A comprehensive analysis of RAS-effector interactions reveals interaction hotspots and new binding partners
2021; Elsevier BV; Volume: 296; Linguagem: Inglês
10.1016/j.jbc.2021.100626
ISSN1083-351X
AutoresSoheila Rezaei Adariani, Neda S. Kazemein Jasemi, Farhad Bazgir, Christoph Wittich, Ehsan Amin, Claus A. M. Seidel, Radovan Dvorský, Mohammad Reza Ahmadian,
Tópico(s)Protein Kinase Regulation and GTPase Signaling
ResumoRAS effectors specifically interact with GTP-bound RAS proteins to link extracellular signals to downstream signaling pathways. These interactions rely on two types of domains, called RAS-binding (RB) and RAS association (RA) domains, which share common structural characteristics. Although the molecular nature of RAS-effector interactions is well-studied for some proteins, most of the RA/RB-domain-containing proteins remain largely uncharacterized. Here, we searched through human proteome databases, extracting 41 RA domains in 39 proteins and 16 RB domains in 14 proteins, each of which can specifically select at least one of the 25 members in the RAS family. We next comprehensively investigated the sequence–structure–function relationship between different representatives of the RAS family, including HRAS, RRAS, RALA, RAP1B, RAP2A, RHEB1, and RIT1, with all members of RA domain family proteins (RASSFs) and the RB-domain-containing CRAF. The binding affinity for RAS-effector interactions, determined using fluorescence polarization, broadly ranged between high (0.3 μM) and very low (500 μM) affinities, raising interesting questions about the consequence of these variable binding affinities in the regulation of signaling events. Sequence and structural alignments pointed to two interaction hotspots in the RA/RB domains, consisting of an average of 19 RAS-binding residues. Moreover, we found novel interactions between RRAS1, RIT1, and RALA and RASSF7, RASSF9, and RASSF1, respectively, which were systematically explored in sequence–structure–property relationship analysis, and validated by mutational analysis. These data provide a set of distinct functional properties and putative biological roles that should now be investigated in the cellular context. RAS effectors specifically interact with GTP-bound RAS proteins to link extracellular signals to downstream signaling pathways. These interactions rely on two types of domains, called RAS-binding (RB) and RAS association (RA) domains, which share common structural characteristics. Although the molecular nature of RAS-effector interactions is well-studied for some proteins, most of the RA/RB-domain-containing proteins remain largely uncharacterized. Here, we searched through human proteome databases, extracting 41 RA domains in 39 proteins and 16 RB domains in 14 proteins, each of which can specifically select at least one of the 25 members in the RAS family. We next comprehensively investigated the sequence–structure–function relationship between different representatives of the RAS family, including HRAS, RRAS, RALA, RAP1B, RAP2A, RHEB1, and RIT1, with all members of RA domain family proteins (RASSFs) and the RB-domain-containing CRAF. The binding affinity for RAS-effector interactions, determined using fluorescence polarization, broadly ranged between high (0.3 μM) and very low (500 μM) affinities, raising interesting questions about the consequence of these variable binding affinities in the regulation of signaling events. Sequence and structural alignments pointed to two interaction hotspots in the RA/RB domains, consisting of an average of 19 RAS-binding residues. Moreover, we found novel interactions between RRAS1, RIT1, and RALA and RASSF7, RASSF9, and RASSF1, respectively, which were systematically explored in sequence–structure–property relationship analysis, and