Journal: bioRxiv

Article Title: Differential TIM-3 glycosylation enables specific dual targeting CAR-T therapy in acute myeloid leukemia

doi: 10.64898/2026.04.22.719217

Figure Lengend Snippet: (A) Frequency (left) and mean fluorescence intensity (MFI; right) of TIM-3 expression on bulk AML blasts compared with residual healthy bone marrow (BM) cells from the same patient (n=25). (B) TIM-3 expression in the LSC-enriched CD34 + CD38 - population (n=17). (C) Schematic of the third-generation TIM-3.CAR containing CD28-OX40 costimulatory domains cloned into the pT4-transposon vector. (D) Transduction efficiency of TIM-3.CAR-CIK cells compared to non-transduced (NT) CIK cells (mean 71.24±16.99, n = 13). See also Figure S1A . (E) Fold increase of TIM-3.CAR-CIK cells compared to CD33.CAR-CIK and NT cells at day 21 of culture (n=9). (F) Viability of TIM-3.CAR-CIK cells, CD33.CAR-CIK and NT cells at days 4, 6 and 8 post-transduction, assessed by flow cytometry. See also Figure S1B . (G) Frequency of CD3 + TIM-3 + cells within NT, TIM3.CAR- and CD33.CAR-CIK populations over the first 8 days of CIK differentiation (n=7 donors for TIM-3.CAR-CIK and NT cells, n=4 donors for CD33.CAR-CIK cells). See also Figure S1C . Data are shown as individual values with mean ± standard deviation (SD). Significance was assessed using paired t test (A, D) or repeated-measures two-way ANOVA with Bonferroni’s post hoc test (E-G). ns, not significant, *** p = 0.0001 and **** p < 0.0001. See also Figure S1 for longitudinal TIM-3 and TIM-3.CAR expression and CD4 + and CD8 + subset analysis.

Article Snippet: Membranes were probed with anti-human TIM-3 antibody (TIM-3-cmAb) (1:250; R&D Systems, MAB23652), a recombinant monoclonal antibody derived from the TIM-3.CAR scFv (TIM-3 scFv-mAb) (1:500; GENEWIZ), biotinylated Aleuria aurantia lectin (AAL; 1:3000; Vector Laboratories, B-1395-1), and biotinylated Ricinus communis agglutinin I (RCA I; 1:3000; Vector Laboratories, B-1085-1).

Techniques: Fluorescence, Expressing, Clone Assay, Plasmid Preparation, Transduction, Flow Cytometry, Standard Deviation

Journal: bioRxiv

Article Title: Differential TIM-3 glycosylation enables specific dual targeting CAR-T therapy in acute myeloid leukemia

doi: 10.64898/2026.04.22.719217

Figure Lengend Snippet: (A) Long-term killing assay of TIM-3.CAR-CIK cells against four primary AML samples compared with NT cells. Blasts survival was assessed by flow cytometry (E:T 1:10 and 1:50, n = 8 donors). See also Figure S2D . (B) Survival of TIM-3 + primary AML blasts (n=3) after 7-day co-culture with TIM-3.CAR-CIK or NT cells (E:T 1:10 and 1:50, n = 4 donors). See also Figure S2E . (C) Recovery of the LSC-enriched CD34 + CD38 - population (n = 9) and ( D ) of GPR56 + blasts (n = 6) after long-term co-culture. (E) Proliferation of TIM-3.CAR-CIK cells assessed by Ki67 staining after 72 hours co-culture with AML blasts (E:T 1:1, n = 7). See also Figure S2G . (F) Cytokine production (IFN-γ, IL-2) after 5 hours co-culture of TIM-3.CAR-CIK or NT cells with primary AML blasts (E:T 1:3, n = 9). See also Figure S2H . Data are presented as individual values and mean ± SD. Statistics were calculated with repeated-measures two-way ANOVA with Bonferroni’s post hoc test. ns, not significant; *p = 0.01, **p < 0.001, ***p = 0.0001 and ****p < 0.0001. See also Figure S2 for TIM-3.CAR validation in KASUMI-3, AML cell line.

Article Snippet: Membranes were probed with anti-human TIM-3 antibody (TIM-3-cmAb) (1:250; R&D Systems, MAB23652), a recombinant monoclonal antibody derived from the TIM-3.CAR scFv (TIM-3 scFv-mAb) (1:500; GENEWIZ), biotinylated Aleuria aurantia lectin (AAL; 1:3000; Vector Laboratories, B-1395-1), and biotinylated Ricinus communis agglutinin I (RCA I; 1:3000; Vector Laboratories, B-1085-1).

