key: cord-0780245-mvs8bmk3
authors: Zhang, Ting; Warden, Antony R.; Li, Yiyang; Ding, Xianting
title: Progress and applications of mass cytometry in sketching immune landscapes
date: 2020-10-14
journal: Clin Transl Med
DOI: 10.1002/ctm2.206
sha: 0a2c81d66d7fad2dafb4cce20e2189590a7cf4a6
doc_id: 780245
cord_uid: mvs8bmk3

Recently emerged mass cytometry (cytometry by time‐of‐flight [CyTOF]) technology permits the identification and quantification of inherently diverse cellular systems, and the simultaneous measurement of functional attributes at the single‐cell resolution. By virtue of its multiplex ability with limited need for compensation, CyTOF has led a critical role in immunological research fields. Here, we present an overview of CyTOF, including the introduction of CyTOF principle and advantages that make it a standalone tool in deciphering immune mysteries. We then discuss the functional assays, introduce the bioinformatics to interpret the data yield via CyTOF, and depict the emerging clinical and research applications of CyTOF technology in sketching immune landscape in a wide variety of diseases.

The immune plasticity is expressed in the remarkably heterogeneous milieus, including circulation and different tissues, during the homeostasis or during the various disease states. 1 An array of adaptive responses, including training, priming, exhaustion, and tolerance, has been distinguished in adaptive and innate immunocytes. Precise classification of cell subpopulations with overlapping phenotypes and simultaneous interrogation of the phenotypic and functional properties of single cells in basal state and after their exposure to exogenous stimuli are of great significance.

Fluorescence-based cytometry techniques have dominated for decades in immune system studies at the single-cell resolution. However, the number of parameters available and complex compensation processes for spectral overlap limit its applications. The developments of fluorescence-based cytometry have been propelled by recent technological advances. Notably, cytometry by timeof-flight (CyTOF), also known as mass cytometry, is a novel combination of flow cytometry and mass spectrometry, which excels in multiparametric single-cell analysis. CyTOF, through the utility of rare-earth metal-tagged antibodies, inductively coupled plasma ionization, and time-of-flight detector, allows simultaneous characterization of up to 50 parameters per cell. Therefore, CyTOF provides significant possibilities toward the identification of disease attributes in cell populations, the orchestrated interplay amongst protean immune cells, as well as the F I G U R E 1 Overview of CyTOF analysis molecular immunological signatures that underlie clinical manifestations.

Here, we briefly introduce the principle, advantages, and current limitations of CyTOF. We then describe current progresses in functional assay developments leveraging CyTOF and the developed bioinformatic techniques and pipelines in CyTOF data analysis. Next, we discuss recent applications of CyTOF in basic and clinical research of immune profiling in the fields of cancer, immunotherapy, autoimmune diseases, infective diseases, cardiovascular diseases, transplantation, and neuroscience. Finally, we comment on the perspectives of CyTOF's rapid entry into research and clinical settings.

The technical aspects of the CyTOF system have been described in detail by Bendall et al, 2 as depicted in Figure 1 . In essence, instead of fluorophores, antibodies are labeled with stable heavy metal isotopes, mainly of the lanthanide series, which are naturally absent in biological systems. Each isotope's readout can be correlated with a specific antibody probe, representing the antigen levels within individual cells. Sample preparation and staining for CyTOF are similar to fluorescence flow cytometry, except that cells are typically fixed prior to analysis. Cells are then introduced into the CyTOF analyzer and nebulized into droplets, which are vaporized, atomized, ionized, and then accelerated toward a mass spectrometer via electrical potential. The current time-of-flight (TOF) detector of commercialized CyTOF is tuned for a mass window of approximately Da, and the remaining atoms are filtered by a quadrupole to increase sensitivity and minimize unwanted signals. The filtered ion clouds are analyzed with TOF detector.

rDespite being generally considered an alteration based on fluorescence flow cytometry, mass cytometry differs from fluorescence cytometry by several aspects including dimensionality, sensitivity, cell throughput, and so on ( Table 1) . The clever switch from fluorescence readout to mass spectrometry detection of heavy metal-tagged antibody probes has given CyTOF several unique advantages. First, the increment in dimensionality represents a groundbreaking breakthrough. In fluorescence flow cytometry, the number of detection parameters is limited by the overlapping of fluorescence signals. This problem is especially prominent when performing antigen-specific T cells screening, chromatin modification profiling, and simultaneous detection of RNAs and proteins. Presently, up to 45 parameters can be simultaneously and reliably quantified with the CyTOF technique. In theory, CyTOF is capable of isotropic discrimination of more than 100 elemental masses, which suggests even higher multiplex capability in future applications. This enables CyTOF profiling and screening in a comprehensive and system-wide manner. Second, given the precise isotopic discrimination of mass spectrometer, the channels in TOF have little cross talk and the need for compensation is limited, 3 which to a great extent circumvents the time-consuming and laborious compensation issues in conventional fluorescence flow cytometry. Next, the detection of these metal isotopes, which are absent or extremely rare in biological systems, would fundamentally resolve autofluorescence issues and background noises problems. In addition, with unique palladium-based barcode labels, 4 palladium-tagged β2-microglobulin-based barcode label combinations, 5 or ratiometry-based CD45 barcode labels, 6 individual samples from a large cohort can be pooled together for further analysis to eliminate batch effect. The commercial palladium-based barcode labels adopt a 6-choose-3 barcoding scheme and allow 20 samples barcoded together, whereas the palladium-tagged β2-microglobulin-based barcode label combinations adopt a 5-choose-2 scheme and enable 10 samples together. It may take multiple barcode sets in large-scale experiments. The ratiometrybased CD45 barcode labels based on three metals and three ratios can barcode 19 samples and as the ratio levels and mass tags numbers increase, the maximum potential barcoding capability would expand exponentially. This would make barcoding and consistent staining for large-cohort study feasible, but at the expense of detection channel numbers. The rate of cell acquiring and analysis of CyTOF instrument is ∼500 cells/s, which permits the analysis of millions of cells in 1 h. More importantly, CyTOF enables in-depth analysis, including exploration of signaling pathway alterations on archival samples, such as curated formalin-fixed paraffin-embedded tissues (FFPEs). 7 With the advent of imaging mass cytometry (IMC), even spatial information and cell interactions in curated FFPEs can be obtained. 8 Last, coupled with advanced computational tools and well-established pipelines for high-dimensional data analysis, CyTOF facilitates the visualization of immunocytes and their networks.

A large body of work has been exerted on the establishment and recapitulation of conventional fluorescence flow cytometry assays. For instance, the carrier strategy that significantly reduces required sample amount in fluorescence cytometry has been successfully adapted in CyTOF technique to enable the analysis of rare and precious clinical samples. 9 Limitations and constraints of the CyTOF technique, however, still exist, as compared with flow cytometry (Table 1) . First, cells cannot be recovered for further functional analysis using CyTOF technique, as cells are fixed and ultimately ionized. Second, the sensitivity of CyTOF is 10-fold less than that of flow cytometry, limited by the chelating polymer used in CyTOF to attach metal reporter ions. 2 Further, a sizable portion of cells is lost during the sample treatment and instrument processing, resulting in less than 50% of cells available for analyses by CyTOF. Flow cytometry, however, can measure over 95% of cells in a sample. 10 In addition, as mass cytometry is mostly dependent on antibodies, careful antibody panel design and validation are required to ensure accurate and specific detection of all targets. Experienced and extensive labor are needed, especially when processing larger panels. Last, both the instrument and the reagents are expensive. The low accessibility of CyTOF instrument and high expense of the assay currently limit its wide utilization.

