Human dermal CD14⁺ cells are a transient population of monocyte-derived macrophages

Abstract

Dendritic cells (DCs), monocytes, and macrophages are leukocytes with critical roles in immunity and tolerance. The DC network is evolutionarily conserved; the homologs of human tissue CD141(hi)XCR1⁺ CLEC9A⁺ DCs and CD1c⁺ DCs are murine CD103⁺ DCs and CD64⁻ CD11b⁺ DCs. In addition, human tissues also contain CD14⁺ cells, currently designated as DCs, with an as-yet unknown murine counterpart. Here we have demonstrated that human dermal CD14⁺ cells are a tissue-resident population of monocyte-derived macrophages with a short half-life of <6 days. The decline and reconstitution kinetics of human blood CD14⁺ monocytes and dermal CD14⁺ cells in vivo supported their precursor-progeny relationship. The murine homologs of human dermal CD14⁺ cells are CD11b⁺ CD64⁺ monocyte-derived macrophages. Human and mouse monocytes and macrophages were defined by highly conserved gene transcripts, which were distinct from DCs. The demonstration of monocyte-derived macrophages in the steady state in human tissue supports a conserved organization of human and mouse mononuclear phagocyte system.

Graphical abstract

Figure 1

Tissue CD14⁺ Cells Are Phenotypically Related to Blood Monocytes and Tissue Macrophages (A) Flow cytometry of enzymatically digested skin. Gating strategy used to identify tissue macrophages (AF⁺, purple gate), CD14⁺cells (blue gate), CD141⁺DCs (red gate), and CD1c⁺ DCs (green gate) is shown. (A, lower panel) Overlay dot plots of CFSE-labeled purified blood CD14⁺ monocytes (cyan) cultured with enzymatically digested skin (red) and phenotypically compared to CD14⁺ cells (blue). Bidirectional arrows depict equivalent cells. Corresponding plots in middle panel and SSC versus HLA-DR from top panel is shown as red. Representative data from at least four skin donors are shown. (B) Relative expression of selected antigens on blood CD14⁺ monocytes and CD1c⁺ DCs, skin CD14⁺ cells, CD1c⁺ DCs, and macrophages compared to isotype control (gray). Representative data from at least three donors are shown. (C) Relative expression of ZBTB46, DCSIGN, LYVE1, F13A1, IL1A, GGT5 mRNA by skin CD14⁺ cells, CD1c⁺ DC, CD141⁺ DC, macrophages, and blood CD14⁺ monocytes. Composite data from six donors is shown, mean ± SEM, ^∗p < 0.05, Mann-Whitney U test. (D) Pseudocolor images of whole-mount skin immunostained for LYVE-1 (green), CD209 (DCSIGN) (red), and FXIIIa (blue). White arrows identify LYVE-1⁻, DC-SIGN⁺, FXIIIa^−/lo cells corresponding to CD14⁺ cells. Scale bar represents 50 μm. Representative image from at least four donors is shown.

Figure 2

Skin CD14⁺ Cells Are Derived from CD14⁺ Blood Monocytes (A) Absolute count of blood CD14⁺ and CD16⁺ monocyte subsets and CD1c⁺ DCs upon conditioning and up to 14 days after HSCT. HC, healthy controls. Data from 17 patients and 15–20 HC are shown, mean ± SEM. (B) Frequency of skin CD14⁺ cells, CD1c⁺ DCs and macrophages upon conditioning and up to 14 days after HSCT as a % of nucleated cells. Data from 17 patients are shown. Mean ± SEM. A maximum of two skin biopsies per patient were taken at different time points and were collagenase digested. Bottom panel depicts representative dot plots of skin flow-cytometry analysis. (C) Phenotype and morphology of CD14⁺ blood monocytes after culture with medium alone, HUVECS or fibroblasts for 0–3 days. Scale bar represents 10 μm. Representative data from nine different donors is shown. Overlay histogram of DC-SIGN and CD16 expression (blue) compared to isotype control (gray).

