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

Use of Single-Cell -Omic Technologies to Study the Gastrointestinal Tract and Diseases, From Single Cell Identities to Patient Features

Background:Although microvascular changes are the earliest histopathologic manifestation of systemic sclerosis (SSc), the vascular pathophysiology and the microenvironment within vascular niche remains poorly understood.Objectives:Here, we employed three different spatial omics...

Clinical successes with immune check-point blockers have demonstrated the potency of the immune system in controlling cancers, most strikingly in Hodgkin lymphoma (HL), where overall response rates to PD1/L1 inhibitors approach 90%. Complete or durable responses, however, are uncommon, therefore targeting the PD1/L1 axis alone is not sufficient. Recent work analyzing the spatial arrangement of PD1 and PDL1 expressing cells has given us new insight into the mechanism of action of PD1/L1 inhibitors, however this work limited itself to studying a single check point marker on a subset of cells. We hypothesize that comprehensive profiling of the frequency and spatial arrangement of immune cells in the Hodgkin lymphoma tumor immune microenvironment (TME) will provide new insights into the mechanism of checkpoint blockers and identify novel targets for immune therapy. Until now, multiparameter spatial analysis of the immune microenvironment was limited by technical challenges. Flow and mass cytometry are able to identify immune subsets of interest but spatial information is lost. Multiplex tissue imaging methods are limited to 6-8 simultaneous markers and cannot capture the full complexity of the immune phenotypes. The Fluidigm Hyperion imaging mass cytometry (IMC) system combines a CyTOF mass cytometer with a laser ablation system allowing for 40+ parameter simultaneous immunophenotyping on a single slide of FFPE tissue, with sub-cellular resolution. We have developed a panel of 34 antibodies that allow for comprehensive characterization of CD4, CD8 and myeloid cells components in the TME of Hodgkin lymphoma using IMC. Here we report on our spatial analysis of TIM3 and LAG3 expressing CD4+ lymphocytes. Our data suggests LAG3+CD4+ and TIM3+CD4+ lymphocytes had shorter mean nearest distance to PDL1+Hodgkin Reed-Sternberg (HRS) cells upon comparison to PDL1- HRS cells (t-test, p=1.703e-08,p=1.126e-14). Future studies should explore multiple exhausted marker models that seeks to further understand the presence of simultaneous exhaustion signals in the niche environment. These data suggest that therapies that target TIM3 and/or LAG3 should be tested in Hodgkin Lymphoma and that spatial analysis of immune subsets by IMC should be explored as selective and pharmacodynamic biomarkers. Citation Format: Anthony R. Colombo, Monirath Hav, Erik Gerdtsson, Jose Bisnesto-Villasboas, Stephen Ansell, James Hicks, Peter Kuhn, Akil Merchant. Revisiting immune exhaustion in Hodgkin’s lymphoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1189.

Clinical successes with immune check-point blockers have demonstrated the potency of the immune system in controlling cancers, most strikingly in Hodgkin lymphoma (HL), where overall response rates to PD1/L1 inhibitors approach 90%. Complete or durable responses, however, are uncommon, therefore targeting the PD1/L1 axis alone is not sufficient. Recent work analyzing the spatial arrangement of PD1 and PDL1 expressing cells has given us new insight into the mechanism of action of PD1/L1 inhibitors, however this work limited itself to studying a single check point marker on a subset of cells. We hypothesize that comprehensive profiling of the frequency and spatial arrangement of immune cells in the Hodgkin lymphoma tumor immune microenvironment (TME) will provide new insights into the mechanism of checkpoint blockers and identify novel targets for immune therapy. Until now, multiparameter spatial analysis of the immune microenvironment was limited by technical challenges. Flow and mass cytometry are able to identify immune subsets of interest but spatial information is lost. Multiplex tissue imaging methods are limited to 6-8 simultaneous markers and cannot capture the full complexity of the immune phenotypes. The Fluidigm Hyperion imaging mass cytometry (IMC) system combines a CyTOF mass cytometer with a laser ablation system allowing for 40+ parameter simultaneous immunophenotyping on a single slide of FFPE tissue, with sub-cellular resolution. We have developed a panel of 34 antibodies that allow for comprehensive characterization of CD4, CD8 and myeloid cells components in the TME of Hodgkin lymphoma using IMC. Here we report on our spatial analysis of TIM3 and LAG3 expressing CD4+ lymphocytes. Our data suggests LAG3+CD4+ and TIM3+CD4+ lymphocytes had shorter mean nearest distance to PDL1+Hodgkin Reed-Sternberg (HRS) cells upon comparison to PDL1- HRS cells (t-test, p=1.703e-08,p=1.126e-14). Future studies should explore multiple exhausted marker models that seeks to further understand the presence of simultaneous exhaustion signals in the niche environment. These data suggest that therapies that target TIM3 and/or LAG3 should be tested in Hodgkin Lymphoma and that spatial analysis of immune subsets by IMC should be explored as selective and pharmacodynamic biomarkers. Citation Format: Anthony R. Colombo, Monirath Hav, Erik Gerdtsson, Jose Bisnesto-Villasboas, Stephen Ansell, James Hicks, Peter Kuhn, Akil Merchant. Revisiting immune exhaustion in Hodgkin’s lymphoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1189.

