Structural insight into the specificity of the B3DNA-binding domains provided by the co-crystalstructure of the C-terminal fragment of BfiIrestriction enzymeDmitrij Golovenko Elena Manakova Linas Zakrys Mindaugas Zaremba

Giedrius Sasnauskas Saulius Grazulis and Virginijus Siksnys

Department of ProteinndashDNA Interactions Institute of Biotechnology Vilnius University Graici uno 8 LT-02241Vilnius Lithuania

Received November 15 2013 Revised and Accepted December 11 2013

ABSTRACT

The B3 DNA-binding domains (DBDs) of plant tran-scription factors (TF) and DBDs of EcoRII and BfiIrestriction endonucleases (EcoRII-N and BfiI-C)share a common structural fold classified as theDNA-binding pseudobarrel The B3 DBDs in theplant TFs recognize a diverse set of target se-quences The only available co-crystal structure ofthe B3-like DBD is that of EcoRII-N (recognitionsequence 50-CCTGG-30) In order to understand thestructural and molecular mechanisms of specificityof B3 DBDs we have solved the crystal structure ofBfiI-C (recognition sequence 50-ACTGGG-30) com-plexed with 12-bp cognate oligoduplex Structuralcomparison of BfiI-CndashDNA and EcoRII-NndashDNAcomplexes reveals a conserved DNA-binding modeand a conserved pattern of interactions with thephosphodiester backbone The determinants of thetarget specificity are located in the loops thatemanate from the conserved structural core TheBfiI-CndashDNA structure presented here expands arange of templates for modeling of the DNA-boundcomplexes of the B3 family of plant TFs

INTRODUCTION

Structural studies of plant transcription factors (TFs) andtype II restriction endonucleases (REases) revealed an un-expected link between these seemingly unrelated proteinfamilies Despite the lack of sequence similarity the B3DNA-binding domain (DBD) of plant TFs (12) the

effector domain of EcoRII REase [EcoRII-N (3)] andthe DBD of BfiI REase [BfiI-C (4)] share a commonfold classified as the DNA-binding pseudobarrel (SCOPnumber 101935) The conserved core of these proteinsreferred here as lsquoB3-like domainsrsquo is comprised of abarrel-shaped 7-stranded b-sheet capped by a-helices atboth ends (56) (Figure 1) The genes encoding B3-likedomains are widespread in plant kingdom from greenalgae to flowering plants For example in Arabidopsisalone there are 118 B3 domain containing proteins thatoften are involved in the hormone response pathways (1)An atomic view of DNA recognition by the B3-like

proteins was provided by the crystal structure ofEcoRII-NndashDNA complex (7) In EcoRII-N the b2 b5and b4 b-strands located at one edge of the pseudobarrelthe a-helix a2 and the connecting loops make a wrench-like cleft that approaches DNA from the major grooveside (7) The DNA-binding interface contains two lsquorecog-nition armsrsquo which interact with different parts of the50-CCTGG-30 target site the N-arm (strand b2 and helixa2) contacts the 50-terminal part of the recognitionsequence while the C-arm (strands b5 and b4) interactswith the 30-terminal part of the recognition site Structuralcomparison of other B3-like domains in the DNA-freeform indicates a conserved structural core and a wrench-like DNA-binding cleft moreover a number of positivelycharged EcoRII residues involved in DNA-backboneinteractions are conserved in this cleft (47) Takentogether these findings suggest that the DNA-bindingmode identified in EcoRII-NndashDNA crystal structure isconserved in other B3-like domains (7) Analysis of theDNA-binding interface of the Arabidopsis thalianaB3-family TFs RAV1 (8) and VRN1 (9) supports thishypothesis

To whom correspondence should be addressed Tel +370 5260 2108 Fax +370 5260 2116 Email siksnysibtltPresent addressesDmitrij Golovenko Department of Structural Biology The Weizmann Institute of Science 76100 Rehovot IsraelLinas Zakrys Thermo Fisher Scientific Graici uno 8 LT-02241 Vilnius Lithuania

Published online 13 January 2014 Nucleic Acids Research 2014 Vol 42 No 6 4113ndash4122doi101093nargkt1368

The Author(s) 2014 Published by Oxford University PressThis is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (httpcreativecommonsorglicensesby-nc30) which permits non-commercial re-use distribution and reproduction in any medium provided the original work is properly cited For commercialre-use please contact journalspermissionsoupcom

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

The B3-like domains display remarkable plasticityin terms of the target specificity they interact withnon-specific pentanucleotide and hexanucleotide DNAsequences (Table 1) The crystal structure of EcoRII-Nprovided a structural mechanism for the pentanucleotidesequence recognition however it remained to be estab-lished how B3-like domains adapt a conserved structuralfold for the recognition of different DNA sites In order tounderstand the structural and molecular mechanisms ofspecificity of B3 DBDs we have solved the crystal struc-ture of the C-terminal domain (BfiI-C) of BfiI REasefrom Bacillus firmus complexed with a 12-bp cognateoligoduplex BfiI-C belongs to the B3-like domain familyand interacts with the hexanucleotide sequence50-ACTGGG-30 which partially overlaps with theEcoRII-N site 50-CCTGG-30 (the overlapping base pairsare underlined) Comparison of BfiI-CndashDNA and EcoRII-NndashDNA complexes revealed a structurally conservedpattern of phosphodiester backbone contacts andconserved amino acid residues in the DNA-binding cleftTaken together the BfiI-CndashDNA structure presented herereveals how a conserved structural core is adapted for therecognition of variable DNA sequences and expandsthe template range for modeling of the DNA-boundcomplexes of the B3 family of plant TFs

