Brain, Vol. 123, No. 4, 816-827,
April 2000
© 2000 Oxford University Press
Mediodorsal thalamic function in scene memory in rhesus monkeys
Department of Experimental Psychology, Oxford University, Oxford, UK
Correspondence to:
D. Gaffan, Department of Experimental Psychology, South Parks Road, Oxford OX1 3UD, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Three monkeys were trained preoperatively in a scene memory task which is analogous, in some ways, to human episodic memory. The same animals were also trained in object–reward association memory. Following bilateral ablations of almost the entire magnocellular division of the mediodorsal thalamic nucleus, the animals were impaired both in scene memory and in object–reward association memory. These results, combined with recent results in object recognition memory from monkeys with mediodorsal thalamic lesions, show that the impairment produced by this lesion is more general, affecting a broader range of memory tasks, than the impairment which is produced in monkeys by lesions restricted to the hippocampus–fornix–mamillary system. It is also more severe than the effect of lesions limited to the medial part of the magnocellular division of the mediodorsal thalamic nucleus. These findings extend the evidence that the magnocellular division of the mediodorsal thalamic nucleus has an important and general role in memory, and they are consistent with the proposal that lesions of the magnocellular division of that nucleus have their effect by disrupting the function of prefrontal cortex.
amnesia; prefrontal; macaque; episodic; Korsakoff
S1–S3 = subjects 1–3
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The role of the magnocellular division of the mediodorsal thalamic nucleus in human memory has remained uncertain despite many attempts to correlate damage with memory loss in amnesic patients with diencephalic lesions, beginning with the large study by Victor and colleagues, who found mediodorsal thalamic lesions in all their Korsakoff amnesic patients (Victor et al., 1971
, 1989). Dense amnesia can be observed in alcoholic Korsakoff syndrome in the absence of any lesion in the mediodorsal thalamic nucleus (Mair et al., 1979; Mayes et al., 1988). However, as Mayes and colleagues pointed out, this does not rule out the possibility that mediodorsal thalamic lesions are sufficient to cause amnesia (Mayes et al., 1988). Because of this uncertainty, several investigators have turned to the study of memory in macaque monkeys (rhesus or cynomolgus) with experimentally produced mediodorsal thalamic lesions. These can be made without significant damage to other structures such as the mamillothalamic tract and the internal medullary lamina, which are frequently involved in clinical cases of thalamic amnesia (Cramon et al., 1985).
Macaques with lesions in the magnocellular division of the mediodorsal thalamic nucleus have shown significant memory impairment in visual object recognition memory, assessed by the delayed non-matching-to-sample or delayed matching-to-sample tasks either in manually operated apparatus (Aggleton and Mishkin, 1983; Zola-Morgan and Squire, 1985) or in computer-controlled apparatus (Parker et al., 1997). However, none of the impairments seen in these studies was as severe as that which is seen in some Korsakoff patients in visual object recognition memory (Warrington and Taylor, 1973). The study by Parker and colleagues compared the effects of mediodorsal thalamic lesions with the effects of ablations of the entorhinal and perirhinal cortical areas in the anterior basal temporal lobe, which are known to be important for normal object recognition memory (Meunier et al., 1990, 1993; Gaffan and Murray, 1992; Eacott et al., 1994; Buckley et al., 1997; Buckley and Gaffan, 2000). The temporal cortical removals had a much more severe effect than mediodorsal thalamic ablations (Parker et al., 1997).
However, the ablations of the magnocellular division of the mediodorsal nucleus in the study by Parker and colleagues were subtotal, being intended to remove only the medial part of the magnocellular division (Parker et al., 1997). Neurons in the macaque entorhinal and perirhinal cortex send axons to the mediodorsal nucleus, and this projection has been heavily emphasized in a recent influential account of the role of the mediodorsal nucleus in memory (Aggleton and Brown, 1999). These axons, however, arrive exclusively in the medial part of the magnocellular division of the nucleus (Russchen et al., 1987), and therefore the very mild effect of ablations restricted to this medial part of the division (Parker et al., 1997) argues that the direct projection from entorhinal and perirhinal cortex to the mediodorsal nucleus, which is only a light projection (Aggleton et al., 1986; Russchen et al., 1987), is not of central importance in object recognition memory. The function of the whole magnocellular division may be much more important for normal memory than is the light direct projection it receives from the temporal cortex. Parker and Gaffan proposed that the main function of the magnocellular mediodorsal nucleus is in its reciprocal interaction with prefrontal cortex (Parker and Gaffan, 1998a). The prefrontal cortex of the frontal lobe is not only the dominant recipient of efferent information from the magnocellular mediodorsal nucleus, but also the dominant source of afferent information to it. These afferent projections from the prefrontal cortex to the magnocellular mediodorsal nucleus are far heavier than the afferent projections from the temporal cortex to the magnocellular mediodorsal nucleus (Russchen et al., 1987, their Fig. 5).
