human luminex discovery assays  (R&D Systems)

Journal: Frontiers in Immunology

Article Title: Exploratory transcriptomic analysis of mouse articular cartilage in response to tissue inhibitor of metalloproteinase 3 identifies inflammation-associated gene expression changes

doi: 10.3389/fimmu.2026.1794078

Figure Lengend Snippet: Experimental design. Femoral head articular cartilage samples from 16 mice were harvested in pairs and each pair was placed in one well of a 96-well plate with serum-free Opti-MEM supplemented with 1% penicillin-streptomycin, maintained under standard normoxic conditions (21% O 2 , 5% CO 2 , 37°C). After three days, half of the cultures were transferred to a hypoxia incubator (3% O 2 ). Following 24 hours of normoxic or hypoxic incubation, cultures were treated with recombinant human TIMP-3 (100 nM; 2.6 µg/ml) or vehicle for 20 hours, and RNA was extracted for RNA sequencing (n = 4 per group). Group labels: NC, normoxia control; NT, normoxia TIMP-3; HC, hypoxia control; HT, hypoxia TIMP-3.

Article Snippet: After 24 hours, cultures were treated with recombinant human TIMP-3 (R&D Systems, Abingdon, Oxon, UK) at 100 nM, (2.6 μg/ml) or vehicle control for 20 hours under normoxia or hypoxia.

Techniques: Incubation, Recombinant, RNA Sequencing, Control

Journal: Frontiers in Immunology

Article Title: Exploratory transcriptomic analysis of mouse articular cartilage in response to tissue inhibitor of metalloproteinase 3 identifies inflammation-associated gene expression changes

doi: 10.3389/fimmu.2026.1794078

Figure Lengend Snippet: TIMP-3 upregulates Saa3 gene expression in cartilage under normoxia and hypoxia. Articular cartilage explants were cultured and processed for RNA-seq as described in the legend to Table 2 . (A) Venn diagram showing overlap of genes differentially regulated in response to TIMP-3 under normoxia (Norm; 21% O 2 ) and hypoxia (Hyp; 3% O 2 ). (B) RT-qPCR validation of Saa3 . RNA-seq (left) and RT-qPCR (right) data shown as log 2 FC versus the mean of the controls (mean ± SD, n = 4). RNA-seq: log 2 FC were calculated from normalized CPM values; § FDR < 0.05 (hypoxia effect), # P < 0.01 (TIMP-3 effect) by quasi-likelihood F-test applied to raw counts (edgeR). RT-qPCR: *P < 0.05 by unpaired two-sided Welch’s t-test on log 2 FC values (-ΔΔCq). Ctrl, control; FC, fold change.

Article Snippet: After 24 hours, cultures were treated with recombinant human TIMP-3 (R&D Systems, Abingdon, Oxon, UK) at 100 nM, (2.6 μg/ml) or vehicle control for 20 hours under normoxia or hypoxia.

Techniques: Gene Expression, Cell Culture, RNA Sequencing, Quantitative RT-PCR, Biomarker Discovery, Control

Journal: Frontiers in Immunology

Article Title: Exploratory transcriptomic analysis of mouse articular cartilage in response to tissue inhibitor of metalloproteinase 3 identifies inflammation-associated gene expression changes

