CD5 deletion enhances the antitumor activity of adoptive T cell therapies

Abstract

Most patients treated with US Food and Drug Administration (FDA)-approved chimeric antigen receptor (CAR) T cells eventually experience disease progression. Furthermore, CAR T cells have not been curative against solid cancers and several hematological malignancies such as T cell lymphomas, which have very poor prognoses. One of the main barriers to the clinical success of adoptive T cell immunotherapies is CAR T cell dysfunction and lack of expansion and/or persistence after infusion. In this study, we found that CD5 inhibits CAR T cell activation and that knockout (KO) of CD5 using CRISPR-Cas9 enhances the antitumor effect of CAR T cells in multiple hematological and solid cancer models. Mechanistically, CD5 KO drives increased T cell effector function with enhanced cytotoxicity, in vivo expansion, and persistence, without apparent toxicity in preclinical models. These findings indicate that CD5 is a critical inhibitor of T cell function and a potential clinical target for enhancing T cell therapies.

Fig. 1.. CD5 KO enhances CART5 phenotype as compared with mock KO CART5 cells.

(A) CD5 KO CAR T expansion protocol. (B) Representative bar graph showing expression of CD5 on the groups of engineered T cells on day 8 of expansion and representative bar graph showing MFI of CD5 on the same cells. n = 2 donors, two technical replicates. (C) Schematic of mock KO and CD5 KO CART5 cells. (D) Expansion of bead-stimulated T cells shown in population doublings (log₂-transformed fold change of total cell counts). n = 4 donors. (E) Representative bar graph showing the expression of CAR5 on the groups of engineered T cells on day 8 of expansion and representative bar graph showing MFI of CAR5 on the same cells. n = 2 donors, two technical replicates. (F) T cell memory phenotypes of each engineered T cell groups at the end of expansion. T_naïve, naïve T cells, CD45RA⁺CCR7⁺; T_CM, central memory T cells, CD45RA⁻CCR7⁺; T_EM, effector memory T cells, CD45RA⁻CCR7⁻; T_EMRA, effector memory T cells reexpressing CD45RA, CD45RA⁺CCR7⁻. n = 4 donors. (G) Expression of T cell exhaustion markers PD-1 and LAG-3 on each engineered T cell group at the end of expansion in CD8⁺ and CD4⁺ cells. n = 4 donors. One-way ANOVA was performed with Tukey correction for multiple comparisons; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Fig. 2.. CD5 KO in CART5 T cells improves engineered T cell product compared with mock KO CART5 cells.

(A) Percent cytotoxicity of UTD or CART5 cells against CD5⁺ primary Sézary or primary T-ALL cells after 48 hours. (B) UTD and CART5 cells labeled with CellTrace Violet were cocultured with irradiated Jurkat cell line at a 1:1 E:T ratio for 5 days. Left: Representative flow cytometry histograms. Right: Compilation of MFI of three technical replicates. Representative data of two independent experiments are shown. (C) NSG mice were engrafted with 1 × 10⁶ Jurkat cells (intravenously) on day −7 and randomized to receive 1 × 10⁶ CD5 KO CART5 or mock KO CART5 cells or an equivalent number of UTD controls (intravenously) on day 0. Bioluminescence imaging of tumor burden in each NSG mouse (n = 5 mice per group) engrafted with Jurkat cell line. The bolded line represents the median of each group. Representative data of two independent experiments are shown. (D) The survival rate of each treatment group in xenograft T cell neoplasm model is shown in a Kaplan-Meier survival curve. (E) Left: Absolute cell counts of hCD3⁺ T cells in 100 μl of mouse blood at day 8. Right: Absolute cell counts of hCD3⁺ T cells in 100 μl of mouse blood over time. The bolded line represents the median of each group. (F) NSG mice were engrafted with 1 × 10⁶ TH20 cells on day −7 (intravenously) and randomized to receive 1 × 10⁶ CD5 KO CART5 cells, mock KO CART5 cells, or UTD controls (intravenously) on day 0. Left: Bioluminescence imaging of tumor burden in each NSG mouse (n = 5 or 6 mice per group) engrafted with primary PDX cells. The bolded line represents the median of each group. Right: Bioluminescence flux of tumor burden of mice treated with mock KO CART5 cells or CD5 KO CART5 cells on day 41. (G) The survival rate of each treatment group in primary PDX model is shown in a Kaplan-Meier survival curve. Student’s t test was used to compare two groups; in analyses where multiple groups were compared, one-way ANOVA was performed with Tukey correction. Survival curves were compared using the log-rank (Mantel-Cox) test. **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Fig. 3.. CD5 KO enhances CART19 therapy in B-ALL and CARTmeso therapy in solid tumor model of pancreatic ductal adenocarcinoma.

