Symposium Schedule ESCP 2026
preliminary schedule
Mass spectrometry (MS)-based single-cell proteomics (SCP) enables proteome-wide analysis at single-cell resolution, offering insights into cellular heterogeneity, biological processes, and disease mechanisms. However, conventional nanoLC-MS workflows are constrained by long gradients and low throughput, limiting their application to small cohorts. Here, we present a multicolumn nanoLC–MS platform that achieves 2.88-minute separation windows with 100% duty cycles at nanoflow rates, enabling the analysis of up to 500 single cells per day with minimal additional hardware. The system provides stable peptide separation, negligible carryover, and robust retention-time reproducibility across 1,200 consecutive injections. Over the course of this study, we successfully analyzed more than 4,000 samples at nearly 500 samples per day (SPD) throughput. We will provide performance metrics and use cases for the system.
[1] Department of Proteomics and Signal Transduction, Max Planck Institute of Biochemistry, Martinsried, Germany.
[2] ETH Zürich, Department for Biosystems Science and Engineering (D-BSSE), Basel, Switzerland
[3] Friedrich Miescher Institute for Biomedical Research, Fabrikstrasse 24, 4056 Basel, Switzerland
[4] Department of Medical Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden
The relationship between a cell’s transcriptome and proteome sits at the core of molecular biology, yet the central dogma’s expectation that protein abundance follows mRNA expression has proven far more complex. Decades of large-scale comparisons have shown that transcript abundance explains only a fraction of the variance in protein levels – the remainder shaped by translation rate, mRNA stability, and protein turnover. The transcriptome and proteome are therefore complementary layers of cellular state that cannot be reliably inferred from one another. Resolving how they relate has been limited by a fundamental gap: while single-cell RNA sequencing is routine, single-cell proteomics is still emerging. Existing strategies to pair the two modalities either average the proteome over many cells, rely on a handful of antibody-based epitopes, or measure each layer in different cells – leaving same-cell, deep, scalable measurement of both layers out of reach. Here, we introduce a workflow that simultaneously quantifies the proteome and transcriptome of the same single cell. By physically separating RNA from protein rather than splitting the lysate, we combine an adapted Smart-seq3 RNA-seq protocol with an oil-overlay single-cell proteomics method and acquisition on the Orbitrap Astral Zoom. Across thousands of cells, we routinely quantify up to 4,000 protein groups and 15,000 transcripts per cell — matching plate-based scRNA-seq in transcriptomic depth while delivering a deep matched single-cell proteome. We apply the workflow along a gradient of biological complexity. In HeLa cells as a steady-state reference, we provide a direct single-cell view of the central dogma, establishing a core transcript–protein correlatome and showing that protein abundance is markedly less variable across cells than transcript abundance. Sampling mouse embryonic stem cells along the naive-to-primed pluripotency trajectory, we find these steady-state relationships partially break down: we resolve discrete cell states, capture the timing of transcription factor turnover at scale, and uncover distinct modes of RNA–protein co-regulation that partition across functional gene classes. Together, these results cast the transcriptome as what a cell can do and the proteome as what a cell is doing – a distinction now directly measurable in the same cell across systems of increasing biological complexity.
[1]Department of Proteomics and Signal Transduction, Max Planck Institute of Biochemistry, Martinsried, Germany.
[2]Department of Cellular Biochemistry, Max Planck Institute of Biochemistry, Martinsried, Germany.
[3]Lipid cell biology lab, UPDANGELO, Institute of Bioengineering (IBI), Ecole Polytechnique Fédérale de Lausanne (EPFL), 1015 Lausanne, Switzerland
[4]Center for Neuropathology and Prion Research, Faculty of Medicine, LMU Munich, Germany
[5]Biomolecular Screening Core Facility and Technology Platform, EPFL, 1015 Lausanne, Switzerland.
