At present no specific investigations are examining how specific brain neurons are damaged during homeostatic disturbance throughout motivational behavior. Such damage may consist of changes in a cell’s morphology, ion metabolism, electrical activity, and other factors that prevent normal functioning of the cell, and these may eventually lead to cell death. Direct evidence of damage, such as cellular swelling, loss of integrity of the membrane, etc. is difficult to reveal in vivo. Nevertheless, the correlation between an outcome of motivational excitations and cellular damage is sometimes so obvious that indirect evidence is presented in many studies. On the one hand, there are the evidences of cellular damage during motivational behavior, and on the other hand, artificially induced damage provokes or augments motivation. As well, the same alterations of cellular homeostasis lead to damage and to a rise in motivations. Accordingly, excitation, the basic property of nerve action (and apparently the basis of motivation), causes damage. Specific examples are given below. We tried to detect primary signals, that initiate motivations and specific metabolic pathways that are typical for development of motivations.
Defensive motivations
The origin of defensive motivations is associated with a pain, fear, anxiety, aversion, disgust, and other signals of menacing danger. Traditionally, the defensive motivation is considered as negative. Defensive or negative motivations may be displayed in the active or passive form: one may passively avoid or actively liquidate the danger. In fact, any motivation is negative and needs satisfaction, which always has a positive meaning.
Attentive studying of neuronal activity during aversive and defensive reactions has revealed a wide distribution of excitations in the brain, but specific aspects of defensive behavior correspond to activity of relatively local areas within the brain. Defensive motivations are first of all connected with the amygdala [297, 556]. Different types of fear-conditioned behavior are mediated by separate nuclei within the amygdala: the central nucleus mediates suppression of behavior elicited by a conditioned fear stimulus, while the basolat-eral amygdala mediates avoidance of conditioned aversive stimulus [624, 556]. Some features of defensive behavior are also connected with other areas of brain. For instance, neurons in the rostral ventromedial medulla exert a facil-itatory influence on spinal nociception "on-cells" and secondary hyperalgesia in acute inflammation [1368], neurons in the spinal cord contribute to chronic injury-induced neuropathic pain [1161], activation of neurons within the peri-aqueductal gray in the midbrain may be induced by a predator [7] and direct activation of these neurons leads to the fear reaction [838]. In addition, transmission and sensation of primary events for the development of avoidance and their effects are broadly scattered within high neural centers of the brain, particularly in the cortex, hippocampus and cerebellum [117, 840, 556, 628]. Different areas of brain are activated during early and late phases of acquisition of fear conditioning, during extinction and recall and, in particular, the ventral hippocampus may play a role in resolving conflicts, and specifically may be relevant to "safety signaling" [556]. Thus, the neurons that feel danger or pain, the neurons controlling action directed to avoidance and the neurons that feel the coming of safety, these are commonly different cells.
Aversive stimuli are inalienable attributes of defensive motivation. A distinctive feature of aversive stimuli is their resistance to habituation and responses to such stimuli even increase after repetition [248]. The origin of defensive motivation is directly connected with the danger of injury. Pain is important, although not an obligatory component of defensive motivation. It is induced by tissue-damaging stimuli and they evoke injury in the central neural system itself [840], that is, defensive motivation is initiated by means of damage-related events in the brain, not only in the body7.
Glutamate receptor activation is an essential component of both the fear reaction and hyperalgesia8. Such an impact to alterations of cellular metabolism inevitably leads to deviation of the ion equilibrium from the optimum to excitotoxicity. Moreover, defensive motivation and negative experience left material traces in the brain, which are observed after cell injury9.
Secondary effects of aversive stimuli increase the number of gap junctions, enhance cyclic AMP concentration [297, 373, 844, 1293] and produce the inflammatory mediators, such as prostaglandins, adenosine, etc. Cyclic AMP pathways are activated and also augmented during neuropathic pain [1161] and by a predator [7]. Inhibition of the cyclic AMP signaling pathway in-creases anxiety-like behaviors [936, 1296, 87], while augmentation of the cyclic AMP system produces hyperalgesia [1150] and antagonizes the analgesic response to morphine [121]. Fear reaction and inhibitory avoidance, such as aversive stimuli, are connected also with activation of synthesis of nitric oxide [838] and decrease in Na+, K+-ATPase activity [1365]. For instance, spread of the pathophysiology of neuropathic pain includes upregulation of nitric oxide pathways in axotomized neurons [1439, 1381]. These secondary consequences of aversive stimuli are usually observed during intensification of homeostatic compensation.
