Vasopressin is a cardiovascular neuropeptide with a wide spectrum of actions exerted in the brain and peripheral organs [58, 59, 60, 61, 62, 63]. In the brain, neurons synthetizing AVP are present mainly in the supraoptic (SON), paraventricular (PVN), and suprachiasmatic (SCN) nucleus. Axons of these neurons project mainly to the posterior pituitary where they release their product to the blood. In addition, several axonal projections innervate brain regions encompassing the nucleus ambiguous, NTS, lateral habenular nucleus, nucleus basalis of Meynert, substantia nigra, ventral hippocampus, central gray, and spinal cord [58, 59, 60, 61]. Among them are the regions controlling both the cardiovascular system and pain, such as the noradrenergic/adrenergic A2/C2 region, nucleus ambiguous, dorsal vagal complex, and subnuclei of the solitary tract [64, 65, 66]. The parvocellular neurons of the PVN send projections to the spinal cord [66, 67]. Secretion of vasopressin is stimulated by hyperosmolality, hyperthermia, hypotension, angiotensin II (Ang II), stress, pain, and inflammation. It also shows distinct diurnal rhythmicity [65, 68, 69, 70, 71, 72, 73]. Stress, pain, and inflammation play a particularly significant role in myocardial infarction and heart failure [73].

Vasopressin activates V1 receptors (V1R; subtypes: V1aR, V1bR) and V2 receptors (V2R) that are located in the brain areas involved in the regulation of pain and cardiovascular functions (Table 1). They are also present in several peripheral organs, including the cardiovascular system, lungs, kidney, liver, and gastrointestinal system [61, 74, 75, 76, 77, 78, 79, 80].

Strong evidence indicates that vasopressin has multiple associations with the regulation of pain. Studies have shown that, in the rat, mechanical or thermal stimuli inducing pain cause activation of vasopressin neurons in the hypothalamo-neurohypophyseal system and release of AVP. Furthermore, it has been shown that pain can be eliminated by pre-application of a V1aR antagonist [68]. Studies on rats suggest that the nociceptive ascending pathway to the SON encompasses the caudal ventrolateral medulla and noradrenergic A1 region [70].

Centrally mediated analgesic effects of AVP. Systemic or intracerebroventricular (ICV) application of AVP elevates the pain threshold in the rat, and this effect can be eliminated by pretreatment with a V1R antagonist [(dPTyr(Me)AVP] but not the opiate antagonist naloxone [81, 82]. Similarly, intrathecal (it) administration of AVP to the spinal cord produces antinociception, which is not mediated by opiates, but can be eliminated by blockade of V1R and is absent in V1aR knock-out mice [82, 83, 84, 85]. Experiments on rats showed that intravenously applied AVP elicits dose-related effects on nociceptive processing in the spinal cord. Namely, it exerts an antinociceptive effect, expressed by reduction of action potentials mediated by C-type nociceptive fibers, although it elicits an opposite action when it is applied at higher doses. Both effects can be abolished by administration of a V1aR antagonist [86].

Strong evidence indicates that there are several sites in the brain where AVP may interact with neurons engaged in regulation of pain (Fig. 1; Table 1). Thus far, the most essential role has been attributed to the amygdala, PAG, and caudate nucleus (CdN). In the amygdala, vasopressin acting on V1a receptors also enhances the nocifensive reflexes related to pain [87, 88]. Experiments on rats showed that administration of AVP to the PAG increases pain threshold through stimulation of V1aR and also induces local elevation of leucine-enkephalin (L-Elk), methionine-enkephalin, and serotonin (5-HT). Moreover, the analgesic effects of AVP in PAG could be abolished by blockade of opiate receptors with naloxone [89], which indicates that, in specific brain regions, AVP interacts with opiates in regulation of pain. It should be noted that the PAG plays a significant role in regulation of cardiovascular function [90].

Fig. 1.

Associations of vasopressin, angiotensins and endothelin with regulation of pain. PVN, paraventricular nucleus; AVP, arginine vasopressin; ANGs, angiotensins; ET, endothelin; Ins, insula; Cing, cingulate gyrus; CVLM, caudal ventrolateral medulla; DMVNc, dorsal ventromedial nucleus of the vagus; NcAmb, nucleus ambiguous; NTS, nucleus tractus solitarius; PAG, periaqueductal grey; PFR, periphornical region; RVLM, rostral ventrolateral medulla; V1aR, vasopressin receptor; AT1aR, AT2R, angiotensin II receptors; ETAR, endothelin receptor; V1a, vasopressin V1a of type 1a; AT1a, angiotensin II of type 1a; AT2, angiotensin II of type 2; Mas, angiotensin-(1-7).

