Transcript Slide 1

Direct-method
SAD phasing
Early attempts
1965: Fan, H. F. Chinese Physics, 1429.
1966: Karle, J. Acta Cryst. 21, 273.
1970: Hazell, A. C. Nature 227, 269.
1973: Sikka, S. K. Acta Cryst. A29, 211.1985
1978: Heinerman, J. J. L., Krabbendam, H., Kroon, J.
& Spek, A. L. Acta Cryst. A34, 447.
1982: Hauptman, H. A. Acta Cryst. A38, 632.
1983: Giacovazzo, C. Acta Cryst. A39, 585.
1985: Fan, H. F. and Gu, Y. X. Acta Cryst. A41, 280.
1993: Kyriakidis, C. E., Peschar, R. and Schenk, H.
Acta Cryst. A49, 557.
1998: Fan, H. F. et al. in Direct Methods for Solving
Macromolecular Structures, Kluwer
Academic Publishes, pp. 479-485.
Direct-method phasing of the 2Å
experimental SAD data of the protein aPP
Avian Pancreatic Polypeptide
Space group: C2
Unit cell:
a=34.18, b=32.92, c=28.44Å
b=105.3o
Protein atoms in ASU: 301
Resolution limit: 2.0Å
Anomalous scatterer: Hg, Zn
Wavelength: 1.542Å (Cu-Ka)
Df” = 7.686, 0.678
Phasing: direct methods
Acta Cryst. (1990). A46, 935.
Direct phasing of the 3Å SAD data of the
protein core streptavidin
Space group: I222
Unit cell:
a=95.27; b=105.41, c= 47.56Å
Protein atoms in ASU: 1836
Resolution limit: 3.0Å
Anomalous scatterer: Se
l = 0.9795Å; Df” = 3.663
Phasing:
direct methods +
solvent flattening +
non-crystallographic
symmetry averaging
Acta Cryst. (1995). D51, 342.
OASIS
(The first edition, 2000)
A computer program for breaking
the phase ambiguity in
One-wavelength Anomalous Scattering or
Single Isomorphous Replacement
(Substitution) protein data.
http://www.ccp4.ac.uk/dist/html/oasis.html
Mlphare
The first example of
+solving
dm an unknown
protein by directmethod phasing of
2.1Å SAD data
Rusticyanin,
MW: 16.8 kDa; SG: P21;
a=32.43, b=60.68,
c=38.01Å ; b=107.82o ;
Anomalous scatterer: Cu
Acta Cryst. (1999). D55,
1620-1622.
Oasis
+ dm
Acta Cryst. D58, 1-9 (2002).
Gd-HPDO3A, a complex to obtain high-phasingpower heavy-atom derivatives for SAD and MAD
experiments: results with
tetragonal hen egg-white lysozyme
Éric Girard, Laurent Chantalat, Jean Vicat and Richard Kahn
Laboratoire de Cristallographie Macromoléculaire, Institut de
Biologie Structurale J.-P. Ebel CEA-CNRS-UJF, 41 Rue Jules
Horowitz, 38027 Grenoble CEDEX 01, France
OASIS
OASIS
+ DM
J. Mol. Biol. 348, 951–959 (2005)
Crystal Structures of Fms1 and its Complex
with Spermine Reveal Substrate Specificity
Qingqiu Huang1, Qun Liu1 and Quan Hao1,2
1MacCHESS
at the
Cornell High Energy
Synchrotron Source,
Cornell University
Ithaca, NY 148538001, USA
2Institute of Physics,
Chinese Academy of
Sciences, Beijing
100080, China
Science, Vol. 306, Issue 5693, 104-107 (2004)
Basis for structural diversity
in homologous RNAs
Andrey S. Krasilnikov*, Yinghua Xiao*, Tao Pan†, and Alfonso Mondragón*
* Department of Biochemistry, Molecular Biology and Cell Biology,
Northwestern University, Evanston, IL 60208, USA
† Department of Biochemistry and Molecular Biology, University of
Chicago, 920 East 58th Street, Chicago, IL 60637, USA
For SAD phasing, the positions of the first 3 Ba+2 sites were
identified using SOLVE, 6 more Ba+2 sites were identified and
added during heavy atom refinement with SHARP. The phase
ambiguity in SAD phasing was partially resolved using OASIS and
solvent flattening with SOLOMON as implemented in SHARP.
Further improvement of the phases was achieved by doing iterative
cycles of phase refinement incorporating phase information from
partially built models followed by solvent flattening.
構造生物 Vol.10 No.1 2004 年2 月発行
CrKα線を用いたSAD 法による位相決定
理学電機(株) 山野昭人、佐藤貴久、長谷川智一
初期位相の決定はSHARP やOASIS が成績が良い。MLPHARE でも可能との
事だ が、筆者の 使用法の問 題だと思われるが 、これまで 成功した経
験がない。
図2 はサウマチンの解析例である。1
フレームを0.5度振動、1 分露光で180
度分測定したイメージをHKL2000 で処
理。15-3Åのデータを用いてSHELX97
を実行。8 個のS-S の位置と1 個の硫
黄原子の位置全てが決定できた。これ
らの座標を基にOASIS を実行。初期位
相を決定した。DM およびSOLOMON に
より位相を改良した。
OASIS-2004
A direct-method program for
ab initio phasing and reciprocal-space
fragment extension with SAD/SIR data
Institute of Physics, CAS, Beijing, P.R. China
http://cryst.iphy.ac.cn
Difficult SAD phasing
1. SAD phasing at Bijvoet ratio
~ 0.56%
2. An originally unknown protein
with 1206 residues in the ASU
solved automatically using CrKa sulfur-SAD data
OASIS-2004
application
Contoured at 1s
Xylanase
Space group: P21
Unit cell: a = 41.07, b = 67.14, c = 50.81Å
b = 113.5o
Resolution limit: 1.75Å; Multiplicity: 15.9
Anomalous scatterer: S (5 )
X-rays: synchrotron radiation
l = 1.488Å; D f ” = 0.52
Bijvoet ratio: <|DF |>/<F > = 0.56%
Phasing: OASIS-2004 + DM (Cowtan)
Model building: RESOLVE BUILD &
ARP/wARP found 299 of the total 303
residues at the 6th cycle of iteration
Data courtesy of Dr. Z. Dauter,
National Cancer Institute, USA
OASIS-2004
application
Space group: P21212
Unit cell:
a = 100.2639
b = 108.9852
c = 114.6272Å
Number of residues
in the ASU: 1206
Resolution range:
50.00-2.01Å
Multiplicity: 20.9
Anomalous scatterer: S (22)
Wavelength:
l = 2.291Å; Df ” = 1.14
Bijvoet ratio: <|DF|>/<F> = 1.16%
Phasing:
OASIS-2004 + DM (Cowtan)
Model building:
RESOLVE BUILD & ARP/wARP
TT0570
ARP/wARP found 1153 of the total 1206 residues
after 2 cycles of iteration
Data courtesy of Professor Isao Tanaka & Dr. Nobuhisa Watanabe
Graduate School of Science, Hokkaido University, Japan
Features of
OASIS-2004
1. Better initial SAD phases
 h   "h  D  h
P  D  h  
1
2

