Correlated STORM-homoFRET imaging reveals highly heterogeneous membrane receptor structures

Abstract

Mapping the self-organization and spatial distribution of membrane proteins is key to understanding their function. Developing methods that can provide insight into correlations between membrane protein colocalization and interactions is challenging. We report here on a correlated stochastic optical reconstruction microscopy/homoFRET imaging approach for resolving the nanoscale distribution and oligomeric state of membrane proteins. Using live cell homoFRET imaging of carcinoembryonic antigen-related cellular adhesion molecule 1, a cell-surface receptor known to exist in a complex equilibrium between monomer and dimer/oligomer states, we revealed highly heterogeneous diffraction-limited structures on the surface of HeLa cells. Furthermore, correlated super-resolved stochastic optical reconstruction microscopy imaging showed that these structures comprised a complex mixture and spatial distribution of self-associated carcinoembryonic antigen-related cellular adhesion molecule 1 molecules. In conclusion, this correlated approach provides a compelling strategy for addressing challenging questions about the interplay between membrane protein concentration, distribution, interaction, clustering, and function. Mapping the self-organization and spatial distribution of membrane proteins is key to understanding their function. Developing methods that can provide insight into correlations between membrane protein colocalization and interactions is challenging. We report here on a correlated stochastic optical reconstruction microscopy/homoFRET imaging approach for resolving the nanoscale distribution and oligomeric state of membrane proteins. Using live cell homoFRET imaging of carcinoembryonic antigen-related cellular adhesion molecule 1, a cell-surface receptor known to exist in a complex equilibrium between monomer and dimer/oligomer states, we revealed highly heterogeneous diffraction-limited structures on the surface of HeLa cells. Furthermore, correlated super-resolved stochastic optical reconstruction microscopy imaging showed that these structures comprised a complex mixture and spatial distribution of self-associated carcinoembryonic antigen-related cellular adhesion molecule 1 molecules. In conclusion, this correlated approach provides a compelling strategy for addressing challenging questions about the interplay between membrane protein concentration, distribution, interaction, clustering, and function. Identifying the factors that govern the spatial-temporal distribution and interaction of membrane proteins is critical for addressing key questions in cell biology. These include determining the structural underpinnings of intercellular engagement and the mechanisms of effector signaling upon receptor activation. Identifying the key structure-function relationships relies upon quantifying protein dynamics and association state (1Kellner R.R. Baier C.J. Willig K.I. Hell S.W. Barrantes F.J. Nanoscale organization of nicotinic acetylcholine receptors revealed by stimulated emission depletion microscopy.Neuroscience. 2007; 144: 135-143Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 2Heller J. Rusakov D. Subcellular distribution and trafficking of astroglial receptors monitored with super-resolution microscopy.Glia. 2015; 63: E451-E452Google Scholar, 3Eggeling C. Studying plasma membrane bioactivity with super-resolution STED microscopy.FEBS J. 2015; 282: 51Google Scholar, 4Zhuang X. Illuminating biology at the nanoscale with super-resolution fluorescence microscopy.FEBS J. 2015; 282: 1PubMed Google Scholar). Techniques such as Förster resonance energy transfer (FRET) (5Piston D.W. Rizzo M.A. FRET by fluorescence polarization microscopy.Methods Cell Biol. 2008; 85: 415-430Crossref PubMed Scopus (61) Google Scholar, 6Rizzo M.A. Piston D.W. High-contrast imaging of fluorescent protein FRET by fluorescence polarization microscopy.Biophys. J. 2005; 88: L14-L16Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar, 7Roy R. Hohng S. Ha T. A practical guide to single-molecule FRET.Nat. Met. 2008; 5: 507-516Crossref PubMed Scopus (1545) Google Scholar) and fluorescence correlation spectroscopy (8Machan R. Wohland T. Recent applications of fluorescence correlation spectroscopy in live systems.FEBS Lett. 2014; 588: 3571-3584Crossref PubMed Scopus (87) Google Scholar, 9Singh A.P. Wohland T. Applications of imaging fluorescence correlation spectroscopy.Curr. Opin. Chem. Biol. 2014; 20: 29-35Crossref PubMed Scopus (27) Google Scholar) are often employed to quantify and spatially map protein–protein interactions, clustering, self-association, and dynamics. One commonly used approach for characterizing self-association is homoFRET, which occurs due to energy transfer between identical fluorophores that can act as either donors or acceptors. In homoFRET, nonradiative energy transfer between identical fluorophores located