[6]Weill Cornell Medicine Qatar, Education City, Qatar Foundation, Doha, Qatar
[7]Department of Neurology, Weill Cornell Medicine, New York, AL, United States of America
Spatial proteomics now resolves proteomes in situ, and Deep Visual Proteomics (DVP) reaches single-cell resolution by coupling AI-guided image analysis, automated laser microdissection, and low-input mass spectrometry. Many structures of biological interest, however, form below the single cell and are averaged away by bulk and single-cell methods alike, while indirect alternatives (proximity labelling, affinity pulldowns, insoluble-fraction biochemistry) discard spatial context. We developed Aggregate DVP, a spatial proteomics workflow that extends DVP to the subcellular scale and profiles a morphologically defined structure against a matched host fraction microdissected from the same sample. Custom Cellpose and CellProfiler models segment the target and its surrounding compartment from high-content or whole-slide images; contours are excised by automated laser microdissection at ~1,000 shapes per hour into 384-well plates, prepared by miniaturized low-input handling, and analysed on an Orbitrap Astral. From as few as 100 microdissected shapes the workflow quantifies up to 7,290 proteins, the matched within-sample design controlling individual, fixation, and batch variation in a single measurement. We further built the first annotation-free DVP workflow for clinical chromogenic immunohistochemistry, color-deconvolving DAB and counterstain channels with dvp-io and spatialdata to threshold targets without manual annotation. The same microdissected material also yields the phosphoproteome (nanoPhos): from the four spatial fractions we quantified 4,565 localized phosphosites and, by normalizing each site to its parent protein, separated phosphorylation changes from protein recruitment. Pure aggregates carried a distinct, strongly hyperphosphorylated phosphoproteome, and the diagnostic α-synuclein pS129 mark rose ~1,000-fold from cytosol into the inclusion even though the inclusion lacked the writing kinases, showing it captures the phosphorylated form rather than generating it. Applied across HEK293 cells, iPSC-derived dopaminergic neurons, and FFPE patient brain, Aggregate DVP reads α-synuclein inclusions directly and resolved which proteins are structure-resident rather than merely abundant. Needing only a structure that can be stained and recognized by shape, it generalizes from single cells to any spatially resolvable subcellular compartment and modification layer.
[1] Institute of Microanalytical Systems, Department of Chemistry, Zhejiang University, Hangzhou, 310058, China
[2] ZJU-Hangzhou Global Scientific and Technological Innovation Center, Single Cell Proteome Research Center, Hangzhou 311200, China
Proteome analysis at the single-cell level poses significant technical challenges. As early as 2004, the presenter’s group participated in the preliminary attempts to achieve single-cell proteomic (SCP) analysis using microfluidic chip and mass spectrometry techniques, but these efforts were unsuccessful. In 2013, we developed the sequential operation droplet array (SODA) technique [1] for performing automated picoliter to nanoliter-scale droplet manipulation, analysis and screening, as well as single-cell reverse transcription qPCR assay [2]. Since 2014, we collaborated with Dr. Catherine Wong at the National Center for Protein Research in Shanghai to conduct single-cell proteomics using the SODA and label-free mass spectrometry techniques. We developed an oil-air-droplet (OAD) chip-based system to perform complex multi-step sample pretreatment for single cells in in-situ nanoliter-scale droplet microreactors, including single cell capture, cell lysis, cell protein reduction, alkylation and digestion, as well as the injection of the droplet sample in a LC-MS/MS system. In 2018, we reported the results of single-cell proteomic analyses of single HeLa cells and mouse oocytes for the first time [3], which demonstrated that the microfluidic in-situ droplet microreactors offers significant advantages in minimizing sample loss during the multi-step pretreatment process of single-cell samples. Based on the SODA technique, we developed an automated single-cell sorting, capture and pretreatment instrument, capable of automatically completing the microscopic imaging, identification, capture, and droplet generation of the target cells, as well as multi-step nanoliter-scale sample pretreatment operation, forming a pick-up single-cell proteome analysis (PiSPA) workflow for deep SCP analysis (up to 3000 proteins per tumor cell) [4]. Recently, we further optimized and streamlined the PiSPA workflow [5] and adapted to higher-sensitivity mass spectrometers, enabling an identification level up to 6,000–7,000 proteins/cell. We applied the above techniques and systems in SCP analysis of immune cell-tumor cell interactions [6], tumor cell resistance [7], and circulating tumor cells (CTCs) in clinical samples. By coupling the PiSPA approach with other single-cell omics methods, we also achieved transcriptomic-proteomic [8], proteomic-metabolic [9,10], and transcriptomic-proteomic-metabolic analyses of single cell individuals.