validated by mutational analysis. These data provide a set of distinct functional properties and putative biological roles that should now be investigated in the cellular context. RAS family proteins control activities of multiple signaling pathways and consequently a wide array of cellular processes, including survival, growth, adhesion, migration, and differentiation (1Jaiswal M. Dvorsky R. Amin E. Risse S.L. Fansa E.K. Zhang S.C. Taha M.S. Gauhar A.R. Nakhaei-Rad S. Kordes C. Koessmeier K.T. Cirstea I.C. Olayioye M.A. Haussinger D. Ahmadian M.R. Functional cross-talk between ras and rho pathways: A Ras-specific GTPase-activating protein (p120RasGAP) competitively inhibits the RhoGAP activity of deleted in liver cancer (DLC) tumor suppressor by masking the catalytic arginine finger.J. Biol. Chem. 2014; 289: 6839-6849Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar). Any dysregulation of these pathways leads, thus, to cancer, developmental disorders, metabolic and cardiovascular diseases (2Simanshu D.K. Nissley D.V. McCormick F. RAS proteins and their regulators in human disease.Cell. 2017; 170: 17-33Abstract Full Text Full Text PDF PubMed Scopus (463) Google Scholar). Signal transduction implies a physical association of RAS proteins with a spectrum of functionally diverse downstream effectors, e.g., CRAF, PI3Kα, TIAM1, RALGDS, PLCε, and RASSF5 (3Gutierrez-Erlandsson S. Herrero-Vidal P. Fernandez-Alfara M. Hernandez-Garcia S. Gonzalo-Flores S. Mudarra-Rubio A. Fresno M. Cubelos B. R-RAS2 overexpression in tumors of the human central nervous system.Mol. Cancer. 2013; 12: 127Crossref PubMed Scopus (14) Google Scholar, 4Karnoub A.E. Weinberg R.A. Ras oncogenes: Split personalities.Nat. Rev. Mol. Cell Biol. 2008; 9: 517-531Crossref PubMed Scopus (1010) Google Scholar, 5Herrmann C. Ras–effector interactions: After one decade.Curr. Opin. Struct. Biol. 2003; 13: 122-129Crossref PubMed Scopus (0) Google Scholar, 6Nakhaei-Rad S. Nakhaeizadeh H. Götze S. Kordes C. Sawitza I. Hoffmann M.J. Franke M. Schulz W.A. Scheller J. Piekorz R.P. The role of embryonic stem cell-expressed RAS (ERAS) in the maintenance of quiescent hepatic stellate cells.J. Biol. Chem. 2016; 291: 8399-8413Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar, 7Castellano E. Downward J. Role of RAS in the regulation of PI 3-kinase.Curr Top Microbiol Immunol. 2010; 346: 143-169Crossref PubMed Google Scholar, 8Chan J.J. Flatters D. Rodrigues-Lima F. Yan J. Thalassinos K. Katan M. Comparative analysis of interactions of RASSF1-10.Adv. Biol. Regul. 2013; 53: 190-201Crossref PubMed Scopus (0) Google Scholar, 9Bunney T.D. Katan M. PLC regulation: Emerging pictures for molecular mechanisms.Trends Biochem. Sci. 2011; 36: 88-96Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar, 10Ferro E. Trabalzini L. RalGDS family members couple Ras to Ral signalling and that's not all.Cell Signal. 2010; 22: 1804-1810Crossref PubMed Scopus (53) Google Scholar, 11Rajalingam K. Schreck R. Rapp U.R. Albert Š. Ras oncogenes and their downstream targets.Biochim. Biophys. Acta. 2007; 1773: 1177-1195Crossref PubMed Scopus (306) Google Scholar). RAS-effector interaction essentially requires RAS association with membranes (12Nussinov R. Tsai C.-J. Muratcioglu S. Jang H. Gursoy A. Keskin O. Principles of K-Ras effector organization and the role of oncogenic K-Ras in cancer initiation through G1 cell cycle deregulation.Expert Rev. Proteomics. 