SAXS and SANS Facilities and Experiments
Jill Trewhella, The University of Sydney
EMBO Global Exchange Lecture Course
April 30, 2011
The small-angle scattering experiment
FT
X-ray Sources
 CuK


(1.54 Å) emission from
Sealed X-ray tubes, practically need a line
source geometry for protein work
Rotating Anode source, allows for point source
geometry and hence simpler analysis
 Synchrotrons



Tunable (anomalous scattering)
High Brilliance
Excellent for rapid scanning of conditions, very
low protein concentrations, time resolved
experiments, etc
Lab-based
X-ray sources
Point versus line collimation
Slit ‘smearing’
Anton Paar SAXSess (line source)
at the University of Sydney
X-ray scattering samples

5-60 minute measurement times using lab based
sources
 Protein concentrations 1-10 mg/ml
 Sample volumes in the order of 20-30 µL
 Experiments using high intensity synchrotron
instrumentation take a few seconds or minutes and
particle concentrations can be more dilute (by at least an
order of magnitude), but radiation damage can be
limiting; free radical scavengers are helpful (DTT, TCEP,
ascorbate)
 Need a perfectly matched solvent blank; preferably a last
step dialysate or column filtrate
 Sample cells are made of ultra-thin quartz or mica, must
be able to measure sample and solvent background in
same cell, identically positioned in the beam
Neutron Sources
 SANS
requires ‘cold’ source; thermal neutrons
are passed through a liquid hydrogen
moderator to slow them (generally to ~4-6 Å)


Reactors are steady state sources of cold neutrons
that are collimated to provide a narrow wavelength
band (/ ~10%)
Spallation sources produce pulses of cold neutrons
and instruments are designed to use time of flight
so that all wavelengths in a given pulse can be
used which compensates at least partially for
relatively the low time-averaged neutron fluxes
Recall: the basic scattering equation
For
an ensemble of identical, randomly oriented
particles, the intensity of coherently, elastically
scattered radiation is dependant only upon the
magnitude of q, and can be expressed as:
I (q)  N V  P(q)S (q)
2
N = molecules/unit volume
V = molecular volume
contrast, the scattering density difference
  (r)  s = between
the scattering particle and solvent
P(q) = form factor  particle shape
S(q) = structure factor  inter-particle correlation distances



Inter-particle distance correlations
between charged molecules
D
-
-
D
D
D
D
-
D
-
D
….. gives a non-unity S(q) term
D
-
Sample requirements for small-angle scattering
determination of particle shape
 Highly
purified samples containing monodisperse, identical particles without
significant inter-particle distance
correlations (S(q) = 1)


Use a final gel filtration step in the purification
immediately prior to measurement to
eliminate any aggregates
Us DLS to evaluate samples for potential
aggregates (mass fraction
aggregates<0.01%)
Essential preliminary
small-angle scattering experiments

Explore the concentration dependence of the
small-angle X-ray scattering to determine if S(q)
 1.
 If S(q)  1, adjust the solution conditions by
changing pH, salt concentration, or decreasing
particle concentration to eliminate
 Determine the particle mass, molecular volume,
and overall shapes of the components and their
complex (Guinier and P(r) analyses, shape
restoration)
Recall: Guinier Analysis
showed that a plot (lnI(q) vs q2)
gives a straight line of slope Rg2/3 and I(0)
intercept that can be interpreted in terms
of the concentration, contrast and volume
of the scattering particle.
2
 Rg 
2

