Post on 18-Feb-2022
BRAINA JOURNAL OF NEUROLOGY
Gudden’s ventral tegmental nucleus is vital formemory: re-evaluating diencephalic inputs foramnesiaSeralynne D. Vann
School of Psychology, Cardiff University, Cardiff, CF10 3YG, UK
Correspondence to: Dr Seralynne Vann,
School of Psychology,
Cardiff University,
Cardiff, CF10 3YG,
UK
E-mail: vannsd@cardiff.ac.uk
Mammillary body atrophy is present in a number of neurological conditions and recent clinical findings highlight the importance
of these nuclei for memory. While most accounts of diencephalic amnesia emphasize the functional importance of the
hippocampal projections to the mammillary bodies, the present study tested the importance of the other major input to the
mammillary bodies, the projections from the ventral tegmental nucleus of Gudden (VTNg). Although the VTNg, and its projec-
tions to the mammillary bodies, is present across species, the size and location of this structure has made it an extremely
difficult structure to assess in primates. The effects of selective, neurotoxic lesions of the VTNg were, therefore, assessed in rats.
The animals with these lesions were impaired on a series of spatial learning tasks, namely delayed-matching-to-place in the
water maze, T-maze alternation and working memory in the radial arm maze. Normal performance on these tasks is dependent
on those brain structures (e.g. hippocampus and mammillary bodies) that are now assumed to cause anterograde amnesia when
damaged in humans. In contrast, the same rats with ventral tegmental nucleus lesions performed normally on two control tasks:
the acquisition and subsequent reversal of an egocentric discrimination task and a visually cued task in the water maze.
This study provides the first clear evidence that the VTNg is critical for memory, and consequently indicates that diencepha-
lic–hippocampal models of memory should be extended to incorporate the limbic midbrain.
Keywords: amnesia; dorsal tegmental nucleus; mammillary body; rat
Abbreviations: AchE = acetylcholinesterase; ITI = inter-trial interval; VTNg = ventral tegmental nucleus of Gudden
IntroductionConvergent evidence from both clinical (Van der Werf et al.,
2000; Carlesimo et al., 2007; Tsivilis et al., 2008; Vann et al.,
2009) and animal (Parker and Gaffan, 1997; Vann and
Aggleton, 2003) research increasingly shows that mammillary
body dysfunction is a major contributor to diencephalic amnesia.
Mammillary body atrophy has now been reported in a number of
conditions, including Korsakoff’s syndrome (Victor et al., 1989;
Kopelman, 1995), colloid cysts in the third ventricle (Denby
et al., 2009), Alzheimer’s disease (Callen et al., 2001;
Copenhaver et al., 2006; Grossi et al., 1989), schizophrenia
(Bernstein et al., 2007; Briess et al., 1998), heart failure (Kumar
et al., 2009) and sleep apnoea (Kumar et al., 2008), highlighting
the importance of understanding the functions of this structure.
Explanations for the importance of the mammillary bodies to
memory have inevitably focused on their dense inputs from the
hippocampal formation (Allen and Hopkins, 1989), with the
doi:10.1093/brain/awp175 Brain 2009: 132; 2372–2384 | 2372
Received February 20, 2009. Revised April 29, 2009. Accepted May 22, 2009. Advance Access publication July 14, 2009
� The Author (2009). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: journals.permissions@oxfordjournals.org
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
mammillary bodies seen as primarily relaying information to the
anterior thalamic nuclei (Delay and Brion, 1969; Gaffan, 1992;
Aggleton and Brown, 1999). Such explanations fail, however, to
account for why there should be a mammillary body relay, given
that there are dense, direct projections from the hippocampus to
the anterior thalamus (Poletti and Creswell, 1977). The explana-
tion is almost certain to come from understanding the functions
of the non-hippocampal inputs to the mammillary bodies, with
priority given to those inputs that do not also project to the
anterior thalamic nuclei.
The two principal inputs to the mammillary bodies are from
the hippocampal formation and the tegmental nuclei of Gudden
(Allen and Hopkins, 1989). The ventral tegmental nucleus of
Gudden (VTNg) has dense, reciprocal connections with the
medial mammillary nucleus, while the dorsal tegmental nucleus
of Gudden (DTNg) is densely interconnected with the lateral
mammillary nucleus. These two, parallel pathways have very
different electrophysiological properties. Cells in the VTNg fire
rhythmically, and highly coherently, in theta bursts (Kocsis et al.,
2001), while the DTNg contains head-direction and angular-
velocity cells (Bassett and Taube, 2001; Sharp et al., 2001;
Bassett et al., 2007). Consequently, it has been supposed that
there are two parallel systems associated with mammillary body
function, i.e. a ‘lateral’ head-direction system and a ‘medial’ theta
system (Vann and Aggleton, 2004). The present study focused
on the ‘medial’ system as only the medial mammillary nucleus is
consistently atrophied in the amnesic Korsakoff’s syndrome (Victor
et al., 1989; Kopelman, 1995). Furthermore, lesion studies in
rodents reinforce the importance of the medial nucleus for spatial
learning and memory (Vann, 2005; Vann and Albasser, 2009).
Finally, there is a report of a man with amnesia that was attributed
to pathology in the VTNg area (Goldberg et al., 1981). The lack of
projections from VTNg to the anterior thalamic nuclei adds to the
potential significance of this pathway for mammillary body
function.
Despite recent discoveries regarding the electrophysiological
properties of the VTNg (Bassant and Poindessous-Jazat, 2001;
Kocsis et al., 2001), there are no lesion studies that have selec-
tively tested the functional importance of this nucleus. Previous
behavioural evidence (Lorens et al., 1975; Wirtshafter and Asin,
1983; Asin and Fibiger, 1984) has come from studies targeted at
adjacent areas (e.g. the median raphe nuclei) and none have used
axon-sparing methods to assess performance on standard spatial
memory tasks. The size and location of the VTNg almost comple-
tely prohibits experimental studies of these nuclei in the primate
brain, where they again project heavily to the mammillary bodies
(R.C. Saunders, unpublished results). The present study, therefore,
examined the impact of selective VTNg lesions in the rat.
The effects of cytotoxic VTNg lesions were assessed on spatial
memory tasks (water maze, T-maze and radial arm maze)
known to be sensitive to both mammillary body and hippocampal
lesions (Sziklas and Petrides, 1993; Cassel et al., 1998; Bannerman
et al., 2002; Vann and Aggleton, 2003). These tasks are informa-
tive as tests of spatial memory in animals are considered to tap
core elements of episodic memory (Aggleton and Pearce, 2001;
Burgess et al., 2002). Immunohistochemical procedures helped to
determine the selectivity of the lesions.
Materials and Methods
Subjects and surgerySubjects were 20 male, pigmented rats (Dark Agouti strain; Harlan,
Bicester, UK) weighing between 238 and 268 g at the time of surgery.
Animals were housed in pairs under diurnal light conditions (14-h
light/10-h dark) and testing was carried out during the light phase.
Animals were given free access to water throughout. All experiments
were carried out in accordance with UK Animals (Scientific Procedures)
Act, 1986 and associated guidelines.
Animals were deeply anaesthetized by intraperitoneal injection
of sodium pentobarbital (60 mg/kg). The 12 rats receiving Gudden’s
ventral tegmental nuclei lesions (VTNx) were then placed in a stereo-
taxic headholder (David Kopf Instruments, Tujunga, CA, USA) with the
nose bar at 0.0, and the scalp was cut and retracted to expose the
skull. The lesions were made by injecting 0.12 M N-methyl-D-aspartate
(NMDA; Sigma Chemical Company Ltd, UK), dissolved in phosphate
buffer (pH 7.2). Injections were made in one site per hemisphere using
a 1 ml Hamilton syringe. The stereotaxic ‘arm’ was set at 20� from
vertical so that the needle enters the same hemisphere for both injec-
tions. The stereotaxic co-ordinates of the lesion placements relative to
bregma were anteroposterior (AP) �8.8, lateral (L) +2.6/+3 and the
depth, from top of cortex, was �7.2 mm. Target co-ordinates had
been determined by prior retrograde tracing studies (injections of
Fluoro-Gold) in separate rats in which those tegmental neurons
that project to the mammillary bodies were identified (Fig. 1).
