Journal: bioRxiv

Article Title: In vivo CRISPR-based screen identifies ZC3H12C as a mediator of CAR-T cell dysfunction in solid tumors

doi: 10.64898/2026.04.30.721530

Figure Lengend Snippet: NSG mice bearing SKOV3 tumors were treated with a single dose of 2×10 6 HER2-28z or CD19-28z CAR+ T cells (Control CAR-T cell). A ) Tumors were measured at indicated time points. Data are plotted as mean ±SEM of tumor volume (control CAR-T, n=6 tumors) or as individual tumors (HER2-CAR-T, n=7 tumors). Representative of n>5 experiments using different healthy donors. Timepoints at which CAR-TILs were isolated from tumors for use in downstream analyses are indicated. B ) Violin plots showing mean fluorescence intensity of PD-1 (right panel) and Ki67 (left panel) in effective compared to dysfunctional CAR-TILs. Each dot represents a tumor (n=4-5 healthy donors). C ) Ex vivo real-time cytotoxicity analysis of isolated CAR-TILs at effective (upper panel) and dysfunctional (bottom panel) phases against SKOV3 tumor cells (E:T=5:1). Data are plotted as mean ±SEM of normalized cell index of a representative experiment using one healthy donor. Time when T cells were added is indicated. D ) Analysis of IFNγ production by CAR-TILs after 24h of ex vivo co-culture with SKOV3 tumor cells (E:T=3:1). The PD-1/PD-L1 axis was inhibited by the addition of PD-L1 blocking antibody Durvalumab. Data are plotted as mean ±SEM (n=3 in effective, n=2 in dysfunctional and dysfunctional + durvalumab). Each dot represents a tumor used for isolating CAR-TILs. **p<0.01 by one-way ANOVA with Tukey’s multiple comparisons test. E ) IFNγ secretion by CAR-TILs after overnight stimulation with PMA/Ionomycin. Each dot represents a tumor used for isolating CAR-TILs (n=2 healthy donors). Fold change of effective vs dysfunctional is indicated. B and E ) **p<0.01, ***p<0.0001 by two-tailed unpaired t-test. Fold change relative to effective (Ki67, IFNγ) or dysfunctional (PD-1) is shown. F ) Principal component analysis of HER2-28z CAR-T samples from preinfusion, effective, and dysfunctional phases included in the RNA-seq analysis. Number of differentially expressed genes (p-adj. value<0.05 and logFc<1 or logFc<-1) between groups is indicated (n=7 healthy donors for preinfusion; n=6 for effective; n=5 for dysfunctional). G ) Ingenuity Pathway Analysis (IPA) of upregulated genes in dysfunctional compared to preinfusion samples. Significant selected pathways are ranked by p-value of enrichment. Orange bars denote NK cell-related pathways, and red bars indicate pathways of T cell exhaustion. H ) Heatmap showing the expression level of 17 TIL marker genes in preinfusion, effective and dysfunctional samples.

Article Snippet: S.G. is an inventor on patents related to CAR-T cell therapy, filed by the University of Pennsylvania and licensed to Novartis and Tmunity, and has received commercial research funding from Gilead.

Techniques: Control, Isolation, Fluorescence, Ex Vivo, Co-Culture Assay, Blocking Assay, Two Tailed Test, RNA Sequencing, Expressing, Marker

