Crystallization By Surface Engineering Methodology And Application

Crystallization By Surface Crystallization By Surface EngineeringEngineering

Methodology And ApplicationMethodology And ApplicationDavid CooperDavid Cooper

Zygmunt Derewenda Zygmunt Derewenda LaboratoryLaboratory

Aims to promote Aims to promote crystallization by altering crystallization by altering surface features that inhibit surface features that inhibit crystallization crystallization

Lysines and glutamates are Lysines and glutamates are primarily responsible for the primarily responsible for the “entropy shield”“entropy shield”

Candidate Proteins:Candidate Proteins: Soluble and Purify WellSoluble and Purify Well Difficult to crystallize or Difficult to crystallize or

diffract poorlydiffract poorly Contain a cluster of Contain a cluster of

highly-entropic residueshighly-entropic residues

Surface Entropy ReductionSurface Entropy Reduction

Lysine GlutamateRotamers Rotamers

GGcrystcryst = = HHcrystcryst – T ( – T (SSproteinprotein + + SSsolventsolvent))crystcryst

Previous Successes w/ Previous Successes w/ RhoGDIRhoGDI

Meets all SER criteriaMeets all SER criteria Rich in lysines 10.1% and glutamates 7.9% Rich in lysines 10.1% and glutamates 7.9%

average incidence of 7.2% and 3.7%, respectivelyaverage incidence of 7.2% and 3.7%, respectively It took years to get a poorly diffracting wild-type It took years to get a poorly diffracting wild-type

crystalcrystal

Other Successes in Our LabOther Successes in Our Lab The RGSL domain of PDZRhoGEFThe RGSL domain of PDZRhoGEF

Longenecker KL, et al. & Derewenda ZS. Longenecker KL, et al. & Derewenda ZS. Structure Structure (2001) 9:559- (2001) 9:559-6969

The LcrV antigen of the plague-causing bacterium The LcrV antigen of the plague-causing bacterium Yersinia pestisYersinia pestisDerewenda, U. et al. & Waugh, D.S. Derewenda, U. et al. & Waugh, D.S. StructureStructure (2001) 9:559-69 (2001) 9:559-69

Product of the Product of the YkoFYkoF B. subtilisB. subtilis gene gene Devedjiev, Y. et al. & Derewenda, Z.S. Devedjiev, Y. et al. & Derewenda, Z.S. J Mol BiolJ Mol Biol (2004) 343:395- (2004) 343:395-

406406 Product of the Product of the YdeNYdeN B. subtilisB. subtilis gene gene

Janda, I. et al. & Derewenda, Z.S. Janda, I. et al. & Derewenda, Z.S. Acta CrystActa Cryst (2004) D60: 1101- (2004) D60: 1101-11071107

Product of the Product of the Hsp33Hsp33 B. subtilisB. subtilis gene gene Janda, I. et al. & Derewenda, Z.S. Janda, I. et al. & Derewenda, Z.S. StructureStructure (2004) 12:1901-1907 (2004) 12:1901-1907

The product of the The product of the YkuDYkuD B. subtilisB. subtilis gene gene Bielnicki, J. et al. & Derewenda, Z.S. Bielnicki, J. et al. & Derewenda, Z.S. ProteinsProteins (2006) 1:144-51 (2006) 1:144-51

Human Doublecortin N-terminal domainHuman Doublecortin N-terminal domainCierpicki, T. Cierpicki, T. et alet al, & , & Derewenda, Z.S. Derewenda, Z.S. ProteinsProteins (2006) 1:874-82 (2006) 1:874-82

The Ohr protein of The Ohr protein of B. subtilisB. subtilisCooper, D. et al. & Derewenda, Z.S. in preparationCooper, D. et al. & Derewenda, Z.S. in preparation

Human NudC C-terminal domainHuman NudC C-terminal domainZheng, M. Zheng, M. et alet al. & Derewenda, Z.S. in preparation. & Derewenda, Z.S. in preparation

APC1446 -- Crystals diffracting to 3.0 APC1446 -- Crystals diffracting to 3.0 Å, but unsolved.Å, but unsolved.

