Mean
Lower 99%
Upper 99%
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor1
Pseudorandom
Blocked1
Excitor1
Pseudorandom
Blocked1
Excitor1
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked1
Excitor1
Pseudorandom
Blocked1
Excitor1
Pseudorandom
Blocked1
Excitor1
Pseudorandom
1051 ⫾ 56
1091 ⫾ 52
1004 ⫾ 67
792 ⫾ 39
830 ⫾ 36
762 ⫾ 46
803 ⫾ 34
813 ⫾ 31
730 ⫾ 40
1084 ⫾ 52
1122 ⫾ 48
946 ⫾ 62
977 ⫾ 46
1011 ⫾ 42
876 ⫾ 55
660 ⫾ 30
668 ⫾ 28
640 ⫾ 36
645 ⫾ 28
666 ⫾ 26
617 ⫾ 33
746 ⫾ 35
745 ⫾ 33
704 ⫾ 42
778 ⫾ 38
793 ⫾ 35
764 ⫾ 45
729 ⫾ 39
780 ⴞ 36
641 ⫾ 46
1605 ⴞ 74
1598 ⴞ 69
1175 ⫾ 89
916 ⴞ 41
944 ⴞ 38
773 ⫾ 49
765 ⫾ 49
733 ⫾ 46
598 ⫾ 59
747 ⫾ 43
796 ⫾ 40
718 ⫾ 51
820 ⫾ 46
922 ⫾ 42
927 ⫾ 55
1028 ⴞ 50
1079 ⴞ 47
835 ⫾ 60
724 ⴞ 34
750 ⴞ 31
565 ⫾ 41
767 ⴞ 37
799 ⴞ 35
591 ⫾ 45
888
940
809
680
726
628
705
722
612
932
981
765
844
887
716
573
587
536
563
589
518
644
650
582
667
690
631
616
675
506
1388
1397
916
797
834
631
622
600
427
622
680
568
686
798
768
881
943
660
625
659
447
658
697
461
1214
1243
1199
905
935
897
902
905
848
1236
1263
1127
1111
1135
1035
748
749
745
728
742
715
849
840
827
889
895
896
843
886
777
1821
1799
1434
1035
1054
915
908
866
769
872
913
867
953
1046
1087
1175
1216
1011
823
842
683
876
900
722
FDG units are in nCi/g tissue wet weight. Regional mean ⫾ SE and 99% confidence intervals are shown for each group. Values are expressed as estimated
marginal means following ANCOVA analysis using white-matter readings as a covariate. Superscript number and boldface indicate group mean greater than
pseudorandom (P ⬍ 0.01). There were no auditory structures with significant differences between the blocked and excitor groups.
FDG uptake in structures with tone-blocked versus
-excitor differences
The comparison of mean FDG uptake in extra-auditory areas
showed significant differences at P ⬍ 0.01 between toneblocked and -excitor groups in three regions (Table 4 and Table
5, I). FDG uptake in the medial prefrontal cortex [MFC, also
known as prelimbic (Zilles and Wree 1985)] was largest in the
tone-excitor group and smallest in the tone-blocked group.
J Neurophysiol • VOL
This region showed the largest mean percent difference (25%)
between tone-blocked and -excitor groups. However, the toneexcitor mean was not significantly greater than the pseudorandom mean, suggesting that the MFC effect involved not just
FDG increase in the tone-excitor group but also some decrease
in the tone-blocked group (Fig. 5).
The spinal trigeminal nucleus (Sp5I) and the external cuneate nucleus (ECu) were the only regions found with greater
means in the tone-blocked group than the tone-excitor group.
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
Lateral lemniscus nucleus-dorsal (DLL)
Group
816
TABLE
D. JONES AND F. GONZALEZ-LIMA
4. FDG uptake in the extra-auditory structures
Structure
Medial frontal cortex (MFC)
Lateral frontal cortex (LFC)
Sulcal frontal cortex (SFC)
Cingulate cortex, anterior (CgA)
Insular cortex (Ins)
Medial septal nucleus (MS)
Accumbens (ACB)
Vertical diagonal band nucleus (VDB)
Parietal cortex, anterior (Par1a)
Caudate putamen, rostral (rCPU)
Medial preoptic area (MPO)
Horizontal diagonal band (HDB)
Anterior hypothalamic nucleus (AHy)
Paraventricular hypothalamic nucleus (PVH)
Cingulate cortex, posterior (CgP)
Mediodorsal thalamic nucleus (MD)
Ventroposterior thalamic nucleus (VPL)
Central amygdaloid nucleus (CeA)
Basolateral amygdaloid nucleus (BLA)
Medial amygdaloid nucleus (MeA)
Ventromedial hypothalamic nucleus (VMH)
Posterior parietal cortex (Par1p)
Mean
Lower 99%
Upper 99%
Blocked
Excitor1
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor1,3
Pseudorandom
Blocked
Excitor1,3
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor1
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked3
Excitor3
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor1,3
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor3
Pseudorandom
870 ⫾ 57
1086 ⴞ 53
943 ⫾ 68
899 ⫾ 40
990 ⫾ 37
878 ⫾ 48
1090 ⫾ 54
1219 ⫾ 51
1102 ⫾ 65
904 ⫾ 42
1073 ⴞ 39
892 ⫾ 51
678 ⫾ 32
771 ⴞ 29
630 ⫾ 38
692 ⫾ 34
765 ⫾ 31
695 ⫾ 40
524 ⫾ 28
603 ⫾ 26
508 ⫾ 33
586 ⫾ 29
668 ⴞ 27
520 ⫾ 35
682 ⫾ 21
756 ⴞ 20
683 ⫾ 25
921 ⫾ 47
979 ⫾ 44
938 ⫾ 56
854 ⴞ 22
871 ⴞ 21
762 ⫾ 27
529 ⫾ 41
666 ⴞ 38
456 ⫾ 49
817 ⫾ 30
859 ⫾ 28
780 ⫾ 36
470 ⫾ 42
603 ⴞ 39
437 ⫾ 50
514 ⫾ 35
616 ⴞ 32
455 ⫾ 41
1128 ⫾ 69
1204 ⴞ 64
932 ⫾ 82
912 ⫾ 44
1015 ⴞ 41
792 ⫾ 53
