University of Groningen
Structural and biochemical characterization of Roco proteins Terheyden, Susanne
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Terheyden, S. (2018). Structural and biochemical characterization of Roco proteins. University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 5
Structural insights into the Roco G-protein cycle
Susanne Terheyden, Sibel Uzuncayir, Janelle Lauer, Egon Deyaert, Wim Versées, Alfred Wittinghofer and Arjan Kortholt
ST, AW, and AK designed the experiments. ST and SU purified and crystallized the proteins from Mb. ST did the biochemical assays, solved and refined the structures. JL performed the HDX experiments. ED and WV provided the Ct LRR-RocCOR structure for the structural comparison. ST and AK wrote the manuscript. ST, AW and AK read and edited the manuscript.
Structural insights into the Roco G-protein cycle
S.Terheyden1,2, S. Uzuncayir1, J. Lauer3, E. Deyaert4,5, W. Versées4,5, A. Wittinghofer2, A.Kortholt1
1
Department of Cell Biochemistry, University of Groningen, Groningen 9747 AG, The Netherlands
2
Structural Biology Group, Max-Planck-Institute for Molecular Physiology, 44227 Dortmund, Germany
3
Max-Planck-Institute of Molecular Cell Biology and Genetics, Pfotenhauerstraße 108, 01307 Dresden, Germany
4
Structural Biology Research Center, VIB, Pleinlaan 2, 1050 Brussels, Belgium
5
Structural Biology Brussels, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium
Abstract
The Roco protein family has come into the focus of research when the human Roco protein LRRK2 (Leucine-rich repeat kinase 2) was discovered to be the most frequent cause of late onset Parkinson’s disease (PD). Roco proteins are characterized by the RocCOR domain tandem, consisting of a Ras-of-complex proteins (Roc) G-domain and a C-terminal-of-Roc (COR) dimerization domain. Although our understanding of the function and regulation mechanism of Roco proteins is still preliminary, evidence is accumulating that the Roc domain is not a classical G-domain but follows an unconventional activation mechanism. Structural data on Roco proteins are rare but would be very valuable in order to describe the activation mechanism and regulation of the RocCOR tandem domain. Here we report for the first time a nucleotide bound structure of the full RocCOR tandem from
Methanosarcina barkeri (Mb). Comparison of the GppNHp and GDP bound states reveal
no major differences in the switch region, in contrast to conformational changes reported for classical small G-proteins such as Ras, thus underscoring an unconventional switch mechanism of Roco proteins. Moreover the structure of a mutation homologous to the PD associated mutation Y1699C in LRRK2 could be solved. Structure and biochemical analysis lead to the assumption of a change in the flexibility of the mutant protein.
Introduction
Roco proteins are found in eukaryotic organism but also in a number of bacteria. They are characterized by a conserved RocCOR tandem, consisting of a Ras-of-complex proteins (Roc) G-domain and a C-terminal-of-Roc (COR) dimerization domain. Except for the RocCOR tandem, the domain composition varies quite substantially. However, almost all Roco proteins possess an upstream leucine-rich repeats (LRR) domain for protein-protein interactions, while many eukaryotic Roco proteins contain a kinase domain (Bosgraaf and Van Haastert, 2003; Marín, 2006). Originally described in 2003 (Bosgraaf and Van Haastert, 2003), the Roco protein family came into the focus of research when mutations in LRRK2, a human Roco protein, were shown to segregate with Parkinson’s disease (PD) (Zimprich et al., 2004). PD is a progressive motor disorder, caused by the degeneration of dopaminergic neurons in the midbrain. LRRK2 mutations account for 5-6 % of familial PD cases, and are identified as a risk factor for sporadic forms of the disease (Gilks et al., 2005)
Previous studies suggest that the Roc domain is not a classical G-protein but follows an unconventional activation mechanism (chapter 2-4). Although the Roc domains of LRRK2 and the Roco protein family members belongs to the class of small G-proteins (Bosgraaf and Van Haastert, 2003) which switch between an active GTP- and inactive GDP-bound state, they seem to function without the help of guanine nucleotide exchange factors (GEFs). GEFs reduce the nucleotide affinity of classical G-proteins from the sub-nanomolar to the micromolar range to allow fast dissociation of nucleotide and thus facilitate nucleotide exchange (Vetter and Wittinghofer, 2001; Wittinghofer and Vetter, 2011). A number of studies from our and other labs have found that Roco proteins have a relatively low affinity for nucleotides (in the micromolar range) corresponding to a fast dissociation [7–11, chapter 4]. Another difference to classical G-proteins may lie within the catalytic machinery itself: In order to switch off, classical G-proteins require certain residues to be provided by a GTPase Activating Protein (GAP) to complement the catalytic machinery for a fast GTP hydrolysis reaction (Vetter and Wittinghofer, 2001; Wittinghofer and Vetter, 2011). For LRRK2 and other Roco proteins it has been suggested that they don’t require GAPs but are able to hydrolyse GTP within the dimer (Gotthardt et al., 2008; Rudi et al., 2015a; Terheyden et al., 2015).
We also have shown that the Roco protein from Chlorobium tepidum (Ct) cycles between a monomer and dimer within a catalytically relevant time scale in a nucleotide dependent manner (Deyaert et al., 2017a) implying that dimerization might have an important role in the hydrolysis mechanism. The structure of the Ct protein has shown that RocCOR is a dimer (Gotthardt et al., 2008) leading to the hypothesis that Roco proteins and also LRRK2 belong to the GAD (G-proteins activated by nucleotide-dependent dimerization) class of molecular switches (Gotthardt et al., 2008; Gasper et al., 2009; Nixon-abell et al., 2016). However, in this structure the RocCOR tandem is free of G-nucleotide and one Roc domain is not resolved. For LRRK2 the structure of the Roc domain alone has been solved bound to GDP (PDB:2ZEJ (Deng et al., 2008)), which unfortunately showed a swapped dimer where part of the G-domain is detached. Whether or not such a dimer is relevant for LRRK2 or is an artefact of crystallization remains to be shown. We have previously published the structure of a truncated RocCOR tandem, RocCOR∆C (PDB: 4WNR), bound to GDP (Terheyden et al., 2015) which is a monomer, highlighting the assumption that the full COR domain is important for dimerization. However, with these constructs, a structural comparison between GTP- and GDP-bound states of the Roc domain was not possible due to the lack of suitable crystals. Therefore we set out to obtain structural evidence for these states in order to understand the activation mechanism of this unconventional G-protein. Here we show that the comparison of the GppNHp (as a non-hydrolysable GTP analogue) and GDP bound states reveal no major differences in the switch region. This is in contrast to conformational changes reported for classical small G-proteins such as Ras, but is in agreements with a previous study that shows that LRRK2 kinase activity is neither enhanced in the presence of GDP nor GppNHp but by the addition of GTP (chapter 4).