Techniques: Flow Cytometry, Co-Culture Assay, Staining, Biomarker Discovery

Journal: bioRxiv

Article Title: Differential TIM-3 glycosylation enables specific dual targeting CAR-T therapy in acute myeloid leukemia

doi: 10.64898/2026.04.22.719217

Figure Lengend Snippet: (A) TIM-3 expression on KASUMI-3 cells, primary AML blasts, and healthy immune subsets (CIK cells, monocytes, NK cells) assessed by flow cytometry using QuantiBRITE beads. REH (ALL cell line) served as negative control. (B) Short-term killing assay of TIM-3.CAR-CIK cells against CIK (n = 11) or KASUMI-3 (n = 8) cells compared with NT cells. Target cell lysis was evaluated by flow cytometry (E:T 5:1). (C) Short-term killing assay of TIM-3.CAR-CIK cells against monocytes (n = 8) or NK cells (n = 8) compared with NT (E:T 5:1). KASUMI-3 (n = 4) were included as positive control. (D) Immunoblot analysis of TIM-3 in lysates from monocytes, CIK cells, and KASUMI-3 cells following enzymatic treatment with PNGase F or broad neuraminidase, probed with a commercial anti–TIM-3 antibody (TIM-3-cmAb). GAPDH, loading control. Glycan symbols follow SNFG. (E) TIM-3 immunoprecipitates from monocytes, CIK cells, and KASUMI-3 cells treated with PNGase F or O- glycosidase and analyzed by immunoblot with TIM-3-cmAb and lectin far-western with Aleuria aurantia lectin (AAL; fucosylated epitopes). TGX stain-free total protein signal is shown as a loading/normalization control. (F) KASUMI-3 cells treated with vehicle (mock) or the fucosylation inhibitor 2F-peracetyl-fucose (SGN-2FF), followed by PNGase F or neuraminidase treatment and immunoblot/lectin probing with TIM-3-cmAb and AAL. See also Figure S3A . (G) Short-term killing assay of TIM-3.CAR-CIK cells against untreated or SGN-2FF-treated KASUMI-3 cells at various E:T ratios (5:1, 1:1, 0.5:1, 0.25:1 and 0.125:1, n = 8). (H) Affinity kinetics (left) and binding avidity at 1000 pN force (right) of TIM-3.CAR-CIK cells to untreated or defucosylated KASUMI-3 by LUMICKS analysis (n = 6). Immunoblot experiments (D-F) were repeated in three independent biological replicates with similar results. Data are presented as individual values and mean ± SD. Statistical significance was determined with repeated-measures two-way ANOVA with Bonferroni’s post hoc test (B, C) or using paired t test (G, H). ns, not significant; *p = 0.01, **p < 0.001, ***p = 0.0001 and ****p < 0.0001. Illustrations were created with Biorender.com. See also Figure S3 for loading-matched TIM-3 immunoprecipitation controls.

Article Snippet: Membranes were probed with anti-human TIM-3 antibody (TIM-3-cmAb) (1:250; R&D Systems, MAB23652), a recombinant monoclonal antibody derived from the TIM-3.CAR scFv (TIM-3 scFv-mAb) (1:500; GENEWIZ), biotinylated Aleuria aurantia lectin (AAL; 1:3000; Vector Laboratories, B-1395-1), and biotinylated Ricinus communis agglutinin I (RCA I; 1:3000; Vector Laboratories, B-1085-1).

Techniques: Expressing, Flow Cytometry, Negative Control, Lysis, Positive Control, Western Blot, Control, Glycoproteomics, Staining, Binding Assay, Immunoprecipitation

Journal: bioRxiv

Article Title: Differential TIM-3 glycosylation enables specific dual targeting CAR-T therapy in acute myeloid leukemia