2015; 12: 669-682Crossref PubMed Scopus (31) Google Scholar), and its activation by specific regulatory proteins (e.g., guanine nucleotide exchange factors or GEFs), leading to the formation of GTP-bound, active RAS (13Ahearn I.M. Haigis K. Bar-Sagi D. Philips M.R. Regulating the regulator: Post-translational modification of RAS.Nat. Rev. Mol. Cell Biol. 2012; 13: 39Crossref Scopus (351) Google Scholar, 14Hennig A. Markwart R. Esparza-Franco M.A. Ladds G. Rubio I. Ras activation revisited: Role of GEF and GAP systems.Biol. Chem. 2015; 396: 831-848Crossref PubMed Scopus (46) Google Scholar, 15Fischer A. Hekman M. Kuhlmann J. Rubio I. Wiese S. Rapp U.R. B-and C-RAF display essential differences in their binding to Ras the isotype-specific N terminus of B-RAF facilitates Ras binding.J. Biol. Chem. 2007; 282: 26503-26516Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar). Notably, RAS proteins change their conformation mainly at two mobile regions, designated as a switch I (residues 30–40) and switch II (residues 60–68) (16Mott H.R. Owen D. Structures of Ras superfamily effector complexes: What have we learnt in two decades?.Crit. Rev. Biochem. Mol. Biol. 2015; 50: 85-133Crossref PubMed Scopus (52) Google Scholar, 17Vetter I.R. Wittinghofer A. The guanine nucleotide-binding switch in three dimensions.Science. 2001; 294: 1299-1304Crossref PubMed Scopus (1254) Google Scholar, 18Filchtinski D. Sharabi O. Rüppel A. Vetter I.R. Herrmann C. Shifman J.M. What makes Ras an efficient molecular switch: A computational, biophysical, and structural study of Ras-GDP interactions with mutants of Raf.J. Mol. Biol. 2010; 399: 422-435Crossref PubMed Scopus (45) Google Scholar). Only in GTP-bound form, the switch regions of the RAS proteins provide a platform for the association of the effector proteins (19Erijman A. Shifman J.M. RAS/effector interactions from structural and biophysical perspective.Mini Rev. Med. Chem. 2016; 16: 370-375Crossref PubMed Google Scholar, 20Nassar N. Horn G. Herrmann C. Block C. Janknecht R. Wittinghofer A. Ras/Rap effector specificity determined by charge reversal.Nat. Struct. Biol. 1996; 3: 723-729Crossref PubMed Scopus (173) Google Scholar). To date, two types of domains, the RAS-binding (RB) and RAS association (RA) domains, have been defined for various effectors. They are comprised of 80 to 100 amino acids and have a similar ubiquitin-like topology (8Chan J.J. Flatters D. Rodrigues-Lima F. Yan J. Thalassinos K. Katan M. Comparative analysis of interactions of RASSF1-10.Adv. Biol. Regul. 2013; 53: 190-201Crossref PubMed Scopus (0) Google Scholar, 21Nakhaeizadeh H. Amin E. Nakhaei-Rad S. Dvorsky R. Ahmadian M.R. The RAS-effector interface: Isoform-specific differences in the effector binding regions.PLoS One. 2016; 11e0167145Crossref PubMed Scopus (26) Google Scholar, 22Repasky G.A. Chenette E.J. Der C.J. Renewing the conspiracy theory debate: Does Raf function alone to mediate Ras oncogenesis?.Trends Cell Biol. 2004; 14: 639-647Abstract Full Text Full Text PDF PubMed Scopus (233) Google Scholar, 23Wohlgemuth S. Kiel C. Krämer A. Serrano L. Wittinghofer F. Herrmann C. Recognizing and defining true Ras binding domains I: Biochemical analysis.J. Mol. Biol. 2005; 348: 741-758Crossref PubMed Scopus (132) Google Scholar, 24Dhanaraman T. Singh S. Killoran R.C. Singh A. Xu X. Shifman J.M. Smith M.J. RASSF effectors couple diverse RAS subfamily GTPases to the Hippo pathway.Sci. Signal. 2020; 13eabb477Crossref Scopus (2) Google Scholar). Considering different RAS effectors, RB and RA domain interactions with RAS proteins do not exhibit the same mode of interaction between different RAS effectors. However, CRAF RB and RALGDS RA domains share a similar structure and contact the switch I region via a similar binding mode (25Goddard T.D. Huang C.C. Meng E.C. Pettersen E.F. Couch G.S. Morris J.H. Ferrin T.E. UCSF ChimeraX: Meeting modern challenges in visualization and analysis.Protein Sci. 2018; 27: 14-25Crossref PubMed Scopus (784) Google Scholar, 26Nassar N. Horn G. Herrmann C.A. Scherer A. McCormick F. Wittinghofer A. The 2.2 Å crystal structure of the Ras-binding domain of the serine/threonine kinase c-Raf1 in complex with RaplA and a GTP analogue.Nature. 