Abstract

Combining neurohistology and mRNA expression maps as a guide for anatomical alignments of the mammilary region of the human hypothalamus

A. MEDINA; D. M. KROLEWSKI; I. A. KERMAN; R. BERNARD; E. G. JONES; W. E. BUNNEY; H. AKIL; S. J. WATSON
Society for Neuroscience. 2007.

Hypothalamic nuclei in humans are not as sharply defined as they are in other species like rodents, and the need for a method to guide the process of obtaining specific regional samples has grown with the availability of precision microdissection techniques such as Laser Capture Microdissection (LCM) and the possibility of performing gene expression profiling of discrete areas. Here we propose the use of a combination of classic histological techniques with in situ hybridization (ISH) to facilitate a detailed process of anatomical alignment in order to acquire representative samples from homologous regions across subjects. Materials and Methods: Human brain samples were collected by the Brain Donor Program at the University of California, Irvine. The brains were extracted during autopsy and sliced in the coronal plane in to 0.75 cm slabs . The slabs were then frozen and stored at -80°C. Prior to cryostat sectioning, smaller blocks containing the hypothalamus were sawed away from the proper slabs. 10 µm sections were cut and in situ hybridization (ISH) was performed in sections 500 µm apart. Four transcripts were selected for this study: melanin-concentrating hormone (MCH),orexin, vasoactive intestinal peptide (VIP) and histidine decarboxylase (HDC). After ISH, slides were processed for modified Kluver's stain. Digitalized images of both ISH and Kluver's stain slides were then overlayed using Adobe® Photoshop® CS 8.0. ISH images were also processed to create a 3D reconstruction of the mRNA signals distribution using Volocity 2.6 software (Improvision Inc, Lexington, MA). Results: The obtained images contain the mammillary region from the rostral level of the tuberomammillary nucleus to the caudal mammillary body at the mammillo- tegmental tract level. The images were organized in plates with anatomical landmarks pointed on each plate. MCH and orexin signals were observed lateral and dorsomedial to both, the fornix and the mammillary body, with orexin predominating at the level of the dorsomedial hypothalamus and MCH prevailing at the posterior hypothalamic area. HDC signal was observed ventromedial and ventrolateral to the fornix at the level of the tuberomammillary nucleus and medial and lateral to the mammillary body at the caudal region. When the observations were extended to the paraventricular region, it was noticeable how all three signals were visible as far as the level of the paraventricular and supraoptic nuclei. VIP was observed above the mammillary body in the caudal edge of the lateral hypothalamic area.