Downloaded from learnmem.cshlp.

org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Object recognition memory and the rodent hippocampus

Nicola J. Broadbent, Stephane Gaskin, Larry R. Squire, et al.

Learn. Mem. 2010 17: 5-11

Access the most recent version at doi:10.1101/lm.1650110

References This article cites 21 articles, 9 of which can be accessed free at:
http://learnmem.cshlp.org/content/17/1/5.full.html#ref-list-1

Article cited in:

http://learnmem.cshlp.org/content/17/1/5.full.html#related-urls
Email alerting Receive free email alerts when new articles cite this article - sign up in the box at the
service top right corner of the article or click here

To subscribe to Learning & Memory go to:

http://learnmem.cshlp.org/subscriptions

Copyright © 2010 Cold Spring Harbor Laboratory Press

 Downloaded from learnmem.cshlp.org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Research

Object recognition memory and the rodent

hippocampus
Nicola J. Broadbent,1 Stephane Gaskin,2 Larry R. Squire,1,3,4,5 and Robert E. Clark1,3,6
1
Department of Psychiatry, University of California, La Jolla, California 92093, USA; 2Department of Psychology, Concordia
University, Montreal, Quebec H4B 1R6, Canada; 3Veterans Affairs Medical Center, San Diego, California 92161, USA;
4
Department of Neurosciences, University of California, La Jolla, California 92093, USA; 5Department of Psychology, University of
California, La Jolla, California 92093, USA

In rodents, the novel object recognition task (NOR) has become a benchmark task for assessing recognition memory.
Yet, despite its widespread use, a consensus has not developed about which brain structures are important for task
performance. We assessed both the anterograde and retrograde effects of hippocampal lesions on performance in the
NOR task. Rats received 12 5-min exposures to two identical objects and then received either bilateral lesions of the
hippocampus or sham surgery 1 d, 4 wk, or 8 wk after the final exposure. On a retention test 2 wk after surgery, the 1-d
and 4-wk hippocampal lesion groups exhibited impaired object recognition memory. In contrast, the 8-wk hippocampal
lesion group performed similarly to controls, and both groups exhibited a preference for the novel object. These same
rats were then given four postoperative tests using unique object pairs and a 3-h delay between the exposure phase and
the test phase. Hippocampal lesions produced moderate and reliable memory impairment. The results suggest that the
hippocampus is important for object recognition memory.

Recognition memory refers to the ability to judge a previously retrograde memory that has involved the NOR task, recognition
encountered item as familiar and depends on the integrity of the memory was impaired when a 5-wk interval intervened between
medial temporal lobe (Squire et al. 2007). Tasks that assess training and hippocampal surgery (Gaskin et al. 2003). It remains
recognition memory (and object recognition memory in particu- possible that memory might be spared if a longer delay was
lar) have become increasingly useful tools for basic and preclini- imposed between training and surgery.
cal research investigating the neural basis of memory (Winters The aim of the present study was to assess both the antero-
et al. 2008). Perhaps the best known of these tasks is the novel grade and retrograde effects of hippocampal lesions on recogni-
object recognition task (NOR) (also known as the visual paired- tion memory using the NOR task. To thoroughly assess the effects
comparison task in studies with humans and monkeys). of hippocampal lesions we used (1) large groups of animals, (2)
Studies of the NOR task in humans with hippocampal multiple tests of NOR memory, (3) a scoring method that allowed
damage (McKee and Squire 1993; Pascalis et al. 2004) and in object preference to be determined on a second-by-second basis
monkeys with selective damage to the hippocampus (Pascalis and during the recognition tests, and (4) a novel training protocol that
Bachevalier 1999; Zola et al. 2000; Nemanic et al. 2004) have permitted the evaluation of recognition memory even after a re-
resulted in clear and consistent findings. Damage limited to the tention interval as long as 10 wk.
hippocampus is sufficient to produce impaired recognition mem-
ory (Squire et al. 2007, Box 1). In rats and mice, the NOR task has Results
become particularly popular and is currently a benchmark task for
assessing recognition memory. Yet despite its widespread use in Neurohistological findings
rodents, the findings are rather mixed. For example, in the rat, Figure 1 illustrates the smallest (black) and largest (stippled)
although there is agreement that the perirhinal cortex is critically extents of the hippocampal lesion for each of the training-surgery
important for normal NOR performance, there is less agreement lesion groups. All rats sustained bilateral damage to all the cell
about the hippocampus (for review, see Winters et al. 2008). fields of the hippocampus. In cases where the lesion was not
Although some of the discrepancies between studies may be complete at a particular level of the hippocampus, the sparing was
attributed to differences in lesion size and in the length of the typically restricted to the most medial aspect of the dorsal dentate
retention delay (Broadbent et al. 2004), these factors cannot gyrus or dorsal CA1 cell field, or the ventral-most region of the
account for all the findings (Squire et al. 2007). hippocampus. In all rats there was damage to the cortex and to the
Whereas most studies have investigated the effects of hip- fimbria overlying the dorsal hippocampus, which was associated
pocampal lesions on postoperative NOR performance, there is with the placement of the Hamilton syringe during surgery. There
also interest in the effects of hippocampal lesions on memory was no evidence of damage to the amygdala. Because impaired
for previously encountered objects. For a number of tasks, hippo- performance on the NOR task requires large hippocampal lesions
campal lesions produce temporally graded retrograde amnesia, (Broadbent et al. 2004), two rats with less than 75% total hippo-
such that memory acquired recently is impaired and memory campal damage were removed from further analysis. A summary
acquired more remotely is spared (for review, see Squire et al. 2004; of the lesion size for each of the hippocampal groups is shown in
Frankland and Bontempi 2005). In the case of the single study of Table 1.

