Distributions of pro‐vasopressin expressing and pro‐vasopressin deficient CRH neurons in the paraventricular hypothalamic nucleus of colchicine‐treated normal and adrenalectomized rats

Abstract

The corticotropin‐releasing hormone (CRH) neurosecretory system in normal rats consists of two major subpopulations of parvicellular neurons in the hypothalamic paraventricular nucleus distinguished by the presence or absence of coexistent vasopressin precursor (pro‐AVP)‐derived peptides. These neurons project to the external zone of the median eminence, where the two subtypes of axons (CRH+/AVP+ and CRH+/AVP‐) were previously found to be approximately equal in number. The present study was undertaken (1) to determine whether the relative numbers of pro‐AVP expressing and pro‐AVP deficient perikarya in the paraventricular nucleus corresponded to what we previously found for the axons in the median eminence, (2) to map the two cell types throughout the entire paraventricular nucleus to determine whether significant differences existed in their distributions, and (3) to ascertain whether or not the pro‐AVP deficient subpopulation expressed pro‐AVP after adrenalectomy. Postembedding electron microscopic immunocytochemistry on serial ultrathin sections was used to identify the peptide phenotypes of perikarya in the paraventricular nucleus in normal rats and 7 days after adrenalectomy with and without colchicine treatment. The peptide phenotypes of neuronal perikarya in the paraventricular nucleus were identified by using antibodies to CRH, AVP, neurophysin (NP), the C‐terminal glycopeptide of pro‐AVP (GP), and oxytocin‐associated neurophysin (NP_OT). Groups of three serial coronal ultrathin sections were analyzed at 200‐μm intervals throughout the entire rostrocaudal extent of the paraventricular nucleus. The sections in each group were stained for CRH, a pro‐AVP‐derived peptide (AVP, NP or GP), and NP_OT, respectively. Parvicellular CRH neurons were defined as CRH‐positive cells, approximately 10 μm in diameter, that did not contain detectable NP_OT. Pro‐AVP expressing cells were defined as staining positively for AVP, GP, or NP and negatively for NP_OT. Approximately equal numbers of pro‐AVP expressing (“NP_AVP+”) and pro‐AVP deficient (“NP_AVP‐”) parvicellular CRH neurons were found within the paraventricular nucleus of colchicine‐treated normal rats, and the two subtypes were distributed differently within the paraventricular nucleus. Although the pro‐AVP expressing CRH cells stained intensely for NP and GP, staining for AVP was quite variable and difficult to quantify in colchicine‐treated normal animals. The CRH+/NP_AVP+ neurons were highly concentrated in the dorsal third of the medial parvicellular subdivision of the paraventricular nucleus, in a region roughly midway between the rostral and caudal ends of the paraventricular nucleus, while the CRH+/NP_AVP‐ neurons were more spread out in the medial parvicellular subdivision in both the dorsoventral and the rostrocaudal axes. Adrenalectomy had no effect on the total number of CRH cells or their distribution in colchicine‐treated rats. However, more than 90% of the parvicellular, NP_OT negative CRH neurons in the paraventricular nucleus stained intensely for AVP, NP, or GP after adrenalectomy and colchicine treatment. The different distributions of the two subpopulations within the paraventricular nucleus in colchicine‐treated normal rats suggest possible differences in the nature of their afferent inputs, and the changes induced by adrenalectomy indicate that normal circulating levels of glucocorticoids are sufficient to inhibit expression of pro‐AVP in one subpopulation but not the other. However, both subpopulations are regulated by glucocorticoids, as evidenced by the increased staining for AVP in both subpopulations after adrenalectomy. The present results are consistent with our hypothesis of functionally distinct subpopulations of CRH neurons that could modulate the ratio of AVP to CRH in response to specific physiological conditions.