Journal: bioRxiv

Article Title: In vivo CRISPR-based screen identifies ZC3H12C as a mediator of CAR-T cell dysfunction in solid tumors

doi: 10.64898/2026.04.30.721530

Figure Lengend Snippet: A ) Overlap of upregulated genes across comparisons between preinfusion, effective and dysfunctional CAR-TILs. A custom CRISPR/Cas9 library was generated by selecting 100 genes from early, late, and progressive dysfunction timepoints (indicated by circles). B ) Gene knockout enrichment following the in vivo screen was analyzed from Amplicon-seq data using the MAGeCK algorithm (n=3 healthy donors). The correlation between MAGeCK score and fold change is shown. Significantly enriched genes (MAGeCK score>1 and p-value<0.05) are indicated in black. Candidate genes used for further validation are highlighted. Safe harbour AAVS1 and ROSA26, and positive controls PD-1 and LAG-3 are also depicted. For simplicity of the plot, only genes with a MAGeCK score>1 are shown. C-F ) NSG bearing pre-established SKOV3 tumors were treated with a single dose of 2×10 6 untransduced T cells (UTD), anti-HER2-28z CAR+ T cells knocked out for candidate genes ZC3H12C or TG, or AAVS1. C ) Histogram of CAR expression in CD8+ T cells used for in vivo assessment of antitumor effect. D ) CRISPR/Cas9 efficiency represented as KO score at the end of primary T cell expansion. E ) Tumor volume of mice treated with KO CAR-T cells was analyzed at indicated timepoints. F ) Change in tumor volume on day 30 after CAR-T therapy versus baseline is plotted for individual tumors. G-J ) In a second experiment, the anti-tumor efficacy of ITGB8- or RGS2-KO CAR-T was evaluated as in (C-F). Histogram of CAR expression in CD8+ CAR-T cells used for this experiment ( G ), KO score ( H ), tumor volume of treated mice ( I ) and change in tumor volume at day 26 after CAR-T therapy ( J ). ( E and I ) Data are plotted as mean ±SEM of tumor volume (n=10-12 tumors per group). *p<0.05 by two-way ANOVA with Tukey’s multiple comparisons test at day 26 after CAR-T therapy. ( F and J ) *p<0.05, **p<0.01 by one-way ANOVA at indicated timepoint. K-L ) Analysis of total T cell concentration in the blood of animals treated with ZC3H12C-KO, TG-KO, ITGB8-KO, control AAVS1-KO CAR-T or UTD T cells (two independent experiments) at day 14 ( K ) and 21 ( L ) after T cell injection. Data are plotted as mean ±SEM. Each dot represents a mouse. Fold change relative to AAVS1-KO group is indicated. **p<0.01, ***p<0.001 by one-way ANOVA with Tukey’s multiple comparisons test.

Article Snippet: S.G. is an inventor on patents related to CAR-T cell therapy, filed by the University of Pennsylvania and licensed to Novartis and Tmunity, and has received commercial research funding from Gilead.

Techniques: CRISPR, Generated, Gene Knockout, In Vivo, Amplification, Biomarker Discovery, Expressing, Concentration Assay, Control, Injection

Journal: bioRxiv

Article Title: In vivo CRISPR-based screen identifies ZC3H12C as a mediator of CAR-T cell dysfunction in solid tumors

doi: 10.64898/2026.04.30.721530

Figure Lengend Snippet: A-B ) Uniform manifold approximation and projection (UMAP) plot showing preinfusion CAR-T cells, and effective and dysfunctional CAR-TILs isolated from SKOV3 tumors. Data from one healthy donor independently analyzed. Similar results were observed in three healthy donors. Data are shown with merged timepoints (A) and separated timepoints (B). C ) Violin plots showing expression of ZC3H12C within clusters identified by single-cell RNA-seq analysis. D ) Genome browser tracks showing ATAC-seq signal at ZC3H12C, TOX and PD-1 regulatory regions within each of the clusters represented in A. E ) Motif analysis was performed to identify regulatory elements around the transcription start site (TSS) of differentially expressed genes in ZC3H12C-positive cells. Selected motifs for transcription factors associated with exhaustion from Top24 enriched motifs are shown. F ) T cell subtype classification integrated with the signature dataset reported by Zheng and colleagues . G ) UMAP plots of preinfusion CAR-T cells and effective and dysfunctional CAR-TILs color-coded by the density of expression of ZC3H12C, TG or ITGB8. H ) Dot plot showing expression of ZC3H12C and selected exhaustion-associated genes in meta-clusters identified in CD8+ T cells from Zheng, et al. 2021 . Exhaustion clusters are highlighted in bold. Color indicates effect size (ES) and size indicates the level of statistical significance.

