Immunocluster: A Computational Tool to Explore the Immune Profile and Cellular Heterogeneity of Hematological Diseases Using Liquid and Imaging Mass, and Flow Cytometry Data

Introduction Our understanding of 'immunome" and its role in the pathogenesis and clinical outcomes of hematological diseases has significantly improved in recent years, which highlights the importance of multidimensional cytometry techniques in investigating immune response both in clinical and research settings. Liquid mass cytometry (LMC), imaging mass cytometry (IMC) and flow cytometry (FC) are powerful techniques for immunophenotyping, biomarker discovery, and patients' immune-monitoring. These techniques provide the ability to profile over 40 markers per cell, resulting in large amounts of data which could be challenging to analyze and interpret. Therefore, we developed ImmunoCluster, an easy-to-use open-source computational pipeline, to explore high dimensionality single-cell cytometry datasets. Case studies Here we describe three examples of the implementation of the ImmunoCluster tool, which is a self-contained R package (accessed via GitHub: https://github.com/kordastilab/ImmunoCluster). Previously published LMC data from 15 leukemia patients 30 and 90 days after bone marrow transplantation (BMT) were used to test ImmunoCluster's ability to reproduce results. Post-BMT 3/15 patients suffered acute graft versus host disease (GvHD). For IMC data a lymph node section from a diffuse large B-cell lymphoma (DLBCL) patient. Finally, FC data from BM of 7 healthy donors (HDs) taken during hip surgery. The pipeline provides tools to allow researchers to follow a workflow which guides them through experimental design, data analyses and interpretation, to publishable graphics identifying differences in phenotype and abundance of cells between conditions (Figure 1A). ImmunoCluster comprises of three core computational stages: Stage 1. Data import and quality control In the experimental design stage, the high dimensional dataset is imported into ImmunoCluster with an associated metadata file, this included timepoints, and response to treatment. All data was stored within a SingleCellExperiment (SCE) object, a data container in which you can store/retrieve information such as metadata and UMAP/tSNE coordinates (Figure 1B). Initial exploratory visualization of the data, e.g. Multidimensional scaling (MDS) plots, and heatmaps showing marker expression for each patient were created and metadata used for annotation. Stage 2. Dimensionality reduction and unsupervised clustering We used three dimensionality reduction tools: MDS, uniform manifold approximation and projection (UMAP), and t-Distributed Stochastic Neighbor Embedding (tSNE). Two clustering algorithms: an ensemble clustering method of FlowSOM and Consensus clustering; and PhenoGraph. The aim of these algorithms were to assign all cells to clusters corresponding to true cell types. Stage 3. Annotation and differential testing Tools exploring cluster marker expression via projection onto UMAP/tSNE, or heatmaps, aided the identification of cell types and phenotypically distinct clusters. Metadata were used to annotate figures, allowing for visualization of the distribution of cell islands between different conditions/timepoints. Tools such as median marker expression, hierarchical clustered heatmaps, and box plots of cell cluster abundance were applied. Statistically significant differences between conditions were identified. Results We successfully replicated the findings from Hartmann et al., 24 cell populations were identified (Figure 2A-B). Significant differences between memory B-cells (FDR p=4.38 x 10-3), naïve B-cells (FDR p=1.35 x 10-2), and naïve CD4+ T-cells (FDR p=3.47 x 10-2) were identified between the GvHD and none (Figure 2C). From the FC HD CD4+ BM population data we were able to identify regulatory T cells (Tregs), including subpopulations, Treg A and Treg B (as low as 0.7% (0.1-2.0) and 0.9% (0.2-2.4), respectively). A marker expression ranking tool was applied to the DLBCL patient IMC data. We identified the majority of Ki-67 high population were proliferating tumor cells (84%), and the Ki-67 low population consisted of a heterogeneous collection of immune cell populations (Figure 3). These case studies show that ImmunoCluster could help clinicians and researchers with varying experience in computational biology to drive their projects from experimental design, wet lab/clinical trial, through to the data analysis process and visualization. Disclosures McLornan: JAZZ PHARMA: Honoraria, Speakers Bureau; NOVARTIS: Honoraria, Speakers Bureau; CELGENE: Honoraria, Speakers Bureau. Harrison:Gilead Sciences: Honoraria, Speakers Bureau; Incyte Corporation: Speakers Bureau; Janssen: Speakers Bureau; Sierra Oncology: Honoraria; Novartis: Honoraria, Research Funding, Speakers Bureau; AOP Orphan Pharmaceuticals: Honoraria; Shire: Honoraria, Speakers Bureau; Promedior: Honoraria; Roche: Honoraria; Celgene: Honoraria, Research Funding, Speakers Bureau; CTI Biopharma Corp: Honoraria, Speakers Bureau. Kordasti:Celgene: Research Funding; Novartis: Research Funding; Alexion: Honoraria.

Download Full-text

Identification of Functionally Primitive and Immunophenotypically Distinct Subpopulations in Secondary Acute Myeloid Leukemia By Mass Cytometry

Blood ◽

10.1182/blood.v128.22.4278.4278 ◽

2016 ◽

Vol 128 (22) ◽

pp. 4278-4278

Author(s):

Shovik Bandyopadhyay ◽

Liyang Yu ◽

Daniel A.C. Fisher ◽

Olga Malkova ◽

Stephen T. Oh

Keyword(s):

