peroxidase determined by paramagnetic NMRc and cytochrome cSolution structure and dynamics of the complex between cytochrome

Alexander N. Volkov, Jonathan A. R. Worrall, Elodie Holtzmann, and Marcellus Ubbink

doi:10.1073/pnas.0603551103 published online Dec 4, 2006; PNAS

This information is current as of December 2006.

Supplementary Material www.pnas.org/cgi/content/full/0603551103/DC1

Supplementary material can be found at:

www.pnas.org#otherarticlesThis article has been cited by other articles:

E-mail Alerts. click hereat the top right corner of the article or

Receive free email alerts when new articles cite this article - sign up in the box

Rights & Permissions www.pnas.org/misc/rightperm.shtml

To reproduce this article in part (figures, tables) or in entirety, see:

Reprints www.pnas.org/misc/reprints.shtml

To order reprints, see:

Notes:

Solution structure and dynamics of the complexbetween cytochrome c and cytochrome c peroxidasedetermined by paramagnetic NMRAlexander N. Volkov, Jonathan A. R. Worrall*, Elodie Holtzmann†, and Marcellus Ubbink‡

Leiden Institute of Chemistry, Leiden University, Gorlaeus Laboratories, P.O. Box 9502, 2300 RA Leiden, The Netherlands

Edited by Brian M. Hoffman, Northwestern University, Evanston, IL, and approved October 18, 2006 (received for review May 1, 2006)

The physiological complex of yeast cytochrome c peroxidase andiso-1-cytochrome c is a paradigm for biological electron transfer.Using paramagnetic NMR spectroscopy, we have determined theconformation of the protein complex in solution, which is shownto be very similar to that observed in the crystal structure [PelletierH, Kraut J (1992) Science 258:1748–1755]. Our results support theview that this transient electron transfer complex is dynamic. Thesolution structure represents the dominant protein–protein orien-tation, which, according to our estimates, is occupied for >70% ofthe lifetime of the complex, with the rest of the time spent in thedynamic encounter state. Based on the observed paramagneticeffects, we have delineated the conformational space sampled bythe protein molecules during the dynamic part of the interaction,providing experimental support for the theoretical predictions ofthe classical Brownian dynamics study [Northrup SH, Boles JO,Reynolds JCL (1988) Science 241:67–70]. Our findings corroboratethe dynamic behavior of this complex and offer an insight into themechanism of the protein complex formation in solution.

electron transfer � encounter state � transient complex � spin label �paramagnetic relaxation enhancement

The process of protein complex formation can be described bya two-step model, in which a short-lived, dynamic encounter

complex precedes a dominant, well defined state (Fig. 1). Theformer enables proteins to undergo reduced-dimensionalitysearch of the optimal binding geometry, thereby acceleratingmolecular association as compared with 3D diffusion (1). Fastmolecular association is essential for protein–protein complexesthat require high turnover rates, like those involved in electrontransfer (ET) in photosynthesis, respiration, and other metabolicprocesses (2). The physiological complex of yeast iso-1-cytochrome c (Cc) and yeast cytochrome c peroxidase (CcP) isa paradigm for the intermolecular ET (3) and is one of the fewtransient ET complexes for which a crystal structure has beensolved (4). A recent study has confirmed that the protein–proteinorientation observed in the crystal is ET-active (5). However, ithas remained a matter of debate whether this structure repre-sents the only form in solution (6–8). According to severalstudies, the complex is dynamic, and the crystal structure mightrepresent only a subpopulation of protein orientations (9–15).Recent studies show that the photoinduced ET between Zn-substituted CcP and Cc, both in the crystal (8, 16) and in solution(17), occurs with faster backward than forward rates, indicatingthat the complex is present in multiple forms, only a few of whichare ET-active (17).

Characterization of the binding interface in the dynamicencounter state (Fig. 1B) has so far proven to be elusive.Investigation of protein complexes by using x-ray crystallographyor conventional NMR spectroscopy addresses only the single-orientation species (Fig. 1C), and the only way to visualize thedynamic state is offered by theoretical modeling studies (11, 18).In the recent, elegant work of Clore and coworkers (19, 20), itwas shown how paramagnetic relaxation can be applied to studythe dynamic state of protein–DNA complexes. We report on the

application of an analogous experimental approach that allowsus to define both the dominant protein–protein orientation (Fig.1C) and the conformational space sampled by the proteins in thedynamic encounter complex (Fig. 1B).

