MIT Reports to the President 1995-96


The Center for Learning and Memory was established in May 1994 as an interdepartmental research center between the Department of Brain and Cognitive Sciences and Department of Biology. The Center's primary research interest is to study the mechanisms underlying learning and memory using multifaceted approaches. Susumu Tonegawa was appointed as the first Director of the Center in May 1994. Matthew A. Wilson joined as an Assistant Professor on July 1, 1994. William G. Quinn who has been a faculty member in Department of Brain and Cognitive Sciences since July 1, 1994 joined the Center on April 1, 1995.



Tonegawa's laboratory has continued to utilize the gene knockout technology and transgenic technology to study (1) molecular and cellular mechanisms underlying synaptic plasticity in the hippocampus and the cerebellum, and (2) the roles of the various types of synaptic plasticity in declarative or procedural learning and memory. In addition, Tonegawa's laboratory has developed experimental technologies that allow a regionally and/or temporally restricted gene knockout.

The molecular mechanisms underlying the long term potentiation (LTP) at one of the major hippocampal synapses, Schaffer colateral-CA1 synapses, was investigated by applying electrophysiological techniques to the hippocampal slices derived from a mouse deficient in aCaMKII gene. Tonegawa's laboratory had previously shown that LTP induction is severely impaired in these slices. Tonegawa's laboratory now found that various inhibitors of protein phosphatases "rescue" LTP in the aCaMKII mutant slices, indicating that LTP induction is subtly regulated by phosphorylation or dephosphorylation of key proteins.

Dr. Tonegawa's laboratory has also been investigating the cellular and molecular mechanisms that operate in cerebellum and underlie motor learning and motor coordination. For this purpose Dr. Tonegawa's laboratory has produced and analyzed two mutant mouse strains, one deficient in a glutamate receptor mGluR1 and the other a protein kinase PKCg. By analyzing these mutant mice by electrophysiology and behavioral experiments, Tonegawa's group found that a long term depression (LTD), a form of synaptic plasticity present at the parallel fiber-Purkinje cell synapses is crucially involved in motor learning. They also found that a failure in eliminating surplus climbing fibers onto Purkinje cells during the neonatal neural development leads to motor discoordination.

Finally, Dr. Tonegawa's laboratory has exploited a phage-derived DNA recombination system called the Cre-loxP system to develop tissue- or cell type-restricted gene knockout technology. Using this technology, Dr. Tonegawa's laboratory produced a mutant mouse in which the gene encoding the NMDA class of glutamate receptors is knocked out only in the pyramidal cells in the CA1 area of the hippocampus. They demonstrated that these mutant mice are deficient in LTP specifically in the CA1 area and are incapable of acquiring spatial memory. This is hitherto the cleanest evidence that supports the notion that hippocampal LTP is a required cellular mechanism for a declarative memory.

Using this and other genetically engineered mice, Tonegawa's laboratory is collaborating with Wilson's laboratory to learn more about the molecular or cellular changes that accompany at the levels of single and ensembles of synapses as the animal acquires spatial memory.


Quinn's laboratory pioneered the use of transgenes to disrupt learning and memory. Nevertheless, in the long term they believe that the forward-genetic approach offers the surest path to genuinely novel information about learning. This approach involves (a) random induction of single-gene mutations, (b) behavioral selection among these to find the ones that affect learning or memory, (c) further behavioral characterization of those mutant fly strains that look promising, and (d) molecular cloning of the altered gene, to deduce enzyme function from sequence.

Among the mutants they have isolated, amnesiac and radish are currently the most interesting. Radish flies show nearly normal learning ability followed by steady memory decay. Radish is almost extremely deficient in consolidated (anesthesia-resistant) memory (ARM) but is normal in protein-synthesis-dependent, very long term memory (VLTM). Radish's memory defect is in striking contrast to flies with a blocking CREB transcription factor., which has normal ARM but lacks VLTM. If, as seems likely, results from Drosophila mutants can be generalized across animal phyla, then researches into long-term memory phases have been studying two entirely separate storage processes without knowing it. Work with Drosophila may provide the key to the molecular mechanisms underlying both these processes. Protein-synthesis-dependent VLTM appears to require the action of cyclic-AMP-activated transcription factors of the CREB family, and thus is amenable to conventional molecular analysis. The other long lasting memory process, ARM, is more mysterious.

