Second messengers and cell survival
Second messengers, in any cell, evidently, execute general basic functions, but particular functions may appear and disappear in specific cells. Cellular damage and damage-related processes are closely connected with the systems of second messengers, Ca2+ and cyclic AMP, which participate in signal trans-duction within a cell and control a variety of processes that are important for survival. Injury increases cyclic AMP and Ca2+ concentrations within a cell. First of all, the effects of cell damage are strictly related to Ca2+ homeostasis [1236]. Damage of different origins injures cells through Ca2+ influx, while inhibition of endogenous Ca2+ suppresses damage [642], Ermak et al. 2002, Linda et al. 2004). Some chemical factors, known by their detrimental action, also affect damage through stimulation of intracellular Ca2+. NMDA enlarges damage through enhancement of Ca2+ [1128, 10]. Retrograde messengers, arachidonic acid, NO and their metabolites, may affect damage by means of distortion of Ca2+ homeostasis [23, 388, 1236]. Augmentation of intracellular Ca?+ may also be evoked by cytokines [339, 798]. On the other hand, Ca2+ elevation is not an obligatory attribute of damage. For instance, neuronal swelling may be independent of the elevation of intracellular Ca2+ [540]. Usually Ca2+ penetrates a cell as the result of activation of excitable membrane, but intracellular Ca2+ may mediate damage independently of change in excitability [123]. The spread of Ca?+ in tissue depends on neural-glia interactions. Ca2+ diffuses through gap junctions to evoke Ca2+ signals in neighboring unstimulated astrocytes, elevates Ca2+ in adjacent neurons and thus aggravates damage [381, 763, 526].
Ca2+ concentration in astrocytes may be quickly elevated because of Ca2+ outflow from intracellular stores [538], and treatment of astrocytes with inositol trisphosphate, controlling Ca2+ influx from intracellular stores, protects astrocytes against oxidative stress [1362]. Recently a new metabolic path for protective activity of neurons through a\ receptors was discovered. These receptors are intracellular proteins on the membrane of the endoplasmatic reticulum of neurons and they modulate Ca2+ mobilization from inositol trisphosphate-gated intracellular pools. The remarkable ability of a\ receptor ligands is to block or delay the deep depolarization that accurs immediately after stroke, although they have excitatory effects at low concentrations. Such a deep depolarization, in the absence of a\ receptor agonists, inevitably leads to cell death [36]. This collapse in membrane function results from failure of the Na+, K+ pump and, in the given case, protection of neurons independent of glutamate receptor mediation, since extracellular glutamate begins to accumulate only after occurrence of the depolarization [808]. Gap junctions in astrocytes are involved in mediating intercellular Ca2+ signaling throughout the glial syncytium, but they are controlled by neurons [878, 1047]. In response to neuronal activity, Ca2+ concentration in astrocytes rises and can lead to glutamate-dependent Ca?+ elevations in adjacent neurons and, hence, although Ca2+ ion is a typical second and not a retrograde messenger, these ions may affect remote cells [964]. Depending upon local Ca2+ concentration, astrocyte Ca2+ waves can either increase or diminish nearby neuronal activity [885] and thus align a level the difference between the states of near neurons. Rapid Ca2+ elevation within astrocytes may, probably, protect them, but injure adjacent cells. Nevertheless, the velocity of Ca2+ signal propagation is in the order of tens of micrometers per second, i.e. two to three orders of magnitude slower than AP propagation [160]. Therefore, regulation of intracellular Ca2+ cannot participate in the on-line control of performance, but may exert a slower influence on behavior.