MATERIALS AND METHODS

DNA oligoduplexes

HPLC-grade oligonucleotides were purchased fromMetabion (Martinsried Germany) The oligoduplex 1212(topbottom strand sequences 50-AGCACTGGGTCG-3030-TCGTGACCCAGC-50 respectively the BfiI siteunderlined) for generation of the minimal BfiI-CndashDNAcomplex was assembled as described in (7) Prior to anneal-ing the bottom strands of oligoduplexes 16SP (50-AGCGTAGCACTGGGCT-3030-TCGCATCGTGACCCGA-50)and 16NS (50-AGCGTAGCCCGGGGCT-3030-TCGCAT

CGGGCCCCGA-50) used for DNA-binding experimentswere radiolabeled using [g-33P]ATP (Hartmann AnalyticBraunschweig Germany) and T4 polynucleotide kinase(Thermo Fisher Scientific Vilnius Lithuania)

Preparation of the BfiI-C domain

BfiI mutant K107A was expressed in Escherichia colistrain ER2566 carrying plasmids pET21b-BfiIR65-K107A and pBfiIM91 and purified as described previ-ously (15) Protein concentration was determined bymeasuring the absorbance at 280 nm and are expressedin terms of dimer using the extinction coefficient of99700Mcm (calculated using the ProtParam toolhttpwebexpasyorgprotparam) The BfiI-CndashDNAcomplex was obtained by limited proteolysis of the full-length BfiIndashDNA complex The full-length K107A BfiImutant was mixed with oligoduplex 1212 at the molarprotein dimerDNA ratio 122 [the stoichiometry of theBfiI-DNA complex is 12 (15)] in a buffer containing10mM TrisndashHCl (pH 75 at 25C) 200mM KCl 1mMDTT 2mM calcium acetate and 10 glycerolThermolysin was added at 1100 ww proteaseproteinratio Limited proteolysis was performed at 37C for20 h and reaction terminated as described in (14) Thereaction mixture was diluted two-fold with 10mM TrisndashHCl (pH 75 at 25C) to reduce the salt concentrationand loaded onto a 1-ml Heparin Sepharose column

Figure 1 B3 and B3-like DBDs of the pseudobarrel fold (SCOP number 101935) (A) Apo DBD (BfiI-C) of the BfiI restriction enzyme [residues193ndash358 of B chain PDB ID 2C1L (4)] (B) The effector domain of the EcoRII restriction enzyme (EcoRII-N) in the DNA-bound form [PDB ID3HQF (7)] (C) The NMR structure of the RAV1-B3 domain [model 1 in PDB ID 1WID (5)] Common structural core made of 7 b-strands iscolored in magenta DNA backbone in (B) is depicted as a black double-helix

Table 1 B3 and B3-like proteins and their recognition sequences

Protein Recognition site Reference

VRN1-B3 non-specific (9)RAV1-B3 50-CACCTG-30 (10)ARF1-B3 50-TGTCTC-30 (11)ABI3-B3 50-CATGCA-30 (12)EcoRII-N 50-CCTGG-30 (713)BfiI-C 50-ACTGGG-30 (14)

4114 Nucleic Acids Research 2014 Vol 42 No 6

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

(GE Healthcare) The BfiI-CndashDNA complex was collectedas the flowthrough fraction and was subsequently loadedonto Mono Q 550 GL column (Amersham PharmaciaBiotech) and eluted using KCl gradient in 20mM TrisndashHCl (pH 75 at 25C) 1mM EDTA and 10 glycerol Allpurification steps were carried out at room temperatureFractions containing the BfiI-CndashDNA complex weredialyzed against 10mM TrisndashHCl (pH 75 at 25C)100mM KCl 1mM DTT 10 glycerol and storedat 4C Concentration of BfiI-C in the complex wasdetermined by Bradford assay using the apo BfiI-Cdomain that was purified previously (14) as a reference

Crystallization data collection andstructure determination

BfiI-C complex with oligoduplex 1212 was concentratedto 17mgml using a Centricon concentrator (Millipore)with 3 kDa MW cut-off Crystals were grown by thesitting drop vapor diffusion method at 19C by mixing05 ml of the complex solution with 05 ml of the crystalliza-tion buffer 18 of the lsquoIndexrsquo screen (Hampton Research)which contained 049M NaH2PO4 091M K2HPO4 (pH69 at 25C) A crystal belonging to the spacegroup P65appeared after 3 years Prior to flash-cooling the crystalwas transferred into a 31 mixture of the reservoir bufferwith glycerol X-ray diffraction data were collected usingan in-house Rigaku RU-H3R rotating anode generatorImages were processed using MOSFLM (1617) andSCALA (18) Initial phases were obtained by the molecu-lar replacement using BfiI-C residues 200ndash347 from thefull-length apo BfiI structure [PDB ID 2C1L (4)]Structure was initially refined using REFMAC (19) thefinal refinement step was performed with PHENIX (20)COOT (21) was used for a model inspection Data collec-tion and refinement statistics are presented in Table 2

After molecular replacement several DNA phosphateswere immediately visible in the electron density map Afterimprovement of the protein model we were able to buildall 12 bp of the duplex [atomic DNA coordinates weregenerated by 3D-DART (22)] However it was still notpossible to distinguish between purines and pyrimidinesand to determine the absolute orientation of DNA inthe DNA-binding cleft (the asymmetric nature of the rec-ognition sequence rules out overlapping of alternativeDNA orientations) To solve this problem we producedtwo models of BfiI-C with alternative DNA orientationsBoth models were subjected to 20 cycles of positional re-finement in REFMAC (19) and the model having betterrefinement statistics (Rfree=026 versus Rfree=032 in thealternative model) was selected as the one having thecorrect duplex orientation The final structure refined toan Rfree=0226 (Table 2) has a clear electron density foreach recognition sequence base pair and PuPy bases areeasily distinguishable in both duplexes within the crystal-lographic asymmetric unit (Supplementary Figure S1)