6
One-day hippocampal lesion group
Corresponding author.
E-mail reclark@ucsd.edu; fax (858) 534-1569. Fifteen rats had lesions to the hippocampus that involved 75%–
Article is online at http://www.learnmem.org/cgi/doi/10.1101/lm.1650110. 100% of total hippocampal volume (mean lesion size = 91%). All

17:5–11 5 Learning & Memory

ISSN 1549-5485/10; www.learnmem.org
 Downloaded from learnmem.cshlp.org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Recognition memory and the hippocampus

the training-surgery intervals (1 d, 4 wk, and 8 wk). We calculated

the preference for the novel object during the first 15 sec (Fig. 2A)
and also during the first 30 sec (Fig. 2B) of cumulative object
exploration.
The control groups exhibited a strong preference for the
novel object by both measures (one-sample, two-tailed t-tests,
15 sec: 1-d group, 72.6 6 4.2%; 4-wk group, 77.6 6 3.9%; 8-wk
group, 65.3 6 4.3%; 30 sec: 1-d group, 72.2 6 3.1%; 4-wk group,
75.1 6 3.7%; 8-wk group, 62.6 6 4.3%; P < 0.01). A one-way
ANOVA for the training-surgery interval failed to reach signifi-
cance for the control groups (30 sec: F(2,44) = 3.1, P = 0.06). There
was a difference between the 4-wk and 8-wk control groups (t(29) =
2.2, P = 0.04); however, the 1-d group did not differ from either the
4-wk or 8-wk groups. Accordingly, though the 8-wk group per-
formed numerically the poorest, there was not strong evidence for
forgetting in intact animals across the three training-surgery
intervals.
Neither the 1-d lesion group nor the 4-wk lesion group
exhibited a significant preference for the novel object (15 sec:
Figure 1. Reconstructions of coronal sections through the hippocam-
pus showing the smallest (black) and largest (stippled) lesions for each of Hippocampal 1-d, 61.0 6 7.5% [t = 1.47, P = 0.17]; hippocampal
the three hippocampal lesion groups (1-d, 4-wk, and 8-wk training- 4-wk, 55.9 6 4.6% [t = 1.3, P = 0.23]; 30 sec: hippocampal 1-d,
surgery intervals). Numbers (right) represent the distance (mm) posterior 58.0 6 7.3% [t = 1.0, P = 0.29]; hippocampal 4-wk, 56.4 6 3.9% [t =
to bregma. 1.6, P = 0.12]; chance = 50%). In contrast to the 1-d and 4-wk
group, only the 8-wk hippocampal group exhibited a significant
preference for the novel object (15 sec: 64.5 6 4.8%, P < 0.01; 30
rats also had damage to the subiculum (25%–89%, average = 70%). sec: 62.1 6 5.8%, P = 0.057). Indeed, the 8-wk lesion group
For eight rats, damage extended beyond the hippocampus to performed similarly to the 8-wk control group (15 sec: t(28) =
encroach bilaterally upon the entorhinal cortex in the anterior- 0.13, P > 0.1; 30 sec: t(28) = 0.72, P > 0.1).
most sections. In two rats, thinning of the anterior-most region of To compare the groups at each training-surgery interval we
the perirhinal cortex was noted, but direct damage to the area first conducted a two-way ANOVA (lesion group by training-
could not be confirmed. surgery interval), followed by planned comparisons (Fisher’s
PLSD) for the 15- and 30-sec time bins. For the 15-sec measure
Four-week hippocampal lesion group (Fig. 2A), there was a group effect (F(1,84) = 7.7, P < 0.01), but no
Fifteen rats had lesions to the hippocampus that involved 87%– effect of training-surgery interval or interaction (P > 0.12). Planned
99% of total hippocampal volume (mean lesion size = 96%). All comparisons for the 1-d group (P = 0.18) and 8-wk group (P = 0.9)
rats also had damage to the subiculum (6%–88%, average = 61%). did not reach significance. The 4-wk lesion group was impaired
In 14 rats the lesion encroached bilaterally upon the anterior-most relative to the control group (P = 0.001). For the 30-sec measure
regions of the entorhinal cortex, and in one rat there was uni- (Fig. 2B), there was also a group effect (F(1,84) = 7.9, P < 0.01), but
lateral entorhinal damage. In one rat, thinning of the anterior- no effect of training-surgery interval or interaction (P > 0.12).
most region of the perirhinal cortex was noted, but direct damage Planned comparisons for the 1-d group (P = 0.078) and 8-wk group
to the area could not be confirmed.

