Fig 2: Luteolin affects pGlu modification of CD47 and the cell surface binding of SIRPα.A. Cell surface pGlu of CD47 using anti-CD47 CC2C6 monoclonal antibody in H929 cells following treatment with 10 μM luteolin for 72 h. The relative median fluorescence intensity (MFI) was displayed relative to DMSO control averages.B.Cell surface expression of CD47 using anti-CD47 B6H12 monoclonal antibody in H929 cells following treatment with 10 μM luteolin for 72 h. The relative median fluorescence intensity (MFI) was displayed relative to DMSO control averages.C. Endogenous transcriptional mRNA level of CD47 in H929 cells was demonstrated by Quantitative Real-Time PCR assay.D Endogenous protein level of CD47 in H929 cells was demonstrated by western-blotting assay. The presented data is a representative image from three independent experiments with similar results.E and F. FACS analysis of cell surface binding of SIRPα in H929 cells after treatment of PQ529 (10 μM) or luteolin (10 μM) for 48 h. The relative median fluorescence intensity (MFI) was calculated using DMSO sample as control. Error bars represent standard deviation from three independent experiments (*p < 0.05, **p < 0.01, ***p < 0.001).

Fig 3: Targeted epigenetic silencing (CRISPRoff) with the T.gonfio library.a, Alignment of the T.gonfio sgRNA sequences to the CRISPRoff targeting window. b, An example of flow cytometry measurement of CD151 in HEK293T cells after exposure (10 days) to pools of three separated sgRNAs (3-sgRNA, used in the CRISPRoff study) or the respective T.gonfio qgRNA. ‘Off’ represents the CRISPRoff plasmid; ‘off-mut’, CRISPRoff mutant carrying a catalytically inactive version of the DNA methyltransferase. c, Quantification of the percentage of cells with ITGB1, CD81 and CD151 silencing 10 days after co-transfection of the CRISPRoff or CRISPRoff-mutant plasmids with pools of three single sgRNAs (3-sg) or the qgRNA plasmid from the T.gonfio library in HEK293T cells. Each dot represents an independent biological repeat of the assay. Data are presented as mean ± s.e.m. d, An example of a flow cytometry measurement of the percentage of cells with IFNGR1 silencing 10 days post-CRISPR knockout with the qgRNA plasmid from the T.spiezzo library or CRISPRoff with the qgRNA plasmid from the T.gonfio library in HEK293T cells. e, Quantification of the percentage of cells with CD47, IFNGR1 and MCAM silencing 10 days post-CRISPR knockout with qgRNA plasmids from the T.spiezzo library or CRISPRoff with qgRNA plasmids from the T.gonfio library in HEK293T cells. Each dot represents a biological repeat of the assay. Data are presented as mean ± s.e.m. f, A schematic of pairwise pooled genome-wide CRISPRoff screens to determine the effectiveness of T.gonfio for gene silencing at the genome scale in HEK293T cells. Created with BioRender.com. g, Mean phenotype scores (γ) for essential and non-essential genes from screens with the CRISPRoff or the T.gonfio library. N = 2 biological repeats. The dashed lines indicate the median and interquartile range. h, SSMD* value measuring the separation between essential and non-essential genes for each repeat of the screens performed with the CRISPRoff or T.gonfio pool libraries. i, Plots of true and false positive rates for the distinction between essential and non-essential genes for screens with the CRISPRoff or the T.gonfio library. AUC, area under the curve. j, Lentiviral template switching between sgRNA2 and sgRNA3 in pre-pooled versus post-pooled T.gonfio libraries. T0 and end: baseline and endpoint as defined in the text. k, Mean phenotype scores (γ) for essential versus non-essential genes in screens with pre-pooled versus post-pooled T.gonfio libraries (2 biological repeats); IQR, interquartile range. l, Mean phenotype scores (γ) for essential genes in post-pooled versus pre-pooled screens. IQRs for each screen were depicted. m, SSMD* between essential and non-essential genes for screens performed with the pre-pooled versus post-pooled library. In g, k and l, values above 0.2 or below −0.6 were set to 0.2 and −0.6, respectively.Source data

Fig 4: Peptide 4N1K engagement on CD47 increased integrin α4β1 activation and function on human Jurkat cells. (A) CD47, α4 integrin, and cell nucleus were probed and stained by an APC conjugated anti-CD47 antibody, a FITC conjugated anti-α4 antibody or Hoechst staining and their merge signal confirmed the interaction between CD47 and α4 integrin on the cell surface (400×). (B–D). Western-blot analysis of the bands for CD47, α4, and β1 integrin subunit after immunoprecipitation with an anti-CD47 (B), anti-α4 (C), or anti-β1 (D) antibody. Cell lysate which was negatively labeled on the left was precipitated with an isotype control antibody. The whole-cell lysate on the right was named input. (E) Peptide 4N1K treatment increased the adhesion of human Jurkat cells to immobilized VCAM-1 (n=3, three replicate samples for each experimental condition). Cells in the control sample were not incubated with peptides. Typical photos of adhesion cells under various treatment conditions (×100) were shown on the left. (F) Peptide 4N1K treatment increased the migration of human Jurkat cells to immobilized VCAM-1 with SDF-1 as a chemoattractant (n=3, three replicate samples for each experimental condition). Typical photos of migrated cells under various treatment conditions (×200) were shown on the left. (G–J) Peptide 4N1K increased the percent of “extended form” integrin α4β1 with detection by an anti-active form α4β1 antibody. Cells were treated with peptide 4KGG (I), 4N1K (J), or without peptide treatment (H) and then were probed with an anti-active form integrin α4β1 antibody and thereafter a FITC-labeled secondary antibody. Human Jurkat cells incubated with an isotype antibody were used as an isotype control sample (G). Statistical analysis of up-regulation of integrin α4β1 active form was shown in (K) (n=3, three detections for the same experimental condition). Data show mean ± SEM. **P < 0.01. Total expression of integrin α4β1 on Jurkat cells was confirmed with the use of a FITC conjugated rabbit anti-human integrin α4β1 monoclonal antibody (M). Jurkat cells incubated with an isotype antibody were used as an isotype control sample (L).

Fig 5: The therapeutic effect of local injection of CD47 antibody, PKA inhibitor, G protein inhibitor, or Src kinase inhibitor on arthritis development. Six groups of wild-type rats were included (n=5) and for panels (A–D) the solid black line stands for the normal rat group, the solid orange line for the model group, the dashed red line for the CD47 antibody treatment group, the solid blue line for the PKA inhibitor (H89) treatment group, the dashed purple line for the G protein inhibitor treatment group and the green line for the Src kinase inhibitor (PP2) treatment group. (A–D). Swelling of hind left paws (A) and right paws (B), arthritis score (C) and clinic score (D) of rats during the experiment were shown. On day 28 of the animal experiment, rats were sacrificed and synovial tissues of the hind left paws were taken and used as samples for western-blot analysis. Photos of western-blot for CD47, SIRP-α, TSP-1, integrin subunit α4, αM, αv, αL, β1, β2, and β3 under PKA inhibitor H89, G protein inhibitor PTX, and Src kinase inhibitor PP2 treatment (E) and CD47 antibody treatment (F) were shown. Photos of western-blot for markers of T cells (CD3), neutrophils (Ly6G), macrophages (CD68), and B cells (CD45RA) in synovial membranes of rats under PKA inhibitor H89, G protein inhibitor PTX, and Src kinase inhibitor PP2 treatment (G) and CD47 antibody treatment (H) were shown.