Board Of Directors ◽

Myeloid Leukemia ◽

Research Funding ◽

Cell Types ◽

Cell Populations ◽

Specific Cell ◽

Mass Cytometry ◽

Advisory Committees ◽

Secondary Acute Myeloid Leukemia ◽

Acute Myeloid

Abstract Introduction: Mass cytometry is a powerful tool for analyzing cellular networks, with the ability to generate massive data sets encompassing > 40 parameters measured simultaneously at the single cell level. Various groups have created a variety of platforms to analyze this high dimensional data in unique and efficient ways. These tools have a range of applications: from using phenotypic similarities to cluster cells, stratifying unique signaling subpopulations based on observed stimulation responses, mapping the developmental trajectory of cell types, among many others. We have previously utilized mass cytometry to characterize NFkB hyperactivation in myeloproliferative neoplasms. Here we applied mass cytometric analysis to a cohort of patients with secondary acute myeloid leukemia (sAML) following a history of chronic MPN. The objective of this work was to identify populations of functionally primitive leukemic cells, relying not only on traditional immunophenotypic designations (which can vary considerably in sAML), but also by inferring functional status based on the presence or absence of cytokine hypersensitivity and constitutively active signaling in specific cell populations. Results: Dimensionality reduction and clustering analysis by viSNE and SPADE identified multiple cell subsets outside the hematopoietic stem/progenitor cell (HSPC) compartment that exhibited overt thrombopoietin (TPO) sensitivity, while healthy controls had highly localized responses largely restricted to the HSPC compartment. Using Phenograph, ten sAML metaclusters were identified containing cells from six sAML patients analyzed. One of these metaclusters represented a distinct subpopulation of CD61+ CD34- CD38- CD45lo cells with variable CD90 and CD11b expression. This subpopulation of CD61+ cells was not identified by manual gating, and exhibited significantly greater STAT3/STAT5 phosphorylation in response to TPO than did lineage-negative CD34+ CD38- cells in five out of six (83%) AML patients examined. In addition, substantially elevated basal STAT3 phosphorylation in this population was hypersensitive to TPO and largely resistant to ex vivo ruxolitnib. The classify function of Phenograph was utilized to determine whether the cytokine hypersensitivity observed in the viSNE and SPADE analysis could be entirely accounted for by the aforementioned CD61+ CD34- CD38- CD45lo population. The signaling responses highly predictive of specific cell types were identified, which were used to assess the functional status of sAML cells compared to healthy Lin- CD34+ CD38- cells. By this approach, sAML cells were found to exhibit significant incongruity between surface cell type designation and functional designation. Furthermore, functionally primitive cells displayed a spectrum of myeloid surface markers, suggesting that restricting analysis to a subset of strictly surface-defined cells would potentially obscure populations of interest. Conclusions: Our analysis revealed a distinct, previously undescribed population of CD61+ CD34- CD38- CD45lo cells in sAML. While the biological relevance of this population requires validation by functional assays, this result demonstrates that immunophenotypic changes in traditional surface-marker-defined populations may conceal important cell populations. These cells, and other functionally primitive but mature-designated cells could be relevant to studying sAML disease pathogenesis, progression, and/or response to therapy. This study further demonstrates the potential for mass cytometry to elucidate rare leukemic subpopulations in highly heterogeneous tumors. Disclosures Oh: Gilead: Membership on an entity's Board of Directors or advisory committees, Research Funding; Incyte Corporation: Membership on an entity's Board of Directors or advisory committees, Research Funding; Janssen: Research Funding; CTI: Research Funding.

Download Full-text

Immune Microenvironment Analysis of Bone Marrow By Mass Cytometry and RNA Sequencing in Multiple Myeloma Patients Treated with Daratumumab and Durvalumab

Blood ◽

10.1182/blood-2018-99-114453 ◽

2018 ◽

Vol 132 (Supplement 1) ◽

pp. 3296-3296 ◽

Cited By ~ 1

Author(s):

Frances Seymour ◽

Mary H Young ◽

Mark Tometsko ◽

Jamie Cavenagh ◽

Ethan G. Thompson ◽

...

Keyword(s):