907 ⫾ 49
938 ⫾ 46
835 ⫾ 59
432 ⫾ 19
470 ⫾ 18
409 ⫾ 23
762 ⫾ 58
777 ⫾ 54
742 ⫾ 69
452 ⫾ 25
516 ⴞ 24
428 ⫾ 30
452 ⫾ 27
503 ⴞ 25
402 ⫾ 32
940 ⫾ 39
1010 ⴞ 36
870 ⫾ 47
703
931
744
782
881
738
930
1072
912
781
958
743
585
685
520
593
674
579
443
528
411
502
590
419
620
699
609
783
851
773
789
810
683
409
555
314
730
778
676
349
490
292
413
522
335
926
1017
691
783
897
639
763
805
663
377
419
343
593
620
540
378
448
339
373
430
308
825
904
732
1036
1241
1142
1016
1098
1018
1249
1367
1292
1029
1188
1040
770
857
741
790
856
813
605
678
605
671
747
621
744
813
756
1060
1107
1103
920
932
840
648
777
599
904
939
883
591
715
582
616
710
576
1329
1391
1172
1040
1134
946
1051
1072
1006
488
522
475
930
933
943
526
585
516
531
576
496
1055
1117
1007
J Neurophysiol • VOL
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
Lateral septal nucleus (LS)
Group
NEUROIMAGING OF CONDITIONING AND BLOCKING EFFECTS
TABLE
817
4. (continued)
Structure
Lateral habenula (Hb)
Hippocampus CA1, rostral (rCA1)
Hippocampus CA3, rostral (rCA3)
Dentate gyrus (DG)
Perirhinal cortex (Per)
Caudate putamen, caudal (cCPU)
Retrosplenial cortex (RSpl)
Presubiculum (Psub)
Subiculum (Sub)
Hippocampus CA1, caudal (cCA1)
Hippocampus CA2, caudal (cCA2)
Hippocampus CA3, caudal (cCA3)
Cerebellar hemisphere, lateral (CHL)
Cerebellar hemisphere, medial (CHM)
Deep cerebellar nucleus (DpCb)
Medial vestibular nucleus (MVe)
Solitary tract nucleus (Sol)
External cuneate nucleus (ECu)
Spinal trigeminal nucleus (Sp5I)
Lateral reticular nucleus (LRt)
Magnocellular reticular nucleus (MCRt)
Mean
Lower 99%
Upper 99%
Blocked
Excitor
Pseudorandom
Blocked3
Excitor3
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked3
Excitor3
Pseudorandom
Blocked
Excitor1,3
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked3
Excitor3
Pseudorandom
Blocked
Excitor3
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked3
Excitor3
Pseudorandom
Blocked2
Excitor
Pseudorandom
Blocked2
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
Blocked
Excitor
Pseudorandom
976 ⫾ 55
1023 ⫾ 51
997 ⫾ 66
710 ⴞ 33
707 ⴞ 30
546 ⫾ 39
693 ⫾ 30
693 ⫾ 28
606 ⫾ 36
620 ⫾ 25
662 ⫾ 24
587 ⫾ 30
792 ⴞ 41
830 ⴞ 38
631 ⫾ 49
708 ⫾ 41
840 ⴞ 38
569 ⫾ 49
718 ⫾ 32
689 ⫾ 29
653 ⫾ 38
1105 ⫾ 51
1072 ⫾ 47
1005 ⫾ 61
892 ⫾ 36
888 ⫾ 34
884 ⫾ 44
847 ⫾ 40
788 ⫾ 37
820 ⫾ 48
611 ⫾ 25
603 ⫾ 23
571 ⫾ 30
639 ⫾ 26
625 ⫾ 24
560 ⫾ 31
703 ⫾ 32
693 ⫾ 30
605 ⫾ 38
805 ⴞ 18
788 ⴞ 17
646 ⫾ 22
622 ⫾ 13
639 ⴞ 12
581 ⫾ 16
900 ⫾ 24
916 ⫾ 22
915 ⫾ 29
1031 ⫾ 31
1047 ⫾ 29
989 ⫾ 37
627 ⴞ 12
627 ⴞ 12
577 ⫾ 15
738 ⴞ 23
673 ⫾ 21
672 ⫾ 27
758 ⴞ 22
692 ⫾ 21
723 ⫾ 27
631 ⫾ 27
618 ⫾ 25
585 ⫾ 31
672 ⫾ 16
669 ⫾ 15
650 ⫾ 20
815
874
804
614
619
433
605
611
501
546
593
498
673
720
489
590
729
427
626
604
542
957
934
828
786
789
756
731
680
681
537
535
483
563
555
469
609
606
493
752
738
582
583
602
534
830
849
831
940
962
880
591
594
533
671
612
593
693
631
645
554
546
492
624
625
593
1138
1173
1190
805
795
660
781
775
712
694
730
675
911
940
773
827
950
711
810
775
763
1254
1210
1183
999
986
1011
963
896
959
685
671
659
715
696
651
797
780
717
859
838
711
661
675
627
971
980
1000
1122
1131
1098
663
661
620
804
735
751
823
752
801
709
689
677
720
714
707
FDG units are in nCi/g tissue wet weight. Regional mean ⫾ SE and 99% confidence intervals are shown for each group. Values are expressed as estimated
marginal means following ANCOVA analysis using white matter readings as a covariate. Superscript numbers and boldface indicate: 1, excitor greater than
blocked; 2, blocked greater than excitor; and 3, indicated group greater than pseudorandom (P ⬍ 0.01).
J Neurophysiol • VOL
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
Entorhinal cortex (Ent)
Group
818
D. JONES AND F. GONZALEZ-LIMA
5. Summary of extra-auditory regions with effects of
conditioning
appeared similar in the three groups with one exception. The
correlation between the anterior parietal cortex (US somatosensory cortex) and the lateral septum was ⫺0.98 in the toneblocked group. This was significantly different from the corresponding values in the tone-excitor and pseudorandom
groups (0.60 and ⫺0.24, respectively). Other nonsignificant
results are summarized here by presenting the mean and range
of correlation coefficients (r) in the three groups: tone-blocked
mean ⫽ 0.38 (range 0.95– 0.01), tone-excitor mean ⫽ 0.46
(range 0.91– 0.04), and pseudorandom mean ⫽ 0.39 (range
0.91– 0.03).