within the requisite Förster distance results in depolarized emission upon excitation with linearly polarized light. As such, this method reports on interactions that occur over distances less than 10 nm and with transfer efficiencies dependent on the fluorophore, local structure, and extent of conformational flexibility (5Piston D.W. Rizzo M.A. FRET by fluorescence polarization microscopy.Methods Cell Biol. 2008; 85: 415-430Crossref PubMed Scopus (61) Google Scholar, 6Rizzo M.A. Piston D.W. High-contrast imaging of fluorescent protein FRET by fluorescence polarization microscopy.Biophys. J. 2005; 88: L14-L16Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar, 10Cameron W.D. Bui C.V. Hutchinson A. Loppnau P. Gräslund S. Rocheleau J.V. Apollo-NADP(+): a spectrally tunable family of genetically encoded sensors for NADP(+).Nat. Met. 2016; 13: 352-358Crossref PubMed Scopus (83) Google Scholar). The result of homoFRET imaging is a diffraction-limited pixel-wise anisotropy map that reports the spatial distribution of self-association. Although FRET-based approaches provide insights into the proximity of two interacting fluorophores, the diffraction limit classically limits the extent of spatial localization. Developments in super-resolution fluorescence microscopy or single molecule localization microscopy (SMLM) are allowing researchers to resolve objects with a localization precision of 20 to 100 nm, often revealing subdiffraction limit structural features (11Hell S.W. Wichmann J. Breaking the diffraction resolution limit by stimulated-emission - stimulated-emission-depletion fluorescence microscopy.Opt. Lett. 1994; 19: 780-782Crossref PubMed Scopus (4546) Google Scholar, 12Klar T.A. Engel E. Hell S.W. Breaking Abbe's diffraction resolution limit in fluorescence microscopy with stimulated emission depletion beams of various shapes.Phys. Rev. E Stat. Nonlin Soft Matter Phys. 2001; 64: 066613Crossref PubMed Scopus (59) Google Scholar). These techniques can be subdivided into two categories: ensemble imaging approaches, which include stimulated emission depletion (1Kellner R.R. Baier C.J. Willig K.I. Hell S.W. Barrantes F.J. Nanoscale organization of nicotinic acetylcholine receptors revealed by stimulated emission depletion microscopy.Neuroscience. 2007; 144: 135-143Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar, 13Klar T.A. Dyba M. Hell S.W. Stimulated emission depletion microscopy with an offset depleting beam.Appl. Phys. Lett. 2001; 78: 393-395Crossref Scopus (39) Google Scholar, 14Westphal V. Blanca C.M. Dyba M. Kastrup L. Hell S.W. Laser-diode-stimulated emission depletion microscopy.Appl. Phys. Lett. 2003; 82: 3125-3127Crossref Scopus (60) Google Scholar, 15Hein B. Willig K.I. Hell S.W. Stimulated emission depletion (STED) nanoscopy of a fluorescent protein-labeled organelle inside a living cell.Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 14271-14276Crossref PubMed Scopus (377) Google Scholar, 16Hell S.W. Sahl S.J. Bates M. Zhuang X. Heintzmann R. Booth M.J. et al.The 2015 super-resolution microscopy roadmap.J. Phys. D-Applied Phys. 2015; 48: 443001Crossref Scopus (267) Google Scholar) and structured illumination microscopies (16Hell S.W. Sahl S.J. Bates M. Zhuang X. Heintzmann R. Booth M.J. et al.The 2015 super-resolution microscopy roadmap.J. Phys. D-Applied Phys. 2015; 48: 443001Crossref Scopus (267) Google Scholar, 17Ball G. Demmerle J. Kaufmann R. Davis I. Dobbie I.M. Schermelleh L. SIMcheck: a toolbox for successful super-resolution structured illumination microscopy.Sci. Rep. 2015; 5: 15915Crossref PubMed Scopus (197) Google Scholar, 18Moerner W.E. Nobel Lecture: single-molecule spectroscopy, imaging, and photocontrol: foundations for super-resolution microscopy.Rev. Mod. Phys. 2015; 87: 1183-1212Crossref Scopus (69) Google Scholar, 19Tang Y.Q. Dai L. Zhang X. Li J. Hendriks J. Fan X. et al.SNSMIL, a real-time single molecule identification and localization algorithm for super-resolution fluorescence microscopy.Sci. Rep. 2015; 5: 11073Crossref PubMed Scopus (19) Google Scholar), and SMLM, which includes stochastic optical reconstruction microscopy (STORM) (20Rust M.J. Bates M. Zhuang X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM).Nat. Met. 2006; 3: 793-795Crossref PubMed Scopus (5919) Google Scholar, 21Benke A. Manley S. Live-cell dSTORM of cellular DNA based on direct DNA labeling.Chembiochem. 2012; 13: 298-301Crossref PubMed Scopus (62) Google Scholar, 22Lehmann M. Gottschalk B. Puchkov D. Schmieder P. Schwagerus S. Hackenberger C.P. et al.Multicolor caged dSTORM resolves the ultrastructure of synaptic vesicles in the brain.Angew. Chem. Int. Ed. Engl. 2015; 54: 13230-13235Crossref PubMed Scopus (24) Google Scholar) and photoactivated localization microscopy (23Betzig E. Patterson G.H. Sougrat R. Lindwasser O.W. Olenych S. Bonifacino J.S. et al.Imaging intracellular fluorescent proteins at nanometer resolution.Science. 