Unveiling the interaction between the molecular motor Myosin Vc and the small GTPase Rab3A

Journal Pre-proof Unveiling the interaction between the molecular motor Myosin Vc and the small GTPase Rab3A

Luciano G. Dolce, Norihiko Ohbayashi, Daniel F.C. da Silva, Allan J.R. Ferrari, Renan A.S. Pirolla, Ana C. de A.P. Schwarzer, Leticia M. Zanphorlin, Lucelia Cabral, Mariana Fioramonte, Carlos H.I. Ramos, Fabio Cesar Gozzo, Mitsunori Fukuda, Priscila O. de Giuseppe, Mário T. Murakami PII:

S1874-3919(19)30321-5

DOI:

https://doi.org/10.1016/j.jprot.2019.103549

Reference:

JPROT 103549

To appear in:

Journal of Proteomics

Received date:

16 July 2019

Revised date:

25 September 2019

Accepted date:

8 October 2019

Please cite this article as: L.G. Dolce, N. Ohbayashi, D.F.C. da Silva, et al., Unveiling the interaction between the molecular motor Myosin Vc and the small GTPase Rab3A, Journal of Proteomics (2018), https://doi.org/10.1016/j.jprot.2019.103549

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2018 Published by Elsevier.

Journal Pre-proof Unveiling the interaction between the molecular motor Myosin Vc and the small GTPase Rab3A

Luciano G. Dolce1,2, Norihiko Ohbayashi5,6, Daniel F. C. da Silva3, Allan J. R. Ferrari4, Renan A. S. Pirolla3,4, Ana C. de A. P. Schwarzer2, Leticia M. Zanphorlin3, Lucelia Cabral3, Mariana Fioramonte4, Carlos H. I. Ramos7, Fabio Cesar Gozzo4, Mitsunori Fukuda5, Priscila O. de Giuseppe2,3,** [email protected] and Mário T. Murakami2,3,* [email protected] Graduate Program in Functional and Molecular Biology, Institute of Biology,

of

1

University of Campinas, Campinas, SP, Brazil

Brazilian Biosciences National Laboratory (LNBio), Brazilian Center for Research in

ro

2

Energy and Materials (CNPEM), Zip Code 13083-100, Campinas, Sao Paulo, Brazil Brazilian Biorenewables National Laboratory (LNBR), Brazilian Center for Research

-p

3

in Energy and Materials (CNPEM), Zip Code 13083-100, Campinas, Sao Paulo, Brazil 4

re

Dalton Mass Spectrometry Laboratory, Institute of Chemistry, University of

Campinas, Zip Code 13083-970, Campinas, Sao Paulo, Brazil Laboratory of Membrane Trafficking Mechanisms, Department of Integrative Life

lP

5

Sciences, Graduate School of Life Sciences, Tohoku University, Aobayama, Aoba-ku, 6

na

Sendai, Miyagi 980-8578, Japan

Department of Physiological Chemistry, Faculty of Medicine and Graduate School of

Jo ur

Comprehensive Human Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8575, Japan 7

Institute of Chemistry, University of Campinas, Zip Code 13083-970, Campinas, Sao

Paulo, Brazil *

Correspondence to: Mario Tyago Murakami, Brazilian Biorenewables National

Laboratory (LNBR), Brazilian Center for Research in Energy and Materials (CNPEM), Zip Code 13083-100, Campinas, Sao Paulo, Brazil. **

Correspondence to: Priscila Oliveira de Giuseppe, Brazilian Biorenewables National

Laboratory (LNBR), Brazilian Center for Research in Energy and Materials (CNPEM), Zip Code 13083-100, Campinas, Sao Paulo, Brazil.

Journal Pre-proof Abstract

Vertebrates usually have three class V myosin paralogues (MyoV) to control membrane trafficking in the actin-rich cell cortex, but their functional overlapping or differentiation through cargoes selectivity is yet only partially understood. In this work, we reveal that the globular tail domain of MyoVc binds to the active form of small GTPase Rab3A with nanomolar affinity, a feature shared with MyoVa but not with MyoVb. Using molecular docking analyses guided by chemical cross-linking restraints, we propose a model to explain how Rab3A selectively recognizes MyoVa and MyoVc

of

via a distinct binding site from that used by Rab11A. The MyoVa/c binding interface involves multiple residues from both lobules (I and II) and the short helix at the α2-α3

ro

link region, which is conserved between MyoVa and MyoVc, but not in MyoVb. This motif is also responsible for the selective binding of RILPL2 by MyoVa and potentially

-p

MyoVc. Together, these findings support the selective recruitment of MyoVa and

re

MyoVc to exocytic pathways via Rab3A and expand our knowledge about the

lP

functional evolution of class V myosins.

Significance

na

Hormone secretion, neurotransmitter release, and cytoplasm membrane recycling are examples of processes that rely on the interaction of molecular motors and Rab

Jo ur

GTPases to regulate the intracellular trafficking and tethering of vesicles. Defects in these proteins may cause neurological impairment, immunodeficiency, and other severe disorders, being fatal in some cases. Despite their crucial roles, little is known about how these molecular motors are selectively recruited by specific members of the large family of Rab GTPases. In this study, we unveil the interaction between the actin-based molecular motor Myosin Vc and the small GTPase Rab3A, a key coordinator of vesicle trafficking and exocytosis in mammalian cells. Moreover, we propose a model for their recognition and demonstrate that Rab3A specifically binds to the globular tail of Myosins Va and Vc, but not of Myosin Vb, advancing our knowledge about the molecular basis for the selective recruitment of class V myosins by Rab GTPases.

