A Single-Cell Atlas of the Tumor and Immune Ecosystem of Human Breast Cancer.

Breast cancer is a heterogeneous disease. Tumor cells and associated healthy cells form ecosystems that determine disease progression and response to therapy. To characterize features of breast cancer ecosystems and their associations with clinical data, we analyzed 144 human breast tumor and 50 non-tumor tissue samples using mass cytometry. The expression of 73 proteins in 26 million cells was evaluated using tumor and immune cell-centric antibody panels. Tumors displayed individuality in tumor cell composition, including phenotypic abnormalities and phenotype dominance. Relationship analyses between tumor and immune cells revealed characteristics of ecosystems related to immunosuppression and poor prognosis. High frequencies of PD-L1+ tumor-associated macrophages and exhausted T cells were found in high-grade ER+ and ER- tumors. This large-scale, single-cell atlas deepens our understanding of breast tumor ecosystems and suggests that ecosystem-based patient classification will facilitate identification of individuals for precision medicine approaches targeting the tumor and its immunoenvironment.

Matching single cells across modalities with contrastive learning and optimal transport.

Understanding the interactions between the biomolecules that govern cellular behaviors remains an emergent question in biology. Recent advances in single-cell technologies have enabled the simultaneous quantification of multiple biomolecules in the same cell, opening new avenues for understanding cellular complexity and heterogeneity. Still, the resulting multimodal single-cell datasets present unique challenges arising from the high dimensionality and multiple sources of acquisition noise. Computational methods able to match cells across different modalities offer an appealing alternative towards this goal. In this work, we propose MatchCLOT, a novel method for modality matching inspired by recent promising developments in contrastive learning and optimal transport. MatchCLOT uses contrastive learning to learn a common representation between two modalities and applies entropic optimal transport as an approximate maximum weight bipartite matching algorithm. Our model obtains state-of-the-art performance on two curated benchmarking datasets and an independent test dataset, improving the top scoring method by 26.1% while preserving the underlying biological structure of the multimodal data. Importantly, MatchCLOT offers high gains in computational time and memory that, in contrast to existing methods, allows it to scale well with the number of cells. As single-cell datasets become increasingly large, MatchCLOT offers an accurate and efficient solution to the problem of modality matching.

ATHENA: analysis of tumor heterogeneity from spatial omics measurements.

SUMMARY: Tumor heterogeneity has emerged as a fundamental property of most human cancers, with broad implications for diagnosis and treatment. Recently, spatial omics have enabled spatial tumor profiling, however computational resources that exploit the measurements to quantify tumor heterogeneity in a spatially aware manner are largely missing. We present ATHENA (Analysis of Tumor HEterogeNeity from spAtial omics measurements), a computational framework that facilitates the visualization, processing and analysis of tumor heterogeneity from spatial omics measurements. ATHENA uses graph representations of tumors and bundles together a large collection of established and novel heterogeneity scores that quantify different aspects of the complexity of tumor ecosystems. AVAILABILITY AND IMPLEMENTATION: ATHENA is available as a Python package under an open-source license at: https://github.com/AI4SCR/ATHENA. Detailed documentation and step-by-step tutorials with example datasets are also available at: https://ai4scr.github.io/ATHENA/. The data presented in this article are publicly available on Figshare at https://figshare.com/articles/dataset/zurich_pkl/19617642/2. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Inference of protein kinetics by stochastic modeling and simulation of fluorescence recovery after photobleaching experiments.

MOTIVATION: Fluorescence recovery after photobleaching (FRAP) is a functional live cell imaging technique that permits the exploration of protein dynamics in living cells. To extract kinetic parameters from FRAP data, a number of analytical models have been developed. Simplifications are inherent in these models, which may lead to inexhaustive or inaccurate exploitation of the experimental data. An appealing alternative is offered by the simulation of biological processes in realistic environments at a particle level. However, inference of kinetic parameters using simulation-based models is still limited. RESULTS: We introduce and demonstrate a new method for the inference of kinetic parameter values from FRAP data. A small number of in silico FRAP experiments is used to construct a mapping from FRAP recovery curves to the parameters of the underlying protein kinetics. Parameter estimates from experimental data can then be computed by applying the mapping to the observed recovery curves. A bootstrap process is used to investigate identifiability of the physical parameters and determine confidence regions for their estimates. Our method circumvents the computational burden of seeking the best-fitting parameters via iterative simulation. After validation on synthetic data, the method is applied to the analysis of the nuclear proteins Cdt1, PCNA and GFPnls. Parameter estimation results from several experimental samples are in accordance with previous findings, but also allow us to discuss identifiability issues as well as cell-to-cell variability of the protein kinetics. IMPLEMENTATION: All methods were implemented in MATLAB R2011b. Monte Carlo simulations were run on the HPC cluster Brutus of ETH Zurich. CONTACT: lygeros@control.ee.ethz.ch or lygerou@med.upatras.gr SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Multi-step loading of human minichromosome maintenance proteins in live human cells.

