CD112 Supports Lymphatic Migration of Human Dermal Dendritic Cells

Abstract

Dendritic cell (DC) migration from peripheral tissues via afferent lymphatic vessels to draining lymph nodes (dLNs) is important for the organism's immune regulation and immune protection. Several lymphatic endothelial cell (LEC)-expressed adhesion molecules have thus far been found to support transmigration and movement within the lymphatic vasculature. In this study, we investigated the contribution of CD112, an adhesion molecule that we recently found to be highly expressed in murine LECs, to this process. Performing in vitro assays in the murine system, we found that transmigration of bone marrow-derived dendritic cells (BM-DCs) across or adhesion to murine LEC monolayers was reduced when CD112 was absent on LECs, DCs, or both cell types, suggesting the involvement of homophilic CD112-CD112 interactions. While CD112 was highly expressed in murine dermal LECs, CD112 levels were low in endogenous murine dermal DCs and BM-DCs. This might explain why we observed no defect in the in vivo lymphatic migration of adoptively transferred BM-DCs or endogenous DCs from the skin to dLNs. Compared to murine DCs, human monocyte-derived DCs expressed higher CD112 levels, and their migration across human CD112-expressing LECs was significantly reduced upon CD112 blockade. CD112 expression was also readily detected in endogenous human dermal DCs and LECs by flow cytometry and immunofluorescence. Upon incubating human skin punch biopsies in the presence of CD112-blocking antibodies, DC emigration from the tissue into the culture medium was significantly reduced, indicating impaired lymphatic migration. Overall, our data reveal a contribution of CD112 to human DC migration.

Keywords: CD112; dendritic cells; human; lymphatic endothelial cells; lymphatic migration; nectin-2.

Figure 1

CD112 is expressed in BM-DCs and LECs and supports DC transmigration. (A) Flow cytometry analysis of immature (−LPS) and LPS-matured (+LPS) BM-DCs (gated on live/single cells). (B) Summary of the delta mean fluorescent intensity (∆MFI; specific-isotype staining) values of CD112 expression of 11 independent experiments. (C–F) FACS analysis of CD112 expression in (C) LPS-matured BM-DCs and (E) primary LN-LECs, derived from WT and CD112 KO mice. (D,F) Summary of the ∆MFI values of CD112 expression of 4–6 independent experiments. Data points of the same experiment in (B,D,F) are connected by a line, and the mean ΔMFI values are indicated by horizontal lines. (G) Set up of the transmigration experiments to investigate the transmigration of BM-DCs (WT or KO) across an LEC monolayer (WT or KO). (H) Impact of ICAM-1 blockade on transmigration of WT BM-DCs. (I,J) Impact of loss of CD112 in either (I) LECs or (J) BM-DCs on transmigration. (K) Impact of simultaneous loss of CD112 in LECs and BM-DCs on transmigration. For each condition in (H–K), one representative experiment with n = 3 technical replicates is shown on the left, and a summary of the averages of 4 independent experiments (biological replicates, each experiment in a different color) is shown on the right. Data points of the same experiment are connected by a line. (L) Adhesion assay of WT and KO BM-DCs to WT or KO lymphatic endothelium. The pool of two independent experiments with three replicates per condition is shown (each dot represents a sample). # BM-DCs: number of BM-DCs. Data in all graphs show mean ± standard error of the mean (SEM). * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns: not significant.

Figure 2

CD112 expression is high in LECs but low in DCs present in murine skin. (A,B) FACS analysis was performed to detect CD112 expression in dermal LECs and BECs. (A) Depiction of the gating strategy in one representative experiment. (B) Summary of the delta mean fluorescent intensity (∆MFI; specific-isotype staining) values of CD112 expression observed in 5 independent experiments. (C–G) Impact of TPA-induced skin inflammation on the expression of CD112 in LECs. (C) Schematic depiction of the experiment: Inflammation was induced in the murine ear skin by topical application of TPA and the ear skin and draining auricular LNs analyzed 24 h later. (D–G) FACS analyses were performed to quantify CD112 expression levels in LECs present in control or inflamed tissues. (D,E) Analysis of murine ear skin and (F,G) auricular LN single-cell suspensions. (E,G) The summary of ∆ MFI values was recorded in 5–6 different experiments performed in one control (CTL) and one TPA-inflamed (TPA) ear skin. (H,I) FACS gating and quantification of CD112 expression in DCs present in CTL and TPA-inflamed ear skin. (H) Gating strategy and (I) summary of ∆MFI values recorded in 3 different experiments. (J–P) Crawl-out experiments. (J) Schematic depiction of the experiment performed to evaluate CD112 expression in (K–M) DCs that had emigrated from murine ear skin into the culture medium or in (N–P) DCs that had remained in the cultured ear skin at the end of the experiment. Representative (K,N) FACS dot plots (gating on single/live cells), identifying DCs as MHCII⁺CD11c⁺ cells. (L,O) Representative histogram plots showing CD112 expression in WT and KO DCs as well as the corresponding fluorescence minus one (FMO) control. (M,P) Summary of ∆MFI values (defined as specific staining—FMO) recorded in 4 different experiments performed with one WT and one KO mouse each. Data points in (B,E,G,I,M,P) of the same experiment are connected by a line.

