Research report
Humans with hippocampus damage display severe spatial memory impairments in a virtual Morris water task

https://doi.org/10.1016/S0166-4328(01)00399-0Get rights and content

Abstract

For nonhumans, it has been shown that the hippocampus (HPC) is critical for spatial memory. We tested patients with unilateral HPC resections on a virtual analogue of a classic spatial task to assess HPC functioning in nonhumans: the Morris water task. We found that when humans are required to use spatial cues to navigate to a hidden escape platform in a pool, patients with HPC resections display severe impairments in spatial navigation relative to age-matched controls and age-matched patients who have had extra-HPC resections. This effect occurred for every patient tested and was evident regardless of side of surgery. Hence, it is apparent across species and irrespective of which hemisphere is damaged that the human HPC is critical for spatial/relational memory.

Introduction

The ability to find objects, recall previous locations and navigate throughout the world is dependent upon spatial learning and memory. There is a myriad of converging evidence that suggests that for nonhumans, the hippocampal formation appears to be the critical anatomical structure necessary for these spatial abilities [3], [27], [28], [33], [35] (but see Refs. [7], [39]). With humans, the data also suggest that the hippocampus is involved in spatial memory. However, it seems that only damage to the right hippocampal formation results in spatial memory impairments [4], [6], [15], [16], [31], [34].

In examining the tasks used to assess spatial memory across species, there are a number of factors that make comparisons difficult. Specifically, egocentric memory is often examined in human tasks, such as maze learning [6] and object location [34], while allocentric memory is examined in nonhuman tasks, such as the Morris water task [23] and 8-arm radial mazes [29]. In addition, even within species the definition of ‘spatial’ varies greatly across tasks, such that face recognition [22], memory for abstract designs [16] and egocentric maze learning [6] are all grouped together as tests of spatial memory.

Attempts to reconcile some of these differences have been made in recent reports by Maguire et al. [19] using virtual environments during PET to examine hippocampal contributions during navigation through a virtual town. Their results indicate that both hippocampi are active during virtual navigation, but only the right hippocampal activation predicts navigation accuracy. These reports take an important step toward bridging these paradigmatic gaps. However, it is unclear from these studies how patients with hippocampus damage would perform in these virtual tasks. Specifically, the tasks consist of navigating from landmark to landmark and this may not require an allocentric strategy at all. In fact, rats with hippocampus damage are unimpaired in using landmarks to locate a goal [24]. Additionally, the procedure in these studies consists of having a subject learn the virtual environment outside the PET scanner and hence, important information about hippocampal contributions during learning is lost. Lastly, whereas this paradigm has the benefit of mimicking real life navigation, it is unable to be adapted properly for use with nonhumans.

An alternative method of using virtual environments to examine hippocampus contributions to spatial learning and memory is to take advantage of the rich paradigms that have been extensively developed for testing rodents. Specifically, one of the most reliable tests used to assess spatial memory in rodents is the Morris water task [23]. In this task, rats are required to use the relationship among distal cues to navigate to a hidden escape platform in a pool of milky water. Normal rats learn to find the platform readily and efficiently. However, rats with hippocampus damage show severe impairments in being able to use a spatial strategy to navigate to this escape platform [24], [35]. This task is advantageous in that no single landmark allows for accurate navigation, since the rat must start at different locations within the pool and some landmarks are initially hidden from view, depending on the starting point. Rather, it is utilizing the relationships of the various landmarks in the room that allows for accurate navigation.

To examine the effects of hippocampus damage on navigation in humans, we developed a virtual version of the Morris water task (vMWT) to assess spatial memory in humans with unilateral hippocampal resections [2]. This task is parametrically and directly modeled, in number of trials, relative room dimensions, probe trials, swimming speed and dependent variables, after the Morris water task used in our laboratory to reveal impairments in rats with hippocampus damage [25]. This is advantageous in that we are able to assess not only allocentric memory, but also the learning process. Additionally, because all the parameters are similar, we are better able to make comparisons across species to help form a more unitary picture of the role of the hippocampus in learning and memory. Moreover, this allows for a solitary task to be used to assess hippocampal functioning across species.

Section snippets

Participants

Our test group consisted of ten patients who had undergone unilateral hippocampal removals (five left-sided, five right-sided) as a component of temporal lobe surgery to treat medically intractable epilepsy. Four of the left hemisphere and two of the right hemisphere patients received a selective amygdalohippocampectomy. The other four patients received an anterior temporal lobectomy. All patients with HPC damage were recruited from the Montreal Neurological Institute and are referred to as the

Results

In examining the latency to find the hidden platform during the 40 training trials, using a repeated measures MANOVA with Trial as the repeating factor, there exists a significant group effect, F(2, 20)=4.21, P=0.03. Specifically, patients with HPC damage were significantly slower to find the hidden platform than either the tumor group, F(1, 14)=6.03, P=0.028 or the unoperated group, F(1, 15)=8.06, P=0.012. However, the unoperated group and the tumor group did not differ from each other, F(1,

Discussion

The current results indicate that, in humans, spatial learning/memory impairments in a virtual Morris water task occur following unilateral HPC removal and are independent of the side removed. Moreover, because the tumor and hippocampal group do not differ on the Rey–Osterrieth complex figure copy and recall task, this result cannot be attributed to a global learning/memory impairment.

Interestingly, this spatial learning/memory impairment does not seem to be lateralized. The patients with left

Acknowledgements

We wish to thank Brenda Milner and Joelle Crane for their invaluable contributions to this project. This work was partially funded by a Minority Biomedical Research Support and Student Research Allocation Committee award to RSA.

References (39)

  • D.G. Mumby et al.

    Retrograde amnesia and selective damage to the hippocampal formation: memory for places and object discriminations

    Behav. Brain Res.

    (1999)
  • J.A. O'Keefe et al.

    The hippocampus as a spatial map: preliminary evidence from unit activity in freely moving rats

    Brain Res.

    (1971)
  • K.P. Ossenkopp et al.

    Spatial learning in an enclosed eight-arm maze in rats with sodium arsinilate-induced labryinthectomies

    Behav. Neural. Biol.

    (1993)
  • S. Pigott et al.

    Memory for different aspects of complex visual scenes after unilateral-temporal or frontal-lobe resection

    Neuropsychologia

    (1993)
  • M.L. Smith et al.

    The role of the right hippocampus in the recall of spatial memory

    Neuropsychologia

    (1981)
  • R.J. Sutherland et al.

    Spatial mapping: definitive disruption by hippocampal and frontal cortex damage in the rat

    Neurosci. Letts.

    (1982)
  • R.J. Sutherland et al.

    Hippocampus, amygdala, and memory deficits in rats

    Behav. Brain Res.

    (1990)
  • G.K. Aguirre et al.

    The parahippocampus subserves topographical learning in man

    Cereb. Cort.

    (1996)
  • V. Bingman

    Unimpaired acquisition of spatial reference memory, but impaired homing performance in hippocampal ablated pigeons

    Behav. Brain Res.

    (1988)
  • Cited by (330)

    View all citing articles on Scopus
    View full text