Neurobiology of Learning, Emotion, and Affect

Neurobiology of Learning, Emotion, and Affect

John Madden IV

New York, NY: Raven Press, 368 pp, 1991

This book comes at a time of progress and exciting new discoveries in the field of research involving the neural basis of learning, emotion and affect. The study of learning and memory has flourished for many years within the neuroscientific community. Although long neglected by all but a handful of neuroscientists, emotion and affect have begun to attract attention. A clear summary of the status of this recent work is therefore particularly welcomed.

The book is divided into three parts. The first part is devoted to recent advances in the research of the neural mechanisms of fear conditioning, in both invertebrates and vertebrates. Most exteroceptive stimuli we encounter in our life are affectively “neutral”. However, they can take on emotional properties and elicit emotional reactions, such as defence or approach responses, through association with other stimuli or events that are affectively charged. It is possible to study experimentally the ways in which the brain forms such associations through the use of Pavlovian conditioning techniques, whereby a “neutral” stimulus, the conditioned, or conditional stimulus (CS), is paired with a biologically significant (affectively charged) stimulus, the unconditional or unconditioned stimulus (US).

Pavlovian conditioning procedures have been used to study the biology of learning in many species. Whether or not it is reasonable to speak of Pavlovian conditioning as a process of emotional learning is debatable for invertebrates, many vertebrates, and even many mammalian species, if emotion is restricted to subjective, experiential factors. However, if by emotion we also mean the neural mechanisms by which stimuli are evaluated for their significance (LeDoux 1990), it is possible to view all animals as engaging in a form of emotional processing. What is different about humans and possibly higher primates is that the emotional processing of stimulus significance becomes represented as conscious content. This view is advantageous in that it places emotional processes on a continuum and allows for studies of emotion throughout the animal kingdom.

In the Aplysia model, described in detail by Hawkins, the cellular mechanisms and neural loci responsible for associative learning are, without a doubt, understood more than in any other animal model. Are the same basic mechanisms discovered in this simple model in operation also in more complex organisms? Hawkins believes that classical conditioning in Aplysia, Hermissenda, cat and rabbit involve the same neural mechanisms. In Aplysia, the US plays a modulatory role in stimulating the facilitator neurons. The convergence of this input with the appropriate CS input (the one that was paired with the US) increases presynaptic facilitation and produces classical conditioning. A similar mechanism is acting in vertebrate models as well, Hawkins suggests. The aminergic and cholinergic systems in the vertebrate nervous system can behave like the facilitator neurons of Aplysia. One challenge for the future is to test this simple and attractive theory.

Another elegant invertebrate model is Drosophila melanogaster. An advantage of this model is the possibility of using genetic tools. In recent years, the isolation of mutations that affect an aversive conditioning task in the fruit fly has considerably increased our understanding of the molecular mechanisms of learning and memory. Tim Tully reminds us that with Drosophila research, we probably have the only evidence that short-term and long-term memory are indeed two genetically distinct processes.

Our understanding of the molecular and underlying elementary neural mechanisms which make learning possible in an individual is far less developed in vertebrate models. In recent years, however, simple vertebrate model systems have been worked out, and important neural structures and pathways essential in learning have been shown. It now appears that different types of learning are mediated by different but, in some cases, overlapping neural circuits.

One promising vertebrate model is the conditioned fear-potentiated startle paradigm. In this behavioral model, the conditioned stimulus (usually a light) is paired with a shock and startle response elicited by a noise burst in either the presence or absence of the light. If the startle response is greater when elicited in the presence of the light, the fear-potentiated startle has occurred.

The role of the amygdala in this reflex has been demonstrated by Davis and collaborators; in this book they review the current state of affairs. The amygdala is the centerpiece of the neural pathway involved in fear-potentiated startle, as well as in other fear conditioning situations (Kapp et al 1990; LeDoux 1990). Davis and collaborators show that the activation of the amygdala from the visual conditioned stimulus pathway triggers the startle reflex pathway. The central nucleus of the amygdala carries this out through its connection with the nucleus reticularis pontis caudalis (a nucleus in the startle circuit). Davis and colleagues believe that the visual input comes to the amygdala through cortical pathways. However, findings from our laboratory suggest that visual fear conditioning can be mediated by subcortical visual inputs to the amygdala (LeDoux 1990).

