Cellular Proteomic Profiling Using Proximity Labeling by TurboID-NES in Microglial and Neuronal Cell Lines

Abstract

•Cytosolic TurboID biotinylates >50% of the proteome in microglia and neuronal cells.•TurboID-NES has minimal impacts on cellular proteomic composition and function.•About 1340 proteins labeled by TurboID differentiate microglia from neurons in vitro.•TurboID proteomic profiling captures microglial activation by lipopolysaccharide. Different brain cell types play distinct roles in brain development and disease. Molecular characterization of cell-specific mechanisms using cell type–specific approaches at the protein (proteomic) level can provide biological and therapeutic insights. To overcome the barriers of conventional isolation-based methods for cell type–specific proteomics, in vivo proteomic labeling with proximity-dependent biotinylation of cytosolic proteins using biotin ligase TurboID, coupled with mass spectrometry (MS) of labeled proteins, emerged as a powerful strategy for cell type–specific proteomics in the native state of cells without the need for cellular isolation. To complement in vivo proximity labeling approaches, in vitro studies are needed to ensure that cellular proteomes using the TurboID approach are representative of the whole-cell proteome and capture cellular responses to stimuli without disruption of cellular processes. To address this, we generated murine neuroblastoma (N2A) and microglial (BV2) lines stably expressing cytosolic TurboID to biotinylate the cellular proteome for downstream purification and analysis using MS. TurboID-mediated biotinylation captured 59% of BV2 and 65% of N2A proteomes under homeostatic conditions. TurboID labeled endolysosome, translation, vesicle, and signaling proteins in BV2 microglia and synaptic, neuron projection, and microtubule proteins in N2A neurons. TurboID expression and biotinylation minimally impacted homeostatic cellular proteomes of BV2 and N2A cells and did not affect lipopolysaccharide-mediated cytokine production or resting cellular respiration in BV2 cells. MS analysis of the microglial biotin-labeled proteins captured the impact of lipopolysaccharide treatment (>500 differentially abundant proteins) including increased canonical proinflammatory proteins (Il1a, Irg1, and Oasl1) and decreased anti-inflammatory proteins (Arg1 and Mgl2). Different brain cell types play distinct roles in brain development and disease. Molecular characterization of cell-specific mechanisms using cell type–specific approaches at the protein (proteomic) level can provide biological and therapeutic insights. To overcome the barriers of conventional isolation-based methods for cell type–specific proteomics, in vivo proteomic labeling with proximity-dependent biotinylation of cytosolic proteins using biotin ligase TurboID, coupled with mass spectrometry (MS) of labeled proteins, emerged as a powerful strategy for cell type–specific proteomics in the native state of cells without the need for cellular isolation. To complement in vivo proximity labeling approaches, in vitro studies are needed to ensure that cellular proteomes using the TurboID approach are representative of the whole-cell proteome and capture cellular responses to stimuli without disruption of cellular processes. To address this, we generated murine neuroblastoma (N2A) and microglial (BV2) lines stably expressing cytosolic TurboID to biotinylate the cellular proteome for downstream purification and analysis using MS. TurboID-mediated biotinylation captured 59% of BV2 and 65% of N2A proteomes under homeostatic conditions. TurboID labeled endolysosome, translation, vesicle, and signaling proteins in BV2 microglia and synaptic, neuron projection, and microtubule proteins in N2A neurons. TurboID expression and biotinylation minimally impacted homeostatic cellular proteomes of BV2 and N2A cells and did not affect lipopolysaccharide-mediated cytokine production or resting cellular respiration in BV2 cells. MS analysis of the microglial biotin-labeled proteins captured the impact of lipopolysaccharide treatment (>500 differentially abundant proteins) including increased canonical proinflammatory proteins (Il1a, Irg1, and Oasl1) and decreased anti-inflammatory