Combinatorial in silico approach for cancer-associated 4Fe-4S protein discovery

Iron-sulfur (Fe-S) proteins play vital roles in multiple cellular processes, including mediating redox balance as well as DNA replication and repair. Given the role of Fe-S cofactors in genome maintenance, mutations in such metalloproteins could be associated with cancer. Nevertheless, only a few cancer-associated Fe-S proteins have been identified. In vitro, Fe-S cluster is susceptible to degradation in oxic environment. It could also be replaced by other metal ions during protein purification, mis-labelled as zinc finger or Zn-containing proteins. In silico, bioinorganic Fe-S cluster lacks unique sequence characteristics that distinguish itself from other metal-coordination sites, making motif prediction based solely on protein sequence difficult. Thus, in this study, three traits have been employed to discover putative cancer-associated 4Fe-4S proteins. Here, we have analyzed the human proteome via a three-pronged approach: (i) the presence of a triamino acid motif, (ii) the geometric arrangements of the cysteines, and (iii) the mutations of cancer-associated cysteines. In addition to MUTYH, a known 4Fe-4S human DNA glycosylase, 21 novel proteins were discovered as potential cancer-associated 4Fe-4S proteins. While 6 receptor proteins and 3 growth factors have been identified as potential targets in this study, 5 histone lysine methyltransferases with SET domains were also predicted to contain 4Fe-4S metallocofactors. This work provides insights for rational adjustments in experimental design and novel cancer biomarker discovery.