Yu. I. Aleksandrov

Learning and memory: traditional and systems approaches.

The aims of the present work were to consider the characteristics of learning and memory from the point of view of a systems approach and to compare this view with the traditional approach. Neuron activity is regarded not as a response to the synaptic influx resulting in excitation but as a means of altering the cell’s relationship with its environment, whose “action” is to eliminate discordance between the cell’s “needs” and its microenvironment. The neuronal mechanisms of learning and consolidation of memory are regarded not as formation of a stable increase in the efficiency of synaptic transmission in circuits of connected neurons, but as a system genesis event which confers new system specializations on neurons which do not have to be directly connected synaptically. The roles of the processes of selection, reconsolidatory modification of previously formed memories, gene activation, neurogenesis, and apoptosis in systems genesis occurring both in normal and pathological conditions are discussed. Individual development is regarded as a sequence of system genesis events. The systems approach is applied to the phenomenon of long-term potentiation. In conclusion, a scheme including different types and stages of memory formation is presented.

Long-Term Potentiation and Evoked Spike Responses in the Cingulate Cortex of Freely Mobile Rats

Long-term potentiation of synaptic efficiency is regarded as a major candidate for the role of the physiological mechanism of long-term memory. However, the limited development of concepts of the cellular and subcellular characteristics of the induction of long-term potentiation in animals in conditions of free behavior does not correspond to the importance of this question. The present study was undertaken to determine whether the characteristics of potentiation in the cingulate cortex in response to stimulation of fibers of the subiculo-cingulate tract are truly long-term, i.e., develop through all known phases and last at least 24 h, in freely moving animals. In addition, the study aims included identification of the effects of application of blockers of different types of glutamate receptors on the development of long-term potentiation and identification of the characteristics of spike responses of single cingulate cortex neurons to stimulation of the subiculo-cingulate tract. Long-term potentiation, lasting more than 24 h, was obtained in freely moving adult rats not treated with GABA blockers. Injection of glutamate NMDA synapse blockers led to significant decreases in evoked cingulate cortex potentials in response to test stimulation. Activatory short-latency spike responses were characterized by a low probability of spike generation, and this increased with increases in the stimulation current. These data demonstrated that it is methodologically possible to compare, in freely moving rats, the involvement of individual neurons in the mechanisms involved in learning one or another type of adaptive behavior and the dynamics of their evoked spike activity during the formation of long-term potentiation.

Neuron activity in the anterolateral motor cortex in operant food-acquiring and alcohol-acquiring behavior

The interactions of the neuronal mechanisms of food-acquiring behavior and newly formed operant alcohol-acquiring behavior were studied by recording the activity of individual neurons in the anterolateral area of the motor cortex in chronically alcoholized rabbits. Adult animals learned food-acquiring behavior in a cage with two feeders and two pedals, in the corners (the food in the feeders was presented after pressing the corresponding pedal). After nine months of chronic alcoholization, the same rabbits learned an alcohol-acquiring behavior in the same experimental cage (gelatin capsules filled with 15% ethanol solution were placed in the feeders instead of food). Analysis of neuron activity showed that the set of neurons involved in supporting food-acquiring and alcohol-acquiring behaviors overlapped, though not completely. These experiments not only help us understand the neuronal mechanisms of the newly formed and the previously formed behaviors, but also facilitate the development of concepts of the similarity of the neuronal mechanisms of long-term memory and long-term modifications of the nervous system, occurring in conditions of repeated intake of addictive substances.