Pain increases release of AVP in the CdN and this effect is associated with stimulation of Ach release, which can be eliminated by application of AVP antagonists (V1ant, V2ant) [91, 92]. There is also evidence that AVP released from PVN neurons may regulate pain through the action exerted in the nucleus raphe magnus [92].

Systemically mediated effects of vasopressin. Systemically released AVP reaches the heart and coronary vessels and can regulate cardiac metabolism and blood oxygenation [59]. V1a receptors have been identified in the human heart, and their expression increases in patients with heart failure [93]. Earlier experiments performed on dogs have demonstrated that AVP acting on coronary vessels causes vasodilation of the epicardial vessels that is mediated by nitric oxide; however, it also elicits vasoconstriction and resistance of coronary vessels. Both effects require activation of V1aR (Table 1) [94].

A recent study in humans undergoing cardiopulmonary procedures revealed significant differences in responsiveness of coronary arterioles to vasopressin in patients with diabetes mellitus and in non-diabetic subjects [95]. Specifically, vessels from patients with poorly controlled diabetes mellitus responded with significantly stronger vasoconstriction than vessels from non-diabetic subjects. The difference could be eliminated by application of a V1aR antagonist (SR49059). In addition, cardioplegic arrest was followed by greater elevation of V1aR in vessels of diabetic patients and resulted in elevation of V1bR in the atrial myocardium of the diabetic subjects, while it was not effective in the atria of non-diabetic subjects [95].

Clinical evidence for the importance of AVP in pain regulation. Several years ago, it was reported that patients in the emergency department complaining of pain have significantly higher plasma AVP levels than control subjects [96]. Subsequently, other studies provided evidence that blood AVP and copeptin concentrations are significantly elevated in patients with acute myocardial infarction. Accordingly, it has been proposed that measuring copeptin concentration may help to identify patients with low and high risk of mortality [97, 98, 99].

Interaction of vasopressin with angiotensin, endothelin and other cardiovascular peptides. Cardiovascular disorders generated by cardiac infarction and/or cardiac failure cause release of several other cardiovascular compounds, which presumably interact with vasopressin in the regulation of pain. In this review, we have focused mainly on components of the renin-angiotensin system (RAS) and endothelin because the knowledge of their role in regulation of pain is relatively well advanced. Several components of RAS, such as angiotensinogen, angiotensin converting enzyme (ACE), and angiotensin receptors (AT1aR, AT2R), have been identified in the sensory dorsal root ganglia and in the brain structures involved in the regulation of pain [71, 100, 101, 102, 103, 104, 105]. Some of them are co-localized with substance P and calcitonin-gene related peptide, which are involved in the regulation of pain [105]. Studies on mice have shown that administration of angiotensin II intensifies nociceptive behavior through activation of angiotensin receptors (AT1R) and activation of p38MAPK [104]. In the rat model of peripheral neuropathy, administration of losartan attenuates allodynia and reduces expression of inflammatory proteins (Tumor necrosis factor-$\alpha$ (TNF-$\alpha$), Tumor necrosis factor receptor 1 (TNFR1), Glial fibrillary acidic protein (GFAP)) in the dorsal root ganglia. Furthermore, blockade of AT1R by administration of telmisartan in Wistar rats with diabetic neuropathy significantly elevates the mechanical nociceptive threshold, which is reduced in diabetic subjects [102, 103]. In contrast, administration of angiotensin-(1-7) [Ang-(1-7)] induces antinociceptive behavior, which is associated with p38 mitogen activated protein kinases (p38MAPK) inhibition (Table 1). In addition, it reduces nociceptive responses to substance P (SP) and N-methyl-D-aspartate receptor (NMDA), which co-localize in the spinal cord with MAS oncogene receptor, which is a G protein coupled receptor activated by Ang-(1-7) [106, 107].

Cooperation of AVP, Ang II, and endothelin-1 (ET-1) with the autonomic nervous system in the regulation of cardiac pain is summarized in Fig. 1 Beneficial antinociceptive effect of stimulation of AT2R has been described in inflammatory and neuropathic pain [108].

