Mass cytometry: a powerful tool for dissecting the immune landscape

Available online at www.sciencedirect.com

ScienceDirect Mass cytometry: a powerful tool for dissecting the immune landscape Yannick Simoni1, Melissa Hui Yen Chng1, Shamin Li1, Michael Fehlings2 and Evan W Newell1 Advancement in methodologies for single cell analysis has historically been a major driver of progress in immunology. Currently, high dimensional flow cytometry, mass cytometry and various forms of single cell sequencing-based analysis methods are being widely adopted to expose the staggering heterogeneity of immune cells in many contexts. Here, we focus on mass cytometry, a form of flow cytometry that allows for simultaneous interrogation of more than 40 different marker molecules, including cytokines and transcription factors, without the need for spectral compensation. We argue that mass cytometry occupies an important niche within the landscape of single-cell analysis platforms that enables the efficient and in-depth study of diverse immune cell subsets with an ability to zoom-in on myeloid and lymphoid compartments in various tissues in health and disease. We further discuss the unique features of mass cytometry that are favorable for combining multiplex peptide-MHC multimer technology and phenotypic characterization of antigen specific T cells. By referring to recent studies revealing the complexities of tumor immune infiltrates, we highlight the particular importance of this technology for studying cancer in the context of cancer immunotherapy. Finally, we provide thoughts on current technical limitations and how we imagine these being overcome. Addresses 1 Agency for Science, Technology and Research (A*STAR), Singapore Immunology Network (SIgN), Singapore 2 Immunoscape Pte. Ltd., Immunos Building, Singapore Corresponding author: Newell, Evan W ([email protected]. edu.sg)

Current Opinion in Immunology 2018, 51:187–196 This review comes from a themed issue on Lymphocyte development and activation Edited by Stephen Nutt and Joanna Groom

https://doi.org/10.1016/j.coi.2018.03.023 0952-7915/ã 2018 Elsevier Ltd. All rights reserved.

the development of high dimensional fluorescent flow cytometry, mass cytometry, single cell mRNA sequencing and other methods for high throughput single cell analysis, cellular subset analysis has entered a new high dimensional era that has exceeded the limits of traditional methods for defining, characterizing and quantifying immune cell subsets [1–3]. The development of mass cytometry (a.k.a. CyTOF — Cytometry by Time-OfFlight) and the sudden increment in the number of parameters that can be reliably measured on individual cells was especially pertinent in exposing the limitations of traditional ways of defining immune cells [4–6]. With the development of many new visualizations, clustering and other high dimensional analysis methods, we are in the process of harnessing these challenging datasets [1,7]. Nonetheless, extracting the underlying meaning or functional utility of each of these countless cell ‘subsets’ remains a challenge. However, even without attempting to dissect and define, broader concepts and immunologically relevant hypotheses can be developed and tested from these complex datasets. Here we examine the strengths and weaknesses of the various single cell analysis methodologies compared with mass cytometry. From there, we can comment on applications that we think best leverage mass cytometry and discuss strategies for the effective use of this method alone or in conjunction with other single-cell analysis methods. We then discuss progress in the use of mass cytometry to explore all types of lymphoid and myeloid cells across tissues in health and disease, in both mouse and humans. Next we describe the many benefits of mass cytometry as it is applied for the study of antigen-specific T cells through the use of peptide-MHC multimers. We highlight a number of studies that show the importance of mass cytometry as a tool for the study of cancer immunology that could lead to improvements in cancer immunotherapy. Lastly, we comment on how we envisage that the most important limitations of current single analysis methods might be overcome as sequencing-based methods continue to rapidly progress.

The place for mass cytometry among cutting edge single cell analysis methods Introduction Exemplified by the ubiquitous use of flow cytometry in cellular immunology, single-cell analysis has long been an obsession of cellular immunologists. In recent years, with www.sciencedirect.com

As alluded to above, progress in single cell analysis methodologies and the accompanying computational methods used to handle the resulting data have seen remarkable progress in recent years. One of the most Current Opinion in Immunology 2018, 51:187–196

188 Lymphocyte development and activation

Figure 1

Flow cytometry

Mass cytometry

Heavy metal

Fluorochrome

Cells labelling

REAP-seq, Ab-seq, CITE-seq ATGATCGATCGAAA

DNA barcode

Antibody

Flow-cytometer (Fluorescence)

Mass-cytometer (mass spectrometry)

REAP-seq (DNA sequencing)

Up to 30

up to 40

currently 82

Sensitivity

High

A bit lower

High

Accessibility

Easy

Moderate

Difficult

Data output

Proteins

Proteins

Proteins and mRNA

10 000 cells/second

300 cells/second

4000 cells/experiment

Method of detection Number of parameters

Cell throughout Cost

+

+++

+++

(exotic fluorochromes +++) Current Opinion in Immunology

High dimensional single cell analysis at protein level. Comparison between flow-cytometry, mass cytometry and oligonucleotide-tagging based approaches (REAP-seq/At-seq/CITE-seq).

exciting phases of progress has been related to the development of single cell mRNA and other sequencing-based single cell genomics methods. For all single cell sequencing based methodologies, cellular throughput remains a significant limitation as compared with flow cytometry based methods. The development, applications and future of these have been exceptionally well described in recent reviews [8,9]. Therefore, we will focus our attention only on methods that allow for single cell analysis of protein expression. Along these lines, the past decade has also witnessed remarkable progress, going from 3 to 6 protein markers detected by flow cytometers (i.e. FACScan, BD Canto) all the way to oligo-tagging based methods, which have been used to measure the expression of as many as 82 proteins expression simultaneously at single cell level (with the theoretical maximum being much higher) [10]. All these tools use antibodies for marker detection which are coupled either to a fluorochrome (flow-cytometry), a metal (mass cytometry) or a DNA-oligonucleotide (e.g. Ab-seq [11], REAP-seq [10] or CITE-Seq [12]) (Figure 1). Although mass spectrometry-based single cell proteomics are also being developed, these methodologies are still in their infancy and are severely limited in cellular throughput and sensitivity for low abundance proteins [13]. Data acquired using these different methods can be easily analyzed using standard bi-axial representation and gating strategies to identify different cell populations and marker expression profiles (using software such as FlowJo1 or BD Current Opinion in Immunology 2018, 51:187–196

