Isolation and Characterization of Synaptic and Nonsynaptic Mitochondria from Mammalian Brain

Until the mid-1960s, studies on the metabolism of brain mitochondria were hampered by the lack of suitable methodologies for brain mitochondrial isolation. In the majority of the studies up to this time, crude mitochondrial preparations were isolated from mammalian brain homogenates using differential centrifugation techniques that were adapted from those that had originally been designed for isolating liver mitochondria. During the early 196Os, two groups of workers [see Whittaker (1969 and 1984) and De Robertis and de Lores Arnaiz (1969) for discussions] independently designed relatively elaborate subcellular fractionation procedures, whereby the crude mitochondrial fraction derived from brain homogenates was subfractionated, using sucrose density gradients, into three or more discrete fractions, including pinched-off nerve-ending particles or “synaptosomes” and myelinated axon fragments (usually referred to as “myelin”), in addition to “free” (i.e., nonsynaptic) mitochondria. Thus, the structural heterogeneity of the crude mitochondrial fraction derived from brain homogenates was revealed. Consequently, many of the metabolic properties (e.g., glycolysis) that were erroneously attributed to isolated brain mitochondria in the studies in that era could be accounted for by the presence of vesicular and other membranous particles (e.g., synaptosomes, myelin, and micro-somes) that contain cytosolic material to a greater or lesser extent [see Clark and Nicklas (1970) for a discussion].