Fig. 1
Identification of circSHKBP1 as a biomarker for GC. a Cluster heat map showing the differentially expressed circRNAs in paired human GC tissues and normal tissues (n = 8). The circRNAs were classified according to the Pearson correlation test. b Volcano plot showing circRNAs that changed significantly between GC tissues and matched normal tissues. c Fold change of 5 most upregulated and downregulated circRNAs in 10 paired GC tumors and normal tissues detected by qRT-PCR. d Structure of circSHKBP1 and ALU elements in the flanking sequence. e Back-splicing sequence and junction site of circSHKBP1. f Representatives showing the expression of circSHKBP1 (red) in GC tumors and matched normal tissues detected by ISH. g Survival analysis of GC patients according to circSHKBP1 expression detected by ISH (n = 152, median survival (months): low expression vs. high expression = 50 vs. 22). The median expression level of circSHKBP1 was used as the cutoff value. Log-rank tests were used to determine statistical significance. h and i Expression of circSHKBP1 in GC tumors was 2.31 times higher than that in normal tissues (n = 72, cases: upregulated vs. downregulated = 53 vs. 19, P = 0.0008). j Level of circSHKBP1 in serum exosomes of GC patients (n = 20) and healthy controls (HC) (n = 20), as determined by qRT-PCR. k. Correlation of circSHKBP1 expression between GC tumors and serum exosomes (Y = 6.041*X + 0.03336, R2 = 0.6793, P < 0.0001). l Levels of exosomal circSHKBP1 before and after gastrectomy (n = 12). Quantitative data from three independent experiments are shown as the mean ± SD (error bars). *P < 0.05, **P < 0.01, ***P < 0.001 (Student’s t-test)