Bio SAXS

Increasingly, BioSAXS is becoming an indispensable tool in molecular and structural biology. Because the technique is applicable to a very wide range of solution conditions (concentration, pH, ionic strength, temperature, additives, etc.) and because it gives information on systems without crystals, even disordered systems, it has become an important tool for gleaning structural information early in the research process, for diagnosing problems, and for understanding structure and association under physiological conditions.

• radius of gyration (typically 2% accuracy)
• molecular weight (typically 10% accuracy)
• maximum intra-particle distance
• low-resolution particle shape
• degree of folding, denaturation, or disorder

• determining physiological oligomeric state
• validating proposed models of complexes
• building complexes from monomers or known fragments
• studying protein-protein interaction under different solution conditions
• modeling missing loops and domains
• refining homology models
• categorizing discrete folded and unfolded states
• finding volume fractions in mixtures

• aggregation
• molecular complex too large (data not collected at low-enough q)
• solution too dilute (no signal)
• sample is a mixture *
• denaturation
• buffer mismatch
• improper normalization in subtraction
• contrast problems
• radiation damage
• bubbles

How big is your molecule? This is an important question in BioSAXS. The maximum diameter of your molecule will determine how long the sample-to-detector distance needs to be. This can be a limiting factor.

 The Shannon Sampling Theorem suggests that objects having a diameter of D_max should be measured down to at least q = π/D_max, preferably lower. When computing R_g directly from a Guinier plot it is customary to use data with q < 1.3/R_g. Guinier and Fournet, in their 1955 book "Small-Angle Scattering of X-rays", recommend that q_min< 1.3/(2R_g). For some shapes, such as rods, it may be necessary to collect data to even lower q values to get an accurate R_g, but since the shape is not known a priori, the 1.3 convention is widely accepted.

The current beamline configuration at F2 (as of 12/2010) allows for the determination of objects having R_g < 100 Å and Dmax < 330 Å (q_min > 0.009 Å). G1 station can achieve somewhat larger sizes using a longer beampipe. Users with larger systems and 300 Å should consult MacCHESS staff for details. 

Actual time is hard to estimate because there are so many variables. Experienced users can collect faster than novice users.

Currently on F2, you can expect single exposure times of at least 180 sec. You'll need at least three dilutions for each protein in addition to one or more buffer exposures. You also have to factor in one rinse time for the capillary (maybe 2-5 min), and sample preparation and dilution time (you should centrifuge concentrated protein samples for 5-15 minutes before dilution). So, on F2 you should count on needing at least 30 min for each different protein sample including its buffer and dilutions. On G1, which is 20-30 times brighter than F2, you can expect 1-10 second exposures, so the collection time per protein might fall to 15 minutes.

We don't yet know how many samples an experienced user group can collect in 24 hr since the robotic system is so new. Probably 20 samples on F2 is realistic. Possibly 50 or more on G1.  

Protein solutions should be monodisperse and concentrated to at least 10 mg/ml if possible without aggregation. One strategy is to concentrate some sample, but also bring some dilute. Even with pure proteins it may be necessary to experiment to find the best conditions under which the protein is monodisperse and well folded.

A total sample volume of 50 microliters is the minimum advisable volume to prepare. 100 microliters is a comfortable amount. This will allow enough for a series of dilutions, tests to determine the best exposure time, and any possible sample loading problems. With a flow cell, the more protein you bring, the better the signal will be. For modest size proteins, don't expect to get much usable signal below 0.8mg/ml. For much larger proteins (>60 kDa), the concentration limit will improve. Intensity at q=0 increases as R⁶ (R = radius), so signal near the beamstop can get very bright for very large objects. But intensity also falls off as q^-4 (Porod’s law), so the usable q_max for large systems may also decline.

Solutions that are too concentrated exhibit concentration-related distortions of the small-angle part of the scattering curve. You can see this in lysozyme stronger than 10 mg/ml and Bovine Serum Albumin (BSA) stronger than 5 mg/ml.  

It is advisable to collect data at several (at least 3) different concentrations and extrapolate to infinite dilution if necessary. Alternatively, you can combine a dilute curve (small-angle part) with a concentrated curve (wide-angle part). Concentration does not effect the wide angle part of the scattering curve.

Users must also provide a matched buffer solution for background subtraction. This should be the exact same buffer used in the original sample preparation if possible. It is good to change buffer if you don't know the original composition exactly. Prepare plenty of extra buffer for sample dilutions and for rinsing sample cells (bring at least1 ml if you can).

Review articles

(1)Jacques, D. A. and J. Trewhella (2010) "Small-angle scattering for structural biology—Expanding the frontier while avoiding the pitfalls." Protein Science 19(4): 642-657

(2) Putnam, C. D., M. Hammel, et al. (2007). "X-ray solution scattering (SAXS) combined with crystallography and computation: defining accurate macromolecular structures, conformations and assemblies in solution." Quarterly Reviews of Biophysics 40(3): 191-285.

(3) Mertens, H. D. T. and D. I. Svergun (2010). "Structural characterization of proteins and complexes using small-angle X-ray solution scattering." Journal of Structural Biology 172(1): 128-141.

(4) Svergun, D. I. and M. H. J. Koch (2003). "Small-angle scattering studies of biological macromolecules in solution." Reports on Progress in Physics 66(10): 1735-1782.

Classic Texts

Feigin, L. A. and D. I. Svergun (1987). Structure Analysis by Small-Angle X-ray and Neutron Scattering, Plenum Press/Springer.

Glatter, O. and Kratky (1982). Small-angle X-ray Scattering. London, Academic Press.

Guinier, A. and G. Fourne (1955). Small-Angle Scattering of X-Rays. New York, John Wiley and Sons.

Useful online resources

bl1231.als.lbl.gov/saxs_protocols/saxs_data_collection_evaluatio.php: SIBYLS beamline (ALS)

bioisis.net/tutorial: A useful introductory tutorial in BioSAXS can be found at BIOISIS (ALS)

www.saxier.org/forum/index.php: The Small Angle X-ray Scattering Initiative for Europe (Saxier) also hosts a forum where you can ask questions about BioSAXS processing

Software

sourceforge.net/projects/bioxtasraw/: "BioXTAS RAW" the processing software we use at MacCHESS

www.embl-hamburg.de/ExternalInfo/Research/Sax/software.html: Biological Small Angle Scattering Group (EMBL Hamburg)

situs.biomachina.org/tutorial_saxs.html#visual: Biomachina