Eight-week hippocampal lesion group

Fourteen rats had lesions to the hippo-
campus that involved 75%–98% of total
hippocampal volume (mean lesion size =
90%). Eleven rats also had damage to the
subiculum (12%–74%, average = 40%). In
seven rats, the lesion encroached bilater-
ally upon the anterior-most regions of
the entorhinal cortex, and three rats had
unilateral entorhinal damage. An addi-
tional two rats had thinning of the ante-
rior-most region of the perirhinal cortex,
but direct damage to the area could not
be confirmed, and a further two rats had
very minor damage to the dorsal thala-
mus (one unilateral and one bilateral).
Figure 2. Postoperative retention at three training-surgery intervals. Rats received either sham
Behavioral findings surgeries or bilateral hippocampal lesions 1 d (CON = 16, H = 15), 4 wk (CON = 16, H = 14) or 8 wk
(CON = 16, H = 14) after training. (A) The cumulative preference for the novel object after 15 sec of
Postoperative retention of three object exploration. The control groups at each training-surgery interval exhibited a preference for the
training-surgery intervals novel object, whereas only the 8-wk hippocampal lesion group exhibited this preference. (B) The
cumulative preference for the novel object after 30 sec of object exploration. As in A, the control groups
Figure 2A,B shows the preference for the at each training-surgery interval exhibited a preference for the novel object. However, unlike the 15 sec
novel object exhibited by the control and analysis (A), the 8-wk hippocampal lesion group fell just short of performing above chance (P = 0.057).
hippocampal lesion groups for each of Asterisks indicate group differences (P < 0.05).

www.learnmem.org 6 Learning & Memory

 Downloaded from learnmem.cshlp.org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Recognition memory and the hippocampus

(P = 0.95) did not reach significance. The 4-wk lesion group was interval groups were considered separately, negative correlations
impaired relative to the control group (P = 0.01). were observed in all three sham groups (sham 1-d, r = 0.70, P <
We also analyzed performance using both measures, but now 0.01; 4-wk, r = 0.40, P = 0.16; 8-wk, r = 0.62, P < 0.01). Thus, the
excluding those rats that had thinning of the perirhinal cortical animals that initially explored the objects the least were the
region (unilateral or bilateral). Removing these animals did not animals that subsequently exhibited the strongest preference for
change the overall pattern of findings. For the 15-sec measure (Fig. the novel object. As might be expected, for the 1-d and 4-wk
2A), the 1-d hippocampal group and the 4-wk hippocampal group hippocampal groups, there was no relationship between the
exhibited no preference for the novel object (P = 0.11 and 0.30, average amount of object exploration during familiarization and
respectively), and the 4-wk hippocampal group was impaired rel- subsequent novel object preference (presumably because object
ative to controls (t(27) = 3.6, P < 0.01). In contrast, the 8-wk hip- preference was at chance for these two groups). However, the 8-wk
pocampal group exhibited a significant preference for the novel hippocampal group did exhibit a preference for the novel object,
object (P < 0.05) and performed similarly to controls (t(26) = 0.3, P = and this group did exhibit a relationship between the average
0.8). The 30-sec measure also indicated the 1-d and 4-wk hippo- amount of object exploration during familiarization and novel
campal lesion groups performed at chance (P = 0.22 and 0.19, object preference (r = 0.59, P < 0.02).
respectively). The major difference in results, when the two rats Together, these findings indicate that the less time animals
with perirhinal thinning were excluded, was that by the 30-sec spent exploring the objects during the familiarization phase, the
measure the 8-wk hippocampal group was not significantly above stronger was the novel object preference during the test phase. The
chance (58.9 6 6.3%; t(11) = 1.4, P = 0.19). implication is that animals that learned about the familiar objects
Last, we conducted a two-way ANOVA involving the hippo- more effectively became less interested in the objects across the
campal lesion and control groups and the three training surgery multiple familiarization episodes (12 episodes across 4 d) than
intervals (1-d, 4-wk, and 8-wk). The pattern of findings did not animals that learned about the objects less efficiently. The finding
differ when we based the analysis on the 15- or 30-sec time bins, or that this relationship was found in animals that received hippo-
when we either included or excluded the animals with perirhinal campal lesions 8 wk after training (but not in animals that received
thinning. Accordingly, we present only an analysis from the hippocampal lesions earlier) provides further evidence of normal