TABLE
I. Blocking effects (unique to blocking group)
1. Blocked lower than excitor
Medial frontal cortex (MFC)
2. Blocked greater than excitor
Spinal trigeminal nucleus (Sp5I)
External cuneate nucleus (ECu)
II. Contiguity effects (common to excitor and blocked groups)
1. Excitor and blocked both greater than pseudorandom
Caudate putamen, rostral (rCPU)
Perirhinal cortex (Per)
Hippocampus CA1, rostral (rCA1)
Cerebellar hemisphere, lateral (CHL)
Solitary tract nucleus (Sol)
III. Excitatory effects (unique to excitor group)
1. Excitor greater than both blocked and pseudorandom
Insular cortex (Ins)
Cingulate cortex, anterior (CgA)
Vertical diagonal band nucleus (VDB)
Anterior hypothalamic nucleus (AHy)
Caudate putamen, caudal (cCPU)
2. Excitor greater than pseudorandom
Accumbens (ACB)
Medial preoptic area (MPO)
Paraventricular hypothalamic nucleus (PVH)
Mediodorsal thalamic nucleus (MD)
Medial amygdaloid nucleus (MeA)
Cingulate cortex, posterior (CgP)
Ventromedial hypothalamic nucleus (VMH)
Posterior parietal cortex (PPC)
Cerebellar hemisphere, medial (CHM)
DISCUSSION
However, the tone-excitor and pseudorandom groups showed
similar means in Sp5I and ECu. Therefore these two structures
are somatosensory medullary regions with FDG increases
unique to the tone-blocked group, indicating an active role of
US somatosensory pathways in the blocking effect.
FDG uptake in structures with group means greater
than pseudorandom
These structures can be subdivided into two categories:
regions with common effects in the tone-blocked and -excitor
groups (i.e., both groups greater than pseudorandom; Table 5,
II) and regions with effects unique to the tone-excitor group
(Table 5, III). The first category included five regions: rostral
caudate putamen (rCPU), perirhinal cortex (Per), CA1 of dorsal hippocampus (rCA1), lateral cerebellar hemisphere (CHL),
and solitary tract nucleus (Sol). These effects were similar to
those seen in auditory regions (Table 3) and suggest common
blocked-excitor conditioning effects based on their CS-US
contiguity not present in the pseudorandom group. Other regions showing greater means in the tone-excitor group as
compared with both the tone-blocked and pseudorandom
groups were the insular (Ins) and anterior cingulate cortex
(CgA), vertical diagonal band nucleus (VDB, with P ⫽ 0.01 in
tone-excitor vs. pseudorandom), anterior hypothalamic nucleus
(AHy), and caudoventral caudate putamen (cCPU). Other regions listed in Table 5 showed FDG uptake increases unique to
the tone-excitor group, which may be based on the CERs
evoked by the tone in this group only.
Covariance analysis
Although there were many group differences in mean activational effects, the inter-regional correlations of FDG uptake
J Neurophysiol • VOL
FIG. 5. Examples of differences in fluorodeoxyglucose (FDG) incorporation in the medial (MFC) area of the prefrontal cortex (A–C, left) and corresponding profiles through a representative level of the autoradiograph (A–C,
right). D: a schematic of the area (box) and level (dotted line) at which
readings for the profile were taken. Although darker (more FDG) labeling may
be seen in this region by the unaided eye, pictures of autoradiographs are
examples only and as such some individual variation may be seen. The results
are not merely based on visual inspection of autoradiographs but also on
quantitative group data collected from the actual films with a calibrated
densitometer as reported in Tables 3 and 4.
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
This study served in localizing brain regions displaying
Pavlovian conditioning effects on FDG uptake in response to
the same physical tone. Brain activity differences are proposed
to result from group differences in previous training because all
subjects received the same tone stimulation after FDG admin-
NEUROIMAGING OF CONDITIONING AND BLOCKING EFFECTS
Blocking effects
There were two types of blocking effects on the brain,
involving higher (frontal) and lower (medullary) regions that
may be engaged in complementary behavioral mechanisms
related to the blocking phenomenon. One type of brain effect
was localized to the medial prefrontal cortex. This region
showed a 25% mean difference between the tone-blocked and
-excitor groups. While the tone-excitor mean was greater than
the tone-blocked mean, the tone-excitor was not significantly
greater than the pseudorandom. Therefore the data suggest that
the blocked-excitor difference in the medial prefrontal cortex
was not only due to increased FDG uptake in the tone-excitor
group, but also it involved reduced activity in the tone-blocked
group. Anatomically, the identified medial prefrontal cortex
region sends a projection to the anterior cingulate (Beckstead
1979) and thereby can influence Pavlovian excitatory effects
found in the anterior cingulate. This effect in the medial
J Neurophysiol • VOL
prefrontal cortex is consistent with previous studies of this
region, which has well-documented roles in behavioral inhibition and selective attention that may be relevant for blocking.
This effect may be also of clinical significance because schizophrenics characterized with abnormal FDG uptake in the prefrontal cortex are unable to develop the blocking effect on
Pavlovian conditioning (Jones et al. 1997).
The second type of brain effect of blocking was the activation of two subcortical somatosensory nuclei, the spinal trigeminal and the external cuneate nuclei in the medulla. These
are major relay nuclei for the two main somatosensory pathways that transmit the ascending effects of the aversive US
(electric shock) from the periphery to the brain: the trigeminal
pain pathway (spinal trigeminal nucleus) and the dorsal column-medial lemniscus tactile pathway (external cuneate nucleus). Interestingly, our analysis of inter-regional correlations
of FDG uptake revealed only the correlation between the
primary somatosensory cortex (anterior parietal cortex) and
lateral septal nucleus to be different in the tone-blocked group
as compared with the other two groups. Therefore it may be
concluded that a blocked tone activates somatosensory US
pathways in a way different from the same tone as an excitor.
How these blocking effects on the brain relate to theoretical
mechanisms derived from behavioral studies is discussed in
relation to three classes of theories that emphasize either CS,
US, or CR mechanisms.
Mackintosh (1983) proposed a CS-based theoretical mechanism that explains blocking through decreased or selective
attention shifted away from a stimulus that has no added
predictive value during phase II of compound training (the tone
in this experiment). An attractive contribution for the frontal
and anterior cingulate areas in blocking is one of selective
attention (Mesulam 1990; Posner et al. 1987). The medial
prefrontal and CgA cortices have been implicated in attentional
processing by a wide array of imaging and neurophysiological
studies. We argue that the absence of blocking effects in the
auditory structures rules out a CS processing failure account,
but this is true only if it is believed that associate effects in
primary sensory structures must be attentional in character.