Keywords: molecular motor, myosin V, globular tail domain, Rab3A, protein complex, membrane trafficking

Journal Pre-proof Abbreviations AGC, automatic gain control; DMF, dimethylformamide; DSS, disuccinimidyl suberate; DTT, dithiothreitol; ESI, electron spray ionization; GTD, globular tail domain; HRP, horseradish

peroxidase;

IPTG,

isopropyl

β-D-1-thiogalactopyranoside;

MALS,

multiangle light scattering; MyoV, class V myosin; MST, microscale thermophoresis; O.D., optical density; TCEP, tris(2-carboxyethyl)phosphine; RMSD., root mean square deviation; SEC, size exclusion chromatography; XL/MS, chemical cross-linking coupled to mass spectrometry.

of

Introduction

ro

Class V Myosins (MyoV) are motor proteins that tether or transport several

-p

cargoes, such as organelles, mRNAs and vesicles on actin filament tracks [1], playing key roles in exocytosis, cell polarity, cilia assembly and plasma membrane homeostasis

re

[2-6]. In humans, three paralogous genes encode for class V myosins: MYO5A, MYO5B, and MYO5C [7]. The three genes are expressed in several organs, being MYO5A most

lP

abundantly expressed in the brain, MYO5B in the intestine and MYO5C in endocrine tissues [8, 9]. Defects in MYO5A have been associated with the Griscelli Syndrome type

na

1, characterized by neurological disorders and partial albinism [10]. Loss of MYO5B function causes severe neonatal diarrhea due to microvillus atrophy [11]. However, no

Jo ur

genetic disease has been associated with MYO5C so far, which hampers a better understanding of the physiological roles of this molecular motor. Structurally, the proteins MyoVa, MyoVb, and MyoVc share a similar domain architecture composed by a motor domain called head, which hydrolyzes ATP and binds to actin; a neck, which controls the step size of 36 nm; a coiled-coil region, to form homodimers; and a tail, which harbors a globular domain involved in cargo recognition and self-regulation [12-17]. Some molecular mechanisms of MyoV recruitment have been elucidated for MyoVa and MyoVb and involve the binding of protein partners to the globular tail domain [18-20]. These studies revealed, for example, how melanophilin specifically recruits MyoVa to melanosomes and how Rab11 can attach MyoVa or MyoVb to recycling membrane compartments. However, the molecular basis for MyoVc recruitment is still poorly understood. Functional studies indicate that MyoVc localizes in secretory granules[21] and late secretory vesicles[22], playing a role in several secretory pathways, such as

Journal Pre-proof transferrin receptor transport[8]; biogenesis and secretion of melanosomes[23] and polarization of epithelial cells[24]. Yeast two-hybrid and cellular colocalization experiments suggest that Rab8A[8], Rab10[25], Rab32 and Rab38[23] are putative adaptor proteins that interact with the medial tail of MyoVc, but none of these interactions have been structurally characterized. Moreover, molecular partners that selectively recognize the globular tail domain (GTD) of MyoVc are still elusive, since the partners of this domain reported so far, Rab7A and Spir-2 seem to recognize the three mammalian MyoV [18, 23]. In this work, we reveal that the small GTPase Rab3A binds to the globular tail

of

domain of MyoVc. Our studies demonstrate that the constitutively active state of Rab3A interacts with MyoVc-GTD and MyoVa-GTD, but not with MyoVb-GTD, indicating a

ro

potential functional overlap between MyoVa and MyoVc motors. Using molecular

-p

docking analyses guided by chemical cross-linking restraints, we propose a model for the complex between Rab3A and the globular tails of MyoVa and MyoVc, providing

re

structural insights about the selective recruitment of MyoVa and MyoVc to membranes

GST Pull-down Assay

na

Materials and Methods

lP

via Rab3A.

Jo ur

COS-7 cells were cultured in Dulbecco’s modified Eagle’s medium (Wako Pure Chemical Industries, Ltd., Osaka, Japan) supplemented with 10% fetal bovine serum, 100 U/mL penicillin G and 100 μg/mL streptomycin, at 37 °C and 5% CO2. COS-7 cells (1 x 106 cells/10-cm dish) were transfected with pEF-FLAG-MyoVb-tail (amino acids 1090-1818) or pEF-FLAG-MyoVc-tail (amino acids 1054-1742) [26] by using Lipofectamine-LTX Plus (Invitrogen) according to the manufacturer’s instructions. At 36 hr after the transfection, the cells were harvested and lysed with a cell lysis buffer (50 mM HEPES-KOH, pH7.2, 150 mM NaCl, 1 mM MgCl2, 1% Triton X-100, and complete EDTA-free protease inhibitor cocktail (Roche)). Glutathione-Sepharose beads (wet vol. 10 µL; GE Healthcare) coupled with GST-Rabs [27, 28] (2 µg each) were treated with the cell lysis buffer containing 0.5 mM GTPγS + 2.5 mM EDTA on ice for 15 min, then treated with 10 mM MgCl2 on ice for 15 min, and incubated with FLAGMyo-tail (Vb/Vc)-expressing COS-7 cell lysates at 4 °C for 2 hr. The beads were washed three times with a washing buffer (50 mM HEPES-KOH, pH7.2, 150 mM

Journal Pre-proof NaCl, 1 mM MgCl2, and 0.2% Triton X-100) and the FLAG-Myo-tail proteins bound to the beads were analyzed by 10% SDS–PAGE followed by immunoblotting with HRPconjugated anti-FLAG tag antibody (1:10,000 dilution, Sigma) and anti-GST antibody (1:5000 dilution, Santa Cruz).

Protein Expression and Purification The proteins 6xHis-MyoV-GTD (Va and Vb) were produced in the Escherichia coli strain BL21(DE3)∆SlyD harboring the plasmid pRARE2, in selective Lysogeny

of

Broth, by adding 0.1 mM IPTG at the culture (O.D.600 nm ~ 0.5) and incubating the cells at 25 ºC for 4 hr, 175 rpm. The protein 6xHis-MyoVc-GTD was produced in the same

ro

E. coli strain grown in selective Terrific Broth, supplemented with 0.5 mM IPTG after O.D.600 nm ~ 1, and then incubated at 20 ºC for 16 hr, 175 rpm. The target proteins were

-p

purified by two chromatographic steps (affinity chromatography and SEC) as described

re

in [13]. The size-exclusion chromatography was carried out in the interaction buffer (10 mM HEPES, 150 mM NaCl, 5% glycerol, 1 mM MgCl2, 0.5 mM TCEP, pH 7.5).

