Pii: S0196-9781(01)00522-8

Adrenomedullin gene products have been localized to neurons in brain that innervate sites known to be important in the regulation of cardiovascular function. Those sites also have been demonstrated to possess receptors for the peptide and central administrations of adrenomedullin (AM) and proadrenomedullin N-terminal 20 peptide (PAMP) elevate blood pressure and heart rate in both conscious and anesthetized animals. The accumulated evidence points to a role of the sympathetic nervous system in these cardiovascular effects. These sympathostimulatory actions of AM and PAMP have been hypothesized to be cardioprotective in nature and to reflect the central nervous system (CNS) equivalent of the direct cardiostimulatory effects of the peptides in the periphery. This review summarizes the most recent data on the CNS actions of the adrenomedullin gene-derived peptides and suggests future strategies for the elucidation of the physiologic relevance of the already demonstrated, pharmacologic actions of these peptides. © 2001 Elsevier Science Inc. All rights reserved. 1. Localization of adrenomedullin (AM)-derived peptides and AM receptors in autonomic centers in brain By far the most comprehensive description of AM-like immunoreactivity in brain was published by Serrano and colleagues [26]. They identified AM-positive neurons throughout rat brain, but particularly in hypothalamus, thalamus, amygdala, midbrain, and medulla, all sites known to be important in the organization of autonomic tone [33]. In the hypothalamus Serrano and colleagues [26] expanded upon the earlier report by Ueta and coworkers [38] demonstrating co-localization of AM with vasopressin (AVP) and oxytocin (OT) in cells of the paraventricular (PVN) and supraoptic (SON) nuclei. Cell bodies containing AM were also identified [26] in the bed nuclei of the anterior commissure and stria terminalis, the lateral preoptic area and lateral hypothalamus. Immunopositive nerve fibers were identified throughout the hypothalamus, but in particular in the dorsomedial hypothalamic area (DMH), the arcuate nucleus, and the PVN and SON. In thalamus the major cell concentrations positive for AM were in the paraventricular thalamic area, the mediodorsal nuclei, and the zona incerta. In caudal regions, AM-like immunoreactive cell bodies were localized to the ventral tegmental area and substantia nigra of the midbrain. In the pons and medulla AM immunoreactivity was observed in neurons of the trigeminal nuclei, the locus coeruleus, raphe, nucleus ambiguus, intermediate reticular nucleus and the dorsal motor nucleus of the vagus. Nerve fibers were found throughout the medulla but particularly in area postrema (AP). In human brains, AM-immunoreactivity was reported in cells of the PVN and SON, as well in the infundibular nucleus, the human homolog of the arcuate nucleus in rat [25] and membrane preparations from all major regions of human brain displayed specific binding to iodinated AM [32]. Scatchard analysis conducted in membranes from cerebral cortex revealed a dissociation constant of 0.17 nM and maximal binding of 99.3 fmol/mg protein. Binding was not competed by calcitonin gene related peptide (CGRP) suggesting the presence of an AM-specific receptor. Abbreviations: A II angiotensin II; ACTH adrenocorticotropin; AM adrenomedullin; AP area postrema; AVP arginine vasopressin; CGRP calcitonin gene related peptide; CNS central nervous system; DMH dorsomedial hypothalamic area; FLI Fos-like immunoreactivity; i.v. intravenous; i.c.v. intracerebroventricular; LPS lipopolysaccharide; NP sodium nitroprusside; NTS nucleus tractus solitarius; OT oxytocin; PAMP proadrenomedullin N-terminal 20 peptide; PVN hypothalamic; paraventricular nucleus; SFO subfornical organ; SON supraoptic nucleus. * Corresponding author. Tel.: 1314-577-8633; fax: 1-314-5778554. E-mail address: [email protected] (W. Samson). Peptides 22 (2001) 1803–1807 0196-9781/01/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved. PII: S0196-9781(01)00522-8 2. What regulates CNS Production of AM? Shan and Krukoff [29] recently reported the presence of preproadrenomedullin mRNA in rat brain using in situ hybridization histochemistry. AM mRNA was detected in several autonomic centers in brain. Importantly, these authors presented the first evidence for regulation of regional gene expression demonstrating that lipopolysaccharide (LPS) treatment decreased transcript levels in PVN (magno and parvocellular components), SON, dorsal motor nucleus of vagus, and AP. Restraint stress lowered transcript level in PVN, SON, nucleus of the solitary tract (NTS), dorsal motor nucleus of the vagus, area postrema and subfornical organ (SFO). Water deprivation reduced AM mRNA levels in PVN and SON only. The AM gene is transcribed in both neurons and glia [12], and it is not clear that the changes observed by Shan and Krukoff [29] were limited to neuronal versus glial cell populations; however, their important results demonstrate at least pathologic regulation of the gene in brain and further draw attention to the anatomic framework that suggests a role for the peptides in the control of autonomic function. 