Materials and Methods
Purification of Roco proteinsThe Methanosarcina barkeri (Mb) RocCOR tandem (aa 287-790, Uniprot: Q46A62) and the indicated SER mutants were cloned into a modified pProExHTb vector with N-terminal TEV cleavage site. The constructs were expressed in Rosetta 2 (DE3) cells using the LEX bioreactor (harbinger biotech) initially at 37°C later at 25°C. At an OD600 between 1 and 2
the temperature was lowered to 20°C and expression was induced over night with 0.1 mM IPTG. Cells were harvested and resuspended in buffer containing 30 mM HEPES pH 7.5, 150 mM NaCl, 3 mM ß-mercaptoethanol and 5 % Glycerol with the addition of lysozyme, DNase and 0.1 mM PMSF. After at least 30 min incubation on ice (stirring), the cells were decomposed with a fluidizer and sonifier. The soluble fraction was then subjected to affinity chromatography (IMAC) using Talon beads. If required anion exchange chromatography was performed. As a final purification step, size exclusion chromatography was used. Isolated proteins were analysed on SDS-page. The protein concentration was determined using the absorption at 280 nm (Nanodrop 2000, Thermo Scientific). Before starting experiments the nucleotide load of each protein batch was determined via reversed phase chromatography coupled to HPLC.
Structure solution, alignments and analysis. Table 1: Crystallization conditions
Protein Concentration Additions Crystallizing condition Experiment
RocCOR GDP Mb Roco2 287-790 572-575 AAA 15 mg/ml 10 mM DTE 10 mM GDP Seeding experiment: seed stock from:
100 mM MES pH 5.5 -6.5 2-2.6 M Ammonium sulfate Growing condition: 1.6 M Sodium Dihydrogen Phosphate 400 mM Di-Potassium Hydrogen Phosphate 100 mM Phosphate-Citrate pH 4.2 (Quiagen JCSG CORE IV H10) 24 well plate Hanging drop --- 96 well plate Sitting drop RocCOR GppNHp 20 mg/ml 10 mM DTE 5 mM GppNHp 100 mM Tris pH 7 200 mM MgCl2 10 %(w/v) PEG 8000 (Quiagen JCSG CORE I C4) 24 well plate Hanging drop RocCOR NF Mb Roco2 287-790 572-575 AAA 590-592 AAA 20 mg/ml 1 mM MgCl2 5 mM AlCl3 50 mM NaF 1 mM GDP 100 mM Tris 7.5 200 mM NaCl 14% PEG 7% MPD 96 well plate Sitting drop RocCOR F652C GDP Mb Roco2 287-790 572-575 AAA 590-592 AAA F652C 15 mg/ml 10 mM DTE 2 mM GDP 100 mM Sodium citrate pH 5.5 2.0 M Ammonium sulfate (Quiagen JCSG CORE III G1)
96 well plate Sitting drop
Table 1 summarizes the conditions used for the crystallization of the different proteins. For cryoprotection cryo oil (perfluoropolyether, MW: 2700) was used. The dataset of nucleotide free (NF) RocCOR was collected at the PETRA III P11 beamline (Deutsches Elektronen-Synchrotron DESY, Hamburg, Germany). The other datasets were collected on beamline X10SA (PXII) at the Swiss Light Source (Paul Scherrer Institute, Villigen, Switzerland). Indexing, integration and scaling was performed with the XDS package (Kabsch, 2010). The structure was solved by molecular replacement using the Roc domain of Mb RocCOR∆C (PDB code 4WNR) as a template with the program Phaser (McCoy et al., 2007) from the CCP4 suite (Winn et al., 2011). The model was built in COOT (Emsley et al., 2010) and refined with REFMAC5 (CCP4 suite) (Murshudov et al., 2011) and Phenix.refine (Adams et al., 2010). Figures were generated in PyMOL (Schrödinger). Structural alignments including RMSD values were generated with the RaptorX (Wang et al., 2011, 2013) and Rapido server (Mosca and Schneider, 2008; Mosca et al., 2008). Crystal contacts and interfaces were analysed using the PISA Webserver (Krissinel and Henrick, 2007).
Equilibrium titrations
To measure the affinity to 3`-O-N-methylanthraniloyl (mant)-labeled GDP and GTPγS nucleotides (1 µM) we performed equilibrium titration experiment. At 20 °C, 1 µM of mant-nucleotide was titrated stepwise in a cuvette containing nucleotide free RocCOR protein in 50 mM Tris pH 7.5, 150 mM NaCl, 5 mM MgCl2 and 5% Glycerol (v/v).
Fluorescence-anisotropy was measured using the FluoroMax-4 (Horiba Scientific, excitation: 366 nm, emission 445 nm). The signal was recorded over at least 5 min. The dissociation constant (KD) was then calculated by fitting a quadratic equation to the data
using GraFit 5.0 (Erithacus software). The error represents the standard error of the mean from at least two measurements.
Steady-state kinetic measurements of GTP hydrolysis
In order to measure Michaelis-Menten constants (KM) and catalysis rates (kcat) of each
proteins’ hydrolysis reaction, a HPLC (Themo Ultimate 3000) based assay was used. A reversed phase C18 column was employed to detect GDP and GTP content (Eberth and Ahmadian, 2009). 1 µM of protein (Buffer : 50 mM Tris pH 7.5, 150 mM NaCl, 20 mM MgCl2, 5% Glycerol) was incubated at 20°C with different amounts of GTP. At different
time points 5 µl of this solution was analysed for GTP and GDP content (in %). Linear rates of GDP production were plotted against GTP concentration. The Michaelis-Menten equation was fitted using GraFit 5.0 (Erithacus software). The error represents the standard error of the mean from at least two measurements.