The prefrontal cortex is now known to be of very general importance in memory in macaques, not just in specialized tests of conditional learning or working memory. Extensive bilaterally symmetrical frontal ablations, of a kind rarely if ever seen clinically, produced a profound impairment in the simplest form of object–reward association memory (Parker and Gaffan, 1998b) as well as in object recognition memory (Kowalska et al., 1991; Meunier et al., 1997). Unlike the projections to the mediodorsal nucleus from perirhinal and entorhinal cortex, prefrontal reciprocal projections are not restricted to the medial part of the magnocellular division but involve the whole of the magnocellular division, which in turn projects to the whole of the prefrontal cortex (Russchen et al., 1987). Parker and Gaffan therefore argued that a removal of the entire magnocellular division would produce a widespread disruption of frontal cortex function and thus have a functional effect qualitatively similar to that of frontal cortex removal, though one would expect it to be milder since the thalamic lesion leaves the cortex itself intact (Parker and Gaffan, 1998a). To test this idea, and to investigate further the route of interaction between perirhinal cortex and frontal cortex in object recognition memory, they examined the effect in this task of crossed unilateral removals of perirhinal cortex in one hemisphere and frontal cortex in the other hemisphere, and they compared this with the effect of crossed unilateral lesions of perirhinal cortex in one hemisphere and the entire magnocellular division of the mediodorsal nucleus in the other hemisphere. The outcome was that the animals with unilateral frontal cortex removal were significantly more impaired than those with unilateral mediodorsal thalamic lesions when both were crossed with unilateral perirhinal cortex ablations. Both of these surgical procedures, however, produced an impairment in object recognition memory much more severe than that observed in the study by Parker and colleagues, in which thalamic lesions were limited to the projection area of the entorhinal and perirhinal cortex within the magnocellular mediodorsal nucleus (Parker et al., 1997). Since a unilateral lesion alone, either temporal frontal or thalamic, has no detectable effect on recognition memory in the monkey so long as the contralateral hemisphere is intact (Parker and Gaffan, 1998a), the implication of these results is that the entire magnocellular division is more important in recognition memory than is the medial part of it.
These results from object recognition memory suggest that the magnocellular division of the mediodorsal nucleus may, like the prefrontal cortex, have an essential and very general role in normal memory. To test this idea further, in the present experiment we removed the entire magnocellular division bilaterally after training monkeys preoperatively in an `object-in-place' scene memory task (Gaffan, 1994). This scene memory task requires the monkey to remember an event (either the delivery of food reward or its absence) associated with a particular object, but the object is embedded in a particular background scene. The task is psychologically similar in some ways to memory for the daily events of human life, since human memory for discrete events in daily life is similarly embedded in specific contexts, `episodic memory' (Tulving, 1983). The monkey scene memory task is also similar to human memory in that it elicits powerful memory performance from normal monkeys; the animals are able to learn 20 new scenes in one or two trials. The effects of lesions in the mediodorsal thalamus have not previously been investigated in the scene memory task. We compared the effects of removing the entire magnocellular division of the mediodorsal thalamic nucleus with the effects seen in an earlier experiment which used the same task, tested in the same apparatus and in the same way as in the present study. The earlier experiment investigated the effects of combined transection of the anterior temporal stem, amygdala and fornix (Gaffan et al., 2000). This combination of temporal lobe transections produced dense amnesia in the monkey, analogous to human dense amnesia after medial temporal lobe lesions; the rationale for our experiment with this combination of transections was derived partly from earlier monkey experiments and partly from recent evidence that the anterior temporal stem is interrupted, together with a lesion of the amygdala and the hippocampus, in the densely amnesic patient H.M. (see panels H, I and J of Fig. 2 in the report by Corkin et al., 1997). Therefore, the present comparison of this effect with the effect of complete magnocellular mediodorsal thalamic lesions tested the possibility that these thalamic lesions might also produce dense amnesia.
|
We also compared the present results from scene memory with previous results in the same task after mamillary body lesions (Parker and Gaffan, 1997b
), since the relative importance of mamillary body lesions and mediodorsal nucleus lesions in human diencephalic amnesia has been the subject of long debate (Delay and Brion, 1969; Victor et al., 1971, 1989; Mair et al., 1979; Mayes et al., 1988; Aggleton and Brown, 1999).
The monkeys were also trained preoperatively in an object–reward association memory task. This is an important task for assessing amnesia in the monkey because human amnesic patients are impaired in this task (Oscar-Berman and Zola-Morgan, 1980; E. A. Gaffan et al., 1990, 1991); however, several experiments attempting to produce amnesia in the monkey have discovered no impairment in this task (reviewed by Gaffan, 1998). Previous results after mediodorsal thalamic lesions from tasks of this kind, tested in the same computer-controlled apparatus as was used in the present experiment (Gaffan and Murray, 1990; Gaffan and Watkins, 1991), led us to expect that this task would be impaired by mediodorsal thalamic lesions. The version of the object–reward association task we employed here had the advantage, however, that the magnitude of the impairment could be directly compared with that which followed combined section of the anterior temporal stem, amygdala and fornix, since the same version of the task had been used in the earlier study (Gaffan et al., 2000). In addition, in the object–reward association task we assessed postoperative retention of memories which had been acquired preoperatively.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Subjects
These were three experimentally naïve female rhesus monkeys, S1, S2 and S3. At the time of surgery they weighed, on average, 3.2 kg. Experiments were carried out under licence issued by the Home Office, UK.
Surgery
On the completion of preoperative training all three subjects were operated upon to remove the entire magnocellular division of the mediodorsal thalamic nucleus bilaterally. The operations were carried out under barbiturate anaesthesia with the aid of an operating microscope. Following sagittal incision of the skin and galea, a D-shaped bone flap was turned over the right hemisphere and the midline. The dura mater over the posterior part of the right hemisphere was incised and retracted to the midline. The right hemisphere was retracted with a brain spoon and the splenium of the corpus callosum was cut in the midline with a glass aspirator. The tela choroidea, which lies between the splenium and the thalamus, was cut in the midline with the application of cautery, using a metal aspirator that was insulated to the tip. The posterior commissure, the third ventricle posterior to the thalamus and the most posterior 5 mm of the midline thalamus were exposed. Next, the most posterior 5 mm of the massa intermedia was cut in the midline with the glass aspirator. Tissue adjacent to the cut surface of the massa intermedia was then removed bilaterally by aspiration with the metal aspirator to an estimated depth of 2 mm. The anteroposterior extent of the ablation was 5 mm.
