DoMY-Seq: A yeast two-hybrid–based technique for precision mapping of protein–protein interaction motifs

Interactions between proteins are fundamental for every biological process and especially important in cell signaling pathways. Biochemical techniques that evaluate these protein–protein interactions (PPIs), such as vitro pull downs coimmunoprecipitations, have become popular most laboratories essential to identify validate novel protein binding partners. Most PPIs occur through small domains or motifs, which challenging laborious map by using standard biochemical approaches because they generally require the cloning of several truncation mutants. Moreover, classical methodologies provide limited resolution interacting interface. Here, we describe development an alternative technique overcome limitations termed “Protein Domain mapping Yeast 2 Hybrid-Next Generation Sequencing” (DoMY-Seq), leverages both yeast two-hybrid next-generation sequencing techniques. In brief, our approach involves creating a library fragments derived from open reading frame interest enriching reporter system. Next-generation is then subsequently employed read sequence fragment, yielding high-resolution plot We optimized DoMY-Seq taking advantage well-described high-affinity interaction KRAS CRAF, domain on this other protein-interacting pairs, including CRAF-MEK1, RIT1-RGL3, p53-MDM2. Thus, provides unbiased method rapidly involved advancing use technology. defined physical contact two more proteins. mediated either covalent noncovalent bonds can also be transient stable, depending biophysical characteristics interaction. Several classes chemical known participate include, but not to, hydrogen, ionic, hydrophobic, disulfide bonds, Van der Waals forces (1Bonetta L. Protein-protein interactions: interactome under construction.Nature. 2010; 468: 851-854Crossref PubMed Scopus (142) Google Scholar, 2Kiel C. Beltrao P. Serrano Analyzing networks structural information.Annu. Rev. Biochem. 2008; 77: 415-441Crossref (73) Scholar). Such necessary virtually all processes, mediating, example, recognition substrates during signal transduction reactions, engagement receptors ligands, host–pathogen interactions, regulation metabolism, proteolysis, epigenetic code, nucleic acid mechanisms, among many others (3Yu H. Tardivo Tam S. Weiner E. Gebreab F. Fan Svrzikapa N. Hirozane-Kishikawa T. Rietman Yang X. Sahalie J. Salehi-Ashtiani K. Hao Cusick M.E. Hill D.E. et al.Next-generation generate datasets.Nat. Methods. 2011; 8: 478-480Crossref (164) only normal physiology pathophysiology, where loss gain lead diseased phenotype (4Ryan D.P. Matthews J.M. human disease.Curr. Opin. Struct. Biol. 2005; 15: 441-446Crossref (207) For instance, some oncoproteins, KRAS, increased affinity effector able propagate downstream leads proliferation survival (5Simanshu D.K. Nissley D.V. McCormick RAS their regulators disease.Cell. 2017; 170: 17-33Abstract Full Text PDF (515) Elucidating PPI motifs critical characterizing used define unique domains, loss-of-function mutations, assist drug discovery studies. Given increasing small-molecule therapeutics disrupt PPIs, improve speed, quality, highly sought after (6Modell A.E. Blosser S.L. Arora P.S. Systematic targeting protein-protein interactions.Trends Pharmacol. Sci. 2016; 37: 702-713Abstract (82) been established study proteins, commonly being pull-down assays, coimmunoprecipitation, (Y2H) technique, fluorescence-based mass spectrometry-based (7Miura An overview current methods confirm interactions.Protein Pept. Lett. 2018; 25: 728-733Crossref (13) motif identification, randomly truncated forms widely employed. These generated based predicted bioinformatically arbitrarily within protein, giving rise overlapping nonoverlapping tagged down coimmunoprecipitation cultured cells (8Xing Wallmeroth Berendzen K.W. Grefen Techniques analysis vivo.Plant Physiol. 171: 727-758PubMed This has disadvantages, low-throughput expressing fragments, biased when designing low assay, given difficulty assessing shorter than 5 kDa. expression unequal, arbitrary cuts interfere with interphase. Hence, identified, follow-up studies desirable minimal motif. To technical conceptual challenges, developed uses Y2H technology means (NGS). Our procedure rapid, unbiased, reliable platform determine specific interactor. order develop suitable rapid took technology, traditionally allows qualitative determination bait high-content potential preys (9Mehla Caufield J.H. Uetz The system: tool interactions.Cold Spring Harb. Protoc. 