(6 character code)

Welcome to BIOISIS ... the resource for macromolecular SAXS.
Experiment of the Month
Splicing factor complex of phosphorylated (P)SF1 residues 1-255 and U2AF65 residues 148-375
Low_res_thumbnail Human SF1 ULM-KH-QUA2, which is phosphorylated on Ser80Ser82, bound to U2AF65 RRM1-RRM2-UHM in a splicing factor complex.

Recent Posts


There were some exciting and diverse membrane protein SAS talks this morning. Two different talks focused on modeling lipid interactions around an oligomeric membrane protein: Javier Perez described his useful SEC-SAXS setup at the SWING beamline, which measures RI and UV alongside the SAXS to obtain stoichiometry information about detergent-protein complexes. He took several useful strategies to model the Aquaporin-0-lipid structure, and validate his models. In the same session, Shuo Qian also wanted to understand how detergents wrap around the protein photosystem 1, and is making great progress using SANS methods. Cecile Fradin spoke about her fluorescence imaging, SANS and AFM studies to characterize the pore forming mechanism involving Bax and Bak proteins’ interactions with mitochondrial membrane, leading to apoptosis. She used a series of contrast matching experiments and vesicle of a variety of lipids to visualize protein vs lipid contributions to the pore forming process and test her “mushroom vs. umbrella” model. Wei Liu of the Cherezov group described how SAXS on lipid matrices is aiding their abilities to define and characterize new lipid mesophases for membrane protein crystallization. Their LCPs form in 5 minutes, and a goal is to characterize structural parameters such as size of the water channel for different matrices in high throughput. Andrew Whitten described his SAXS, SANS and cross linking studies to characterize interactions of Munc18-1:Syntaxins, which can form open and closed states relevant to a synaptic vesicle fusion mechanism. Shuo Qian also gave an overview of the Bio-SANS/CSMB user facility at Oak Ridge National Lab. There are 2 SANS stations from the reactor, with large q-range and are are also setting up an onsite SAXS station for testing SANS samples. The facility also has ongoing GI-SANS efforts to characterize membrane structure around proteins. They also have a biodeuteration Lab (user facility) for preparing contrast matching samples and deuterated lipids for extraction from E. coli.


Greg Hura from SIBYLS beamline at ALS/LBNL talked about learning new biology with high-throughput SAXS by getting back to hypothesis testing and the scientific method. How? He described several applications of his new Volatility-of-ratio metric applied to comparison heatmaps, which SAXS users can access at Michal Hammel from SIBYLS beamline presented a talk, “Prediction of Protein Crystallizibility from SAXS” aka “You Never Know.” Kratky plots and Porod exponents did not help him in these efforts. His main summary: the worst looking SAXS profiles may indicate crystallizable constructs. Bragg peaks (even at low concentrations) can indicate nucleation and be informative as to positioning within complexes. SAXS is a tool of comparison, but there is a high level of false positive/false negative. Philip Anfinrud brought up the notion of undoing the orientational averaging in SAXS by polarizing the system via illuminating a chromatically active protein coupled to 100 ps timeresolved SAXS.


Dr. Trewhella also re-introduced the ongoing efforts toward developing SAXS guidelines for evaluation and publication, and encouraged the SAXS community to get involved with the discussion at:


Jill Trewhella spoke about her ongoing work on understanding molecular events in muscle relaxation/contraction via cardiac myosin binding protein c and its assembly with actin. Due to the large size of these filaments, her team had to start from a 2D model to look at the cross sectional information (she developed a “2D bead modeling variant of DAMMIN”), and then optimize rotation/translation of components in the 3D model. In the cell, only one cMyBP-C tends to bind per 30 actin units, but in her case, they are looking at saturated actin, which has provided beneficial symmetry information. EM studies showed agreement with the SAS model interaction of the actin central core to the C0 subunit of within ~17 Å, which for a model of this complexity, seemed great. She also addressed species differences in the proteins, as the “PA” linker length between domains of cMyBP-c apparently correlates with heart rate. By forming these filaments in the NMR tube, she also confirmed the C0 residue-level positioning. So she has now used MX, SAXS, SANS, EM and NMR information to validate her model, since, as she put it, actin is very sticky, so why should such an intricate model be believed?