lP

GST-Rab3AQ81L [27, 28] was expressed in the E. coli strain BL21(DE3)∆SlyD harboring the plasmid pRARE2, in selective Lysogeny Broth, supplementing the culture

na

with 0.1 mM IPTG at O.D.600 nm ~ 1.5, and incubating the cells at 18 ºC, for 16 hr, 180 rpm. The cell lysate (in lysis buffer containing 50 mM HEPES, 500 mM NaCl, 5% glycerol, 1 mM MgCl2, 0.5 mM TCEP, pH 7.5 supplemented with 0.1 mg/mL

Jo ur

lysozyme, and SIGMAFAST™ Protease Inhibitor Tablets (Sigma-Aldrich)) was clarified by centrifugation (30.000 x g, 50 min), and injected in two Glutathione Sepharose GSTrap FF 5 mL columns coupled in tandem (GE Healthcare) using a FPLC system (GE Healthcare). The columns were washed with 100 ml of wash buffer (50 mM HEPES, 500 mM NaCl, 5% glycerol, 1 mM MgCl2 and 0.5 mM TCEP, at pH 7.5), and incubated with the GTP buffer (wash buffer + 1 mM GTP and 2.5 mM EDTA). After 30 min of incubation, the columns were rinsed with the wash buffer + 10 mM MgCl2, to stabilize the complex between Rab3AQ81L and GTP (Rab3AQ81L-GTP), and finally incubated with 0.03 mg/mL Thrombin for 16 h, at 4 °C, to remove the GST tag. After the proteolysis, Rab3AQ81L-GTP was eluted from the column with wash buffer, and further purified on a HiLoad 16/600 Superdex 200 column (GE Healthcare) equilibrated with the interaction buffer.

Journal Pre-proof SEC and SEC-MALS For SEC analysis, MyoVc-GTD at 5 mg/mL and Rab3AQ81L-GTP at 1.5 mg/mL were incubated for 30 min at 4 ºC in buffer (10 mM HEPES, 200 mM NaCl, 5% glycerol and 0.5 mM TCEP, at pH 7.5) and injected onto a HiLoad 16/600 Superdex 200 column (GE Healthcare) coupled to an FPLC system (GE Healthcare) using a flow rate of 0.5 mL/min. Input and eluted samples were analyzed by 13% SDS-PAGE stained with Coomassie Brilliant Blue [29]. The SEC-MALS experiment was performed in an ÄKTA purifier (GE Healthcare) equipment coupled to MALS detectors miniDAWN™ TREOS and

of

Optilab® T-rEX, using a Superdex 200 HR 10/300 GL column (GE Healthcare). A mixture of MyoVc-GTD (15 µM) and Rab3AQ81L-GTP (55 µM), as well as the individual

ro

protein samples at 29 µM (MyoVc-GTD) or 55 µM (Rab3AQ81L-GTP), were injected in

-p

the column equilibrated with the interaction buffer using a 0.5 mL/min flow. The data

lP

Microscale Thermophoresis (MST)

re

were processed using the ASTRA V software (Wyatt Technology).

For the MST assays, the 6xHis-MyoVa-GTD, 6xHis-MyoVb-GTD, 6xHis-

na

MyoVc-GTD, and Rab3A samples were purified by SEC (Superdex 200 or 75, 16/600) in the interaction buffer. 6xHis-tagged proteins, at a final concentration of 100 nM, were

Jo ur

incubated with 50 nM RED-tris-NTA probe (NanoTemper) according to the kit manual. After 30 min of incubation at room temperature, the samples were centrifuged for 10 min, 15,000 x g, 4 °C.

MST measurements were made using 50 nM of the labeled 6xHis-MyoVa-GTD, 6xHis-MyoVb-GTD or 6xHis-MyoVc-GTD and a serial dilution of Rab3AQ81L-GTP. Samples previously centrifuged at 15,000 x g, 4 °C for 5 min were loaded onto Monolith™ NT.115 MST Premium capillaries (NanoTemper Technologies) and the thermophoresis data were measured on a Monolith™ NT.115 (NanoTemper Technologies) equipment using the intrinsic fluorescence of RED-tris-NTA probe. The intensity of the LED was adjusted to 40%, and the intensity of MST was varied in two experiments, 40% and 60%.

Journal Pre-proof Chemical Cross-linking Coupled to Mass Spectrometry (XL/MS) Cross-linking reactions were performed as described previously [30]. The protein complex (7.5 nmol) was incubated in 20 mM HEPES buffer (pH 7.5) with DSS (stock at 10 mg/mL in DMF) in a 1:100 molar ratio (complex:DSS) for 2 hr at 27 °C and the cross-linking reaction was quenched with 100 mM ammonium bicarbonate (pH 8.0). Reduction of cysteine residues was performed using 10 mM dithiothreitol (DTT) for 30 min at 60 °C followed by alkylation with 30 mM iodoacetamide for 30 min at room temperature. Each sample was digested by trypsin (1:30 m/m trypsin:complex) at 37 °C overnight, generating peptides that were analyzed by nanoLC coupled with an

of

Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific). Peptides were separated on a nanoLC system (Ultimate 3000 RSLCnano,

ro

Thermo Fisher Scientific) operating in a reverse-phase mode at a flow rate of 300

-p

nL/min using a 77 min gradient of mobile phase A (5% acetonitrile and 0.1% formic acid in water) and B (95% acetonitrile and 0.1% formic acid in water). The system was

re

connected to a nano-ESI source coupled to Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific). The spray voltage was set to 2.2 kV with a capillary

lP

temperature of 275 °C. Each MS scan was acquired in the orbitrap over a mass range of m/z 400–2000 at a resolution of 120 K, maximum injection time 50 ms, automatic gain

na

control (AGC) target 1E6 and followed by a data-dependent acquisition mode. The most intense signals (Top20) in the mass spectrum were selected for higher-energy collisional

Jo ur

dissociation (HCD) applying stepped normalized collision energy (sNCE) of 30−40. Fragments were detected by Orbitrap at a resolution of 15 K, maximum injection time 200 ms and AGC target 2E5. Cross-linked peptides were identified using the SIM-XL [31] software and the following search parameters: XL reaction site K-K, K-S and K-Nterm; mass accuracy of 20 ppm in precursor ion, trypsin fully specific protease and a maximum of 3 missed cleavages, carbamidomethylation of cysteine as fixed modification and oxidation of methionine as variable modification. The initial identification was manually validated assessing: (1) unassigned intense fragment ions; (2) dubious cross-linking site due to lack of flanking fragment ions; (3) abnormal error distribution, and (4) insufficient fragmentation of a chain.