For example, disturbance of intracellular Ca2+ may be important in affective states [1190] and play a crucial role in the pathogenesis of Alzheimer’s disease [1144]. In bipolar disorder patients, abnormalities in Ca2+ homeosta-sis were found and these abnormalities may be linked to disturbances in the function of G proteins that mediate cyclic AMP signaling [358]. Modulation of intracellular Ca2+ through a1 receptors is a necessary component of cocaine and ethanol-induced motivational effects, and above all, activation of a1 receptors decreases anxiety. They act as antidepressants and are involved in learning, response to stress, addiction and pain perception [808]. The acute and repeated stimulations of a a1 receptor subtype improve behavioral depression [1283]. Thus, Ca2+ homeostasis directly or through activation of a1 receptors participates in fine-tuning of attributes of behavior, and may by likened to fine-tuning the regulation of the mood. At the same time, impairment of second messenger systems does not lead to immediate death, but provokes delicate changes in behavior.
A second messenger cyclic AMP such as Ca2+ ions, is tightly connected with processes of damage-protection. However, Ca2+ is an important modulator of damage, while cyclic AMP mediates protective effects in cells. Cyclic AMP, as Ca2+ ions, affect various intracellular processes2 and can penetrate through gap junctions, too, but a wave of cyclic AMP cannot spread very quickly, since intracellular stores for cyclic AMP are absent. Cyclic GMP and cyclic AMP pass equally well through various hemichannels, whereas other nucleotides including AMP, ADP, ATP, cyclic TMP, and cyclic CMP are not effective [483]. The cyclic AMP role in cell protection was investigated especially carefully. Cyclic AMP molecules are important in many biological processes and in any one of cell types. It is used for intracellular signal trans-duction, such as, say, transferring the effects of hormones that cannot move through the cell membrane. Cyclic AMP connects events in outer and in inner environments and with cyclic AMP cells are capable of distinguishing different outer factors (for instance, stress and mutagens) [812], activating protein kinases, increasing expression of a large number of genes and controlling functions of ion channels. Augmentation of the second messenger cyclic AMP converts a cell to activation of developmental plasticity and causes, for instance, the formation of new synapses between rat hippocampal neurons in primary culture [402] and induces morphological changes within Aplysia sensory neurons [1195], dependly on protein synthesis [909]. Thus, cyclic AMP metabolic pathways regulate cellular homeostasis by means of activation or repression of transcription in response to extracellular signals [510].
In bacteria, cyclic AMP is a chemoattractant and controls mating. In mul-ticellular animals, cyclic AMP is very important for neural function, but, for sure, in this case it is not concerned with mating only, although cyclic AMP is activated by steroid sex hormones [1146] and cyclic AMP pathways determine sexual differentiation of the brain. An example of this divergence is the response of cyclic AMP system to GABA. In the male rat brain, GABA action leads to increased phosphorylation of cyclic AMP response element binding protein, whereas GABA action in the female brain leads to a decrease in phosphorylation of this protein [68]. The general function of cyclic AMP is to support cellular activity during stressful or injuring impacts. In neurons, cyclic AMP increases excitability and allows intensive functioning when it is necessary. Factually, in bacteria, cyclic AMP plays a similar role, since it increases expression of enzymes that can supply energy independently of glucose, when the glucose-dependent pathway of ATP synthesis is exhausted. According with the "traditional" role of cyclic AMP, as the substance that counteracts stressful conditions, injury of any kind evokes a prominent enhancement of intracel-lular cyclic AMP in neurons and this is intimately associated with neuronal defense3. In spite of this protective action, cyclic AMP increases excitability of neurons without increasing symptoms of excitotoxicity. In most cases, the treatments that excite neurons lead to their damage, while the treatments that inhibit neural cells protect them. This is not the case for treatments affecting the cyclic AMP system. Axonal injury elevates cyclic AMP and increases excitability of sensory neurons of Aplysia, in the hippocampus and other brain areas [1195, 1159]. In its turn, activation of the regulatory cyclic AMP transduction cascade produces acute sensitization in rat sensory neurons [147], increases potential-dependent Na+ current and excitability with a latency of some minutes after acute cyclic AMP activation [330, 1377]. Activation of cyclic AMP pathways potentiates not only excitability but also augments postsynaptic potentials in neurons of various brain areas [1159]. Beside enhancement of gap junctions, excitability and synaptic efficacy, cyclic AMP activates other neuronal functions and favored expression of a mature ion channel pattern in astrocytes. Nonetheless, knockout mice with deletion of binding protein of cyclic AMP-responsive element performed normally in the Morris-water maze and fear conditioning [1316].
