Figure 1.
Immunophenotyping (a) and gene-expression (b) of Eμ-ALL01 cells.
Immunophenotyping was performed with the antibodies listed on the axes. Cells were first gated within a live lymphoid gate on a scatter plot. The bottom panels are isotype (iso) or fluorescence minus one (FMO) controls. (b) RT-PCR was performed for the listed genes and electrophoresed through a 1% agarose gel and stained with Ethidium Bromide. No-template (NT) and Cd8, which is expressed usually in T cells but not B cells, are included as negative controls. MWT refers to the 1 Kb Plus DNA ladder (Invitrogen). The triangles point to the 1 Kb (top) and 100 bp fragments.
Figure 2.
Progressive disease in mice after injection with Eμ-ALL01 tumor cells.
(a) Retro-orbital blood was collected from 5 mice on the days indicated after Eμ-ALL01 injection and analyzed serially over time with an ACT diff cell counter (Beckman Coulter, Brea, CA). White cell count (WBC), Hemoglobin (Hgb), and Platelets were measured and statistically analyzed with t-tests, which were non-significant for Day 0 and Day 14 means. Comparison of the Day 0 and Day 32 means by t-tests were all significant (p values are noted on graph). Error bars are the standard error of the mean (SEM). The mice were sacrificed 4 weeks after injection with Eμ-ALL01 because of clinical deterioration and retro-orbital blood, bone marrow, and spleen were harvested for anatomical (b) and cellular (c) analyses, with representative images displayed. (b) Blood and tissue were used to prepare peripheral smears (left-panel) and H&E stained slides of sections of the bone marrow (middle-panel) and spleen (right-panel). (c) Single-cell suspensions were prepared from bone marrow (BM) and spleen (SP) and then incubated with antibodies specific for B220 and IgM. The panels on the top are cells isolated from disease-free, wild-type C57BL/6 mice (B6) and the panels on the bottom were injected with the Eμ-ALL01 cells. Cells displayed have been gated on live lymphoid cells.
Figure 3.
Generation and in vitro function of the m1928z CAR.
(a) Schematic of the genetic construct GL-2A-m1928z for the reporter gene (GFP) and CAR (m1928z). Depicted are the packaging signal ψ, splice donor (SD), splice acceptor (SA), the VH and VL regions of the scFV, and the extracellular (EC), transmembrane (TM), and cytosolic (C) regions. (b) GFP-expression in mouse T cells after transduction with GL-2A-m1928z retroviral supernatant (right-panel). The left-panel displays untransduced (UNT) T cells as a control. The gene-transfer efficiency, estimated as the GFP+ population, is derived from a single experiment with a double-transduction of a bulk population of mouse T splenocytes as described [22]. (c) Cytotoxic T lymphocyte antigen-specific killing was evaluated with a Chromium release assay [39]. Target cells (EL4-mCD19) are EL4 cells retrovirally transduced with mouse CD19. Effector cells are T cells transduced with GL-2A-m1928z. Control effector cells were transduced with GL-2A-h1928z since it is identical to GL-2A-m1928z except for the scFv, which is derived from an anti-human CD19 antibody [5]. Effector to target ratio (x-axis) are based on the number of GFP+ T cells to EL4-mCD19 cells and were performed in triplicate. Killing efficiency (y-axis) was calculated as described [39]. (d) Cytokine secretion by m1928z and h1928z T cells was evaluated after stimulation with 3T3-mCD19 cells. Stimulation with 3T3-mCD19 cells was performed in triplicate and supernatants were obtained 1 day after stimulation. Each supernatant was evaluated for cytokines also in triplicate (no dilution, 3×dilution, and 9×dilution). Error bars represent the SEM.
Figure 4.
Adoptive transfer of CD19 CAR-targeted T cells into mice with leukemia.
(a) Flow cytometry of bulk mouse T splenocytes double-transduced with GL-2A-m1928z. (b) Survival curve of mice injected with Eμ-ALL01 and not treated (No Rx), treated only with cyclophosphamide (CTX), or treated with cyclophosphamide (100-200 mg/kg IP) and CD19 CAR-targeted T cells (CTX+m1928z), displayed in (a). This data (n = 28) is pooled from 2 independent experiments. Another study (Supplemental Figure S2) confirms that cyclophosphamide and control T cells mediate no survival advantage. (c) Flow cytometry of cells from the BM, spleen (SP), and lymph node (LN) of a single mouse treated with cyclophosphamide but sacrificed due to clinical deterioration. Splenocytes were also stained with isotype antibodies (ISO) as a control.
