In Search of Memory: The Emergence of a New Science of Mind (29 page)

This training procedure gave Quinn and Dudai a way of identifying the flies that lacked the ability to remember that odor 1 is accompanied by a shock. By 1974 they had screened thousands of flies and isolated the first mutant with a defect in short-term memory. Benzer called the mutant
dunce
. In 1981 Benzer’s student Duncan Byers, following up on the work in
Aplysia
, began to examine the cyclic AMP pathway in
dunce
and found a mutation in the gene responsible for disposing of cyclic AMP. As a result, the fly accumulates too much of the substance; its synapses presumably become saturated, making them insensitive to further change and preventing them from functioning optimally. Other mutations in memory genes were subsequently identified. They, too, involve the cyclic AMP pathway.

THE MUTUALLY REINFORCING RESULTS IN
APLYSIA
AND
Drosophila
—two very different experimental animals examined for different types of learning using different approaches—were vastly reassuring. Together, they made it clear that the cellular mechanisms underlying simple forms of implicit memory are likely to be the same in many animal species, including in people, and in many different forms of learning because those mechanisms have been conserved through evolution. Biochemistry and, later, molecular biology would be powerful tools for revealing common features in the biological machinery of different organisms.

The discoveries in
Aplysia
and
Drosophila
also reinforced an important biological principle: evolution does not require new, specialized molecules to produce a new adaptive mechanism. The cyclic AMP pathway is not unique to memory storage. As Sutherland had shown, it is not even unique to neurons: the gut, the kidney, and the liver all make use of the cyclic AMP pathway to produce persistent metabolic changes. In fact, of all the known second messengers, the cyclic AMP system is probably the most primitive. It is the most important, and in some cases the only second-messenger system found in single-celled organisms such as the bacterium
E. coli
, in which it signals hunger. Thus the biochemical actions underlying memory did not arise specifically to support memory. Rather, neurons simply recruited an efficient signaling system employed for other purposes in other cells and used it to produce the changes in synaptic strength required for memory storage.

As the molecular geneticist François Jacob has pointed out, evolution is not an original designer that sets out to solve new problems with completely new sets of solutions. Evolution is a tinkerer. It uses the same collection of genes time and again in slightly different ways. It works by varying existing conditions, by sifting through random mutations in gene structure that give rise to slightly different variations of a protein or to variations in the way that protein is deployed in cells. Most mutations are neutral or even detrimental and do not survive the test of time. Only the rare mutation that enhances an individual’s survival and reproductive capacities is likely to be retained. As Jacob writes:

In living organisms, new capabilities are achieved by modifying existing molecules slightly and adjusting their interaction with other existing molecules. Because human mental processes have long been thought to be unique, some early students of the brain expected to find many new classes of proteins lurking in our gray matter. Instead, science has found surprisingly few proteins that are truly unique to the human brain and no signaling systems that are unique to it. Almost all of the proteins in the brain have relatives that serve similar purposes in other cells of the body. This is true even of proteins used in processes that are unique to the brain, such as the proteins that serve as receptors for neurotransmitters. All life, including the substrate of our thoughts and memories, is composed of the same building blocks.

I SUMMARIZED THE FIRST COHERENT INSIGHTS INTO THE CELL
biology of short-term memory in a book entitled
Cellular Basis of Behavior
, published in 1976. In it I spelled out my belief—almost as a manifesto—that to understand behavior, one had to apply to it the same type of radical reductionist approach that had proved so effective in other areas of biology. At about the same time, Steve Kuffler and John Nicholls published
From Neuron to Brain
, a book that emphasizes the power of the cellular approach. They used cell biology to explain how nerve cells work and how they form circuits in the brain, and I used cell biology to connect the brain to behavior. Steve also sensed that connection and saw that the field of neurobiology was poised to take another big step.

I was therefore particularly pleased that in August 1980 Steve and I had a chance to travel together. We were both invited to Vienna to be inducted as honorary members of the Austrian Physiological Society. Steve had fled Vienna in 1938. We were introduced to the medical faculty of the University of Vienna by Wilhelm Auerwald, a pretentious academic who had accomplished little scientifically and who acted as if nothing out of the ordinary had caused these two sons of Vienna to flee the country. The professor blithely remarked that Kuffler had attended medical school in Vienna and that I had lived in Severingasse, literally around the corner from the university. His silence regarding our actual experiences in Vienna spoke volumes. Neither Steve nor I responded to his comments.