Coordinates and structure factors are deposited underPDB ID 3ZI5 DNA geometry in BfiI-CndashDNA structurewas analysed with CURVES (23) The contact surfacesburied between the two molecules were calculated usingNACCESS (24) ProteinndashDNA contacts were analysed by

NUCPLOT (25) Graphical representations wereproduced with MolScript (2627)

BfiI mutants

His-tagged full-length BfiI mutant variants were generatedby QuickChange (28) or lsquomegaprimerrsquo methods (29) usingthe plasmid pET15b-BfiIR65 as the template for PCRThe E coli strain ER2267 carrying the plasmidpBfiIM91 was used as a transformation host The muta-tions were confirmed by DNA sequencing of the entiregene The proteins were purified using Ni2+-chargedHiTrap HP and HiTrap Heparin HP columns (GEHealthcare) as described in (15) to gt90 (50 for theY227A) homogeneity as estimated by SDS-PAGEConcentrations of the WT and mutant proteins weredetermined as described above using extinction coeffi-cients for each mutant calculated by the ProtParam tool(httpwebexpasyorgprotparam) and are expressed interms of dimer The K340A mutant was not isolated dueto a very low expression level

Electrophoretic mobility shift assay

DNA binding by the WT BfiI and mutants was analysedby the electrophoretic mobility shift assay (EMSA) usingthe 16-bp specific and non-specific DNA duplexes 16SPand 16NS respectively DNA (final concentration 1 nM)was incubated with BfiI (final concentrations varied from1 to 1000 nM of dimer) for 10min in 20 ml of the bindingbuffer containing 40mM Trisndashacetate (pH 83 at 25C)

Table 2 Diffraction data and structure refinement statistics

Data collectiona

Spacegroup P65Unit cell a=17518 b=17518

c=3579 A= =900 =120

Resolution A (final shell) 5057ndash32 (337ndash32)Reflections unique (total) 10809 (75356)Completeness () overall (final shell) 100 (100)II overall (final shell) 133 (72)Rmerge overall (final shell) b 0147 (0240)B(iso) from Wilson (A2) 244

Refinementc

Resolution range (A) 438ndash32Number of protein atoms 2642Number of DNA atoms 972Rcryst (Rfree) 0176 (0226)RMS bonds (A)angles () 00030743Average B factors (A2) total 491Main chain 489Side chains 496Solvent 288

Ramachandran plotFavored 9604Allowed 396Outliers 000

aDataset was collected at 100KbRmerge frac14

Ph

Pnhifrac141 Ihh i Ihieth THORN

2=P

h

Pnhifrac141 I

2hi where Ihi is an intensity

value of i-th measurement of reflection h h frac14 hkleth THORN sum h runsover all measured reflections and Ihh i is an average measured inten-sity of the reflection h Number nh is a number of measurements ofreflection hcTest set size 96

Nucleic Acids Research 2014 Vol 42 No 6 4115

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

01mM EDTA 01mgml BSA and 10 glycerol at roomtemperature Free DNA and proteinndashDNA complexeswere separated by electrophoresis using 6ndash8 acryl-amide gels (291 acrylamidebisacrylamide in 40mMTrisndashacetate pH 83 at 25C and 01mM EDTA)Radiolabeled DNA was detected and quantified usingthe Cyclone phosphorimager and the OptiQuantsoftware (Packard Instrument) The association constantKA (KA= 1KD where KD is the dissociation constant)values were calculated as described previously (13)

Activity measurements

DNA cleavage activity of WT BfiI and mutants wasmeasured by incubating different amounts of purifiedproteins (varied from 2 to 1000 nM) with 1 mg of DNAin 50 ml of the Y+TangoTM buffer (Thermo FisherScientific) for 1 h at 37C and analysing the DNAcleavage products by agarose gel electrophoresis Oneunit of BfiI is defined as the smallest amount of proteinthat completely digests DNA The specific activity ofWT BfiI equals 100 000 unitsmg protein To demonstratethat the decrease of catalytic activity results only from thepoint mutations in the C-domain all mutant proteinswere subjected to limited proteolysis with thermolysinwhich liberates the protease-resistant catalyticN-terminal domain (14) the resultant N-terminaldomains in all cases demonstrated the expectednon-specific DNA cleavage

RESULTS

Overall structure of the BfiI-CndashDNA complex

The asymmetric unit of the crystal contains two BfiI-CndashDNA complexes (Supplementary Figure S2A)Despite the relatively large contact surface between theprotein chains (amp1000A2) this interaction seems to be bio-logically irrelevant since isolated BfiI-C is a monomerthat binds only one DNA copy (14) Both monomers inthe asymmetric unit are nearly identical and can besuperimposed with an RMSD of 013A including proteinside chains The DNA bound in the complex is BndashDNAform In the crystal the DNA oligoduplexes form a con-tinuous left-handed pseudohelix due to the stacking inter-actions between the symmetry-related oligoduplexes(Supplementary Figure S2B) The secondary structure ofthe DNA-bound BfiI-C domain is similar to that of apo-BfiI-C (PDB ID 2C1L residues 193ndash358) with the onlyexception of the 310 helix (residues 232ndash237) that ismissing in the apo-form

DNA recognition by BfiI-C

The wrench-like DNA-binding cleft of BfiI-C is similar tothat of EcoRII-N (7) (Figures 1A and B) The N-arm(residues 210ndash231) in the BfiI-C binding cleft includesa-helix a6 b-strand b11 and the connecting loop whilethe C-arm (residues 266ndash287) is comprised of anti-parallelb14-15 b-strands and the connecting loop (Figure 2A) Incontrast to EcoRII-N DNA contacts made by the BfiI-Cextend beyond the N- and C-arms and include additional

structural elements namely the N-loop connectingb-strands b12 and b13 (residues 245ndash252 Figure 2A)and the C-loop connecting the b-strands b18 and b19(residues 335-341 Figure 2A)