Results and DiscussionSolution Structure of the Complex. The concept of the approach isthat NMR resonance intensities of one of the proteins in thecomplex are affected by a paramagnetic spin label covalentlyattached to the other protein (Fig. 2). The paramagnetic effectsare converted into distance restraints (21–23), which can be usedto calculate protein–protein orientations within the complex(24). Five single-cysteine CcP variants were prepared (Fig. 3A)and labeled with a paramagnetic nitroxide spin label, (1-oxyl-2,2,5,5-tetramethyl-3-pyrroline-3-methyl)-methanethiosulfon-ate (MTSL). For three of them (N38C, N200C, and T288C), thisresults in a broadening of backbone amide resonances of Ccwhen bound to CcP, whereas for the other two (S263C andT137C), no effects are observed [see supporting information (SI)Fig. 6]. A set of experimental restraints from the spin-labeledcomplexes was used for the subsequent docking calculations (SITable 1). Reproducibly, a single set of orientations was found(Fig. 3B). Comparison of the best solution structure with thecrystal structure of the complex shows that the two are very

Author contributions: A.N.V., J.A.R.W., and M.U. designed research; A.N.V., J.A.R.W., andE.H. performed research; A.N.V., J.A.R.W., and M.U. analyzed data; and A.N.V. and M.U.wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS direct submission.

Abbreviations: ET, electron transfer; Cc, yeast iso-1-cytochrome c; CcP, yeast cytochrome cperoxidase; PRE, paramagnetic relaxation enhancement; MTSL, (1-oxyl-2,2,5,5-tetramethyl-3-pyrroline-3-methyl)-methanethiosulfonate; MTS, (1-acetyl-2,2,5,5-tetramethyl-3-pyrroline-3-methyl)-methanethiosulfonate; HSQC, heteronuclear single-quantum coherence.

Data deposition: The atomic coordinates have been deposited in the Protein Data Bank,www.pdb.org (PDB ID code 2GB8).

*Present address: Department of Biochemistry, University of Cambridge, Tennis CourtRoad, Old Addenbrookes Site, Cambridge CB2 1GA, United Kingdom.

†Present address: Faculty of Pharmacy, Universite Paris 5, 4 Avenue de l’Observatoire, 75006Paris, France.

‡To whom correspondence should be addressed. E-mail: m.ubbink@chem.leidenuniv.nl.

This article contains supporting information online at www.pnas.org/cgi/content/full/0603551103/DC1.

© 2006 by The National Academy of Sciences of the USA

Fig. 1. Model for the formation of a protein complex. Free proteins (A)associate to form an encounter complex (B) consisting of an ensemble of proteinorientations, which is in equilibrium with a single-orientation complex (C).

www.pnas.org�cgi�doi�10.1073�pnas.0603551103 PNAS � December 12, 2006 � vol. 103 � no. 50 � 18945–18950

BIO

PHYS

ICS

similar, with a root mean square deviation (rmsd) of the Ccbackbone atoms of 2.2 Å after superposition of CcP (Fig. 3C).This finding provides direct evidence that the crystal structurerepresents the dominant form in solution.

The approach used here provides information about theconformation of the proteins within the complex but gives nodetails on the orientation of the side chains. To establish whetherthe starting coordinate set influences the final structure, wecompared the docking solution obtained with the coordinates offree Cc (28) and CcP (29) with those taken from the crystalstructure (4). The two complexes exhibit an rmsd of Cc backboneatoms of 4.2 Å. With the coordinates of free CcP and those ofCc taken from the crystal structure, the difference disappears(rmsd 0.70 Å), indicating that the coordinates of free Cc affectthe outcome of the docking. This effect may be attributable tothe Cc side chain of Q16. In free Cc, Q16 protrudes from thesurface, preventing a close approach of the proteins and causinga lateral translation within the docked complex as compared withthe crystal structure. In the latter, the Q16 side chain is foldedback to form a hydrogen bond with its own backbone (4), which,

rather than being an artifact of crystallization, is induced by thebinding (30). Apparently, this rearrangement is crucial for theformation of the correct complex.

