Direct cholinergic activation of numerous areas of the forebrain and diencephalic limbic structures consistently induced polydipsia, i.e., significant water ingestion in rats that have previously drunk to satiation ^{[1][2][3][4][5][6][7][8][9]}[1,2,3,4,5,6,7,8,9]. The drinking response was induced by muscarinic compounds such as carbachol, pilocarpine, muscarine, or physostigmine, was dose-dependent, and reversed by intracerebral atropine or systemic application of centrally active atropine salt, as well as, by scopolamine or methylatropine, but not by hexamethonium or d-tubocurarine ^{[8][10][11][12][13][14][15][16]}[8,10,11,12,13,14,15,16]. Any method of drug delivery, e.g., in crystalline form, by solution injection, or by iontophoretic application, caused comparable responses ^{[1][2][4][5][16][17]}[1,2,4,5,16,17]. For intracerebral injections, the drinking response was also not affected by the solvent used (water, saline, or sucrose solutions) [15].

In the studies of the cholinergic drinking response, carbachol was used most frequently because of its stability and long-lasting effects. The carbachol-induced drinking response in rats was specific to this neurochemical activation [18], and drinking was not successfully induced by electrical stimulation of the brain sites, from which carbachol-induced drinking. Carbachol-induced drinking response was consistently reproduced for the rat brain; however, it was not obtained from the brains of monkeys ^[19][20][19,20], cat ^[21][22][21,22], guinea pig [9], chicken, or dove ^[23][24][23,24], while a marginal and unreliable drinking response was reported for dogs after intracerebroventricular injection of carbachol [25]. In the rabbit, a limited increase in drinking after carbachol was obtained from the supraoptic nucleus only [26], and in the sheep brain, drinking was obtained with only 14% efficacy [27]. Intracerebroventricular injection of carbachol in newborn rat pups had no effect on drinking before the 4th postnatal day [28]. The carbachol-induced drinking was not a “blind”, reflexive drinking, but it was augmented when a palatable saline solution was presented and decreased with unpalatable concentrated solutions [17]; and the response could be conditioned [29].

2. The Ascending Mesolimbic Cholinergic System for Aversive Arousal

The existence of two functionally discernable arousal system has been suggested long time ago ^[30][89]. One of these systems (named Arousal System I) was cortically oriented and associated with arousal in response to all environmental stimuli, while the other (Arousal System II) was related to the limbic system and associated with incentive-related stimuli ^[30][89]. The operation of the limbic system as an arousal system was, however, unclear. It was further complicated by very elaborate models of its arousing functions ^[31][90]. Moreover, neurochemical, and behavioral studies suggested inclusion of many ascending transmitter systems (noradrenalin, acetylcholine, dopamine, and serotonin) in the regulation of arousal, and the cholinergic component was postulated to be responsible for stimulus detection ^[32][91]. Further studies provided descriptions of numerous ascending projections based on many neurotransmitters, such as glutamatergic, cholinergic, noradrenergic, dopaminergic, serotonergic, histaminergic, and orexinergic systems, which influence the activity of the neocortex, basal forebrain, and activity of the limbic system ^[33][85]. This multitude of extensive ascending projections led to the conclusion of the existence of multiple arousal systems, which are interrelated and have combined arousing action on the brain ^[34][35][36][92,93,94]. The projections of the cholinergic component of the ascending afferents, particularly to the limbic structures were neglected and interpreted in terms of only cortical arousal ^[37][95]. All these hypotheses were not clearly distinguishing between cognitive arousal and emotional arousal, although, the emotional component of arousal associated with the effects of intracerebral carbachol was suggested some time ago by Grossman ^[38][96], who first reported carbachol-induced drinking response. He stated that “it seems that the diencephalon may contain a very diffuse system of cholinergic fibers and cells, which seem to be part of the neural mechanisms which control affective reactions” ^[38][96] (p. 79). Many decades of thorough studies with pharmacological approach to emission of rat ultrasonic vocalization provided a new model that explained affective arousal and its relationship to the aversive or appetitive nature of these responses. Vocalization expresses animal emotionality ^[39][40][41][97,98,99] and is particularly suitable for studying emotional behavior and for developing new treatments for affective disorders ^[42][100]. Extensive functional mapping and anatomical tracing studies (e.g., ^[43][44][45][61,101,102]), revealed a two-component emotional arousal system that was based on the ascending cholinergic projections for aversive arousal and on the ascending dopaminergic projections for appetitive arousal ^[46][47][103,104]. This ascending acetylcholine system originates from cholinergic perikarya located mostly in the medial part of LDT ^[48][82] and has been termed ascending mesolimbic cholinergic system for aversive emotional arousal ^[49][50][105,106]. It innervates predominantly the medial diencephalic and basomedial forebrain regions ^[51][52][53][53,54,56] and is reaching to the medial prefrontal cortex ^[54][107]. In addition to that, it has been shown using c-Fos labeling and immunohistochemical staining of cholinergic cell bodies for ChAT, that the cholinergic neurons of the LDT are active during initiation of aversive responses with emission of alarming 22 kHz vocalizations in rats ^[55][108]. Functional mapping of vocal responses that were induced by intracerebral injections of carbachol in the rat and cat brains revealed an extensive subcortical system of cholinoceptive limbic regions ^[44][56][101,109] that were comparable to the extensive regions, from which the drinking response was obtained in the rat. Thus, it may be postulated that the drinking response (polydipsia) induced in rats by carbachol was caused by activation of the emotional aversive arousal system, which initiated an emotional state, and was not associated with the regulation of thirst. This hypothesis would explain most problems the researchers had when they tried explaining the carbachol-induced drinking response as a homeostatic regulation of thirst. The ascending mesolimbic cholinergic system for aversive emotional arousal has diffused nature and single LDT neuron innervates many structures on the ipsilateral and contralateral side of the brain ^{[51][53][57][58]}[53,56,69,110]. It was hypothesized long time ago that the forebrain limbic regions form reciprocal loops with the midbrain limbic regions ^[59][111], so the forebrain limbic structures with terminal fields of the ascending cholinergic projections and the LDT are reciprocally interconnected. Hence, activation of the terminal fields by carbachol would reciprocally activate the source neurons of the LDT ^[55][59][108,111]. This was directly proven by studies using c-Fos for labeling active neurons and ChAT immunohistochemical labeling of cholinergic neurons in the LDT. Injection of carbachol into forebrain sites activated cholinergic neurons within the LDT, i.e., the source neurons of the ascending projections ^[55][108]. This important finding may explain why injection of atropine, a cholinergic muscarinic antagonist, into the brain sites different than the original site of carbachol-activated drinking response (including contralateral sites) were able, at least partially, to antagonize the response, presumably by blocking the reciprocal connections.

