|Home | About | Journals | Submit | Contact Us | Français|
The hormone corticotropin (ACTH) is employed as therapy for diverse neurological disorders, but the mechanisms for its efficacy remain unknown. ACTH promotes the release of adrenal steroids (glucocorticoids), and most ACTH effects on the central nervous system (CNS) have been attributed to activation of glucocorticoid receptors. However, in several human disorders, ACTH has therapeutic actions that differ qualitatively or quantitatively from those of steroids. This study tested the hypothesis that ACTH directly influences limbic neurons via the recently characterized melanocortin receptors and focused on the effects of ACTH on the expression of corticotropin-releasing hormone (CRH), a neuropeptide involved in neuroimmune functions and in certain developmental seizures. The results demonstrated that ACTH potently reduced CRH expression in amygdala neurons. This down-regulation was not abolished by experimental elimination of steroids or by blocking their receptors and was reproduced by a centrally administered ACTH fragment that does not promote steroid release. Importantly, selective blocking of melanocortin receptors prevented ACTH-induced down-regulation of CRH expression. Taken together, these data indicate that ACTH activates central melanocortin receptors to modulate CRH gene expression in amygdala, supporting the notion that direct, steroid-independent actions of ACTH may account for some of its established clinical effects on the CNS.
The hormone corticotropin (ACTH) has been used extensively to treat neurological disorders. In some auto-immune disorders (eg, opsoclonus-myoclonus) and in certain developmental seizure disorders (eg, infantile spasms), ACTH possesses increased efficacy or a different spectrum of effects compared to adrenal steroids.1–4 However, the mechanisms for these singular actions of ACTH on the central nervous system (CNS) have remained unresolved.1,5
Native ACTH is synthesized in the pituitary and functions to stimulate the adrenal cortex to release steroids into the blood stream as part of the neuroendocrine stress response. Neurons containing ACTH have been localized to the CNS, particularly the hypothalamus, and ACTH-immunoreactive cell bodies or fibers have also been described in amygdala, cerebral cortex, brainstem, and cerebellum.6,7 In contrast to pituitary ACTH, the functions of CNS ACTH have not been well-defined. Evidence from both human and animal studies has suggested that CNS ACTH may function as a neurotransmitter/neuromodulator.1,8 Indeed, central physiological roles for ACTH, including modulation of learning and memory processes and facilitation of arousal states, have been suggested, but the mechanisms for these actions of ACTH have remained unclear.1,8 Characterization of the melanocortin receptor family, consisting of several members possessing binding affinity for ACTH,9,10 has yielded insight regarding the likely sites and mechanisms of ACTH effects. However, the downstream effects of activation of these receptors and the specific molecular changes that can lead to the clinical effects of ACTH on CNS function have not been elucidated.
Pituitary ACTH synthesis and secretion are regulated by the peptide corticotropin-releasing hormone (CRH), and both hormones participate in the neuroendocrine response to stress. Furthermore, ACTH reduces CRH gene expression in hypothalamus as part of the neuroendocrine feedback loop of the stress response.11 ACTH and CRH are found in close proximity in the amygdala, and complex interactions between the ACTH- and CRH-expressing neuronal systems have been suggested:12–14 For example, ACTH-immunoreactive (ir) fibers have been demonstrated in the amygdaloid central nucleus, a major limbic locus of CRH-expressing neurons. Could regulation of CRH expression account for some of the actions of ACTH on CNS function?
Indeed, down-regulation of CRH by ACTH provides an attractive mechanism for some of the clinical effects of the latter. CRH has been implicated in neuroimmune functions,15–17 in learning and memory,18,19 and in developmental seizures.20,21 In the clinical context, CRH modulates immune function in a complex manner,15,22 and regulation of its actions may mediate some of the efficacy of ACTH in autoimmune disorders. Importantly, ACTH is much more potent than steroids in infantile spasms.3,4 Whereas the etiology of this disorder remains entirely unclear, evidence from human23,24 and animal20,21 data supports a potential role of stress-activated proconvulsant actions of CRH on limbic neurons. Thus, the all-or-none effects of ACTH on these seizures may involve reduction of CRH availability at limbic synapses.21,25
The goal of the current studies was to examine whether ACTH directly influenced CRH gene expression in the immature rat amygdala. Because steroids also regulate CRH expression in this region, this study focused on discerning steroid-independent effects, testing the hypothesis that ACTH suppresses CRH expression in amygdaloid central nucleus (ACe) directly, via activation of central melanocortin receptors.