Ongoing improvements to CyTOF approaches continue to open new opportunities for implementing various functional assays to tackle the complexities of cellular immunology. Here, we provide an overview of CyTOFbased functional assays.

CyTOF can multiplex up to 45 cellular markers with limited need for spectral overlap compensation, opening up a post-fluorescence era of cytometry well suited for deep phenotyping of cells in complex systems. Genetically similar or even identical cells that play distinct roles in disease pathogenesis could be distinguished with CyTOF, hosting important implications for personalized medicine. Being able to quantitatively probe nearly any feature (Figure 2A) , CyTOF can characterize various cell types 11 including rare cells across the immune cell continuum. 12

Intracellular cytokine profiling offers novel perspectives on immune activation states. The higher multiparameter capacity in CyTOF renders it an attractive instrument in intracellular cytokine staining assays ( Figure 2B ). Vendrame et al utilized CyTOF to evaluate the effects of cytokines on the natural killer (NK) cells and revealed that interleukin (IL)-12/IL-15/IL-18 stimulation induced dramatically increased interferon-γ (IFN-γ) expression in NK cells. 13 Doyle et al characterized plasmacytoid dendritic cells (pDCs) in the liver and peripheral blood of patients with hepatitis C virus (HCV) infection and demonstrated that liver pDCs were polyfunctional and capable of producing abundant IFN-γ and other immune modulators during chronic HCV infection. 14 With ever-growing reports utilizing CyTOF to determine intracellular cytokines, we believe that CyTOF will likely act as an indispensable tool in immune cell function studies.

Cellular circuits sense environmental stimuli and accordingly attune the signaling network to enable key decisions regarding cellular response. Using metal-chelated antibodies targeting phosphorylated proteins, CyTOF enables interrogation of signal propagation within individual cells. Studies have demonstrated the utility of CyTOF to assess cell signaling states in kinetic or time point ( Figure 2C ). 15, 16 Shinko et al have provided an optimized protocol of phosphorylated signaling proteins staining for clinical blood samples. 17 Combined with lineage markers and stimulations or inhibitions, the signaling states and cellular responsiveness are comprehensively evaluated. 16, 18, 19 

Cell size and volume are fundamental characteristics that impact the structure and functions of any given cell type.

Fluorescence flow cytometry uses the metrics of light scatter properties (forward and side scatter intensity) to determine cell volume and size. Stern et al established wheat germ agglutinin-based and Osmium tetroxide-based plasma membrane staining to gauge the size of mammalian cells. 20 Rapsomaniki et al identified the ruthe-nium complex ASCQ_Ru as robust marker for quantifying cell volume. 21 These stains ( Figure 2D ) substitute for light scatter properties evaluated in fluorescence flow cytometry and expand the range of parameters measured via CyTOF.

Discriminating cellular viability is critical in biological sample analysis, especially in functional studies such as intracellular signaling or drug responses, as it enables the removal of nonviable cells. Platinum-based covalent viability regent, 22 Cisplatin, was developed to label cells for CyTOF discrimination of live/dead ratios, on the basis that Cisplatin preferentially labels nonviable cells ( Figure 2E ).

Cell cycle alterations are important aspects in tumor progression, developmental biology, and immune modulation. Behbehani et al developed a novel CyTOF approach to delineate cell cycle stages, based on iododeoxyuridine to mark cells in the S phase, together with antibodies against phosphorylated retinoblastoma, cyclin B1, cyclin A, and phosphorylated histone H3 to define G0, G1, G2, and M phases ( Figure 2F ). 23 Protocols of cell cycle identification were well developed and detailed. 24 Utilizing this cell cycle identification method, researches have revealed cell cycle differences that mediated chemotherapy sensitivities of acute myeloid leukemia 25 and erythropoiesis impairment in telomerase knockout mice. 26 

Good et al provided a dye dilution protocol for cell proliferation tracing across time and states, which could be widely applied in directing studies of cellular differentiation. 27 A dilution assay of carboxyfluorescein succinimidyl ester (CFSE) was adapted and metal-labeled anti-fluorescein isothiocyanate antibody was employed to track CFSE signal changes ( Figure 2G ). Dividing cells that pass half CFSE signals to individual offspring cell act as a proxy for cell division counting. The proliferation tracing method helps uncouple the phenotypic and functional transitions in tandem with the sequence of cellular differentiation. Good et al used 23 markers and the proliferation tracing approach to track single naïve human T cells. A map of cell variations during naïve T cell expansion was built and revealed that undivided cells represent a large portion of phenotype diversity.

The ratio of drug-bound receptor to total receptor on individual cells is known as receptor occupancy. It represents a response biomarker for treatment of therapeutic monoclonal antibodies. Punet-Ortiz et al 28 employed a multiparametric quantitative flow cytometry to monitor CD49d receptor occupancy on peripheral blood mononuclear cells (PBMCs) of 19 multiple sclerosis patients receiving natalizumab therapy for a 6-month follow up. They used this index of CD49d receptor occupancy to determine a safe, personalized regimen, and proposed an optimized CyTOFbased receptor occupancy measurement. 29 CyTOF enables the measurement of receptor occupancy in conjunction with more markers within wide varieties of cell types in a more reliable and reproducible manner. It would be valuable in drug pharmacodynamics or immune suppression fields, replacing fluorescence flow cytometry. Huse et al emphasized that receptor occupancy determination by CyTOF expands the clinical cytometry toolbox and introduced three basic formats of CyTOF-based receptor occupancy assays ( Figure 2H ). 30

To better develop vaccines and targeted therapies for autoimmune diseases and cancer, cognate-specific antigens and binding affinity of antigen-binding T cells need to be comprehensively characterized. Studies have emphasized the adaptability of mass cytometry to profile T-cell responses that are antigen specific and screen T-cell reactivity against various major histocompatibility complex (MHC)-class-restricted epitopes. A modified peptide-MHC class I tetramer was conjugated to metal chelating polymers, allowing multidimensional analyses by CyTOF ( Figure 2I ). 31 Coupled with a combinatorial staining approach, Newell et al simultaneously tested hundreds of neoantigens in cancer. 32-34 Antigen-specific CD8 + T cells were screened with highly multiplexed combinatorial tetramer staining in a cohort of 14 nonsmall cell lung carcinoma patients before and after atezolizumab treatment, 33 including eight responder patients and six with progressive disease status. The results indicated enrichment of low-differentiated effector neoantigen-specific CD8 + T cells in responders to atezolizumab treatment. Although lower affinity between T-cell receptor and peptide-MHC II and lower frequencies of antigen-specific CD4 + T cells render peptide-MHC II tetramers screening by CyTOF challenging, peptide-MHC II tetramer staining in PBMCs and tissue-derived cell suspensions was achieved after optimization. 35,36

Histone modifications are fundamental to proteomic epigenetic regulation. Recently published studies employed CyTOF to investigate chromatin modification. 37,38 Cheung et al proposed epigenetic landscape profiling using CyTOF (EpiTOF), a highly multiplexed form to analyze histone modifications and identify dysregulations associated with immune-mediated diseases at the single-cell resolution ( Figure 2J ). 38 Strategies of chromatin marks manipulation were established, such as ectopic overexpression or CRISPR-or RNAi-mediated chromatin-modifying enzymes depletion, to validate and select antibodies for EpiTOF. Two antibodies that are able to recognize total histone proteins were integrated into EpiTOF for variation control of antibody background, nuclear epitope accessibility, and histone expression. The investigators applied EpiTOF to examine 24 healthy cytomegalovirusseronegative subjects, focusing on 22 major immune subsets and examined the cellular levels of four histone variants in eight classes of histone modifications. The results indicated that chromatin variations increased with age, which are largely driven by nonheritable factors, and cell-type-specific chromatin mask profiles predicted identity of immune cells.