3. The Set of Symptoms of Aversive Arousal and Their Biological Role

Experiments with direct cholinergic stimulation of the ascending cholinergic projections have revealed a vast array of defensive responses comprising both somatic and autonomic symptoms. They include general arousal, preparation of the motor and sensory systems for immediate action such as fight or escape, activation of the autonomic and endocrine systems, and emergency activation of bodily resources. Intracerebral injection of carbachol to terminal fields of the ascending mesolimbic cholinergic arousal system caused an increase in attention, vigilance, in the number of eye movements, dilation of pupils, increase in respiration rate and muscle tension, as studied in the cat ^[60][112], and caused a concurrent decrease in locomotor activity, increased immobility (freezing response), and induced a general behavioral inhibition both in cats and rats ^[61][62][113,114]. These symptoms were accompanied by sustained emission of aversive vocalizations, 22 kHz alarm calls in the rat ^[44][63][101,115] and threatening growling vocalizations in the cat ^[60][112]. The clusters of all the effects were interpreted as behavioral symptoms of anxiety, and emission of 22 kHz calls in rats as an evolutionary equivalent of crying in humans ^[64][116]. Intracerebral carbachol also increased activity of the sympathetic nervous system, increased the mean arterial blood pressure and heart rate ^[65][66][67][117,118,119], and caused an increase in glycogen synthesis and hyperglycemia ^[68][69][120,121]. Intracerebroventricular carbachol increased serum corticosterone levels in a dose-dependent way and increased hypothalamic noradrenaline levels ^[70][71][122,123]. Additionally, in general, carbachol elicited hormonal and metabolic responses like those to moderate stress, as studied in dogs ^[72][124]. Intrahypothalamic-preoptic injections of carbachol significantly increased core body temperature in normal ambient temperatures ^[73][125] but lowered hypothalamic temperature, suggesting a redistribution of blood ^[74][126]. The development of stress response with a set of physiological changes prepares the animal for response to approaching danger and potential fight with an opponent or response to other harm to the body. This preparatory behavior also includes a decrease in feeling pain (increase in pain threshold) as preparation for harmful danger. Intrahypothalamic injection of carbachol in rats decreased formalin-induced pain in a dose-dependent manner ^[75][127]. Antinociception was also obtained from other brain sites, and an increased threshold for pain was obtained by stimulation of 119 sites of the midbrain and diencephalon with carbachol ^[76][128]. Antinociception was frequently associated with hyperexcitability to non-noxious stimuli ^[76][128]. Thus, the animal felt less pain but was more sensitive to any external stimulation. This response is expected at high emotional arousal state, particularly when the animal may expect violent struggle and fight. It was found recently that cholinergically induced antinociception may be further regulated by the orexinergic system ^[77][129]. The set of symptoms induced by cholinergic activation of the ascending mesolimbic cholinergic system for aversive emotional arousal allows for better understanding what emotional arousal entails. The widespread effects of this system are quickly changing the entire state of the organism from a usual functioning at rest to an alarm mode of varying intensity by activating the somatic, autonomic, and endocrine systems. The results presented in previous subsections indicate that this defensive alarm response is somehow associated with the drinking response. How drinking response, would contribute to this defensive behavior?