(A) NSG mice were engrafted with 1 × 10⁶ Nalm6 cells (intravenous) on day −5 and injected with 1 × 10⁶ mock or CD5 KO CART19 cells (intravenous) on day 0. Bioluminescence imaging of tumor burden in each NSG mouse (n = 4 mice per group) engrafted with Nalm6 B-ALL cell line. The bolded line represents the median of each group. Representative data of two independent experiments are shown. (B) The survival rate of each treatment group in xenograft B-ALL model is shown in a Kaplan-Meier survival curve. (C) NSG mice (n = 5 mice per group) were engrafted with 1 × 10⁶ Nalm6 cells (intravenously) on day −6 and injected with 1 × 10⁶ mock or CD5 KO CART19 cells (intravenously) on day 0. Absolute cell counts of hCD3⁺ T cells in 100 μl of mouse blood at day 9. (D) NSG mice were engrafted with 15 × 10⁶ HDLM2 HL cells (subcutaneously) on day −62 and injected with 0.25 × 10⁶ mock or CD5 KO CART30 cells (intravenously) on day 0. Left: Tumor volume (in mm³) in each NSG mouse (n = 6 or 7 mice per group) engrafted with HDLM2 HL cell line. The bolded line represents the median of each group. Right: Tumor volume of mice on day 63. (E) NSG mice were engrafted with 2 × 10⁶ AsPC1 cells (subcutaneously) on day −27 and injected with 0.75 × 10⁶ mock or CD5 KO CARTmeso cells (intravenously) on day 0. Tumor volume (in mm³) in each NSG mouse (n = 7 mice per group) engrafted with AsPC1 PDAC cell line. The bolded line represents the median of each group. Representative data of two independent experiments are shown. (F) Bioluminescence imaging of tumor burden in each NSG mouse. The bolded line represents the median of each group. (G) Absolute cell counts of hCD45⁺ T cells in 100 μl of mouse blood at day 58. (H) NSG mice were engrafted with 2 × 10⁶ AsPC1 PDAC cells (subcutaneously) on day −23 and injected with 0.2 × 10⁶ mock or CD5 KO CARTmeso cells (intravenously) on day 0. Tumor volume (in mm³) in each NSG mouse (n = 4 or 5 mice per group) engrafted with AsPC1 PDAC cell line. The bolded line represents the median of each group. (I) The survival rate of each treatment group in xenograft PDAC model is shown in a Kaplan-Meier survival curve. (J) Absolute cell counts of hCD45⁺ hCD3⁺ T cells in 100 μl of mouse blood over time. The bolded line represents the median of each group. (K) Tumor volume (in mm³) in control NSG mice (n = 3) or mice treated with CD5 KO CARTmeso (n = 2) and rechallenged with AsPC1 cells on day 74. The bolded line represents the median of each group. (L) Left: Integrated GFP intensity of GFP⁺ PC3 cancer cells across 60 hours with the indicated engineered T cell treatment (E:T = 0.5:1). Right: Integrated GFP intensity of GFP⁺ PC3 cancer cells treated with mock or CD5 KO CART–HER2 at the 60-hour time point. (M) Left: Integrated GFP intensity of GFP⁺ DM6 cancer cells across 60 hours with the indicated engineered T cell treatment (E:T= 5:1). Right: Integrated GFP intensity of GFP⁺ DM6 cancer cells treated with TRAC or TRAC CD5 KO TCR-GP100 at the 60-hour time point. Student’s t test was used to compare two groups; in analyses where multiple groups were compared, one-way ANOVA was performed with Tukey correction. Survival curves were compared using the log-rank (Mantel-Cox) test. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. MOS, median overall survival; ns, not significant.

