Structural Details of Light Activation of the LOV2-based Photoswitch

Nov 4, 2014 - (8) After the final gel filtration step in buffer A (10 mM Tris, pH 8.5, 20 mM NaCl, 5 mM MgCl2, and 2 mM DTE) concentrated aliquots wer...
0 downloads 0 Views 4MB Size
Download PDF
Articles pubs.acs.org/acschemicalbiology

Structural Details of Light Activation of the LOV2-based Photoswitch PA-Rac1 Andreas Winkler,† Thomas R. M. Barends,† Anikó Udvarhelyi,† Daniel Lenherr-Frey,† Lukas Lomb,† Andreas Menzel,‡ and Ilme Schlichting*,† †

Department of Biomolecular Mechanisms, Max Planck Institute for Medical Research, Jahnstrasse 29, 69120 Heidelberg, Germany Paul Scherrer Institute, PSI, 5232 Villigen, Switzerland

‡

ACS Chem. Biol. 2015.10:502-509. Downloaded from pubs.acs.org by KAOHSIUNG MEDICAL UNIV on 07/11/18. For personal use only.

S Supporting Information *

ABSTRACT: Optical control of cellular processes is an emerging approach for studying biological systems, affording control with high spatial and temporal resolution. Specifically designed artificial photoswitches add an interesting extension to naturally occurring light-regulated functionalities. However, despite a great deal of structural information, the generation of new tools cannot be based fully on rational design yet; in many cases design is limited by our understanding of molecular details of light activation and signal transduction. Our biochemical and biophysical studies on the established optogenetic tool PA-Rac1, the photoactivatable small GTPase Rac1, reveal how unexpected details of the sensor−effector interface, such as metal coordination, significantly affect functionally important structural elements of this photoswitch. Together with solution scattering experiments, our results favor differences in the population of pre-existing conformations as the underlying allosteric activation mechanism of PA-Rac1, rather than the assumed release of the Rac1 domain from the caging photoreceptor domain. These results have implications for the design of new optogenetic tools and highlight the importance of including molecular details of the sensor−effector interface, which is however difficult to assess during the initial design of novel artificial photoswitches.

I

available tools conveys the notion that the design of these systems should be straightforward. However, the generation of novel tools is far from a rational approach and usually requires substantial screening efforts. In the case of PA-Rac1, the availability of a crystal structure8 has raised hopes that this system might serve as a blueprint for caging related GTPases. Unexpectedly, however, the successful concept of the designed photoswitch PA-Rac1 could not even be transferred to closely related GTPases. In the case of Cdc42, it was found that mutations at the sensor−effector interface could stabilize the caged dark-state conformation but still did not render this tool practically useful. For the related GTPase, RhoA, no functional photoswitch could be obtained despite extensive efforts based on the concepts learned from PA-Rac1 (personal communication, K. Hahn, UNC Chapel Hill). This experience shows that the successful design of artificial photoswitches should consider not only the activation mechanisms of the light sensor or the native functionalities of the regulated protein but also the molecular details of the engineered sensor−effector interface and the structural dynamics of the dark−light transition. Therefore, we decided to reinvestigate the LOV2−Rac1 interface of PA-Rac1 and to

nterest in light-regulated proteins has recently surged due to the emerging technique of optogenetics: the use of genetically encoded proteins to target biological systems for optical control of cellular processes. Prominent examples are channelrhodopsin1 and halorhodopsin2 and light-activated adenylate cyclase,3,4 targeting ion flux and production of the second messenger cyclic adenosine monophosphate (cAMP), respectively. However, the demand for light-controlled systems goes beyond the possibilities provided by these and other naturally occurring photoreceptors. For this reason, a range of artificial photoswitches have been designed to expand the optogenetic toolbox (recently reviewed in ref 5). A frequently applied photoreceptor in this approach is the LOV2 domain of phototropin 1 from Avena sativa.6−15 Early NMR studies of this photoswitch demonstrated the light-induced unfolding of the C-terminal α-helical region (termed Jα),16 and based on this concept, a variety of optogenetic tools were developed following the concept of “caged” functionalities. Currently available LOV2-based tools allow, for example, light control of DNA binding7 and modulation of enzymatic6 or peptidic9,13,15 activities, as well as light regulated protein degradation12 or induction of apoptosis.10 Another light switch designed using the LOV2 domain, the photoactivatable GTPase Rac1 (PA-Rac1), was successfully employed to control the motility of living cells through the light-regulated interaction of Rac1 with the p21-activated kinase (PAK).8 The diversity of © 2014 American Chemical Society

Received: September 16, 2014 Accepted: November 4, 2014 Published: November 4, 2014 502