Additional Paramagnetic Effects. Analysis of the solution structureshows that most of the restraints are satisfied (Fig. 4). Theviolated Cc residues are nonrandomly distributed. For instance,the spin label attached to N38C CcP exerts paramagnetic effectson Cc residues that map on the back of the protein (red in Fig.4A), whereas the spin label at N200C CcP strongly affects theN-terminal helix (red in Fig. 4B). In the solution structure, theseresidues are located too far away from the spin label to beaffected. The ‘‘additional’’ effects, shown in Fig. 4, are repro-ducible and suggest that the calculated orientation is not suffi-cient to explain all observed paramagnetic effects. These addi-tional effects could have several causes: (i) nonspecificinteractions caused by random collisions; (ii) binding of morethan one Cc to CcP (31–33); or (iii) dynamics within the complex.It should be noted that a 2:1 complex of Cc and CcP (possibilityii) is not expected to be observed under the present experimental

Fig. 2. Intermolecular paramagnetic effects. (A and B) Parts of the 15N-1H HSQC spectra of Cc bound to CcP N200C modified with a diamagnetic (black, A) orparamagnetic (red, B) label. The asterisks indicate resonances broadened beyond detection. (C) A slice through the amide resonance of K54 in the 1H dimension,marked with a dotted line in A and B. The lines marked Ipara and Idia indicate the intensities of the resonance in the paramagnetic and diamagnetic complexes,respectively.

Fig. 3. Solution structure of the Cc–CcP complex. (A) The residues of CcP used for mutagenesis are labeled and shown as red sticks. Ribbon representation ofthe complex was drawn from the crystal structure [PDB entry 2PCC (4)]. (B) Ensemble of best 20 solution structures. The average rmsd from the best structurefor Cc backbone atoms is 0.7 Å. (C) Comparison of the solution and crystal structures. In A–C, CcP is at the bottom, and heme groups for both proteins are shownin sticks. In B and C, the CcP molecules are superimposed; the labels indicate protein termini. In C, the Cc backbones for the best solution and crystal structuresare shown in red and gray, respectively. The backbone rmsd between the two Cc molecules is 2.2 Å. All images of the proteins used in this report were madewith MOLSCRIPT (25) and rendered in Raster3D (26), except for Fig. 5, which was made with Swiss-PDB (27) and POV-Ray 3.6 (www.povray.org).

18946 � www.pnas.org�cgi�doi�10.1073�pnas.0603551103 Volkov et al.

conditions (0.1 M NaCl) (33). The first two possibilities imply thepresence of distinct complexes in which Cc is bound withdifferent affinities. If this is the case, lowering Cc concentrationin a sample with a constant concentration of CcP would resultin a decrease of the fraction of the weaker complexes. Thisdecrease would lead to a reduction of the contributions from theweaker complexes to the paramagnetic relaxation enhancements(PREs), resulting in fewer violations and a nonuniform concen-tration-dependence of PREs. On the contrary, in the case ofdynamics within the complex, lowering the Cc concentrationshould not influence the contribution of the dynamic encounterstate to the PREs.

Control experiments in which Cc concentration was five timeslower than in the experiments described above, although theconcentration of CcP remained the same, were carried out forN200C CcP. This variant was selected because the spin label

attached at this position produces the largest violations (Fig. 4B).Under the conditions used, the fraction of Cc available for theweaker interactions decreases from 0.10 to 0.015 (see Materialsand Methods). However, the results show the same pattern forthe observed paramagnetic effect as those for the 1:1 proteinratio, with the largest deviation of the restraint target distanceof �2 Å for several residues (SI Fig. 6F). Therefore, we concludethat additional paramagnetic effects are independent of the Ccconcentration in the range used. This finding is in agreementwith multiple protein–protein orientations within the complex.

Dynamics Within the Complex. The presence of multiple orienta-tions within Cc–CcP complex previously has been suggested byseveral studies (9–15, 17). A complex comprising an equilibriumbetween a well defined form and a dynamic state (Fig. 1)adequately could explain the additional paramagnetic effects.Because of the sixth-power distance-dependence of the para-magnetic relaxation, a dynamic state will contribute to the PREsfor the residues that get close to the spin label, even if this stateis populated for only a few percent of the time. Thus, theobserved paramagnetic effect represents a sum of PRE contri-butions from all protein orientations within the complex, cor-responding to the combination of the dominant complex and themultiple forms in the dynamic ensemble (Fig. 1 B and C; also seeMaterials and Methods).

We have made an estimate for the lower limit of the fractionof time spent by the proteins in the dominant orientation ( fdom),based on the assumption that the effects for all residues withsatisfied restraints arise mostly from this orientation. To obtainan actual PRE (R2

para) for these residues, the observed PRE(R2, obs

para ) must be multiplied by 1/fdom (see Eq. 3 in Materials andMethods). Protein docking performed with the R2

para valuescorresponding to fdom � 0.7 results in an energy increase for allsolutions attributable to both van der Waals and restraintviolations. To satisfy the shorter distance restraints resultingfrom low fdom values, Cc should come unrealistically close toMTSL and, accordingly, CcP atoms, resulting in solutions thatare sterically forbidden. From these calculations, we concludethat the proteins spend �70% of the lifetime of the complex inthe dominant orientation. This finding is consistent with otherstudies that have proposed the existence of a single-orientationcomplex in solution (10, 34, 35). In particular, large NMRchemical-shift perturbations of Cc resonances upon complexformation suggest the presence of a dominant orientation (30)because for more dynamic protein complexes, the perturbationstend to be smaller (36, 37).