Analytical Ultracentrifugation
Sedimentation velocity (SV-AUC) experiments on Mb RocCOR were carried out at 20°C in in the presence or absence of 100 µM GDP or 200 µM GppNHp in buffer containing 50 mM Tris pH 7.5, 150 mM NaCl, 5 mM MgCl2, 5% Glycerol. The absorbance at 280 nm
was measured on a Beckman Coulter ProteomLabTM XL-I analytical ultracentrifuge. Samples were prepared at a concentration of 0.5 mg/ml (NF and GppNHp; 8.3 µM) and 0.8 mg/ml (GDP; 13.3 µM). Standard double sector centrepieces were used. The cells were scanned every minute and in total 200 scans were collected. The data was analysed using SEDFIT 15.01b (Schuck, 2000) with the continuous distribution model c(s). Solution density ρ, viscosity η and partial specific volumes 𝑣𝑣̅ were calculated using SEDNTERP (Laue et al., 1992) (ρ =1.02061 g/l; η = 0.01197 kg/(s*m), 𝑣𝑣̅ (Mb RocCOR) = 0.74366. The c(s) analysis was carried out with an s range of 0 to 15 with a resolution of 100 and a confidence level of 0.68. In all cases, fits were good, with root mean square deviation (RMSD) values ranging from 0.012 to 0.031 AU. Results were prepared for publication using GUSSI 1.2.1 (Brautigam, 2015).
Hydrogen Deuterium Exchange Mass Spectrometry (HDX-MS)
HDX-MS was performed as previously described (Kan et al., 2011; Walters et al., 2012; He et al., 2015). Proteins (1 uM) were diluted 6:4 with 8M urea, 1% trifluoroacetic acid and passed over an immobilized pepsin column (2.1 mm x 30 mm, ThermoFisher Scientific) in 0.1% trifluoroacetic acid at 15 °C. Peptides are captured on a reversed-phase C8 cartridge, desalted and separated by a Zorbax 300SB-C18 column (Agilent) at 1 °C using a 5-40% acetonitrile gradient containing 0.1% formic acid over 10 min and electrosprayed directly into an Orbitrap mass spectrometer (LTQ-Orbitrap XL, ThermoFisher Scientific) with a T-piece split flow setup (1:400). Data were collected in profile mode with source parameters: spray voltage 3.4kV, capillary voltage 40V, tube lens 170V, capillary temperature 170 °C. MS/MS CID fragment ions were detected in centroid
mode with an AGC target value of 104. CID fragmentation was 35% normalized collision energy (NCE) for 30 ms at Q of 0.25. HCD fragmentation NCE was 35eV. Peptides were identified using Mascot (Matrix Science) and manually verified to remove ambiguous peptides. For measurement of deuterium uptake, 10uM protein is pre-incubated with 1mM desired nucleotide, where applicable and diluted 1:9 in 20mM HEPES (pH 7.5), 100mM NaCl, 5mM MgCl2 prepared with deuterated solvent. Samples were incubated for 0, 60 and 900s at 22 °C followed by the aforementioned digestion, desalting, separation and mass spectrometry steps. The intensity weighted average m/z value of a peptide’s isotopic envelope is compared plus and minus deuteration using the HDX workbench software platform (Pascal et al., 2012). Individual peptides are verified by manual inspection. Data are visualized using Pymol. Deuterium uptake is normalized for back-exchange when necessary by comparing deuterium uptake to a sample incubated in 6M urea in deuterated buffer for 12-18h at room temperature and processed as indicated above.
Results and Discussion
The Mb RocCOR tandem containing surface mutations was crystallized in nucleotide free and bound to different nucleotides
To better understand the activation mechanism of the RocCOR tandem we set out to crystallize a Roco protein in its GTP- and GDP-bound state. However all initial attempts to crystallize the bacterial RocCOR constructs failed. Therefore, we used a surface entropy reduction (SER) approach on the RocCOR tandem of Mb Roco2. Surface residues were mutated from lysine or glutamate to alanine in order to reduce the loss of entropy and increase chances of crystallization. As shown in figure 1A, surface patches of three of these residues were determined by the SER-server (Goldschmidt et al., 2007). A single mutant with the patch 572-575 AAA was generated and a double mutant with two alanine patches (572-575 AAA and 590-592 AAA). Both lie in the N-terminal COR subdomain (COR-A).
To be sure that these SER mutations don’t affect the proteins activity, we performed nucleotide affinity and GTPase measurements (figure 1B, 1C and table S1). The SER mutants don’t show significant differences in nucleotide binding. The differences are less than 4 fold (e.g. 1.8 and 6.95 µM for the wild type (wt) protein and the single mutant, respectively). Furthermore Michaelis-Menten kinetics of the GTPase reaction were determined to monitor possible changes in activity upon mutation. KM values show only
minor differences between wild-type and the SER mutants (wt: 353 ± 129 µM (chapter 4), single: 970 ± 302 µM, double: 991 ± 496 µM), whereas the kcat values of both mutated
proteins are approximately half of that of wild-type (wt: 0.063 ± 0.006 min-1 (Wauters et al., 2017), single: 0.033 ± 0.007 min-1, double: 0.027 ± 0.013 min-1). Together, these data show that the SER mutations do not severely affect the enzymatic activity of the Roc-COR tandem domain and thus can be used for crystallization studies.
Apo, GTP- and GDP-bound states are structurally very similar
Crystals of nucleotide bound protein could be obtained using the single patch SER mutant whereas the double patch mutant was used for the nucleotide free protein (table 2). The nucleotide free and GppNHp bound form crystallized with two monomers in the asymmetric unit (ASU) in space group P 21 21 21, whereas the GDP form crystallized as 12 molecules in the ASU in P 31 1 2. The nucleotide bound structures diffract to a
resolution of 3.25 Å, whereas the nucleotide free form gave diffraction to 2.1 Å. The Roc domain shows the canonical G-domain fold including the five conserved sequence motifs (Wittinghofer and Vetter, 2011) as previously described (Terheyden et al., 2015). The COR domain displays the same fold as previously described (Gotthardt et al., 2008; Terheyden et al., 2015). The COR domain is L-shaped, winding around the Roc domain and can be divided into COR-A (grey, N-terminal subdomain) and COR-B (orange, C-terminal) subdomains. As found for previous bacterial homologues RocCOR forms a tight non swapped dimer where dimerization is mediated by COR-B of both protomers (Figure 1D shows the RocCOR dimer bound to GppNHp). The N-terminal subdomain (COR-A) consisting of two helical subdomains and the C-terminal subdomain (COR-B) forming the dimerization interface with a beta-alpha-beta fold.
