CD19/BAFF-R dual-targeted CAR T cells for the treatment of mixed antigen-negative variants of acute lymphoblastic leukemia

Abstract

Chimeric antigen receptor (CAR) T cells targeting CD19 mediate potent antitumor effects in B-cell malignancies including acute lymphoblastic leukemia (ALL), but antigen loss remains the major cause of treatment failure. To mitigate antigen escape and potentially improve the durability of remission, we developed a dual-targeting approach using an optimized, bispecific CAR construct that targets both CD19 and BAFF-R. CD19/BAFF-R dual CAR T cells exhibited antigen-specific cytokine release, degranulation, and cytotoxicity against both CD19-/- and BAFF-R-/- variant human ALL cells in vitro. Immunodeficient mice engrafted with mixed CD19-/- and BAFF-R-/- variant ALL cells and treated with a single dose of CD19/BAFF-R dual CAR T cells experienced complete eradication of both CD19-/- and BAFF-R-/- ALL variants, whereas mice treated with monospecific CD19 or BAFF-R CAR T cells succumbed to outgrowths of CD19-/BAFF-R+ or CD19+/BAFF-R- tumors, respectively. Further, CD19/BAFF-R dual CAR T cells showed prolonged in vivo persistence, raising the possibility that these cells may have the potential to promote durable remissions. Together, our data support clinical translation of BAFF-R/CD19 dual CAR T cells to treat ALL.

Figure 1.. Development of prototype BAFF-R/CD19 dual-targeting CAR T cells.

A. Schematics depict the arrangement of BAFF-R and CD19 targeting scFvs designed for each dual-targeting CAR construct. B. Human CD8 naïve cells from healthy human donors were enriched, activated and transduced with lentiviral vector encoding 4–1BB costimulatory molecules and GFP for single CAR or EGFR for dual CAR. After 7 days in vitro expansion, GFP+ or EGFR+ CAR T cells were enriched and expanded for another 7 days before functional assay. FACS analysis plots show CAR positive T cells following enrichment. CAR reporter genes were gated on T cells: GFP-positive single-targeting CAR T cells (top panel); EGFR positive dual CARs (lower panel). C. FACS histogram show BAFF-R or CD19 expression in Nalm-6 ALL tumor lines, including wild-type, CD19−/−, and BAFF-R−/− variants. Cells were stained by BAFF-R-APC, CD19-APC or isotype control antibodies. D. Calculated specific lysis are plotted from a cytotoxicity assay against Nalm-6 ALL tumor lines, including wild-type (WT), CD19−/−, and BAFF-R−/− at 10:1 of E:T ratios. Effector CAR T cells were CD19-BAFF-R(t) and CD19-BAFF-R(l) dual-targeting CARs and BAFF-R and CD19 single-targeting CARs. Mock T cells were used as negative control. Experiments were conducted in triplicate and analyzed by a Student’s t-test. ***P<0.001, ****P<0.0001, NS: not statistically significant.

Figure 2.. Functional evaluation of CD19/BAFF-R dual-targeting CAR T cells.

Human CD4 and CD8 naïve cells were enriched, activated and transduced with lentiviral vectors encoding CD19-BAFF-R(t) and CD19-BAFF-R(l) CARs separately. After 7 days in vitro expansion, EGFR+ CAR T cells were enriched and expanded for another 7 days before functional analysis. A. FACS analysis plots show CD4 and CD8 CAR positive T cells following the EGFR enrichment. Purity of CAR T cells is depicted. B. Calculated specific lysis are plotted from a cytotoxicity assay against Nalm-6 CD19−/−and Nalm-6 BAFF-R−/− variants at 10:1 E:T ratio. Effector CARs are CD19-BAFF-R(t) and CD19-BAFF-R(l) dual-CARs for both CD4 and CD8 CAR T cells. Mock T cells were used as negative control. Experiments were conducted in triplicate and analyzed with a Student’s t-test. **P<0.01, ***P<0.001. C. Bioluminescent imaging of NSG Mice (n=5/group in one experiment) inoculated with mixed tumor (0.1 ×10⁶ Nalm-6-CD19−/− plus 0.25 ×10⁶ Nalm-6-BAFF-R−/−). Mice were administered with a single infusion of 2.5 ×10⁶ CD4 T cells + 1 ×10⁶of CD8 T cells, either CD19-BAFF-R(t) or CD19-BAFF-R(l) dual-targeting CAR T cells on day 10. Mock T cells from the same donor were used as control. D. Flux (photons/sec) was determined by measuring bioluminescence every 10–20 days E. Kaplan-Meier survival curve show the results after monitoring mice for up to 90 days. Log-rank test: **p<0.01.