FACSDivaTM). Recently, new tools that incorporate high dimensional analysis methods allow a deeper understanding of the data, by offering multiple new functions and features, such as dimensionality reduction for visualization, clustering for quantifying cellular subsets and cellular trajectory analysis to infer changes in marker expression associated with differentiation processes [2]. Here we focus our attention on comparing the strengths and weaknesses of three attractive state-of-the-art approaches. First is the most recent generation of fluorescent flow cytometers, such as spectral-analyzer based flow cytometers [14] or the BD FACSymphony, which uses 5 lasers, numerous photomultiplier tube detectors and currently offers the possibility to analyze up to 30 different antibody markers [15]. The second is the CyTOFTM mass cytometer that allows detection of up to 40 different antibody markers on each cell [4]. Lastly, we consider approaches that use oligonucleotides as tags for antibodies, which offer the potential for a much higher degree of multiplexing capacity. One early study demonstrated the utility of this approach and employed NanostringTM technology as a means to quantify each tag [16]. More recently, several groups have used DNA sequencing of single cells to quantify the binding of antibodies on each cell (e.g. Abseq [11], REAP-seq [10] or CITE-Seq [12]). One such study impressively employed an antibody panel to probe 82 different markers simultaneously. These approaches have the additional advantage of www.sciencedirect.com

Profiling immune cells with mass cytometry Simoni et al. 189

allowing simultaneous analysis of the mRNA profiles of the same cells [10]. Despite having the same purpose of analyzing markers expression at a cellular level, these three techniques differ by several aspects, such as dimensionality, sensitivity, accessibility and overall throughput (Figure 1). Compared to flow cytometry, mass cytometry increased dimensionality to measuring up to 40 parameters (Figure 1). Avoiding auto-fluorescence and compensation issues without losing sensitivity is another big asset. Conversely, these advantages are dampened by the cell loss during the whole experimental procedure and acquisition time that is 50 times slower than by flow cytometry (see detailed review [3]). Still, mass cytometry permits analysis of millions of cells in a matter of hours. Using DNA-tagged antibodies that are resolved using deep sequencing provides an interesting new alternative with even more parameters than mass cytometry. Although not yet as well validated for each marker, protein expression profiles revealed by DNA-tagged antibodies are consistent with those analyzed by flow cytometry [10,12]. In addition to the potentially very high number of antibody probes that can be used simultaneously, this approach is particularly exciting because it permits directs assessments of complex relationships between cellular protein and mRNA expression profiles on individual cells. However, in addition to the new need for validation of each oligo-tagged antibody, this new technique is not yet a user-friendly protocol with requirement of time consuming deep sequencing and associated computational analysis. It is also unclear how easy it will be for the bandwidth of sequencing to accommodate the analysis of many high abundance markers in parallel with low abundance ones, which might necessitate further panel-specific optimizations. In terms of cellular throughput, sequencing-based approaches are severely lagging and limited both by the throughput of instrumentation involved and by the time and costs of deep sequencing. As of now, only thousands of cells can be analyzed in an experiment, and routes for scaling this up are expensive. These limitations and other considerations are summarized in Figure 1. Based on this comparison, we believe that the future of sequencing-based methodologies is especially promising, assuming cost, throughput and other limitations can be overcome. In the meantime, by objective assessment, mass cytometry, with its current ability to accurately measure dozens of parameters on millions of cells collected within a reasonable timeframe as well as the confidence that panel design and crosstalk between channels will have minor effects on the data, is a good choice for many applications. As described below for peptideMHC multimer staining, these advantages and minimal variation in the intensity of metal tags gives mass cytometry a major advantage over flow cytometry when it www.sciencedirect.com

comes to multiplex applications, such as combinatorial tetramer staining [17] or cellular barcoding [18]. In addition, less crosstalk between channels means that fewer assumptions need to be made when designing panels. For this reason, we think mass cytometry is particularly powerful for the deep dissection of individual cell subsets where it is not possible to assume mutually exclusive patterns of marker expression, which is often the case when designing fluorescent flow cytometry panels. Thus, although it may seem that a major advantage of mass cytometry would be to interrogate many different cell types (e.g. lymphocytes and myeloid cells) from a single sample at once, many of the studies we discuss here mainly focused on separate cell compartments, thus deciphering an unprecedented degree of phenotypic heterogeneity and intermediates within well-established cell subsets [2,3,5].

Revealing myeloid and lymphoid cell diversity With the aid of innovative bioinformatics algorithms analyzing high-dimensional data sets (t-SNE, SPADE, PhenoGraph), mass cytometry has been used to comprehensively study the diversity of immune cells [5] and we have focused much of our attention on the analysis of T cell subsets [6] (see t-SNE example in Figure 2). As previously discussed in detail [1,2], the high dimensional perspective of cellular heterogeneity provides a broader and unbiased perspective that is free from the blinders imposed by predetermined gating strategies. This forces us to account for all types of cells and exposes arbitrary distinctions between cell ‘subsets’ that are often better described as extremes along continuums of phenotypic diversity. To evaluate the utility of this approach in a complex system, we attempted to broadly describe mouse myeloid cells across many tissues. Myeloid cells are interesting for this type of analysis due to their plasticity, and large degree of overlap and variation in markers typically used to define their subsets. Our goal was to work toward unifying definitions and to evaluate the relative distinctness or relatedness of each. In one study, our group demonstrated that mass cytometry, in combination with t-SNE analysis allowed an unambiguous and unbiased characterization of the myeloid system across eight different mouse tissues [1]. Similar approaches have gone further to more accurately reveal the heterogeneity of dendritic cells subsets [19] and monocytes [20] across mouse and human tissues. For instance, using mass cytometry to study dendritic cell precursors (pre-DC) in humans, with our help, Florent Ginhoux’s group identified different subpopulations of pre-DC exhibiting functional differences and found that pre-DC share surface markers with plasmacytoid dendritic cells (pDC) but have distinct functional properties that were previously attributed to pDC [21]. The use of mass cytometry has also further supported their recent findings in human fetus where they identified a subset of dendritic cells that promotes prenatal T cell immune-suppression [22]. Current Opinion in Immunology 2018, 51:187–196

190 Lymphocyte development and activation

Figure 2

(a) Gated on live CD45+ CD8 T cells

CD3

TCR Vα7.2

CD4

PD-1

CTLA-4

CD27

CD127 CD25 Naive CD8 T cells Tγδ

(b)