The lesions involved bilateral injections of 0.18ml NMDA made over
5 min, the needle was then left in situ for a further 5 min. After
the injections of NMDA, animals were injected with 0.05 ml of a
respiratory stimulant (Millophyline, Arnolds Vetinary Products Ltd,
Shrewsbury, UK). At the completion of all surgeries, the skin was
sutured and an antibiotic powder (Acramide: Dales Pharmaceuticals,
Skipton, UK) was applied topically. Animals also received subcutaneous
injections of 5 ml glucose and were given paracetamol and sucrose in
their drinking water for 3 days post-surgery; all animals recovered well
following surgery. The eight animals acting as controls received the
same procedure and drugs as the animals receiving lesions, which
involved the needle being lowered into the same co-ordinates
but no injection was made.
Apparatus and behavioural training
Experiment 1—spontaneous locomotor activity
Animals were food-deprived overnight before being placed in novel
test cages (56 cm�39 cm �19 cm) in a novel room. Activity was mea-
sured using pairs of photobeams situated 20 cm apart and 18 cm from
the end of the cage (Paul Fray Ltd, Cambridge, UK). The total num-
bers of beam breaks were recorded. Data were gathered in 16 intervals
of 5 min each.
Experiment 2—delayed-matching-to-place in thewater maze
The water maze (200 cm diameter, 60 cm deep) was made of
white fibre glass and mounted 58 cm above the floor. The pool
was filled with water (24� 1�C), made opaque by the addition of
non-toxic emulsion (Opacifier, Chesham Chemicals, Harrow, UK).
An escape platform (10 cm diameter, 2 cm below water surface)
could be placed in the pool. The pool was in a room measuring
440�400 cm. Lighting was provided by four floor-mounted spotlights
Ventral tegmental nuclei and memory Brain 2009: 132; 2372–2384 | 2373
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
(500 W) and the room contained salient visual cues such as geometric
shapes and high-contrast stimuli on the walls. There was a curtain
hanging from the ceiling around the pool that could be opened or
closed. Swim paths were tracked with a video camera suspended
directly above each pool. Data were collected and analysed on-line
with an HVS Image analyser (HVS Image Ltd, Hampton, UK) connected
to a computer that used Water-maze Software (Edinburgh, UK).
For this experiment, both food and water were available ad libitum.
The animals were trained on a working memory task (‘delayed
matching-to-place’) in the water maze. For this, 12 platform positions,
which varied in their distance from the pool perimeter, were used
along with eight possible start positions. Animals received 2 days of
pre-training with four swims a day. For this, the curtain was drawn
closed around the pool and both the start position and platform
position were changed for every forced swim. Each swim was
terminated when the animal either located the submerged platform
or after 120 s had elapsed. If the animal had not located the platform
at the end of 120 s, it was guided there by the experimenter and then
had to remain on the platform for 30 s. The animals were transported
between the holding room and water maze in an opaque, aluminium
travelling box. They were also placed in the opaque holding box
in-between each trial.
For the 16 days of actual training, the curtain was removed from
around the pool, i.e. room cues were visible. The location of the plat-
form remained constant across the four trials of a given day but varied
between days. The same start position was used for the first two
trials of each session but was then varied for the remaining two
trials. This made it possible to match distances for trials 1 and 2, the
key comparison. Each trial terminated when the animal had either
located the platform or 120 s had elapsed. The animals were then
left on the platform for 30 s. The next trial began almost immediately
afterwards, giving an inter-trial interval (ITI) of �15 s.
Experiment 3—reinforced spatial alternation in theT-maze
Testing was carried out in a modifiable four-arm (cross-shaped) maze.
The four arms (70 cm long, 10 cm wide) were made of wood, while
the walls (17 cm high) were made of clear Perspex. At any time, one
of the arms could be blocked off to form a T-shaped maze. Aluminium
Figure 1 (A) Diagram of coronal sections showing the level and position of the ventral tegmental nucleus, taken from Paxinos
and Watson (1997); (B) VTNx following an injection of Fluoro-Gold into the mammillary bodies in a surgical sham; (C) AChE-stained
sections from a VTNx lesion animal; (D) AChE-stained sections from a surgical sham at the level of the Gudden’s VTNx; (E) Nissl-
stained section from a VTNx lesion animal; (F) Nissl-stained section showing the ventral tegmental of nuclei in a surgical sham.
All arrows point to the Gudden’s VTNx. Scale bars = 1 mm.
2374 | Brain 2009: 132; 2372–2384 S. D. Vann
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
barriers could be positioned �25 cm from the end of each arm to
create a start area. For this experiment, the location of the start arm
remained constant such that the T-maze was in the same orientation
throughout testing. The maze was supported by two stands (94 cm
high), and was situated in a rectangular room (280�300� 240 cm)
with salient visual cues.
Animals were food deprived to 485% of their free-feeding body
weight. Each animal was given 7 days of 5 min pre-training in order
to train them to run reliably down the stem of the maze and to
find food pellets in the food wells in both arms; rats were prevented
from ‘turning’ (e.g. going from start arm to either choice arm) in the
acquisition stage by blocking off the arms and the stem and placing
the rats in each part of the T-maze separately. Following pre-training,
the acquisition phase began.
At the start of each acquisition trial, which consisted of two stages,
two food pellets (45 mg; Noyes Purified Rodent Diet, Lancaster, NH,
USA) were placed in each food well and an aluminium block was
placed at the neck of the T-maze so closing off one arm. As a con-
sequence, on each ‘sample run’ the animal was forced to enter the
open arm where it was allowed to eat the food at the end of the arm.
The animal was then picked up and confined in the start box for a
delay of 10 s, during which the aluminium block was removed. The
door to the start arm was then opened and the animal allowed a free
choice between the two arms of the T-maze. On this ‘choice run’, the
animal was considered to have chosen the correct arm if it had alter-
nated, i.e. had entered the arm not previously entered on the ‘sample
run’ and would then be allowed to eat the food reward before being
returned to the holding box. If the animal made an incorrect choice,
i.e. returned to the arm visited on the ‘sample run’, the rat was con-
fined to that arm for �5 s before being returned to the travelling box.
The rats were tested in groups of four with each animal having one
trial in turn so that the ITI was �4 min. The animals received six trials a
day for a total of 6 days.
The acquisition phase was immediately followed by five sessions on
a test of continuous alternation. At the start of each of these sessions,
the rat was forced to either the right or left arm and this initial sample
run was rewarded with two pellets. This sample run was immediately
followed by 12 consecutive massed trials in which the correct choice
was always the opposite arm to the one chosen on the previous trial;
the ITI was 15 s and there were no correction trials.
This test was completed with four final continuous alternation
sessions, each of 12 trials. After each arm choice the maze was rotated
by 90�, either clockwise or anticlockwise, so that the previous choice
arm became the start arm for the subsequent trial. This manipulation
was to control for the use of intra-maze cues such as odour trails.
Experiment 4—radial arm maze
Testing was carried out in an eight-arm radial maze. The maze
consisted of an octagonal central platform (34 cm diameter) and
eight equally spaced radial arms (87 cm long, 10 cm wide). The base
of the central platform and the arms were made of wood, whereas
panels of clear Perspex (24 cm high) formed the walls of the arms.
At the start of each arm was a clear Perspex guillotine door
(12 cm high) attached to a pulley. The maze was positioned in a
room (255�330�260 cm), which contained salient visual cues such
as geometric shapes and high-contrast stimuli on the walls.
Animals were maintained on restricted feeding at 485%, of their
free-feeding body weight. Pre-training for the radial arm maze began
10 days after the completion of testing in the T-maze and involved
two habituation sessions where the animals were allowed to explore
the maze freely for 5 min with the guillotine doors raised and food
pellets (45 mg; Noyes Purified Rodent Diet) scattered down the arms.
The animals were then trained on the standard radial arm maze task
(see below). A time limit of 10 min was placed on each session.