Distribution of Behaviorally Specialized Neurons and Expression of Transcription Factor c-Fos in the Rat Cerebral Cortex during Learning

Studies of learning are now performed at very differentlevels, including the neuronal and molecular biological lev-els. Integration of data obtained at these two levels canfacilitate the development of overall concepts of processesoccurring in the brain during learning.Data obtained using neurophysiological methods pro-vide evidence that learning is mediated by a process of“behavioral specialization” of silent reserve neurons [1, 3,5, 7]. According to the systems-selection theory of the for-mation of a new behavioral act, learning involves the for-mation of a new functional system, i.e., systemogenesis; atthe neuronal level, this corresponds to the formation of neu-ron specializations for this system [7]. Neuron specializa-tion consists of the appearance of activation of previously“silent” neurons every time the relevant formed behavioralact takes place.It has repeatedly been demonstrated that differentbrain structures are characterized by different patterns ofbehavioral neuron specializations [1]. Thus, the motor cor-tex is dominated by neurons specialized with regard to sys-tems formed at the early stages of individual development:so-called old system neurons, for example, “movement”neurons or “food taking” neurons. The cingulate cortex isdominated by neurons specialized with regard to new sys-tems formed when animals learn in an experimental cage,for example, “pedal-pressing” neurons [4]. Comparison ofthe patterns of specialization in the retrosplenial area of thecingulate cortex and in the anterolateral area of the motorcortex in rabbits shows that the number of “new” neurons inthe former is an order of magnitude greater than in the lat-ter. It has also been demonstrated that the patterns of spe-cialization in the cingulate cortex are similar in rabbits andrats: “new” neurons predominate in both species [3]. Fur-ther studies reported by Gavrilov et al. have demonstrated asimilarity between the patterns of specialization in themotor cortex of rats and rabbits. Thus, the ratio of the pat-terns of neuron specializations in rats is the same as that inrabbits.The processes underlying the formation of neuron spe-cialization are evidently based on long-term changes in cellfunctions and cell connections, which must require activa-tion of gene expression. Learning has been shown to inducea cascade of molecular rearrangements in neurons, andexpression of early genes has been shown to be one of thecritical elements of these modifications [2]. Induction of thec-fos gene during formation of a new behavior, this beingone of the main members of the immediate early gene fam-ily, varies in different brain structures. Zhu et al. [9] haveestablished that the set of structures activated in the ratbrain, demonstrated by immunohistochemical mapping,coincides with the set of structures identified as activated byrecording of neuron activity during presentation of familiarand unfamiliar objects. These points suggest that differ-ences in the levels of gene expression between structuresmay be associated with the different contributions of thesebrain structures to the process of neuron specialization dur-ing learning. Confirmation of this suggestion and improve-ment of our understanding of the molecular biological basesof the formation of neuron specialization require compar-isons of the numbers of neurons expressing early genes in agiven brain structure during learning with the level ofinvolvement of that structure in the formation of new spe-

Role of goal in determination of neuronal activity of the rabbit motor and visual cortical areas

Conclusions1.In different behavioral acts (grasping a carrot and plastic), which can be characterized as the same movements in the same environment, the compositions of the activated neurons of the motor and visual cortical areas differ. The occurrence of activations of cortical neurons in behavior is not determined uniquely by the parameters of the movements and environment. This ambiguity is explained by a change in the motor and receptive fields of the neurons.2.The “total pictures” of the activity of neurons both of the motor and visual cortical areas (percent of neurons activated in each of the stages of the behavioral act) are similar in the behavioral acts of grasping plastic and carrot.

Neuron Activity in the Retrosplenial Cortex of the Rat at the Early and Late Stages of Memory Consolidation

Activity was recorded from single neurons in the retrosplenial cortex during performance of an operant food-procuring behavior in two groups of rats; in the first six days after training to this behavior (group 1) and one week later, during which the animal did not perform the learned skill (group 2). At the same time, these groups showed no significant difference in the percentages of neurons specialized with respect to the learned behavior; in group 1, 40% of the cells of this category showed activation occurring in 80–90% of performances of the specific act, and not in all (100%), which was significantly different from the proportion of such cells (4%) in animals of group 2. All neurons with less than 100% activation at the early post-training stage were specialized with respect to the most recent act in the training history: approach to and pressing of the pedal. It is suggested that at the first stages of consolidation of the operant skill, its realization may occur by means of a variable set of cells in the retrosplenial cortex specialized with relative to the system of new behavioral acts.