The helix preceding the Roc domain (α0-helix, yellow) folds onto COR-A. Figure 1E
shows one protomer (monomer) alone. Figure 1F highlights the four patches that contact each other in the dimer. According to the PISA server (Krissinel and Henrick, 2007), the total area of this interface is 1175 Å2. It is composed of 4 patches: Patch 1 (yellow) and 2 (red) with 3 and 7 residues respectively lie in the Roc domain of the protein whereas the remaining 30 residues are contributed from COR-B (patch 3, blue and patch 4, green). Patch 4 interacts with patch 4 in the other protomer whereas patch 3 interacts with patch 1, 2 and 3 from the other protomer. As indicated in figure 1, the density for the different nucleotides is clearly visible in the nucleotide binding pocket and is distinct (fig. 1 G and H), confirming that the different ligands are bound. In the nucleotide free form, the P-loop is not formed but is part of helix 1 (figure 1I). Figure 1J shows a scheme of the proteins interactions with GppNHp. Classical interactions with the p-loop (G333, K334, T335), switch I (T351), G4 (N/TKxD, here NKID, D433) and G5 (SAK, here SCK, C460) motifs can be found (Wittinghofer and Vetter, 2011). Interestingly direct binding to switch II cannot be observed. The DxxG motif is more than 6 Å away from the γ-phosphate. The lack of the interaction between the γ-phosphate and switch II might explain the rather low hydrolysis rate. The unusual conformation of switch II might also result in poorly coordinated magnesium, which could influence both hydrolysis rate and nucleotide affinities. This however is difficult to judge since the position of the magnesium ion is not well defined at this resolution. However, our Roco structures reveal significant difference to classical small G-proteins in the GTP-bound state, which might explain the different hydrolysis mechanism (chapter 4).
Table 2: Data collection and refinement statistics
Data collection GppNHp GDP Nucleotide free
Space Group P 21 21 21 (19) P 31 1 2 (151) P 21 21 21 (19) Cell dimensions a, b, c (Å) α, β, γ (°) 73, 114.3, 151.2 90, 90, 90 206, 206, 366.8 90, 90, 120 71.8, 111.8, 151.8 90, 90, 90 Resolution (Å) 46-3.28 (3.48-3.28) 49.6-3.23 (3.43-3.23) 47.27-2.10 (2.26-2.10) Rmeas 0.23 (1.19) 0.25 (1.87) 0.15 (1.75) CC1/2 99.2 (59.7) 99.8 (60.0) 99.7 (54.1) I/σ 7.8 (1.5) 10.8 (1.6) 7.5 (1.1) % Completeness 99.1 (95.2) 99.4 (96.3) 99.94 (100) Wavelength (Å) 0.91638 0.91638 0.979218
rotation range per image
(°) 0.25 0.1 0.1
total rotation range (°) 180 180 360
Refinement
No. of reflections 128517 1440573 873446
No. of unique reflections 19788 140985 72029
R, Rfree 0.274, 0.302 0.226, 0.266 0.209, 0.253 No. atoms Protein 7092 82004 (including hydrogens) 7940 Ligand/ion 66 452 17 Water - - 156 B-factor Protein 87 72 36 Ligand/ion 98 96 35 Water - - 34 R.m.s. deviations Bond length (Å) 0.002 0.014 0.016 Bond angles (°) 0.64 1.52 1.75 Number of molecules /ASU 2 12 2 Ramachandran outliers 0.2% 0.4 % 0.1% Clashscore 5 22 2
Figure 1: A: Surface representation of the RocCOR tandem. SER mutant patches are
shown in red (572-575 AAA) and green (590-592 AAA). B+C: Nucleotide affinities (KDs)
measured with fluorescence polarization in an equilibrium titration experiment using mant-labelled GTPγS (as a GTP analogue) and mant-GDP: The SER mutants show no difference in binding compared to the wt protein. D-F: Nucleotide binding pockets of GDP, GppNHp
bound and nucleotide free Mb RocCOR tandem structures. The density for the nucleotides is well defined. G+H: Front and bottom view of the three overlaid Mb RocCOR tandem structures. Differences are highlighted in colour: nucleotide free in pink, bound to GDP in cyan and bound to GppNHP in green I: Topology diagram of the Roc domain. Differences between the three structures (P-loop, switch I and II and G5-motif) are highlighted in red. J: Close up view of the P-loop. In the nucleotide free structure, the P-loop is not properly formed but part of helix 1. K: Overlay of all monomers from one asymmetric unit of RocCOR-GDP highlighting the poorly ordered switch II region.
Structural alignments of the three structures with the Raptor and Rapido show that there are no major differences in the overall folding (table 3). Remarkably the GDP and GppNHp bound structures (and NF) crystallized in different space groups. That proteins are able to crystallize in different space groups is known for a long time and it is even known that sometimes a certain conformation will only form crystals in a certain space group, however, in this case all analysable crystals grown in a condition with GppNHp were found to be space group P 21 21 21 whereas in the presence of GDP only P 31 1 2 crystals were formed. Therefore this might indicate that the structural composition of the protein solution changed with respect to the nucleotide. A more detailed comparison of the three structures reveals local differences (highlighted in colour, NF: pink, GDP: cyan, GppNHp: green) within the G-domain loops and on the surface of the COR-A subdomain (figure 1K and 1L). The differences on the surface of the protein in the COR-A domain originate probably from the different crystal packing: Most of the helix shifts in COR-A are observed in the GDP-state which crystallized in a hexagonal lattice (P 31 1 2) in contrast to NF and GppNHp in an orthorhombic lattice (P 21 21 21). However, switch I, switch II and the G5-motif, which are not located on the surface, show variations and therefore are more likely to originate from the nature of the different nucleotide states (figure 1M). In the NF state, the P-loop of one monomer is not properly formed but integrates into helix 1 (figure 1N). Despite the better resolution of 2.1 Å, switch I is not visible in both monomers, suggesting this loop is flexible. In the GDP state, switch II is rather disordered and only partly visible (figure 1O). Meaningful conformational changes in the switch I or switch II region, displaying a defined active and inactive state, could not be detected at this resolution.
Nevertheless, this work, for the first time presents a comparison between the nucleotide free, GDP and GppNHp bound states of a RocCOR tandem. The comparison reveals no major differences in the structures, but local differences in the dynamics of the switch
regions. Furthermore that fact that the GDP- and GppNHp-bound states crystallized in different space groups under the same conditions seems to indicate structural changes and/or dynamics in the surface of the proteins in solution. These findings might point towards a different G-protein cycle of Roco proteins compared to small G-proteins, in that the GDP- and GppNHp-states in Roco proteins do not represent an inactive versus active conformation, respectively. Results from previous studies support this idea: Rudi et al. (Rudi et al., 2015a) show that the binding of these nucleotides changes the equilibrium of existing conformations or the dynamics/flexibility of the protein rather than defined conformational states. It was also shown for LRRK2 in chapter 4 that only the presence of GTP enhances LRRK2 kinase activity in contrast to GppNHp and GDP.