Article Snippet: S.G. is an inventor on patents related to CAR-T cell therapy, filed by the University of Pennsylvania and licensed to Novartis and Tmunity, and has received commercial research funding from Gilead.

Techniques: Isolation, Expressing, Single Cell, RNA Sequencing

Journal: bioRxiv

Article Title: In vivo CRISPR-based screen identifies ZC3H12C as a mediator of CAR-T cell dysfunction in solid tumors

doi: 10.64898/2026.04.30.721530

Figure Lengend Snippet: A) Schematic diagram illustrating the in vitro assay used to induce dysfunction through repetitive stimulation of CAR-T cells with antigen. B ) Population doublings of AAVS1-KO (left panel) and ZC3H12C-KO (right panel) CAR-T cells during in vitro restimulation with target cells. Each color represents a different healthy donor. C ) Cytotoxicity of AAVS1-KO and ZC3H12-KO CAR-T cells against fresh HCC1954 tumor cells (E:T=1:1) was measured at indicated timepoints during the restimulation assay by using a real-time assay. Mean ±SEM of normalized cell index of one representative donor for each timepoint is shown. Time when T cells were added is indicated. D ) TNF-α, IL-2, and IFNγ production by AAVS1-KO and ZC3H12C-KO CAR-T cells after 24h co-culture with fresh HCC1954 tumor cells (E:T=1:3) was measured at indicated timepoints of in vitro restimulation with tumor cells. Data are plotted as mean ±SEM. Each dot represents a healthy donor (n=5 for day 0 and day 7; n=4 for day 14). Fold increase of cytokine secretion by ZC3H12C-KO versus AAVS1-KO CAR-T cells is indicated. *p<0.05, **p<0.01 by two-tailed paired T-test. E ) Representative flow cytometry plots of intracellular TNF-α and IFNγ staining in AAVS1-KO and ZC3H12C-KO CAR-T cells (gated on live/CD45+) after 24h co-culture with fresh SKOV3 cells (E:T=1:3) on day 7 after repetitive antigen stimulation in vitro (top). Frequencies of single and double-positive T cells for IFN-γ and TNF-α are represented as absolute numbers (bottom). Data are plotted as mean ±SD. Each dot represents a healthy donor (n=5). *p<0.05, **p<0.01 by two-tailed paired T-test. Fold change of ZC3H12C-KO relative to AAVS1-KO is indicated (bottom).

Article Snippet: S.G. is an inventor on patents related to CAR-T cell therapy, filed by the University of Pennsylvania and licensed to Novartis and Tmunity, and has received commercial research funding from Gilead.

Techniques: In Vitro, Co-Culture Assay, Two Tailed Test, Flow Cytometry, Staining

Journal: bioRxiv

Article Title: In vivo CRISPR-based screen identifies ZC3H12C as a mediator of CAR-T cell dysfunction in solid tumors

