Discovery of a σ1 receptor antagonist by combination of unbiased cell painting and thermal proteome profiling

•Morphological profiling can identify bioactive compounds and suggest putative targets•Thermal proteome of small molecules usually identifies multiple targets•Targets be prioritized by combining morphological thermal profiling•Thereby, a σ1 receptor antagonist was discovered Phenotypic screening for is typically combined with affinity-based chemical proteomics to uncover the respective molecular targets. However, such assays explored bioactivity are biased toward monitored phenotype, target identification often requires derivatization hit compound. In contrast, unbiased cellular approaches record hundreds parameters upon compound perturbation map in broader biological context may link profile or mode action. Herein we report discovery diaminopyrimidine DP68 as Sigma 1 (σ1) using Cell Painting assay profiling. Our results highlight that integration complementary enable both detection molecules. The novel at heart biology research. this endeavor, typically, libraries screened bioactivity, cell-based assays. Such chemically diverse contain plethora biologically relevant compounds, potentially covering vast range bioactivities. target-based phenotypic only focus on certain protein process, respectively (Eggert, 2013Eggert U.S. why how small-molecule screens.Nat. Chem. Biol. 2013; 9: 206-209Crossref PubMed Scopus (101) Google Scholar). Thus, detect fraction within library. (e.g., probing gene expression morphology) sense. A prominent example (CPA), which six dyes multiplexed visualize changes induced High-content imaging automated image analysis employed alterations generate fingerprint each Thereby, CPA able broad bioactivities provide valuable information about targets actions comparison fingerprints determined annotated reference (Bray et al., 2016Bray M.-A. Singh S. Han H. Davis C.T. Borgeson B. Hartland C. Kost-Alimova M. Gustafsdottir S.M. Gibson C.C. Carpenter A.E. Painting, high-content image-based fluorescent dyes.Nat. Protoc. 2016; 11: 1757-1774Crossref (247) Scholar; Christoforow 2019Christoforow A. Wilke J. Binici Pahl Ostermann Sievers Waldmann Design, synthesis, pyrano-furo-pyridone pseudo natural products.Angew. Int. Ed. 2019; 58: 14715-14723Crossref (35) Sievers, 2019Pahl tool new matter.in: Ziegler Systems Chemical Biology. Methods Protocols. Springer New York, 2019: 115-126Crossref (14) Zimmermann 2019Zimmermann Akbarzadeh Otte F. Strohmann Sankar M.G. Kumar K. scaffold-diversity synthesis intriguing cyclic sulfonamides.Chemistry. 25: 15498-15503Crossref Foley 2020Foley D.J. Zinken Corkery D. Laraia L. Wu Y.-W. Phenotyping reveals pseudo-natural-product autophagy inhibitor.Angew. 2020; 59: 12470-12476Crossref (19) 2020Laraia Garivet G. Kaiser N. Müller Pinkert T. al.Image-based lysosomotropic, iron-sequestering 5721-5729Crossref (25) Schneidewind 2020Schneidewind Brause Burhop Mejuch Morphological common action different targets.Chembiochem. 21: 3197-3207Crossref (11) For an in-depth characterization screening, mediates needs identified validated. To end, broadly applied method knowledge underlying structure–activity relationship (SAR) affinity probes (Ziegler 2013Ziegler Pries V. Hedberg Target molecules: finding needle haystack.Angew. 