Once-per-cell cycle replication is regulated through the assembly onto chromatin of multisubunit protein complexes that license DNA for a further round of replication. Licensing consists of the loading of the hexameric MCM2-7 complex onto chromatin during G1 phase and is dependent on the licensing factor Cdt1. In vitro experiments have suggested a two-step binding mode for minichromosome maintenance (MCM) proteins, with transient initial interactions converted to stable chromatin loading. Here, we assess MCM loading in live human cells using an in vivo licensing assay on the basis of fluorescence recovery after photobleaching of GFP-tagged MCM protein subunits through the cell cycle. We show that, in telophase, MCM2 and MCM4 maintain transient interactions with chromatin, exhibiting kinetics similar to Cdt1. These are converted to stable interactions from early G1 phase. The immobile fraction of MCM2 and MCM4 increases during G1 phase, suggestive of reiterative licensing. In late G1 phase, a large fraction of MCM proteins are loaded onto chromatin, with maximal licensing observed just prior to S phase onset. Fluorescence loss in photobleaching experiments show subnuclear concentrations of MCM-chromatin interactions that differ as G1 phase progresses and do not colocalize with sites of DNA synthesis in S phase.

ScLinear predicts protein abundance at single-cell resolution.

Single-cell multi-omics have transformed biomedical research and present exciting machine learning opportunities. We present scLinear, a linear regression-based approach that predicts single-cell protein abundance based on RNA expression. ScLinear is vastly more efficient than state-of-the-art methodologies, without compromising its accuracy. ScLinear is interpretable and accurately generalizes in unseen single-cell and spatial transcriptomics data. Importantly, we offer a critical view in using complex algorithms ignoring simpler, faster, and more efficient approaches.

easyFRAP: an interactive, easy-to-use tool for qualitative and quantitative analysis of FRAP data.

SUMMARY: We present easyFRAP, a versatile tool that assists quantitative and qualitative analysis of fluorescence recovery after photobleaching (FRAP) data. The user can handle simultaneously large data sets of raw data, visualize fluorescence recovery curves, exclude low quality data, perform data normalization, extract quantitative parameters, perform batch analysis and save the resulting data and figures for further use. Our tool is implemented as a single-screen Graphical User Interface (GUI) and is highly interactive, as it permits parameterization and visual data quality assessment at various points during the analysis. AVAILABILITY: easyFRAP is free software, available under the General Public License (GPL). Executable and source files, supplementary material and sample data sets can be downloaded at: ccl.med.upatras.gr/easyfrap.html.

scQUEST: Quantifying tumor ecosystem heterogeneity from mass or flow cytometry data.

With mass and flow cytometry, millions of single-cell profiles with dozens of parameters can be measured to comprehensively characterize complex tumor ecosystems. Here, we present scQUEST, an open-source Python library for cell type identification and quantification of tumor ecosystem heterogeneity in patient cohorts. We provide a step-by-step protocol on the application of scQUEST on our previously generated human breast cancer single-cell atlas using mass cytometry and discuss how it can be adapted and extended for other datasets and analyses. For complete details on the use and execution of this protocol, please refer to Wagner et al. (2019).

Quantification of tumor heterogeneity: from data acquisition to metric generation.

Tumors are unique and complex ecosystems, in which heterogeneous cell subpopulations with variable molecular profiles, aggressiveness, and proliferation potential coexist and interact. Understanding how heterogeneity influences tumor progression has important clinical implications for improving diagnosis, prognosis, and treatment response prediction. Several recent innovations in data acquisition methods and computational metrics have enabled the quantification of spatiotemporal heterogeneity across different scales of tumor organization. Here, we summarize the most promising efforts from a common experimental and computational perspective, discussing their advantages, shortcomings, and challenges. With personalized medicine entering a new era of unprecedented opportunities, our vision is that of future workflows integrating across modalities, scales, and dimensions to capture intricate aspects of the tumor ecosystem and to open new avenues for improved patient care.

Spatial protein heterogeneity analysis in frozen tissues to evaluate tumor heterogeneity.

A new workflow for protein-based tumor heterogeneity probing in tissues is here presented. Tumor heterogeneity is believed to be key for therapy failure and differences in prognosis in cancer patients. Comprehending tumor heterogeneity, especially at the protein level, is critical for tracking tumor evolution, and showing the presence of different phenotypical variants and their location with respect to tissue architecture. Although a variety of techniques is available for quantifying protein expression, the heterogeneity observed in the tissue is rarely addressed. The proposed method is validated in breast cancer fresh-frozen tissues derived from five patients. Protein expression is quantified on the tissue regions of interest (ROI) with a resolution of up to 100 µm in diameter. High heterogeneity values across the analyzed patients in proteins such as cytokeratin 7, ß-actin and epidermal growth factor receptor (EGFR) using a Shannon entropy analysis are observed. Additionally, ROIs are clustered according to their expression levels, showing their location in the tissue section, and highlighting that similar phenotypical variants are not always located in neighboring regions. Interestingly, a patient with a phenotype related to increased aggressiveness of the tumor presents a unique protein expression pattern. In summary, a workflow for the localized extraction and protein analysis of regions of interest from frozen tissues, enabling the evaluation of tumor heterogeneity at the protein level is presented.