PBMC

GranzymeB

CD39

ICOS

MAIT

CD103 Treg

Treg

Naive Activated

CCR7

Perforin

Cytotoxic CD8 T

CD8α

CD161

TCRγδ

MAIT cells

CD45RO

TIGIT

CD69 CD4 T cells

KLRG1

2B4 (CD244)

CD38 CD8 T cells

Patient A – Colorectal cancer Stage IV Colon Primary tumor (colon)

CD161

Metastasis (liver)

CD8 T cells

t-SNE2

Tγδ

MAIT cells

CD4 T cells

t-SNE1 Tγδ

MAIT

CD4 T cells

CD8 T cells

Current Opinion in Immunology

Profiling human T lymphocyte populations by mass cytometry. (a) Gating strategy to simultaneously analyze different T cell populations by mass cytometry: T gamma/delta (TCRgd+ CD3+), Mucosal Associated Invariant T cells (MAIT) (TCRgd CD3+ CD161+ TCR Va7.2+), T alpha/beta (TCRgd CD3+ CD4+ or CD8+). For each population, cytotoxic cytokines (granzymeB, perforin), activation markers (CCR7, CD45RO), or transcription factors (FoxP3, T-bet) can be analyzed. Histograms display various other surface markers expressed by Tgd (red), MAIT (blue), Treg (dark green), CD4+ (light green) and CD8+ T cells (orange) from tumor samples. Naive CD8+ T cells from blood of the same patient were used as controls (grey). (b) As an alternative method to assess T cell profiles, we performed t-SNE (t-distribution Stochastic Neighbor Embedding) analysis. t-SNE maps high dimensional data into low dimensional space by making a pair-wise comparisons of cellular phenotypes to optimally plot similar cells nearby to one another [1,52,70]. For our analysis, we used t-SNE to reduce 40 parameters (dimensions) into two dimensions (t-SNE1 and tSNE2).

Beyond studies of myeloid cells at steady state, the unbiased approach appears to be especially critical when attempting to make sense of myeloid cell diversity in the context of inflammation. The high plasticity of myeloid cells and their ability to rapidly change marker expression Current Opinion in Immunology 2018, 51:187–196

in response to inflammatory signals makes these cells especially difficult to study. Thus, as has been recently demonstrated [23,24], we think that mass cytometry’s ability to rapidly profile a wide range of cells sampled across states of inflammation in various contexts is www.sciencedirect.com

Profiling immune cells with mass cytometry Simoni et al. 191

especially valuable when used in conjunction with fatemapping methodologies that allow for incorporation of information related to the ontogeny of these cells [25]. We have also been active in the use of mass cytometry for its ability to provide an in-depth picture of the lymphoid system. By interrogating CD8+ T cells (including cells with various T cell antigen-specificities) with a panel consisting of 25 marker molecules that are differently expressed in humans, we explored the complex relationships between cellular phenotype, antigen-specificity and the capacity to express a wide range of effector molecules [6]. Later, we developed an alternative visualization, called One-SENSE, that allows for more direct high dimensional assessment of relationships between such categories of markers [26]. With One-SENSE, measured parameters are grouped into predefined categories (e.g. co-stimulatory markers, trafficking markers or exhaustion markers) and cells are projected onto a space composed of one dimension for each category. Each dimension in OneSENSE has biological meaning that can be easily annotated with binned heat plots. For instance, our group was able to simultaneously assess T cell trafficking and function across eight different human tissues and to further reveal tissue-specific immune signatures [27]. Although only a few patient samples were evaluated for each tissue in this study, the high resolution provided of the range of possible lymphocyte phenotypes is complementary to other studies that have more comprehensively evaluated human derived tissues from many more donors [28,29]. Data from the mass cytometry study have been deposited public use (https://flowrepository.org/id/ for FR-FCM-ZZTM) and we recommend using these data to evaluate any lymphocyte subset of interest as a way to refine hypotheses prior to initiating further study. More specifically, analyzing T-cell phenotypes and function through mass cytometry has also allowed to compare different subsets of Follicular Helper T (Tfh) cells in blood and tonsils, and to identify a particular memory Tfh cell subset that accumulates during childhood in tonsils [30]. In general and as mentioned above, mass cytometry analyses of T cell populations highlight the challenges of neatly segregating cell subsets that often overlap as continuum of related phenotypic profiles. The heterogeneity of Innate Lymphoid Cells (ILC) across different human tissues was also explored by mass cytometry, which facilitated unravelling phenotypic characteristics of ILC subsets and to highlight their uniqueness in humans as compared to the murine system [31– 34]. We highlight this study as an example of a situation where the strengths of mass cytometry are uniquely emphasized. For one, in this case, mass cytometry was used as a way to more precisely define ILC subsets and more accurately exclude the possibility of contaminating cells, which requires more parameters than were easily available to fluorescent flow cytometry. Secondly, by www.sciencedirect.com

combining this with a broad perspective afforded by tSNE, we were able to more objectively validate existing ILC subset definitions and call into question the utility of the helper-type ILC1 cell subset definition [35]. Thus, while most high dimensional analysis studies provide broad perspectives and evidence for many new cell subsets, this study was unique to report a reduction of useful ILC subsets. It should be noted that these data are also publicly available and should be useful for investigators interested in ILCs derived from any of the 12 human tissues evaluated (https://flowrepository.org/id/ FR-FCM-ZYZX). In summary, the advances of mass cytometry for the analysis of immune cells diversity are apparent and more issues regarding immune populations in specific pathological conditions are waiting to be addressed.