The data suggest the intriguing alternative that the acoustic
associative changes are preattentive and that attentional modulation is only exhibited in association cortical areas such as
the prefrontal cortex and anterior cingulate. In other words, if
the association cortical component is compromised while the
acoustic processing component is intact, CS processing may
still fail.
However, the similar excitatory effects of the tone CS on the
auditory system in both tone-blocked and -excitor groups argue
against an attention shift away from the blocked tone as suggested by popular views of blocking. The enhanced effect of
the blocked tone in the auditory system provides neural evidence against CS/attentional explanations of blocking. Some of
the inattention accounts of blocking, including that of Kamin
(1969), imply that the CS is ignored or somehow rejected so
that no proper CS-US association may be formed during blocking. These accounts have been referred to as CS processingfailure explanations of blocking (Mackintosh 1983). However,
our findings of similar FDG increases to the tone in the
auditory system of tone-blocked and -excitor groups do not
support this explanation of blocking. It does appear that the
blocked tone is as effective at producing conditioned activation
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
istration. The main focus of this study is to describe FDG
differences between the two conditioning groups (tone-blocked
vs. tone-excitor), while the pseudorandom group provides a
control for unconditioned tone effects. In both the tone-blocked
and -excitor groups, the subjects’ experience with the tone was
identical. At the time of FDG testing, the resulting differences
in behavior and concomitant changes in FDG incorporation are
most likely due to each groups history with the light presentations. While the tone was contiguous with the US during
phase II training in both conditioning groups, the tone did not
elicit a conditioned response in the tone-blocked group as the
light was predictive of the US following phase I training. In the
tone-excitor group, this was not the case; during phase I, the
light did not predict the occurrence of the US, and both the
light and tone came to elicit a conditioned response following
phase II training. The pseudorandom control group received
unconditioned stimuli identical to both experimental groups
but lacked contiguity and CER effects. Subjects in this control
group were not exposed to paired light-US presentations (similar to the tone-excitor group) and pseudorandom subjects did
not display a CER to the tone. This absence of a CER was also
seen in the tone-blocked group.
The brains of tone-blocked and -excitor rats differed in their
average metabolic responses in certain regions that may be
implicated in the Kamin blocking effect on Pavlovian conditioning. Hence this study provides a first large-scale map of
brain regions underlying the blocking effect. However, all
brain regions identified as having group differences are not
necessarily components of the blocking effect per se. Conditioning effects that result from a blocking training protocol
may include various associative and nonassociative mechanisms. Inferring other functional aspects of FDG uptake differences remains an interpretive challenge (Gonzalez-Lima
1992), but one that offers an opportunity to integrate relevant
research into a systems-level model. In light of previous studies, we attempt to do just this, proposing neural pathways that
may underlie three major conditioning effects within the constraints of our experimental design: blocking effects unique to
the tone-blocked group; contiguity effects common to the
tone-blocked and -excitor groups; and excitatory effects unique
to the tone-excitor group. Table 5 provides a summary of these
effects.
819
820
D. JONES AND F. GONZALEZ-LIMA
J Neurophysiol • VOL
from the interplay of these mechanisms in the brain as they
both contribute to the same behavioral outcome. This general
conclusion is consistent with the contiguity and excitatory
effects discussed in the next sections as well as our previous
FDG mapping studies of learning that indicate that the rat brain
uses a combined strategy to modify behavior as a result of
experience (Gonzalez-Lima 1992).
Contiguity effects
The tone-blocked and -excitor groups showed a number of
similar FDG effects relative to the pseudorandom group. These
similar effects were interpreted as the result of tone-shock
(CS-US) contiguity common to the conditioned groups but
absent in the pseudorandom group. These common effects are
unlikely due to the elicitation of a CER because, while the
tone-excitor group showed a CER, the tone-blocked and pseudorandom groups did not. Similarly the common effects cannot
be accounted for by the number of stimuli because they were
the same in all groups. Thus the most likely explanation is that
tone-shock pairings changed the way certain brain regions
responded to the tone as compared with pseudorandom tone
presentations.
This interpretation is reinforced by the finding of strong
FDG uptake increases in auditory regions such as the inferior
colliculus and auditory cortex in response to the tone in the
tone-blocked and -excitor groups. It has been a consistent
finding in our FDG studies since 1984 (reviewed in GonzalezLima and McIntosh 1996) that pairing of a tone with shock or
another aversive US leads to a subsequent greater metabolic
response to the tone in auditory structures. This is the case in
the rat with tone-shock pairings as well as in humans during
eyeblink conditioning to a tone (Molchan et al. 1994).
Similar CS-US contiguity effects are also present in the
current FDG study between experimental groups in certain
extra-auditory regions (Table 5): rCPU, Per, rCA1, CHL, and
Sol. These regions showed FDG increases similar but lower
than those seen in auditory regions. These effects may indicate
conditioning effects based on the CS-US contiguity not present
in the pseudorandom group. We have previously reported that
pairing of a tone with US reticular stimulation leads to FDG
uptake increases in caudate putamen and CA1, but the other
regions were not examined in that study (Gonzalez-Lima and
Scheich 1986). However, in a subsequent FDG study using
tone-shock pairings, a tone conditioned as an excitor compared
with a tone conditioned as a Pavlovian inhibitor revealed
differences in Per and cerebellum (McIntosh and GonzalezLima 1994). In that study, the excitor effects on the brain could
be accounted for by either CS-US pairing or CER expression.
The present results aid in the interpretation of our previous
studies because CS-US pairing and CER expression is separated across groups. The conditioning effects in these regions
are likely due to the common factor of CS-US contiguity in the
blocked and excitor groups rather than CER expression, which
is absent in blocked and inhibitor conditions.
Other studies involving learned associations have consistently implicated these regions we attributed to contiguity
effects, like the rabbit CA1 and cerebellum in eyeblink conditioning (Lavond et al. 1993) or the monkey Per in delayed
nonmatching to sample and object retention memory (ZolaMorgan et al. 1989). In the rat Per, lesion studies have also
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
of auditory regions as the tone conditioned as an excitor.
Therefore blocking is not likely the result of simple reductions
in CS processing pathways, and it would be hard to reconcile
a role of decreased CS processing as part of the blocking effect.
