The descriptive term gap junction originally arose from the electron micrograph studies of Revel and Karnovsky in 1967 (). This morphological name referred to a cell-cell contact area where two cell membranes were bridged by a specialized membrane protein complex. Gap junctions are found in almost all tissues where cells abut each other. The “gap” is a distinct morphological feature whereby heavy metal stains are able to intercalate into a space between the two joined cell membranes, as opposed to the situation in tight junctions, where the membranes fuse and the stain cannot penetrate in between the cells. Subsequent electron microscopy revealed that gap junctions consist of tens to thousands of membrane channels (also called intercellular channels) containing proteins generally called connexins. Each membrane channel was shown to contain two connexons, the hexamer of connexins, which dock at their extracellular surfaces. In essence, the gap junction membrane channel is a dimer of two hexamers joined together in the gap region. In this way, the membrane channel extends across both cell membranes. Both freeze-fracture and thin section electron micrographs showed that these gap junctional plaques contained quasicrystalline areas of the membrane channels. During the next decade, intensive efforts by many researchers were directed toward purifying the gap junction plaques first seen in these micrographs for both biochemical and structural analysis. Many man-years were spent developing and refining isolation protocols that reliably gave pure preparations of gap junctions, which were used for structure determination by electron crystallography () and X-ray diffraction (). The specimens that are considered most desirable for electron crystallography are thin, uniform two-dimensional crystals. Early on, gap junctions were looked on as a tantalizing specimen, not only because of their biological importance, but also because they form quasicrystals in the cell and with limited detergent treatments could be induced into well-ordered twodimensional crystals (). It is clear from the structural work that has been done to date that the critical step in the structure determination is the isolation of crystalline material in sufficient and pure quantities (a problem that structural biologists also refer to as “no crystal, no grant”). In liver, the gap junction arrays are estimated to cover only ~0.1% of the surface area of a hepatocyte, therefore, one generally begins with large amounts of starting material to end with a small quantity of pure gap junctions (, , ). Liver, lens and heart tissue (, , ) have classically been the organs of choice for isolating gap junctions because the majority of the tissue contains only one or two types of connexins and starting material can be obtained in large quantities. The situation has improved with the advent of molecular biology techniques for expression of the connexins in tissue culture systems such as the baby hamster kidney (BHK) cell line () and the insect cell line derived from Spodoptera (,). In these cell lines, the gap junctions can be overexpressed, isolated and purified with larger yields than from native tissues. Recent advances in the three-dimensional structure determination at ~7Å have occurred because of the development of overexpression systems and generation of recombinant connexin material (see ref. and Chapter 4 in this volume) as well as improved electron crystallographic structure determination methods. These EM crystallographic methods have been greatly enhanced by both improvements in electron microscopes as well as in computer reconstruction techniques (see Chapter 4 and also ref. ).