Fig. 4.. CD5 deletion enhances CAR signaling.

(A) Schematic describing the inhibitory role of CD5 in T cell activation: Upon activation, CD5 recruits several mediators to the cell membrane, including SHP1, CBL, and CBL-B. CBL ubiquitinates and promotes the degradation of PLCγ1, reducing total protein levels. SHP1 dephosphorylates LAT, an upstream positive regulator of PLCγ1, whereas CBL-B ubiquitinates and promotes its degradation. (B) Principal components analysis shows distinct grouping of mock KO and CD5 KO CART5 cells. n = 2 donors. (C) Gene set enrichment analysis identifies calcium-dependent events and DAG and IP₃ signaling as enriched pathways within CD5 KO CART5 cells. NES, normalized enrichment score. (D) For each indicated marker, (left) fold change of CD5 KO CART19 MFI/mock KO CART19 MFI of phosphorylated marker at indicated time points. (Right) MFI values of phosphorylated marker in mock KO CART19 and CD5 KO CART19 cells in two separate donors. (E) NSG mice were engrafted with 5 × 10⁶ OCI-Ly18 DLBCL cells (subcutaneously) on day −10 and injected with 4 × 10⁶ mock or CD5 KO CART19 cells (intravenously) on day 0. Bioluminescence imaging of tumor burden in each NSG mouse (n = 3 to 8 mice per group) engrafted with OCI-Ly18 cell line. The bolded line represents the median of each group. (F) Cell type annotations defined by colors indicating different cell types in the dataset. (G) Top: Uniform manifold approximation and projection (UMAP) of all clusters among mock KO CART19 and CD5 KO CART19 populations. Bottom: Volcano plot of differentially expressed genes in cluster 7 compared with all other clusters and gene set enrichment analysis of top pathways enriched in cluster 7. (H) Top enriched pathways defined by gene set enrichment analysis of all CD5 KO CART19 cells and the associated genes. (I) NSG mice were engrafted with 1 × 10⁶ Nalm6 B-ALL cells (intravenously) on day −6 and injected with 0.75 × 10⁶ mock or CD5 KO or SHP1 KO CART19 cells (intravenously) on day 0. Left: Bioluminescence imaging of tumor burden in each NSG mouse (n = 4 to 6 mice per group) engrafted with Nalm6 B-ALL cell line. The bolded line represents the median of each group. Right: The survival rate of each treatment group in xenograft B-ALL model is shown in a Kaplan-Meier survival curve [compared using log-rank (Mantel-Cox) test]. (J) NSG mice were engrafted with 2 × 10⁶ A20 BCL cells (subcutaneously) on day −22, injected with cyclophosphamide (100 mg/kg, intraperitoneally) on day −1, and injected with 0.3 × 10⁶ mock or mCD5 KO mCART19 cells (intravenously) on day 0. Left: Tumor volume (in mm³) in each BALB/c mouse (n = 4 to 7 mice per group) engrafted with A20 murine lymphoma cell line. The bolded line represents the median of each group. Middle: Tumor volume (in mm³) in each BALB/c mouse engrafted with A20 murine lymphoma on day 10 after CAR T cell injections. Right: The survival rate of each treatment group in xenograft A20 model is shown in a Kaplan-Meier survival curve (compared using the Gehan-Breslow-Wilcoxon test). Student’s t test was used to compare two groups; one-way ANOVA was performed with Tukey correction for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Fig. 5.. Comparison of CD5 KO with PD-1 KO in CAR T therapy.