REFERENCES 1. Y. Zhu, Y-X. Zhang, L-F. Cai, Q. Fang, Sequential operation droplet array: an automated microfluidic platform for picoliter-scale liquid handling, analysis, and screening, Anal. Chem. 2013, 85, 6723. 2. Y. Zhu, Y-X. Zhang, W-W. Liu, Y. Ma, Q. Fang, B. Yao, Printing 2-dimentional droplet array for single-cell reverse transcription quantitative PCR assay with a microfluidic robot, Sci. Rep., 2015, 5, 9551. 3. Z-Y. Li, M. Huang, X-K. Wang, Y. Zhu, J-S. Li, C. C. L. Wong, Q. Fang, Nanoliter-scale oil-air-droplet chip-based single cell proteomic analysis, Anal. Chem., 2018, 90, 5430. 4. Y. Wang, Z-Y. Guan, S-W. Shi, Y-R. Jiang, J. Zhang, Y. Yang, Q. Wu, J. Wu, J-B. Chen, W-X. Ying, Q-Q. Xu, Q-X. Fan, H-F. Wang, L. Zhou, J. Fang J-Z. Pan, Q. Fang, Pick-up single-cell proteomic analysis for quantifying up to 3000 proteins in a mammalian cell. Nat. Commun., 2024, 15, 1279. 5. J. Wang, Y. Huang, F. Lu, Q. Xu, Z. Yang, Y. Jiang, S. Shi, J. Pan, Y. Yang, Q. Fang, Benchmarking informatics workflows for data-independent acquisition single-cell proteomics. Nat. Commun., 2025, 16, 10276. 6. Q-Q. Xu, Y-R. Jiang, J-B. Chen, J. Wu, Y-X. Chen, Q-X. Fan, H-F. Wang, Y. Yang, J-Z. Pan, Q. Fang, Single cell-pair proteomics for decoding immune-cancer cell interactions. Adv. Sci., 2025, 2414769. 7. Q-X. Fan, Y-R. Jiang, J-B. Chen, Y-X. Chen, Q-Q. Xu, J. Wu, H-F. Wang, Y. Yang, J-Z. Pan, Q. Fang, Integrated single-cell proteomic and morphometric analysis reveals heterogeneous drug-resistant subpopulations. Anal. Chem., 2025, 97, 15028. 8. Y-R. Jiang, L. Zhu, L-R. Cao, Q. Wu, J-B. Chen, Y. Wang, J. Wu, T-Y. Zhang, Z-L. Wang, Z-Y. Guan, Q-Q. Xu, Q-X. Fan, S-W. Shi, H-F. Wang, J-Z. Pan, X-D. Fu, Y. Wang, Q. Fang, Simultaneous deep transcriptome and proteome profiling in a single mouse oocyte, Cell Rep., 2023, 42, 113455. 9. J. Wu, Q-Q. Xu, Y-R. Jiang, J-B. Chen, W-X. Ying, Q-X. Fan, H-F. Wang, Y. Wang, S-W. Shi, J-Z. Pan, Q. Fang, One-shot single-cell proteome and metabolome analysis strategy for the same single cell, Anal. Chem., 2024, 96, 5499. 10. J. Wu, F-H. Lu, L-Y. Tao, Y-L. Wang, Y. Huang, Y. Wang, Y. Yang, J-Z. Pan, X. Liang, Q. Fang, Simultaneous in-depth single-cell proteomic and metabolomic analysis, Anal. Chem. 2026, 98, 13495.
[1] Spatial Proteomics Group, Max-Delbrück-Center for Molecular Medicine in the Helmholtz Association (MDC), Berlin, Germany
[2] German Cancer Consortium (DKTK), Partner Site Berlin, Heidelberg, Germany
[3] Institute of Pathology, Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Berlin, Germany
[4] Institute of Pathology, Ludwig Maximilians University Hospital Munich, Munich, Germany
[5] German Cancer Consortium (DKTK), Partner Site Munich, and German Cancer Research Center (DKFZ), Heidelberg, Germany
[6] BIFOLD – Berlin Institute for the Foundations of Learning and Data, Berlin, Germany.
We developed an open-source protocol combining multiplex immunofluorescence (mIF) with Deep Visual Proteomics (DVP) to analyze formalin-fixed paraffin-embedded (FFPE) human tissue samples. This method, applicable to diverse research questions, integrates iterative antibody-based staining using open-source reagents and imaging systems with image analysis and laser microdissection, followed by ultra-sensitive mass spectrometry. Whole-slide imaging enables the analysis of large tissue sections or tissue microarrays, supporting cohort-scale applications. We applied this workflow using a 29-marker panel to profile the tumor microenvironment in a 30 mm head and neck cancer tissue section. Spatial analysis identified distinct cellular neighborhoods, including three cancer-associated neighborhoods. Following laser microdissection of selected microregions and mass spectrometry, we found that the three spatially defined cancer neighborhoods harbor fundamentally different molecular programs and therapeutic target profiles, indicating that each neighborhood may require a distinct treatment strategy. We further scaled the approach down to single-cell measurements and observed that molecular properties were preserved after seven cycles of cyclic immunofluorescence. To demonstrate cohort scalability, we applied the workflow to a lung cancer cohort comprising 400 tissue microarray cores, validating previously identified cellular neighborhoods and performing deep proteomic profiling to characterize their molecular features and identify candidate biomarkers for patient stratification. Together, this open-source workflow enables scalable, spatially resolved deep proteomic analysis across tissues and diseases.
