Most cancer cells are driven by multiple oncogenic pathways. What is worse, these pathways can interact with each other, endowing cancer cells to become resistant to therapies. Identifying a target that can simultaneously hit multiple oncogenic pathways and determining how this target influences tumorigenesis would improve the efficiency of cancer treatments.
Figure.1. Schematic representation of selected Furin substrates involved in maintaining hallmarks of CRC.
Furin, the first discovered and best-characterized proprotein convertase member, activates a variety of precursor proteins carboxyterminal of specific basic amino acid motifs1. Since its discovery in 1990s, multiple groups including Abdel-Majid Khatib´s lab and John Creemers´ lab devote time and effort to identifying the potential Furin substrates and exploring the role of Furin in human diseases. So far, Furin is predicted to have more than 100 substrates which range from cancer associated oncoproteins to coronaviruses related spike proteins2. More importantly, many of these substrates are involved in maintaining hallmarks of cancer (Figure.1). It is not surprising that Furin has been proposed as a potential target for treating human cancers. Unexpectedly, when we checked Furin expression in pan-cancer cell types, we found that not all types of cancer cells show higher expression level of Furin compared to matched normal tissues, indicating that Furin does not always contribute to carcinogenesis (Figure.2). For instance , Furin inhibition is not beneficial for inhibiting tumorigenic properties of liver cancer3,4.
Figure.2. The expression level of Furin in both tumor samples and matched normal tissues in pan-cancer cell types. T: tumors; N: normal tissue. Abbreviation of cancer types can be found in the TCGA database.
As a potential oncogene, we found that Furin inactivation simultaneously impairs processing of known pro-oncoproteins including IGF1R, IR, c-MET, TGF-β1 and NOTCH1 in colorectal cancer (CRC) cells. However, Furin inactivation leads to different phenotypes in CRC cells. Further analysis identified that Furin inactivation contributes to inhibiting tumorigenic properties in KRAS or BRAF driven CRC cells. This is further supported by the fact that Furin expression is positively correlated with KRAS and BRAF expression in CRC cells and Furin inactivation results in inhibition of MAPK and PI3K pathways in CRC cells with KRAS or BRAF mutations. In contrast, Furin inactivation leads to aggressive phenotypes in CRC cells without KRAS and BRAF mutations. We identified genes involved in activating the ERK-MAPK pathway, such as PTGS2 (also known as COX2), whose upregulation enhance tumorigenic properties in these CRC cells. Therefore, we propose that Furin inactivation is beneficial for treating KRAS/BRAF driven CRC. These findings not only open up a new avenue to treat devastating human cancers by simultaneously regulating multiple oncogenic pathways, but also provide a potential target for the tumor-promoting role of Furin inactivation. Furthermore, the mutant KRAS or BRAF could be a potential biomarker for the beneficial role of Furin inactivation in treating CRC. Of course, more work is still needed before Furin inhibition can be used for clinical applications. For example, what are the long-term side effects of Furin inhibitors in vivo?
1.Creemers JWM, Khatib A-M. Knock-out mouse models of proprotein convertases: unique functions or redundancy? Front Biosci 2008; 13: 4960–4971.
2.Tian S, Huang Q, Fang Y, Wu J. FurinDB: A Database of 20-residue furin cleavage site motifs, substrates and their associated drugs. Int J Mol Sci 2011; 12: 1060–1065.
3.Huang YH, Lin KH, Liao CH, Lai MW, Tseng YH, Yeh CT. Furin overexpression suppresses tumor growth and predicts a better postoperative disease-free survival in hepatocellular carcinoma. PLoS One 2012; 7: 1–10.
4.Declercq J, Brouwers B, Pruniau VPEG, Stijnen P, Tuand K, Meulemans S et al. Liver-specific inactivation of the proprotein convertase FURIN leads to increased hepatocellular carcinoma growth. Biomed Res Int 2015; 2015: 148651.