Nevertheless, even after chronic enhancement of the cyclic AMP activation (up to 7 day), the ability of sensory neurons to be sensitized is maintained and does not appear to be down-regulated [147]. Cyclic AMP-evoked increase in excitability allows intensive functioning of neurons [812] and does not lead to cell death. Considering that excitotoxicity is an important reason for cell damage, cyclic AMP excitation itself is not the reason for neuronal protection. A pharmacologically reinforced cyclic AMP signaling in rat glial cell cultures depresses oxygen radical formation in microglia and the release of cytokines which mediate oxidative damage and secondary astrocyte activation [1096]. Evidently, turning on the cyclic AMP cascade ensures that the neural system can compensate for detrimental consequences of injury, but the resulting steady state is not the component of previous steady state recovery in a neural system and rather accompanies only a transfer to a new stable state. Cursory evaluation of complexity of these biochemical events (which, in reality, are much more complicated) shows that cyclic AMP cannot participate in rapid on-line processes, but may be important during reorganization of cell states during homeostatic survival.
Different types of adenylyl cyclases and protein kinases may have specific (even opposite) effect on cell functions during acute and chronic action [889] and, hence, the same molecules of cyclic AMP may lead to different outcomes and this is important for the organization of behavior. Nevertheless, when evaluating a possible role of the cyclic AMP system in behavior, one must realize that the protective result of cyclic AMP influences is too prolonged in time. The question is how this activity participates in a mental function. The target gene activation in response to cyclic AMP is typically observed after several minutes. The protection begins slowly (at least in minutes and usually much longer), it needs continuous metabolic and energetic support, and evidently, weakly depends (if at all) on real-time activity and learning, and some more memory. It had been supposed, that cyclic AMP pathways were critically involved in memory formation (LTP is implied) [20, 298]. The cyclic AMP system and intracellular Ca2+, as with other pathways connected with damage-protection may execute its function only as a slow or middle-time regulator of behavior. For example, a homeostatic control of neural activity is considered to be monitored through activity-dependent changes in second messenger AMP [296]. Also, rest in Drosophila, which has common features with mammalian sleep such as prolonged immobility and decreased sensory sensitivity, is connected with cyclic AMP signaling. Activation of cyclic AMP pathways is inversely related to the duration of rest, thus demonstrating a cyclic AMP role in waking and rest homeostasis [534]. Some forms of behavior, which are connected with chronic damage and the depolarization of neurons such as pathologic drug-dependence, are also determined by the cyclic AMP system. Opioid [1362] and cocaine [850] dependencies develop through activation of the cyclic AMP pathway, impairment of Na+,K+-ATPase activity and depolarization. Cyclic AMP pathways may be also concerned in recovery of functions [953].
Intercellular protection by retrograde messengers and cytokines
Besides second messengers, retrograde messengers also exert a direct influence to intracellular metabolism and are serve by the means for coordination of adjacent neuron activities and neuro-glial interactions [1128]. These tools provide intercellular transmission. Small molecules of retrograde messengers penetrate the cellular membrane and exert their influence apart from a synap-tic transmission. Then the systems of second and retrograde messengers interact. For instance, cyclic GMP mediates the modulation of transmitter release by NO [991]. Some conventional neurotransmitters, such as GABA, dopamine and neuropeptides may also reach the intercellular space, but they interact with chemoreceptors. Retrograde messengers are released from regions on the postsynaptic cell that do not display typical morphological specializations for secretion (e.g. vesicles, active zones, chemoreceptors). In principle, the messenger can act on any part of the presynaptic cell: its terminals, axon, dendrites or soma. Retrograde messengers, endogenous cannabinoids, lipids (arachidonic acid) and the gases NO and CO are membrane-permeable and may penetrate the outside of a mother cell and within adjacent cells in any location, but usually cannabinoids act at the presynaptic axon terminal, while the gases NO and CO act globally.