Histology
At the completion of all behavioural testing, the monkeys were deeply anaesthetized and transcardially perfused with physiological saline followed by 10% formalin. The brains were extracted and cut on a freezing microtome in 50 µm sections in the coronal plane. The sections were stained with cresyl violet, mounted on slides and coverslipped.
To estimate the extent of the removals we plotted the missing tissue in the stained sections on to drawings of the normal thalamus, using the drawings and the normal photomicrographs published by Gaffan and Murray (Gaffan and Murray, 1990). The results are shown in Fig. 1 and illustrative stained sections from each animal are shown in Fig. 2
. The removals were almost complete except for the anteroventral parts of the magnocellular division in S1 and S2, and the most posterior part of the division in all three animals. The fornix was intact in S2 and S3, while S1 showed unilateral fornix damage.
|
Apparatus
The computer-controlled test apparatus was identical to that which has previously been described in detail by Gaffan (Gaffan, 1994
). Briefly, it consisted of a large touch-sensitive colour monitor on which the visual stimuli were displayed, and dispensers for food rewards.
Stimulus material
Scenes
These were identical to those previously described in detail by Gaffan (Gaffan, 1994, Experiment 5). Briefly, a set of unique scenes was generated for each test session. They were artificially constructed by an algorithm based on a random number generator. Each scene had a complex background, unique to that scene, on which two foreground objects, selected from a large population of such objects, were superimposed; one of the two foreground objects was the correct one for the monkey to touch and the other was incorrect.
Objects
Objects were constructed by the same algorithm as described earlier (Parker and Gaffan, 1998a). Each object was a digram, on average 35 mm high and 45 mm wide, made by abutting two coloured typographic characters side-by-side. Objects were displayed in pairs with their left–right position on the screen varying randomly from trial to trial.
Procedure
1. Scenes
In each daily session a list of 20 new scenes that the monkey had not seen before was presented for 160 trials in total. Each scene was presented once in each of eight successive blocks of 20 trials within the session. The order of presentation of the 20 scenes was the same in each of the eight blocks. On each trial, the display remained on the screen until the animal touched the screen. If the touch was to the positive foreground object, the positive object flashed on and off in the background scene, then the screen blanked and simultaneously one reward pellet (190 mg) was dispensed into the food hopper; an inter-trial interval of 10 s then began, ending with the presentation of the next trial in the session. If the touch was to the negative foreground object, the screen blanked and an inter-trial interval of 20 s began. In blocks 2–8 in each session, choice of the negative object was followed by the next scheduled trial in the session, without any correction trial for the scene in which an error had been committed; a correction trial was only presented after any error in the first block of 20 trials in each session, that is, the first run through that list of scenes. A correction trial consisted of the re-presentation of the same scene with only one foreground object in it, the positive object; when the positive object was touched, the flashing and food followed as in the main trials of the task. For the final scene in the session the food reward was a very large food reward and for this trial alone in blocks 2–8 a correction trial was presented if an error was made, thus ensuring that the large food reward would always be obtained. Within each day, the monkey's performance began at chance in the first 20 trials, when the monkey had no information as to which was the correct object to choose, and improved over the subsequent 140 trials (seven blocks of one trial each for the 20 scenes). Thus, the proficiency of within-day learning can be expressed by the average percentage error in trial blocks 2–8. Touching the screen during the inter-trial interval reset the interval. Touching any part of a displayed scene that was not one of the foreground objects in that scene resulted in screen blanking, an inter-trial interval and the presentation again of the same scene for the same trial. These `inaccurate' touches were recorded separately from errors (that is, choices of the negative object) and were made on less than 1% of trials by all animals.
2. Object–reward association learning
On any trial two objects were displayed on the left and right sides of the screen, the objects being assigned to these two positions at random on each trial. Any such pair of objects constituted a reward association learning problem, one object having been designated the correct (rewarded) one in the pair and the other the wrong one. The monkey chose one object by touching it and both objects then disappeared. If the chosen object was the correct one a food reward (190 mg) was dispensed. During the inter-trial interval of 10 s any touch to the screen reset the interval. Sessions continued until 100 correct choices had been made and the last correct choice was rewarded with a very large food reward. Objects were learned in sets of 10 problems (object pairs) concurrently. In successive runs of 10 trials each problem was presented once. The order of problems within a 10-trial run was random. The animal continued daily sessions with a set of 10 until a criterion was met of 90% correct choices in the whole session, in sessions after the first session with that set, or of 90% correct choices after the first 10 trials, in the first session with that set. The day after reaching criterion on one set, the animal began training with a new set. Animals learned 10 sets to criterion preoperatively. Postoperatively the animals learned five further sets to criterion in the same way.
3. Overtraining and retention test of a set of 100 object pairs
The 10 sets of 10 object pairs which were learned to criterion preoperatively were subsequently combined into a single set of 100 pairs of objects. These were presented for concurrent learning of object–reward associations in the same way as before and with the same reward associations (rewarded or unrewarded) as in initial training. The first 100 trials of each session presented each pair once in random order. The pairs were then presented again in a new random order. The session ended when 100 rewards had been earned; the last reward was a very large food reward. Preoperatively, training continued in daily sessions until the animal attained an overall mean of 95% correct choices in 20 successive daily sessions. Postoperatively, the same 100 object pairs were presented once each as a retention test.