doi: 10.64898/2026.04.22.719217

Figure Lengend Snippet: (A) Immunoblot profiling of TIM-3 glycoforms in monocytes, CIK cells, and KASUMI-3 lysates using a recombinant scFv-derived monoclonal antibody (TIM-3scFv-mAb) following enzymatic treatment with PNGase F or broad neuraminidase. GAPDH, loading control. (B) TIM-3 immunoprecipitates from healthy monocytes, KASUMI-3 cells, and primary AML blasts treated with neuraminidase and/or PNGase F and analyzed by lectin and antibody probing: Ricinus communis agglutinin I (RCA-I; terminal β-galactose/LacNAc motifs), CA19-9 (sialyl-Lewis A), CSLEX1 (sialyl-Lewis X), and TIM-3scFv-mAb. See also Figure S3B . (C) High-resolution immunoblot of TIM-3 species detected by TIM-3scFv-mAb in CIK cells, primary AML blasts, and KASUMI-3 cells. GAPDH, loading control. See also Figure S3C . (D) RT-qPCR expression profiling of glycosyltransferases (FUT7, FUT8, ST3GAL3, ST3GAL4, ST3GAL6) in monocytes, KASUMI-3 cells, and primary AML blasts. Data are plotted as fold-change relative to monocytes and normalized to 18S RNA; individual points denote biological samples where applicable. (E) Schematic model summarizing a glycoform-biased recognition framework in which AML-associated remodeling of TIM-3 N -glycans contributes to preferential TIM-3.CAR recognition of AML-enriched TIM-3 glycoforms. Representative N -glycan structures are proposed for TIM-3 in AML blasts, monocytes and CIK cells based on enzymatic perturbation and lectin/antibody probing. Sugar moieties drawn with dashed outlines indicate features not directly resolved/assigned. Glycan symbols follow SNFG. Immunoblot and lectin/antibody blot experiments (A-C) were repeated in three independent biological replicates with similar results. Illustrations were created with Biorender.com. See also Figure S3 for additional lectin/antibody probing of TIM-3 glycoforms and terminal galactose exposure.

Article Snippet: Membranes were probed with anti-human TIM-3 antibody (TIM-3-cmAb) (1:250; R&D Systems, MAB23652), a recombinant monoclonal antibody derived from the TIM-3.CAR scFv (TIM-3 scFv-mAb) (1:500; GENEWIZ), biotinylated Aleuria aurantia lectin (AAL; 1:3000; Vector Laboratories, B-1395-1), and biotinylated Ricinus communis agglutinin I (RCA I; 1:3000; Vector Laboratories, B-1085-1).

Techniques: Western Blot, Recombinant, Derivative Assay, Control, Quantitative RT-PCR, Expressing, Glycoproteomics

Journal: bioRxiv

Article Title: Differential TIM-3 glycosylation enables specific dual targeting CAR-T therapy in acute myeloid leukemia

doi: 10.64898/2026.04.22.719217

Figure Lengend Snippet: (A) Schematic of the xenograft KASUMI-3 AML model. (B) Representative flow cytometry plots of hCD45 + CD33 + (up) and of hCD45 + TIM-3 + cells (down) in the BM of CTR or TIM-3.CAR treated mice at sacrifice. ( C-E ) Frequencies of hCD33 + and hTIM-3 + cells in the (C) BM, (D), spleen and (E) peripheral blood (PB) at sacrifice. Illustrations were created with Biorender.com. Results represent three independent experiments using TIM-3.CAR-CIK cells generated from 3 different donors. Data are presented as individual values and mean ± SD. Statistical significance was determined by unpaired t test. *p = 0.01, **p < 0.001 and ****p < 0.0001.

Article Snippet: Membranes were probed with anti-human TIM-3 antibody (TIM-3-cmAb) (1:250; R&D Systems, MAB23652), a recombinant monoclonal antibody derived from the TIM-3.CAR scFv (TIM-3 scFv-mAb) (1:500; GENEWIZ), biotinylated Aleuria aurantia lectin (AAL; 1:3000; Vector Laboratories, B-1395-1), and biotinylated Ricinus communis agglutinin I (RCA I; 1:3000; Vector Laboratories, B-1085-1).

Techniques: Flow Cytometry, Generated

Journal: bioRxiv

Article Title: Differential TIM-3 glycosylation enables specific dual targeting CAR-T therapy in acute myeloid leukemia

doi: 10.64898/2026.04.22.719217

Figure Lengend Snippet: (A) Schematic of IF-BETTER gate strategy showing dual antigen recognition of CD33 + /TIM-3 + target cell by CD33.CAR/TIM-3.CCR and TIM3.CAR/CD33.CCR-CIK cells. (B) Co-distribution of CD33 and TIM-3 expression (MFI) on bulk AML (top) and LSC-enriched CD34 + CD38 - population (bottom). Each dot represents a distinct patient (n=44 patients). (C) Schematics of next-generation Dual CD33.CAR/TIM-3.CCR and TIM-3.CAR/CD33.CCR constructs. CAR molecules are second-generation, carrying CD28 co-stimulatory domain, while CCR molecules present 4-1BB as co-stimulus. Both constructs were cloned into a pT4-transposon vector. See also Figure S4A, B . (D) Long-term killing assay (E:T 1:10) of all CAR-CIK cells against primary AML blasts (n=8 blasts) compared to NT cells. Blasts survival was determined by flow cytometry (n=13 donors). (E) Recovery of LSC-enriched CD34 + CD38 - population (n=8 patient samples) after 7 days co-culture with all CAR-CIK cells (E:T 1:10), compared to NT cells (n=13). Data are presented as individual values and the mean ± SD. Statistical significance was determined by one-way ANOVA test. ** p<0.01, **** p<0.0001. Illustrations were created with Biorender.com. See also Figure S4 for expression and phenotypic characterization of Dual CAR constructs, Figure S5 for Dual CAR-CIK cell activity against AML cell lines and Figure S6 for Dual CAR-CIK cell off-tumor toxicity against healthy immune and hematopoietic cells.