A system of retrograde messengers is involved in the regulation of vital brain functions and they can, depending on conditions, cause an inhibition of cyclic AMP or a sharp rise in its concentration [402]. Potent chemical control of protection is accomplished by cannabinoids, which are the primary psychoactive component of the cannabis plant (marijuana), but they are naturally produced in the bodies of animals and their receptors, coupled with the G proteins, are widely presented in the brain [23, 402, 1138, 1285]. The cannabinoid system is currently thoroughly studied retrograde signal system in the brain [23]. Natural cannabinoids are derived from arachidonic acid and their amount increases during damage. They are synthesized for actual needs and are not stored for later use. Activation of cannabinoid receptors temporarily reduces the amount of conventional neurotransmitter released and this permits the postsynaptic cell to control its own synaptic inputs.
The effect of the cannabinoid secretion depends on the nature of the conventional transmitter that it controls, since it may reduce both excitatory and inhibitory synaptic inputs. Cannabinoids cause an inhibition of cyclic AMP, but concurrent stimulation of cannabinoids together with dopamine (as evidently happens during behavioral reinforcement) leads to a sharp rise in cyclic AMP concentration, reduces the secretion of NO and the inflammatory cytokine from microglia, and facilitates neurogenesis [402]. Canabinoids are potent regulator of mood and cognition. The sensitization of cannabi-noid receptor-mediated G protein signaling in the prefrontal cortex is one of the factors in the pathophysiology of suicide and post-mortem studies have shown a higher density of the cannabinoid receptor in the prefrontal cortex, striatum and anterior cingulate cortex of schizophrenics [1305]. Canabinoids, as other substances affecting damage-protection, regulate many life-important processes: anxiety, pain, epilepsia, insomnia, depression, enhance appetite and produce hypothermia, which reduces brain damage [610]. Arachidonic acid, NO and their metabolites have a number of common targets on which they may exert similar or opposite actions, and have a crucial role in the regulation of inflammation, immune responses and cell viability [23, 1431]. Arachi-donic acid is found to be cytotoxic at concentrations that overlap physiological ones and evoked necrosis [985]. At present, good evidence exists for inhibition of voltage-gated Ca?+ channels, a direct action on the synaptic release machinery at the presynaptic terminal and to the opening of K + channels on dendrites The retrograde messenger of a global action, free radical NO is a small molecule and it can relatively quickly (milliseconds to seconds) spread to adjacent cells. Therefore, there is short-term activity-dependent retrograde signaling [23] and this may be a reason for parallel modulation of synaptic efficacy and excitability during learning. A release of NO affects numerous other cells in the environment without regard to their functions and synaptic relationship to the releasing cell [23]. The balance of NO levels in the tissue may be crucial for orienting microglial reactions towards neuroprotection or neurotoxicity and it is likely that a low NO promotes protection, while a high NO results into neurodegeneration, since enhancement of NO concentrations is toxic to surrounding neurons [848, 749, 360]4. Evidently, during the reversible stage of damage, when NO concentration is small enough this retrograde messenger protects adjacent cells, but after damage, NO concentration increases and kills neighboring cells. By the way, taking into account that NO is a retrograde messenger, it is natural to suppose that the neuron emitting NO is injured itself, but protects (even if at the beginning) neighbor cells, since NO concentrations reaching these cells have to be smaller.