BfiI-C approaches DNA from the major groove andmakes an extensive set of contacts with the DNAbackbone (Supplementary Table S1 and Figure S3A)More specifically a number of residues in the N-loopand the R272 residue in the C-arm make contacts to thephosphates on the top DNA strand of the target site whilethe residues in the C-loop and within (or in the vicinity) ofthe N-arm interact with the phosphates on the bottomstrand

The sequence-specific BfiI-CndashDNA interactions areprovided by the amino acid residues located in theN- and C-arms Residues in the N-arm are involved inhydrogen bonding interactions with the bases at the50-end of the site (green box in Figure 2B) whereas theresidues in the C-arm make specific contacts with the basesat the 30-end of the site (orange box in Figure 2B) In totalBfiI-C employs nine amino acid residues to make H-bondsand vdW contacts to the bases of the recognition sitein the major groove (Figure 2C) and uses a sole R247residue located in the N-loop for minor grooveinteractions

Comparison of apo- and DNA-bound forms of BfiI-C

The conformational changes of BfiI-C occurring uponDNA binding are limited to the N-arm N-loop andC-arm regions (Supplementary Figure S4) The loop inthe N-arm (residues 216ndash223) moves towards the DNAbackbone and brings the DNA-recognition residue T225closer to the first two base pairs of the recognition site(Figure 2C) Upon DNA-binding unstructured region(amino acid residues 232ndash237) folds into a 310 helix andthe a-helix a6 (residues 208ndash215) moves closer to the DNAbackbone implying a favorable dipole interaction(Supplementary Figure S4)

The b-strands b14-15 of the C-arm move in respect tothe b-barrel core by pulling away from the b18 strandThis conformational change brings the N280 residue5A closer to the DNA bases and enables a base-specific contact to the penultimate C base Significant con-formational change upon DNA binding occurs also in theN-loop (residues 245ndash252) which is shifted towards DNAalong with the C-arm (Supplementary Figure S4) As aresult the R247 residue moves 75A towards DNAmaking a H-bond to the 50-terminal adenine from theminor groove side (Figure 2C)

Mutational analysis of DNA-contacting residues

BfiI-CndashDNA co-crystal structure was solved at therelatively low 32A resolution therefore DNA-bindingresidues predicted by the structural analysis were sub-jected to the site-directed mutagenesis to verify their func-tional importance (Table 3) Three sets of amino acidresidues were analysed (i) residues that make directH-bonds or non-polar contacts to the bases of the recog-nition site in the major and the minor grooves (Figure 2C)(ii) positively charged amino acid residues and N245 that

4116 Nucleic Acids Research 2014 Vol 42 No 6

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

interact with the DNA phosphates (SupplementaryTable S1) (iii) residues T252 and Q226 that are conservedin the BfiI family enzymes (Supplementary Figure S5) andoverlap with EcoRII-N residues Q37 and N61 involved inDNA binding The selected residues were replaced byalanine in the context of full-length BfiI mutant proteinswere isolated and DNA binding and cleavage abilities

evaluated by EMSA (representative data provided inSupplementary Figure S6) and DNA cleavage assayrespectively Mutations with respect to their effect onDNA binding and cleavage clustered into three groups(i) lsquounimportantrsquo mutations that lead to decrease of thecognate DNA-binding affinity (KA) or specific cleavageactivity up to 10-fold (ii) lsquoimportantrsquo mutations that

Figure 2 DNA recognition by BfiI-C and EcoRII-N (A) The view of the BfiI-CndashDNA complex along the long DNA axis (left) and the side view(right) The DNA-recognition site is colored dark grey The secondary structure elements of the N-arm (a-helix a6 b-strand b11) and the C-arm(b-strands b14 and b15) are colored green and orange respectively Spheres of the matching colors represent the Ca atoms of the DNA-recognitionresidues from the N- and C-arms The additional DNA-recognition element N-loop (residues 245ndash252) is colored blue and the C-loop (residues335ndash341) is red A region of the top DNA strand (nucleotides A4-G7) and adjacent recognition residues are shown against their mFO-DFC

SIGMAA-weighted-electron density contoured at 20 s level (B) The sequence and numbering of the cognate 1212 oligoduplex used in thisstudy DNA bases that interact with the N- and C-arms are boxed in green and orange respectively (C) Recognition of individual base pairs byBfiI-C Panels for the individual base pairs are arranged following the top strand ACTGGG in the 5030 direction The N- and C-arm residue labelsare colored as in panel (A) (D and E) The sequence and numbering of the cognate EcoRII-N oligoduplex and the recognition of individual basepairs by EcoRII-N [PDB ID 3HQF (7)] The residue labels and boxes encircling EcoRII-N sequence elements are colored as in panels (B and C)