Animals were tested until they had completed 15 sessions in Stage 1
and six sessions in Stage 2.
Stage 1 (Sessions 1–15) was the standard working memory version
of the radial arm maze task (Olton et al., 1978) where the animals’
optimal strategy was to retrieve the reward pellets from all eight arms
without re-entering any previously entered arms. At the start of a trial,
all eight arms were baited with two food pellets. The animal would
make an arm choice and then return to the central platform and all
the doors were closed for �10 s before they were opened again,
permitting the animal to make another choice. This continued until
all eight arms had been visited or 10 min had elapsed. Only trials
where animals made a minimum of eight arm choices were included
in the analyses. The number of sequential choice responses was
calculated, which is when the animals’ successive choices involve
immediately adjacent arms in a constant direction. It is measured by
giving the animal a score of +1 (clockwise) or �1 (anticlockwise) if the
arm is immediately adjacent to previous choice and 0 for any other
arm choice. A higher absolute score would therefore reflect the use of
a sequential response strategy (Olton and Samuelson, 1976; Ennaceur
and Aggleton, 1997). To compensate for overall differences in arm
entries across the two groups, the overall sequential choice score
was divided by total arm choices to give a ratio.
Stage 2 (Sessions 16–21) was to test for the possible use of intra-
maze cues in performing the task. The start of the session was as
before, but after the animal had made four different arm choices it
was kept in the centre of the maze while the maze was rotated by 45�
(clockwise/anticlockwise on alternate days), and the remaining food
pellets moved so that they were still in the same allocentric locations
but the actual arms had changed. The session then continued until
all reward pellets had been retrieved.
Experiment 5—egocentric discrimination in thecross-maze
The egocentric task started the day following completion of the radial
arm maze task and was carried out in the same room, using the same
apparatus, as the T-maze task. For this task all four arms of the maze,
and arm locations, were used. Each rat received 12 consecutive trials
a day. One of the four arms was randomly selected as the start arm
for each rat, and the opposite arm was blocked off, thereby creating
a T-maze. One of the arms of the resultant T-maze was baited with a
single food pellet (45 mg; Noyes Purified Rodent Diet), so that the rat
was rewarded for always selecting the arm in a given direction (always
left or always right). After reaching the end of the chosen arm, the rat
was placed at the end of another randomly chosen arm, and the
process was repeated. The direction to be rewarded was decided on
the first trial of the first session, when neither arm had been baited.
The direction that the rat chose was noted and for all continuing
trials the rat was rewarded only if it turned in the opposite direction.
This was to ensure that rats were not being rewarded for following
any prior bias. When rats had reached a criterion level of 30 or more
correct responses over three consecutive sessions (36 trials), they were
then moved onto reversal learning. For this, animals were now
rewarded for turning in the opposite direction to that which was
rewarded during acquisition. All rats were tested on this reversal
stage for 12 days.
Experiment 6—visual-cued task in the water maze
Animals were returned to ad libitum feeding following completion of
the egocentric discrimination task; after 4 days, testing began on the
Ventral tegmental nuclei and memory Brain 2009: 132; 2372–2384 | 2375
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
visual task in the water maze. The same pool was used as for the
previous water maze experiment, but with the curtain pulled closed
throughout; there was a metal ‘stick’ (14 cm high, 1 cm diameter,
with alternating black and white tape around the circumference)
attached to the edge of the platform. Animals received four trials
a day for 4 days. The platform positions and start positions varied
across trials.
Experiment 7—delayed-matching-to-place in thewater maze
In order to determine whether performance in the last two experi-
ments was a result of functional recovery, animals were re-tested on
the delayed-matching-to-position task in the water maze for 4 days.
Exactly the same procedure was used as earlier in the study.
Histological proceduresOn completion of the experiments, rats were deeply anaesthetized
with sodium pentobarbital (60 mg/kg, Euthatal, Rhone Merieux, UK)
and transcardially perfused with 0.1 M phosphate buffer saline (PBS)
followed by 4% paraformaldehyde (PFA) in 0.1 M PBS. The brains
were removed and postfixed in PFA for 4 h and then transferred to
25% sucrose overnight at room temperature with rotation. Sections
were cut at 40 mm on a freezing microtome in the coronal plane.
A one-in-three series of sections was mounted onto gelatine-coated
slides and stained with cresyl violet, a Nissl stain and a further
two series (also one-in-three sections) were collected in PBS to be
processed for either serotonin or the neuronal nuclei protein (NeuN).
NeuN and serotonin immunostainingIn a subset of eight animals with lesions, a series of sections was
processed for the immunohistochemical demonstration of NeuN to
enable a more accurate evaluation of nuclei in the vicinity of the
lesions (Jongen-Relo and Feldon, 2002); in the remaining six animals,
a series was processed for serotonin to assess the integrity of the
nearby raphe nuclei. Free-floating sections were rinsed in 0.1 M
PBST (PBS with 0.2% Triton X-100) and treated with 0.3% H2O2 in
0.1 M PBST for 10 min to suppress endogenous peroxidase activity.
Sections were rinsed four times with 0.1 M PBST for 10 min each.
Sections were subsequently incubated for 48 h at 4�C in the mono-
clonal anti-NeuN serum (1:5000; Chemicon, Temecula, CA, USA)
diluted in PBST or anti-serotonin antibody (1:20000, ab66047,
AbCam, UK). After rinsing four times in 0.1 M PBST (10 min each),
sections were incubated for 2 h at room temperature in biotinylated
horse anti-mouse serum (5 ml per ml PBST, Vector Laboratories,
Peterborough, UK) in secondary diluent consisting of normal horse
serum (15 ml per ml PBST, Vector Laboratories, Peterborough, UK) in
0.1 M PBST. After four rinses in 0.1 M PBST (10 min each), sections
were incubated for 1 h in an avidin–biotin–horseradish peroxidase
complex (1:200, ABC-Elite, Vector Laboratories, Peterborough, UK)
in PBST. Following four rinses in 0.1 M PBST and two rinses in
0.05 M Tris buffer, sections were placed for 1–2 min in a chromagen
solution consisting of 0.05% diaminobenzidine (Sigma Chemical
Company Ltd), buffer solution and 0.01% H2O2 (DAB substrate kit;
Vector Laboratories, Burlingame, CA, USA). The reaction was visually
monitored and stopped in rinses of cold 0.1 M PBS. The sections were
mounted and dried on gelatin-coated slides, and then dehydrated in
ascending alcohols and xylene before being cover-slipped using DPX
mounting medium.
Acetylcholinesterase staining—modified Koelle methodTissue from a subset of six rats was processed for acetylcholinesterase
(AChE). Sections were mounted onto gelatin-coated glass slides and
dried overnight in a slide oven. Slides were then immersed in incuba-
tion medium (copper sulphate 0.781 g/l, glycine 0.75 g/l and sodium
acetate 2.88 g/l dissolved in distilled water and adjusted to pH 5),
then acetylthiocholine iodide, 1.15 g/l, and ethopropazine, 0.05 g/l,
were added and the incubating solutions left overnight. Slides were
then washed four times in distilled water and then placed in sulphide
solution (10 g/l dissolved in distilled water and adjusted to pH 7.5).
When an appropriate colour had developed, the sections were washed
a further four times in distilled water. Sections were dried and then
cover-slipped as described above.
Image capturing and assessmentImages were captured using a Q Imaging MicroPublisher 3.3 RTV
camera attached to a Zeiss Axiostar Plus microscope. Intensity mea-
sures for the AChE-stained sections were acquired on a Macintosh
computer using the public domain NIH Image programme (developed
at the US National Institutes of Health and available on the Internet
at http://rsb.info.nih.gov/nih-image/). To assess the integrity of the
adjacent laterodorsal tegmental nuclei, AChE intensity measures were
taken for the anteroventral thalamic nuclei and interpeduncular
nucleus. The laterodorsal tegmental nucleus is the sole source of
anterior thalamic cholinergic innervation, as demonstrated by lesion
studies, and the laterodorsal tegmental nuclei are also a major
source of interpeduncular cholinergic inputs (Satoh and Fibiger,
1986; Sikes and Vogt, 1987). Intensity measures were taken for the
anteroventral or interpeduncular nuclei across four sections (i.e. eight
total measures for the anteroventral thalamic nuclei for both
hemispheres) and for white matter adjacent to the nuclei for each
section, the stria medullaris of the thalamus and the cerebral peduncle,
respectively. The white matter value was then subtracted from the
nucleus value, for each section separately, and the mean values
calculated.
