Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Multiplex genetic engineering employing miRNA-mediated gene silencing as an alternative to CRISPR/Cas9 genome editing in mesothelin-directed CAR (M5CAR) T cells. (a) , The effects of miRNA-mediated silencing versus CRISPR/Cas9-based knockout on multiple targets (CD3ζ and β2M) were compared in T cells expressing the M5CAR (M5). The second generation M5CAR is composed of a human Msln-binding scFv (pink) and CD8a hinge and transmembrane (orange) domains fused to 4-1BB (blue) and CD3ζ (red) cytoplasmic signaling domains. (b) Screening and selection of the most effective silencing constructs. Five novel target sequences were designed for both CD3ζ and β2M ( Supplementary Table 2 ). Each sequence was individually cloned into optimized gene silencing constructs, delivered to T cells via lentiviral transduction, and assessed for silencing efficiency by flow cytometry, using mCherry expression as a reporter for transduced cells. CD3ζ_T1, CD3ζ_T2 and β2M_T5 miRNAs were selected and incorporated into the same silencing construct for CAR T cell engineering. This strategy aimed to achieve complete silencing of the TCR complex and a tuned silencing to retain low surface levels of the MHC-I complex. (c) The miRNA expression construct for silencing of CD3ζ and β2M was cloned into the vector coding for the M5CAR in a bidirectional promoter configuration, resulting in dual CD3ζ and β2M silenced (S) M5CAR T cells (S CD3ζ&β2M M5). This design optimized both miRNAs and CAR expression. CRISPR/Cas9 editing employed selected sgRNAs to engineer dual CD3ζ and β2M knockout (KO) M5CAR T cells (KO CD3ζ&β2M M5). Both single and dual CD3ζ and β2M S or KO M5CAR T cells were produced. P1: elongation factor 1 short (EFs) promoter; S: spacer; P2 and P3: human phosphoglycerate kinase (hPGK) promoter.

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: Multiplex Assay, CRISPR, Knock-Out, Expressing, Binding Assay, Selection, Construct, Sequencing, Clone Assay, Transduction, Flow Cytometry, Plasmid Preparation, Produced

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Screening of miRNA target sequences for silencing of TCR and MHC-I expression in primary T cells. (a–c) , Five miRNA target sequences (T1 to T5) against transcripts of CD3ζ (a, b) or β2M (c) were screened, and silencing efficiency was evaluated on transduced mCherry-expressing T cells. CD3ζ_T1, CD3ζ_T2 and β2M_T5 miRNAs were selected. In (b) , CD3ζ_T1 and CD3ζ_T2 miRNAs selected in (a) were combined (CD3ζ_T2_T1) for TCR optimal silencing. Representative histograms reporting TCR (via TCRαβ) (a, b) , and MHC-I (via MHC-I) (c) expression levels and their relative expression percentage are shown. N= 2 (a, c) , 3 (b) T cell donors/condition. Each symbol represents a different T cell donor. Non-parametric tests were used for statistical analysis, but the small sample size prevented statistical significance-driven conclusions. Error bars represent mean (SD).

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: Expressing

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Enhanced expansion, viability, and better-preserved memory phenotypes in silenced M5CAR T cells versus CRISPR/Cas9-edited ones. (a) , Fold expansion of M5CAR T cells expressed as the ratio of the final number of cells to the initial number of cells. For KO M5CAR T cell groups, the cell number after nucleofection was used as the initial value. M5CAR T cells, M5; M5CAR T cells with scrambled miRNA (scrambled M5); M5CAR T cells with miRNAs for single silenced (S) CD3ζ (S CD3ζ M5) or β2M (S β2M M5) transcripts; M5CAR T cells with miRNAs for combined silencing of CD3ζ and β2M (S CD3ζ&β2M M5); Mock-edited M5CAR T cells (Mock M5); M5CAR T cells with sgRNAs for single knockout (KO) of CD3ζ (KO CD3ζ M5) or β2M (KO β2M M5) genes; M5CAR T cells with combined sgRNAs for dual KO of CD3ζ and β2M genes (KO CD3ζ&β2M M5CAR). (b) , Cell death of M5CAR T cells on day 5 and at the end of expansion, measured using the LIVE/DEAD™ Fixable Dead Cell Stain Kit. (c) , T cell memory subsets in M5CAR+ T cells by CD45RO and CCR7 markers on day 5 and at the end of expansion. Naïve-like, N; Central memory, CM; effector memory, EM; terminally differentiated, TD. N= 3 T cell donors/group. In (a) , each symbol represents a different T cell donor (◼ ND587; ▲; ND610; ● ND561). Non-parametric tests were used for statistical analysis, but the small sample size prevented statistical significance-driven conclusions. Error bars represent mean (SD) (a) and mean (SEM) (b, c) .