Conformational Space of the Encounter Complex. It is impossible toidentify the orientations constituting the dynamic encounterstate without knowing how many of these comprise the ensembleand what fraction of the lifetime of the complex is spent in eachof them. However, we can define the conformational spaceoccupied by Cc in the encounter complex (Fig. 5). This definitionis achieved by mapping out the areas around those spin labelsthat give rise to additional PREs (N38C, N200C, and T288C) andthose that do not affect Cc atoms at all (S263C and T137C). Inthe dynamic state, Cc must come close to the spin labels thatcause additional PREs (the extent of approach indicated by thered area in Fig. 5) and be far from those showing no effects (theareas to be avoided are colored blue in Fig. 5). The remainingspace around CcP also could harbor Cc molecules; however,those would not contribute to the observed paramagnetic effects.

In the dynamic state, the binding interface on CcP is localizedaround the Cc position in the dominant orientation. This findingis in agreement with the view that an encounter complexfacilitates formation of the dominant one via preorientation ofthe protein molecules and reduced-dimensionality search (1, 9).The CcP binding surface contains a part of the negatively

Fig. 4. Violation analysis of the best solution structure for the Cc–CcPcomplex. (Left) Graphs illustrate the distances from the Cc backbone amideprotons in the best solution structure (red circles) and crystal structure (redline) to the averaged position of the oxygen atom of the spin label attachedto CcP at N38C (A), N200C (B), and T288C (C). The black dotted and solid linesindicate the PRE-derived distances and error margins, respectively, used in thestructure calculations. (Right) Cartoon representations of the best solutionstructure, indicating the residues with satisfied (blue) and violated (red)restraints. Heme groups for both proteins are in cyan. For each of the spin-label positions, four oxygen atoms, representing the conformational freedomof the spin label, are shown as orange spheres. These were used for ensembleaveraging in the structure calculations (see Materials and Methods).

Volkov et al. PNAS � December 12, 2006 � vol. 103 � no. 50 � 18947

BIO

PHYS

ICS

charged region predicted to interact with Cc by a classicalBrownian dynamics study (11), suggesting that electrostaticattraction plays a dominant role in determining the nature of theencounter complex. Clearly, the binding interface on CcP islarger than that observed in the crystal structure (4), with anearlier NMR chemical-shift perturbation analysis showing thesame to be true for Cc (30). Several reports described CcPbinding effects on the rear of Cc in solution (13, 14, 30), and arecent study of complexes between CcP and different variants ofCc has reported crystal structures in which Cc is positioned withits back toward CcP (16). Such binding modes, sampled in thedynamic state of the complex, can explain the paramagneticeffects at the back of Cc (Fig. 4A).

Implications for Intermolecular ET. As we show in this work, thedominant form of the Cc–CcP complex in solution is the sameas the protein–protein orientation observed in the crystal struc-ture. If the dominant form of the complex also is the mostET-reactive, as indeed was suggested by Poulos and coworkers(5), this could explain similar ET rates observed in the crystalstructure (8) and in solution (17). However, the forward ETbetween Cc and CcP is likely to be conformationally gated,implying that the most favorable binding geometry is not nec-essarily the most ET-active (8, 17). In fact, there might be amultitude of reactive ET configurations as suggested by the studyof variant Cc–CcP complexes (16). In this case, the ET may berationalized by the protein–protein orientations, some of whichcould be ET-active, sampled in the dynamic encounter state ofthe complex. The conformational space of the encounter statedelineated in this work therefore could represent the extent ofconformational f luctuations in Cc–CcP complex that modulatethe ET in solution (17).

Summary. By defining the conformational space of an ensembleof protein orientations and characterizing the dominant struc-ture of the complex, we were able to address both sides of theequilibrium of the complex formation between Cc and CcP (Fig.1). The dominant orientation of the protein complex in solutionis the same as that in the crystal structure. This form is inequilibrium with a dynamic state that contributes to the observedintermolecular effects. The experimental approach outlined inthe present report will be useful for study of other biological

macromolecular complexes with a particular focus on the role ofdynamics in biomolecular interactions.

