Immunotargeted lesions of paraventricular CRF and AVP neurons in developing rats reveal the pattern of maturation of these systems and their functional importance.

Pituitary ACTH secretion in the rat is controlled by a number of hypothalamic secretagogues, like CRF and AVP and by inhibitory feedback provided by glucocorticoids. During development, little is known about the precise regulation of ACTH release by hypothalamic neuropeptides and glucocorticoids. We used immunotargeted chemical PVN lesions to investigate the role of CRF and AVP neurons of the hypothalamic paraventricular nucleus (PVN) in the control of ACTH secretion in neonatal rats under basal conditions and 5 days after adrenalectomy (ADX). Neonates aged day (d) 4 or d14 were injected over the PVN with ricin A toxin associated with either non-specific antibodies (IgG/Tx), or monoclonal antibodies directed towards CRF (CRF/Tx) or AVP (AVP/Tx). Rats from each group received either sham surgery (SHAM) or were adrenalectomized (ADX). Pups were sacrificed 5 days after PVN treatment and adrenal surgery (d9 or 19). Plasma ACTH and corticosterone (B) levels were measured by RIAs. Changes in CRF and AVP expression in the PVN and other brain regions were determined by immunohistochemistry (ICC) and in situ hybridization. Injection of the toxin associated with IgGs did not have non specific effects on body weight gain, neuropeptide expression or plasma ACTH and B secretion compared to intact, uninjected rats. Lesions of CRF or AVP neurons greatly reduced peptide expression and mRNA levels in the PVN and median eminence at both ages. However, the specificity of the lesion was greater in older than in young pups. At both ages, we observed a dissociation between the morphological effects of the lesions and hormonal responses. In d14-19 pups, CRF and AVP lesions prevented ADX-induced changes in mRNA levels and peptide expression but did not reduce ACTH secretion under basal or stimulated (post ADX) conditions. However, CRF and AVP lesions increased the expression of CRF in the central amygdala and the bed nucleus of the stria terminalis. Lesions with AVP also stimulated CRF expression in the PVN. Thus, these compensatory changes could take over some of the hypophysiotropic actions of the damaged PVN neurons. In young pups (d4-9), we did not observe the typical increase in CRF and AVP mRNA levels and peptide expression found after ADX in older pups or adults. Lesions of the CRF neurons also affected the AVP system and reciprocally. We suggest that this could be explained by a high degree of colocalization of CRF and AVP observed in parvocellular and small, immature magnocellular neurons in young pups. The lesions did not affect basal or ADX-induced ACTH secretion, suggesting that during the early neonatal period, the pituitary is the major site of glucocorticoid inhibitory feedback on ACTH secretion and that the hypothalamus does not exert a tonic control over basal pituitary secretion. These results unravel ontogenetical differences in the regulation of ACTH secretion by hypothalamic CRF and AVP. During the first 10 days of life, within the adrenal stress hyporesponsive period, hypothalamic CRF and AVP neurons are not sensitive to glucocorticoid feedback and basal ACTH secretion appears to be relatively independent from hypothalamic input. After the second week of life, maturation of glucocorticoid receptors, neuronal phenotype and connections of the PVN to other brain structures (bed nucleus of the stria terminalis, central amygdala) allows for the full expression of corticosterone effect on hypothalamic neurons and for compensatory changes to occur following lesions. These results emphasize the extraordinary capacity of the developing central nervous system to adapt to changes in functionning of some neuronal areas critical for homeostatic balance and the important potential role of intra-hypothalamic and extrahypothalamic relationships in maintaining control over ACTH and glucocorticoid production during development.