To enable the codetection of RNA and protein signatures at the single-cell resolution, Frei et al developed the proximity ligation assay for RNA (PLAYR) method, which is a proximity ligation assay by CyTOF for highly multiplexed transcript quantification ( Figure 2K ). 39 PLAYR adopts pairs of DNA oligonucleotide probes that include a region to hybridize target transcripts and another region as a template to bind and circularize two additional oligonucleotides, which are ligated and amplified through rolling circle amplification. The amplified products of each probe pair are detected using oligonucleotides labeled with mass tags. Another novel CyTOF-based mRNA transcript and protein codetection method, termed metal in situ hybridization (MISH), was developed via combination of CyTOF and RNAscope@ platform. 40 The technique enables signal amplification through hybridizing RNA-specific probes and binding with amplified sequence-targeting metal-labeled probes. Both PLAYR and MISH are compatible with routine immunostaining, and simultaneous transcripts quantification number is only limited by the number of reporters that can be conjugated to the oligonucleotides. Notably, in both methods, the measurement of RNAs will occupy the mass channels, which means the total number of simultaneously codetected RNAs and protein targets depends on the number of mass channels (up to 45 at present). Frei et al 39 gave an example of simultaneous multiplexed profiling of protein and transcripts. They used PLAYR to monitor the induction of eight cytokine transcripts and 18 protein epitopes in PBMCs after stimulation with lipopolysaccharide and revealed correlations between the functional capacity of each cell and its protein marker expression.

Geisen et al used CyTOF to image tissue samples to acquire spatial proteomics. 8 The proposed technique, IMC, uses laser ablation with a resolution of 1 μm to generate plumes of tissue sections that are aerosolized, atomized, ionized, and carried to the mass spectrometry detector by an inert gas stream ( Figure 2L ). IMC is considered as a landmark development, as it allows for simultaneous profiling of up to 50 parameters in tandem with cell interaction and spatial information at subcellular resolution. Since its introduction, IMC is being rapidly adopted for various applications. 41-47 Damond et al employed IMC to investigate 1581 islets from four nondiabetic, four patients of onset type 1 diabetes, and four patients of long-duration type 1 diabetes. 44 They depicted the progression of human type 1 diabetes and revealed an alteration of β cells phenotype prior to its destruction.

Prior to analysis, a stringent beads-based data normalization is required 48 to correct variations in instrument performance caused by drift and build-up of cellular debris. For samples run individually or in multiple barcode sets, a batch adjustment technique is required. Schuyler et al 49 flexibly adopted several standard normalization methods, such as per channel quantile normalization, and recommended combining statistical testing results of conditions within batches via Fisher's method. Shaham et al 50 presented a deep learning approach, based on distributionmatching residual networks, to effectively attenuate batch effects. In both methods, technical replicates, as anchors or references, are included in each run, allowing direct estimation and adjustment of batch effect. CyTOF achieves analysis of high cell throughput (up to millions of cells) and high dimensions, which pose challenges toward its analysis. Traditional user-guided manual gating on bivariate plots has been proven as subjective, cumbersome, and inefficient. Novel computational analyses that are able to visualize and explain CyTOF-generated data have been developed ( Table 2) . Each of these algorithmic tools is designed with distinct goals and advantages. Of these tools, dimensionality reduction is utilized to organize complex data into recognizable patterns and thus provides an overview of the data. Principal component analysis (PCA) 51 is a commonly employed dimensionality reduction technique for visualizing the multidimensional data. 52 With linear transformation, PCA can reduce dimensionality; however, biological systems can contain nonlinear relationships. Therefore, nonlinear dimensionality reduction techniques that avoid representation of overcrowding have gradually gained attraction. Nonlinear dimensionality reduction techniques employed in CyTOF data analysis include Isomap, 53 Diffusion Map, 54 t-distributed stochastic neighborhood embedding (t-SNE/viSNE), 55 and uniform manifold approximation and projection (UMAP). 56 Of these, t-SNE can efficiently present the local data structure and is commonly used in single-cell data analysis. But the limitations of t-SNE include global information loss, long computational time, and inability to provide meaningful results. Compared to t-SNE, UMAP can preserve more data structure during a shorter run time.

Other computational tools further define cell clusters. Clustering approaches can be divided into unsupervised [57] [58] [59] [60] [61] [62] and supervised 63, 64 categories. Unsupervised clustering methods detect cell clusters mainly based on protein expression profiles from a single, multiple, or combined biological sample. Unsupervised clustering detects cell subsets in a data-driven and unbiased manner that facilitates exploratory analysis of previously unknown cell subpopulations. Defined clusters can be individually analyzed or compared across samples in different biological conditions. For example, ACCENSE combines t-SNE mapping with discrete cell clusters identification. 57 With a peak-detection algorithm that identifies local maxima, ACCENSE is able to partition a two-dimensional t-SNE map. Clusters are decided by a user-specified P-value threshold. Corresponding expression profiles are generated with built-in functions for each detected cluster. ACCENSE produces color-coded and density-partitioned cluster maps, facilitating visualization and comparison. Supervised clustering methods usually depend on biological or clinical variables that describe each sample, such as disease status or clinical outcome. The external information can be used to train an interpretable prediction model. For instance, defined clusters that correlate with clinical outcomes can be considered as biomarkers. Automated cell-type discovery and classification (ACDC) 63 and linear discriminant analysis (LDA) are typical supervised clustering methods. ACDC utilizes biological knowledge of a marker × cell type annotation table to guide learning algorithms. Trained with predetermined TA B L E 2 Comparisons between flow cytometry and mass cytometry methods

Classification Description Ref.