doi: 10.3389/fimmu.2026.1794078

Figure Lengend Snippet: RNA-seq analysis of TIMP-3–treated cartilage under normoxia. (A) MA plot showing differential expression between TIMP-3–treated and control samples, with log 2 FC plotted against average log 2 CPM expression (n = 4). Grey: no change, red: upregulated, blue: downregulated transcripts (P < 0.01 by quasi-likelihood F-test in edgeR, |log 2 FC| > 0.58). The 10 most highly expressed regulated genes are labelled. Complete gene lists are in . (B) Heatmap of differentially expressed genes (log 2 FC versus the mean of the controls). Genes are ordered by hierarchical clustering. Sample labels: NC1-4: normoxia controls; NT1-4: normoxia TIMP-3–treated. (C) Sole significantly enriched pathway (FDR < 0.05) from DAVID analysis of 26 upregulated genes. As only one KEGG pathway passed the significance threshold, it is shown individually together with its enrichment score (Fold enrichment) and FDR for visual consistency within the multipanel figure. (D) RT-qPCR validation. RNA-seq (left) and RT-qPCR (right) data shown as log 2 FC versus the mean of the controls (mean ± SD, n = 4). RNA-seq: log 2 FC were calculated from normalized CPM values; § FDR < 0.05 (hypoxia effect), # P < 0.01 (TIMP-3 effect) by quasi-likelihood F-test applied to raw counts (edgeR). RT-qPCR: *P < 0.05, **P < 0.01, ***P < 0.001 by unpaired two-sided Welch’s t-test on log 2 FC values (-ΔΔCq). CPM, counts per million; Ctrl, control; FC, fold change; Norm, normoxia; Hyp, hypoxia.

Article Snippet: After 24 hours, cultures were treated with recombinant human TIMP-3 (R&D Systems, Abingdon, Oxon, UK) at 100 nM, (2.6 μg/ml) or vehicle control for 20 hours under normoxia or hypoxia.

Techniques: RNA Sequencing, Quantitative Proteomics, Control, Expressing, Quantitative RT-PCR, Biomarker Discovery

Journal: Frontiers in Immunology

Article Title: Exploratory transcriptomic analysis of mouse articular cartilage in response to tissue inhibitor of metalloproteinase 3 identifies inflammation-associated gene expression changes

doi: 10.3389/fimmu.2026.1794078

Figure Lengend Snippet: RNA-seq analysis of TIMP-3–treated cartilage under hypoxia. (A) MA plot showing differential expression between TIMP-3–treated and control articular cartilage, with log 2 FC plotted against average log 2 CPM expression (n = 4). Grey: no change, red: upregulated, blue: downregulated (P < 0.01, |log 2 FC| > 0.58). The 10 most highly expressed regulated genes are labelled. G1 and G2 correspond to E430024I08Rik and AC127578.1 respectively. Complete gene lists are in . (B) Heatmap of differentially expressed genes (log 2 FC versus the mean of the controls). Genes are ordered by hierarchical clustering. Sample labels: HC1-4: hypoxia controls; HT1-4: hypoxia TIMP-3-treated. (C) Protein–protein interaction (PPI) network corresponding to genes downregulated by TIMP-3 under hypoxia (P <0.01 by quasi-likelihood F-test in edgeR, |log 2 FC| > 0.58), generated using STRING (default interaction score ≥ 0.400). All nodes represent the initially filtered gene list and are included to show the network context and highlight that only Pbk and Racgap1 display a documented interaction. Line colors indicate evidence type: green, text mining; pink, experimental; black, co-expression. Combined interaction score: 0.711. (D) RT-qPCR validation. RNA-seq (left) and RT-qPCR (right) data shown as log 2 FC versus the mean of the controls (mean ± SD, n = 4). RNA-seq: log 2 FC were calculated from normalized CPM values; § FDR < 0.05 (hypoxia effect), # P < 0.01 (TIMP-3 effect) by quasi-likelihood F-test applied to raw counts (edgeR); RT-qPCR: *P < 0.05, **P < 0.01 by unpaired two-sided Welch’s t-test on log 2 FC values (-ΔΔCq). CPM, counts per million; Ctrl, control.

Article Snippet: After 24 hours, cultures were treated with recombinant human TIMP-3 (R&D Systems, Abingdon, Oxon, UK) at 100 nM, (2.6 μg/ml) or vehicle control for 20 hours under normoxia or hypoxia.

