Crystallographic study of calcium ion pump of sarcoplasmic reticulum

对肌浆网（SR）钙离子泵（Ca^<2+>-ATPase）进行了晶体学研究，以阐明主动转运的结构基础。在我们开始这个项目时，我们知道两种类型的CA^<2+>-ATPase晶体：一种是在EGTA和钒酸盐存在下形成的天然SR膜的管状晶体;另一种是在高浓度钙存在下形成的由重构膜片组成的小的三维晶体。两者都很小，适合电子晶体学。使用管状晶体，在8 A分辨率下获得了三种状态的三维结构：无ATP，含铬-ATP和含毒胡萝卜素（最后一种是与纽约大学斯托克斯实验室合作）。具有结合毒胡萝卜素的一个被最好地解析，并且清楚地显示出从二级结构预测（发表在Nature上）预期的10个跨膜螺旋。具有铬ATP的一个指示ATP结合位点的位置（发表在Biophys. J.）。为此，我们制定了一系列广泛的计划。沿着小3D晶体的b轴的投影沿着与来自管状晶体的3D结构的比较揭示了归因于钙结合的大规模构象变化（Biophys. J.）。通过改进结晶条件，我们成功地制备了这种离子泵的有序板状晶体。最薄的可以用电子分析，最厚的可以用X射线晶体学分析。我们已经开发了一种旋转照相机来收集电子衍射强度，并了解到即使是最薄的晶体也具有与最厚的晶体几乎相同的晶格常数。因此，我们直接从电子显微照片中收集相位信息，并获得了一组非常有用的某些反射相位。我们现在正在寻找良好的同晶重金属衍生物，以解决在原子分辨率的结构。

Crystallographic study of the calcium ion pump (Ca^<2+>-ATPase) of sarcoplasmic reticulum(SR)has been performed to elucidate the structural basis of active transport. Two types of crystal of CA^<2+>-ATPase were known when we started this project : one was tubular crystal of native SR membrane formed in the presence of EGTA and vanadate ; the other was small three-dimensional crystals consisting of reconstituted membrane sheets formed in the presence of high concentration of calcium. Both were small and suitable for electron crystallography. Using tubular crystals, three-dimensional structures were obtained for three states at 8 A resolution : no ATP, with chromium-ATP and with thapsigargin (the last one was in collaboration with Stokes Lab at NY University). The one with bound thapsigargin was best resolved and clearly showed 10 transmembrane helices expected from secondary structure predictions (published in Nature). The one with chromium ATP indicated the position of ATP-binding site(published in Biophys. J.). We have developed an extensive set of programs for this purpose. Comparison of the projection along the b-axis of the small 3D crystals and 3D structure from tubular crystals revealed large scale conformational changes attributable to the binding of calcium (Biophys. J.).By improving the crystallisation conditions, we succeeded in preparing well ordered plate-like crystals of this ion pump. The thinnest ones could be analysed by electron and the thickest ones by X-ray crystallography. We have developed a rotation camera for collecting electron diffraction intensities and learned that even the thinnest crystals have virtually the same lattice constants with the thickest ones. We therefore proceeded to collect phase information directly from the electron micrographs and obtained a very useful set of phases for certain reflections. We are now looking for good isomorphous heavy metal derivatives for solving the structure at an atomic resolution.