Two days later, we took a boat down the Danube from Vienna to Budapest, where we attended the International Meeting of Physiologists. It was the last important meeting Steve attended. He gave a superb lecture. Shortly thereafter, in October 1980, he died of a heart attack at his weekend home in Woods Hole, Massachusetts, having just come back from a long swim.

Like most of the neural science community, I was shattered when I heard the news. We were all indebted to and in some ways dependent upon him. Jack McMahan, one of Steve’s most devoted students, described the reaction many of us felt: “How could he do this to us?”

I was president of the Society for Neuroscience that year and was responsible with the program committee for organizing the annual meeting in November. The meeting was held in Los Angeles just a few weeks after Steve’s death, and about ten thousand neuroscientists attended. David Hubel delivered a remarkable eulogy. Accompanied by slides, he illustrated how prescient, insightful, and generous Steve had been and how much he had meant to us all. I don’t think anyone on the American scene since then has been as influential or as beloved as Steve Kuffler. Jack McMahan organized a posthumous volume in his honor, and in my contribution I stated, “In writing this piece, I sense how much he is still here. Next to Alden Spencer, there is no colleague in science that I have lost that I think of and miss more.”

The death of Steve Kuffler marked the end of an era, an era in which the neural science community was still relatively small and focused on the cell as the unit of brain organization. Steve’s death coincided with the merger of molecular biology and neural science, a step that expanded dramatically both the scope of the field and the number of scientists in it. My own work reflects this change: to a large degree I ended my cellular and biochemical studies of learning and memory in 1980. By that time it was becoming clear to me that the increase in cyclic AMP and the enhancement of transmitter release produced by serotonin in response to a single learning trial lasts only minutes. Long-term facilitation lasting days and weeks must involve something more, perhaps changes in the expression of genes as well as anatomical changes. So I turned to the study of genes.

I was ready for this step. Long-term memory was beginning to fire my imagination. How can one remember events from childhood for the whole of one’s life? Denise’s mother, Sara Bystryn, who imbued Denise and her brother, Jean-Claude, as well as their spouses and children, with her taste in decorative arts—art nouveau furniture, vases, and lamps—rarely spoke to me about my science. But she must have somehow sensed that I was ready to tackle genes and long-term memory.

On my fiftieth birthday, November 7, 1979, she bought me a beautiful Viennese vase by Teplist (figure 16–5) and gave it to me with the following note:

Sara Bystryn had defined my task.

16–5 The Teplist Vase.
(From Eric Kandel’s personal collection.)

I
n reflecting on his genetic studies of bacteria, François Jacob distinguished between two categories of scientific investigation: day science and night science. Day science is rational, logical, and pragmatic, carried forward by precisely designed experiments. “Day science employs reasoning that meshes like gears, and achieves results with the force of certainty,” Jacob wrote. Night science, on the other hand, “is a sort of workshop of the possible, where are elaborated what will become the building materials of science. Where hypotheses take the form of vague presentiments, of hazy sensations.”

By the mid-1980s, I felt that our studies of short-term memory in
Aplysia
were edging toward the threshold of day science. We had succeeded in tracing a simple learned response in
Aplysia
to the neurons and synapses that mediate it and had found that learning gives rise to short-term memory by producing transient changes in the strength of existing synaptic connections between sensory and motor neurons. Those short-term changes are mediated by proteins and other molecules already present at the synapse. We had discovered that cyclic AMP and protein kinase A enhance the release of glutamate from the terminals of the sensory neurons, and that this enhanced release is a key element in short-term memory formation. In brief, we had in
Aplysia
an experimental system whose molecular components we could manipulate experimentally in a logical way.

But a central mystery in the molecular biology of memory storage remained: How are short-term memories transformed into enduring, long-term memories? This mystery became for me a subject of night science: of romantic musings and unconnected ideas, of months of considering how we might pursue the solution through day science experiments.

Jimmy Schwartz and I had found that long-term memory formation depends upon the synthesis of new proteins. I had a hunch that long-term memory, which involves enduring changes in synaptic strength, could be tracked to changes in the genetic machinery of sensory neurons. Pursuing this vague idea meant carrying our analysis of memory formation even deeper into the molecular labyrinth of the neuron: to the nucleus of the cell, where genes reside and where their activity is controlled.

In my late-night musings, I dreamed of taking the next step, of using the newly developed techniques of molecular biology to listen in on the dialogue between sensory neurons’ genes and their synapses. This next step could not have come at a more opportune time. By 1980, molecular biology had become the dominant and unifying force within biology. It would soon extend its influence to neural science and help create a new science of mind.

