Hybrid Sensor Using Gold Nanoparticles and Conjugated Polyelectrolytes for Studying Sequence Rule in Protein–DNA Interactions

Protein-DNA interactions play center roles in many biological processes. Studying sequence specific protein-DNA interactions and revealing sequence rules require sensitive and quantitative methodologies that are capable of capturing subtle affinity difference with high accuracy and in a high throughput manner. In this study, double stranded DNA-conjugated gold nanoparticles (dsDNA-AuNPs) and water-soluble conjugated polyelectrolytes (CPEs) are used as cooperative sensing elements to construct a suit of hybrid sensors for detecting protein-DNA interactions, exploiting the differential Förster resonance energy transfer (FRET) with and without protein binding. Through a proper selection of CPEs in terms of charge properties relative to the charge of dsDNA-AuNPs and emission wavelengths relative to the AuNP extinction peak, the hybrid sensors can be constructed into "light-on", "light-off", and "two-way" models. Protein binding can be detected by fluorescence recovery, fluorescence quenching, or both ways, respectively. The "two-way" sensor allows for detection of proteins of any charge properties or unknown charge properties. With estrogen receptor (ERα and ERβ), their consensus DNA (5'-GGTCAnnnTGACC-5') element, and all 15 possible singly mutated elements (i.e., 3 possible base substitutions at each of 1 to 5 positions from left to right of the 5' end half site, GGTCA), we have demonstrated the accuracy of the hybrids sensors for determination of binding affinity constant, binding stoichiometry, and site- and nucleotide-specific binding energy matrix. The in vitro binding energy determined by the hybrid sensors correlates very well with the energy matrix computed from in vivo genome-wide ERα binding data using Thermodynamic Modeling of ChIP-Seq (rank correlation coefficient 0.98). The high degree of correlation of the in vitro energy matrix versus the in vivo matrix renders the new method a highly reliable alternative for understanding in vivo protein binding in the whole genome.