CXCL16-positive dendritic cells enhance invariant natural killer T cell-dependent IFNγ production and tumor control

Figures & data

Figure 1. CXCL16 expression on DCs, B cells and macrophages following iNKT cell stimulation with α-GalCer. Spleen and liver cells were harvested 0–72 h after mice were treated intraperitoneally with the glycolipid α-GalCer (4 μg). CXCL16 expression was analyzed on different immune cells by flow cytometry. (A) Representative histograms show CXCL16 expression on liver DCs (CD11c⁺ MHCII⁺), B cells (CD19⁺ B220⁺) and macrophages (CD11b⁺ F4/80⁺) under basal conditions and 18 h after α-GalCer treatment. (B) Aggregate data show CXCL16 expression on liver and spleen cell populations under basal conditions and 18 h after α-GalCer treatment (n = 3–8 per group). (C) Time course of CXCL16 expression on splenic DCs from wild-type and Jα18^−/− mice following glycolipid stimulation. (n = 4 per time point). *p < 0.05 compared to baseline. †p < 0.05 compared to wild-type.

Figure 2. In vitro iNKT cell activation in the presence and absence of CXCL16. Liver mononuclear cells were cultured for 2 h in wells coated with 0, 1 or 5 μg/mL anti-CD3, with or without 100 ng/mL of recombinant CXCL16. iNKT cells (CD1d tetramer⁺ TCRβ⁺) were stained to examine intracellular (A) IFNγ and (B) IL-4 production, and cell surface expression of (C) CD40L and (D) CD69 by flow cytometry (n = 3 per group). Sorted human iNKT cells (5 × 10⁴ TCRβ⁺Vα24Jα18⁺) were cultured in wells coated with 0, 1 or 5 μg/mL anti-CD3, with or without 100 ng/mL of recombinant CXCL16. After 24 h supernatants were harvested to measure (E) IFNγ and (F) IL-4 production (n = 5 healthy donors). *p < 0.05 compared with anti-CD3 alone.

Figure 3. In vitro cytokine responses of iNKT cells stimulated with glycolipid-loaded CXCL16^hi or CXCL16^neg DCs. CD11c⁺ DCs were enriched from the spleen by magnetic sorting and loaded overnight with α-GalCer (200 ng/mL). DCs were sorted into CXCL16^hi and CXCL16^neg subsets and incubated with sorted iNKT cells (CD1d-tetramer⁺ TCRβ⁺) for 24 h, at a ratio of 1:2, DCs to iNKT cells. Cytokine levels in culture supernatants were examined using a cytokine array (n = 5–6 per group). *p < 0.05 compared with CXCL16^hi DCs.

Figure 4. In vivo cytokine responses following adoptive transfer of α-GalCer-loaded CXCL16^hi or CXCL16^neg DCs. CD11c⁺ DCs were enriched from splenocytes by magnetic sorting and loaded overnight with α-GalCer (200 ng/mL). DCs were sorted into CXCL16^hi and CXCL16^neg subsets and injected i.v. into wild-type or Jα18^−/− mice. Unloaded DCs and Jα18^−/− mice were used as controls. Serum cytokine levels were analyzed via a multiplex cytokine array. (A) IFNγ, IL-4, and IL-12p-70 levels were measured at 2, 6 and 18 h after DC transfer (n = 5–10 per group). *p < 0.05 compared with 2 h time point, †p < 0.05 compared to CXCL16^hi DC. (B) Representative plots showing expression of CD86, CD80, CD40, I-A, and CD1d on CXCL16^hi DCs and CXCL16^neg DCs following isolation and after overnight culture. Receptor expression is also shown for the CD86^hi gated population. (C) Serum levels of IFNγ and IL-4 were measured by ELISA 18 h after adoptive transfer of α-GalCer-loaded CD86^hi CXCL16^hi or CD86^hi CXCL16^neg DCs. (n = 4 per group). *p < 0.05 compared with CXCL16^hi DC. (D) Serum IFNγ, IL-4, and IL-12p70 levels were measured by ELISA 18 h after adoptive transfer of splenic CXCL16^+/+ or CXCL16^−/− DCs (n = 6 per group). *p < 0.05 compared to CXCL16^+/+ DCs.

Figure 5. In vivo profiling of intracellular IFNγ and T-bet expression in iNKT cells following adoptive transfer of α-GalCer-loaded CXCL16^+/+ or CXCL16^−/− DCs. Mice were stimulated by adoptive transfer of α-GalCer-loaded CXCL16^+/+ or CXCL16^−/− BMDCs. The number of (A) total iNKT cells, (B) IFNγ⁺ iNKT cells and (C) T-bet⁺ iNKT cells was examined by flow cytometry 2 h or 72 h later (n = 3–6 per group). *p < 0.05 compared to unloaded DCs, †p < 0.05 compared to CXCL16^+/+ DCs.

Figure 6. Expression and role of CXCR6 in NK cell transactivation. (A) Surface expression of CXCR6 was examined on NK cells (NK1.1⁺ TCRβ⁻) and iNKT cells (CD1d-tetramer⁺ TCRβ⁺) isolated from spleen and liver using a chimeric CXCL16-Fc construct (n = 4 per group). (B and C) Expanded iNKT cells (CD1d tetramer⁺ TCRβ⁺) from wild-type donor mice were adoptively transferred (i.v. 1 × 10⁷) into recipient Jα18^−/− or Jα18^−/− CXCR6^−/− mice. Twenty-four hours later, control and iNKT cell-reconstituted mice were stimulated with α-GalCer-loaded CXCL16^hi BMDCs or unloaded control DCs (i.v. 2 × 10⁵). (B) Serum cytokine levels and (C) NK cell intracellular IFNγ staining were examined 18 h following stimulation (n = 3 per group). *p < 0.05 compared to no DC stimulation.

Figure 7. Control of metastatic B16 melanoma lesions in the liver and lung via transfer of glycolipid-loaded DCs. (A) To induce liver metastasis, wild-type mice were inoculated in the spleen with 2.5 × 10⁵ B16 melanoma cells. Five days later, mice were injected i.v. with 2 × 10⁵ α-GalCer-loaded CXCL16^hi or CXCL16^neg BMDCs, or unloaded control BMDCs generated from wild-type mice (n = 14–20 per group). *p < 0.05 compared to unloaded DCs, †p < 0.05 compared to CXCL16^neg DCs. (B) In independent experiments, 2 × 10⁵ α-GalCer-loaded BMDCs generated from wild-type or CXCL16^−/− mice were delivered 5 d after B16 inoculation (n = 9–10 per group). (C) To induce lung metastasis, wild-type mice were inoculated i.v. with 2.5 × 10⁵ B16 melanoma cells. Three days later, mice were injected i.v. with 2 × 10⁵ α-GalCer-loaded BMDCs generated from wild-type or CXCL16^−/− mice (n = 4–6 per group). Liver and lung metastasis were examined 14 d after tumor cell inoculation using image analysis software to calculate tumor coverage. (D) Serum ALT levels were measured in tumor-bearing mice following administration of wild-type or CXCL16^−/− DCs (n = 4–6 per group). *p < 0.05 compared to unloaded DCs, †p < 0.05 compared to CXCL16^−/− DCs.