Bone Marrow ◽

Flow Cytometry ◽

T Cells ◽

Nk Cells ◽

Research Funding ◽

Cell Populations ◽

Mass Cytometry ◽

Immune Microenvironment ◽

Employment Equity ◽

Equity Ownership

Abstract Introduction Relapsed and refractory multiple myeloma (RRMM) remains a challenging disease to treat due to its heterogeneity and complexity. There is an urgent need for novel combination strategies, including immunotherapy. The study of the tumour and immune microenvironment before and after treatment with combination therapy is a crucial part of understanding the underpinning of disease response. Methods Longitudinal samples of bone marrow aspirates and whole blood were collected from a phase II clinical trial, MEDI4736-MM-003 (NCT02807454) where daratumumab and durvalumab naïve patients were exposed simultaneously to both these drugs. A combination of mass cytometry (CyTOF), RNAseq and flow cytometry were performed on a subset of samples from these subjects. Specifically, paired bone marrow mononuclear cells (BMMC) samples from nine patients taken at screening and six weeks post-treatment were analysed by mass cytometry (CyTOF) using a 37-marker pan-immune panel that included both lineage and functional intracellular/extracellular markers. In addition, whole blood sample specimens were collected at screening and on treatment (8, 15, 30, and 45 days after treatment) and analysed by flow cytometry. Flow cytometry panels were designed to allow interrogation of the abundance and activation status of immune cell subsets. Finally, RNA from bone marrow aspirates at screening and C2D15 were analysed by RNA sequencing. Expression profiles from the aspirates were used to estimate cell proportions by computational deconvolution. Individual cell types in these microenvironments were estimated using the DCQ algorithm and a gene expression signature matrix based on the published LM22 leukocyte matrix (Newman et al., 2015) augmented with 5 bone marrow- and myeloma-specific cell types. Results In a heavily pre-treated population with RRMM, treatment with durvalumab and daratumumab leads to shifts in a number of key immunological populations when compared to pre-treatment. In the bone marrow, CD8 and CD4 populations rise (by CyTOF and RNAseq), while NK, DC and B cell populations fall (by CyTOF). In the bone marrow within CD8+ T lymphocyte populations, we observed a post-treatment rise in markers of degranulation (granzyme p=0.0195, perforin p=0.0078, Wilcoxon signed-rank test). This is also accompanied by a fall in PD1 expression (p=0.0078) and rise in the co-stimulatory receptor DNAM1 (p=0.0273). These changes are most marked on cells with an effector memory CD45RA+ CD8+ T cell phenotype. In the blood, similar to the bone marrow, CD8+ T cells proliferate over the course of treatment (flow cytometry). A fall in both naïve and active NK cell populations is seen following treatment in bone marrow. NK cells express high levels of CD38 and are therefore depleted by daratumumab. Those NK cells which remain have an active phenotype with increased expression of TNFa (p=0.0039) and IFNg (p=0.0195) following treatment. Across the time points sampled in peripheral blood, NK cells were also decreased and those that remained were proliferating. Dendritic cells with a tolerogenic phenotype can be identified prior to treatment and are seen to fall in abundance following treatment with durvalumab and daratumumab. Conclusions The combination of durvalumab and daratumumab leads to several immune microenvironment changes that biologically portend clinical effect. We see increases in the abundance of cell populations with functional anti-tumour activity, including granzyme B+ CD8 T cells and a reduction in PD1high T cells. Despite the treatment expectedly reducing NK cell numbers, many functionally competent NK cells remain, as evidenced by the presence of anti-tumour cytokines. This combination strategy also reduces immunosuppressive tolerogenic DCs, which suppress CD4 and CD8 T cell activity. Taken together, this suggests that this chemotherapy free, doublet treatment has the potential to up-regulate anti-tumour immunological responses, which may restore immunosurveillance mechanisms critically needed in these highly refractory patients. Disclosures Seymour: Celgene: Research Funding. Young:Celgene Corporation: Employment, Equity Ownership. Tometsko:Celgene Corporation: Employment, Equity Ownership. Cavenagh:Celgene: Honoraria, Research Funding, Speakers Bureau; Janssen: Honoraria, Speakers Bureau; Takeda: Research Funding, Speakers Bureau; Novartis: Honoraria, Speakers Bureau; Amgen: Honoraria, Speakers Bureau. Thompson:Celgene Corporation: Employment, Equity Ownership. Whalen:Celgene Corporation: Employment, Equity Ownership. Danziger:Celgene Corporation: Employment, Equity Ownership. Fitch:Celgene Corporation: Employment, Equity Ownership. Fox:Celgene Corporation: Employment, Equity Ownership. Dervan:Celgene Corporation: Employment, Equity Ownership. Foy:Celgene Corporation: Employment, Equity Ownership. Newhall:Celgene Corporation: Employment, Equity Ownership. Gribben:Acerta Pharma: Honoraria, Research Funding; Cancer Research UK: Research Funding; TG Therapeutics: Honoraria; Roche: Honoraria; NIH: Research Funding; Medical Research Council: Research Funding; Celgene: Consultancy, Honoraria, Research Funding; Abbvie: Honoraria; Kite: Honoraria; Pharmacyclics: Honoraria; Novartis: Honoraria; Janssen: Honoraria, Research Funding; Wellcome Trust: Research Funding; Unum: Equity Ownership.

Download Full-text

ROR-1 Is a Highly Discriminative Marker in Flow Cytometric Minimal Residual Disease (MRD) Detection in Chronic Lymphocytic Leukemia (CLL)

Blood ◽

10.1182/blood.v128.22.3197.3197 ◽

2016 ◽

Vol 128 (22) ◽

pp. 3197-3197 ◽

Cited By ~ 2

Author(s):

Michaela Patz ◽

Barbara Pentok ◽

Kathrin Cremer ◽

Stefanie Linnartz ◽

Esther Lilienweiss ◽

...

Keyword(s):

Flow Cytometry ◽

B Cell ◽

Research Funding ◽

Residual Disease ◽

Lymphocytic Leukemia ◽

Cell Populations ◽

Discrimination Analysis ◽

Flow Cytometric ◽

Correct Discrimination ◽

Minimal Residual

Abstract Introduction:With the advent of new potent therapies for chronic lymphocytic leukemia (CLL) minimal residual disease (MRD) detection becomes increasingly important to assess remission depth. While molecular MRD detection for CLL remains laborious and time consuming flow cytometry is a fast, economic and sensitive method in detecting low frequencies of CLL cells. The usefulness of the antigens CD81, CD5, CD20, CD43 and CD79b has been previously described for this purpose. ROR-1 has recently been identified as a signature gene in CLL and mantle cell lymphoma. The potential utility of ROR-1 in flow cytometric minimal residual cell analysis has not been evaluated yet. Methods: 10 normal samples and 77 remnants of randomly selected samples from diagnosed patients undergoing CLL therapy were analyzed by flow cytometry. A customized dry formulation of an antibody panel was used, comprising antibodies directed against CD5, CD19, CD20, CD43, CD45, CD79b, CD81 and ROR-1 (DuraClone RE CLB). Linearity, repeatability and inter-operator variability of data analysis of the method were examined. B cell populations comprising at least 50 positive events (46 normal B cell populations, 25 CLL populations, paired and unpaired) were analyzed for their expression profile as assessed by respective mean fluorescence intensities of the antibody labels within classified populations. The expression profiles were subject to supervised discrimination analysis (DA). Results: Between124,000 and 2,122,000 (683,000 ± 450,000) CD45+ events were acquired from the 87 samples. The background of cells with a CLL-like phenotype in the normal samples was determined as <0.001% of CD45+ events. Linearity was confirmed in the range from 1% to 0.0025%. The Repeatability analysis and the inter-operator variability showed concordance with typical Poisson distribution characteristics. The 46 populations with a typical normal B cell phenotype ranged from 0.014% to 9.592% with an average of 2.45% ± 2.75 of CD45+ events. The 25 populations with a classical or non-classical CLL phenotype ranged from 0.007% to 5.459% with an average of 1.41% ± 1.65 of CD45+ events. Posterior discrimination analysis revealed 100% correct discrimination for CLL populations and 96% correct discrimination for normal populations when relying on ROR-1 expression alone in CD19+CD45+ B cells. This result was only surpassed by the complete antibody combination (100% / 100%) but not by any other of the markers, neither in single use nor in combination Conclusion: The 8-color dry flow cytometry panel comprising CD5, CD19, CD20, CD43, CD45, CD79b, CD81 and ROR-1 demonstrated sensitive, linear and specific detection of residual CLL cells in a relevant low range of frequency. ROR-1 revealed to be a highly discriminative marker in the analysis of residual CLL cells by flow cytometry. Utilizing this flow cytometry approach, MRD detection showing sensitivity comparable to molecular techniques can be achieved in CLL. Disclosures Hallek: AbbVIe: Consultancy, Honoraria; Mundipharma: Consultancy, Honoraria; Glaxo-SmithKline: Consultancy, Honoraria; Gilead: Consultancy, Honoraria; Janssen: Consultancy, Honoraria, Speakers Bureau; Pharmacyclics: Consultancy, Speakers Bureau; Celgene: Consultancy, Honoraria; Roche: Consultancy, Research Funding, Speakers Bureau. Kreuzer:Gilead Sciences: Consultancy, Honoraria, Research Funding, Speakers Bureau; Roche Pharma GmbH and Mundipharma GmbH: Consultancy, Honoraria, Research Funding, Speakers Bureau.