Journal Pre-proof Molecular Docking As MyoVc crystal structure does not contain the residues 1524-1544 (PDB 4L8T), 100 full-length structure candidates were generated using the hybridization protocol [32] in Rosetta and fragments of known structures to build the missing region [33]. The lowest scored model of MyoVc and the crystal structure of Rab3A were used as inputs for the docking protocol. Starting configuration for docking was generated by manually orienting both chains in PyMol [34] in order that cross-linked residues identified in the chemical cross-linking experiment were placed in the interface of interaction and close in space (Cβ-Cβ atoms < 20 Å). RosettaDock [35] was applied to produce candidates of

of

interaction. The docking protocol in Rosetta works in two stages aiming to simulate the physical encounter of proteins and maximize interactions that result in the complex

ro

formation. In the first stage, named global docking, a rigid body translation-rotation of

-p

one of the protein partners is used to sample possible interaction modes using a low resolution, centroid-based, representation of side-chains. In the second stage, named

re

local docking, centroid mode is converted to full-atom mode representation, small random rigid body perturbations take place followed by side chain minimization that

lP

aims to optimize local interactions. Our protocol decoupled these two stages. Rosetta energy function was modified to include information from the two constraints from the

na

chemical cross-linking experiment with statistical limits modeling the experimental data as described previously [36, 37]. 50,000 models were generated using the global

Jo ur

docking stage with 8 Å translation and 10 Å rotation perturbation parameters. Topolink [36, 38] was applied to evaluate and validate models consistent with experimental data. 8,000 validated models proceeded to the local docking stage using default parameters (3 Å translation and 0.1 Å rotation perturbation parameters). For each input model, 10 output models were generated. All resulting models were minimized using the Rosetta Relax application [39] with default parameters to further optimize protein complex geometry and total score. This last step can work by introducing small backbone conformation perturbations which can occur upon binding. Ultimately, the lowest scored models were assumed to best represent the physical interaction. Twenty lowest scored models were visually inspected for convergence (< 3Å pairwise RMSD). In the case of divergence, 500 lowest scored models were submitted to local docking protocol (10 output models for each input) followed by minimization. After 3 iterations, we observed the convergence of our strategy. The final subset of models was validated with Topolink resulting in 19 models, 17 of which were within an R.M.S.D. of 1.7 Å,

Journal Pre-proof revealing a preference for a specific biding mode validated by the experimental constraints.

Results Rab3A specifically interacts with MyoVc-GTD and MyoVa-GTD To systematically investigate the ability of MyoVc to interact with Rab proteins, we performed a GST pull-down screen using MyoVc tail tagged with FLAG against 41member proteins of the Rab family (GTPS-bound active form) fused to GST (Figure

of

1). In this assay, MyoVc tail was more efficiently pulled down by Rab3A, followed by Rab37 (Figure 1). No interaction with Rab7 was detected, in contrast to previously

ro

reported data using yeast two-hybrid assays [23]. As a control, we performed the same assay for the MyoVb tail, which was pulled down by Rab11A and Rab8A, as expected

-p

[20, 40], but not by Rab3A, indicating the selective recognition of MyoV paralogs by

Jo ur

na

lP

re

specific Rabs.

Figure 1: GST pull-down screen of 41 GST-tagged Rabs incubated with FLAG-tagged MyoVb or MyoVc tails. Note that Rab3A and Rab37 specifically pulled down MyoVc tail, whereas Rab11A and Rab8A only recognized the MyoVb tail.

Next, to investigate whether the GTD of MyoVc tail was the domain recognized by Rab3A, we performed binding assays in vitro using individually expressed and purified MyoVc-GTD and Rab3AQ81L-GTP, a mutant of Rab3A that displays lower GTP hydrolysis rate, thus favoring the GTP-bound conformation responsible for recruiting protein effectors [41]. Microscale thermophoresis (MST) data revealed that MyoVcGTD binds to Rab3AQ81L-GTP with a Kd of 390 ± 20 nM (Figure 2a). Since Rab3A has been previously reported to interact with the tail of MyoVa [42], we also studied the interaction between MyoVa-GTD and Rab3AQ81L-GTP. The Kd estimated for the MyoVa-

Journal Pre-proof GTD:Rab3AQ81L-GTP complex was 284 ± 33 nM (Figure 2b), which was very similar to that involving MyoVc-GTD, despite the relatively low sequence identity shared between these two paralogous GTDs (64%). Corroborating the pull-down results, Rab3AQ81L-GTP did not interact with MyoVb-GTD in MST assays (Figure 2c), showing

Jo ur

na

lP

re

-p

ro

of

that this small GTPase specifically recognizes MyoVa and MyoVc GTDs.

Figure 2. Rab3AQ81L-GTP binds to MyoVc and MyoVa GTDs but not in MyoVbGTD. MST experiments of Rab3AQ81L-GTP against (a) MyoVc-GTD; (b) MyoVa-GTD; (c) MyoVb-GTD. The Kd was determined using the Hill1 fit (black) from Origin 8.0. Fnorm = normalized fluorescence. Data are presented as mean ± SD (error bars) from triplicates.

Journal Pre-proof Structural basis for selective recognition of MyoVc and MyoVa by Rab3A In solution analyses using size-exclusion chromatography coupled to Multiangle light scattering (SEC-MALS) showed that free Rab3AQ81L-GTP and free MyoVcGTD are monomers in solution and form a complex of 67 ± 9 kDa, compatible with the theoretical mass of one molecule of MyoVc-GTD (49 kDa) and one molecule of Rab3AQ81L (25 kDa) (Figure 3a-c). Although the elution peak of the protein complex overlaps with that of MyoVc-GTD alone, SDS-PAGE analysis of SEC fractions confirmed that a subpopulation of Rab3AQ81L-GTP co-elutes with MyoVc-GTD,

Jo ur

na

lP

re

-p

ro

of

corroborating our SEC-MALS results (Figure 3d and e).