Table 3: RMSD values (in Å) obtained by comparing the different structures (and E, F, G,
H, I and J chains in case of GDP) to each other. The values are obtained by RaptorX (upper case) and the Rapido server rigid body alignment (italic letters, lower case).
GppNHp GDP Nucleotide free GH IJ GDP EF 0.81 0.78 1.22 1.0 1.04 1.03 GH 1.33 1.24 0.0 1.18 - Nucleotide free 1.07 1 1.33 1.4 1.29 1.26 -
Roc and COR domains have multiple conformations
A comparison with the other published structure of the RocCOR tandem, the nucleotide free RocCOR of Ct (3DPU), and our nucleotide free Mb RocCOR with the RaptorX server gave a RMSD value of 4.55 Å, indicating that there are significant differences between these structures. Although, one monomer and the C-terminal part of the COR domain (COR-B) are overlaying rather well (see figure 2A), the COR-A of the second protomer is shifted (figure 2B). A similar movement of the COR-A domain is visible when comparing the Mb RocCOR tandem with an unpublished structure of the complete Ct LRR-RocCOR module (nucleotide free) (Deyaert et al., 2017b) (figure 2D). The alignment (RaptorX) gave an overall RMSD value of 6.81 Å. One protomer and COR-B of the second protomer
overlay quite well (figure 2C) whereas the conformations of COR-A and Roc (Chain B) show large shifts (figure 2D and E). The Roc domain is shifted inwards in Ct LRR-RocCOR making contact to the other Roc domain highlighted by the yellow arrow in figure 2E. A detailed analysis of the interaction interfaces generated by PISA reveals that
Ct LRR-RocCOR and Mb RocCOR display similar interaction patches between the Roc
and COR domain (see figure S1). However, the Ct Roc domain interaction (patch 3) is much more pronounced. One general observation is that, although residues are not very conserved, the size of the interaction patches is rather similar (except for patch 3) and they are located at almost the same positions. Similarly, when comparing Mb RocCOR-GDP (Chains GH) with the published Mb RocCOR∆C (4WNR, also GDP bound), also the COR-A parts don’t align well (figure 2F-H). The overall RMSD value is 4.22 Å.
Together this shows that the Roc, COR-A and COR-B can obtain multiple conformations, which might suggest they are flexible with respect to each other.
Figure 2 A+B: Comparison of the Mb RocCOR (NF, in green and yellow colours) with the
Ct RocCOR tandem structure (PDB: 3DPU; NF, in blue colours). A: front view; B: 30° rotated to the left. One protomer, and both B domains overlay nicely whereas COR-A of the second protomer is shifted. The second Roc domain of Ct RocCOR is not resolved. The scheme in the upper right corner shows the domain movement schematically. C-E: Comparison of the Mb RocCOR (NF) with the unpublished Ct LRR-RocCOR structure. D: Enlargement of the overlayed second protomers‘ COR domains. E: Bottom view, highlighting only the Roc domains. Again one protomer and both COR-B domains overlay nicely, while COR-A (D) and Roc of the second protomer are shifted (E). F-H:
Comparison of the the Mb RocCOR (GDP, in green) with the Mb RocCOR∆C (PDB: 4WNR, in blue). Here the Roc domain overlays well (G) whereas the COR-A is shifted.
Mb RocCOR undergoes similar conformational changes as Ct RocCOR without
monomerizing
To investigate the dynamic changes of the protein upon nucleotide binding, Hydrogen/Deuterium Exchange Mass Spectrometry (HDX-MS) experiments were performed in the presence of GDP or GppNHp. HDX experiments monitor the exchange rate of the amide hydrogen in a peptide bond through the uptake of deuterium (Konermann et al., 2011). Fast exchange points towards high surface accessibility. Through this exchange, the comparison between two states can visualize conformational changes. Figure 3A highlights the regions in the Mb RocCOR tandem that are more dynamic (warm colours) and shielded (cold colours) in the GppNHp state. A moderate increase in flexibility is observed in the COR-B and COR-A parts whereas major changes in dynamics are visible in the α0-helix (N-terminus) and more interestingly in the linker region between
Roc and COR domain. This again highlights the importance of the position of COR-A towards the Roc domain. Moreover the Roc domains, especially the nucleotide binding pockets are shielded in the GppNHp state. The most drastic change towards more structure/shielding can be seen in the switch II region of the protein. Unpublished HDX-MS data for the Ct Roco protein shows a very similar trend (personal communication). These result suggest, that Mb RocCOR experiences a similar conformational change compared to Ct Roco where the Roc domains come close to each other. Since the Ct Roco protein is reported to monomerize upon GppNHp binding (Deyaert et al., 2017a), we next set out to investigate the dimerization behaviour of the protein by sedimentation velocity analytical ultracentrifugation (SV-AUC) experiments (figure 3B and S2A-C). Since nucleotides itself give an absorption signal at 280 nm, the measurements with GDP and GppNHp show an increased background noise compared to the nucleotide free form (S2A-C). Nevertheless, the data of all experiments could be fitted with RMSD values ranging from 0.012 to 0.031 AU. In all three nucleotide states, the main component has a sedimentation coefficient around 4.5 S which leads to an estimated molecular weight of around 120 kDa, considering the fitted frictional ratios (figure 3B and table 4). This matches the dimeric form of the protein. The frictional ratio is a measure of the shape of
the molecule. The value 1 indicates a perfect sphere whereas a very elongated protein has a value of 2. Here, the ratio was estimated by the program SEDFIT during data analysis. Minor peaks for the sedimentation coefficient could be detected that possibly correspond to monomers (in NF state) or multimers (in GDP and GppNHp state). However, these fractions are rather small (less than 10%) and could possibly originate from minor contaminations, therefore the significance is questionable. We conclude that under the conditions used and in contrast to Ct Roco, the Mb RocCOR tandem does not monomerize upon GppNHp binding. However, the calculated frictional ratio reveals a possible conformational change of the GppNHp state. It changes from around 1.4 to 1.65, which suggests a shift to a more elongated form.