Figure 3

Skin CD14⁺ Cells Are Transcriptionally Aligned to Human Monocytes and Macrophages (A) Principal component analysis of CD141⁺ DCs, CD1c⁺ DCs, CD14⁺ cells, monocyte subsets, macrophages, and pDCs. Each symbol represents an individual sample. Rectangles depict skin subset and circles depict blood subset. Data from three to eight independent blood and skin donors are shown. (B) CMAP enrichment scores for skin CD14⁺ cells signature compared to human skin and blood monocytes and DC subsets. Each symbol represents an individual sample. Enrichment scores were significant at p < 0.001 for CD14⁺ cell signature compared to other subsets. (C) Heatmap showing 106 genes which were 2-log fold up (red) or downregulated (blue) in human monocyte-macrophages compared to DCs. p < 0.001 for each gene. Each row represents one sample. (D) Scatterplot comparing genes that were >1.5 log fold up or downregulated in both human and mouse monocyte-macrophages (blue dots and square) compared to DCs (red dots and square). p < 0.001 for each transcript.

Figure 4

Spontaneous Migration of Skin CD14⁺ Cells Does Not Occur via Lymphatic Vessels (A) Left panel shows gating strategy used to identify CD14⁺ cells and CD1c⁺CD1a⁺ DCs from live, CD45⁺, HLA-DR⁺ cells isolated by digestion (top) and spontaneous migration from skin explants (bottom). Right panel shows relative expression of CCR7 by dermal CD14⁺ and CD1c⁺ DC isolated by digestion, migration, and migration in the presence of LPS, TNF-α, and IL-1β. Representative data from at least five donors is shown. (B) Frequency (as % of HLA-DR⁺ cells) of dermal CD14⁺ cells, CD1c⁺ DCs, CD141⁺ DCs, and macrophages in skin explant medium after 0–72 hr of culture (n = 6, mean ± SEM). Solid line represents migrated cells, and dotted line represents cells in digested skin remnant. (C) Pseudocolor immunofluorescence whole-mount microscopy of human skin (T0 = freshly harvested, T48 = 48 hr culture of skin explant ex vivo). Top left panel; T0, depicts distribution of HLA-DR⁺ (green) and DCSIGN⁺ (red) cells outside LYVE-1⁺ lymphatics (blue). Top right panel: DCSIGN⁺ (red) cells outside LYVE-1⁺ lymphatics (blue). Lower panel; T48 depicts HLA-DR⁺ (green) cells mainly located within lymphatic vessels but DCSIGN⁺ cells are outside the lymphatic vessels. Bottom right; close up three-dimensional reconstruction of boxed area in left panel. Representative image from six donors is shown. (D) Heat map showing the expression of 96 genes, analyzed by Taqman Low Density Array, by CD14⁺ cells and CD1c⁺ DCs isolated by migration and enzymatic skin digestion. Data from five skin donors (pairs indicated by Roman numerals) are shown.

Figure 5

CD14⁺ Cells Are Potent Inducers of Memory T Cell Response (A) Naive T cell stimulation: proliferation of allogeneic naive CD4⁺ T cells, (determined by CFSE dilution) after coculture for 6 days with CD14⁺ cells, CD14⁺CD1c⁺ cells, macrophages, CD1c⁺ DCs from skin and CD14⁺ monocytes from blood (n = 5, mean ± SEM). (B) Memory T cell stimulation: Intracellular expression of IL-17, IL-22, IL-4, and IFN-γ by PMA and ionomycin restimulated autologous CFSE-labeled bulk CD4⁺ T cells following coculture with skin DC and macrophage subsets pulsed with Candida albicans (n = 6, mean ± SEM).

Figure 6

Murine Homolog of Human Dermal CD14⁺ Cells (A) CMAP enrichment scores for the signatures of human dermal CD14⁺ cells, CD1c⁺ DCs, macrophages, and CD14⁺ blood monocytes compared with murine CD11b⁺ dermal cell populations found in steady state and upon contact sensitization with DNFB. P1, tissue monocytes; P2 and P3, dermal monocyte-derived DC-like cells; P4 and P5, dermal macrophages, as described in (Tamoutounour et al., 2013). All enrichment scores were significant at p < 0.001. (B) Flow cytometry analysis of S100a4xRosaYFP-flox mouse dermal ear cell suspension. Values in contour plots indicate percentage of cells in the respective gates. %YFP⁻ cells of the different populations are shown in histogram (lower panel). P1, tissue monocytes; P2 and P3, dermal monocyte-derived DC-like cells; P4 and P5, dermal macrophages; as described in (Tamoutounour et al., 2013). Data shown is representative of six individually analyzed mice from two independent experiments.