1995; 375: 554Crossref PubMed Google Scholar). In contrast, PI3Kα RB, RASSF5 RA, and PLCε RA domains do not share sequence and structural similarity but commonly associate with the switch regions, particularly switch I (27Bunney T.D. Harris R. Gandarillas N.L. Josephs M.B. Roe S.M. Sorli S.C. Paterson H.F. Rodrigues-Lima F. Esposito D. Ponting C.P. Structural and mechanistic insights into ras association domains of phospholipase C epsilon.Mol. Cell. 2006; 21: 495-507Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar, 28Stieglitz B. Bee C. Schwarz D. Yildiz Ö. Moshnikova A. Khokhlatchev A. Herrmann C. Novel type of Ras effector interaction established between tumour suppressor NORE1A and Ras switch II..EMBO J. 2008; 27: 1995-2005Crossref PubMed Scopus (0) Google Scholar, 29Pacold M.E. Suire S. Perisic O. Lara-Gonzalez S. Davis C.T. Walker E.H. Hawkins P.T. Stephens L. Eccleston J.F. Williams R.L. Crystal structure and functional analysis of Ras binding to its effector phosphoinositide 3-kinase γ.Cell. 2000; 103: 931-944Abstract Full Text Full Text PDF PubMed Google Scholar). RAS-effector interaction strikingly shares a similar binding mode adopted by three components: two antiparallel β-sheets of the RA/RB domains and the RAS switch I region, respectively, and the first α-helix of the RA/RB domains (30Smith M.J. Ottoni E. Ishiyama N. Goudreault M. Haman A. Meyer C. Tucholska M. Gasmi-Seabrook G. Menezes S. Laister R.C. Evolution of AF6-RAS association and its implications in mixed-lineage leukemia.Nat. Commun. 2017; 8: 1-13Crossref PubMed Scopus (0) Google Scholar). In this study, we conducted an in-depth database search in the human proteome and extracted 57 RA/RB domains. We used ten RASSF RA domains to analyze their interactions with seven representatives of the RAS proteins family, including HRAS, RRAS1, RAP1B, RAP2A, RALA, RIT1, and RHEB1. CRAF RB domain was used as control. The binding analysis was performed under the same conditions using fluorescence polarization. Obtained dissociation constants (Kd) with a broad range (0.3–500 μM) along with a matrix for a potential interaction of 25 RAS proteins and 57 RA/RB domains provide us a detailed view of the sequence–structure–property relationships of RAS-effector binding capabilities. Mining in the UniProt database led to the extraction of 130 RB and 145 RA-domain-containing proteins, respectively. In a parallel search using HMMER, 127 RB and 164 RA-domain-containing proteins were extracted. These numbers were reduced to 46 RB and 97 RA-domain-containing proteins by excluding proteins containing RHO-binding domains, mitochondrial proton/calcium antiporter domain, and receptors. In the last step, all isoforms with identical sequences of the RB and RA domains were excluded using multiple sequence alignments generated with the ClustalW algorithm. This approach identified a total number of 16 RB domains in 14 RB-domain-containing proteins and 41 RA domains in 39 RA-domain-containing proteins, (Fig. S1; Tables S1 and S2). Both types of RAS effector domains share sequence identity of 10.5% and 9.2% and sequence similarity of 25.5% and 20.2% (Figs. S2 and S3). The direct interaction of different RA-domain-containing proteins with RAS proteins has been comprehensively analyzed (23Wohlgemuth S. Kiel C. Krämer A. Serrano L. Wittinghofer F. Herrmann C. Recognizing and defining true Ras binding domains I: Biochemical analysis.J. Mol. Biol. 2005; 348: 741-758Crossref PubMed Scopus (132) Google Scholar, 31Kiel C. Wohlgemuth S. Rousseau F. Schymkowitz J. Ferkinghoff-Borg J. Wittinghofer F. Serrano L. Recognizing and defining true Ras binding domains II: In silico prediction based on homology modelling and energy calculations.J. Mol. Biol. 2005; 348: 759-775Crossref PubMed Scopus (89) Google Scholar). However, the majority of proteins with a RA domain remain uncharacterized (Table S1). The RAS association domain family (RASSF), which controls a broad range of signaling pathways (8Chan J.J. Flatters D. Rodrigues-Lima F. Yan J. Thalassinos K. Katan M. Comparative analysis of interactions of RASSF1-10.Adv. Biol. Regul. 