HOW DID MOLECULAR BIOLOGY, PARTICULARLY MOLECULAR
genetics, get to be so important? The emergence of molecular biology and its initial influence can be traced to the 1850s, when Gregor Mendel first realized that hereditary information is passed from parent to offspring by means of discrete biological units we now call genes. In about 1915 Thomas Hunt Morgan discovered in fruit flies that each gene resides at a specific site, or locus, on the chromosomes. In flies and other higher organisms, the chromosomes are paired: one comes from the mother, the other from the father. The offspring thus receives one copy of each gene from each of its two parents. In 1942 the Austrian-born theoretical physicist Erwin Schrödinger gave a series of lectures in Dublin that was later published in a little volume entitled
What Is Life
? In that book he noted that it is the differences in their genes that distinguish one animal species from another and human beings from other animals. Genes, Schrödinger wrote, endow organisms with their distinctive features; they code biological information in a stable form so that it can be copied and transmitted reliably from generation to generation. Thus when a paired chromosome separates, as it does during cell division, the genes on each chromosome must be copied exactly into genes on the new chromosome. Life’s key processes—the storing and passing on of biological information from one generation to the next—are carried out through the replication of chromosomes and the expression of genes.

Schrödinger’s ideas caught the attention of physicists and brought a number of them into biology. In addition, his ideas helped transform biochemistry, one of the core areas of biology, from a discipline concerned with enzymes and the transformation of energy (that is, with how energy is produced and utilized in the cell) to a discipline concerned with the transformation of information (how information is copied, transmitted, and modified within the cell). Viewed from this new perspective, the importance of chromosomes and genes is that they are carriers of biological information. By 1949 it was already clear that a number of neurological diseases, such as Huntington’s and Parkinson’s, as well as several mental illnesses, including schizophrenia and depression, had genetic components. The nature of the gene therefore became the central question for all of biology, including, ultimately, the biology of the brain.

What is the nature of the gene? Of what is it made? In 1944 Oswald Avery, Maclyn McCarty, and Colin MacLeod at the Rockefeller Institute made the breakthrough discovery that genes are not proteins as many biologists had thought, but instead are made of deoxyribonucleic acid (DNA).

Nine years later, in the April 25, 1953, issue of
Nature
, James Watson and Francis Crick described their now historic model of the structure of DNA. With the help of X-ray photographs taken by the structural biologists Rosalind Franklin and Maurice Wilkins, Watson and Crick were able to infer that DNA is composed of two long strands wound around each other in the form of a spiral, or helix. Knowing that each strand in this double helix is made up of four small, repeating units called nucleotide bases—adenine, thymine, guanine, and cytosine—Watson and Crick assumed that the four nucleotides are the information-carrying elements of the gene. This led them to the striking discovery that the two strands of DNA are complementary and that the nucleotide bases on one strand of DNA form pairs with specific nucleotide bases on the other strand: adenine (A) in one stand pairing and binding only to thymine (T) in the other, and guanine (G) in one strand pairing and binding only to cytosine (C) in the other. The pairing of nucleotide bases at multiple points along their length holds the two strands together.

Watson and Crick’s discovery put Schrödinger’s ideas into a molecular framework, and molecular biology took off. The essential operation of genes, as Schrödinger pointed out, is replication. Watson and Crick ended their classic paper with the now famous sentence, “It has not escaped our notice that the specific pairing we have postulated immediately suggests a copying mechanism for genetic material.”

The double helix model illustrates how gene replication works. When the two strands of DNA unwind during replication, each parent strand acts as a template for the formation of a complementary daughter strand. Since the sequence of the information-containing nucleotides on the parent strand is given, it follows that the sequence on the daughter strand will also be given: A will bind to T and G to C. The daughter strand can then serve as a template for the formation of still another strand. In this way, multiple copies of DNA can be replicated faithfully as a cell divides and the copies can be distributed to daughter cells. This pattern extends to all the cells of an organism, including the sperm and the egg, thus enabling the organism as a whole to be replicated from generation to generation.

Taking their cue from gene replication, Watson and Crick further suggested a mechanism for protein synthesis. Since each gene directs the production of a particular protein, they reasoned that the sequence of nucleotide bases in each gene carries the code for protein production. As in gene replication, the genetic code for proteins is “read out” by making a complementary copy of the nucleotide bases in a strand of DNA. But in protein synthesis, later work showed, the code is carried by an intermediary molecule called messenger RNA (ribonucleic acid). Like DNA, messenger RNA is a nucleic acid made up of four nucleotides. Three of them—adenine, guanine, and cytosine—are identical to the nucleotides in DNA, but the fourth, uracil, is unique to RNA and replaces thymine. When the two strands of DNA in a gene separate, one of the strands is copied into messenger RNA. The sequence of nucleotides in messenger RNA is later translated into protein. Watson and Crick thus formulated the central dogma of molecular biology: DNA makes RNA, and RNA makes protein.

