Scientists identify pan-cancer biomarkers

Genomics and transcriptomics have successfully identified many therapeutic targets for cancer. However, changes in protein abundances and their chemical modifications can also drive tumor progression. To consider this additional dimension, Guo-sheng Hu, Zao-zao Zheng, Yao-hui He, Du-chuang Wang and colleagues at Xiamen University analyzed RNA and protein data from thousands of patients with 13 cancer types. They published their findings in Molecular & Cellular Proteomics. Using bioinformatics tools, they identified upregulated and downregulated genes specific to each cancer type as well as genes common to most types analyzed.

Tom Ellenberger, Washington University School of Medicine in St. Louis, via Wikimedia Commons

DNA ligase encircles the double helix to repair a broken strand of DNA.

The team discovered that tissue-specific genes were downregulated at both RNA and protein levels in all cancer types, indicating a loss of tissue identity. They showed that many genes involved in messenger RNA splicing, interferon pathway, fatty acid metabolism, and complement coagulation cascade, were dysregulated across several cancer types. The authors also found that ADH1B, the alcohol dehydrogenase that converts ethanol to acetaldehyde, was significantly downregulated in all cancer types. Conversely, the ribonucleotide regulatory subunit RRM2 was overexpressed. These proteins are examples of potential pan-cancer biomarkers, which can be used to discern cancer tissues from normal cells and potentially inform novel therapeutic strategies.

Effective cancer treatment also requires knowledge of the tumor’s stage of progression. To identify biomarkers for each tumor stage, the team analyzed how the cancer proteome changed throughout tumor progression. They used these findings to build models for tumor stage classification of several cancer types based on these biomarkers. They further constructed prognostic risk stratification models for corresponding cancer types based on dysregulated genes. They found that these models, when combined with the tumor-node-metastasis classification system, predicted cancer patient prognosis more accurately than either approach individually.

Protein-based approaches like these could be the key to better understanding cancer mechanisms and developing better treatments. Inhibitor drugs targeting RRM2 and other differentially expressed proteins identified in this study could be used to treat a range of cancer types and will be investigated further in future studies.