2013; 53: 190-201Crossref PubMed Scopus (0) Google Scholar, 32Donninger H. Schmidt M.L. Mezzanotte J. Barnoud T. Clark G.J. Ras signaling through RASSF proteins.Semin. Cell Dev. Biol. 2016; 58: 86-95Crossref PubMed Scopus (55) Google Scholar), is the largest RA-domain-containing protein family (Fig. 1). Their RA domains differently interact with classical RAS proteins (8Chan J.J. Flatters D. Rodrigues-Lima F. Yan J. Thalassinos K. Katan M. Comparative analysis of interactions of RASSF1-10.Adv. Biol. Regul. 2013; 53: 190-201Crossref PubMed Scopus (0) Google Scholar, 24Dhanaraman T. Singh S. Killoran R.C. Singh A. Xu X. Shifman J.M. Smith M.J. RASSF effectors couple diverse RAS subfamily GTPases to the Hippo pathway.Sci. Signal. 2020; 13eabb477Crossref Scopus (2) Google Scholar). Among them, only the interaction of RASSF1 and RASSF5/NORE1 RA domains has been characterized quantitatively so far (23Wohlgemuth S. Kiel C. Krämer A. Serrano L. Wittinghofer F. Herrmann C. Recognizing and defining true Ras binding domains I: Biochemical analysis.J. Mol. Biol. 2005; 348: 741-758Crossref PubMed Scopus (132) Google Scholar, 31Kiel C. Wohlgemuth S. Rousseau F. Schymkowitz J. Ferkinghoff-Borg J. Wittinghofer F. Serrano L. Recognizing and defining true Ras binding domains II: In silico prediction based on homology modelling and energy calculations.J. Mol. Biol. 2005; 348: 759-775Crossref PubMed Scopus (89) Google Scholar). Other characterized RA-domain-containing proteins, including RALGDS-like proteins, PLCε, AF6, RIN1/2, and PDZGEF1/2, regulate diverse cellular processes. They share high structural similarity and exhibit differential selectivity for HRAS and RAP1B (23Wohlgemuth S. Kiel C. Krämer A. Serrano L. Wittinghofer F. Herrmann C. Recognizing and defining true Ras binding domains I: Biochemical analysis.J. Mol. Biol. 2005; 348: 741-758Crossref PubMed Scopus (132) Google Scholar, 31Kiel C. Wohlgemuth S. Rousseau F. Schymkowitz J. Ferkinghoff-Borg J. Wittinghofer F. Serrano L. Recognizing and defining true Ras binding domains II: In silico prediction based on homology modelling and energy calculations.J. Mol. Biol. 2005; 348: 759-775Crossref PubMed Scopus (89) Google Scholar). RB-domain-containing proteins are mostly kinases (Table S2). The serine/threonine RAF kinase family proteins (A/B/CRAF; (33Rezaei Adariani S. Buchholzer M. Akbarzadeh M. Nakhaei-Rad S. Dvorsky R. Ahmadian M.R. Structural snapshots of RAF kinase interactions.Biochem. Soc. Trans. 2018; 46: 1393-1406Crossref PubMed Scopus (12) Google Scholar)) activate the MEK-ERK axis and control cell proliferation and differentiation (34Haghighi F. Dahlmann J. Nakhaei-Rad S. Lang A. Kutschka I. Zenker M. Kensah G. Piekorz R.P. Ahmadian M.R. bFGF-mediated pluripotency maintenance in human induced pluripotent stem cells is associated with NRAS-MAPK signaling.Cell Commun. Signal. 2018; 16: 96Crossref PubMed Scopus (0) Google Scholar, 35Desideri E. Cavallo A.L. Baccarini M. Alike but different: RAF paralogs and their signaling outputs.Cell. 2015; 161: 967-970Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar). PI3Kα generates phosphatidylinositol (3,4,5)-trisphosphate (PIP3) and regulates cell growth, cell survival, cytoskeleton reorganization, and metabolism (36Castellano E. Downward J. RAS interaction with PI3K: More than just another effector pathway.Genes Cancer. 