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509

ACS Chemical Biology

Articles

a central element of the functionally important switch I region of this GTPase.18 While calcium binding is due to high Ca2+ concentrations in the crystallization conditions, PISA19 analysis indicates that the presence of calcium at the interface stabilizes the corresponding complex interface as indicated by an increased complexation significance score of the individual domains. To experimentally test this predicted calcium-induced stabilization, we performed thermal denaturation scans analyzed by CD spectroscopy. In the presence of 10 mM CaCl2, we observed an increase in melting temperature by 2 °C together with an increase in the cooperativity of unfolding (Supplementary Figure 1a, Supporting Information). This supports a role of calcium in stabilizing the PA-Rac1 assembly, and therefore, we probed the affinity of calcium binding by microscale thermophoresis (MST, Supplementary Figure 1b, Supporting Information). The measured dissociation constant is 460 ± 90 μM, which is higher than the free calcium concentration in the cytosol of cells. However, the site of action of PA-Rac1 to induce lamellipodia is the plasma membrane, where local concentrations of Ca2+ have been reported to be in the high micromolar range.20 Importantly, this high concentration is required for orchestrating actin regulation and the migration behavior of cells,21,22 which is directly related to the physiological target function of PA-Rac1. The PA-Rac1 system was designed based on NMR data16 of the isolated LOV2 domain that suggested an unwinding of the Jα helix upon light activation. Thus, it was expected that the sensor−effector interface of PA-Rac1 dissociates upon lightinduced unraveling of the Jα helix, exposing the Rac1 binding site for interaction with the Cdc42/Rac interactive binding (CRIB) motif of PAK8 and other Rac effectors. To test this model and to identify functionally relevant details of the Rac1− LOV2 interface, we performed hydrogen−deuterium exchange measurements coupled to mass spectrometry (HDX-MS) comparing deuterium uptake of the isolated LOV2 domain (LOVmf,d,; see naming convention in the Methods section) with deuteration of PA-Rac1 in the dark (PAmf,d, Figure 2a). The measurements show reduced deuterium incorporation of the Jα helix in the PA-Rac1 complex due to interactions of the LOV domain with Rac1 (Figure 2b) and, in addition, reveal a reduction in conformational dynamics of the structural element providing the cysteinyl ligand (Cys450) of the flavin photoproduct and the preceding Dα helix (Figure 2c), which is part of the sensor−effector interface. Interestingly, the bimodal exchange behavior of Jα peptides, indicated by the broad distribution of deuterated species with two distinct peaks (Figure 2b, lower panels), indicates a characteristic exchange regime, which corresponds to two conformational states with significantly different deuterium exchange kinetics (EX1 kinetics).24 While the rapid deuterium uptake of Jα residues of the isolated LOV2 domain is in line with results obtained by NMR spectroscopy16 and is indicative of the inherent flexibility of this structural element, the conformational state with reduced deuterium incorporation likely corresponds to the caged assembly of LOV2 and Rac1. This reveals that the Jα helix of PA-Rac1 significantly populates an unstructured conformation that features deuterium incorporation kinetics resembling those of the isolated LOV2 domain. Interestingly, the exchange kinetics of the N-terminal A′α helix appear to be related to those of Jα, which supports the functional interplay of these structural elements as suggested previously.25,26 The length of A′α is significantly reduced in the artificial PA-Rac1 switch and compromised by a

characterize the effect of light activation. Our results demonstrate that efficient caging of the effector domain is influenced significantly by molecular details of the artificially generated sensor−effector interface. In addition, we provide structural details of light-activated PA-Rac1, which indicate only moderate global structural changes instead of the assumed release of the caged Rac1 domain from LOV2 upon illumination. This provides important insight in the conceptual rationale for designing functional photoswitches. Results from time-resolved solution scattering studies and molecular details of the LOV2−Rac1 interface in PA-Rac1 may also be important for understanding phototropin function in general.

■

RESULTS AND DISCUSSION Since it was previously observed that mutations at the LOV core-Jα9,17 or sensor−effector8 interface can substantially affect the dynamic range of LOV2-based systems and hence influence the success of these photoswitches, we had a closer look at the molecular interface of LOV2 and Rac1. Interestingly, one of the published PA-Rac1 structures8 (pdb 2wkp) features a metal ion in a prominent position of the sensor−effector interface (Figure 1). This calcium ion is coordinated by two carbonyl oxygens of LOV2 residues Glu443 and Leu446 and by the Asp581 side chain of PA-Rac1 (Asp38 Rac1 numbering, Figure 1b), which is

Figure 1. Structural details of PA-Rac1. (a) Overview of the PA-Rac1 structure (pdb 2wkp)8 showing the N-terminal LOV2 core domain in blue, the Jα helix in red, and Rac1 in green. Flavin mononucleotide (FMN) and guanosine triphosphate (GTP) are shown as stick models in orange and gray, respectively. Magnesium and calcium ions are shown as gray and purple spheres. (b) Close-up of the calcium binding site at the sensor−effector interface. Coordinating residues that provide four of the eight oxygen ligands are shown as stick models. The remaining water ligands are represented as red spheres. The isomorphous Fo−Fo difference electron density map calculated from structure factors of 2wkp−2wkq8 contoured at 5σ is shown in magenta. 503

