Protein Production and Crystallization Workshop

Structural Genomics of Pathogenic Protozoa (SGPP)

Crystal Growth LabLori Anderson

www.sgpp.org

SGPP Crystal Growth Pipeline

Protein Production UnitUniversity of Washington

Seattle Washington

Protein Production UnitUniversity of Rochester

Rochester New York

Crystal ScreeningHauptman Woodward

Medical Research InstituteBuffalo New York

Crystal Growth University of Washington

Seattle Washington

Structure Determination Unit 1University of Washington

Seattle Washington

Structure Determination Unit 2University of Washington

Seattle Washington

Flash Frozen Protein

Images

Crystals

SGPP Crystal Growth Protein Totals

• 333 Protein Samples Received of 192 Different Targets, one antibody and one target-antibody complex

Species # of Targets

Total # of Samples

# of SelMet Targets

# of SelMet Samples

L. aethiopica 4 4 0 0

L. braziliensis 2 4 1 1

L. donovani 8 10 2 2

L. guyaneis 1 1 1 1

L. infantum 4 7 0 0

L. major 139 242 50 61

L. mexicana 5 8 2 2

L. tarentolae 3 3 0 0

L. tropica 4 7 0 0

P. falciparum 22 47 6 9

Totals 192 333 62 76

SGPP Crystal Growth Lab Seattle

• Crystals from 26 targets sent downstream to the Structure Determination Units

• Crystals sent downstream represent 4 different species– 24 L. major – 2 P. falciparum – 1 L. donovani– 1 L. braziliensis

• 8 structures to date – 5 structures deposited into PDB

• 4 L. major • 1 P. falciparum

Robots in Seattle SCU

Oryx-6 Hydra II Plus 1

Acapella

RoboDesign

MicroScope II

Refined Optimization

Matrices Made by Hand

Initial Optimization or Screen

Database and Image Archive

Harvestable Crystals

Crystallographer Review Manual Scoring

Structure Determination UnitsLarry Desoto

SGPP Crystal Growth Bottleneck:Making Solutions

• Optimization matrices made by hand • Hydra II Plus 1 uses 96 deep well plates• 96 well plate set-up takes over 2 hours to prepare

SOLUTION:

Alchemist I

• Designed by Robodesign• Anticipated arrival April 2004• Set-up 96 deep well plate in 20

minutes

Where Will We Be?

Oryx-6 Hydra II Plus 1

Acapella

RoboDesign

MicroScope II

Refined Optimization Matrices Made by Hand

Initial Optimization or Screen

Database and Image Archive

Structure Determination Units

Crystal Track Designs Matrices

Larry Desoto

Crystallographer Review Manual Scoring

Harvestable Crystals

SGPP Crystal Growth Lab:Special Projects

Problem: How do we increase the success for getting crystals?

SGPP Approches

• Small Molecule Co-Crystallants; Erkang Fan, Christophe Verlinde UW

• Antibodies; Mark Sullivan UR

Co-crystallants and Crystallization (I)A Case Study

• Pfal008421WES; Uridine phosphorylase-like– Acyclovir is a potential inhibitor of Uridine phosphorylase-like

• Screening of the SulMet protein in absence of Acyclovir – SulMet produced 132 Hits at HWI– Optimization produced SulMet crystals that were ~35µm X 35µm

• Screening of the SelMet protein in absence Acyclovir – SelMet had 21 Hits at HWI– Optimization produced SelMet crystals that were ~50µm X 50µm– SDU #2 was able to mount SelMet crystals and obtain diffraction

at ALS

• 5 generations of optimization of protein without Acyclovir, DID NOT produce larger crystals

Co-crystallants and Crystallization (II)

• SulMet plus Acyclovir – SulMet-Acyclovir complex produced 235 hits

at HWI– One generation of optimization in Seattle

produced 100µm X 100µm crystals

• The best of the Pfal008421WES/Acyclovir complex hits were unique and DID NOT overlap with the initial SulMet hits

Co-crystallants and Crystallization (III)

0.1 M Ammonium Sulfate0.1 MOPS pH 6.525% PEG 3350Drop size 1ul + 1ul

Pfal008421WES Native Pfal008421WES Native plus Acyclovir

0.2 M Ammonium Sulfate0.1 M HEPES pH 7.525% PEG 33502.1mM AcyclovirDrop Size 0.4ul + 0.4ul

UNMOUNTABLE MOUNTABLE

SGPP Crystal Growth Lab:Special Projects

• DLS and Proteolysis– Problem: Is there any way to predict which

proteins will crystallize?

DLS and Limited Proteolysis (LP): Definitions (1)

Good

Medium

Bad

NS-nonstable proteinSF-stable fragmentSP-stable proteinSSF-superstable fragmentSSP-superstable protein

NS SF SP SSF SSP

0 1 24 0 1 24 0 1 24 0 1 240 1 24

DLS Limited Proteolysis

Oleksandr Kalyuzhniy

DLS and Limited Proteolysis (LP) as a Predictor of Crystallization (2)

0%

20%

40%

60%

80%

100%

Medium

Bad

Good

Medium 6 19 11 36

Bad 10 20 17 47

Good 29 29 25 83

Definite Questionable No hits Total

Hauptman Woodward Institute Results (July 2003-February 2004)Total………………….166 samples (100%)Definite……………….45 samples (27%)Questionable………...68 samples (41%)No hits………………..53 samples (32%)

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

NS

SF

SP

SSF

SSP

NS 6 10 21 37

SF 4 4 5 13

SP 5 6 6 17

SSF 2 6 1 9

SSP 10 5 5 20

Def inite Questionable No hits Total

166 samples, 141 ORFs 96 samples, 86 ORFs

DLS Limited Proteolysis

Oleksandr Kalyuzhniy

DLS and LP as a Predictor for Good Diffraction (3)

0%

20%

40%

60%

80%

100%

Medium

Bad

Good

Medium 1 0

Bad 3 0

Good 20 11

Harvestable Diffractors

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

NS

Stable

NS 1 0

Stable 18 7

Harvestable Diffractors

Seattle Optimization Results:• Harvestable Crystals-- 24 Targets• Good diffractors-- 11 Targets

DLS Limited Proteolysis

Oleksandr Kalyuzhniy

Crystal Growth Summary

• 333 Protein Samples Received– 192 Targets

• 26 targets sent to SDU• Crystal Track Database has been implemented • Alchemist Liquid Handling Gradient Maker

anticipated arrival April 2004• Known co-crystallant greatly improved

crystallization• If your protein has good DLS and is “stable”,

then the crystal is more likely to get good diffraction

Acknowledgments

• Thanks to all SGPP members especially:– The SGPP Protein Production Units– The SGPP HWI Crystal Screening – The SGPP Antibody Unit– The SGPP Co-crystallant Unit – The SGPP Structure Determination Units– NIGMS AND NIAID