ln I (q)  ln I (0)  q
 3 


 Guinier
I (0)  N (V )
2
Recall: I(q) and P(r) related by Fourier Transform
Fourier transform must be done using indirect methods
due to finite q-range measured; quality samples and data
give well behaved transforms with certain characteristics
Sample
Protein
conc.a
(mg/ml)
Rg (Å)
NL1-638-(A&B)
1.8
42.5 ± 0.4
NL1/NX complex
3.6
46.8 ± 0.2
Porod
Volumeb
(103 x Å3)
Calculated
Volumec
(103 x Å3)
MWd
(kDa)
130
209 ± 20
198
130/151/144
155
*
275
181/201/199
Dmax
(Å)
SSRL data
University of Utah SAXSess instrument data
13.2
38.3 ± 0.3
7.6
40.1 ± 0.5
4.1
41.4 ± 0.6
Inf. dilution
42.4 ± 0.6
130
208 ± 14
198
130/151/144
NL1-638
3.3
42.7 ± 0.7
130
250 ± 19
220
136/166/160
NL1-691
3.8
51.8 ± 1.0
165
255 ± 26
257
148/189/185
NL2-615
3.7
40.6 ± 0.6
130
178 ± 7
193
135/146/140
NL3-639
1.2
40.3 ± 0.7
130
164 ± 12
190
128/144/138
NL4-619
3.4
42.1 ± 0.6
135
199 ± 7
200
132/140/145
NL1/NX complex
19.7
40.9 ± 0.3
15.7
40.8 ± 0.2
9.8
44.0 ± 0.3
8.7
43.7 ± 0.4
6.6
44.5 ± 0.4
4.5
45.2 ± 0.5
3.8
47.7 ± 0.9
155
*
Inf. dilution
47.7 ± 0.8
155
*
275
181/201/199
NL1-638-(A&B)
Sample
Rg (Å)
Vol (Å3)
Vol (Å3)
Experimental Calculated
55
50
45
Complex
47.1 ± 0.7
228,139 ± 9,965
265,851
NL1-638-Δ(A&B)
42.19 ± 0.7
184,172 ± 7,778
199,261
P(r) arbitrary units
40
Complex
NL1-638-Δ(A&B)
35
30
25
20
15
10
5
0
0
10
20
30
40
50
60
70
80
90 100 110 120 130 140 150
Distance (Angstroms)
Determining the size of your
scattering particle
 Place
data on an absolute scale (water
scattering) and use:
Orthaber et al. (2000) J. Appl. Cryst. 33, 218
 Use
a known mono-disperse protein
scatterer (such as lysozyme) and use:
Krigbaum and Kugler (1970) Biochemistry 9, 1216

Fischer et al. (2010) The molecular weight of proteins in
solution can be determined from a single SAXS
measurement on a relative scale. J. Appl. Cryst. 43,101
If you scale your data so that I(0) = 1, then:
2
V
Q
2
where
Q

2
I(q)q
dq

0
In practice we can only calculate:
q min
Q'   I(q)q2q
q max

So Fischer et al calculated correction factors to relate the
‘apparent’ volume using Q’ to the actual volume based on
1148 unique, known structures
and their model profiles.

Planning the
neutron scattering experiment
 Choose
your data collection strategy
(solvent matching or contrast variation?)
 Determine how much sample is needed
 Decide which subunit to label
 What deuteration level is needed in the
labeling subunit
 See MULCh*
http://www.mmb/usyd.edu.au/NCVWeb/
*MULCh, Whitten et al, accepted J. Appl. Cryst. 2007
MULCh

ModULes for the analysis of neutron Contrast
variation data



Contrast, computes neutron contrasts of the
components of a complex
Rg, analyses the contrast dependence of the radius of
gyration to yield information relating to the size and
disposition of the labelled and unlabeled components
in a complex
Compost, decomposes the contrast variation data into
composite scattering functions containing information
on the shape of the lab\led and unlabeled
components and their dispositions
Solvent matching
 Best
used when you are interested in the
shape of one component in a complex,
possibly how it changes upon ligand
binding or complex formation.
 Requires enough of the component to be
solvent matched to complete a contrast
variation series to determine required
%D2O (~4 x 200-300 L, ~5 mg/ml).
 Requires 200-300 L of the labeled
complex at 5-10mg/ml.
Solvent Match Point Determination
Front view
90°
Side view
90°
Apical view
G99S
Co-refinement of the 
neurexin positions and
orientations with respect
to NL1 give a model
against the X-ray and
neutron data gives us a
model that we can map
autism-linked mutations
K378R
R451C
V403M
Comoletti, Grishaev, Whitten et al.
Structure 15, 693-705, 2007.
Superposition of solution scattering and
crystal structure for NL-NX
Contrast variation

To determine the shapes and
dispositions of labeled and unlabelled
components in a complex
 Requires  5 x 200-300L (= 1 –
1.5mL) of your labeled complex at  5
mg/ml .
 Deuteration level in labeled protein
depends upon its size.