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509

ACS Chemical Biology

Articles

above (Figure 2b), calcium coordination also affects the conformational dynamics of peptides in helical LOV elements surrounding the flavin cofactor and in the switch I region of Rac1 (Figure 3a). Interestingly, while both of the latter regions are in close proximity to the calcium binding site (Figure 1b), the effect of illumination on the switch I region differs in the presence of the metal ion (Figure 3b,c). The observed increase of conformational dynamics in this region upon calcium binding likely influences the dynamic range of this lightregulated system since the effect on deuterium incorporation is more pronounced in the dark state of PA-Rac1 than in the light-activated state (see Figure 3a and Supplementary Figure 2c, Supporting Information). Importantly, light activation of PA-Rac1 results in reversal of deuteration characteristics of the caged Jα conformation (Figure 3b,c and Figure 4a−c), and due to the initial stabilization upon metal coordination, this effect is stronger for calcium-bound PA-Rac1. Interestingly, this amplification of the conformational changes upon light irradiation in the presence of Ca2+ propagates also to other functionally important elements such as the switch II region of PA-Rac1 (Figure 4d; residues 602−617, Rac1 numbering 59− 74) and the LOV2−Rac1 linker region (Figure 4e). In addition, illumination also results in increased conformational dynamics of the Gβ and Hβ regions preceding the functionally important Phe494 residue of LOV2.30 This β-sheet region is also of central importance for LOV signaling in general, as recently reviewed in ref 31. Plotting the observed changes in deuterium incorporation onto the structure of PA-Rac1 illustrates the localized effects of metal binding (Figure 4a) and illumination (Figure 4b) on the LOV2−Rac1 interface. While this could be interpreted in terms of the original mechanism of light-induced interface dissociation, the lack of full reversal in deuteration characteristics compared with that of the isolated LOV domain, e.g. in helices Dα and Eα, strongly suggests that LOV2 and Rac1 do interact even in the light-activated state. This is further supported by a weak but significant effect of calcium binding on interface elements such the domain linker and switch I regions in the light state of PA-Rac1 (Supplementary Figure 2c, Supporting Information). An overview of all HDX-MS peptides is shown in Supplementary Figures 3−5, Supporting Information. Animations of the time-dependence of changes in deuterium incorporation corresponding to Figure 4a,b are shown in Supplementary Movies 1 and 2, respectively, Supporting Information. To probe the effect of illumination on the structure of PARac1, we performed small-angle X-ray scattering (SAXS) experiments. Envelope reconstructions of dark-state data are in very good agreement with the crystal structure (Figure 5a). Upon light activation, no change in oligomerization state was observed, and the light-state envelope appears only slightly more elongated (Figure 5b), which indicates a subtle increase of independent two domain characteristics but no complete dissociation of the LOV2−Rac1 interface (Supplementary Figure 6a−c, Supporting Information). Control experiments show full reversibility of this light−dark transition with darkstate recovery kinetics similar to previous reports8 (Supplementary Figure 7a,b, Supporting Information). Interestingly, the millisecond time-resolved light−dark difference spectra (Figure 5c) show no direct transition from dark to light state but feature characteristics of a transient intermediate. The extracted compound spectra after singular value decomposition (SVD) analysis correspond to the dominating light−dark difference spectrum and to a transient low-q depression (Figure

Figure 2. HDX-MS measurements of LOV2 and PA-Rac1. (a) Overview of LOV2 peptides and changes in their relative deuterium incorporation (ΔDrel) upon interaction with Rac1 (Drel of PAmf,d − Drel of LOVmf,d). Each box corresponds to one peptide and contains five different colors according to the legend for the incubation times of 10, 45, 180, 900, and 3600 s (bottom up); see example peptide in the top left corner. MS/MS confirmed peptides are indicated by white diamonds. Secondary structure elements are taken from DSSP (Define Secondary Structure of Proteins) analysis of PA-Rac1 (pdb 2wkp).8 (b) Deuterium uptake curves of a representative Jα peptide with Drel plotted against the deuteration time for three different experiments. The lower panels show the software estimated abundance distribution of individual deuterated species on a scale from undeuterated to all exchangeable amides deuterated.23 (c) Deuterium uptake curves of the peptide including helices Dα and Eα with Drel plotted against the deuteration time for three different experiments. Drel values are shown as the mean of three independent measurements, and error bars correspond to the standard deviation.

cloning artifact, which might explain the comparatively small amplitude of the signals. However, it was recently shown for longer LOV2 constructs that this element is involved in dimerization, suggesting that the observed Jα−A′α coupling might be functionally relevant in the context of full-length phototropin. Since calcium binding involves residues of the functionally relevant switch I region at the sensor−effector interface of PA-Rac1, we addressed the influence of the metal ion on the caging properties of PA-Rac1. Importantly, the inclusion of calcium during HDX measurements of PA-Rac1 (PACa,d) results in a substantial increase of the conformational substate corresponding to the caged assembly (Figure 2b and Supplementary Figure 2a, Supporting Information). Considering the importance of the Jα helix for LOV2-based photoswitches,17,25,27−29 this suggests a functional role of metal binding at the sensor−effector interface of PA-Rac1, since no stabilizing effect of metal binding is observed in the isolated LOV2 domain (Supplementary Figure 2b, Supporting Information). In terms of the functioning of PA-Rac1, the observed influence of calcium on caging raises the question whether the metal coordination also influences other functionally important regions. Therefore, we compared HDX-MS experiments of metal-free and calcium-bound PA-Rac1 in the dark and during constant blue light illumination. In addition to the pronounced effect of metal binding on Jα peptides in the dark discussed 504

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509

ACS Chemical Biology

Articles

Figure 3. Effect of metal binding and illumination on full-length PA-Rac1. Each box corresponds to one peptide and contains five different colors according to the legend for the incubation times of 10, 45, 180, 900, and 3600 s (bottom up). MS/MS confirmed peptides are marked with diamonds. Secondary structure elements are taken from DSSP analysis of PA-Rac1 (pdb 2wkp). (a) Effect of calcium binding on PA-Rac1 (Drel of PACa,d − Drel of PAmf,d). (b) Illumination of PA-Rac1 in the presence of calcium (Drel of PACa,l − Drel of PACa,d). (c) Illumination of metal-free PARac1 (Drel of PAmf,l − Drel of PAmf,d).

