Introduction Investigations of the biological basis of memory have proceeded simultaneously at many different levels: individual neurons and synapses, systems of neurons, whole brains, and whole behaving organisms. Isolated cells, slabs of brain tissue, live invertebrates (such as sea slugs), nonhuman vertebrates (often rats), and even awake human patients have been studied. Several strategies have been used to reveal the location of and mechanisms underlying the engram, or memory trace. In one approach, after an animal subject has learned a task very well, lesions are made in specific regions of the brain. If only memory of the task is impaired, the damaged structure is implicated in the memory process.
Trauma, stroke, disease, and even deliberate surgery may produce “natural” lesions in human patients. Resultant amnesia, or memory loss, can be correlated with the damaged structures using magnetic resonance imaging (MRI), positron emission tomography (PET), computed tomography (CT), and other scanning techniques, and with behavioral performance on standardized psychological and neuropsychological tests.
Another approach is to record changes in the nervous system that occur at the same time as a memory process. If such changes occur only when memory is formed, then they may play a role in the process. Such functional changes in the human (and animal) brain may be seen using single-cell or multicell electrophysiology, electroencephalography, and evoked-potential electrical recording techniques. Researchers have studied memory formation by looking at the chemical composition of the brain (for example, neurotransmitters and protein synthesis).
Finally, various techniques may be used to disrupt the memory formation (consolidation) process shortly before or after a subject has learned a task (that is, started to develop permanent memories). Successful disruption may point to the underlying nature of the process. For example, formation of permanent memories can be prevented by giving direct electrical stimulation to the amygdala (a specific brain area associated with emotion and motivation) or enhanced by stimulation of the reticular formation (associated with alertness and waking), suggesting different roles for these areas.
Neuropsychology, a specialty combining clinical neurology and behavioral analysis, has used both “natural” and experimental lesion techniques to explore memory processes. According to some researchers, for the purposes of neuropsychological analysis, memory should be divided into two categories: declarative and procedural. Declarative memories are facts that one can consciously recollect, while procedural memories are skills or operations that one does not have to think about consciously and that are not linked to a particular time or place. Brain damage does not usually impair procedural knowledge. Global permanent anterograde amnesia
(inability to form new permanent memories) and temporary retrograde amnesia(inability to remember past events) for declarative knowledge, however, have been correlated consistently with bilateral (left and right side) lesions of both cortical (the highest and most newly evolved) and subcortical (the lower and older) regions of the brain. Memory problems of chronic alcoholics (Wernicke-Korsakoff syndrome), demented patients (Alzheimer’s disease), and patients with some strokes and aneurysms have all been associated with damage to these areas.
Deficits in long-term memory produced by temporal cortex lesions (lesions of the lateral brain area) depend on the type of material that is presented: Left lesions interfere with verbal material, while right lesions interfere with nonverbal material. It does not matter by what sensory modality the material was presented or what modality was used to test for its retention. By contrast, lesions of the frontal cortex (the largest and most forward brain area) interfere with only certain components of memory: memory for the order of things and events, short-term memory for the location of things in space, normal resistance to distraction during learning, and the ability to learn new material without being confused by old material. Unusual and discrete types of amnesia can result from damage to other specific cortical areas, such as the parietal, posterior temporal, and occipital lobes. For example, people may be unable to remember and recognize colors, faces, the names of objects, or the location of an object in the environment.
In contrast to this whole-brain approach, neurobiological investigations have focused on short-, intermediate-, and long-term cellular memory mechanisms. The sensory memory, the shortest memory, persists for about 0.5 second and depends on the activity of reverberating circuits. These circuits are loops of interconnected neurons arranged so that stimulating one will activate each successive one, including, eventually, the first one again. The net effect of this arrangement is that the entire loop stays active—and the memory trace of the initial stimulation persists—long after the initial stimulation has ended.