[1] School of Biomedical Engineering, University of British Columbia, Vancouver, British Columbia, Canada
[2] BC Children’s Hospital Research Institute, Vancouver, British Columbia, Canada
[3] Vancouver Prostate Centre, Vancouver, British Columbia, Canada
Spatially resolved proteomics of defined cell populations in live cultures remains technically challenging due to limitations such as shallow proteome coverage and sample loss during collection. Furthermore, most spatial technologies do not allow sequential capture of regions across timepoints from a single culture. To address these gaps, we developed MaSSProt, a spatial proteomics workflow that integrates microfluidic probe (MFP)-based, near single-cell-resolution spatial sampling with sensitive dia-PASEF acquisition to enable deep, spatiotemporally resolved proteomics on microscale regions of live cultures.
The MFP localises lysis reagents on open surfaces with micrometer precision, enabling spatially restricted sampling from live cell monolayers and tissue sections without enclosing cells in closed channels. Footprint areas corresponding to ~10–300 cells are achieved by tuning injection-to-aspiration flow ratios, microchannel dimensions and probe-to-surface distances. Scaled-down MFP head designs (25 µm injection channels) bring footprints to the <10-cell scale, with a clear path toward single-cell resolution. A DDM-passivated, low-dead-volume collection path limits loss to <7% for peptides and <14% for proteins relative to direct-lysis controls – competitive with leading low-input proteomics platforms.
MaSSProt was validated in a PC-3 prostate cancer scratch assay, sampling migratory wound-edge and non-migratory bulk populations. Approximately 4,000 proteins were identified from footprints of as few as 12 cells (CV <20%, n=5 technical replicates). Differential abundance analysis revealed significant enrichment of pro-metastatic proteins (UTS2, CCBE1, TIMP1) and RNA-binding proteins (PTBP2, RTRAF) in migratory cells, consistent with known invasion biology. This demonstrates the platform’s ability to resolve biologically meaningful proteomic differences between spatially distinct subpopulations.
At the conference, we will present quantitative benchmarks of MaSSProt’s spatial resolution, depth and robustness, as well as ongoing progress toward single-cell sensitivity, positioning MaSSProt as a versatile tool for spatially and temporally resolved proteomics across cancer biology and beyond.
Peripheral blood mononuclear cells (PBMCs) provide minimally invasive and accessible biofluid-derived sources for monitoring systemic immune alterations. However, their functional proteomic remodeling during cancer progression remains unclear. In this study, we established a bedside-to-bench and label-free mass spectrometry-based single-cell proteomics (SCP) workflow for the proteomic analysis of individual cells. We applied this workflow to cryopreserved PBMCs from two healthy donors (controls) and patients with head and neck squamous cell carcinoma (HNSCC) at distinct stages of local metastasis progression. The workflow generated proteomic profiles for 619 circulating immune cells, quantifying up to 1,600 protein groups per individual cell. Unsupervised clustering, followed by correlation of cluster marker proteins with a publicly available bulk-sorted proteomics dataset, enabled the annotation of major immune populations, including monocytes, CD4⁺ and CD8⁺ T cells, NK cells, and B cells. Furthermore, an exploratory trajectory analysis revealed progressive immune remodeling from controls toward non-metastatic and metastatic disease, accompanied by increased monocyte representation and a relative reduction in lymphocytes. A monocyte state detected in patients with nodal metastasis exhibited enrichment of interferon-stimulated proteins, HLA molecules, myeloid immunosuppressive markers, and cellular stress-response proteins. In parallel, circulating lymphocytes, including CD8⁺ T cells, displayed progressive disruption of processes associated with activation, cytotoxicity and degranulation. Collectively, these findings demonstrate that SCP of cryopreserved PBMCs can reveal clinically relevant systemic immune changes associated with HNSCC progression, highlighting its potential for minimally invasive monitoring of immune cell states.
[1] German Cancer Research Center (DKFZ), Heidelberg, Germany;
[2] Heidelberg University, Medical Faculty, Heidelberg, Germany;
[3] EMBL, Heidelberg, Germany
Single-cell proteomics (SCP) is moving from cell-type profiling toward reconstruction of therapeutic adaptation, but resistance models remain limited by proteome depth, throughput and functional validation. Here, we establish a multi-layer SCP framework to resolve KRAS inhibitor resistance from deep label-free proteome states, protein turnover and genetic dependencies.