proteins (Arg1 and Mgl2). The brain is a complex organ possessing heterogeneous populations of neurons, glia, and vascular cells. The orchestration of interactions within cell types (cell autonomous) and between cell types (noncell autonomous) support higher level processes critical to development, aging, and neurodegeneration. Protein-level analyses using mass spectrometry (MS) expand upon other systems-level analyses, including genomics and transcriptomics. Specifically, proteomics can profile total protein abundances, identify post-translational modifications, and resolve protein-level changes occurring in subcellular compartments. A central challenge to neuroproteomics is the difficulty in obtaining cell type–specific proteomes from brain tissue. Traditional approaches to isolating cell type–specific proteomes for MS including fluorescence-activated cell sorting and magnetic-activated cell sorting require fresh brain tissue, and the harsh and laborious processing itself poses challenges (1Hulett H.R. Bonner W.A. Barrett J. Herzenberg L.A. Cell sorting: automated separation of mammalian cells as a function of intracellular fluorescence.Science. 1969; 166: 747-749Crossref PubMed Scopus (2) Google Scholar, 2Schmitz B. Radbruch A. Kümmel T. Wickenhauser C. Korb H. Hansmann M.L. et al.Magnetic activated cell sorting (MACS) — a new immunomagnetic method for megakaryocytic cell isolation: comparison of different separation techniques.Eur. J. Haematol. 1994; 52: 267-275Crossref PubMed Scopus (112) Google Scholar). A majority of adult neurons do not survive the isolation process, and sampling bias for healthier non-neuronal brain cells able to withstand the isolation process limits proteomic profiling in disease states. In addition, contamination from proteins derived from nontarget cell types persists. The challenges to maintaining cellular integrity with isolation methods motivated the field to innovate novel methods of applying cell type–specific labeling to in vitro and in vivo systems. One approach to achieve cell type–specific proteomic labeling uses BioOrthogonal Non-Canonical Amino acid Tagging (BONCAT) in which a mutated methionyl-tRNA synthetase incorporates azidonorleucine, a methionine analog, into newly synthesized peptides (3Tanrikulu I.C. Schmitt E. Mechulam Y. Goddard III, W.A. Tirrell D.A. Discovery of Escherichia coli methionyl-tRNA synthetase mutants for efficient labeling of proteins with azidonorleucine in vivo.Proc. Natl. Acad. Sci. U. S. A. 2009; 106: 15285-15290Crossref PubMed Scopus (98) Google Scholar, 4Link A.J. Vink M.K.S. Agard N.J. Prescher J.A. Bertozzi C.R. Tirrell D.A. Discovery of aminoacyl-tRNA synthetase activity through cell-surface display of noncanonical amino acids.Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 10180-10185Crossref PubMed Scopus (151) Google Scholar, 5Dieterich D.C. Link A.J. Graumann J. Tirrell D.A. Schuman E.M. Selective identification of newly synthesized proteins in mammalian cells using bioorthogonal noncanonical amino acid tagging (BONCAT).Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 9482-9487Crossref PubMed Scopus (587) Google Scholar). Subsequently, cell lysates or brain homogenates undergo click chemistry with biotin-alkyne to biotinylate the azidonorleucine-containing peptides. By driving methionyl-tRNA synthetase expression under a cell type–specific promoter and enriching biotinylated peptides by streptavidin affinity purification (AP), BONCAT can purify cell type–specific newly translated proteins (6Kunkle B.W. Grenier-Boley B. Sims R. Bis J.C. Damotte V. Naj A.C. et al.Genetic meta-analysis of diagnosed Alzheimer's disease identifies new risk loci and implicates Aβ, tau, immunity and lipid processing.Nat. Genet. 2019; 51: 414-430Crossref PubMed Scopus (1281) Google Scholar, 7Alvarez-Castelao B. Schanzenbächer C.T. Hanus C. Glock C. Tom Dieck S. Dörrbaum A.R. et al.Cell-type-specific metabolic labeling of nascent proteomes in vivo.Nat. Biotechnol. 