Our previous study of blocking effects on the auditory
system using cytochrome oxidase histochemistry (Poremba et
al. 1997) may shed some light on how blocking develops
during training. But there has to be a clear understanding of the
different kind of information provided by our cytochrome
oxidase study. The main difference is that the cytochrome
oxidase study reflects effects of all stimuli through out training
(i.e., tone, shock, and light). In contrast, the present FDG study
reflects responses to the tone after training (i.e., during a
posttraining FDG uptake period with tone CS alone). There is
no analogous posttraining uptake period in a cytochrome oxidase study, so any differences in cytochrome oxidase at the
time of brain histochemistry reflect cumulative CS and US
effects building-up during the entire training period. Interestingly, cytochrome oxidase suggests the auditory system is
activated during blocking only in regions receiving anatomical
connections from US somatosensory pathways: dorsal cochlear
nucleus, inferior colliculus, and secondary auditory cortex
(Poremba et al. 1997). So it is unlikely that blocking develops
due to a lack of tone-shock CS-US association in the brain as
implied by some CS inattention accounts of this phenomenon
(Mackintosh 1983). Together these studies point to the US
influence in the brain as more relevant than the CS influence
during blocking.
The neural activational effect on US pathways can be interpreted as support of the most well-known behavioral theory
that deals with the blocking effect (Rescorla and Wagner
1972). The Rescorla-Wagner model of learning suggests that
blocking involves the modification of US pathways. That is,
there may be a limit in the amount of learning that a US can
support, and the blocking effect may result from saturating
activity in US relay pathways. The specific neuroimaging
finding of differences in subcortical somatosensory nuclei
seems to support this US-based theoretical mechanism, at least
in part. Thus we derive support for the Rescorla-Wagner model
from the enhanced activation of the somatosensory areas, suggesting that this model predicts that blocking is due to modification of the US pathway. Yet the Rescorla-Wagner model
suggests that a given US supports a fixed amount of associative
strength and blocking occurs because some or all of the available strength was preinvested in the light-shock association.
These dynamics seem to be a matter of interference or associative competition rather than US efficacy per se.
The suppressive effect found in the medial frontal cortex
suggests that blocking may also involve CR mechanisms. A
third theoretical mechanism based on behavioral studies suggests that blocking involves a conditioned response inhibition
(Miller et al. 1993). That is, tone-shock contiguity may support
a CS-US association, but the CER is inhibited after phase II
training in blocking. To the extent that the medial prefrontal
cortex is involved in CR inhibition, our neural findings support
CR mechanisms that emphasize response failure in blocking
(Balaz et al. 1982; Miller et al. 1993; Schachtman et al. 1983).
Therefore the brain regions modified in this study suggest a
combined strategy with higher (frontal) and lower (medullary)
regions engaged in two complementary mechanisms of CR
inhibition and US saturation. Blocking effects may then result
NEUROIMAGING OF CONDITIONING AND BLOCKING EFFECTS
shown that it is a key structure in the circuit for conditioned
fear-potentiated acoustic startle (Rosen et al. 1992). Finally,
the solitary nucleus has been implicated in conditioning effects
by both neuronal recording (Giza et al. 1997) and lesion studies
(Grigson et al. 1997). In particular, aversive conditioning effects are greater than preference conditioning effects in the rat
solitary nucleus (Giza et al. 1997), and these activation effects
do not disappear after conditioned response extinction (McCaughey et al. 1997). Together with our data, previous studies
support the conclusion that these regions likely mediate associative learning effects rather than effects due to conditioned
response expression. Results from the current experiment are
uniquely suited to support such a conclusion as CS-US pairings
are similar across experimental groups while the presence of a
CER is separated by groups.
Excitatory effects
J Neurophysiol • VOL
A region showing consistent tone-evoked excitatory effects
in our FDG conditioning studies is the cCPU. These effects
occur whether an aversive (Gonzalez-Lima and Scheich 1986)
or an appetitive US (Gonzalez-Lima 1992) is used, as long as
a tone is the CS. We have proposed that this region links the
auditory system with a striatal motor response system (Gonzalez-Lima 1989). The rat cCPU receives direct projections from
both auditory cortex (Roger and Arnault 1989) and medial
geniculate (LeDoux et al. 1990). Thus the cCPU contains a
secondary auditory projection field that is an anatomical continuation of the primary auditory pathways into the neostriatum. The contribution of this auditory striatal field to auditory
Pavlovian conditioning has been examined with lesion studies
that showed that its damage eliminates tone-evoked conditioned responses in rats (Iwata et al. 1986). Thus the cCPU may
contribute a conditioned striatal motor component to the CER
elicited by the tone in the tone-excitor group.
In addition, there were multiple distributed effects on regional metabolism in other regions that were similar in toneexcitor and -blocked groups but that showed FDG uptake
increases in the tone-excitor group as compared with the pseudorandom group (listed in Table 5). These other regions showing excitatory effects limited to the tone-excitor group have
been observed in our previous FDG studies. For example, we
identified these regions in a study of CER elicited by a tone
conditioned to reticular stimulation (Gonzalez-Lima and
Scheich 1986) as well as in a study using tone-shock stimuli
(McIntosh and Gonzalez-Lima 1994) comparing a tone as
excitor (elicits CER) versus the same tone as inhibitor (inhibits
CER). Pavlovian conditioned excitation and inhibition affected
the same circuits, but this effect was through a shift in sign in
the correlative activity between the regions from positive in the
tone-excitor condition to negative in the tone-inhibitor condition (McIntosh and Gonzalez-Lima 1994). Thus it appears that
Pavlovian conditioned inhibition and the Kamin blocking effect are mediated by different brain mechanisms.
It also appears that lesion studies aimed at impairing a
conditioned response provide an incomplete picture of the
effects of Pavlovian conditioning in the intact brain. For example, damaging a relatively small number of brain stem
regions may impair a simple conditioned response involving
eyeblink or nictitating membrane closure (Lavond et al. 1986).
While lesions of the amygdala and related regions may impair
the expression of a CER (Iwata et al. 1986), it appears that
large-scale networks are affected by tone-shock conditioning in
the intact brain.
For example, the medial frontal cortex may have important
functional contributions to tone blocking effects on the identified excitatory brain circuits. The cortical limbic circuit receives auditory CS information, and can influence the basal
forebrain through medial frontal cortex projections to the accumbens (Beckstead 1979; Phillipson and Griffths 1985) and
through its reciprocal projection with the VDB (Saper 1984).
The efferent projection from the medial frontal cortex to the
cCPU (Beckstead 1979) may serve to compare different levels
of auditory CS processing. Auditory CS information can enter
this system at various levels. The strongest of these are the
projections from the medial geniculate and auditory cortex to
the caudal caudate putamen (LeDoux et al. 1985), but the
medial geniculate also has efferents to the medial amygdaloid
nucleus (Ottersen and Ben-Ari 1979). Auditory CS information
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
Effects unique to the tone-excitor group (tone elicits CER)
are referred to as excitatory effects in Pavlovian terminology.