1

tanh sin D  h    
 c 
2
 m m 

 h' h  h' h , h' sin   3  D  h' best  D  h  h' bes t    sin  h
 h'

1

tan( D  h best )  2 P ( D  h ) 
sin D  h


2


cos D  h
1
m h  exp   s
2
h
2
  
 2
1
1
2    2  P  D  h       1  cos 2 D  h   cos 2 D  h 
2
2 
  

Phase information
available in SAD
Bimodal distribution
of SAD
Sim
distribution
 " D 
Peaked at
   "
p
2
"
The phase of
F”
Cochran
distribution
 " D 
Peaked at
any where
from 0 to 2p
Two different kinds
of initial SAD phases
PSim  PBimodal
Sim-modified
phases
PSim PCochran
P+-modified
phases
P 
P+
+
P+ 
Histone Methyltransferase Set 7/9
Space group: P212121
Unit cell: a = 66.09, b = 82.83, c = 116.15Å
Number of residues in ASU: 560
Number of independent reflections: 16352
Resolution limit: 2.8Å
Multiplicity: 3.8
Anomalous scatterer: Se(12)
l = 0.9794Å; Df’ = -7.5, Df” = 6.5
Bijvoet ratio: <|DF|>/<F> = 7.03%
SAD phasing by OASIS-2004 + DM
Data provided by Dr. S. J. Gamblin
and Dr. B. Xiao
Cover figure of Acta Cryst. D60, Part 11 (2004)
Comparison of the two kinds of initial phases
using 4 typical reflections from the protein
histone methyltransferase SET 7/9
Comparison on cumulative phase errors
sorted in descending order of Fobs
Number of reflections
Errors of Sim-modified
phases ( o )
1500
57.1
45.8
3000
57.1
49.1
4500
56.5
50.0
6000
57.0
51.2
7500
57. 8
52.9
9000
58.7
54.1
10500
59.4
55.6
12000
60.8
56.9
13500
61.9
58.4
15000
63.4
60.2
16352
65.2
62.3
Errors of P+-modified
phases ( o )
2. Inclusion and auto balance of the
lack-of-closure error in the direct-method
phasing formula
P  D  h  