Figure 3. SEC-MALS analyses of (a) Rab3AQ81L-GTP; (b) MyoVc-GTD and (c) the mixture of Rab3AQ81L-GTP and MyoVc-GTD. (d) SEC analysis of a mixture containing Rab3AQ81L-GTP and MyoVc-GTD loaded into a Superdex 200 pg 16/600 column. (e) SDS-PAGE analysis of samples from the input (I) and fractions from the peaks 1 and 2 indicated in the chromatogram of panel d. Note that part of the Rab3AQ81L-GTP population co-eluted with MyoVc-GTD in peak 1. M = molecular mass standard.

Journal Pre-proof

As attempts to crystallize this complex were unsuccessful, we employed molecular docking analyses guided by experimental data from chemical cross-linking assays to characterize the complex interface. From the list of cross-links identified by mass spectrometry and validated, six were intermolecular cross-links connecting the residues

MyoVc1620-Rab3A12,

MyoVc1618-Rab3AN-terminus,

MyoVc1618-Rab3A62,

MyoVc1383-Rab3AN-terminus, MyoVc1620-Rab3A72, MyoVc1690-Rab3AN-terminus (Figure S1, Table S1). However, only two constraints (MyoVc1618-Rab3A62 and MyoVc1620Rab3A72) were selected to guide our molecular docking analysis, since the other

of

intermolecular cross-links were located in disordered regions of Rab3A not present in the available crystallographic structures (Figure 4) [41]. The molecular docking analysis

ro

guided by the MyoVc1618-Rab3A62 and MyoVc1620-Rab3A72 cross-links converged to 17

Jo ur

na

lP

re

-p

solutions of lower energy with an all-atom RMSD of 1.8 Å (Figure 5).

Figure 4. Chemical cross-links identified by mass spectrometry provide distance constraints to model the Rab3AMyoVc-GTD complex. Panels a and b show MS/MS

Journal Pre-proof (product-ion scans) spectra of two DSS cross-links between Rab3AQ81L-GTP and MyoVcGTD identified with SIM-XL software [31]. In blue are highlighted the fragment ions referent to the longer peptides (α-chain) whereas in red are highlighted those from the shorter peptide (β-chain) composing the cross-linked molecule. The fragment ions were generated by higher energy collisional dissociation of precursor ions. Dashed lines

Jo ur

na

lP

re

-p

ro

of

indicate which residues were connected by the chemical cross-linker.

Figure 5. Molecular docking models guided by chemical cross-linking restraints converged to similar solutions. a. Schematic representation of MyoVc-GTD and Rab3A sequences labeling the residues linked by the two chemical cross-links (green and red lines) identified by MS/MS analysis and selected to guide modeling. b. Model of lowest energy representative of 17 convergent docking solutions (R.M.S.D = 1.7 Å) highlighting the topological distances (solvent-accessible paths, red and green lines) of the chemical cross-links found between MyoVc-GTD and Rab3A peptides. Numbers

Journal Pre-proof represent topological distances in Å between the Cβ of cross-linked residues. N = Nterminus and C = C-terminus. Molecular docking data indicate that Rab3AGTP likely binds to the convex region of MyoVc-GTD formed by the α2-α3 link-region, the helices α7 and α9, and part of the phospho-loop (Figure 6). The same region, specifically the α2-α3 link-region, was addressed to play a role in the selective binding of RILPL2 by MyoV paralogues [13, 19], supporting our results. Indeed, this region is divergent between MyoVa and MyoVb but is conserved between MyoVa and MyoVc, even though MyoVa-GTD is more

of

similar to MyoVb-GTD (71% sequence identity) than to MyoVc-GTD (64 % sequence identity), reinforcing our finding.

ro

To gather insight about the mechanisms involved in the selective recognition of

-p

MyoVc and Va by Rab3A, we mapped the residues conserved exclusively between MyoVc and Va in the predicted binding site and found six sites that might be crucial for

re

Rab3A-binding: V1387, N1390, R1385, Q1564, D1619, S1550 (MyoVc-numbering). These residues are absent (V1387) or replaced by residues containing shorter side-

lP

chains in MyoVb, probably compromising the formation of hydrophobic (V1387) and polar contacts with Rab3A (Figure 6). The probable site on the Rab3A surface where

na

MyoVc-GTD binds include residues from Switch I, α1, α5, β2 and β3 (Figure 7a and b). Most of them are conserved or semi-conserved in Rab3B, Rab3C, and Rab3D, but

Jo ur

divergent in Rab11A (Figure 7c). These results agree with the capacity of active mutants of Rab3B, Rab3C and Rab3D to bind to MyoVc-GTD, which is not observed for Rab11A in the pull-down assays (Figure 7d).

re

-p

ro

of

Journal Pre-proof

Jo ur

na

lP

Figure 6. Rab3A-binding site in MyoVc predicted from molecular docking analyses guided by chemical cross-linking restraints. a. The surface of MyoVc-GTD highlighting the most probable region of Rab3A-binding colored according to the frequency that these residues appeared buried at the complex interface of the most convergent models (n=17, R.M.S.D. = 1.7 Å). On the right, the Rab3A-binding score is presented as a function of the MyoVc-GTD primary structure and the sequences of the hot-spots are shown and color-coded according to their Rab3A-binding scores. Gray residues = no interaction. b. Cartoon representation of MyoVc-GTD highlighting residues from the α2-α3 link-region and from helices α7 and α9 that are identical in MyoVc and MyoVa but divergent in MyoVb. Although some replacements are semiconservative (K1620R, L1552I), most of them gave rise to shorter side-chains in MyoVb that probably compromise the formation of polar contacts with Rab3A. Note that there is no residue equivalent to V1387 in MyoVb, which might represent a loss of hydrophobic contacts required for Rab3A-binding.