That is, they refer to conditioned response excitation in a
behavioral sense rather than excitation in a neuronal electrophysiological sense. The tone is a CS excitor when it elicits a
conditioned response as in the tone-excitor group but not in the
tone-blocked or pseudorandom groups. There were five regions
with FDG uptake in the tone-excitor group greater than the
other two groups (i.e., tone-excitor ⬎ tone-blocked and pseudorandom in Table 5): the Ins, CgA, VDB, AHy, and cCPU.
The simplest explanation for these excitatory effects is that
they are due to CER components rather than any CS-US
contiguity effects common to tone-excitor and -blocked
groups. However, these excitatory effects are not simply the
result of nonspecific motor or arousal effects because relevant
regions of the primary motor system (lateral frontal cortex) and
reticular system (lateral and magnocellular reticular nuclei)
showed no such effects. Each of the affected regions appeared
more specifically related to the CER. First, there is mounting
evidence that the Ins mediates visceral reactions and avoidance
responses to aversive stimuli. For example, lesions or reversible inactivation of the rat Ins produce amnesia for inhibitory
avoidance (Bermudez-Rattoni et al. 1991). The Ins receives
taste and visceral information from the thalamus and sends
direct projections to the Sol along with the anterior hypothalamus (van der Kooy et al. 1984). Thus the insular and anterior
hypothalamic projections to the Sol may mediate visceral components of the CER in the tone-excitor group.
Similarly, the role of the CgA in avoidance responses to
learned aversive stimuli has been well documented in a series
of elegant studies by Gabriel and collaborators. For example,
Gabriel (1990) has shown that a locomotive avoidance response to a tone that signals a shock is retarded after lesions of
the cingulate cortex in rabbits. Interestingly, early-developing
training-induced unit activity in the posterior cingulate (area
29) was absent in rabbits with anterior cingulate (area 24)
lesions, indicating that the anterior cingulate is a source of
early-developing plasticity in avoidance conditioning (Gabriel
et al. 1991). Presumably Pavlovian CERs develop early in
avoidance conditioning before an instrumental response is observed. Therefore the anterior cingulate change in the toneexcitor group may reflect Pavlovian excitatory effects sustaining CERs to tones paired with shock.
821
822
D. JONES AND F. GONZALEZ-LIMA
can also reach the medial preoptic area indirectly through the
medial amygdaloid nucleus (Conrad and Pfaff 1976; McDonald 1987). The medial amygdaloid nucleus may play a role
in stimulus aversive emotional value. C-fos immunoreactivity
is reported to increase in the medial amygdala of rats subjected
to tail pinch (Smith et al. 1996), and this nucleus has been
shown to be important in regulation of antinociceptive processes (Werka 1997). This would suggest that the role of the
medial amygdaloid nucleus be related to US emotional properties evoked by the CS. It is clear from this and previous
neuroimaging studies that large-scale networks are involved in
the behavioral effects of tone-shock Pavlovian conditioning in
intact brains (McIntosh and Gonzalez-Lima 1998). Therefore
Pavlovian conditioning effects are of various kinds and are
pervasive throughout the brain of intact organisms, including
humans (Molchan et al. 1994).
REFERENCES
ASHER HB. Causal Modeling. London: Sage Publications, 1983.
BALAZ MA, GUTSIN P, CACHEIRO H, AND MILLER RR. Blocking as a retrieval
failure: reactivation of associations to a blocked stimulus. Q J Exp Psychol
A 34b: 99 –113, 1982.
BECKSTEAD RM. An autoradiographic examination of corticocortical and subcortical projections of the mediodorsal-projection (prefrontal) cortex of the
rat. J Comp Neurol 184: 43– 46, 1979.
BERMUDEZ-RATTONI F AND MCGAUGH JL. Insular cortex and amygdala lesions
differentially affect acquisition on inhibitory avoidance and conditioned
taste aversion. Brain Res 549: 165–170, 1991.
CONRAD LC AND PFAFF DW. Efferents from medial basal forebrain and
hypothalamus in the rat. I. An autoradiographic study of the medial preoptica area. J Comp Neurol 169: 185–219, 1976.
DOMJAN M, CUSATO B, AND VILLARREAL R. Pavlovian feed-forward mechanisms in the control of social behavior. Behav Brain Sci 23: 235–249, 2000.
FANSELOW MS. Pavlovian conditioning, negative feedback and blocking;
mechanisms that regulate association formation. Neuron 20: 625– 627, 1998.
GABRIEL M. Functions of anterior and posterior cingulate cortex during avoidance learning in rabbits. Prog Brain Res 85: 467– 482, 1990.
GABRIEL M, KUBOTA Y, SPARENBORG S, STRAUBE K, AND VOGT BA. Effects of
cingulate cortical lesions on avoidance learning and training-induced unit
activity in rabbits. Exp Brain Res 86: 585– 600, 1991.
GALLO M AND CANDIDO A. Dorsal hippocampal lesions impair blocking but
not latent inhibition of taste aversion learning in rats. Behav Neurosci 109:
413– 425, 1995.
GARRUD P, MACKINTOSH NJ, COTTON MM, AND GELDON J. Successful overshadowing and blocking in hippocampectomized rats. Behav Brain Res 12:
39 –53, 1984.
GIFTAKIS JE AND TAIT RW. Blocking of the rabbit’s classically conditioned
nictitating membrane response: effects of modifications of contextual associative strength. Learn Motiv 29: 23– 48, 1998.
GIZA BK, ACKROFF K, MCCAUGHEY SA, SCLAFANI A, AND SCOTT TR. Preference conditioning alters taste responses in the nucleus of the solitary tract of
the rat. Am J Physiol Regulatory Integrative Comp Physiol 273: R1230 –
R1240, 1997.
GONZALEZ-LIMA F. Functional brain circuitry related to arousal and learning in
rats. In: Visuomotor Coordination: Amphibians, Comparisons, Models and
Robots, edited by Ewert J-P and Arbib MA. New York: Plenum, 1989, p.
729 –765.
