In Search of Memory: The Emergence of a New Science of Mind (34 page)

These two features make the new variant of the prion ideally designed for memory storage. Self-perpetuation of a protein that is critical for local protein synthesis allows information to be stored selectively and in perpetuity at one synapse, and not, Kausik soon discovered, at the many others that a neuron makes with its target cells.

Beyond discovering the new prion’s relevance to the persistence of memory or even to the functioning of the brain, Kausik and I had found two new biological features of prions. First, a normal physiological signal—serotonin—is critical for converting CPEB from one form to another. Second, CPEB is the first self-propagating form of a prion known to serve a physiological function—in this case, perpetuation of synaptic facilitation and memory storage. In all other cases previously studied, the self-propagating form either causes disease and death by killing nerve cells or, more rarely, is inactive.

We have come to believe that Kausik’s discovery may be just the tip of a new biological iceberg. In principle, this mechanism—activation of a nonheritable, self-perpetuating change in a protein—could operate in many other biological contexts, including development and gene transcription.

This exciting finding in my laboratory illustrates that basic science can be like a good mystery novel with surprising twists: some new, astonishing process lurks in an undiscovered corner of life and is later found to have wide significance. This particular finding was unusual in that the molecular processes underlying a group of strange brain diseases are now seen also to underlie long-term memory, a fundamental aspect of healthy brain function. Usually, basic biology contributes to our understanding of disease states, not vice versa.

IN RETROSPECT, OUR WORK ON LONG-TERM SENSITIZATION AND
the discovery of the prionlike mechanism brought to the forefront three new principles that relate not only to
Aplysia
but to memory storage in all animals, including people. First, activating long-term memory requires the switching on of genes. Second, there is a biological constraint on what experiences get stored in memory. To switch on the genes for long-term memory, CREB-1 proteins must be activated and CREB-2 proteins, which suppress the memory-enhancing genes, must be inactivated. Since people do not remember everything they have learned—nor would anyone want to—it is clear that the genes that encode suppressor proteins set a high threshold for converting short-term to long-term memory. It is for this reason that we remember only certain events and experiences for the long run. Most things we simply forget. Removing that biological constraint triggers the switch to long-term memory. The genes activated by CREB-1 are required for new synaptic growth. The fact that a gene must be switched on to form long-term memory shows clearly that genes are not simply determinants of behavior but are also responsive to environmental stimulation, such as learning.

Finally, the growth and maintenance of new synaptic terminals makes memory persist. Thus, if you remember anything of this book, it will be because your brain is slightly different after you have finished reading it. This ability to grow new synaptic connections as a result of experience appears to have been conserved throughout evolution. As an example, in people, as in simpler animals, the cortical maps of the body surface are subject to constant modification in response to changing input from sensory pathways.

W
hen I first began to study the biological basis of memory, I focused on the memory storage that ensues from the three simplest forms of learning: habituation, sensitization, and classical conditioning. I found that when a simple motor behavior is modified by learning, those modifications directly affect the neural circuit responsible for the behavior, altering the strength of preexisting connections. Once stored in the neural circuit, that memory can be recalled immediately.

This finding gave us our first insight into the biology of implicit memory, a form of memory that is not recalled consciously. Implicit memory is responsible not only for simple perceptual and motor skills but also, in principle, for the pirouettes of Margot Fonteyn, the trumpeting technique of Wynton Marsalis, the accurate ground strokes of Andre Agassi, and the leg movements of an adolescent riding a bicycle. Implicit memory guides us through well-established routines that are not consciously controlled.

The more complex memory that had inspired me initially—the explicit memory for people, objects, and places—is consciously recalled and can typically be expressed in images or words. Explicit memory is far more sophisticated than the simple reflex I had studied in
Aplysia
. It depends on the elaborate neural circuitry of the hippocampus and the medial temporal lobe, and it has many more possible storage sites.