1

2
1

2
tanh sin D  h   m h' m h  h' h , h' sin   3  D  h'best  D  h  h'best    sin  h 
 h'

m h  exp   s
s
2
h



1
1
 

2    2  P     1  cos 2 D  h   cos 2 D  h 
2
2 


  

2
2
h
 ns 
D Fh
2 Fh "
2
2
exp   s
2
h
2 
1
2
1 2

Automatic solution of protein
structures
by a single run of
RESOLVE
(Build only)
and/or
OASIS-2004 + DM +
ARP/wARP
OASIS-2004
application
Contoured at 1s
Pepstatin-insenstive carboxylproteinase
Space group: P62
Unit cell: a = b = 97.31, c = 82.94Å, g = 120o
Resolution limit: 1.8Å; Multiplicity: 5.45
Anomalous scatterer: Br (13)
X-rays: synchrotron radiation
l = 0.9191Å; D f ” = 5.0
Bijvoet ratio: <|D F |>/<F > = 7.06%
Phasing: OASIS-2004 + DM (Cowtan)
Model building:
ARP/wARP found 358 of the total 372 residues
Data courtesy of Dr. Z. Dauter,
National Cancer Institute, USA
OASIS-2004
application
Contoured at 1s
Porcine Pancreatic Elastase
Space group: P212121
Unit cell: a = 50.2, b = 58.1, c = 74.3Å
Resolution limit: 1.94Å;
Total rotation range: 360o
Anomalous scatterer: Xe (1)
X-rays: synchrotron radiation
l = 2.1Å; D f ” = 11.8
Bijvoet ratio: <|D F |>/<F > = 5.76%
Phasing: OASIS-2004 + DM (Cowtan)
Model building:
ARP/wARP found 236 of the total 240
residues
Data courtesy of Dr. M. S. Weiss, EMBL
Hamburg Outstation, c/o DESY, Germany
OASIS-2004
application
Contoured at 1s
Lysozyme
Space group: P43212
Unit cell: a = 78.81, c = 36.80Å
Atoms in the asymmetric unit: 1001
Resolution limit: 1.53Å; Multiplicity: 23
Anomalous scatterer: S (10), Cl (7)
X-rays: synchrotron radiation
l = 1.54Å; D f ” = 0.56, 0.70
Bijvoet ratio: <|D F |>/<F > = 1.55%
Phasing: OASIS-2004 + DM (Cowtan)
Model building:
ARP/wARP found 128 of the total 129
residues
Data courtesy of Dr. Z. Dauter,
National Cancer Institute, USA
OASIS-2004
application
YfbpA
Space group: P212121
Unit cell: a = 42.792,
b = 54.134,
c = 115.222Å
Resolution range: 57.74 – 1.42Å
Anomalous scatterer: Fe (1)
Wavelength: 1.542Å
Df ” = 3.20
<|DF|>/<F> ~ 1.4%
Phased by:
OASIS + DM (Cowtan)
Automatic model building by:
ARP/wARP
Data provided by Dr. Cheng Yang
Rigaku/MSC, USA
302 residues found automatically
3. Dual-space fragment extension
Partial
model
Reciprocal-space
fragment extension by
OASIS-2004 + DM
Real-space
fragment extension by
RESOLVE BUILD
and/or ARP/wARP
No
P  D  h 



  tanh  sin D  h    
2 2



1
1
OK?
Yes
End


  m h' m h  h' h , h' sin   3  D  h' best  D  h  h'best    si n  h  
 h'