There is a general consensus about the involvement of the amygdala in mediating the acquisition of autonomic responses in the aversive classical conditioning. These responses are defined as “non-specific” and include heart rate, generalized motor activity (freezing, startle response) and skin resistance. They all develop rapidly after a few trials, sometimes requiring just a single contingent pairing. “Specific” responses, on the other hand, are discrete, skeletal muscle responses elicited by specific aversive stimuli. They are much slower to develop and are more specifically adaptive for the organism.

Steinmetz and Thompson describe how the cerebellum, in particular the interpositus nucleus, is an important site for the acquisition of these discrete, specific responses. Using a multi-technique approach (recording, stimulation, lesion studies) they developed a detailed anatomical map of the circuitry involved in the adaptation to aversive events for these behavioral responses. According to Steinmetz and Thompson, the cerebellum plays an “informational” role in the learning of the organism. It is activated in conjunction with the aversive system that involves higher brain regions (the amygdala, for example). These two systems in the brain are distinct, but they interact in the adaptive learning phase. The US pathway activates both systems. And while the cerebellum appears to play an important role in learning the specific response, recent studies also suggest that the amygdala is involved in the early phases of learning. The final attempts to relate these ideas to the popular Rescorla-Wagner learning theory.

Part II of the book deals with experimental situations in which the animal is allowed little or no control over the aversive stimulation How do the subjects cope with stressful events when species-specific defense responses are not helpful? What kind of behavioral and biochemical changes will stressful events elicit? Results from these experimental studies are particularly valuable for their clinical implications. Weiss, using an uncontrollable shock paradigm, suggests in fact that stressed animals show symptoms closely corresponding to those developed by depressed individuals. Moreover, the author proposes that stressful events (for example, uncontrollable shock) elicit both anxiety and behavioral depression. In his analysis, Weiss shows how the locus coeruleus seems to be a key structure in the neurochemical unbalance produced by the uncontrollable shock.

Cognitive deficits may also occur as part of the response to stress. Animals that are exposed to shocks that they cannot avoid or escape later fail to escape shock in a situation in which escape is possible. They also fall well behind control animals (that were allowed to escape or avoid shock in the first phase of the experiment) in Y-maze or water-maze learning tests. Some researchers have explained these results in terms of a simple motoric impairment: the shocked animals learn a coping response in the new situation, but they cannot perform it. Maier rejects this hypothesis; his work suggests that inescapable shock gives rise to learning deficits that cannot be explained by a motor activity deficit.

However, the impairment is not the result of associational learning between the first phase of the experiment and later test situations, as it was suggested in the early explanations of this phenomenon. What is impaired in these subjects, according to Maier, is the capacity to attend the salient external cues; the deficit is then a cognitive one, not an associative one (the learned helplessness hypothesis). Whether this cognitive deficit will be connected to some neural loci or biochemical systems will have to be determined in future studies.

More complex issues of depression and human affect disorders, which involve multiple interactions between emotional and cognitive systems operating at conscious and unconscious levels, are difficult to probe with vertebrates too distant from us in the evolutionary ladder and in social habits. Suomi describes a primate model of affective disorders, the separation model. Humans share over 90% of non-replicated DNA material with higher primates. In addition, the behavior of these animals is characterized by advanced and dynamic social interactions among members of the same species. Rhesus monkeys, in particular, have been studied extensively in the wild and in laboratory settings. Stressful events caused by social relationships among individuals closely resemble human social life. Separation from an attachment object, such as a mother or a loved conspecific, produces profound behavioral and physiological effects, both in the wild and in the laboratory. The evidence of individual differences in these animals makes this model particularly interesting. Not all subjects respond to the same social separation in the same manner. There is also consistency in this behavior in individuals. The same monkeys “at risk” are more readily aroused by and behaviorally fearful of other stressful events. Having individuated the groups “at risk” in the rhesus monkeys population, it may be possible to forecast or even prevent the affective disorder. The “separation/risk” model in monkeys, as Suomi defines it, seems to be extremely promising for improving our understanding of the psychobiology of human affective disorders.

