Potential applications of gene expression profiles obtained from circulating extracellular vesicles in breast cancer

作     者：Aritra Gupta Siddharth Bhardwaj Sayan Ghorai Rosina Ahmed Sanjit Agarwal Geetashree Mukherjee Kartiki V Desai 

作者机构：Biotechnology Research Innovation Council-National Institute of Biomedical Genomics (BRIC-NIBMG) Kalyani India. Regional Centre for Biotechnology PhD Program India. Tata Medical Centre 14 MAR (DH Block) New Town Rajarhat Kolkata 700160 India. 

出 版 物：《The journal of liquid biopsy》 

年 卷 期：2025年第7卷

页      面：100287页

主　　题：Diagnosis Liquid biopsy Transcriptomics 

摘      要：Background:Liquid biopsy-based biomarkers offer several advantages since they are minimally invasive, can be useful in longitudinal monitoring of the disease and have higher patient compliance. We describe a protocol using minimal volumes of archival and prospective serum/plasma samples to define the RNA contents of EVs and discuss its benefits and limitations. Methods:RNA-seq analysis of matched tumor biopsy, circulating EVs from breast cancer patients (EV-C, n = 26) and healthy donors (EV-H, n = 4) was performed and differentially expressed genes were validated by RT-PCR in a separate series of samples (EV-C, n = 32 and EV-H, n = 22). A total of 84 samples were studied. Results:RNA-seq data from 500 μl serum samples yielded more than 17000 genes, of which 320 were DEGs (adjusted p value ≤ 0.05) between EV-C and EV-H samples. Pathways for Myc V1, reactive oxygen species, angiogenesis, allograft rejection and Interferon regulated genes were over-represented in EV-C samples. Computational deconvolution algorithms for cell signatures identified immune cells such as Th1 and memory T-cells, endothelial cells, and osteoblasts from the stromal compartment as significant. Top 6 genes were validated by qRT-PCR in all samples (n = 84) and they consistently and correctly classified cancer and healthy groups. An independent set of 374 and 640 DEGs could segregate ER positive/ER negative and non-metastatic versus metastatic samples, respectively. EVs from metastatic samples had higher variability in gene expression patterns whereas those from non-metastatic samples showed a better correlation. Conclusion:By using low serum amounts successfully for EV transcriptomics, we demonstrate that a minimally invasive technique could be converted to a microinvasive format. These data lay the foundation for EV RNA based biomarker discovery for segregating breast cancers.