Besides the above-mentioned direct pathways affecting damage, there are substances that modulate damage-protection. Powerful participants in the processes of damage-protection are cytokines. Cytokines are peptides that are produced by every cell type in the body: neurons, glia, astrocytes, etc. [120], but predominantly by activated immune cells such as microglia and are involved in the amplification or reduction of inflammatory reactions5. However, this enhanced expression is not the cause for neuronal damage. Defiantly, mice, lacking receptors of tumor necrosis factor, demonstrated increased neuronal degeneration [190]. The the colony stimulating factor cytokine also demonstrates "protective" reaction [120]. As part of the scope of our topic, it is important to consider manners of cytokine influences, since the same substances modulate slow modifications of behavior. So, although maximal expression of cytokine interleukin-1 occurs five days after experimental cerebral ischemia [467] and this time robustly exceeds the time that is necessary for completion of mental functions, cytokines may participate in cerebral activity. In low doses, (picomolar range) interleukin-1 elicits a slow-wave sleep [1046]. At the same time, it may contribute to healing before reaching its maximal concentration. Some cytokines even more potently affect a mental activity, in particular the cytokines neurotensin and thyrotropin-releasing hormone. Although neurotensin is a proinflammatory neuropeptide [215] and it is elevated in plasma during damage [340], it stimulates intestinal wound healing [172], minimizes tissue damage [179], reduces oxidative damage [63] and counteracts ischemic damage [1239]. Thyrotropin-releasing hormone also exerts protection against damage [150, 600, 1302] and, moreover, it has been shown to be able to reverse damage to glia and neurons [955]. Both neurotensin and thyrotropin-releasing hormone protect cells against damage in physiological doses. Nevertheless, these cytokines inversely affect awareness: neurotensin decreases arousal and calm animals without sedating them [155], while thyrotropin-releasing hormone wakes animals from anesthesia [989]. Bath application of thyrotropin-releasing hormone resulted in a transient cessation of spindle waves, which is prominent during slow-wave sleep and rhythmic burst firing [168], and maybe this is one of the causes of its awakening action. Evidently, thyrotropin releasing hormone strains defensive functions6, while neurotensin is probably connected with the reinforcement system of brain7. Neurotensin and thyrotropin-releasing hormone, evidently, protect cells by different means and, in particular, they differently affect neuronal activity. Neurotensin, at the first decreases in membrane resistance by the activation of a cationic conductance, depolarizes and activates neurons (100 nM – 1 ^M) and after that evokes inhibition, at time scales of a few minutes [104, 231, 905]. Primary depolarization depends on suppression of K+ conductance, but did not depend on Na+ potential-sensitive channels, GABA receptors and ionotropic and metabotropic gluta-mate receptors [231, 905]. The first AP in the response usually has a larger amplitude, than following APs, as usually happened after depolarization-evoked excitation under normal conditions, when excitatory effects were compensated by manually hyperpolarizing the cell to control values [104]. In addition, neu-rotensin increased the frequency of both excitatory and inhibitory spontaneous synaptic currents and this effect depended on presynaptic potential-sensitive Na+ channels [905].
Thus, during damage, neurotensin evidently augments homeostatic activity of cells. It evokes strong, but short-lived excitation and prolonged inhibition. Neurotensin probably stresses the compensation and affect of neuronal activity in a conventional way. That is, neurotensine directly counteracts injurious factors to recover an initial state and, probably, counteracts a basic component of damage, high activity. Damage-protection is similarly dependent on the influence of neurotensin and on a shift of the membrane potential if one considers only weak linear deviations, when compensation recovers the lost equilibrium. High neurotensin concentration decreases Na+,K~-ATPase activity [917], suppresses the hyperpolarizing of potassium currents, activates a cationic conductance [104, 231] and stimulates inositol-trisphosphate production that provokes Ca2+ release from intracellular stores [763]. This cannot be the primary cause of protection, and, oppositely, can promote development of excitotoxicity, but prolonged action is inhibitory concerning both neuronal activity and common physiological functions.