Figure 4. Comparison of structural elements involved in metal binding and illumination of PA-Rac1. (a, b) Structures of PA-Rac1 colored according to the differences in relative deuteration levels upon calcium coordination in the dark and illumination in the presence of calcium ions, respectively. Coloring corresponds to the raw data shown in Figure 3 for the 45 s time point according to ref 23. (c) Deuterium uptake plot of a Jα peptide with Drel plotted against the deuteration time for three different experiments. (d) Deuterium uptake curves of a representative switch II region peptide. (e) Deuterium uptake plot of a LOV2−Rac1 linker peptide.

5d). Additional experiments are required to better characterize this intermediate species, which is populated on the millisecond time scale and might correspond to a spectroscopically silent LOV2 intermediate discussed previously32−34 in the context of a full-length photoswitch. The light−dark difference spectrum

on the other hand is remarkably similar to the calculated difference curves of substructures belonging to a specific normal mode calculated for the PA-Rac1 structure (Supplementary Figure 8a,b, Supporting Information). Extended substructures of the corresponding clam-shell opening also fit 505

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509

ACS Chemical Biology

Articles

that the functionality of the system may be calcium dependent, since the effector interaction regions of Rac1 overlap with the Ca2+-binding site of PA-Rac1 (Supplementary Figure 9, Supporting Information). However, calcium concentrations are tightly regulated in cells and can differ in various compartments making it difficult to control the dynamic range of PA-Rac1 by modulating Ca2+-levels. Future experiments addressing the calcium dependency of the affinities of various Rac1 effectors could, however, help in understanding the molecular basis with respect to the difficulties of caging other small GTPases. The characteristic bimodal exchange regime of Jα peptides supports an underlying activation mechanism of PA-Rac1 with a population shift of two pre-existing conformations that is characteristic for a two-state allosteric behavior37 and in line with previous results obtained for the isolated LOV2 domain.38 In this context, the classical “caging” description of LOV2-based optogenetic tools might appear misleading, since the observed caging effect strongly depends on system specific details and the experimental conditions. In fact, molecular details of the activation mechanism are not known for many LOV2containing optogenetic tools, and therefore it appears difficult to differentiate between systems designed based on the idea of allosteric modulation6,7,39 or caging of specific sequence elements.9,11,12 Since functionality in the context of optogenetic tools greatly depends on molecular properties of the interaction interface, a regulation mechanism following modern concepts of two-state allosteric behavior37,40 also rationalizes the issues encountered in caging related GTPases. In contrast to the full dissociation model and the concept of uncaging upon illumination, such an allosteric mechanism is clearly more difficult to translate from one system to another. However, the versatility of the LOV domain in naturally occurring systems41 and the concept of allosteric surface sites for regulation of diverse functionalities6,39 should enable the design of many interesting artificial photoswitches in the future, even though rational approaches might appear difficult. In fact, there is a substantial potential for improving the dynamic range of initially inefficient LOV2-based photoswitches by modulating the flavin environment,42 by altering the LOV core interaction with Jα9,17 or by considering molecular details of the sensor−effector interface.8,9 Identifying the appropriate mutations can, however, be difficult from a rational perspective.43 While functional tools based on the same photoreceptor will have parallels in their activation mechanisms, the molecular details of signal transduction to the various effector domains can differ substantially. The characterization of these molecular mechanisms might also provide interesting insight into the functioning of the photoreceptor in its natural context, revealing how molecular details of the sensor−effector interface can significantly modulate the functionality of the photoswitch.

Figure 5. Solution scattering data of PA-Rac1. (a, b) Envelope reconstructions of dark- and light-state solution scattering curves, respectively. Superimposed are structural models of PA-Rac1. The dark-state envelope is aligned to the structure shown in Figure 1a. Panel b features the most open substructure of the normal mode shown in Supplementary Figure 8, Supporting Information. (c) Experimental time-resolved difference spectra are shown for selected time points as indicated in the legend. (d) Light minus dark difference spectrum (red) obtained from SVD analysis of time-resolved solution scattering experiments. The low-q depression spectrum (black) corresponds to a transiently populated species, required to fit the time-dependent SAXS spectra. The inset shows a close-up of the low-q region on the same scale as Supplementary Figure 7b, Supporting Information.

the light-state envelope reconstructions better (Figure 5b). In combination with the HDX data, this finding supports an activation model of PA-Rac1 where an increased flexibility of the Jα helix results in a more open sensor−effector interface exhibiting increased conformational dynamics that promote the interaction with the CRIB domain of PAK. Along this line, the importance of conformational dynamics in determining binding affinities is emerging as a general concept of protein functioning.35,36 Conclusions and Implications. PA-Rac1 is the archetype of a rationally designed photoactivatable effector protein that enables precise spatiotemporal photocontrol of protein activity in living cells. It consists of a fusion of Rac1 and the photoreactive LOV2 domain from phototropin, and its design was based on the idea that Rac1 interactions are sterically blocked until irradiation unwinds the Jα helix linking the LOV and Rac1 domains.8 The difficulties in caging related small GTPases suggested that this model may be too naive and prompted us both to revisit the Rac1−LOV2 interface and to investigate its light-induced structural changes. Our study reveals a previously undescribed serendipitously introduced calcium binding site at the Rac1−LOV2 interface and demonstrates its importance for functional aspects of this photoswitch. Solution scattering and HDX-MS studies support the idea that the sensor−effector interface does not fully dissociate upon light activation and suggest that a modulation of the conformational dynamics of the Rac1 interaction site with downstream effectors is sufficient to obtain a functional photoswitch. This observation not only helps explain the difficulties in caging other GTPases but is of general importance for the design of LOV2-based artificial photoreceptors for applications in optogenetics. As far as the in vivo application of PA-Rac1 is concerned, the results also suggest