Techniques: RNA Sequencing, Quantitative Proteomics, Control, Expressing, Generated, Quantitative RT-PCR, Biomarker Discovery

Journal: Cell Reports Medicine

Article Title: Multimodal profiling of pancreatic cancer reveals a TIMP-1-dominated secretory profile determining pro-tumor immunoinstruction in human cancers

doi: 10.1016/j.xcrm.2025.102546

Figure Lengend Snippet: snRNA-seq identifies epithelial-subtype-specific CISS upregulation toward TIMP-1 hi basal-like PDAC (A) UMAP embedding of PDAC patient tumor ( n = 17) snRNA-seq and post-hoc cell-type annotation. CAF, cancer-associated fibroblasts; VSMC, vascular smooth muscle cells. (B) Selected marker genes ( Z scores) and proportion of positive cells in indicated cell types. (C) Non-epithelial cell UMAP embeddings and post-hoc cell-type annotation (see B–S3F). (D) Epithelial cell UMAP embeddings and post-hoc cell-type annotation. (E) Selected marker genes ( Z scores) and proportion of positive cells in epithelial subsets. (F and G) RNA velocity analysis to infer transcriptional dynamics in acinar/ductal cells (F) and cell state progression across neoplastic/malignant cells (G). Statistics (G) by Kruskal-Wallis and Dunn’s test compared to PanIN. (H) CISS expression in epithelial subsets and induction compared to acinar cells by Kruskal-Wallis and Dunn’s test. (I) (Left) CISS factor expression in epithelial subsets normalized to acinar (mean ±95% CI). (Right) Heatmap and statistics of CISS factors by Kruskal-Wallis and Dunn’s test. (J) Normalized TIMP1 expression on the UMAP embedding. (K) (Left) TIMP1 prevalence within the CISS by comparing norm. CISS expression with TIMP1 (red) or without TIMP1 (black) (mean ±95% CI). Statistics by Mann-Whitney tests. (Right) Volcano plot of TIMP1 prevalence across epithelial clusters identified TIMP1 lo classical, TIMP1 int basal-like, and TIMP1 hi basal-like cancer cells. Groups identified by log2 fold-changes and significance of TIMP1 prevalences within CISS pattern (left). (L) TIMP1 expression and proportion of individual cell subtypes within epithelial compartment across patient tumors. For all statistics: n.s., non-significant; ∗ p < 0.05, ∗∗ p < 0.01; ∗∗∗ p < 0.001; ∗∗∗∗ p < 0.0001.

Article Snippet: TIMP-1 anti-human mAb (rabbit) , Cell Signaling , CAT# 8946; clone D10E6; RRID: AB_10891805.

Techniques: Marker, Expressing, MANN-WHITNEY

Journal: Cell Reports Medicine

Article Title: Multimodal profiling of pancreatic cancer reveals a TIMP-1-dominated secretory profile determining pro-tumor immunoinstruction in human cancers