52: 2744-2792Crossref (334) Since SAR studies fail site tolerates modification, label-free techniques, like (TPP), powerful alternatives (Mateus 2017Mateus Määttä T.A. Savitski M.M. Thermal profiling: assessment state through heat-induced stability changes.Proteome Sci. 2017; 15: 13Crossref (63) Reckzeh 2019Reckzeh E.S. Brockmeyer Metz Janning P. engagement levels.in: 73-98Crossref (7) TPP, proteins investigated presence absence molecule. As binding molecule force into more (or less) thermally stable conformation, change indicates protein–small interaction. Herein, describe combination TPP. We validate means radioligand displacement demonstrate via shift (CETSA). antagonizes mouse model mechanical allodynia. TPP its target. profiled approximately 10,000 in-house U-2 OS cells. Additionally, set 3,900 (reference compounds) tested. employs stain compartments components acquired images were subsequently subjected pipeline, extracted 579 quantified parameter obtained DMSO-treated cells (control). parameter, Z score calculated difference between measured value control median divided absolute deviation (MAD) controls. ≥ ±3 considered significant parameter. scores all summarized fingerprint. measure induction values calculating percentage significantly changed (Christoforow 34.7% concentration μM (Figure 1A 1B). two fingerprints, biosimilarity correlation distance calculation. 3 10 display high similarity stepwise increase induction. induces concentration-dependent manner. Interestingly, displayed (>85%) 265 1C). majority these G protein-coupled receptors (GPCRs), especially dopamine serotonin receptors. While widely expressed central nervous system, most them not (Table S1). partners narrow down potential targets, HeLa cell lysates 1). Fifteen reproducible melting temperature (ΔTm) thus regarded partners.Table 1List altered behavior DP68Protein nameGeneΔTm [°C]Sigma non-opioid intracellular 1SIGMAR1−4.60 ± 0.46Genetic suppressor element 1GSE1−3.92 1.74Bis(5-adenosyl)-triphosphataseFHIT−3.39 0.73O-phosphoseryl-tRNA(Sec) selenium transferaseSEPSECS−3.28 0.87ADP-ribosylation factor-like 6-interacting 4ARL6IP4−3.12 1.46Putative hydroxypyruvate isomeraseHYI−2.99 0.89Histone acetyltransferase p300EP300−2.96 1.20Insulin-like growth factor receptorIGF1R−2.36 0.64Sushi domain-containing 5SUSD5−2.33 0.68SRE-binding cleavage-activating proteinSCAP−2.12 0.40Gamma-interferon-inducible lysosomal thiol reductaseIFI302.18 0.64MORF4 family-associated 1MRFAP12.37 0.45Cysteine-rich EGF-like domain 1CRELD13.18 1.42Tetraspanin-4TSPAN43.32 0.96Trypsin-3PRSS35.31 2.35TPP performed lysates. Proteins displaying ΔTm 2°C ≤ −2°C (test versus DMSO) over replicates hits. Mean SD; N = 1, n 3. Open table tab list contains opipramol, L-687,384, perphenazine, siramesine, carbetapentane, σ (Tam Cook, 1984Tam S.W. Cook opiates antipsychotic drugs mutually inhibit (+)-3H SKF 10,047 3Hhaloperidol guinea pig brain membranes.Proc. Natl. Acad. U S 1984; 81: 5618-5621Crossref (386) Rao 1990Rao T.S. Cler J.A. Mick S.J. Dilworth V.M. Contreras P.C. Iyengar Wood P.L. Neurochemical dopaminergic effects potent sigma ligand, vivo.Neuropharmacology. 1990; 29: 1191-1197Crossref (30) Hudkins DeHaven-Hudkins, 1991Hudkins R.L. DeHaven-Hudkins D.L. M1 muscarinic antagonists interact recognition sites.Life 1991; 49: 1229-1235Crossref (51) McLarnon 1994McLarnon Sawyer Church NMDA-induced currents cultured rat hippocampal pyramidal neurons.Neurosci. Lett. 1994; 174: 181-184Crossref (9) Perregaard 1995Perregaard Moltzen E.K. Meier E. Sanchez Ligands subnanomolar preference σ2 site. 1. 