2011; 2: 261-274Crossref PubMed Scopus (419) Google Scholar). RGS12/14, which usually act as inactivators of Gα proteins (37Ross E.M. Wilkie T.M. GTPase-activating proteins for heterotrimeric G proteins: regulators of G protein signaling (RGS) and RGS-like proteins.Annu. Rev. Biochem. 2000; 69: 795-827Crossref PubMed Scopus (886) Google Scholar), physically interact with various members of the RAS family. They appear to facilitate the assembly of the components of the MAPK pathway through direct association with activated HRAS (38Willard F.S. Willard M.D. Kimple A.J. Soundararajan M. Oestreich E.A. Li X. Sowa N.A. Kimple R.J. Doyle D.A. Der C.J. Regulator of G-protein signaling 14 (RGS14) is a selective H-Ras effector.PLoS One. 2009; 4e4884Crossref PubMed Scopus (32) Google Scholar). TIAM1/2, which act as specific GEFs for the RHO family proteins and control cell migration (39Malliri A. Collard J.G. Role of Rho-family proteins in cell adhesion and cancer.Curr. Opin. Cell Biol. 2003; 15: 583-589Crossref PubMed Scopus (139) Google Scholar, 40Rooney C. White G. Nazgiewicz A. Woodcock S.A. Anderson K.I. Ballestrem C. Malliri A. The Rac activator STEF (Tiam2) regulates cell migration by microtubule-mediated focal adhesion disassembly.EMBO Rep. 2010; 11: 292-298Crossref PubMed Scopus (61) Google Scholar), have been suggested to recognize activated RAS proteins (41Yamauchi J. Miyamoto Y. Tanoue A. Shooter E.M. Chan J.R. Ras activation of a Rac1 exchange factor, Tiam1, mediates neurotrophin-3-induced Schwann cell migration.Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 14889-14894Crossref PubMed Scopus (89) Google Scholar). However, their direct interaction with RAS proteins has not been shown to date (23Wohlgemuth S. Kiel C. Krämer A. Serrano L. Wittinghofer F. Herrmann C. Recognizing and defining true Ras binding domains I: Biochemical analysis.J. Mol. Biol. 2005; 348: 741-758Crossref PubMed Scopus (132) Google Scholar). Moreover, a few proteins, reported as RAS effectors, do not apparently contain an RA/RB domain (Table S3). To determine the binding capability between the effector domains and diverse proteins of the RAS family, the following proteins were selected for this study: (i) all ten RASSF family proteins as representative RA-domain-containing effector proteins; (ii) CRAF RB domain (Fig. 1) was used as a representative of the RB-domain-containing proteins; and (iii) the RAS family includes 23 genes coding for at least 25 proteins, which share, considering their G domains, sequence identity of 48.6% and sequence similarity of 61.5% (Fig. S4). Based on sequence identity, structure, and function of their G domains, the RAS proteins were divided into eight paralog groups (Table S4): RAS, RRAS, RAP, RAL, RIT, RHEB, RASD, and DIRAS (42Nakhaei-Rad S. Haghighi F. Nouri P. Rezaei Adariani S. Lissy J. Kazemein Jasemi N.S. Dvorsky R. Ahmadian M.R. Structural fingerprints, interactions, and signaling networks of RAS family proteins beyond RAS isoforms.Crit. Rev. Biochem. Mol. Biol. 2018; 53: 130-156Crossref PubMed Scopus (16) Google Scholar). RAS-related proteins RASLs, RERG, RERGL, NKIRAS1/2 were excluded from this list and study due to their major sequence deviations. To monitor binding we applied a fluorescence polarization assay (21Nakhaeizadeh H. Amin E. Nakhaei-Rad S. Dvorsky R. Ahmadian M.R. The RAS-effector interface: Isoform-specific differences in the effector binding regions.PLoS One. 2016; 11e0167145Crossref PubMed Scopus (26) Google Scholar) to determine the dissociation constants (Kd) for the RAS-effector interactions. For this, we prepared HRAS, RRAS, RAP1B, RAP2A, RALA, RIT1, and RHEB1 in complex with a nonhydrolyzable, fluorescent analog of GTP, called mGppNHp. Representatives of RASD and DIRAS groups