Funded Grants

Learning, storing and retrieval: An improved approach to dissecting hippocampal function

Memory has been described as "the glue that binds our mental life, the scaffolding that holds our personal history and that makes it possible to grow and change throughout life. When memory is lost ...we lose the ability to re-create our past, and as a result we lose our connection with ourselves and with others." (Squire and Kandel, 1999, p. ix). It is for these reasons that disturbances of memory have such profound effects on people's lives. Be it Alzheimer or Huntington disease, or any other of the injuries or diseases that can affect memory, memory impairment is very disabling and disruptive for both patients and their families.

Questions about the nature of memory have been posed since antiquity. However, only at the end of the 19th century did inquiry move beyond clinical observation and philosophical discussion to become an experimental science. And only in the latter half of the 20th century, with the emergence of neuroscience, did it become possible to ask biological questions about how memory is acquired, organized and stored.

In the 1950s, critical insights regarding the neuroscience of memory came from a patient who became known as H.M. At the age of 27 H.M. underwent surgery to remove the inner surface of the temporal lobes, including the hippocampus, in an attempt relieve severe epilepsy. Though this procedure did help his epilepsy, it also left him with a devastating memory impairment. The most salient feature of this impairment can be described as profound forgetfulness.

Since the effects on memory of medial temporal lobe resection were described in patient H.M., a considerable body of work has been directed toward understanding how the hippocampus and related structures in the medial temporal lobe contribute to memory function. This work, which has focused especially on the hippocampus, has been carried out mainly in amnesic patients, monkeys, and rodents. One major fording is that the role of the hippocampus in memory is selective, in that it is involved in only one kind of memory (variously termed declarative, explicit, or relational).

A phenomenon that has been especially important to this tradition of work is temporally-graded retrograde amnesia. Temporally-graded retrograde amnesia refers to the finding that (declarative) information learned long before the onset of memory impairment is often spared relative to information learned more recently. A good deal of progress has been made in understanding this phenomenon through neuropsychological testing of patients with brain damage. While studies of patients have improved the understanding of temporally-graded retrograde amnesia, it is difficult to study retrograde amnesia in patient populations for two reasons. First, studies with patients necessarily rely on retrospective methods and imperfect tests. Specifically, with retrospective tests, it is difficult to compare performance across past time periods because one cannot guarantee that the information from each time period has been learned to the same extent. Second, the most informative studies are those where detailed neuropathological and neuropsychological information have been obtained. However, the precise locus and extent of damage in amnesic patients is often difficult to determine eves with brain imaging techniques. Moreover, the damage seldom honors anatomical boundaries, so that structures in addition to the hippocampus are often damaged as well. These difficulties can be overcome by using experimental animals, where the locus and extent of damage can be precisely controlled and where the extent of learning can be carefully controlled by the use of prospective tests.

Previous work in animals has relied on conventional (permanent) lesions made at different times before or after learning. Yet, the conventional lesion approach cannot identify the specific and unique times that a structure is important for function. It has been pointed out before that reversible lesion techniques have many advantages over the conventional lesion method. Indeed, reversible lesion techniques are now beginning to be developed and applied to the problem of memory using both pharmacological and genetic methods. I am proposing to study memory using a water-soluble version of CNQX (an AMPA/kainite antagonist), which can be introduced into the rat hippocampus by means of surgically implanted osmotic minipumps. The infusion can occur continuously for up to one week (or longer if a second pump is implanted).

The fundamental questions I am asking are: exactly when relative to learning, storage, and retrieval is the hippocampus important for memory? Is it needed at the time of learning and for some limited time after learning? Would transient disruption of hippocampal function at any point during this interval impair retention? If the disruption is effective, does extending the period of disruption cause more severe disruption? If hippocampal neural activity is needed during the retention interval, is there also a requirement for NMDA-dependent synaptic plasticity? Is there a fundamental difference between spatial and nonspatial memory? The studies I have proposed here, when completed, should provide important insights into when and how hippocampal function supports the establishment of long-term memory. Clarifying these issues is critical to understanding how memory is organized and stored in the brain. Improved understanding of the biology of memory should help in the development of preventions and treatments for conditions that affect memory.

Though the use of animals to study memory is a powerful tool in and of itself, I believe it is critical to use what we know from studies of memory-impaired patients to guide our experiments with animals and to interpret the results obtained from studies of animals within the broader context of what we know from studies of humans. As an example, work with humans has suggested that hippocampal damage impairs declarative memory, that is, memory for facts and events. This impairment includes an impaired ability to remember the spatial layouts of previously encountered environments, but this spatial impairment is not a salient feature of the phenotype in humans. However, in the rat, the study of hippocampal function grew out of the study of spatial memory. For example, the discovery of place cells in the hippocampus (cells that predictably and reliably change their firing rates when a rat enters a specific area in an environment) was tremendously influential. Further work demonstrated that hippocampal damage in the rat profoundly impairs spatial memory. These results led to the idea that the hippocampus is responsible for forming and storing spatial maps of the environment, and many proposals about hippocampal function emphasize spatial cognition. Perhaps one reason why ideas about spatial memory in the rat have been so prominent is that spatial memory is readily trained and measured in the rat. Yet, there are now several different memory tests designed for the rat that measure nonspatial memory, and I have implemented a number of these tasks in my laboratory. Moreover, because I believe that tests of both spatial and nonspatial memory are critical for accurately assessing how the hippocampus functions, in the present proposal I have outlined studies of both spatial and nonspatial memory.

In summary, I have proposed studies of memory and hippocampus in the rat as the first step in a new program of research that will study systematically the effects of reversible lesions. The work will be guided by what is known from work with humans and will attempt to address issues that are part of a long, and to some extent a polarized, tradition of work. In the sense that I am studying rodents but within an intellectual context that was developed from human cognition, I believe that the proposed work will be eminently useful for bridging brain, mind, and behavior.