doi: 10.1016/j.xcrm.2025.102546

Figure Lengend Snippet: CISS-prevalent TIMP-1 is causal for PDAC-cell-induced NK cell suppression (A) PCA integrating epithelial (red) and immune (blue) cell fractions, fractions of CISS-expressing epithelial cells, and epithelial TIMP1 . (B) Pearson correlation between PCA-standardized variables (A) (∗ p < 0.05, ∗∗ p < 0.01; ∗∗∗ p < 0.001; ∗∗∗∗ p < 0.0001). (C) Comparison of correlations between epithelial and immune subtypes (A and B) by paired Student’s t test between indicated groups. (D–F) Workflow (D) to identify canonical pathways [GSEA; (E)] and immune cell profiles [CIBERSORTx; (F)] correlated with TIMP1 expression (TCGA-PAAD). For GSEA, genes were ranked by correlation coefficients (Spearman) with TIMP1 expression. Reference gene set: C2:CP. (G–I) Workflow (G) for MIA PaCa-2-cell-derived secreted factors suppressing NK cell killing of K562 targets (H) and degranulation (I) at indicated effector-to-target ratios (E:T). Comparison to control media [no cancer conditioning; shared with (O,P)] by unpaired Student’s t tests. (J) Workflow for CRISPR-Cas9-based TIMP-1 knockout in MIA PaCa-2 cells and RNA-seq. (K) PCA of RNA-seq data [ n = 3 per cell line; (J)]. (L) TIMP-1-dependent DEGs (DESeq2) in MIA PaCa-2 cells. Intersection shows DEGs between TIMP-1 WT and TIMP-1 KO (1 and 2) cells, independent of CRISPR-Cas9 (CRISPR Control). (M) GSEA of TIMP-1-dependent biological processes in MIA PaCa-2 cells. log2 fold-changes of DEGs (L) between means of TIMP-1-competent (“TIMP-1 WT”/“ CRISPR Control”) cells and TIMP-1-deficient (“TIMP-1 KO 1/2”) cells. Reference gene set: GO:BP. Enriched gene sets (FDR q < 0.05) were categorized (see G). (N–P) Workflow (N) of TIMP-1-dependent NK cell suppression of CRISPR-Cas9-derived MIA PaCa-2 cell lines (J) on K562 target cell killing (O) and degranulation (P) at indicated E:T. Statistics by one-way ANOVA and Tukey test for indicated groups. Data in (H, I, K, O, and P) showing biological replicates as box and whiskers plots (H, I, O, and P). (D, G, J, and N) Created with BioRender.com .

Article Snippet: TIMP-1 anti-human mAb (rabbit) , Cell Signaling , CAT# 8946; clone D10E6; RRID: AB_10891805.

Techniques: Expressing, Comparison, Derivative Assay, Control, CRISPR, Knock-Out, RNA Sequencing

Journal: Cell Reports Medicine

Article Title: Multimodal profiling of pancreatic cancer reveals a TIMP-1-dominated secretory profile determining pro-tumor immunoinstruction in human cancers

doi: 10.1016/j.xcrm.2025.102546

Figure Lengend Snippet: CISS-prevalent TIMP-1 is sufficient to suppress cytotoxic capacity in NK cells via CD74 signaling (A–F) Workflow (A) to identify NK cell clusters in TUM Cohort (B, C, and F) and Steele Cohort (D and E). (B and D) NK cell UMAP embeddings, post-hoc annotations, and selected marker genes ( Z scores) (C and D, bottom). ctx, cytotoxicity. KLRC, killer cell lectin like receptor C gene family. (E and F, right) GZMB and PRF1 expression on the UMAP (B and D). (F, left) RNA velocity to infer transcriptional dynamics in NK cell clusters. (G–I) NK cell granzyme B and perforin levels upon exposure to (G) rhWT-TIMP-1 (granzyme B: UNT, n = 17; 50 ng/mL, n = 6; 100 ng/mL, n = 6; 250 ng/mL, n = 10; 500 ng/mL, n = 14; perforin: UNT, n = 12; 50 ng/mL, n = 6; 100 ng/mL, n = 6; 250 ng/mL, n = 6; 500 ng/mL, n = 11). (H) 500 ng/mL WT-TIMP-1 vs. equimolar N-TIMP-1 (granzyme B: UNT, n = 32; WT-TIMP-1, n = 30; N-TIMP-1, n = 24; perforin: UNT, n = 27; WT-TIMP-1, n = 27; N-TIMP-1, n = 24); (I) 500 ng/mL WT-TIMP-1 vs. equimolar N-TIMP-1 with or without α-CD74 antibody milatuzumab or immunoglobulin G (IgG) control (granzyme B: IgG alone, n = 23; IgG + WT-TIMP-1, n = 25; IgG + N-TIMP-1, n = 20; α-CD74 alone, n = 26; α-CD74 + WT-TIMP-1, n = 24; α-CD74 + N-TIMP-1, n = 20; perforin: IgG alone, n = 19; IgG + WT-TIMP-1, n = 20; IgG + N-TIMP-1, n = 20; α-CD74 alone, n = 19; α-CD74 + WT-TIMP-1, n = 20; α-CD74 + N-TIMP-1, n = 20). Data (G–I) pooled from four independent experiments show biological replicates (box and whiskers plots) derived from six healthy donors. Statistics by one-way ANOVA and Dunnett test (G) or one-way ANOVA and Tukey test (H and I) across indicated groups. (J and K) Proportions of NK (J) and epithelial (K) subsets within samples (TUM Cohort). (L) PCA of NK and epithelial subsets (J and K), NK cell CD74 , and cancer cell TIMP1 expression (small dots) across patients (large dots). Exp., expression. (M) Spearman correlation between TIMP1 hi basal-like cancer cells and ctx hi NK cells. (N) Inferred cancer cell-to-NK cell signaling by CellChat in tumors (J and K) (also see H). (A) Created with BioRender.com .