Article Snippet: Membranes were probed with anti-human TIM-3 antibody (TIM-3-cmAb) (1:250; R&D Systems, MAB23652), a recombinant monoclonal antibody derived from the TIM-3.CAR scFv (TIM-3 scFv-mAb) (1:500; GENEWIZ), biotinylated Aleuria aurantia lectin (AAL; 1:3000; Vector Laboratories, B-1395-1), and biotinylated Ricinus communis agglutinin I (RCA I; 1:3000; Vector Laboratories, B-1085-1).

Techniques: Expressing, Construct, Clone Assay, Plasmid Preparation, Flow Cytometry, Co-Culture Assay, Activity Assay

Journal: bioRxiv

Article Title: Differential TIM-3 glycosylation enables specific dual targeting CAR-T therapy in acute myeloid leukemia

doi: 10.64898/2026.04.22.719217

Figure Lengend Snippet: ( A, F ) Schematic of KG-1 TIM-3 + low (A) and high (F) burden in vivo experimental design. ( B, G ) Representative FACS plots of hCD45 + CD33 + cells in the BM of CTR and CAR-treated mice in low-(B) and high-burden (G) settings. ( C, H ) Summary of BM hCD33 + cell percentages for all groups. ( D, I ) Analysis of TIM-3 expression on residual hCD33 + cells in BM of CTR and treated mice. ( E, J ) Leukemic burden in spleen (left) and PB (right) evaluated by flow cytometry, in low- (E) and (J) high-burden models. Illustrations were created with Biorender.com. Results represent two independent experiments using CAR-CIK cells generated from 2 donors. Data are presented as individual values and the mean ± SD. Statistical significance was determined by one-way ANOVA. **p < 0.001, ***p = 0.0001 and ****p < 0.0001.

Article Snippet: Membranes were probed with anti-human TIM-3 antibody (TIM-3-cmAb) (1:250; R&D Systems, MAB23652), a recombinant monoclonal antibody derived from the TIM-3.CAR scFv (TIM-3 scFv-mAb) (1:500; GENEWIZ), biotinylated Aleuria aurantia lectin (AAL; 1:3000; Vector Laboratories, B-1395-1), and biotinylated Ricinus communis agglutinin I (RCA I; 1:3000; Vector Laboratories, B-1085-1).

Techniques: In Vivo, Expressing, Flow Cytometry, Generated

Journal: bioRxiv

Article Title: Single-Cell Profiling Reveals Developmental Trajectories and identifies SYK and TIM3 as Targets in some T Cell Lymphomas

doi: 10.64898/2026.03.27.714741

Figure Lengend Snippet: (A) A UMAP showing the integration of tumour and benign cells of the TME, and healthy lymph node data from two public studies (PMID35027729, PMID39566559). Cells are coloured according to the data source, and scANVI-predicted cell types are shown in (B), (C) and (D). Images of an ALK+ ALCL case colored by Cell2location-predicted cell type abundances in each Visium spot. (C) Predictions for the locations of malignant cells, TIM3+ T cells (T_TIM3+) and cytotoxic T cells (T_CD8+_cytotoxic), (D) Predictions for the locations of dark zone B cells (B_GC_DZ), light zone B cells (B_GC_LZ) and follicular dendritic cells (FDC). (E) A dotplot showing non-negative matrix factorisation (NMF) factors, where each factor is a group of co-localized cell types. In the dotplot, the size and color density represent the loading of each cell type in each factor. (F) Staining of TIM3 and CD30 expression on two representative ALK+ ALCL cases (G) Summary of the % TIM3+ infiltrating T lymphocytes in each of the 9 ALK+ ALCL cases analysed.

Article Snippet: Following heat-mediated antigen retrieval, consecutive sections were stained for TIM3 (MAB23652, R&D Systems) and CD30 (M0751, Agilent Dako) both at a 1:50 dilution.

Techniques: Staining, Expressing