Not only pain, but also the state of anxiety is related to cell damage. In depressive patients, the prefrontal cortex loses neurons, while the number of cells in the hypothalamus increases [1009]. Growth of neurons as well as their death can be a direct result of damage to them. Anxiety-relevant substances, serotonin, GABA, corticotrophin-releasing hormone, interferon y, and neural adhesion molecules play important roles in neuronal development, proliferation, cell-to-cell communication and intracellular signaling [1353]. Thus, defensive motivations are directly connected with cell damage, while avoidance of defensive motivation protects brain cells. Moreover, not only does negative motivation induce neuronal damage, but damage induced by the administration of cytokines [343] and excitatory amino acids [906] also increases defensive withdrawal. Neuronal-glial interactions, resulting in an immediate release of glutamate and discussed above as damaging, are broadly involved in the creation of pathological pain and passive avoidance learning [697, 1326]. Glial cells participate in defensive behavior. Both microglia (in the initial phases) and astrocytes (in the later phase) are important in neuropathic pain. Brain injury and some pain states leads to hypertrophy of glial cells. Concomitantly, production of a variety of chemokines, and other pain-producing substances increases. Their expression sensitizes primary afferents and dorsal horn neurons and therefore may contribute to neuropathic pain after peripheral nerve injury. Prevention of glial activation by pharmacological means produces an antinociceptive effect, while spinal administration of the pain mediator substance P produces robust activation of microglia [824]. Cells of microglia sense homeostatic disturbances and even relatively minor deviations from normal neuronal activity [1109]. The microglial response to danger might result from membrane breakdown products, the extracellular presence of cytosolic compounds, abnormally processed or aggregated proteins, or an abnormal abundance of the excitatory transmitter, glutamate. The executive functions of microglia can change not only in magnitude but also in quality and variability of microglial activity. Their activity is not a reflection of stimulus strength or persistence; rather, it is determined largely by the nature and context of the stimuli and the intracellular signal transduction pathways that they activate [1109].
Brain possesses special means that counteracts these negative senses in the system of endogenous opioids. Danger in the environment compels a living being to active actions, but strong negative feelings may prevent defensive behavior. Although pain or fear signal about a threat, when the danger appears in the surroundings, these feeling must not preclude an avoidance of danger. Therefore, after revelation of a threat, temporal relief allows the brain to work against damage. The opioid system performs this function well [910, 1293, 141]. Endogenous opioids work both for pain relief and to ameliorate negative feeling.
Respiratory motivation
Breathing is usually an unaware action. It continues during sleep and during a fall in consciousness, but breathing during alert behavior is, as a rule also unconscious and is controlled by an autonomous central pattern generator of rhythm. Respiratory motivation sharply increases and becomes conscious after a disturbance of gas exchanges, oxygen and carbon dioxide. Thus, respiration exhibits an example of parallel occurrence of two or more motivations, at least till they become conscious. Breathing is necessary for oxidative phos-phorylation that uses energy released by the oxidation of nutrients in order to produce molecules of ATP that provide a highly efficient way of storing energy, used for metabolic needs. The substrate for respiratory motivation lies in the pre-Botzinger complex, a cluster of interneurons in the ventrolateral medulla and in the brainstem medulla. Neurons in the ventral medulla contain chemoreceptors, regulating breathing,, while the central command mechanism for breathing rhythm lies in the hypothalamus [378]. Preinspiratory neurons in the medulla are opioid-insensitive, while neurons in the pre-Botzinger complex are opioid-sensitive.
Respiratory motivation is connected to a decrease in the pH of the brain-stem and an increase in gap junctions [1220, 125, 81, 187, 926]. Both CO2 and pH sensors regulate breathing and both intra- and extracellular pH appear to serve as the proximate stimulus for central chemoreception. pH is sensed by many proteins which are located in low resistance gap junctions. Locations of the central chemoreceptors only partially coincide with location of the central pattern generator for breathing [378]. Neurons controlling inspiration and respiration are different cells, and the sensory neurons for respiratory motivation are, evidently, not the same cells that produce breathing.
In the basal state, all types of respiratory neurons in the ventral respiratory group of rats receive GABAergic and glutamatergic inputs. These neurons display periodic waves of inhibitory and excitatory postsynaptic potentials [907]. Although both glutamatergic and GABAergic mechanisms are involved in setting the resting respiratory rhythm, excitatory amino acids play a leading role in respiration [438]. During a shortage of oxygen, glutamatergic neurons of the ventral medulla are vigorously activated by CO2 and by acidification, whereas serotonergic neurons are not [867]. Inspiratory bursts in the pre-Botzinger complex in mice depend on a Ca2+-activated cation current linked to glutamate receptors [930]. During a powerful rise in respiratory motivation, such as in the acute ventilatory response to hypoxia, Ca2+ channels are activated, intracellular Ca2+ is accumulated, taurine falls and glutamate release is enhanced in the medulla as well as at the other sites associated with respiration [558]. We may conclude that shortage of oxygen leads to disturbance of the energetic balance in specific cells and, as a consequence, to Ca?+ dependent damage of these neurons. In correspondence with the phased change of neuronal activity after damage, the ventilatory response to acute hypoxia in a mammal is biphasic, an initial hyperventilatory response is followed by a reduction in ventilation within 23 min below the peak level [558]. In particular, agonists to excitatory amino acid glutamate in the doses in which it facilitated respiration, destroy the respiration-related neurons in the nucleus parabrachialis [542].
Experimentally evoked hypoxia is the main method of cell death investigation. Direct indications of excitotoxic damage are observed during the respiratory drive [125, 187, 849]. Opioids, which protect neurons against damage, suppress respiration, while substance P stimulates respiratory frequency [81, 960]. Metabolic coupling between the glia and neurons is necessary for maintaining rhythmic respiratory activity in the medulla [154, 189, 557, 622]. Dynamic change in baroreceptor-sympathetic coupling is observed in the central pattern generator during the respiratory cycle [455]. Probably, not only in the case of pathologically augmented ventilatory response, but during normal respiration, inspiratory and expiratory phases are connected with the cyclic changes of injurious and protective influences.