■

METHODS

Sample Preparation. PA-Rac1 was expressed and purified as described previously.8 After the final gel filtration step in buffer A (10 mM Tris, pH 8.5, 20 mM NaCl, 5 mM MgCl2, and 2 mM DTE) concentrated aliquots were stored at −80 °C until further use. For experiments addressing the effect of calcium binding, the protein was buffer exchanged to buffer B (10 mM Tris, pH 8.5, 20 mM NaCl, and 2 mM DTE) using NAP5 columns (GE Healthcare) according to the manufacturer’s instructions. This gel filtration step does not remove the magnesium ion coordinated by GTP and Rac1 residues, as evidenced previously by the crystal structure of buffer-exchanged 506

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509

ACS Chemical Biology

Articles

protein.8 The isolated LOV2 domain construct encompasses residues 392−546 of PA-Rac1. It was PCR amplified from the PA-Rac1 construct in pQE-308 using the primer pair 5′-GCTGGGATCCTTGGCTACTACACTTGAACG-3′ and 5′-GCTAGCTAAGCTTTAAAGTTCTTTTGCCGCC-3′ and subsequently cloned into pQE-30 using the restriction enzymes BamHI and HindIII (bold face). Protein expression and purification followed the procedures described for PARac1 using buffer B for the final gel filtration step. Microscale Thermophoresis (MST). The affinity of calcium to buffer exchanged PA-Rac1 was determined by MST using the Monolith NT.115 instrument (Nanotemper). PA-Rac1 was randomly labeled at amine positions using the red fluorescent dye NT647 (Nanotemper) according to the instructions provided by the manufacturer; the labeling reaction was performed in 10 mM HEPES, pH 7.5, 20 mM NaCl buffer. Different concentrations of CaCl2 (125 mM to 3.8 nM) in buffer B were mixed with equivalent volumes of labeled PA-Rac1 in buffer B supplemented with 0.1% (w/ v) Tween-20 and 0.2 mg mL−1 bovine serum albumin, resulting in a final PA-Rac1 concentration of 100 nM. Standard treated capillaries (Nanotemper) were used for all experiments. Data from two independent experiments and three capillary positions each were averaged and evaluated using the quadratic equation of the law of mass action. Circular Dichroism (CD) Spectroscopy. Thermal unfolding of PA-Rac1 in the absence and presence of calcium was followed at 223 nm using a J-810 spectropolarimeter (Jasco). Measurements were performed with 13 μM protein in 10 mM MES, pH 7.5, 1 mM DTE, using cuvettes with 1 mm path length and a heating rate of 1 °C min−1. CaCl2 (10 mM) was included in the buffer to address the influence of calcium binding. HDX-MS. We performed six deuterium labeling experiments to address the influence of calcium binding to PA-Rac1, the effect of illumination of the photoswitch, and the caging properties of the LOV2 domain. Individual experiments are referred to as PAmf,d, PAmf,l, PACa,d, PACa,l, LOVmf,d, and LOVCa,d to differentiate between experiments performed without calcium (metal-free, mf), in the presence of calcium (Ca) and under dark (d) or blue light (l) conditions. Protein samples were prepared under safe light conditions at final concentrations of 200 μM for PA-Rac1 and 200 μM for LOV2. Buffer B was used for metal-free experiments, while addressing calcium binding required the addition of CaCl2 to a final concentration of 10 mM (buffer C). Aliquots of 2 μL were equilibrated at 20 °C under dark or blue light conditions for 1 min prior to the labeling reaction. A royal blue (455 nm) collimated LED lamp (Thorlabs) providing a light intensity of 1 mW cm−2 at the sample position was used to populate the light activated state. The corresponding light and temperature conditions were maintained during the labeling reactions, which were prepared in triplicate for all experiments. Deuterium incorporation was initiated by addition of 38 μL of buffer BD or CD (corresponding to buffers B and C prepared with D2O and considering the D2O correction for a final pD of 8.5 at 20 °C) and 6 μL aliquots were removed after 10 s, 45 s, 3 min, 15 min, and 60 min. The labeling step was terminated by quenching with 56 μL of ice-cold 200 mM ammonium formic acid, pH 2.6, and 55 μL was injected into a cooled HPLC setup, as described previously.44 Briefly, deuterated samples were digested on an immobilized pepsin column (Poroszyme, Life Technologies) operated at 10 °C, and resulting peptides were desalted on a 2 cm C18 guard column (Discovery Bio C18, Sigma). Separation of peptides was achieved during a 7 min acetonitrile gradient (15−50%) in the presence of 0.6% (v/v) formic acid on a reversed phase column (XR ODS 75 mm × 3 mm, 2.2 μm; Shimadzu). Eluting peptides were infused into a maXis electrospray ionization−ultra high resolution−time-of-flight mass spectrometer (Bruker). Deuterium incorporation was analyzed and quantified using the Hexicon 2 software package23 (http://hx2.mpimfheidelberg.mpg.de). Small Angle X-ray Scattering (SAXS). Time-resolved solution scattering experiments were performed at the cSAXS beamline X12SA (Swiss Light Source, Villigen, Switzerland). Protein concentrations in the range of 170 to 690 μM of PA-Rac1 were measured in buffer A.