Figure 3.. Antitumor activity of CD19-BAFF-R(l) dual-targeting CAR T cells manufactured in a clinically relevant platform.

A. Naïve and central memory T cells (Tn/mem) were enriched, activated and transduced, and expanded for 14 days. Growth curve plots the expansion of CD19-BAFF-R(l) dual CAR T cells. Non-transduced T cells (Non-CAR) were isolated from the same donor and cultured in parallel. B. FACS analysis of CAR transduction efficiency of CD19-BAFF-R(l) dual-CAR T cells at MOI=2. Samples were analyzed at the end of CAR production on day 14 and analyzed with anti-EGFR-APC antibody. C. Bar Graph shows ELISA measurements of IFN-γ released by CD19-BAFF-R(l) dual CAR T cells. Dual CAR T cells were co-cultured for 24 hours with either BAFF-R−/− or CD19−/− Nalm-6 B-ALL cell lines at 10:1 E:T ratio. IFN-γ was measured from sampled supernatant by ELISA. Mock T cells from the same donor and irrelevant HL-60 cell line were used as controls. Experiments were conducted in triplicate and analyzed by a Student’s t-test; ***P<0.001, ****P<0.0001, NS: not statistically significant. D. FACS plots of CD19-BAFF-R(l) dual CAR T-cell functional potency as measured by a CD107a degranulation assay. 19-BAFFF-R(l) dual CAR T cells were co-incubated with either BAFF-R−/− or CD19−/− Nalm-6 B-ALL cell lines. Non-transduced T cells from the same donor (Non-CAR) and irrelevant HL-60 cell line were used as controls. Percentages of CD107a from gated CD4 and CD8 CAR are presented. E. Specific lysis induced by CAR T cells against Z138 MCL tumor lines, including wild-type (WT) and CD19−/− at 10:1 of E:T ratios. Effector CAR T cells were CD19-BAFF-R(l) dual-targeting and CD19 single-targeting CARs. Mock T cells were used as negative control. Experiments were conducted in triplicate and analyzed by a Student’s t-test. ****P<0.0001, ***P<0.001, **P<0.01. F. Bioluminescent imaging of NSG Mice (n=5/group in one experiment) challenged with mixed tumor (0.1 × 10⁶ Nalm-6-CD19−/− plus 0.1 × 10⁶ Nalm-6-BAFF-R−/−). Mice were administered with a single infusion (1 × 10⁶ CAR T cells) of CD19-BAFF-R (l) dual-targeting CAR T cells on day 9. Mock T cells at the respective total dose levels (Mock, 2.5 × 10⁶ T cells) from the same donor were used as control. G. Flux (photons/sec) was determined by measuring bioluminescence every 10–20 days. H. Kaplan-Meier survival curve show the results after monitoring mice for up to 100 days. Log-rank test: **p<0.01, between CD19-BAFF-R (l) CAR and Mock T cells.

Figure 4.. Comparison of CD19-BAFF-R(l) dual-targeting CAR with CD19 and BAFF-R single CAR in a mixed tumor model.