Microglia, the resident innate immune cells of the central nervous system (CNS), play an important role in brain development and homoeostasis, as well as in neuroinflammatory, neurodegenerative and psychiatric diseases. Studies in animal models have been used to determine the origin and development of microglia, and how these cells alter their transcriptional and phenotypic signatures during CNS pathology. However, little is known about their human counterparts. Recent studies in human brain samples have harnessed the power of multiplexed single‐cell technologies such as single‐cell RNA sequencing (scRNA‐seq) and mass cytometry (cytometry by time‐of‐flight [CyTOF]) to provide a comprehensive molecular view of human microglia in healthy and diseased brains. CyTOF is a powerful tool to study high‐dimensional protein expression of human microglia (huMG) at the single‐cell level. This technology widens the possibilities of high‐throughput quantification (of over 60 targeted molecules) at a single‐cell resolution. CyTOF can be combined with scRNA‐seq for comprehensive analysis, as it allows single‐cell analysis of post‐translational modifications of proteins, which provides insights into cell signalling dynamics in targeted cells. In addition, imaging mass cytometry (IMC) has recently become commercially available, and will be useful for analysing multiple cell types in human brain sections. IMC leverages mass spectrometry to acquire spatial data of cell–cell interactions on tissue sections, using (theoretically) over 40 markers at the same time. In this review, we summarise recent studies of huMG using CyTOF and IMC analyses. The uses and limitations as well as future directions of these technologies are discussed.

Analysis of multiplexed assays is highly important for clinical diagnostics and other analytical applications. Mass cytometry enables multi-dimensional, single-cell analysis of cell type and state. In mass cytometry, the rare earth metals used as reporters on antibodies allow determination of marker expression in individual cells. Barcode-based bioassays for CyTOF are able to encode and decode for different experimental conditions or samples within the same experiment, facilitating progress in producing straightforward and consistent results. Herein, an integrated protocol for automated sample preparation for barcoding used in conjunction with mass cytometry for clinical bioanalysis samples is described; we offer results of our work with barcoding protocol optimization. In addition, we present some points to be considered in order to minimize the variability of quantitative mass cytometry measurements. For example, we discuss the importance of having multiple populations during titration of the antibodies and effect of storage and shipping of labelled samples on the stability of staining for purposes of CyTOF analysis. Data quality is not affected when labelled samples are stored either frozen or at 4°C and used within 10days; we observed that cell loss is greater if cells are washed with deionized water prior to shipment or are shipped in lower concentration. Once the labelled samples for CyTOF are suspended in deionized water, the analysis should be performed expeditiously, preferably within the first hour. Damage can be minimized if the cells are resuspended in phosphate-buffered saline (PBS) rather than deionized water while waiting for data acquisition.

Highly multiplexed imaging technology is a powerful tool to facilitate understanding the composition and interactions of cells in tumor microenvironments at subcellular resolution, which is crucial for both basic research and clinical applications. Imaging mass cytometry (IMC), a multiplex imaging method recently introduced, can measure up to 100 markers simultaneously in one tissue section by using a high-resolution laser with a mass cytometer. However, due to its high resolution and large number of channels, how to process and interpret the image data from IMC remains a key challenge to its further applications. Accurate and reliable single cell segmentation is the first and a critical step to process IMC image data. Unfortunately, existing segmentation pipelines either produce inaccurate cell segmentation results or require manual annotation, which is very time consuming. Here, we developed Dice-XMBD1, a Deep learnIng-based Cell sEgmentation algorithm for tissue multiplexed imaging data. In comparison with other state-of-the-art cell segmentation methods currently used for IMC images, Dice-XMBD generates more accurate single cell masks efficiently on IMC images produced with different nuclear, membrane, and cytoplasm markers. All codes and datasets are available at https://github.com/xmuyulab/Dice-XMBD.