Samples were mounted as described previously.45 Briefly, 1 mm diameter quartz capillaries were filled under red-light conditions and kept at 10 °C throughout the experiment. X-ray exposures of 30 ms, using 12.4 keV photons and a beam size of 250 μm, were recorded on a PILATUS 2M detector operated at a readout time of 5 ms and positioned 2.2 m from the sample. Forty exposures totaling 1.4 s were performed per capillary position, and ∼65 positions were sampled in 500 μm steps along the capillary. During measurements, the sample was illuminated with a blue laser (3.5 mW at 443 nm at the capillary position; 56RCS laser series, Melles Griot) line-focused to a width of 250 μm overlapping with the X-ray beam interaction volume of the capillary. After 200 ms delay time, the laser was turned on for 200 ms and then switched off for the remaining scan time. The overall protocol was repeated 4 times per capillary after a 7 min waiting time between iterations to allow for dark recovery of the protein sample. This measurement protocol without laser irradiation and waiting times was also performed with buffer alone for subsequent buffer subtraction. Guinier plots of each capillary scan were used to assess sample quality and stability upon X-ray and laser irradiation. No signs of aggregation or sample deterioration were observed following the protocol described above. For analysis of the SAXS data, all diffraction images acquired from five independent capillaries, each at the lowest PA-Rac1 concentration used in the experiments, were azimuthally integrated and averaged in line with the acquisition protocol after correction for the buffer contribution. Difference spectra were calculated by subtracting the averaged first five dark exposures from each acquisition. This resulted in difference signals for 35 time points after onset of illumination averaged from a total of 1299 images each. Analysis of the timeresolved difference spectra using SVD with the boundary condition of non-negative amplitudes for individual components (Global Kinetik Explorer,46 KinTek Corporation) was used for extracting the characteristic light-state difference radial distribution. In addition, a transient intermediate on the milliseconds time scale corresponding to a low-q depression was observed. The scattering vector q is defined as q = 4π sin(θ) λ−1. For control experiments addressing the reversibility of the light-dark transition and the dark-state recovery of PA-Rac1, seven exposures were performed per capillary position including an initial dark-state spectrum followed by a 500 ms blue light illumination and six time points during the dark-state recovery (16, 32, 64, 128, 256, and 512 s). This measurement protocol without laser irradiation and delay times was also performed for buffer alone to allow subsequent buffer subtraction. Data analysis followed the outline described above adapted to the differences in the experimental protocol. Envelope reconstructions based on chain-like dummy residues were calculated with GASBORi 2.247 after evaluating the particle distance distribution function using GNOM 4.6.48 Standard settings of GNOM were used, and the maximal particle diameters for dark- and light-state data were set to 94 and 104 Å, respectively, since these values provided a stable solution for the regularization parameter ALPHA. GASBOR options included P1 symmetry and 340 dummy residues in the asymmetric unit. Normal Mode Analysis (NMA). Normal modes of PA-Rac1 were calculated using the NOMAD-ref Web server.49 The wild-type PARac1 crystal structure (pdb 2wkp)8 was used as input with alternate conformations removed and ligands (FMN, GTP, Mg2+) in the pdb file redefined as ATOM entries. Radial density distributions of the resulting substructures were calculated using CRYSOL 2.850 with standard parameters and compared with the experimental scattering curves.

■

ASSOCIATED CONTENT

S Supporting Information *

Calcium binding to PA-Rac1, comparison of HDX characteristics of LOV2 peptides, overview of peptides evaluated during HDX analysis of LOV2 and PA-Rac1, SAXS data of dark- and light-state PA-Rac1, control experiments for the PA-Rac1 solution scattering studies, NMA of PA-Rac1, comparison of 507