Other, more enduring memories, lasting from days to years, are thought to reside in synaptic (neuron-to-neuron) connections and to result from neuronal plasticity—the creation of new synaptic connections or increased capacity or efficiency of old ones. Many mechanisms for this have been proposed. One possibility is that learning causes neurons to sprout new terminals and make new connections directly (this is termed synaptic turnover or reactive synaptogenesis). Another possibility is that learning liberates blocked connections, which can then respond to incoming signals (the Calpain-Fodrin theory). Dendritic branching may also be important. If memories are stored in synaptic connections, brains with thickly branched dendrites could store more memories than those with thinly branched dendrites. It is known, for example, that brains of healthy elderly people have more dendritic branches than brains of younger people or of adults with some types of memory disorders. Moreover, animals raised in stimulating learning environments develop more branches and a greater brain mass than do less stimulated control animals. Learning-induced changes in the shape of dendrites may also promote memory: Stubby ones transmit information more readily than long, thin ones.
Among the most important proponents of animal models in the elucidation of the physiological basis for memory has been Eric Kandel. Kandel, a professor at Columbia University in New York, was awarded the Nobel Prize in Physiology or Medicine in 2000 for his discoveries in the molecular basis for memory. Kandel has carried out much of his work studying the nervous system in the sea slug Aplysia. This organism contains relatively few nerve cells (approximately twenty thousand); its neural circuitry is simple by comparison with more evolved organisms such as human beings (containing approximately one trillion nerve cells), and thus it serves as an ideal laboratory animal in the study of memory. Behavioral changes in the animal may involve fewer than one hundred nerve cells.
Kandel tested various forms of stimuli on the organism and observed changes in the responding gill withdrawal reflex as the basis for “memory.” Since the reflex would remain for various periods of time, it represented a primitive form of memory. An amplification of the synapses connecting sensory nerve cells to motor neurons could be detected as a molecular response to stimuli and represented memory development.
If the stimulus was weak, a form of “short-term memory” would develop. In this case, the reflex lasted only a short period of time. If the stimulus was stronger, “memory” would last for weeks. At the molecular level, the basis for memory was found to be the relative levels of neurotransmitters that would be released at the synapses. Short-term memory was represented by calcium, which originated from specific ion channels, with resultant release of higher levels of neurotransmitters at the synapses. Kandel found phosphates were joined to certain channel proteins, resulting in amplification of the response.
Formation of long-term memory had an analogous mechanism utilizing phosphorylation reactions. The concentration of an enzyme, protein kinase A (PKA), also involved in phosphorylation of protein targets, was increased in neurons following higher levels of stimuli.
Activity of this enzyme was related to formation of a second molecule within the cell, cyclic adenosine monophosphate (cAMP). The result of increasing the activity of PKA and cAMP was to stimulate the cell to increase levels of proteins in the synapse, with the effect of increasing function of that synapse. Another key protein activated by cAMP was called the cAMP response element binding protein (CREB). If synthesis of new proteins was blocked using drugs, no long-term memory would result. Kandel summarized his work in the statement that all memory is “located in the synapse.”
Role of Neurotransmitters The brain and body produce a number of substances that have the ability to modify memory in everyday life. Catecholamines, brain neurotransmitters released during emotional states, seem to facilitate memory storage, and damage to brain structures that secrete catecholamines impairs memory. Stress-produced hormones from the pituitary and adrenal glands can also alter memory formation. Even the endorphins, the body’s own morphinelike substances released during stress and involved in pain reduction, may play a modulatory role. Drugs can also influence memory function by altering nervous system activity. For example, stimulants of the central nervous system, such as amphetamine, can enhance memory formation, while depressants of the system, such as barbiturates and morphine, can interfere with it.
Kandel believes that the neurotransmitter serotonin is particularly important in regulating activity of both cAMP and PKA, thereby playing a critical role in memory. Addition of serotonin resulted in an increase in excitability of the synapses, similar to that of adding cAMP directly. Serotonin itself was found to cause an increase in cAMP levels.
The same neurotransmitter, serotonin, was found to be involved in development of both short-term and long-term memory. When a synapse is activated by serotonin, a signal results that activates cAMP and the CREB protein. Newly synthesized proteins move to the terminals of the cells. Only those synapses bound by serotonin undergo development and growth.
The different forms of memory are the result of different forms of stimuli. Short-term memory results from a synapse-specific increase in the level of neurotransmitter. Preexisting proteins are modified (phosphorylated), and synaptic connections involve a relatively small number of neurons. In contrast, long-term memory results from activation of a protein pathway and requires new protein synthesis. In addition, significantly more connections are formed with other nerve synapses.
