Stimulated emission depletion (STED) microscopy enables super-resolution imaging of complex biological samples in 3D, in large volumes, and live. However, molecular quantification with STED has remained underexplored. Here, we present a straightforward approach for quantitative STED that enables molecule counting. For this purpose, we designed DNA-fluorophore labels that enable signal amplification and allow for reliable intensity-based quantitative imaging. We demonstrate accurate molecule counting on DNA origami. Furthermore, we visualized and quantified EGF receptor monomers and dimers in cells. In summary, we introduce a robust, fast, and easy-to-implement tool for quantitative STED microscopy with single-protein resolution.

Major histocompatibility complex class I (MHC I) molecules present antigenic peptides to cytotoxic T cells, a process central to immune surveillance. However, the nanoscale spatial organization of peptide-MHC I (pMHC I) on human dendritic cells (DCs), key initiators of cytotoxic T cell responses, remains largely unexplored. Here, we combine high-affinity soluble T cell receptors with DNA-based point accumulation for imaging in nanoscale topography (DNA-PAINT) to quantitatively map and count defined pMHC I complexes at single-molecule resolution on HLA-A*02:01-expressing cells and primary human monocyte-derived DCs. We found no evidence for higher-order pMHC I nanoclusters under conditions of extracellular peptide exchange or physiological intracellular loading. Instead, detected signals correspond to individual pMHC I complexes. Notably, DC differentiation and activation modulate pMHC I surface abundance and spatial compartmentalization. These findings refine current models of antigen presentation by emphasizing regulation through surface density and spatial distribution, and establish a quantitative framework for epitope-specific, single-molecule quantification of antigen presentation in human immune cells.

We developed a fully-automated smart fluorescence microscopy workflow that combines fast confocal microscopy, morphological object screening, event-triggered and targeted 3D super-resolution STED microscopy and quantitative image analysis. This workflow was tailored to search for and identify rare objects or events in cells, and to subsequently direct the full performance of the microscope towards a nanoscale characterization of these objects. Using this workflow, we measured the 3D nano-morphology of paraspeckles, a phase-separated membrane-less organelle (MLO) located in the nucleus of eukaryotic cells. Furthermore, we applied this workflow to detect cell organelle contact sites in living cells. This smart microscopy approach is resource-efficient, enables targeted and high-throughput characterization of rare objects and events, and is transferable to a large variety of cellular structures.

Super-resolution microscopy with DNA-fluorophore labels is primed for multi-target imaging of cell biological samples. However, direct interaction with the sample is required to exchange or add DNA-fluorophore labels in each imaging round, which can impair the accuracy of the imaging data at the nanometer scale. To bypass this requirement, we introduce PhotoPAINT, a wash-free method that employs DNA oligonucleotides equipped with photocaging groups. Irradiation with light removes these photo-modulatable groups and changes the hybridization properties of DNA labels, enabling light-modulated targeting. We demonstrate this concept by imaging various cellular targets with confocal microscopy, single-molecule localization microscopy (SMLM), and stimulated emission depletion (STED) microscopy.

Abstract Horizontal gene transfer in Acinetobacter baumannii has been attributed to conjugative pili, transduction, and natural transformation. Here, we describe a previously unrecognized type of chromosomal DNA exchange mediated by cell-envelope conduits that establish direct cytoplasmic continuity between neighboring A. baumannii cells. Cryo-electron tomography reveals that these conduits, which were found in multiple clinical isolates, comprise an outer membrane, a peptidoglycan layer, and an inner membrane. The ∼65 nm-thick conduits contain two ∼2 nm-thick filaments, which super-resolution microscopy identifies as DNA. Interstrain horizontal gene transfer assays coupled with whole-genome sequencing demonstrate the transfer and homologous recombination of chromosomal segments up to 1.1 Mbp, corresponding to as much as 27% of the A. baumannii genome. Together, these findings provide direct structural and functional evidence for a conduit-mediated large-scale chromosomal DNA exchange, thus expanding the known repertoire of horizontal gene transfer strategies that may contribute to the remarkable genomic plasticity of A. baumannii .