2017; 35: 1196-1201Crossref PubMed Scopus (108) Google Scholar). One advantage of this strategy lies in its ability to label and purify low-abundant and newly synthesized proteins. A limitation may be low proteomic depth and biases toward proteins with high turnover. The BONCAT approach has thus far been applied to characterize excitatory and inhibitory neurons of mice and rats in both in vivo, ex vivo, and in vitro contexts (6Kunkle B.W. Grenier-Boley B. Sims R. Bis J.C. Damotte V. Naj A.C. et al.Genetic meta-analysis of diagnosed Alzheimer's disease identifies new risk loci and implicates Aβ, tau, immunity and lipid processing.Nat. Genet. 2019; 51: 414-430Crossref PubMed Scopus (1281) Google Scholar, 7Alvarez-Castelao B. Schanzenbächer C.T. Hanus C. Glock C. Tom Dieck S. Dörrbaum A.R. et al.Cell-type-specific metabolic labeling of nascent proteomes in vivo.Nat. Biotechnol. 2017; 35: 1196-1201Crossref PubMed Scopus (108) Google Scholar, 8Hodas J.J. Nehring A. Höche N. Sweredoski M.J. Pielot R. Hess S. et al.Dopaminergic modulation of the hippocampal neuropil proteome identified by bioorthogonal noncanonical amino acid tagging (BONCAT).Proteomics. 2012; 12: 2464-2476Crossref PubMed Scopus (53) Google Scholar, 9Di Paolo A. Farias J. Garat J. Macklin A. Ignatchenko V. Kislinger T. et al.Rat sciatic nerve axoplasm proteome is enriched with ribosomal proteins during regeneration processes.J. Proteome Res. 2021; 20: 2506-2520Crossref PubMed Scopus (7) Google Scholar). To date, extension to other neuronal and glial cell types has not yet been published. In contrast to the BONCAT approach that labels only newly synthesized proteins, proximity-labeling techniques rely on biotin ligases, which biotinylate nearby interactors. BioID is a promiscuous biotin ligase engineered from the site-specific biotin ligase, BirA, endogenously produced by Escherichia coli (10Choi-Rhee E. Schulman H. Cronan J.E. Promiscuous protein biotinylation by Escherichia coli biotin protein ligase.Protein Sci. 2004; PubMed Scopus Google Scholar). BioID is this for in vivo the of biotin labeling to A. A.C. et of complex PubMed Scopus Google Scholar). to the of BioID by in the to J.A. T. et proximity labeling in cells and with Biotechnol. PubMed Scopus Google Scholar). biotin ligase, TurboID, can and biotinylate proteomes in cells and without cellular in as as in cell J.A. T. et proximity labeling in cells and with Biotechnol. PubMed Scopus Google Scholar, labeling in mammalian cells with TurboID and PubMed Scopus Google Scholar). in its TurboID has been to proteins of to protein to a subcellular of and has been of the to label cytosolic proteins of the using proximity Cell 2021; PubMed Scopus Google Scholar, biotin in for proteome and 2021; PubMed Google Scholar, S. A. et analysis of by proximity 2021; PubMed Google Scholar, A. D.C. labeling of protein and proteomes in by 2019; Scopus Google Scholar, T. cell-surface proximity labeling in Res. 2021; PubMed Scopus Google Scholar, H. R. S. biotin for and biotinylation and its to TurboID 2021; PubMed Scopus Google Scholar). In addition, to proteins to in which of TurboID in the of labeling in mammalian cells with TurboID and PubMed Scopus Google Scholar, S. T. T. et proximity labeling in Natl. Acad. Sci. U. S. A. Scopus Google Scholar). under a cell type–specific both TurboID and can label distinct proteomes in native state for downstream affinity capture and MS. which and biotinylation of cell type–specific to has been applied to proteome between and neurons in vivo T. cell-surface proximity labeling in Res. 2021; PubMed Scopus Google Scholar, T. A. et of of in PubMed Scopus Google Scholar). A novel for expression of TurboID with a to resolve proteomic of neurons and in adult brain T. A. et of of in PubMed Scopus Google Scholar, S. S. R. S. et biotin labeling in vivo neuronal and proteomic in PubMed Scopus Google Scholar). cell type–specific in vivo biotinylation of proteins as a approach to resolve distinct cellular proteomes in different in vivo homeostatic and In of in vivo proximity labeling of for cell type–specific proteomics, is to the of cytosolic biotinylation on and cellular processes in mammalian cells. is critical to characterize the of cytosolic proteins labeled by proteins are of whole-cell and identify biases of the In to support the of to label cytosolic proteins that are to cellular is to TurboID-mediated proteomics can differentiate distinct cell as neurons and we need to proteomic changes by stimuli can be