Understanding Memory Loss Sometimes an anecdotal observation leads to new understanding of human behavior. Schizophrenia is a severe mental illness, characterized by thought disorder and incoherent speech, that has been extremely resistant to cure and treatment. By chance, the Italian psychiatrist Ugo Cerletti, practicing in the late 1930s, observed that electric shock applied to the heads of pigs in the local slaughterhouse made the pigs easier to manage. This inspired him to apply a small voltage to the temples of one of his schizophrenic patients in the hope that improvement would result. Nothing appeared to change, so he announced his intention to increase the voltage. At this point, to Cerletti’s amazement, the patient protested loudly and with perfectly coherent speech. Encouraged by this improvement, Cerletti gave another shock. This time, unfortunately, the patient became unconscious because of a massive brain seizure and, on awakening, experienced retrograde amnesia—he was unable to recall what had happened to him in the recent and sometimes distant past.
Following this, shock treatment, or electroconvulsive therapy (ECT), became quite popular for treating schizophrenia and was tried for every sort of mental illness. It became clear over the years, however, that only patients with mood disorders (particularly certain severely depressed patients) showed consistent benefits. In suicidal patients, ECT became the life-saving treatment of choice.
Transient and even permanent memory loss associated with ECT was largely ignored by clinicians, but this amnesiac effect became the focus of experimental research with animals. In a typical experiment, rats were trained on a simple learning task. Electroconvulsive shock—electrical shock to the brain—was given either immediately or at various intervals after training. Memory for the task was then tested a day or so later. Memory was poorest if the shocks were given right after training. This observation led to the conclusion that memory does not form instantaneously but takes time to “consolidate,” a major development in memory research. Further investigation determined that, in animals, the body convulsions produced by the shocks were not themselves responsible for the retrograde amnesiac effects. Moreover, a variety of brain structures were found to produce amnesia without seizures. Indeed, stimulation of one structure (the hippocampus) disturbed long-term memory, while stimulation of another (the reticular formation) interfered with short-term memory, indicating different roles for each in the memory process.
This consolidation-disruption research strategy has become one of the major approaches used by memory investigators. Experimental work such as this has also led to improvements in ECT: Shocks are given to a smaller brain region; voltages are lower and better controlled (although seizures are still produced); muscle relaxants and sedatives are administered to avoid the hazards of convulsions; drugs are given to minimize heart-rate and blood-pressure changes; and oxygen is given to lessen memory loss.
The interplay between clinical and experimental work is also seen in research on
Alzheimer’s disease, a slowly debilitating and life-threatening disease that affects 6 percent of the adult population. Once thought to be an inevitable consequence of aging, or senility, caused by multiple “ministrokes,” it is now recognized as a distinct and specific disorder. It begins insidiously, with difficulties of concentration, followed by increasing troubles with problem solving, speaking, learning, and remembering. Patients become apathetic and disoriented; even the ability to recognize loved ones is eventually lost.
Much of psychology’s meager understanding of the basis for the memory impairment associated with this disease has come from basic research with animals. For example, studies of patients and animals with lesions in the hippocampus suggested early that this was an important site for memory formation but not for permanent storage. The hippocampus also contains mechanisms for neural plasticity. This is significant because postmortem studies of Alzheimer’s patients’ brains have revealed a selective loss of cells going to the hippocampus and cells in the neocortex, the highest brain area. Many researchers believe that both the number and shape of dendrites and the growth of synaptic endings are important for memory formation. The number of dendrites is reduced in Alzheimer’s patients. Moreover, even normal cells contain abnormal strands (neurofibrillary tangles) inside and tangled masses (neuritic plaques) outside the nerve cells.
Basic animal research has also shown that acetylcholine, an important brain neurotransmitter, plays an important modulatory role in memory. Drugs that interfere with acetylcholine function can produce some symptoms of dementia (cognitive impairment, including memory deficits) in normal human subjects. Moreover, if acetylcholine neurons are transplanted from fetal (still-developing) rat brains to the brains of old rats, the old rats regain some of their youthful ability to learn and remember. In view of these findings, it is significant that both acetylcholine and the enzyme needed to make it (choline acetyltransferase) are reduced in Alzheimer’s patients. Indeed, the greater the reduction, the worse the patient’s symptoms. This observation, coupled with results from animal studies, has encouraged clinicians to treat the memory and other cognitive problems of Alzheimer’s patients with drugs that enhance acetylcholine function—so far, however, with limited success.