3-(ω-aminoalkyl)-1H-indoles.J. Med. 1995; 38: 1998-2008Crossref (111) (σ1 receptor) showed strongest destabilization −4.60°C 0.46°C. selected further validation studies. one members family. It various tissues originally assigned opioid (Tesei 2018Tesei Cortesi Zamagni Arienti Pignatta Zanoni Paolillo Curti Rui Rossi al.Sigma endoplasmic reticulum stress “gatekeepers” their modulators emerging weapons fight against cancer.Front. Pharmacol. 2018; 711Crossref (38) GPCR but reticulum-residing transmembrane fulfilling chaperone functions, known bind drugs; e.g., ligands An manual annotation revealed actually 24 2). next interaction treatment observed 3A). strengthen observation, investigate σ-receptor selectivity, σ1- σ2-selective ligands, namely [3H]-(+)-pentazocine [3H]-1,3-di-(2-tolyl)guanidine, 3B). line results, displaced ligand Ki 799 nM 90 CETSA (Martinez Molina Nordlund, 2016Martinez Nordlund assay: biophysical situ drug mechanistic biomarker studies.Annu. Rev. Toxicol. 56: 141-161Crossref (157) treated DMSO prior harvesting gradient heat treatment. strong (ΔTm −4.33°C 1.54°C) 3C). These findings selectively binds engages involved pain modulation, reduce allodynia (pain due stimulus does normally provoke pain), antiallodynic effect reversed agonists (Merlos 2017Merlos Romero Zamanillo Plata-Salamán Vela J.M. Sigma-1 pain.in: Kim F.J. Pasternak G.W. Proteins. Evolution Concept Receptors. Springer, 2017: 131-161Crossref (28) assess whether DP68-σ1 has physiological consequences, evaluated intraplantar injection capsaicin (Entrena 2009aEntrena Cobos E.J. Nieto F.R. Cendán C.M. Baeyens Del Pozo Antagonism haloperidol metabolites hypersensitivity mice: role sigma-1 receptors.Psychopharmacology. 2009; 205: 21-33Crossref (52) Like BD-1063 4B) haloperidol, properties 4A) DP68, nM, lower potency terms compared reported 100-fold higher affinity, explain differences (Matsumoto 2001Matsumoto R.R. McCracken K.A. Friedman M.J. Pouw de Costa B.R. Bowen W.D. Conformationally restricted analogs BD1008 antisense oligodeoxynucleotide targeting produce anti-cocaine mice.Eur. 2001; 419: 163-174Crossref (70) Entrena Furthermore, pre-treatment PRE-084 4D), high-affinity agonist, partially influence 4B). BD-1063, alleviation previously 2009bEntrena Gris essential capsaicin-induced hypersensitivity: selective knockout mice.PAIN. 143: 252Abstract Full Text PDF (127) receptor. 241 substances modes addition, another seven present total set, >80% while five did show any activity assay, including agonist SKF-10047 allosteric modulator phenytoin S1) (Hayashi Su, 2004Hayashi Su T.-P. ligands: neuropsychiatric disorders.CNS drugs. 2004; 18: 269-284Crossref (271) 2005Cobos Phenytoin differentially modulates brain.Synapse. 2005; 55: 192-195Crossref (61) Among CPA, fluoxetine imipramine) progesterone found This raises question σ1-modulatory reflected fingerprints. Alternatively, could consequence other characteristics shared among compounds. Many system (CNS) lipophilic basic amines, lead accumulation increased pH; i.e., lysosomotropism (Marceau 2012Marceau Bawolak M.-T. Lodge R. Bouthillier Gagné-Henley Gaudreault R.C. Morissette Cation trapping acidic compartments: beyond concept lysosomotropic drugs.Toxicol. Appl. 2012; 259: 1-12Crossref (83) Banks Greig, 2019Banks W.A. Greig N.H. Small therapeutics: old challenges, directions, philosophic divide.Future 489-493Crossref (15) 4-propylmorpholine moiety falls category. Therefore, tested LysoTracker DND-99 S2). dye accumulates lysosomes low pH agents pH, decreased fluorescence. Indeed, reduced fluorescence intensity staining, points pH. Biosimilarity CNS-targeting fluphenazine, metoclopramide, procaine, nitrogen, detected (Gustafsdottir 2013Gustafsdottir Ljosa