To induce drug resistance, MiaPaCa-2 cells were exposed to Sotorasib (AMG-510) over 20 weeks of dose escalation. We generated a longitudinal label-free SCP dataset across ~3,000 single cells collected at 8 time points, quantifying >6,500 protein groups per cell and ~9,000 protein groups across the dataset. This depth enabled measurement of signaling, metabolism, proteostasis, chromatin regulation and epithelial-mesenchymal state without multiplexed labels. Static protein abundance analysis was extended to protein turnover analysis via pSILAC-SCP with 16 h heavy-amino-acid labeling for single-cell protein renewal (~1,200 single cells), bulk AHA-pulsed nascent proteomics (~10,000 proteins) for early translational adaptation, Ubiquitin-Trap pulldown for ubiquitin remodeling, and genome-wide CRISPR screening for dependencies. Local entropy, pseudotime, hdWGCNA, GSVA, TF inference and XGBoost converted deep SCP measurements into resistance trajectories and predictive cell-state features.
Label-free SCP revealed a progressive decrease in local proteomic entropy, consistent with selection of pre-existing persister-like cells. Pseudotime and hdWGCNA resolved an ordered transition toward a mesenchymal, stress-adapted state with TF rewiring. AHA proteomics identified metabolic and chromatin remodeling as early resistance programs. CRISPR screening identified the ubiquitin-proteasome system as a major dependency, including DCAF4, WAC and TRIM8. pSILAC-SCP showed reduced global protein renewal despite maintained proteasome activity, supporting selective ubiquitin-mediated remodeling rather than proteome shutdown. Resistant cells progressively lost ferroptosis defense proteins, including GPX4 and SLC7A11, exposing ferroptosis as a collateral vulnerability. Sensitivity was rescued by Ferrostatin-1, Liproxstatin-1 and Deferiprone and reproduced in KRAS G12D models treated with MRTX1133.
Together, ultra-deep label-free SCP, turnover proteomics, and perturbation screening transform drug resistance from an endpoint phenotype into a quantitative single-cell trajectory and reveal actionable vulnerabilities.
This potential is illustrated through established and emerging computational approaches. MLMarker1, a tissue-of-origin classifier trained on public bulk proteomics data, can be directly applied to single-cell measurements. In parallel, a foundation model trained on public proteomics and transcriptomics datasets, illustrates how the broad integration of publicly available data may support single-cell proteome interpretation beyond the scope of task-specific classifiers such as MLMarker.
However, the transition from models trained on bulk data to models developed specifically for single-cell proteomics depends on the availability of reliable and well-annotated single-cell datasets in the public domain. As the volume of these data increases, harmonized quality-control criteria and standardized annotation become essential for their reuse, integration, and incorporation into AI model development. Public data and single-cell proteomics therefore form a reciprocal relationship: existing datasets provide context for interpreting individual cells, whereas high-quality single-cell datasets expand and refine the resources available for model training.
This relationship is examined in the context of the HUPO Single-Cell Proteomics Initiative, including an ongoing interlaboratory benchmarking effort to establish shared standards for single-cell proteomics data quality and annotation2. Ultimately, the ability of AI models to interpret single-cell proteomics and of single-cell proteomics to improve those models, will depend on maintaining FAIR data practices throughout this cycle.
[1] Claeys, T.; van Puyenbroeck, S.; Gevaert, K.; Martens, L. MLMarker: A Machine Learning Framework for Tissue Inference and Biomarker Discovery. Genome Biol 2026, 27 (1), 207. https://doi.org/10.1186/s13059-026-04125-8.
[2] Puyenbroeck, S. van; Claeys, T.; Seth, A.; Rijal, J.-B.; Keller, C.; Lin, L.; Mayer, R.; Matzinger, M.; Han, I.; Fernandez, P. A.; Petrosius, V.; Boyle, B.; Rivera, K.; Tourniaire, G.; Rosenberger, F. A.; Martens, L.; Carr, S. A.; Dong, Z.; Vegvari, A.; Carapito, C.; Kelly, R.; Mechtler, K.; Budnik, B.; Schoof, E. M.; Ctortecka, C. Defining Quality Control Standards for Single-Cell Proteomics by Inter-Laboratory Benchmarking. bioRxiv July 14, 2026, p 2026.07.13.738155. https://doi.org/10.64898/2026.07.13.738155.
As single cell proteomics transitions from a specialized technology to a widely adopted approach, the data volume created by our community has the potential to broadly characterize cellular activity and responses to both genetic and environmental perturbations. As a repository of single cell data, an Atlas is designed for intense data re-use – both by machine learning experts and also by non-ML users. We anticipate that a single-cell proteomics Atlas could be leveraged into GPT-like AI frameworks to build a foundational knowledge of cellular biology, allowing researchers to interactively explore potential experimental designs and likely results. With the recent improvements in proteome coverage and throughput, single cell proteomics has the potential to begin creating Atlas-scale datasets, quantifying thousands of proteins from hundreds of thousands of cells. However, creating this community resource requires significant planning to achieve the desired use-cases. Here, we discuss how data can be generated, aggregated and analyzed to create an Atlas capable of modeling cell biology. We focus on the unique aspects of proteomics data that can add to the single cell ecosystem.