30-sec time bin and include animals with perirhinal thinning. memory retention in this group.
There was an effect for group (F(1,84) = 7.9, P < 0.01), but no effect
of training-surgery interval (F(2,84) = 0.3, P > 0.1) or interaction Postoperative training and testing
(F(2,84) = 1.9, P > 0.1). A one-way ANOVA for training-surgery Figure 3A,B shows the percent preference for the novel object for
interval did not reach significance for the control group (F(2,44) = each of the four tests of object recognition and the mean
3.1, P = 0.06), or the lesion group (F(2,40) = 0.3, P = 0.78). performance for both 15 sec of cumulative object exploration
Accordingly, we were unable to reliably document forgetting in (A) and 30 sec of cumulative object exploration (B). On each day
the control groups or an increase in performance in the lesion the interval between familiarization and testing was 3 h.
group across the three training-surgery intervals. Both the control group and the hippocampal group exhibited
a preference for the novel object on each of the four tests of the
NOR task. This finding was obtained using both the 15- and 30-sec
Object exploration during the familiarization phase measures and irrespective of whether rats with apparent perirhinal
During the familiarization phase, each rat was allowed to explore thinning were included in the analysis (P < 0.05). The important
the objects for 5 min on 12 different occasions (three times each finding was that the hippocampal group exhibited less preference
day for 4 d). We calculated the total amount of time each rat spent for the novel object than the controls across the 4 d of testing
exploring the objects during these 5-min periods. The sham (15 sec: t(89) = 2.4, P < 0.05; 30 sec: t(89) = 2.3, P < 0.05). Figure 4
groups and the to-be-lesion groups spent a similar amount of shows the cumulative percent preference for the novel object
total time exploring the objects (162 6 9 and 169 6 8 sec, across 30 sec of object exploration for each group and averaged
respectively; t(88) = 0.6; P > 0.1). Further analysis of exploratory across the 4 d. A group difference emerged by the fourth second of
behavior during the familiarization phase revealed two interesting object exploration (P < 0.05), and this difference was maintained
findings. First, although the various groups exhibited a similar throughout the preference test. All data points for both groups
amount of total exploration, there was a considerable range of were above chance (P < 0.05). This analysis indicates that a robust
exploration across animals (overall mean = 166 6 6 sec; range = group difference in novel object preference emerged rapidly. The
64–315 sec). Second, there was a striking reduction in object overall pattern of exploration across the 30-sec test was similar in
exploration across the 4 d of familiarization (day 1 = 66 6 3 sec; the two groups, but the control group exhibited a much stronger
day 4 = 29 6 2 sec). This variability allowed us to explore the preference for the novel object than did the hippocampal group.
relationship between amount of exploration during the familiar- Importantly, significant group differences were not observed
ization phase and subsequent object recognition memory. Would on all of the individual test days. In fact for the 15-sec time bin
animals that had explored the objects the most thoroughly also comparisons, all of the individual test day comparisons failed to
exhibit the strongest preference for the novel object (because they reach significance (all P > 0.1). For the 30-sec time bin analysis, the
had the most experience with the original objects)? Alternatively, day 3 and day 1 comparisons were significant or marginally
perhaps the animals who had explored the objects the least would significant, respectively (day 3, t(89) = 2.3, P = 0.02; day 1, t(89) =
exhibit the strongest preference for the novel object (because they 1.9, P = 0.059), whereas the day 2 and day 4 comparisons were not
had learned about the original objects so well that they quickly significant (P > 0.1).
found them familiar and as a result explored them less overall).
Object exploration during the familiarization phase
The relationship between object exploration during the familiarization phase The animals with hippocampal lesions spent significantly more
and subsequent recognition memory 1 d, 4 wk, and 8 wk later time exploring the objects during the 15-min familiarization
Overall (three training-surgery intervals combined for the sham phases during all four test days (all P < 0.02). The lesion group,
animals), the correlation between the average amount of object on average, spent 80.0 6 27.0 sec in each familiarization phase,
exploration during familiarization and novel object preference whereas the control group on average spent 56.0 6 20.1 sec in each
was 0.57 (P < 0.001). When the individual training-surgery familiarization phase (t(89) = 4.7, P < 0.0001). One interpretation

