Bethesda, March 4th 2009

Semi-automatic structure solution with

HKL-3000

Structural Biology

Bethesda, March 4th 2009

Semi-automatic structure solution with

HKL-3000

Structural Biology

Bethesda, March 4th 2009

Structural biology bottlneck

I have a dream:To convert high throughput, high output SG center into, not necessary high throughput, but very high output center.

Can we use sequence to predict: solubility, crystallization, ….. structure, … function, ….

Bethesda, March 4th 2009

HKL/HKL-2000 vs. HKL-3000

HKL/HKL-2000 h, k, l, F, σMethods in Enzymology (1997) 275: 307-326. 

Times cited: 18018

HKL-3000 partially build model Acta Cryst. D (2006) 62: 859-66. 

Times cited: 77

PDB deposits: >710 

Bethesda, March 4th 2009

Integrated system – HKL-3000current status

HKL-2000, DENZO, SCALEPACK SHELXD, SHELXE MLPHARE, CYCLMLPHARE DM, REFMAC SOLVE, RESOLVE ARP/WARP STATBUILD O, COOT,CCP4

Interactive development:Integrated approach shows how the choice in one step influences subsequent steps.

Bethesda, March 4th 2009

HKL-3000 six mouse click programEngage your brain !

Bethesda, March 4th 2009

or Automatic structure solution

HKL

Bethesda, March 4th 2009

Why to engage my brain ?One little detail …

O-ring

Bethesda, March 4th 2009

IntegrationHKL-3000 is a pipeline component

Other systems provide additionalcontextual up- and downstreaminformation

Pepdbexpression & purification

Xtaldbcrystallization

HKL-3000Data collection,

diffraction & structure solution

Wetlabchemicals & solutions

Bethesda, March 4th 2009

Project setup Check against PDB during data collection

Bethesda, March 4th 2009

Completeness – overlapsMulti-crystal strategy

Bethesda, March 4th 2009

Priceless crystals

Star of Africa Crystallography lab

Bethesda, March 4th 2009

Scaling

Signal from Sulfur

Bethesda, March 4th 2009

Feedback of possible hardware, sample or experiment problems

Detector Goniostat Cooling

Crystal decay Experiment Twinning …..

Bethesda, March 4th 2009

Possible Pseudo-translational Symmetry

Bethesda, March 4th 2009

SAD or MAD phasing

Bethesda, March 4th 2009

Substructure solution

Data management in HKL-3000 - substructure solution

Bethesda, March 4th 2009

Substructure enantiomorph determination

Bethesda, March 4th 2009

Double conformations

Bethesda, March 4th 2009

Optimization of solvent content worthwhile but slow

automaticoptimization

manual

Bethesda, March 4th 2009

NCS identification and use

From sites From map

Bethesda, March 4th 2009

Model building and preliminary refinement (10 minutes, 190aa)

Secondary structure after each cycle

Unidentified density -> Xtaldb

Bethesda, March 4th 2009

TM1086 (APC4579)

10 Se-Met/282 aa35.1kDa

I222a= 290Å, b= 300Å, c=316Å

~81% solvent content = 17 molecules in AU~38% solvent content = 57 molecules in AU

Bethesda, March 4th 2009

TM1086 (APC4579)

282 aa35.1kDaP3221

a= 165Å, b= 165Å, c=98Å~78% solvent content = 3 molecules in AU~40% solvent content = 8 molecules in AU

Bethesda, March 4th 2009

TM1086 (APC4579)

Bethesda, March 4th 2009

I222 crystal form

~210 Å

30 chains in AU

8460 aa

1.05MDa

Bethesda, March 4th 2009

APC81521 ( ChangSoo)

Final model with right targetAPC81521

Attempt to use automatic building with wrong targetAPC81501

To figure out the right target at the resolution of 2.5Ǻ1. Manually build helixes and stands;2. Submit this model to dali to find out several similar structures in PDB;3. Search targetDB(MCSG) using the sequence of these structures to find out the right target.

MCSG sequence from electron density map

Bethesda, March 4th 2009

2as8 1.95Å resolution, R=18.8/17.4, Rfree=25.2/23.1

Bethesda, March 4th 2009

# of PDB deposits with given # of Mg ions

1 1273

2 943

3 210

4 358

5-10 274

11-50 80

51-100 33

>101 52

# of Mg # of deposits

Bethesda, March 4th 2009

Four deposits - 1896 Mg ions

Mg-O distance distribution

0

5

10

15

20

25

30

35

40

1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2

Mg - O distance (Å)

%

2a68,2a69,1iw7,1smy

CSD

HR PDB

Bethesda, March 4th 2009

Unknown ligands

Bethesda, March 4th 2009

Completeness of data and completeness of the model (3bqs)

Bethesda, March 4th 2009

3bqs 1ze0

Acta Cryst. D 63(2007)348-54

Bethesda, March 4th 2009

Molecular Replacement

SGCZoledronic acid

in Cryptosporidium parvum

Arun MalhotraRNase T

Bethesda, March 4th 2009

Directions for a future

Using NCS in structure solution

Using high-resolution native data with low-resolution phasing data

The boundaries can be moved by improving diffraction data quality or by using more elaborate methods

Resolution [Å]4.0

Unique content in ASU [%] = = (1 – solvent content)/NCS factor

No methods to interpret such maps

Arp/wARP builds

most of the model Atomicity

based phasing e.g. ME

Challenging projects

70

40

50

60

10

30

20

2.0 1.52.53.03.5

Bethesda, March 4th 2009

People involved

Wladek Minor• Marcin Cymborowski • Maks Chruszcz• Matt Zimmerman• Heping Zheng• Marek Grabowski

Zbyszek Otwinowski Dominika Borek Andrzej Kudlicki

Andrzej Joachimiak• Rongguang Zhan• Youngchang Kim• Marianne Cuff• Chris Lazarski and SBC/MCSG staff

Zbyszek DauterTom Terwilliger Solve/ResolveGeorge Sheldrick Shelxd/ShelxePaul Emsley CootGarib Murshudov Refmac

Grants: NIH GM53163, GM62414, GM74942NIAID HHSN272200700058C DOE, NCI, HKL Research

Bethesda, March 4th 2009

Possible pseudo-translational symmetry