What Structural Elements make a G-Quadruplex Physiologically Viable?

Abstract

As much as 375,000 potential G-Quadruplex forming sequences (PQS) have been identified in the human genome nevertheless conclusive demonstration of existence and relevance of these structures in vivo has been challenging. Along with the abundance of the PQS, a variety of proteins have also been demonstrated to efficiently destabilize G-Quadruplex (GQ) structures in vitro, suggesting the possibility that most GQ forming PQS are rapidly identified and unfolded by these proteins. In an attempt to perform a systematic study of GQ-protein interactions we selected Replication Protein A (RPA) as a model system and studied the interactions of RPA with 10 different GQ constructs with varying loop lengths and number of tetrad layers. RPA is the most abundant single strand DNA binding protein in eukaryotes, is involved in DNA replication and repair along with telomere maintenance, making it an ideal model system for this study. Single molecule FRET and bulk biophysical methods were used to perform these studies. Our results demonstrate a systematic increase in the stability of GQ structures against RPA unfolding as the loop lengths are shortened or the number of tetrad layers is increased. Certain GQ constructs were completely unfolded at RPA concentrations several orders of magnitude less than physiological RPA concentration, suggesting that they would not be viable in vivo. Interestingly, the time it takes for an RPA molecule to unfold a GQ construct does not depend on the number of layers the GQ has, suggesting that binding of RPA to GQ is the rate limiting step and that binding affinity decreases as the number of layers is increased. The implications of our data in terms of a molecular understanding of RPA-GQ interactions will be discussed along with our preliminary attempts to model this system.