www.learnmem.org 7 Learning & Memory

 Downloaded from learnmem.cshlp.org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Recognition memory and the hippocampus

tion phase exploration and novel object preference was not as

striking when the retention interval was only 3 h as when the
retention interval ranged from 1 d to 8 wk. Though it is unclear
what factor might account for this difference, we note that during
the familiarization periods for the long-term retention tests,
animals accrued much more contact time with the objects than
during the single 15-min familiarization phase used for the 3-h
delay (166 6 6 sec vs. 68 6 3 sec, respectively, t(88) = 19.2, P < 0.01).
Perhaps more experience with the objects is required before a clear
relationship between object exploration and subsequent object
preference emerges.

Discussion
Rats were given 5-min exposures to objects on 12 different
occasions during 4 d and then they received either bilateral lesions
of the hippocampus or sham surgery 1 d, 4 wk, or 8 wk after the
final 5-min exposure. On a retention test 2 wk after surgery, the
1-d and 4-wk hippocampal lesion groups exhibited impaired ob-
ject recognition memory. In contrast, the 8-wk hippocampal
lesion group performed similarly to controls, and both groups
exhibited a significant preference for the novel object. When these
same groups of rats were then given four postoperative tests with
the same task using unique object pairs, the hippocampal lesion
group was moderately impaired when performance was averaged
across the four tests.

Postoperative recognition memory

The present study assessed the anterograde effects of hippocampal
Figure 3. Postoperative training and testing of NOR on four different lesions on recognition memory using a large group of animals and
days in sham operated animals (CON, n = 47) and animals with multiple behavioral tests. As a consequence, we were able to
hippocampal lesions (H, n = 44). Performance was scored over 15 sec obtain a reliable and robust measure of performance. Even though
of cumulative object exploration (A) and over 30 sec of cumulative object
a significant deficit did not emerge on most individual test days,
exploration (B). The left side of each panel (line graphs) shows the percent
preference for the novel object on four different days for both the a reliable impairment was detected when performance was aver-
hippocampal lesion group (black circles) and the sham group (white aged across the four days of testing. The finding that individual
circles). The right side of each panel (bar graphs) shows the 4-d mean for NOR tests did not usually detect impaired performance in the
the hippocampal lesion group (black bar) and the sham group (white animals with hippocampal lesions is important because it suggests
bar). Asterisks indicate group differences (* P = 0.059, ** P < 0.05). that a single test of object recognition memory will often be
insufficiently sensitive to reveal impaired recognition memory
after hippocampal damage. One notable difference between the
of this finding is that rats with hippocampal lesions returned to present study and others that reported no impairment (see
and explored the objects more than the control group because
they had memory impairment. Accordingly, the objects became
familiar to rats with hippocampal lesions at a slower rate than
control rats (i.e., the objects remained interesting for the lesioned
rats for a longer time than they did for the control rats). The
increased exploration of the lesion group could also be related to
a general increase in activity and exploration, though increased
exploration itself could of course be indicative of memory impair-
ment.