Temperature regulation
Warm-blooded beings maintain their body temperature within certain limits, that is, decrease or increase body heat. In both cold and hot environments, a thermal homeostasis spends effort for keeping temperature constant. The temperature of a body may be regulated not only by a change of metabolic rate, but by a choice of the optimal behavior: animals search for a warm or cool place, decrease or increase a motion and irrigate or sun themselves. A behavioral means instead of homeostatic regulation is used also by coldblooded animals.
Precise thermoregulation is mediated by the preoptic and anterior hy-pothalamic areas. Local warming and cooling of this zone leads to lowering or raising body temperature. There are cold-sensitive effector neurons for heat production and warm-sensitive effector neurons for heat loss. Warm sensitivity results from inherent changes in cellular activity, while cold sensitivity is primarily due to inhibitory synaptic input from nearby warm sensitive neurons [498]. There are thermoresponsive Na+ conductances in warm sensitive neurons and some thermosensitive channels are also responsive to low pH. Several types of channels are tuned to various temperature setpoints. Since some neurons are sensitive to warming, while others to cooling, and optimal temperature may vary for different neurons, initiation of a motivation may be a united activity of a group of cells. In particular, temperature regulation is controlled by a system of neurons and not by a dispersed assemblage of autonomous regulators.
Temperature regulation is dependent on cellular events connected with damage. Cytokines cause a rise in body temperature and a clear correlation exists between thermosensitivity in the ventromedial preoptic area and neural responses to some cytokines [498]. Angiotensin II, a motivationally-relevant substance, connected with water metabolism also activates preoptic neurons [1243]. These neurons, which sensitive to temperature and osmolarity, are located side by side. Termoregulation, which is related to liquid homeosta-sis, also alters membrane properties. Warming decreased the amplitudes of the action potentials in the central termosensitive neurons and affected Na+ channels [498]. Changes in the local environment of the hypothalamus (i.e. osmolality, glucose concentration, or the presence of reproductive steroids), have the greatest affect on the activity of warm sensitive neurons [498]. Probably, neurons that are sensitive to primary signals of homeostatic disturbance are susceptible to transient damage and this may be a cause of interaction between different motivations.
Drinking motivation
Thirst is one of the most important biological motivations. Nevertheless, investigations of drinking motivation are still almost at the beginning of their efforts. In particular, for our knowledge, nobody has compared central drinking mechanisms for animals living in water and in an air environment. This comparison could clarify how a morphologic substrate of motivation depends on the accessibility of reward, but this is a task for future investigations.
The need for liquid evokes dehydration and induces thirst. Circumventric-ular organs have a leading role in the control of body fluid regulation, and have significant involvement also in temperature control, feeding behavior and reproduction [268]. Initiation of drinking depends on the hypothalamic sub-region of circumventricular organs, the subfornical organ, the paraventrical nucleus, the median preoptic nucleus and the arcuate nucleus [430]. It is considered highly unlikely that separate subpopulations of neurons exist in each of the circumventricular organs which subserve the control of each physiological system. Rather, it would seem likely that neurons in these circumven-tricular organs contribute to integrated control of complex interwoven physiological systems [268]. The neurons here are sensitive to angiotensin II and apelin, motivationally-relevant substances that stimulate drinking motivation and have effects on fluid and electrolyte homeostasis, osmotic pressure, body fluid homeostasis and they lose water during thirst [732, 1243, 25, 1018, 430]. Salt loading and dehydration increase the level of galanin [703]. Central os-moreceptors are located in the magnocellular neurons of the circumventrisular organ and they are sensitive to glutamate and Na+ [573]. During dehydration, astrocytes change their coupling with neurons in the hypothalamus [697].
Change in osmolarity affects also neurons out of the thirst center, for instance in the hippocampus [567], but magnitude of the response is usually low.
Neurons in the thirst centers are subjected to deleterious influences during hyperosmotic conditions. They are activated due to a decrease in the hyper-polarizing influences of taurine, which participates in the compensation of cellular injury in neurons and the activation of glutamate afferents (from sensory osmoreceptors). Taurine efflux may be triggered as a response to damage-induced cell swelling [1085] and participates in the controls of osmotic balance, heart rhythm, respiration, blood pressure, body temperature, motor behavior, feeding behavior, alcohol addiction, and learning [29, 154, 189, 619, 1090]. Unfortunately, study of the physiological role of taurine is now in its infancy.
Besides a fall in taurine, Na+ channels augment damage, and mechanore-ceptors, which are activated during cell swelling, transmite osmotic perturbations into electrical signals [573, 697, 1307]. Ca2+ channels play important roles in the angiotensin II-induced drinking behavior [1434], but large doses of angiotensin II (^100 nM) protect dopaminergic neurons against neurotoxicity [495]. Thirst centers support an optimal level of osmolarity and both hyperosmotic and hypoosmotic conditions increased the cytosolic Ca2+ level [88]. In rats that feel chronic thirst, an excitability of magnocellular neurose-cretory cells in the supraoptic nucleus of the hypothalamus was enhanced [1215]. In such rats were found long-term changes in Na+ channel expression and an increase in transient Na+ current [498]. Thus, during drinking motivation, neurons in the thirst centers display alterations of cell metabolism which change their liquid and further ionic homeostasis. This alteration is, apparently, typical for cell injury.
