iNKT cells coordinate immune pathways to enable engraftment in nonconditioned hosts

(A) Flow cytometric analysis of murine bone marrow samples collected at 3 mo posttransplantation showing staining of the CD34-negative human population for myeloid lineage markers (CD33 and CD123). Plot on right shows numbers of human myeloid cells detected in one femur bone of mice in the indicated treatment groups, with bars showing the median for each group (symbol below dotted line indicates no specifically detectable cells). Results aggregated from nine independent experiments; P-values calculated using a two-tailed Mann–Whitney test. (B) Staining for B-lineage cells, and aggregated data for numbers of B cells in bone marrow. (C) Staining for T cells and NK cells, and aggregated data for numbers of T cells in bone marrow. (D) Aggregated data for phenotypic analysis of T cells detected in bone marrow of mice that received cord blood mononuclear cells alone (light symbols) compared with cord blood mononuclear cells + invariant natural killer T cells (dark symbols).

(A) Human compartment in bone marrow of NSG mice transplanted with CBMCs and iNKT cells show significantly greater ratio of B cells to myeloid lineage cells than in NBSGW mice transplanted with purified CD34⁺ HSPCs. (B) Flow cytometric analysis of human cells in bone marrow at the indicated time points posttransplantation showing the emergence of B-lineage cells. (C) Mean (±SEM) numbers of human B cells detected in murine bone marrow (left plot) or spleen (right plot) at the indicated times postengraftment. Data are aggregated from six independent experiments; with each symbol representing the average of three to eight mice. (D) Flow cytometric analysis of human cells in spleen of an NSG mouse given CBMCs and iNKTs at 6 mo posttransplantation. (E) Immunohistochemical analysis of serial tissue sections from the spleen of a mouse transplanted 10 mo earlier with human CBMCs and iNKT cells. Left panel shows co-staining for human CD20 (brown color) and Bcl-6 (dark purple color; examples of Bcl-6⁺ cells indicated by yellow arrows). Middle panel shows staining for human IgM (μ chain). Right panel shows staining for human IgG (γ chain). Images from light microscopic analysis at 10× magnification. (F) Plasma samples were collected at 5–9 mo posttransplantation, and tested using an ELISA specific for human Igκ light chain. Total amounts of Igκ were estimated by comparing plasma sample titers to a pooled human AB serum standard, and multiplying by previously determined values for human immunoglobulin (Cassidy & Nordby, 1975). Each symbol represents the mean of three replicate analyses of a plasma sample from an individual mouse. Data aggregated from three independent experiments; P-value calculated by two-tailed Mann–Whitney test.

(A) Analysis of the number of total human cells detected in one femur of NSG mice transplanted with 0.5–1 × 10⁵ purified CD34⁺ cells alone or in combination with 1 × 10⁶ allogeneic CD4⁺ iNKT cells. (B) Analysis of the amount of cell–cell interaction between iNKT cells and the three most abundant cell types found in cord blood. iNKT cells were labeled with a red dye and placed into tissue culture wells with equivalent numbers of purified cord blood monocytes, T cells, or B cells that had been labeled with a green dye. Fluorescence microscopic images were taken at 20× magnification every 60 min using an IncuCyte Live Cell imaging system. The plot shows the mean ± SEM area of fluorescence overlap calculated from three replicate wells for each co-culture condition. (C) Flow cytometric staining of CD1d (filled histograms) compared with isotype (dotted lines) on cord blood T cells (upper panel) or monocytes (lower panel). (D) NSG mice were transplanted with 0.5–1 × 10⁵ purified CD34⁺ cells in combination with the indicated cell types. Left plot shows number of human CD34⁺ cells per femur, and right plot shows number of lineage⁺ human cells (excluding T cells) per femur at 3 mo posttransplantation. Results aggregated from three independent experiments; P-values calculated using a two-tailed Mann–Whitney test.

(A) Purified CD34⁺ HSPCs were seeded at 2.5 × 10⁴ cells/well and cultured in the lower wells of transwell plates in medium containing a hematopoietic cytokine cocktail. Where indicated, iNKT cells were added to the transwell inserts alone or in combination with CD3/CD28 Dynabeads. After 10 d, the number of lineage⁺ cells in the lower transwell was quantitated. Plot shows aggregated results from six independent experiments with bars at the means and P-values calculated by a two-tailed Mann–Whitney test. (B) CD4⁺ iNKT cells were exposed to plate-bound recombinant CD1d–Fc fusion protein that had been pretreated with the indicated concentrations of α-GalCer lipid antigen or with vehicle alone. Plot shows amounts of IL-3 (left axis) or GM-CSF (right axis) detected in the culture supernatant after 24 h. Symbols show means of three replicate wells, with error bars showing standard deviations (not always large enough to be visible). (C) iNKT cells, isolated cord T cells, and/or autologous cord monocytes were cultured alone or in the indicated combinations for 48 h, and amounts of GM-CSF and IL-3 in the culture supernatants were quantitated. Each symbol represents the mean amount of cytokine detected from two to three replicate wells of an independent analysis, with bars showing means of aggregated results. P-values were calculated using a two-tailed Mann–Whitney test for each condition in comparison to iNKT cells alone. (D) Analysis of human GM-CSF and IL-3 levels in bone marrow or blood plasma of NSG mice transplanted with cord blood mononuclear cells alone or cord blood mononuclear cells + iNKTs. Each symbol represents the mean amount of cytokine detected from an individual mouse; P-values calculated using a two-tailed unpaired t test. (E) Purified CD34⁺ HSPCs were seeded at 2.5 × 10⁴ cells/well and cultured in the lower wells of transwell plates in medium containing a hematopoietic cytokine cocktail. Where indicated, iNKT cells, cord monocytes, and/or autologous cord T cells were added to the transwell inserts. After 10 d, the number of lineage⁺ cells in the lower transwell was quantitated. Plot shows aggregated results from five independent experiments with bars at the means. P-values were calculated using a two-tailed Mann–Whitney test for each transwell culture condition in comparison to HSPCs cultured alone.