Article Snippet: TIMP-1 anti-human mAb (rabbit) , Cell Signaling , CAT# 8946; clone D10E6; RRID: AB_10891805.

Techniques: Marker, Expressing, Control, Derivative Assay

Journal: Cell Reports Medicine

Article Title: Multimodal profiling of pancreatic cancer reveals a TIMP-1-dominated secretory profile determining pro-tumor immunoinstruction in human cancers

doi: 10.1016/j.xcrm.2025.102546

Figure Lengend Snippet: TIMP-1-dependent suppression of NK cell mTOR signaling links PDAC immunosuppression to clinical risk profiles (A–C) Workflow (A) to identify enriched pathways in NK_ctx hi C1 cluster vs. all other NK clusters in TUM Cohort (B) and Steele Cohort (C). DEGs (adj. p < 0.05) by Wilcoxon rank-sum test and auROC analysis. Mean odds ratios by pathway enrichment using Enrichr (Hallmark reference gene sets; see A and S6B , ). AKT, protein kinase B; IL-2, interleukin-2; mTORC1, mechanistic target of rapamycin complex 1; PI3K, phosphoinositide 3-kinase; STAT5, signal transducer and activator of transcription 5; UV, ultraviolet. (D–K) Workflow (D) to identify TIMP-1-dependent MIA PaCa-2-cell-mediated suppression of mTOR-signaling and IL-2 responses in NK cells, using co-culture [3 h (F); 24h (E and G)] or cancer-cell-conditioned media [24 h (H and I); 72 h (J and K)]. NK cell suppression with or without IL-2 activation assessed by (E) K562 killing (E:T 1:3; killing for 3 h); (F–H) p-mTOR(Ser2448)/p-S6(Ser235/236) signaling; (I) intracellular NK cell IFN-γ and TNF-α; (J) cell growth; (K) neutral lipid content; norm. to IL-2-free controls (H–J). Statistics between indicated groups: one-way ANOVA and Dunnett test (E–G, upper), one-way ANOVA and Tukey test (H, J, and K), unpaired Student’s t tests (E–G, lower; I). (L and M) NK cell (L) p-mTOR and (M) p-S6 levels upon exposure to rhWT-TIMP-1. Statistics by two-way ANOVA and Dunnett test. Data (E–M) shown as biological replicates [box and whiskers plots (E–I, and K–M) or mean ± SEM (J)]. (N–P) Cox-regression-based TIMP1 /NK risk score for recurrence-free (RFS) or metastasis-free survival (MFS) (TCGA-PAAD, n = 137), (N) For MFS, patient hazard ratios (HRs), and risk groups separated by quartiles. Statistics between linear predictors of TIMP1 expression and NK cell activity by Spearman correlation. (O) For MFS, survival probabilities by Kaplan-Meier curves (±95% CI). Statistics: global differences by KONP test, restricted mean survival time (RMST; τ = 0.9) between high (H) and low (L) risk groups by Wald test. (P) Heatmaps showing HRs (upper) and significance (lower) for the TIMP1 /NK score, both factors individually, and CISS by Cox regression analyses and G squared log likelihood ratio (∗ p < 0.05; ∗∗ p < 0.01). (A and D) Created with BioRender.com .