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509

ACS Chemical Biology

Articles

(13) Schmidt, D., Tillberg, P. W., Chen, F., and Boyden, E. S. (2014) A fully genetically encoded protein architecture for optical control of peptide ligand concentration. Nat. Commun. 5, No. 3019. (14) Fukuda, N., Matsuda, T., and Nagai, T. (2014) Optical control of the Ca2+ concentration in a live specimen with a genetically encoded Ca2+-releasing molecular tool. ACS Chem. Biol. 9, 1197−1203. (15) Yi, J. J., Wang, H., Vilela, M., Danuser, G., and Hahn, K. M. (2014) Manipulation of endogenous kinase activity in living cells using photoswitchable inhibitory peptides. ACS Synth. Biol., DOI: 10.1021/ sb5001356. (16) Harper, S. M., Neil, L. C., and Gardner, K. H. (2003) Structural basis of a phototropin light switch. Science 301, 1541−1544. (17) Strickland, D., Yao, X., Gawlak, G., Rosen, M. K., Gardner, K. H., and Sosnick, T. R. (2010) Rationally improving LOV domainbased photoswitches. Nat. Methods 7, 623−626. (18) Bustelo, X. R., Sauzeau, V., and Berenjeno, I. M. (2007) GTPbinding proteins of the Rho/Rac family: Regulation, effectors and functions in vivo. BioEssays 29, 356−370. (19) Krissinel, E., and Henrick, K. (2007) Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 372, 774−797. (20) Tay, L. H., Dick, I. E., Yang, W., Mank, M., Griesbeck, O., and Yue, D. T. (2012) Nanodomain Ca2+ of Ca2+ channels detected by a tethered genetically encoded Ca2+ sensor. Nat. Commun. 3, No. 778. (21) Evans, J. H., and Falke, J. J. (2007) Ca2+ influx is an essential component of the positive-feedback loop that maintains leading-edge structure and activity in macrophages. Proc. Natl. Acad. Sci. U. S. A. 104, 16176−16181. (22) Wei, C., Wang, X., Chen, M., Ouyang, K., Song, L. S., and Cheng, H. (2009) Calcium flickers steer cell migration. Nature 457, 901−905. (23) Lindner, R., Lou, X., Reinstein, J., Shoeman, R. L., Hamprecht, F. A., and Winkler, A. (2014) Hexicon 2: Automated processing of hydrogen-deuterium exchange mass spectrometry data with improved deuteration distribution estimation. J. Am. Soc. Mass Spectrom. 25, 1018−1028. (24) Konermann, L., Pan, J., and Liu, Y. H. (2011) Hydrogen exchange mass spectrometry for studying protein structure and dynamics. Chem. Soc. Rev. 40, 1224−1234. (25) Zayner, J. P., Antoniou, C., and Sosnick, T. R. (2012) The amino-terminal helix modulates light-activated conformational changes in AsLOV2. J. Mol. Biol. 419, 61−74. (26) Freddolino, P. L., Gardner, K. H., and Schulten, K. (2013) Signaling mechanisms of LOV domains: New insights from molecular dynamics studies. Photochem. Photobiol. Sci. 12, 1158−1170. (27) Harper, S. M., Christie, J. M., and Gardner, K. H. (2004) Disruption of the LOV-Jα helix interaction activates phototropin kinase activity. Biochemistry 43, 16184−16192. (28) Halavaty, A. S., and Moffat, K. (2007) N- and C-terminal flanking regions modulate light-induced signal transduction in the LOV2 domain of the blue light sensor phototropin 1 from Avena sativa. Biochemistry 46, 14001−14009. (29) Peter, E., Dick, B., and Baeurle, S. A. (2010) Mechanism of signal transduction of the LOV2-Jα photosensor from Avena sativa. Nat. Commun. 1, No. 122. (30) Yamamoto, A., Iwata, T., Tokutomi, S., and Kandori, H. (2008) Role of Phe1010 in light-induced structural changes of the neo1-LOV2 domain of Adiantum. Biochemistry 47, 922−928. (31) Zoltowski, B. D., and Gardner, K. H. (2011) Tripping the light fantastic: Blue-light photoreceptors as examples of environmentally modulated protein−protein interactions. Biochemistry 50, 4−16. (32) Corchnoy, S. B., Swartz, T. E., Lewis, J. W., Szundi, I., Briggs, W. R., and Bogomolni, R. A. (2003) Intramolecular proton transfers and structural changes during the photocycle of the LOV2 domain of phototropin 1. J. Biol. Chem. 278, 724−731. (33) Eitoku, T., Nakasone, Y., Matsuoka, D., Tokutomi, S., and Terazima, M. (2005) Conformational dynamics of phototropin 2 LOV2 domain with the linker upon photoexcitation. J. Am. Chem. Soc. 127, 13238−13244.

Rac-GTPase interactions with different effectors, and movies showing animation of time-dependent changes in deuterium incorporation upon calcium binding in the dark and upon bluelight illumination in the presence of calcium. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*I.S. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank R. Lindner for input regarding the Hexicon 2 analysis software and C. Roome for providing excellent IT support. We are grateful to J. Reinstein for critical comments regarding the SVD analysis of time-resolved SAXS data. We also thank K. Hahn for personal communications related to the design of LOV2-caged GTPases and for critical reading of the manuscript. We acknowledge financial support by the German Research Foundation (FOR1279 to I.S.) and the Max Planck Society.

■

REFERENCES

(1) Boyden, E. S., Zhang, F., Bamberg, E., Nagel, G., and Deisseroth, K. (2005) Millisecond-timescale, genetically targeted optical control of neural activity. Nat. Neurosci. 8, 1263−1268. (2) Gradinaru, V., Thompson, K. R., and Deisseroth, K. (2008) eNpHR: A Natronomonas halorhodopsin enhanced for optogenetic applications. Brain Cell Biol. 36, 129−139. (3) Ryu, M. H., Moskvin, O. V., Siltberg-Liberles, J., and Gomelsky, M. (2010) Natural and engineered photoactivated nucleotidyl cyclases for optogenetic applications. J. Biol. Chem. 285, 41501−41508. (4) Stierl, M., Stumpf, P., Udwari, D., Gueta, R., Hagedorn, R., Losi, A., Gärtner, W., Petereit, L., Efetova, M., Schwarzel, M., Oertner, T. G., Nagel, G., and Hegemann, P. (2011) Light modulation of cellular cAMP by a small bacterial photoactivated adenylyl cyclase, bPAC, of the soil bacterium Beggiatoa. J. Biol. Chem. 286, 1181−1188. (5) Pathak, G. P., Vrana, J. D., and Tucker, C. L. (2013) Optogenetic control of cell function using engineered photoreceptors. Biol. Cell 105, 59−72. (6) Lee, J., Natarajan, M., Nashine, V. C., Socolich, M., Vo, T., Russ, W. P., Benkovic, S. J., and Ranganathan, R. (2008) Surface sites for engineering allosteric control in proteins. Science 322, 438−442. (7) Strickland, D., Moffat, K., and Sosnick, T. R. (2008) Lightactivated DNA binding in a designed allosteric protein. Proc. Natl. Acad. Sci. U. S. A. 105, 10709−10714. (8) Wu, Y. I., Frey, D., Lungu, O. I., Jaehrig, A., Schlichting, I., Kuhlman, B., and Hahn, K. M. (2009) A genetically encoded photoactivatable Rac controls the motility of living cells. Nature 461, 104−108. (9) Lungu, O. I., Hallett, R. A., Choi, E. J., Aiken, M. J., Hahn, K. M., and Kuhlman, B. (2012) Designing photoswitchable peptides using the AsLOV2 domain. Chem. Biol. 19, 507−517. (10) Mills, E., Chen, X., Pham, E., Wong, S., and Truong, K. (2012) Engineering a photoactivated caspase-7 for rapid induction of apoptosis. ACS Synth. Biol. 1, 75−82. (11) Strickland, D., Lin, Y., Wagner, E., Hope, C. M., Zayner, J., Antoniou, C., Sosnick, T. R., Weiss, E. L., and Glotzer, M. (2012) TULIPs: Tunable, light-controlled interacting protein tags for cell biology. Nat. Methods 9, 379−384. (12) Renicke, C., Schuster, D., Usherenko, S., Essen, L. O., and Taxis, C. (2013) A LOV2 domain-based optogenetic tool to control protein degradation and cellular function. Chem. Biol. 20, 619−626. 508