The relationship of object exploration during the familiarization phase

to recognition memory 3 h later
We next examined the relationship between the amount of object
exploration during the familiarization phase and preference for
the novel object during the test phase three hours later. For the
control group, there was no overall relationship between famil-
iarization and test phases (r = 0.08, P > 0.1), although a negative
relationship did appear on test day 4 (r = 0.38, P < 0.01). The
hippocampal lesion group did exhibit a relationship between
object exploration time and novel object preference (r = 0.33, Figure 4. Cumulative percent preference for the novel objects across
P < 0.05), but this finding was driven primarily by five poor- 30 sec of object exploration averaged across four different days for the
hippocampal lesion group and the control group. The pattern of
performing rats (of the 44 total) that explored the objects more
performance indicates that a group difference emerged by the fourth
during the familiarization than all the other rats. When these rats second of object exploration and this difference was maintained through-
were removed from the analysis, the correlation did not approach out the remainder of the preference test. All points for both groups were
significance (P > 0.1). Thus the relationship between familiariza- above chance. Asterisks indicate group differences (P < 0.05).

www.learnmem.org 8 Learning & Memory

 Downloaded from learnmem.cshlp.org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Recognition memory and the hippocampus

Table 1. Mean percent of total hippocampal damage (and the two conditions (3 h vs. 1 d or 4 wk). This observation is also
standard error) for the hippocampal lesion groups (1-d, 4-wk, 8-wk) consistent with earlier work showing that, whereas rats trained
and tested after hippocampal lesions can exhibit mild memory
Training-surgery interval Hippocampal damage
impairment, or even no memory impairment, rats trained before
1-d 91.5 6 2.3% (range, 75%–100%) a hippocampal lesion can exhibit severe impairment (e.g., context
4-wk 96 6 1.0% (range, 87%–99%) fear conditioning, Maren et al. [1997]; visual discrimination
8-wk 90 6 1.7% (range, 75%–98%) learning, Driscoll et al. [2005]; Broadbent et al. [2007]; Epp et al.
[2008]; NOR, Gaskin et al. [2003]). This pattern of findings
suggests that under normal conditions the hippocampus is criti-
cally involved in learning, such that subsequent hippocampal
Winters et al. 2008) is that our study involved a considerably larger damage severely impairs memory. In contrast, when the hippo-
number of rats with hippocampal lesions than did any of the campus is damaged before learning, the task can sometimes be
earlier studies (n = 6–17 rats with hippocampal lesions vs. n = 44 acquired using alternative strategies that depend on other brain
rats with hippocampal lesions in the present study). Given that regions.
the recognition memory impairment in the present study was The severe retention impairment in animals that received
moderate and that the performance measure in this task can range lesions 1 d or 4 wk after training is consistent with the single
from 0% to 100% even in control animals, detecting impairment previous study that evaluated postoperative retention of preoper-
may require both multiple test trials and large group sizes. ative memory using the NOR task (Gaskin et al. 2003). In that
Additionally, we have previously suggested that lesion size and study, rats received hippocampal lesions either 48 h or 5 wk after
delay length are critical factors influencing whether impaired training, and recognition memory was severely impaired.
object recognition memory is detected after hippocampal damage In contrast to the findings in the 1-d and 4-wk groups, remote
(Broadbent et al. 2004). Hippocampal lesions are more likely to memory for the familiar object was intact when hippocampal
result in impaired recognition memory when the lesion size is lesions were made 8 wk after training. However, this finding
large (>75%; Broadbent et al. 2004) and when the delay length is should be interpreted with caution. The strongest evidence for
sufficiently long (>10 min; Clark et al. 2000). Although the present temporal gradients of retrograde amnesia comes from studies
study cannot rule out the possible contribution of a nonmemory where the performances of the lesion groups differ as a function
performance deficit, previous work using the NOR task has ruled of training-surgery interval, such that remote-memory perfor-
against this possibility by showing a delay-dependent memory mance is significantly better than recent-memory performance
impairment following hippocampal lesions (e.g., Clark et al. 2000). (e.g., Clark et al. 2002). In the present case, performance in the
It has been suggested that hippocampal lesions impair object 8-wk group was significantly above chance (unlike performance in
recognition when spatial or contextual cues are visible to rats the 1-d and 4-wk groups), but the 8-wk group did not perform
during testing in large open-field arenas (Winters et al. 2004; significantly better than the 1-d or 4-wk lesion groups. Further, the
Forwood et al. 2005). Further, large arenas might exacerbate interaction term from a two-way ANOVA for lesion group and
impaired exploratory behavior arising from hippocampal lesions, training-surgery interval was not significant. Accordingly, we
thus resulting in poor performance. Yet, we have found memory cannot reach an unambiguous, strong conclusion regarding tem-
impairment after hippocampal lesions both with small arenas that poral gradients from these findings. Although this feature of the
block most external contextual cues and minimize nonobject data tempers our conclusions about the sparing of remote memory
exploratory behavior (Broadbent et al. 2004), as well as with large after hippocampal lesions, there is an additional reason to suppose
arenas where external contextual cues were visible (Clark et al. that the 8-wk group is different from the 1-d and 4-wk groups.
2000). Thus, impaired performance on the NOR task need not Specifically, the 8-wk group, like the control group, exhibited
depend on the presence of contextual cues or impaired explor- a negative relationship between total exploration time during the
atory behavior. familiarization phase and NOR performance (8-wk hippocampal
To summarize, it appears that impaired recognition memory group, r = 0.59; control group, r = 0.62). That is, the animals
is most likely detected after hippocampal damage when large that explored the objects most during the familiarization phase
numbers of rats are used to assess performance, when recognition were the ones who later exhibited the weakest preference for the
memory is assessed with multiple independent tests, when the novel object. The 1-d and 4-wk lesion groups did not exhibit this
hippocampal lesion is essentially complete, and when the delays relationship.
are longer than several minutes. It is worth mentioning another report that found a positive
Although we did observe impairment, it is notable that relationship, rather than a negative relationship, between explo-
animals with hippocampal lesions performed above chance on ration time during the familiarization phase and NOR perfor-
each test day, and that only a moderately severe impairment mance (Albasser et al. 2009). The difference between that study
was observed. Accordingly, structures in addition to the hippo- and the present study is that in Albasser et al. (2009), rats had
campus must be capable of supporting recognition memory, even restricted exposure times during familiarization training (mean =
across retention intervals as long as 3 h. Thus, although the 64 sec), whereas in our study rats had extended exposure times
hippocampus makes an important contribution, structures such (mean = 162 sec). Interestingly, in both studies, the animals that
as the perirhinal cortex can support performance to some extent accumulated 80–120 sec of object exploration during the famil-
(Winters et al. 2004, 2008; Winters and Bussey 2005). iarization phase performed the best in the subsequent NOR test.
Animals that explored less than this amount performed poorly on
Postoperative retention of preoperative memory the NOR test (Albasser et al. 2009), and animals that explored
Rats that received hippocampal lesions 1 d or 4 wk after training more than this amount also performed poorly (present study). We
exhibited no memory for the previously encountered familiar suggest that poor NOR performance can occur either because
object (i.e., performance was at chance). When the lesion was animals insufficiently explore the objects during the familiariza-
made 8 wk after training, performance was intact. The deficit tion phase (Albasser et al. 2009) or because some animals tend to
in the 1-d and 4-wk groups was more severe than the deficit in encode the objects inefficiently. These rats will tend to explore the
recognition memory detected postoperatively. This difference in objects more during the familiarization phase and perform poorly
severity might be related to the length of the retention delay for during the test.