Research and Memory During the 1880s, the fact that learning, memory, and forgetting operate according to laws was revealed by Hermann Ebbinghaus’s laboratory investigations in Über das Gedächtnis (1885; Memory, 1913). Human amnesia and its implications for memory organization were chronicled by Theodule-Armand Ribot in Les Maladies de la mémoire (1881; The Diseases of Memory, 1883). Sergei Korsakoff first documented alcohol-induced amnesia (Korsakoff syndrome), and the psychologist William James, in his classic work
The Principles of Psychology
(1890), distinguished primary and secondary memory, key concepts in the search for the engram, or memory trace.
The search for a mechanism began in earnest in the 1930s with the systematic experimental animal (as opposed to purely clinical and correlational) studies in the neuropsychology laboratory of Karl Lashley. Using lesions, he attempted to localize memory in particular brain structures, but he failed. This led him to conclude that the brain is equipotential: that the memory engram is distributed throughout the brain or is at least distributed equally throughout functional subunits of the brain.
Donald O. Hebb, who proposed the notion of reverberating circuits in the 1940s, was one in a long line of researchers who, in opposition to Lashley, believed that memory would be found in specific neural circuits. The role of the middle temporal cortex, in particular, was inadvertently revealed by William Scoville in the 1950s. In an effort to eradicate epileptic seizures in his now-famous patient H. M., he performed a bilateral temporal lobe resection. This resulted in permanent anterograde amnesia. Some researchers now propose that both distributed and localized accounts of memory are valid. That is, memory, in the general sense, may be distributed widely throughout the brain, but different areas may store different components of memory. The anatomical details of this theory remain to be worked out.
Rapid advances in computer-based neural network techniques seem to hold particular promise for yielding useful models of how memory works. Data from basic research on cellular memory mechanisms are translated into mathematical formulations. These in turn are transformed into computer programs that simulate or mimic the observed function of selected subsets of neural nets. Comparing the behavior of computer-simulated nets to actual nets permits refinement of the hypothesized mechanisms. Ideas emerging from this work—for example, the notion of parallel distributed processing (PDP)—will provide work for neurobiologists for years to come.
The possible role played by memory suppressor genes is among the newest areas of research into the regulation of memory. Most prior research centered on positive control of memory formation at the molecular level—the activation of genes for formation of neural circuits and resultant memory. Kandel has also found evidence for a negative control, using the products of memory suppressor genes. Products of these genes inhibit development of the synapse and prevent memory formation.
If indeed such regulation exists, pharmacological agents targeted at suppressor proteins might serve to increase development of memory. This would provide additional targets for reversal, or at least slowing, of the memory loss associated with neurodegenerative illnesses such as Alzheimer’s disease.
Bibliography
Abel, Ted, Kelsey Martin, Dusan Bartsch, and Eric Kandel. “Memory Suppressor Genes: Inhibitory Constraints on the Storage of Long-Term Memory.” Science 279 (1998): 338–41. Print.
Allman, William F. Apprentices of Wonder: Inside the Neural Network Revolution. New York: Bantam, 1990. Print.
Carlson, Neil R. Foundations of Physiological Psychology. 7th ed. Boston: Allyn, 2008. Print.
Eichenbaum, Howard, and Neal Cohen. From Conditioning to Conscious Recollection: Memory Systems and the Brain. New York: Oxford UP, 2004. Print.
Kandel, Eric. “The Molecular Biology of Memory Storage: A Dialogue between Genes and Synapses.” Science 294.5544 (2001): 1030–38. Print.
Radvansky, Gabriel A. Human Memory. 2nd ed. Boston: Allyn, 2011. Print.
Sweatt, J. David. Mechanisms of Memory. 2nd ed. Boston: Elsevier, 2010. Print.
Taylor, Annette Kujawski, ed. Encyclopedia of Human Memory. Santa Barbara: Greenwood, 2013. Print.