Download Full-text

Simultaneous Monitoring of Drug Responses on Distinct Hematopoietic Cell Populations Allow Assessment of Direct and Indirect Cytotoxic Effects of Targeted Therapies

Blood ◽

10.1182/blood.v128.22.3515.3515 ◽

2016 ◽

Vol 128 (22) ◽

pp. 3515-3515

Author(s):

Muntasir M Majumder ◽

Aino Maija Leppä ◽

Caroline A Heckman

Keyword(s):

T Cells ◽

B Cells ◽

Nk Cells ◽

Research Funding ◽

Plasma Cells ◽

Cell Types ◽

Cell Populations ◽

Specific Cell ◽

Drug Responses ◽

Nk T Cells

Abstract Introduction Off-target cytotoxicity resulting in severe side effects and compromising patient survival often hampers the development of new cancer therapeutics. Understanding the complete drug response landscape of different cell populations is crucial to identify drugs that selectively eradicate the malignant cell population, but spare healthy cells. Here, we developed a high content, no wash, multi-parametric flow cytometry based assay that enables testing of blood cancer patient samples and simultaneously monitors the effects of several drugs on 11 hematopoietic cell types. The assay can be used to i) dissect malignant from healthy cell responses and predict off-target effects; ii) assess drug effects on immune cell subsets; iii) identify drugs that can potentially be repositioned to new blood cancer indications. Methods Mononuclear cells were prepared from bone marrow aspirates of 7 multiple myeloma (MM) and 3 acute myeloid leukemia (AML) patients plus the peripheral blood from a healthy donor, which were collected following informed consent and in compliance with the Declaration of Helsinki. Optimal cell density, antibody dilutions, incubation time, and wash versus no wash assay conditions for the selected antibody panels were determined. Cells were incubated at a density of 2 million cells/ml in either 96- or 384-well plates for 3 days. The antibodies were tested in two panels to study the effects of 6 drugs in 5 dilutions (1-10000 nM) (clofarabine, bortezomib, dexamethasone, navitoclax, venetoclax and omipalisib) on 11 cell populations, namely hematopoietic stem cells (HSCs) (CD34+CD38-), common progenitor cells (CPCs) (CD34+CD38+), monocytes (CD14+), B cells (CD45+CD19+), cytotoxic T cells (CD45+CD3+CD8+), T helper cells (CD45+CD3+CD4+), NK-T cells (CD45+CD3+CD56+), NK cells (CD45+CD56+CD3-), clonal plasma cells (CD138+CD38+), other plasma cells (CD138+CD38-) and granulocytes (CD45+, SSC++). Annexin-V and 7AAD were used to distinguish live cell populations from apoptotic and dead cells. After 1 h incubation with antibodies, the plates were read with the iQue Screener PLUS instrument (Intellicyt). Counts for each population were used to generate four parameter nonlinear regression fitted dose response curves with GraphPad Prism 7. Three samples were tested in duplicate to assess reproducibility. Results To decrease the complexity of the assay, we tested all antibodies under wash and no wash conditions, and found that results from both conditions were comparable. To minimize the amount of sample needed as well as maximize the number of drugs tested and cell populations that can be detected, we set up the assay in both 96- and 384-well plates. The assay was highly reproducible when samples were tested in replicate and was scalable to a 384-well format without compromising sensitivity to detect rare populations such as plasma cells. Due to the differentiation of immature cells to specialized cell types, the drug responses of specific populations tended to drift. HSCs (CD34+CD38-) were shown to be refractory to the tested drugs compared to CPCs characterized as (CD34+CD38+) and other cell types. Interestingly, the proteasome inhibitor bortezomib was cytotoxic to all cell populations except for CD138+CD38- plasma cells. Clofarabine, a nucleoside analog used to treat ALL, effectively targeted CPC, NK and B cells, while HSCs and plasma cells were resistant. The glucocorticoid and immunosuppressive drug dexamethasone specifically targeted B and NK cells compared to T cell populations (CD8+, CD4+), while NK-T cells were modestly sensitive. The cell population response patterns were similar in samples derived from MM, AML and healthy individuals, highlighting that the drug responses are highly cell type specific. Summary Using a high content, multi-parametric assay, we could rapidly assess the effect of several drugs on specific cell populations in individual patient samples. Our results demonstrate that many drugs preferentially affect different hematological cell lineages. Although heterogeneity was observed between individual patients, the pattern of cytotoxic response exhibited by specific cell types was consistent among samples derived from MM, AML and healthy donors. The assay will be useful to identify drugs with maximal on-target and minimal off-target specificity, and can potentially be used to guide treatment decision and predict patient response Disclosures Heckman: Celgene: Research Funding; Pfizer: Research Funding.