(A) NSG mice were transplanted with 5 × 10⁶ total cord blood mononuclear cells (CBMCs) or 3 × 10⁶ T-cell–depleted CBMCs. Left plot shows number of CD34⁺ human cells and right plot shows number of lineage⁺ human cells (excluding T cells) in one femur at 3 mo posttransplantation. Results aggregated from seven independent experiments; P-values calculated by two-tailed Mann–Whitney test. (B) Purified HLA-A2⁺ CD34⁺ HSPCs were cultured alone or in the presence of CD34-depleted HLA-A2^- CBMCs in medium containing a hematopoietic cytokine cocktail. Where indicated, anti-CD3, and anti-CD28 antibodies were added to activate the cord T cells, and anti-IFN-γ or anti–TNF-α–blocking antibodies were added. Plot shows the fold expansion of HLA-A2⁺ cells after 7 d (mean ± SD of three replicate cultures), with P-values calculated using a two-tailed unpaired t test. (C) Samples of blood plasma were collected at the indicated times from mice transplanted with CBMCs alone or with iNKT cells, and analyzed for human IFN-γ by ELISA. Symbols represent mean ± SEM from 8–14 individual mice; data aggregated from four independent experiments. **Indicates P = 0.0033 as determined by two-tailed Mann–Whitney test. (D) Cord blood T cells were labeled with CTV dye and co-cultured with autologous monocytes alone, or with monocytes and iNKT cells, in medium containing IL-7 and IL-2 to drive T-cell expansion. Left plot shows expansion index of the cord T cells after 3–7 d, as determined by flow cytometric analysis of CTV staining. Middle and right plots show percentage of cord T cells staining positively for TNF-α (middle) or IFN-γ (right) after PMA and ionomycin stimulation. P-values calculated by two-tailed nonparametric paired t test. (E) NSG mice were transplanted with isolated cord T cells and autologous monocytes in the presence or absence of allogeneic iNKT cells. Samples of blood plasma were collected at the indicated times posttransplantation and analyzed for human IFN-γ by ELISA; symbols represent mean ± SEM from 5 mice tested in parallel.

(A) Isolated CD4⁺ cord blood T cells were cultured for 3–5 d with anti-CD3 and anti-CD28 antibodies in the presence or absence of transwell inserts containing iNKT cells and cord blood monocytes. Plot shows percentage of T cells staining positively for intracellular IFN-γ after PMA/ionomycin stimulation; paired symbols show results from independent experiments. P-value calculated by two-tailed nonparametric paired t test. (B) Flow cytometric staining of cord T cells for PGE₂ receptors EP3 and EP4. Filled histograms show staining by specific antibodies; dotted line shows isotype control antibody. (C) iNKT cells and cord blood monocytes were co-cultured for 24 h and PGE₂ in the supernatant was quantitated using an enzyme assay. Symbols represent means from independent experiments. (D) Cord blood T cells were cultured for 3–4 d with anti-CD3 and anti-CD28 antibodies in medium containing 500 pg/ml PGE₂, then stimulated with PMA/ionomycin. Plot shows the IFN-γ⁺ T cells as a percent of the response by parallel cultures of cord T cells that were not exposed to PGE₂. Symbols show results from independent experiments. P-value calculated by two-tailed one sample t test. (E) Cord blood T cells were cultured for 3–5 d with anti-CD3 and anti-CD28 antibodies in the presence of transwell inserts containing iNKT cells and cord blood monocytes, then stimulated with PMA/ionomycin. Before co-culture, the monocytes were pretreated with dexamethasone (500 ng/ml) or Cox-2 inhibitor (NS-398, 10 μM), or the cord T cells were pretreated with an inhibitor of EP4 (L-161,982, 10 μM). Plot shows IFN-γ production from the transwell co-cultures as a percent of the response by positive control T cells that were cultured with anti-CD3/CD28 alone (dashed line). Bars represent mean ± SEM of two to three independent experiments. (F) NSG mice were transplanted with cord blood mononuclear cells and iNKT cells and blood samples were removed at the indicated times and tested for human IFN-γ by ELISA. Where indicated, the monocytes were isolated and treated with 500 pg/ml dexamethasone, then washed and added back to the cord blood mononuclear cells before injection. Symbols represent mean ± SEM from five mice tested in parallel.