Examination of the biological role of i-Motif and G-quadruplex nucleic acid structures

Abstract

It is well known that B-form DNA is not the only structure formed in the genome. In addition to the genetic information encoded in the nucleotide sequence of genomic DNA, non-B-form structures provide a second layer of information to regulate genome functions. G-quadruplex (G4) and i-Motif (i-M) are two four-stranded non-B-form structures formed in the guanine (G)- rich and cytosine (C)-rich regions of the genome, respectively. Recently, G4 DNA was visualized in human cells. Despite strong in vitro evidence that C-rich sequences can fold into i-M structures, the in vivo existence of these structures in the human genome has remained elusive to date. I have generated an antibody fragment (called i-Mab) that recognizes i-M structures with exquisite selectivity and affinity. For the first time, by visualizing i-M structures using i-Mab, I have shown that these structures are dynamically formed in the nuclei of human cells. I demonstrate that the in vivo formation of i-Ms is cell-cycle and pH dependent. In addition, I provide evidence that telomeres might adopt i-M structures. These results support the notion that i-Ms, in concert with G4s, could impart biologically relevant roles in the functioning of the genome. In addition, I used published whole cancer genome sequence data to find mutations in cancer patients that overlap potential RNA G4-forming sequences in 5' untranslated regions (UTRs) of mRNAs. Using RNAfold software, I assessed the effect of these mutations on the thermodynamic stability of RNA G4s in the context of full-length 5' UTRs. Of the 217 identified mutations, I found that 33 are predicted to destabilize and 21 predicted to stabilize RNA G4. I experimentally validated the effect of destabilizing mutations in the 5' UTRs of BCL2 and CXCL14 and one stabilizing mutation in the 5' UTR of TAOK2. These mutations resulted in an increase or a decrease in translation of these mRNAs, respectively. These findings suggest that mutations that modulate the G4 stability in the noncoding regions could act as cancer driver mutations, which present an opportunity for early cancer diagnosis using individual sequencing.