Preoperative training sequence
Each animal was first given introductory training in easier versions of the scenes task, as described in detail elsewhere (Gaffan, 1994). They were then trained for 45 sessions in the scene learning task as described above (procedure 1). Next they learned 10 sets of 10 object–reward association problems to criterion (procedure 2). Finally they learned all 100 object–reward association pairs to criterion (procedure 3).
Postoperative training sequence
The animals were first tested for retention of the 100 overtrained object–reward association problems (procedure 3). Next they were trained to criterion in five successive sets of 10 object–reward association problems (procedure 2). Finally they were tested for 12 sessions in scene learning (procedure 1).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Preoperative
Learning in the last 10 preoperative sessions of the scene memory task is shown in Fig. 3
and Table 1. It can be seen that learning of new lists of 20 scenes was proficient at the end of preoperative training. Learning in the last five of the 10 preoperatively learned sets of 10 object–reward association problems is presented in Fig. 4 and Table 2, showing that learning at the end of preoperative training in this task was also proficient. In overtraining of 100 object–reward association problems preoperatively, the animals took an average of 31 sessions, including the criterial sessions, to meet the criterion of an overall mean of 95% correct choices in 20 successive daily sessions.
|
|
|
|
Postoperative: retention of object–reward associations
The animals were tested for retention of 100 overtrained object–reward association problems by presenting each problem for one trial. In these 100 trials S1 made 86 correct choices, S2 96 and S3 83. This performance is significantly greater than chance for each animal, using a binomial test (P < 0.05). The average performance was 88.3% correct.
Postoperative: new learning of object–reward associations
This is shown in Fig. 4. The animals made more errors postoperatively than preoperatively [t(2) = 5.127, P < 0.05, one-tailed]. However, as the figure also shows, they were not so severely impaired postoperatively as the animals in a previous study using exactly the same task (Gaffan et al., 2000) after combined section of temporal stem, amygdala and fornix [t(5) = 3.036, P < 0.05, two-tailed]. Results from individual sets of 10 problems are presented in Table 2, where it can be seen that the postoperative impairment was stable across sets of problems.
Postoperative: scene memory
This is shown in Fig. 3. The animals made more errors postoperatively than preoperatively [t(2) = 6.252, P < 0.05, one-tailed]. However, as the figure also shows, they were not so severely impaired postoperatively as the animals in a previous study using exactly the same task (Gaffan et al., 2000) after combined section of temporal stem, amygdala and fornix [t(4) = 7.468, P < 0.05, two-tailed]. The severity of impairment was similar to that seen in an earlier study using exactly the same task with mamillary body lesions, as shown in the figure. Results from the newly reported individual animals are presented numerically in Table 1.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Lesions of almost the entire magnocellular division of the mediodorsal thalamic nucleus produced a substantial impairment of postoperative new learning, both in scene memory (Fig. 3
) and in object–reward association memory (Fig. 4). These results support the idea that the magnocellular division of the mediodorsal nucleus has a general role in memory, not limited to object recognition memory. However, retention of preoperatively overtrained object–reward associations was at a high level, 88% correct, showing that the monkeys were able to retrieve some information from preoperative learning. In principle, the effects observed could be due to the removal of midline structures lying between the mediodorsal nuclei of the two sides of the thalamus, since these midline structures were always included in the lesion, as detailed in Fig. 1. In practice, however, this is unlikely, since Gaffan and Murray included in their experiment a group (the three monkeys in their group MD/A after the first operation) with lesions that removed these midline structures but left the mediodorsal nucleus intact on one side, and these animals were unchanged in learning ability (Gaffan and Murray, 1990). Nevertheless, the present results cannot rule out the possibility that the involvement of the midline thalamus, in combination with the bilateral removal of the magnocellular division of the mediodorsal nucleus, contributed to the severity of the impairment in the present animals. To assess the severity of the presently observed anterograde amnesic effects, and to relate mediodorsal thalamic function in memory to that of other structures, these effects need to be compared quantitatively and qualitatively with the effects of other lesions which produce amnesia in monkeys.
Any bilateral lesion in a connected series of structures including the hippocampus, fornix, mamillary bodies and anterior thalamus (Delay and Brion, 1969) produces a significant amnesia in monkeys as revealed by the scene memory task (Gaffan, 1994; Parker and Gaffan, 1997a, b; Murray et al., 1998). Furthermore, the scene memory task is not an overtly spatial task and in this and some other ways it is similar to human memory for events in life (Gaffan, 1992, 1994). These results support the idea of the functional unity of the Delay–Brion system in episodic memory (Delay and Brion, 1969). However, the effects of these lesions do not amount to an analogue of human dense amnesia. Object recognition memory, for example, can be performed normally by monkeys with selective lesions in the hippocampus (Murray and Mishkin, 1998) or fornix (Gaffan et al., 1984a; Parker and Gaffan, 1998a). Many further examples of unimpaired learning and memory tasks in monkeys with lesions in the Delay–Brion system have been reviewed elsewhere (Gaffan, 1998; Aggleton and Brown, 1999). Similarly, patients with complete fornix lesions show significant impairments in some memory tasks on psychometric testing (McMackin et al., 1995; Aggleton et al., 2000), and in the present scene memory task they show a level of impairment similar to that seen in monkeys with experimental fornix transection (Aggleton et al., 2000); however, these patients can perform at a normal level (McMackin et al., 1995) in an object recognition memory test which was specifically designed as a test of amnesia, the Warrington Recognition Memory Test (Warrington, 1984). The fact that, in the series reported by Victor and colleagues, at least four patients with subsequently verified mamillary body lesions were diagnosed on clinical examination as not suffering from amnesia (Victor et al., 1989, p. 113) also supports the idea that lesions in the Delay–Brion system do not produce dense amnesia, even though we assume that those four patients would have revealed some memory impairment if tested psychometrically. This view of the role in memory of the Delay–Brion system, as subserving human episodic memory and its analogues in animals, but as providing less than the whole explanation of dense amnesia, is also strongly supported by experiments in rats (Aggleton and Brown, 1999).