Explicit memory is highly individual. Some people live with such memories all the time. Virginia Woolf falls into this category. Her memories of childhood were always at the edge of her consciousness, ready to be summoned up and incorporated into everyday moments, and she had an exquisite ability to describe the details of her recalled experiences. Thus, years after the death of her mother, Woolf ’s memory of her was still fresh:

Other people call up their past life only occasionally. Periodically, I think back and recall the two police officers coming to our apartment and ordering us to leave on the day of Kristallnacht. When this memory enters my consciousness, I can once again see and feel their presence. I can visualize the worried expression on my mother’s face, feel the anxiety in my body, and perceive the confidence in my brother’s actions while retrieving his coin and stamp collections. Once I place these memories in the context of the spatial layout of our small apartment, the remaining details emerge in my mind with surprising clarity.

Remembering such details of an event is like recalling a dream or watching a movie in which we play a part. We can even recall past emotional states, though often in a much simplified form. To this day I remember some of the emotional context of my romantic encounter with our housekeeper Mitzi.

As Tennessee Williams wrote in
The Milk Train Doesn’t Stop Here Anymore
, describing what we now call explicit memory, “Has it ever struck you…that life is all memory, except for the one present moment that goes by you so quickly you hardly catch it going? It’s really all memory…except for each passing moment.”

For all of us, explicit memory makes it possible to leap across space and time and conjure up events and emotional states that have vanished into the past yet somehow continue to live on in our minds. But recalling a memory episodically—no matter how important the memory—is not like simply turning to a photograph in an album. Recall of memory is a creative process. What the brain stores is thought to be only a core memory. Upon recall, this core memory is then elaborated upon and reconstructed, with subtractions, additions, elaborations, and distortions. What biological processes enable me to review my own history with such emotional vividness?

ON REACHING MY SIXTIETH BIRTHDAY, I FINALLY GATHERED THE
courage to return to the study of the hippocampus and explicit memory. I had long been curious to see whether some of the basic molecular principles we had learned from a simple reflex circuit in
Aplysia
also applied to the complex neural circuits of the mammalian brain. By 1989, three major breakthroughs had made it feasible to explore this question in the laboratory.

The first was the discovery that the pyramidal cells of the hippocampus play a critical role in an animal’s perception of its spatial environment. The second was the discovery of a remarkable synaptic strengthening mechanism in the hippocampus called long-term potentiation. Many researchers thought this mechanism might underlie explicit memory. The third breakthrough, and the one most immediately relevant to my own molecular approach to learning, was the invention of powerful new methodologies for modifying mice genetically. My colleagues and I would adapt those methods to the brain in an attempt to explore explicit memory in the hippocampus in the same molecular detail as we had studied implicit memory in
Aplysia
.

The new era of research in the hippocampus began in 1971, when John O’Keefe at University College, London made an amazing discovery about how the hippocampus processes sensory information. He found that neurons in the hippocampus of the rat register information not about a single sensory modality—sight, sound, touch, or pain—but about the space surrounding the animal, a modality that depends on information from several senses. He went on to show that the hippocampus of rats contains a representation—a map—of external space and that the units of that map are the pyramidal cells of the hippocampus, which process information about place. In fact, the pattern of action potentials in these neurons is so distinctively related to a particular area of space that O’Keefe referred to them as “place cells.” Soon after O’Keefe’s discovery, experiments with rodents showed that damage to the hippocampus severely compromises the animals’ ability to learn a task that relies on spatial information. This finding indicated that the spatial map plays a central role in spatial cognition, our awareness of the environment around us.

Since space involves information acquired through several sensory modalities, it raised the questions: How are these modalities brought together? How is the spatial map established? Once established, how is the spatial map maintained?

The first clue to the answers emerged in 1973, when Terje Lømo and Tim Bliss, postdoctoral students in Per Andersen’s laboratory in Oslo, discovered that the neuronal pathways leading to the hippocampus of rabbits can be strengthened by a brief burst of neural activity. Lømo and Bliss were unaware of O’Keefe’s work and did not attempt to examine the functioning of the hippocampus in the context of memory or a specific behavior, as we had done with
Aplysia
’s gill-withdrawal reflex. Instead, they adopted an approach similar to the one Ladislav Tauc and I had first taken in 1962: they developed a neural analog of learning. Rather than basing their neural analog on conventional behavioral paradigms, such as habituation, sensitization, or classical conditioning, they based it on neuronal activity per se. They applied a very rapid train of electrical stimuli (100 impulses per second) to a neuronal pathway leading to the hippocampus and found that the synaptic connections in that pathway were strengthened for several hours to one or more days. Lømo and Bliss called this form of synaptic facilitation long-term potentiation.