Nucleic Acids Research 2014 Vol 42 No 6 4117

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

lead to decrease of the cognate DNA-binding affinity (KA)or specific cleavage activity by 1ndash2 orders of magnitudeand (iii) lsquoessentialrsquo mutations that lead to decrease of thecognate DNA-binding affinity (KA) or specific cleavageactivity by more than two orders of magnitude (Table 3)Nearly all N- and C-arm residues that make direct

contacts to the DNA bases are either essential or import-ant (Figure 2C and Table 3) and also show a high degreeof conservation in the BfiI family (SupplementaryFigure S5) Two of these residues namely Y227 andD282 are involved in H-bond interactions with the N4

amino groups of cytosines that become methylatedby the BfiI methyltransferases [Figure 2C (30)] Itseems that disruption of these interactions by theN4-methylation or mutations (Y227A or D282A) com-pletely abolishes the BfiI cleavage The only discrepancybetween DNA binding and cleavage data is observedfor the R247A mutation it is essential for binding butunimportant for cleavage Presumably removal of thispositively charged residue destabilized the BfiIndashDNAcomplex to such an extent that it could no longer beresolved in the non-equilibrium EMSA experiment butthe complex was sufficiently long-lived to enable DNAcleavage under the DNA cleavage assay conditionsA number of residues making contacts with the DNAbackbone are also important for BfiI function the mostcritical being the highly conserved C-arm residue R272that makes H-bond to the 5rsquo-pACTGGG-30 phosphate(Figure 3) In contrast the Q223 residue from theN-arm is unimportant for the BfiI function implyingthat it does not make any contact to the methyl groupof the T base in the bottom DNA strand (Figure 2C)

Residues Q226 and T252 though they do have structuralcounterparts in BfiI family enzymes and EcoRII are notimportant for BfiI (Table 3)

DISCUSSION

Close structural similarity between B3 domains of plantTFs and the effector domain of EcoRII REase suggeststhat B3 domains either diverged from a common ancestoror were horizontally transferred into higher plants fromsymbiotic or pathogenic bacteria (1233) Target sites forplant TFs of the B3 superfamily vary both in length andsequence (Table 1) suggesting that the DNA-bindingpseudobarrel fold exhibits a structural plasticity thatenables recognition of different DNA sequences within aconserved structural core The molecular mechanisms ofthe target site specificity of the plant TFs yet has to beestablished since 3D structures of plant B3 domains aresolved in the absence of DNA [PDB IDs 1WID (5) 4I1K(9) 1YEL (6)] The crystal structure of the EcoRII effectordomain bound to the oligoduplex containing a 5-bp targetsequence provided a first structural template for modelingof DNA complexes of plant TFs (79) The BfiI-CndashDNAstructure presented here reveals for the first time the struc-tural mechanism of the B3-like domain adaption for adifferent DNA target and provides a new template forthe modeling of plant TFs

Conserved DNA orientation in the B3-like domains

The fold of B3-like domains is termed a lsquopseudobarrelrsquobecause of the missing connectivity between two strands

Table 3 Mutational analysis of BfiI-C residues

Mutated residue Location Contact with DNA DNA-bindingability ()a

Impact on DNAbinding

Specificactivity ()b

Impact onDNA cleavage

Direct contacts to DNA basesR212 N-arm H-bonds to G8 (bottom) and A7 2 Important lt05 EssentialQ223 N-arm vdW contact to T9 100 Unimportant 100 UnimportantT225 N-arm H-bonds to A4 and G8 (bottom) 10 Important 10 ImportantY227 N-arm H-bond to C5 (top) 5 Important lt05 EssentialW229 N-arm vdW contact to C6 1 Essential 3 ImportantR247 N-loop H-bond to A4 lt04 Essential 20 UnimportantE276 C-arm vdW contact to T6 10 Important 1 EssentialN279 C-arm H-bond to G7 lt02 Essential lt05 EssentialN280 C-arm H-bonds to G8 (top) and C4 lt02 Essential lt05 EssentialD282 C-arm H-bonds to C5 (bottom) and C6 lt02 Essential lt05 EssentialR284 C-arm H-bonds to T6 and G7 lt02 Essential lt1 Essential

Contacts with the DNA phosphatesN245 N-loop 50-ApCTGGG-30 100 Unimportant 20 UnimportantK250 N-loop 50-ACTpGGG-30 2 Important 10 ImportantR272 C-arm 50-pACTGGG-30 1 Essential 1 EssentialR291 Close to C-arm 50-pNACTGGG-30 2 Important 5 ImportantK340c C-loop 30-TGACCpC-50 ndd nd nd nd

Other residues conserved in BfiI-C-related proteinsQ226 N-arm ndash 50 unimportant 20 unimportantT252 N-loop ndash nd nd 100 unimportant

aThe DNA-binding ability is expressed in percent () as the ratio of the proteinndashDNA association constant KA of full-length BfiI mutants relative tothe KA of WT BfiI dimer [(5 108)M]bThe specific activity of BfiI mutants is expressed in percent () relative to the activity of WT BfiIcDue to very low expression level the BfiI K340A mutant could not be purifieddnd not determined

4118 Nucleic Acids Research 2014 Vol 42 No 6

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

Figure 3 Sequence and structure elements involved in proteinndashDNA interactions in B3-like domains (A) Structure-based multiple sequence align-ment of BfiI-C (PDB ID 2C1L chain A) EcoRII-N (PDB ID 3HQF) RAV1-B3 (PDB ID 1WID model 1) VRN1-B3 (PDB ID 4I1K chain A) andAt1g16640-B3 (PDB ID 1YEL model 1) was generated by MultiProt and Staccato (31) Residues and secondary structure elements are numberedaccording to the BfiI-CndashDNA structure BfiI-C DNA-binding elements N-arm N-loop C-arm and C-loop are marked by green blue orange andred stripes respectively The (DE)XR motif of BfiI-C and EcoRII-N responsible for recognition of the 50-TGG-30 trinucleotide is marked by blackboxes and black circles Residues contacting the lsquoclamprsquo phosphates are marked by magenta boxes and asterisks The figure was generated usingESPRIPT (32) (B) Interaction of B3-like domains with DNA BfiI-C (this study) EcoRII-N [PDB ID 3HQF (7)] and RAV1-B3 [PDB ID 1WIDinteractions with DNA according to Yamasaki et al (58)] The bases in the recognition sites are colored yellow orange green blue and red islandsencircle residues from the N-arm C-arm N-loop and C-loop respectively Residues contacting DNA phosphate oxygen atoms are depicted in theproximity of the corresponding phosphates lsquoClamprsquo phosphates are colored in cyan