Table 4: Results from analytical ultracentrifugation sedimentation velocity experiments
for Mb RocCOR in the presence of different nucleotides.
Frictional ratio coefficient c(s) [S] Sedimentation Calculated molecular weight [kDa]
NF 1.4 Peak 1: 4.6 120 (76%) Peak 2: 2.7 52.5 (1.3%) GDP 1.45 Peak 1: 4.4 118 (73%) Peak 2: 6.7 218 (81%) GppNHp 1.65 Peak 1: 4.2 131 (60%) Peak 2: 6.9 279 (9.7%)
The PD associated mutation F652C shows no structural differences in the GDP-state
The residue F652 corresponds to Y1699 in LRRK2. LRRK2 Y1699C shows a reduced autophosphorylation of residues S910/935/955/973 (Nichols et al., 2010) but increased autophosphorylation at S1292 and highly elevated phosphorylation of the kinase’ substrates Rab8 and Rab10 (Steger et al., 2016). Y1699C GTPase activity is reduced in LRRK2 (Daniëls et al., 2011a). Also for Ct RocCOR it has been shown that the corresponding mutation (Y804C) has reduced GTPase activity (Gotthardt et al., 2008). Here, we set out to investigate if the corresponding mutation in the Mb RocCOR tandem (F652C) has an effect on the Roc structure and activity.
Biochemical analysis of mutant F652C shows that the catalytic rate of the hydrolysis reaction kcat is almost half (0.035 ± 0.007 min-1) of that of wild-type (table S1). The KM
remains in the same range (466 ± 261 µM) as well as the affinities for GDP and GTP (KD(GDP)= 5.08 ± 2.34 µM, KD(GTPγS)= 0.71 ± 0.05 µM). Although the effect on kcat is
not very pronounced, this might still be significant, since in the LRRK2 field it has been reported previously that mutations have rather moderate effects (2-fold/3-fold) on LRRK2 kinase and GTPase activity in vitro (Greggio and Cookson, 2009).
In the next step, we solved the structure of the mutated Mb RocCOR tandem in the presence of GDP. The space group, unit cell and asymmetric unit are identical to the Mb RocCOR-GDP structure (table 5). The density of the mutation (F652C) in the COR domain is clearly visible in most molecules (figure 3D) and is located in the hydrophobic interface between Roc and COR domain (figure 3C). This interface connects the Roc domain of one protomer (switch II) with its COR domain but also the COR domain of the other protomer. R1441 also lies within this interface. When all 12 monomers of the F652C mutant are compared with each other (RaptorX), the deviation ranges between 0.76 and 1.21 Å (table S1), showing that all molecules are very similar in the ASU. Moreover, several different chains of the Mb RocCOR-GDP were compared to Mb RocCOR F652C-GDP again with the RaptorX alignment server. The obtained RMSD values are listed in table S2. They lie between 0.81 and 1.37 Å, demonstrating that their overall structure doesn’t change (figure 3E). Since the mutation effects activity the change might be more pronounced in a GppNHp or transition state structure (e.g. GDP-AlFx). However, crystals for those structures could not be obtained. Considering the location of F562 in the interface it is to be expected that the mutation creates an effect on protein dynamics which is difficult to visualize with a crystal structure but would require an analysis of protein flexibility using for example HDX-MS.
Table 5: Data collection and refinement statistics for Mb RocCOR F652C bound to GDP
Data collection and processing F652C GDP
Space Group P31 1 2 (151) Cell dimensions a, b, c (Å) α, β, γ (°) 207, 207, 365 90, 90, 120 Resolution (Å) 49.5-3.09 (3.28-3.09) Rmeas 0.198 (1.636) CC1/2 99.8 (62.8) I/σ 11.7 (1.57) % Data completeness 99.6 (97.5) Wavelength (Å) 0.91883
rotation range per image (°) 0.25
total rotation range (°) 360
Refinement
No. of reflections 3332145
No. of unique reflections 315400
R, Rfree 0.225, 0.27
No. atoms 85672 (including hydrogens)
Protein 85336 Ligand/ion 515 Water - B-factor Protein 72 Ligand/ion 109 Water - R.m.s. deviations Bond length (Å) 0.2 Bond angles (°) 1.66 No. Of mileccules/ASU 12 Ramachandran outliers 0.2% Clashscore 17
Figure 3 A: Results from HDX-MS experiment comparing Mb RocCOR bound to GDP
with GppNHp. Warm colours highlight dynamic regions. Cold colours indicate shielding from the solvent upon GppNHp binding. B: Results from SV-AUC experiments: The sedimentation coefficient doesn't change significantly in the presence of either GDP or GppNHp compared the NF state, indicating that the oligomerization state doesn't change. According to calculations the molecular weight of these fractions corresponds to a dimer (120 kDa). C: The density for the F652C mutation is well defined. The residues lies in the hydrophobic interface between Roc and COR domain also in close proximity to the second protomer. D: Interface between Roc and COR domains in the GppNHp bound dimer. Switch II is highlighted in grey. Residues H415, F419 and F652 which correspond to PD mutated residues N1437, R1441 and Y1699 in LRRK2, are shown in magenta. E: Overlay of the Mb RocCOR F652C-GDP structure with the wt bound to GDP. The overlay is almost perfect.
Conclusion
All in all, this work presents solid structural insight into the RocCOR tandem bound to different nucleotide states. It shows minor differences in the switch regions and increased flexibility of the domain linkages. Our data highlights the importance of the position and flexibility of Roc, COR-A and COR-B. Interestingly, the switch II region lies in this hydrophobic interphase between the Roc and COR domain, showing the highest shielding upon GppNHp binding in the HDX-MS experiment. Also several of the LRRK2 residues mutated in PD (N1437, R1441 and Y1699) lie in this hydrophobic interphase (figure 3C), suggesting an important role of this interface and switch II in the regulation of Roc activity. Switch II might present the link that mediates the information about the nucleotide state from the Roc domain towards the COR domain. However, the data are not sufficient to explain the hydrolysis mechanism or the G-protein cycle of the RocCOR tandem. For a better understanding a transition state structure, represented by the GDP AlFx mimic, is still missing and might shed light on the question on how the protein catalyses the hydrolysis reaction. Furthermore, the dynamic changes of the protein in solution need to be investigated further in order to understand the actual conformational changes that are induced by GTP binding and hydrolysis.