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: CRISPR, Knock-Out, Staining

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Comparison of multiplex silencing versus deletion of TCR and MHC-I complexes in M5CAR T cells before and after TCRαβ depletion. (a) , Representative histograms, and median fluorescence intensity, reporting TCR (via CD3ϵ and TCRαβ) and MHC-I (via β2M) expression levels in silenced (S) and knockout (KO) M5CAR+ T cells. (b) , Representative dot plots of dual CD3ζ and β2M S M5CAR T cells (S CD3ζ&β2M M5) before (top row) and after (bottom row) TCRαβ depletion. M5CAR (M5) expression is shown on the x-axis, with TCR (via CD3ϵ and TCRαβ) and MHC-I (via β2M) complexes on the y-axis. (c) , Relative expression percentages of TCR and MHC complexes in S M5CAR T cells and in M5CAR T cells at the end of expansion, normalized to scramble M5CAR T cells. In S CD3ζ&β2M M5CAR T cells, mean (SD) percentages were 13.1% (3.6) for CD3ϵ, 8.9% (6.7) for TCRαβ, and 32.3% (9.3) for β2M, which result to surface silencing efficiencies of 86.9% (3.6) and 91.1% (6.7) for TCR via CD3ϵ and TCRαβ, respectively, and 67.7% (9.3) for MHC-I via β2M. (d) , Expression of TCR (via CD3ϵ and TCRαβ) and MHC-I (via β2M) complexes and M5CAR (M5) after TCRαβ depletion in M5CAR T cells with either single CD3ζ (S CD3ζ M5) or dual CD3ζ and β2M (S CD3ζ&β2M M5) silencing. (e) , Representative dot plots of dual CD3ζ and β2M KO M5CAR T cells (KO CD3ζ&β2M M5) before (top row) and after (bottom row) TCRαβ depletion. (f) , Expression of TCR and MHC complexes on KO M5CAR T cells at the end of expansion. Low percentages of KO CD3ζ&β2M M5CAR T cells still expressed CD3ϵ, TCRαβ, and β2M, corresponding to mean (SD) KO efficiencies of 90.9 (3.2) %, 90.3 (2.4) %, and 81.6 (3.2) % for TCR via CD3ϵ and TCRαβ, and for MHC-I, respectively. (g) , Expression of TCR (via CD3ϵ and TCRαβ) and MHC-I (via β2M) complexes and M5CAR (M5) after TCRαβ depletion in M5CAR T cells with either single CD3ζ (KO CD3ζ M5) or dual CD3ζ and β2M (KO CD3ζ&β2M M5) gene deletion. N= 4-6 (c) , 3 (d, f, g) T cell donors/group. Each symbol represents a different T cell donor (◼ ND587; ▲; ND610; ● ND561; ◇ ND541; ○ ND224; △ ND582). P value (*p ≤ 0.05) was determined by Wilcoxon Mann-Whitney test. In (d, f, g) , non-parametric tests were used for statistical analysis, but the small sample size prevented statistical significance-driven conclusions. Error bars represent mean (SD).

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: Comparison, Multiplex Assay, Fluorescence, Expressing, Knock-Out, MANN-WHITNEY

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: MiRNA-mediated silencing achieves a higher yield and purer population of non-alloreactive and hypoimmunogenic M5CAR T cells compared to CRISPR/Cas9-mediated editing. This table illustrates the yield and characteristics of the final product when starting with an equal number of T cells (e.g., 2 x 10 6 cells) for both silenced (S) CD3ζ&β2M M5CAR T cells (S CD3ζ&β2M M5) and knockout (KO) CD3ζ&β2M M5CAR T cells (KO CD3ζ&β2M M5). Factors considered include cell loss after nucleofection (NF) in KO CD3ζ&β2M M5CAR T cells, population doublings (PDs), transduction efficiency, and yield of TCRαβ depletion, with values based on the average from five unique T cell donors. The derived cell numbers are approximations based on these factors and rounding. For S CD3ζ&β2M M5CAR T cells, the final product would likely be a pure population of TCRαβ-depleted CAR+ T cells with MHC-I silencing. In contrast, for KO CD3ζ&β2M M5CAR T cells, the final product would consist of TCRαβ-depleted cells, including both CAR+ and non-transduced T cells, all carrying the KO. A subset of these cells would still express β2M, with the proportions of CAR+ T cells and β2M+ T cells determined by the efficiencies of transduction and KO, respectively.