Nucleic Acids Research 2014 Vol 42 No 6 4119

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

(b10 and b11 in BfiI-C and b1 and b2 in EcoRII-N) in theotherwise barrel-shaped 7-stranded b-sheet which in thecase of BfiI-C and EcoRII-N overlap with a RMSD lt2Afor 75 Ca atoms [calculated using MultiProt (31)] Thelsquoopenrsquo edge of the b-sheet forms the concave wrench-likesurface that fits into the DNA major groove and providesbase-specific contacts The spatial positioning of thelsquopseudobarrelrsquo in respect to the DNA long-helical axisin the BfiI-CndashDNA and EcoRII-NndashDNA structures isconserved (Supplementary Figure S3A and B) suggestinga similar DNA orientation in respect to the protein corefor other B3-like and B3 domains including RAV1-B3[PDB ID 1WID (5)] At1g14660-B3 [PDB ID 1YEL(6)] and VRN1-B3 [PDB ID 4I1K (9)] A recent high-resolution crystal structure and DNA-binding analysisof plant TF VRN1 are consistent with a similar DNA-binding mode (9)

Plasticity of the DNA-binding interface of the B3-likedomains enables recognition of different sequences

The crystal structure of BfiI-C presented here and EcoRII-N structure solved by us earlier (7) provide a uniqueopportunity to compare structural and molecular mechan-isms of sequence recognition by two B3-like proteins inter-acting with target sites of different length and sequenceBfiI-C is specific for the asymmetric 6-bp sequence50-ACTGGG-30 while EcoRII-N recognizes a partiallyoverlapping pseudopalindromic 5-bp sequence50-CCWGG-30 (W=A or T the overlapping bases areunderlined) Most of the sequence specific and DNAbackbone contacts in BfiI-C and EcoRII-N are made byresidues located in the loops that extend from the concaveb-sheet BfiI-C and EcoRII-N share two conserved struc-tural elements called N- and C-arms that make base-specific contacts to DNA in both proteins The N-armsof both proteins contribute amino acid residues for the spe-cific interactions with bases at the 50-half of the recogni-tion site while the residues located in the C-arm contactsthe bases in the 30-half of the target site (Figure 2B and D)Intriguingly BfiI-C and EcoRII-N bind their partiallyoverlapping recognition sites in the same orientationMoreover though BfiI-C and EcoRII-N recognize thefirst CG base in the overlapping region differently bothproteins make a similar set of contacts to the 50-TGG-30

trinucleotide For example both proteins use a structur-ally conservative C-arm arginine (R284 in BfiI-C R98 inEcoRII-N) to make a H-bond to the O4 atom of theT base and a conserved carboxylate (D282 in BfiI-CE96 in EcoRII-N) to make a H-bond to the N4 atom ofthe cytosine in the last GC base pair (Figure 2C and E)Most of the sequence-specific contacts in BfiI are

provided by amino acid residues located in the N- andC-arms Interestingly the N-arm of BfiI-C is 8 and11ndash13 amino acid residues longer than structural equiva-lents in EcoRII-N and other B3-like domains respectively(Figure 3A) Despite of the length differences the N-armof BfiI-C makes a similar number of base-specific contactsas the N-arm of EcoRII-N (Figure 3B) The extra part ofthe loop connecting the a-helix a6 and b-strand b11 in theBfiI-C N-arm points away from DNA and does not

contribute to the interactions with DNA Presumablythe length of the N-arm in B3 domains is also sufficientto secure base-specific contacts to DNA (Figure 3)

BfiI-C also contains additional structural elementsnamely N- and C-loops that are involved in interactionswith DNA Amino acid residues in the N-loop contributeto the base-specific interactions at the 50-end of the tar-get site while amino acid residues in the C-loop makeDNAndashbackbone contacts at the 30-end of the recogni-tion sequence The proteinndashDNA interface area in theBfiI-CndashDNA complex therefore is larger than in theEcoRII-NndashDNA complex [2800A2 versus 2200A2 (7)]The structural equivalents of the N- and C-loops inEcoRII-N and the N-loop in B3 domains from plantTFs are shorter than in BfiI-C and therefore cannotmake direct contacts to DNA (Figure 3A)

The BfiI-C recognition sequence is extended to 6 bp byan extra GC base pair at the 30-terminus in the respect tothe 5-bp EcoRII-N target Surprisingly BfiI-C recognizesthe extended recognition site using a 5 residues shorter andmore compact C-arm (Figure 3A) than EcoRII-N Theside chains of DNA-facing amino acids on the C-armloop are shorter in BfiI-C than structurally equivalentresidues in EcoRII-N (D282 and N280 in BfiI-C versusE96 and R94 in EcoRII-N Figure 4) BfiI-C positionsthis compact cluster of amino acids close to the bases atthe 30-end of the recognition site 50-ACTGGG-30 andmakes direct contacts to all three 30-terminal GC basepairs (Figures 3B and 4) In contrast equivalent residuesin EcoRII-N (R94 and E96) make direct contacts onlyto the terminal GC base pair (underlined) within the50-CCTGG-30 target sequence In summary BfiI-Ctrades two non-specific contacts made by the C-arm ofEcoRII-N for a direct base-specific H-bond (N280 mainchain oxygen to the N4 atom of the terminal C base inthe bottom-strand) The loss of the C-arm andDNAndashbackbone contacts in BfiI-C is compensated by anincreased number of non-specific contacts coming fromother structural elements (Figure 3 and SupplementaryTable S1) The C-arm in RAV1-B3 is only 1 amino acidshorter than in BfiI-C and would be consistent withspecific binding of RAV1-B3 to a 6-bp recognition site(Figure 3) On the other hand the non-specific DNA-binding protein VRN1-B3 (9) has a C-arm that is 3residues shorter than in BfiI-C It would be interestingto see whether the length of N- and C-arms of B3-likedomains correlate with the DNA-binding specificity andrecognition sequence length

