OMIP‐044: 28‐color immunophenotyping of the human dendritic cell compartment

This work describes the first 30-parameter immunophenotyping of the human dendritic cell (DC) compartment using fluorescent-based flow cytometry. The optimized panel allows for simultaneous detection of 21 myeloid-centric markers distinguishing all canonical DC subsets, with parallel enumeration of monocytes, T and B cells as well as NK cells. Thus, this panel will be useful for extensive phenotyping of immune cells from a variety of human samples limited in size. This 28-color panel has been developed for an extensive phenotyping of antigen-presenting cells (APCs) in human blood and tissue samples (Table 1). 15 markers associated with changes in dendritic cell (DC) function were selected based on the current literature. Five additional markers were used to pregate on canonical DC subpopulations, namely CD141+ cross-presenting conventional DC1 (cDC1), CD1c+ conventional DC2 (cDC2), CD141- CD1c- cDCs as well as CD123+ plasmacytoid DCs (pDC). Furthermore, seven lineage markers were included for parallel enumeration of CD14+ monocytes, B cells, NK cells as well as CD4+ and CD8+ αβ T cells with basic phenotyping of their differentiation status. The panel has been tested on cryopreserved peripheral blood mononuclear cells (PBMCs) and allows further inclusion of DC antigens at the cost of pan-phenotyping markers. While adaptive immune cells such as T and B lymphocytes are widely studied as the essential effector cells of the immune system, APCs are the upstream gatekeepers initiating and shaping virtually any adaptive immune response 1, 2. The past decade has revealed considerable functional specialization within the murine and human dendritic cell (DC) compartment 3-5. However, only recently studies have focused on the myeloid populations present in human nonlymphoid tissues, showing similarities as well as functional differences to the murine system 6. These advancements have largely been fueled by technical developments in the field of flow cytometry 7 and single-cell RNA-sequencing (sc-RNAseq) 8. Importantly, increasing evidence suggests that studying DC biology in the context of autoimmune diseases 9 as well as cancer progression might reveal novel approaches for therapeutic interventions 10, 11. In particular, via their presence in nonlymphoid tissues and their expression of co-stimulatory or co-inhibitory receptors, DCs are considered critical for shaping either a tolerogenic or pro-inflammatory local immune milieu 2, 12. We took advantage of the availability of 30-parameter fluorescent flow cytometry instrumentation to develop an in-depth phenotyping panel of the human DC compartment (Table 2) based on the most recent sc-RNAseq based classifications 8, 13. Thus, we designed the panel to provide a flow cytometry-based approach to examine the DC subsets reported in these sc-RNAseq-based studies. CD45 was included to allow clear separation of hematopoietic cells from contaminating stroma in nonlymphoid tissue samples. For gating on established DC subsets the markers CD11c, HLA-DR, CD141 (also known as BDCA-3, marks cDC1), CD1c (BDCA-1, marks cDC2), and CD123 (pDC) were used 14, 15. A set of 15 surface molecules was compiled, each of which has previously been reported to be a relevant readout for DC function either in inflammatory disorders or in the context of anti-tumor immunity: the costimulatory molecules CD40, CD80, and CD86, the low-affinity Fc-receptors CD16 (FcγrIIIa) and CD32 (FcγrIIa/b), the chemokine receptors CCR7 and CX3CR1, the scavenger receptor CD163, the complement receptor CD11b, as well as CD85k (also known as ILT-3), BTLA (CD272, also known as B and T lymphocyte attenuator), CD26, CD38, and CD172 (also known as Sirpα) 16-20. To obtain a more complete assessment of the immune compartment in rare and valuable human tissue samples, the lineage markers CD3 (pan T cells), CD14 (classical monocytes), CD19 (B cells), CD56 (NK cells) as well as CD4 and CD8 were included. Thus, in conjunction with CCR7 and CD45RA, this panel allows the parallel assessment of the T cell compartment for differentiation status 21. Of note, CCR7 has also been shown to be essential for DC migration to draining lymph nodes 22. The detailed steps of panel development as well as different tested reagent combinations are described in the Supporting Information. Briefly, the initial in silico planning of marker assignment was based on a spillover spreading matrix calculated for the instrument used 23, followed by iterative refinement to minimize the loss of resolution in detectors with low-expression antigens. All reagents used were titrated either to saturation or toward a selected optimal concentration. Of note, the panel was designed to allow the inclusion of two additional DC markers (namely the co-inhibitory molecules PD-L1 and PD-L2, which are not expressed on blood DCs but only after stimulation or on nonlymphoid tissue DCs) at the cost of the T cell markers CD4 and CD8. Furthermore, the two markers assigned to the V-450 and B-515 detectors (CX3CR1 and