were not applied due to their physical instability in vitro. Small-sized RB and RA domains were fused to maltose-binding protein (MBP, 42 kDa) to increase their overall molecular weight and to ensure a homogeneous monomeric form of the fusion proteins. Figure 1 shows SDS gels for all purified proteins used in this study. Increasing concentrations of MBP-fused effector proteins were titrated to RAS•mGppNHp proteins to assess the binding capability of the respective interaction pairs. We observed a significant change in fluorescence polarization for the majority of the measurements (Figs. S5 and S6A). However, evaluated Kd values ranged from 0.3 to more than 500 μM. These data are summarized in Table S5 and illustrated in Figure 2. Under these experimental conditions, the CRAF RB domain revealed the highest affinity for HRAS and RRAS1 while the RASSF5 RA domain exhibited a relatively high affinity for HRAS, RAP1B, and RAP2A (Fig. 2, A and B, green bars). The intermediate affinities were obtained for the interaction of the CRAF RB domain with RAP1B as well as RASSF1 with RAP1B, RAP2A and RALA, RASSF9 with RIT1 and RASSF7 with RRAS1 (Fig. 2, A and B; blue bars). The majority of the interaction pairs showed, however, low and very low affinities (Fig. 2B, red and black bars, respectively). Among them, RHEB notably revealed the majority of low-affinity interactions. No binding was observed for 12 pairwise interactions. Purified MBP, which was titrated to HRAS•mGppNHp as a negative control, exhibited no interaction (Fig. S7A). The reproducibility of the fluorescence polarization measurements was assessed by determining the Kd value for the interaction between HRAS•mGppNHp with RASSF1-RA in three different experiments. To understand the atomic interactions between RAS and effector proteins and explain observed variable affinities, we analyzed various structures of RAS-effector protein complexes. To date, 13 structures of RAS-effector protein complexes exist in the PDB (Table S6). As some of them contain more than one complex in the unit cell, there were altogether 19 complexes available for the analysis. In order to map atomic interactions responsible for observed variable affinities, we have extracted information about interacting interface from all of the abovementioned complex structures (Figs. S8 and S9) and combined them with their sequence alignments (Figs. S2–S4). It is important to note that some amino acids, aligned according to the sequence, were quite distant in the space. Therefore, we edited the sequence alignment to synchronize it with the structural alignment. Our python code finally took sequence alignments with PDB files of complex structures as inputs and calculated all interaction pairs in analyzed complex structures in the form of an interaction matrix. The resultant matrix comprehensively relates the interacting residues on both sides of the complexes, with RAS paralogs as rows and the RA/RB domains as columns (Fig. 3). All numbering in this study is based on HRAS on the one side and CRAF and RASSF5, for RB and RA domains respectively, on the other side. Each element of the matrix that can be accounted for a "hotspot" relates one homologous residue from RAS proteins to one homologous residue from the RA/RB domains. The number value of this element, ranging from 0 to 19, represents the number of complex structures in which these residues interact (Fig. 3). Thus, 0 means that these two residues do not contact each other in any structure while a maximal value 19 means that this particular interaction exists in all analyzed complex structures of the RAS-RA/RB domains. We have sorted the residues at both sides of the matrix according to their conservation versus variability. As can be seen in Figs. S4 and S9, the majority