Targeting Metabolic Deregulation Landscapes in Breast Cancer Subtypes

Abstract

Metabolic deregulation is an emergent hallmark of cancer. Altered patterns of metabolic pathways result in exacerbated synthesis of macromolecules, increased proliferation, and resistance to treatment via alteration of drug processing. In addition, molecular heterogeneity creates a barrier to therapeutic options. In breast cancer, this broad variation in molecular metabolism constitutes, simultaneously, a source of prognostic and therapeutic challenges and a doorway to novel interventions. In this work, we investigated the metabolic deregulation landscapes in breast cancer molecular subtypes. Such landscapes are the regulatory signatures behind subtype-specific metabolic features. n = 735 breast cancer samples of the Luminal A, Luminal B, Her2+, and Basal subtypes, as well as n = 113 healthy breast tissue samples were analyzed. By means of a single-sample-based algorithm, deregulation for all metabolic pathways in every sample was determined. Deregulation levels match almost perfectly with the molecular classification, indicating that metabolic anomalies are closely associated with gene-expression signatures. Luminal B tumors are the most deregulated but are also the ones with higher within-subtype variance. We argued that this variation may underlie the fact that Luminal B tumors usually present the worst prognosis, a high rate of recurrence, and the lowest response to treatment in the long term. Finally, we designed a therapeutic scheme to regulate purine metabolism in breast cancer, independently of the molecular subtype. This scheme is founded on a computational tool that provides a set of FDA-approved drugs to target pathway-specific differentially expressed genes. By providing metabolic deregulation patterns at the single-sample level in breast cancer subtypes, we have been able to further characterize tumor behavior. This approach, together with targeted therapy, may open novel avenues for the design of personalized diagnostic, prognostic, and therapeutic strategies.

Loss of Long Distance Co-Expression in Lung Cancer

“…Genomic information for gene stable ID, chromosome/scaffold name, gene start (bp), gene end (bp), gene% GC content, and gene type was mapped using BioMart database (version GRCh38.p12). This data pre-processing pipeline has been previously implemented to analyze RNA-Seq data from breast cancer (Drago-García et al, 2017 ; Espinal-Enriquez et al, 2017 ; Dorantes-Gilardi et al, 2020 ; García-Cortés et al, 2020 ; Serrano-Carbajal et al, 2020 ) and clear cell renal carcinoma (Zamora-Fuentes et al, 2020 ).…”

Gene Expression and Co-expression Networks Are Strongly Altered Through Stages in Clear Cell Renal Carcinoma

“…(1) pre-normalization quality control, (2) batch and bias corrections (normalization) and (3) post-normalization quality control. Data pre-processing was conducted as previously (Drago-García et al, 2017 ; Espinal-Enriquez et al, 2017 ; de Anda-Jáuregui et al, 2019b , c ; García-Cortés et al, 2020 ; Serrano-Carbajal et al, 2020 ). Briefly, we assessed (a) biotype abundances, to assure that samples contained protein coding genes.…”

Luminal A Breast Cancer Co-expression Network: Structural and Functional Alterations

“…There are multiple studies where GCNs are constructed and important aspects of the connectivity structure are analyzed to identify genes prognosis markers (Hsu et al, 2019), metabolic deregulation (Serrano-Carbajal et al, 2020), and differences in transcriptional profiles (van Dam et al, 2018).…”