Thyrotropin-releasing hormone elicited a transient hyperpolarization (of a few seconds) of the cell membrane and an increase in the membrane conductance to K +, followed by an enhancement of the generation of action potentials due to Ca2+ entry from the extracellular space. In addition, the input resistance of the cell membrane increases during the facilitation (few minutes) [928, 339]. Facilitation stimulates the release of prolactin and growth hormone, which usually protect cells. The hyperpolarizing response was transient despite the continuous leakage of thyrotropin-releasing hormone from the pipette. The enhancement of the spike generation was not due to membrane depolarization, since thyrotropin-releasing hormone causes a burst of action potentials without any detectable change in the resting membrane potential and the first AP after hyperpolarization was small, like during the damage depicted in Fig. 2.1. Hyperpolarization also tended to be gradually reduced by repeated administration of thyrotropin-releasing hormone, and the second application of thyrotropin-releasing hormone immediately enhanced the spike generation without eliciting the hyperpolarization. Action of thyrotropin-releasing hormone to an inhibitory neuron differed from action to excitatory ones. Thyrotropin-releasing hormone application to the GABAergic thalamocortical or hippocampal neurons resulted in depolarization, and increased excitability as well as increase the AP firing frequency and membrane resistance. This effect is mediated by a decrease in K + conductance [168, 310]. The mechanism of the facilitatory action of thyrotropin-releasing hormone is different from the mechanisms of conventional excitatory neurotransmitters or from shifts of membrane potential [928]. However, thyrotropin-releasing hormone exerts a classical excitatory influence, compatible with the effects of change in membrane potential to inhibitory GABAergic neurons and thus increases a direct inhibitory protection. However, such classical influences are distorted during interaction with the excitatory neurons, and the correspondence between excitability, membrane potential and spike amplitude testifies more to a complex non-linear effect of thyrotropin-releasing hormone than to compensatory processes. It produces transient inhibition and protracted excitation. Protection evoked by the thyrotropin-releasing hormone may be understood as a distortion of excitable membrane properties and reminds one of alterations that are observed during damage-related activation of the cyclic AMP system and other protective excitatory mechanisms. APs arise without depolarization during an increase in membrane resistance and, at the same time, amplitudes of spikes decrease. This ought to prevent the development of excitotoxicity. In addition, thyrotropin-releasing hormone triggers intracel-lular Ca?+ release, which opens Ca2+-activated K channels, evokes transient hyperpolarization and does not directly modulate Ca2+ channel activity [339]. Nevertheless, while such alterations delay the harmful consequence of damage and extend the amount of time required to search for a new homeostatic equilibrium, these alterations do not make the tissue healthy.
Protection through a detoured route
The first acute reaction of a neuron to damage is usually an excitation. The "reaction to damage" by its meaning ought to be clearly defensive and an excitation, as its primary meaning, is a protective action. Of course, not every excitation in an organism has an adaptive character, and it may acquire pathological character. As we already have described, strong excitation leads to damage and further excitation. This excessive excitation is a symptom of cell damage and factors that powerfully excite neurons lead to damage, too. Nevertheless, activation of the cyclic AMP system induces excitation and generates a protective response. The cyclic AMP cascade is not the only example of protective activity that is accompanied by an excitatory influence. There are also a few other factors that protect brain through excitation. Some cy-tokines, for example, interleukin-1 [467], granulocytemacrophage colony stimulating factor [120], thyrotropin-releasing hormone [928, 310] and anesthetic fentanyl (which decreases injury despite it activates epileptoid activity) [862] also excite and protect neurons. Evidently, protective processes in the given cases are not reduced to counteraction between inhibition and excessive excitation. We may suppose that compensation of damage through augmentation of excitation is connected with a search for a new point of equilibrium that does not coincide with initial normal conditions. The general sense of acute excitatory reaction of an organism or its cells may be not only a defense, but may constitute an attack, too. The efforts that are directed to compensation of injury are targeted not only to the outer environment, but also to a homeostatic recovery of one of the possible stable points of the inner state. Therefore, one may say figuratively that protection through excitation is more reminiscent of aggression, while defense may be correlated with recovery of the initial state.
