Advertisement

Enhancing the antitumor functions of invariant natural killer T cells using a soluble CD1d-CD19 fusion protein

Rupali Das, Peng Guan, Susan J. Wiener, Nishant P. Patel, Trevor G. Gohl, Elizabeth Evans, Maurice Zauderer and Kim E. Nichols

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    Design and binding specificity of the CD1d-CD19 fusion protein. (A) Schematic depicting the genetic fusion of human β2m with the sCD1d and the single chain of the human anti-CD19 antibody (CD19 scFV). DNA fragments were produced by polymerase chain reaction, including sequences for flexible glycine-serine rich linker (G10S3). A 6-histidine tag was added at the C terminus for Ni-NTA purification. (B) Illustration of the antigen-loaded fusion protein. (C) Depiction of how the CD1d-CD19 fusion “links” iNKT cells to tumor cells in a CD19-specific, yet CD1d-independent, manner (right). Once loaded with agonistic iNKT-directed lipids such as αGC, it is anticipated that the fusion will engage the iNKT TCR, induce iNKT activation, and promote lysis of CD19+CD1d tumor cells (see visual abstract). (D) CD1d expression on the cell lines used in this study, as determined by staining target cells with a phycoerythrin-labeled isotype, control and anti-CD1d antibody. (E) Binding of the CD1d-CD19 fusion protein to CD1d cell lines used in this study and detection of the fusion by a phycoerythrin-labeled anti-CD1d antibody. Data from >4 independent experiments are shown. Ab, antibody.

  • Figure 2.

    αGC-loaded CD1d-CD19 fusion promotes hu-iNKT cell activation in vitro. (A) hu-iNKT cells were expanded from PBMCs as detailed in “Materials and methods.” (Top left) Representative density plot showing % human CD1d tetramer (Cd1dtet)+CD3+ cells in the expanded PBMC cultures after 7 days. (Top right) Representative density plot showing that hu-iNKT cells can be successfully expanded from PBMCs and isolated to >98% purity. (Bottom left) Representative density plot showing percent CD4+ and percent CD8+ cells gated on iNKT cells from the PBMC cultures described previously. (Bottom right) Frequencies of CD4+, CD8+, and CD4CD8 (DN) iNKT cells. Data are pooled from 6 independent donors. Each symbol in the graphs corresponds to 1 donor. (B) Freshly isolated hu-iNKT cells were cultured in plates coated with αGC-loaded or unloaded CD1d-CD19 fusion (1.0 μg/mL). After 24 hours, cells were harvested and analyzed for expression of activation markers (CD25 and CD69) by flow cytometry. (C) hu-iNKT cells were labeled with 250 nM of CFSE on day 0 and cultured as described. After 4 days, cells were harvested and analyzed for cell proliferation by flow cytometry. Numbers in the histograms indicate MFI. (D) Culture supernatants from experiments in panel A were analyzed for the levels of cytokines by MILLIPLEX MAP Human High Sensitivity T Cell Magnetic Bead Panel kit. Data are presented as mean ± standard error of the mean (SEM) and compiled from independent experiments using hu-iNKT cells from 10 normal donors. Significance was determined by unpaired 2-tailed Student t test. *P < .05, **P < .01. Ag, αGC; DN, double negative; GM-CSF, granulocyte-macrophage colony-stimulating factor; ns, not significant.

  • Figure 3.

    αGC-loaded CD1d-CD19 fusion promotes hu-iNKT cell-mediated transactivation of other immune cells in vitro. Freshly isolated hu-iNKT cells were mixed with autologous PBMCs at a 1:1 ratio (2 × 105 PBMCs: 2 × 105 NKT cells). Control wells had PBMCs only (4 × 105 cells). Cells were added to wells precoated with αGC-loaded or unloaded CD1d-CD19 fusion (1 μg/mL). After 24 hours, cells were harvested and analyzed for surface expression of HLA and costimulatory molecules on CD19+ B cells (A) and CD11c+ myeloid cells (B) by flow cytometry. (C) From the same experiments, culture supernatants were harvested and analyzed for cytokines as in Figure 2. Data are presented as mean ± SEM and compiled from independent experiments using autologous PBMCs and purified iNKT cells from 3 normal human donors. CD69 (D) and CD107a (E) expression was determined on different immune cell populations by flow cytometry. Data shown in panels A-B and D-E are representative histograms from 3 independent experiments. Numbers in the histograms indicate MFI. Significance was determined by unpaired 2-tailed Student t test. **P < .01.

  • Figure 4.