Materials and MethodsProtein Preparation. Isotopically enriched 15N Cc (T-5A/C102T)was expressed in Escherichia coli and purified as previouslydescribed (38, 39). To prepare CcP mutants, its single nativecysteine (C128) was changed into alanine, and another cysteineresidue was introduced at a desired position on the proteinsurface (Fig. 3A). Site-directed mutagenesis was carried out byusing the Quik Change PCR protocol (Stratagene, La Jolla, CA)with the plasmid CCP(MKT) as a template (40). All constructswere verified by DNA sequencing. Mutant proteins were ex-pressed and purified analogously to the wild type (40), except forthe gel-filtration step, which was carried out in the presence of1 mM DTT.

Purified single-cysteine CcP mutants were incubated with 10mM DTT in 0.1 M Tris�HCl (pH 8.0)/0.1 M NaCl for 2 h at roomtemperature to reduce intermolecular disulfide bonds. DTT wasremoved by passing the CcP solution through a PD-10 column(Amersham Pharmacia, Uppsala, Sweden). The resulting mo-nomeric protein was reacted with a 7- to 10-fold excess ofparamagnetic (MTSL) or diamagnetic control (MTS) label (SIFig. 7) and incubated overnight at room temperature. MTSL andMTS [(1-acetyl-2,2,5,5-tetramethyl-3-pyrroline-3-methyl)-methanethiosulfonate] were purchased from Toronto ResearchChemicals (North York, ON, Canada) and used without furtherpurification. Upon completion of the reaction, protein solutionwas passed through a PD-10 column to remove any unreactedlabel, exchanged into 20 mM NaPi/100 mM NaCl buffer (pH6.0), and concentrated by using Amicon Ultra concentrators(Millipore, Billerica, MA). In each case, the yield of labeling wasclose to 100% as estimated from double-integrated EPR spectraof mutant CcP-MTSL and a control sample containing a knownamount of free MTSL.

Throughout the study, we used ferric Cc and five-coordinatedhigh-spin ferric CcP with previously reported purity criteria (30).Concentrations of Cc and CcP were determined according to theoptical absorbance peaks at 410 nm (� � 106.1 mM�1 cm�1) and408 nm (� � 98 mM�1 cm�1), respectively (30).

NMR Samples and Experiments. NMR samples contained 0.3–0.5mM 1:1 complex of 15N Cc with wild-type CcP, mutant CcP-

Fig. 5. Conformational space occupied in the Cc–CcP encounter complex. The best solution structure of the complex in the same orientation as that in Fig. 3B and C is shown in a stereoview. CcP and Cc are in light gray and yellow ribbons, respectively, with heme groups in green and spin-label positions indicated byorange spheres. The red surface defines the area that must contain the center of mass of Cc for at least 3% of the lifetime of the complex (outer perimeter ofthe red area) or less (moving from the periphery toward a spin label). The blue surface corresponds to the area that cannot be occupied by the Cc center of massfor longer than 3% of the lifetime of the complex (outer perimeter of the blue area) or less (moving to the center toward a spin label).

18948 � www.pnas.org�cgi�doi�10.1073�pnas.0603551103 Volkov et al.

MTSL, or CcP-MTS in 20 mM NaPi/100 mM NaCl (pH 6.0)/6%D2O for lock and 0.1 mM CH3CO15NH2 as an internal reference.For the control experiment with low Cc concentration, describedin Results and Discussion, samples of 0.3 mM CcP-MTS orCcP-MTSL with 0.06 mM 15N Cc were used. The pH of thesamples was adjusted to 6.00 � 0.05 with small aliquots of 0.1 MHCl or 0.1 M NaOH. Measurements were performed at 301 Kon a Bruker DMX600 spectrometer equipped with a triple-resonance gradient probe or, in case of the samples containing0.06 mM 15N Cc, a TCI-Z-GRAD CryoProbe (Bruker,Karlsruhe, Germany). 2D 15N-1H heteronuclear single-quantumcoherence (HSQC) spectra were obtained with 2,048 and 512points in the direct and indirect dimensions, respectively, andspectral widths of 32 ppm (15N) and 16 ppm (1H). All data wereprocessed with the Azara suite of programs (provided by WayneBoucher and the Department of Biochemistry, University ofCambridge, Cambridge, U.K.) and analyzed in Ansig for Win-dows (41, 42). Assignments of the 15N and 1H nuclei of Cc weretaken from a previous report (30). In the present work, a numberof Cc amides (A3, F10, H26, L32-H33, H39, N56, N63, M64, E66,Y74, K79-A81, G83-G84, K86, E88, and L94) were either notobserved or not analyzed because of spectral overlap.