Linear, principle component analysis, orthogonal transformation. 51 Isomap Nonlinear, spectral clustering, geodesic distance. 53 Diffusion Map Nonlinear, spectral clustering, diffusion distance. 54 t-SNE/viSNE Nonlinear, t-distributed stochastic neighborhood embedding, attraction/repulsion balance. 55 UMAP Nonlinear, uniform manifold approximation and projection, identify user-specified number of neighbors to build high-dimensional manifolds. 56 ACCENSE Unsupervised clustering t-SNE, kernel-based density estimation, peak-finding, and partitioning. 57 Phenograph k-nearest neighbors (k-NN) detection, community detection, and Jaccard similarity coefficient. 58 Xshift Weighted k-NN density estimation and density-ascending path-based clustering. 59 FlowSOM Self-organizing map, minimal spanning tree-based nodes connection, and consensus of hierarchical meta-clustering. 60 DEPECHE Penalized k-means clustering. 61 SPADE Density-normalization, spanning tree progression analysis, and hierarchical/agglomerative clustering. 62 ACDC Semi-supervised clustering Community detection of landmark points. Cells and random walker-based clustering. 63 LDA Linear discriminant analysis. 64 Citrus

Hierarchically clustering, regularized supervised learning algorithms, nearest shrunken centroid methods, and lasso regularized logistic regression. 67 Wanderlust Differentiation trajectory determination

Ensemble of k I-nearest neighbor graphs, shortest path distance-based trajectory construction, and waypoints-based iteratively trajectory refinement. Furthermore, some computational tools provide statistical testing or infer differentiation trajectories. Citrus 67 (cluster identification, characterization, and regression) aims to identify cell abundance or cellular traits associated with disease conditions used to categorize samples. Conditions generally include healthy control or patients, before or after/with or without therapeutic intervention, cancerous or paracancerous tissue, responsive or irresponsive to treatment, and good or poor prognosis. In addition to hierarchically clustering of phenotypically similar cells, Citrus employs regularized supervised learning algorithms including L1-Penalized Regression (glmnet) and Prediction Analysis for Microarrays to identify predictive cluster features. Gaudillière et al applied Citrus and identified a signature that correlated with patient recovery after hip surgery. 68 Wanderlust sequentially orders cells based on their developmental trajectory and enables inference of the differentiation course. First, the data are transformed into an ensemble of graphs, then waypoints are randomly selected. Next, an orientation trajectory is calculated based on a user-defined "initiate" cell, which is further refined by waypoint cells. Wanderlust produces a trajectory over an average of all graphs. Within the results, each marker can be plotted against the trajectory axis to further examine trends. Wanderlust was used to build and identify novel transitional phenotypes in B-cell development by extending a trajectory map from hematopoietic stem cells to naïve B cells. 69 Novel tools that are faster and containing more features are being continuously developed and proposed. In addition, proper combination of computational approaches may provide more insights into the interpretation of the true complexity of biological systems. Therefore, comprehensive understanding of the functionality, the strengths, as well as the limitations of available algorithms is critical to the selection of optimal analytical tool for specific goals.

The benefits of CyTOF are obvious in the immunology field. In combination with novel bioinformatics techniques, CyTOF allows for the systematic investigations of fundamental questions on immune pathogens and immune mechanisms underlying clinical disease manifestations. A wide range of applications for either mice or humans including studies of cancer, immunotherapy, autoimmune diseases, infective diseases, cardiovascular disease, and neuron science are described below.

With relative high cellular throughput and high dimensions, CyTOF is an ideal and potent technique that reveals intratumor heterogeneity in a variety of tumors, including breast, 70 Table 3 .

Beyond the analysis of phenotypic markers, 88 CyTOF allows simultaneous measurement of cell signaling process 7,85,92,93 through the analysis of protein phosphorylation and neoantigen-specific T-cell pools. In addition to peripheral blood profiling, 81 single-cell suspensions dissociated from organs or tumor tissues 75, 79, 82, 88 can also disentangle alterations in the local immune networks. It is evident that CyTOF can potentially identify valuable phenotypic and functional variations, including cytokines and phosphor-signaling alterations, in the course of tumor development, progression, and metastasis. This enables researchers to identify defined populations for deep causative mechanism studies. Moreover, CyTOF permits biomarker discovery for disease diagnosis or prognosis prediction, because multiple parameters can be measured at the single-cell resolution from clinical samples of different conditions. For example, Good et al utilized CyTOF to simultaneously quantify 35 phenotypic and signaling-associated proteins in the B-cell development of 60 primary diagnostic patients with B-cell precursor acute lymphoblastic leukemia. 93 They employed machine learning to analyze the highdimensional data and identified activated and responsive pre-B-cell receptor signaling in pre-BI cells, and activated mTOR signaling in pro-BII sunsets was sufficient to predict patient relapse at diagnosis. By following a data-driven approach, this study provides a framework for applying CyTOF in human cancer diagnosis and prognosis prediction.

The heterogeneous and suboptimal clinical responses to immunotherapy treatment highlight the need for deep profiling of systemic immune networks to support studies of pathogenesis, disease tracking, therapeutic targets identification, and treatment selection. Spitzer et al 98 demonstrated the utility of CyTOF to profile systemic immune orchestrations with multiparameters in peripheral blood and tissue, including tumor, lymph node, spleen, and bone marrow. They studied three groups of mice with spontaneous model of triple-negative breast cancer: untreated, effective, and ineffective immunotherapy. The investigators sought to define immune system alterations in the tumor environments between effective and ineffective treatment group. Network analysis identified that CD4 + T cells initiated immunotherapy response and conferred protection against new tumors. Researches further confirmed the functional differences of T-cell compartments in response to immunotherapy, both in peripheral blood 99 and in tumor infiltrates. 100 Although current immune checkpoint therapies target mainly lymphoid compartments, researches have started to focus on the role of myeloid compartments in response of immunotherapy (Table 4) . [101] [102] [103] In addition, in combination with immunotherapy and by elevating the efficacy and duration of immune responses, CyTOF offers an attractive promise in developing more effective treatment schemes. Beyrend et al used CyTOF to decipher the rational design of combination immunotherapy and concluded that PD-L1 blockade therapy was enhanced by therapeutically co-targeting activating and inhibitory (LAG3/PD-1) molecules. 104 Chua et al focused on the synergy between radiotherapy and immune checkpoint blockade and found that expansion of activated Ki-67 + CD8 + T cells may account for its synergism relationship. 105 Although challenging, CyTOF profiling should offer novel insights into personalized immunotherapy.

Systemic activation of inflammatory cells plays a critical role in disease severity, progression, and therapy response of patients with autoimmune diseases. A deep understanding of heterogeneities in inflammatory states of individual patients during the disease course can contribute to therapeutic decisions. Many studies have used CyTOF to examine the pathogenesis of autoimmune diseases (Table 4) . 18 

Circulating CXCR3 + CD4 + T, CXCR3 + CD8 + T, and CXCR3 + NK cells were upregulated in severe patients and restored to normal levels after mesenchymal stem cell transplantation. 147 Helminth infection PBMC 37-plex Human type 2 and regulatory networks were heterogeneous in helminth-infected patients. 148 Malaria infection PBMC 29-plex Approximately 80% of mature B cells that expanded after acute infection expressed CD11c. 149 Cardiovascular disease

Aorta 35-plex Aortic leukocyte system is as complex as that in lymphoid organs. 150 Human atherosclerotic plaques

