Fig. 6: SWINGN controls a pro-oncogenic hub predictive of cancer status.

a Correlation plot showing SWINGN and PDGFRB or COL1A1 expression levels (as FPKM+ pseudocount values) in LUSC samples (n = 552) from TCGA database. Correlation p-values are calculated using a t-distribution. b RT-qPCR analysis of H226 cells stably expressing dCAS9-VP64 fusion protein and infected with lentiviral vectors carrying two independent sgRNAs targeting SWINGN promoter or an empty vector. Two different primer sets were used to detect SWINGN transcript. Graph shows mean ± SD of normalized values from two independent experiments. c RT-qPCR analysis of H226 cells treated with siRNAs depleting PDGFRB (siRNA#2) or COL1A1 (siRNA#1) alone or in combination with GAS6 knockdown (siRNA#2). Graphs show mean ± SD of three independent experiments. RNA levels were normalized to CTRL siRNA conditions. d Western blot analysis of H226 cells depleted of PDGFRB (siRNA#2) or COL1A1 (siRNA#1) alone or in combination with GAS6 (siRNA#2). Picture refers to a representative experiment out of two performed. e MTS proliferation assay of H226 cells treated with siRNAs depleting PDGFRB or COL1A1 alone or in combination with GAS6 knockdown at the indicated time points. Absorbance values were normalized to day 0. Graph shows mean ± SEM of three independent experiments. Significance has been calculated comparing each condition to CTRL siRNA sample. Conditions of PDGFRB and COL1A1 knockdown in combination with GAS6 were also compared to single knockdown samples. f K-means clustering of 552 LUSC_TCGA samples according to the expression of SWINGN gene signature (27 genes). Clusters are represented as red-labeled cluster#1 (comprising 13 tumor samples and 49 adjacent healthy tissues) and gray-colored cluster #2 (including 490 cancer tissues). g Box plot illustrating average expression of SWINGN gene signature in each LUSC_TCGA sample grouped according to cluster division. Squared dots represent healthy tissues while triangular points refer to tumor status. Statistics was calculated by Student’s t-test. Source data underlying b–e panels are provided as a Source Data file. Student’s t-test p-values are summarized as follows: * < 0.05; ** < 0.01; *** < 0.001.