and captured by the biotinylation of cytosolic proteins. are critical for proteomic from proximity labeling studies that to label the cellular proteome in mammalian in vitro and in To under we generated neuroblastoma (N2A) and microglial (BV2) cell lines that stably to label the cellular proteome the the and of cytosolic proteomic labeling by in N2A and BV2 under resting and lipopolysaccharide using MS. that expression the ability to resolve distinct cellular proteomes at the MS level under homeostatic and A of and is and and in a new and and in in a new N2A and BV2 cells in with high and and with The cells at and The cells a to a of In for MS cells in cells expressing in for The and is a from using a E. coli to the cells on into into and by of into of To the of of by the The with the cells for on the by for for and on for of to on a to at at for using the the The into and N2A and BV2 cells with a of a and a a generated in a of we with of and In biological of cell the at of or or of the with fresh and the cells on of the with of for a of the of the other with fresh to cells. this, the cells and with of the by the of cells with a and A majority of cells of of the cells the cells in in of the at of and we did not a in between and BV2 and N2A cells with or cells with or both and biotin the for at at to cellular and in a using of cells and with fresh to and cell on cells and at for at to a fresh on Cell with of cell in of and to for at at on Cell in with and without Cell lysates at in to and cell cell lysates at for at The to a fresh The protein of the cell lysates by acid using the biological for analyses and in to MS In to of protein from cell lysates in a to and using the The for at in at or at the with with by with with and with The for at in a of streptavidin to biotinylated proteins and to The as the as with for cells S. S. R. S. et biotin labeling in vivo neuronal and proteomic in PubMed Scopus Google Scholar). BV2 and N2A cells at cells and in a cells biotin treatment for the of cells with to with in for and in for at cells and in a of in in for at on the To cells of in in at cells a with for with a using analysis with from methods R. S. Y. of by with 2017; PubMed Scopus Google Scholar, J.A. S. of expression in different brain of PubMed Scopus Google Scholar). we the of with or biotinylation or biotinylated proteins, from the total of or streptavidin using the cells for both and or for using the The total of cells of the of a cell with of and or for BV2 cells for cells. The total for for BV2 cells challenge cells. The total for BV2 cells cells. The total for for BV2 cells cells. and BV2 and N2A cells in on and to with and Cell with a cell in with of to Cell to and at at for The and the in To the of the cell to a fresh The of cell at at for and of and to cell Cell on for at at for The to a fresh as The cell of high and for on for on by at for at at to fresh and at To the of subcellular of protein to with as a as a to and to and and for at and to ensure the of the for of for streptavidin of protein as and the protein and for studies to for MS analysis to the of biotinylated proteins as as ensure the of for biotinylated proteins and from the as labeling in mammalian cells with TurboID and PubMed Scopus Google Scholar, S. S. R. S. et biotin labeling in vivo neuronal and proteomic in PubMed Scopus Google of protein derived from and BV2 and N2A lysates affinity of streptavidin to of to the on for at a the from the To of fresh of The at on a on the and the and at the the biotinylated proteins of at with of for of for in of in of in and with for on a on the using to the as the with may The in of to a new with of into of of to a new and on a for The The and with of protein with at for and for of biotinylated proteins streptavidin and for to minimal between and To the and biotinylation of proteins in the of protein from a to for and of protein on a to for and biotinylated proteins in and acid for at in a on a The for a and applied at with in the on the of biotinylated proteins and of by with and for MS to from J.J. et of proteins in brain and disease in Alzheimer's 2021; Scopus Google Scholar, et proteomic analysis of Alzheimer's disease brain and changes in with microglia and PubMed Scopus