GONZALEZ-LIMA F. Brain imaging of auditory learning functions in rats:
studies with fluorodeoxyglucose autoradiography and cytochrome oxidase
histochemistry. In: Advances in Metabolic Mapping Techniques for Brain
Imaging of Behavioral and Learning Functions, edited by Gonzalez-Lima F,
Finkenstadt Th, and Scheich H. Boston, MA: Kluwer, 1992, p. 39 –110.
GONZALEZ-LIMA F AND AGUDO J. Functional reorganization of neural auditory
maps by differential learning. Neuroreport 1: 161–164, 1990.
J Neurophysiol • VOL
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
We thank M. Domjan for advice concerning the behavioral studies and A.
Crane for advice on the manuscript.
This work was supported by National Institute of Neurological Disorders
and Stroke Grant RO1 NS-37755 to F. Gonzalez-Lima.
GONZALEZ-LIMA F AND CADA A. Cytochrome oxidase atlas of rat brain. In:
Cytochrome Oxidase in Neuronal Metabolism and Alzheimer’s Disease,
edited by Gonzalez-Lima F. New York: Plenum, 1998, p. 263–280.
GONZALEZ-LIMA F, FINKENSTADT T, AND EWERT J-P. Neural substrates for
long-term habituation of the acoustic startle reflex in rats: a 2-deoxyglucose
study. Neurosci Lett 96: 151–156, 1989.
GONZALEZ-LIMA F AND MCINTOSH AR. Conceptual and methodological issues
in the interpretation of brain-behavior relationships. In: Developmental
Neuroimaging: Mapping the Development of Brain and Behavior, edited by
Thatcher RW, Reid Lyon G, Ramsey J, and Krasnegor N. Orlando, FL:
Academic, 1996, p. 235–253.
GONZALEZ-LIMA F AND SCHEICH H. Neural substrates for tone-conditioned
bradycardia demonstrated with 2-deoxyglucose. I. Activation of auditory
nuclei. Behav Brain 14: 213–233, 1984.
GONZALEZ-LIMA F AND SCHEICH H. Ascending reticular activating system in
the rat: a 2-deoxyglucose study. Brain Res 344: 70 – 88, 1985.
GONZALEZ-LIMA F AND SCHEICH H. Classical conditioning of tone-signaled
bradycardia modifies 2-deoxyglucose uptake patterns in cortex, thalamus,
habenula, caudate-putamen and hippocampal formation. Brain Res 363:
239 –256, 1986.
GRIGSON PS, SHIMURA T, AND NORGREN R. Brain stem lesions and gustatory
function. II. The role of the nucleus of the solitary tract in Na⫹ appetite,
conditioned taste aversion, and conditioned odor aversion in rats. Behav
Neurosci 111: 169 –179, 1997.
HOLLAND PC AND GALLAGHER M. Effects of amygdala central nucleus lesions
on blocking and unblocking. Behav Neurosci 107: 235–245, 1993.
IWATA J, LEDOUX JE, MEELEY MP, ARNERIC S, AND REIS DJ. Intrinsic neurons
in the amygdaloid field projected to by the medial geniculate body mediate
emotional responses conditioned to acoustic stimuli. Brain Res 383: 195–
214, 1986.
JONES SH, HEMSLEY D, BALL S, AND SERRA A. Disruption of the Kamin
blocking effect in schizophrenia and in normal subjects following amphetamine. Behav Brain Res 88: 103–114, 1997.
JÖRESKOG KG AND SÖRBOM D. LISREL 7 User’s Reference Guide. Mooresville: Scientific Software International, 1989.
KAMIN LJ. Predictability, surprise, attention, and conditioning. In: Punishment
and Aversive Behavior, edited by Campbell BA and Church RM. New York:
Appleton-Century-Crofts, 1969, p. 242–259.
KIM JJ, KRUPA DJ, AND THOMPSON RF. Inhibitory cerebello-olivary projections
and blocking effect in classical conditioning. Science 279: 570 –573, 1998.
KINKAIDE PS AND WALLY RE. Visual evoked potentials during stimulus
selection in eyelid conditioning in the rabbit (Oryctolagus cuniculuc).
J Comp Physiol Psychol 87: 481– 494, 1974.
LAVOND DG, KIM JJ, AND THOMPSON RF. Mammalian brain substrates of
aversive classical conditioning. Annu Rev Psychol 44: 317–342, 1993.
LEDOUX JE AND RUGGIERO DA. Topographic organization of neurons in the
acoustic thalamus that project to the amygdala. J Neurosci 10: 1043–1054,
1990.
LEDOUX JE, RUGGIERO DA, AND REIS DJ. Projections to the subcortical
forebrain from anatomically defined regions of the medial geniculate body
in the rat. J Comp Neurol 242: 182–213, 1985.
MACKINTOSH NJ. Conditioning and Associative Learning. New York: Oxford
Univ. Press, 1983.
MCALONAN K, BROWN VJ, AND BOWMAN EM. Thalamic reticular nucleus
activation reflects attentional gating during classical conditioning. J Neurosci 20: 8897– 8901, 2000.
MCCAUGHEY SA, GIZA BK, SCOTT TR, AND SCOTT TR. Extinction of a
conditioned taste aversion in rats. II. Neural effects in the nucleus of the
solitary tract. Physiol Behav 61: 373–379, 1997.
MCDONALD AJ. Organization of amygdaloid projections to the mediodorsal
thalamus and prefrontal cortex: a fluorescence retrograde transport study in
the rat. J Comp Neurol 262: 46 –58, 1987.
MCINTOSH AR AND GONZALEZ-LIMA F. Network analysis of functional auditory pathways mapped with fluorodeoxyglucose: associative effects of a
tone conditioned as a Pavlovian excitor or inhibitor. Brain Res 627: 129 –
140, 1993.
MCINTOSH AR AND GONZALEZ-LIMA F. Network interactions among limbic
cortices, basal forebrain, and cerebellum differentiate a tone conditioned as
a Pavlovian excitor or inhibitor: fluorodeoxyglucose mapping and covariance structural modeling. J Neurophysiol 72: 1717–1733, 1994.
MCINTOSH AR AND GONZALEZ-LIMA F. Functional network interactions between parallel auditory pathways during Pavlovian conditioned inhibition.
Brain Res 683: 228 –241, 1995.
NEUROIMAGING OF CONDITIONING AND BLOCKING EFFECTS
J Neurophysiol • VOL
ROGER M AND ARNAULT P. Anatomical study of the connections of the primary
auditory area in the rat. J Comp Neurol 287: 339 –356, 1989.