    αGC-loaded fusion promotes hu-iNKT cell cytotoxicity in vitro. (A-B) Freshly isolated hu-iNKT cells were mixed with autologous PBMCs and then added to wells pre-coated with αGC-loaded or unloaded CD1d-CD19 fusion as in Figure 3. After 24 hours, cells were harvested and analyzed for different immune cell populations by flow cytometry. (A) Representative density plots showing percent CD19+ B, percent CD4+, and percent CD8+ cells in PBMCs + iNKT cell mix after culture on plate-bound unloaded or loaded fusion protein (1 μg/mL) for 24 hours. (B) Frequencies of B (CD19+), CD4+, and CD8+ cells in PBMCs and iNKT cell cocultures plated on either unloaded or αGC-loaded CD1d-CD19 fusion protein. Data are pooled from 4 independent experiments. Each symbol in the graphs corresponds to 1 donor. (C) Freshly isolated human were plated with 51Cr-labeled EBV-LCLs at varying effector:target (E:T) ratios with no stimulus (no Ag), PBS44, or unloaded (ULFP) or αGC-loaded fusion (LFP, 1 μg/mL). After 16 to 18 hours, culture supernatants were collected, radioactivity quantified, and percent specific lysis calculated. Depicted are results obtained using 3 representative EBV-LCLs. Data are from 1 of 6 independent experiments. (D) Mean percent increase ± SEM in cytolysis of the EBV-LCLs in the presence of unloaded or αGC-loaded fusion compared with cytolysis in the presence of PBS44. Data are averaged from 6 EBV-LCL lines. (E) hu-iNKTs were plated with nonradiolabeled EBV-LCLs at varying E:T = 20:1 (40 × 103 NKT cells and 2 × 103 EBV-LCLs) ratio with no stimulus, or the unloaded or αGC-loaded CD1d-CD19 fusion (1 μg/mL). After 16 to 18 hours, supernatants were collected and IFN-γ levels were measured by enzyme-linked immunosorbent assay. Representative data (mean ± standard deviation [SD]) from 1 of 3 experiments is shown. (F) Data are as in panel D, except that in these experiments Jurkat (human, left) or EL4 (murine, right) T cells were used as targets. (G-H) Freshly isolated hu-iNKT cells were added to wells coated with plate-bound αGC-loaded or unloaded CD1d-CD19 fusion (1 μg/mL) or left untreated. (G) After 24 hours, cells were harvested and analyzed for intracellular levels of lytic molecule (granzyme B) and degranulation (CD107a). Data are representative of 3 independent experiments. Numbers in the histograms indicate MFI. (H) Compiled data (mean ± SD) from 3 independent experiments showing fold increase in granzyme B (GrB) and CD107a expression on iNKT cells plated on αGC-loaded CD1d-CD19 fusion compared with those plated on unloaded fusion. Significance in panels B, D, and H was determined by unpaired 2-tailed Student t test; and in panels C and F by 2-way analysis of variance (ANOVA) test. *P < .05, **P < .01.

  • Figure 5.

    Antigen-loaded but not unloaded CD1d-CD19 fusion activates murine iNKT cell functions in vivo. B6 mice were injected intraperitoneally with PBS44 (4 μg), αGC-loaded or unloaded CD1d-CD19 fusion (20 μg), or left untreated. After 2 (A) or 18 (B) hours, splenic and hepatic iNKT cells producing IFN-γ (A-B) or IL-4 (C) directly ex vivo were analyzed by intracellular cytokine staining and flow cytometry. Data are representative of 3 experiments with 1 to 2 mice analyzed per condition. Numbers in the histograms indicate MFI. (D) Serum was collected at different time points as indicated on the graphs and IFN-γ levels were measured by enzyme-linked immunosorbent assay. Data represent the mean ± SEM from 3 independent experiments for each time point indicated. Significance was determined by unpaired 2-tailed Student t test. *P < .05, **P < .01.

  • Figure 6.

    αGC-loaded but not the unloaded CD1d-CD19 fusion promotes murine iNKT cell immunomodulatory functions in vivo. B6 mice were injected intraperitoneally with PBS44 (4 μg), αGC-loaded or unloaded CD1d-CD19 fusion (20 μg), or left untreated. (A) After 18 hours, splenocytes were analyzed for TCR+, CD4+CD8+, NK1.1+, B220+, and CD11c+ cells expressing CD69. Surface expression of the costimulatory molecule CD86 and MHC II on splenic CD19+ B cells (B) and CD11c+ myeloid cells (C) was assessed by flow cytometry. (D) Splenocytes were stained with NK1.1 and TCRβ antibodies and the NK1.1+TCRβ cells producing IFN-γ directly ex vivo were identified by intracellular staining and flow cytometry. Data are representative of 3 experiments with 1 to 2 mice analyzed per condition. Numbers in the histograms indicate MFI.

  • Figure 7.

    Treatment with the αGC-loaded CD1d-CD19 fusion limits tumor growth in vivo. (A) Freshly isolated hu-iNKT cells were plated with 51Cr-labeled MC38-CD19+ at varying E:T ratios with no stimulus (no Ag), PBS44, or ULFP or LFP (1 μg/mL). After 16 to 18 hours, culture supernatants were collected, radioactivity quantified, and percent specific lysis calculated. Depicted are results obtained using iNKT cells from 3 independent healthy donors. (B) Mice were grafted subcutaneously with 700 000 MC38-CD19+ cells, and treatment was started 72 hours later. Mice were treated every 3 days for a total of 6 doses CD1d-CD19 with either phosphate-buffered saline (control), αGC-loaded or unloaded CD1d-CD19 fusion (40 μg), or equivalent amounts of αGC-loaded or unloaded irrelevant Cd1d-Her2 fusion (40 μg). Mice were analyzed after 3 weeks. (C) Excised tumors from 1 representative experiment. ▲, no visible or palpable tumors. Weight (D) and volume (E) of excised tumors. Data in panels D-E are presented as mean ± SD of 4 to 5 mice per group from 1 of 2 independent experiments. Significance was determined by unpaired 2-tailed Student t test (D) or 2-way ANOVA (A,E). *P < .05, **P < .01.