Feeding motivation
Performance of physiological functions needs energy. Plants receive energy from the sun. Daily appearance of the sun does not depend on plants, but sometimes even plants actively search for energy, as does the sunflower. Animals obtain energy from food and have to search for the source of food. Therefore, the search for food is an example of intended actions. Food consumption recovers exhausted energetic homeostasis. During voluntary feeding, intertrial intervals, which are an indicator of the level of hunger, are directly related to homeostatic needs (dependence on levels of glucose and insulin, etc.) of the animal to consume food [50].
Thus, for feeding motivation, the metabolic signal is the energetic misbal-ance of the organism. Energy metabolism [285, 892] and glucose, the main source for ATP production [201], decreases spontaneously for several minutes preceding the onset of a meal. Fasting suppresses oxygen consumption and, in the short-term, feeding increases it [499]. Neurons in the motivation center of feeding use glucose in their own metabolism [201, 1191] and glucose deficiency in the plasma stimulates eating and damages neurons [1341].
Metabolic sensing neurons control the energy homeostasis and integrate a variety of metabolic, humoral and neural inputs from the periphery. Such neurons, originally called "glucosensing", also respond to fatty acids, hormones and metabolites from the periphery [726].
Feeding motivation is mainly connected with the lateral hypothalamus [1341], but other brain areas are also involved in feeding behavior, particularly, the orbitofrontal cortex, insular cortex and amygdala. Nevertheless, participation of the lateral hypothalamus is the most prominent. At present, more than ten motivationally-relevant substances are known that regulate feeding motivation and satiation [174]. For example, neuropeptide Y and insuline-like receptor signaling systems are essential for the dynamic regulation of (even) noxious food intake [1363]. Insulin receptors are ultimately expressed in the arcuate nucleus [131] and food restriction increases the synthesis of neuropeptide Y there, but not in brain areas outside of the arcuate nucleus [970].
It was shown that neurons in the feeding centers generate activity in correspondence with the feeding cycle [50]. The most feeding-related neurons preferentially responded to a unique phase within a cycle (hunger-satiety-hunger), but neuronal populations integrated single-unit information, which reflected the animal’s motivational state across the entire cycle. Mean population firing rates can more efficiently represent phases of a feeding cycle than individual neurons. It was demonstrated that the average performance of individual neurons in the lateral hypothalamus during voluntary eating to satiety was in better correlation with the performance of the entire corresponding ensemble than the performance of neurons in the orbitofrontal cortex, insular cortex or amygdala [50]. Hence, taking into consideration that the large majority of satiety-modulated neurons preferentially responded to a unique phase of the feeding cycle, feeding behavior goes on through activity of a majority of neurons, which are organized in the system, while each one neuron executes narrow functions and does not control entire feeding cycle. This means that feeding motivation of an entire animal is based on an interaction between different neurons and is maintained by the neural network. If a whole motivation would be organized on the foundation of elemental motivations of neurons, one may expect that all phases of motivational behavior be displayed at the level of a neuron. Sometimes this is true, but this is not the rule. We had been noting a non-coincidence between behaviors of separate neurons with phases of an entire motivational behavior, describing other motivations.
Why doesn’t each feeding neuron participate in the entire feeding cycle? A whole feeding behavior is probably too complex for a single cell. Even if a motivational behavior for each neuron is a closed-logic cycle and consists of detection of a metabolic flaw, completion of a goal-directed action and consumption of the reward, these phases of single neuron behavior might not coincide with the phases of an entire behavior. For instance, action of the neuron may be directed to particular side of whole behavior of the animal, while reward may be received as an opioid or GABA signal and not as a detection of true reward from the environment. Nevertheless, although whole feeding motivation (and not only feeding motivation) is organized as the activity of an entire network, elemental motivation of neurons, we will demonstrate, plays an important role in a behavior.
A large body of data indicates that neuronal damage is involved in feeding motivation. The level of motivation is connected with the status of excitation in the specific neurons. Stress and pain can also paradoxically stimulate eating [910]. Neurons in the feeding-related brain areas displayed higher activity levels during hunger and decreased activity levels during satiety [50]. Injection of low doses of excitatory amino acids into the lateral hypothalamus elicits eating [1168] at the same concentration at which they induce neuronal damage [801] and increase energy expenditure [65]. Glutamate stimulates appetite so efficiently that it is used in commercial fabrication of food, in spite of its excitotoxicity. Glutamate was ultimately removed from American baby food products, but it is still present in many other prepared foods [1045].
Initiation of feeding is dependent upon the activation of the NMDA [717] or AMPA [1168] glutamate receptors in the lateral and perifornical hypothalamus. Treatment of the nucleus accumbens core with the antagonist to gluta-mate NMDA receptors impairs response-reinforcement learning in the acquisition of a simple lever-press task to obtain food [615]. Once the rats learned the task, the antagonist had no effect, demonstrating the requirement of receptor-dependent plasticity in the early stages of learning. After blockage of NMDA receptors in the nucleus accumbens core, rats had normal feeding and loco-motor responses and were capable of acquiring stimulus-reward associations. Hence, glutamate receptors in the nucleus accumbens core are not absolutely necessary for feeding, but their blockage weakens feeding motivation, so that animals do not make efforts to feed.
