Dendritic Cells (DC) Facilitate Detachment of Squamous Carcinoma Cells (SCC), While SCC Promote an Immature CD16(+) DC Phenotype and Control DC Migration

Abstract

In the inflammatory mucosal microenvironment of head and neck SCC (HNSCC), DC express CD16 and are usually in direct contact with tumor cells. Mucosal and inflammation-associated DC develop from monocytes, and monocyte-derived DC are used in HNSCC immunotherapy. However, beyond apoptotic tumor cell uptake and presentation of tumor antigens by DC, HNSCC cell interactions with DC are poorly understood. Using co-cultures of monocyte-derived DC and two established HNSCC cell lines that represent well- and poorly-differentiated SCC, respectively, we found that carcinoma cells induced significant increases in CD16 expression on DC while promoting a CD1a(+)CD86(dim) immature phenotype, similar to that observed in HNSCC specimens. Moreover, HNSCC cells affected steady-state and CCL21-induced migration of DC, and these effects were donor-dependent. The CCL21-induced migration directly correlated with HNSCC-mediated effects on CCR7 and CD38 expression on DC-SIGN-high DC. The dominant pattern seen in six out of nine donors was the increase in steady-state and CCL21-induced DC migration in co-cultures with HNSCC, while the reverse pattern, i.e., decreased DC migration in co-cultures with SCC, was identified in two donors. A split in migratory DC behavior, i.e. increase with one HNSCC cell line and a decrease with the second cell line, was observed in one donor. Remarkably, the numbers of live detached HNSCC cells were orders of magnitude higher in DC-HNSCC co-cultures than in parallel HNSCC cell cultures without DC. This study provides novel insights into the effects of DC-HNSCC interactions relevant to the tumor microenvironment.

Fig. 1

a Well (WD) or moderately differentiated (MD), oral SCC contain CD16⁺ cells with dendritic morphology. Sections of oral SCC specimens stained by IHC with HRP-DAB detection show brown CD16⁺ cells. b Cell lines selected for the study. Primary keratinocytes HTE1163, well-differentiated HNSCC 1483, and poorly differentiated HNSCC Cal27 cells were grown in Permanox slide-mounted chambers, formalin-fixed and stained by IHC for cell cycle/proliferation marker Ki67. c Detailed schematic of experimental set-up. d Phenotypes of Donor 1 DC. Monocytes differentiated with GM-CSF+IL-4 produced the attached (DC-SIGN-low) and detached (DC-SIGN-high) subsets, evaluated by flow cytometry. Numbers represent the mean fluorescence intensity (MFI) of antibody binding. NOTE: Surface phenotypes of DC from Donors 1, 2 and 3 are compared in Table 1

Fig. 2

Dendritic cells affect viability of detached HNSCC cells. HNSCC cells were plated alone or combined at 1:1 ratios with DC-SIGN-low or DC-SIGN-high DC and cultured for 3 days. Floating cell populations were stained with anti-CD11c–PECy5, Annexin V-FITC (AnnV-FITC) and propidium iodide (PI) and analyzed by flow cytometry. Dot plots from Donor 1 DC subsets with Cal27 (a) and with 1483 (c) are shown. CD11c was used to distinguish between live DC and SCC populations. Only the PI-positive SCC cells non-specifically stained with the DC-specific anti-CD11c antibody. (Supplementary Figure 1). Fold enrichment of total viable floating HNSCC cells (Cal27 and 1483, respectively) in the presence of each DC subset for all donors (mean ± SEM from two independent experiments for each donor/HNSCC cell line combination) relative to HNSCC cells cultured alone (b and d)

Fig. 3

HNSCC cells affect DC phenotype. Donor 1 DC-SIGN-low and DC-SIGN-high DC subsets were cultured for 3-days alone or with HNSCC cell lines Cal27 (a) or 1483 (b). Resulting plastic-attached and floating populations were analyzed by flow cytometry. The floating populations gated on the CD11c⁺ DC and MFI of antibody binding are shown. Data are representative of at least two independent experiments. NOTE: Statistical analysis for Donors 1, 2 and 3 is shown in Table 2. c) Donor 1 DC were cultured alone or in co-cultures, as indicated, with or without 200 ng/ml LPS for 3 days. Supernatants were stored frozen at −80°C followed by ELISA. “None”—control, no HNSCC cells (only DC+/− LPS). Results are representative of at least three independent experiments

Fig. 4

Cal27 cells affect DC migration and the expression of CCR7 and CD38 on DC. DC-SIGN-low and DC-SIGN-high subsets from Donors 1, 2, and 3 were each co-cultured for 3-days with HNSCC cell line Cal27. Floating populations from the co-cultures were analyzed for CCR7 and CD38 expression by flow cytometry (a, c, e), as well as tested in migration assays (b, d, f) to examine steady-state migration (Media) and CCL21-induced migration. a, c, e The MFI of antibody binding is in the upper right corner of each panel, where the top number is MFI on DC cultured alone and the lower bold number is the MFI on DC from the co-cultures with HNSCC cells. b, d, f Error bars = standard error for DC alone; and standard deviation for DC-Cal27 co-cultures. The data are representative of two independent experiments for each donor. NOTE: Statistical analysis of DC migration for Donors 1, 2 and 3 are shown in Table 3

Fig. 5

1483 cells affect DC migration and the expression of CCR7 and CD38 on DC. All procedures with 1483 cells were the same as with Cal27 cells (please see legend to Fig. 4). The data are representative of two independent experiments for each donor. NOTE: Statistical analysis of DC migration for Donors 1, 2 and 3 are shown in Table 3

Fig. 6

Persistence of the two identified migration patterns in SCC co-cultures with DC from multiple donors. DC-SIGN-high subsets from Donors 4–9 were each co-cultured for 3-days with HNSCC cell line Cal27 or 1483 respectively. Floating populations from the co-cultures were tested in migration assays with and without CCL21. Error bars represent standard deviations of triplicate ELISA measurements. Donors 4-6 were between 50 and 65 years of age; Donors 7-9 were between 25 and 35 years of age