re

-p

ro

of

Journal Pre-proof

Figure 7. The most probable binding site of MyoVc-GTD on Rab3A surface is

lP

conserved in Rab3 paralogues but divergent in Rab11A. a. The surface of Rab3A highlighting the most probable region of MyoVc-GTD binding colored according to the

na

frequency that these residues appeared buried at the complex interface of the most convergent models (n=17, R.M.S.D. = 1.7 Å). In the right, the MyoVc-binding score is

Jo ur

presented as a function of the Rab3A primary structure and the sequences of the hotspots are shown and color-coded according to their MyoVc-binding scores. Gray residues = no interaction. b. Cartoon representation of Rab3A showing GTP and Mg2+as spheres and coloring residues from Switch I, α1, α5, β2, and β3 according to their MyoVc-binding scores. c. Sequence alignment of Rab3A with Rab3B, Rab3C, Rab3D and Rab11A highlighting the residues probably involved in MyoVc-binding (orange boxes) and their conservation (black), semi-conservation (grey) or non-conservation (red) in the compared sequences. d. Western blotting analysis of pull-down assays of FLAG-MyoVc-tail with active forms of GST-Rabs.

Journal Pre-proof

Discussion In this work, we reveal MyoVc as a novel effector of Rab3A and provide structural insights on how this small GTPase selectively recognizes MyoVc and MyoVa, but not MyoVb. This finding advances previous studies about the interaction between Rab3A and MyoVa tail [42, 43] and indicates a functional overlap between MyoVa and MyoVc in Rab3A binding, which is consistent with the similar roles reported for these myosins in the trafficking and exocytosis of secretory granules [21,

of

44].

ro

Besides confirming a previous work showing the direct interaction between MyoVa tail and Rab3A-GTP in vitro [43], our results demonstrate that the GTD of

-p

MyoVa is sufficient to bind to the GTP-bound form of Rab3A. Interestingly, one of the Rab3A effectors, the protein rabphilin-3A, also seems to interact with the tail of MyoVa

re

[44]. The interface between Rab3A and rabphilin-3A is known with atomic detail [41]

lP

but how rabphilin-3A interacts with MyoVa remains unclear. Our models for the Rab3A:MyoVc-GTD complex indicate that the MyoVc(a)-binding site in Rab3A is

na

distinct from that of rabphilin-3A (PDB 1ZBD, [41]). In this context, our data provide a framework for future studies aiming to investigate whether these proteins can form a tripartite complex as do MyoVa, melanophilin, and Rab27A [45].

Jo ur

According to our in vitro studies, the GTP-bound form of Rab3A binds to the GTD of MyoVa and MyoVc with nanomolar affinity, in a similar range to that reported for complexes between MyoVb-GTD and Rab11GTP [20]. However, unlike Rab11A, which recognizes MyoVa and MyoVb [18, 20], Rab3A uses another binding site in the GTD to selectively recruit MyoVa and MyoVc. The few studies about MyoVc published so far have evidenced a role for MyoVc in the exocytosis and trafficking of secretory vesicles and the importance of its GTD for targeting to secretory granules [21, 22, 46]. However, the molecular recruiter of MyoVc-GTD in such cargoes was elusive so far. Thus, the data presented here provide a possible mechanism for the recruitment of MyoVc to secretory vesicles via binding to Rab3A, unraveling a link between this actin-based molecular motor and a key coordinator of vesicle trafficking and exocytosis in mammalian cells [47-49].

Journal Pre-proof

Acknowledgments We would like to thank the Brazilian Biosciences National Laboratory (LNBio) for the provision of time on the LEC facility. This work was supported by FAPESP #2014/09720-9 to MTM, #2016/13195-2 to AJRF and #2014/00584-5 to LGD and CNPq #478059/2009-4 and 486841/2012-0 to MTM and #150552/2017-3 to LC.

Jo ur

na

lP

re

-p

ro

of

Author contributions Luciano G. Dolce – Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Validation, Visualization, Writing (original draft), Writing (review & editing) Norihiko Ohbayashi - Conceptualization, Data curation, Investigation, Methodology, Validation, Visualization, Writing (review & editing) Daniel F. C. da Silva - Conceptualization, Data curation, Investigation, Methodology, Validation, Writing (review & editing) Allan J. R. Ferrari - Data curation, Formal analysis, Investigation, Methodology, Validation, Writing (original draft), Writing (review & editing) Renan A. S. Pirolla – Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Writing (original draft), Writing (review & editing) Ana C. de A. P. Schwarzer - Data curation, Investigation, Validation, Writing (review & editing) Leticia M. Zanphorlin - Formal analysis, Investigation, Validation, Writing (review & editing) Lucelia Cabral - Funding acquisition, Investigation, Writing (review & editing) Mariana Fioramonte – Methodology, Writing (review & editing) Carlos H. I. Ramos – Resources, Methodology, Writing (review & editing) Fabio Cesar Gozzo - Resources, Methodology, Writing (review & editing) Mitsunori Fukuda – Conceptualization, Resources, Supervision, Methodology, Writing (review & editing) Priscila O. de Giuseppe - Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Supervision, Visualization, Writing (original draft), Writing (review & editing) Mário T. Murakami – Conceptualization, Funding Acquisition, Project administration, Resources, Supervision, Writing (review & editing)

References [1] C. Desnos, S. Huet, F. Darchen, 'Should I stay or should I go?': myosin V function in organelle trafficking, Biol Cell 99(8) (2007) 411-23. [2] R. Rudolf, C.M. Bittins, H.H. Gerdes, The role of myosin V in exocytosis and synaptic plasticity, J Neurochem 116(2) (2011) 177-91. [3] J.A. Hammer, 3rd, W. Wagner, Functions of class V myosins in neurons, J Biol Chem 288(40) (2013) 28428-34. [4] Sonal, J. Sidhaye, M. Phatak, S. Banerjee, A. Mulay, O. Deshpande, S. Bhide, T. Jacob, I. Gehring, C. Nuesslein-Volhard, M. Sonawane, Myosin Vb mediated plasma membrane