The impairment in scene memory after lesions of the magnocellular division of the mediodorsal thalamic nucleus (Fig. 3) was as severe as the impairments in this task which have been seen previously after lesions in the Delay–Brion system. The comparison with mamillary body lesions, presented in Fig. 3, is especially relevant to human diencephalic amnesia, but the impairments in the same task after fornix transection or lesions of the anterior thalamic nuclei were also very similar in magnitude (Parker and Gaffan, 1997a, b; the latter study also included a group of monkeys with a control operation, showing that the postoperative scene learning rate of these control monkeys was identical to their preoperative rate). Furthermore, results from other memory tasks show that the effects of lesions in the magnocellular division of the mediodorsal thalamic nucleus are more general than the effects of lesions in the Delay–Brion system in monkeys. In object recognition memory in manually operated apparatus, the positive effects of mediodorsal thalamic lesions reported by Aggleton and Mishkin (Aggleton and Mishkin, 1983) can be directly compared with the negative effects of selective hippocampal lesions (combined with amygdala lesions) reported by Murray and Mishkin (Murray and Mishkin, 1998) or the very mild effects of fornix transection reported by Bachevalier and colleagues (Bachevalier et al., 1985), all three experiments having been conducted in the same apparatus and with the same procedure. Similarly, in object recognition memory in the computer-controlled apparatus, disconnection of the perirhinal cortex from the entire magnocellular division of the mediodorsal nucleus produced a severe impairment in which the monkeys could reliably remember only one object at a time, like the patients of Warrington and Taylor (Warrington and Taylor, 1973), while monkeys with transection of the fornix (combined with aspiration amygdalectomy) recovered normal performance postoperatively in the same task (Parker and Gaffan, 1998a). In object–reward association memory, fornix transection or selective hippocampal lesions produced only a transitory effect (Gaffan and Harrison, 1984; Murray et al., 1998) and after a moderate amount of postoperative training fornix-transected monkeys showed normal memory for object–reward associations (Gaffan et al., 1984b). (These three experiments were all conducted in manually operated apparatus, but an unpublished study by Gaffan and Harrison confirmed that fornix transection does not impair object–reward association learning in the automated apparatus.) This contrasts with the substantial and stable impairment in object–reward association memory caused by lesions in the magnocellular division of the mediodorsal thalamic nucleus, both in the present study (Table 2 and Fig. 4) and in the experiments by Gaffan and Murray (Gaffan and Murray, 1990) and Gaffan and Watkins (Gaffan and Watkins, 1991). Thus, the effects in object recognition memory and in object–reward association memory allow the firm conclusion that, in monkeys, complete lesions of the magnocellular division of the mediodorsal thalamic nucleus produce a more general memory impairment than lesions in the Delay–Brion system. Combined with the present evidence that they also produce an equal impairment in scene memory with that which is produced by a lesion in the Delay–Brion system, the inference is that a lesion of the magnocellular division of the mediodorsal nucleus must contribute substantially to the memory impairments of those patients who have a complete lesion in this division combined with a lesion in the Delay–Brion system. This is not an infrequent combination of lesions, since Victor and colleagues (Victor, 1989, p. 90, their Tables 6–9) categorized 42% (16 out of 38) of their amnesic patients as having a severe lesion in the mediodorsal nucleus combined with a mamillary lesion, and also noted (p. 77) that the most severe lesions they observed in the mediodorsal thalamic nucleus always involved the entire magnocellular division.
On the other hand, these effects in the monkey of lesions of the entire magnocellular division of the mediodorsal thalamic nucleus are not so severe as to resemble human dense amnesia. The animals with these lesions were able to learn object–reward associations and scenes, and they did so at a rate only two or three times more slowly than they had done preoperatively (Figs 3 and 4). Furthermore, the animals were much less impaired in those two tasks than animals with combined section of fornix, amygdala and anterior temporal stem, a model of human dense amnesia after medial temporal lesions (Gaffan et al., 2000). The comparisons presented in Fig. 3 are therefore consistent with the idea that dense amnesia after diencephalic lesions, comparable in severity with amnesia after medial temporal lesions, results only from a combined lesion both in the mediodorsal nucleus and in some diencephalic part of the Delay–Brion system, either the mamillary bodies, the mamillothalamic tract or the anterior thalamic nuclei, and not from either a mediodorsal thalamic lesion alone or a Delay–Brion system lesion alone (Aggleton and Mishkin, 1983; Cramon et al., 1985; Graff-Radford et al., 1990; Aggleton and Brown, 1999). According to this idea, a combined lesion both in the mediodorsal nucleus and in the Delay–Brion system should produce a dense general amnesia in the monkey.
