Fig 2: Spatial patterns of PD-L1+ cells in mouse and human SCCs demonstrate that infiltrating leukocytes are the primary source of PD-L1 within tumors.a Percentage of PD-L1+ A223 and LY2 cells after stimulation with 2 ng/ml recombinant TGFß (orange triangles, n = 4) or PBS vehicle (black circles, n = 4) as measured by flow cytometry. b tdTomato expression (purple bars: positive, blue bars: negative) in PD-L1+ live cells isolated from mTmG mice bearing A223 tumors and treated with TGFß trap control isotype control 3×/week for 1 week, as quantified by flow cytometry (n = 5 per treatment group). c Analysis of CD11b and CD11c expression of tdTomato+ cells from c as quantified by flow cytometry (CD11b+ CD11c-: blue bars, CD11b+ CD11c+: purple bars, CD11b-: orange bars, n = 5 per treatment group). d Representative tumor from multispectral imaging analysis of 23 human HNSCC tumors stained for pan-cytokeratin (cyan), CD68 (magenta), CD11c (yellow), PD-L1 (red), and DAPI (blue). Arrowheads represent example phenotypes using the same color code. e Table of total number of cells of each indicated phenotype across all samples in (d) (horizontal bars), and counts of cells that overlap one or more phenotypes (vertical bars). The central matrix displays the combinations graphically with connected dots indicating overlapping phenotypes. Unpaired two-tailed t tests were performed for (a–c), and all error bars represent the SEM. Raw unstacked values in (b) and (c) are represented by the indicated symbols.

Fig 3: BEA promotes transcriptional changes associated with chemokine and cytokine production and TLR signaling pathway activation but are distinct from LPS stimulation. MHC IIhigh CD11c+ BMDCs sorted by flow cytometry were followed by stimulation with BEA (5 µM) or LPS (1 µg/ml) alone or BEA combined with LPS for 4 hours. Results shown are from one experiment containing 4 independent biological replicates (cells from 4 mice per group, n=4) for each condition. (A) PCA of the quadruplicate biological replicates. (B) Heatmap showing expression profile for 4,015 genes that were found to be significantly regulated in at least one of the comparisons using untreated as baseline condition. (C) Enriched Reactome pathways (upper plot) and biological process (lower plot) in DEGs in BEA treated BMDCs compared with untreated BMDCs. (D) Cytoscape representation of significantly enriched signatures in BEA treated BMDCs compared with untreated BMDCs.

Fig 4: Loss of PD-L1 Impairs Dermal Dendritic Cell (dDC) Migration following Poly(I:C)(A) Experimental design for (B)–(E), in which black and white lines indicate chimerism.(B) Representative flow plots showing FITC+ dDCs (CD11c+, MHC2hi, and B220-) in dLNs of WT and Pdl1-/- chimeric mice.(C) Number of FITC+ dDCs in the dLN 24 h after intradermal poly(I:C).(D) PD-L1 expression on FITC+ dDCs in the dLN.(E) CD80 expression on FITC+ dDCs in the dLN.(F) Experimental design for (G)–(I). WT and Pdl1-/- bone marrow-derived dendritic cells (BMDCs) co-injected in the footpad of WT mice 24 h after poly(I:C) (5 µg/footpad).(G) Representative flow plots of CD11c+ cells before transfer and from the popliteal dLN 24 h after transfer.(H) Ratio of Pdl1-/- BMDCs to WT BMDCs found in the dLN 24 h after transfer into WT hosts.(I) Number of WT and Pdl1-/- BMDCs in the dLN 24 h after transfer.For (B–E), data are shown from axillary and brachial LNs combined, n = 5–7 mice in data shown, repeated four times with similar results. For (G–I), data are shown from two popliteal LNs, n = 4 mice in data shown, repeated three times with similar results. Statistical analysis was done using a one-way ANOVA (C–E), an unpaired Student’s t test (H), or a paired Student’s t test (I). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Error bars indicate standard error of the mean. See also Figure S2.

Fig 5: Loss of PD-L1 Leads to a Reduction in the Number of Dendritic Cells (DCs) in the Draining Lymph Node (dLN) following Poly(I:C)(A) Flow cytometry gating strategy for conventional DC subsets and myeloid cells in the LN from WT and Pdl1-/- mice. PD-L1 expression shown is from naive mice and mice 24 h following poly(I:C) injection.(B) Quantification of PD-L1 expression on WT conventional dendritic cell (cDC) 1 and cDC2 at baseline and 24 h following poly(I:C), as well as Pdl1-/- DCs.(C) Number of cDC1 and cDC2 in the dLN of WT and Pdl1-/- before and 24 h after intradermal poly(I:C) injection (10 µg/mouse). Data are shown from axillary and brachial LNs combined.(D) Number of CD11b+ CD11c- cells in the dLN of WT and Pdl1-/- mice. Data are shown from axillary and brachial LNs combined.(E) Number and percentage of CD11c-high cells in the skin from the site of injection of WT and Pdl1-/- mice based on weight.For (B–E), n = 3 mice per group. Data are shown from one experiment that was repeated four times with similar results. Statistical analysis was done using a one-way ANOVA. *p < 0.05; **p < 0.01; ****p < 0.0001; n.s., p > 0.05. Error bars indicate standard error of the mean. See also Figure S1.