Plaque/PBMC 37-plex The atherosclerotic plaque are dominated by T cells and macrophages. 151 Transplantation Kidney transplantation PBMC 34-plex PD-1 + CD57 − exhausted T cells increased after lymphocyte-depleting induction treatment, which correlated with better allograft function. 154 Kidney transplantation PBMC 33-plex Frequencies of transitional B cell and regulatory T cell at the baseline could discern between responders and nonresponders. 155 Pediatric liver transplantation PBMC 22-plex In operationally tolerant patients, the CD4 + CD5 + CD25 + CD38 −/lo CD45RA − cells were upregulated in comparison with patients of low immunosuppression levels. 156 Neuroscience Normal mouse brain Brain/PBMC 44-plex CD44 discriminates infiltrating and resident myeloid cells in the brain. 157 Postmortem human brain Brain∖PBMC 57-plex Regional specific heterogeneity existed in human microglia. 158 Homeostasis, epilepsy, or tumors Brain 37-plex/36-plex A unique glioma-associated microglia was identified. 159 Aging, Alzheimer's disease, and multiple sclerosis Brain 43-plex Central nervous system border-associated macrophages were distinguished by CD38 and major histocompatibility complex (MHC) II and all microglia are homogenously affected in neuroinflammatory disease. 160 Neurodegeneration Brain 33-plex Repopulated microglia showed IFN regulatory factor 7-driven activation pattern. 161 Alzheimer's disease PBMC 21-plex Increased CD8 + TEM cells were observed in Alzheimer's disease. 162 Refractory epilepsy and autoimmune encephalitis PBMC 40-plex Patients with refractory epilepsy and autoimmune encephalitis displayed CD4 + and CD8 + T cell subsets alterations and unbalanced proinflammatory IL-17 production. Refractory epilepsy patients uniquely showed NK cells alteration. 163 Acute stroke PBMC 38-plex Increased signal transducers and activators of transcription 3 (STAT3) signaling in innate immune cells in the acute phase, increased cAMP-response element binding protein signaling in adaptive immune cells during the intermediate phase, and increased neutrophils and immunoglobulin M (IgM) + B cells in the late phase were observed. 164 Abbreviations: BM, bone marrow; PBMC, peripheral blood mononuclear cells.

depicted with synovitis in small-and medium-sized joints, 18,106-109 exploring both the peripheral blood 18, 107 and synovium tissue samples. 106, 108 In one study, 106 the investigators examined synovitis of patients with rheumatoid arthritis through a 36-plex CyTOF panel specific to activated T cells. The results identified upregulated PD-1 hi CXCR5 − CD4 + T cells in patients, indicating a key functional role of CD4 + T subsets in rheumatoid arthritis. O'Gorman and coworkers focused on specific chemokine signature in Systemic lupus erythematosus (SLE) patients and identified Toll-like receptor activation. 112 Similarly, Slight-Webb's groups applied CyTOF to reveal the STAT3 phosphorylation reduction after mycophenolate mofetil treatment in SLE. 113 Another study, with the aid of CyTOF, interrogated the circulatory reservoir of CD4 + subsets in juvenile idiopathic arthritis patients undergoing TNFalpha therapy withdrawal and found putative subsets prior to withdrawal that discriminated relapse from remission. 111 Researchers have also provided comprehensive overview of distinct immune signatures in multiple autoimmune conditions, including psoriasis, [114] [115] [116] neuroinflammation, 117 and multiple sclerosis. 118 Together, these results provided a strong foundation for CyTOF studies with increased dimensionality to characterize central immune mediators in various autoimmune disorders.

Infection caused by a vast majority of microorganisms induces profound immune responses that involve innate and adaptive immune subsets. Recently, a plethora of studies applied CyTOF to investigate pathogen-specific immunological signatures in infectious diseases (Table 4) . Revealing bacteria's physiology and pathogenicity and the specific immune compartments that steer the immunological reactions to infection is crucial for vaccine development, diagnostic, and tailoring of treatment schemes. CyTOF has been gradually integrated into the research of bacteria-associated diseases to identify pathogen-specific immune signatures and characterize response disparity of leukocytes to vaccine. Rudolph et al studied T-cell responses to HLA-E-restricted Salmonella enterica serovar Typhi antigen before and after Ty21a vaccination. 119 Several multifunctional gut-homing effector memory T and effector memory CD45RA + T cells were more abundant in adult patients, compared with younger children. Napolitani et al performed another CyTOF analysis of Salmonella infection-induced T-cell responses and reconstructed the Salmonella serovar-shaped repertoire of circulating effector CD4 + T cells. 120 CyTOF also supported the investigations of immune factors associated with Mycobacterium tuberculosis 121, 122 and Streptococcus pneumoniae 123 infection. Viral infections pose a constant challenge to the hosts' immune system. Researchers have applied CyTOF to explore the immune alterations of patients infected with influenza, 124 HIV, [125] [126] [127] [128] [129] [130] [131] [132] [133] Japanese Encephalitis, 134 Ebola, 135 Gammaherpesvirus, 136 chikungunya, 137 hepatitis B, 138, 139 as well as the mosquito-borne human viral pathogens, including dengue [140] [141] [142] and Zika, 143, 144 and elucidated the fates of immune cells across viral infections. The use of CyTOF has also supported recent findings in COVID-19 pathogenesis and immune perturbations, 145, 146 where the results showed immunosuppression and dysfunction in PBMCs of COVID-19-infected patients. Leng et al investigated the inflammatory responses to SARS-coronavirus-2 (SARS-COV-2) in patients with COVID-19 after treatment with ACE2-mesenchymal stem cells. 147 A key conclusion of this study is that treatment played a vital immune modulation role to reverse the functional failure of lymphocytes.

Some headway has also been made in employing CyTOF to investigate parasites-induced immune regulatory networks. By profiling type-2 immune response through a 37-marker CyTOF analysis, Ruiter et al revealed detailed insights into the spectrum of immunomodulatory effects of helminth infection. 148 Heathy Indonesians, Indonesians infected with soil-transmitted helminths, and healthy Europeans that are not normally exposed to helminths were profiled. The profiling was conducted prior to and 1 year after deworming. The investigators found that immune signatures in Europeans and Indonesians were distinct and identified both Th2 and rare ILC2 cells, which expanded and acted as sources of type 2 cytokines in helminth-infected patients. Only Th2 cells decreased after deworming, whereas the functional activity of both Th2 and ILC2s declined after anthelmintic treatment. Another study applied CyTOF to profile B-cell compartments in Plasmodium falciparum malaria-infected patients. 149 In general, the understanding of immune subsets contributions in antibacterial, antiviral, and antiparasites defense by CyTOF profiling helps to define correlations of protective immune factors and guide effective vaccine development.

Atherosclerosis is affected by the interaction of pro-and anti-inflammatory factors in the aorta. The immune system's heterogeneity provides an effective defense against various pathogens. The CyTOF technique helps to define the multifaceted contributions of immune defense within circulation and plaques (Table 4) . Recently, Winkels's group demonstrated the phenotypic diversity of leukocytes from aortas of healthy and atherosclerotic mice via single-cell RNA sequencing (scSeq) and CyTOF with a panel of 35 markers. The results revealed three principle B-cell subsets that exhibited varied functional pathway responses, based on marker expressions of CD43 and CD220. 150 Similarly, the Fernandez's group adopted 37-plex CyTOF combined with scSeq and revealed a single-cell immune landscape within human carotid artery plaques. Their analysis shows that, in symptomatic patients, plaques were characterized by distinct T-cell subsets presenting markers of activation, differentiation, and exhaustion, whereas in plaques of asymptomatic patients, T cells and macrophages were activated and exhibited IL-1β signaling. 151 In essence, CyTOF technique is anticipated to accelerate discovery of underlying immunopathogenic factors for tailored cardiovascular immunotherapies.