2006; 313: 1642-1645Crossref PubMed Scopus (6554) Google Scholar). A key limitation for SMLM lies in determining whether the localized structures are in fact interacting or are simply clustered in a noninteracting fashion. This is further complicated by the very nature of the SMLM acquisition and image processing algorithms, which typically rely on stochastic excitation of, and emission from, individual fluorophores. This challenge prompted us to develop an integrated approach that would marry the high spatial resolution of SMLM with self-association data provided by homoFRET. Since homoFRET images are diffraction-limited, the pixel-wise r-values can only report on the average association state within the pixel, and not the true spatial distribution. In principle, correlating super-resolution and homoFRET images would provide new insights into the relationships between oligomeric state and spatial distribution (24Oreopoulos J. Gray-Owen S.D. Yip C.M. High density or urban sprawl: what works best in biology?.ACS Nano. 2017; 11: 1131-1135Crossref PubMed Scopus (3) Google Scholar). Here we describe the methodology behind a correlative STORM/homoFRET technique (Figs. 1 and S1). When applied to membrane or membrane-associated proteins, this correlated approach can reveal their nanoscale distribution and oligomerization or self-association state. For the purposes of the present work, we focused on the prototype member of the carcinoembryonic antigen-related cellular adhesion molecule (CEACAM) family, which are cell surface glycoproteins involved in homophilic and heterophilic intercellular interactions involved in cellular growth, differentiation, tumorigenesis, inflammation, and infection (25Gray-Owen S.D. Blumberg R.S. CEACAM1: contact-dependent control of immunity.Nat. Rev. Immunol. 2006; 6: 433-446Crossref PubMed Scopus (390) Google Scholar). Despite CEACAMs’ importance in health and disease, the nature of CEACAM interactions and their mechanisms of signaling remain poorly understood. The full-length splice variant of CEACAM1 has an extracellular domain consisting of an amino-terminal IgV-like domain and three IgC2-like domains, with all these domains being heavily glycosylated (26Nagaishi T. Pao L. Lin S.H. Iijima H. Kaser A. Qiao S.W. et al.SHP1 phosphatase-dependent T cell inhibition by CEACAM1 adhesion molecule isoforms.Immunity. 2006; 25: 769-781Abstract Full Text Full Text PDF PubMed Scopus (107) Google Scholar, 27Chen C.J. Kirshner J. Sherman M.A. Hu W. Nguyen T. Shively J.E. Mutation analysis of the short cytoplasmic domain of the cell-cell adhesion molecule CEACAM1 identifies residues that orchestrate and Biol. Chem. 2007; 282: Full Text Full Text PDF PubMed Scopus Google Scholar). This protein is into the membrane a and a cytoplasmic domain that two these (25Gray-Owen S.D. Blumberg R.S. CEACAM1: contact-dependent control of immunity.Nat. Rev. Immunol. 2006; 6: 433-446Crossref PubMed Scopus (390) Google Scholar). has known that CEACAM1 can exist as a a or and that oligomeric signaling these are A. C.M. Yip C.M. et signaling in the carcinoembryonic antigen-related cellular adhesion molecule 1 oligomeric state to control cell adhesion Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar, E. et al.The CEACAM1 domain and and is for of CEACAM1 Cell Biol. PubMed Scopus Google Scholar, E. B. adhesion and of and of and Cell Biol. PubMed Scopus Google Scholar). Furthermore, the dynamics and between these oligomeric remain poorly These questions prompted us to develop a correlative approach for characterizing the nanoscale distribution of and on the cell In this homoFRET imaging a single domain with The HeLa are of the cell that allowing of individual family to function. has that can between the and in cell that this protein A. C.M. Yip C.M. et signaling in the carcinoembryonic antigen-related cellular adhesion molecule 1 oligomeric state to control cell adhesion Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar, I. H. M. B. for of cell-cell adhesion molecule in J. PubMed Scopus Google Scholar). for these we used HeLa with a CEACAM1 with a the of fluorescence homoFRET HeLa and the due to the nature of and we anisotropy to what would be for a this and all the to for of the that and In we that the pixel-wise anisotropy map is of the and as a between the two emission Live-cell imaging of in HeLa revealed the of and the comprised of or oligomeric CEACAM1 with anisotropy to the high with anisotropy to are not into in the homoFRET as the anisotropy two the with to are a for all The of that the equilibrium is by the of the cell interactions with and biology In we homoFRET on a CEACAM1 known to be to These revealed a to anisotropy for this We homoFRET on a CEACAM1 of interactions A. C.M. Yip C.M. et signaling in the carcinoembryonic antigen-related cellular adhesion molecule 1 oligomeric state to control cell adhesion Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). In this the distribution is of in and Although not the of this these insights the fact that