Cellular immunophenotyping in human and primate tissues during healthy conditions and Ebola and Nipah infections

Abstract

We developed a 29-color spectral cytometry panel to enhance nonhuman primate (NHP) models for cross-reactive immunophenotyping. This panel is suitable for biosafety level 4 (BSL-4) viruses and can be used with both human and NHP samples in BSL-2 research settings. Tissues from humans, rhesus monkeys (RhMs), crab-eating macaques (CEMs), and green monkeys (GMs) were stained with a 29-color immunophenotyping panel requiring only 2 clone substitutions. Comparable staining was observed for all samples. Unbiased analysis showed acceptable overlap in T cell phenotypes across samples, with differences in human and NHP B cells and granulocytes. In CEMs, most circulating CD8+ T cells were from effector memory cells, with significantly higher levels than in humans, RhMs, and GMs. Analysis of samples from various anatomical sites revealed distinct location-specific phenotypes. In Nipah virus-exposed animals, splenocytes showed a substantial increase in IgM+ B cells and a reduction in effector memory CD8+ T cells compared with unexposed controls. Lymph nodes from Ebola virus-exposed animals showed a loss of CXCR3+CD8+ T cells versus unexposed controls. This panel may guide the development of additional multicolor panels in preclinical and clinical settings and may increase understanding of the pathogenesis of diseases caused by emerging and reemerging viruses.

Keywords: Cell migration/adhesion; Immunology; Infectious disease; T cells.

Figure 1. Cross-species comparisons.

(A) UMAP dimensional reduction of PBMCs from human and 3 NHP species demonstrating overlap of major populations. Species identity is color coded. (B) FlowSOM metaclusters (k = 17) projected onto UMAP plot from A. (C) Proportion of total live CD45⁺ PBMCs for each species represented by each metacluster from B. (D) Proportion of CD8⁺ T cells made comprised of central memory (CD28⁺CD95⁺) T cells, effector memory (CD28^–CD95⁺) T cells, and naive (CD28⁺CD95^–) T cells. Significance is shown in proportion among species. (E) Projections of central memory, effector memory, and naive CD8⁺ T cells onto UMAP dimensional reduction, with division by metaclustering as defined in B from 4 different species. (F) Proportion of CD4⁺ T cells made comprised of central memory T cells, effector memory T cells, and naive T cells. Significance is shown in proportion among species. (G) Projections of central memory, effector memory, and naive CD4⁺ T cells onto UMAP dimensional reduction, with division by metaclustering as defined in B from 4 different species. (H) Proportion of CD8⁺ T cells that are central memory plotted against human-equivalent age. (I) Proportion of CD8⁺ T cells that are effector memory plotted against human-equivalent age. (J) Proportion of CD4⁺ T cells that are naive plotted against human-equivalent age. The asterisks indicate significant differences among species as calculated by 2-way ANOVA with Šidák’s multiple comparison analysis. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. The number of green monkeys (GM), crab-eating macaques (CEM), humans, and rhesus monkeys (RhM) are 2, 3, 2, and 3, respectively, for A, C, and E–G and are 4, 3, 4, and 3, respectively, for D, F, and H–J. NHP, nonhuman primate; UMAP, Uniform Manifold Approximation and Projection.

Figure 2. Comparisons of leukocytes from multiple tissues in RhMs.

(A) UMAP dimensional reduction was performed on live CD45⁺ cells from liver, Mes LN, PBMCs, and spleen of RhMs, revealing overlap of major populations. (B) The UMAP plot from A shows the FlowSOM metaclusters (k = 21) projected onto it. (C) Proportions of total cells in each tissue belonging to each metacluster are displayed using column graphs with means. (D) Fold change of the proportion of cells in a tissue compared with that in PBMCs for 21 metaclusters plotted on a logarithmic scale. The color scale indicates fold change values, with yellow representing a 2-fold increase and blue representing a 2-fold decrease. (E) FlowSOM analysis produced a heatmap with 21 metaclusters. Each row represents a unique metacluster, and columns represent analyzed markers. Mean fluorescence intensity values for metaclusters are represented by color scaling for each marker independently. Mes LN, mesenteric lymph node; RhMs, rhesus monkeys; UMAP, Uniform Manifold Approximation and Projection.

Figure 3. Changes in splenocytes phenotype between exposed and unexposed GMs.

(A) UMAP dimension reduction of live, CD45⁺ splenocytes from NiV-exposed GMs (n =4) versus splenocytes from unexposed GMs (n = 5). (B) The UMAP plot from A shows the FlowSOM metaclusters (k = 26) projected onto it. (C) Proportions of total cells from each group belonging to each metacluster are displayed using column graphs with means. Note the following metacluster for specific cell types: MC-7, IgM⁺ B cells, MC-14, CD28^–CD95⁺, CD8⁺ effector memory T cells. (D) FlowSOM analysis produced a heatmap with 26 metaclusters. Each row represents a unique metacluster, and columns represent analyzed markers. Mean fluorescence intensity values for metaclusters are represented by color scaling for each marker independently. The asterisks indicate significant differences among species as calculated by 2-way ANOVA with Šidák’s multiple comparison analysis. ****P < 0.0001. GMs, green monkeys; NiV, Nipah virus; UMAP, Uniform Manifold Approximation and Projection.

Figure 4. Comparison of Mes LN cell phenotypes in RhMs with and without EBOV exposure.

(A) UMAP dimension reduction of live CD45⁺ cells from Mes LN of EBOV-exposed versus unexposed animals (n =2). (B) The UMAP plot from A shows the FlowSOM metaclusters (k = 21) projected onto it. (C) Proportion of total cells from each group belonging to each metacluster are displayed using column graphs with means. (D) Expression of proliferation marker (Ki67), T cell exhaustion marker (PD-1), early activation marker (CD69), and late activation marker (HLA-DR) in Mes LN CD4⁺ T cells from EBOV-exposed versus unexposed RhMs. (E) Expression of Ki-67, PD-1, HLA-DR, and CD69 on CD8⁺ T cells in EBOV-exposed versus unexposed RhMs. (F) Representative dot plots of CD8⁺ T cells expressing CXCR3⁺ cells are shown for unexposed (left panel) and EBOV-exposed (right panel) samples. Circled areas represent CXCR3⁺CD8⁺ T cells. (G) The frequency of CXCR3⁺CD8⁺ T cells is presented for EBOV-exposed and unexposed RhMs in columns with means. (H) The proportion of CD20⁺ B cells, CD11b⁺ macrophages, and CD11b⁺CD11c⁺ DCs in Mes LN from EBOV-exposed versus unexposed RhMs. (I) The proportion of CD20⁺ B cells that are CD95⁺, CD38^–, HLA-DR⁺, and IgM^– in Mes LN from EBOV-exposed versus unexposed RhMs in columns with means (n =2). EBOV, Ebola virus; Mes LN, mesenteric lymph node; RhMs, rhesus monkeys; UMAP, Uniform Manifold Approximation and Projection.