ACTH (Acthargel; Rhone-Poulenc Rorer, Collegeville, PA) was administered into the peritoneal cavity (i.p.) at a high dose (80 I.U./kg), because of the limited blood–brain barrier penetration of the peptide.26,27 ACTH4–10, an analog that binds melanocortin receptors but does not induce steroid secretion from the adrenal, was a gift from Dr. J. Rivier. ACTH4–10 was dissolved in 0.9% saline and infused into the cerebral ventricles (i.c.v.) in a dose of 1 μg. RU 38486 (a gift from Roussel UCLAF, Romainville, France) was dissolved in 2% ethanol/98% polyethelene-glycol and injected i.p. (100 mg/kg).28 SHU9119 (courtesy of Dr. K. Yagaloff, Roche) was dissolved in 0.9% saline and infused i.c.v. (1 nmol).29
Several experiments were carried out to test the hypothesis that ACTH alters CRH gene expression in ACe independently of the effects of steroids. These experiments were performed on developing rats (postnatal days 9–11) for several reasons: 1) ACTH is used frequently and in large doses in human infants during a stage of brain development generally equivalent to that of the immature rat in the second postnatal week.30 2) Previous work from our laboratory has characterized the levels and regulation of CRH mRNA expression in ACe during this age.21,31 Experimental groups were chosen to manipulate independently levels of ACTH and of circulating steroids (glucocorticoids). These groups and the respective hormonal profiles of each are schematized in the Table.
For obvious technical reasons, experiments were performed in “batches.” However, each included both experimental animals and littermate controls, and tissue processing and data analysis were performed to include all experiments.
Offspring of timed-pregnancy Sprague Dawley-derived rats (Zivic-Miller, Zelienople, PA) were used in this study. Dams were maintained in an NIH-approved facility on a 12 hour light/dark cycle (lights on at 7 AM). Cages were monitored every 12 hours for the presence of pups, and the date of birth was considered day 0; litters were culled within 24 hours of delivery to 12 pups. Rats were killed by decapitation, and brains were rapidly removed and frozen on dry ice for subsequent in situ hybridization. Trunk blood was collected at the time of sacrifice for measurement of plasma ACTH and corticosterone (the major glucocorticoid in rat) levels using commercial radioimmunoassay (RIA) kits (Inc-star, Stillwater, MN, and ICN, Irvine, CA, respectively). All experiments were carried out according to NIH guidelines for the care of experimental animals and were approved by the Institutional Animal Care Committee.
Surgery was performed with the animals under halothane anesthesia 24 hours prior to experiments, to permit adequate recovery from the procedure. To eliminate endogenous steroids, bilateral adrenalectomy was performed via dorsal incisions. Sham adrenalectomy consisted of opening the peritoneal cavity without removal of the adrenals. Incisions were closed using acrylic glue (Krazy Glue; Elmer’s Products, Inc., Columbus, OH). The completeness of adrenalectomy was verified by visual inspection as well as by hormonal measurements: Animals with corticosterone levels >0.2 μg/dl were considered not fully adrenalectomized and were eliminated from the study. For i.c.v. infusions, stainless steel cannulae were implanted into the lateral cerebral ventricle using an infant-rat stereotaxic apparatus (Kopf, Tujunga, CA) and stereotaxic coordinates (P 0.7 mm, L 2 mm, and V 3.3 mm from bregma), as published elsewhere.32,33 When both adrenalectomy and cannula were required, surgical procedures were performed sequentially, during a single anesthesia session.
In situ hybridization (ISH) histochemistry was performed as described previously.31 Briefly, 20 μm coronal sections were collected on gelatin-coated slides and stored at −80°C. Sections were thawed, air dried, fixed in paraformaldehyde, dehydrated, and rehydrated through graded ethanols, then exposed to 0.25% acetic anhydride in 0.1 M triethanolamine (pH 8) and dehydrated. Prehybridization and hybridization steps were performed at 40°C in a humidified chamber. Following a 1 hour prehybridization, sections were hybridized overnight (20 hours) with a deoxyoligonucleotide probe complementary to the coding region of CRH mRNA and 3′-end-labeled with 35S-dATP. Sections were then washed and apposed to film (Hyperfilm β-Max; Amersham, Arlington Heights, IL) for 7–14 days. Selected sections were also dipped in emulsion (NTB-2; Eastman Kodak, Rochester, NY) and exposed for 3 or 4 weeks.