This combined-lesion hypothesis faces the insuperable objection, however, that dense amnesia in patients can be observed after diencephalic lesions which spare the mediodorsal nucleus (Mair et al., 1979; Mayes et al., 1988). Victor and colleagues (Victor et al., 1989) have claimed that the most extreme medial segment of the magnocellular division of the mediodorsal nucleus may have been slightly involved in the lesion in the cases reported by Mair and colleagues (Mair et al., 1979). Whatever the validity of that claim, either in relation to the patients reported by Mair and colleagues or in relation to those reported by Mayes et al. (Mayes et al., 1988), it appears unable to explain these patients' dense amnesia, since lesions restricted to the medial half of the mediodorsal nucleus had only a very mild effect in the monkey (Parker et al., 1997).
Furthermore, there is an inference from psychometric data that there was not unrevealed pathology in the mediodorsal nucleus in the patients of Mair and colleagues (Mair et al., 1979) and Mayes and colleagues (Mayes et al., 1988). In the introduction we reviewed anatomical and behavioural data in favour of the proposal that the main function of the magnocellular mediodorsal nucleus is its reciprocal interaction with prefrontal cortex, and that a lesion of the nucleus has its effect by disrupting frontal cortex function (Parker and Gaffan, 1998a). This idea is further supported by psychometric data reported by Hodges and McCarthy. They studied a patient who became densely amnesic after a thalamic infarction which destroyed the mamillothalamic tract and the magnocellular mediodorsal nucleus bilaterally. In addition to severe memory impairment, the patient was also severely impaired in four tests sensitive to frontal cortex damage (the Wisconsin Card-Sorting Test, Verbal Fluency Test, Trail-Making Test and cognitive estimates), confirming a disruption of frontal cortex function in this patient, even though the frontal cortex was not directly damaged (Hodges and McCarthy, 1993). However, in the report by Mayes and colleagues the densely amnesic patient J.W. was normal on tests sensitive to frontal cortex damage (Mayes et al., 1987; their other patient, B.C., was mildly impaired on these tests but had frontal cortex atrophy in addition to the diencephalic lesion). Mair and colleagues did not report data from their patients on tests sensitive to frontal cortex damage, but it is nevertheless perfectly clear that these patients did not have this kind of impairment, since Mair and colleagues emphasize that their patients were chosen for detailed study specifically on the basis that their amnesia was pure and not associated with any other cognitive deficits (Mair et al., 1979, p. 778). The inference is, therefore, that these patients and J.W. did not have unrevealed pathology in the mediodorsal nucleus, since this would have induced an impairment on tests sensitive to frontal cortex damage, as the known pathology did in the patient of Hodges and McCarthy (Hodges and McCarthy, 1993).
Many Korsakoff patients, however, do show impairments in tests that are sensitive to frontal cortex damage (Kopelman, 1991) and it is reasonable to suppose, in light of the report by Victor and colleagues (Victor et al., 1989), that many, perhaps all, of the patients who do so have lesions in the magnocellular division of the mediodorsal nucleus. It is possible therefore that dense amnesia in many Korsakoff patients, including the patients reported by Victor and colleagues (Victor et al., 1989), did indeed result from combined lesions in the mediodorsal nucleus and the mamillary bodies, as the combined-lesion hypothesis proposes (Aggleton and Mishkin, 1983). However, dense and pure amnesia in Korsakoff syndrome, in the absence of signs of disrupted frontal cortex function and with full neuropathological investigation, has only been documented in patients with the mediodorsal nucleus intact (Mair et al., 1979; Mayes et al., 1988). This latter kind of Korsakoff patient requires a different explanation from that which the combined-lesion hypothesis offers. There could be individual variation among patients in their susceptibility to memory impairment after discrete lesions in the Delay–Brion system, with a minority of patients showing dense amnesia after such a lesion. If this is the case then monkey experiments can shed little light on the phenomenon, since it has not been observed in any monkey. Alternatively, there could be some other additional disorder, not a mediodorsal thalamic lesion, in these patients—for example, a disorder of the cholinergic and adrenergic ascending systems, as suggested by Mayes and colleagues (Mayes et al., 1988).
In summary, the present study has given further evidence that a lesion of the magnocellular mediodorsal thalamic nucleus in the monkey produces a broad range of substantial memory impairments, consistent with our proposal that it disrupts the function of frontal cortex. It has also added to the evidence that the effects of this lesion in the monkey are not so severe as to resemble human dense amnesia. Furthermore, together with the other recent studies we have discussed, it has suggested a modified version of the combined-lesion hypothesis (Aggleton and Mishkin, 1983) for diencephalic amnesia. We propose that one subset of patients with dense diencephalic amnesia have a lesion in the Delay–Brion system combined with a lesion in the magnocellular mediodorsal nucleus, and that amnesia in these patients is mixed with other cognitive effects, similar to those that are produced by damage to prefrontal cortex, while a different subset of patients with dense diencephalic amnesia have a lesion in the Delay–Brion system and not in the magnocellular mediodorsal nucleus, and amnesia in these patients is pure.
|
Acknowledgments |
|---|
This research was supported by the UK Medical Research Council.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Aggleton JP, Brown MW. Episodic memory, amnesia, and the hippocampal-anterior thalamic axis. Behav Brain Sci 1999; 22: 425–89.[Web of Science][Medline]
Aggleton JP, Mishkin M. Memory impairments following restricted medial thalamic lesions in monkeys. Exp Brain Res 1983; 52: 199–209.[Web of Science][Medline]
Aggleton JP, Saunders RC. The relationships between temporal lobe and diencephalic structures implicated in anterograde amnesia. [Review]. Memory 1997; 5: 49–71.[Web of Science][Medline]
Aggleton JP, Desimone R, Mishkin M. The origin, course and termination of the hippocampothalamic projections in the macaque. J Comp Neurol 1986; 243: 409–21.[Web of Science][Medline]
Aggleton JP, McMackin D, Carpenter K, Hornak J, Kapur N, Wiles CM, et al. Differential cognitive effects of colloid cysts in the third ventricle that spare or compromise the fornix. Brain. In press 2000.
