Site(s) of Negative Feedback in the Hypothalamic-Pituitary-Adrenocortical Axis

Feedback is primarily exerted in the brain under normal conditions, with only moderate increases in mean corticosteroid levels or low-magnitude acute stressors. Although the CRH and AVP neurons in the PVN are plentifully supplied with GRs, this site does not appear to be inhibited directly by the naturally occurring corticosteroids.

Instead, the corticosteroids appear to inhibit the activity of neuronal pathways that eventually impinge on CRH and AVP neurons. At high sustained corticosteroid levels, there is also clearly negative feedback on anterior pituitary corticotrophs.

Feedback in the Brain. Microinfusions of low amounts of MR and GR antagonists into the cerebral ventricles or into specific brain sites augment the magnitude of ACTH responses to a variety of stressors, strongly suggesting that stress-induced glucocorticoid secretion decreases the stimulatory input to the PVN during both the fast- and intermediate-feedback time domains.

Small crystalline implants of corticosterone placed chronically into prefrontal cortex, the preoptic area, and the dorsal hippocampus have been shown to reduce stress or adrenalectomy-induced ACTH secretion. However, it appears that steroids implanted into brain must bathe those neurons that are activated by the specific stressor used to excite the HPA axis. That is, there appear to be distributed sites at which steroids may inhibit HPA responses to stressors, not a single feedback site that globally inhibits PVN neuronal responses to stressors.

Corticosteroids, acting through both MRs and GRs, have been shown to alter neuronal properties and activity in the hippocampus, and antagonists infused into both the amygdala and hippocampus have been shown to alter conditioned behaviors. However, it has not yet been directly demonstrated that implants in these sites alter the amount of ACTH secreted in response to a stressor.

It may well be that corticosteroid receptors in these sites have little to do with modulating CRH and AVP secretion directly. In the brain, it appears that for feedback effects of steroid implants to be observed on ACTH secretion the steroid must act on sites that are also activated by stress; the feedback action is selective and specific.

Feedback at the Pituitary. Corticotrophs contain GRs but not MRs, and the direct inhibitory effects of corticosteroids are evident at the anterior pituitary. The concentrations of the naturally occurring steroids required to effect the inhibition of ACTH secretion are relatively high. This is probably because corticotrophs contain the high-affinity CBG corticosteroid binder apparently taken up from the circulation.

The CBG in the corticotrophs sequesters the natural corticosteroids and high concentrations (10^-8 M) are required to observe inhibition, although dexamethasone, which does not bind to CBG, is a more effective inhibitory agent.

Moreover, corticotrophs tend to maintain sensitivity to AVP, although sensitivity to CRH decreases when corticosteroid levels are elevated. The relative insensitivity of the corticotroph to corticosteroid inhibition is useful because it allows the corticotroph to be the site of last resort, requiring the highest levels of signal, for the global inhibitory control of the HPA axis.