Conservation of DNAndashbackbone contacts

Analysis the DNAndashbackbone contacts in the BfiI-C andEcoRII-N complexes revealed conserved interactions withphosphate groups referred here as lsquoclamprsquo phosphates atthe 50-ends of the top and the bottom DNA strands(Figure 3B and Supplementary Figure S3) The C-armof BfiI-C is fixed to the 50-pACTGGG-30 phosphategroup via a positively charged side chain of R272residue (Figure 3B) that is spatially equivalent to theEcoRII-N R81 residue that interacts with the top strandphosphate 50-pCCTGG-30 The impaired DNA-binding

4120 Nucleic Acids Research 2014 Vol 42 No 6

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

affinity of the R272A BfiI mutant (Table 3) and spatialconservation of this arginine residue in all available B3-like domain structures (Figure 3A) underscores import-ance of this backbone contact for all B3-like domainsBfiI lacks the direct structural equivalent to the EcoRII-N residue K23 which binds the bottom strand phosphate30-GGACCp-50 and is conserved in RAV1-B3 At1g16440-B3 (Figure 3A) and many other B3 domains (5) Howevera similar DNAndashbackbone contact to the clamp phosphate30-TGACCpC-50 on the bottom strand is provided by theBfiI-C residue K340 which resides on a different struc-tural element (Figure 3 and Supplementary Figure S3)We propose that interactions of the conserved positivelycharged residues with lsquoclamprsquo phosphates on the top andbottom DNA strands help to position the DNA moleculein the binding cleft in the conserved orientation andpromote formation of base-specific contacts This obser-vation is in line with the recent analysis of crystal struc-tures of DNA-binding protein complexes (34)

CONCLUSIONS

The crystal structure of the BfiI REase DBD (BfiI-C) incomplex with DNA provides a first glimpse into the mech-anism of 6-bp sequence recognition by a B3-like proteinThe BfiI-CndashDNA structure confirms the conserved DNA-binding mode inferred previously for B3-like domains (7)and reveals a conserved set of non-specific and specificinteractions with DNA Two positively charged BfiI-Cresidues namely K340 and R272 make contacts to thelsquoclamprsquo DNA phosphates at the opposite termini of theDNA-recognition site thereby anchoring the wrench-likeDNA-binding surface in the DNA major groove Spatialconservation of these residues in majority of B3 do-mains implies a similar DNA lsquoclampingrsquo mechanismFurthermore both BfiI-C and EcoRII-N use N- andC-arms for making base-specific contacts to their 6- and5-bp recognition sequences respectively The amino acid

residues located in the loops within the C-arms determinethe length and specificity of the target site The loops in theN- and C-arms of plant B3 domains show great variabilityin length and sequence consistent with the diversity of theirDNA-recognition sequences The BfiI-CndashDNA structurepresented here opens the way for modeling of DNA-bound B3 domains of plant TFs using a novel template

ACCESSION NUMBERS

3ZI5

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

The authors are grateful to Dr Ar unas Lagunaviciusfor sharing BfiI plasmids and data on R246A andR247A mutants Authors acknowledge Egle_Rudokiene_ forsequencing and Dr Zita Liutkevici ute_ for the mass spec-trometry Authors also thank Dr Garib Murshudov andDr Haim Rozenberg for suggestions and help during thestructure refinement

FUNDING

Research Council of Lithuania [MIP-0862011 to DGand EM] European Union Research Potential CallFP7-REGPOT-2009-1 (245721 lsquoMoBiLirsquo project)Funding for open access charge Research Council ofLithuania

Conflict of interest statement None declared

Figure 4 Recognition of the 30-terminal nucleotides by BfiI-C and EcoRII-N The C-arms of BfiI-C and EcoRII-N are orange and pink respect-ively the top DNA strand is white the bottom DNA strand is grey Only the last 3 bp of the BfiI recognition site and the overlapping base pairsfrom the EcoRII-NndashDNA structure (PDB ID 3HQF) are shown

Nucleic Acids Research 2014 Vol 42 No 6 4121

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from

REFERENCES

1 SwaminathanK PetersonK and JackT (2008) The plant B3superfamily Trends Plant Sci 13 647ndash655

2 RomanelEAC SchragoCG CounagoRM RussoCAMand Alves-FerreiraM (2009) Evolution of the B3 DNA bindingsuperfamily new insights into REM family gene diversificationPLoS One 4 e5791

3 ZhouXE WangY ReuterM MuckeM KrugerDHMeehanEJ and ChenL (2004) Crystal structure of type IIErestriction endonuclease EcoRII reveals an autoinhibitionmechanism by a novel effector-binding fold J Mol Biol 335307ndash319

4 GrazulisS ManakovaE RoessleM BochtlerMTamulaitieneG HuberR and SiksnysV (2005) Structure of themetal-independent restriction enzyme BfiI reveals fusion of aspecific DNA-binding domain with a nonspecific nuclease ProcNatl Acad Sci USA 102 15797ndash15802

5 YamasakiK KigawaT InoueM TatenoM YamasakiTYabukiT AokiM SekiE MatsudaT TomoY et al (2004)Solution structure of the B3 DNA binding domain of theArabidopsis cold-responsive transcription factor RAV1 PlantCell 16 3448ndash3459

6 WaltnerJK PetersonFC LytleBL and VolkmanBF (2005)Structure of the B3 domain from Arabidopsis thaliana proteinAt1g16640 Protein Sci 14 2478ndash2483