ROSEN JB, HITCHCOCK JM, FALLS WA, CAMPEAU S, AND DAVIS M. Lesions of
the perirhinal cortex but not of the frontal, medial prefrontal, visual, or
insular cortex block fear-potentiated startle using a visual conditioned stimulus. J Neurosci 12: 4624 – 4633, 1992.
SAPER CB. Organization of cerebral cortical afferent systems in the rat. II.
Magnocellular basal nucleus. J Comp Neurol 222: 313–342, 1984.
SCHACHTMAN TR, GEE JL, KASPROW WJ, AND MILLER RR. Reminder-induced
recovery from blocking as a function of the number of compound trials.
Learn Motiv 14: 154 –164, 1983.
SCHEICH H, STARK H, ZUSCHRATTER W, OHL FW, AND SIMONIS CE. Some
functions of primary auditory cortex in learning and memory formation. Adv
Neurol 73: 179 –193, 1997.
SHAO J AND DONGSHENG T. The Jacknife and Bootstrap. New York: Springer
Verlag, 1995.
SMITH WJ, STEWART J, AND PFAUS JG. Tail pinch induces FOS immunoreactivity within several regions of the male rat brain: effects of age. Physiol
Behav 61: 717–723, 1996.
SOKOLOFF L. Imaging techniques in studies of neural functions. In: Advances
in Metabolic Mapping Techniques for Brain Imaging of Behavioral and
Learning Functions, edited by Gonzalez-Lima F, Finkenstadt T, and Scheich
H. Boston, MA: Kluwer Academic, 1992, p. 1–38.
SOLOMON PR. Role of the hippocampus in blocking and conditioned inhibition
of the rabbit’s nictitating membrane response. J Comp Physiol Psychol 91:
407– 417, 1977.
TOGA AW AND COLLINS RC. Glucose metabolism increases in visual pathways
following habituation. Physiol Behav 27: 825– 834, 1981.
VAN DER KOOY D, KODA LY, MCGINTY JF, GERFEN CR, AND BLOOM FE. The
organization of projections from the cortex, amygdala, and hypothalamus to
the nucleus of the solitary tract in rat. J Comp Neurol 224: 1–24, 1984.
WEINBERGER NW. Dynamic regulation of receptive fields and maps in the adult
sensory cortex. Annu Rev Neurosci 18: 129 –158, 1995.
WEINBERGER NW AND DIAMOND DM. Physiological plasticity in auditory
cortex: rapid induction by learning. Prog Neurobiol 29: 1–55, 1987.
WERKA T. The effects of the medial and cortical amygdala lesions on poststress analgesia in rats. Behav Brain Res 86: 59 – 65, 1997.
ZILLES K AND WREE A. Cortex: areal and laminar structure. In: The Rat
Nervous System, edited by Paxinos G. New York: Academic, 1985, p.
375– 416.
ZOLA-MORGAN S, SQUIRE LR, AMARAL DG, AND SUZUKI WA. Lesions of
perirhinal and parahippocampal cortex that spare the amygdala and hippocampal formation produce severe memory impairment. J Neurosci 9:
4355– 4370, 1989.
86 • AUGUST 2001 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.3 on May 15, 2017
MCINTOSH AR AND GONZALEZ-LIMA F. Large-scale functional connectivity in
associative learning: interrelations of the rat auditory, visual, and limbic
systems. J Neurophysiol 80: 3148 –3162, 1998.
MESULAM MM. Large-scale neurocognitive networks and distributed processing for attention, language, and memory. Ann Neurol 28: 597– 613, 1990.
MILLER JS, JAGIELO JA, AND SPEAR NE. The influence of retention interval on
the US preexposure effect: changes in contextual blocking over time. Learn
Motiv 376 –394: 1993.
MOLCHAN SE, SUNDERLAND T, MCINTOSH AR, HERSCOVITCH P, AND SCHREURS
BGA. Functional anatomical study of associative learning in humans. Proc
Natl Acad Sci USA 91: 8122– 8126, 1994.
NAIR HP AND GONZALEZ-LIMA F. Extinction of behavior in infant rats: development of functional coupling between septal, hippocampal, and ventral
tegmental regions. J Neurosci 19: 8646 – 8655, 1999.
NEUENSCHWANDER-EL MASSIOUI N, DUTRIEUX G, AND EDELINE JM. Conditioned hippocampal cellular response to a behaviorally silent conditioned
stimulus. Behav Neurosci 105: 313–325, 1991.
NOBREGA JN. Brain metabolic mapping and behavior: assessing the effects of
early developmental experiences in adult animals. In: Advances in Metabolic
Mapping Techniques for Brain Imaging of Behavioral and Learning Functions, edited by Gonzalez-Lima F, Finkenstadt Th, and Scheich H. Boston,
MA: Kluwer Academic, 1992, p. 125–149.
OTTERSEN OP AND BEN-ARI Y. Afferent connections to the amygdaloid complex of the rat and cat. J Comp Neurol 187: 401– 424, 1979.
PAXINOS G AND WATSON C. The Rat Brain in Sterotaxic Coordinates. San
Diego, CA: Academic, 1986.
PEARCE JM AND HALL G. A model for Pavlovian learning: variations in the
effectiveness of conditioned but not of the unconditioned stimuli. Psychol
Rev 87: 552, 1980.
PHILLIPSON OT AND GRIFFITHS AC. The topographic order of inputs to nucleus
accumbens in the rat. Neuroscience 16: 275–296, 1985.
POREMBA A, JONES DA, AND GONZALEZ-LIMA F. Metabolic effects of blocking
tone conditioning on the rat auditory system. Neurobiol Learn Mem 68:
154 –171, 1997.
POREMBA A, JONES DA, AND GONZALEZ-LIMA F. Classical conditioning modifies cytochrome oxidase activity in the auditory system. Eur J Neurosci 10:
3035–3043, 1998.
POSNER MI, INHOFF AW, FRIEDRICH FJ, AND COHEN A. Isolating attentional
systems: a cognitive-anatomical analysis. Psychobiology 15: 107–121,
1987.
RESCORLA RA AND WAGNER AR. A theory of Pavlovian conditioning: variations in the effectiveness of reinforcement and nonreinforcement. In: Classical Conditioning II, edited by Black AH and Prokasy WF. New York:
Appleton-Century-Crofts, 1972, p. 64 –98.
RICKERT EJ, BENNETT TL, LANE P, AND FRENCH J. Hippocampectomy and the
attenuation of blocking. Behav Biology 22: 147–160, 1978.
823