2015; 2015: 425-430PubMed 10Brückner A. Polge Lentze Auerbach D. Schlattner U. two-hybrid, powerful systems biology.Int. Mol. 2009; 10: 2763-2788Crossref (260) complemented generating prey contained thousands random (ORF) interactor interest. was achieved shearing subcloning DNA reference plasmid encoded preparation protocols (See Experimental procedures Fig. S1). Next, fragment library, following into auxotrophic binary system, combined opposite mating type containing selected diploid exhibited activation appropriate factors. After selection process, isolated enriched populations compositions were analyzed Illumina NGS protocol. bound reads aligned plasmid. exact location determined at single amino-acid overlap saturating amount (Fig. 1A). generation encoding interest, partners (the 4A splice isoform) prototype PPI. fact, first described screening largely occurs N-terminal (aa 56–131) Ras-binding (RBD) CRAF effector-binding 32–40) GTP-dependent manner (11Vojtek A.B. Hollenberg S.M. Cooper J.A. Mammalian Ras interacts directly serine/threonine kinase Raf.Cell. 1993; 74: 205-214Abstract (1612) 12Scheffler J.E. Waugh D.S. Bekesi Kiefer S.E. LoSardo Neri Prinzo K.M. Tsao K.L. Wegrzynski B. Emerson S.D. Characterization 78-residue c-Raf-1 comprises Ras-GTP.J. Chem. 1994; 269: 22340-22346Abstract facilitate interaction, started mutant G12V, exhibits high levels loaded GTP eukaryotic 13Gibbs J.B. Schaber M.D. Marshall M.S. Scolnick E.M. Sigal I.S. Identification guanine nucleotides ras-encoded growing cells.J. 1987; 262: 10426-10429Abstract Scholar) confirmed G12V mated carrying full-length vector 1B). compatibility libraries template. establish length produced approximately 1000 bp, 700 500 350 200 against 1C). size did impact number positive colonies media, represents degree bait. Only 200-bp yielded lower longer fragments. interactors, recommend testing prior proceeding assay avoid short (<200 bp). 350- 500-bp expected encode 100- 150-aa ?75 aa, decided select conditions further analysis. tested whether orientation assay. swapped (pGAD) (pGBKT7) ORFs pGBKT7 pGAD plasmids, respectively. no differences colony formation detected, picked direct (pGAD-CRAF) reverse (pGBKT7-CRAF) Sanger identity interactors. could align in-frame sequences RBD, suggesting 1D). address specificity toward bait, introduced separate point mutations namely T35A E37G substitutions, render unable bind effectors (14White M.A. Nicolette Minden Polverino Aelst L.V. Karin M. Wigler M.H. Multiple ras functions contribute mammalian transformation.Cell. 1995; 80: 533-541Abstract (617) found mutants produce colonies, confirming 1E). Similar questioned certain performance Because requires interact different oncogenic shown associated (15Der C.J. Finkel G.M. Biological properties rasH genes mutated codon 61.Cell. 1986; 44: 167-176Abstract (392) Scholar), addressed distinct yield higher colonies. assays show WT, G12C, G12D, G13D, Q61L mutants, expressed cells, those S2A), reported previously (13Gibbs cellular trafficking RAF paralogs, nuclear localization result improved yields prenylation sites C180 C186 (16Lin D.T.S. Davis N.G. Conibear Targeting palmitoylation/depalmitoylation cycle cancer.Biochem. Soc. Trans. 45: 913-921Crossref (27) C186S mutation CAAX box, abrogates farnesylation media S2B). line previous experiments encodes G domain, thereby nuclei, takes place (17Khazak V. Golemis E.A. Weber Development screen Ras-Raf inhibitors.Methods 310: 253-271Crossref (8) Despite activation, find sufficient advantages variant. observed background empty control panels. Since DBD very level, likely reflects nonspecific activation. However, consistently (see 1, B E), it does assays. Nevertheless, undertaking experiments, recommended test level control, upon prolonged selective growth, signal-to-noise ratio. performed high-throughput amplifying adaptors present low- high-stringency (LTH [-Leu/-Trp/-His] LTHA [-Leu/-Trp/-His/-Adenine], respectively). Upon analysis, tested, average 300 400 consistent initial 2A). As fragmentation preparation, six frames expected, designated ?3 ?2 ?1 +1 +2 +3 bp. While out do any identifiable peptide before selection, (designated bp) peptides mapped selection. examined distribution similar preselection (x¯ = 16.67 ± 0.8%) (+2 LTH 2B). indicates prey–bait place, activating prototrophic reporter, mainly CRAF. data obtained (“peak calling”) clear enrichment region RBD CRD domain. When considered peak coverage strictly aa 58 128 (70 aa), located 