Sokolnicki K.L. Wilson Walpita Kemp Seiler K.P. Carrel H.A. Golub T.R. Schreiber S.L. al.Multiplex cytological states.PLoS One. 8: e80999Crossref (117) One these, several highly biosimilar DP68; however, do caused receptor-targeting activity. least partly provoked properties. pharmacophore Glennon, 2005Glennon R.A. Pharmacophore binding: application "deconstruction-reconstruction-elaboration" approach.Mini 5: 927-940Crossref (134) Scholar comprises amine flanked hydrophobic moieties, matches motif. Dissecting effect, interaction, triggers phenotype challenging. Approaches candidates require subsequent validation. prioritization (i.e., orthogonal deconvolution strategies) substantially shorten confirmation process. Moreover, protein-centric techniques will best complemented also cover non-protein space (Schneidewind Wilkinson 2020Wilkinson I.V.L. Terstappen G.C. Russell A.J. Combining experimental strategies successful deconvolution.Drug Discov. Today. 1998-2005Crossref Here, single alone, mechanisms exhibits non-protein-related (lysosomotropism). decisive conclusive individual strategies. conclusion, approach alone suggested narrowed likely one, successfully study highlights identification. Identification key objective Small-molecule assessed just portion space. fashion. enables descriptors suggests based generation hypotheses hampered lack selectivity references. Subsequent still required, and, thus, derivatization. interactions mapped technique, calls approaches. Integration couple proteome-wide mapping Using approach, derivative partners, led methods promises spur Tabled 1REAGENT RESOURCESOURCEIDENTIFIERAntibodiesGoat anti-mouse, infrared 800CW-labeledLI-COR BiosciencesRRID:AB_621842; Cat#926-32210Mouse monoclonal anti-SIGMAR1Santa CruzRRID:AB_2270263; Cat#sc-166392Chemicals, peptides recombinant proteinsChemical: [3H]-(+)-pentazocineARCCat#ART 1585; CAS: 359-83-1Chemical: [3H]di-o-tolylguanidineARCCat#ART 0624; 97-39-2Chemical: Alexa Fluor™ 568 PhalloidinThermo Fisher ScientificCat#A12380; 1333973-24-2Chemical: BD-1063TocrisCat#0883; 150208-28-9Chemical: Concanavalin A, 488 ConjugateThermo ScientificCat#C11252; N/AChemical: Hoechst 33342InvitrogenCat#H1399; 23491-52-3Chemical: Red DND-99Thermo ScientificCat#L7528; 23-1946-72-8Chemical: MitoTracker Deep RedThermo ScientificCat#M22426; 873315-86-7Chemical: NP-40 alternativeMerck-CalbiochemCat#492016; 9016-45-9Chemical: PRE-084TocrisCat#0589; 138847-85-5Chemical: Recombinant trypsinSigma-AldrichCat#3708969001; SYTO™ 14 Green Fluorescent Nucleic Acid StainThermo ScientificCat#S7576; 173080-70-1Chemical: Triethylammonium bicarbonate (TEAB) bufferSigma-AldrichCat#T7408; 15715-58-9Chemical: Tris(2-carboxyethyl)phosphine (TCEP)Sigma-AldrichCat#20491; 51805-45-9Chemical: Wheat Germ Agglutinin, 594 ScientificCat#W11262; N/ACritical commercial assaysBlocking buffer (PBS-based)LI-COR BiosciencesCat#927-70001SuperSignal West Pico ChemiluminescentThermo ScientificCat#10177533TMT10plex™ Isobaric Mass Tag Labelling ReagentsThermo ScientificCat#90110Experimental models: lineHuman: (female)CLS GmbHRRID:CVCL_0030; Cat#300364Human: (female)DSMZ GmbHRRID:CVCL_0042; Cat#ACC 57Software algorithmsCellProfiler(Carpenter 2006Carpenter Jones Lamprecht M.R. Clarke Kang I.H. Friman O. Guertin D.A. Chang J.H. Lindquist Moffat al.CellProfiler: software identifying quantifying phenotypes.Genome 2006; 7: 1-11Crossref (3178) Scholar)RRID:SCR_007358; http://cellprofiler.orgFiji ImageJ 1.52(Schindelin 2012Schindelin Arganda-Carreras I. Frise Kaynig Longair Pietzsch P