Fig 1.
Production and testing of CD123-specific CARs with chimeric scFvs.
(A) Schematic diagrams of conventional and chimeric scFv specific for CD123. CARs 1 to 4: CD123-specific CARs generated by fusing VL and VH chains of mAbs specific to CD123. CARs 5–9: Chimeric scFvs created by mix-and-matching VL and VH chains. The scFvs of CARs 1–9 were fused to the signaling domains of CD28 and CD3ζ via CD8α hinge and TM domains. CAR-10 was derived by fusing the chimeric scFv from CAR 6 to the CD3ζ and CD28 endo-domains via the IgG4 hinge and CD28 TM domains. (B) Expansion kinetics of CARs 1–4 (left) and CARs 5–10 (right) over a period of 28 days from day 1 following electroporation of SB CAR plasmids. Data are pooled from 3 donors; graph displays mean ± SEM (C) CAR expression on Day 21 after electroporation. CAR expression was detected by CD123 recombinant protein fused to Fc followed by serial staining with fluorescence-labeled anti-Fc and anti-CD3 antibodies. (D) in vitro lysis of CD123+ target cells Nalm 6, TF1, 293T-parental cells, CD123-transfected 293T cells, and 123neg by CAR+ T cells. Histograms represent the mean ± SEM, n = 3. (E) CAR+ T cell killing of BM-derived target cells. Mononuclear cells were isolated from normal human bone marrow samples and sorted for expression of lineage markers into lineage-positive (Lin+) and lineage-negative (Linneg) groups. The latter presumable reflects the HSC pool. The BM-derived cells were then labeled with PKH-26 and incubated with CAR+ T cells for 2 days before vitality was assessed by flow cytometry. The percent lysis compared to controls is shown. Histograms represent the mean ± SEM of 3 replicates. (F) Interferon-γ release by CAR+ T cells after exposure to CD123. Day 28 CD123-specific CAR+ T cells were incubated for 24 hours with Nalm-6 cells (CD123+), 293T cells (CD123neg), or alone, then the supernatant tested for cytokine expression using Biolegend plex Th1 cytokine capture beads, measured by flow cytometry. Results for IFN-γ are shown; other cytokines were not detectible over background. Histograms represent mean ± SEM for 2 replicates from 2 different experiments.
Fig 2.
Production and testing of CD123 specific CAR+ T cells with CD28 vs. CD137 costimulatory domains.
(A) CAR expression in CD123-CD28 (middle) and CD123-CD137 (right) CAR+ and CARneg T cells (left) on day 1and day 35 after electroporation and co-culture on AaPC, Clone 1-CD123. T cells were detected with fluorescence-labeled anti-CD3, and CAR expression with fluorescence-labeled Fc-specific antibody binding to the IgG4 hinge. (B) Expression of iCasp9 in CAR+ T cells. The same cells tested in (A) were examined by flow cytometry for surface expression of IgG4 hinge by Fc binding, and intracellularly for iCasp9 expression. (C) Expansion kinetics of CD123-CD28 (left) and CD123-CD137 CARs (right). Total CD3+ cells and CD3+CAR+ T cells expanded on AaPC Clone 1-CD123 over a period of 35 days. Graph shows the mean ± SEM of three donors. (D) CAR copy number was determined on day 28 using primers and probes specific for the CD28 transmembrane and IgG4 hinge regions. CARneg and CAR+ Jurkat cells were used as negative and positive controls respectively, Histograms represent the mean ± SEM, n = 3
Fig 3.
Immunophenotype of CAR+ T cells with CD28 or CD137 costimulatory domains.
(A) Flow analysis of memory markers on CD3+CAR+ T cells. Flow cytometry histograms are representative images from one of three donors tested for each memory marker vs. CD45RA (Y axis). (B) Expression of CD4, CD8, and CD56 is shown as in (A) (C) Histograms represent the percentage of CAR+ T cells expressing each memory or exhaustion marker (mean ± SEM, n = 3). (D) Histograms represent the percentage of CAR+ T cells in each subset, based on flow cytometry phenotype: TNaïve (CD45RA+, CD62L+, CD95-, CCR7+), TEMRA (CD45RA+, CD62Lneg, CD95neg, CCR7neg), TEM (CD45RAneg, CD62Lneg, CD95+, CCR7neg) and TCM (CD45RAneg, CD62L+, CD95+, CCR7+) in CD123-CD28 CAR+ T cells (black bars) and CD123-CD137 CAR+ T cells (grey bars) (n = 3). Fig 2E, 2F and 2G display the quantitation of mRNA transcripts of lymphocyte genes expressed in CAR T cells as analyzed by non-enzymatic digital multiplex array. (E) Transcriptional profile of activation-, co-stimulation- and exhaustion-related genes. (F) Transcriptional profile of genes associated with differentiation phenotype and memory stage (G) Transcriptional profile of genes for cytokine receptors and markers associated with effector function.
Fig 4.
in vitro lysis of AML tumor cell lines and primary AML samples.
(A) Overlay histograms display the flow cytometric analysis of CD123 expression on AML cell lines MV4-11, Molm-13, TF1, OCI-AML3, EL4-Parental and EL4-CD123. Isotype control is shown in grey, and specific staining by the unfilled black line. (B) Specific lysis of CD123-CD28 and CD123-CD137 CAR+ T cells against AML cell lines EL4, CD123neg OCI-Ly19, MV4-11, TF1, EL4-CD123, Molm-13, and OCI-AML3 assessed with a 4 hour chromium release assay. Histograms represent mean ± SEM, n = 3 (C) Flow cytometric analysis of CD123 expression on primary AML samples used in the co-culture assay depicted in (D). Lysis of PKH-26 labeled primary AML cells by CD123-CD28 or CD123-CD137 CAR T cells at 1:1 ratio for 72 hours. CD19-specific CAR+ T cells were used as a negative control.
Fig 5.
Comparison of costimulatory domains for the treatment of AML using CD123-specific CAR+ T cells in a murine model.
(A) Schematic of the TF1 xenograft model. 2.5 × 106 TF1-effLuc-mKate cells were injected intravenously into NSG mice on day 0. On Day 5, tumor engraftment was quantified using non-invasive bioluminescence imaging (BLI), and mice were randomly divided into 3 groups: untreated (control), CD123-CD28-treated, or CD123-CD137-treated. CAR-treated mice were given infusions of T cells followed by IL-2 treatment and BLI on day 5, 11 and 20. Untreated mice received no T cells. (B) BLI images of mice display an overlay of luciferase activity, using the color scale shown on the right, displayed over the white-light image of the mice. (C) Histograms represent the luciferase activity measured by BLI for each group (** p < 0.01). (D) Kaplan-Meier curves display the survival analysis of xenograft mice treated with CD123-specific CAR T cells (** p < 0.01).
Fig 6.
Efficacy of CD123-specific CAR+ T cells for the treatment of B-ALL in a murine model.
A) In vitro lysis of B-ALL cell lines by CD123-specific CAR+ T cells measured with a 4 hour chromium release assay (B) Schematic of the RCH-ACV B-ALL xenograft model. The experimental design is similar to that shown in 5A, but T cells and cytokines were given on days 7, 14 and 21, with imaging weekly. (C) BLI imaging of the CAR-treated and untreated groups on day 28. Images are displayed as in 5B. (D) Luciferase activity measured by BLI in the CAR-treated group compared with the untreated group. (E) Kaplan-Meier curves display the survival analysis of xenograft mice treated with CD123-specific CAR T cells compared with untreated mice. ** p < 0.01.