[1] Proteomics Center of Excellence, Chemistry of Life Processes Institute, Northwestern University, Evanston, IL
[2] Department of Surgery, Division of Organ Transplant, Feinberg School of Medicine, Northwestern University, Chicago, IL
[3] Comprehensive Transplant Center, Feinberg School of Medicine, Northwestern University, Chicago, IL
[4] University of California Riverside, Riverside, CA
[5] Department of Chemistry, Northwestern University, Evanston, IL
[6] Division of Nephrology, Department of Medicine, College of Medicine, University of Illinois Chicago, Chicago, IL
[7] Department of Microbiology-Immunology and Pediatrics, Feinberg School of Medicine, Northwestern University, Chicago, IL
[8] Department of Molecular Biosciences, Northwestern University, Evanston, IL
Single-cell proteomics has achieved significant coverage gains, yet most methods cannot resolve protein isoforms, post-translational modifications (PTMs), and processing variants, the proteoforms that drive cellular function. This limitation is critical for the objective classification of donor organ quality that is imprinted in proteoform-level changes in signaling, protein cleavage, and PTMs that are not yet fully understood. We combine individual ion mass spectrometry (I2MS) with nanospray Desorption Electrospray Ionization for rapid, proteoform-level screening from intact tissues to thousands of dissociated single cells, bridging high-throughput population screening with molecular depth to resolve cellular heterogeneity. Single-cell arrays generated using cellenONE enable regular cell deposition, increasing single-cell Proteoform Imaging Mass Spectrometry (scPiMS) throughput 4-fold to ~7,000 cells/day. Image-based isolation supplements proteoforms with cell-specific characteristics including size and morphology while removing debris critical for heterogeneous primary cells and frozen dissociated tissue. A custom nanoDESI source continuously extracts soluble proteins (~7-seconds per cell), generating proteoform landscapes that are queried via intact mass tag search against experimental-specific and pan-human databases. To meet kidney transplant demand, the tissue pool has expanded from gold-standard living to deceased donors, yet 30% are discarded due to lack of reliable biomarkers. Tissue-level PiMS of fresh kidney biopsies distinguished living donors by metabolic signatures, while deceased and discarded tissues showed elevated stress markers that correlate with clinical outcome of the recipient (i.e., abnormal biopsy or rejection). scPiMS resolved donor-specific heterogeneity and proteoform variability depending on the cellular microenvironment within the tissue. Pre- and post-transplant peripheral blood mononuclear cell analysis validated oxidative stress and immune activation in recipients of poor-quality kidneys, consistently showing elevated candidate biomarkers: ACTG1, acetylated CRYAB, PARK7, and S100A4. This workflow bridges rapid population-wide screening with sensitive proteoform analysis, providing point-of-care biomarker candidates for objective kidney assessment that could rescue viable organs while protecting patients from compromised grafts.
[1] Section for Medical Biotechnology, Department of Bioengineering, Technical University of Denmark, Denmark
Single-cell proteomics by mass spectrometry (scp-MS) has emerged as a powerful complement to transcriptomics, enabling direct quantification of protein expression and eventually post-translational modifications. In scp-MS every processing step is a potential source of irreversible sensitivity loss, and therefore minimising processing volumes and optimising reagent compositions are critical to maximising proteome coverage from individual cells. Here we present scTAP (single-cell TFE-Aided Proteomics), a scalable, label-free, one-pot workflow processing up to 374 single cells in parallel on a hydrophobic glass slide using the cellenONE platform. The master mix, comprising trypsin, buffer, and trifluoroethanol (TFE), is dispensed with high stability and reproducibility, in contrast to detergent-based formulations that frequently cause dispensing inconsistencies. Benchmarked against our established 384-well plate-based workflow, scTAP at 15 nL improved precursor and protein group identifications in single HEK293 cells by 20% and 17%, respectively. At the subcellular level, scTAP identified 40% more nuclear, 34% more membrane-associated, and 27% more cytoplasmic proteins on average. Assessed across four chromatographic speeds using WISH-DIA acquisition without carrier-based boosting in DIA-NN 2.5, scTAP yielded mean protein group identifications of 3579, 3810, 3456, and 3063 at 120, 145, 160, and 200 samples-per-day, respectively. We further conducted a systematic investigation of the effects of processing volume, trypsin concentration, and total trypsin amount on sensitivity. Using 100 pg of HeLa digest standard incubated on the hydrophobic glass slide, reducing the processing volume from 1000 to 15 nL alone yielded an approximately 30% gain in identifications at both the precursor and protein group level. Importantly, excess trypsin used in single-cell proteomics workflows markedly enhances peptide recovery also in higher processing volume and we conclude that the enzyme itself and its autolysis fragments act as a carrier that competes with cell-derived peptides for adsorption to hydrophobic contact surfaces. Taken together, our data demonstrate that in single-cell processing, volume miniaturisation and trypsin concentration are both critical determinants of digestion efficiency and peptide adsorption but that total trypsin amount represents an equally important and previously underappreciated parameter for maximising recovery.