**MCSG Targets****MCSG Targets**

Crystallization often involves Crystallization often involves mutation clustersmutation clusters

RGSL domain Of PDZ-RhoGEF

Ykof LcrV

Hsp33

The recurrence of crystal contacts involving the mutations seems to validate the hypothesis that crystallization is facilitated by surface entropy reduction.

Publications by other groups reporting crystallization of novel Publications by other groups reporting crystallization of novel proteins (proteins (green)green), or preparations of higher quality crystal forms , or preparations of higher quality crystal forms

((greygrey) of proteins previously crystallized, by the surface ) of proteins previously crystallized, by the surface engineering approach as of Sept 2006engineering approach as of Sept 2006

The CUE:ubiquitin complexThe CUE:ubiquitin complex Prag G et al., & Hurley JH, Cell (2003) 113:609-20

Unactivated insulin-like growth factor-1 receptor kinaseUnactivated insulin-like growth factor-1 receptor kinaseMunshi, S. et al. & Kuo, L.C. Acta Cryst (2003) D59:1725-1730

Human choline acetyltransferaseHuman choline acetyltransferaseKim, A-R., et al. & Shilton, B. H. Acta Cryst (2005) D61, 1306-1310

Activated factor XI in complex with benzamidineActivated factor XI in complex with benzamidineJin, L., et al. & Strickler, J.E. Acta Cryst (2005) D61:1418-1425

Axon guidance protein MICALAxon guidance protein MICALNadella, M., et al. & Amzel, M.L. PNAS (2005) 102:16830-16835

Functionally intact Hsc70 chaperoneFunctionally intact Hsc70 chaperoneJiang, J., et al. & Sousa, R. Molecular Cell (2005) 20:513-524

L-rhamnulose kinase from E. coliL-rhamnulose kinase from E. coliGrueninger D, & Schultz, G.E. J Mol Biol (2006) 359:787-797

T4 vertex gp24 protein T4 vertex gp24 protein Boeshans, K.M.., et al. & Ahvazi, B. Protein Expr Purif (2006) 49:235-43

Borrelia burgdorferi outer surface protein ABorrelia burgdorferi outer surface protein AMakabe, K., et al. & Koide, S. Protein Science, (2006) 15:1907-1914

SH2 domain from the SH2-B murine adapter proteinSH2 domain from the SH2-B murine adapter proteinHu, J., & Hubbard, S.R J Mol Biol, (2006) 361:69-79

Mycoplasma arthriditisMycoplasma arthriditis-derived mitogen-derived mitogenGuo, Y., et al., & Li, H. J., Acta Cryst (2006) F62:238-241

•3 Mutations: E583A, E584A, W593H3 Mutations: E583A, E584A, W593H•2 Structures2 Structures

•2HDV -- SH2 domain of SH2-B2HDV -- SH2 domain of SH2-B2.0 2.0 Å resolution with Å resolution with 2 copies in the ASU2 copies in the ASU

•2HDX – SH2 domain of SH2-B with Jak2 2HDX – SH2 domain of SH2-B with Jak2 PhosphopeptidePhosphopeptide

2.3 2.3 Å resolution with Å resolution with 6 copies in the ASU6 copies in the ASU

““Crystals of wild-type SH2-BSH2, either unliganded or Crystals of wild-type SH2-BSH2, either unliganded or liganded, were not obtained.”liganded, were not obtained.”