Bachevalier J, Saunders RC, Mishkin M. Visual recognition in monkeys: effects of transection of fornix. Exp Brain Res 1985; 57: 547–53.[Web of Science][Medline]
Buckley MJ, Gaffan D. The hippocampus, temporal cortex and memory in the monkey. In: Bolhuis JJ, editor. Brain mechanisms of perception, learning and memory: essays for Gabriel Horn. Oxford: Oxford University Press. In press 2000.
Buckley MJ, Gaffan D, Murray EA. Functional double dissociation between two inferior temporal cortical areas: perirhinal cortex versus middle temporal gyrus. J Neurophysiol 1997; 77: 587–98.
Corkin S, Amaral DG, Gonzalez RG, Johnson KA, Hyman BT. H.M.'s medial temporal lobe lesion: findings from magnetic resonance imaging. J Neurosci 1997; 17: 3964–79.
Cramon DY von, Hebel N, Schuri U. A contribution to the anatomical basis of thalamic amnesia. Brain 1985; 108: 993–1008.
Delay J, Brion S. Le syndrome de Korsakoff. Paris: Masson; 1969.
Eacott MJ, Gaffan D, Murray EA. Preserved recognition memory for small sets, and impaired stimulus identification for large sets, following rhinal cortex ablations in monkeys. Eur J Neurosci 1994; 6: 1466–78.[Web of Science][Medline]
Gaffan D. Amnesia for complex naturalistic scenes and for objects following fornix transection in the rhesus monkey. Eur J Neurosci 1992; 4: 381–8.[Web of Science][Medline]
Gaffan D. Scene-specific memory for objects: a model of episodic memory impairment in monkeys with fornix transection. J Cogn Neurosci 1994; 6: 305–20.[Web of Science]
Gaffan D. Idiothetic input into object-place configuration as the contribution to memory of the monkey and human hippocampus: a review. [Review]. Exp Brain Res 1998; 123: 201–9.[Web of Science][Medline]
Gaffan D, Harrison S. Reversal learning by fornix-transected monkeys. Q J Exp Psychol 1984; 36B: 223–34.[Web of Science]
Gaffan D, Murray EA. Amygdalar interaction with the mediodorsal nucleus of the thalamus and the ventromedial prefrontal cortex in stimulus-reward associative learning in the monkey. J Neurosci 1990; 10: 3479–93.[Abstract]
Gaffan D, Murray EA. Monkeys (Macaca fascicularis) with rhinal cortex ablations succeed in object discrimination learning despite 24-hr intertrial intervals and fail at matching to sample despite double sample presentations. Behav Neurosci 1992; 106: 30–8.[Web of Science][Medline]
Gaffan D, Watkins S. Mediodorsal thalamic lesions impair long-term visual associative memory in macaques. Eur J Neurosci 1991; 3: 615–20.[Web of Science][Medline]
Gaffan D, Gaffan EA, Harrison S. Effects of fornix transection on spontaneous and trained non-matching by monkeys. Q J Exp Psychol 1984a; 36B: 285–303.[Web of Science]
Gaffan D, Saunders RC, Gaffan EA, Harrison S, Shields C, Owen MJ. Effects of fornix transection upon associative memory in monkeys: role of the hippocampus in learned action. Q J Exp Psychol 1984b; 36B: 173–221.
Gaffan D, Parker A, Easton A. Dense amnesia in the monkey after transection of fornix, amygdala and anterior temporal stem. Neuropsychologia 2000. In press 2000.
Gaffan EA, Aggleton JP, Gaffan D, Shaw C. Concurrent and sequential pattern discrimination learning by patients with Korsakoff amnesia. Cortex 1990; 26: 381–97.[Web of Science][Medline]
Gaffan EA, Gaffan D, Hodges JR. Amnesia following damage to the left fornix and to other sites: a comparative study. Brain 1991; 114: 1297–313.
Graff-Radford NR, Tramel N, Van Hoesen GW, Brandt JP. Diencephalic amnesia. Brain 1990; 113: 1–25.
Hodges JR, McCarthy RA. Autobiographical amnesia resulting from bilateral paramedian thalamic infarction: a case study in cognitive neurobiology. Brain 1993; 116: 921–40.
Kopelman MD. Frontal dysfunction and memory deficits in the alcoholic Korsakoff syndrome and Alzheimer-type dementia. Brain 1991; 114: 117–37.
Kowalska DM, Bachevalier J, Mishkin M. The role of the inferior prefrontal convexity in performance of delayed nonmatching-to-sample. Neuropsychologia 1991; 29: 583–600.[Web of Science][Medline]
Mair WG, Warrington EK, Weiskrantz L. Memory disorder in Korsakoff's psychosis: a neuropathological and neuropsychological investigation of two cases. Brain 1979; 102: 749–83.
Mayes AR, Meudell PR, Mann D, Pickering A. Location of lesions in Korsakoff's syndrome: neuropsychological and neuropathological data on two patients. Cortex 1988; 24: 367–88.[Web of Science][Medline]
McMackin D, Cockburn J, Anslow P, Gaffan D. Correlation of fornix damage with memory impairment in six cases of colloid cyst removal. Acta Neurochir (Wien) 1995; 135: 12–8.[Medline]
Meunier M, Murray EA, Bachevalier J, Mishkin M. Effects of perirhinal cortical lesions on visual recognition memory in rhesus monkeys [abstract]. Soc Neurosci Abstr 1990; 16: 616.