$ These authors contributed equally
[1]Centre for Proteome Research, Department of Biochemistry, Cell & Systems Biology, Institute of Systems, Molecular & Integrative Biology, Biosciences Building, Crown Street, University of Liverpool, Liverpool, L69 7ZB, UK
[2]Centre for Cell Imaging, Liverpool Shared Research Facilities, Biosciences Building, Crown Street, University of Liverpool, Liverpool, L69 7ZB, UK
[3]The Pirbright Institute, Guildford, Surrey, UK
The reliability of single-cell proteomics (SCP) is intrinsically linked to the fidelity of cell isolation; however, the identification of co-isolated cells (doublets or multiplets) remains a persistent challenge for single-cell (sc-) omics. While single-cell transcriptomics has established probabilistic frameworks for doublet detection, these methods are ill-suited for the sparser throughput of SCP datasets. This work presents scpImaging, a novel computational pipeline that repurposes the latent microscopy data routinely generated during cellenONE-based sample preparation to provide deterministic quality control (QC) and high-content phenotypic profiling. We demonstrate that standard proteomic quality metrics for SCP (e.g., peptide count, signal intensity) paradoxically favour retaining doublets. In contrast, scpImaging applies machine-learning-based cell segmentation (Cellpose-SAM) to identify and exclude these artefacts with high precision. Furthermore, the framework integrates morphological metrics generated using the open-source CellProfiler, with proteomic abundance data, enabling joint analysis that links cell shape and texture to molecular cell state. Provided as an open-source R package, scpImaging offers a scalable, automated solution for enhancing data integrity and adding a phenotypic dimension to SCP experiments without increasing experimental cost or complexity.
Ann-Christin Branscheid [1,4], Artür Manukyan [1], Luiz Gustavo Teixeira Alves [1], Maximilian Gerwien [1,2], Robert Kerridge [1,2], Markus Landthaler [1,2], Jana Wolf [1,5], Matthias Selbach [1,6]
[1] Max Delbrück Center for Molecular Medicine,13125 Berlin, Germany
[2] Humboldt-Universität zu Berlin, 10117 Berlin, Germany
[3] German Cancer Consortium (DKTK), partner site Berlin, a partnership between DKFZ and
Charité University Medicine Berlin, Germany
[4] Technische Universität Berlin, 10623 Berlin, Germany
[5] Freie Universität Berlin, 14195 Berlin, Germany
[6] Charité-Universitätsmedizin Berlin, 10117 Berlin, Germany
Even cells of the same type growing in the same environment show cell-to-cell differences in protein abundance, a phenomenon known as gene expression noise. This variability can be decomposed into intrinsic components, reflecting molecular randomness, and extrinsic components, arising from differences in cellular state. While gene expression noise has been studied genome-wide in microbes, its global organization remains largely unknown in mammalian cells. Here, we develop a spike-in-based stable isotope single-cell proteomics approach that enables robust quantification of protein-level gene expression noise across thousands of human proteins. We find that protein noise scales inversely with abundance until reaching a plateau, consistent with an extrinsic noise floor and conserved scaling principles observed in bacteria and yeast. Cell cycle stage and cell size contribute substantially to protein variability but do not fully account for the observed heterogeneity. Gene-specific features such as mRNA and protein half-lives and translation efficiency show only weak associations with protein noise, and variability at the mRNA level is a weak predictor of protein variability. Instead, protein noise is largely extrinsic, with coordinated variation across proteins encoding biologically organized cellular states. Consistently, coordinated proteome programs predict intercellular differences in proteome dynamics, linking protein variability to cellular function. Together, these results provide a proteome-wide view of gene expression noise in mammalian cells, establishing that protein-level variability encodes structured and functionally relevant differences in cellular state.
