Astbury Centre University of Leeds Lessons Learned from

Astbury Centre, University of Leeds Lessons Learned from Crystallising Protein DNA Complexes. Synergy between High-Throughput and Novel Crystallisation Techniques Jonathan Hadden jmh@bmb. leeds. ac. uk Astbury Centre for Structural Molecular Biology University of Leeds

Astbury Centre, University of Leeds Overview of Talk • Introduction to the project • The High-throughput approach we took (twist due to funding) • Problems we encountered • Optimisation approaches used to overcome the problems

Astbury Centre, University of Leeds Personal perspective on crystallisation • Increasing throughput has very much driven developments in protein crystallisation over past 10 years • Fantastic benefits: faster setup, less sample required, more experiments • Increasing throughput usually means significant miniaturisation • Resultant scale-up required during optimisation • Scale-up particularly important for poorly diffracting crystals or those sensitive to radiation damage. E. g. Protein/DNA complexes or Membrane proteins. Where some of the largest challengers are • BIG CRYSTALS REQUIRED – BUT OUR DROPS ARE GETTING SMALLER!

Astbury Centre, University of Leeds Trying to Crystallise a Holliday Junction Resolving enzyme complex since 1984 A Holliday junction is a four-way DNA junction and is a central intermediate in recombination Recombination (a ubiquitous biological process) The exchange of one DNA sequence with an other or The incorporation of one DNA sequence into an other Many examples of recombination in nature Exchange of homologous DNA during meiosis in diploid organisms Generation of diversity in Immunoglobulin genes Integration of bacteriophage DNA into host DNA

Astbury Centre, University of Leeds Classical (Holliday) scheme for homologous recombination Strands nicked Holliday Junction Branch migration Strand invasion Endo I or Splice Recombination Patch Recombination Junction resolution

Astbury Centre, University of Leeds Structure of a Holliday junction Structure of a Resolving enzyme Ortiz-Lombardia et al. 1999, NSB 6, 913 -917 Hadden et al. 2002, EMBO J. 21, 3505 -3515 Endo I Cut

Astbury Centre, University of Leeds The Approach • Purify protein DNA complexes – limited quantities for crystallisation • Use a mutant Endo I that binds junction but does not cleave • High-throughput crystallisation – 960 condition screen • Protein constant • Varied the DNA sequence – up to 100 sequences • Potentially 96, 000 trials!

Astbury Centre, University of Leeds Problem Prepare 250 nl + 250 nl crystallisation experiments in 96 well vapour diffusion plates at minimal cost with technology available 8 years ago Solution Use a Douglas instruments ORYX 6 robot

Astbury Centre, University of Leeds Another problem The robot is comparatively slow - 18 minutes to set up a tray Evaporation of 500 nl drops is a big problem Another solution Cover drops with light (1 c. S) silicone oil. The oil will initially protect the drop but eventually evaporate into the sealed atmosphere

Astbury Centre, University of Leeds

Astbury Centre, University of Leeds Add screening solutions to plate using 8 -channel pipette Aspirate screening solution, protein already loaded Dispense 0. 25µl of protein and screening solution Dispense 5 µl of light silicone oil onto drop Seal with film and wait for oil to evaporate

Astbury Centre, University of Leeds

Astbury Centre, University of Leeds Some hits!

Astbury Centre, University of Leeds We needed to grow much larger crystals • Optimise in 24 -well plates 1μl+1μl or 2μl+2μl drops • Quite often conventional methods failed dismally! • “All or nothing effect” (lots of small crystals or non at all) • Used two methods to successfully overcome these problems - on most occasions Equilibration rate control Seeding • However, neither performed conventionally

Astbury Centre, University of Leeds Equilibration rate control • Modify the rate of equilibration in a vapour diffusion experiment • Traditionally – Temperature Size of drop • More recently- Use a barrier between drop and reservoir • Bobs gadget (Equilipro) • Silicone fluid on reservoir (Naomi Chaen)

Equilibration rate control – silicone fluid Astbury Centre, University of Leeds • Silicone fluid on reservoir Early stages - Equilibration of drop / unsaturated air space Silicone Fluid Significantly better results • Silicone fluid on drop Early stages – protected drop Silicone Fluid

Astbury Centre, University of Leeds Equilibration rate control Variables • Viscosity of fluid – 1 c. S or 5 c. S works well • Volume of fluid - 20μl is good in a standard microbridge with a 2μl-4μl drop • Pre-equilibration time (time between setting up drop and covering drop with oil - 0 Hours) • PPT concentration – usually +0 to +3% works well SIGNIFICANTLY LARGER CRYSTALS FOR ~80% OF TARGETS!

The remaining 20% - Seeding Astbury Centre, University of Leeds Higher-throughput seeding • Simple approach - streak seeding • Fast turn round - find optimal conditions quickly • Used drop pre-incubation times to control crystal size, quality, numbers. • Well solutions and protein concentration identical to those used to grow original hits

Streak Seeding - Practical Setup Astbury Centre, University of Leeds Start with conditions that originally grew original hits Setup a whole tray of drops – two drops per well The numbers indicate drop pre-incubation times in hours At the predefined times manually streak seed into the four drops sequentially

T 7 endonuclease I / Holliday junction complex Astbury Centre, University of Leeds 24 hour pre-incubation + 12 hours growth

T 7 endonuclease I / Holliday junction complex Astbury Centre, University of Leeds 4 hour pre-incubation + 12 hours growth NB 3 hour pre-incubation no crystals!

T 7 endonuclease I / Holliday junction complex Astbury Centre, University of Leeds 4 hour pre-incubation + 72 hours growth 300μm

Astbury Centre, University of Leeds Conclusions • Vapour diffusion using Oryx 6 + silicone fluid works exceptionally well • Most ‘hits’ can be easily optimised and scaled-up using Equilbration rate diffusion or Seeding

Astbury Centre, University of Leeds

Astbury Centre, University of Leeds Endonuclease I Holliday Junction Complex Structure, 3. 1Å To be Published in Nature 4 th October 2007

Acknowledgements Astbury Centre, University of Leeds and Dundee Anne-Cécile Déclais Steve Carr David Lilley Synchrotrons ESRF Grenoble SRS Daresbury Support Wellcome Trust Simon Phillips Cancer Research UK BBSRC

Astbury Centre, University of Leeds What is silicone oil / fluid? CH 3 -Si-O -Si-CH 3 n. CH 3 Polydimethylsiloxane Defined by viscosity measured in centistokes The viscosity / volatility can be varied by changing n

Unequal arm Length The Approach Astbury Centre, University of Leeds Arm H Oligo Z 5’ • Many DNA variations – cater for up to 100 G C T 25 A T G C A T C G G C T A A T G C C G A T T A G C C C 25 G 5 C • Started project about 8 years ago G A 5 Hairpin loops C C • 960 condition screen T G Cut A A • Use Vapour diffusion T 1 G A T C G 10 • Finance for robotics about £ 30, 000 29 Arm B T 15 T 20 20 T T T 15 Arm R T Oligo Z 3’ T 29 10 A 1 Oligo Y 5’ Arm X Sequence Cut

Astbury Centre, University of Leeds Seeding Higher-throughput seeding • About 3/4 of targets optimised by setting up one tray! • 3 -4 hours pre-incubation often best QUICK, EASY, FAST