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: CRISPR, Knock-Out, Transduction, Derivative Assay

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Efficient in vitro cytotoxicity by TCR and MHC-I dual silenced (S CD3ζ&β2M) M5CAR T cells. (a–c) , Viability of AsPC-1 tumor cells after co-culture with silenced (S) and knockout (KO) M5CAR T cells measured by bioluminescence imaging at a 48-hour endpoint (A, B) and in real-time over more than 5 days of co-culture (C) . N= 3 T cell donors/group. Each symbol represents a different T cell donor (◼ ND587; ▲; ND610; ● ND561). Non-parametric tests were used for statistical analysis, but the small sample size prevented statistical significance-driven conclusions. Error bars represent mean (SD).

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: In Vitro, Co-Culture Assay, Knock-Out, Imaging

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Improved metastatic control of S CD3ζ&β2M M5CAR T cells relative to KO CD3ζ&β2M M5CAR T cells, with similar antitumor effects on primary tumors. (a) , Schematic of in vivo experiments evaluating the anti-tumor efficacy of silenced (S) and knockout (KO) M5CAR T-cell therapy against PDAC. Animals were injected subcutaneously (sc) with 2 x 10 6 AsPC-1 CBG GFP cells. Once an average tumor volume of 150–200 mm 3 was reached, 1 x 10 6 M5CAR+ T cells were injected via tail vein (iv). The following T cell groups were tested, each matched for total T-cell dose injected by NTD normalization: NTD, M5CAR T cells, S CD3ζ&β2M M5CAR T cells and KO CD3ζ&β2M M5CAR T cells. No TCRαβ-depleted M5CAR T cells were used. Two independent experiments were conducted, with a total of five different T cell donors, the first using ND587 and ND610 (n= 8–9 mice/group each donor), and the second ND584, ND569, and ND627 (n= 6–7 mice/group each donor). These donors were selected based on the characterization of M5CAR T cells after six days of co-culture with tumor cells, aiming to encompass all the variability detected among T cell donors and M5CAR T cell groups. Mice were euthanized per protocol if tumor dimensions exceeded 20 mm, ulceration covered over 90% of the tumor surface, or body weight loss exceeded 20%. A fixed endpoint of 100 days for the first experiment and 66 days for the second experiment after CAR T-cell injection was set for the remaining mice. (b, c) , Tumor volume over time, either normalized to baseline volume at day -1 from CAR T cell injection (top rows) or presented as non-normalized individual mouse curves for the S CD3ζ&β2M and KO CD3ζ&β2M M5CAR T cell groups (bottom rows), up to the point when a representative number of mice remained in each group, as indicated on the x-axis. Non-normalized individual mouse curves are shown only for the S CD3ζ&β2M and KO CD3ζ&β2M M5CAR T cell groups to provide a more detailed visualization of tumor progression in the two groups under direct comparison, beyond the time point at which normalized plots including all groups become less informative due to the decreasing number of surviving animals. (d) , In the second experiment, necropsy was performed on day 66 after CAR T cell injection. The histograms for each treatment group show the percentage of mice carrying metastases, with characteristics and tissue distribution indicated by the numeric legend (1 to 6) as follows: (1) n<2 small metastases in one lung lobe; (2) n>2 small metastases in one lung lobe; (3) n>2 small lung metastases in both lung lobes; (4) large lung metastases; (5) metastases in the flank of the primary tumor extending toward the upper arm, located subcutaneously outside the peritoneum; (6) spleen metastases. N= 10, 15, 16, and 18 mice for NTD, control M5CAR T cell, KO CD3ζ&β2M M5CAR T cell, and S CD3ζ&β2M M5CAR T cell groups, respectively. P values (*p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001) were calculated by mixed-effects model with a Gaussian distribution and an identity link. Error bars represent mean (SEM).