Semiquantitative analysis of CRH mRNA ISH signal was performed on digitized films using the ImageTool software program (UTHSC, San Antonio, TX) as described elsewhere.31,34 For balanced comparison among the different experimental groups, five anatomically matched sections per region per brain were analyzed. Unbiased methods for section sampling have been described in detail elsewhere,34 and all analyses were performed without knowledge of treatment group (blindly). 14C standards were used to ascertain that analyzed signals fell in the linear range, and values are expressed as relative optical density units (ROD). Differences among groups were determined using a one-way analysis of variance (ANOVA) or Student’s t test, as appropriate (Prism GraphPad, San Diego, CA), and significance levels were set at p < 0.05. Further analysis used Tukey’s multiple comparison posthoc tests to determine the source of the detected significance in the ANOVAs.
CRH mRNA levels in ACe were significantly reduced by systemic administration of ACTH: 82 ± 4 and 53 ± 12 ROD in the control and ACTH groups, respectively (a 35% reduction; t = 2.67; df = 23; p < 0.02). Figure 1 shows CRH mRNA expression in representative photomicrographs of coronal brain sections derived from both groups; Figure 2A depicts semi-quantitative analyses of the intensity of the CRH mRNA signal.
ACTH, given systemically, led to a significant reduction in CRH mRNA expression in ACe also in the absence of endogenous steroids (Fig 2A,C). Thus, in adrenalectomized rats—as in intact littermates—ACTH resulted in a significant (33%) suppression of CRH mRNA levels in ACe compared to vehicle-injected, sham-operated controls (56 ± 8 vs. 84 ± 4 ROD units, respectively; t = 3.1; df = 18; p < 0.01), indicating that endogenous steroids were not required for this effect. It should be noted that adrenalectomy—eliminating steroids and the consequent negative feedback on ACTH secretion—resulted in high levels of intrinsic, endogenous ACTH (Fig 2B). These ACTH levels, by themselves, were sufficient to down-regulate significantly CRH mRNA in ACe (34% reduction from sham-operated controls; see Fig 2A,B; t = 2.3; df = 18; p < 0.05). No further suppression of CRH expression resulted from adding exogenous ACTH to these animals, indicating that the high endogenous ACTH levels after adrenalectomy maximally activated melanocortin receptors, and further reduction of CRH expression was not possible via this mechanism. Additional reduction of CRH mRNA levels required an independent mechanism, i.e., blocking of the enhancing action of glucocorticoid receptor activation (see below).
The expected changes in plasma stress hormones were observed after the experimental interventions reported here (Fig 2B,C). ACTH levels were significantly elevated in animals given exogenous ACTH (695 ± 50.7 pg/ml) and in adrenalectomized animals (296.3 ± 53.4 pg/ml) compared to controls (118.5 ± 16.37 pg/ml). In addition, additive effects of administration of exogenous ACTH and adrenalectomy were apparent, leading to plasma ACTH levels of 1475 ± 165.2 pg/ml, significantly higher than those of either treatment alone [F(24) = 12; p < 0.0001, one-way ANOVA]. Plasma corticosterone levels increased significantly after ACTH administration (6.8 ± 0.9 vs. 1.05 ± 0.19 μg/dl in controls), whereas corticosterone levels in adrenalectomized animals were negligible [F(28) = 42.1; p < 0.0001, one-way ANOVA].
Similarly to systemically administered ACTH, the analog ACTH4–10 infused i.c.v. resulted in a significant down-regulation of CRH mRNA expression in ACe [45% reduction from i.c.v. vehicle-infused control levels (Fig 3); t = 4.04; p < 0.001]. The dose of ACTH4–10 used here was relatively small and was therefore unlikely to reach the adrenal when given i.c.v. In addition, this ACTH analog does not stimulate corticosterone secretion from the adrenal.1 Indeed, as shown in Figure 3, plasma corticosterone levels of ACTH4–10-infused rats were not higher than those of vehicle-treated controls, whereas systemic ACTH led to a marked increase in plasma corticosterone at the time of sacrifice (see Fig 2C). These data suggest that ACTH4–10 (and ACTH) reduced CRH expression in ACe via a central rather than a steroid-mediated peripheral mechanism.