Feeding behavior cannot be explained by direct interaction of synaptic processes with an excitable membrane, as, for example, reactions to signals, and simple reflexes are explained by transmission of excitations from one neuron to others. Participation of slow metabolic processes, second and retrograde messengers is a necessary link in organization of feeding. Feeding behavior activates the cyclic AMP system, G proteins, NO, the axonal sprouting and neural-glial interactions10. We may suppose that feeding behavior is connected with the metabolic pathways related to a tension of physiological functions such as those that are activated during development and damage-protection. Feeding motivation, evidently, evokes a transient cell injury in feeding centers, while starvation can be connected with irreversible neuronal damage.
Sexual motivation
Sexual motivation is not an obligatory attribute of individual life, as is evident from longevity after castration. Correspondingly, the origin of sexual motivation is determined by the hormonal status of the organism and by outer stimuli. Initiation of sexual motivation is not really caused by any need of the organism and does not depend on homeostasis [1344]. Its metabolic signal acts as if external to the brain: it is generated by neurosteroids, without a metabolic mismatch of vitally important demands, but creating its own mismatch and demand. Steroid hormones affect metabolism and distress homeostasis of specific neurons and the recovery of homeostasis occurs after sexual satisfaction. Nevertheless, although sexual behavior is not directly connected to any metabolic conflict, the brain sites identified as potential areas for control of sexual function have been implicated in a variety of homeostatic functions [822]. By the way, strictly speaking, with reference to an origin of motivation, sexual and defensive motivations are close enough, since defensive motivation also arises without exhausting inner resources and metabolic disturbance, but the disturbance comes later after action of an external, with respect to brain, force. May be this is the reason why both these motivations have common executive centers, the amygdala [297, 948, 279, 383, 703]. Nevertheless, although much of the scientific literature stresses the role of the amygdala in negative affects (e.g. fear, anxiety and disgust) and defensive learning, amygdala is involved also in positive and appetitive learning [552], and neurons representing positive and negative values (licking and blinking behavior) did not demonstrate clear anatomical clustering [944].
Although the basic mechanisms of sexual motivation (and any other motivation) are similar for both high and primitive animals, sometimes one considers that animals do not feel pleasure from sex and mate only for reproduction, with the exception of the high mammalian. We think this looks somewhat narcissistic. Certainly, we cannot examine what an animal experiences. May be, animals do not have pleasure from feeding, too, and feed only for filling up their supply of ATP. For instance, execution of sexual behavior can function as a reward [9]. After all, one cannot even objectively check what his spouse feels, and must resort to believe her words.
Sexual behavior in mammalian is mediated by the medial preoptic nucleus, the amygdala, the piriform area, the arcuate nucleus, and the red nucleus [136, 288, 846]. The neurons of sexual centers demonstrate powerful excitation during mating [846, 1134]. The medial preoptic area is a critical regulatory site for the control of male sexual behavior [333]. Neurons in a medial preoptic area of males containing androgen receptors are active during mating and a third of these neurons contain estrogen receptors, too [802, 496], which also influence male copulatory behavior [243]. The medial preoptic area is crucially involved in consummatory aspects of sexual behavior. By contrast, ventral striatal mechanisms primarily affect appetitive sexual responses [369] and thus again demonstrate that different aspects and phases of motivations are controlled by different brain areas. Sexual motivation increases also in neuronal excitability in the hippocampus [697], which is affected also during other motivations. Male copulatory behavior in the mollusk increases excitability of neurons in the anterior lobe, enhances AP broadening and activates inwards currents that are carried by Na+ and by Ca?+ ions [1154]. Powerful excitation in sexual centers may evoke excitotoxic damage.
Influences of sex hormones to neurons are in correspondence with the effects of natural sexual motivation. Sex hormones affect development of sexual motivation and influence the reception of pain and anxiety, indicating their injurious actions [431, 1053]. Steroid sex hormones modulate the activity of receptors of excitatory amino acids, GABA receptors, voltage-gated Ca2+ channels [118, 822], increase the number of neuronal-somatic gap junctions [441] and play a regulatory role in expression of the calcium-binding protein calbindin expression in sexual centers [1186]. The sex hormone estrogen also directly acts on damage-protection processes, such as ordinary sexual behavior. It also influences neurogenesis, the balance between neuronal survival and death, cell migration, dendritic and neurite growth, synaptogenesis, and glial differentiation and neuronal morphology, and alters membrane channel activity [1346].
Sex hormones, which are in fact multiple growth hormones, intervene in the processes of injury-protection of various brain cells and can disrupt metabolism [392, 1174, 1327]. The short-term effect of estradiol on neurons is non-protective and excitatory [118], while slower (several hours) actions of estradiol are thought to be protective in nature [1347]. Protection by estradiol follows late deleterious receptor-mediated effects [901]. Testosterone also augments neuronal damage after axotomy in young male rats, but protects neurons of female rats [1388]. Middle doses of testosterone (1 – 100 nM) also have neuroprotective actions [978].