doi: 10.64898/2026.04.30.721530

Figure Lengend Snippet: A ) Schematic representing the experimental approach used to analyze the ex vivo function of ZC3H12C-KO CAR-TILs. B ) Analysis of CD8+ T cell infiltration in tumors treated with AAVS1-KO or ZC3H12C-KO CAR-T cells on day 17 after infusion. Representative flow cytometry plots showing CD45 and HER2 staining after excluding dead cells (left panel). Violin plots showing the frequency of live/CD45+ cells in tumors treated with AAVS1-KO or ZC3H12C-KO CAR-T cells (right panel). Each dot represents a tumor (n=2 healthy donors, 7-8 tumors per group). Fold change relative to AAVS1 is indicated. C ) Frequency of PD-1+ LAG-3+ TIM3+ CD8+ CAR-TILs (gated as live, CD45+). *p<0.05 by two-tailed unpaired t-test. D ) Ex vivo cytotoxicity of AAVS1-KO and ZC3H12C-KO CAR-TILs was analyzed against SKOV3 tumor cells (E:T=5:1) using a real-time assay. Time when T cells were added is indicated. Data are plotted as mean ±SEM of normalized index of two duplicates (n=1). E ) Cytokine production by AAVS1-KO and ZC3H12C-KO CAR-TILs after a 24h ex vivo co-culture with SKOV3 tumor cells (E:T=3:1). Data are plotted as mean ±SEM. Each dot represents a healthy donor (n=4). *p<0.05, **p<0.01 by two-tailed paired t-test. Fold change of ZC3H12C-KO relative to AAVS1-KO is indicated.

Article Snippet: S.G. is an inventor on patents related to CAR-T cell therapy, filed by the University of Pennsylvania and licensed to Novartis and Tmunity, and has received commercial research funding from Gilead.

Techniques: Ex Vivo, Flow Cytometry, Staining, Two Tailed Test, Co-Culture Assay

Journal: bioRxiv

Article Title: In vivo CRISPR-based screen identifies ZC3H12C as a mediator of CAR-T cell dysfunction in solid tumors

doi: 10.64898/2026.04.30.721530

Figure Lengend Snippet: A-C ) NSG mice bearing subcutaneous L55-CD19 tumors were treated with a single dose of 4×10 6 control unstransduced (UTD) T cells, AAVS1-KO or ZC3H12C-KO CAR+ T cells. A ) Tumors were measured at indicated timepoints. Data are plotted as mean ±SEM of tumor volume (n=8-10 tumors per group). *p<0.05, by two-way ANOVA with Tukey’s multiple comparisons test at day 27. B ) Change in tumor volume on day 31 after CAR-T cell therapy. **p<0.01 by two-tailed unpaired t-test. C ) The concentration of total T cells was determined in the blood of animals treated with ZC3H12C-KO, AAVS1-KO ARI-0001 CAR-T cells, or UTD T cells, at day 14 (left panel) and day 21 (right panel) after treatment. Data are plotted as mean ±SD. Each dot represents a mouse (n=1 healthy donor, 8-10 tumors per group). ****p<0.0001 by one-way ANOVA with Tukey’s multiple comparisons test. Fold change relative to ZC3H12C-KO group is indicated. D ) Schematic diagram illustrating the in vitro assay using hematological cancer cells. E ) T cell proliferation after an in vitro 7-day co-culture of CAR-T cells with NALM-6 or Ramos cell lines (E:T=1:2) analyzed by flow cytometry. For each cell line, T cell is represented as absolute numbers (left panel) or fold change compared to AAVS1-KO CAR-T cell group (right panel) (n=1 healthy donor). F-G ) NSG mice were intravenously injected with Nalm6-GFP-luciferase cells. After 7 days, mice were randomized and treated with 2×10 6 CAR+ T cells, including UTD control T cells, AAVS1-KO or ZC3H12C-KO ARI-0001 (n=10 mice per group, 2 healthy donors). F ) The Kaplan-Meier survival curves of treated mice are plotted and median survival days calculated. *p<0.05 by Log-rank (Mantel-Cox) test. G ) T cell persistence in the blood of treated animals was determined on days 7 and 14 after CAR-T cell therapy. Each dot represents a mouse. *p<0.05 by one-way ANOVA with Tukey’s multiple comparisons test. Fold change relative to ZC3H12C-KO group is indicated.

Article Snippet: S.G. is an inventor on patents related to CAR-T cell therapy, filed by the University of Pennsylvania and licensed to Novartis and Tmunity, and has received commercial research funding from Gilead.

Techniques: Control, Two Tailed Test, Concentration Assay, In Vitro, Co-Culture Assay, Flow Cytometry, Injection, Luciferase