The spatiotemporal organization of cytokine receptor signaling complexes is currently controversially debated. While originally dimerization of two or more receptor subunits by simultaneous interaction with the ligand has been proposed, pre-dimerization and pre-clustering has emerged as an alternative concept. We have devised single molecule fluorescence imaging techniques for quantifying cytokine receptor assembly at the plasma membrane of living cells under close to physiological conditions, i.e. low cell surface receptor expression levels (1).

Binding of fluorescent IFN to endogenous IFNAR in HeLa cells at saturating concentrations.

For this purpose, we have employed posttranslational enzymatic labeling as well as nanobodies for efficient labeling with photostable fluorescence dyes. Using dual-color total internal reflection fluorescence (TIRF) microscopy, direct visualization of diffusion and interaction of the entire population of receptors was achieved in the plasma membrane of living cells (1, 2). Single molecule localization provides unique opportunities for quantitative evaluation of spatial co-organization of receptor subunits.

Single molecule localization-based analyses for quantifying the spatiotemporal receptor organization in the plasma membrane.

Detailed studies on the Type I Interferon (IFN) receptor excluded receptor predimerization or pre-organization in absence of the ligand. However, efficient receptor dimerization was observed upon addition of IFN (1).

Transient dimerization observed in presence IFN furthermore confirmed dynamic equilibrium between binary and ternary complexes.

Transient formation of an individual IFNAR1-IFNAR2 dimer observed by single molecule co-locomotion analysis.

Two-step dimerization of the IFN receptor and formation of a dynamic ternary complex.

Meanwhile, we have confirmed ligand-induced dimerization for several other cytokine receptors including the heterodimeric type II interleukin 4 (IL-4) receptor (2) and the homodimeric erythropoietin (Epo) receptor (3), both members of the class I cytokine receptor family. Surrogate ligands based on receptor-dimerizing diabodies were found to similarly activate signaling (3). These results establish a key role of ligand-induced receptor dimerization for cytokine receptor activation.

Role of USP18 in the dimerization efficiency of IFN variants and mutants with different IFNAR1 binding affinities.

Several cytokine receptors bind different ligands, which can differentially activate cellular responses. The IFN receptor is a paradigm for such receptor plasticity, with 15 different ligands being recognized by a single cell surface receptor. Detailed studies of IFN recognition by its receptor subunits IFNAR1 and IFNAR2 revealed that binding affinity rather than structural differences are responsible for differential IFN activities, suggesting regulation at the level of receptor assembly (1-3). Indeed, systematic tailoring of IFN activities was shown to be possible by engineering its receptor binding affinities (2). Recently, negative feedback regulation by the ubiquitin-specific protease USP18 was found to be a key determinant for differential IFN activity (4). It turned out that USP18 negatively regulates receptor dimerization, probably by binding to the subunit IFNAR2 and thus interfering with possible interactions between the JAKs, as observed by single molecule dimerization experiments (5).

Based on this insight, we have proposed that differential IFN activities are caused by a temporal change in potency, which is determined by a complex interplay of different feedback mechanisms (6, 7).

Feedback mechanisms controlling differential interferon activities, which are characterized by different affinity-potency relationships observed for short-term and long-term cellular responses (6).

Reconstitution into polymer-supported membranes (PSM) is a promising approach for studying membrane protein interaction and conformations in a controlled lipid environment by advanced imaging techniques. Based on a dense PEG polymer brush and hydrophobic anchoring groups for capturing protein liposomes, we have developed a robust approach for PSM assembly on glass-type surfaces, which is compatible with single molecule fluorescence imaging, fluorescence correlation spectroscopy and atomic force microscopy (1). Rapid diffusion of a large variety of transmembrane proteins reconstituted into these PSM was found, as well as retained functionality with respect to protein-protein interactions (1, 2).

As the lipid composition can be controlled in these PSM, separation of ld and lo lipid phases is readily achieved (3). Moreover, micropatterned organization of anchoring groups provides unique possibilities to control PSM formation and lipid phase-separation in a spatially resolved manner (3).

Spatially controlled assembly of phase-separated PSM based on micropatterned palmitic and oleic acid moieties.

As only very minor protein quantities are typically demanded for single molecule applications, production and rapid transfer from mammalian cells is possible (4).

Rapid transfer of transmembrane receptors expressed in HeLa cells into polymer-supported membranes for single molecule fluorescence imaging.

Based on this approach, ligand-induced heterodimerization of the IFN receptor could be quantified in a reconstituted system (4).

Single molecule dimerization of transmembrane IFNAR1 and IFNAR2 in polymer-supported membranes.

Next to transmembrane receptors, reconstitution of integral membrane proteins was achieved including beta-barrel outer membrane proteins, for which spontaneous aggregation into clusters was observed (5).

Micropatterned phase-separated PSMs open exciting perspectives for bioanalytical applications. Based on membrane tethering of His-tagged proteins via tris-NTA lipid, phase partitioning upon dimerization was exploited as a readout for protein-protein interaction (6). Thus, weak binding affinities up to 1 mM could be determined for proteins directly captured to surfaces from mammalian cell lysates (6).