Progestins and androgens alter sexual receptivity and anxiety through rapid membrane effects and by modulating GABA^ receptors [431, 1053]. For instance, excitatory action of acute administration of estrogen (female sex hormone) is displayed as a rapid decrease (< 10 min) GABA inhibition of neurons located within the insular cortex [1221], hippocampus [1013, 714] and hypothalamic ventromedial nucleus [667].
After initiation, sexual motivation advances, as for any other simple motivation, through deleterious reorganization of cell metabolism. Increase in glutamate levels in the medial preoptic area can elicit genital reflexes even in anesthetized rats (without consciousness), while glutamate receptor antagonists impair copulation [333]. Rapid (non-genomic) initial stimulation of membrane receptors by steroids alters their coupling with G protein and changes production of the second and retrograde messengers, thus changing sexual arousability [118, 200]. Nitric oxide also increases penile erection in anesthetized rats and NMDA-induced erection and yawning are mediated by increased NO synthesis in the paraventricular nucleus [1232]. Correspondingly, inhibition of nitric oxide synthesis eliminates male-like behavior and the deficit was principally in mounting, suggesting that sexual motivational systems were affected rather than consummatory mechanisms [1073]. Retrograde messenger nitric oxide [786] and excitotoxic agents [335] enhance both sexual behavior and neuronal damage in rats. Sexual behavior connected with alteration of homeostasis: in the preoptic area of males, neuronal excitability and Na+,K+-ATPase activity increases with the progression of mating [802]. Sexual motivation leads to astrocyte swelling [697]. In males, chronic testosterone treatment enhances the Na+,K+-ATPase activity in the preop-tic area, whereas castration decreased it and none of these manipulations had effects in the cerebral cortex [802]. As a consequence, the complex of metabolic pathways connected with sexual arousal indicates a development of cellular damage during sexual motivation.
In some aspects, sex hormones affect differently the male and the female and this is not unexpected, in light of their different hormonal status. In the male rat brain, GABA action leads to increased cyclic AMP activity, whereas GABA action in the female brain leads to the opposite effect, while at birth, males have an increased activity of the cyclic AMP system compared to females [68]. Nevertheless, this difference needs to be explained11. Thus, in spite of fractional and scattered data related to the mechanism of sexual behavior, there are many indications to a connection of sexual motivation with neurohormone-induced cell damage.
Artificial motivations: drug-dependence, self-administration and self-stimulation
Besides natural motivations, there are artificial motivations, which, on the one hand, have a significant role in human life and, on the other hand, give the opportunity to investigate the mechanism of a motivation. Certainly, these phenomena are observed in animals, too. The most important of artificial motivations is drug dependence. Some substances, for instance opioids, induce euphoria, but repeating their implementation leads to the origin of dependence, the necessity of recurring their administration and to the withdrawal syndrome after refusal of further drug treatment [887]. Drugs of abuse, as a rule, produce both euphoric (reward) and despairing (negative withdrawal) reactions. Primary euphoria and subsequent abstinence syndrome are the main reasons for drug addiction. The difference between acute and chronic actions of drugs of abuse is presumably the development of overcompensation that we have considered in topic 2.
The syndromes of drug addiction are extraordinarily similar for any drug, although drugs of abuse are highly diverse substances. Each drug binds to its selective protein target in the brain and elicits selective physiological effects upon acute administration. The common actions on brain reward circuits include brain’s endogenous opioid, GABA system, dopaminergic neurons and cannabinoid systems. Both inhibition of nucleus accumbens neurons and moderate to strong stimulation of dopaminergic transmission are distinctive features of diverse classes of abusive drugs [981]. All these metabolic pathways are interconnected. There are subregion-specific differences in the proportion of dopaminergic and GABAergic neurons [911]. Thus, the brain is non-homogenous in respect to sensitivity for drugs of abuse, as we demonstrated in reference to other motivations. However, certain drugs of abuse can induce cross-tolerance and cross-sensitization to one another with respect to their rewarding effects [887]. During a self-administration session, distinct but overlapping subpopulations of neurons in the rat medial prefrontal cortex and nucleus accumbens become active during operant responding for different rewarding substances (cocaine and heroin), which, obviously, have different targets [225]. Therefore, even if receptors for different drugs differ, some important properties of drug-dependence mechanisms are closely related.
The mechanism of drug dependence was brought to light after the discovery of endogenous opioids, endorphins and enkephalins, the central mechanism of the reward system of brain. Thus, artificially administrated morphine, an exogenous opioid, affects the same areas of a brain that are affected by endogenous opioids under natural conditions. Beside the substances that imitate action of endogenous opioids, there are stimulants of intermediate structures with the end-point link of endogenous opioids, such as cocaine, amphetamine, caffeine, alcohol, etc. For instance, the psychostimulants cocaine and amphetamine affect the dopaminergic neurons, which participate in regulation of the opioid system [282, 450].
Targets for drugs of abuse are located first of all in the nucleus accumbens, in pyramidal cells of the medial prefrontal cortex, which are primarily involved in the reward circuitry, and also in the hippocampus, in the locus coeruleus, in the ventral tegmental area and some other brain zones [1034, 911].