CyTOF can also investigate the immune system perturbations after transplantation and patients' heterogeneities in response to specific immunosuppressive regimes (Table 4) . 152, 153 One elegant implementation of CyTOF in transplantation research is the work of Fribourg et al 154 ; through characterizing 26 kidney transplant recipients via serially collected PBMCs (before, 3 m, and 6 m after transplantation) using a panel of 35 immune markers, they defined a distinct role of T-cell exhaustion in induction therapy responses and allograft functions. They delineated PD-1 + CD57 − exhausted T cells that correlated with better allograft function, which had low ATP production and cytokine secretion. This helps to identify T-cell exhaustion-associated factors for risk assessment. Allograft biopsies can be examined with the CyTOF technique at unprecedented resolution. CyTOF has also supported the identification of immune subsets that correlated with desensitization therapy results in patients undergoing sensitized kidney transplantation 155 and a distinct immune profile that suggests operational tolerance in pediatric liver transplantation. 156 In addition, CyTOF enables the investigations of protective and pathogenic immunity involved in transplants. As such, CyTOF analysis can link phenotype characteristics to variable posttransplant conditions, reveal the underlying mechanisms, and thus dramatically support the need to create and experiment with new strategies against posttransplant diseases.

The brain consists of highly dynamic and complex microenvironments that are populated with immune cells. The depiction of immune systems within the naïve or impaired brain compartment uncovers the diverse roles of immunomodulatory in central nervous system homeostasis. Korin et al characterized the immune system of the naïve mouse brain, with comparison to immune system in circulation, via CyTOF with 44 surface markers.

The results show that most infiltrating leukocytes were mainly present at the brain boundaries, such as meninges and choroid plexus, and that CD44 distinguished resident and infiltrating immune cells. 157 To better define human microglia cell repertoire, CyTOF was applied to investigate postmortem human microglia isolated from nine donors 158 and interestingly revealed that mucin-like hormone receptor 1 was expressed in human microglia, whereas it was not expressed in the monocytes and myeloid cells of circulation and cerebrospinal fluid.

Researches also provided a nuanced comparison of the human brain's immune states during homeostasis, aging, and disease through the integration of CyTOF and other omic techniques and demonstrated that microglia subsets exhibited disease-specific transformations in the brain (Table 4 ). [159] [160] [161] [162] [163] Another interesting study focused on the impact of immune response to stroke on long-term cognitive disability. 164 With the approach of CyTOF, serial blood sample from patients was collected and analyzed over the course of a year to functionally and comprehensively characterize the immune response to stroke and its correlation with cognitive functioning between 90 and 365 days poststroke. The results confirmed a significant correlation between immune response measured during the stroke trajectories, which demonstrated the utility of CyTOF in the clinical prognosis and prediction of stroke. In summary, these studies depict the confounding utility of CyTOF in understanding immune orchestrations in human brain.

Recent progress and applications illuminate the salient features and the prospects of CyTOF in sketching the immune landscape. In essence, CyTOF can picture both innate 77, 137, 165, 166 and adaptive 123, 167 immune landscapes, which includes numerous phenotypically and functionally heterogeneous cell subsets of lymphoid and myeloid lineages that are involved in adequate surveillance and pathogens killing. Panoramic views of systemic immunity involving circulation and infiltration can also be obtained. 98 In addition to horizontal comparison among groups of health and different disease statuses, CyTOF also supports longitudinal profiling, 164 with the advancement in normalization 48, 168 and batch effect minimizing. 49, 50 CyTOF can also work in conjunction with other techniques including single-cell genome and transcriptome profiling 169, 170 and bioinformatic pipelines. 171, 172 For instance, Zheng et al conducted CyTOF profiling of immune microenvironment in hepatocellular carcinoma and revealed that leading-edge regions exhibited an increase of tumor-associated CD4/CD8 double-positive T (DPT) cells, which synergistically expressed PD-1/HLA-DR/ICOS/CD45RO. 173 Single-cell RNA-seq was employed to characterize DPT cells and specifically identified PD-1 high DPT cluster derived from intratumoral CD8 + T cells. Instead of only investigating single aspects of a cell subset, the joint analysis of multiple omics offers the opportunity to provide comprehensive insights into coordinated cellular process across different omic layers.

In general, the current progress made by CyTOF techniques mainly includes identifying cell ancestry to reveal the trajectory of cell differentiation and find reliable and specific immune signatures in patient stratification and treatment. CyTOF has facilitated the biospecimens analysis across large cohorts and permits the identification of ideal immunological biomarkers that are stable, reproducible, and easily measured, to reflect features underlying pathophysiology, 174,175 disease progression, 95 or treatment schemes. 176 We expect CyTOF to be established in pharmacological research and function as a powerful drug screening tool. Cell heterogeneities pose challenges for treatment selection. 177 However, this heterogeneity is often obscured in conventional screening methods that yield average measurements of bulk populations, motivating the need for a high-throughput single-cell technique. CyTOF facilitates the screening of signaling network and dynamics of baseline or stimulated status. This creates a detailed response profile of the drug-perturbed immune system and aids in the examination of drugs' impact on the immune system. 178 The understanding of drugs and its relationships with the immune system confers information for treating diseases, minimizing unwanted side effects, facilitates individual-tailored medicine, and optimizes combination therapies. 179 In addition, we expect CyTOF to enter clinical laboratories and gradually play a dominant role in immune compartment profiling. By virtue of its capacity of comprehensive and systematic immune characterization, CyTOF would play an indispensable role in routine evaluation of global health.

However, critical challenges remain. Several aspects of CyTOF require careful adaptation and routinization before its adaption in routine laboratory and clinical settings. Notable among these aspects are the standardization of single-cell preparation techniques, rigorous antibody validation, sensitivity enhancement of antibody tags, a paradigm shift in pattern linking of data to relevant clinical outcomes, and even more parameters in routine use. Once the challenges in scalability, sensitivity, reproducibility, and reliability are overcome, it seems likely that in the near future, in tandem with its wider accessibility in pharmaceutical companies and clinical departments, CyTOF will assume the center stage to sketch immune landscape for varieties of diseases. We envision it playing a more confounding role in drug and vaccine development and immune demystification.

Researchers across biological fields have advanced the understanding of immunocyte heterogeneities involved in malignant diseases. By utilizing the increased dimensionality, CyTOF provides opportunities toward unlocking the mysteries surrounding immunopathogenesis and immune responses, which underlines clinical manifestation and shows immense potential for clinical laboratorial diagnosis, therapy efficacy monitoring, as well as treatment strategies. Ongoing improvements on technical aspects, including marker scalability and mass tag barcoding approaches for eliminating batch effects, facilitate the adaptation of CyTOF for routine applications in research and clinical laboratories. Coupled with innovative data analysis pipelines, application of CyTOF should accelerate the progress in fundamental and clinical immunology and advance the study and application of precision medicine.

TZ and XD were involved in the conception, design, and drafting of the manuscript. TZ, AW, YL, and XD were involved in the editing and revision of the manuscript. All authors have read and approved the final manuscript.

This research was supported by the National Natural Science Foundation of China (81871448), Shanghai Municipal Science and Technology (2017SHZDZX01), and Shanghai Municipal Education Commission (ZXWF082101).

The authors declare no conflict of interest. 