Alejandro J Brenes*, Rupert L Mayer*, Agata N Makar*, Patricia Coelho, Pranvera Sadiku, Sarah R Walmsley, Manuel Matzinger*, Karl Metchler* and Alex von Kriegsheim*
Introduction:
Mass spectrometry-based single cell proteomics (SCP) is rapidly emerging as a powerful approach for biological research, with applications extending beyond in-vitro cancer cell lines. Recent advances make it possible to apply SCP to ex-vivo human cells from tissues such as the brain and pancreas, as well as to technically challenging immune populations such as neutrophils. However, these analyses remain more challenging and typically result in reduced proteomic coverage. To support the development of robust workflows for SCP data acquisition and analysis, we systematically evaluated multiple DIA search engines, search engine settings, the inclusion of high-load library samples in single-cell search spaces, the impact of contaminants, and the quantitative properties of identified proteins. These comparisons were performed across two major instrumentation platforms, Orbitrap Astral and timsTOF SCP, and across A549, RKO cells and neutrophils, three cell types differing in size and protein content.
Methods:
Single cells along with different high-load samples were sorted into 384 well plates, using the CellenOne for RKO cells and A549 cells, and FACS for neutrophils. For neutrophils and RKO cells the high-load libraries were composed of 5, 10, 25 and 50 cells. For A549 cells it was 5, 11, 20, and 40 cells. The single cells and high-load libraries were analysed on an Orbitrap Astral with FAIMS at 50 samples per day (SPD) throughput and on a timsTOF SCP (tSCP) at 40 SPD throughput. The data were then searched with both Spectronaut 20.3 and DIA-NN 2.3 & 2.5.
Results:
Our work shows that stringent software parameters have limited reductions in proteome coverage and should be widely used. We also show instrument specific effects, where high-load libraries with stringent parameters show no improved identification potential on the Astral on DIA-NN 2.3 or Spectronaut, but show some potential with DIA-NN 2.5. Higher-load libraries also showed contradictory effects on tSCP data with significant reductions in protein identifications when >10 cells are used with Spectronaut 20.3 and significant improvements when >10 cells are used with DIA-NN 2.3 & 2.5. We show that as Astral has significantly higher sensitivity than the tSCP, especially pronounced for small cells, but at the same time show the tSCP has better quant for small cells. We also show how skin contaminants that are shared in proteome of human cells can skew quant and show how to identify these for single cell experiments.
Falk Schlaudraff, Leica Microsystems
Transferring regions of interest from upstream imaging into a laser microdissection (LMD) system is a critical step in spatial and single-cell workflows and it directly shapes throughput and reproducibility. This talk maps the full spectrum of ROI transfer strategies, from fully manual navigation to API-driven automation. Step by step, we introduce orientation via the specimen overview, loading and aligning original images including multiplexed data, stage-position memory for rapid navigation, and automated alignment. Attendees leave with a practical framework for choosing the right approach for their sample and can experience each method hands-on at our booth.
Introduction Complex cellular systems such as cancers, like Multiple Myeloma, present a wide cellular heterogeneity. While proteolysis-targeting chimeras (PROTACs) have revolutionized targeted protein degradation, molecular glues such as Cereblon Modulating Drugs (CELMoDs) are emerging as a nextgeneration therapeutic strategy enabling the selective degradation of disease-causing proteins such as GSPT1. Bulk proteomics involves analyzing peptides from thousands of cells, providing an averaged image of the expressed proteome. Conversely, Single-Cell Proteomics (SCP), while remaining analytically challenging, opens promising perspectives and application fields with an unprecedented level of granularity. Here, we developed a dual bulk and SCP approach to elucidate CeLMODs drugs effects and dose responses. Methodology Single myeloma cells were sorted and digested using a CellenONE cell sorter (Cellenion®, Lyon). LC and DIA-PASEF MS acquisition methods were optimized on short Aurora (IonOptiks) C18-RP (75µmx50mm, 1.7µm) columns using an EVOSEP coupled to a TimsTOF Ultra 2 (Bruker Daltonics) operated at 80 to 120 samples per day. Data treatment was performed with DIA-NN (v 1.9.2 and v2.2.0) and Spectronaut (v18). Results High proteome coverage was achieved consistently exceeding 8000 proteins in bulk and 3000 proteins per single cell. The downregulation of GSPT1 by the CeLMODs has been finely characterized at both bulk and single cell levels, demonstrating its specific degradation upon treatment conditions. Equally important, a high selectivity towards CeLMODs treatments has been demonstrated as no off-targets were detected at both bulk and single cell levels. Following these promising results and the unprecedented sensitivity enabled by our robust SCP workflow, drug cellular internalization levels are currently investigated using reporter peptides to decipher and quantify the drug-dose response at the single-cell level. Conclusion Our high performing single cell proteomics workflow demonstrates the high selectivity and specificity of molecular glue drugs towards targeted degradation of GSPT1 and enables quantifying drug-dose responses at the single cell level
requires separate registration, find all details here: https://scverse.org/sc-proteomics2026/