Google Scholar, S. T. H. et proteomics of microglia identifies novel Alzheimer's PubMed Scopus Google Scholar, V. et approach identifies in and 2017; PubMed Scopus Google Scholar, J. J.J. et proteomic approach the disease and in PubMed Scopus Google Scholar, Y. et analysis of disease brain proteomic changes not at the PubMed Scopus Google Scholar, J.J. et analysis of the brain proteome in and Scopus Google Scholar, et analysis of the brain proteome in using Proteome Res. 2012; PubMed Scopus (53) Google Scholar). of protein from cell with at for on a by with at for on a in with of to to with of at to the of to to of and at to the for a of acid and to the to The using a in of to a of the during a on a A of the during a in the on a To of at on a The at on a with the of of to the the with to the to The using the on the One using with using the MS with a in using with a of MS to MS at a of to and higher at a of isolation and to for within a isolation MS of and using the into the murine methionine and of and of peptides with to in the a mass of and the of The mass for is and the mass for is protein and at of proteins identified on the of a can be in for the proteins identified in the and for the proteins identified in the for proteins identified on the of a that the of from into and a can be at and the and T. C. The to and Proteome Res. 2021; 20: PubMed Scopus Google Scholar). To generated by we analysis as a expression analysis to identify in protein between analyses identified proteins with within and and analysis enriched of proteins. and for to TurboID to for in TurboID and in on of that for from of MS and using The on cytokine profiling with using the function from the C.R. C. et PubMed Scopus Google Scholar). in using within including and N2A BV2 and BV2 BV2 within including BV2 BV2 proteins as with for The of changes in to using the of proteins within a for biological within a the within The responses to between and with using as To identify of differentially enriched proteins or with we The method to the of for by using and analysis analyses as generated using analysis of using with the of to identify of proteins biotinylated by that by cell by using enriched N2A and BV2 proteins and proteins, as and the of proteins identified in the as the and as using to identify in proteins A. T. et for the 2019; 103: PubMed Scopus Google Scholar). To of with into for The and of and and and the in risk loci with disease disease and from a from S. cell proteomics is to resolve disease Scopus Google Scholar). To or biotin impacts function in BV2 we and using a in a In a cell or BV2 cells in of and at in cell for cells to the of the of and cells in the cell at and of cells a and to or as a for The at and using of to of the without from the and of to to The in at in a for to the to cells with of and in a for at with of of and of and and the BV2 cells using the in with for and using and cells using the of and to cell profiling as with S. S. R. S. et biotin labeling in vivo neuronal and proteomic in PubMed Scopus Google Scholar). from of and BV2 and N2A cells or The cytokine and The from cytokine from the cytokine and to To for as S. E.M. analysis of the and to brain Scopus Google and of total on a The as cell or challenge or expression or and or biological for the BV2 in which biological a total of and in the The of biological within a analysis not by the function in maintaining the and and (2) to contamination of biotin-labeled proteins by TurboID in for analysis is in and analyses, expression analyses, a the including a by a and a The to biotinylation to the generated this and BV2 microglia and N2A neuroblastoma cells by with for at and cell sorting of cells. cells in for to stably and lines in to cell lines with we and to biotinylation of proteins in cell lysates and cytosolic of and biotinylated proteins and for biotinylated proteins identified in which endogenously biotinylated as and R. in vivo of biotinylated cell PubMed Scopus Google Scholar). The in and the of the and the biotinylation of cytosolic proteins in both and cells. cytosolic and of in BV2 and N2A cell we cell lysates for and BV2 and N2A cells of or of in to biotin for as a to BV2 cells or of on lysates in with lysates streptavidin