Results
Histological analysisA stringent criterion was adopted for final inclusion in the study,
and as a consequence the final groups comprised seven Gudden’s
VTNx animals and eight surgical shams (Sham). Histological anal-
yses were performed blind to the individual animal’s behavioural
performance. In the excluded lesion animals, there was either
appreciable sparing of the VTNg (three animals) or the lesions
were complete but extended into additional structures (two
animals). Of the final seven rats, the cellular lesions of Gudden’s
ventral tegmental nucleus appeared complete in six of the animals,
while the seventh had a very small amount of sparing in the
caudal-most aspect of the nucleus.
Lesions were assessed using Nissl, NeuN and AChE-stained
sections (Fig. 1), and the loss of the large, distinctive cells in the
ventral tegmental nucleus could be clearly seen. In all animals,
there was a very small amount of damage to the medial-most
2376 | Brain 2009: 132; 2372–2384 S. D. Vann
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
part of the oral part of the reticular pontine nucleus; this damage
was always unilateral and on the contralateral side of the brain
to that in which the needle was inserted to make the lesion. The
raphe nuclei were assessed using NeuN, Nissl and serotonin-
stained sections. The dorsal and median raphe nuclei were densely
stained in both lesion and control animals (Supplementary Fig. 1),
and there did not appear to be any loss of raphe neurons at any
level other than that of the VTNg. At just this level, there was a
restricted loss of cells in the most dorsal part of the median raphe
nuclei, i.e. limited to just that part directly adjacent to the ventral
tegmental nucleus. There was no evidence that the lesions
extended into either the dorsal tegmental nuclei of Gudden or
the laterodorsal tegmental nuclei, both of which stained densely
for AChE. Although neurons in the laterodorsal and dorsal teg-
mental nuclei appeared intact, the overall volume of the structures
appeared smaller than in the surgical shams, which is consistent
with previous reports of their retrograde degeneration (Briggs and
Kaelber, 1971). Inspection of the mammillary bodies in the VTNx
lesion animals revealed them to be severely atrophied, with
the medial nuclei being most affected. These changes are again
consistent with degeneration reported in previous anatomical
studies (e.g. Briggs and Kaelber, 1971). Assessment of the
mammillary nuclei with NeuN staining suggests that this atrophy
may reflect some loss of neurons and not just fibres.
In four animals, 0.06ml Fluoro-Gold (Fluorochrome, LLC,
4% solution in distilled water) was injected into the mammillary
bodies to assess the extent of the VTNx in normal animals (Fig. 1)
and completeness of the lesion. These injections confirmed the
findings from the Nissl and NeuN-stained sections.
AChE assessmentTo confirm that the behavioural deficits were not due to the loss
of acetylcholine projections from the nearby laterodorsal tegmen-
tal nuclei, the cholinergic projection from this structure was
assessed using AChE-stained sections (Supplementary Fig. 2).
The integrity of this cholinergic projection was demonstrated by
a lack of difference in staining in either the anteroventral thalamic
nuclei or the rostral part of the interpeduncular nuclei. The
mean grey values (�SD) for the anteroventral thalamic nuclei
were: VTNx: 93.1� 7.13, 105.8� 14.77, 95.8� 5.99; Sham:
88.5� 5.83, 97.8� 6.20. The mean density (�SD) for the
interpeduncular nucleus (rostral part) were: VTNx: 122.7� 4.38,
132.9� 14.30, 115.0� 9.97; Sham: 115.0� 5.91, 128.4� 9.71.
Experiment 1—spontaneous locomotoractivityAnimals’ activity was measured in 5 min intervals across 80 min
(Fig. 2). The VTNx group was significantly more active than the
Sham group [F(1,13) = 29.94, P50.0005]. There was a main
effect of time [F(15,195) = 24.20, P50.0001] reflecting the ani-
mals’ habituation to their surroundings. In addition, there was a
significant group� time interaction [F(15,195) = 3.85, P50.0001]
as the groups only differed in activity levels during the first 30 min,
and final 5 min, of the test phase.
Experiment 2—delayed-matching-to-place in the water mazeConsistent with the hyperactivity seen in the VTNx animals in
Experiment 1, the VTNx group also had significantly faster swim
speeds than the Sham group in the water maze [F(1,13) = 84.14,
P50.0001; mean� SD: VTNx = 31.1� 5.3; Sham = 19.5� 4.9];
the VTNx group’s swim speeds remained the same throughout
training (F51), whereas the swim speeds of the Sham group
reduced across training [F(15,195) = 2.54, P = 0.002]. Due
to these differences in swim speeds only path-length data were
analysed. An analysis of variance (ANOVA) using the factors group
(2)� day (16)� trial (4) revealed that the VTNx group had signifi-
cantly longer path lengths [F(1,13) = 17.07, P50.005; Fig. 3].
There was no effect of day [F(15,195) = 1.36, P40.1] or group�
day interaction [F(15,195) = 1.29, P40.2]. There was a significant
trial effect [F(3,39) = 9.93, P50.0005] showing some improvement
by both groups over the trials. Although there was no group� trial
interaction [F(3,39) = 2.28, P = 0.09], analysis of the simple effects
revealed no group difference at Trial 1 (P40.05) but significant
group differences at Trials 2 –4 (all P50.01).
Experiment 3—reinforced spatialalternation in the T-mazeFor the 6 days of initial acquisition, where each trial was separated
by an ITI of �4 min and comprised a sample and test run, an
analysis carried out on the total correct choices revealed a signif-
icant group difference [F(1,13) = 14.7, P50.005] as the VTNx
group made fewer correct choices than the Shams (Fig. 4A).
There was no effect of day (F51) or group� day interaction
[F(5,65) = 1.47, P40.2] reflecting the stable performance
of both the groups across the acquisition stage.
Figure 2 Spontaneous activity (Experiment 1). Mean number
of beam breaks (�SE) in 5 min bins over 80 min.
Ventral tegmental nuclei and memory Brain 2009: 132; 2372–2384 | 2377
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
The animals were then tested for 5 days (12 trials a day) on the
continuous alternation procedure, which is designed to increase
levels of proactive interference (Fig. 4B). A comparison of the
number of correct choices again revealed that the VTNx group
was significantly impaired [F(1,13) = 17.01, P50.005]. There was
no overall effect of day [F(4,52) = 1.02, P40.4] although there
was a borderline group�day interaction [F(4,52) = 2.54,
P = 0.050] as the VTNx group’s performance improved over
the test days [F(4,52) = 3.06, P50.05], whereas the Shams’
performance stayed stable (F51).
The third condition assessed continuous alternation performance
but with the maze being rotated in-between trials to discourage
the use of intra-maze cues (Fig. 4B). Again, there was a significant
group difference with the VTNx group making fewer correct
choices than the Shams [F(1,13) = 8.42, P50.05]. There was no
effect of day (F51) or group�day interaction (F51) showing
that the performance of both groups was stable across test days.
To determine the effect of maze rotation on animals’
performance, an analysis was carried out on the final 2 days of
continuous alternation and on the first 2 days of continuous
alternation with maze rotation. The overall performance of the
VTNx group was significantly worse than that of the Shams
[F(1,13) = 10.16, P50.01], and there was a significant effect of
the rotation manipulation on task performance [F(1,13) = 20.20,
P50.001]. However, the lack of group� rotation interaction
(F51) indicates that the lesion group was no more sensitive
to the manipulation than the control group, although there was
a potential scaling effect.