(A) NSG mice were engrafted with 1 × 10⁶ Nalm6 B-ALL cells (intravenously) on day −6 and injected with 0.35 × 10⁶ mock or CD5 KO or PD-1 KO CART19 cells (intravenously) on day 0. Bioluminescence imaging of tumor burden in each NSG mouse (n = 5 mice per group) engrafted with Nalm6 B-ALL cell line. The bolded line represents the median of each group. (B) The survival rate of each treatment group in xenograft B-ALL model is shown in a Kaplan-Meier survival curve. (C) Absolute cell counts of hCD45⁺hCD3⁺ T cells in 100 μl of mouse blood at day 14. (D) NSG MHC I/II DKO mice were engrafted with 2 × 10⁶ AsPC1 PDAC cells (subcutaneously) on day −27 and injected with 0.5 × 10⁶ mock or CD5 KO or PD-1 KO CARTmeso cells (intravenously) on day 0. Tumor volume (in mm³) in each NSG-MHC I/II DKO mouse (n = 5 mice per group) engrafted with AsPC1 PDAC cell line. The bolded line represents the median of each group. (E) Bioluminescence imaging of tumor burden in each NSG-MHC I/II DKO mouse. The bolded line represents the median of each group. (F) The survival rate of each treatment group in xenograft PDAC model is shown in a Kaplan-Meier survival curve. (G) Left: Absolute cell counts of hCD45⁺hCD3⁺ T cells in 100 μl of mouse blood at day 13. Right: Absolute cell counts of hCD45⁺hCD3⁺ T cells in 100 μl of mouse blood over time. The bolded line represents the median of each group. One-way ANOVA was performed with Tukey correction for multiple comparisons; survival curves were compared using the log-rank (Mantel-Cox) test. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Fig. 6.. Real-world survival data of CD5 expression in T cells.

(A) CD5 expression in CAR T cell infusion products of patients treated with the FDA-approved CART19 product tisagenlecleucel [complete responders (CR): n = 5 patients, 29,811 total cells; nonresponders (NR): n = 8 patients, 30,949 total cells]. (B) CD5 expression in CAR T cells on day 7 of patients treated with the FDA-approved CART19 product axicabtagene ciloleucel (CR: n = 5 patients, 2727 total cells; NR: n = 6 patients, 6374 total cells). Student’s t test was used to compare two groups. ****P < 0.0001.

Fig. 7.. Development of a clinical-grade, rapidly manufactured CD5 KO CART5 product.

(A) Rapid CD5 KO CAR T manufacturing protocol. T cells are electroporated and stimulated both on day 0 and transduced with CAR lentivirus on day 1, and the final product is frozen on day 5. (B) CD5 expression, CAR5 expression, and population doublings were monitored and recorded for 4 to 5 days in CD5 KO CART5 cells manufactured by either the rapid-manufacturing or the conventional-manufacturing protocol. (C) T cell memory phenotypes of each engineered T cell group after thaw. (D) Heat map of z score normalized levels of CD107a, granulocyte-macrophage colony-stimulating factor (GM-CSF), IFN-γ, IL-2, and tumor necrosis factor–α in CD5 KO CART5_rapid and CD5 KO CART5_conv T cells, with and without 4-hour activation by coculture with Jurkat (E:T = 1:1). (E) Percent cytotoxicity of UTD or CART5 cells against CD5⁺ Jurkat T-ALL cells after 72 hours. (F) NSG mice were engrafted with 1 × 10⁶ Jurkat T-ALL cells (intravenously) on day −7 and injected with 0.5 × 10⁶ CD5 KO CART5_rapid or CD5 KO CART5_conv cells (intravenously) on day 0. Bioluminescence imaging of tumor burden in each NSG mouse (n = 5 mice per group) engrafted with Jurkat T-ALL cell line. The bolded line represents the median of each group. (G) The survival rate of each treatment group in xenograft T-ALL model is shown in a Kaplan-Meier survival curve. (H) Left: Absolute cell counts of hCD45⁺hCD3⁺ T cells in 100 μl of mouse blood on day 14. Right: Absolute cell counts of hCD45⁺hCD3⁺ T cells in 100 μl of mouse blood over time. The bolded line represents the median of each group. One-way ANOVA was performed with Tukey correction for multiple comparisons; survival curves were compared using the log-rank (Mantel-Cox) test. **P < 0.01.