7 GolovenkoD ManakovaE TamulaitieneG GrazulisS andSiksnysV (2009) Structural mechanisms for the 5rsquo-CCWGGsequence recognition by the N- and C-terminal domains ofEcoRII Nucleic Acids Res 37 6613ndash6624

8 YamasakiK KigawaT SekiM ShinozakiK andYokoyamaS (2013) DNA-binding domains of plant-specifictranscription factors structure function and evolution TrendsPlant Sci 18 267ndash276

9 KingGJ ChansonAH McCallumEJ Ohme-TakagiMByrielK HillJM MartinJL and MylneJS (2013) TheArabidopsis B3 Domain Protein VERNALIZATION1 (VRN1) IsInvolved in Processes Essential for Development with Structuraland Mutational Studies Revealing Its DNA-binding SurfaceJ Biol Chem 288 3198ndash3207

10 KagayaY OhmiyaK and HattoriT (1999) RAV1 a novelDNA-binding protein binds to bipartite recognition sequencethrough two distinct DNA-binding domains uniquely found inhigher plants Nucleic Acids Res 27 470ndash478

11 UlmasovT HagenG and GuilfoyleTJ (1997) ARF1 atranscription factor that binds to auxin response elementsScience 276 1865ndash1868

12 SuzukiM KaoCY and McCartyDR (1997) The conserved B3domain of VIVIPAROUS1 has a cooperative DNA bindingactivity Plant Cell 9 799ndash807

13 TamulaitisG MuckeM and SiksnysV (2006) Biochemical andmutational analysis of EcoRII functional domains revealsevolutionary links between restriction enzymes FEBS Lett 5801665ndash1671

14 ZarembaM UrbankeC HalfordSE and SiksnysV (2004)Generation of the BfiI restriction endonuclease from the fusion ofa DNA recognition domain to a non-specific nuclease from thephospholipase D superfamily J Mol Biol 336 81ndash92

15 LagunaviciusA SasnauskasG HalfordSE and SiksnysV(2003) The metal-independent type IIs restriction enzyme BfiI is a

dimer that binds two DNA sites but has only one catalyticcentre J Mol Biol 326 1051ndash1064

16 LeslieA (1992) Recent changes to the MOSFLM package forprocessing film and image plate data Joint CCP4+ ESF-EAMCB Newsletter on Protein Crystallography 26 27ndash33

17 LeslieAGW (2006) The integration of macromoleculardiffraction data Acta Crystallogr D Biol Crystallogr 62 48ndash57

18 EvansP (2006) Scaling and assessment of data quality ActaCrystallogr D Biol Crystallogr 62 72ndash82

19 MurshudovGN VaginAA and DodsonEJ (1997) Refinementof macromolecular structures by the maximum-likelihood methodActa Crystallogr D Biol Crystallogr 53 240ndash255

20 AdamsPD AfoninePV BunkocziG ChenVB DavisIWEcholsN HeaddJJ HungL-W KapralGJ Grosse-KunstleveRW et al (2010) PHENIX a comprehensive Python-based system for macromolecular structure solution ActaCrystallogr D Biol Crystallogr 66 213ndash221

21 EmsleyP and CowtanK (2004) Coot model-building tools formolecular graphics Acta Crystallogr D Biol Crystallogr 602126ndash2132

22 van DijkM and BonvinAMJJ (2009) 3D-DART a DNAstructure modelling server Nucleic Acids Res 37 W235ndashW239

23 LaveryR and SklenarH (1989) Defining the structure ofirregular nucleic acids conventions and principles J BiomolStruct Dyn 6 655ndash667

24 HubbardSJ and ThorntonJ(1993) lsquoNACCESSrsquo ComputerProgram

25 LuscombeNM LaskowskiRA and ThorntonJM (1997)NUCPLOT a program to generate schematic diagrams ofprotein-nucleic acid interactions Nucleic Acids Res 254940ndash4945

26 KraulisPJ (1991) MOLSCRIPT a program to produce bothdetailed and schematic plots of protein structures J Appl Cryst24 946ndash950

27 MerrittEA and BaconDJ (1997) Raster3D Photorealisticmolecular graphics Methods Enzymol 277 505ndash524

28 ZhengL BaumannU and ReymondJ-L (2004) An efficientone-step site-directed and site-saturation mutagenesis protocolNucleic Acids Res 32 e115

29 BarikS (1995) Site-directed mutagenesis by double polymerasechain reaction Mol Biotechnol 3 1ndash7

30 SapranauskasR SasnauskasG LagunaviciusA VilkaitisGLubysA and SiksnysV (2000) Novel subtype of type IIsrestriction enzymes BfiI endonuclease exhibits similarities to theEDTA-resistant nuclease Nuc of Salmonella typhimurium J BiolChem 275 30878ndash30885

31 ShatskyM NussinovR and WolfsonHJ (2004) A method forsimultaneous alignment of multiple protein structures Proteins56 143ndash156

32 GouetP RobertX and CourcelleE (2003) ESPriptENDscriptExtracting and rendering sequence and 3D information fromatomic structures of proteins Nucleic Acids Res 31 3320ndash3323

33 YamasakiK KigawaT InoueM WatanabeS TatenoMSekiM ShinozakiK and YokoyamaS (2008) Structures andevolutionary origins of plant-specific transcription factorDNA-binding domains Plant Physiol Biochem 46 394ndash401

34 RohsR JinX WestSM JoshiR HonigB and MannRS(2010) Origins of specificity in protein-DNA recognition AnnuRev Biochem 79 233ndash269

4122 Nucleic Acids Research 2014 Vol 42 No 6

at Vilnius U

niversity on September 26 2016

httpnaroxfordjournalsorgD

ownloaded from