Addressing T-cell antigen specificity Fundamental to our understanding of adaptive immune responses, T cell recognition of peptide presented in the context of MHC forms the basis of immunological memory. The broad utility of the latest methods for identifying and profiling antigen-specific T cells has been extensively reviewed [36]. In the context of infectious diseases, the study of antigen-specific T cell responses holds promise for the development of more accurate and broadly applicable correlates of protection that can be used to speed the development or improvement of T cell based vaccination strategies. This task can be further complicated in instances where variations in the infectious organism’s genome occur due to genetic mutations or the emergence of new subtypes as consequences of antigenic shifts and this demands more efficient means for the identification of novel targets that can induce immunity. In the context of cancer, tumor-specific antigens (i.e. antigens arising through tumor-specific mutations, neoantigens) have been linked to the clinical activity of cancer immunotherapy by a CD8+ T cell dependent mechanism and are now considered promising targets for the creation of synthetic cancer vaccines or the autologous transfer of such enriched tumor-specific T cells in combination with checkpoint blockade immunotherapy to control different human cancers [37]. Profiles of such tumor-specific cells also hold promise to improve on biomarkers and various forms of personalized therapy as we have discussed [38]. Depending on the type of cancer, the mutational burden can comprise up to hundreds or thousands of non-synonymous mutations within expressed genes that can result in a very high cancer-specific neoantigen load [39]. Although combinations of state of the art exome sequencing and antigen prediction algorithms have been shown to provide a means of narrowing down the potentially vast number of likely relevant target candidates [37], the real existence of such antigens is still guesswork and detection can be encumbered by the limited availability of patient samples. This is further complicated by the very low numbers of such neoantigen-specific T cells [40]. Hence, Current Opinion in Immunology 2018, 51:187–196

192 Lymphocyte development and activation

the problem that needs to be solved is how do we identify antigen-responsive T cells and their cognate antigens in a high throughput fashion while overcoming the potential signal to noise issues that arise with detecting rare cells? T cell epitopes can be identified by stimulating T cells with different antigens that induce cytokine secretion or cell proliferation. However, these methods are of low sensitivity and bind to antigens that do not elicit an expected response [41,42]. Staining cells with MHC class I tetramers represent a more direct means of measurement [43]. Since only a physical binding interaction is required, pMHC multimers can identify antigens regardless of activation state or functional capacities of the T cell [44] and can be used to detect very rare T cell receptor specificities [45]. In addition, since in vitro stimulation is not required to detect such cells, pMHC multimers are an excellent tool for studying ex vivo unmanipulated phenotypic profiles of antigen-specific T cells [43]. High throughput capacities of modern flow cytometry devices have allowed for rapid acquisition of large cell numbers and the invention of tetramer multiplexing approaches [46,47]. However, spectral overlap and variations in the performance of fluorescently tagged reagents still dampen the number of possible combinations and make the deconvolution of flow cytometry-based multiplexing difficult. An additional difficulty comes from the variation in physical properties (e.g. propensity for non-specific binding) for each of the numerous possible fluorophores that can be used for fluorescent flow cytometry. The use of mass cytometry can overcome these lowresolution limitations and probe for several CD8+ T-cell antigen specificities while still retaining the possibility to track very low frequencies of antigen-specific cells [6]. Due to isotopically pure reporter elements together with a mass independent signal intensity, tetramers consisting of streptavidin molecules conjugated with different multiatom metal tags each have very similar physical properties and therefore show similar staining intensities, which facilitates a clear segregation of tetramer positive cells (i.e. antigen-specific cells) from the pool of tetramernegative T cells (Figure 3). The availability of up to 40 parameters with minimal channel crosstalk further facilitates tailoring high multiplexing approaches without the need for compensation. For instance, our lab routinely uses a triple coding approach using 10 out of 40 metals to screen for up to 120 different epitope candidates in mice and men [48,49]. By increasing the metal channels dedicated to the tetramers as well as the coding strategy, the numbers of target epitopes can be vastly elevated (e.g. Figure 3). Although a triple or higher coding approach is accompanied by a loss in the signal intensity for each individual metal used, it also exponentially mitigates the background noise to signal ratio [48] and allows the detection of a very small minority of antigen-specific cells in small sample sizes. Non-specific T-cell binding can be Current Opinion in Immunology 2018, 51:187–196

identified and corrected for through the incorporation of a second set of tetramers with an alternative combinatorial metal-coding scheme for each peptide used. Since variations in the signal intensities for the metal labelled tetramers are inconsequential, the correspondence of frequencies of tetramer binding cells detected in both configurations serves as an indicator for assessing and obviating cross-reactivity and false positive events. Lastly, the large number of remaining channels dedicated to phenotypic or functional markers (e.g. up to 30 when using 10 channels for tetramers) allows for further control of background noise and false positives. Cells identified as tetramer positive can be compared against tetramer-irrelevant cell subsets (e.g. non peptide-MHC-class I tetramer restricted cell subsets) or classic phenotypic profiles (e.g. naive T cells) to determine if the cells express skewed phenotypic profiles that are associated with antigen experience. This becomes particularly important when probing samples for rare antigen-specific T cell populations and limitations in sample volumes hinder the detection of a solid and reliable number of real cell events. In summary, the high dimensional capacities of mass cytometry can be leveraged for the establishment of robust criteria to guide the validation of bona-fide antigen-specific T cells while using a high throughput probing of hundreds to thousands of tetramers, including (i) multi-tetramer coding schemes to increase specificity, (ii) two configuration staining correspondence and cutoff thresholds for a minimum number of cells to be detected in each configuration, (iii) assessment of background pMHC tetramer staining on cells not restricted to MHC class I molecules (such as CD4+ T cells), (iv) cell profiling to assess phenotype skewing that is associated with antigen-experienced cells.

A caveat of using mass cytometry for the screening of antigen-specific T cell is the low cell throughput and the loss of cells during staining and acquisition. Although this is constantly being improved, smaller sample sizes still remain a challenge and ultimately, the total number of CD8+ T cells that reach the mass detector defines the overall threshold for the detection limit of antigen-specific T cell frequencies. In addition, new technological developments are on the rise which may overcome the limitations of mass cytometry. For example, new generation flow cytometry devices like the BD FACSymphony have reduced spectral overlap and more parameters without the same cell loss rate, while pMHC tetramer oligomer tagging has opened up the number of antigens that can be screened at any one time by leaps and bounds, with one study screening more than 1000 peptides [50]. Although so far employed using cells stained in bulk, this method www.sciencedirect.com

Profiling immune cells with mass cytometry Simoni et al. 193

Figure 3

Streptavidin 164 Streptavidin 172

CD127 (IL-7R)

Streptavidin 169

CD103

CCR7

CD45RO

CD69

Streptavidin 164

Streptavidin 175

KLRG1

Streptavidin 165

Streptavidin 159

Streptavidin 173

Streptavidin 170

Streptavidin 167

Streptavidin 161 Streptavidin 154

CD3

Streptavidin 173

CD8

CD8 T cells

Streptavidin 158

Multiplex tetramer staining by mass-cytometry (PBMC – individual #1)

PD-1

TIGIT

Current Opinion in Immunology

Multiplex tetramer staining performed by mass cytometry. Example of multiplex tetramer staining performed in PBMC from one individual (HLAA*24:02 positive) using a combination of 283 different tetramers, each tetramer having an unique code of three different streptavidins (13 choose 3 = 283). Each tetramer positive population (specific for different peptides) are represented with a different color code.

should also be applicable to be used in conjunction with DNA. Nonetheless, for the detection of rare antigen specificities, mass cytometry is still the current state of the art and has been designated as technology to watch in 2018 for the generation of neoantigen-derived anti-cancer vaccines and multidimensional cell profiling [51].