Acknowledgements
I would like to thank Arsen Petrovic for his assistance with the SV-AUC experiments. Also I thank the X-ray community of the Max-Planck Institute of molecular Physiology in Dortmund especially Ingrid Vetter, Eckhard Hofmann (Ruhr-University Bochum), Raphael Gasper-Schönenbrücher, Katja Gotthardt, Matthias Müller, Arthur Porfetye, Eldar Zent and Amrita Rai. Moreover, this research was supported by the Michael J. Fox Foundation for Parkinson’s Research.
References
Adams, P.D., Afonine, P. V., Bunkóczi, G., Chen, V.B., Davis, I.W., Echols, N., Headd, J.J., Hung, L.W., Kapral, G.J., Grosse-Kunstleve, R.W., et al. (2010). PHENIX: A comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. Sect. D Biol. Crystallogr. 66, 213–221.
Bosgraaf, L., and Van Haastert, P.J.M. (2003). Roc, a Ras/GTPase domain in complex proteins. Biochim. Biophys. Acta - Mol. Cell Res. 1643, 5–10.
Brautigam, C.A. (2015). Chapter Five – Calculations and Publication-Quality Illustrations for Analytical Ultracentrifugation Data. In Methods in Enzymology, pp. 109–133.
Daniëls, V., Baekelandt, V., and Taymans, J.-M. (2011). On the road to leucine-rich repeat kinase 2 signalling: evidence from cellular and in vivo studies. Neurosignals. 19, 1–15. Deng, J., Lewis, P. a, Greggio, E., Sluch, E., Beilina, A., and Cookson, M.R. (2008). Structure of the ROC domain from the Parkinson’s disease-associated leucine-rich repeat kinase 2 reveals a dimeric GTPase. Proc. Natl. Acad. Sci. U. S. A. 105, 1499–1504.
Deyaert, E., Wauters, L., Guaitoli, G., Konijnenberg, A., Leemans, M., Terheyden, S., Petrovic, A., Gallardo, R., Nederveen-Schippers, L.M., Athanasopoulos, P.S., et al. (2017a). A homologue of the Parkinson’s disease-associated protein LRRK2 undergoes a monomer-dimer transition during GTP turnover. Nat. Commun. 8, 1008.
Deyaert, E., Kortholt, A., and Versées, W. (2017b). The LRR-Roc-COR module of the Chlorobium tepidum Roco protein: Crystallization and X-ray crystallographic analysis. Acta Crystallogr. Sect. Struct. Biol. Commun. 73, 520–524.
Eberth, A., and Ahmadian, M.R. (2009). In Vitro GEF and GAP Assays. Curr. Protoc. Cell Biol. 1–25.
Emsley, P., Lohkamp, B., Scott, W.G., and Cowtan, K. (2010). Features and development of Coot. Acta Crystallogr. Sect. D Biol. Crystallogr. 66, 486–501.
Gasper, R., Meyer, S., Gotthardt, K., and Sirajuddin, M. (2009). It takes two to tango: regulation of G proteins by dimerization. Nat. Rev. Mol. Cell Biol. 10, 423–429.
Gilks, W.P., Abou-Sleiman, P.M., Gandhi, S., Jain, S., Singleton, A., Lees, A.J., Shaw, K., Bhatia, K.P., Bonifati, V., Quinn, N.P., et al. (2005). A common LRRK2 mutation in idiopathic Parkinson’s disease. Lancet 365, 415–416.
Goldschmidt, L., Cooper, D.R., Derewenda, Z.S., and Eisenberg, D. (2007). Toward rational protein crystallization: A Web server for the design of crystallizable protein variants. Protein Sci. 16, 1569–1576.
Gotthardt, K., Weyand, M., Kortholt, A., Van Haastert, P.J.M., and Wittinghofer, A. (2008). Structure of the Roc-COR domain tandem of C. tepidum, a prokaryotic homologue of the human LRRK2 Parkinson kinase. EMBO J. 27, 2239–2249.
Greggio, E., and Cookson, M.R. (2009). Leucine-rich repeat kinase 2 mutations and Parkinson’s disease: three questions. ASN Neuro 1, 13–24.
He, W., Bai, G., Zhou, H., Wei, N., White, N.M., Lauer, J., Liu, H., Shi, Y., Dumitru, C.D., Lettieri, K., et al. (2015). CMT2D neuropathy is linked to the neomorphic binding activity of glycyl-tRNA synthetase. Nature 526, 710–714.
Ito, G., Okai, T., Fujino, G., Takeda, K., Ichijo, H., Katada, T., and Iwatsubo, T. (2007). GTP binding is essential to the protein kinase activity of LRRK2, a causative gene product for familial Parkinson’s disease. Biochemistry 46, 1380–1388.
Jebelli, J.D., Dihanich, S., Civiero, L., Manzoni, C., Greggio, E., and Lewis, P. a (2012). GTP binding and intramolecular regulation by the ROC domain of Death Associated Protein Kinase 1. Sci. Rep. 2, 695.
Kabsch, W. (2010). Xds. Acta Crystallogr. Sect. D Biol. Crystallogr. 66, 125–132.
Kan, Z.Y., Mayne, L., Sevugan Chetty, P., and Englander, S.W. (2011). ExMS: Data analysis for HX-MS experiments. J. Am. Soc. Mass Spectrom. 22, 1906–1915.
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.
Krissinel, E., and Henrick, K. (2007). Inference of Macromolecular Assemblies from Crystalline State. J. Mol. Biol. 372, 774–797.
Laue, T.M., Shah, B.D., Ridgeway, T.M., and Pelletier, S.L. (1992). Analytical Ultracentrifugation in Biochemistry and Polymer Science. In Royal Society of Chemistry, pp. 90–125.
Liao, J., Wu, C.-X., Burlak, C., Zhang, S., Sahm, H., Wang, M., Zhang, Z.-Y., Vogel, K.W., Federici, M., Riddle, S.M., et al. (2014). Parkinson disease-associated mutation R1441H in LRRK2 prolongs the “active state” of its GTPase domain. Proc. Natl. Acad. Sci. U. S. A. 111, 4055–4060.
Marín, I. (2006). The Parkinson disease gene LRRK2: evolutionary and structural insights. Mol. Biol. Evol. 23, 2423–2433.
McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C., and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674.
Mosca, R., and Schneider, T.R. (2008). RAPIDO: a web server for the alignment of protein structures in the presence of conformational changes. Nucleic Acids Res. 36, 42–46.