Just as for any other motivation, the motivation to receive the next dose of a drug may be put at the basis of an instrumental behavior. Self-administration of drugs of abuse is a well established fact. Both the euphoric effect of the drug and a withdrawal syndrome compel one to search for new doses of the drug. This corresponds to the development of any other motivation: a negative sensation leads to a search for reward.
The acute and chronic introduction of drugs acts oppositely to activity of main metabolic systems participating in processes of damage-protection, such as cyclic AMP system, Na+,K+-ATPase activity and GABA system12.
Chronic exposure to many drugs of abuse causes an increase in excitability of dopamine neurons and impairs the dopamine system, mediated via increases in AMPA glutamate receptor responsiveness and augmentation of the cyclic AMP system [887]. Second messenger systems participate in the development of drug addiction [850, 1296, 673, 887, 770, 851] and daily morphine administration downregulates the inhibitory Gj protein and increases cyclic AMP in motivational areas [282, 887]. Alteration of the cyclic AMP system was not uniformly distributed within the brain. Chronic decreased cyclic AMP activity in the shell of the nucleus accumbens (presumably its rewarding site) may be involved in producing positive affective states, but on the other hand, decreased cyclic AMP function in the nucleus of the amygdala (presumably the negative centers) and medial nucleus of the amygdala may be involved in producing and maintaining negative affective states (anxiety-like withdrawal symptoms) [935, 887]. Both rewarding or aversive stimuli cause short-term increases in cyclic AMP activity in the nucleus accumbens, but this is not found in other brain areas [888].
We have to remind ourselves that activation of the cyclic AMP system after damage does not mean necessarily the deleterious role of this system, but rather is evidence of tense activity, directed to protection. Injured as a result of drug dependence, neurons, probably try to survive and augment the cyclic AMP system. For example, artificial inhibition of the cyclic AMP signaling pathway increased alcohol-drinking behaviors [936].
Cyclic AMP signaling, it is supposed, plays a role in morphine withdrawal without participating in the rewarding properties induced by morphine, cocaine and natural reward (food) [1296]. Opioid dependence is heavily investigated. Acute administration of opioids, as a rule, lowers neuronal excitation, reduces cell damage, and this is pleasant for the subject, although the effect depends on the type of opioid receptor and on drug concentration [1014, 1149, 1145]. Opioids protect against ischemia and hypoxia, too. Opi-oid induced short-term protection and chronic damage is displayed in various brain areas, even in the visual cortex. Neurons of morphine-dependent animals are depolarized relative to the naive animal [718]. Chronic morphine withdrawal increases spontaneous activity, lowers signal-to-noise ratios and produces weaker orientation and direction selectivity in visual cortical cells that were significantly improved by morphine re-exposure [141]. Compared with the cells recorded within 3 h before morphine injection, the cells recorded within 3 h after injection showed improvement of parameter activities. Repeated administration of a narcotic drug overprotects neurons and forces the home-ostasis to recover equilibrium, to counteract the protection, increase damage and depart equilibrium to injury [1002, 1149].
In response to drugs of abuse, many genes change their expression and many such genes also change their expression in response to a wide range of other stimuli or emotional triggers such as stress, fear, attention and arousal, motor activity and aggression [1026]. Drugs of abuse are strong triggers of pathological motivation, because of loss of corresponding control and they promote pathology-related alterations in the neural tissue. The amount of striatal ribonucleic acids for voltage-dependent K + channels during drug dependence in addicts is reduced, the number of Ca2+channel proteins in adrenal-derived cells and in the cerebral cortex is upregulated [608], the a1 receptors are activated [807] and NMDA receptors are supertuned. Participation of intracellular Ca2+ may depend on a predisposition to drug addiction. A significant increase in the high voltage activated Ca2+ currents during the withdrawal period was observed in the withdrawal seizure-prone mouse strain, while those of the resistant strain showed no significant enhancement [967]. All these cellular alterations increase excitability and make the neuron more sensitive to injury.
A classic explanation for addiction is that use of a drug causes compensatory physiological adaptations [1345]. Acute administration of drugs of abuse produces euphoric reactions (rewards), while their chronic administration generates despairing (negative withdrawal) reactions. Correspondingly, reactions of neural tissue to chronic and acute introductions of these drugs are also opposite, and acute action of physiologically relevant doses in naive animals is usually inhibitory, at the same time as their chronic action increases excitability. In addition, acute action of the drug in drug abusers evokes also short-term inhibition and this, evidently, is connected with the origin of a euphoric feeling.
When cellular homeostasis compensates for this overprotection, neuronal damage increases and this compels one to take new doses of opioids. This property is not an exclusive attribute of narcotic drugs. For instance, whereas a single injection of thyrotropin-releasing hormone decreases short-term food and water intake in rats, repeated daily treatments stimulate water intake but not food intake [239]. Studying of opioid addiction gives us a complete example of behavioral, physiological and biochemical patterns of drug-dependence. This convincingly complements scattered data obtained during study of other drugs.