4. Biological Role of Aversive Arousal-Induced Drinking Response

Thus far, carbachol-indued polydipsia did not serve the regulation of thirst but it could be a defensive measure induced by the activation of the ascending mesolimbic aversive arousal system. Rats overloaded their bodies with water regardless of homeostatic water balance needs. Moreover, drinking an additional load of water should increase diuresis as a mechanism for returning body water content to the original level. Instead, it was reported that injection of carbachol into the medial preoptic area induced antidiuresis, i.e., increased and retained water content in the body in a dose-dependent manner but controlling the electrolyte content ^[78][130]. A similar effect of antidiuresis induced by carbachol was recently confirmed and could be reversed by injection of catalase inhibitor into the rat septum ^[79][131]. These observations indicate that the biological role of the cholinergically activated drinking response is to increase water content in the body at least for the duration of the potential danger or duration of the aversive state. What benefits of higher water content in the body would be during an emotional defensive state? There are several possible explanations to this question. Firstly, higher content of water in the organism would increase blood pressure ^[80][81][132,133]. This will increase blood perfusion in all organs, particularly in muscles and increase their oxygenation. Together with increased heart rate and glucose level in the blood (see Section 6 above), higher content of water would prepare the rat’s muscular system for increased effort. Another explanation might be that drinking additional fluids would increase overall blood volume as a defensive measure in anticipation of blood loss during intraspecies aggressive fight, although, fights among rats are highly ritualized and usually not associated with significant blood loss ^[82][134]. The excessive drinking response (polydipsia) might be only an occasional auxiliary behavior since rats might not have access to drinking water in face of a sudden danger in the natural environment. The most plausible alternative explanation is that the drinking response in rats caused by activation of the ascending aversive arousal system is an anxiety-induced response. It was suggested in earlier studies that the carbachol-induced responses represent anxiety type of aversive behavior, which is associated with stress response ^[64][116]. Thus, if water is available, drinking would represent anxiety-related response. If this explanation is correct, then, more species would show anxiety/stress-induced drinking response. It was observed, however, that in many species studied for cholinergically induced drinking, the drinking response was not present (see Section 1). Anxiety- or stress-induced water drinking might not be a universal response, but some species should demonstrate it. Significant polydipsia was recently shown after chronic mild stress in mice without changes in sucrose intake ^[83][135]. Social stress in mice caused the development polydipsia or stress-indued overhydration with demonstrated trends of anxiety-like behavior ^[84][85][136,137]. Anxiety is a documented factor in this behavior both in animals and humans. Polydipsia in humans is usually unrelated to the homeostatic regulation of water intake and is commonly observed in psychiatric patients ^[86][87][138,139]. A recent review has emphasized that polydipsic overhydration happens both in rodents and humans in response to anxiety and social isolation ^[88][140]. These results agree with the conclusion reached for the carbachol-induced drinking response in the rat. Long-lasting and severe polydipsia may, however, lead in humans to detrimental conditions termed “water intoxication” with secondary hyponatremia ^[88][89][140,141], so anxiety-induced polydipsia should be of short duration in physiological conditions. Thus, what benefit the organism has by a limited water overload in an anxiety situation? It seems that the polydipsic behavior might be performed as a soothing procedure that decreases the aversiveness of the situation, so as a rewarding maneuver that contributes to alleviating the feeling of anxiety. Primary polydipsia was already suggested to occur as a stress-reducing behavior ^[90][142]. Counteracting the state of anxiety could happen by activation of the mesolimbic dopaminergic system originating from the ventral tegmental area and terminating in the nucleus accumbens. The ascending cholinergic fibers were reported to synapse directly onto the dopaminergic neurons in the ventral tegmental area ^[91][144] and stimulation of the LDT neurons caused release of dopamine in the nucleus accumbens, a terminal field of the ascending dopaminergic projections ^[92][93][145,146]. These cholinergic connections to the dopamine neurons in the ventral tegmental area were suggested serving as a natural breaking mechanism during prolonged anxiety conditions. Studies of rat ultrasonic vocalization have shown that after induction of an intensive anxiety response by carbachol injected into the lateral septum with prolonged emission of 22 kHz vocalizations, a rebound effect has spontaneously occurred after several minutes (over 300 s). The rebound was characterized by sporadic emission of 50 kHz calls that were driven by the dopaminergic system, because they were antagonized by systemic haloperidol ^[94][147]. In other words, after several minutes of persistent cholinergically induced state, a rebound of dopaminergically driven episodes started gradually increasing up to the termination of the anxiety response. Cholinergic stimulation of dopaminergic VTA neurons is known to increase reward-seeking behavior and energize motor and locomotor behavior ^[95][148].