Calcium-mediated Evidence from the Changes in Gap Junction Low Angle X-ray Pattern Structure:

Rat liver gap junctions were isolated in Ca2+-free media and analyzed in controlled environments by x-ray diffraction of partially oriented pellets. Different treatments of the same preparations were compared. The ordered hexagonal lattices gave rise to detail that was sensitive to low Ca 2+ concentrations (0.05 mM), but not to Mg 2+ (up to 0.16 mM) or pH (between 6.0 and 8.0). The major Ca2+-mediated responses were reductions in the intensity of the (1, 0) peak and in the off-equatorial contributions to the (2, 1) peak, and changes of scale equivalent to a decrease (,-,2%) in lattice dimension, but an increase (---4%) in the dimension perpendicular to the lattice. A simple structural interpretation of these findings is that Ca 2+ induces the subunits of the channel-forming assembly, the connexon, to align more nearly parallel to the channel, thereby causing the connexon to become slightly longer and more radially compact. The rearrangement is of the same nature as one found under less physiological circumstances by electron microscopy (Unwin, P. N. T., and G. Zampighi, 1980, Nature (Lond.)., 283:545-549), and may be part of a coordinated mechanism by which the channel closes. The gap junction is a region of contact between neighboring animal cells that allows the exchange of small molecules and ions between their interiors (4, 6). The channels responsible for this exchange lie along the hexad axes of hexameric membrane proteins called connexons (2), which link, in pairs, the apposed plasma membranes. In living tissue the connexon responds to local changes in intracellular free Ca 2+ concentration (10, 11), and/or pH (11, 13) and membrane potential (3, 9), thereby regulating the channel's permeability. The mechanism of the transition between the open and closed states of the channel is not known. However, studies of isolated, hexagonally ordered, gap junctions suggest it may involve a change in conformation of the connexon in the space between the two plasma membranes (1, 7) or, alternatively, a coordinated rearrangement of the connexon subunits around the channel (14). As a first step toward understanding the physiological significance of these connexon configurations, we examined the structural response of isolated junctions to changes in ionic environment. Response was assessed quantitatively by dividing each preparation in two, exposing the separate portions to different solutions, and then comparing x-ray diffraction patterns recorded from each portion. Pair-wise analysis provided a sensitive measure of structural state, making the effect of small variabilities between preparations relatively unimportant. Results showed that the isolated junctions change their structure on exposure to low Ca 2÷ concentrations. We report on the properties of this transition, its relation to one previously described (14), and its possible relevance to living tissue. MATERIALS AND METHODS Chemicals: Lubrol WX and Arsenazo IlI were obtained from Sigma Chemical Co. (St. Louis, MO); sodium deoxycholate from Calbiochem-Behring Corp. (La Jolla, CA); and HLlabeled sodium deoxycholate (4 Ci/mmol) from New England Nuclear (Boston, MA). Sucrose gradients were made using ultrapure sucrose from Schwartz/Mann (Orangeburg, NY). All other chemicals were reagent grade. All solutions were made up with doubled distilled water. Isolation: The isolation procedure was similar to that described in reference 15. We used 20-30 rats (Sprague-Dawley, Madison, WI), of ~ 160 g body wt, for each set of experiments. Following excision, their livers (total wet weight: 150-180 g) were pooled and washed in saline. They were transferred to Medium A (0.5 mM EGTA, 1 mM MgCI2, 50 eM phenylmethylsulphonyl fluoride, 4% [wt/wt] sucrose, 10 mM sodium bicarbonate, pH 8.0), cut into small pieces, and homogenized in 20-ml portions in a Potter-Elvehjem glass homogenizer with a loose-fitting teflon pestle. The homogenate was pooled, diluted with Medium A to 4 1, and left to stand for 10 min. It was then filtered two times through four layers of surgieal gauze and subsequently through two layers of nylon mesh having 50-em pores. The filtered homogenate was centrifuged at 4,000 rpm for 20 rain (JA-10 rotor, Beckman Instruments, Inc., Pain Alto, CA) and the supernatant was discarded. The top, slightly pink layers of the pellets were collected, pooled, and homogenized using a tight-fitting pestle. The resulting 80-ml suspension was brought to a sucrose concentration of 43% (wt/wt) with 60% (wt/wt) sucrose in Medium A, and adjusted to pH THE JOURNAL OF CELL BIOLOGY • VOLUME 97 NOVEMBER 1983 1459-1466 © The Rockefeller University Press • 0021-9525183111/1459/08 $1.0