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: Control, In Vivo, Knock-Out, Injection, Co-Culture Assay, Comparison

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Greater persistence of S CD3ζ&β2M M5CAR T cells than KO CD3ζ&β2M M5CAR T cells and stable target gene silencing in vivo . (a–c) , Concentration of CD4+ and CD8+ silenced (S) and knockout (KO) M5CAR (M5)+ T cells in mouse blood on days 16 (a) and 26 (b) , and in the spleen on day 66 (c) after injection, for all specified donors. (d) , M5CAR MFI of M5CAR T cells detected in the blood at days 16 and 26, and in the spleen on day 66 after injection. For the blood, n= 8–9 mice/group for ND587 and ND610 and n= 6–7 mice/group for ND584, ND569, and ND627. For the spleen, n= 9, 12, and 12 mice for control M5CAR T cell, S CD3ζ&β2M M5CAR T cell, and KO CD3ζ&β2M M5CAR T cell groups, respectively. (e) , Dot plots of human CD45+ cells (first and second row) and M5CAR+ T cells (third row) detected in representative blood samples on day 26 after injection in S CD3ζ&β2M M5CAR T cell (S CD3ζ&β2M M5) or KO CD3ζ&β2M M5CAR T cell (KO CD3ζ&β2M M5) groups, from the same T cell donor. (f, g) , Dot plots of human CD45+ cells (first and second row) and M5CAR+ T cells (third row) analyzed in representative spleen samples on day 66 after injection in S CD3ζ&β2M M5CAR T cell [S CD3ζ&β2M M5; (f) ], or KO CD3ζ&β2M M5CAR T cell [KO CD3ζ&β2M M5; (g) ] groups, each compared with their respective donor-derived control M5CAR T cells (M5). For (e, f, g) , first and second rows are dot plots showing the expression of TCR (via CD3ϵ) and MHC-I (via β2M), respectively, along with M5CAR (M5), gated on the human CD45+ cell population. The third row shows dot plots of CD3ϵ and β2M expression gated on the M5CAR+ T cell population. Each symbol represents a different T cell donor (◼ ND587; ▲; ND610; ▼ND584; ◓ ND569; ⬣ ND627). P values (*p ≤ 0.05; **p ≤ 0.01) were determined by Wilcoxon Mann-Whitney test. In [ (c, d) spleen], non-parametric tests were used for statistical analysis, but the small sample size prevented statistical significance-driven conclusions. Error bars represent mean (SD).

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: In Vivo, Concentration Assay, Knock-Out, Injection, Control, Derivative Assay, Expressing, MANN-WHITNEY

Journal: Frontiers in Immunology

Article Title: Multiplex engineering using microRNA-mediated gene silencing in CAR T cells

doi: 10.3389/fimmu.2025.1647433

Figure Lengend Snippet: Higher protection of S CD3ζ&β2M M5CAR T cells against allogeneic NK cells than KO CD3ζ&β2M M5CAR T cells, with similar alloreactivity prevention from allogeneic PBMCs. (a, b) Live CalceinAM high expressing TCRαβ-depleted silenced (S) and knockout (KO) M5CAR T target cells were assessed after a 4-hour co-culture with allo-NK cells at NK:T ratios of 12:1 (a) and 6:1 (b) . Data are presented as the ratio of the percentage of viable target cells in the co-culture condition to the percentage of viable target cells at baseline. In (b) , protection from allo-NK was correlated to the levels of MHC-I (via β2M) silencing across different TCRαβ-depleted M5CAR T cell donors. (c) , Proliferation of allo-PBMC responders, with or without NK cells, measured by 3 H-thymidine incorporation after six days of co-culture with irradiated TCRαβ-depleted M5CAR T cells as stimulators at a 1:1 E:T ratio. Each point in the graphs represents an independent experiment. Three M5CAR T cell donors (◼ ND587; ▲; ND610; ● ND561) and four NK cell donors (for the 12:1 ratio) and nine NK cell donors (for the 6:1 ratio) were used in panel (a) . Three M5CAR T cell donors (◼ ND587; ▲; ND610; ● ND561) and three allo-PBMC donors (for conditions with NK cells) or two allo-PBMC donors (for conditions without NK cells) were used in panel (c) . P values (*p ≤ 0.05; **p ≤ 0.01) were calculated by Wilcoxon Mann-Whitney test. Error bars represent median (IQR) (a) and mean (SD) (b, c) .

Article Snippet: CRISPR single guide RNA (sgRNA) targeting CD3ζ were designed using multiple softwares: Benchling, CRISPick (Broad Institute), IDT, ChopChop.

Techniques: Expressing, Knock-Out, Co-Culture Assay, Irradiation, MANN-WHITNEY