Although it was originally thought that the sole locus of hypocretin production was in hypothalamus, some of the peptide may be produced in testis.32'38 Whether the hypocretin apparently produced in testis, or peptide produced in brain, gains access to the general circulation in sufficient amounts to act as a true hormone is unknown. However, hypocretin receptors are present in a variety of peripheral tissues, including the adrenal gland.33'42-45 In vitro studies have demonstrated significant actions of the hypocretins on adrenal steroid and catecholamine release. Cultured rat zona fasciculata and reticularis cells released increased amounts of corticosterone under basal, but not stimulated conditions.46 Orexin A also increased cAMP formation in these cells and the effect on corticosterone release was blocked by a protein kinase A inhibitor. No significant effects of orexin on basal or agonist stimulated aldosterone secretion were observed in zona glomerulosa cell cultures. These authors also reported that s.c. injection of orexin A stimulated corticosterone, but not aldosterone secretion in conscious rats.46 In another publication, these same authors reported that s.c. injection of orexin A stimulated ACTH, corticosterone and aldosterone secretion in rats.47 They then followed those acute administration studies with a report that prolonged, systemic orexin administration (seven days) failed to alter adrenal weight or morphology but resulted in elevated concentrations of corticosterone and aldosterone, but not ACTH, in plasma.48 Others have contributed similar observations.49
Using dispersed cell preparations of human adrenal tissues, Mazzochi and colleagues reported that orexin A but not orexin B stimulated basal cortisol secretion and enhanced the secretory response to angiotensin II, endothelin and ACTH.42 No significant effects were observed on aldosterone or catecholamine release from adrenal slices. As was the case in their studies on rat tissues, both hypocretin receptors were detected in human adrenal gland extracts, in this case the mRNA for the hcrtrl was most abundant. Similarly, as in rat tissues, the effect of orexin A appeared to be via activation of adenylyl cyclase. These same investigators also demonstrated the expression of the hcrtr2, but not the hcrtrl, gene in human pheochromocytomas and the ability of orexin A and orexin B to stimulate catecholamine release from these cells. The hcrtr2-mediated effect of the orexins was mediated not by adenylyl cyclase, but instead via activation of PLC and PKC.42 Opposing results were reported in an immortalized, rat pheochromocytoma cell line, PC12 cells.50 Here orexin A and B reduced PACAP-stimulated transcription of the gene for tyrosine hydroxylase, the rate limiting step of catecholamine synthesis, and inhibited PACAP-induced dopamine secretion, probably via an action on the hcrtr2 receptor. These effects were mediated at least in part via activation of adenylyl cyclase.
How does one reconcile these apparently disparate results? Could it be merely a species difference? Are the effects in vitro dependent on culture conditions? One thing is certain, when administered systemically in vivo, the hypocretins appear to increase glucocorticoid secretion. Is this pharmacology or do these results indicate a direct, physiologically relevant role for the peptides in adrenal function? Is there enough hypocretin produced in adrenal gland, or delivered via the general circulation, to exert autocrine or endocrine effects in the gland? Is there a relationship between these findings and the ability of glucocorticoids to increase hypocretin gene expression in hypothalamic neurons?51 Are the secretory dynamics of mineralocorticoids and glucocorticoids altered in genetic models of orexin absence, and if so do those alterations support the stimulatory effects of the peptides advanced by the studies of Malendowicz and colleagues?
6. AN INTEGRATIVE VIEW OF THE ACTIONS OF THE HYPOCRETINS IN THE HYPOTHALAMO-PITUITARY-ADRENAL AXIS (Figure 1)
There is consensus that the hypocretins act in brain to stimulate the HPA axis, elevate sympathetic tone, and elicit behaviors typical of the stress reaction. These effects are not unique to the hypocretins. Several other peptides exert similar effects in brain to activate the HPA axis and sympathetic tone, while stimulating ingestive and locomotor activities.52-55 Is it possible that the ability of the hypocretins to stimulate the HPA axis by an action in brain is a reflection of, but not the primary response to, the generalized stress response?56 Is the effect of orexin any more important than the response of the HPA axis to those other peptides? Genetic models lend some insight into this question.
If the direct, pharmacological effects of the hypocretins to stimulate glucocorticoid or mineralocorticoid secretion are physiologically relevant, then one would expect that knockout mice in which the gene was deleted might manifest altered electrolyte status and metabolism. However, these animals at 13-15 weeks of age displayed normal plasma electrolytes and blood glucose levels.57 Hypocretin expression was already absent in the embryos, and thus endogenous hypocretin is not essential for development of the adrenal gland or for the ability to shift from a liquid diet to solid food at weaning. Interestingly, even though these animals have never "seen" orexin, their cataplexic phenotype can be rescued by ectopic hypocretin expression or acute peptide administration,58 thus the hypocretin receptors remain functional.
These animals have been useful for the examination of the physiologic relevance of the sympathostimulatory and behavioral actions of the hypocretins. In a resident-intruder model of emotional stress, the orexin knockout mice displayed a significantly lower cardiovascular and behavioral response to the introduction of the intruder, or when the knockout mouse was the intruder, to the novelty of the paradigm.59 These observations, taken together with the fact that basal blood pressures were lower in the knockout mice, strongly suggests that the effects of exogenous hypocretin on autonomic function27,56,60 have a physiologic correlate. It would be very interesting to measure circulating corticosterone levels in those animals to determine, in a similar manner, the role of endogenous hypocretin in the control of HPA axis activity. We would predict that basal corticosterone levels in these animals would be elevated and that hormone secretion in response to emotional stress would be enhanced, reflecting the loss of hypocretin's inhibitory effect in the pituitary gland.36,61
The transgenic model that perhaps best mirrors human narcolepsy is the ataxin-3 expressing mouse62 and now rat.30 These animals express normal levels of the hypocretins until early adulthood, when expression of the truncated Machado-Joseph disease gene (driven by the orexin promotor) product causes lethality in neurons expressing the protein. The mice display as adults the behavioral phenotype most similar to human narcolepsy and in addition develop late-onset obesity, and what appears to be insulin resistance. Transgenic mice ate less food, but consistently gained more weight than wild type littermates. The decreased food intake may have been due to a loss of the orexigenic action of the hypocretins. These transgenic animals did display less spontaneous motor activity than controls during the dark phase, when the cataplexic attacks were most prevalent, and therefore a decreased metabolic rate may have contributed to the weight gain, although his was not directly assessed. This is suggested by the observation that in normal animals orexin stimulates oxygen consumption, an effect that reflects increased metabolic rate.63
The orexin/ataxin-3 mice present a potential model for the physical appearance of many human narcoleptics, in whom obesity and type 2 diabetes are common.64,65 This combination of obesity and insulin resistant diabetes is also a common feature of Cushing's Syndrome patients, who suffer from excess production of cortisol, and display an obese body habitus and insulin resistance secondary to the elevated cortisol levels.66 Could it be that in the absence of orexin's actions in pituitary gland to "buffer" the release of ACTH in response to endogenous CRH, more corticosterone is secreted in the knockout mice and, that in human narcoleptics some of the increased weight gain and insulin resistance are due to a similar loss of orexin action? This in fact is our working hypothesis and we are beginning a collaborative study in a human population of narcoleptics to examine this possibility.
Was this article helpful?