A. Naïve and central memory T cells (Tn/mem) cells were enriched, activated, transduced, and expanded for 13 days. FACS analysis plots CD107a positive CD4 or CD8 CAR T cells (CAR gated) comparing CD19-BAFF-R(l) dual CAR, BAFF-R single CAR, and CD19 single CAR degranulation against Nalm-6 CD19−/−BAFF-R+/+ and Nalm-6 CD19+/+BAFF-R−/− B-ALL lines. Mock T cells from the same donor were used as a control. B. Graph shows ELISA measurements of IFN-γ released by CD19-BAFF-R(l) dual CAR T cells compared to BAFF-R and CD19 single CAR T cells. CAR T cells were co-incubated with either BAFF-R−/− or CD19−/− Nalm-6 B-ALL lines. IFN-γ was measured from sampled supernatant. Experiments were conducted in triplicate and analyzed by a Student’s t-test; **P<0.01, ***P<0.001, ****P<0.0001, NS: not statistically significant. C. Specific lysis induced by CAR T cells against Nalm-6 WT, BAFF-R−/− or CD19−/− Nalm-6 B-ALL cell lines. Effector CAR T cells were CD19-BAFF-R(l) dual-targeting CARs and BAFF-R and CD19 single-targeting CARs at 10:1, 2:1, and 0.5:1 E:T ratios. Mock T cells were used as negative control. Experiments were conducted in triplicate and analyzed by a Student’s t-test. ****P<0.0001, ****P<0.0001, ***P<0.001, **P<0.01, NS: not statistically significant. D. Bioluminescent imaging of NSG Mice (n=5/group in one experiment) challenged with mixed tumor (0.1 ×10⁶ Nalm-6-CD19−/− plus 0.1 ×10⁶ Nalm-6-BAFF-R−/−). Mice were administered a single infusion of (1 × 10⁶ CAR T cells), BAFF-R, CD19 single-targeting CAR T cells or CD19-BAFF-R(l) dual-targeting CAR T cells on day 9. Mock T cells from the same donor were used as control. E. Flux (photons/sec) was determined by measuring bioluminescence every two weeks. F. Kaplan-Meier survival curve show the results after monitoring mice for up to 100 days. Log-rank test: **p<0.01, between dual CAR and CD19, BAFF-R single CAR.

Figure 5.. Analysis of residual tumor and CAR T cells in mouse peripheral blood.

Blood was collected 46–50 days post tumor challenge and analyzed with multicolor flow cytometry. A. FACS plots depict the gating strategy used to identify CAR-T and tumor cells in mouse peripheral blood. Viable singlets were gated for CD3+ EGFR+ CAR-T cells and CD10+ tumor cells. B. FACS plots of CD3+ T cell or CD10+ tumor cells in PBS, Mock, CD19 Single CAR, BAFF-R Single CAR and CD19-BAFF-R (l) Dual CAR treated mouse blood. One mouse is displayed as representative of each cohort. C. The dot graph shows the cumulative data of CD3+ T cell or CD10+ tumor cells from five mice in each cohort. Data were analyzed by a Student’s t-test; ****P<0.0001, NS, not significant. D. Lineage analysis of tumor cells for CD19 and BAFF-R expression based off CD10 gating in PBS, Mock, BAFF-R single CAR, CD-19 single CAR, and CD19-BAFF-R(l) Dual CAR groups. E. Cumulative data of CD19 and BAFF-R expression from five mice in each cohort are presented. Data were analyzed by a Student’s t-test; ****P<0.0001, NS, not significant. F. Percentages of CAR (EGFR) T cells from CD3 gate are depicted from one representative mouse. FACS analysis plots from one representative mouse are displayed for each cohort. E. Cumulative data of CAR T cells from 5 mice per cohort are presented. Data were analyzed by a Student’s t-test; *P<0.05, NS: not statistically significant.

Figure 6. Dual CAR T cells exhibited efficient cytotoxicity against primary ALL cells

A. Seven blood samples from patients with ALL were collected and CD19 and BAFF-R expression were analyzed with flow cytometry. N=3 for CD19+BAFF-R+; N=1 for CD19+BAFF-R-; N=2 CD19-BAFF-R+; N=1 for CD19+BAFF-R+. B-C. T cell-depleted patient samples were co-cultured with BAFF-R and CD19 monospecific or CD19-BAFF-R(l) dual CAR T cells at 1:2 ratio for 6 hours. CD107a expression was analyzed with flow cytometry. Representative data of CD107a expression under EGFR (CAR) gated population and accumulative data are presented. Experiments were conducted in triplicate for each patient specimen. D. Specific lysis of CAR T cells against primary ALL cells. Effector CAR T cells were CD19-BAFF-R(l) dual or BAFF-R and CD19 monospecific CAR T cells at 10:1 E:T. Mock T cells were used as negative control. Experiments were conducted in triplicate and analyzed by a Student’s t-test. ****P<0.0001, ****P<0.0001, ***P<0.001, **P<0.01, *P<0.05, NS: not statistically significant.