Evidently, the role of any neurotransmitter is not limited by its influence to membrane potential. Synaptic influences also induce alteration in concentration of intracellular Ca2+ and other second and retrograde messengers. As a result, they affect intimate properties of brain cells, damage, protection, homeostasis and proliferation. The possibility of so versatile a control of damage-protection using synaptic process suggests that such alterations of the neuronal state somehow depend on physiological activity of brain and may perform mental functions not connected with unequivocal pathology. These intimate properties of cells somehow participate in information processing.
Flexible dependence of damage from heterogeneous parameters (or, more exactly, the absence of a general rule for a correspondence between the state of damage and the specific magnitude of particular parameters) may be demonstrated by the example of the injurious influences of volatile anesthetics. The most voltage-gated ion channels, glutamate receptors and many of second messenger systems are relatively insensitive to volatile anesthetics in clinically relevant concentrations, however, general anesthetics potentiate the GABA^ and glycine receptors [394, 408, 1050], hyperpolarize neurons and decrease in excitability [52]. Therefore, anesthetics do not evoke damage, and, oppositely, protect cells. Nevertheless, some gaseous anesthetics halothane, desflu-rane and isoflurane in anesthetic doses cause cytotoxic effects, contribute to tissue injury and reduce antioxidant defense mechanisms in cells of various lines [693, 1147]. Besides, preconditioning with the volatile anesthetics, when they were administered before severe damage, reduces following neuronal injury caused by overstimulation of glutamate receptors and by ischemic events [126, 302, 130, 354, 6] that are typical for preliminary actions of deleterious agents (see further). Certainly, volatile anesthetics reduce gap junctions [109, 257, 1332] and this may be the reason for their deleterious affects. By the way, connexin hemichannels are blocked by hyperpolarization of the plasma membrane [260], although hyperpolarization itself protects neurons. Does this mean that gap junction conductivity is the most important of parameters? This is not true, since in other cases other parameters are vital. In addition, the state of gap junctions affects the spread of damage in tissue, rather than damage itself.
Our description demonstrates that any known particular characteristics of an injured cell such as augmentation of intracellular Ca2+, increase in excitability, cell volume, free radical enhancement, etc. cannot be considered as the features of damage and the opposite changes cannot be considered as the sign of protection. Damage comes when the possibility for compensation is exhausted. Augmentation of the cyclic AMP system in tissue is the evidence of compensational processes participating in the modification of many other metabolic pathways: an increase in gap junctional coupling, O2 uptake potentiation [716], gene expression involved in the regulation of Ca2+ homeostasis [44], etc. In particular, the cyclic AMP system regulates the basic instrument of ion homeostasis of cells, Na+ ,K+-ATPase. Homeostasis of the sodium pump can be enhanced [1143] or inhibited [691, 1364] by the cyclic AMP system and the effect may be different in different areas of brain [105]. Amplification of cyclic AMP does not routinely convert a sick cell into healthy one. This second messenger works intensively for repair. So, acute enhancement of cyclic AMP (20 min) by exposure of cyclic AMP agonist produced a sensitization of rat sensory neurons and a release of immunoreactive substance P, which (such as the cyclic AMP) produces protection [668, 866, 364] and excites neural cells. At the same time, long-term exposure of the agonist does not appear to downregulate its ability to augment substance P release, to increase cyclic AMP production or to be sensitized [147]. That is, despite chronic treatment, neurons are capable of undergoing a long-term sensitization that does not downregulate, but requires the continual presence of a sensitizing agent. Production of cyclic AMP is not evidently the result of damage or protection. This is an expendable substance needed for functioning of the compensatory process. Yet, it is necessary to remark that the cyclic AMP system is not unique in organization of protection. Other examples are some neuro-transmitters, retrograde messengers and cytokines. The cyclic AMP system is rather the best investigated.