2HDV – Unliganded SH2-B2HDV – Unliganded SH2-BE563A, E564A, W573HE563A, E564A, W573H

2HDX – SH2-B with JAK2pY8132HDX – SH2-B with JAK2pY813

Purpose of this studyPurpose of this study Evaluate effect of various target residuesEvaluate effect of various target residues

Alanine, Histidine, Serine, Threonine, TyrosineAlanine, Histidine, Serine, Threonine, Tyrosine Learn how to select sitesLearn how to select sites Evaluate Alternative Reservoir ScreeningEvaluate Alternative Reservoir Screening

Our Screening ProcessOur Screening Process Protein concentration ~ 15 mg/ml Protein concentration ~ 15 mg/ml Standard Screen Standard Screen

Drops of Super Screen reagent + proteinDrops of Super Screen reagent + proteinOur Super Screen is very similar to JCSG+ Our Super Screen is very similar to JCSG+

Reservoir is 100 Reservoir is 100 l of Super Screen reagentl of Super Screen reagent

““Salt” Screen Salt” Screen Drops of Super Screen reagent + proteinDrops of Super Screen reagent + protein Reservoir is 100 Reservoir is 100 l of 1.5 M NaCll of 1.5 M NaCl

The Wild-Type RhoGDI The Wild-Type RhoGDI Failed to crystallize in the Standard Screen Failed to crystallize in the Standard Screen 1 hit in the Salt screen1 hit in the Salt screen

The Mutation ClustersThe Mutation Clusters

A B C D

E F G H I

Discontinued:Discontinued:Solubility Solubility ProblemsProblems

Ala His Ser Thr Tyr Totals

Mutations commonly form crystal Mutations commonly form crystal contactscontacts

AS

CT

DY

FA

FH

Other Target Residues?Other Target Residues? Phospholipase APhospholipase A22

Not intentionally a SER structureNot intentionally a SER structure

““Biochemical studies of bovine pancreatic Biochemical studies of bovine pancreatic PLA2 (Rogers et al., 1998; Yu et al., 2000) PLA2 (Rogers et al., 1998; Yu et al., 2000) suggested that the lysine-to-methionine suggested that the lysine-to-methionine substitution of the residues 53, 56, 120 substitution of the residues 53, 56, 120 and 121 eliminates the anionic interface and 121 eliminates the anionic interface preference of the wild type.”preference of the wild type.”

K53M,K53M, K56M, K120M, and K121M K56M, K120M, and K121M Improvement of resolution (1.5Improvement of resolution (1.5Å to 1.1 Å)Å to 1.1 Å)

Suggests that SER could alsoSuggests that SER could also

be used to introduce be used to introduce

methionines for phasingmethionines for phasing

purposes.purposes.

K Sekar, et al & M.-D Tsai. Acta Cryst D62:717

Subtle effects of the Subtle effects of the mutationsmutations

This trimer of dimers This trimer of dimers occurs with as little as occurs with as little as 2 and as many as 6 2 and as many as 6 monomers in the ASUmonomers in the ASU

EA1R32 a=b=129.6 c=166.6

EA2C2 a=132.2 b=131.4 c=92.5

This packing is This packing is seen with one seen with one or two or two monomers in monomers in the ASU. the ASU. Notice the Notice the same dimer. same dimer.

CYP3121a=b=77.3 c=171.7

FHP3221a=b=75.2 c=91.6

Observations from this Observations from this studystudy Alanines, Tyrosine and Threonines are good Alanines, Tyrosine and Threonines are good

choices for the target residue.choices for the target residue. Performing traditional and alternative Performing traditional and alternative

reservoir screening greatly increases the reservoir screening greatly increases the chances of getting a hit and greatly increases chances of getting a hit and greatly increases the number of conditions that give hits.the number of conditions that give hits.

The as yet unexplainable preference for a The as yet unexplainable preference for a particular target residue for some mutation particular target residue for some mutation clusters suggests that multiple target residues clusters suggests that multiple target residues should be tried for each cluster.should be tried for each cluster.

Recommendation: Before pursuing mutations, Recommendation: Before pursuing mutations, try alternative reservoir screening. If that try alternative reservoir screening. If that fails, use the server to pick a mutation site fails, use the server to pick a mutation site and use 2 or more target residues.and use 2 or more target residues.