Figure 3

Loss of CD112 does not impact the in vivo migration of adoptively transferred or endogenous DCs to dLNs. (A–D) Adoptive transfer experiment. (A) Scheme of the experiment. (B) Gating strategy to identify fluorescently labeled adoptively transferred BM-DCs in popliteal LNs. (C) The ratio of KO–WT DCs recovered from popliteal LNs draining control (CTL) or CHS-inflamed (CHS) footpads of WT or KO mice. (D–J) FITC painting experiment. (D) Scheme of the experiment. (E) ΔEar thickness, defined as the difference between the ear thickness measured at the start and at the end of the experiment. (F) Cellularity and (G) weight of the ear-draining auricular LN at the end of the experiment. (H) Gating strategy to identify and quantify the number (#) of (I) all CD11c⁺MHCII^hi migratory DCs (mDCs) and (J) FITC⁺ mDCs. Summaries of three (A–D) and two (D–J) independent experiments, each with 2–7 mice per condition, are shown. Each dot represents one mouse. Mann–Whitney t-test was used. Red bars in all graphs show the mean. ns: not significant.

Figure 4

Blockade of CD112 decreases in vitro transmigration of human moDCs across human dermal LEC monolayers. (A–C) Analysis of CD112, DNAM-1, TIGIT and CD113 expression in in vitro-differentiated (A) immature (−LPS) and (B) LPS-matured (+LPS) human moDCs. LPS was added 24 h prior to FACS analysis. Representative FACS plots are shown in (A,B). (C) Summary of the delta mean fluorescent intensity (∆MFI; defined as specific-isotype staining) values recorded for each corresponding marker in 3–6 independent experiments (biological replicates). Data points of the same experiment are connected by a line, and the means of the ΔMFI values are indicated by horizontal red lines. (D,E) Analysis of CD112, DNAM-1, TIGIT and CD113 expression in primary human dermal LECs. (D) Representative FACS histograms recorded upon gating on CD31⁺podoplanin⁺ cells, and (E) summary of the MFI values recorded for all markers and corresponding isotype controls in 4–5 independent experiments performed on LECs from two different donors. Data points of the same experiment are connected by a line, and the means of the MFI values are indicated by horizontal red lines. (F–I) Transmigration experiments involving human moDCs and human dermal LECs, performed in the presence/absence of (F,G) αICAM-1 or of (H,I) αCD112 or the corresponding isotype controls; (F–I) The number of transmigrated DCs (# DCs) was assessed. (F,H) show representative results from one representative experiment with n = 6 technical replicates per condition. (G,I) show the summaries of four independent experiments (i.e., different biological replicates, shown with different colors) with 3–6 replicates per condition. The averages from each experiment are connected by a line. The standard error of the mean (SEM) is shown; the Mann–Whitney t-test was used. * p < 0.05; ** p < 0.01.

Figure 5

CD112 is expressed by DCs and LECs in human skin. (A–D) FACS-based analysis of CD112 expression in endothelial cells and DCs present in human skin. (A,C) Gating strategy used to detect CD112 expression in (A) BECs and LECs and (C) DCs. (B,D) Summary of mean fluorescent intensity (MFI) values of CD112 expression in (B) LEC and BECs or (D) HLA-DR⁺ CD86⁺ DCs in 2 independent experiments (i.e., different biological replicates) was analyzed. Data points of the same experiment are connected by a line. (E,F) Confocal images of human skin sections depicting (E) CD112 expression (white) by dendritic cells (examples indicated by white arrows), identified as HLA-DR⁺ (green) and CD11c⁺ (red). Scale bar = 100 μm (F) CD112 expression (white) by lymphatic vessels, LYVE-1 (green) and PLVAP (red). Scale bar = 100 μm. (G) Top: Gating strategy and Bottom: representative histogram plot showing CD112 expression on DCs that had emigrated from a human breast skin punch biopsy. (H) Crawl-out experiments from punch biopsies derived from either breast or abdominal skin were performed in the presence of a CD112-blocking antibody or media/isotype control (CTL) in the culture medium. Top: Representative FACS gating plot from abdominal skin. Bottom: Quantification of emigrated HLA-DR+CD86⁺ DCs. Pooled data from 5 independent experiments with 4–10 punches per condition are shown. (I) Crawl-out experiment from abdominal skin punch biopsies to verify the expression of CD112-binding partners DNAM-1, TIGIT and CD113 on human DCs, identified as live, HLA-DR⁺ cells. Representative stainings from one out of three independent experiments are shown. The mean and standard deviation (SD) are shown in (H). Mann–Whitney t-test was used. ** p < 0.01.