Characterizing immune cell heterogeneity in oncology In recent years, mass cytometry technology together with computational analysis approaches has emerged as a powerful tool to reveal new insights into cancer biology www.sciencedirect.com

[52,53] (Figure 2). In this context, mass cytometry has been used to characterize immune cell compositions in lymphoma, brain, lung, colorectal, renal and liver cancer [54–57]. Beyond the analysis of cellular surface markers, mass cytometry further allows evaluating cell signaling processes through the analysis of protein phosphorylation (e.g. STAT5, NFkB) [5,18]. In human acute myeloid leukemia (AML), this demonstrated tumor cell heterogeneity as well as a distinct pattern of signaling activation between AML samples [52,58,59]. Using a similar approach, Nolan’s group identified AML profiles that correlated with survival in independent cohorts [59]. Current Opinion in Immunology 2018, 51:187–196

194 Lymphocyte development and activation

Recently, the group of Miriam Merad revealed immune cell heterogeneity in human lung tumors. Their analysis shows that tumor immune-signatures are independent of the cancer stage but a significant change in innate immune cells occurs in early lung cancer, that is characterized by a reduced number of NK cells and CD141+ dendritic cell subsets, but an enrichment of PPARg macrophages [60]. Similarly, Bodenmiller’s group showed heterogeneity in immune cells in renal cell carcinoma and identified a specific immune signature that is linked to a shorter progression-free survival [54]. Beyond the characterization of tumor infiltrating immune cells, mass cytometry has also been used to monitor changes in immune cell subsets during and after immunotherapy. By using mouse models, Engleman’s group showed that activation of the immune system in response to immunotherapy can be monitored in peripheral blood samples. They identified a CD4+ T subset (CD44+ CD69+ CD62L CD27low T-bet+) in peripheral blood that correlated with tumor protection and further confirmed these results by identifying a similar population in human melanoma patients that were treated with antiCTLA4 as immune checkpoint inhibitor [61]. Similarly, Allison’s group showed that anti-PD1 and anti-CTL4 therapy in tumor bearing mice induce distinct cellular mechanism, but both target exhausted T cells. Interestingly, they identified a subset of ICOS+ TH1-like CD4 T cells that are specifically expanded following anti-CTLA4 treatment [62]. In human glioblastoma (GBM), mass cytometry analysis reveal that anti-PD-1 treatment shifts the profile of Treg cells toward an exhausted cell profile, similar to what has been observed for Treg cells infiltrating GBM tumors [63]. By profiling peripheral blood in human melanoma patients treated with anti-PD-1, multiple groups have identified cell populations that could facilitate the identification and classification of responder versus non-responder patients to anti-PD-1 treatment [64–66]. Focusing again on antigen-specific T cells, our group showed by using a mass cytometry based multiplex pMHC class I tetramer staining approach that tumorspecific CD8+ TILs display a highly heterogeneous phenotype that changes in response to anti-CTLA-4 treatment in tumors, but not in the periphery of tumor bearing mice [49]. From all of these studies, it is clear that immune responses in cancer are extremely diverse both on the cellular level within tumors from individual patients and between patients, even with otherwise similar clinical classifications. Although challenging, we believe that insights gained from methods such as mass cytometry will continue to lead to the identification of quantifiable cellular profiles that should enable many aspects of cancer immunotherapy.

Perspectives With relatively high cellular throughput and depth of analysis, mass cytometry currently fills an important niche Current Opinion in Immunology 2018, 51:187–196

for the study and better understanding of the highly complex cellular immune system. Nonetheless, comparatively slow cellular throughput and increased loss of cells during acquisition as compared to fluorescence-based flow cytometry limits the scope of applications and may make mass cytometry less feasible for the analysis of large cohorts or small-sized samples. With continued advances in fluorescence flow cytometry on one side and single cell sequencing-based methods on the other, we anticipate that mass cytometry’s niche within the single cell analysis space will continue to narrow in long term future. However, as we highlight here, one particularly advantageous application for mass cytometry is the study of T cell antigen specificity. Furthermore, ongoing improvements on technical aspects, such as advanced fluidic systems, mass tag barcoding approaches [18] or enrichment procedures, however, will continue to significantly reduce these impairments. The scalability of markers that can be interrogated by mass cytometry is limited by the number of available metal isotopes and there is still room for improvement in this aspect as well. Thus, we hope that there is continued investment in the improvement of the fundamentals of this important technology. Much investment is already being made in mass cytometry-like methods of highly multiplexed tissue imaging [67–69] which, further enhances the possibilities of elucidating and imaging phenotypic profiles cells in tissues and particularly in human tumor samples [67]. With opportunity to synergize with standard mass cytometry and other single-cell analysis methods, including analysis of T cell antigen specificity, this added dimensionality might be a milestone for the improvement in biomarker discovery the understanding of cellular function.

Conflict of interest statement E.W.N. is a board director and shareholder of immunoSCAPE Pte.Ltd. M.F. is Director, Scientific Affairs and shareholder of immunoSCAPE Pte. Ltd. All other authors declare no competing financial interests.

References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as:  of special interest  of outstanding interest 1.

Becher B, Schlitzer A, Chen J, Mair F, Sumatoh HR, Teng KW, Low D, Ruedl C, Riccardi-Castagnoli P, Poidinger M et al.: Highdimensional analysis of the murine myeloid cell system. Nat Immunol 2014, 15:1181-1189.

2.

Newell EW, Cheng Y: Mass cytometry: blessed with the curse of dimensionality. Nat Immunol 2016, 17:890-895.