It soon emerged that long-term potentiation occurs in all three of the pathways within the hippocampus and that it is not a unitary process. Instead, long-term potentiation describes a family of slightly different mechanisms, each of which increases the strength of the synapse in response to different rates and patterns of stimulation. Long-term potentiation is analogous to long-term facilitation of the connections between sensory and motor neurons in
Aplysia
in that it enhances the strength of synaptic connections. But whereas long-term facilitation in
Aplysia
strengthens synapses heterosynaptically, by means of a modulatory transmitter acting on the homosynaptic pathway, many of long-term potentiation can be initiated merely by means of homosynaptic activity. As we and others found later, however, neuromodulators are usually recruited to switch short-term homosynaptic plasticity into long-term heterosynaptic plasticity.

In the early 1980s, Andersen simplified Lømo and Bliss’s research methodology greatly by removing the hippocampus from the brain of a rat, cutting it into slices, and placing the slices in an experimental dish. This enabled him to observe the several neural pathways in a specific segment of the hippocampus. Amazingly, such brain slices can function for hours when prepared properly. With this advance, researchers could analyze the biochemistry of long-term potentiation and observe the effects of drugs blocking various signaling components.

Key molecules involved in long-term potentiation began to emerge from such studies. In the 1960s David Curtis collaborating with Geoffrey Watkins discovered that glutamate, a common amino acid, is the major excitatory transmitter in the vertebrate brain (just as it is in the invertebrate brain, we later discovered). Watkins and Graham Collingridge then found that glutamate acts on two different types of ionotropic receptors in the hippocampus, the AMPA receptor and the NMDA receptor. The AMPA receptor mediates normal synaptic transmission and responds to an individual action potential in the presynaptic neuron. The NMDA receptor, on the other hand, responds only to extraordinarily rapid trains of stimuli and is required for long-term potentiation.

When a postsynaptic neuron is stimulated repeatedly, as in Bliss and Lømo’s experiments, the AMPA receptor generates a powerful synaptic potential that depolarizes the cell membrane by as much as 20 or 30 millivolts. This depolarization causes an ion channel in the NMDA receptor to open, allowing calcium to flow into the cell. Roger Nicoll at the University of California, San Francisco and Gary Lynch at the University of California, Irvine discovered independently that the flow of calcium ions into the postsynaptic cell acts as a second messenger (much as cyclic AMP does), triggering long-term potentiation. Thus the NMDA receptor can translate the electrical signal of the synaptic potential into a biochemical signal.

These biochemical reactions are important because they trigger molecular signals that can be broadcast throughout the cell and thus contribute to long-lasting synaptic modifications. Specifically, calcium activates a kinase (called the calcium, calmodulin-dependent protein kinase) that increases synaptic strength for about an hour. Nicoll went on to show that the calcium influx and the activation of this kinase lead to the strengthening of synaptic connections by causing additional AMPA receptors to be assembled and inserted into the membrane of the postsynaptic cell.

The analysis of how the NMDA receptor functions elicited great excitement among neuroscientists, for it showed that the receptor acts as a coincidence detector. It allows calcium ions to flow through its channel if and only if it detects the coincidence of two neural events, one presynaptic and the other postsynaptic. The presynaptic neuron must be active and release glutamate,
and
the AMPA receptor in the postsynaptic cell must bind glutamate and depolarize the cell. Only then will the NMDA receptors become active and allow calcium to flow into the cell, thereby triggering long-term potentiation. Interestingly, in 1949 the psychologist D. O. Hebb had predicted that some kind of neural coincidence detector would be present in the brain during learning: “When an axon of cell A…excites cell B and repeatedly or persistently takes part in its firing, some growth process or metabolic changes take place in one or both cells so that A’s efficiency is increased.”

Aristotle, and subsequently the British empiricist philosophers and many other thinkers, had proposed that learning and memory are somehow the result of mind’s ability to associate and form some lasting mental connection between two ideas or stimuli. With the discovery of the NMDA receptor and long-term potentiation, neuroscientists had unearthed a molecular and cellular process that could well carry out this associative process.