The next step was to crack the genetic code, the rules whereby the nucleotides in messenger RNA are translated into the amino acids of protein, including proteins important for memory storage. Attempts to do this began in earnest in 1956, when Crick and Sydney Brenner focused on how the four nucleotides in DNA could code for the twenty amino acids that combine to form proteins. A one-to-one system, with each nucleotide coding for a single amino acid, would yield only four amino acids. A code using different pairs of nucleotides would yield only sixteen amino acids. To produce twenty unique amino acids, Brenner argued, the system would have to be based on triplets—that is, on combinations of three nucleotides. However, triplets of nucleotides yield not twenty, but sixty-four combinations. Brenner therefore suggested that a code based on triplets is degenerate (redundant), meaning that more than one triplet of nucleotides encodes the same amino acid.

In 1961 Brenner and Crick proved that the genetic code consists of a series of nucleotide triplets, each of which contains the instructions for forming a unique amino acid. But they did not show which triplets code for which amino acids. That was revealed later in the same year by Marshall Nirenberg at NIH and by Har Gohind Khorana at the University of Wisconsin. They tested Brenner and Crick’s idea biochemically and cracked the genetic code by describing the specific combinations of nucleotides that code for each amino acid.

In the late 1970s Walter Gilbert at Harvard and Frederick Sanger in Cambridge, England, developed a new biochemical technique that made it possible to sequence DNA rapidly, that is, to read segments of the nucleotide sequences in DNA with relative ease and thus to determine what protein a given gene encodes. This proved a remarkable advance. It enabled scientists to observe that the same stretches of DNA occur in different genes and encode identical or similar regions in a variety of proteins. These recognizable regions, called domains, mediate the same biological function, regardless of the protein in which they occur. Thus, by merely looking at some of the nucleotide sequences that make up a gene, scientists could determine important aspects of how the protein encoded by that gene would work, whether the protein was a kinase, an ion channel, or a receptor, for example. Furthermore, by comparing the sequence of amino acids in different proteins, they could recognize similarities between proteins encountered in very different contexts, such as in different cells of the body or even in vastly different organisms.

From these sequences and comparisons of them, a blueprint emerged of how cells work and how they signal one another, forming a conceptual framework for understanding many of life’s processes. In particular, these studies revealed once again that different cells—indeed, different organisms—are made out of the same material. All multicellular organisms have the enzyme that synthesizes cyclic AMP; they all have kinases, ion channels, and on and on. In fact, half of the genes expressed in the human genome are present in much simpler invertebrate animals such as the worm
C. elegans
, the fly
Drosophila
, and the snail
Aplysia
. The mouse has more than 90 percent and the higher apes 98 percent of the coding sequences of the human genome.

A KEY ADVANCE IN MOLECULAR BIOLOGY THAT FOLLOWED DNA
sequencing, and the one that brought me into the field, was the emergence of recombinant DNA and gene cloning, techniques that make it possible to identify genes, including those expressed in the brain, and to determine their function. The first step is to isolate from a person, a mouse, or a snail the gene one wishes to study—that is, the segment of DNA that codes for a particular protein. One does this by locating the gene on the chromosome and then snipping it out with molecular scissors—enzymes that cut the DNA at appropriate spots.

The next step is to make many copies of the gene, a process known as cloning. In cloning, the ends of the excised gene are stitched to stretches of DNA from another organism, such as a bacterium, creating what is known as recombinant DNA—recombinant because a gene snipped from one organism’s DNA is recombined with the genome from another organism. The genome of a bacterium divides every twenty minutes or so, producing large numbers of identical copies of the original gene. The final step is to decipher the protein that the gene encodes. This is done by reading the sequence of nucleotides or molecular building blocks, in the gene.

In 1972 Paul Berg of Stanford University succeeded in creating the first recombinant DNA molecule, and in 1973 Herbert Boyer of the University of California, San Francisco and Stanley Cohen of Stanford University elaborated on Berg’s technique to develop gene cloning. By 1980 Boyer had spliced the human insulin gene into a bacterium, a feat that gave rise to an unlimited amount of human insulin and thereby created the biotechnology industry. Jim Watson, the co-discoverer of the structure of DNA, would write of these achievements as playing God: