The cytochrome f–plastocyanin complex as a model to study transient interactions between redox proteins

Review The cytochrome f–plastocyanin complex as a model to study transient interactions between redox proteins Isabel Cruz-Gallardo, Irene Díaz-Moreno, Antonio Díaz-Quintana, Miguel A. De la Rosa ⇑ Instituto de Bioquímica Vegetal y Fotosíntesis, Centro de Investigaciones Científicas, Isla de la Cartuja (cicCartuja), Universidad de Sevilla-CSIC, Avda. Américo Vespucio 49, Sevilla 41092, Spain article info Article history: Received 3 July 2011 Revised 8 August 2011 Accepted 24 August 2011 Available online 30 August 2011 Edited by Miguel Teixeira and Ricardo O. Louro This article is dedicated to the memory of Prof. António V. Xavier Keywords: Cytochrome f Electron transfer Plastocyanin Transient complex Photosynthesis abstract Transient complexes, with a lifetime ranging between microseconds and seconds, are essential for biochemical reactions requiring a fast turnover. That is the case of the interactions between proteins engaged in electron transfer reactions, which are involved in relevant physiological processes such as respiration and photosynthesis. In the latter, the copper protein plastocyanin acts as a soluble carrier transferring electrons between the two membrane-embedded complexes cytochrome b 6 f and photosystem I. Here we review the combination of experimental efforts in the literature to unveil the functional and structural features of the complex between cytochrome fand plastocyanin, which have widely been used as a suitable model for analyzing transient redox interactions. Ó 2011 Federation of European Biochemical Societies. Published by Elsevier B.V. 1. Introduction Protein–protein interactions are key processes in the proper operation of living cells. Different kinds of interactions can be distinguished on the basis of protein binding affinities. Complexes characterized by low binding affinities occur in transient protein–protein interactions, which display high dissociation equilibrium constants ðK D Þ, within the range of l M–mM, or lifetimes in the order of milliseconds [1,2]. Transient complexes are typical of physiological processes that require a compromise between binding specificity and turnover to run at an appropriate rate [3]. On the basis of a two-step model [4], the formation of a final complex first entails a primary unspecific recognition as an encounter that can be transiently stabilized to yield a productive complex as an outcome. Electron transfer (ET) reactions are excellent examples of transient complexes. In fact, soluble proteins mediate redox exchange between large membrane complexes in the photosynthetic and respiratory electron transport chains via short-lived interactions. For instance, in oxygenic photosynthesis there is a soluble metalloprotein that shuttles electrons from cytochrome f(Cf), which is a component of the membrane-embedded cytochrome b 6 f(Cb 6 f) complex, to P700, which is the special chlorophyll pair of Photosystem I (PSI) [5–7]. Either plastocyanin (Pc) or cytochrome c 6 (Cc 6 ) can play such a role of electron carrier between Cfand PSI. Higher plants only contain Pc, whereas most cyanobacteria and green algae synthesize either Pc or Cc 6 depending on the relative availability of copper and iron, their respective cofactor metals [8,9]. Cfis anchored to the thylakoid membrane, within the Cb 6 f complex, by a C-terminal transmembrane helix leaving a 28-kDa N-terminal soluble portion exposed to the lumen with a clear two-domain structure. The large domain harbors the heme group, and the small domain possesses a patch of charged residues. Cfis considered an unusual c-type cytochrome because of its bsheet-based structure, elongated form and particular heme axial 0014-5793 Ó2011 Federation of European Biochemical Societies. Published by Elsevier B.V. doi:10.1016/j.febslet.2011.08.035 Abbreviations: BD, brownian dynamics; Cb 6 f, cytochrome b 6 f;Cc 6 , cytochrome c 6 ;Cf, cytochrome f; CSP, chemical-shift perturbations; ET, electron transfer; EXAFS, extended X-ray absorption fine structure; k 2 , bimolecular rate constant; K D , dissociation equilibrium constant; MD, molecular dynamics; NIR, nitrite reductase; NMR, nuclear magnetic resonance; Pc, plastocyanin; PCS, pseudocontact shifts; pI, isoelectric point; PRE, paramagnetic relaxation enhancement; PSI, Photosystem I; RMSD, root mean square deviation; WT, wild type; XANES, X-ray absorption near edge structure; XAS, X-ray absorption spectrometry ⇑ Corresponding author. Fax: +34 954460065. E-mail address: [email protected] (M.A. De la Rosa). FEBS Letters 586 (2012) 646–652 journal homepage: www.FEBSLetters.org Open access under CC BY-NC-ND license. Open access under CC BY-NC-ND license. coordination with the N-terminus Tyr1 [10,11]. Pc, in its turn, is an 11-kDa cupredoxin with a b-barrel structure formed by eight bstrands and a small a -helix, along with a copper centre coordinated by two histidines, one methionine and one cysteine [12]. The overall structures of both Cfand Pc are highly conserved from cyanobacteria to higher plants [13–15], but striking differences occur in some of their physical properties. The surface near the heme moiety in Cfis mainly hydrophobic for all organisms. However, a remarkable basic ridge is found in the small domain for higher plants and green algae turning to acidic in cyanobacteria. On the other hand, Pc relies on two functional sites: a hydrophobic patch surrounding the solvent-accessible histidine copper ligand, or the so-called ‘‘site 1’’, and an electrostatically charged surface area, the so-called ‘‘site 2’’, whose nature varies from one organism to another. Actually, site 2 is mainly acidic in plants and green algae, whereas it ranges from acidic to basic in cyanobacteria. This feature is a key for the dynamics of the complex and is thus herein reviewed. The Cf–Pc complex has extensively been studied in the last years as a model to understand the nature of protein–protein interactions in ET chains. The main goal of this article is to briefly review not only our current understanding of the mechanism of ET in transient complexes [16], but also the different techniques used to analyze the structural features of such complexes [17] in an ample set of prokaryotic and eukaryotic organisms. 2. Kinetics within the Cf–Pc complex Kinetic analyses provide information about the ET reaction mechanism, namely the limiting step and the nature of the partner interactions. Some years ago, our group proposed three different kinetic mechanisms to analyze the well-related interactions of PSI with Pc and Cc 6 from a wide range of photosynthetic organisms [18–20]. These kinetic models can also be applied to the redox interactions between Cfand the two soluble carriers Pc and Cc 6 [21–23]. Some of the currently available experimental data for the interaction between Cfand Pc are summarized in Table 1, where the values for the following constants are presented: bimolecular rate constant for complex formation ðk 2 Þ, equilibrium constant for association between partners ðK A Þ, and effective electron transfer rate constant ðk 0 et Þ. The interaction between Cfand Cc 6 cannot be analyzed because of spectral overlapping of the two heme proteins. Electrostatics determines the kinetics of the Cf–Pc reaction. In plants, such effect on the ET rate was mainly established by analyzing the interactions of lysine mutants at the basic ridge of turnip Cf with spinach and pea Pc [21]. Those charged residues at the basic Cfpatch are crucial for the in vitro interaction with Pc. Lysines 58, 65 and 187 of Cfdirectly interact with acidic residues at site 2 of Pc. The electrostatic attraction between the two partners results in a bell-shaped reaction rate dependence on ionic strength [16] (Fig. 1,upper panel). This behavior has been explained by assuming that the complex gets ‘locked’ in a non-productive electrostatic orientation at low ionic strength. As the ionic strength increases, the rearrangement of both proteins into the complex takes place in order to achieve a well-oriented and productive complex, which corresponds to the maximum of the bell. Further increase in ionic strength leads to complex dissociation [6]. It has recently been proposed that the expression ‘locked complex’ should be avoided as such a non-productive state at low ionic strength could rather correspond to a set of non-productive orientations slightly different in energy [4]. In vitro kinetic analyses of the interaction between Cfand Pc in cyanobacteria have mainly been addressed in the mesophilic Nostoc [22,24] and thermophilic Phormidium [23,25,26] species. The electron transfer rate constant between the two Nostoc proteins monotonically decreases with increasing ionic strength (Fig. 1,middle panel). The charge mutations at site 2 reduce the ability of Pc to oxidize Cf[22], whereas the mutations at the hydrophobic patch decrease its reactivity towards the heme protein. It is worth to mention that the charge mutation of Arg93 – a residue of Pc located at the interface between sites 1 and 2 – drastically diminishes the reaction rate, so revealing the crucial role of such amino acid in ET within the complex (Table 1). In contrast, the charge replacements at the small domain of Cfhardly affect the interaction with Pc [24] (Table 1), thereby suggesting that the specificity in the Cf–Pc interaction is mostly determined by the cupredoxin. This has been corroborated by NMR studies of the mixed Cf–Pc complexes between Nostoc and Phormidium cyanobacterial proteins [27]. The ET rate constant between the wild-type (WT) forms of the two Phormidium proteins slightly depends on ionic strength (Fig. 1,lower panel), but site-directed mutagenesis of certain charged residues of both proteins revealed a clear influence over the reaction rate (Table 1)[23,25]. Actually, a deeper analysis of the Cf–Pc complex by continuum electrostatics shows that net coulombic forces between protein charges are slightly repulsive [28]. Assuming a diffusion controlled reaction, the electrostatic forces do influence encounter complex formation. The process also involves specific hydrophobic interactions of aromatic residues in the N-terminal peptide of Cf[26]. Altogether, the major contribution to k 2 in higher plants and Nostoc cyanobacterium is electrostatics, which leads to the encounter between proteins and to the formation of a well-defined Table 1 Kinetic data for the redox reaction between WT and mutant forms of Cfand Pc. CfPc k 2 (10 7 M 1 s 1 )k 0 et (10 3 s 1 )K A (10 3 M 1 ) References Turnip WT Spinach WT 17.6 ± 2.2 3 a –[21,22] Turnip K187E Spinach WT 2.5 ± 0.3 – – [21] Turnip WT Pea WT 17.5 ± 0.3 – 6.9 ± 0.18 [21] Turnip K65Q Pea WT 3.5 ± 0.1 – 0.55 ± 0.38 [21] Phormidium WT Phormidium WT 4.7 b –0.3 [23,39] Phormidium WT Phormidium D44A 6.0 b ––[23] Phormidium WT Phormidium R93E 1.0 b ––[23] Phormidium D63A Phormidium WT 3.1 b ––[25] Nostoc WT Nostoc WT – 13.4 26 ± 1 [22,27] Nostoc WT Nostoc D54K – 25.5 – [22] Nostoc WT Nostoc R93E – 1.8 – [22] Nostoc D64A Nostoc WT – 13.3 – [24] Phormidium WT Nostoc WT – – 12 ± 1 [27] Prochlorothrix WT Prochlorothrix WT 20 – 25 ± 2 [40] k 2 , bimolecular rate constant for complex formation; K A , equilibrium constant for association between partners; k 0 et , effective electron transfer rate constant. a French bean Pc was employed. b The overall error of k 2 with all Phormidium variants is estimated to be 65%. I. Cruz-Gallardo et al. / FEBS Letters 586 (2012) 646–652 647 functional complex. The opposite is valid for the thermophilic Phormidium cyanobacterium, in which non-polar interactions play the main role because of the small net contribution of electrostatics [23,26,28]. Nevertheless, the hydrophobic interactions are essential for ET in all complexes, since the reorganization of water molecules at the interface should impair the charge transfer process. In fact, the exclusion of water molecules yields a productive complex within which the ET takes place. Intriguingly, the reaction rates reach similar values under physiological salt concentration and temperature ðk 2 10 8 M 1 s 1 Þdespite the apparent differences among organisms [23]. In vivo approaches are quite scarce. In the alga Chlamydomonas reinhardtii [29,30], neutralization of Cfbasic ridge has a negligible effect on the rate of protein re-oxidation, suggesting that electrostatic interactions within the Cf–Pc complex are not so relevant in vivo in this organism. Experimental conditions mimicking viscosity, macromolecular crowding and temperature have also been explored, but no relevant conclusions can be reached [31]. 3. Defining the relative orientation between Cfand Pc The structure of the Cf–Pc complex, as well as that of the Cf–Cc 6 one, has been analyzed by nuclear magnetic resonance (NMR) spectroscopy [17,32] by using the chemical and pseudo-contact shifts (PCS) experienced by the Pc signals as restraints for rigidbody docking calculations. Diamagnetic, chemical-shift perturbations (CSP) are caused by changes in the chemical environment suffered by the atoms located at the binding interface. Structural restrains derived from CSP do not provide precise geometrical information about angles and distances between the two partners. However, measuring PCS renders geometrical restraints that may be sufficient to define the relative orientation of both proteins. PCS restraints are provided by the intrinsic paramagnetic probe of Cf, which supplies the oxidized iron atom (Fe 3+ ) from the heme phorphyrin ring. PCS depend on the axial and rhombic components of the magnetic susceptibility tensor and the angles that the vector joining the heme Fe atom of Cfto the target amide proton in Pc forms with them. Moreover, PCS are inversely proportional to the cubed distance between any particular nucleus of Pc and the paramagnetic center. Structure elucidation of the Cf–Pc complex for different organisms shown in this section fits well with the orientation of Cfwithin the Cb 6 fcomplex as determined by X-ray crystallography [11,33]. Some similar structural features, like interface areas of ca. 600– 850 Å 2 per protein, are shared by the Cf–Pc complexes from prokaryotic and eukaryotic organisms. All the cases studied show that Cu-ligand His87 in Pc is close to iron-coordinating Tyr1 in Cf, thereby providing an efficient electron transfer pathway [34]. However, remarkable differences are found not only when comparing the structures of plant and cyanobacterial complexes, but also when the comparisons are made between complexes from cyanobacteria (Fig. 1). Plant heterologous complexes between Cffrom turnip and Pc from several other sources like spinach, parsley or poplar have been solved [35–37]. Their overall structures show a binding mode named side-on (Fig. 1,upper panel) in which the hydrophobic patches of the two proteins keep in close contact with each other and the complementary charged regions are juxtaposed. In other words, both Pc sites 1 and 2 are involved in the binding interface. The Fe–Cu distance varies from 10.9 Å in spinach to 13.9 Å in poplar and 13.0 Å in parsley. Actually, poplar and parsley Pc are slightly tilted with regard to the position of spinach Pc [36]. Within the poplar Cf–Pc complex, the approach of the side-chain of His87 from Pc to the heme ring at Cfis restricted by the loop containing this copper ligand. In contrast, the parsley Cf–Pc model shows a rotation of Pc, with respect to the spinach, in a direction opposite to that in poplar complex [37]. The cyanobacterial complexes have been analyzed using both Pc and Cffrom the same source. The structure of the Cf–Pc complex from Nostoc [38] shows a relative orientation that resembles the plant side-on fashion, but with the charges reversed. Actually, the positive charged residues at site 2 of Nostoc Pc – which indeed are determinant in the isoelectric point (pI) of 8.4 for the whole protein – are keeping salt bridges with negative residues at the hinge and small domain of Cf. However, the pIof plant Pc is ca. 4.0 and its negatively charged site 2 interacts with the positive charged residues of Cf. The hydrophobic interactions involving the two metal centers of both proteins are also present (Fig. 1,middle panel). In contrast, Pc in the Phormidium complex is oriented in ahead-on manner relative to Cf, with only site 1 (hydrophobic patch) making contact to Cf(Fig. 1,lower panel). In Phormidium, Cflacks the typical basic ridge, whereas Pc has fewer charged residues at site 2 [39] than Pc from plants and shows a dipole moment aligned with the main molecule axis. This results in the attraction of the copper site towards Cfand repulsion of site 2. The resulting net coulombic term between the two proteins is repulsive, but is mostly compensated by solvent polarization phenomena to yield a small ionic strength effect [28]. The Phormidium complex shows a highly dynamic nature that could explain the lower precision of the solved structure – with an RMSD of 3.7 Å – compared with other Cf–Pc models. It could also be explained by differences in Fig. 1. Comparison of Cf–Pc complexes from different organisms. Right, Superposition of the average structure, with the lowest energy value, calculated from the 10 best NMR structures of each complex. Left, profile of the ionic strength dependence of the observed rate constant in each case. Upper, poplar [36];middle,Nostoc [22,38]; and lower,Phormidium [25,39].Cfand Pc ribbons are in red and blue, respectively. The heme porphyrin ring is depicted in green sticks, whereas the copper and iron atoms are grey spheres. 648 I. Cruz-Gallardo et al. / FEBS Letters 586 (2012) 646–652 the dynamic behavior of the complexes, a fact that is especially significant in those organisms that inhabit niches at high temperatures. Crossed complexes between proteins from Nostoc and Phormidium revealed that it must be a low net electrostatic contribution in the latter complex [27], in agreement with results on ET kinetics. Although early NMR data [39] suggested the absence of electrostatic interactions, several transient conformations involving electrostatics could remain invisible by NMR, at least in the absence of paramagnetic relaxation enhancement measurements (PRE). Therefore, only the long-life conformation into the complex is detectable. In addition, studies of such mixed complexes show that the differences in interactions are mainly attributable to the surface properties of Pc [27]. The NMR analyses (PCS measurements) show that the Cf–Pc complex from Prochlorothrix hollandica, unlike the Nostoc Cf–Pc complex, exhibits a side-on orientation, in which Pc site 2 is not interacting with the small domain of Cf, thereby suggesting a dynamic nature that resembles the Phormidium complex [40]. An overall view of structural features for the well-defined oriented Cf–Pc complexes postulates a predominant side-on conformation highly dependent on electrostatics. As the electrostatic contribution becomes lower, the orientation of Pc relative to Cf shifts from side-on to head-on. All these observations are also in agreement with the ET proposed kinetic mechanisms. 4. Metal cofactors within the Cf–Pc complex Although there are a lot of experimental data on the interaction between the two redox partners, the information on the effect of binding on the metal cofactors and their properties is rather scarce. In fact, the redox potential of Pc decreases in 30 mV upon binding to Cf[41]. Noteworthy, the way on how metal sites can adapt to changes in the protein matrix and modulate the ET reaction highly contributes to understand the transient complexes involved in this kind of processes. The cyanobacterial Nostoc Cf–Pc complex has been analyzed by X-ray absorption spectroscopy (XAS). Both Fe and Cu K-edge XAS measurements of free and bound redox proteins have been studied in solution, using either the oxidized or reduced species [42,43]. In Cf, the Fe atom of the heme group is axially coordinated by two nitrogen atoms belonging to Tyr1 (N) and His26 (N  2 ). Such atypical coordination geometry involves the N-terminal tyrosine, which also takes part in hydrophobic interactions with Pc, according to the structural elucidations of this transient complex [38]. XAS measurements reveal that the Fe coordination geometry remains unaltered upon binding to Pc. Getting a deeper insight into the data, a slight distortion of the metal (Fe 2þ ) center geometry seems to happen when reduced Cfbinds to reduced Pc. The resulting geometry is closer to that of oxidized Cf, either free or bound to oxidized Pc. Because of the smaller size of Fe 3þ compared to Fe 2þ , the first fits better than the second into the heme group as Fe 2þ is slightly out of the plane [44]. However, Fe 2þ could be driven back to the ring upon complex formation. Iron-N-terminus nitrogen bond in Cfis strong enough to prevent distortions over the coordination geometry of the metal center, so facilitating a stable binding site to Pc and enhancing ET. The copper atom in Pc is coordinated by two nitrogen atoms from His39 and His92, as well as by two sulphur atoms from Cys89 and Met97. All of them are well-conserved residues placed at a protein loop that is at the interface with Cf[35,38,39]. Cu–K edge XAS shows a remarkable distortion in the trigonal pyramidal geometry of the copper coordination sphere upon binding to Cf, regardless of its redox state (Fig. 2,upper and middle panel). The main evidence for such a distortion is the contribution of the S d atom from Met97 to the extended X-ray absorption fine structure (EXAFS) wave, indicating that the mobility of this side-chain is substantially restrained upon binding of Pc to Cf. Actually, the resulting tetrahedral structure of the copper center within the Cf–Pc complex exhibits a shorter Cu–Sd(Met97) distance with respect Fig. 2. Distortion of the copper geometry of Pc upon Cfbinding in Phormidium. Upper, EXAFS spectra at the Cu K-edge of oxidized Pc upon binding to oxidized Cf (black) and fitting in the k space either considering that the contribution of S d (Met97) is negligible (red, v 2 : 16.97) or taking it into account (blue, v 2 : 11.3). Middle, Fourier Transform modules of free Pc (blue) and Pc-bound to Cf(red), both in the reduced state. The best fits for the data are depicted in dotted lines following the same color-code. The contribution of the S d (Met97) atom is considered for the fit in R space. Lower, XANES region of the Cu K-edge XAS spectra for free Pc (blue) and Pc-bound to Cf, both in the oxidized state (red). I. Cruz-Gallardo et al. / FEBS Letters 586 (2012) 646–652 649 to the crystallographic structure of free Pc [45]. Noteworthy, the Cu–Sd(Met97) distance of Cf–bound Pc by EXAFS resembles that of the crystal structure of oxidized nitrite reductase (NIR) [46], which belongs to the so-called ‘‘perturbed’’ copper centers. In addition, the electronic density around the copper atom increases when Pc binds to Cfin their oxidized states, according to the data from the X-ray absorption near edge structure (XANES) region (Fig. 2, lower panel). In fact, the redox potential of Pc becomes more negative upon binding to Cf[41]. As a result, the driving force for ET between both metalloproteins is significantly decreased by the Cf–Pc interaction. In addition, the observed geometrical changes in the first coordination sphere of copper within the Cf–Pc complex modulate the electron coupling along the different ET pathways and hence the redox reaction. 5. Theoretical approaches The resulting data from analyzing the weak Cf–Pc complex by means of theoretical methods add useful information to our current knowledge of transient interactions. The Cf–Pc complexes of a wide range of organisms, from cyanobacteria to plants, have been studied by Brownian Dynamics (BD) simulations using crystal or NMR structures. The role of electrostatic interactions in higher plant and algal complexes has been confirmed by BD simulations that are in good agreement with the k 2 values for ET reactions previously reported [21,23,25]. The complex between turnip Cfand spinach Pc [47] and that between Cfand Pc from the green alga Chlamydomonas reinhardtii [48] have both been studied by performing systematic mutations of charged residues at the protein interfaces. These data support the model of an electrostatically driven encounter complex as a previous step of a final productive complex wherein efficient ET occurs. BD analyses with Phormidium Cfand Pc from different sources [49] show considerable variations in electrostatic forces, a finding that highlights the key role of a few charged residues close to His92 (His87 in plants and algae) for the complex interface. However, the computational studies show that Pc is highly mobile in all these complexes, opposite to its behavior in plant complexes. Still, hydrophobic interactions remain highly significant in all organisms. Fig. 3. Molecular dynamics of the Phormidium Cf–Pc complex. (A) Drift of Pc around Cfalong the MD trajectory in Ref. [28]. The dihedral first angle projections of the trajectories of iron from Cf(brown dots), copper (blue dots) and center of mass of Pc (grey dots) are represented with respect to the main axes and mass center of the large domain of Cf. (B) Interaction of the copper site and Arg93 in Pc with negative residues of Cfupon approaching of the two partners. Pc and copper center are in purple, and Cfin gold. The side chains of charged residues and the heme porphyrin ring are depicted in sticks following the same color-pattern. (C) Coordinate covariance matrix of a trajectory corresponding to the Cf–Pc complex obtained after aligning main chain atoms of the large domain of Cf. Residues 1–105 correspond to Pc, and 106–354 to CfPositive values are shown in blue, whereas negative are in red. (D) Analysis of continuum electrostatics averaged along different time intervals of a MD computation of the Cf–Pc complex. The insert summarizes the thermodynamic cycle used in the computation of the different terms in which D G ele-bind stands for the sum of free energies that account for distorting the free monomers to get the conformation they adopt in the complex, D G ele-strain , and their subsequent rigid-body association, D G ele-rigid . D G ele-bind is represented by white bars; D G ele-strain by red-filled bars; D G ele-rigid bars are colored in grey. 650 I. Cruz-Gallardo et al. / FEBS Letters 586 (2012) 646–652 BD analyses are quite sensitive to small changes in conformation of the proteins that are being modeled. For instance, the Cf– Pc complex with spinach Pc adopted two different conformations when using distinct structural data sources [50]. On the basis of such finding, it was suggested that Pc may assume distinct conformations in solution so as to yield a productive ET complex. BD and other docking approaches treat proteins as rigid bodies [40], whereas Molecular Dynamics (MD) calculations provide additional information about internal protein motions. Despite all the kinetic, structural and theoretical studies performed on the Phormidium Cf–Pc complex, there is certain controversy regarding NMR and functional data. Hence, MD simulations combined to continuum electrostatic calculations on this particular interaction [28] try to harmonize the two sets of data. The MD trajectories reveal that Pc tilts towards the small domain of Cfapproaching site 2 to the heme protein (Fig. 3A). This involves the interaction of the positive charges of the copper site and Arg93 with the negatively charged residues at the loop of Cfsurrounding the heme group (Fig. 3B). Thus, the relative orientation of Pc respect to Cfin the Phormidium complex can be redefined as a tilted head-on conformation (see Fig. 1,lower panel). However, the repulsive forces make both proteins swing in an opposite but concerted manner within the transient Cf–Pc complex, as inferred from the negative covariance between motions of Pc and those of the small domain of Cf (Fig. 3C). Interestingly, the conformation of Cfis strained upon binding of its partner and relaxes upon release. Although there are no direct contacts between Pc and the small domain of Cf, continuum electrostatic calculations indicate that the long-range electrostatic interactions between them are responsible for straining the conformation of the partners (Fig. 3D). The thermodynamic cycle shown in Fig. 3D reveals that D G ele-bind can be split in two terms if any one of the partners changes within the complex. The first term is the ‘‘strain energy’’ ð D G ele-strain Þ, which accounts for the cost of structural changes needed for docking; the second one is the ‘‘rigid binding’’ ð D G ele-rigid Þcomponent, which represents the binding energy of the strained partners. Fig. 3D also shows that the time intervals corresponding to the two major ensembles of conformations (from 0 to 3.7 ns, and from 7 to 13 ns) are characterized by a positive D G ele-strain term and by large strain energy values ( D G ele-bind ). However, D G ele-bind becomes negative, with a concomitant drop in D G ele-strain , in the time interval from 4 to 6 ns corresponding to the transition between these two major populations. An explanation for this binding energy decrease comes from the reaction field term that accounts for solvent polarization phenomena. In addition, the electrostatic strain may play a key role in breaking the complex off after the attraction between the copper surroundings and Cfis weakened upon charge transfer. 6. Conclusion and outlook A multidisciplinary effort has been made in the last years to understand the behavior of transient ET complexes and, in particular, of the short-lived Cf–Pc complex from different organisms. In most cases, the functional and structural approaches coincide in postulating the relevance of long-range electrostatic interactions in vitro. The existence of an encounter complex may indeed play an important role in the Cf–Pc complexes, a finding that should be further explored in detail. Some controversy still exists over experimental data interpretation in the cyanobacterial model system from Phormidium, but theoretical simulations are providing valuable information to conciliate them. Actually, recent modeling developments could help to widen the scope of this subject. The bottleneck is to unveil the relevance of these data in vivo. Specifically, the role played by electrostatics on the ET process inside crowded cells needs further research. Acknowledgements The authors wish to thank the Spanish Ministry of Science and Innovation (BFU2009-07190) and the Andalusian Government (BIO198 and P08-CVI-3876) for financial support. References [1] Perkins, J.R., Diboun, I., Dessailly, B.H., Lees, J.C. and Orengo, C. (2010) Transient protein–protein interactions: structural, functional, and network properties. Structure 18, 1233–1243. [2] Janin, J. (2000) Kinetics and thermodynamics of protein–protein interactions in: Protein–protein recognition (Kleanthous, C., Ed.), pp. 1–32, Oxford University Press, Oxford. [3] Bashir, Q., Scanu, S. and Ubbink, M. (2011) Dynamics in electron transfer protein complexes. FEBS J. 278, 1391–1400. [4] Ubbink, M. (2009) The courtship of proteins: Understanding the encounter complex. FEBS Lett. 583, 1060–1066. [5] Allen, J.F. (2004) Cytochrome b 6 f: structure for signalling and vectorial metabolism. Trends Plant Sci. 9, 130–137. [6] Hervás, M., Navarro, J.A. and De la Rosa, M.A. (2003) Electron transfer between membrane complexes and soluble proteins in photosynthesis. Acc. Chem. Res. 36, 798–805. [7] Sun, J., Xu, W., Navarro, J.A., Hervás, M., De la Rosa, M.A. and Chitnis, P.R. (1999) Oxidizing side of the cyanobacterial photosystem I: evidence for interaction between the electron donor proteins and a lumenal surface helix of the PsaB subunit. J. Biol. Chem. 274, 19048–19054. [8] De la Rosa, M.A., Navarro, J.A., Hervás, M., De la Cerda, B., Molina-Heredia, F.P., Balme, A., Murdoch, P.S., Díaz-Moreno, I., Durán, R.V. and Hervás, M. (2002) An evolutionary analysis of the reaction mechanisms of photosystem I reduction by cytochrome c 6 and plastocyanin. Bioelectrochemistry 55, 41–45. [9] Durán, R., Hervás, M., De la Rosa, M.A. and Navarro, J.A. (2004) The efficient functioning of photosynthesis and respiration in Synechocystis sp. PCC 6803 strictly requires the presence of either cytochrome c6 or plastocyanin. J. Biol. Chem. 279, 7229–7233. [10] Martinez, S.E., Huang, D., Szczepaniak, A., Cramer, W.A. and Smith, J.L. (1994) Crystal structure of chloroplast cytochrome f reveals a novel cytochrome fold and unexpected haem ligation. Structure 2, 95–105. [11] Stroebel, D., Choquet, Y., Popot, J.L. and Picot, D. (2003) An atypical haem in the cytochrome b 6 fcomplex. Nature 426, 413–418. [12] Coleman, P.M., Freeman, H.C., Guss, J.M., Murata, M., Norris, V.A., Ramshaw, J.A.M. and Venkatappa, M.P. (1978) X-ray crystal structure analysis of plastocyanin at 2.7 Aresolution. Nature 272, 319–324. [13] Carrel, C.J., Schlarb, B.G., Bendall, D.S., Howe, C.J., Cramer, W.A. and Smith, J.L. (1999) Structure of the soluble domain of cytochrome f from cyanobacterium Phormidium laminosum. Biochemistry 38, 9590–9599. [14] Bond, C.S., Bendall, D.S., Freeman, H.C., Guss, J.M., Howe, C.J., Wagner, M.J. and Wilce, M.C.J. (1999) The structure of plastocyanin from the cyanobacterium Phormidium laminosum. Acta Cryst. D55, 414–421. [15] Chi, Y.I., Huang, L.S., Zhang, Z., Fernández-Velasco, J.G. and Berry, E.A. (2000) Xray structure of a truncated form of cytochrome f from Chlamydomonas reindhardtii. Biochemistry 39, 7689–7701. [16] Hope, A.B. (2000) Electron transfers amongst cytochrome f, plastocyanin and photosystem I: kinetics and mechanisms. Biochim. Biophys. Acta 1456, 5–26. [17] Ubbink, M. (2004) Complexes of photosynthetic redox proteins studied by NMR. Photosynth. Res. 81, 277–287. [18] Hervás, M., Navarro, J.A., Díaz, A., Bottin, H. and De la Rosa, M.A. (1995) Laserflash kinetic analysis of the fast electron transfer from plastocyanin and cytochrome c 6 to photosystem I. Experimental evidence on the evolution of the reaction mechanism. Biochemistry 34, 11321–11326. [19] Hervás, M., Navarro, J.A., Díaz, A. and De la Rosa, M.A. (1996) A comparative thermodynamic analysis by laser-flash absorption spectroscopy of photosystem I reduction by plastocyanin and cytochrome c6 in Anabaena PCC 7119, Synechocystis PCC 6803, and spinach. Biochemistry 35, 2693–2698. [20] De la Cerda, B., Navarro, J.A., Hervás, M. and De la Rosa, M.A. (1997) Changes in the reaction mechanism of electron transfer from plastocyanin to photosystem I in the cyanobacterium Synechocystis sp. PCC 6803 as induced by site-directed mutagenesis of the copper protein. Biochemistry 36, 10125– 10130. [21] Gong, X.S., Wen, J.Q., Fisher, N.E., Young, S., Howe, C.J., Bendall, D.S. and Gray, J.C. (2000) The role of individual lysine residues in the basic patch on turnip cytochrome f for electrostatic interactions with plastocyanin in vitro. Eur. J. Biochem. 267, 3461–3468. [22] Albarrán, C., Navarro, J.A., Molina-Heredia, F.P., Murdoch, P.S., De la Rosa, M.A. and Hervás, M. (2005) Laser flash-induced kinetic analysis of cytochrome f oxidation by wild-type and mutant plastocyanin from the cyanobacterium Nostoc sp. PCC 7119. Biochemistry 44, 11601–11607. [23] Schlarb-Ridley, B.G., Bendall, D.S. and Howe, C.J. (2002) Role of electrostatics in the interaction between cytochrome f and plastocyanin of the cyanobacterium Phormidium laminosum. Biochemistry 41, 3279–3285. [24] Albarrán, C., Navarro, J.A., De la Rosa, M.A. and Hervás, M. (2007) The specificity in the interaction between cytochrome f and plastocyanin from the cyanobacterium Nostoc sp. PCC 7119 is mainly determined by the copper protein. Biochemistry 46, 997–1003. I. Cruz-Gallardo et al. / FEBS Letters 586 (2012) 646–652 651 [25] Hart, S.E., Schlarb-Ridley, B.G., Delon, C., Bendall, D.S. and Howe, C.J. (2003) Role of charges on cytochrome f from the cyanobacterium Phormidium laminosum in its interaction with plastocyanin. Biochemistry 42, 4829–4836. [26] Schlarb-Ridley, B.G., Bendall, D.S. and Howe, C.J. (2003) Relation between interface properties and kinetics of electron transfer in the interaction of cytochrome f and plastocyanin from plants and the cyanobacterium Phormidium laminosum. Biochemistry 42, 4057–4063. [27] Díaz-Moreno, I., Díaz-Quintana, A., De la Rosa, M.A., Crowley, P.B. and Ubbink, M. (2005) Different modes of interaction in cyanobacterial complexes of plastocyanin and cytochrome. Biochemistry 44, 3176–3183. [28] Díaz-Moreno, I., MuñozLópez, F.J., Frutos-Beltrán, E., De la Rosa, M.A. and Díaz-Quintana, A. (2009) Electrostatic strain and concerted motions in the transient complex between plastocyanin and cytochrome f from the cyanobacterium Phormidium laminosum. Bioelectrochemistry 77, 43–52. [29] Soriano, G.M., Ponamarev, M.V., Piskorowski, R.A. and Cramer, W.A. (1998) Identification of the basic residues of cytochrome f responsible for electrostatic docking interactions with plastocyanin in vivo: relevance to the electron transfer in vivo. Biochemistry 37, 15120–15128. [30] Soriano, G.M., Ponamarev, M.V., Tae, G.S. and Cramer, W.A. (1996) Effect of interdomain basic region of cytochrome f on its redox reactions in vivo. Biochemistry 35, 14590–14598. [31] Schlarb-Ridley, B.G., Mi, H., Teale, W.D., Meyer, V.S., Howe, C.J. and Bendall, D.S. (2005) Implications of the effects of viscosity, macromolecular crowding, and temperature for the transient interactions between cytochrome f and plastocyanin from the cyanobacterium Phormidium laminosum. Biochemistry 44, 6232–6238. [32] Crowley, P.B., Díaz-Quintana, A., Molina-Heredia, F.P., Nieto, P., Sutter, M., Haehnel, W., De la Rosa, M.A. and Ubbink, M. (2002) The interactions of cyanobacterial cytochrome c 6 and cytochrome f, characterized by NMR. J. Biol. Chem. 277, 48685–48689. [33] Kurisu, G., Zhang, H., Smith, J.L. and Cramer, W.A. (2003) Structure of the cytochrome b6f complex of oxygenic photosynthesis: tuning the cavity. Science 302, 1009–1014. [34] Díaz-Quintana, A., Hervás, M., Navarro, J.A. and De la Rosa, M.A. (2008) Plastocyanin and cytochrome c6: the soluble electron carriers between the cytochrome b6f complex and photosystem I in: Structure of photosynthetic proteins (Fromme, P., Ed.), pp. 181–200, Wiley-VCH, Weinheim, Germany. [35] Ubbink, M., Ejdeback, M., Karlsson, B.G. and Bendall, D.S. (1998) The structure of the complex of plastocyanin and cytochrome f, determined by paramagnetic NMR and restrained rigid-body molecular dynamics. Structure 6, 323–335. [36] Lange, C., Cornvik, T., Díaz-Moreno, I. and Ubbink, M. (2005) The transient complex of poplar plastocyanin with cytochrome f: effects of ionic strength and pH. Biochim. Biophys. Acta 1707, 179–188. [37] Crowley, P.B., Hunter, D.M., Sato, K., McFarlane, W. and Dennison, C. (2004) The parsley plastocyanin-turnip cytochrome f complex: a structurally distorted but kinetically functional acidic patch. Biochem. J. 378, 45–51. [38] Díaz-Moreno, I., Díaz-Quintana, A., De la Rosa, M.A. and Ubbink, M. (2005) Structure of the complex between plastocyanin and cytochrome f from the cyanobacterium Nostoc sp. PCC 7119 as determined by paramagnetic NMR. J. Biol. Chem. 280, 18908–18915. [39] Crowley, P.B., Otting, G., Schlarb-Ridley, B.G., Canters, G.W. and Ubbink, M. (2001) Hydrophobic interactions in a cyanobacterial plastocyanin-cytochrome f complex. J. Am. Chem. Soc. 123, 10444–10453. [40] Hulsker, R., Baranova, M.V., Bullerjahn, G.S. and Ubbink, M. (2008) Dynamics in the transient complex of plastocyanin and cytochrome f from Prochlorothrix hollandica. J. Am. Chem. Soc. 130, 1985–1991. [41] Malkin, R., Knaff, D.B. and Bearden, A.J. (1973) The oxidation–reduction potential of membrane-bound chloroplast plastocyanin and cytochrome f. Biochim. Biophys. Acta 305, 675–678. [42] Díaz-Moreno, I., Díaz-Moreno, S., Subías, G., De la Rosa, M.A. and DíazQuintana, A. (2006) The atypical iron-coordination geometry of cytochrome f remains unchanged upon binding to plastocyanin, as inferred by XAS. Photosynth. Res. 90, 23–28. [43] Díaz-Moreno, I., Díaz-Quintana, A., Díaz-Moreno, S., Subías, G. and De la Rosa, M.A. (2006) Transient binding of plastocyanin to its physiological redox partners modifies the copper site geometry. FEBS Lett. 580, 6187–6194. [44] Schnackenberg, J., Than, M.E., Mann, K., Wiegand, G., Huber, R. and Reuter, W. (1999) Aminoacid sequence, crystallization and structure determination of reduced and oxidized cytochrome c 6 from the green alga Scenedesmus obliquus. J. Mol. Biol. 290, 1019–1030. [45] Schmidt, L., Christensen, H.E.M. and Harris, P. (2006) Structure of plastocyanin from the cyanobacterium Anabaena variabilis. Acta Cryst. D 62, 1022–1029. [46] Adman, E.T., Godden, J.W. and Turley, S. (1995) The structure of copper-nitrite reductase from Achromobacter cycloclastes at five pH values, with NO 2 bound and with type II copper depleted. J. Biol. Chem. 270, 27458–27474. [47] De Rienzo, F., Gabdoulline, R.R., Menziani, M.C., De Benedetti, P.G. and Wade, R.C. (2001) Electrostatic analysis and Brownian dynamics simulation of the association of plastocyanin and cytochrome f. Biophys. J. 81, 3090–3104. [48] Haddadian, E.J. and Gross, E.L. (2005) Brownian dynamics study of cytochrome f interactions with cytochrome c 6 and plastocyanin in Chlamydomonas reinhardtii:plastocyaninandcytochromec 6 mutants.Biophys.J.88,2823–2839. [49] Gross, E.L. and Rosenberg, A. (2006) A Brownian dynamics study of the interaction of Phormidium cytochrome f with various cyanobacterial plastocyanins. Biophys. J. 90, 366–380. [50] Gross, E.L. (2007) A Brownian dynamics computational study of the interaction of spinach plastocyanin with turnip cytochrome f: the importance of plastocyanin conformational changes. Photosynth. Res. 94, 411–422. 652 I. Cruz-Gallardo et al. / FEBS Letters 586 (2012) 646–652