ARM is legitimate long-term memory because it persists up to four days after spaced training in Drosophila. It is entirely independent of the transcriptional events that subserve VLTM. At present, the only clear path to understanding the molecular mechanism underlying ARM is by cloning the radish gene whose mutation specifically abolishes it. Quinn's lab has mapped the radish gene to an interval of 140 kilobases on the X-chromosome. They have carried out a molecular walk through the region in phages and cosmid transforming vectors, and have generated the appropriate marked radish host for transformation rescue experiments. Quinn's lab is currently analyzing transcription patterns in the region. His lab is, in effect, analyzing radish's genetic defect as other laboratories have mapped and analyzed human genetic diseases. These experiments are anticipated to take at least another year but cloning the radish gene, with its entree to the mechanistic nature of ARM, seems worth the long effort.

Apparently nature has elaborated even more mechanisms to prolong the effects of experience beyond the duration of simple second-messenger responses. The memory mutant amnesiac -- very deficient in memory measured 30-60 minutes after training and reduced in both ARM and VLTM -- is altered in a gene for an apparent pre-pro neuropeptide. The most exciting consequence of this discovery is the possibility of a mammalian homologue of the amnesiac's neuropeptide. After unsuccessful "short-cut" attempts to find such a homology with PCR reactions and expression vectors, Quinn's lab is cloning and sequencing the gene from related fly species in order to determine the most conserved portions of the gene to use for more enlightened homology searches. Finding an amnesiac homologue in mammals would have direct implications in the study of human memory processes and perhaps in human therapeutics, because neuropeptides are diffusible, extracellularly-acting compounds whose action can be altered or mimicked by drugs.


Wilson's lab has successfully adapted their multiple electrode recording techniques to the mouse and have completed their initial characterization of neuronal activity in the hippocampus of the wild-type animal. This effort has been extended to several lines of genetic leaning mutants which have been made available through the Tonegawa lab. This combination of genetic, neurophysiological, and behavioral approaches promises to establish a firm link between molecular mechanisms of learning and memory and the processes of higher cognition.

In an effort to extend the cognitive scope of this work, projects are now underway examining the interactions between hippocampus and prefrontal cortical areas during tasks involving the coordination of memory and behavioral decision-making. In a related project examining long-term memory consolidation, successful simultaneous recordings have been made in hippocampus and visual neocortex of the rat, revealing the nature of these interactions for the first time.

Projects engaged in the development of microstimulation paradigms for the direct manipulation of hippocampal memory formation, reactivation, and consolidation have begun.


At least two collaborative projects among the three laboratories were established as a result of the discussion group meetings. The Wilson and Tonegawa laboratories began extending Dr. Wilson's multielectrode approach, developed for rats, for application to mice. This collaboration will allow the analysis of memory mutants that are either already available or will be produced in the Tonegawa laboratory by Wilson's multielectrode approach. Such experiments may be helpful in linking experience-induced changes in the activities of ensembles of neurons to the activity of specific gene products. Presently Dr. Tonegawa's graduate student, Thomas McHugh, is working in Dr. Wilson's laboratory to develop the mouse multielectrode analysis system. Recently Dr. Tonegawa's postdoctoral fellow, Pato Huerta, is also working in part in Dr. Wilson's laboratory.

A second collaboration between the Quinn and Tonegawa laboratories is in the exploratory stages. Dr. Quinn's laboratory previously demonstrated that the Drosophila amnesiac mutant harbors an impairment in long-term memory and has recently shown that the gene encodes a neuropeptide. Dr. Quinn's laboratory is currently cloning the mouse homologue of this gene. If successful, the Tonegawa lab will produce a mouse and analyze the mouse mutant in which the amnesia gene is knocked out.


Interactions are primarily with members of the Department of Brain and Cognitive Sciences. For instance, Tonegawa's laboratory has a collaboration with Professor Ann Graybiel's laboratory on the characterization of mutant mice deficient in dopamine receptors, and with Professor Jerry Schneider's laboratory on the mechanism of the control of nerve regeneration. In addition, Tonegawa had several scientific discussions with Professor Earl Miller on the mechanism of working memory and Professor Emilio Bizzi on motor learning and motor control. Wilson's laboratory is also collaboration with Professor Ann Graybiel's laboratory recording synaptic activity in the striatum.


Menicon Co. LTD., a Japanese manufacturer of contact lenses in Japan, gave a $2.5-million endowment to recruit a professor to the Center for Learning and Memory. A ceremony was held on January 29, 1996, to officially announce the endowment with dignitaries from MIT and Japan participating. Menicon President Kyoichi Tanaka presented the gift to the Center for Learning and Memory and MIT to commemorate the opening of the Menicon Integrated Research Laboratory in Japan.

Susumu Tonegawa

MIT Reports to the President 1995-96