Although substantial evidence shows that acute myocardial infarction causes strong activation of RAS and elevation of plasma Ang II concentration [71, 109, 110], not many studies have assessed the role of RAS in regulation of pain in cardiac pathology. Thus far, it has been shown that administration of ACE inhibitors intensifies pain in fibromyalgia; however, it is likely that this effect is mediated by decreased degradation of bradykinin [101]. ACE inhibitors proved to be effective in the treatment of cardiac syndrome X [9, 111] and were able to reduce pain during exercise [112].

Plasma ET-1 level is significantly elevated in patients with angina and myocardial infarction [113, 114, 115] and its level markedly increases during coronary artery spasm [116, 117]. There is also evidence that acute myocardial infarction induces release of endothelin from the heart, as elevation of ET-1 is higher in coronary blood than in systemic blood [118]. Experiments on cats suggest that myocardial infarction causes activation of endothelin receptor of type 1A (ET1A) receptors on cardiac sensory neurons of the thoracic root ganglia [119]. Patients suffering from chest pain associated with normal coronary arteriography have higher levels of plasma endothelin during the treadmill exercise test than control subjects. The latter finding suggests that endothelin may influence intensity of chest pain in patients with syndrome X [9, 120].

Systemic levels of other factors, such as orexin [121, 122], brain natriuretic peptide [123], and insulin [124], are also elevated during pain and it is likely that they may participate in the regulation of cardiac ischemic pain.

Multiple studies draw attention to a diurnal periodicity of cardiac pain sensation. Onset of angina pain in patients with myocardial infarction shows a circadian rhythm with peaks between 06:00 h and 12:00 h and 20:00 h and 23:00 h [125, 126, 127, 128]. At present, the origin of the circadian periodicity of pain sensation is not fully elucidated, but there is evidence that the periodicity may follow the circadian rhythmicity of other brain functions, including secretion of neuroendocrine factors such as melatonin, vasopressin, components of RAS, and endothelin. Secretion and action of cardiovascular hormones shows circadian variability and effectiveness of a pharmacological medication may depend on the time of application of the treatment [128, 129, 130, 131, 132, 133].

There is strong evidence that vasopressin plays an essential role in synchronization of pain sensation and cardiovascular parameters with environmental light [60, 134]. Vasopressin is present in retinal cells projecting to the suprachiasmatic nucleus (SCN) and application of light evokes its release in this nucleus. Furthermore, secretion of AVP in the SCN manifests circadian rhythmicity, showing a peak in the early morning and a decline in the late afternoon [135, 136]. There is also evidence that vasopressin significantly contributes to the pacemaker function of the SCN. Neurons expressing AVP form a large population of SCN neurons, while the AVP promoter, which is a target for CLOCK/BMAL1 transcription factors, participates in generation of the circadian rhythm [134, 137, 138]. Experiments on mice provided evidence that AVP resets the clock function of the SCN through actions exerted on V1a/V1b receptors [139]. In humans, plasma vasopressin concentration shows a nocturnal increase with a peak between 24:00 h and 04:00 h [67, 140]. The nocturnal elevation of plasma AVP concentration is markedly attenuated in elderly subjects [141]. In post-stroke patients and patients with cardiac failure, the circadian rhythm of AVP secretion is abolished [141].

It is likely that vasopressin cooperates with Ang II and endothelin in the regulation of circadian rhythmicity. For instance, the circadian pattern of blood pressure fluctuations and pressor responses to restrain stress are significantly affected by systemic infusion of Ang II [142]. It appears that Ang II, acting on angiotensin II receptor of type 2 (AT2R), exerts a short-lasting antinociceptive effect, whereas prolonged stimulation of these receptors increases nociception at the beginning and end of the light phase. Both effects are attenuated by application of an AT2 antagonist (ditrifluoroacetate (PD 123319)) [143]. Inverted diurnal pain sensation was found in spontaneously hypertensive rats (SHR) and this dissimilarity could be corrected by blockade of AT1R with losartan [144].

Evidence is emerging that ET-1 may also belong to a family of cardiovascular peptides showing clear circadian rhythmicity. In the rat, expression of ET-1 mRNA in the SCN shows a peak around 04:00 h, whereas in the heart and lungs, it culminates between 12:00 h and 20:00 h [145]. Studies in mice suggest that plasma ET-1 level peaks during the night active period [146]. In human subjects, circulating ET-1 circadian peaks are present in the morning and afternoon [146, 147, 148]. The vasodilatory responses of endothelin and acetylcholine to iontophoresis show two peaks at 20:00 h and 08:00 h [147].