A deep profiler's guide to cytometry

Compensation of signal spillover in suspension and imaging mass cytometry

Palladium-based mass tag cell barcoding with a doublet-filtering scheme and single-cell deconvolution algorithm

Surface barcoding of live PBMC for multiplexed mass cytometry

A mass-ratiometrybased CD45 barcoding method for mass cytometry detection

Impaired coordination between signaling pathways is revealed in human colorectal cancer using single-cell mass cytometry of archival tissue blocks

Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry

A carrier strategy for mass cytometry analysis of small numbers of cells

Mass cytometry: a highly multiplexed single-cell technology for advancing drug development

Phenotype molding of T cells in colorectal cancer by single-cell analysis

Visual analysis of mass cytometry data by hierarchical stochastic neighbour embedding reveals rare cell types

Mass cytometry analytical approaches reveal cytokineinduced changes in natural killer cells

Individual liver plasmacytoid dendritic cells are capable of producing IFNalpha and multiple additional cytokines during chronic HCV infection

Single-cell mass cytometry of differential immune and drug responses across a human hematopoietic continuum

Multiplexed mass cytometry profiling of cellular states perturbed by smallmolecule regulators

Staining of phosphorylated signalling markers protocol for mass cytometry

Candidate markers for stratification and classification in rheumatoid arthritis

Mass cytometry analysis reveals hyperactive NF Kappa B signaling in myelofibrosis and secondary acute myeloid leukemia

Cell size assays for mass cytometry

CellCycleTRACER accounts for cell cycle and volume in mass cytometry data

A platinum-based covalent viability reagent for single-cell mass cytometry

Single-cell mass cytometry adapted to measurements of the cell cycle

Mass cytometric cell cycle analysis

Mass cytometric functional profiling of acute myeloid leukemia defines cell-cycle and immunophenotypic properties that correlate with known responses to therapy

Reversibility of defective hematopoiesis caused by telomere shortening in telomerase knockout mice

Proliferation tracing with single-cell mass cytometry optimizes generation of stem cell memory-like T cells

Monitoring CD49d receptor occupancy: a method to optimize and personalize natalizumab therapy in multiple sclerosis patients

Optimization of receptor occupancy assays in mass cytometry: standardization across channels with QSC beads

Expanding the clinical cytometry toolbox-receptor occupancy by mass cytometry

Combinatorial tetramer staining and mass cytometry analysis facilitate T-cell epitope mapping and characterization

Cross-species singlecell analysis of pancreatic ductal adenocarcinoma reveals antigen-presenting cancer-associated fibroblasts

Multiplexed imaging of immune cells in staged multiple sclerosis lesions by mass cytometry. eLife

Normalization of mass cytometry data with bead standards

Minimizing batch effects in mass cytometry data

Removal of batch effects using distribution-matching residual networks

Analysis of a complex of statistical variables into principle components

Cytometry by time-of-flight shows combinatorial cytokine expression and virus-specific cell niches within a continuum of CD8+ T cell phenotypes

A global geometric framework for nonlinear dimensionality reduction

Diffusion maps for high-dimensional single-cell analysis of differentiation data

viSNE enables visualization of high dimensional single-cell data and reveals phenotypic heterogeneity of leukemia

Dimensionality reduction for visualizing single-cell data using UMAP

Automatic Classification of Cellular Expression by Nonlinear Stochastic Embedding (ACCENSE)

Data-driven phenotypic dissection of AML reveals progenitor-like cells that correlate with prognosis

Automated mapping of phenotype space with single-cell data

FlowSOM: using self-organizing maps for visualization and interpretation of cytometry data

Determination of essential phenotypic elements of clusters in high-dimensional entities-DEPECHE

Visualization and cellular hierarchy inference of single-cell data using SPADE

Automated cell type discovery and classification through knowledge transfer

Predicting cell populations in single cell mass cytometry data

Comparison of clustering methods for high-dimensional single-cell flow and mass cytometry data

A comparison framework and guideline of clustering methods for mass cytometry data

Automated identification of stratifying signatures in cellular subpopulations

Clinical recovery from surgery correlates with single-cell immune signatures

Single-cell trajectory detection uncovers progression and regulatory coordination in human B cell development

A single-cell atlas of the tumor and immune ecosystem of human breast cancer

Single cell mass cytometry revealed the immunomodulatory effect of cisplatin via downregulation of splenic CD44+, IL-17A+ MDSCs and promotion of circulating IFN-gamma+ myeloid cells in the 4T1 metastatic breast cancer model

Single-cell map of diverse immune phenotypes in the breast tumor microenvironment

Cdh1 and Pik3ca mutations cooperate to induce immune-related invasive lobular carcinoma of the breast

Highdimensional single cell mapping of cerium distribution in the lung immune microenvironment of an active smoker. Cytometry B Clin Cytom

Expression analysis and significance of PD-1, LAG-3, and TIM-3 in human non-small cell lung cancer using spatially resolved and multiparametric single-cell analysis

Mapping lung cancer epithelial-mesenchymal transition states and trajectories with single-cell resolution

Innate immune landscape in early lung adenocarcinoma by paired single-cell analyses

Intratumoral HPV16-specific T cells constitute a type I-oriented tumor microenvironment to improve survival in HPV16-driven oropharyngeal cancer

Immune profiling of human tumors identifies CD73 as a combinatorial target in glioblastoma

Metronomic capecitabine as an immune modulator in glioblastoma patients reduces myeloid-derived suppressor cells

Global immune fingerprinting in glioblastoma patient peripheral blood reveals immune-suppression signatures associated with prognosis

Single-cell mapping of human brain cancer reveals tumor-specific instruction of tissue-invading leukocytes

Expression of PD-1 by T cells in malignant glioma patients reflects exhaustion and activation

Immunocyte profiling using singlecell mass cytometry reveals EpCAM(+) CD4(+) T cells abnormal in colon cancer

Mass cytometry reveals a sustained reduction in CD16(+) natural killer cells following chemotherapy in colorectal cancer patients

Regulatory T cells in ovarian cancer are characterized by a highly activated phenotype distinct from that in melanoma

Commonly occurring cell subsets in high-grade serous ovarian tumors identified by single-cell mass cytometry

An immune atlas of clear cell renal cell carcinoma

High-dimensional single-cell proteomics analysis reveals the landscape of immune cells and stem-like cells in renal tumors

High-dimensional analyses reveal a distinct role of T-cell subsets in the immune microenvironment of gastric cancer

Application value of CyTOF 2 mass cytometer technology at single-cell level in human gastric cancer cells

Identification of functionally primitive and immunophenotypically distinct subpopulations in secondary acute myeloid leukemia by mass cytometry

Single-cell developmental classification of B cell precursor acute lymphoblastic leukemia at diagnosis reveals predictors of relapse

Mass cytometry defines distinct immune profile in germinal center B-cell lymphomas

Mass cytometry analysis reveals that specific intratumoral CD4(+) T cell subsets correlate with patient survival in follicular lymphoma

Establishing high dimensional immune signatures from peripheral blood via mass cytometry in a discovery cohort of stage IV melanoma patients

GCN2 drives macrophage and MDSC function and immunosuppression in the tumor microenvironment

Systemic immunity is required for effective cancer immunotherapy

T-cell invigoration to tumour burden ratio associated with anti-PD-1 response

Distinct cellular mechanisms underlie anti-CTLA-4 and anti-PD-1 checkpoint blockade

High-dimensional single-cell analysis predicts response to anti-PD-1 immunotherapy

High-dimensional analysis delineates myeloid and lymphoid compartment remodeling during successful immune-checkpoint cancer therapy

Distinct predictive biomarker candidates for response to anti-CTLA-4 and anti-PD-1 immunotherapy in melanoma patients