Receptor tyrosine kinase signaling is initiated by extracellular ligand binding, which drives the formation of membrane-protein assemblies that activate intracellular signal transduction. Accurately resolving the molecular composition of these assemblies in situ remains challenging due to their nanoscale dimensions and intrinsic heterogeneity. Here, we introduce a single-molecule super-resolution imaging and analysis workflow designed to resolve and quantitatively characterize individual membrane-protein assembly sites in cells. We apply this approach to the nanoscale organization of the epidermal growth factor receptor (EGFR) and its adaptor protein Grb2 following stimulation with the native ligand epidermal growth factor. As activation progresses, we observe a reduction in EGFR density at the plasma membrane, a progressive accumulation of Grb2 at EGFR assembly sites, and an increase in both dimeric and higher-order oligomeric EGFR. The experimental and analytical framework presented here is broadly applicable to the study of diverse membrane-protein assemblies.

Bacterial chromosomes are spatiotemporally organized and sensitive to environmental changes. However, the mechanisms underlying chromosome configuration and reorganization are not fully understood. Here, we use single-molecule localization microscopy and live-cell imaging to show that the Escherichia coli nucleoid adopts a condensed, membrane-proximal configuration during rapid growth. Drug treatment induces a rapid collapse of the nucleoid from an apparently membrane-bound state within 10 min of halting transcription and translation. This hints toward an active role of transertion (coupled transcription, translation, and membrane insertion) in nucleoid organization, while cell wall synthesis inhibitors only affect nucleoid organization during morphological changes. Further, we provide evidence that the nucleoid spatially correlates with elongasomes in unperturbed cells, suggesting that large membrane-bound complexes might be hotspots for transertion. The observed correlation diminishes in cells with changed cell geometry or upon inhibition of protein biosynthesis. Replication inhibition experiments, as well as multi-drug treatments highlight the role of entropic effects and transcription in nucleoid condensation and positioning. Thus, our results indicate that transcription and translation, possibly in the context of transertion, act as a principal organizer of the bacterial nucleoid, and show that an altered metabolic state and antibiotic treatment lead to major changes in the spatial organization of the nucleoid.

We present Rho-tag and SiR-tag, engineered protein tags derived from bacterial multidrug-resistance proteins that bind unsubstituted (silicon-) rhodamines with nanomolar affinity, enabling fast, reversible, and fluorogenic protein labeling. In live cells, Rho-tag labeling occurs within seconds — faster than HaloTag7 — and the tags are compatible with super-resolution methods like STED, SMLM, and MINFLUX. The high specificity of Rho-tag and SiR-tag for unsubstituted rhodamines allows their use alongside HaloTag7 and SNAP-tag. In vivo applications are demonstrated by efficient neuronal labeling in zebrafish larvae. ### Competing Interest Statement The authors have declared no competing interest. Max Planck Society, https://ror.org/01hhn8329 Deutsche Forschungsgemeinschaft, TRR 186, SFB1507 (project-id: 450648163) and INST 161/778-1 FUGG.

Advances in microscopy imaging enable researchers to visualize structures at the nanoscale level thereby unraveling intricate details of biological organization. However, challenges such as image noise, photobleaching of fluorophores, and low tolerability of biological samples to high light doses remain, restricting temporal resolutions and experiment durations. Reduced laser doses enable longer measurements at the cost of lower resolution and increased noise, which hinders accurate downstream analyses. Here we train a denoising diffusion probabilistic model (DDPM) to predict high-resolution images by conditioning the model on low-resolution information. Additionally, the probabilistic aspect of the DDPM allows for repeated generation of images that tend to further increase the signal-to-noise ratio. We show that our model achieves a performance that is better or similar to the previously best-performing methods, across four highly diverse datasets. Importantly, while any of the previous methods show competitive performance for some, but not all datasets, our method consistently achieves high performance across all four data sets, suggesting high generalizability. Our code and datasets are available at https://github.com/kaschube-lab/ddpm_highres_microscopy, https://doi.org/10.5281/zenodo.14178277, https://doi.org/10.5281/zenodo.14215837.