to for biotinylated proteins lysates and with to biotinylation of proteins in cell lines and to streptavidin and with and from streptavidin for biotinylation using streptavidin and to minimal of proteins in In BV2 treatment in proteins The and for not only the of to biotinylate proteins by treatment the ability to proteins biotinylated by did not protein in the or the of N2A the of biotinylated proteins in BV2 and N2A cell we for and are in we for of proteomic from and A and for are in and for are BV2 BV2 BV2 BV2 N2A N2A N2A and N2A of the high of in the that only captures endogenously biotinylated proteins, we on or BV2 and N2A In the of of proteomes and captured cell between BV2 and N2A cells without from or treatment The can be in In of for of the and captured BV2 N2A cell for and captured the of only in BV2 cells and not in N2A cells The can be in analyses of as as proteomes that the cell between BV2 and N2A expression or proteomic the majority of in proteomes from cell as as within the BV2 cells. To the proteomic and between proteomes or N2A and and or N2A we analysis identified distinct using the method for identified of proteins abundant in proteomes and of proteins in between and and we identified of proteins distinct to which biotinylation for and we identified a of proteins that are in between cells identified proteins in BV2 of which biotinylated or MS identified proteins in N2A of which biotinylated a total of proteins or using of the with the proteome of proteins identified in the using and the and of cellular Sci. 2019; PubMed Scopus Google Scholar, S. of and Res. PubMed Scopus Google Scholar, Y. and cellular Res. 2021; PubMed Scopus Google Scholar, J.A. H. et for the of The Genet. PubMed Scopus Google Scholar, enriching a Res. 2021; PubMed Scopus Google Scholar). of the proteins) of cell and metabolic proteins and proteins derived from to biotinylation by in both cell In proteins) proteins abundant in both and a of proteins labeled by in both cell of of and proteins, that biotinylated and in both cell the in with and labeled proteins in is a function to both microglial and neuronal we that the of proteins labeled by not by cell In contrast to and that from we identified cell type–specific and that abundant in N2A or BV2 of or the a high of proteins between and N2A cells. The proteins enriched in cell and proteins, by the of N2A cells. the abundant in and from BV2 cells. The enriched in proteins with cell activation and translation, and as be for BV2 microglial cells. and cell type–specific proteomic only in the not in the may be proteins that are at low in the and are enriched by labeling and proteomic of N2A and BV2 cell types that by the approach and are at the level of of to cellular compartments. proteins) abundant in N2A with BV2 and enriched in proteins in with in N2A cells. on the other abundant in BV2 and enriched in ribosomal and with a in and cells as BV2 The of and that biotinylation can proteins in N2A and BV2 which are not at the proteomic analyses of N2A and BV2 proteomes with proteomes that proteome representative of the In addition, biotinylated of proteins mammalian cell types as as proteins to microglial or neuronal cell of biotinylated proteins captured changes at the cell type–specific proteomic in the proteomes as with the and of in mammalian cells impact cellular and and for in vivo of To expression impacts cellular under both homeostatic and we and for cytokine profiling in to and we the activity of BV2 cells. To the that expression itself impacts the we BV2 and N2A BV2 proteins and N2A proteins, including with expression of total proteins. The and BV2 or N2A proteomes can be in and proteomic that expression minimally impacted proteomes of BV2 and N2A cells under resting conditions. The of the did not in the the increased from the BV2 and and and TurboID The decreased proteins with TurboID in BV2 ligase and cell the impact of TurboID expression on N2A N2A the the increased proteins TurboID, and The decreased proteins with TurboID in N2A protein and TurboID in BV2 and N2A cells impacts the of and of the proteins identified in the a of proteins are impacted by TurboID, we did yet in proteins in both cell including in BV2 and

Topics & Concepts