In silico analysis of DNA re-replication across a complete genome reveals cell-to-cell heterogeneity and genome plasticity.

DNA replication is a complex and remarkably robust process: despite its inherent uncertainty, manifested through stochastic replication timing at a single-cell level, multiple control mechanisms ensure its accurate and timely completion across a population. Disruptions in these mechanisms lead to DNA re-replication, closely connected to genomic instability and oncogenesis. Here, we present a stochastic hybrid model of DNA re-replication that accurately portrays the interplay between discrete dynamics, continuous dynamics and uncertainty. Using experimental data on the fission yeast genome, model simulations show how different regions respond to re-replication and permit insight into the key mechanisms affecting re-replication dynamics. Simulated and experimental population-level profiles exhibit a good correlation along the genome, robust to model parameters, validating our approach. At a single-cell level, copy numbers of individual loci are affected by intrinsic properties of each locus, in cis effects from adjoining loci and in trans effects from distant loci. In silico analysis and single-cell imaging reveal that cell-to-cell heterogeneity is inherent in re-replication and can lead to genome plasticity and a plethora of genotypic variations.

CellCycleTRACER accounts for cell cycle and volume in mass cytometry data.

Recent studies have shown that cell cycle and cell volume are confounding factors when studying biological phenomena in single cells. Here we present a combined experimental and computational method, CellCycleTRACER, to account for these factors in mass cytometry data. CellCycleTRACER is applied to mass cytometry data collected on three different cell types during a TNFα stimulation time-course. CellCycleTRACER reveals signaling relationships and cell heterogeneity that were otherwise masked.

Analysis of Protein Kinetics Using Fluorescence Recovery After Photobleaching (FRAP).

Fluorescence recovery after photobleaching (FRAP) is a cutting-edge live-cell functional imaging technique that enables the exploration of protein dynamics in individual cells and thus permits the elucidation of protein mobility, function, and interactions at a single-cell level. During a typical FRAP experiment, fluorescent molecules in a defined region of interest within the cell are bleached by a short and powerful laser pulse, while the recovery of the fluorescence in the region is monitored over time by time-lapse microscopy. FRAP experimental setup and image acquisition involve a number of steps that need to be carefully executed to avoid technical artifacts. Equally important is the subsequent computational analysis of FRAP raw data, to derive quantitative information on protein diffusion and binding parameters. Here we present an integrated in vivo and in silico protocol for the analysis of protein kinetics using FRAP. We focus on the most commonly encountered challenges and technical or computational pitfalls and their troubleshooting so that valid and robust insight into protein dynamics within living cells is gained.

IPP Complex Reinforces Adhesion by Relaying Tension-Dependent Signals to Inhibit Integrin Turnover.

Cytoskeleton-mediated forces regulate the assembly and function of integrin adhesions; however, the underlying mechanisms remain unclear. The tripartite IPP complex, comprising ILK, Parvin, and PINCH, mediates the integrin-actin link at Drosophila embryo muscle attachment sites (MASs). Here, we demonstrate a developmentally earlier function for the IPP complex: to reinforce integrin-extracellular matrix (ECM) adhesion in response to tension. In IPP-complex mutants, the integrin-ECM linkage at MASs breaks in response to intense muscle contractility. Mechanistically, the IPP complex is required to relay force-elicited signals that decelerate integrin turnover at the plasma membrane so that the integrin immobile fraction is adequate to withstand tension. Epistasis analysis shows that alleviation of muscle contractility, downregulation of endocytosis, and enhanced integrin binding to the ECM are sufficient to restore integrin-ECM adhesion and maintain integrin-adhesome organization in IPP-complex mutants. Our findings reveal a role for the IPP complex as an essential mechanosensitive regulatory switch of integrin turnover in vivo.

CK1δ restrains lipin-1 induction, lipid droplet formation and cell proliferation under hypoxia by reducing HIF-1α/ARNT complex formation.

Proliferation of cells under hypoxia is facilitated by metabolic adaptation, mediated by the transcriptional activator Hypoxia Inducible Factor-1 (HIF-1). HIF-1α, the inducible subunit of HIF-1 is regulated by oxygen as well as by oxygen-independent mechanisms involving phosphorylation. We have previously shown that CK1δ phosphorylates HIF-1α in its N-terminus and reduces its affinity for its heterodimerization partner ARNT. To investigate the importance of this mechanism for cell proliferation under hypoxia, we visually monitored HIF-1α interactions within the cell nucleus using the in situ proximity ligation assay (PLA) and fluorescence recovery after photobleaching (FRAP). Both methods show that CK1δ-dependent modification of HIF-1α impairs the formation of a chromatin binding HIF-1 complex. This is confirmed by analyzing expression of lipin-1, a direct target of HIF-1 that mediates hypoxic neutral lipid accumulation. Inhibition of CK1δ increases lipid droplet formation and proliferation of both cancer and normal cells specifically under hypoxia and in an HIF-1α- and lipin-1-dependent manner. These data reveal a novel role for CK1δ in regulating lipid metabolism and, through it, cell adaptation to low oxygen conditions.