Analysis of neuronal networks: a review of techniques for labeling axonal projections

In order to analyze connections between neurons in the vertebrate central nervous system, methods have been developed to label a given population of axons of known origin so that they can be differentiated from other, non-labeled structures. Three such methods are reviewed here: experimentally induced orthograde (Wallerian) degeneration, axon transport of radioactive proteins demonstrated by autoradiography, and axon transport of macromolecules that can be reacted histochemically to yield a visible reaction product. Each of the methods has particular strengths and weaknesses. Degeneration methods may differentiate between different functional classes of axons which have different fiber diameters. However, degeneration distorts the morphology of axon terminals, making them more difficult to interpret, and degenerating terminals may be removed rapidly by phagocytosis. Autoradiography of radioactive terminals preserves normal fine structure, but the necessary exposure times extend the method by weeks or months, and care must be exercised to distinguish labeled axons from other structures exhibiting background or transneuronal radioactivity. Histochemical methods, such as those used to demonstrate horseradish peroxidase conjugated to wheat germ lectin (WGA-HRP), are sensitive and rapid, but the injection site must be carefully characterized, and the presence of transneuronal label may make interpretation of the results difficult. Experimental methods of axonal labeling have been invaluable in studying neuronal networks. Each of the methods described here may be of particular value, given the nature of the system to be analyzed.