Experiment 4—radial arm maze(acquisition and rotation)Both total errors made and number of correct entries in the first
eight choices were analysed. For the first 15 days (Stage 1), an
analysis of the number of correct entries in the first eight choices
revealed a significant difference between the sham and lesion
groups [F(1,13) = 36.3, P50.0001] as the lesion group made
fewer correct entries (Fig. 5A). There was a significant effect of
day [F(14,182) = 12.2, P50.001] but no group� day interaction
(F51), again reflecting the improvement of both groups across
training. An analysis using the number of errors also revealed a
significant main effect of group [F(1,13) = 54.7, P50.001] as
the VTNx animals made more errors than the Shams (Fig. 5B).
Figure 4 Reinforced alternation in the T-maze (Experiment 3).
(A) Mean correct choices (�SE) during the six acquisition ses-
sions where each trial comprised a sample run and a choice run
and an ITI of �4 min (six trials per session); (B) Mean correct
choices (�SE) for five sessions of continuous alternation with
12 trials per session and subsequent four sessions when the
maze was rotated between choices.
Figure 3 Delayed-matching-to-place in the water maze
(Experiment 2). Mean path length (�SE) to find the hidden
platform across the four trials of a day. Data are presented in
two blocks (Days 1–8 and Days 9–16).
2378 | Brain 2009: 132; 2372–2384 S. D. Vann
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
Again, there was a main effect day [F(14, 182) = 6.58, P50.001]
but no group� day interaction (F51) reflecting the improvement
in both groups’ performance across test days. An analysis of
sequential choice responses showed no group difference (F51)
or overall effect of the day [F(14,182) = 1.07, P40.3], although
there was a significant group� day interaction [F(14,182) = 2.77,
P50.001] as the ratio of sequential choice responses was greater
for the Shams during the first half of training but a higher
proportion of sequential choice responses was made by the
lesion group in the last 5 days of task acquisition (Fig. 5C),
which coincided with them making fewer errors (Fig. 5B).
The next 6 days of testing (Stage 2) involved trials with a
rotation of the maze after the first four choices. This was to tax
the use of extra-maze cues. The VTNx group made significantly
more errors [F(1,13) = 22.9, P50.001; Fig. 5A] and fewer correct
entries in the first eight choices [F(1,13) = 28.6, P50.001; Fig. 5B].
There was neither an effect of day using either measure [errors,
F51; entries, F(5,65) = 1.03, P40.4] nor was there a group�day
interaction (both F51).
To determine the effect of maze rotation on the rats’ perfor-
mance, analyses were carried out using the last 3 days of standard
radial arm maze acquisition and the first 3 days of rotation. The
VTNx group was impaired overall, which is reflected by a main
effect of group using both errors [F(1,13) = 38.0, P50.001] and
correct entries [F(1,13) = 32.9, P50.001]. Overall performance on
the radial arm maze task was worse when maze rotation was
included, as measured by both total errors [F(1,13) = 12.6,
P50.005] and correct entries in first eight [F(1,13) = 26.3,
P50.001]. The lack of a clear group� condition interaction
using either errors [F(1,13) = 3.72, P = 0.076] or correct entries
(F51) indicates that the lesion group was not selectively disrupted
by the maze rotation, although there is a potential scaling effect.
Experiment 5—egocentricdiscrimination in the cross-mazeAnalyses of the total number of errors made by the animals before
reaching criterion of 30/36 revealed no difference between the
two groups (t51; Fig. 6A). Animals were subsequently tested
for 12 days on reversal learning where they were reinforced for
turning in the opposite direction to that rewarded during the
acquisition stage (Fig. 6B). There was no overall group difference
in the number of errors made (F51) and no group�day interac-
tion (F51), reflecting the equivalent improvement of both groups
across sessions.
Experiment 6—visually cued task inthe water mazeAnimals were tested for 4 days on a visually cued version of the
water maze task during which the platform position was indicated
by the presence of a black and white ‘stick’. There was no group
difference in swim paths to reach the platform [F(1,13) = 2.88,
P40.1; Fig. 7]. While there were significant main effects of trial
[F(3,39) = 10.9, P50.001] and day [F(3,39) = 17.2, P50.001],
there were no group� trial (F51) or group� day (F51)
Figure 5 Radial arm maze task (Experiment 4). (A) Mean
number of correct entries (�SE) in first eight arm choices.
First five blocks represent acquisition of the tasks and the final
two blocks include rotation of the maze; (B) mean number of
errors (�SE) during acquisition (first five blocks) and rotation
(final two blocks); (C) ratio of sequential choice responses (�SE)
during task acquisition.
Ventral tegmental nuclei and memory Brain 2009: 132; 2372–2384 | 2379
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
interactions, reflecting the similar patterns of performance across
both groups.
Experiment 7—delayed-matching-to-place in the water mazeTo determine whether the null effects in Experiments 5 and
6 might be a result of functional recovery, Experiment 2 was
repeated so that animals were tested for a final 4 days on
the ‘delayed-matching-to-place’ task (Fig. 7). The VTNx group’s
performance was similar to that originally found in Experiment 2
as they required significantly longer swim paths to reach hidden
platforms [F(1,13) = 11.02,P50.01]. There was also a main effect
of trial [F(3,39) = 10.78, P50.0001] reflecting reduced path
lengths across trials as animals learnt the position of the platform
within a session. There was also a significant group� trial interac-
tion [F(3,39) = 3.05, P = 0.04] as the groups only differed on Trial
2 [F(1,44) = 18.66, P50.001] but not Trial 3 [F(1,44) = 2.32,
P = 0.14] or Trial 4 (F51), whereas there was a borderline
difference at Trial 1 [F(1,44) = 4.46, P = 0.05].
DiscussionWhile current theories of diencephalic amnesia emphasize the con-
tribution of the hippocampal–fornix–mammillary body pathway
(Delay and Brion, 1969; Gaffan, 1992; Aggleton and Brown,
1999), the present study provides clear evidence that a quite
separate input to the mammillary bodies has a key role in support-
ing learning and memory. Selective, excitotoxic lesions of the
VTNg impaired the performance on spatial memory tasks in the
water maze, T-maze and radial arm maze that are all sensitive to
mammillary body and mammillothalamic tract lesions (Sziklas and
Petrides, 1993; Vann and Aggleton, 2003), as well as to lesions of
the anterior thalamic nuclei (e.g. Warburton et al., 1997) and
hippocampus (e.g. Cassel et al., 1998; Bannerman et al., 2002).
The impairments following VTNg lesions were not due to sensori-
motor disturbances, gross learning impairments or an inability to
adapt behaviour as the same rats were unimpaired on a visually
cued task in the water maze, and showed unimpaired acquisition
and reversal learning of an egocentric discrimination task. The
VTNg lesions also led to hyperactivity in a novel environment
and in the water maze, as signified by increased swimming
speeds. This hyperactivity is consistent with previous findings
that mammillary body lesions, mammillotegmental tract lesions,
mammillothalamic tract lesions and supramammillary nucleus
Figure 6 Egocentric discrimination in the cross-maze
(Experiment 5). (A) Mean errors (�SE) to reach criterion of
30/36 correct choices during task acquisition; (B) mean errors
(�SE) made during the 12 sessions of reversal learning.
Figure 7 Left-hand panel: mean path lengths (�SE) to find
visually cued platform in the water maze; right-hand panel:
mean path lengths (�SE) to find the hidden platform in a
delayed-matching-to-place task in the water maze.
2380 | Brain 2009: 132; 2372–2384 S. D. Vann
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
lesions can produce hyperactivity (Field et al., 1978; Pan and
McNaughton, 2002; Vann and Aggleton, unpublished results).
There is also a striking parallel with the hyperactivity so frequently
reported after hippocampal lesions (Gray and McNaughton,
1983). It is, however, most unlikely that this hyperactivity was
responsible for the spatial impairments following VTNg lesions as
very clear water maze deficits were observed using path length as
opposed to latency. Furthermore, this hyperactivity does not
explain the normal acquisition of the egocentric task in a cross-
maze when contrasted with the marked deficits on T-maze
alternation.
The principal goal of the study was to test whether the VTNg
is necessary for the normal performance of spatial memory tasks
that are sensitive to hippocampal and mammillary body damage.