Mosca, R., Brannetti, B., and Schneider, T.R. (2008). Alignment of protein structures in the presence of domain motions. BMC Bioinformatics 9, 352.
Murshudov, G.N., Skubák, P., Lebedev, A.A., Pannu, N.S., Steiner, R.A., Nicholls, R.A., Winn, M.D., Long, F., and Vagin, A.A. (2011). REFMAC5 for the refinement of macromolecular crystal structures. Acta Crystallogr. Sect. D Biol. Crystallogr. 67, 355– 367.
Nichols, R.J., Dzamko, N., Morrice, N.A., Campbell, D.G., Deak, M., Ordureau, A., Macartney, T., Tong, Y., Shen, J., Prescott, A.R., et al. (2010). 14-3-3 binding to LRRK2 is disrupted by multiple Parkinson’s disease-associated mutations and regulates cytoplasmic localization. Biochem. J. 430, 393–404.
Nixon-abell, J., Berwick, D.C., and Harvey, K. (2016). L ’ RRK de Triomphe : a solution for LRRK2 GTPase activity ? Biochem. Soc. Trans. 44, 1625–1634.
Pascal, B.D., Willis, S., Lauer, J.L., Landgraf, R.R., West, G.M., Marciano, D., Novick, S., Goswami, D., and Chalmers, M.J. (2012). HDX Workbench: Software for the Analysis of H/D Exchange MS Data. J Am Soc Mass Spectrom 23.
Rudi, K., Ho, F.Y., Gilsbach, B.K., Pots, H., Wittinghofer, A., Kortholt, A., and Klare, J.P. (2015). Conformational heterogeneity of the Roc domains in C. tepidum Roc-COR and implications for human LRRK2 Parkinson mutations. Biosci. Rep. 35.
Schrödinger The PyMOL Molecular Graphics System. Version 2.0 Schrödinger, LLC. Schuck, P. (2000). Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and lamm equation modeling. Biophys. J. 78, 1606–1619.
Steger, M., Tonelli, F., Ito, G., Davies, P., Trost, M., Vetter, M., Wachter, S., Lorentzen, E., Duddy, G., Wilson, S., et al. (2016). Phosphoproteomics reveals that Parkinson’s disease kinase LRRK2 regulates a subset of Rab GTPases. Elife 5.
Terheyden, S., Ho, F.Y., Gilsbach, B.K., Wittinghofer, A., and Kortholt, A. (2015). Revisiting the Roco G-protein cycle. Biochem. J. 465, 139–147.
Vetter, I.R., and Wittinghofer, a (2001). The guanine nucleotide-binding switch in three dimensions. Science 294, 1299–1304.
Walters, B.T., Ricciuti, A., Mayne, L., and Englander, S.W. (2012). Minimizing Back Exchange in the Hydrogen Exchange Experiment. J Am Soc Mass Spectrom 23, 1–9. Wang, S., Peng, J., and Xu, J. (2011). Alignment of distantly related protein structures: Algorithm, bound and implications to homology modeling. Bioinformatics 27, 2537–2545. Wang, S., Ma, J., Peng, J., and Xu, J. (2013). Protein structure alignment beyond spatial proximity. Sci. Rep. 3, 1448.
Winn, M.D., Ballard, C.C., Cowtan, K.D., Dodson, E.J., Emsley, P., Evans, P.R., Keegan, R.M., Krissinel, E.B., Leslie, A.G.W., McCoy, A., et al. (2011). Overview of the CCP4 suite and current developments. Acta Crystallogr. Sect. D Biol. Crystallogr. 67, 235–242. Wittinghofer, A., and Vetter, I.R. (2011). Structure-function relationships of the G domain, a canonical switch motif. Annu. Rev. Biochem. 80, 943–971.
Zimprich, A., Biskup, S., Leitner, P., Lichtner, P., Farrer, M., Lincoln, S., Kachergus, J., Hulihan, M., Uitti, R.J., Calne, D.B., et al. (2004). Mutations in LRRK2 Cause Autosomal-Dominant Parkinsonism with Pleomorphic Pathology. Neuron 44, 601–607.
Supplementary Information
Table S1: Affinities (KD) and Michaelis-Menten kinetic constants (KM and kcat) of the wt
Mb RocCOR, SER mutants and F652C.
Protein GDP [µM] GTPγS [µM] GppNHP [µM] KM [µM] kcat [min -1] Wildtype (wt) 1.8 (Terheyden et al., 2014) 0.18 ± 0.08 (chapter 4) 8.1 (Terheyden et al., 2014) 353 ± 129 (chapter 4) 0.063 ± 0.006 (chapter 4) single 572-575 AAA 6.95 ± 0.95 (n=2) 0.64 ± 0.14 (n=2) - 972 ± 302 (n=2) 0.033 ± 0.007 (n=2) double 572-575 AAA/590-592 AAA 1.85 ± 0.85 (n=2) 0.47 ± 0.013 (n=2) - 991± 496 (n=2) 0.027 ± 0.0013 (n=2) F652C 5.08 ± 2.34 (n=3) 0.71 ± 0.05 (n=3) - 466 ± 261 (n=3) 0.035 ± 0.007 (n=3)
Table S2: RMSD values (Å) of the comparison of Mb RocCOR-GDP (different chains)
and Mb RocCOR F652C-GDP (by RaptorX).
GDP F652C EF GH IJ AB 1.11 1.13 0.85 CD 0.81 1.37 1.1 EF 1.04 1.24 0.83 GH 0.81 1.15 1.09 IJ 0.86 1.25 1.32 KL 0.83 1.18 1.2
Table S3: RMSD values (Å) of the comparison between the different dimers of Mb
RocCOR F652C-GDP (by Raptor X).
Chains AB CD EF GH IJ CD 1.14 - - - - EF 0.96 1.06 - - - GH 1.04 0.91 0.98 - - IJ 1.1 0.86 1.06 0.76 - KL 1.21 0.91 1.19 0.88 0.95
Figure S1: Excerpt of the sequence alignment of the Ct Roco protein with Mb Roco2, Hs
Ras and Hs LRRK2. The interaction patches from the Ct LRR-RocCOR and Mb RocCOR structure are highlighted. Patch C-COR interacts with C-COR, Patch 1 and 3 only with Patch 2 and Patch 2 with Patch 1 and Patch 3. Although most patches are in the same area, Ct Patch 3 (interaction between the Roc domains) is much smaller in Mb RocCOR.
Figure S2 A-C: Plot of the fit, residuals bitmap and residuals from the SV-AUC
