Prolonged cognate NK–DC interaction results in cytolysis of allogeneic DCs, visualized in dLNs by two-photon imaging. A, Velocities of individual NK cells from LNs under steady-state (syngeneic DCs; hatched bars) and activating (allogeneic DCs; ■) conditions. Arrows indicate mean NK cell velocities (syn-DCs: mean velocity = 9.7 ± 0.4 μm/min, n = 101 cells; allo-DCs: mean velocity = 10.4 ± 0.3 μm/min, n = 125 cells). B, Individual NK cell velocities versus confinement ratios (total length of track divided by the distance between the start and end point) were calculated from DC dLNs (syn-DCs: black dots; allo-DCs: red dots; n = 101). C, Distribution of NK–DC contact durations. An interaction was defined as a contact lasting >1 min. ■, Exact contact times; hatched bars, cellular interactions that could not be viewed in their entirety (i.e., movie began/ended with cells associated or conjugates moved out of imaging field prior to detachment). Arrows indicate mean NK cell contact time (minutes) with syn-DCs (4 min 52 s ± 36 s; n = 83; blue bars) and with allo-DCs when the interaction was noncytolytic (5 min 54 s ± 42 s; n = 84; yellow bars) or cytolytic (20 min 20 s ± 4 min 17 s; n = 28; red bars). *p < 0.0001. D, Time-lapse images of a lytic NK–DC interaction (yellow arrows). An individual NK cell (red) is seen engaging an individual allo-DC (green) in the presence of a syn-DC (blue). Original magnification ×20. Scale bar, 10 μm.

Allogeneic DCs fail to prime T cells or establish T cell recall responses through direct presentation in vivo. A, In vivo Ag-specific proliferative responses of adoptively transferred CFSE-labeled (OT2 × BALB/c.Thy1.1)F₁ T cells following immunization with syngeneic F1 OVA-pulsed DCs (dashed line), allogeneic B6 OVA-pulsed DCs (solid line), or PBS (dotted line) alone (left panel histogram). Percentage of proliferated cells within each dLN was calculated as follows: 100 (number of CFSE-diluted cells within each dLN ÷ total number of CFSE-diluted cells from all dLNs from each individual animal) (right panel). Equal numbers of LN cells were collected for each condition (n = 3 mice). B, In vitro proliferative responses from CFSE-labeled (OT2 × BALB/c. Thy1.1)F₁ T cells stimulated with syngeneic F1 OVA-pulsed DCs (dashed line) or allogeneic B6 OVA-pulsed DCs (solid line) (left panel histogram) and percent proliferation were calculated as follows: 100 (CFSE-diluted cells ÷ the total number of cells) (right panel). Experiments performed with parallel DC cultures used in A. C, DTH responses in B6 (upper and middle panels) or F₁ (lower panel) mice immunized with Ag-pulsed syngeneic or allogeneic DCs. After 7 d, mice were challenged with cognate Ag, and ear swelling was measured the following day (n = 4–5 mice/group). Data are representative of three independent experiments.

Direct presentation by donor DCs is insufficient to trigger acute allograft rejection. A, Cardiac survival of WT BALB/c donor allografts placed heterotopically into the abdomen of CD11c-DTR⁺ (dashed line; mean survival time [MST] = 35 ± 5.7 d; n = 6) or CD11c-DTR⁻ (dotted line; MST = 7.2 ± 0.5 d; n = 6) BM chimeras and treated i.p. every other day with 100 ng of DT for the duration of the study. Control CD11c-DTR⁺ chimeras (solid line; MST = 6.8 ± 0.5 d; n = 5) were left untreated (no DT). *p < 0.0001. B, CD11c-DTR⁺ (solid line; MST = 8.6 ± 0.2 d; n = 7) or CD11c-DTR⁻ (dotted line; MST = 7.5 ± 0.2 d; n = 6) BALB/c donors were treated i.p. with 100 ng of DT. The following day, cardiac allografts were harvested and transplanted heterotopically into the abdomen of DT-treated WT B6 recipients. WT B6 recipients were treated with DT to ensure depletion of any donor-derived passenger DCs that may have remained within the transplanted allograft. MST, mean survival time.

NK cells are sufficient to reject allogeneic DCs within dLNs. A, Experimental design timeline indicating phenotyping, cell labeling, and recovery of BMDCs from LNs. From left: 1, FACS analysis of BMDCs following LPS-induced maturation showing isotype controls (dashed lines) and DC markers (solid lines) as indicated (left panel histograms); 2, CFSE-labeled congenic (B6.Ly5.1⁺) and allogeneic (BALB/c.Ly5.1⁻) BMDCs mixed at a 1:1 ratio (middle panel histogram); and, 3, following injection into footpads or pinnae, recovery from B6.Ly5.2 WT or Rag^−/− LNs. dLNs were harvested and stained with Ly5.1 Ab, and the percentage of allogeneic (Ly5.1⁻) DCs persisting at the indicated time was determined by FACS analysis using a CFSE⁺ DC gate (right panel histograms). B and D, Kinetics of allogeneic DC rejection in WT (B) and Rag^−/− (D) recipients. C, Rejection of allogeneic DCs in WT recipients following excision of injection site to prevent continued DC migration. E, Rejection of DCs in NK cell-replete (isotype) or NK cell-deficient (anti-NK1.1) Rag^−/− recipients. Each circle represents an individual mouse (n = 3 mice per time point). Data are representative of three (B) or two (C–E) separate experiments. WT, wild-type.

In vivo priming of 4C T cells occurs preferentially by indirect presentation. A, Snapshot visualization of endogenous CD11c-EYFP⁺ DCs (green arrow), allogeneic BMDCs (orange arrow) and 4C T cells (blue arrow). Original magnification ×20. Scale bar, 20 μm. B, Individual 4C T cell velocities and confinement ratios in LN-draining allogeneic DCs (red dots; n = 68) or unimmunized control LNs (black dots; n = 63). C, 4C T cell interactions with endogenous CD11c-EYFP⁺ (■; 103 of 112) and allogeneic (dotted bar; 9 of 112) DCs were determined. A 4C: DC conjugate was defined as an individual interaction lasting >10 min.