PD-L1 blockade engages tumor-infiltrating lymphocytes to co-express targetable activating and inhibitory receptors

High-dimensional characterization of the systemic immune landscape informs on synergism between radiotherapy and immune checkpoint blockade

Pathologically expanded peripheral T helper cell subset drives B cells in rheumatoid arthritis

Mixed-effects association of single cells identifies an expanded effector CD4(+) T cell subset in rheumatoid arthritis

Defining inflammatory cell states in rheumatoid arthritis joint synovial tissues by integrating single-cell transcriptomics and mass cytometry

Characterization of phenotypes and functional activities of leukocytes from rheumatoid arthritis patients by mass cytometry

Cytometry by time of flight identifies distinct signatures in patients with systemic sclerosis, systemic lupus erythematosus and Sjogrens syndrome

Immunome perturbation is present in patients with juvenile idiopathic arthritis who are in remission and will relapse upon anti-TNF alpha withdrawal

Single-cell systems-level analysis of human Toll-like receptor activation defines a chemokine signature in patients with systemic lupus erythematosus

Mycophenolate mofetil reduces STAT3 phosphorylation in systemic lupus erythematosus patients

Lymphocyte mass cytometry identifies a CD3-CD4+ cell subset with a potential role in psoriasis

T-cell phenotyping uncovers systemic features of atopic dermatitis and psoriasis

Multiparameter immune profiling of peripheral blood mononuclear cells by multiplexed single-cell mass cytometry in patients with early multiple sclerosis

Single-cell mass cytometry reveals distinct populations of brain myeloid cells in mouse neuroinflammation and neurodegeneration models

GM-CSF and CXCR4 define a T helper cell signature in multiple sclerosis

Age-associated heterogeneity of Ty21a-induced T cell responses to HLA-E restricted salmonella typhi antigen presentation

Clonal analysis of Salmonella-specific effector T cells reveals serovar-specific and cross-reactive T cell responses

Metformin alters human host responses to mycobacterium tuberculosis in healthy subjects

A multi-cohort study of the immune factors associated with M. tuberculosis infection outcomes

Innate and adaptive nasal mucosal immune responses following experimental human pneumococcal colonization

Differential induction of IFN-alpha and modulation of CD112 and CD54 expression govern the magnitude of NK cell IFN-gamma response to influenza A viruses

Mass cytometric analysis of HIV entry, replication, and remodeling in tissue CD4+ T cells

Comprehensive mass cytometry analysis of cell cycle, activation, and coinhibitory receptors expression in CD4 T cells from healthy and HIVinfected individuals

IL-15 regulates susceptibility of CD4(+) T cells to HIV infection

The receptor repertoire and functional profile of follicular T cells in HIV-infected lymph nodes

Mass cytometry identifies distinct CD4+ T cell clusters distinguishing HIV-1-infected patients according to antiretroviral therapy initiation

Mass cytometry analysis reveals complex cell-state modifications of blood myeloid cells during HIV infection

Characterization of leukocytes from HIV-ART patients using combined cytometric profiles of 72 cell markers

Upregulation of chitinase 1 in alveolar macrophages of HIV-infected smokers

Treated HIV infection alters phenotype but not HIV-specific function of peripheral blood natural killer cells

PD1(+)CCR2(+)CD8(+) T cells infiltrate the central nervous system during acute japanese encephalitis virus infection

Immunologic timeline of Ebola virus disease and recovery in humans

High-dimensional characterization of IL-10 production and IL-10-dependent regulation during primary gammaherpesvirus infection

Comprehensive innate immune profiling of chikungunya virus infection in pediatric cases

Effects of hepatitis B surface antigen on virus-specific and global T cells in patients with chronic HBV infection

Distinct phenotype and function of circulating Vdelta1+ and Vdelta2+ gammadeltaT-cells in acute and chronic hepatitis B

Single cell immune profiling of dengue virus patients reveals intact immune responses to Zika virus with enrichment of innate immune signatures

Dengue-specific CD8+ T cell subsets display specialized transcriptomic and TCR profiles

Large-scale HLA tetramer tracking of T cells during dengue infection reveals broad acute activation and differentiation into two memory cell fates

Comprehensive immunoprofiling of pediatric zika reveals key role for monocytes in the acute phase and no effect of prior dengue virus infection

Arumugaswami V. Zika virus mucosal infection provides protective immunity

High-dimensional immune profiling by mass cytometry revealed immunosuppression and dysfunction of immunity in COVID-19 patients

Systemslevel immunomonitoring from acute to recovery phase of severe COVID-19

Transplantation of ACE2(-) mesenchymal stem cells improves the outcome of patients with COVID-19 pneumonia

Helminth infections drive heterogeneity in human type 2 and regulatory cells

B cell profiling in malaria reveals expansion and remodeling of CD11c(+) B cell subsets

Atlas of the immune cell repertoire in mouse atherosclerosis defined by single-cell RNA-sequencing and mass cytometry

Singlecell immune landscape of human atherosclerotic plaques

Applying mass cytometry to the analysis of lymphoid populations in transplantation

Single cell immune profiling in transplantation research

T-cell exhaustion correlates with improved outcomes in kidney transplant recipients

Immune profiles to predict response to desensitization therapy in highly HLA-sensitized kidney transplant candidates

Mass cytometry reveals a distinct immunoprofile of operational tolerance in pediatric liver transplantation

High-dimensional, single-cell characterization of the brain's immune compartment

Human microglia regional heterogeneity and phenotypes determined by multiplexed single-cell mass cytometry

Mapping microglia states in the human brain through the integration of highdimensional techniques

High-Dimensional single-cell mapping of central nervous system immune cells reveals distinct myeloid subsets in health, aging, and disease

Acute microglia ablation induces neurodegeneration in the somatosensory system

Clonally expanded CD8 T cells patrol the cerebrospinal fluid in Alzheimer's disease

Pro-inflammatory, IL-17 pathways dominate the architecture of the immunome in pediatric refractory epilepsy

A year-long immune profile of the systemic response in acute stroke survivors

Mass cytometry reveals innate lymphoid cell differentiation pathways in the human fetal intestine

In depth comparative phenotyping of blood innate myeloid leukocytes from healthy humans and macaques using mass cytometry. Cytometry A

Innate and adaptive immune correlates of chronic and self-limiting Epstein-Barr virus (EBV) DNAemia in solid-organ transplant recipients

CytoNorm: a normalization algorithm for cytometry data

Defining human cardiac transcription factor hierarchies using integrated singlecell heterogeneity analysis

Mapping the human DC lineage through the integration of high-dimensional techniques

Evaluation of integrative clustering methods for the analysis of multiomics data

Fast, sensitive and accurate integration of single-cell data with Harmony

Trajectory and functional analysis of PD-1(high) CD4(+)CD8(+) T cells in hepatocellular carcinoma by single-cell cytometry and transcriptome sequencing

High dimensional immune biomarkers demonstrate differences in phenotypes and endotypes in food allergy and asthma

Sensitive detection of rare diseaseassociated cell subsets via representation learning

Immune activation underlies a sustained clinical response to Yttrium-90 radioembolisation in hepatocellular carcinoma

Sivakumar S. Tumor heterogeneity: does it matter?

Mapping the effects of drugs on the immune system

DRUG-NEM: optimizing drug combinations using single-cell perturbation response to account for intratumoral heterogeneity