www.learnmem.org 9 Learning & Memory

 Downloaded from learnmem.cshlp.org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Recognition memory and the hippocampus

In summary, the present findings indicate that the hippo- with Lambda. For the lesion group (H), bilateral excitotoxic
campus is important for object recognition memory as assessed by hippocampal lesions were produced by local microinjections of
the NOR task. Hippocampal lesions produce a moderate and ibotenate acid (IBO; Biosearch Technologies). IBO was dissolved in
reliable anterograde memory impairment. This finding appeared 0.1 M phosphate-buffered saline to provide a solution with
to depend on using large groups of animals and multiple tests, a concentration of 10 mg/mL, pH 7.4. IBO was injected at a rate
of 0.1 mL/min with a 10 mL Hamilton syringe mounted on
which may explain why several studies have not observed im- a stereotaxic frame and held with a Kopf Microinjector (model
paired recognition memory in the NOR task after hippocampal 5000). The syringe needle was lowered to the target coordinate
damage (e.g., Winters et al. 2004; Forwood et al. 2005; Mumby and left in place for 1 min before beginning the injection.
et al. 2005). Furthermore, the findings suggest that the hippocam- Following the injection, the syringe needle was left in place for
pus may play a time-limited role in memory for this task, though a further 2 min to reduce the spread of IBO up the needle tract. For
our data on this point do not permit a strong conclusion. Together the lesion group, a total of 0.51 mL of IBO was injected into 18 sites
with previous work involving rodents, nonhuman primates, and within each hippocampus (for coordinates, see Clark et al. 2000).
memory-impaired patients, the present study provides strong The procedure for the sham-operated control (CON) group was the
support for the conclusion that normal recognition performance same as for the lesion groups, with the exception that the dura was
not punctured, the syringe needle was not lowered into the cortex,
depends on the integrity of the hippocampus (Squire et al. 2007). and no IBO was injected. Once awake and responsive, each rat was
returned to its home cage in the colony room for a 14-d recovery
period.
Methods
NOR test
Subjects
Retention testing began 15–16 d following surgery. Each rat was
Subjects were 94 male, Long–Evans rats weighing between 300 first rehabituated to the testing area by being placed in the empty
and 350 g at the beginning of the study. Rats were individually box for 1 min. The rat was then removed, two objects (one novel
housed and maintained on a 12:12 h light:dark cycle. Food and object and a copy of the object from the familiarization phase)
water were freely available. Rats were randomly assigned to receive were placed in the box and the rat was allowed to explore the
either bilateral lesions of the hippocampus (H) or sham surgeries objects for 10 min. Object exploration was later scored from video
(CON). Surgery occurred either 1 d after completion of training recordings of each trial by an experimenter who was blind to the
(H = 16; CON = 16); 4 wk after training (H = 15; CON = 16) or 8 wk group membership of the rats during testing and during off-line
after training (H = 15; CON = 16). data analysis. Object exploration was scored when the rat’s nose
was within 1 cm of the object and the vibrissae were moving (see
Apparatus Clark et al. 2000). Object exploration was not scored when the rat
The NOR task was conducted in an opaque plastic box measuring used the object to rear upward with the nose of the rat facing the
35 cm 3 41.5 cm 3 50 cm high. Stimuli consisted of ceramic or ceiling. Preference for the novel object was expressed as the
plastic objects that varied in color and size (width = 7.6–8.9 cm; percent time that a rat spent exploring the novel object (compared
height = 7.5–12.7 cm). Three identical copies of each object were to the familiar object). Which object served as the novel object
available. The objects were secured to the floor of the box using and the left/right position of the novel object were counter-
Velcro strips situated ;9 cm apart. A video camera mounted on the balanced within each group.
wall directly above the box was used to record the testing session To characterize postoperative retention, preference for the
for off-line analysis. Overhead fluorescent lighting illuminated the novel object was evaluated during the first 15 sec of cumulative
box. object exploration, as well as during the first 30 sec of cumulative
object exploration. Previous work has shown (e.g., Clark et al.
2000) that the strongest novel object preference scores tend to
Procedure occur early in the test phase (i.e., during the time when the novel
object is still relatively novel—as opposed to later in the test trial
Habituation when continuing exploration of the novel object makes it more
Rats were acclimated to the testing room and chamber for two familiar).
consecutive days prior to testing (45 min in the testing room and 5
min to explore the empty box). Anterograde memory testing
Beginning 5–16 d after the retention test, the same rats were given
Object familiarization four additional NOR tests. Rats received one session per day for
Rats were given 4 d of familiarization with the two identical 4 d with novel pairs of objects in each session. Each daily session
sample objects (three exposures each day). On each day of testing, was conducted as follows: The rat was first acclimated to the
rats were acclimated to the testing room for 45 min and then testing room for 45 min and then placed in the empty box for
placed in the empty box for 1 min. Then the rat was removed and 1 min. Then the rat was removed, and two identical objects were
two identical objects were placed centrally 9 cm apart. The rat was placed centrally 9 cm apart. The rat was then allowed to explore
then placed back in the box and allowed to explore for 5 min. the box and the objects for 15 min. After a delay of 3 h, the rat was
Once all rats in the group were familiarized with the objects, the returned to an empty box for 1 min, and then reintroduced to the
same procedure was repeated two more times on the same day box in the presence of two objects: a novel object and a copy of the
(mean interval between daily exposures about 60 min). Alto- previously encountered object. Preference for the novel object was
gether, rats were repeatedly exposed to the same two identical expressed as the percent time that a rat spent exploring the novel
objects for a total of 12, 5-min familiarization exposures distrib- object (compared to the familiar object).
uted across 4 d. The same sample objects were used for each of the
three training-surgery interval groups. Histology
At completion of testing, the rats were administered an overdose
Surgery of sodium pentobarbital and perfused transcardially with buffered
Surgery designed to remove the entire hippocampus was con- 0.9% NaCl solution followed by 10% formaldehyde solution (in
ducted 24–36 h (designated as the 1-d group), 4 wk, or 8 wk after 0.1 M phosphate buffer). The brains were then removed and
the final familiarization day. Anesthesia was maintained through- cryoprotected in 20% glycerol/10% formaldehyde. Coronal sec-
out surgery with isoflurane gas (0.8%–2.0% isoflurane delivered in tions (50 mm) were cut with a freezing microtome beginning at the
O2 at 1 L/min). The rat was placed in a Kopf stereotaxic in- level of the anterior commissure and continuing caudally through
strument, and the incisor bar was adjusted until Bregma was level the length of the hippocampus. Every fifth section was mounted

www.learnmem.org 10 Learning & Memory

 Downloaded from learnmem.cshlp.org on February 23, 2013 - Published by Cold Spring Harbor Laboratory Press