Further insight into the mechanism by which ACTH down-regulated CRH expression in ACe was gained by administering the melanocortin receptor-4 blocker SHU9119 and the glucocorticoid receptor antagonist RU 38486. As is evident from Figure 4A, in the setting of high endogenous plasma ACTH levels achieved by adrenalectomy, administration of SHU9119 blocked ACTH-induced reduction of CRH mRNA in ACe [90 ± 11 ROD in (SHU+ Adx) vs. 55 ± 12 ROD in adrenalectomy alone]. In contrast, when RU 38486 was given to intact animals, together with ACTH, it failed to block the effects of ACTH administration on CRH mRNA expression in ACe [30 ± 9 ROD in (ACTH + RU 38486), a 63% reduction from i.p.-injected controls; t = 5.3; df = 8; p < 0.001; Fig 4B]. These data indicate that activation of melanocortin receptors, but not of glucocorticoid receptors, is required for the actions of ACTH on CRH expression in ACe.
The direct, steroid-independent down-regulation of CRH mRNA expression by ACTH was specific to ACe. Analysis of ACTH and steroid plasma levels and the corresponding CRH mRNA levels in hypothalamus and hippocampus (Fig 5) revealed distinct regional patterns: In hypothalamic paraventricular nucleus (PVN), CRH mRNA levels correlated inversely with plasma glucocorticoids, consistently with previous reports.35,36 ACTH administration, inducing high plasma corticosterone levels, reduced CRH mRNA expression 40% in PVN (290 ± 40 ROD in ACTH-treated vs. 480 ± 30 ROD in controls), whereas adrenalectomy, eliminating plasma glucocorticoids, enhanced PVN CRH mRNA levels 108% [830 ± 150 ROD in adrenalectomy vs. 400 ± 30 ROD in sham adrenalectomy controls; F(24) = 6.8; p < 0.01, one-way ANOVA]. No consistent correlation of PVN CRH mRNA expression with plasma ACTH levels was noted.
In hippocampus, exogenous ACTH did not alter CRH mRNA expression in any of the analyzed fields (CA1, CA3, and dentate gyrus; Fig 5). Steroid elimination via adrenalectomy significantly decreased CRH mRNA levels in CA1 neurons [F(25) = 10.44; p < 0.0001, one-way ANOVA], known to express a high density of glucocorticoid receptors, but not in the granule cell layer of the dentate gyrus.
This study demonstrates direct actions of systemically or centrally administered ACTH on limbic neurons that are independent of adrenal steroids. ACTH-induced reduction of CRH expression in amygdala may provide a mechanism for the efficacy of ACTH in several human disorders. Importantly, these effects of ACTH involve activation of specific melanocortin receptors, which may provide new and important therapeutic targets.
ACTH has been used for treating a variety of neurological disorders. An established effect of this hormone includes activation of adrenal steroid secretion, and the latter has generally been considered as the major mechanism of action of ACTH. However, release of adrenal steroids leading to activation of steroid receptors cannot fully explain the spectrum and magnitude of the therapeutic effects of ACTH.
For example, both ACTH and steroids have been shown to be effective in treating certain childhood seizures,37,38 particularly infantile spasms.3,20,39 However, for this disorder, ACTH is a significantly more powerful agent than glucocorticoids.3,4 Indeed, both uncontrolled3 and prospective, blinded, and controlled studies4 have shown that ACTH in very large doses is significantly superior to large doses of steroids, controlling the seizures and normalizing the EEG in 86–88% of individuals compared to 35–40%.3,4,39 It should be noted that the levels of plasma steroids achieved after administration of the typical doses of prednisone should be sufficient to saturate neuronal glucocorticoid receptors.40 In addition, most studies demonstrate the superior efficacy of higher compared to relatively lower doses of ACTH3,4,41,42 (but see Hrachovy et al43), although these “lower” doses suffice to stimulate maximally secretion of steroids from the adrenal.
These clinical data raise several issues: First, if ACTH acts via steroids, then maximal doses of both agents should have similar effects. Second, once doses of ACTH that suffice to release endogenous steroids maximally are used, then higher doses should not have additional benefits. The enhanced potency of ACTH compared to steroids and the superiority of extremely high doses of ACTH compared to lower one are consistent with direct, steroid-independent actions of ACTH within the CNS. In addition, the poor penetration of systemically given ACTH through the blood–brain barrier26,27 may underlie the requirement for large systemic doses.
