Assaying structural changes after nerve damage, an essay on quantitative morphology

Application of the physical disector to the central nervous system: estimation of the total number of neurons in subdivisions of the rat hippocampus

Stereology is a group of mathematical and statistical methods that allows the extrapolation of three-dimensional structural information from two-dimensional sections (or slices). This allows researchers to derive important quantitative structural information, such as the volume, surface area or numbers of particular particles (e.g. cells) within defined regional boundaries. The need for such quantitative information in biology is of particular importance when evaluating the influence of various experimental treatments on specific organs, tissues and cells in the body. Knowledge of such changes has given important insights into the neural substrates that may be responsible for the functional and behavioral consequences of a disparate range of experimental treatments. Here, we describe some of these methods as applied to quantifying the total numbers of cells in defined regions of the hippocampal formation. The methods used for this evaluation were, first, the Cavalieri principle, which was used to determine the volumes of the various subdivisions of the rat hippocampus, and, second, the 'physical disector' method, which was used to estimate the numerical density of neurons within each subdivision. Once these values were derived, it was but a simple task to multiply them together to obtain estimates for the total numbers of cells in the given hippocampal region. We found that 16-and 30-day-old normal male rats had 176 800 and 152 700 pyramidal cells in the CA1 region, respectively. This decrease in the neuronal number was statistically significant. However, in the CA2 + CA3 region, there were approximately 169 300 and 149 600 pyramidal cells in 16- and 30-day-old normal male rats, respectively, which was not significantly different. In the dentate gyrus, there were approximately 36 700 neurons in the hilus region and 483 000 granule cells in the granule cell layer, irrespective of the age of the rats. There were no significant differences between these estimates of hilus neurons and granule cells.

Design-based stereology in neuroscience

Quantitative morphology of the CNS has recently undergone major developments. In particular, several new approaches, known as design-based stereologic methods, have become available and have been successfully applied to neuromorphological research. However, much confusion and uncertainty remains about the meaning, implications, and advantages of these design-based stereologic methods. The objective of this review is to provide some clarification. It does not comprise a full description of all stereologic methods available. Rather, it is written by users for users, provides the reader with a guided tour through the relevant literature. It has been the experience of the authors that most neuroscientists potentially interested in design-based stereology need to analyze volumes of brain regions, numbers of cells (neurons, glial cells) within these brain regions, mean volumes (nuclear, perikaryal) of these cells, length densities of linear biological structures such as vessels and nerve fibers within brain regions, and the cytoarchitecture of brain regions (i.e. the spatial distribution of cells within a region of interest). Therefore, a comprehensive introduction to design-based stereologic methods for estimating these parameters is provided. It is demonstrated that results obtained with design-based stereology are representative for the entire brain region of interest, and are independent of the size, shape, spatial orientation, and spatial distribution of the cells to be investigated. Also, it is shown that bias (i.e. systematic error) in results obtained with design-based stereology can be limited to a minimum, and that it is possible to assess the variability of these results. These characteristics establish the advantages of design-based stereologic methods in quantitative neuromorphology.

Rapid estimates of neuron number in the confocal microscope combined with in situ hybridisation and immunocytochemistry

A rapid counting protocol is described which combines the optical disector and Cavalieri methods on non-embedded slices of fixed tissue viewed in the confocal laser scanning microscope. By eliminating the embedding stage and avoiding the need to align adjacent sections in the z plane for counting, considerable time savings are gained over physical disector methods. It also allows the remaining non-embedded sections to be used for other purposes, such as in situ hybridisation and immunocytochemistry. Starting with fixed brainstem, it was possible in less than 2 h to determine the total number of motoneurons in both facial nuclei of an adult Sprague-Dawley rat. This method revealed that the normal facial nucleus contained approximately 3200 motoneurons (n=12 rats). One month following facial nerve avulsion (n=4 rats), mean numbers of motoneurons were reduced by 75%. Using intervening sections, changes in neuronal number were compared with changes in in situ hybridisation signal and immunostaining.