Journal Pre-proof

Jo ur

na

lP

re

-p

ro

of

homeostasis regulates peridermal cell size and maintains tissue homeostasis in the zebrafish epidermis, PLoS Genet 10(9) (2014) e1004614. [5] L.H. Assis, R.M. Silva-Junior, L.G. Dolce, M.R. Alborghetti, R.V. Honorato, A.F. Nascimento, T.D. Melo-Hanchuk, D.M. Trindade, C.C. Tonoli, C.T. Santos, P.S. Oliveira, R.E. Larson, J. Kobarg, E.M. Espreafico, P.O. Giuseppe, M.T. Murakami, The molecular motor Myosin Va interacts with the cilia-centrosomal protein RPGRIP1L, Sci Rep 7 (2017) 43692. [6] C.T. Wu, H.Y. Chen, T.K. Tang, Myosin-Va is required for preciliary vesicle transportation to the mother centriole during ciliogenesis, Nat Cell Biol 20(2) (2018) 175-185. [7] J.S. Berg, B.C. Powell, R.E. Cheney, A Millennial Myosin Census, Mol Biol Cell2001, pp. 78094. [8] O.C. Rodriguez, R.E. Cheney, Human myosin-Vc is a novel class V myosin expressed in epithelial cells, J Cell Sci 115(Pt 5) (2002) 991-1004. [9] M. Uhlen, P. Oksvold, L. Fagerberg, E. Lundberg, K. Jonasson, M. Forsberg, M. Zwahlen, C. Kampf, K. Wester, S. Hober, H. Wernerus, L. Bjorling, F. Ponten, Towards a knowledge-based Human Protein Atlas, Nat Biotechnol 28(12) (2010) 1248-50. [10] E. Pastural, F.J. Barrat, R. Dufourcq-Lagelouse, S. Certain, O. Sanal, N. Jabado, R. Seger, C. Griscelli, A. Fischer, G. de Saint Basile, Griscelli disease maps to chromosome 15q21 and is associated with mutations in the myosin-Va gene, Nat Genet 16(3) (1997) 289-92. [11] T. Muller, M.W. Hess, N. Schiefermeier, K. Pfaller, H.L. Ebner, P. Heinz-Erian, H. Ponstingl, J. Partsch, B. Rollinghoff, H. Kohler, T. Berger, H. Lenhartz, B. Schlenck, R.J. Houwen, C.J. Taylor, H. Zoller, S. Lechner, O. Goulet, G. Utermann, F.M. Ruemmele, L.A. Huber, A.R. Janecke, MYO5B mutations cause microvillus inclusion disease and disrupt epithelial cell polarity, Nat Genet 40(10) (2008) 1163-5. [12] K.M. Trybus, Myosin V from head to tail, Cell Mol Life Sci 65(9) (2008) 1378-89. [13] A.F. Nascimento, D.M. Trindade, C.C. Tonoli, P.O. de Giuseppe, L.H. Assis, R.V. Honorato, P.S. de Oliveira, P. Mahajan, N.A. Burgess-Brown, F. von Delft, R.E. Larson, M.T. Murakami, Structural insights into functional overlapping and differentiation among myosin V motors, J Biol Chem 288(47) (2013) 34131-45. [14] X.D. Li, H.S. Jung, Q. Wang, R. Ikebe, R. Craig, M. Ikebe, The globular tail domain puts on the brake to stop the ATPase cycle of myosin Va, Proc Natl Acad Sci U S A 105(4) (2008) 1140-5. [15] M. Ikebe, Regulation of the function of mammalian myosin and its conformational change, Biochem Biophys Res Commun 369(1) (2008) 157-64. [16] J.F. Li, A. Nebenfuhr, The tail that wags the dog: the globular tail domain defines the function of myosin V/XI, Traffic 9(3) (2008) 290-8. [17] N. Zhang, L.L. Yao, X.D. Li, Regulation of class V myosin, Cell Mol Life Sci (2017). [18] O. Pylypenko, T. Welz, J. Tittel, M. Kollmar, F. Chardon, G. Malherbe, S. Weiss, C.I. Michel, A. Samol-Wolf, A.T. Grasskamp, A. Hume, B. Goud, B. Baron, P. England, M.A. Titus, P. Schwille, T. Weidemann, A. Houdusse, E. Kerkhoff, Coordinated recruitment of Spir actin nucleators and myosin V motors to Rab11 vesicle membranes, Elife 5 (2016). [19] Z. Wei, X. Liu, C. Yu, M. Zhang, Structural basis of cargo recognitions for class V myosins, Proc Natl Acad Sci U S A 110(28) (2013) 11314-9. [20] O. Pylypenko, W. Attanda, C. Gauquelin, M. Lahmani, D. Coulibaly, B. Baron, S. Hoos, M.A. Titus, P. England, A.M. Houdusse, Structural basis of myosin V Rab GTPase-dependent cargo recognition, Proc Natl Acad Sci U S A 110(51) (2013) 20443-8. [21] D.T. Jacobs, R. Weigert, K.D. Grode, J.G. Donaldson, R.E. Cheney, Myosin Vc is a molecular motor that functions in secretory granule trafficking, Mol Biol Cell 20(21) (2009) 4471-88. [22] R.R. Marchelletta, D.T. Jacobs, J.E. Schechter, R.E. Cheney, S.F. Hamm-Alvarez, The class V myosin motor, myosin 5c, localizes to mature secretory vesicles and facilitates exocytosis in lacrimal acini, Am J Physiol Cell Physiol 295(1) (2008) C13-28. [23] J.J. Bultema, J.A. Boyle, P.B. Malenke, F.E. Martin, E.C. Dell'Angelica, R.E. Cheney, S.M. Di Pietro, Myosin Vc Interacts with Rab32 and Rab38 Proteins and Works in the Biogenesis and Secretion of Melanosomes*, J Biol Chem 289(48) (2014) 33513-28.