Deep neural networks have led to significant advancements in microscopy image generation and analysis. In single-molecule localization based super-resolution microscopy, neural networks are capable of predicting fluorophore positions from high-density emitter data, thus reducing acquisition time, and increasing imaging throughput. However, neural network-based solutions in localization microscopy require intensive human intervention and computation expertise to address the compromise between model performance and its generalization. For example, researchers manually tune parameters to generate training images that are similar to their experimental data; thus, for every change in the experimental conditions, a new training set should be manually tuned, and a new model should be trained. Here, we introduce AutoDS and AutoDS3D, two software programs for reconstruction of single-molecule super-resolution microscopy data that are based on Deep-STORM and DeepSTORM3D, that significantly reduce human intervention from the analysis process by automatically extracting the experimental parameters from the imaging raw data. In the 2D case, AutoDS selects the optimal model for the analysis out of a set of pre-trained models, hence, completely removing user supervision from the process. In the 3D case, we improve the computation efficiency of DeepSTORM3D and integrate the lengthy workflow into a graphic user interface that enables image reconstruction with a single click. Ultimately, we demonstrate superior performance of both pipelines compared to Deep-STORM and DeepSTORM3D for single-molecule imaging data of complex biological samples, while significantly reducing the manual labor and computation time.

Resolving the nanoscale organization of viral and host proteins is important to understanding virion assembly and infectivity. Here, we present a robust framework for multiplexed optical 3D super-resolution microscopy of human immunodeficiency virus type 1 (HIV-1) particles using minimal fluorescence photon flux (MINFLUX) nanoscopy and DNA point accumulation for imaging in nanoscale topography (DNA-PAINT), achieving isotropic localization precision below 10 nm for five target proteins. First, we assessed linkage errors introduced by different labeling strategies by employing the HIV-1 matrix layer as a reference structure. We then extended the approach to display five viral and host proteins and mapped the spatial organization of tetraspanin proteins CD9 and CD81 in single virus-like particles. For accurate visualization and quantitation of multicolor 3D MINFLUX imaging data, we developed the analysis workflow and software matFLUX. The approach presented here enables high-resolution spatial mapping of protein components within individual virus particles and is generally applicable to the study of nanoscale architectures in 3D.

Rhodamine dyes conjugated to targeting ligands can yield exceptionally bright fluorescent probes for live-cell imaging. However, the limited permeability of such rhodamine derivatives restricts their broader applications, particularly in vivo. Here, we present Rho-tag and SiR-tag, engineered protein tags derived from bacterial multidrug-resistant proteins that bind unsubstituted (silicon) rhodamines with nanomolar affinity. Unsubstituted (silicon) rhodamines readily cross membranes and enable rapid, reversible, and fluorogenic labeling of the tags in mammalian cells within seconds. The labeling of Rho-tag and SiR-tag is compatible with various super-resolution imaging methods and allows their use alongside self-labeling tags, such as HaloTag7 and SNAP-tag. The high affinity and specificity of both tags, combined with the permeability and outstanding spectroscopic properties of rhodamines, make them particularly attractive for in vivo bioimaging, as demonstrated by efficient fluorescent labeling inC. elegansembryos and zebrafish larvae.

Small subcellular organelles orchestrate key cellular functions. How biomolecules are spatially organized within these assemblies is poorly understood. Here, we report an automated super-resolution imaging and analysis workflow that integrates confocal microscopy, morphological object screening, targeted 3D super-resolution STED microscopy and quantitative image analysis. Using this smart microscopy workflow, we target the 3D organization of NEAT1, an architectural RNA that constitutes the structural backbone of paraspeckles, a membraneless nuclear organelle. Using site-specific labeling, morphological sorting and particle averaging, we reconstruct the morphological space of paraspeckles along their development cycle from over 10,000 individual particles. Applying spherical harmonics analysis, we report so-far unknown heterotypes of NEAT1 RNA organization. By integrating multi-positional labeling, we determine the coarse conformation of NEAT1 within the organelle and show that the 3' end forms a loop-like structure at the surface of the paraspeckle. Our study reveals key structural features of paraspeckle structure and growth, as well as the molecular organization of its scaffolding RNA.

Stimulated emission depletion (STED) microscopy enables super-resolution imaging of complex biological samples in 3D, in large volumes, and live. However, molecular quantification with STED has remained underexplored. Here, we present a straightforward approach for quantitative STED that enables molecule counting. For this purpose, we designed DNA-fluorophore labels that enable signal amplification and allow for reliable intensity-based quantitative imaging. We demonstrate accurate molecule counting on DNA origami. Furthermore, we visualized and quantified EGF receptor monomers and dimers in cells. In summary, we introduce a robust, fast, and easy-to-implement tool for quantitative STED microscopy with single-protein resolution. ### Competing Interest Statement The authors have declared no competing interest.

