Fig 1: Transcriptional level of DEGs in OSCC.(A–H) The transcription level of DEGs in module cluster 1 according to GEPIA. The expression levels of CSTA, PPL, EVPL, SPRR3 and TGM1 with P-value < 0.05 in student’s t test (two tailed) were significantly differentially expressed between tumor and normal tissues in HNSCC (n.s represents not significant). (I–M) The differentially transcript level of candidate genes in module cluster 1 in GSE30784 dataset. The expression levels of CSTA, PPL, EVPL, SPRR3 and TGM1 with P-value < 0.001 were significantly differentially expressed between tumor and normal tissues in OSCC by student’s t test (two tailed).
Fig 2: Prognostic significance.(A) Validation was performed for tuning parameter selection through the LASSO regression analysis in GSE42743. LASSO, the least absolute shrinkage and selection operator. (B) Elaboration for LASSO coefficient profiles of prognostic RNAs. (C) K–M survival analysis based on GSE42743 dataset according to SPRR3 expression level. (D) Validation was performed for tuning parameter selection through the LASSO regression model in TCGA. (E) Elaboration for LASSO coefficient profiles of prognostic RNAs. (F) K–M survival analysis based on TCGA-OSCC dataset according to SPRR3 expression level.
Fig 3: DEGs between OSCC and adjacent tissues.(A) Volcano plot exhibits the high-throughput RNA-seq result. Among them, 85 genes were up-regulated (red plots) and 144 were down-regulated (green plots). SPRR3 was marked in the figure. (B) Heatmap of RNA sequencing result, red for up-regulated genes and blue for down-regulated genes. SPRR3 was marked in the figure. (C) GO enrichment of the DEGs. The DEGs are mostly correlated with keratinization, keratinocyte differentiation, ECM organization, cell adhesion and migration. (D) KEGG enrichment of the DEGs. The DEGs were enriched in xenobiotic metabolism, drug metabolism and chemical carcinogenesis, etc. (E) GSEA in hallmark enrichment, NES, P-value and FDR were listed respectively. The DEGs were enriched in epithelial mesenchymal transition, K-Ras signaling down and TGF-β signaling, etc. (F) GSEA in GO enrichment. The DEGs were enriched in DNA repication, epidermal cell differentiation, ECM disassembly and positve regulation of cell activation, etc. (G) GSEA in KEGG enrichment. The DEGs were enriched in drug metabolism, xenobiotic metabolism, focal adhesion, P53 signaling and tight junction, etc.
Fig 4: GSEA analysis of SPRR3 in TCGA database.(A) SPRR3 was enriched in apical surface, fatty acid metabolism, K-Ras signaling down and xenobiotic metabolism in hallmark terms.(B) SPRR3 was enriched in alcohol metabolic, drug metabolic, epidermis development, fatty acid derivative process and response to xenobiotic stimulus in GO terms. (C) SPRR3 was enriched in drug metabolism cytochrome P450, fatty acid metabolism, glycolysis gluconeogenesis, metabolism of xenobiotics by cytochrome and VEGF signaling pathway in KEGG terms. (D) Pearson correlation analysis between DEGs in GSE30784 dataset. SPRR3 was positively correlated with other DEGs in module cluster 1 (CSTA, DSG1, EVPL, IVL, PPL, TGM1) and negatively correlated with DEGs in Module cluster 2 (PLAU, COL10A1, COL3A1, COL4A6, COL5A3), and it was positively correlated with ADH1C, ADH7, ALDH1A1, ALDH1A3, ALDH3B2, CYP2E1, CDH1, TJP1, it was negatively correlated with CDH2, FN1, SNAI1, SNAI2 and TGFBI. (E) Pearson correlation analysis between DEGs in TCGA-OSCC dataset. SPRR3 was positively correlated with other DEGs in module cluster 1 and negatively correlated with DEGs in Module cluster 2, and it was positively correlated with ADH1C, ADH7, ALDH1A1, ALDH1A3, ALDH3B2, CYP2E1, CDH1, TJP1, it was negatively correlated with CDH2, FN1, SNAI1, SNAI2 and TGFBI. The gauge on the right refers to the “r” of Pearson correlation test. Asterisk (*), double asterisk (**) and triple asterisk (***) stand for P-value < 0.05, P-value < 0.01 and P-value < 0.001, respectively.
Fig 5: Clinical correlation of SPRR3 expression level in OSCC.(A) The expression of SPRR3 in TCGA-OSCC database, the expression level of SPRR3 was significantly lower in tumor group than that in normal group by student’s t test (two tailed). (B) In GSE30784, ROC curve was created to verify the diagnostic value of SPRR3. AUC = 0.920, P-value < 0.001. (C) In ROC curve analysis from TCGA database, AUC was 0.731 and P-value = 0.001. (D–H) Two-tailed student’s t test and one-way ANOVA test suggested that SPRR3 was differentially expressed in groups in accordance of whether the patients had alcohol consumption (P = 0.0139), the histological grade (P-value < 0.0001), N stage (P-value = 0.0076) of the patients, and whether the patients had lymphovascular invasion (P-value = 0.0102), perineural invasion (P-value = 0.0068) respectively. (I) Protein expression level of SPRR3 in well differentiated OSCC tissue. (J) Protein expression level of SPRR3 in moderate differentiated OSCC tissue. (K) Protein expression level of SPRR3 in poor differentiated OSCC tissue. (L) Protein expression level of SPRR3 in adjacent tissue. The image on the left in the group was taken at the low power field (100*), the right was taken at the high-power field (400*). The black line represents the length of 100 nm.
Supplier Page from Abcam for Anti-SPRR3 antibody