Figure 5

Expression analyses of PGC-1α and glutamine pathway in breast cancer patients. (A) Expression of PGC-1α across the different breast cancer clinical subtypes in The Cancer Genome Atlas (TCGA) data set, where crosses represent outliers for this subtype. Expression is given in log2-normalized reads per kilobase of exon model per million mapped reads (RPKM). The number of samples for Basal-like, ERBB2-enriched, Luminal A, Luminal B, and Normal-like is 87, 53, 201, 110, and 7, respectively. PGC-1α expression is higher in basal than luminal A and B subtypes, higher in luminal A subtype than luminal B subtype, and higher in ERBB2-enriched than luminal B subtype. P <1E-13 for Kruskal-Wallis, *P <0.05 using a posteriori Tukey-Kramer test. (B) (top) Expression of the glutamine metabolism genes is correlated with that of PGC-1α. The table shows the Pearson correlation coefficients between PGC-1α expression and the pathway score of the glutamine cluster. Colors represent the intensity of correlation as per the gradient shown. (bottom) Patient survival is related to glutamine cluster expression. The table shows P values for survival. Higher expression of glutamine cluster correlated with reduced survival in ERBB2-enriched and luminal B subtypes as well as in combined subtypes. Green arrows represent significant differences. P <0.05, logrank test. (C) High expression of glutamine metabolism genes is associated with reduced survival in all patients (all subtypes, left) and ERBB2-enriched patients (right). Kaplan-Meier survival curves represent the differences in survival between patients with high and low expression of the glutamine cluster.