Glycosylation is a crucial biochemical modification of proteins and other biomolecules in cells that generates an exceptional structural and functional diversity. Aberrant glycosylation is implicated in numerous diseases, including neurodegenerative disorders and cancer. While glycan organization at the cell surface is studied extensively, the nanoscale spatial arrangement of glycans inside cells and within organelles remains largely unexplored. Here, super-resolution imaging with fluorophore-labeled lectins, combined with a dedicated multiplexing strategy (Glyco-STORM), was used in thin neuronal tissue sections and permeabilized cells to systematically map glycosylation across multiple organelles and cellular compartments. Lectin markers were identified that enabled the visualization of nanodomains within the endoplasmic reticulum, subdomains along the Golgi axes, diverse glycosylation states in the endolysosomal system, and a polarized lysosomal clathrin coat. At neuronal synapses, mature glycans were found to demarcate the synaptic cleft and subsynaptic tubules adjacent to the postsynaptic density. In summary, Glyco-STORM establishes one of the first comprehensive nanoscale maps of intracellular glycosylation and provides entry points into health- and disease-related studies.

The activation of transmembrane receptors through the binding of external ligands initiates information transfer across the cell membrane. Understanding these processes requires observations in living cells. Given the heterogeneity and lack of synchronization of such events, single‐molecule experiments are required to resolve distinct sub‐populations. Here, single‐molecule FRET microscopy and single‐particle tracking are combined to track the ligand‐induced dimerization and activation of the MET receptor tyrosine kinase in the plasma membrane of living cells. First, using fluorophore‐labeled variants of the MET ligand internalin B (InlB), the lifetime of a ligand‐activated dimeric (MET:InlB) 2 receptor complex is determined to be ≈1 s. Next, diffusion coefficients of monomeric and dimeric MET:InlB complexes are extracted from single‐molecule FRET trajectories, revealing an ≈1.6‐fold slower diffusion of the dimeric receptor compared to the monomeric receptor, accompanied by spatially confined motion. The combination of single‐molecule FRET and single‐particle tracking provides essential biophysical parameters of membrane receptor activation in living cells.

Glycosylation is a crucial biochemical modification of proteins and other biomolecules in cells that generates an exceptional structural and functional diversity. Aberrant glycosylation is implicated in numerous diseases, including neurodegenerative disorders and cancer. Despite its significance, methodological constraints to date have limited the exploration of the nanometer scale spatial arrangement of glycans across entire cells. We developed Glyco-STORM, a super-resolution imaging approach that generates nano-structural maps of cellular glycosylation. Glyco-STORM employs fluorophore-labeled lectins and multiplexed single-molecule super-resolution microscopy, in combination with nanoscale spatial pattern analysis. For example, Glyco-STORM unraveled nanodomains within the endoplasmic reticulum, subdomains along the Golgi axes, and a polarized lysosomal clathrin coat. At synaptic contact sites, mature glycans delineate the synaptic cleft and subsynaptic tubules adjacent to the postsynaptic density. In summary, Glyco-STORM elucidates the spatial arrangement of glycosylation sites from subcellular to molecular levels, revealing the previously obscured glycosylation landscape at nanoscale and establishing a 'spatial glycosylation code' that provides a unique perspective on cellular organization distinct from traditional protein-centric views. ### Competing Interest Statement The authors have declared no competing interest.

Deep neural networks have led to significant advancements in microscopy image generation and analysis. In single-molecule localization-based super-resolution microscopy, neural networks are capable of predicting fluorophore positions from high-density emitter data, thus reducing acquisition time, and increasing imaging throughput. However, neural network-based solutions in localization microscopy require intensive human intervention and often compromise between model performance and its generalization. Researchers have to manually tune simulated training data parameters to resemble their experimental data; thus, for every change in the experimental conditions, a new training set should be manually tuned, and a new model should be trained. Here, we introduce AutoDS and AutoDS3D, two software programs for super-resolution reconstruction of single-molecule localization microscopy data that are based on Deep-STORM and DeepSTORM3D. Our methods significantly reduce human intervention from the analysis process by automatically extracting the experimental parameters from the imaging raw data. In the 2D case, AutoDS selects the optimal model for the analysis out of a set of pre-trained models, hence, completely removing user supervision from the process. In the 3D case, we improve the computation efficiency of DeepSTORM3D and integrate the lengthy workflow into a graphic user interface that enables image reconstruction with a single click. Ultimately, we demonstrate comparable or superior performance of both methods compared to Deep-STORM, DeepSTORM3D, and other state-of-the-art methods, while significantly reducing the manual labor and computation time.