Memory loss is one of the most devastating symptoms of Alzheimer’s disease (AD). But are all of these memories gone forever, or are some of them just irretrievable? And how can we find out? In today’s PLoS Biology, Richard Morris and colleagues at Edinburgh University, Scotland, describe a new type of water maze test that can distinguish memory storage from memory retrieval problems in rats. The technique may be useful to study memory loss in animal models of disease.

Hanging over the field of amnesia research has been the question of whether memory problems are due to truly lost memories or simply poor recollection. To eliminate the latter possibility, research must show that memories are really retrieved and not a result of relearning. Many experiments that test animal memory have trouble distinguishing between these two possibilities. For example, does that rat really recall a mild foot shock or has its memory actually been wiped and it has relearned to avoid it? By adapting the classic Morris water maze, first author Livia de Hoz and colleagues have designed a test that distinguishes retrieval from relearning. The test may open the flood gates for investigation into how the hippocampus and cortex cooperate to store and retrieve memories.

The key to the test is a simple hint. De Hoz and colleagues trained animals to find and climb to safety on a hidden platform in the water maze. After induction of amnesia—a partial or complete hippocampal lesion—they retested the animals before and after briefly showing them the platform again. The idea is to remind the animals of the existence of the platform but not to teach them of its location.

The results were definitive. After partial hippocampal lesions, rats searched aimlessly around the tank until they got the hint, then they began to search more locally, where the platform was originally. The key observation was that even though the hint can come in the opposite quadrant of the tank, the rats still searched the original locale (see image). The results suggest that their memories can indeed be retrieved and the rats are not just relearning the platform position. The results also indicate that the lesions caused retrieval rather than storage problems. In contrast, animals with complete lesions wandered aimlessly around the tank even after getting the hint, which is what would be expected in a case where memories just cannot be retrieved.

Image credit: © De Hoz et al.

Rats trained to find the hidden platform (red dot) in the water maze cannot remember where it is after partial hippocampal lesion (left animal track), but after getting a hint--the platform is briefly raised either in its original location or in the opposite quadrant of the tank (yellow dot)--the animals begin to search more locally. The fact that they search in the original quadrant, irrespective of where the hint is, suggests that they are truly remembering the location and not just relearning.

The authors further eliminated the possibility that the animals were relearning rather than remembering by training them in a totally new environment. While the sham-operated control animals could learn the position of a new platform, those with hippocampal lesions could not.

While the test may or may not be directly applicable to animal models of AD, as the authors write, “the findings open a potential avenue of research into the neural dynamics of memory reactivation and retrieval.” In particular, these findings offer a new means to explore the relationship of the hippocampal-cortex axis to memory. For example, does the hippocampus mediate links between different cortical regions before memories are consolidated, as has been suggested (see review by Squire and Alvarez, 1995), or is it itself the site of long-term storage (see discussion by Moscovitch and Nadel, 1998)? Answers to these questions may help us understand the role the hippocampus may play in Alzheimer’s-related memory losses (see, for example, Eustache et al., 2004).—Tom Fagan


  1. A fundamental question about memory is: What happens to a memory after it is initially stored? Studies of patients with amnesia have found that a strengthening process known as consolidation must be occurring on some of our memories. While some memories fade with time, others appear to become more firmly ingrained. These “consolidated” memories survive damage to the memory systems of the brain, producing the phenomenon of “temporally graded retrograde amnesia,” which is often observed when conversing with an amnesic patient. The patient will often have no difficulty recalling memories from his or her youth, but will have great difficulty remembering more recent events, even those that occurred before the patient became amnesic. For this patient, the consolidation process had not yet run its course on those recent memories and they were lost during the event that produced the amnesia. The more distant memories have had years to become consolidated and they are intact, even after damage to the memory system.

    The mechanism by which the brain consolidates our memories is not known. One hypothesis is that this occurs naturally during sleep. A separate suggestion has been that conscious reminding, or thinking about the memory, is the key event in consolidation. These two ideas are not incompatible; perhaps during sleep, some sort of unconscious reminding occurs to help consolidate memories. But it is difficult to assess the effect of reminding, particularly in animal models of memory dysfunction that are extremely helpful in understanding the neurobiology of memory. The recent report by de Hoz, Martin and Morris (2004) provides an elegant experimental approach to examining the effect of reminding in an animal model of amnesia.

    Rats have good memory for spatial location and quickly learn to perform a task known as a “water maze.” In this task, rats swim in a tub of water to a platform that allows them to “escape” and perch above the water (rats are good swimmers, but find it annoying). The rat’s memory is tested by clouding the water (e.g., with powdered milk) and they show that they remember where the escape platform is by quickly swimming to it, even when it is not visible. Rats whose memory system is impaired perform very poorly at this task; they cannot remember where the platform is in between tests.

    The critical finding of de Hoz et al. (2004) is that rats with a mild memory impairment can be reminded of their past experience and this improves their performance. Rats were first shown the platform at position A. Memory-impaired rats were not very good at remembering to search at A when tested later. A separate group was reminded of their experience with another swim in the maze, this time with the platform at a new place, B. Surprisingly, the effect of this practice trial was to cause the rats to search in position A in a later test. The memory impairment caused the rats to forget the time they escaped at position B, but reminded and strengthened the memory of the time they escaped at place A. This effect did not occur with severe memory impairments, indicating that some memory of searching at A has to survive for the reminding to work.

    This result provides strong evidence that reminding or thinking about a past event strengthens the memory of that event and makes it easier to recall later, even when the brain’s memory system is partially damaged. This suggests that frequent reminding about recent events might help a patient with a memory disorder retain some memories. After onset of a syndrome that impairs memory, such as Alzheimer disease, frequent reminding about events that occurred before onset could help reduce the effect of retrograde amnesia and loss of those memories.

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Retrograde amnesia and memory consolidation: a neurobiological perspective. Curr Opin Neurobiol. 1995 Apr;5(2):169-77. PubMed.
  2. . Consolidation and the hippocampal complex revisited: in defense of the multiple-trace model. Curr Opin Neurobiol. 1998 Apr;8(2):297-300. PubMed.
  3. . 'In the course of time': a PET study of the cerebral substrates of autobiographical amnesia in Alzheimer's disease. Brain. 2004 Jul;127(Pt 7):1549-60. PubMed.

Further Reading

Primary Papers

  1. . Forgetting, reminding, and remembering: the retrieval of lost spatial memory. PLoS Biol. 2004 Aug;2(8):E225. PubMed.