A Success from Screening A Success from Screening AloneAlone

A MCSG abandoned target.Wild-type crystallized only in the salt screen!

Solve / Resolve_build did the trick!

Few Sites are Predicted / Few Sites are Predicted / NeededNeeded

OhrB

Hsp33

Few Sites are Predicted / Few Sites are Predicted / NeededNeeded

Yden

YkoF

Because many proteins have very few potential sites, trying several target residues for one cluster may be the only option.

There are some exceptions . There are some exceptions . . .. .

Like RhoGDI.Like RhoGDI. IA B D/C EF GIA B D/C EF G H H

In this case, you could either use either multiple target residues, multiple clusters or a combination.

Current work:Current work:Server ValidationServer Validation

Test Top Hits for RhoGDI and OthersTest Top Hits for RhoGDI and Others K98,K99,Q100 (no longer the top hit)K98,K99,Q100 (no longer the top hit)

Threonine – 6 standard – 4 salt – 7 uniqueThreonine – 6 standard – 4 salt – 7 unique Tyrosine – No hits.Tyrosine – No hits.

Need to test current tops hits.Need to test current tops hits. New Targets:New Targets:

Selected MCSG targets that could be purified, Selected MCSG targets that could be purified, but failed to crystallize.but failed to crystallize.

Eliminated proteins with close homologues Eliminated proteins with close homologues solvedsolved

Submitted sequences to the server and Submitted sequences to the server and requested those with the highest hits.requested those with the highest hits.

TargetTarget ProteinProtein ResidueResiduess

MW MW

kDAkDA

APC1087Hypothetical UPF0087 Protein Ydzf.SERp Score:4.7

109109 12.12.66

APC1126Yfnk Yetn Protein.SERp Score:5.11

356356 41.41.44

APC1319Putative Fibronectin-binding Protein Yloa Protein.SERp Score:5.36

572572 65.65.55

APC1628Histidinol-phosphatase Holpase.SERp Score:4.46

268268 30.30.55

APC1661Ytaa Protein. Possible Adp-heptose:lps HeptosyltransferaseSERp Score:5.32

357357 41.41.22

APC22720Hypothetical Cytosolic Protein SERp Score:5.94

370370 40.40.55

APC22734Periplasmic Component Of Efflux System SERp Score:4.09

361361 38.38.99

APC22756Signal Transduction Protein TRAP SERp Score:4.12

165165 18.18.99

APC22763Hypothetical Protein BC_1106SERp Score:5.07

154154 17.17.99

APC22990Enterochelin Synthetase Component D SERp Score:6

234234 25.25.88

Extending the Extending the MethodMethod

Multi-domain proteins.Multi-domain proteins.

2CGJ

2CGK

Red and Green denote domains.

Another Multi-domain Example Another Multi-domain Example

K141A, K142AK141A, K142A

For multi-domain proteins of unknown structure, using multiple sites (instead of multiple target residues) may help avoid disrupting domain interfaces.

Extending the Extending the MethodMethod

Mutating Multiple ClustersMutating Multiple Clusters First MutationsFirst Mutations

K48A, K60A, K83A,K196AK48A, K60A, K83A,K196ADidn’t workDidn’t work

Added Added E37S, E45S, K46S,K64S, E37S, E45S, K46S,K64S, E104S, K107S, K239S, E104S, K107S, K239S, E240S, and K254S E240S, and K254S

Too many alanine mutations can decrease Too many alanine mutations can decrease solubility, so using a hydrophilic target residue is solubility, so using a hydrophilic target residue is wise. wise.

Yeah, ISFI

AcknowledgementsAcknowledgements Tomasz BoczekTomasz Boczek Kasia GrelewskaKasia Grelewska Gosia PinkowskaGosia Pinkowska Gosia SikorskaGosia Sikorska Michal ZawadzkiMichal Zawadzki Luki GoldschmidtLuki Goldschmidt David EisenbergDavid Eisenberg Zygmunt DerewendaZygmunt Derewenda