Recent Advances in Protein Powder Diffraction

1
Recent Advances in Protein Powder Diffraction
Reaching for High Resolution in Protein Powder
Diffraction

• R.B. Von Dreele, XSD/IPNS
• Argonne National Laboratory, USA

Thanks Peter Stephens, Peter Lee, US
DOE/OS/BES
2
What is a powder? - polycrystalline mass
All orientations of crystallites possible
Sample 1ml powder of 1mm crystallites - 109
particles
Packing efficiency typically 50 Spaces air,
solvent, etc.
Single crystal reciprocal lattice - smeared
into spherical shells - the overlap problem
lost information
3
Powder diffraction - reciprocal space
Braggs Law
Ewald sphere
Typical 1-D scan
2Q
1/l
s
d
so
Smear in 3D
Spherical reflection shells
4
Radiation damage initial observations _at_NSLS
X3b1 Xtal/analyzer detector
1 day
Problem Severe radiation damage of proteins
Much worse at APS!! Happens in 10-20min!!
5
Faster data collection
Image Plate Detector MAR345
Beam focused to IP surface IP offset 6-10cm up
l0.6Å
Beam stop
Sample (spun, 1x1mm, lt1mg)
350mm
700mm
Guinier geometry
6
Compare image plate with analyzer/detector
11BMB 10min scan
1BM/MAR345 1sec exposure
7
Rings protein pattern (HEWL) X-rays 30s _at_
20kV on MAR345 lt1mg HEWL
Texture free sample no graininess 1mm
perfect powder
Resolution limit 1.85Å
Residual solvent scattering background
Inner most ring d55Å (110) Reflection, lowest
order for tetragonal lysozyme 2Q 0.67deg
Beam stop holder
9000 Fhkl for HEWL gt2Å
(Air, solvent Kapton background subtracted)
8
Powder diffraction from 2D image plates (MAR345)
Best focusing best resolution FIT2D cake
integration to dmin 2Å resolution
0.035o FWHM 300mm beam/pt. spread fxn. 4X
sample contribution!
9
HEWL comparison 30s on 1BM Mar345 vs 10min on
11BM
20keV
30keV
10
Background problem subtract air, liquid Kapton
Before subtraction Gave too small Rwp
(lt0.5) from high background No sensitivity to
structure
After background subtraction Weights Compensat
e for 2D detector integration effect
11
Radiation damage reflection intensities
positions
APS 1BM - 30s exposures 150s delay, 300K Full
sequence wrt NaCl pH effects?? Immediate
changes seen 2 stages? - lt10min gt10min exposure
Focus here a up c down
12
Multiple patterns different lattice strains
Buffer effect
phthalate (ltpH5) Phosphate(gtpH5)
Actually solvent effect?
Focus here Shortest exposure
13
Solvent radiation damage induced lattice
strains for HEWL
Salt pH effects
Radiation exposure 8 loss in 4.5m
Compare DM 0.07M DpH 0.07 (DH-15)
in 4.5m
cf. FWHM 0.035o2Q 0.02 strain _at_18o2Q
14
Analysis Induced lattice strain model obey
Laue symmetry
From d-spacing expression
partial derivatves wrt gij of Braggs law gives
And symmetrized for tetragonal
a-axis strain for tetragonal is (c-axis similar)
Peak shape function 5 in GSAS
15
2.5-2.0Å resolution range NaCl sequence
Obsd calcd powder patterns
0.25M
Peak shifts!
1.25M
Also for rad. dam. just less
16
Profile fit 1/5 patterns Rwp1.84
17
HEWL superposition of 3 determinations
(NaCl,pH5NaCl,pH4 RD) H2O independently detn.
H2O many common positions ( some
not) Variations? RMSD0.4Å (all protein atoms)
18
Structure quality?
Ramachandran plot 90 most favored
Total OMIT map protein H2O
19
HEWL results structure comparisons
Appropriate for 2Å resolution?
20
Compare single pattern result (1JA2) best
single xtal
ERRAT2 atom neighbor analysis - packing
5 NaCl patterns - this work Rad. Dam. Seq. similar
1JA2 - powder PDB
Q57.63
Q98.33
PDB 1IEE
Q94.22
Compare best HEWL single crystal result (low
temp xtal from shuttle!)
Howzat?
21
Overlap factor effect of lattice strain better
effective resolution?
1 pattern
5 patterns
2 patterns
SC 0
Overlap factor, Rij1-Dij/2FWHM 1 if complete
overlap, 0 if no overlap FiSmin(Rij) for
multipatterns
D0
Dgt2FWHM
22
Sample size limit? in situ?

• 1st experiments NSLS X3b1 1 analyzer/detector
  10mg HEWL slurry 6hr scan
• Image plate APS 1BM MAR345 1mg HEWL slurry
  30s exposure
• CCD APS 8BM (now defunct) 15mg HEWL slurry
  10s exposure

2 of 1536 well plate xtal growth test not
particularly optimized corners lt3Å
23
High throughput screening for crystallization
X-ray
8BM 12keV 1536 well plate 1st real expt. 4
plates look for spots/rings? Compare optical
pix ADSC 315 20Mb each pix 5-6
DVDs/plate! ROBOT!!
Craig Ogata, et al. Blessing, et al.
24
Detector development spatial resolution
ADSC 100mm
MAR345 300mm
35mm ideal match with sample
25
Tileable area detectors?
Cover best part of powder pattern Tilted array
avoid blind spots in powder pattern
Cover curved area to match resolution? Not
necessarily spherical!
26
Conclusion data combinations in proteins (at
least HEWL)
Protein powder diffraction ? Image plates
lower powder resolution, but Better intensity
measurement to higher diffr. resolution Induced
lattice strain from RD, pH, salt, etc. variation
? multiple powder patterns ? lattice
variation Recover powder resolution Result
higher powder diffraction resolution ?
better protein structure (including water
molecules) Future smaller samples better
resolution Structure solution multiple data set
extraction of Fo