Imaging mass cytometry (IMC) is able to quantify the expression of dozens of markers at sub-cellular resolution on a single tissue section by combining a novel laser ablation system with mass cytometry. As such, it allows us to gain spatial information and antigen quantification in situ, and can be applied to both snap-frozen and formalin-fixed, paraffin-embedded (FFPE) tissue sections. Herein, we have developed and optimized the immunodetection conditions for a 34-antibody panel for use on human snap-frozen tissue sections. For this, we tested the performance of 80 antibodies. Moreover, we compared tissue drying times, fixation procedures and antibody incubation conditions. We observed that variations in the drying times of tissue sections had little impact on the quality of the images. Fixation with methanol for 5 min at −20°C or 1% paraformaldehyde (PFA) for 5 min at room temperature followed by methanol for 5 min at −20°C were superior to fixation with acetone or PFA only. Finally, we observed that antibody incubation overnight at 4°C yielded more consistent results as compared to staining at room temperature for 5 h. Finally, we used the optimized method for staining of human fetal and adult intestinal tissue samples. We present the tissue architecture and spatial distribution of the stromal cells and immune cells in these samples visualizing blood vessels, the epithelium and lamina propria based on the expression of α-smooth muscle actin (α-SMA), E-Cadherin and Vimentin, while simultaneously revealing the colocalization of T cells, innate lymphoid cells (ILCs), and various myeloid cell subsets in the lamina propria of the human fetal intestine. We expect that this work can aid the scientific community who wish to improve IMC data quality.

BackgroundRenal Cell Carcinoma (RCC) has been characterized as being amongst the most immune infiltrated solid tumors with a highly heterogenous immune landscape. Within spatially organized cellular networks (CNs) of the tumor immune microenvironment (TIME), key immune cell-cell interactions (CCIs) impact immune cell function and organization ultimately impacting the patient’s overall response. Multiple studies have observed an association of tertiary lymphoid structures, a commonly observed spatial CN, with positive clinical outcomes in RCC, however additional CNs and the CCIs that control these networks need to be identified to better harness and potentially reprogram immune responses to improve patient outcomes. The recent explosion of interest in the heterogeneity of the immune landscape in RCC has led to numerous publications using the latest technologies in spatial transcriptomics, proteomics, and metabolomics. However, many of these studies use this data in isolation and therefore, may be hindered by the technological biases inherent in each method. Here, we have developed a novel approach to integrate spatial and single cell multi-omic data harnessing the strengths of each technology to better interrogate CNs that exist in the RCC TIME.MethodsFresh surgical samples were procured at the University Health Network (Toronto, Canada) through the REnal cancer MicroEnvironment DiscoverY (REMEDY) study. Bulk RNA sequencing (RNA-seq), single cell RNA sequencing (scRNA-seq), single cell suspension mass cytometry (SMC), whole transcriptome digital spatial profiling (DSP), and imaging mass cytometry (IMC) was performed on spatially concordant tumor regions across 54 patients. scRNA-seq enabled the identification of immune, stromal, and malignant high-resolution cell states, which informed a tailored antibody panel design for SMC and IMC and served as a reference framework for harmonized cell-type annotation across modalities. This enabled the integration of our transcriptomic and proteomic data to delineate RCC-specific CNs enriched for defined CCIs across unique biological pathways.ResultsUsing this integrative approach, we identified seven high-resolution patient immunophenotypes. To evaluate their prognostic and predictive relevance, we derived representative gene signatures using a linear mixed model (Flash-MM) to interrogate publicly available bulk RNA-seq datasets, including TCGA, JAVELIN, and IMmotion151. This analysis revealed immunophenotype-specific associations with survival following surgery or systemic therapy in univariate models. To explore potential biological mechanisms associated with these divergent clinical outcomes, we incorporated spatial information into traditional CCI analyses and performed pathway analyses to define functional relationships. This revealed that patient subtypes with high lymphoid infiltration exhibit greater spatial heterogeneity, potentially reflecting the coexistence of multiple activated immune pathways. In contrast, patient subtypes with low immune infiltration were enriched in more developmental signaling pathways. In addition to our biological observations, we were also able to assess the technological differences between patient matched samples and compare the ability of each technology to capture inter-patient and intra-patient heterogeneity.ConclusionsCollectively, this work identifies clinically distinct subgroups defined by CNs and details the interpatient cellular heterogeneity that exists in RCC, providing the foundation for future personalized therapeutic interventions against this disease.