Download Full-text

The Ribomethylome Landscape of Hematopoietic System

Blood ◽

10.1182/blood-2020-136522 ◽

2020 ◽

Vol 136 (Supplement 1) ◽

pp. 41-42

Author(s):

Yanyan Wang ◽

Christian Rohde ◽

Fengbiao Zhou ◽

Marco Hennrich ◽

Laura Poisa-Beiro ◽

...

Keyword(s):

Bone Marrow ◽

Research Funding ◽

Cell Types ◽

Cell Populations ◽

Spearman Correlation ◽

Cell Type ◽

Hematopoietic Stem ◽

Healthy Donors ◽

Cell Type Specific ◽

Cellular Identity

Introduction: RNA modifications are emerging as important determinants of cell identity and cell fate. Small nucleolar RNAs (snoRNA) guide pseudouridylation and 2'-O-methylation of RNA species. C/D box snoRNAs are essential for AML1/ETO-induced leukemia (Zhou et al. Nat Cell Biol 2017). Dynamics and relevance of these modifications in hematopoiesis are unknown. Here, we aimed to determine the plasticity of ribosomal 2'-O-methylation (Ribomethylome) patterns in hematopoietic cell populations and the interdependence with snoRNA expression, transcriptomics and proteomics. Methods: Healthy donors (19-86yrs) donated bone marrow and six cell populations were sorted or prepared: Hematopoietic stem/progenitor cell (HPC), Monocyte/macrophage precursor (MON), Granulocytic precursor (GRA), Erythroid precursor (ERY), Lymphocyte progenitor (LYM), and Mesenchymal stem/stromal cell (MSC). Small RNA sequencing and Ribometh-seq data were obtained for 65 and 55 samples, respectively. Data were analyzed together with accompanying RNA-seq and Mass-spec proteomics data which were available for all the specimens. Bioinformatics analyses were based on PCA, tSNE, spearman correlation, paired t-test, GSEA and ANOVA. Results: The analyses of 2'-O-methylation (Ribomethylome) in six bone marrow cell types from healthy donors revealed that ribosomal modifications occurred different during the process of hematopoietic differentiation. Among these sides, HPC and Myeloid lineage showed significant variability between different cell populations. Ribomethylome patterns differed between cell types and PCA analyses indicated that cellular identity was matched with a specific Ribomethylome pattern. Plasticity in Ribomethylomes were most evident for HPC, LYM, GRA and MON which showed high levels of 2'-O-methylation (almost 100% of rRNA methylated) whereas methylation levels in MSC cells were much lower (Spearman correlation<0.4). These findings indicated that Ribomethylome patterns were cell type specific. Using snoRNA sequencing, we showed that snoRNA expression levels differed between the different cell types. C/D box snoRNAs were variably expressed, and the expression differences for SNORD68 and SNORD87 were associated with respective Ribomethylome changes of predicted target sites. We next analyzed the association between specific 2'-O-methylation levels and the levels of protein expression. Only those proteins were included for whom no association between mRNA and total protein levels were observed. Spearman rank analyses suggested that RAB7A, PSME1 involved "antigen processing and presentation" and FLNA, RCC2 involved "cell migration" correlated closely with 2'-O-methylation of the dynamically regulated sites 28S_3723_SNORD87 and 5.8S_14_SNORD71. Conclusion: Our finding based on multi-omics analyses identifies cell type specific Ribomethylomes. Myeloid differentiation is associated with specific Ribomethylome changes. Distinct Ribomethylomes may contribute to cellular identity by directing translation of specific sets of mRNAs. Figure 1: The effects of ribomethylome and protein translation were evident and separated by different cell populations(tSNE). Figure 1 Disclosures Müller-Tidow: Daiichi Sankyo: Research Funding; BiolineRx: Research Funding; Janssen-Cilag GmbH: Speakers Bureau; Pfizer: Research Funding, Speakers Bureau.

Download Full-text

Predicting cell types in single cell mass cytometry data

10.1101/316034 ◽

2018 ◽

Cited By ~ 1

Author(s):

Tamim Abdelaal ◽

Vincent van Unen ◽

Thomas Höllt ◽

Frits Koning ◽

Marcel J.T. Reinders ◽

...

Keyword(s):

Single Cell ◽

High Performance ◽

Cell Mass ◽

Cell Types ◽

Cell Populations ◽

Cell Type ◽

Mass Cytometry ◽

Linear Discriminant ◽

Linear Classifier ◽

Benchmark Datasets

AbstractMotivationMass cytometry (CyTOF) is a valuable technology for high-dimensional analysis at the single cell level. Identification of different cell populations is an important task during the data analysis. Many clustering tools can perform this task, however, they are time consuming, often involve a manual step, and lack reproducibility when new data is included in the analysis. Learning cell types from an annotated set of cells solves these problems. However, currently available mass cytometry classifiers are either complex, dependent on prior knowledge of the cell type markers during the learning process, or can only identify canonical cell types.ResultsWe propose to use a Linear Discriminant Analysis (LDA) classifier to automatically identify cell populations in CyTOF data. LDA shows comparable results with two state-of-the-art algorithms on four benchmark datasets and also outperforms a non-linear classifier such as the k-nearest neighbour classifier. To illustrate its scalability to large datasets with deeply annotated cell subtypes, we apply LDA to a dataset of ~3.5 million cells representing 57 cell types. LDA has high performance on abundant cell types as well as the majority of rare cell types, and provides accurate estimates of cell type frequencies. Further incorporating a rejection option, based on the estimated posterior probabilities, allows LDA to identify cell types that were not encountered during training. Altogether, reproducible prediction of cell type compositions using LDA opens up possibilities to analyse large cohort studies based on mass cytometry data.AvailabilityImplementation is available on GitHub (https://github.com/tabdelaal/CyTOF-Linear-Classifier)[email protected]