3. Localization of single cells responding to AM Two approaches have been taken to identify cells responding to AM. One, which reveals direct effects, is the microinjection of peptide directly onto single neurons within a nuclear group. Using this approach Ferguson has identified cells in AP [2] and hypothalamic PVN(31, and AV Ferguson, personal communication) whose polarization state is altered by direct application of AM. Those cellular effects have physiologic consequences since these microinjections also resulted in significant alterations in blood pressure in these anesthetized rats [2,31]. A second experimental approach is to inject peptide either centrally or peripherally and examine the expression of the immediate early gene product c-fos. Expression of fos is thought to be a reliable indicator of cell activation; however, this approach is not as selective as single cell recording because it might fail to identify cells whose activity was inhibited and it does not identify the direct cellular site of action, instead also the network of cells that may be excited secondary to the action of the peptide. Just the same, this approach, when considered along with the apparent presence of receptor for the peptide injected can localize approximate sites of action. It also may provide insight into the multisynaptic chain of events initiated by the primary action of the injected peptide. Using this approach, Ueta and coworkers [39] described the induction of Fos-like immunoreactivity (FLI)in PVN and SON following intracerebroventricular (i.c.v.) administration of AM. In particular the expression of fos was localized to OT containing neurons and was prevented by pretreatment with the AM antagonist AM 22–52. Having demonstrated the activation of fos in these OT-ergic cells, these authors then conducted extracellular recordings in the PVN and identified direct effects of AM on cells identified as OT-producing neurons. In area postrema, Allen and Ferguson [1] had demonstrated direct cellular effects of microinjection of AM, that translated into significant alterations in blood pressure [31]. Shan and Krukoff [28] reasoned that the AP, a circumventricular organ lacking a blood brain barrier, might be a site where AM of either peripheral or central origin might act. They lowered blood pressure by peripheral infusion of AM or sodium nitroprusside (NP) and examined FLI in PVN, NTS, and AP. In response to similar lowerings of blood pressure more FLI positive cells were identified in PVN, NTS and AP in the AM than in the NP treated animals. No differences were observed between treatment groups in the SFO. These authors reasoned that the greater activation of fos in response to AM-induced hypotension, that was mirrored in magnitude by that of NP, meant that peripherally administered AM was exerting direct effects in brain. AM is produced in choroid plexus [11] and in cerebral microvessels [10] where receptors are present. Additionally, immunoreactive AM is present in cerebrospinal fluid [35]. Thus there may be actions in addition to those exerted solely in the peripheral vasculature. This possibility would not have been ruled out in the final set of experiments conducted by Shan and Krukoff [28]. In that experiment the AP was ablated and then the hypotensive challenge repeated. Fewer FLI positive cells were identified in PVN of AP ablated rats than in sham operated animals. This would suggest that the peripherally administered AM peptide gained access to the CNS via the AP; however, it does not rule out the possibility that AM crosses into brain at other sites and that the changes in FLI observed reflected the AP ablation itself and not altered AM passage into brain. AM may also be acting directly on the baroreceptor afferents themselves. Indeed Fukuhara [6] demonstrated that the baroreceptor activation following intravenous (i.v.) AM was attenuated when compared to that seen after a similar hypotensive challenge caused by NP. Their data would suggest that AM acts in the periphery to desensitize the baroreceptor reflex [6]. On the other hand, subsequent data from the same laboratory [13] indicated that, when administered centrally, AM augments the baroreflex. Thus multiple sites of action of AM within the afferents and the efferents of the baroreflex pathways may exist. These multiple sites of action might, then, confound the interpretation of experimental results comparing the reflex responses to AM and other vasodilators. It is clear, however, that centrally administered AM can stimulate fos expression in PVN and SON [27], and that the activation of neurons in SON and PVN may underlie the stimulatory effect of AM on OT release [39]. Recently, Ferguson and colleagues [31] demonstrated that microinjection of AM into PVN resulted in decreased arterial blood pressure in anesthetized rats, a finding that seems contradictory in light of the multiple studies demonstrating eleva1804 Taylor and Sampson / Peptides 22 (2001) 1803–1807