*2.2.2.1. Silver impregnation techniques*

For the rapid Golgi staining, the hypothalamus, after 1 month's fixation in fresh prepared formalin, was immersed in potassium dichromate for 10 days and in 1% silver nitrate for additional 10 days. Following dehydration in graded alcohol solutions, the specimens were embedded in paraffin and cut, some of them at 100 μ and some at 25 μ, alternatively [54]. Sections of 25 μ were stained also with methylene blue, according to Golgi-Nissl method [57–60]. All the sections were mounted in Entellan (Merck-Millipore, Darmstadt, Germany), between two cover slips and studied in a Zeiss Axiolab Photomicroscope, equipped with digital camera and computer.

We studied extensively the suprachiasmatic (SCN), the supraoptic (SON) and the paraventricular nuclei (PVN) of the hypothalamus [45]. The volume of the nuclei was estimated according to Cavalieri principle [61, 62]. We described the type of dendritic arborization, the morphology of the dendritic branches and spines, and then we estimated the number of dendritic branches, as well as the spinal density, on sections stained according to rapid Golgi, and Golgi-Nissl methods.

#### *2.2.3. Morphometry*

Morphometric studies were performed with an image analyzer (Image J program). The mean surface area of the neurons, as well as the dendritic arborization, was calculated in silver staining [63]. The morphology of the soma and the dendrites was estimated on the basis of the criteria posed by Jacobs et al. [64], concerning the quality of staining of dendrites and the contrast between neurons and neuropile.

The estimation of the in space distribution of the dendritic branches was performed in a centrifugal way according to Uylings et al. [65]. We estimated the diameter of the soma, the length of the dendrites, the number and the type of the dendritic branches, the length of dendritic segments per dendritic order and the spinal density per segment, given that each dendrite which arises from the neuronal body up to the first bifurcation is considered as firstorder dendritic branch.

For the quantitation we applied Image J program, which was properly adjusted for the used microscope (Carl Zeiss Axiolab Photomicroscope). The dendritic arborization was assessed on the basis of the method of concentric cycles introduced by Sholl [66].

The dendritic spines were counted on three sequent segments of the dendritic field. The first segment, 20–30 μm in length, was located on the primary dendrite, the second segment, 20–30 μm in length, on the secondary one and the third segment of 40–50 μm, on the tertiary dendrite.

At the level of electron microscopy we applied the stereological estimation introduced by Nyengaard [67] and West [68–70]. We estimated the number, the length, the surface area, the volume and the spatial distribution for the mitochondria [54, 70] and for the cisternae and the vesicles of the Golgi complex [71].

We estimated also the mean nuclear area, the dendritic profiles of the neurons [72], the spinal density per dendritic segment, the areas of the pre- and postsynaptic terminals [73–75] and the number of synaptic vesicles per presynaptic component [54, 75].

The statistical evaluation of the data was based on the Student t tests. P-values below 0.05 were considered statistically significant, and those bellow 0.01, highly significant.