Comparable motivational behavior may be based also on electrical stimulation of euphoria-related brain areas (self-stimulation) and these zones may be exceptionally narrow. For example, the site in the ventral tegmental area effective for self-stimulation is approximately 1 millimeter across [1137]. The brain substrates underlying stimulation-induced feeding and self-stimulation are found in close proximity, but sometimes do not coincide and the threshold for self-stimulation is usually higher [128].
Chemical substances that augment or inhibit other motivations similarly affect self-stimulation. Application of a dopamine receptor antagonist in the nucleus accumbens abolishes self-stimulation of the lateral hypothalamus [262]. Psychostimulants amplify self-stimulation in a manner suggesting enhanced reward [571]. Microinjection of neurotensin in the ventral tegmental area (which increases dopamine release in the nucleus accumbens) produces a long lasting increase in the rewarding effect of self-stimulation, whereas cholecystokinin, which decreases feeding, has the opposite action. Similarly to drug-dependence, both physical and psychological stressors, which evoke damage in neurons, facilitate the acquisition of drug self-administration [976]. Efficiency of self-stimulation, beside the location of irritation, depends only on intensity of stimulation, exactly as cell damage and LTP initiation depends on the power of excitation. Altering frequency, current, or pulse width produced almost identical changes in self-stimulation responses. These findings show that the neutral network serving self-stimulation simply integrates the amount of charge over time [571] and does not transmit or activate any information within the tissue.
Thus, metabolic pathways of drug-dependence and other simple motivations are coincidental as a whole. The number of astrocytic gap junctions increases during amphetamine withdrawal [913]. Drugs of abuse and electrical self-stimulation given for an extended period produce morphological changes in the dendritic branching and spine density in medium spiny neurons of the nucleus accumbens and in other areas involved in the reward circuitry [1034, 887]. There is evidence for a loss of neurons in many regions of the brain in chronic alcoholics [169, 373]. Even uncomplicated alcoholics who have no specific neurological or hepatic problems show signs of regional brain damage and cognitive dysfunction [518]. The enhancement of synaptic responses after ethanol withdrawal is not a result of increased transmitter release from presynaptic terminals [886], but is the consequence of change in the postsynaptic structures and display altered states of neurons. Therefore, there are a few doubts that addictive drugs in the short-term protect neurons and produce pleasure, while their chronic implementation routinely increases protection and provokes the homeostatic system to compensate this overpro-tection and thus injure cells and produce frustration. For that reason, this homeostatically-induced damage forces one to repeat administration of the drug in order to temporarily protect neurons and receive pleasure.
Motivation to sleep
The motivation to sleep arises after long wakefulness, fatigue or as a result of pharmacological treatments. When sleep begins, a consummatory stage of motivation is commenced and, if this stage is long enough, motivation to sleep ceases. Sleep comprises two main phases: deep, slow wave sleep (sleep spindles with a high amplitude of rhythmic activity in the electroencephalogram), and paradoxical sleep (rapid eye movement sleep with an active electroencephalogram).
As a result of the perceptual overload of the brain cortex, extracellular adenosine concentrations increase in parallel to cerebral oxidation, augment cell damage and trigger a sleep motivation. Sleep starts in the most loaded regions of the cortex and then eventually, after the release of the inflammatory mediator adenosine, reaches the ventrolateral preoptic area of the hypotha-lamus. After that, brain metabolism drops significantly with a drop in brain temperature and triggers a reaction similar to abortive or partial awakening (paradoxical sleep) [873].
The medial preoptic area is a sleep- and thermoregulatory center and lesions of this area produce a sleep decrease with an elevation of body temperature levels [575]. One of the functions of sleep is the regulation of energy consumption and the mechanism of sleep initiation is related to the mechanisms of feeding and temperature regulation: motivations that are also connected with energy exchange [575]. Sleep is dependent on feeding, because of the metabolic consequences of food ingestion. This basic signal is communicated to the vigilance-controlling centers by a cascade of peptidic and non-peptidic messengers that promote wakefulness and hunger, possibly via a hypometabolic action (as in the case of neuropeptide Y or orexins), or somnolence and satiety, possibly via a hypermetabolic action (as in the case of leptin) [891]. Different combinations of chemosensitive sites operate in wakefulness and sleep [378]. GABA^ receptors, amphetamine-like stimulants, amino acids, lipids (retrograde messengers), proteins (cytokines) and the neuropeptide orexin, are known to significantly modulate sleep [897]. The consummatory phase of sleep is connected with inhibitory processes, as we have described when discussing other motivations13.
Extended restlessness involves high general neuronal activity in the brain, sensitizes the standard pool of damage-related pathways and may promote excitotoxic damage14.
As a result, properties of all elemental motivations that we have discussed are comparable and correspond to our idea about the origin of motivation being as a result of deviation of important biological parameters from their optimum and the subsequent development of transient or sometimes permanent cell damage. If homeostatic recovery from damage by means of endogenous resources is impossible as a consequence of exhausting of energy or substances, motivational behavior is the only means for the recovery of healthy equilibrium. We consider that in this case homeostasis undertakes efforts for organization of goal-directed action, such as to establish a new set-point when maintenance of the previous set-point is unattainable. Although the majority of available data correspond to alterations in neurons during the development of motivation (and we mostly paid attention to these processes), neural-glial interaction has been shown to participate in motivational behavior, but this aspect of the problem is poorly studied.