Smaller components require higher levels
of deuteration to be distinguished.
Ideally would like to be able to take data
at the solvent match points for the labeled
and unlabeled components





Measure sample and solvent
blanks at each contrast point (use
a broad range of D2O
concentrations; e.g. 0,20,40, 80,
100% D2O)
Subtract solvent blank data from
sample
Sample to low-q with sufficient
frequency to determine large
distances accurately (min. 15-20
points in the Guinier region)
Measure to high enough q to aid in
checking background subtraction
(q = 0.45 Å-1)
q = 0.01 -.45 is typical range for 10150 kDa particles, usually requires
two detector positions
Effects of incoherent scattering
from 1H on backgrounds


HCaM
measurement was
done in 42% D2O
to solvent match
the HCaM.
Objective was to
see DCaM in
presence of HCaM,
but without
interference from
HCaM
Incoherent
scattering from 1H
is a constant with q

X-ray scattering data from
LacI, with insert showing
Guinier plot with adequate
sampling.
Use Rg (from MULCh) for Sturhman analysis
2
obs
R


R 

  2
2
m
RKinA = 25.40 Å
RSda = 25.3 Å
D
= 27.0 Å
Rm2  f H RH2  f D RD2  f H f D D 2
  (  H   D ) f H f D RH2  RD2  ( f D2  f H2 ) D 2 
  (  H   D ) 2 f H2 f D2 D 2
Use Compost (from
MULCh) to solve for
I(q)11, I(q)22, I(q)12
I2
I1
I12
I (q)   I (q)   I (q)  12 I12 (q)
2
1 11
2
2 22
Use SASREF7 to do rigid body
refinement of the components
against the scattering data (if you
have pdb files for components)
2
2 = 1.27
2 = 0.97
2 = 0.63
2 = 0.56
2 = 0.76
2 = 0.92
2 = 1.12
2 = 0.95
The described protocols allow the deuteration content in recombinant proteins to be
predicted
Incorporation of deuterium up to 86% of
the chemically Non-exchangeable protons
can be obtained by using D2O as the
deuterium source. Complete deuteration can
only be obtained by addition of
perdeuterated carbon source (glucose or
glycerol).
Use mass spec to determine deuteration
levels.
Neutron scattering sample cells

Helma quartz cells (high precision path-length,
suprasil) – need lots of them!
 Banjo-style (280 L per 1 mm path length) or
rectangular (170 L per 1 mm path length) cells
can be used
 Path lengths are only good to 1%, so good idea
to measure sample and solvent background in
the same cell if practical, but experiment
logistics may prohibit that, so often have to
‘fudge’ background subtractions
 High incoherent scattering for 1H means you
always want  1mm 1H2O in the neutron beam to
avoid multiple scattering
Doing a Quality Experiment

After your final gel filtration step, check out your
samples with dynamic light scattering
 Carefully calibrate your concentration assay –
colorimetric assays are almost useless, extinction
coefficient is good if strong enough, quantitative
amino acid analysis can work
 Compare your data to a well characterized standard(s)
 For protein/DNA complexes, standards are more
difficult. Measure the partial specific volume of your
particle if you have enough sample – or use a good
model to calculate it, e.g. see MULCh or
http://geometry.molmovdb.org/NucProt/
Neutrons






Non-ionizing radiation
Penetrating
Wavelength and energies available that are suitable for
probing structures with dimensions 1-1000s Å
Coherent scattering lengths that vary randomly with
atomic weight and large isotope effect for hydrogen –
contrast variation
Large incoherent scattering cross-section for 1H is a
source of noise in small-angle scattering
Interact weakly with matter and are difficult to produce
and detect – therefore should only be used when they
provide information that cannot be otherwise obtained.
Assessing the quality of
small-angle scattering results

Are there instrumental effects unaccounted for?
 Are the scattering particles mono-disperse and identical or is there
a conformational ensemble?
 Do you have dilute solution conditions?
 Do the data show the expected Guinier and Porod behavior?
 Is the P(r) “well-behaved?”
 Are background subtractions accurate?
 Have standards been measured?
 How well characterized is the sample (purity, concentration)
 Are errors appropriately handled – can you rely on 2?
Jacques & Trewhella (2010)
“Small-angle Scattering for
Structural Biology;
Expanding the Frontier
While Avoiding the Pitfalls,”
Protein Science 19, 642-657