Fig 2: Updated SMYD3 interactome. Nodes and edges of SMYD3 functional associations are represented based on STRING Database criteria (https://string-db.org). The gene name of each interactor is indicated in agreement with HGNC nomenclature (Hugo Gene Nomenclature Committee, https://www.genenames.org/). mTOR/BLM/MET/p130 (RBL2) and AMPK subunits are connected to SMYD3 by orange lines corresponding to additional interactions, which include interactions recently identified by our group and by others and the interactions that are described in the present report.

Fig 3: SMYD3 knockdown induces the expression of CD8+ T-cell attracting chemokines and antigen processing machinery molecules in SCCHN cells(A-B) Induction of CXCL9, CXCL10 and CXCL11 mRNA transcripts and protein levels upon SMYD3 knockdown in SCCHN cell lines. HN-6 and HN-SCC-151 cells were transfected with a control siRNA (siNC) and a SMYD3-specifc siRNA (siSMYD3) for a period of 72 hours. 24 hours prior to collection, cells were exposed to 1000U/mL of IFN-ß. (A) SMYD3 knockdown resulted in significant upregulation of CXCL9, CXCL10 and CXCL11 mRNA levels in HN-6 and HN-SCC-151 cells assessed by real-time PCR. (***P < 0.001). mRNA levels were normalized by GAPDH. (B) Effect of SMYD3 knockdown on IFN-ß-mediated release of CXCL9, CXCL10 and CXCL11 chemokines in SCCHN cells. Supernatants collected from HN-6 cells were further assayed for CXCL9, CXCL10 and CXCL11 protein levels by ELISA. Cells treated with siSMYD3 showed increased protein levels of CXCL9, CXCL10 and CXCL11 production compared to siNC treated cells. (*P < 0.05; **P < 0.01; ***P < 0.001). (C-D) Induction of TAP1 mRNA transcripts and protein levels upon SMYD3 knockdown in SCCHN cell lines. HN-6 and HN-SCC-151 cells were transfected with a control siRNA (siNC) and a SMYD3-specifc siRNA (siSMYD3) for a period of 72 hours. 24 hours prior to collection, cells were exposed to 1000U/mL of IFN-ß. (C) Real-time PCR demonstrates upregulation of TAP1 after SMYD3 knockdown (same samples as in (A)) (*P < 0.05; ***P < 0.001). mRNA levels were normalized by GAPDH. (D) Western blotting of nuclear (SMYD3, H3K4me3) and cytoplasmic (TAP1) extracts shows an increase in TAP1 protein levels and attenuation of H3K4me3 levels upon SMYD3 knockdown.

Fig 4: Negative correlations between SMYD3 (A) or SUV39H2 (B) and chemokines or APM molecules(A) SMYD3 mRNA was negatively correlated with mRNA levels of CXCL9, CXCL10 and CXCL11 chemokines, and with mRNA levels of TAPBP, TAP1 and TAP2 using the TCGA database for SCCHN. (B) SUV39H2 mRNA was negatively correlated with mRNA levels of CXCL9, CXCL10 and CXCL11 chemokines, and with mRNA levels of HLA-A, B2M and TAP1 using the TCGA database for SCCHN. Pearson's correlation coefficients r were calculated and significance was defined as P < 0.05. All shown results were statistically significant.

Fig 5: In cellulo validation of SMYD3 interactions identified in silico. (A-E) Validation of SMYD3 interactions in HT-29 CRC cells. Co-immunoprecipitation of endogenous SMYD3 and mTOR (A), BLM (B), MET (C), p130 (D), or RPB1 (E) using specific antibodies. RPB1 was used as a control of our in silico analysis. Input corresponds to 10% of the lysate. Anti-IgGs were used as negative controls. Results are representative of at least three independent experiments. (F) P-tripeptide localization in specific domains of AMPK subunits. (G) Validation of SMYD3-AMPK interaction in gastrointestinal cancer cell lines. Co-immunoprecipitation of endogenous SMYD3 and AMPK using anti-SMYD3 and anti-AMPK antibodies in HT-29 CRC cells, HGC-27 GC cells, HLC-19 HCC cells, and CAPAN-1 PC cells. Input corresponds to 10% of the lysate. Anti-IgGs were used as negative controls. Results are representative of at least three independent experiments. (H) Validation of SMYD3 interaction with phospho-mTOR, phospho-AMPK, and phospho-MET in HT-29 CRC cells. Co-immunoprecipitation of endogenous SMYD3 and phospho-mTOR/phospho-AMPK/phospho-MET using anti-SMYD3 antibodies and antibodies against the phosphorylated form of the interacting proteins. In order to phospho-activate MET, HT-29 CRC cells were serum-starved for 24 h and subsequently treated with 10 ng/ml of HGF for 2 h. Input corresponds to 10% of the lysate. Anti-IgGs were used as negative controls. Results are representative of at least three independent experiments.