There is evidence from this primate model that the noradrenergic system is involved in mediating depressive states. Matthysse reaches the same conclusion in his analysis of mood disorders. In particular, the locus coeruleus, an important source of forebrain noradrenalin, is advanced as a candidate for depression, as Weiss also proposed for the helplessness model. According to Matthysse, this nucleus becomes activated by excited unhappy memories, which are the result of early loss or other traumatic events. Cerebral activation is then reduced by the increased firing of the locus coeruleus. At this point, physostigmine, the cholinergic agent, enters the circuit to produce the physiological symptoms of depression. In Matthysse’s view, while the memories are the primary cause of noradrenergic changes, the biogenic amines are only the effectors in this theory. Matthysse points out that human studies suggest that decreased cerebral activation is produced by physiostigmine in normal subjects (for example, the subjects feel “apathy, slowness of speech and movement”). Physiostigmine physiological symptoms closely resemble those of depression. In his fascinating theory, however, Matthysse does not tell us where these memories may be stored in the brain or why, at a certain point, they overflow to reach the locus coeruleus. Also, it will be interesting to try to define the role of anxiety in this context. Under identical conditions of stress some people respond only with anxiety; do they have the capacity to deal with the unhappy memories activation in a different way? Interesting questions arise from this theory that only future experiments can elucidate.

Control over stressors is a critical factor that influences biological functions which regulate adaptive and maladaptive behaviors. Bandura states that the ability to control stress effects is the principal factor that makes a person cope with stress events. It is the ability to control the stress effects that prevents the release of stress-related hormones or the impairment of the immune system. With the support of human quasi-experimental studies, Bandura describes the effect of being able to gain control over the stressful situation. For instance, catecholamine levels in phobic patients dropped after allowing them to acquire controlling efficacy. Pain tolerance was increased in normal subjects when there was self-efficacy, even in those subjects to whom naloxone was administered to block opioid activation. Self-efficacy mechanisms play an essential role in the individual’s well-being. Bandura’s theory proposes an entire psychological approach to deal with stressful events that deteriorate our biological systems. This approach provides an intelligent stimulus to operate in the sphere of human maladaptive behaviors with “clean” psychological tools but with an attentive eye on the neurobiological domain.

Part III of the book explores broader issues in the field of emotion and affect and presents two models. Gray describes three systems that control emotional behavior. Each of these systems is associated with a particular area in the brain. One of these systems, the behavioral inhibition system, is thought by Gray to be centered around the hippocampal formation. This system is responsive to stimuli associated with punishment or with the conditioned aversive stimuli and is involved in increasing the level of arousal and increasing attention to the environment to cope with the new situation. Anxiolytic drugs affect the septohippocampal system, and lesions of this system lead to a behavioral syndrome similar to that seen after the administration of anxiolytic drugs. These results are the strong-hold of Gray’s theory. The amygdala, a structure that has long been considered to play a pivital role in fear, including fear learning, is not included in the behavioral inhibition system. It is involved instead in the second system, the fight/flight system, which is responsive to unconditioned punishment and non-reward, that is, to issue commands either for fight or for flight depending on the context and type of punishment received. Gray dismisses fear conditioning studies as irrelevant to the problem of anxiety and relegates them to elicitor of a more species-specific type of defence or attack responses. We do not agree with such a view. Nevertheless, Gray’s model is commendable in its breadth and attempt to synthesize divergent findings into a unified theory of anxiety. The second model uses the opponent-process theory to explain a large set of affective phenomena, taken from everyday life and experimental settings. This model states that it is possible to produce acquired motives by non-associative mechanisms alone. The mechanism is repetition, not conditioning (association). Solomon believes that his approach can explain food intake, drug addiction and related phenomena that Pavlovian conditioning is not able to explain. An important question that the author raises is whether or not the opponent-process states are predictable. He is convinced that they are and analyzes the process of food intake and obesity following the logic of opponent-process theory.

Neurobiology of Learning, Emotion, and Affect is a well organized, multi-disciplinary book. It presents a range of approaches and contributions. The literature on behavioral neuroscience, in general, has increased tremendously in the last few years. Moreover, there is a growing interest in research on emotion, which has been somewhat neglected in the age of cognition. This book will help bring research on emotion into the limelight. The book will be very useful as an introduction to research on the biology of emotion and learning and will also be useful as reference point for future research.