Experience of a First, “Whisker-Dependent,” Skill Affects the Induction of c-Fos Expression in Somatosensory Cortex Barrel Field Neurons in Rats on Training to a Second Skill

We report here studies in which a controlled history of sequential learning of two skills was created with the aim of identifying the pattern of activation of neurons for the first skill on formation of the second. Animals were initially trained to an operant drinking behavior requiring use of the whiskers on the left or right side (a “whisker-dependent” skill), which was followed by training to a food-procuring pedal-pressing skill not requiring use of the whiskers. The results showed that training to a food-procuring skill induced c-Fos expression in a significantly larger number of neurons in the barrel field in animals previously trained to the operant drinking (whisker-dependent) skill than in the analogous area of control animals previously trained to a non-operant drinking skill. Our data suggest that activation of c-Fos expression on repeat training also occurred in those neurons which had already become specialized in relation to the first, whisker-dependent, skill.

Specialization of motor cortex neurons in rabbits under normal conditions and after ablation of the visual cortex.

The activity of motor cortex neurons in instrumental food-acquisition behavior is compared in two control rabbits and in three rabbits after bilateral ablation of the visual cortex. Although the same types of neuron specialization were found in the experimental and control animals, their numerical ratio differed markedly in two out of the three experimental rabbits in comparison with the controls: the number of neurons activated in the act of seizing food was halved, while the number of neurons activated in connection with acts of instrumental behavior doubled. The similarity of the processes underlying behavior learning and recovery is discussed.

Long-lived newly formed neurons in the mature brain are involved in the support of learning and memory processes

Using immunohistochemistry, we stained cells with BrdU (for detection of newly formed cells), NeuN (a neuronal specific marker), c-Fos (a marker of neuronal plasticity), and ApoDNA (marker of apoptotic cells) in the cerebellar vermis, dentate gyrus and CA1-CA4 fields of the hippocampus, motor, and retrosplenial cortex of the right and left brain hemispheres in adult rats. Animals were trained in spatial skills in the Morris water maze or were subjected to a soft forced-swimming test 6 months after a 14-day intracerebral administration of BrdU. Significant differences in the amount and composition of the labeled cells in the trained and control rats were found. The relationship between the number of new neural cells and the parameters of the formation of long-term spatial memory was determined. The results indicate that the newly formed neurons with an age of 6 months, as well as the neural cell precursors of the relevant brain structures, are selectively involved in the support of long-term spatial memory.

Dopaminergic mechanisms underlying involvement of the prefrontal cortex and amygdala in acoustique startle modificftion during the action of stimulus associated with positive reinforcement

We studied the neurochemical mechanisms underlying the action of the conditioned stimulus associated with positive reinforcement on the amplitude of the acoustic startle response (ASR). To this aim, we injected a blocker of dopamine D2 receptors, sulpiride, into the rat medial prefrontal cortex or basolateral amygdala during the repeated procedure of ASR habituation. We have shown that, after the injection of saline to control rats into both the prefrontal cortex and amygdala, the change in ASR amplitude after an increase in illumination depends on the preliminary learning of a behavioral skill associated with a conditioned light signal. We found that if an acoustic signal appeared immediately after the change in the illumination the blockage of D2 dopamine receptors in the medial prefrontal cortex induced an increase in ASR amplitude. This disturbed the habituation of animals that were preliminarily subjected to associative training before this procedure. The injection of sulpiride into the basolateral amygdala during the increase in illumination did not affect the dynamics of ASR amplitude. This effect of the drug did not depend on the preliminary acquirement of positive conditioning to light. Our data suggest that D2 dopamine receptors of the medial prefrontal cortex, but not the basolateral amygdala, are involved in the reproduction of previously acquired skills associated with positive reinforcement under conditions of current defensive behavior.