2C). adjacent existence putative additional contacts suggested (18Winkler D.G. Cutler R.E. Drugan J.K. Campbell Morrison residues cysteine-rich Raf-1 activity.J. 1998; 273: 21578-21584Abstract (31) same variant match positions assayed ?500-bp insert cloned reversed S3). site via amounts delivers precise reproducible results. Using experimental approach, leveraged ?350-bp core MEK1, phosphorylated paralogs ARAF, BRAF, (19Lavoie Therrien Regulation kinases ERK signalling.Nat. Cell 16: 281-298Crossref (298) revealed N terminus 3A), included MEK1-binding 310 375 (65 aa). Interestingly, asymmetrical, intradomain 5’ site, while stopped sharply 3’ side. suggest preferential association presence amino precludes MEK1. 700-bp slightly larger remained similar, supporting idea negative effect. BRAF paralog (47% identity) malignancies melanoma, thyroid carcinoma, colorectal cancer, hairy leukemia, well rare disease non–Langerhans histiocytosis (20Karoulia Z. Gavathiotis Poulikakos P.I. New perspectives cancer.Nat. Cancer. 17: 676-691Crossref (161) target, apo form complex molecules, another protomer, 14-3-3 (21Haling J.R. Sudhamsu Yen I. Sideris Sandoval W. Phung Bravo B.J. Giannetti A.M. Peck Masselot Morales Smith Brandhuber Hymowitz S.G. Malek Structure BRAF-MEK reveals activity independent role MAPK signaling.Cancer Cell. 2014; 26: 402-413Abstract (111) 22Thevakumaran Lavoie Critton D.A. Tebben Marinier Sicheri Crystal structure monomer explains basis allosteric regulation.Nat. 22: 37-43Crossref (86) 23Karoulia Wu Y. Ahmed T.A. Xin Q. Bollard Krepler Zhang Bollag G. Herlyn Fagin Lujambio integrated model inhibitor action predicts 30: 485-498Abstract (70) 24Park Rawson Li Kim B.-W. Ficarro S.B. Pino G.G.-D. Sharif Marto Jeon Eck M.J. Architecture autoinhibited active BRAF–MEK1–14-3-3 complexes.Nature. 2019; 575: 545-550Crossref (62) Therefore, CRAF-binding conserved cocrystal BRAF/MEK1 Residues 418 483 3B). R462 K483 identical R354 K375. At P-loop N-lobe 3C). act interface MEK1 (e.g., R462E) A closer look suggests E138 3D). BRAF-MEK1 crystal shows secondary I617, N661, I666, detect corresponding DoMY-Seq. failed capture second asked GTPases domains. purpose, RIT1, GTPase family recently pathogenesis Noonan syndrome, neurodevelopmental disorder (25Aoki Niihori Banjo Okamoto Mizuno Kurosawa Ogata Takada Yano Ando Hoshika Barnett Ohashi Kawame Hasegawa al.Gain-of-function RIT1 cause RAS/MAPK pathway syndrome.Am. Hum. Genet. 2013; 93: 173-180Abstract (194) identified RGL3, Ral exchange factor, (26Shao Andres RalGEF-like candidate rit Ras.J. 2000; 275: 26914-26924Abstract LexA RIT1–RGL3 significant 350-bp RGL3 ability 4A). 614 695 (81 homology-based annotated Ras-association (RA 613–700 RGL3). GTP-loaded (27Shi G.-X. Rit contributes nerve growth factor-induced neuronal differentiation B-Raf-extracellular signal-regulated p38 mitogen-activated cascades.Mol. 830-846Crossref (52) repeated Gal4-based sequence, highlighting consistency systems. Lastly, applied characterize p53 tumor suppressor MDM2 E3 ubiquitin ligase, ubiquitinates proteasomal degradation (28Hafner Bulyk M.L. Jambhekar Lahav multiple mechanisms regulate fate.Nat. 20: 199-210Crossref (257) p53–MDM2 therapeutic variety drugs peptidomimetics destabilize (29Vassilev L.T. Vu B.T. Graves Carvajal Podlaski Filipovic Kong Kammlott Lukacs Klein Fotouhi Liu vivo antagonists MDM2.Science. 2004; 303: 844-848Crossref (3514) 30Li Lozano Molecular pathways: Mdm2 Mdm4 cancer therapy.Clin. Cancer Res. 19: 34-41Crossref (131) readily detectable exploited (31Wong Alfatah Sin M.F. Sim H.M. Verma C.S. Lane Arumugam system characterization inhibitors identifies Mdm2-binding p53.BMC 108Crossref (19) localized relatively 25–109) hydrophobic pocket NH2 15-aa amphipathic crystallography (32Chen Marechal Levine A.J. Mapping mdm-2 domains.Mol. 13: 4107-4114Crossref (609) 33Kussie P.H. Gorina Elenbaas Moreau Pavletich N.P. oncoprotein transactivation domain.Science. 1996; 274: 948-953Crossref (1674) constitutes required, although (34Wallace Worrall Pettersson Hupp T.R. Ball Dual-site E3-ubiquitin ligase activity.Mol. 2006; 23: 251-263Abstract (147) 35Poyurovsky M.V. Katz Laptenko O. Beckerman R. Lokshin Ahn Byeon I.-J.L. Gabizon Mattia Zupnick Brown L.M. Friedler Prives C binds MDM2.Nat. 982-989Crossref (106) MDM2, one ?350 ?500 aa. overlaps MDM2-binding partner orientation. DoMY-Seq, see plateau (between 28 and107, total 80-aa site), corresponds 4B). assume contains determinant before, since maps region. strong middle acidic there pl