Determination of Distance Restraints from PREs. For each observedamide proton of Cc, an MTSL-induced PRE was calculated fromEq. 1 (23):

Ipara

Idia�

R2,dia exp(�tR2,para)R2,dia�R2,para

, [1]

where Ipara and Idia are measured intensities of HSQC peaks ofCc in the complex with CcP-MTSL and CcP-MTS, respectively(SI Fig. 6); R2, dia is the transverse relaxation rate of Cc amideprotons in the complex with CcP-MTS; R2, para is the paramag-netic contribution to the relaxation rate (PRE); and t is the totalpolarization transfer time of the HSQC. For each amide peak,R2, dia was estimated from the width at half-height (��1/2) of itsLorentzian fit in proton dimension by using R2, dia � ���1/2. Forthe residues whose resonances disappear in the paramagneticspectrum, Ipara was estimated from the noise level of the spec-trum. Calculated PRE rates were converted into distances byusing Eq. 2 (23):

r � �6 �2g2�2�c

20R2,para� 4

31 h

2�c2� , [2]

where r is the distance between the unpaired electron of theMTSL and a given amide proton of Cc; �c is the rotationalcorrelation time of the electron-nucleus vector; h and � areproton Larmor frequency and gyromagnetic ratio, respectively;g is the electronic g factor; and � is the Bohr magneton.

Three classes of intermolecular distance restraints, used insubsequent docking calculations, were defined (SI Table 1) (23).Residues that are strongly affected by MTSL and whose reso-nances disappear in the paramagnetic spectrum are restrainedwith only an upper bound. Those not affected by MTSL haveonly a lower limit. Finally, the residues affected by the spin labeland whose resonances are observed in the paramagnetic spectraare restrained with both upper and lower bounds. The marginsof �4 Å for the calculated restraints accommodate the experi-mental error and enhance the convergence of docking solutionsas compared with tighter bounds (23).

Protein Docking. Coordinates of both proteins were taken fromthe Protein Data Bank [PDB ID code 2PCC (4)]. For thedocking of individual proteins, the molecular coordinates fromPDB ID codes 2YCC (28) for Cc and 1ZBY (29) for CcP wereused. Surface cysteine mutations of CcP were introduced in

silico, followed by addition of the MTSL atoms and energyminimization of the labeled protein. To take into accountmobility of the MTSL attached to the surface of CcP, weperformed r�6 ensemble averaging of the intermolecular dis-tance restraints with four different MTSL orientations. Togenerate these orientations, we systematically rotated the at-tached MTSL around five single bonds that join its ring to the C�

atom of the cysteine and retained only the sterically allowedconformers for each mutant. From these, four representativeorientations were selected. Using 10 MTSL orientations pro-duced similar results, consistent with the finding that averagingover a larger set of atoms does not alter the outcome (43).

A set of distance restraints from four MTSL positions (SITable 1) was used to drive docking of the protein complex inXplor-NIH version 2.13 (44). First, a rigid-body docking of theprotein molecules was carried out. Only two energy terms werespecified during this calculation, corresponding to restraints andvan der Waals forces. The latter were defined as repel forces forthe protein atoms, whereas for MTSL atoms they were set tozero. For each run performed, a single cluster of low-energysolutions was consistently produced. During the second step, 30to 40 best structures were subjected to energy minimization andside-chain dynamics with fixed position of backbone atoms forboth proteins and active van der Waals parameters for MTSL.For the refined structures, the entire docking procedure wasrepeated until no further reduction in energy was observed. The20 best structures of the final solution showed an average rmsdfrom the lowest energy structure of 0.7 � 0.2 Å for the backboneatoms of Cc after superposition of CcP molecules (Fig. 3B).

Additional Caveats for Intermolecular Distance Calculations from PRE.The presence of a spin label on the surface of CcP can potentiallyinterfere with Cc binding. For all labeled CcP mutants reportedin this study, chemical-shift perturbations in the complex with15N Cc are equal to those in the wild-type complex at the sameprotein ratio, indicating that an attached spin label does not alterthe binding (data not shown).

The complex is in fast exchange on the NMR time scale;therefore, only one set of Cc resonances is observed in thespectrum. Given an association constant Ka � (1.9 � 0.3)�105

M�1 (obtained by isothermal titration calorimetry at 30°C in 20mM NaPi/0.1 M NaCl, pH 6.0; A.N.V., J.A.R.W., and M.U.,unpublished data), 88–91% of Cc is bound to CcP in a 1:1complex for the protein concentrations used (0.3�0.5 mM).Thus, the observed PRE (R2, obs

para ) represents a fraction (F) of theactual PRE (R2

para), so that R2, obspara � FR2

para, with F �0.88�0.91. The experimentally determined PREs were cor-rected by 1/F, to obtain the R2

para values that were used in furthercalculations.

Although �c is different for each amide proton, the use of asingle value for all nuclei has been justified (21, 22). For thecomplex of Cc and CcP, the upper limit of �c of 16 ns wasestimated from the Einstein–Stokes equation (45); it is in goodagreement with the previously reported value for a proteinsystem of a similar size (23). Considering a wide range of �cvalues used in the literature for conversion of PRE into distances(21–23), two control sets of distance restraints were generatedwith �c of 12 and 4 ns for each mutant. In each case, dockingcalculations resulted in the same best solution as that for therestraints based on �c of 16 ns (data not shown).

To check for possible nonspecific binding effects, a spin labelwas placed at two CcP positions farther away from the Cc bindingsite (Fig. 3A). Cc experiences no paramagnetic effect in thecomplex with spin-labeled T137C CcP, whereas with S263C CcP,only two Cc residues are affected (SI Fig. 6), consistent with thebest solution structure. This finding suggests that a nonspecificprotein binding does not take place in solution.

Volkov et al. PNAS � December 12, 2006 � vol. 103 � no. 50 � 18949

BIO

PHYS

ICS

Conformational Space of the Encounter Complex. For each Ccresidue, the observed paramagnetic effect (R2, obs

para ) represents asum of PRE contributions (R2

para) from n protein orientationswithin the complex, weighted by fraction of time ( f ) spent ineach of them:

R2,obspara � �

i�1

n

f iR2,i

para.

[3]

To delineate the conformational space explored by Cc in thecomplex with CcP as shown in Fig. 5, the largest distancebetween a backbone nucleus of Cc and a spin label that wouldresult in a significant (�7 Hz) R2, obs

para with f � 0.03 was calculatedto be 12 Å. Thus, nuclei that experience additional PREs (Fig.4) must spend at least 3% of the time at this distance from thespin label or less time at a shorter distance. Cc nuclei spend atmost 3% of the time at 12 Å from those spin labels that exert noparamagnetic effects and even less time at shorter distances. Useof a somewhat arbitrary value of f � 0.03 is justified because, dueto r�6 dependence, very similar results are obtained for othervalues for the encounter complex as can be appreciated from SI

Fig. 8. See SI Data Sets 1–3 for the coordinates of the bestsolution structure and the conformational space generated at f �0.01, f � 0.03, and f � 0.10, respectively.

To define the space that must be occupied or is avoided by thecenter of mass of Cc (Fig. 5 and SI Fig. 8), red and blue spheres,respectively, were constructed around the oxygen atom of MTSLwith the radius of 12 Å plus the average distance from the Cccenter of mass to its surface (14 Å). Only the segments of the redand blue spheres that are between 12 and 19 Å from the surfaceof CcP are retained. The former corresponds to the closestapproach of the center of mass of Cc to the surface of CcP,whereas the latter denotes the farthest distance from CcP atwhich the proteins still form a complex. In case of an overlap,blue space is given priority over red. The segments around spinlabels at N38C, N200C, and T288C are in red, whereas thosearound T137C and S263C are in blue.

We thank Prof. R. Boelens for valuable suggestions, Prof. Gerard W.Canters and Dr. Derek S. Bendall for critical reading of the manuscript,and an anonymous reviewer for useful advice. This work was supportedby The Netherlands Organization for Scientific Research (NWO-CWGrants 700.50.514, 700.52.425, and 98S1010).

1. Adam G, Delbruck M (1968) in Structural Chemistry and Molecular Biology, edsRich A, Davidson N (Freeman, San Francisco), pp 198–215.

2. Bendall DS (1996) in Protein Electron Transfer, ed Bendall DS (BIOS, Oxford),pp 43–68.

3. Mathews FS, Mauk AG, Moore GR (2000) in Protein–Protein Recognition, edKleanthous C (Oxford Univ Press, New York), pp 60–101.

4. Pelletier H, Kraut J (1992) Science 258:1748–1755.5. Guo M, Bhaskar B, Li H, Barrows TP, Poulos TL (2004) Proc Natl Acad Sci

USA 101:5940–5945.6. Nocek JM, Zhou JS, de Forest S, Priyadarshy S, Beratan DN, Onuchic JN,

Hoffman BM (1996) Chem Rev 96:2459–2490.7. Erman JE, Vitello LB (2002) Biochim Biophys Acta 1597:193–220.8. Kang SA, Marjavaara PJ, Crane BR (2004) J Am Chem Soc 126:10836–10837.9. McLendon G (1991) Struct Bond 75:159–174.