3.

Bendall SC, Nolan GP, Roederer M, Chattopadhyay PK: A deep profiler’s guide to cytometry. Trends Immunol 2012, 33:323-332.

4.

Bandura DR, Baranov VI, Ornatsky OI, Antonov A, Kinach R, Lou X, Pavlov S, Vorobiev S, Dick JE, Tanner SD: Mass cytometry: technique for real time single cell multitarget immunoassay based on inductively coupled plasma time-of-flight mass spectrometry. Anal Chem 2009, 81:6813-6822. www.sciencedirect.com

Profiling immune cells with mass cytometry Simoni et al. 195

5.

6.

Bendall SC, Simonds EF, Qiu P, Amir el AD, Krutzik PO, Finck R, Bruggner RV, Melamed R, Trejo A, Ornatsky OI et al.: Single-cell mass cytometry of differential immune and drug responses across a human hematopoietic continuum. Science 2011, 332:687-696. Newell EW, Sigal N, Bendall SC, Nolan GP, Davis MM: Cytometry by time-of-flight shows combinatorial cytokine expression and virus-specific cell niches within a continuum of CD8+ T cell phenotypes. Immunity 2012, 36:142-152.

7.

Wagner A, Regev A, Yosef N: Revealing the vectors of cellular identity with single-cell genomics. Nat Biotechnol 2016, 34:1145-1160.

8.

Stubbington MJT, Rozenblatt-Rosen O, Regev A, Teichmann SA: Single-cell transcriptomics to explore the immune system in health and disease. Science 2017, 358:58-63.

9.

Patel AP, Tirosh I, Trombetta JJ, Shalek AK, Gillespie SM, Wakimoto H, Cahill DP, Nahed BV, Curry WT, Martuza RL et al.: Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma. Science 2014, 344:1396-1401.

10. Peterson VM, Zhang KX, Kumar N, Wong J, Li L, Wilson DC,  Moore R, McClanahan TK, Sadekova S, Klappenbach JA: Multiplexed quantification of proteins and transcripts in single cells. Nat Biotechnol 2017. An exciting demonstration of single-cell sequencing of oligonucleotide tagged antibodies and mRNA for simutaneous analysis of single cell protein and gene expression. 11. Shahi P, Kim SC, Haliburton JR, Gartner ZJ, Abate AR: Abseq: ultrahigh-throughput single cell protein profiling with droplet microfluidic barcoding. Sci Rep 2017, 7:44447. 12. Stoeckius M, Hafemeister C, Stephenson W, Houck-Loomis B, Chattopadhyay PK, Swerdlow H, Satija R, Smibert P:  Simultaneous epitope and transcriptome measurement in single cells. Nat Methods 2017, 14:865-868. An exciting demonstration of single-cell sequencing of oligonucleotide tagged antibodies and mRNA for simutaneous analysis of single cell protein and gene expression.

human DC lineage through the integration of highdimensional techniques. Science 2017:356. 22. McGovern N, Shin A, Low G, Low D, Duan K, Yao LJ, Msallam R, Low I, Shadan NB, Sumatoh HR et al.: Human fetal dendritic cells promote prenatal T-cell immune suppression through arginase-2. Nature 2017, 546:662-666. 23. Mrdjen D, Pavlovic A, Hartmann FJ, Schreiner B, Utz SG, Leung BP, Lelios I, Heppner FL, Kipnis J, Merkler D et al.: Highdimensional single-cell mapping of central nervous system immune cells reveals distinct myeloid subsets in health, aging, and disease. Immunity 2018, 48 380–395.e386. 24. Cheng CY, Gutierrez NM, Marzuki MB, Lu X, Foreman TW, Paleja B, Lee B, Balachander A, Chen J, Tsenova L et al.: Host sirtuin 1 regulates mycobacterial immunopathogenesis and represents a therapeutic target against tuberculosis. Sci Immunol 2017:2. 25. Ginhoux F, Schultze JL, Murray PJ, Ochando J, Biswas SK: New insights into the multidimensional concept of macrophage ontogeny, activation and function. Nat Immunol 2016, 17:34-40. 26. Cheng Y, Wong MT, van der Maaten L, Newell EW: Categorical analysis of human T cell heterogeneity with one-dimensional soli-expression by nonlinear stochastic embedding. J Immunol 2016, 196:924-932. 27. Wong MT, Ong DE, Lim FS, Teng KW, McGovern N, Narayanan S,  Ho WQ, Cerny D, Tan HK, Anicete R et al.: A high-dimensional atlas of human T cell diversity reveals tissue-specific trafficking and cytokine signatures. Immunity 2016, 45:442-456. An atlas of lymphocyte phenotypes obtained using two mass cytometry applied on eight different human tissues. All data available on flow repository. 28. Thome JJ, Yudanin N, Ohmura Y, Kubota M, Grinshpun B, Sathaliyawala T, Kato T, Lerner H, Shen Y, Farber DL: Spatial map of human T cell compartmentalization and maintenance over decades of life. Cell 2014, 159:814-828.

13. Zhang L, Vertes A: Single-cell mass spectrometry approaches to explore cellular heterogeneity. Angew Chem Int Ed Engl 2017.

29. Kumar BV, Connors TJ, Farber DL: Human T cell development, localization, and function throughout life. Immunity 2018, 48:202-213.

14. Futamura K, Sekino M, Hata A, Ikebuchi R, Nakanishi Y, Egawa G, Kabashima K, Watanabe T, Furuki M, Tomura M: Novel fullspectral flow cytometry with multiple spectrally-adjacent fluorescent proteins and fluorochromes and visualization of in vivo cellular movement. Cytometry A 2015, 87:830-842.

30. Wong MT, Chen J, Narayanan S, Lin W, Anicete R, Kiaang HT, De Lafaille MA, Poidinger M, Newell EW: Mapping the diversity of follicular helper T cells in human blood and tonsils using highdimensional mass cytometry analysis. Cell Rep 2015, 11:18221833.

15. Perfetto SP, Chattopadhyay PK, Roederer M: Seventeen-colour flow cytometry: unravelling the immune system. Nat Rev Immunol 2004, 4:648-655.

