Development of a customizable mouse backbone spectral flow cytometry panel to delineate immune cell populations in normal and tumor tissues

Abstract

Introduction: In vivo studies of cancer biology and assessment of therapeutic efficacy are critical to advancing cancer research and ultimately improving patient outcomes. Murine cancer models have proven to be an invaluable tool in pre-clinical studies. In this context, multi-parameter flow cytometry is a powerful method for elucidating the profile of immune cells within the tumor microenvironment and/or play a role in hematological diseases. However, designing an appropriate multi-parameter panel to comprehensively profile the increasing diversity of immune cells across different murine tissues can be extremely challenging.

Methods: To address this issue, we designed a panel with 13 fixed markers that define the major immune populations -referred to as the backbone panel- that can be profiled in different tissues but with the option to incorporate up to seven additional fluorochromes, including any marker specific to the study in question.

Results: This backbone panel maintains its resolution across different spectral flow cytometers and organs, both hematopoietic and non-hematopoietic, as well as tumors with complex immune microenvironments.

Discussion: Having a robust backbone that can be easily customized with pre-validated drop-in fluorochromes saves time and resources and brings consistency and standardization, making it a versatile solution for immuno-oncology researchers. In addition, the approach presented here can serve as a guide to develop similar types of customizable backbone panels for different research questions requiring high-parameter flow cytometry panels.

Keywords: backbone panel; immune cells; immunophenotyping; mouse; spectral flow cytometry; tumor microenvironment (TME).

Figure 1

Backbone panel design and gating strategy. (A) Markers used to define the main immune populations as helper T cells, cytotoxic T cells, T regs, B cells, NK cells, pDCs, cDCs, macrophage, monocytes, and neutrophils. (B) Fluorochrome assignment. Drop-in positions are highlighted and suggested fluorochromes are written in italic. Myeloid markers are in blue, lymphoid markers in red. CD45, and Live/Dead dye are in purple.

Figure 2

The backbone panel efficiently resolves the main murine immune populations. (A) Manual gating strategy applied to spleen cells stained with the backbone panel. (B) t-SNE scatter plot overlayed with the manual gated populations (t-SNE iterations = 1000 and k = 30). (C) Colored t-SNE scatterplots showing the expression level and distribution of the backbone markers.

Figure 3

Evaluation of the backbone panel and impact on drop-in fluorochromes. (A) Histograms of single-stained spleen cells overlaid with backbone-stained spleen cells. Single-stained samples for each fluorochrome are in grey. A representative sample stained with the complete backbone panel is colored. (B) Histograms showing the impact of the backbone on the drop-in fluorochromes (FITC, PE, PE-Cy7, APC, BV421, BV605 and BV785). Unstained splenocytes (in gray) overlaid with the backbone-stained splenocytes gated on the lymphoid cells (T, B and NK cells –in pink). (C) Unstained splenocytes (in gray) overlaid with the backbone-stained splenocytes gated on the myeloid cells (CD11b+ cells –in blue).

Figure 4

The performance of the backbone is reproducible across different spectral flow cytometers. (A) Manual gating strategy showing the main immune populations. Wildtype spleen cells were stained with the backbone panel and acquired on Cytek Aurora (in pink –top), Sony ID 7000 (in blue –middle), and BD S6 SE (in green –bottom). (B) Comparison of the frequency of the main immune populations of live CD45⁺ cells shows no significant differences across the three instruments (n=3 mice/instrument). Data are ± mean s.e.m. Statistical analysis was performed using two-way ANOVA with Geisser-Greenhouse correction, followed by Tukey’s multiple comparisons tests to compare the mean values of the immune cell population percentages between the two instruments (all p values were > 0.1.).

Figure 5

The backbone-defined immune populations are unaffected by the addition of drop-in markers. (A) Manual gating strategy applied to cells harvested from Kras^G12C/+; Trp53^fl/fl lung adenocarcinoma derived from a syngeneic mouse model stained with the backbone panel only and backbone plus drop-ins of the tumor microenvironment (TME) panel (i.e., Epcam, CD31, PDPN, PD-1, and Lag-3). (B) Manual gating strategy applied to wildtype spleen cells stained with the backbone panel only and backbone plus drop-ins of the immune cell panel (i.e., CD62L Siglec-F, c-Kit, CD44, TIM-4, PD-1, and Lag-3). (C) Comparison of the frequency of the backbone-defined immune population within live CD45⁺ cells in the presence or absence of drop-ins in lung tumor and spleen samples.

Figure 6

The backbone panel is organ-agnostic. (A) The UMAP scatter plot shows concatenated events from six different organs, and the overlay shows the distribution of the manual gated cells. For the UMAP analysis, each organ sample was downsized to 40,000 of manual gated live CD45⁺ singlet cells. (B) Individual UMAP scatterplot showing the differences between the organs, namely spleen, blood, bone marrow (BM), liver and lung. (C) Distribution of the immune populations frequency within live CD45⁺ cells, manually gated.

Figure 7

The backbone panel is efficient in analyzing the complex pancreatic ductal adenocarcinoma microenvironment. (A) Manual analysis of cells harvested from pancreatic ductal adenocarcinoma (PDAC), pancreata stained with the backbone and the drop-ins of the TME panel (i.e., Epcam, CD31, PDPN, PD-1, and Lag-3). (B) UMAP scatter plot shows concatenated events from the six samples (three normal and three PDAC pancreata, n=3 mice/group). The overlay shows the distribution of normal and tumor cells. For the UMAP analysis, each sample was downsized to 10,000 manually gated from the live CD45+ singlet cells. The colored scatterplot shows the expression level and distribution of the markers with the most relevant differences. (C) Comparison of the manually gated populations between normal and PDAC samples. Bar graphs showing percentages of different cell populations. Data are ± mean s.e.m.; p values from two-tailed unpaired Student’s t-test (ns, non-significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p< 0.0001).

Figure 8

The backbone panel efficiently works in the presence of a tdTomato fluorescent and other drop-in fluorochromes and allows for an immune cell characterization of tdTomato Tet2^KO mice relative to WT. (A) Manual analysis of cells harvested from whole bone marrow of WT (non-tdT) and tdT-expressing Tet2^KO mice (tdT). To immunophenotype Tet2 loss in HSC-Scl-Cre-ER^T mice, we previously crossed them to Tet2^flox/flox to make a phenotyping comparison between Cre⁺ (Tet2^{Knockout (KO)} ) mice –expressing tdT– and their age-matched Cre^- (functionally WT mice) counterparts –lacking tdT. (B) Comparison of the manually gated bone marrow immune cell populations between WT and Tet2^KO mice. Bar graphs showing percentages of myeloid and lymphoid cell populations as percentage of live CD45⁺ cells. Data are ± mean s.e.m.; p values from two-tailed unpaired Student’s t-test (ns, non-significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p< 0.0001). (C) Comparison of effector and memory CD4⁺ and CD8⁺ T cells in the bone marrow between WT and Tet2^KO mice. Data are ± mean s.e.m.; p values from two-tailed unpaired Student’s t-test (ns, non-significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p< 0.0001).