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509

ACS Chemical Biology

Articles

(34) Takeda, K., Nakasone, Y., Zikihara, K., Tokutomi, S., and Terazima, M. (2013) Dynamics of the amino-terminal and carboxylterminal helices of Arabidopsis phototropin 1 LOV2 studied by the transient grating. J. Phys. Chem. B 117, 15606−15613. (35) Boehr, D. D., Nussinov, R., and Wright, P. E. (2009) The role of dynamic conformational ensembles in biomolecular recognition. Nat. Chem. Biol. 5, 789−796. (36) Seo, M. H., Park, J., Kim, E., Hohng, S., and Kim, H. S. (2014) Protein conformational dynamics dictate the binding affinity for a ligand. Nat. Commun. 5, No. 3724. (37) Volkman, B. F., Lipson, D., Wemmer, D. E., and Kern, D. (2001) Two-state allosteric behavior in a single-domain signaling protein. Science 291, 2429−2433. (38) Yao, X., Rosen, M. K., and Gardner, K. H. (2008) Estimation of the available free energy in a LOV2-J alpha photoswitch. Nat. Chem. Biol. 4, 491−497. (39) Reynolds, K. A., McLaughlin, R. N., and Ranganathan, R. (2011) Hot spots for allosteric regulation on protein surfaces. Cell 147, 1564− 1575. (40) Tompa, P. (2014) Multisteric regulation by structural disorder in modular signaling proteins: An extension of the concept of allostery. Chem. Rev. 114, 6715−6732. (41) Krauss, U., Minh, B. Q., Losi, A., Gärtner, W., Eggert, T., von Haeseler, A., and Jaeger, K. E. (2009) Distribution and phylogeny of light-oxygen-voltage-blue-light-signaling proteins in the three kingdoms of life. J. Bacteriol. 191, 7234−7242. (42) Zoltowski, B. D., Vaccaro, B., and Crane, B. R. (2009) Mechanism-based tuning of a LOV domain photoreceptor. Nat. Chem. Biol. 5, 827−834. (43) Zayner, J. P., Antoniou, C., French, A. R., Hause, R. J., Jr., and Sosnick, T. R. (2013) Investigating models of protein function and allostery with a widespread mutational analysis of a light-activated protein. Biophys. J. 105, 1027−1036. (44) Winkler, A., Heintz, U., Lindner, R., Reinstein, J., Shoeman, R. L., and Schlichting, I. (2013) A ternary AppA−PpsR−DNA complex mediates light regulation of photosynthesis-related gene expression. Nat. Struct. Mol. Biol. 20, 859−867. (45) Winkler, A., Udvarhelyi, A., Hartmann, E., Reinstein, J., Menzel, A., Shoeman, R. L., and Schlichting, I. (2014) Characterization of elements involved in allosteric light regulation of phosphodiesterase activity by comparison of different functional BlrP1 states. J. Mol. Biol. 426, 853−868. (46) Johnson, K. A., Simpson, Z. B., and Blom, T. (2009) Global kinetic explorer: A new computer program for dynamic simulation and fitting of kinetic data. Anal. Biochem. 387, 20−29. (47) Svergun, D. I., Petoukhov, M. V., and Koch, M. H. (2001) Determination of domain structure of proteins from X-ray solution scattering. Biophys. J. 80, 2946−2953. (48) Svergun, D. (1992) Determination of the regularization parameter in indirect-transform methods using perceptual criteria. J. Appl. Crystallogr. 25, 495−503. (49) Lindahl, E., Azuara, C., Koehl, P., and Delarue, M. (2006) NOMAD-Ref: Visualization, deformation and refinement of macromolecular structures based on all-atom normal mode analysis. Nucleic Acids Res. 34, W52−56. (50) Svergun, D., Barberato, C., and Koch, M. H. J. (1995) CRYSOL - A program to evaluate X-ray solution scattering of biological macromolecules from atomic coordinates. J. Appl. Crystallogr. 28, 768− 773.

509

dx.doi.org/10.1021/cb500744m | ACS Chem. Biol. 2015, 10, 502−509