Article Snippet: TIMP-1 anti-human mAb (rabbit) , Cell Signaling , CAT# 8946; clone D10E6; RRID: AB_10891805.

Techniques: Co-Culture Assay, Activation Assay, Expressing, Activity Assay

Journal: Cell Reports Medicine

Article Title: Multimodal profiling of pancreatic cancer reveals a TIMP-1-dominated secretory profile determining pro-tumor immunoinstruction in human cancers

doi: 10.1016/j.xcrm.2025.102546

Figure Lengend Snippet: Multikinase inhibition targets TIMP-1 and CISS and enhances NK cell cytotoxicity in TIMP1 hi /CISS hi basal-like PDAC in vivo (A) TIMP1 hi /CISS hi basal-like PDAC in patients based on TIMP1/CISS expression (see ). Statistics by Mann-Whitney tests. (B) Transcription factor targets (TFTs) and kinase perturbations correlated with TIMP1/CISS in basal-like PDAC. (Left) Genes correlating (Spearman; p < 0.05) with TIMP1 expression were ranked by coefficients for GSEA (reference: C3:TFT). DEGs (Wilcoxon rank-sum test and auROC analysis) between TIMP1 hi /CISS hi basal-like cancer and other epithelial cells tested for pathway enrichment by Enrichr (“Kinase perturbations from GEO UP” and “DOWN”). (C) UMAP embedding of TIMP1 expression (see J), ERK activity, and FGFR signaling. ERK activity (TFT:MAPK3_Target_Genes) and FGFR signaling (Reactome_Signaling_by_FGFR) calculated by UCell and correlated to TIMP1 expression (Spearman). (D) Western blot ( n = 3; biological replicates) of intracellular TIMP-1 in MIA PaCa-2 cells upon treatment with trametinib (T) and nintedanib (N). Statistics by one-way ANOVA and Dunnett test. (E) ZIP synergy map of TIMP-1 inhibition (intracellular TIMP-1) in MIA PaCa-2 cells using trametinib or nintedanib (see A). (F) Workflow to assess in vivo effect of trametinib, nintedanib, and a -PDL1 treatment on TIMP1/CISS and NK cells in orthotopic classical and basal-like PDAC transplantation mouse models. (G–J) UMAP embedding of all cells (G) or cancer cells (H) from scRNA-seq of PDAC tumors derived from (F) and post-hoc cell-type annotations (G), cancer cell type (H, upper), or Timp1 expression (H, lower). (I) Timp1 and (J) CISS expression in cancer cells. Statistics by Mann-Whitney tests. (K) CISS factors in PDAC cells upon indicated treatments vs. controls from snRNA-seq data (F–J). Changes in CISS factors calculated by pseudobulk limma-voom workflow. Statistics: one-way ANOVA for matched data (genes) and Dunnett test between indicated groups. (L and M) UMAP embedding of NK cell clusters (M) and proportions across treatments (F,G), assessed by frequencies (M, upper) and PCA (M, lower). (N and O) Pathways enriched (N; Enrichr using Hallmark reference gene set) and cytotoxicity gene expression (O) between treatment- enriched and - reduced NK clusters. Changes in cytotoxicity genes (O) calculated by Wilcoxon rank-sum test and auROC analysis. Statistics by paired Student’s t test. (F) Created with BioRender.com .

Article Snippet: TIMP-1 anti-human mAb (rabbit) , Cell Signaling , CAT# 8946; clone D10E6; RRID: AB_10891805.

Techniques: Inhibition, In Vivo, Expressing, MANN-WHITNEY, Activity Assay, Western Blot, Transplantation Assay, Derivative Assay, Gene Expression