CD26, respectively) can be swapped to almost any other target antigen of interest without the need for additional modifications to the panel, since these two detectors neither collect nor contribute any relevant spillover (Supporting Information Fig. S2). An example gating tree of the full 28-color staining on a typical PBMC sample is shown in Fig. 1, together with selected fluorescence-minus-one (FMO) controls. Additional staining and gating controls are shown in Supporting Information Figs. S3 and S4. The panel performed robustly on different PBMC samples (Supporting Information Fig. S5), and has also successfully been tested on human tissue samples and fresh human blood (data not shown). After pregating on CD45+ live singlets (Fig. 1A), lineage positive cells (CD14, CD3, CD19, CD56) were gated as separate populations and excluded for downstream APC analysis (Fig. 1C), pDCs were identified as CD123+ cells, and conventional DCs as CD123- CD11c+ HLA-DR+ cells (Fig. 1E). Within this gate, cDC1 are CD141+, cDC2 are CD1c+, and the remaining population constitutes double negative DCs (DN-DCs). The differential expression pattern of the phenotypic markers on cDC1, cDC2, DN-DCs, and pDCs is summarized in Fig. 1G, and selected FMOs are shown in Fig. 1F and 1H. Furthermore, this 30-parameter data set allows parallel phenotyping of CD4+ and CD8+ T cells (e.g., using CD45RA, CCR7, and CX3CR1) (Fig. 1D) as well as B cells (e.g., using CD80, CD32, CCR7), and NK cells (e.g., using CD16). Given that exhaustive analysis using manual two-dimensional gating is not feasible 24, 25, several novel approaches for exploratory analysis of high-dimensional single-cell data sets have been reported in the past years, among them t-stochastic neighbor embedding (t-SNE) 26, 27, which can be used in conjunction with the clustering algorithms Phenograph 28 or FlowSOM 29. Though these tools have been developed primarily for an intuitive graphical representation of complex mass cytometry data sets, they can be used for fluorescent flow cytometry data as well, if appropriate quality control is performed 30. An example t-SNE analysis of the CD45+ live events is shown in Supporting Information Fig. 6. Overview gating of an example 30-parameter data set on cryopreserved PBMCs. Cryopreserved PBMCs were obtained from AllCells, stained and subsequently aquired on a BD FACSymphony. Optimal PMT voltages were determined by voltage titration experiments and iterative refinement. All reagents were titrated prior to use. For Figures (C) till (G), files were concatenated from two separate acquisitions. Additional gating and staining controls are shown in Supporting Information Figs. S3–S5. (A) Gating strategy for live CD45+ singlets, after exclusion of debris and doublets. (B) Overview of the 28 markers analyzed within this experiment. (C) Gating strategy for major immune cell lineages (CD14+ classical monocytes, CD19+ B cell and CD3+ T cells). (D) CCR7 and CD45RA expression on CD4+ and CD8+ T cells. (E) After exclusion of CD14+ CD19+ CD3+ events, gating strategy for CD56+ NK cells, CD123+ plasmacytoid DCs, and canonical subsets within conventional CD11c+ HLA-DR+ DCs. Color shading indicates the populations used for histogram overlays. (F) FMO control for CD80 expression on CD19+ B cells. (G) Differential expression of selected markers on the indicated DC subsets: CD123+ pDCs (green), double negative DCs (orange), CD141+ cDC1 (blue) and CD1c+ cDC2 (red) as gated in (E). (H) Overview of selected markers on CD11c+ HLA-DR+ DCs including the corresponding FMO controls (lower panel). BUV, Brilliant Ultraviolet; BV, Brilliant Violet; BB, Brilliant Blue; AF, AlexaFluor; PE, Phycoerythrin; APC, Allophycocyanin. In summary, we report here the first 28-color immunophenotyping focused on the human APC compartment, in parallel yielding additional information about all major immune lineages. We believe that this panel not only provides a valuable resource for future studies of myeloid cell phenotypes, but also allows for flexible exchange of up to four markers, and can be used for a variety of human blood and tissue samples in different disease states. While there are other OMIPs that allow for a broad immunophenotyping of human PBMCs including simple enumeration of DCs (OMIP-023 and OMIP-024) or DC subsets 31, up to date there is no OMIP focusing specifically on the human myeloid compartment, and none that allows for an extensive phenotyping of human DCs as described here. We would like to thank the Flow Cytometry Shared Resource of the FHCRC (Andrew Berger) and the HIV Vaccine Trial Network (in particular Dr. Stephen de Rosa and Karen McLellan) for access to their instruments. We thank Drs. Aaron Tyznik and Suraj Saksena (BD Biosciences) for providing reagents and advice, and Dr. Sabine Spath for critical reading of the manuscript. F.M. is an ISAC scholar. The authors declare no conflict of interest. Additional supporting information may be found in the online version of this article. Supplementary Information Supporting Figures Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.