Mass cytometry (CyTOF) and imaging mass cytometry (IMC), as cutting-edge technologies in single-cell analysis, are capable of detecting more than 40 biomarkers simultaneously on a single cell. However, their sensitivity and multiparameter detection capabilities have been long constrained by the development of metal labeling materials. Meanwhile, as an imaging technique, IMC has suffered from a rather slow data acquisition rate. Here, we present a luminescent PCN-224-OH material that exhibits both fluorescent and mass dual-functionality and is enriched with Zr-OH-/H2O active sites. Without the additional need for complex postmodification or chemical coupling reactions, PCN-224-OH can be directly functionalized with antibodies/aptamers and poly(ethylene glycol) (PEG), resulting in the production of PCN-224-Ab-PEG or PCN-224-Apt-PEG probes. We demonstrated that PCN-224-Ab-PEG was compatible with commercial polymer-based probes but with superior sensitivity and specificity. Meanwhile, since PCN-224-Apt-PEG expressed both fluorescence and mass signals, we could adopt fluorescence signals for rapid tissue section scanning to swiftly identify the regions of interest (ROIs), and then adopt IMC for multiparameter imaging at the specific ROIs. The application of the PCN-224-Apt-PEG probe could significantly reduce the blind IMC scanning time by up to 90% and effectively compensate for IMC's low resolution. This study not only broadens the application scope of luminescent metal-organic frameworks but also offers a potentially novel toolbox for single-cell multiparameter detection.

Imaging Mass Cytometry (IMC) is an expansion of mass cytometry, but rather than analyzing single cells in suspension, it uses laser ablation to generate plumes of particles that are carried to the mass cytometer by a stream of inert gas. Images reconstructed from tissue sections scanned by IMC have a resolution comparable to light microscopy, with the high content of mass cytometry enabled through the use of isotopically labeled probes and ICP-MS detection. Importantly, IMC can be performed on paraffin-embedded tissue sections, so can be applied to the retrospective analysis of patient cohorts whose outcome is known, and eventually to personalized medicine. Since the original description in 2014, IMC has evolved rapidly into a commercial instrument of unprecedented power for the analysis of histological sections. In this Review, we discuss the underlying principles of this new technology, and outline emerging applications of IMC in the analysis of normal and pathological tissues. © 2017 International Society for Advancement of Cytometry.

Mass cytometry (cytometry by time-of-flight detection [CyTOF]) is a bioanalytical technique that enables the identification and quantification of diverse features of cellular systems with single-cell resolution. In suspension mass cytometry, cells are stained with stable heavy-atom isotope-tagged reagents, and then the cells are nebulized into an inductively coupled plasma time-of-flight mass spectrometry (ICP-TOF-MS) instrument. In imaging mass cytometry, a pulsed laser is used to ablate ca. 1 μm2 spots of a tissue section. The plume is then transferred to the CyTOF, generating an image of biomarker expression. Similar measurements are possible with multiplexed ion bean imaging (MIBI). The unit mass resolution of the ICP-TOF-MS detector allows for multiparametric analysis of (in principle) up to 130 different parameters. Currently available reagents, however, allow simultaneous measurement of up to 50 biomarkers. As new reagents are developed, the scope of information that can be obtained by mass cytometry continues to increase, particularly due to the development of new small molecule reagents which enable monitoring of active biochemistry at the cellular level. This review summarizes the history and current state of mass cytometry reagent development and elaborates on areas where there is a need for new reagents. Additionally, this review provides guidelines on how new reagents should be tested and how the data should be presented to make them most meaningful to the mass cytometry user community.

A metal-chelating polymer for chelating zirconium and its use in mass cytometry

Imaging mass cytometry (IMC) is a high-dimensional imaging technology that allows the capture and quantification of up to fifty biomarkers at sub-cellular resolution. Each IMC dimension (‘channel’) corresponds to antibodies coupled to metal-tagged antigens and captures the imaging representation of biomarkers for each cell. Conventional IMC analysis relies on manually segmenting the cells across all IMC channels, which is a time-consuming process and prone to human error. Recent advances in computerized IMC analysis, using deep learning techniques such as convolutional neural networks (CNNs), enable automated segmentation by quantifying cellular structures through a data-driven approach. However, existing CNN-based methods concatenate and integrate IMC image channels at an early stage in the learning process, which potentially limiting the model's ability to retain meaningful correlations among different channels for training. In addition, the cluttered nature and inhomogeneous textures of cellular structures may further complicate the training process, making it even more difficult for CNNs to accurately segment the cell boundaries. In this study, we propose a Stacked Channel Learning Network (SCLN), a two-phase CNN-based method for IMC cell segmentation. SCLN uses a channel embedding approach with two training phases: Phase I generates imaging masks for cellular structures using a pre-trained Res-U-Net model, while Phase II refines the segmentation by incorporating these masks as an additional channel alongside the existing biomarker channels. In addition, SCLN preserves the image representations derived independently from each channel, enabling more detailed channel-level feature analysis. The proposed method was evaluated on the widely used IMC breast cancer METABRIC dataset. The experimental results show that our SCLN achieved a Dice coefficient score of 91.45%, which outperformed the existing segmentation methods by ~ 10%, especially for the challenging studies e.g., cluttered cells and cells with inhomogeneous textures. Our codes can be found at https://github.com/kongnet-djd/SCLN.Graphical

Compensation of Signal Spillover in Suspension and Imaging Mass Cytometry