MACCHESS PROGRAM FOR SOLUTION SAXS AND ENVELOPE PHASING

这个子项目是许多研究子项目中利用 资源由NIH/NCRR资助的中心拨款提供。子项目和 调查员(PI)可能从NIH的另一个来源获得了主要资金， 并因此可以在其他清晰的条目中表示。列出的机构是 该中心不一定是调查人员的机构。 如前面几节所述，这一倡议的合作者提出了具有挑战性的结构性问题，这些问题需要解决方案SAXS和伴随而来的阶段性方法的进步。一个很好的例子是约瑟夫·韦德金德(U·罗切斯特)对人类白细胞中APOBEC3G(A3G)酶功能的结构基础的研究，在该研究中，APOBEC3G(A3G)酶起到了阻止艾滋病毒感染的作用。该团队在结晶A3G方面取得了进展，但需要补充结构信息，最终目标是了解这种重要的蛋白质如何对针对其降解的调控因素做出反应。韦德金德与理查德·吉利兰合作，利用SAXS获得了溶液中分子的大致形状。如果晶体可用，溶液形状信息将被用来确定晶体结构的相态。所提出的阶段化过程需要三个步骤：(1)获得溶液SAXS图谱并确定分子包络。(2)如果存在晶体，则收集标准结晶学数据集，并使用低分辨率相的包络信息。(3)将相位扩展到更高的分辨率，以求解高分辨率结构。具体操作步骤如下。获得SAXS模式并确定分子包膜我们建议优化使用SAXS确定分子包络的实验技术。目前在国际象棋G1光束线上进行SAXS实验的设置如下(Gillilan，2002，未发表；图31)：+G1多层光学摇摆线+1.2埃波长，带~2%带通+量子4 2K(ADSC)探测器@~780 mm样品池+500？m×500？m光束，0.8 mm保护缝+He飞行管，样品端带有Be窗口，探测器端0.5mylar+1 mm路径(80？L)样品池(由英国达累斯伯里J.G.Grossmann提供)+25？m云母窗在样品池散射数据采集和分析中，样品被两个薄的平行窗口包裹在一个小室中。采集了不同浓度的缓冲液和样品的SAXS数据。SAXS数据使用洪新国和Richard Gillilan[108]开发的软件进行处理。数据缩减包括将散射数据归一化为发射光束的强度以及减去缓冲器的背景散射。然后将所有的散射曲线标准化为蛋白质浓度为1 mg/ml的散射曲线。将低角度数据外推到无限稀释，并与在高蛋白质浓度下测量的高角度数据合并，以得到最终的散射曲线。一旦获得蛋白质样品的溶液散射数据，下一步就是从一维散射图案恢复三维包络。将使用Svergan和他的同事开发的两种方法。在第一种广义从头算方法[74,109]中，粒子的角包络函数R=F()，其中(r，)是球坐标，用一系列球谐函数来描述。因此，低分辨率形状由几个参数定义，这些参数是该系列中符合散射数据的系数。这种方法在计算机程序Sasha[75]中实现。研究表明，在某些情况下，可以从散射数据中提取唯一的包络(除了SAXS中所有从头算方法所具有的形状的惯用性质外)。左手和右手都应该进行测试，并在结构确定的后期阶段解决模棱两可的问题，届时可以确定螺旋的手。由球谐函数定义的包络的使用仅限于形状相对简单且没有明显内部空穴的球形粒子。可以使用不同类型的蒙特卡罗搜索和利用模拟退火法从一开始就构建更详细的模型，在该方法中，通过大量紧密堆积的珠子来模拟复合体的形状，这些珠子被移动以尽可能地匹配观察到的散射分布[110-112]。基于蒙特卡罗的模型包含数百或数千个参数，需要谨慎，以避免过度解释。一种常见的方法是对齐由独立形状重建运行产生的一组模型，以获得保留最持久且可能也是最可靠的特征的平均模型，例如使用程序SUPCOMB[113]。粒子对称性，如果已知，提供了非常有用的约束，这可以在程序Sasha和Dammin以及程序GASBOR[110-112]中施加。一旦从SAXS图样确定了分子形状并收集了结晶学数据，就使用FSearch程序[93]中实现的分子置换方法来定位晶体单胞中的包络，从而提供低分辨率相。FSearch(由CCP4分发)可以接受两种包络形式(球谐或基于蒙特卡罗的模型)中的任何一种作为输入搜索模型。研究表明，探测器饱和引起的低分辨率强反射的缺失会大大降低FSearch解的质量[114]。因此，在晶体数据收集实验中要特别注意，以确保低分辨率数据(100-10？)接近完成(通过使用小光束挡板)并且不饱和(通过减少曝光时间)。

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. As described in earlier sections, the collaborators on this initiative have proposed challenging structural problems that demand solution SAXS and concomitant advances in phasing methods. A good example is Joseph Wedekind's (U. Rochester) investigation of the structural basis for function of the APOBEC3G (A3G) enzyme in human white blood cells, where it acts to stave off HIV infections. The team is making progress on crystallizing A3G, but needs complementary structural information with the ultimate goal of learning how this important protein responds to regulatory factors that target it for degradation. Wedekind has teamed up with Richard Gillilan to obtain the general shape of a molecule in solution utilizing SAXS. If and when crystals become available, the solution shape information will be used to phase the crystallographic structures. The proposed phasing procedure requires three steps: (1) Obtain a solution SAXS pattern and determine the molecular envelope. (2) If crystals exist, collect a standard crystallographic data set and use the envelope information for low-resolution phases. (3) Extend phases to higher resolution to solve for the high resolution structure. The procedures are detailed below. Obtain a SAXS Pattern and Determine the Molecular Envelope We propose to optimize experimental techniques to determine molecular envelopes using SAXS. The current setup for SAXS experiments at the CHESS G1 beamline is as follows (Gillilan, 2002, unpublished; Fig. 31): + G1 wiggler line with multilayer optics + 1.2 Angstrom wavelength with ~2% bandpass + Quantum4 2K (ADSC) detector @ ~780mm from sample cell + 500¿m x 500¿m beam with 0.8mm guard slits + He flight tube with Be window on sample end, 0.5mil mylar on detector end + 1mm path (80¿l) sample cell (courtesy of J.G. Grossmann, Daresbury U.K.) + 25¿m mica windows in sample cell Scattering Data Acquisition and Analysis Samples are encapsulated inside a cell sandwiched by two thin parallel windows. SAXS data of buffer and samples at different concentrations are collected. SAXS data are processed using the software developed by Xinguo Hong and Richard Gillilan [108]. The data reduction includes normalization of the scattered data to the intensity of the transmitted beam and subtraction of the background scattering of the buffer. All scattering curves are then standardized to that of a protein concentration of 1 mg/ml. The low angle data will be extrapolated to infinite dilution and merged with the high angle data measured at high protein concentrations to yield final scattering curves. Once the solution scattering data from a protein sample is obtained, the next step is to recover the three-dimensional envelope from the one-dimensional scattering pattern. Two methods developed by Svergun and colleagues will be used. In the first general ab initio approach [74,109], an angular envelope function of the particle, R = F( ), where (r, ) are spherical coordinates, is described by a series of spherical harmonics. The lowresolution shape is thus defined by a few parameters  the coefficients of this series  that fit the scattering data. This approach was implemented in the computer program SASHA [75]. It was demonstrated that, under certain circumstances, a unique envelope can be extracted from the scattering data (except for the handedness of the shape  this ambiguity holds for all ab initio methods in SAXS). Both `left` and `right` hands should be tested and the ambiguity resolved in the later stages of the structure determination, when the hand of helices can be determined. The use of envelopes defined by spherical harmonics is limited to globular particles with relatively simple shapes and without significant internal cavities. More detailed models can be constructed ab initio using different types of Monte Carlo searches and the utilization of a simulated annealing approach in which the shape of the complex is modeled by a large number of close-packed beads, which are moved around so as to match the observed scattering profile as well as possible [110-112]. The Monte-Carlo-based models contain hundreds or thousands of parameters, and caution is required to avoid over-interpretation. A common approach is to align a set of models resulting from independent shape reconstruction runs to obtain an average model that retains the most persistent, and presumably also most reliable, features, e.g. using the program SUPCOMB [113]. Particle symmetry, if known, provides very useful constraints, which can be imposed in the programs SASHA and DAMMIN, and in the program GASBOR [110-112]. Once the molecular shape is determined from the SAXS pattern and crystallographic data are collected, the molecular replacement method implemented in the FSEARCH program [93] is used to locate the envelope in the crystallographic unit cell, thus providing low resolution phases. FSEARCH (distributed by CCP4) can accept either of the two forms of envelope (spherical harmonics or Monte-Carlo-based models) as an input search model. It has been shown that the absence of strong reflections at low resolution caused by saturation at the detector can degrade FSEARCH solutions greatly [114]. Therefore, particular attention is paid in crystallographic data collection experiments to ensure low resolution data (100-10¿) are near complete (by using a small beam stop) and not saturated (by reducing exposure time).