The studies described here support the notion of direct CNS actions of ACTH. Using doses analogous to those used clinically, we show that systemic ACTH influences gene expression in specific brain regions regardless of the presence or absence of adrenal steroids. We further show that these effects are generated by ACTH given either systemically or directly into the cerebral ventricles. These effects persist when glucocorticoid receptors are blocked, but not when a subtype of ACTH receptors (melanocortin receptors) is antagonized, showing that activation of these latter receptors is required and sufficient for these actions of ACTH. These findings are important for several reasons. First, they provide a mechanism for the direct actions of ACTH on neurons, in both therapeutic and physiological contexts. In addition, they demonstrate that ACTH may indeed be a native ligand for melanocortin receptor-4.
Melanocortin receptors, a family of transmembrane G-protein-coupled receptors, have been demonstrated in regions subserving ACTH-mediated functions (eg, substantia nigra for grooming and amygdala for learning and memory). In addition, activation of central melanocortin receptors plays a role in appetitive and grooming behavior.10,44 However, several endogenous members of the proopiomelanocortin family bind to melanocortin receptors, and whether ACTH is an important endogenous ligand of these receptors has not been determined. The current study demonstrates categorically that ACTH may activate melanocortin receptor-4. Indeed, in adrenalectomized rats, the effect of increased endogenous ACTH on CRH expression was abolished by administration of the melanocortin receptor-4 antagonist SHU9119 directly into the CNS. These data strongly support the notion that endogenous ACTH activates central melanocortin receptor-4 to influence CRH synthesis. Furthermore, the receptor implicated here in the effects of ACTH on CRH gene expression, melanocortin receptor-4, has been demonstrated in a number of brain regions, including hippocampus and all three major divisions of the central amygdaloid nucleus.45 It is therefore interesting to note that, unlike the hippocampus (expressing melanocortin receptors but devoid of endogenous ACTH) and the hypothalamus (containing ACTH but lacking melanocortin receptor-4), ACe is a site where both ACTH and melanocortin receptors coexist. Therefore, the current study is consistent with a scenario in which ACTH, released from fiber terminals in ACe, acts locally on melanocortin receptors located on CRH-expressing neurons. ACTH, given in clinical settings should have similar effects.
Interestingly, the 4–10 fragment of ACTH that does not release adrenal steroids also led to the depression of CRH gene expression. Earlier clinical studies have shown that shorter fragments, such as ACTH4–9, that failed to release adrenal steroids were not efficacious in treating infantile spasms.46,47 Currently available information explains the apparent discrepancy: New information about the binding of ACTH fragments to melanocortin receptors shows that, although the 4–10 fragment used here activates these receptors, the 4–9 fragment does not.10,48 Thus, the current data are consistent with activation of melanocortin receptor as the mechanism for the observed effects of ACTH on amygdala neurons.
Modulation of CRH synthesis may prove to be one of the mechanisms by which ACTH exerts its profound and diverse effects on CNS function. For example, CRH has been shown to play an important modulatory role in affective disorders49 and neuroimmune function.15,17,22 Cytokines including interleukins 1 and 2 up-regulate CRH expression in specific brain regions,16,50 and these cytokines, in turn, are induced by stress, by mechanisms that may involve CRH release.17 CRH released from central neurons interacts with immune responses both directly and by activation of steroid release.17,22 In addition, a role for CRH as an excitatory neuromodulator that can provoke severe seizures early in life has been emerging.20,21,51 These seizures originate in the amygdala and involve other limbic regions.51 A hypothesis for the involvement of CRH in infantile spasms has been proposed20,21 and is supported by abnormalities of the CRH-ACTH-steroid axis found in infants with this disorder.23,24 It has been proposed that the myriad etiologies of infantile spasms share the characteristic of activating the responses to stress, thus leading to enhanced CRH expression in the amygdala.20,31 This leads to excessive neuronal excitability25,33 and seizures.21,51 By repressing CRH expression, ACTH may decrease this hyperexcitability, ameliorating infantile spasms.20 Indeed, the time course and all-or-none effects of ACTH are consistent with such an effect.20,39,43,52
In summary, the current study provides a potential mechanism for the efficacy of ACTH for certain neuroimmune and developmental seizure disorders and highlights the ACTH-CRH system as a potential target for development of new therapeutic agents.
Supported by NIH grant RO1 NS28912 and R41 HD34975 (T.Z.B.), a system-wide University of California biotechnology-oriented predoctoral award (K.L.B.) and UCI undergraduate research (UROP) award (N.K.).
We thank Drs Bruce McEwen and Roger Cone for helpful critiques and discussions.