A key feature was the selectivity of the lesions. Although there
was inevitable loss of those few median raphe nuclei cells imme-
diately adjacent to the VTNg, this damage was very restricted as
determined both by Nissl and NeuN staining. Consistent with this
conclusion was evidence that serotonin staining in all other parts
of the raphe nuclei appeared unaffected. Furthermore, the selec-
tive loss of serotonergic fibres/cells in the median raphe nucleus is
not sufficient to disrupt T-maze alternation (Asin and Fibiger,
1984). Likewise, the cholinergic outputs from the nearby latero-
dorsal tegmental nucleus to sites such as the anterior thalamic
nuclei and lateral mammillary nuclei were also shown to be
unaffected by the VTNg lesions. Of special note was the integrity
of the cholinergic projection from the laterodorsal tegmental nuclei
to the interpeducuncular nuclei as this pathway runs in the vicinity
of the lesion site. These findings help to confirm that the lesions
were axon-sparing and, hence, that the spatial memory
impairments were due to a selective loss of VTNg neurons.
A small number of previous lesion studies have unintentionally
included the VTNg, and while these earlier experiments failed to
make selective lesions, their findings are consistent with those
from the present study. For example, electrolytic lesions that
included the VTNg region impaired spatial memory, including
delayed alternation and radial arm maze tasks (Wirtshafter and
Asin, 1983; Asin and Fibiger, 1984), though damage to fibres of
passage and adjacent bundles limited any conclusion. Likewise,
while a study of excitotoxic median raphe nuclei lesions reported
a significant correlation between the extent of VTNg damage and
the number of errors made on a reinforced T-maze task (Asin and
Fibiger, 1984), the consistent presence of extensive raphe cell loss
made it impossible to determine the impact of selective damage to
VTNg. However, the previously reported correlation between
extent of VTNg damage and spatial memory performance (Asin
and Fibiger, 1984) is consistent with the data from those animals
in the present study, which were excluded due to appreciable
VTNg sparing. While only two animals fitted the partial lesion
criteria, making it difficult to draw any strong conclusions, their
performance levels typically fell between those of the controls and
of the complete VTNg lesion group. From both this observation,
and the correlation reported by Asin and Fibiger (1984), it is likely
that even partial damage to the VTNx is sufficient to produce
some degree of mnemonic impairment.
The rationale for the present study assumed that the principal
VTNg lesion effect on spatial learning would be via its interactions
with the medial mammillary nucleus. This reasoning stems from
the density of the reciprocal connections between these two sites,
the evidence of impaired delayed alternation following mammillo-
tegmental tract lesions (Field et al., 1978), and the impact of
mammillary body lesions on such tasks (e.g. Vann and Aggleton,
2003). The VTNg does, however, innervate a number of other
sites, with light projections to the lateral hypothalamus, preoptic
area, medial septum, oral and caudal parts of the reticular pontine
nucleus, median raphe nuclei, supramammillary nucleus and
ventral tegmental area (Petrovicky, 1973; Leichnetz et al., 1989;
Hayakawa et al., 1993). Even though their inputs from the VTNg
appear very light, and in some cases are disputed, a contribution
from the disconnection of these sites cannot be excluded. A
number of these sites are, for example, directly connected with
the hippocampus, e.g. the medial septum, the supramammillary
nucleus, the lateral hypothalamus, the median raphe and ventral
tegmental area. Of those sites within the medial diencephalon, the
supramammillary nucleus is of particular interest as it contains cells
that control the frequency of theta rhythm in the hippocampus,
and damage to this nucleus can induce hyperactivity (Pan and
McNaughton, 2002; 2004). Nevertheless, the finding that supra-
mammillary nucleus lesions can have no apparent effect on either
working or reference memory in a water maze (Pan and
McNaughton, 2002) indicates that the disconnection of this
nucleus is not sufficient to explain the present findings.
The clear implication is that the loss of the reciprocal connec-
tions between VTNg and the medial mammillary nucleus best
accounts for the learning deficits seen in the present study.
Consistent with this explanation is the finding that cytotoxic
lesions of the mammillary bodies produce deficits on the same
array of spatial memory tasks at what appears to be a similar
degree of severity (Vann and Aggleton, 2003). Furthermore,
selective lesion studies indicate that medial mammillary nucleus
damage is primarily responsible for these deficits as selective
lesions of the lateral mammillary nuclei only produce mild, tran-
sient deficits on the spatial memory tasks used in the current
study (Vann, 2005). At the same time, it remains possible that
the loss of other VTNg connections could contribute to the array
of deficits, though only the loss of mammillary body interactions
appears sufficient to account for the learning impairments. The
next step is to consider why these tegmental inputs might be
critical for normal learning.
The dorsal and ventral nucleus of Gudden have very different
anatomical connections and electrophysiological properties; while
the dorsal tegmental nucleus forms part of a ‘lateral’ system
involved in generating head-direction information (Bassett et al.,
2007), the VTNg forms part of a ‘medial’ theta system (Vann and
Aggleton, 2004). The VTNg is reciprocally connected with the
medial mammillary nucleus, and all VTNg cells are thought to
fire rhythmically and highly coherently with hippocampal
theta (Kocsis et al., 2001). Consistent with this finding is the
identification of theta-related cells in the medial mammillary
nucleus (Alonso and Llinas, 1992; Kocsis and Vertes, 1994;
Bland et al., 1995; Kirk et al., 1996; Kirk, 1998). The present
study focused on the VTNg for several reasons: (i) the medial
mammillary nucleus is always affected in the amnesic Korsakoff’s
syndrome (Kopelman, 1995; Victor et al., 1989); (ii) the medial
Ventral tegmental nuclei and memory Brain 2009: 132; 2372–2384 | 2381
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
mammillary body projections to the anterior thalamic nuclei are
necessary for normal hippocampal, prefrontal and retrosplenial
function (Vann and Albasser, 2009); (iii) there is a single report
of a patient with persistent amnesia attributed to pathology in the
VTNg and surrounding area (Goldberg et al., 1981); and (iv) there
is a growing evidence that theta plays an important role in opti-
mizing the conditions for information storage (e.g. Hasselmo
et al., 2002; Hasselmo, 2005).
Although the relative size of VTNg is smaller in primates, the
structure and connections of the mammillary bodies and Gudden’s
tegmental nuclei are remarkably similar across species (Petrovicky,
1973; Irle and Markowitsch, 1982; Hayakawa and Zyo, 1984;
Huang et al., 1992; R.C. Saunders, unpublished results). The pro-
jections from VTNg to medial mammillary bodies use GABA and
appear to form inhibitory synapses (Allen and Hopkins, 1989;
Wirthshafter and Stratford, 1993; Gonzalo-Ruiz et al., 1999),
which led to the proposal that the VTNg acts as an inhibitory
feedback loop that controls the transfer of information from the
hippocampus to the anterior thalamic nuclei (Wirtshafter and
Stratford, 1993). More recently, VTNg function has been linked
to its rhythmic firing properties; one proposal is that it moderates
the hippocampal-driven rhythmic firing in the medial mammillary
(Kocsis et al., 2001; Vertes et al., 2004). A more radical account is
that the VTNg is a generator of hippocampal theta (Bassant and
Poindessous-Jazat, 2001). However, while lesions of the supra-
mammillary nuclei, which may also involve the mammillary
bodies, can disrupt some aspects of hippocampal theta, they do
not eliminate it completely (Thinschmidt et al., 1995; Sharp and
Koester, 2008). As the normal VTNg effects would presumably be
moderated via these structures, it is unlikely that the VTNg is a
generator of hippocampal theta, though it may possibly modulate
hippocampal theta frequency.
The present study focused on those tasks in animals that are
sensitive to damage to the hippocampus, mammillary bodies and
anterior thalamic nuclei (Sziklas and Petrides, 1993; Warburton
et al., 1997; Bannerman et al., 2002; Vann and Aggleton,
2003). There appears to be a close relationship between brain
regions that result in spatial memory impairments when damaged
in animals and those that result in anterograde amnesia/episodic
memory impairments in humans (Aggleton and Pearce, 2001).