Figure 5.
Persistent in vivo B cell targeting and regeneration after CD19 CAR-targeted T cell transfer.
(a) Peripheral blood B cell counts from mice (n = 25) treated with 100 mg/kg IP cyclophosphamide (CTX) and/or CD19 CAR-targeted T cells was measured on the days listed on the x-axis. All groups were infused with Eμ-ALL01 tumor cells and the CTX group is treated with chemotherapy alone, CTX+m1928z group is treated with both chemotherapy and T cells, while the Eμ-ALL01 group is untreated. The red asterisks note time points when no samples were available since all the mice in the Eμ-ALL01 group had died. The green lines mark the range of the 95% confidence interval for the peripheral B cell count in wild-type untreated B6 mice. Error bars are the SEM. Statistical analyses were performed on both treatment groups using t tests for Days 7, 25, 40, and 61. For every one of these comparisons p<0.01. Retro-orbital blood was collected on the indicated days (Day 0 being adoptive transfer of T cells) and analyzed with an ACT diff cell counter and then incubated with anti-B220, anti-CD19, and anti-CD3 antibodies. The number of B cells was calculated as the concentration of cells multiplied by the frequency of CD19+ cells. (b) BM and spleen cells were isolated from a C57BL/6 mouse (B6) as a control, or a single mouse that had been injected with Eμ-ALL01 and treated eight-months prior with cyclophosphamide (100 mg/kg IP) and CD19 CAR-targeted T cells (m1928z). Cells were stained with anti-CD19, anti-CD3, anti-IgM, and anti-CD43 antibodies. In every row, the two far-right panels are derived from the CD19+ gate of the left-panel.
Figure 6.
CD19 CAR-targeted T cell dependence on conditioning chemotherapy and T cell dose.
T cells were transduced with m1928z and infused at increasing doses (noted on axes) into congenic mice (n = 40) that had been conditioned with increasing doses of cyclophosphamide (noted above columns). Peripheral blood B and congenic T cells were evaluated by flow cytometry for B and T cell markers 4 weeks after adoptive transfer. (b and c) A repeat experiment (n = 44) was done while holding cyclophosphamide conditioning constant (300 mg/kg), but with increasing T cell doses. Mice were sacrificed 1 and 5 weeks after adoptive transfer with CD19 CAR-targeted congenic T cells and flow cytometry for B cell (B220) and congenic T cells (Thy1.1/Thy1.2) was performed on single-cell suspensions of the BM and spleen. All counts and percentages were done with Countbright beads (Invitrogen). Statistical significance (p <0.05) was calculated by one-way ANOVA and is denoted by asterisks. For all sections of this figure error bars represent the SEM.
Table 1.
Antigen-dependence of CD19 CAR-targeted T cells in the BM.
Figure 7.
Adoptive transfer of CD8+ CD19 CAR-targeted T cells alone is sufficient for long-term persistence and B cell eradication.
(a) OTI Thy1.2+ T cells were transduced with the m1928z CAR or the m19z CAR, which has a CD3ζ signal transduction domain but lacks the CD28 signal transduction domain. Untransduced T cells (UNT) are mock-transduced cells and are included as a negative control. Donor T cells (5×106) were then adoptively transferred into congenic Thy1.1 mice (n = 19) 1 day after treatment with cyclophosphamide (300 mg/kg). (b) Peripheral blood B and OTI T cell counts were evaluated by flow cytometry 4 weeks after adoptive transfer. One-way ANOVA was done for the treatment groups' (UNT, m19z, m1928z) B cell counts (p = 0.0001) and also OTI T cell counts (p = 0.03). Error bars are the SEM. (c) A few of the mice in each group were sacrificed 7 weeks after adoptive T cell transfer. BM and spleen cells were analyzed by flow cytometry for B and congenic T cell markers. A representative plot is included from each group. Displayed cells were gated on Live and DUMP-negative cells, which were characterized with Mac1, Gr1, NK1.1, and Ter119 antibodies. (d) In another study (n = 17), peripheral blood counts of B and congenic T cells were performed from 3–22 weeks after adoptive transfer. This study included groups of control mice that were untreated wild-type (WT), or treated with cyclophosphamide alone (CTX), or treated with T cells modified with the m19Δz CAR, which lacks any signaling element. Statistical analyses (t tests) of the treatment groups (m19Δz and m1928z) were significant (p<0.05) when done on the B cell counts but not the OTI T cell counts (p = 0.47). Error bars are the SEM. * is Not Done.