Meunier M, Bachevalier J, Mishkin M, Murray EA. Effects on visual recognition of combined and separate ablations of the entorhinal and perirhinal cortex in rhesus monkeys. J Neurosci 1993; 13: 5418–32.[Abstract]
Meunier M, Bachevalier J, Mishkin M. Effects of orbital frontal and anterior cingulate lesions on object and spatial memory in rhesus monkeys. Neuropsychologia 1997; 35: 999–1015.[Web of Science][Medline]
Murray EA, Mishkin M. Object recognition and location memory in monkeys with excitotoxic lesions of the amygdala and hippocampus. J Neurosci 1998; 18: 6568–82.
Murray EA, Baxter MG, Gaffan D. Monkeys with rhinal cortex damage or neurotoxic hippocampal lesions are impaired on spatial scene learning and object reversals. Behav Neurosci 1998; 112: 1291–303.[Web of Science][Medline]
Oscar-Berman M, Zola-Morgan SM. Comparative neuropsychology and Korsakoff's syndrome. II – Two-choice visual discrimination learning. Neuropsychologia 1980; 18: 513–25.[Web of Science][Medline]
Parker A, Gaffan D. The effect of anterior thalamic and cingulate cortex lesions on object-in-place memory in monkeys. Neuropsychologia 1997a; 35: 1093–102.[Web of Science][Medline]
Parker A, Gaffan D. Mamillary body lesions in monkeys impair object-in-place memory: functional unity of the fornix-mamillary system. J Cogn Neurosci 1997b; 9: 512–21.[Web of Science]
Parker A, Gaffan D. Interaction of frontal and perirhinal cortices in visual object recognition memory in monkeys. Eur J Neurosci 1998a; 10: 3044–57.[Web of Science][Medline]
Parker A, Gaffan D. Memory after frontal/temporal disconnection in monkeys: conditional and non-conditional tasks, unilateral and bilateral frontal lesions. Neuropsychologia 1998b; 36: 259–71.[Web of Science][Medline]
Parker A, Eacott MJ, Gaffan D. The recognition memory deficit caused by mediodorsal thalamic lesion in non-human primates: a comparison with rhinal cortex lesion. Eur J Neurosci 1997; 9: 2423–31.[Web of Science][Medline]
Russchen FT, Amaral DG, Price JL. The afferent input to the magnocellular division of the mediodorsal thalamic nucleus in the monkey, Macaca fascicularis. J Comp Neurol 1987; 256: 175–210.[Web of Science][Medline]
Tulving E. Elements of episodic memory. Oxford: Clarendon; 1983.
Victor M, Adams RD, Collins GH. The Wernicke-Korsakoff syndrome. Philadelphia: F. A. Davis; 1971.
Victor M, Adams RD, Collins GH. The Wernicke-Korsakoff syndrome. 2nd ed. Philadelphia: F. A. Davis; 1989.
Warrington EK. Recognition Memory Test. Windsor (UK): NFER-Nelson; 1984.
Warrington EK, Taylor AM. Immediate memory for faces: long- or short-term memory? Q J Exp Psychol 1973; 25: 316–22.[Web of Science][Medline]
Zola-Morgan S, Squire LR. Amnesia in monkeys after lesions of the mediodorsal nucleus of the thalamus. Ann Neurol 1985; 17: 558–64.[Web of Science][Medline]
Received June 15, 1999.
Revised October 6, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. S. Mitchell and D. Gaffan The Magnocellular Mediodorsal Thalamus is Necessary for Memory Acquisition, But Not Retrieval J. Neurosci., January 2, 2008; 28(1): 258 - 263. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Mitchell, M. G. Baxter, and D. Gaffan Dissociable Performance on Scene Learning and Strategy Implementation after Lesions to Magnocellular Mediodorsal Thalamic Nucleus J. Neurosci., October 31, 2007; 27(44): 11888 - 11895. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Mitchell, P. G. F. Browning, and M. G. Baxter Neurotoxic Lesions of the Medial Mediodorsal Nucleus of the Thalamus Disrupt Reinforcer Devaluation Effects in Rhesus Monkeys J. Neurosci., October 17, 2007; 27(42): 11289 - 11295. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. G. Baxter, D. Gaffan, D. A. Kyriazis, and A. S. Mitchell Orbital Prefrontal Cortex Is Required for Object-in-Place Scene Memory But Not Performance of a Strategy Implementation Task J. Neurosci., October 17, 2007; 27(42): 11327 - 11333. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Watanabe and S. Funahashi Neuronal Activity Throughout the Primate Mediodorsal Nucleus of the Thalamus During Oculomotor Delayed-Responses. I. Cue-, Delay-, and Response-Period Activity J Neurophysiol, September 1, 2004; 92(3): 1738 - 1755. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Watanabe and S. Funahashi Neuronal Activity Throughout the Primate Mediodorsal Nucleus of the Thalamus During Oculomotor Delayed-Responses. II. Activity Encoding Visual Versus Motor Signal J Neurophysiol, September 1, 2004; 92(3): 1756 - 1769. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K Tanji, K Suzuki, T Fujii, S Higano, and A Yamadori A case of frontal network amnesia J. Neurol. Neurosurg. Psychiatry, January 1, 2003; 74(1): 106 - 109. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Mitchell, J. C. Dalrymple-Alford, and M. A. Christie Spatial Working Memory and the Brainstem Cholinergic Innervation to the Anterior Thalamus J. Neurosci., March 1, 2002; 22(5): 1922 - 1928. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||