Download Full-text

Immune System Profiling of Waldenstrom Macroglobulinemia (WM) and Immunoglobulin M Monoclonal Gammopathy of Undetermined Significance (IgM MGUS) Using Mass Cytometry (CyTOF)

Blood ◽

10.1182/blood-2018-99-117568 ◽

2018 ◽

Vol 132 (Supplement 1) ◽

pp. 4138-4138

Author(s):

Jonas Paludo ◽

Jose Villasboas Bisneto ◽

Anne J Novak ◽

Morie A. Gertz ◽

Prashant Kapoor ◽

...

Keyword(s):

Immune System ◽

B Cells ◽

Single Cell ◽

Monoclonal Gammopathy ◽

Research Funding ◽

Cell Types ◽

Immunoglobulin M ◽

Mass Cytometry ◽

Entire Cohort ◽

Undetermined Significance

Abstract Introduction Immunoglobulin M monoclonal gammopathy of undetermined significance (IgM MGUS) is a relatively common pre-malignant condition that may progress to Waldenstrom macroglobulinemia (WM), an incurable low-grade non-Hodgkin lymphoma. The acquisition of genomic alterations in a multistep process of progression is likely a central mechanism responsible for the progression of IgM MGUS to WM, leading to increased tumor burden and end-organ damage. The exact role of the immune system in this process is still poorly characterized. The modulation of the immune system by malignant cells leading to anti-lymphoma immunity or immune evasion is a well-stablished principle associated with lymphoma survival and progression, respectively. The understanding of the pathobiology of IgM MGUS and WM has grown significantly over the last few years; however, most of the past research has focused on genetics and flow cytometric analysis of malignant cells. The introduction of CyTOF, a novel platform coupling mass spectometry with single-cell flow cytometry using antibody-metal isotope pairs, has allowed a broader single-cell analysis and the study of functional profiles. We here report the use of CyTOF technology to comprehensively profile the immune system of patients with IgM MGUS and WM. Methods Viably-cryopreserved (DMSO 10%) single-cell suspensions containing peripheral blood mononuclear cells from 15 patients (WM n=9 and IgM MGUS n=6) were used in this study. Recovered cells (1-3 x 106 per sample) were stained with a 30-parameter surface protein CyTOF panel designed to characterize multiple cell types and profile the immune system. Nucleated cellular events were identified using a DNA intercalator conjugated to natural abundance iridium (191Ir and 193Ir). Cisplatin (195Pt) was used for dead-live cell discrimination and calibration beads containing natural abundance cerium (140/142Ce), europium (151/153Eu), holmium (165Ho), and lutetium (175/176Lu) were used for normalization of the instrument signal. CyTOF analysis of normalized files was carried out using the Astrolabe Diagnostics platform. Cell clusters were assigned using the FlowSOM package and differential abundance analysis was performed using the edgeR package. Analysis is adjusted for multiple comparisons. Results The median live single-cell count for the entire cohort was 573,300 (range 50,681 to 869,727), representing a median of 81% (range 39% to 87%) of the total event counts [median 820,800 (range 130,452 to 1,000,000]. The cell count and the proportion of event count was similar in patients with WM or IgM MGUS. T cells represented 45.8% of the total events; granulocytes represented 31.3%; NK cells represented 8.6%; monocytes represented 8.1%; B cells represented 2.4% of the total events for the entire cohort. Figure 1 shows a heat map of different immune cell sub-types proportion per patient sample. A difference was noted in the proportion of monocytes (CD14+, CD16+) and B-cells when comparing patients with IgM MGUS and WM (figure 2). Patients with WM had a higher proportion of monocytes compared to IgM MGUS patients (p=0.05). A higher proportion of B-cells was noted in patients with IgM MGUS when compared to WM (4.2% vs 0.6%, p=0.01). When B-cells were sub-classified as naïve, transitional, non-switched memory, switched memory, and plasmablasts, the proportion of non-switched memory B-cells was higher in IgM MGUS compared to WM (p=0.05). While the results of this pilot study suggest important differences between IgM MGUS and WM as regards the immune repertoire, confirmatory studies are in progress in a larger cohort of patients. Conclusion To our knowledge, this is the first study of mass cytometry in patients with WM and IgM MGUS patients. A higher proportion of monocytes and lower proportion of memory B-cells was noted in patients with WM compared to IgM MGUS. CyTOF study of the immune profile in these patients is feasible and can potentially uncover relationships between cell types not typically associated with the disease progression continuum (IgM MGUS to WM). While definitive conclusion cannot be made using this dataset due to the small sample size, CyTOF is a powerful tool in the study of the immune profile of WM patients. Disclosures Gertz: Amgen: Consultancy; Abbvie: Consultancy; Apellis: Consultancy; annexon: Consultancy; spectrum: Consultancy, Honoraria; Teva: Consultancy; Alnylam: Honoraria; Physicians Education Resource: Consultancy; Ionis: Honoraria; janssen: Consultancy; celgene: Consultancy; Prothena: Honoraria; Research to Practice: Consultancy; Medscape: Consultancy. Kapoor:Celgene: Research Funding; Takeda: Research Funding. Ailawadhi:Takeda: Consultancy; Amgen: Consultancy; Janssen: Consultancy; Celgene: Consultancy; Pharmacyclics: Research Funding. Reeder:Affimed: Research Funding. Ansell:Seattle Genetics: Research Funding; Regeneron: Research Funding; Takeda: Research Funding; Pfizer: Research Funding; Merck & Co: Research Funding; Trillium: Research Funding; Affimed: Research Funding; LAM Therapeutics: Research Funding; Celldex: Research Funding; Bristol-Myers Squibb: Research Funding.

