We then evaluated in vivo DHODH mRNA
We then evaluated in vivo DHODH mRNA dibromide using data from TCGA (Lee, Palm, Grimes, & Ji, 2015). To perform a pan-disease comparison of DHODH expression, log2 TPM expression values were converted to z-scores calculated per patient. Fig. 6 shows that across all 34 diseases (9726 unique patient samples) the median DHODH expression is above a z-score of zero indicating that most patients express DHODH more than the average of all other genes measured. Among these diseases, liver hepatocellular carcinoma (LIHC) patients tended to express DHODH the most (Fig. 6). TCGA disease patient samples were evaluated for reduced survival by comparing survival outcomes for patients with high DHODH expression to those with low DHODH expression. High DHODH expression was associated with reduced survivability in low-grade glioma (LGG) and stomach adenocarcinoma (STAD) patients. High DHODH expression was also found to be associated with increased grade in glioma patients from the TCGA in the Rembrandt and Gravendeel studies (Fig. 7) (Gravendeel et al., 2009; Madhavan et al., 2009; R Core Team, 2018). Within glioma, grade III and IV tumors have higher average mRNA expression of DHODH in comparison to normal and grade I tumors. Therefore, in general, higher grade glioma tumors have a higher DHODH mRNA expression level. In an effort to determine genes that are correlated with DHODH expression in the TCGA patient population, gene set enrichment analysis (GSEA) was used to identify pathways that were enriched with genes that co-express with DHODH. Co-expression of DHODH in patients was evaluated in LGG, STAD, LIHC, colon adenocarcinoma (COAD), rectum adenocarcinoma (READ), and pancreatic adenocarcinoma (PAAD) using a Pearson correlation measure (Fig. 8) (Subramanian et al., 2005). In each of these patient populations, DHODH is co-expressed with genes involved with mitochondrial translation, mitochondrial respiratory complex, the ETC, MYC targets, and translation elongation. These findings are consistent with our above description of DHODH function. To instantiate such a hypothesis, we constructed a protein interaction network using curated protein interactions sourced from Pathway Commons from KEGG, HPRD, and BioGRID that connected DHODH with the co-expressed genes POLD2, PPAN, RRP9, and TUFM (Cerami et al., 2011; Chatr-Aryamontri et al., 2017; Goel, Harsha, Pandey, & Prasad, 2012; Kanehisa, Sato, Kawashima, Furumichi, & Tanabe, 2016). Up to three intermediate interactions were allowed between DHODH and the co-expressed genes, generating a network consisting of 621 edges. Gene set enrichment was performed on the 200 intermediate network proteins using the DAVID web service and we identified eight statistically significant KEGG pathways with FDR adjusted p-values <0.1 which are represented as rectangular colored boxes in Fig. 9C (Dennis Jr et al., 2003; Jiao et al., 2012). Consistent with DHODH canonical function, nucleotide metabolism and related metabolic pathways were significantly enriched. Enrichment of the ribosome-related proteins was also consistent with known PPAN and RRP9 involvement in ribosomal biogenesis. In Fig. 9C, we indicate known direct, known indirect, or unknown connections between DHODH, POLD2, TUFM, PPAN, and RRP9 and the eight enriched pathways based on our review of the literature. TUFM participates in the elongation of a nascent peptide and DHODH provides the substrate for production of UMP, a component of rRNA – this known direct connection is represented by a solid black line. Other gene sets are connected through known indirect interactions (indicated by the black dashed lines) to DHODH co-expressed genes according to the literature. For example, POLD2 and the ribosome both require nucleotides to carry out their function. The stars in Fig. 9C indicate proteins that have been linked to differentiation. Finally, our analysis identified connections that have not previously been reported and are potential novel interactions (grey dashed lines).