Inorganic nanoparticles provide unique physical properties for visualizing and manipulating cellular processes down to the single molecule level. However, biofunctionalization of these nanoparticles for efficient and stoichiometrically defined conjugation with target proteins without biasing their function remains challenging. We have developed approaches for controlling conjugation stoichiometry by exploiting electrostatic steering of coupling reactions (1, 2). Thus, unbiased diffusion of protein diffusion and interaction at the plasma membrane could be probed using monofunctional quantum dots for cell surface receptor labeling.

Spatiotemporal dynamics of IFN-receptor complexes in living cells visualized by dual color quantum dot tracking.

For efficient nanoparticle targeting inside living cells, we have optimized a substrate for fast, covalent reaction with HaloTag fusion proteins (3). On our quest to apply nanoparticles in the cytosol of living cells, we identified autophagy as a critical mechanism for nanoparticle clearance (4). Using massive PEGylation, monofunctionalized stealth nanoparticles were generated, which were efficiently targeted to a protein in the outer mitochondrial membrane (4).

Protein immobilization on solid support opens versatile possibilities for protein interaction analysis. We have developed surface modification for functional protein immobilization based on site-specific protein capturing. For stable, yet reversible protein immobilization we have developed tris(nitrilotriacetic acid) (tris-NTA), which binds oligohististidine-tagged proteins with nanomolar affinity due to multivalent interactions (1, 2). We have established a large repertoire of molecular tools for surface modification and protein labeling based on tris-NTA (3). For increased specificity and stability of protein capturing, this approach was complemented by enzyme-based coupling techniques including enzymatic phosphopantetheinyl transfer as well as the HaloTag (4). For spatially resolved protein immobilization, we have combined these capturing techniques with soft and photolithographic surface patterning techniques, yielding a versatile toolbox for functional protein organization on surfaces (5, 6).

Microtubule (red) transported by motor proteins (green) immobilized via micropattern tris-NTA (5).

A fundamental objective of these approaches is to develop robust tools for interfacing proteins with surfaces while maintaining the physiological context, e.g. lipid environment of membrane proteins or cellular interaction partners. To this end, we developed micropatterned surface architectures for capturing protein in the plasma membrane of living cells (7, 8).

Surface architecture for capturing HaloTag fusion proteins in the plasma membrane of living cells.

Based on this technique protein-protein interactions at membranes such as receptor hetero-dimerization or effector recruitment can be readily detected and quantified in living cells.

Triggered micropatterning of IFNAR1 (green) in the plasma membrane of living cells followed by ligand-induced association of IFNAR2 (red) (8).

For studying cytosolic protein complexes, we have established surface architectures for pull-down of GFP-tagged proteins from single cells (SiCPull). In combination with single molecule TIRF microscopy, we succeeded to quantify stability and stoichiometry of complexes captured by SiCPull.

Concept of single cell pulldown for quantifying the stability and the stoichiometry of protein complexes (9).

Visualization of cellular processes in living cells with highest spatial and temporal resolution plays a key role in modern cell biology. Recent breakthroughs in overcoming the diffraction limitation in far-field fluorescence microscopy have opened exciting perspectives for live cell imaging. We have focused on single molecule localization microscopy (SMLM) techniques such as fluorescence photoactivation localization microscopy (FPALM) and stochastic optical reconstruction microscopy (STORM). By combining FPALM and direct STORM, we could demonstrate triple color superresolution imaging in living cells (1).

Triple-color superresolution imaging in living cells by combining FPALM (PAGFP and PAmCherry) and dSTORM (ATTO655 coupled via the HaloTag) reveals coorganization of receptor subunits (green, red) in the context of the cortical actin cytoskeleton (blue).

While PALM is typically limited to a single read-out cycle, life cell dSTORM opens exciting possibilities for superresolution imaging of cellular nanostructures over prolonged time scales (1).

Time-lapse superresolution imaging of clathrin dynamics at the plasma membrane using dSTORM (ATTO655).

Classic SMLM requires high overexpression of target proteins in order to achieve sufficient density to fulfill the Nyquist criterion, which may affect cellular functions and morphologies. For SMLM at physiological expression levels, we have introduced tracking and localization microscopy (TALM), which is based on a small ensemble of permanently fluorescent species, which successively explore cellular nanostructures by diffusion in the living cell (2, 3). This technique is particularly applicable to membrane proteins as their mobility is compatible with precise localization. Thus, not only morphologies and proteins mobility, but also connectivities or barriers can be identified. Pair correlation applied to TALM imaging thus provides information about spatial and temporal protein organization, which opens exciting possibilities for identifying and quantifying protein clustering and co-clustering (4).

Application of SMLM on biological questions requires access to dedicated image analysis tools not only for robust localization and tracking of individual emitters, but also for versatile quantification of spatial and temporal properties such as (co-)clustering, co-locomotion, complex stoichiometries or morphologies. We are constantly developing an integrated software package providing versatile tools for such analyses. We share these advanced imaging and evaluation techniques within the framework of the DFG-funded Integrated Bioimaging Facility Osnabrück (iBiOs) at the Center of Cellular Nanoanalytics (CellNanOs).