Partial
model
Examples
and
Case study
Lysozyme
S-SAD
Cr-Ka
98%
52%
Cycle 6
0
Cycle 3
0
95%
42%
Azurin
Cu-SAD
Synchrotron l = 0.97Å
Glucose
isomerase
S-SAD
Cu-Ka
99%
25%
Cycle 6
0
0
Cycle 4
52%
97%
Cr-Ka
Se, S-SAD
Alanine racemase
17%
97%
Cycle 0
6
Xylanase S-SAD
Synchrotron
l = 1.49Å
OASIS-2004
application
Contoured at 1s
Xylanase
Space group: P21
Unit cell: a = 41.07, b = 67.14, c = 50.81Å
b = 113.5o
Resolution limit: 1.75Å; Multiplicity: 15.9
Anomalous scatterer: S (5 )
X-rays: synchrotron radiation
l = 1.488Å; D f ” = 0.52
Bijvoet ratio: <|DF |>/<F > = 0.56%
Phasing: OASIS-2004 + DM (Cowtan)
Model building: RESOLVE BUILD &
ARP/wARP found 299 of the total 303
residues at the 6th cycle of iteration
Data courtesy of Dr. Z. Dauter,
National Cancer Institute, USA
Xylanase: average phase error decreased
during dual-space
iteration
Is OASIS
80
necessary
here?
Phase error in degrees
70
Is OASIS
necessary
here?
Yes
60
50
What would it be
without using
RESOLVE
OASIS-2004
(build only)?
40
 DM
30
 Partial model from
RESOLVE BUILD
or ARP/wARP
20
No
10
0
1
2
3
Cycle
4
5
6
OASIS-DM-ARP/wARP Iteration
Xylanase sulfur-SAD phasing
Synchrotron radiation l = 1.49Å, <DF>/<F> = 0.56%
Phase error in degrees
80
70
60
50
 OASIS-2004
40
 DM
 Partial model from
ARP/wARP
30
20
0
2
4
6
8
Cycle
10
12
14
16
Improvement on electron-density map and
automatic model building
Cycle 0
Cycle 3
Cycle 6
Inside direct-method
SAD phasing
SAD formulation
SAD formulation
N
F (h) 
F
( h ) f F' (ihf) ")Fexp(
'( h )i 2pFh"(rh))
(
f

j
j
j
o
o
j
j 1
F * (  h )  F ( h )  F '( h )  F "( h )
o
F
F”
2F "
FF ++
Fo
F* 
- F”
 " h 
F’
F ( h )  F * (  h )  2 F "( h )
SAD formulation
F (h ) 
F
F
F (h ) 
+
F ( h ) exp  i  h  
F


(h )  F (h )

 (h )
D h
 " h 
 D h 
F+
F"
?
 ( h )   "( h ) 
F
D  (h )
2
P+ formula
PC ochr a n  D  h   N exp    h , h ' c os  D  h  b '  
 h'

 h , h '  2s 3s
E E h ' E h  h ' , s n   Z
PC ochran   h   Nh exp
  h , h ' cos   h   h '   h  h '  
3 / 2
2
n
j
 h'
j

b '   3  D  h '  D  h  h ' ,  3    "h   "h '   "h  h '
 h   "h  D  h
PS i m  D  h   N ' exp   co s  D  h   h  
  2 E h E h , p s u , s u   Z s 2 ,  h   'h   " h
PSim  h   N ' exp u cos   h   ' h  
2
u
w here
p denotes the partial structu re;
u denotes the unknow n part of the structure.
P  D  h   PC ochran  D  h   PSim  D  h  
 2p I 0  a  
1
exp a cos  D  h  b  
2
a
2


    h , h ' sin   ' 3  D  h '  D  h  h '    sin  h  
 h'

h
Maximizing P(D ) 

cos  D
' 
 Dh=bD 
   cos 
 

tan b 
h ,h '
h'

h ,h '
3
h'
hh '

h

2
sin   ' 3  D  h '  D  h  h '    sin  h
h'

h'
h ,h '
cos   ' 3  D  h '  D  h  h '    cos  h
tan D  h 

h ,h '
sin   ' 3  D  h '  D  h  h '    sin  h
h'

h ,h '
cos   ' 3  D  h '  D  h  h '    cos  h
h'
1 1
Replacing
Ehexp(ia) with
P  D    
2 2
m
E
exp(i
a
)


h
h
best

tanh sin D 
sin    D   D 

   sin 

h

h
 h '
h , h'
3
h'
h  h'

h


tan D  h 
m
h'
m h  h ' h , h ' sin   ' 3  D  h ' best  D  h  h ' best    sin  h
h'
m
h'
m h  h ' h , h ' cos   ' 3  D  h ' best  D  h  h ' best    cos  h
h'
Fan, H.F. & Gu, Y.X.,
Acta Cryst. A41, 280-284 (1985)
P  D  h  
1
2


1
tanh sin D  h    
 c 
2
 m m 


sin


D


D



sin

 3

h' best
h  h' bes t 
h
 h' h' h  h' h , h'

m h  exp   s
2
h
Fan, H.F. & Gu, Y.X.,
Acta Cryst. A41, 280-284 (1985)
2
  

1
1
  2  P     1  cos 2 D  h   cos 2 D  h 
2
2 
  

2
D  h best   h best   "h
tan( D  h best )  2( P 
1
2
) sin D  h
cos D  h
1 2

Thank you!