of the residues (14 out of 20) on the side of 25 RAS proteins are conserved, nine of which (Q/N25, D/E33, I/V36, E37, D38, S/T39, Y40, R/K41 in switch I, and Y64 in the switch II; HRAS numbering) account for major hotspots (Fig. 3). On the other side, and in contrast, the majority of 19 RAS interacting residues in RA/RB domains are variable and only two distant residues are conserved (R/K59 and K/R84; CRAF numbering; R/K241 and K/R308; RASSF5 numbering) (Fig. 3 and Fig. S9). However, what is striking is the middle cluster of the matrix with the most frequent interactions between the conserved residues in the switch I region of the RAS proteins (β2-strand residues 36–41; HRAS numbering) and the variable residues of the RA/RB domains (β2-strand residues 64–71; CRAF numbering; residues 284–291; RASSF5 numbering) (Fig. 3 and Fig. S9). This cluster adopts an arrangement of intermolecular β-sheet interactions in an antiparallel fashion (Fig. S8). A substantial number of the contacts in this cluster are mediated by main-chain/main-chain interactions, which typically involve hydrogen bonds between the N-H group and the carbonyl oxygen of the amino acids 37 to 39 from the RAS side and positions 66 to 69 (CRAF numbering) and 286 to 289 (RASSF5 numbering) from the side of the RA/RB domains. To prove the impact of the hotspot residues on the selectivity of the RASSF RA domain interactions with RAS family proteins, we selected the weak and strong RAS-RASSF interactions, and substituted 4 to 5 amino acids in the hotspot region (Fig. 3, boxed residues) RASSF2 to RASSF1 as well as RASSF4 and RASSF9 to RASSF5. The variants, RASSF2-to-1, RASSF4-to-5, and RASSF9-to-5 (Fig. 1), were successfully expressed and purified. Their binding affinities for HRAS, RIT1, RALA, RAP2A and RRAS1 were measured using fluorescence polarization (Fig. S10). Remarkable differences in binding affinities of the analyzed RASSF variants are summarized in Figure 4 for comparison. RASSF2-to-1 variant revealed a significant increase of RALA (p < 0.006) and RRAS1 (p < 0.014) binding affinity compared with RASSF2 but declined compared with RASSF1. In contrast, RIT1, which did not show any binding to RASSF2 and a very low affinity to RASSF1, now exhibited a reasonable Kd value of 65 μM for RASSF2-to-1. The RASSF4-to-5 variant, on the one hand, showed a tremendous increase in affinity for HRAS of about 20-fold (p < 0.0118) and, on the other hand, diminished RIT1 property to bind RASSF4 by threefold (p < 0.0351). These data suggest that the hotspot residues favor RASSF4 binding to RIT1, whereas those residues of RASSF5 counteract RIT1 binding. Similarly, the RASSF4-to-5 affinity for RAP2A was increased by 2.5-fold (n.s., p < 0.087), which emphasizes the high-affinity RAP2A-RASSF5 interaction. The RASSF9-to-5 variant showed a 4.5-fold increase in HRAS-binding affinity as compared with RASSF9 (p < 0.008) that can be attributed to the high-affinity interaction of HRAS with RASSF5. The intermediate affinity of RASSF9 for RIT1 of 27 μM is validated by the RASSF9-to-5 variant, which revealed a 5.5-fold higher Kd value (p < 0.005). The interaction of the RASSF9-to-5 variant with RALA was drastically enhanced (Kd = 35 μM) considering the lack of RASSF9 binding to RALA. Our data on residue swapping in RASSF proteins successfully validated the key role of hotspot residues in the RAS-RASSF interaction, particularly RASSF1-RALA, RASSF5-HRAS, and RASSF9-RIT1. To prove physiological relevance of identified RIT1 interactions with RASSF7 and RASSF9, we transfect Human Embryonic Kidney (HEK) 293T cells with human RIT1 and used His-tagged RA domains of RASSF7 and RASSF9 to pull down HA-tagged RIT1 from the cell lysates. As a control, we used lysates of HRAS-transfected cells and His-ta
Referência(s)