Journal Pre-proof

Jo ur

na

lP

re

-p

ro

of

[24] R.E. Farquhar, E. Rodrigues, K.L. Hamilton, The Role of the Cytoskeleton and Myosin-Vc in the Targeting of KCa3.1 to the Basolateral Membrane of Polarized Epithelial Cells, Front Physiol 7 (2016). [25] J.T. Roland, L.A. Lapierre, J.R. Goldenring, Alternative Splicing in Class V Myosins Determines Association with Rab10*S⃞, J Biol Chem 284(2) (2009) 1213-23. [26] M. Fukuda, T.S. Kuroda, Missense mutations in the globular tail of myosin-Va in dilute mice partially impair binding of Slac2-a/melanophilin, J Cell Sci 117(Pt 4) (2004) 583-91. [27] T. Itoh, M. Satoh, E. Kanno, M. Fukuda, Screening for target Rabs of TBC (Tre2/Bub2/Cdc16) domain-containing proteins based on their Rab-binding activity, Genes Cells 11(9) (2006) 1023-37. [28] T. Itoh, N. Fujita, E. Kanno, A. Yamamoto, T. Yoshimori, M. Fukuda, Golgi-resident Small GTPase Rab33B Interacts with Atg16L and Modulates Autophagosome Formation, Mol Biol Cell 19(7) (2008) 2916-25. [29] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227(5259) (1970) 680-5. [30] L.M. Zanphorlin, T.B. Lima, M.J. Wong, T.S. Balbuena, C.A. Minetti, D.P. Remeta, J.C. Young, L.R. Barbosa, F.C. Gozzo, C.H. Ramos, Heat Shock Protein 90 kDa (Hsp90) Has a Second Functional Interaction Site with the Mitochondrial Import Receptor Tom70, J Biol Chem 291(36) (2016) 18620-31. [31] D.B. Lima, T.B. de Lima, T.S. Balbuena, A.G.C. Neves-Ferreira, V.C. Barbosa, F.C. Gozzo, P.C. Carvalho, SIM-XL: A powerful and user-friendly tool for peptide cross-linking analysis, J Proteomics 129 (2015) 51-55. [32] Y. Song, F. DiMaio, R.Y. Wang, D. Kim, C. Miles, T. Brunette, J. Thompson, D. Baker, Highresolution comparative modeling with RosettaCM, Structure 21(10) (2013) 1735-42. [33] D.E. Kim, D. Chivian, D. Baker, Protein structure prediction and analysis using the Robetta server, Nucleic Acids Res 32(Web Server issue) (2004) W526-31. [34] Schrodinger, LLC, The PyMOL Molecular Graphics System, Version 1.8, 2015. [35] S. Chaudhury, M. Berrondo, B.D. Weitzner, P. Muthu, H. Bergman, J.J. Gray, Benchmarking and analysis of protein docking performance in Rosetta v3.2, PLoS One 6(8) (2011) e22477. [36] A.J.R. Ferrari, F.C. Gozzo, L. Martinez, Statistical force-field for structural modeling using chemical cross-linking/mass spectrometry distance constraints, Bioinformatics 35(17) (2019) 3005-3012. [37] A. Kahraman, F. Herzog, A. Leitner, G. Rosenberger, R. Aebersold, L. Malmstrom, Cross-link guided molecular modeling with ROSETTA, PLoS One 8(9) (2013) e73411. [38] A.J.R. Ferrari, M.A. Clasen, L. Kurt, P.C. Carvalho, F.C. Gozzo, L. Martinez, TopoLink: evaluation of structural models using chemical crosslinking distance constraints, Bioinformatics 35(17) (2019) 3169-3170. [39] P. Conway, M.D. Tyka, F. DiMaio, D.E. Konerding, D. Baker, Relaxation of backbone bond geometry improves protein energy landscape modeling, Protein Sci 23(1) (2014) 47-55. [40] Y. Sun, T.T. Chiu, K.P. Foley, P.J. Bilan, A. Klip, Myosin Va mediates Rab8A-regulated GLUT4 vesicle exocytosis in insulin-stimulated muscle cells, Mol Biol Cell 25(7) (2014) 1159-70. [41] C. Ostermeier, A.T. Brunger, Structural basis of Rab effector specificity: crystal structure of the small G protein Rab3A complexed with the effector domain of rabphilin-3A, Cell 96(3) (1999) 363-74. [42] A.J. Lindsay, F. Jollivet, C.P. Horgan, A.R. Khan, G. Raposo, M.W. McCaffrey, B. Goud, Identification and characterization of multiple novel Rab-myosin Va interactions, Mol Biol Cell 24(21) (2013) 3420-34. [43] T. Wollert, A. Patel, Y.L. Lee, D.W. Provance, Jr., V.E. Vought, M.S. Cosgrove, J.A. Mercer, G.M. Langford, Myosin5a tail associates directly with Rab3A-containing compartments in neurons, J Biol Chem 286(16) (2011) 14352-61.

Journal Pre-proof

of

[44] F. Brozzi, F. Diraison, S. Lajus, S. Rajatileka, T. Philips, R. Regazzi, M. Fukuda, P. Verkade, E. Molnar, A. Varadi, Molecular mechanism of myosin Va recruitment to dense core secretory granules, Traffic 13(1) (2012) 54-69. [45] M. Fukuda, T.S. Kuroda, K. Mikoshiba, Slac2-a/melanophilin, the missing link between Rab27 and myosin Va: implications of a tripartite protein complex for melanosome transport, J Biol Chem 277(14) (2002) 12432-6. [46] T.E. Sladewski, E.B. Krementsova, K.M. Trybus, Myosin Vc Is Specialized for Transport on a Secretory Superhighway, Curr Biol 26(16) (2016) 2202-7. [47] O.D. Bello, A.I. Cappa, M. de Paola, M.N. Zanetti, M. Fukuda, R.A. Fissore, L.S. Mayorga, M.A. Michaut, Rab3A, a possible marker of cortical granules, participates in cortical granule exocytosis in mouse eggs, Exp Cell Res 347(1) (2016) 42-51. [48] M.A. Bustos, O. Lucchesi, M.C. Ruete, L.S. Mayorga, C.N. Tomes, Rab27 and Rab3 sequentially regulate human sperm dense-core granule exocytosis, Proc Natl Acad Sci U S A 109(30) (2012) E2057-66. [49] T. Tsuboi, M. Fukuda, Rab3A and Rab27A cooperatively regulate the docking step of dense-core vesicle exocytosis in PC12 cells, J Cell Sci 119(Pt 11) (2006) 2196-203.

Jo ur

na

lP

re

-p

ro

Graphical abstract

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7