Download Full-text

Mass Cytometry Analysis of Dysregulated Cytokine Production and Intracellular Signaling in Myelofibrosis

Blood ◽

10.1182/blood.v128.22.4277.4277 ◽

2016 ◽

Vol 128 (22) ◽

pp. 4277-4277 ◽

Cited By ~ 1

Author(s):

Daniel A.C. Fisher ◽

Cathrine A. Miner ◽

Elizabeth K. Engle ◽

Taylor M. Brost ◽

Olga Malkova ◽

...

Keyword(s):

Board Of Directors ◽

Research Funding ◽

Cytokine Production ◽

Jak2 V617f ◽

Cell Populations ◽

Mass Cytometry ◽

Advisory Committees ◽

Downstream Signaling ◽

Cd34 Cells ◽

Multiple Cell

Abstract Myeloproliferative neoplasms including myelofibrosis (MF) are characterized by a malignant clone containing JAK2 V617F or other mutations leading to unregulated JAK2 kinase activity. MF is characterized by anemia, splenomegaly, bone marrow fibrosis, inflammatory cytokine production, and a propensity for transformation to secondary acute myeloid leukemia. Inhibition of JAK2 with ruxolitinib improves constitutional symptoms and splenomegaly and lowers circulating plasma cytokine levels. However, improvements in anemia, fibrosis, and malignant clonal burden are infrequent. These observations illustrate the need for improved therapy for MF. The objective of this study is to better understand the relationship between dysregulated cytokines and downstream signaling in MF, with the goal of determining how these pathways can be more effectively manipulated for therapeutic benefit. To interrogate altered signaling in MF, we have employed mass cytometry (CyTOF), a method which enables the quantitative analysis of signaling throughout hematopoiesis. In our survey of signaling in MF, hyperactivation of the NFκB signaling pathway was found to be widespread. This finding was corroborated by gene set enrichment analysis (GSEA) of a published gene expression dataset of CD34+ cells from MF versus normal controls (Norfo et al. 2014 Blood). Evidence for NFκB hyperactivation, both by mass cytometry and GSEA, was strongest in JAK2V617F-mutant MF patients. This supports a hypothesis that pronounced NFκB signaling hyperactivation is a consequence of mutant JAK2. In MF, hyperactivated NFκB signaling was widespread among hematopoietic cell populations, including T cells. This distribution suggests that NFκB activation may be partly driven by non-cell-autonomous mechanisms. Consistent with previous studies, plasma TNFα levels were found to be elevated in these patients, suggesting that excessive production of TNFα could result in downstream activation of NFκB across multiple cell populations in a non-cell-autonomous fashion. To further elucidate the etiology of systemic NFκB hyperactivation and understand the interplay of inflammatory cytokines and downstream signaling, we extended our mass cytometry approach to study the cellular distribution of cytokine production in MF. A panel of 22 surface marker antibodies and 12 cytokine antibodies was developed for these experiments. Examination of peripheral blood from two JAK2V617F-mutant MF patients revealed that intracellular levels of several cytokines were constitutively elevated in both MF patients compared to healthy controls. Monocytes produced the highest levels of TNFα among hematopoietic populations, and these were higher in MF versus control monocytes. Supranormal cytokine expression was accentuated by stimulation with PMA/ionomycin or TLR ligands R848 or PAM3CSK4. Incubation with TNFα led to supranormal levels of the cytokines MIP1β and IL-6, in monocytes from one or both patients. Therefore, abnormal production of TNFα by MF patient monocytes could result in overproduction of IL-6, and MIP1β in the same cells. PMA/ionomycin led to above normal production of TNFα from MF Lin-CD34+ cells and CD33+CD34- immature myeloid cells, suggestive that these cells could be hypersensitive to pathophysiologic signaling stimulations in a manner resulting in elevated cytokine production. MF patient T cells also showed hypersensitivity to PMA/ionomycin stimulation, compared with controls, in their production of IFNγ and MIP1β. These cytokines, along with the MF monocyte-overexpressed cytokines TNFα, IL-6, and MIP1β, were found to be elevated in MF patient plasma, consistent with prior studies. These findings imply that multiple cell populations in JAK2 V617F-mutant MF patients overexpress inflammatory cytokines and are hypersensitive to inflammatory insults. The upregulation of cytokines is likely to underlie the systemic hyperactivation of NFκB signaling observed in MF, and could generate non-cell-autonomous effects on the malignant myeloid clone. While NFκB phosphorylation responses to TNFα appear strongest in Lin-CD34+ cells, other cytokines may mediate signaling abnormalities across a variety of cell types. Future experiments will attempt to identify signaling effects of multiple elevated cytokines, which may underlie features of MF that persist despite JAK2 inhibitor therapy. Disclosures Oh: CTI: Research Funding; Janssen: Research Funding; Gilead: Membership on an entity's Board of Directors or advisory committees, Research Funding; Incyte Corporation: Membership on an entity's Board of Directors or advisory committees, Research Funding.

Download Full-text

Evaluation of Cell Types and Morphologies in Sickle Cell Disease with an Imaging Flow Cytometer

Blood ◽

10.1182/blood.v126.23.972.972 ◽

2015 ◽

Vol 126 (23) ◽

pp. 972-972 ◽

Cited By ~ 1

Author(s):

Hasan Tahsin Ozpolat ◽

Tim Chang ◽

Junmei Chen ◽

Xiaoping Wu ◽

Colette Norby ◽

...