31. van Unen V, Li N, Molendijk I, Temurhan M, Hollt T, van der Meulen-de Jong AE, Verspaget HW, Mearin ML, Mulder CJ, van Bergen J et al.: Mass cytometry of the human mucosal immune system identifies tissue- and disease-associated immune subsets. Immunity 2016, 44:1227-1239.

16. Ullal AV, Peterson V, Agasti SS, Tuang S, Juric D, Castro CM, Weissleder R: Cancer cell profiling by barcoding allows multiplexed protein analysis in fine-needle aspirates. Sci Transl Med 2014, 6 219ra219. 17. Newell EW, Sigal N, Nair N, Kidd BA, Greenberg HB, Davis MM: Combinatorial tetramer staining and mass cytometry analysis facilitate T-cell epitope mapping and characterization. Nat Biotechnol 2013, 31:623-629. 18. Bodenmiller B, Zunder ER, Finck R, Chen TJ, Savig ES, Bruggner RV, Simonds EF, Bendall SC, Sachs K, Krutzik PO et al.: Multiplexed mass cytometry profiling of cellular states perturbed by small-molecule regulators. Nat Biotechnol 2012, 30:858-867. 19. Guilliams M, Dutertre CA, Scott CL, McGovern N, Sichien D, Chakarov S, Van Gassen S, Chen J, Poidinger M, De Prijck S et al.: Unsupervised high-dimensional analysis aligns dendritic cells across tissues and species. Immunity 2016, 45:669-684. 20. Roussel M, Ferrell PB Jr, Greenplate AR, Lhomme F, Le Gallou S, Diggins KE, Johnson DB, Irish JM: Mass cytometry deep phenotyping of human mononuclear phagocytes and myeloidderived suppressor cells from human blood and bone marrow. J Leukoc Biol 2017, 102:437-447. 21. See P, Dutertre CA, Chen J, Gunther P, McGovern N, Irac SE, Gunawan M, Beyer M, Handler K, Duan K et al.: Mapping the www.sciencedirect.com

32. van Unen V, Hollt T, Pezzotti N, Li N, Reinders MJT, Eisemann E, Koning F, Vilanova A, Lelieveldt BPF: Visual analysis of mass cytometry data by hierarchical stochastic neighbour embedding reveals rare cell types. Nat Commun 2017, 8:1740. 33. Simoni Y, Fehlings M, Kloverpris HN, McGovern N, Koo SL, Loh CY, Lim S, Kurioka A, Fergusson JR, Tang CL et al.: Human innate lymphoid cell subsets possess tissue-type based heterogeneity in phenotype and frequency. Immunity 2017, 46:148-161. 34. Simoni Y, Newell EW: Dissecting human ILC heterogeneity: more than just three subsets. Immunology 2018, 153:297-303. 35. Simoni Y, Newell EW: Toward meaningful definitions of innatelymphoid-cell subsets. Immunity 2017, 46:760-761. 36. Hadrup SR, Newell EW: Determining T-cell specificity to understand and treat disease. Nature Biomedical Engineering 2017, 1:784-795. 37. Schumacher TN, Schreiber RD: Neoantigens in cancer immunotherapy. Science 2015, 348:69-74. 38. Newell EW, Becht E: High-dimensional profiling of tumorspecific immune responses: asking T cells about what they “see” in cancer. Cancer Immunol Res 2018, 6:2-9. Current Opinion in Immunology 2018, 51:187–196

196 Lymphocyte development and activation

39. Alexandrov LB, Nik-Zainal S, Wedge DC, Aparicio SA, Behjati S, Biankin AV, Bignell GR, Bolli N, Borg A, Borresen-Dale AL et al.: Signatures of mutational processes in human cancer. Nature 2013, 500:415-421.

gradient in hepatocellular carcinoma using high-dimensional proteomic and transcriptomic analyses. Proc Natl Acad Sci U S A 2017, 114:E5900-E5909.

40. Tran E, Robbins PF, Rosenberg SA: ‘Final common pathway’ of human cancer immunotherapy: targeting random somatic mutations. Nat Immunol 2017, 18:255.

56. Baughn LB, Sachs Z, Noble-Orcutt KE, Mitra A, Van Ness BG, Linden MA: Phenotypic and functional characterization of a bortezomib-resistant multiple myeloma cell line by flow and mass cytometry. Leuk Lymphoma 2017, 58:1931-1940.

41. Fiore-Gartland A, Manso BA, Friedrich DP, Gabriel EE, Finak G, Moodie Z, Hertz T, De Rosa SC, Frahm N, Gilbert PB et al.: Pooled-peptide epitope mapping strategies are efficient and highly sensitive: an evaluation of methods for identifying human T cell epitope specificities in large-scale HIV vaccine efficacy trials. PLOS ONE 2016, 11:e0147812.

57. Fisher DAC, Malkova O, Engle EK, Miner CA, Fulbright MC, Behbehani GK, Collins TB, Bandyopadhyay S, Zhou A, Nolan GP et al.: Mass cytometry analysis reveals hyperactive NF Kappa B signaling in myelofibrosis and secondary acute myeloid leukemia. Leukemia 2017, 31:1962-1974.

42. McCutcheon M, Wehner N, Wensky A, Kushner M, Doan S, Hsiao L, Calabresi P, Ha T, Tran TV, Tate KM et al.: A sensitive ELISPOT assay to detect low-frequency human T lymphocytes. J Immunol Methods 1997, 210:149-166.

58. Han L, Qiu P, Zeng Z, Jorgensen JL, Mak DH, Burks JK, Schober W, McQueen TJ, Cortes J, Tanner SD et al.: Single-cell mass cytometry reveals intracellular survival/proliferative signaling in FLT3-ITD-mutated AML stem/progenitor cells. Cytometry A 2015, 87:346-356.