Recognition memory and the hippocampus

and stained with thionin to assess the extent of the lesions. Lesion Forwood SE, Winters BD, Bussey TJ. 2005. Hippocampal lesions that abolish
size estimates were obtained by calculating the percent damage in spatial maze performance spare object recognition at delays of up to
1 mm increments through the anterior–posterior extent of the 48 hours. Hippocampus 15: 347–355.
Frankland PW, Bontempi B. 2005. The organization of recent and remote
hippocampus (4 sections, from 2.80 to 5.80 mm from bregma)
memories. Nat Rev Neurosci 6: 119–130.
(Paxinos and Watson 1998). Each section was assessed under Gaskin S, Tremblay A, Mumby DG. 2003. Retrograde and anterograde
magnification, and the tissue was considered damaged if it was object recognition in rats with hippocampal lesions. Hippocampus 13:
absent or necrotic (i.e., hippocampal tissue was present, but there 962–969.
was no evidence of Nissl staining, or the tissue was gliotic). The Maren S, Aharonov G, Fanselow MS. 1997. Neurotoxic lesions of the dorsal
region damaged was drawn onto a control template for each hippocampus and Pavlovian fear conditioning in rats. Behav Brain Res
section, and the area of damage was calculated using an automated 110: 436–442.
tool in a computer graphics program (Canvas 8, Deneba). The area McKee RD, Squire LR. 1993. On the development of declarative memory.
J Exp Psychol Learn Mem Cogn 19: 397–404.
of damage was then summed across sections and calculated as
Mumby DG, Tremblay A, Lecluse V, Lehmann H. 2005. Hippocampal
a percentage of the total control hippocampal area. damage and anterograde object-recognition in rats after long retention
intervals. Hippocampus 15: 1050–1056.
Nemanic S, Alvarado MC, Bachevalier J. 2004. The hippocampal/
Acknowledgments parahippocampal regions and recognition memory: Insights from
This work was supported by the Medical Research Service of the visual paired comparison versus object-delayed nonmatching in
monkeys. J Neurosci 24: 2013–2026.
Department of Veterans Affairs, the National Institute of Mental
Pascalis OP, Bachevalier J. 1999. Neonatal aspiration lesions of the
Health, the Metropolitan Life Foundation, National Institute of hippocampal formation impair visual recognition memory when
Aging grant (P50 AG05131), the National Science Foundation (SBE assessed by paired-comparison task but not by delayed nonmatching-
0542013 and 237053), the James S. McDonnell Foundation, the to-sample task. Hippocampus 9: 609–616.
Kavli Foundation, and a NSF grant (#SBE-0542013) to the Tempo- Pascalis O, Hunkin NM, Holdstock JS, Isaac CL, Mayes AR. 2004. Visual
ral Dynamics of Learning Center, a NSF Science of Learning Cen- paired comparison performance is impaired in a patient with selective
ter. We thank Susan Davis, Laura Johnson, and Brittany Masatsugu hippocampal lesions and relatively intact item recognition.
for their assistance. Neuropsychologia 42: 1293–1300.
Paxinos G, Watson C. 1998. The rat brain in stereotaxic coordinates, 4th ed.
Academic, San Diego, CA.
Squire LR, Clark RE, Bayley P. 2004. Medial temporal lobe functions
References and memory. In The cognitive neurosciences, 3rd ed. (ed. M Gazzaniga),
Albasser MM, Davies M, Futter JE, Aggleton JP. 2009. Magnitude of object pp. 691–708. MIT, Cambridge, MA.
recognition deficit associated with perirhinal cortex damage in rats: Squire LR, Wixted JT, Clark RE. 2007. Recognition memory and the medial
Effects of varying the lesion extent and the duration of the sample temporal lobe: A new perspective. Nat Rev Neurosci 8: 872–883.
period. Behav Neurosci 123: 115–124. Winters BD, Bussey TJ. 2005. Glutamate receptors in perirhinal cortex
Broadbent NJ, Squire LR, Clark RE. 2004. Spatial memory, recognition mediate encoding, retrieval, and consolidation of object recognition
memory, and the hippocampus. Proc Natl Acad Sci 101: 14515–14520. memory. J Neurosci 25: 4243–4251.
Broadbent NJ, Squire LR, Clark RE. 2007. Rats depend on habit memory for Winters BD, Forwood SE, Cowell RA, Saksida LM, Bussey TJ. 2004. Double
discrimination learning and retention. Learn Mem 14: 145–151. dissociation between the effects of peri-postrhinal cortex and
Clark RE, Zola SM, Squire LR. 2000. Impaired recognition memory in rats hippocampal lesions on tests of object recognition and spatial memory:
after damage to the hippocampus. J Neurosci 20: 8853–8860. Heterogeneity of function within the temporal lobe. J Neurosci 24:
Clark RE, Broadbent NJ, Zola SM, Squire LR. 2002. Anterograde amnesia and 5901–5908.
temporally-graded retrograde amnesia for a nonspatial memory task Winters BD, Saksida LM, Bussey TJ. 2008. Object recognition memory:
following lesions of hippocampus and subiculum. J Neurosci 22: 4663– Neurobiological mechanisms of encoding, consolidation and retrieval.
4669. Neurosci Biobehav Rev 32: 1055–1070.
Driscoll I, Howard SR, Prusky GT, Rudy JW, Sutherland RJ. 2005. Seahorse Zola SM, Squire LR, Teng E, Stefanacci L, Buffalo EA, Clark RE. 2000.
wins all races: Hippocampus participates in both linear and non-linear Impaired recognition memory in monkeys after damage limited to the
visual discrimination learning. Behav Brain Res 164: 29–35. hippocampal region. J Neurosci 20: 451–463.
Epp J, Keith JR, Spanswick SC, Stone JC, Prusky GT, Sutherland RJ. 2008.
Retrograde amnesia for visual memories after hippocampal damage in
rats. Learn Mem 15: 214–221. Received September 28, 2009; accepted in revised form October 5, 2009.

www.learnmem.org 11 Learning & Memory