Keyword(s):

Flow Cytometry ◽

Sickle Cell ◽

Cell Morphology ◽

Research Funding ◽

Hemoglobin Concentration ◽

Cell Types ◽

Blood Film ◽

Hypotonic Solution ◽

Dense Cell ◽

Activation Status

Abstract Sickle cell disease (SCD) is a hemoglobinopathy characterized by vaso-occlusive episodes and hemolysis. Hemoglobin S is prone to polymerize at low oxygen tension, causing the red cell to become sickle shaped, more rigid and sticky. Evaluation of blood cell morphology, counts and activation are important components of the patient evaluation. This is usually accomplished by evaluation of the blood film, performing a complete blood count (CBC), and with the use of flow cytometry. A typical blood film from an SCD patient shows anisocytosis, poikilocytosis, polychromasia, nucleated erythrocytes, sickled cells, and irregular contracted cells. The methods of blood cell evaluation all have disadvantages. Preparation of the blood film is laborious and its evaluation is highly subjective and requires extensive experience. Some CBC counters (e.g., Siemens - ADVIA 2120) are able to detect dense cells (increased hemoglobin content-high MCHC cells) by their volume and hemoglobin concentration after the red blood cells (RBC) are swelled to spheres with a hypotonic solution. Dense cells resist becoming spheres and are detected by their low volume and high hemoglobin concentration. However, the number of dense cells might be underestimated because reversibly sickled cells are capable of undergoing the sphering and will not be detected. In addition, the hypotonic solution can lyse the cells. Finally, RBC counters cannot detect cells on the basis of specific cell markers, which can be used to define cell types and cell morphology and activation status (platelets). Conventional flow cytometry can detect cell markers, but yields little information on morphology and cannot detect dense cells. Here, we used the ImagestreamX Flow Cytometer (Amnis) to analyze SCD blood. In addition to providing information available with conventional cytometers, this instrument provides an image of each cell analyzed, thus allowing for detailed morphological assessment of a large population of cells. We analyzed 5 patients. All were outpatients not suffering from acute complications. Blood was collected by venipuncture into citrate anticoagulant, stained with antibodies or other reagents, and then fixed in 4% paraformaldehyde. We evaluated the blood for cell numbers and morphology, reticulocytes, dense cells, platelet-monocyte aggregates, phosphatidylserine exposure, and platelet activation status. The blood from all of the SCD patients displayed characteristics not found in control blood. We could clearly distinguish RBC morphologies corresponding to sickle cells, dense cell and reticulocytes. Reticulocytes, identified by CD71 positivity, often displayed a "puckered" morphology, as previously seen in electron micrographs. We calculated the percentage of RBCs that were sickled based on the shape ratio of > 2 (length along the long axis/maximum thickness along the short axis). The sickle cell percentage was 1.4±0.5% (normal 0%) out of total RBC population. We also evaluated dense cell morphology after separating the cells on a percoll density gradient. The cells appeared flattened and "deflated", clearly indicating their loss of intracellular fluid. We also analyzed platelet activation status based on staining for P-selectin, the activated form of integrin aIIbb3 (PAC-1 antibody), and phosphatidylserine exposure. Platelets staining positively for these markers also demonstrated morphological evidence of activation: shape change and filopodia extension. Platelet-monocyte aggregates were higher in all of the patients than in controls (0.036% vs 0%) and were easily distinguished from coincident events by morphology. The number of platelets associated with individual monocytes varied from 1 to 3. Other heterotypic cell aggregates were rare. In summary, imaging flow cytometry provides a powerful tool for the analysis of blood in SCD. The technique allows cell population analysis like conventional cytometry, while yielding detailed morphological information on many parameters of relevance in the disease. Further, the morphological assessment avoids many of the potential artifacts arising from blood film preparation and allows an unbiased assessment of the results. Disclosures Konkle: Baxalta: Consultancy, Research Funding; Biogen: Consultancy, Research Funding; CSL Behring: Consultancy, Other: IDMC chair; Pfizer: Other: IDMC member; Octapharma: Research Funding; Novo Nordisk: Consultancy.

Download Full-text

High-Dimensional Single-Cell Quantitative Profiling of Skeletal Muscle Cell Population Dynamics during Regeneration

Cells ◽

10.3390/cells9071723 ◽

2020 ◽

Vol 9 (7) ◽

pp. 1723

Author(s):

Lucia Lisa Petrilli ◽

Filomena Spada ◽

Alessandro Palma ◽

Alessio Reggio ◽

Marco Rosina ◽

...

Keyword(s):

Skeletal Muscle ◽

Single Cells ◽

Cell Types ◽

Interstitial Space ◽

Cell Populations ◽

Wild Type ◽

Mass Cytometry ◽

Pathological Conditions ◽

Comprehensive Picture ◽

Quantitative Profiling

The interstitial space surrounding the skeletal muscle fibers is populated by a variety of mononuclear cell types. Upon acute or chronic insult, these cell populations become activated and initiate finely-orchestrated crosstalk that promotes myofiber repair and regeneration. Mass cytometry is a powerful and highly multiplexed technique for profiling single-cells. Herein, it was used to dissect the dynamics of cell populations in the skeletal muscle in physiological and pathological conditions. Here, we characterized an antibody panel that could be used to identify most of the cell populations in the muscle interstitial space. By exploiting the mass cytometry resolution, we provided a comprehensive picture of the dynamics of the major cell populations that sensed and responded to acute damage in wild type mice and in a mouse model of Duchenne muscular dystrophy. In addition, we revealed the intrinsic heterogeneity of many of these cell populations.

Download Full-text