There is also a strong relationship between those structures (e.g.
hippocampus and retrosplenial cortex) that result in topographical
amnesia in humans and those that result in episodic memory
impairments (e.g. Maguire et al., 1996; Maguire, 2001). These
relationships suggest that the VTNg contributes to episodic
memory and new learning in humans as well as spatial memory.
From the single reported case of a patient with damage to the
VTNg (Goldberg et al., 1981), it is also possible that this structure
contributes to retrograde memory although this has not been
assessed in the present study.
In conclusion, Gudden’s tegmental nuclei were first described
over 100 years ago (Gudden, 1884), yet have remained poorly
understood. The intriguing report of a patient with persistent
amnesia attributed to pathology in the VTNg and surrounding
area (Goldberg et al., 1981) appears to have not been pursued,
presumably because pathology localization was based on
CT imaging and so may well have involved neighbouring sites
and tracts. By producing selective, axon-sparing lesions centred
in the VTNg, the present study not only reveals that this tegmen-
tal nucleus may be vital for an array of memory tasks, but also
suggests that its projections, e.g. to the mammillary bodies, are an
integral element of how medial diencephalic nuclei support
memory.
Supplementary materialSupplementary material is available at Brain online.
AcknowledgementsThanks to John Aggleton, Heather Phillips and Moira Davies.
FundingBiotechnology and Biological Sciences Research Council UK
(BBSRC) David Phillips Research fellowship (BB/B501955/1).
ReferencesAggleton JP, Brown MW. Episodic memory, amnesia, and the
hippocampal-anterior thalamic axis. Behav Brain Sci 1999; 22:
425–44; Discussion 444–89.
Aggleton JP, Pearce JM. Neural systems underlying episodic memory:
insights from animal research. Philos Trans R Soc Lond B Biol Sci
2001; 356: 1467–82.
Allen GV, Hopkins DA. Mamillary body in the rat: topography and
synaptology of projections from the subicular complex, prefrontal
cortex, and midbrain tegmentum. J Comp Neurol 1989; 286: 311–36.
Alonso A, Llinas RR. Electrophysiology of the mammillary complex
in vitro. 2. Medial mammillary neurons. J Neurophysiol 1992; 68:
1321–31.Asin KE, Fibiger HC. Spontaneous and delayed spatial alternation follow-
ing damage to specific neuronal elements within the nucleus medianus
raphe. Behav Brain Res 1984; 13: 241–50.
Bannerman DM, Deacon RM, Offen S, Friswell J, Grubb M, Rawlins JN.
Double dissociation of function within the hippocampus: spatial
memory and hyponeophagia. Behav Neurosci 2002; 116: 884–901.Bassant MH, Poindessous-Jazat F. Ventral tegmental nucleus of Gudden:
a pontine hippocampal theta generator? Hippocampus 2001; 11:
809–13.
Bassett JP, Taube JS. Neural correlates for angular head velocity in the rat
dorsal tegmental nucleus. J Neurosci 2001; 21: 5740–51.
Bassett JP, Tullman ML, Taube JS. Lesions of the tegmentomammillary
circuit in the head direction system disrupt the head direction signal in
the anterior thalamus. J Neurosci 2007; 27: 7564–77.
Bernstein HG, Krause S, Krell D, Dobrowolny H, Wolter M, Stauch R,
et al. Strongly reduced number of parvalbumin-immunoreactive
projection neurons in the mammillary bodies in schizophrenia: further
evidence for limbic neuropathology. Ann N Y Acad Sci 2007; 1096:
120–7.Bland BH, Konopacki J, Kirk IJ, Oddie SD, Dickson CT. Discharge
patterns of hippocampal theta-related cells in the caudal diencephalon
of the urethane-anesthetized rat. J Neurophysiol 1995; 74: 322–33.
Briess D, Cotter D, Doshi R, Everall I. Mamillary body abnormalities in
schizophrenia. Lancet 1998; 352: 789–90.
2382 | Brain 2009: 132; 2372–2384 S. D. Vann
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
Briggs TL, Kaelber WW. Efferent fiber connections of the dorsal and
deep tegmental nuclei of Gudden. An experimental study in the cat.
Brain Res 1971; 29: 17–29.
Burgess N, Maguire EA, O’Keefe J. The human hippocampus and spatial
and episodic memory. Neuron 2002; 35: 625–41.
Callen DJ, Black SE, Gao F, Caldwell CB, Szalai JP. Beyond the
hippocampus: MRI volumetry confirms widespread limbic atrophy in
AD. Neurology 2001; 57: 1669–74.
Carlesimo GA, Serra L, Fadda L, Cherubini A, Bozzali M, Caltagirone C.
Bilateral damage to the mammillo-thalamic tract impairs recollection
but not familiarity in the recognition process: a single case investiga-
tion. Neuropsychologia 2007; 45: 2467–79.
Cassel J-C, Cassel S, Galani R, Kelche C, Will B, Jarrard L. Fimbria-fornix
vs selective hippocampal lesions in rats: effects on locomotor activity
and spatial learning and memory. Neurobiol Learn Mem 1998; 69:
22–45.
Copenhaver BR, Rabin LA, Saykin AJ, Roth RM, Wishart HA,
Flashman LA, et al. The fornix and mammillary bodies in older
adults with Alzheimer’s disease, mild cognitive impairment, and
cognitive complaints: a volumetric MRI study. Psychiatry Res 2006;
147: 93–103.
Delay J, Brion S. Le syndrome de korsakoff. Paris: Mason; 1969.
Denby CE, Vann SD, Tsivilis D, Aggleton JP, Montaldi D, Roberts N,
et al. The frequency and extent of mammillary body atrophy
associated with surgical removal of a colloid cyst. AJNR Am J
Neuroradiol 2009; 30: 736–43.
Ennaceur A, Aggleton JP. The effects of neurotoxic lesions of the peri-
rhinal cortex combined to fornix transection on object recognition
memory in the rat. Behav Brain Res 1997; 88: 181–93.Field TD, Rosenstock J, King EC, Greene E. Behavioral role of the
mammillary efferent system. Brain Res Bull 1978; 3: 451–6.Gaffan D. The role of the hippocampus–fornix–mammillary system in
episodic memory. In: Squire LR and Butters N, editors.
Neuropsychology of memory. New York: The Guildford Press; 1992.
p. 336–46.
Goldberg E, Antin SP, Bilder RM Jr, Gerstman LJ, Hughes JE, Mattis S.
Retrograde amnesia: possible role of mesencephalic reticular activation
in long-term memory. Science 1981; 213: 1392–4.
Gonzalo-Ruiz A, Romero JC, Sanz JM, Morte L. Localization of amino
acids, neuropeptides and cholinergic neurotransmitter markers in iden-
tified projections from the mesencephalic tegmentum to the mammil-
lary nuclei of the rat. J Chem Neuroanat 1999; 16: 117–33.
Gray JA, McNaughton N. Comparison between the behavioural effects
of septal and hippocampal lesions: a review. Neurosci Biobehav Rev
1983; 7: 119–88.
Grossi D, Lopez OL, Martinez AJ. Mamillary bodies in Alzheimer’s
disease. Acta Neurol Scand 1989; 80: 41–5.
Gudden BV. Uber das Corpus mamillare und die sogenannten
Schenkel des Fornix. Allgemeine Zeitschrift fur Psychiatrie 1884; 41:
697–701.
Hasselmo ME. What is the function of hippocampal theta rhythm?
Linking behavioral data to phasic properties of field potential and
unit recording data. Hippocampus 2005; 15: 936–49.Hasselmo ME, Bodelon C, Wyble BP. A proposed function for
hippocampal theta rhythm: separate phases of encoding and
retrieval enhance reversal of prior learning. Neural Comput 2002;
14: 793–817.
Hayakawa T, Ito H, Zyo K. Neuroanatomical study of afferent projections
to the supramammillary nucleus of the rat. Anat Embryol 1993; 188:
139–48.