oligomeric CEACAM1 in the these the of this approach for self-association is a spatial distribution of the surface of with an correlation between CEACAM1 density by and the of CEACAM1 that is in an oligomeric state by anisotropy the diffraction-limited spatial resolution of of CEACAM1 organization within these this we applied imaging to the that by For imaging, is with an single domain E. C.M. A. D. J. H. microscopy of with or in 2015; 88: PubMed Scopus Google Scholar, Bates M. Zhuang X. of fluorophores for in super-resolution Met. PubMed Scopus Google Scholar). A key challenge in the of for imaging lies in the and flexibility of the which are typically nm this can result in in the localization. we to a comprised of the to domain of a This is to nm in and has to to the of and with high and M.A. S.J. of in the of fluorescent 6: Full Text Full Text PDF Scopus Google Scholar). only 1 to fluorophores, which the of the individual imaging and the of can to and or of single This can to the of or in the of a distribution. As by et D. on membrane of a single HeLa and cell by microscopy Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar), in to the of the has to be than that of the In this the distances between individual are to we to the average distances between individual for the and to that we We 10 the for a cell and super-resolved the individual to a analysis on the and the and approaches, three to for The average distances for for of the and nm and nm a These results that with localization that the to the localization of and These results the of strategy for imaging of protein For imaging, the HeLa with and with 1 of for of fluorescence images revealed that the and highly for of three In the is to an which is to a to the cytoplasmic domain of we the of the over a 1 acquisition We this to be the localization precision of These data that this approach provides for highly and correlative STORM/homoFRET homoFRET of can be on live with the and of the cells. these correlation of the anisotropy and anisotropy and of the images that the and by these an or oligomeric The anisotropy revealed in anisotropy upon a to the and the anisotropy to to to in the This that be in nature due to the of This is with that only at an on membrane D. on membrane of a single HeLa and cell by microscopy Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar). and as in to self-association states, clustering, and cell diffraction-limited microscopy revealed that CEACAM1 is either in a or as various and the nature of the CEACAM1 within the analysis on super-resolved CEACAM1 the a of into by the to of two For the distance to the can be a distance distribution cell or of of can be and by the of and and this analysis revealed the of a distribution of the distance of the of and Since the is in the in the nature of we that the and of in the and membrane of HeLa and are as a of density that to with and density for membrane proteins and are to exist in to their to intercellular or to with structures C. A. J. L. et organization and dynamics of the with the in J. 2016; PubMed Scopus Google Scholar, H. imaging and spatial of receptors in Natl. Acad. Sci. U. S. A. 2016; PubMed Scopus Google Scholar, M.J. J. et resolution microscopy that is for spatial organization of and signaling in Scopus Google Scholar, G. The of and in a complex interactions by super-resolution microscopy and single protein Cell 2016; PubMed Scopus Google Scholar). not to this methods results a in within intercellular the we and for of the and cells. are the average for and nm and nm and than These between the and and in their association to the membrane and analysis the diffraction-limited homoFRET image to a image and on the CEACAM1 of and density can be correlated to their with a anisotropy can be with a subdiffraction we the of processing in the of the super-resolution with the homoFRET the fact that the individual that this is this analysis is that with anisotropy a of nanoscale is that anisotropy to be of to be comprised of and These that CEACAM1 into three a distribution of of and and of We that by SMLM P. T. D. M. J. protein in the plasma membrane and correlation Met. PubMed Scopus Google Scholar). In to this we correlated the anisotropy map to map which revealed that the anisotropy at the of the on that are to their This is with of a correlation between self-association and protein The present has that with homoFRET imaging direct of the spatial and oligomeric state of surface membrane proteins. In the of we that this can exist in three clustered states, with self-association consisting of present on the cell of and and of that to be within cell These are not diffraction-limited The between CEACAM1 density and to be in an oligomeric state provides new insight into the dynamics of this The domain of to be within cell membrane that are in and which are to P. V. The CEACAM1 not the cytoplasmic of membrane 2008; PubMed Scopus Google Scholar). that the cytoplasmic domain of