43. Altman JD, Moss PA, Goulder PJ, Barouch DH, McHeyzerWilliams MG, Bell JI, McMichael AJ, Davis MM: Phenotypic analysis of antigen-specific T lymphocytes. Science 1996, 274:94-96. 44. Wedemeyer H, He XS, Nascimbeni M, Davis AR, Greenberg HB, Hoofnagle JH, Liang TJ, Alter H, Rehermann B: Impaired effector function of hepatitis C virus-specific CD8+ T cells in chronic hepatitis C virus infection. J Immunol 2002, 169:3447-3458. 45. Barnes E, Ward SM, Kasprowicz VO, Dusheiko G, Klenerman P, Lucas M: Ultra-sensitive class I tetramer analysis reveals previously undetectable populations of antiviral CD8+ T cells. Eur J Immunol 2004, 34:1570-1577. 46. Newell EW, Klein LO, Yu W, Davis MM: Simultaneous detection of many T-cell specificities using combinatorial tetramer staining. Nat Methods 2009, 6:497-499. 47. Hadrup SR, Bakker AH, Shu CJ, Andersen RS, van Veluw J, Hombrink P, Castermans E, Thor Straten P, Blank C, Haanen JB et al.: Parallel detection of antigen-specific T-cell responses by multidimensional encoding of MHC multimers. Nat Methods 2009, 6:520-526. 48. Newell EW: Higher throughput methods of identifying T cell epitopes for studying outcomes of altered antigen processing and presentation. Front Immunol 2013, 4:430. 49. Fehlings M, Simoni Y, Penny HL, Becht E, Loh CY, Gubin MM,  Ward JP, Wong SC, Schreiber RD, Newell EW: Checkpoint blockade immunotherapy reshapes the high-dimensional phenotypic heterogeneity of murine intratumoural neoantigen-specific CD8+ T cells. Nat Commun 2017, 8:562. Highlighting the use of mass cytometry to simultaneously screen-for and profile tumor-specific T cells in the context of immunotherapy. 50. Bentzen AK, Marquard AM, Lyngaa R, Saini SK, Ramskov S,  Donia M, Such L, Furness AJ, McGranahan N, Rosenthal R et al.: Large-scale detection of antigen-specific T cells using peptide-MHC-I multimers labeled with DNA barcodes. Nat Biotechnol 2016, 34:1037-1045. An exciting demonstration of oligonucleotide tagged peptide-MHC mulitmers for very highly multiplexed screening of antigen-specific T cells. 51. Powell K: Technology to watch in 2018. Nature 2018, 553:531534. 52. Amir el AD, Davis KL, Tadmor MD, Simonds EF, Levine JH, Bendall SC, Shenfeld DK, Krishnaswamy S, Nolan GP, Pe’er D: viSNE enables visualization of high dimensional single-cell data and reveals phenotypic heterogeneity of leukemia. Nat Biotechnol 2013, 31:545-552. 53. Irish JM, Doxie DB: High-dimensional single-cell cancer biology. Curr Top Microbiol Immunol 2014, 377:1-21.

59. Levine JH, Simonds EF, Bendall SC, Davis KL, Amir el AD, Tadmor MD, Litvin O, Fienberg HG, Jager A, Zunder ER et al.: Data-driven phenotypic dissection of AML reveals progenitorlike cells that correlate with prognosis. Cell 2015, 162:184-197. 60. Lavin Y, Kobayashi S, Leader A, Amir ED, Elefant N, Bigenwald C, Remark R, Sweeney R, Becker CD, Levine JH et al.: Innate immune landscape in early lung adenocarcinoma by paired single-cell analyses. Cell 2017, 169 750–765.e717. 61. Spitzer MH, Carmi Y, Reticker-Flynn NE, Kwek SS, Madhireddy D, Martins MM, Gherardini PF, Prestwood TR, Chabon J, Bendall SC et al.: Systemic immunity is required for effective cancer immunotherapy. Cell 2017, 168 487–502.e415. 62. Wei SC, Levine JH, Cogdill AP, Zhao Y, Anang NAS, Andrews MC,  Sharma P, Wang J, Wargo JA, Pe’er D et al.: Distinct cellular mechanisms underlie anti-CTLA-4 and anti-PD-1 checkpoint blockade. Cell 2017, 170 1120–1133.e1117. A simultaneous immune cells deep profiling during checkpoint blockage therapy in cancer. 63. Lowther DE, Goods BA, Lucca LE, Lerner BA, Raddassi K, van Dijk D, Hernandez AL, Duan X, Gunel M, Coric V et al.: PD-1 marks dysfunctional regulatory T cells in malignant gliomas. JCI Insight 2016:1. 64. Huang AC, Postow MA, Orlowski RJ, Mick R, Bengsch B, Manne S, Xu W, Harmon S, Giles JR, Wenz B et al.: T-cell invigoration to tumour burden ratio associated with anti-PD-1 response. Nature 2017, 545:60-65. 65. Kamphorst AO, Pillai RN, Yang S, Nasti TH, Akondy RS, Wieland A, Sica GL, Yu K, Koenig L, Patel NT et al.: Proliferation of PD-1+ CD8 T cells in peripheral blood after PD-1-targeted therapy in lung cancer patients. Proc Natl Acad Sci U S A 2017, 114:49934998. 66. Krieg C, Nowicka M, Guglietta S, Schindler S, Hartmann FJ, Weber LM, Dummer R, Robinson MD, Levesque MP, Becher B: High-dimensional single-cell analysis predicts response to anti-PD-1 immunotherapy. Nat Med 2018, 24:144-153. 67. Angelo M, Bendall SC, Finck R, Hale MB, Hitzman C, Borowsky AD, Levenson RM, Lowe JB, Liu SD, Zhao S et al.: Multiplexed ion beam imaging of human breast tumors. Nat Med 2014, 20:436-442. 68. Wang S, Shan X, Patel U, Huang X, Lu J, Li J, Tao N: Label-free imaging, detection, and mass measurement of single viruses by surface plasmon resonance. Proc Natl Acad Sci U S A 2010, 107:16028-16032.

54. Chevrier S, Levine JH, Zanotelli VRT, Silina K, Schulz D, Bacac M, Ries CH, Ailles L, Jewett MAS, Moch H et al.: An immune atlas of clear cell renal cell carcinoma. Cell 2017, 169 736–749.e718.

69. Giesen C, Wang HA, Schapiro D, Zivanovic N, Jacobs A, Hattendorf B, Schuffler PJ, Grolimund D, Buhmann JM, Brandt S et al.: Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry. Nat Methods 2014, 11:417-422.

55. Chew V, Lai L, Pan L, Lim CJ, Li J, Ong R, Chua C, Leong JY, Lim KH, Toh HC et al.: Delineation of an immunosuppressive

70. Van der Maaten L, Hinton G: Visualizing data using t-SNE. J Mach Learn Res 2008, 9:85.

Current Opinion in Immunology 2018, 51:187–196

www.sciencedirect.com