Hayakawa T, Zyo K. Comparative anatomical study of the tegmento-
mammillary projections in some mammals —a horseradish-peroxidase
study. Brain Res 1984; 300: 335–49.Huang XF, Tork I, Halliday GM, Paxinos G. The dorsal, posterodorsal,
and ventral tegmental nuclei: a cyto- and chemoarchitectonic study in
the human. J Comp Neurol 1992; 318: 117–37.
Irle E, Markowitsch HJ. Connections of the hippocampal formation,
mamillary bodies, anterior thalamus and cingulate cortex. A retrograde
study using horseradish peroxidase in the cat. Exp Brain Res 1982; 47:
79–94.
Jongen-Relo AL, Feldon J. Specific neuronal protein: a new tool for
histological evaluation of excitotoxic lesions. Physiol Behav 2002; 76:
449–56.
Kirk IJ. Frequency modulation of hippocampal theta by the supra-
mammillary nucleus, and other hypothalamo-hippocampal interactions:
mechanisms and functional implications. Neurosci Biobehav Rev 1998;
22: 291–302.
Kirk IJ, Oddie SD, Konopacki J, Bland BH. Evidence for differential
control of posterior hypothalamic, supramammillary, and medial mam-
millary theta-related cellular discharge by ascending and descending
pathways. J Neurosci 1996; 16: 5547–54.Kocsis B, Di Prisco GV, Vertes RP. Theta synchronization in the limbic
system: the role of Gudden’s tegmental nuclei. Eur J Neurosci 2001;
13: 381–8.
Kocsis B, Vertes RP. Characterization of neurons of the supramammillary
nucleus and mammillary body that discharge rhythmically with the
hippocampal theta rhythm in the rat. J Exp Psych Anim Behav Proc
1994; 14: 7040–52.
Kopelman MD. The Korsakoff syndrome. Br J Psychiatry 1995; 166:
154–73.
Kumar R, Birrer BVX, Macey PM, Woo MA, Gupta RK, Yan-Go FL, et al.
Reduced mammillary body volume in patients with obstructive sleep
apnea. Neurosci Lett 2008; 438: 330–4.
Kumar R, Woo MA, Birrer BVX, Macey PM, Fonarow GC, Hamilton MA,
et al. Mammillary bodies and fornix fibers are injured in heart failure.
Neurobiol Dis 2009; 33: 236–42.
Leichnetz GR, Carlton SM, Katayama Y, Gonzalo-Ruiz A, Holstege G,
DeSalles AA, et al. Afferent and efferent connections of the cholino-
ceptive medial pontine reticular formation (region of the ventral
tegmental nucleus) in the cat. Brain Res Bull 1989; 22: 665–88.Lorens SA, Kohler C, Guldberg HC. Lesions in Guddens tegmental nuclei
produce behavioral and 5-HT effects similar to those after raphe
lesions. Pharmacol Biochem Behav 1975; 3: 653–9.
Maguire EA. The retrosplenial contribution to human navigation: a
review of lesion and neuroimaging findings. Scand J Psychol 2001;
42: 225–38.
Maguire EA, Burke T, Phillips J, Staunton H. Topographical disorientation
following unilateral temporal lobe lesions in humans. Neuropsychologia
1996; 34: 993–1001.Olton DS, Samuelson RJ. Remembrance of places passed—spatial
memory in rats. J Exp Psychol Anim Behav Process 1976; 2: 97–116.Olton DS, Walker JA, Gage FH. Hippocampal connections and spatial
discrimination. Brain Res 1978; 139: 295–308.Pan WX, McNaughton N. The role of the medial supramammillary
nucleus in the control of hippocampal theta activity and behaviour
in rats. Eur J Neurosci 2002; 16: 1797–809.
Pan WX, McNaughton N. The supramammillary area: its organization,
functions and relationship to the hippocampus. Prog Neurobiol 2004;
74: 127–66.
Parker A, Gaffan D. Mamillary body lesions in monkeys impair object-in-
place memory: functional unity of the fornix–mamillary system. J Cogn
Neurosci 1997; 9: 512–21.Paxinos G, Watson C. The rat brain in stereotaxic coordinates.
San Diego: Academic; 1997.Petrovicky P. Note on the connections of Gudden’s tegmental nuclei. I.
Efferent ascending connections in the mammillary peduncle. Acta Anat
1973; 86: 165–90.
Poletti CE, Creswell G. Fornix system efferent projections in the squirrel
monkey: an experimental degeneration study. J Comp Neurol 1977;
175: 101–28.
Satoh K, Fibiger HC. Cholinergic neurons of the laterodorsal tegmental
nucleus: efferent and afferent connections. J Comp Neurol 1986; 253:
277–302.Sharp PE, Koester K. Lesions of the mammillary body region alter
hippocampal movement signals and theta frequency: implications for
path integration models. Hippocampus 2008; 18: 862–78.
Ventral tegmental nuclei and memory Brain 2009: 132; 2372–2384 | 2383
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022
Sharp PE, Tinkelman A, Cho JW. Angular velocity and head directionsignals recorded from the dorsal tegmental nucleus of Gudden in the
rat: implications for path integration in the head direction cell circuit.
Behav Neurosci 2001; 115: 571–88.
Sikes RW, Vogt BA. Afferent connections of anterior thalamus in rats:sources and association with muscarinic acetylcholine receptors.
J Comp Neurol 1987; 256: 538–51.
Sziklas V, Petrides M. Memory impairment following lesions to the mam-
millary region of the rat. Eur J Neurosci 1993; 5: 525–40.Thinschmidt JS, Kinney GG, Kocsis B. The supramammillary nucleus:
is it necessary for the mediation of hippocampal theta rhythm?
Neuroscience 1995; 67: 301–12.Tsivilis D, Vann SD, Denby C, Roberts N, Mayes AR, Montaldi D,
et al. A disproportionate role for the fornix and mammillary
bodies in recall versus recognition memory. Nat Neurosci 2008;
11: 834–42.Van der Werf YD, Witter MP, Uylings HB, Jolles J. Neuropsychology of
infarctions in the thalamus: a review. Neuropsychologia 2000; 38:
613–27.
Vann SD. Transient spatial deficit associated with bilateral lesions of thelateral mammillary nuclei. Eur J Neurosci 2005; 21: 820–4.
Vann SD, Aggleton JP. Evidence of a spatial encoding deficit in rats with
lesions of the mammillary bodies or mammillothalamic tract. J Neurosci
2003; 23: 3506–14.
Vann SD, Aggleton JP. The mammillary bodies: two memory systems inone? Nat Rev Neurosci 2004; 5: 35–44.
Vann SD, Albasser MM. Hippocampal, retrosplenial, and prefron-
tal hypoactivity in a model of diencephalic amnesia: evidence
towards an interdependent subcortical-cortical memory network.Hippocampus 2009.
Vann SD, Tsivilis D, Denby CE, Quamme JR, Yonelinas AP, Aggleton JP,
et al. Impaired recollection but spared familiarity in patients with
extended hippocampal system damage revealed by 3 convergentmethods. Proc Natl Acad Sci USA 2009; 106: 5442–7.
Vertes RP, Hoover WB, Viana Di Prisco G. Theta rhythm of the
hippocampus: subcortical control and functional significance. BehavCogn Neurosci Rev 2004; 3: 173–200.
Victor M, Adams RD, Collins GH. The Wernicke–Korsakoff syndrome
and related neurological disorders due to alcoholism and malnutrition.
Philadelphia: F. A. Davis Company; 1989.Warburton EC, Baird AL, Aggleton JP. Assessing the magnitude of the
allocentric spatial deficit associated with complete loss of the anterior
thalamic nuclei in rats. Behav Brain Res 1997; 87: 223–32.
Wirtshafter D, Asin KE. Impaired radial maze performance in rats withelectrolytic median raphe lesions. Exp Neurol 1983; 79: 412–21.
Wirtshafter D, Stratford TR. Evidence for GABAergic projections from the
tegmental nuclei of Gudden to the mammillary body in the rat. Brain
Res 1993; 630: 188–94.
2384 | Brain 2009: 132; 2372–2384 S. D. Vann
Dow
nloaded from https://academ
ic.oup.com/brain/article/132/9/2372/357919 by guest on 18 February 2022