A simple method for determination of electrospray response factors of noncovalent complexes: application to DNA G-quadruplex binding and self-assembly

Introduction
Electrospray mass spectrometry (ESI-MS) is now widely used to investigate non-covalent assemblies, and the power of the method resides in the individual detection of the various stoichiometries present in the mixture. However, to characterize the mixture quantitatively (i.e. to determine the concentration of each constituent), the biggest challenge comes from the fact that electrospray response factors of all constituents can be different, especially for constituents of significantly different masses and charges. Here we present a novel and simple method to determine the response factors of all constituents, using an internal standard. The method can be transposed to a wide variety of biological problems and to different ionization methods.
Experimental
All oligodeoxynucleotides were purchased from Eurogentec (Seraing, Belgium). G-quadruplexes formation is induced by addition of ammonium acetate (150 mM) to the guanine-rich strand. Polythymines (e.g. dT6) with m/z not overlapping with those of the complexes were used as internal standards. The ESI mass spectra were acquired using a Q-TOF Ultima Global (Waters, Manchester, UK) in the negative ion mode, with source parameters optimized for detecting the intact complexes. Quadruplex-ligand binding constants were quantified by titration of the quadruplex with the ligand (BOQ1, Dr. Teulade-Fichou). The kinetics of G-quadruplex self-assembly was monitored on-line, with the starting time defined by the addition of ammonium to the single strand. Peak areas were calculated using MassLynx and data were processed using Mathcad.
Preliminary results
Let us illustrate the method for the characterization of the self-assembly of dCGGTGGT. In NH4OAc, this G-rich strand forms a tetramer within a few hours, but the most stable species is an octamer, which is formed after a few days. The relative intensities of monomer, tetramer and octamer vary with the experimental settings: harsher conditions favor the transmission of the larger complexes. It is therefore foolish to assume that the response factors of the monomer, tetramer and octamer are equal. Using dT6 as an internal reference, we monitored the self-assembly kinetics by recording at several time points the following intensities: I(ref2-), I(G12-), I(G45-) and I(G89-), where Gn indicates a complex containing n G-rich strands. For each species Gn, we define Rn as its response factor relative to that of the reference: [Gn]/[ref] = Rn * I(Gn)/I(ref). At each time point the mass balance equation for the strand concentration must be satisfied: [G]tot = [G1] + 4[G4] + 8[G8]. Substituting with the definition of Rn’s one obtains a set of equations: [G]tot = [ref] * {R1*(I(G1)/I(ref) + 4R4*I(G4)/I(ref) + 8R8*I(G8)/I(ref)}. If the number of equations (time points) is greater than the number of unknown response factors, the least-squares fit to this system of equations is calculated using the Moore-Penrose pseudoinverse of the matrix constituted by the intensity ratios. Once all response factors are known, concentrations are recalculated at each time point, providing detailed characterization of the reaction kinetics, including reaction intermediates. We will also show that the same method can be used to quantify ligand binding using titration of the substrate by the ligand. This will be illustrated with titration of the quadruplex (TGGGGT)4 by ligand BOQ1. This ligand binds to the quadruplex and also produces ligand-bridged quadruplex multimers, which the present method allows quantifying for the first time.