CEACAM1 is to be by family within these domains, is to that either this the of effector proteins CEACAM1 within is not for P. V. The CEACAM1 not the cytoplasmic of membrane 2008; PubMed Scopus Google Scholar), cytoplasmic domain can with the at of S. Li X. cell-cell Biol. 11: PubMed Scopus Google Scholar), of to the in this we results that and within within whether the distribution and oligomeric state of CEACAM1 association with or focused on the full-length variant of the of in of and strategy this approach to all membrane proteins. This method can be to fluorescent proteins. The of to fluorophores that the as as fluorescent the for STORM/homoFRET the of cellular proteins. We the importance of a analysis strategy in to single molecule into In to be the approach as for the of into these localization the in density are of in density that the distribution of The distance or density revealed a distribution for CEACAM1 in the three that diffraction-limited fluorescence microscopy Although of by in of SMLM techniques H. imaging and spatial of receptors in Natl. Acad. Sci. U. S. A. 2016; PubMed Scopus Google Scholar), be that factors such as the of and the of reconstruction and analysis a in is to the dynamics of the that we in the present This further in the nature of CEACAM1 self-association and the of on function. In we are the of correlated approach to the intercellular a of revealing the nanoscale of CEACAM1 at the can to CEACAM1 and and are by intercellular interaction CEACAM family new insights their and the for such a of HeLa in with by by and at with to for and on with and to and for to to with the by 10 of the in of for by the of and at of the 100 is to imaging are for to upon which are to microscopy on a microscopy integrated with an an a high A of used to the to the surface of the of an nm of an nm with a of at the A used to spectrally the A polarization used to between polarized and depolarized emission images are two these we an of for a single cell over a and to 10 that not a in to for the acquisition homoFRET images are as of images is the polarized fluorescence a to the excitation and the depolarized fluorescence a to the excitation images of 1, and an of for for and that in their the various optical of the an is the is anisotropy can be the The of by the of for a 100 are an optical is or into the optical of the as to For of the 1 protein into a of at 10 in which in to the The for 1 at into a at HeLa at in in for 10 with and with for with to residues and with in for 1 The CEACAM1 in for to with The for in with to For and to the used for The into the and to to analysis the on three used to the between and images T. S. M. of single-molecule localization Cell Biol. 2014; PubMed Scopus Google Scholar). and of the and the high of colocalization between and of the of a In to stochastic for is to The of and in This provides that a high for and the of by and a A nm is to a of 20 as the which is used to to an to of fluorophores on in a stochastic over the acquisition a super-resolved to images with an of This of to the of the which single localization distance to be of the localization precision in to be to a resolution to the average localization precision of 20 to nm T. S. M. of single-molecule localization Cell Biol. 2014; PubMed Scopus Google Scholar). are and the localization algorithm localization localization and the of single are based on their localization precision and in to are for cell and We that and in of nm, with a localization precision with a within nm are into a single localization in to for of the or on a The used to for Despite the in and and due to the stochastic nature of the in the of fluorophores protein and imaging the super-resolved localization not an the to the of proteins in a of protein a a of protein distribution and used by on of SMLM techniques with a for analysis and spatial analysis with P. T. D. M. J. protein in the plasma membrane and correlation Met. PubMed Scopus Google Scholar, S.H. Li P. Using Sci. 5: PubMed Scopus Google Scholar). these methods present as analysis which the analysis S.H. Li P. Using Sci. 5: PubMed Scopus Google Scholar, T. G. of the spatial organization of with Scopus Google Scholar, quantify super-resolution images and due to 2012; PubMed Scopus Google Scholar). we which is a method that into E. A. C. A. D. et a method to and quantify super-resolution microscopy Met. 2015; PubMed Scopus Google Scholar). The by all the a of of that what is as a In this is single with their Since a single image would to on the a algorithm to a in a of for a of with We the for this that localization is not into the localization is a as the localization in the super-resolved are not in the distance in data are upon This The that of with the of this We for of single domain and for the analysis We and for on this We of for the of the and We for in cell A. S. and C. A. D. A. D. A. D. A. D. S. and C. and S. and C. S. and C. C. This by of and the and of

Topics & Concepts