10. Moench SJ, Chroni S, Lou BS, Erman JE, Satterlee JD (1992) Biochemistry31:3661–3670.

11. Northrup SH, Boles JO, Reynolds JCL (1988) Science 241:67–70.12. McLendon G, Zhang Q, Wallin SA, Miller RM, Billstone V, Spears KG,

Hoffman BM (1993) J Am Chem Soc 115:3665–3669.13. Jeng MF, Englander SW, Pardue K, Rogalskyj JS, McLendon G (1994) Nat

Struct Biol 1:234–238.14. Yi Q, Erman JE, Satterlee JD (1994) Biochemistry 33:12032–12041.15. Moore GR, Cox MC, Crowe D, Osborne MJ, Rosell FI, Bujons J, Barker PD,

Mauk MR, Mauk AG (1998) Biochem J 332:439–449.16. Kang SA, Crane BR (2005) Proc Natl Acad Sci USA 102:15465–15470.17. Hoffman BM, Celis LM, Cull DA, Patel AD, Seifert JL, Wheeler KE, Wang

J, Yao J, Kurnikov IV, Nocek JM (2005) Proc Natl Acad Sci USA 102:3564–3569.

18. Gabdoulline RR, Wade RC (2001) J Mol Biol 306:1139–1155.19. Iwahara J, Schwieters CD, Clore GM (2004) J Am Chem Soc 126:12800–12808.20. Iwahara J, Clore GM (2006) Nature 440:1227–1230.21. Gillespie JR, Shortle D (1997) J Mol Biol 268:158–169.22. Gaponenko V, Howarth JW, Columbus L, Gasmi-Seabrook G, Yuan J,

Hubbell WL, Rosevear PR (2000) Prot Sci 9:302–309.

23. Battiste JL, Wagner G (2000) Biochemistry 39:5355–5365.24. Ubbink M, Ejdeback M, Karlsson BG, Bendall DS (1998) Structure (London)

6:323–335.25. Kraulis PJ (1991) J Appl Crystallogr 24:946–950.26. Merritt EA, Bacon DJ (1997) Methods Enzymol 277:505–524.27. Guex N, Peitsch MC (1997) Electrophoresis 18:2714–2723.28. Berghuis AM, Brayer GD (1992) J Mol Biol 233:959–976.29. Bonagura CA, Bhaskar B, Shimizu H, Li H, Sundaramoorthy M, McRee DE,

Goodin DB, Poulos TL (2003) Biochemistry 42:5600–5608.30. Worrall JAR, Kolczak U, Canters GW, Ubbink M (2001) Biochemistry

40:7069–7076.31. Zhou JS, Hoffman BM (1994) Science 265:1693–1696.32. Zhou JS, Nocek JM, DeVan ML, Hoffman BM (1995) Science 269:204–207.33. Mauk MR, Ferrer JC, Mauk AG (1994) Biochemistry 33:12609–12614.34. Wang X, Pielak GJ (1999) Biochemistry 38:16876–16881.35. Pielak GJ, Wang X (2001) Biochemistry 40:422–428.36. Worrall JAR, Liu Y, Crowley PB, Nocek JM, Hoffman BM, Ubbink M (2002)

Biochemistry 41:11721–11730.37. Volkov AN, Ferrari D, Worrall JAR, Bonvin AMJJ, Ubbink M (2005) Prot Sci

14:799–811.38. Pollock WBR, Rosell FI, Twitchett MB, Dumont ME, Mauk AG (1998)

Biochemistry 37:6124–6131.39. Morar AS, Kakouras D, Young GB, Boyd J, Pielak GJ (1999) J Biol Inorg Chem

4:220–222.40. Goodin DB, Davidson MG, Roe JA, Mauk AG, Smith M (1991) Biochemistry

30:4953–4962.41. Kraulis PJ (1989) J Magn Reson 84:627–633.42. Helgstrand M, Kraulis PJ, Allard P, Hard T (2000) J Biol NMR 18:329–336.43. Iwahara J, Schwieters CD, Clore GM (2004) J Am Chem Soc 126:5879–5896.44. Schwieters CD, Kuszewski JJ, Tjandra N, Clore GM (2003) J Magn Res

160:65–73.45. Cavanagh J, Fairbrother WJ, Palmer AG, III, Skelton NJ (1995) Protein NMR

Spectroscopy: Principles and Practice (Academic, London), pp 17–19.

18950 � www.pnas.org�cgi�doi�10.1073�pnas.0603551103 Volkov et al.